diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,154043 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 5.0,
+  "eval_steps": 500,
+  "global_step": 22000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00022727272727272727,
+      "grad_norm": 8.425206114804645,
+      "learning_rate": 1.25e-06,
+      "loss": 0.475,
+      "step": 1
+    },
+    {
+      "epoch": 0.00045454545454545455,
+      "grad_norm": 16.703857732700694,
+      "learning_rate": 1.24999999362758e-06,
+      "loss": 0.579,
+      "step": 2
+    },
+    {
+      "epoch": 0.0006818181818181819,
+      "grad_norm": 7.099934793829739,
+      "learning_rate": 1.2499999745103194e-06,
+      "loss": 0.3342,
+      "step": 3
+    },
+    {
+      "epoch": 0.0009090909090909091,
+      "grad_norm": 4.559620656121011,
+      "learning_rate": 1.2499999426482194e-06,
+      "loss": 0.1829,
+      "step": 4
+    },
+    {
+      "epoch": 0.0011363636363636363,
+      "grad_norm": 7.422326419934191,
+      "learning_rate": 1.2499998980412797e-06,
+      "loss": 0.3333,
+      "step": 5
+    },
+    {
+      "epoch": 0.0013636363636363637,
+      "grad_norm": 9.956772117779357,
+      "learning_rate": 1.249999840689502e-06,
+      "loss": 0.3571,
+      "step": 6
+    },
+    {
+      "epoch": 0.001590909090909091,
+      "grad_norm": 5.702703078049519,
+      "learning_rate": 1.2499997705928872e-06,
+      "loss": 0.2781,
+      "step": 7
+    },
+    {
+      "epoch": 0.0018181818181818182,
+      "grad_norm": 10.725659302893323,
+      "learning_rate": 1.2499996877514365e-06,
+      "loss": 0.4169,
+      "step": 8
+    },
+    {
+      "epoch": 0.0020454545454545456,
+      "grad_norm": 4.273175739435004,
+      "learning_rate": 1.2499995921651518e-06,
+      "loss": 0.17,
+      "step": 9
+    },
+    {
+      "epoch": 0.0022727272727272726,
+      "grad_norm": 7.520400898110296,
+      "learning_rate": 1.249999483834035e-06,
+      "loss": 0.2823,
+      "step": 10
+    },
+    {
+      "epoch": 0.0025,
+      "grad_norm": 4.65148821886084,
+      "learning_rate": 1.2499993627580887e-06,
+      "loss": 0.1931,
+      "step": 11
+    },
+    {
+      "epoch": 0.0027272727272727275,
+      "grad_norm": 6.953006290461999,
+      "learning_rate": 1.2499992289373146e-06,
+      "loss": 0.2133,
+      "step": 12
+    },
+    {
+      "epoch": 0.0029545454545454545,
+      "grad_norm": 7.482458571821909,
+      "learning_rate": 1.2499990823717161e-06,
+      "loss": 0.2589,
+      "step": 13
+    },
+    {
+      "epoch": 0.003181818181818182,
+      "grad_norm": 7.613036628092412,
+      "learning_rate": 1.249998923061296e-06,
+      "loss": 0.2075,
+      "step": 14
+    },
+    {
+      "epoch": 0.003409090909090909,
+      "grad_norm": 2.3314591700466054,
+      "learning_rate": 1.2499987510060575e-06,
+      "loss": 0.1283,
+      "step": 15
+    },
+    {
+      "epoch": 0.0036363636363636364,
+      "grad_norm": 3.2990615368241865,
+      "learning_rate": 1.2499985662060041e-06,
+      "loss": 0.1542,
+      "step": 16
+    },
+    {
+      "epoch": 0.003863636363636364,
+      "grad_norm": 1.4304698587146976,
+      "learning_rate": 1.2499983686611394e-06,
+      "loss": 0.0606,
+      "step": 17
+    },
+    {
+      "epoch": 0.004090909090909091,
+      "grad_norm": 3.241554699379774,
+      "learning_rate": 1.2499981583714676e-06,
+      "loss": 0.1733,
+      "step": 18
+    },
+    {
+      "epoch": 0.004318181818181818,
+      "grad_norm": 3.5868841402281575,
+      "learning_rate": 1.249997935336993e-06,
+      "loss": 0.2223,
+      "step": 19
+    },
+    {
+      "epoch": 0.004545454545454545,
+      "grad_norm": 5.976148995061139,
+      "learning_rate": 1.24999769955772e-06,
+      "loss": 0.2827,
+      "step": 20
+    },
+    {
+      "epoch": 0.004772727272727273,
+      "grad_norm": 5.1402702339018065,
+      "learning_rate": 1.249997451033654e-06,
+      "loss": 0.2825,
+      "step": 21
+    },
+    {
+      "epoch": 0.005,
+      "grad_norm": 3.627796703059699,
+      "learning_rate": 1.2499971897647991e-06,
+      "loss": 0.2427,
+      "step": 22
+    },
+    {
+      "epoch": 0.005227272727272727,
+      "grad_norm": 1.234144121093955,
+      "learning_rate": 1.2499969157511616e-06,
+      "loss": 0.0697,
+      "step": 23
+    },
+    {
+      "epoch": 0.005454545454545455,
+      "grad_norm": 4.577978689267598,
+      "learning_rate": 1.2499966289927463e-06,
+      "loss": 0.292,
+      "step": 24
+    },
+    {
+      "epoch": 0.005681818181818182,
+      "grad_norm": 2.0923687727777533,
+      "learning_rate": 1.2499963294895594e-06,
+      "loss": 0.1413,
+      "step": 25
+    },
+    {
+      "epoch": 0.005909090909090909,
+      "grad_norm": 5.99356656479195,
+      "learning_rate": 1.249996017241607e-06,
+      "loss": 0.3025,
+      "step": 26
+    },
+    {
+      "epoch": 0.006136363636363636,
+      "grad_norm": 1.6003841634184675,
+      "learning_rate": 1.2499956922488955e-06,
+      "loss": 0.0913,
+      "step": 27
+    },
+    {
+      "epoch": 0.006363636363636364,
+      "grad_norm": 2.302109421708841,
+      "learning_rate": 1.2499953545114313e-06,
+      "loss": 0.1076,
+      "step": 28
+    },
+    {
+      "epoch": 0.006590909090909091,
+      "grad_norm": 2.5945966971506076,
+      "learning_rate": 1.2499950040292216e-06,
+      "loss": 0.1678,
+      "step": 29
+    },
+    {
+      "epoch": 0.006818181818181818,
+      "grad_norm": 3.5228951231755,
+      "learning_rate": 1.2499946408022733e-06,
+      "loss": 0.2916,
+      "step": 30
+    },
+    {
+      "epoch": 0.007045454545454546,
+      "grad_norm": 2.792304209501951,
+      "learning_rate": 1.249994264830594e-06,
+      "loss": 0.1611,
+      "step": 31
+    },
+    {
+      "epoch": 0.007272727272727273,
+      "grad_norm": 2.353026654958416,
+      "learning_rate": 1.2499938761141914e-06,
+      "loss": 0.1503,
+      "step": 32
+    },
+    {
+      "epoch": 0.0075,
+      "grad_norm": 2.3258363595677394,
+      "learning_rate": 1.2499934746530731e-06,
+      "loss": 0.0987,
+      "step": 33
+    },
+    {
+      "epoch": 0.007727272727272728,
+      "grad_norm": 2.426693678106597,
+      "learning_rate": 1.2499930604472477e-06,
+      "loss": 0.1915,
+      "step": 34
+    },
+    {
+      "epoch": 0.007954545454545454,
+      "grad_norm": 2.02652166158405,
+      "learning_rate": 1.2499926334967234e-06,
+      "loss": 0.1145,
+      "step": 35
+    },
+    {
+      "epoch": 0.008181818181818182,
+      "grad_norm": 3.448090877175573,
+      "learning_rate": 1.2499921938015086e-06,
+      "loss": 0.1857,
+      "step": 36
+    },
+    {
+      "epoch": 0.00840909090909091,
+      "grad_norm": 2.578194172574861,
+      "learning_rate": 1.2499917413616129e-06,
+      "loss": 0.1732,
+      "step": 37
+    },
+    {
+      "epoch": 0.008636363636363636,
+      "grad_norm": 2.0283621665901763,
+      "learning_rate": 1.2499912761770451e-06,
+      "loss": 0.0928,
+      "step": 38
+    },
+    {
+      "epoch": 0.008863636363636363,
+      "grad_norm": 2.209456268698372,
+      "learning_rate": 1.2499907982478151e-06,
+      "loss": 0.1045,
+      "step": 39
+    },
+    {
+      "epoch": 0.00909090909090909,
+      "grad_norm": 2.596781773522846,
+      "learning_rate": 1.2499903075739321e-06,
+      "loss": 0.0881,
+      "step": 40
+    },
+    {
+      "epoch": 0.009318181818181817,
+      "grad_norm": 1.41605170431485,
+      "learning_rate": 1.2499898041554066e-06,
+      "loss": 0.0732,
+      "step": 41
+    },
+    {
+      "epoch": 0.009545454545454546,
+      "grad_norm": 2.266427654157247,
+      "learning_rate": 1.2499892879922484e-06,
+      "loss": 0.1618,
+      "step": 42
+    },
+    {
+      "epoch": 0.009772727272727273,
+      "grad_norm": 1.9247497128404598,
+      "learning_rate": 1.2499887590844683e-06,
+      "loss": 0.1101,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1413568061229467,
+      "learning_rate": 1.249988217432077e-06,
+      "loss": 0.1008,
+      "step": 44
+    },
+    {
+      "epoch": 0.010227272727272727,
+      "grad_norm": 4.1478161104180735,
+      "learning_rate": 1.2499876630350857e-06,
+      "loss": 0.1565,
+      "step": 45
+    },
+    {
+      "epoch": 0.010454545454545454,
+      "grad_norm": 2.3751698213546266,
+      "learning_rate": 1.2499870958935056e-06,
+      "loss": 0.192,
+      "step": 46
+    },
+    {
+      "epoch": 0.010681818181818181,
+      "grad_norm": 2.8648272631896,
+      "learning_rate": 1.2499865160073483e-06,
+      "loss": 0.1784,
+      "step": 47
+    },
+    {
+      "epoch": 0.01090909090909091,
+      "grad_norm": 2.9990128984515123,
+      "learning_rate": 1.2499859233766258e-06,
+      "loss": 0.1286,
+      "step": 48
+    },
+    {
+      "epoch": 0.011136363636363637,
+      "grad_norm": 2.1761181998744568,
+      "learning_rate": 1.2499853180013497e-06,
+      "loss": 0.1692,
+      "step": 49
+    },
+    {
+      "epoch": 0.011363636363636364,
+      "grad_norm": 2.08815293782181,
+      "learning_rate": 1.2499846998815326e-06,
+      "loss": 0.0965,
+      "step": 50
+    },
+    {
+      "epoch": 0.011590909090909091,
+      "grad_norm": 1.7064313120527725,
+      "learning_rate": 1.2499840690171872e-06,
+      "loss": 0.0778,
+      "step": 51
+    },
+    {
+      "epoch": 0.011818181818181818,
+      "grad_norm": 1.4252415333286064,
+      "learning_rate": 1.2499834254083262e-06,
+      "loss": 0.0785,
+      "step": 52
+    },
+    {
+      "epoch": 0.012045454545454545,
+      "grad_norm": 1.841453559562954,
+      "learning_rate": 1.2499827690549629e-06,
+      "loss": 0.1032,
+      "step": 53
+    },
+    {
+      "epoch": 0.012272727272727272,
+      "grad_norm": 1.8797360413824669,
+      "learning_rate": 1.2499820999571105e-06,
+      "loss": 0.0916,
+      "step": 54
+    },
+    {
+      "epoch": 0.0125,
+      "grad_norm": 0.6858851757755645,
+      "learning_rate": 1.249981418114783e-06,
+      "loss": 0.0253,
+      "step": 55
+    },
+    {
+      "epoch": 0.012727272727272728,
+      "grad_norm": 1.9885090750837784,
+      "learning_rate": 1.2499807235279937e-06,
+      "loss": 0.1436,
+      "step": 56
+    },
+    {
+      "epoch": 0.012954545454545455,
+      "grad_norm": 2.2780271360017643,
+      "learning_rate": 1.2499800161967573e-06,
+      "loss": 0.1392,
+      "step": 57
+    },
+    {
+      "epoch": 0.013181818181818182,
+      "grad_norm": 2.6829151727499947,
+      "learning_rate": 1.249979296121088e-06,
+      "loss": 0.1135,
+      "step": 58
+    },
+    {
+      "epoch": 0.013409090909090909,
+      "grad_norm": 1.864267199601082,
+      "learning_rate": 1.2499785633010005e-06,
+      "loss": 0.0913,
+      "step": 59
+    },
+    {
+      "epoch": 0.013636363636363636,
+      "grad_norm": 3.0713924448287013,
+      "learning_rate": 1.2499778177365098e-06,
+      "loss": 0.1564,
+      "step": 60
+    },
+    {
+      "epoch": 0.013863636363636364,
+      "grad_norm": 2.3199758758166973,
+      "learning_rate": 1.249977059427631e-06,
+      "loss": 0.1422,
+      "step": 61
+    },
+    {
+      "epoch": 0.014090909090909091,
+      "grad_norm": 2.34687159665448,
+      "learning_rate": 1.2499762883743795e-06,
+      "loss": 0.1066,
+      "step": 62
+    },
+    {
+      "epoch": 0.014318181818181818,
+      "grad_norm": 2.33181984716656,
+      "learning_rate": 1.2499755045767714e-06,
+      "loss": 0.1264,
+      "step": 63
+    },
+    {
+      "epoch": 0.014545454545454545,
+      "grad_norm": 1.7764690148004718,
+      "learning_rate": 1.2499747080348224e-06,
+      "loss": 0.0764,
+      "step": 64
+    },
+    {
+      "epoch": 0.014772727272727272,
+      "grad_norm": 2.04858612752072,
+      "learning_rate": 1.2499738987485487e-06,
+      "loss": 0.0887,
+      "step": 65
+    },
+    {
+      "epoch": 0.015,
+      "grad_norm": 2.711148414041785,
+      "learning_rate": 1.2499730767179668e-06,
+      "loss": 0.1972,
+      "step": 66
+    },
+    {
+      "epoch": 0.015227272727272726,
+      "grad_norm": 2.007693746794,
+      "learning_rate": 1.2499722419430938e-06,
+      "loss": 0.0699,
+      "step": 67
+    },
+    {
+      "epoch": 0.015454545454545455,
+      "grad_norm": 2.3623544721642116,
+      "learning_rate": 1.2499713944239462e-06,
+      "loss": 0.1025,
+      "step": 68
+    },
+    {
+      "epoch": 0.015681818181818182,
+      "grad_norm": 1.846270686638651,
+      "learning_rate": 1.2499705341605417e-06,
+      "loss": 0.0998,
+      "step": 69
+    },
+    {
+      "epoch": 0.015909090909090907,
+      "grad_norm": 2.8471111843861223,
+      "learning_rate": 1.2499696611528975e-06,
+      "loss": 0.1039,
+      "step": 70
+    },
+    {
+      "epoch": 0.016136363636363636,
+      "grad_norm": 2.0844758442060525,
+      "learning_rate": 1.2499687754010318e-06,
+      "loss": 0.1456,
+      "step": 71
+    },
+    {
+      "epoch": 0.016363636363636365,
+      "grad_norm": 1.7645340566749503,
+      "learning_rate": 1.2499678769049624e-06,
+      "loss": 0.1181,
+      "step": 72
+    },
+    {
+      "epoch": 0.01659090909090909,
+      "grad_norm": 4.631649822026886,
+      "learning_rate": 1.2499669656647076e-06,
+      "loss": 0.1956,
+      "step": 73
+    },
+    {
+      "epoch": 0.01681818181818182,
+      "grad_norm": 1.77394717636909,
+      "learning_rate": 1.249966041680286e-06,
+      "loss": 0.1044,
+      "step": 74
+    },
+    {
+      "epoch": 0.017045454545454544,
+      "grad_norm": 0.9104467273199681,
+      "learning_rate": 1.2499651049517166e-06,
+      "loss": 0.0303,
+      "step": 75
+    },
+    {
+      "epoch": 0.017272727272727273,
+      "grad_norm": 2.489106170574559,
+      "learning_rate": 1.2499641554790185e-06,
+      "loss": 0.1675,
+      "step": 76
+    },
+    {
+      "epoch": 0.0175,
+      "grad_norm": 3.1243175810193464,
+      "learning_rate": 1.2499631932622108e-06,
+      "loss": 0.1638,
+      "step": 77
+    },
+    {
+      "epoch": 0.017727272727272727,
+      "grad_norm": 1.7098376755087212,
+      "learning_rate": 1.2499622183013136e-06,
+      "loss": 0.0705,
+      "step": 78
+    },
+    {
+      "epoch": 0.017954545454545456,
+      "grad_norm": 1.454357663122096,
+      "learning_rate": 1.249961230596346e-06,
+      "loss": 0.0723,
+      "step": 79
+    },
+    {
+      "epoch": 0.01818181818181818,
+      "grad_norm": 1.7139847635692904,
+      "learning_rate": 1.249960230147329e-06,
+      "loss": 0.0535,
+      "step": 80
+    },
+    {
+      "epoch": 0.01840909090909091,
+      "grad_norm": 1.6398250347879888,
+      "learning_rate": 1.2499592169542823e-06,
+      "loss": 0.0956,
+      "step": 81
+    },
+    {
+      "epoch": 0.018636363636363635,
+      "grad_norm": 2.1467671250122122,
+      "learning_rate": 1.2499581910172271e-06,
+      "loss": 0.1512,
+      "step": 82
+    },
+    {
+      "epoch": 0.018863636363636364,
+      "grad_norm": 1.947834593194105,
+      "learning_rate": 1.2499571523361839e-06,
+      "loss": 0.1064,
+      "step": 83
+    },
+    {
+      "epoch": 0.019090909090909092,
+      "grad_norm": 3.3384304047035713,
+      "learning_rate": 1.2499561009111742e-06,
+      "loss": 0.1294,
+      "step": 84
+    },
+    {
+      "epoch": 0.019318181818181818,
+      "grad_norm": 2.9405487457471042,
+      "learning_rate": 1.249955036742219e-06,
+      "loss": 0.2327,
+      "step": 85
+    },
+    {
+      "epoch": 0.019545454545454546,
+      "grad_norm": 3.084034917753596,
+      "learning_rate": 1.2499539598293406e-06,
+      "loss": 0.0959,
+      "step": 86
+    },
+    {
+      "epoch": 0.01977272727272727,
+      "grad_norm": 1.3523080467450024,
+      "learning_rate": 1.2499528701725605e-06,
+      "loss": 0.0912,
+      "step": 87
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.564159196646641,
+      "learning_rate": 1.2499517677719015e-06,
+      "loss": 0.1905,
+      "step": 88
+    },
+    {
+      "epoch": 0.020227272727272726,
+      "grad_norm": 2.0641438977079645,
+      "learning_rate": 1.2499506526273853e-06,
+      "loss": 0.1362,
+      "step": 89
+    },
+    {
+      "epoch": 0.020454545454545454,
+      "grad_norm": 2.8543544691094014,
+      "learning_rate": 1.249949524739035e-06,
+      "loss": 0.1613,
+      "step": 90
+    },
+    {
+      "epoch": 0.020681818181818183,
+      "grad_norm": 2.5920658841582584,
+      "learning_rate": 1.2499483841068736e-06,
+      "loss": 0.1772,
+      "step": 91
+    },
+    {
+      "epoch": 0.02090909090909091,
+      "grad_norm": 3.194994862372547,
+      "learning_rate": 1.2499472307309245e-06,
+      "loss": 0.1675,
+      "step": 92
+    },
+    {
+      "epoch": 0.021136363636363637,
+      "grad_norm": 1.6982227522331572,
+      "learning_rate": 1.249946064611211e-06,
+      "loss": 0.0895,
+      "step": 93
+    },
+    {
+      "epoch": 0.021363636363636362,
+      "grad_norm": 1.7331006445671497,
+      "learning_rate": 1.249944885747757e-06,
+      "loss": 0.0963,
+      "step": 94
+    },
+    {
+      "epoch": 0.02159090909090909,
+      "grad_norm": 3.145097631653776,
+      "learning_rate": 1.2499436941405865e-06,
+      "loss": 0.1139,
+      "step": 95
+    },
+    {
+      "epoch": 0.02181818181818182,
+      "grad_norm": 2.2044363859308658,
+      "learning_rate": 1.2499424897897237e-06,
+      "loss": 0.1396,
+      "step": 96
+    },
+    {
+      "epoch": 0.022045454545454545,
+      "grad_norm": 1.9853438928264655,
+      "learning_rate": 1.2499412726951935e-06,
+      "loss": 0.08,
+      "step": 97
+    },
+    {
+      "epoch": 0.022272727272727274,
+      "grad_norm": 2.198557331332217,
+      "learning_rate": 1.2499400428570202e-06,
+      "loss": 0.0661,
+      "step": 98
+    },
+    {
+      "epoch": 0.0225,
+      "grad_norm": 4.080717454572187,
+      "learning_rate": 1.2499388002752294e-06,
+      "loss": 0.1376,
+      "step": 99
+    },
+    {
+      "epoch": 0.022727272727272728,
+      "grad_norm": 1.5428408331144001,
+      "learning_rate": 1.2499375449498461e-06,
+      "loss": 0.0651,
+      "step": 100
+    },
+    {
+      "epoch": 0.022954545454545453,
+      "grad_norm": 2.600536078400038,
+      "learning_rate": 1.2499362768808958e-06,
+      "loss": 0.0751,
+      "step": 101
+    },
+    {
+      "epoch": 0.023181818181818182,
+      "grad_norm": 2.137513298977496,
+      "learning_rate": 1.2499349960684049e-06,
+      "loss": 0.1131,
+      "step": 102
+    },
+    {
+      "epoch": 0.02340909090909091,
+      "grad_norm": 2.8391249828635443,
+      "learning_rate": 1.249933702512399e-06,
+      "loss": 0.1247,
+      "step": 103
+    },
+    {
+      "epoch": 0.023636363636363636,
+      "grad_norm": 2.355507539024441,
+      "learning_rate": 1.2499323962129046e-06,
+      "loss": 0.1332,
+      "step": 104
+    },
+    {
+      "epoch": 0.023863636363636365,
+      "grad_norm": 2.0182839041589102,
+      "learning_rate": 1.2499310771699485e-06,
+      "loss": 0.0891,
+      "step": 105
+    },
+    {
+      "epoch": 0.02409090909090909,
+      "grad_norm": 1.8400690426452173,
+      "learning_rate": 1.2499297453835574e-06,
+      "loss": 0.1337,
+      "step": 106
+    },
+    {
+      "epoch": 0.02431818181818182,
+      "grad_norm": 2.526361066181947,
+      "learning_rate": 1.2499284008537586e-06,
+      "loss": 0.0891,
+      "step": 107
+    },
+    {
+      "epoch": 0.024545454545454544,
+      "grad_norm": 2.3842886203859415,
+      "learning_rate": 1.2499270435805793e-06,
+      "loss": 0.1126,
+      "step": 108
+    },
+    {
+      "epoch": 0.024772727272727273,
+      "grad_norm": 2.1826463491692567,
+      "learning_rate": 1.2499256735640477e-06,
+      "loss": 0.0836,
+      "step": 109
+    },
+    {
+      "epoch": 0.025,
+      "grad_norm": 1.9757112887727335,
+      "learning_rate": 1.2499242908041912e-06,
+      "loss": 0.1173,
+      "step": 110
+    },
+    {
+      "epoch": 0.025227272727272727,
+      "grad_norm": 2.016627005660986,
+      "learning_rate": 1.2499228953010379e-06,
+      "loss": 0.0647,
+      "step": 111
+    },
+    {
+      "epoch": 0.025454545454545455,
+      "grad_norm": 2.791577238457075,
+      "learning_rate": 1.249921487054617e-06,
+      "loss": 0.0924,
+      "step": 112
+    },
+    {
+      "epoch": 0.02568181818181818,
+      "grad_norm": 0.7414768129941378,
+      "learning_rate": 1.2499200660649562e-06,
+      "loss": 0.0467,
+      "step": 113
+    },
+    {
+      "epoch": 0.02590909090909091,
+      "grad_norm": 2.470550895257173,
+      "learning_rate": 1.2499186323320853e-06,
+      "loss": 0.0903,
+      "step": 114
+    },
+    {
+      "epoch": 0.026136363636363635,
+      "grad_norm": 1.989997460311596,
+      "learning_rate": 1.2499171858560334e-06,
+      "loss": 0.0747,
+      "step": 115
+    },
+    {
+      "epoch": 0.026363636363636363,
+      "grad_norm": 1.304636253648533,
+      "learning_rate": 1.2499157266368298e-06,
+      "loss": 0.106,
+      "step": 116
+    },
+    {
+      "epoch": 0.026590909090909092,
+      "grad_norm": 2.170813182873413,
+      "learning_rate": 1.2499142546745041e-06,
+      "loss": 0.118,
+      "step": 117
+    },
+    {
+      "epoch": 0.026818181818181817,
+      "grad_norm": 1.7079936765084984,
+      "learning_rate": 1.2499127699690865e-06,
+      "loss": 0.0825,
+      "step": 118
+    },
+    {
+      "epoch": 0.027045454545454546,
+      "grad_norm": 1.9935873445273355,
+      "learning_rate": 1.2499112725206075e-06,
+      "loss": 0.0942,
+      "step": 119
+    },
+    {
+      "epoch": 0.02727272727272727,
+      "grad_norm": 1.9698541722045861,
+      "learning_rate": 1.2499097623290974e-06,
+      "loss": 0.0738,
+      "step": 120
+    },
+    {
+      "epoch": 0.0275,
+      "grad_norm": 2.3751522382890737,
+      "learning_rate": 1.2499082393945871e-06,
+      "loss": 0.1569,
+      "step": 121
+    },
+    {
+      "epoch": 0.02772727272727273,
+      "grad_norm": 2.3655077215281177,
+      "learning_rate": 1.2499067037171076e-06,
+      "loss": 0.0848,
+      "step": 122
+    },
+    {
+      "epoch": 0.027954545454545454,
+      "grad_norm": 1.6685493980232267,
+      "learning_rate": 1.24990515529669e-06,
+      "loss": 0.081,
+      "step": 123
+    },
+    {
+      "epoch": 0.028181818181818183,
+      "grad_norm": 1.5229183443726408,
+      "learning_rate": 1.2499035941333662e-06,
+      "loss": 0.0588,
+      "step": 124
+    },
+    {
+      "epoch": 0.028409090909090908,
+      "grad_norm": 4.039603890034486,
+      "learning_rate": 1.2499020202271677e-06,
+      "loss": 0.1518,
+      "step": 125
+    },
+    {
+      "epoch": 0.028636363636363637,
+      "grad_norm": 2.069408513450431,
+      "learning_rate": 1.2499004335781272e-06,
+      "loss": 0.0755,
+      "step": 126
+    },
+    {
+      "epoch": 0.028863636363636362,
+      "grad_norm": 1.3721481584511823,
+      "learning_rate": 1.2498988341862764e-06,
+      "loss": 0.0713,
+      "step": 127
+    },
+    {
+      "epoch": 0.02909090909090909,
+      "grad_norm": 1.6868324923514968,
+      "learning_rate": 1.2498972220516482e-06,
+      "loss": 0.0919,
+      "step": 128
+    },
+    {
+      "epoch": 0.02931818181818182,
+      "grad_norm": 2.2392283318126434,
+      "learning_rate": 1.2498955971742754e-06,
+      "loss": 0.066,
+      "step": 129
+    },
+    {
+      "epoch": 0.029545454545454545,
+      "grad_norm": 1.4128560132521177,
+      "learning_rate": 1.2498939595541915e-06,
+      "loss": 0.1017,
+      "step": 130
+    },
+    {
+      "epoch": 0.029772727272727274,
+      "grad_norm": 1.4535607792449552,
+      "learning_rate": 1.2498923091914293e-06,
+      "loss": 0.0821,
+      "step": 131
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8206341262742942,
+      "learning_rate": 1.2498906460860228e-06,
+      "loss": 0.0836,
+      "step": 132
+    },
+    {
+      "epoch": 0.030227272727272728,
+      "grad_norm": 7.3955525540031255,
+      "learning_rate": 1.2498889702380058e-06,
+      "loss": 0.2209,
+      "step": 133
+    },
+    {
+      "epoch": 0.030454545454545453,
+      "grad_norm": 1.7465762796843254,
+      "learning_rate": 1.2498872816474125e-06,
+      "loss": 0.0793,
+      "step": 134
+    },
+    {
+      "epoch": 0.03068181818181818,
+      "grad_norm": 2.7390507513910394,
+      "learning_rate": 1.2498855803142773e-06,
+      "loss": 0.1136,
+      "step": 135
+    },
+    {
+      "epoch": 0.03090909090909091,
+      "grad_norm": 2.1252806498839534,
+      "learning_rate": 1.249883866238635e-06,
+      "loss": 0.0894,
+      "step": 136
+    },
+    {
+      "epoch": 0.031136363636363636,
+      "grad_norm": 1.7280668714054925,
+      "learning_rate": 1.2498821394205207e-06,
+      "loss": 0.0773,
+      "step": 137
+    },
+    {
+      "epoch": 0.031363636363636364,
+      "grad_norm": 1.0649491990270854,
+      "learning_rate": 1.249880399859969e-06,
+      "loss": 0.0423,
+      "step": 138
+    },
+    {
+      "epoch": 0.03159090909090909,
+      "grad_norm": 1.2294938221743361,
+      "learning_rate": 1.2498786475570163e-06,
+      "loss": 0.063,
+      "step": 139
+    },
+    {
+      "epoch": 0.031818181818181815,
+      "grad_norm": 1.5496095898268643,
+      "learning_rate": 1.2498768825116976e-06,
+      "loss": 0.1103,
+      "step": 140
+    },
+    {
+      "epoch": 0.032045454545454544,
+      "grad_norm": 2.0984875167544668,
+      "learning_rate": 1.2498751047240488e-06,
+      "loss": 0.1442,
+      "step": 141
+    },
+    {
+      "epoch": 0.03227272727272727,
+      "grad_norm": 1.6985936082092827,
+      "learning_rate": 1.2498733141941068e-06,
+      "loss": 0.0925,
+      "step": 142
+    },
+    {
+      "epoch": 0.0325,
+      "grad_norm": 1.9046935066187278,
+      "learning_rate": 1.2498715109219076e-06,
+      "loss": 0.0868,
+      "step": 143
+    },
+    {
+      "epoch": 0.03272727272727273,
+      "grad_norm": 2.5763026674261957,
+      "learning_rate": 1.2498696949074883e-06,
+      "loss": 0.0807,
+      "step": 144
+    },
+    {
+      "epoch": 0.03295454545454545,
+      "grad_norm": 5.018254303558268,
+      "learning_rate": 1.2498678661508856e-06,
+      "loss": 0.1445,
+      "step": 145
+    },
+    {
+      "epoch": 0.03318181818181818,
+      "grad_norm": 2.747862257844138,
+      "learning_rate": 1.2498660246521371e-06,
+      "loss": 0.1741,
+      "step": 146
+    },
+    {
+      "epoch": 0.03340909090909091,
+      "grad_norm": 2.55595877057827,
+      "learning_rate": 1.2498641704112802e-06,
+      "loss": 0.1301,
+      "step": 147
+    },
+    {
+      "epoch": 0.03363636363636364,
+      "grad_norm": 2.8126716646922745,
+      "learning_rate": 1.2498623034283526e-06,
+      "loss": 0.1102,
+      "step": 148
+    },
+    {
+      "epoch": 0.03386363636363637,
+      "grad_norm": 3.081604463671771,
+      "learning_rate": 1.2498604237033927e-06,
+      "loss": 0.175,
+      "step": 149
+    },
+    {
+      "epoch": 0.03409090909090909,
+      "grad_norm": 2.2010244452017274,
+      "learning_rate": 1.2498585312364386e-06,
+      "loss": 0.1104,
+      "step": 150
+    },
+    {
+      "epoch": 0.03431818181818182,
+      "grad_norm": 1.9417322200625715,
+      "learning_rate": 1.2498566260275289e-06,
+      "loss": 0.1266,
+      "step": 151
+    },
+    {
+      "epoch": 0.034545454545454546,
+      "grad_norm": 2.9557238564175226,
+      "learning_rate": 1.2498547080767026e-06,
+      "loss": 0.0694,
+      "step": 152
+    },
+    {
+      "epoch": 0.034772727272727275,
+      "grad_norm": 3.7840363508815273,
+      "learning_rate": 1.2498527773839984e-06,
+      "loss": 0.1266,
+      "step": 153
+    },
+    {
+      "epoch": 0.035,
+      "grad_norm": 1.8482617552341176,
+      "learning_rate": 1.2498508339494561e-06,
+      "loss": 0.0864,
+      "step": 154
+    },
+    {
+      "epoch": 0.035227272727272725,
+      "grad_norm": 2.4543360121326874,
+      "learning_rate": 1.2498488777731152e-06,
+      "loss": 0.0786,
+      "step": 155
+    },
+    {
+      "epoch": 0.035454545454545454,
+      "grad_norm": 3.4823282383888716,
+      "learning_rate": 1.2498469088550156e-06,
+      "loss": 0.1944,
+      "step": 156
+    },
+    {
+      "epoch": 0.03568181818181818,
+      "grad_norm": 2.54803770619922,
+      "learning_rate": 1.2498449271951972e-06,
+      "loss": 0.1017,
+      "step": 157
+    },
+    {
+      "epoch": 0.03590909090909091,
+      "grad_norm": 2.5809136450193124,
+      "learning_rate": 1.249842932793701e-06,
+      "loss": 0.1267,
+      "step": 158
+    },
+    {
+      "epoch": 0.03613636363636363,
+      "grad_norm": 1.851422587892908,
+      "learning_rate": 1.249840925650567e-06,
+      "loss": 0.0899,
+      "step": 159
+    },
+    {
+      "epoch": 0.03636363636363636,
+      "grad_norm": 1.475004184436514,
+      "learning_rate": 1.2498389057658364e-06,
+      "loss": 0.1157,
+      "step": 160
+    },
+    {
+      "epoch": 0.03659090909090909,
+      "grad_norm": 2.7990019256184673,
+      "learning_rate": 1.2498368731395507e-06,
+      "loss": 0.1012,
+      "step": 161
+    },
+    {
+      "epoch": 0.03681818181818182,
+      "grad_norm": 2.181800246998864,
+      "learning_rate": 1.2498348277717509e-06,
+      "loss": 0.0857,
+      "step": 162
+    },
+    {
+      "epoch": 0.03704545454545455,
+      "grad_norm": 1.6332586141941614,
+      "learning_rate": 1.2498327696624787e-06,
+      "loss": 0.0774,
+      "step": 163
+    },
+    {
+      "epoch": 0.03727272727272727,
+      "grad_norm": 1.9927264263870847,
+      "learning_rate": 1.2498306988117765e-06,
+      "loss": 0.1419,
+      "step": 164
+    },
+    {
+      "epoch": 0.0375,
+      "grad_norm": 1.8319460739877114,
+      "learning_rate": 1.2498286152196863e-06,
+      "loss": 0.1171,
+      "step": 165
+    },
+    {
+      "epoch": 0.03772727272727273,
+      "grad_norm": 2.992269631090057,
+      "learning_rate": 1.2498265188862505e-06,
+      "loss": 0.1453,
+      "step": 166
+    },
+    {
+      "epoch": 0.037954545454545456,
+      "grad_norm": 2.844302432802208,
+      "learning_rate": 1.2498244098115118e-06,
+      "loss": 0.1557,
+      "step": 167
+    },
+    {
+      "epoch": 0.038181818181818185,
+      "grad_norm": 3.278661117609225,
+      "learning_rate": 1.2498222879955135e-06,
+      "loss": 0.1527,
+      "step": 168
+    },
+    {
+      "epoch": 0.03840909090909091,
+      "grad_norm": 4.445555455608285,
+      "learning_rate": 1.2498201534382985e-06,
+      "loss": 0.2075,
+      "step": 169
+    },
+    {
+      "epoch": 0.038636363636363635,
+      "grad_norm": 2.483457356533238,
+      "learning_rate": 1.2498180061399108e-06,
+      "loss": 0.0971,
+      "step": 170
+    },
+    {
+      "epoch": 0.038863636363636364,
+      "grad_norm": 2.790420445652532,
+      "learning_rate": 1.2498158461003935e-06,
+      "loss": 0.1253,
+      "step": 171
+    },
+    {
+      "epoch": 0.03909090909090909,
+      "grad_norm": 1.7560217431711485,
+      "learning_rate": 1.2498136733197914e-06,
+      "loss": 0.0884,
+      "step": 172
+    },
+    {
+      "epoch": 0.03931818181818182,
+      "grad_norm": 2.1492245930351688,
+      "learning_rate": 1.2498114877981483e-06,
+      "loss": 0.1128,
+      "step": 173
+    },
+    {
+      "epoch": 0.03954545454545454,
+      "grad_norm": 1.5919068467925894,
+      "learning_rate": 1.2498092895355088e-06,
+      "loss": 0.0804,
+      "step": 174
+    },
+    {
+      "epoch": 0.03977272727272727,
+      "grad_norm": 1.535338022737946,
+      "learning_rate": 1.249807078531918e-06,
+      "loss": 0.0585,
+      "step": 175
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5122790448465122,
+      "learning_rate": 1.2498048547874208e-06,
+      "loss": 0.0641,
+      "step": 176
+    },
+    {
+      "epoch": 0.04022727272727273,
+      "grad_norm": 2.056362755877933,
+      "learning_rate": 1.2498026183020626e-06,
+      "loss": 0.1096,
+      "step": 177
+    },
+    {
+      "epoch": 0.04045454545454545,
+      "grad_norm": 2.2804138898594504,
+      "learning_rate": 1.249800369075889e-06,
+      "loss": 0.1326,
+      "step": 178
+    },
+    {
+      "epoch": 0.04068181818181818,
+      "grad_norm": 2.2174483316912963,
+      "learning_rate": 1.2497981071089457e-06,
+      "loss": 0.077,
+      "step": 179
+    },
+    {
+      "epoch": 0.04090909090909091,
+      "grad_norm": 1.9239300653465918,
+      "learning_rate": 1.2497958324012791e-06,
+      "loss": 0.1238,
+      "step": 180
+    },
+    {
+      "epoch": 0.04113636363636364,
+      "grad_norm": 2.141703003193185,
+      "learning_rate": 1.2497935449529355e-06,
+      "loss": 0.1105,
+      "step": 181
+    },
+    {
+      "epoch": 0.041363636363636366,
+      "grad_norm": 3.308460116599906,
+      "learning_rate": 1.2497912447639616e-06,
+      "loss": 0.1365,
+      "step": 182
+    },
+    {
+      "epoch": 0.04159090909090909,
+      "grad_norm": 4.666500187910259,
+      "learning_rate": 1.249788931834404e-06,
+      "loss": 0.198,
+      "step": 183
+    },
+    {
+      "epoch": 0.04181818181818182,
+      "grad_norm": 1.615050602590198,
+      "learning_rate": 1.2497866061643102e-06,
+      "loss": 0.0575,
+      "step": 184
+    },
+    {
+      "epoch": 0.042045454545454546,
+      "grad_norm": 4.356514522295226,
+      "learning_rate": 1.2497842677537275e-06,
+      "loss": 0.1694,
+      "step": 185
+    },
+    {
+      "epoch": 0.042272727272727274,
+      "grad_norm": 1.920813600028758,
+      "learning_rate": 1.2497819166027035e-06,
+      "loss": 0.1375,
+      "step": 186
+    },
+    {
+      "epoch": 0.0425,
+      "grad_norm": 2.59940977102559,
+      "learning_rate": 1.2497795527112862e-06,
+      "loss": 0.1253,
+      "step": 187
+    },
+    {
+      "epoch": 0.042727272727272725,
+      "grad_norm": 2.356601275047971,
+      "learning_rate": 1.2497771760795238e-06,
+      "loss": 0.0752,
+      "step": 188
+    },
+    {
+      "epoch": 0.042954545454545454,
+      "grad_norm": 2.071449766141187,
+      "learning_rate": 1.2497747867074649e-06,
+      "loss": 0.0771,
+      "step": 189
+    },
+    {
+      "epoch": 0.04318181818181818,
+      "grad_norm": 2.401166790918426,
+      "learning_rate": 1.2497723845951582e-06,
+      "loss": 0.1061,
+      "step": 190
+    },
+    {
+      "epoch": 0.04340909090909091,
+      "grad_norm": 2.560214481755178,
+      "learning_rate": 1.2497699697426523e-06,
+      "loss": 0.1079,
+      "step": 191
+    },
+    {
+      "epoch": 0.04363636363636364,
+      "grad_norm": 4.281690250819271,
+      "learning_rate": 1.2497675421499971e-06,
+      "loss": 0.1465,
+      "step": 192
+    },
+    {
+      "epoch": 0.04386363636363636,
+      "grad_norm": 2.833431138910108,
+      "learning_rate": 1.2497651018172415e-06,
+      "loss": 0.0916,
+      "step": 193
+    },
+    {
+      "epoch": 0.04409090909090909,
+      "grad_norm": 2.2489099362984653,
+      "learning_rate": 1.2497626487444358e-06,
+      "loss": 0.0655,
+      "step": 194
+    },
+    {
+      "epoch": 0.04431818181818182,
+      "grad_norm": 3.5877419186353645,
+      "learning_rate": 1.2497601829316296e-06,
+      "loss": 0.1229,
+      "step": 195
+    },
+    {
+      "epoch": 0.04454545454545455,
+      "grad_norm": 1.2213501618802172,
+      "learning_rate": 1.2497577043788732e-06,
+      "loss": 0.0336,
+      "step": 196
+    },
+    {
+      "epoch": 0.04477272727272727,
+      "grad_norm": 4.076953702090065,
+      "learning_rate": 1.2497552130862173e-06,
+      "loss": 0.1612,
+      "step": 197
+    },
+    {
+      "epoch": 0.045,
+      "grad_norm": 1.563594770601146,
+      "learning_rate": 1.2497527090537126e-06,
+      "loss": 0.0695,
+      "step": 198
+    },
+    {
+      "epoch": 0.04522727272727273,
+      "grad_norm": 2.437574125766786,
+      "learning_rate": 1.2497501922814104e-06,
+      "loss": 0.1189,
+      "step": 199
+    },
+    {
+      "epoch": 0.045454545454545456,
+      "grad_norm": 2.7885626114056037,
+      "learning_rate": 1.249747662769362e-06,
+      "loss": 0.1224,
+      "step": 200
+    },
+    {
+      "epoch": 0.045681818181818185,
+      "grad_norm": 2.1744865890816962,
+      "learning_rate": 1.2497451205176183e-06,
+      "loss": 0.0911,
+      "step": 201
+    },
+    {
+      "epoch": 0.045909090909090906,
+      "grad_norm": 3.1319319844623803,
+      "learning_rate": 1.249742565526232e-06,
+      "loss": 0.1822,
+      "step": 202
+    },
+    {
+      "epoch": 0.046136363636363635,
+      "grad_norm": 3.3244537323424663,
+      "learning_rate": 1.2497399977952549e-06,
+      "loss": 0.1669,
+      "step": 203
+    },
+    {
+      "epoch": 0.046363636363636364,
+      "grad_norm": 2.028526820478898,
+      "learning_rate": 1.2497374173247393e-06,
+      "loss": 0.0709,
+      "step": 204
+    },
+    {
+      "epoch": 0.04659090909090909,
+      "grad_norm": 2.122511834513276,
+      "learning_rate": 1.2497348241147376e-06,
+      "loss": 0.0976,
+      "step": 205
+    },
+    {
+      "epoch": 0.04681818181818182,
+      "grad_norm": 3.840727210648857,
+      "learning_rate": 1.2497322181653032e-06,
+      "loss": 0.2493,
+      "step": 206
+    },
+    {
+      "epoch": 0.04704545454545454,
+      "grad_norm": 2.54371290019698,
+      "learning_rate": 1.2497295994764888e-06,
+      "loss": 0.085,
+      "step": 207
+    },
+    {
+      "epoch": 0.04727272727272727,
+      "grad_norm": 2.487488937050158,
+      "learning_rate": 1.249726968048348e-06,
+      "loss": 0.095,
+      "step": 208
+    },
+    {
+      "epoch": 0.0475,
+      "grad_norm": 1.934020143763007,
+      "learning_rate": 1.2497243238809345e-06,
+      "loss": 0.1052,
+      "step": 209
+    },
+    {
+      "epoch": 0.04772727272727273,
+      "grad_norm": 2.1609220039110566,
+      "learning_rate": 1.249721666974302e-06,
+      "loss": 0.066,
+      "step": 210
+    },
+    {
+      "epoch": 0.04795454545454545,
+      "grad_norm": 1.836340754105955,
+      "learning_rate": 1.249718997328505e-06,
+      "loss": 0.0892,
+      "step": 211
+    },
+    {
+      "epoch": 0.04818181818181818,
+      "grad_norm": 1.4928885032828618,
+      "learning_rate": 1.2497163149435978e-06,
+      "loss": 0.0715,
+      "step": 212
+    },
+    {
+      "epoch": 0.04840909090909091,
+      "grad_norm": 2.106951108086813,
+      "learning_rate": 1.2497136198196349e-06,
+      "loss": 0.0889,
+      "step": 213
+    },
+    {
+      "epoch": 0.04863636363636364,
+      "grad_norm": 2.5932231997284565,
+      "learning_rate": 1.2497109119566716e-06,
+      "loss": 0.1419,
+      "step": 214
+    },
+    {
+      "epoch": 0.048863636363636366,
+      "grad_norm": 1.833938053838939,
+      "learning_rate": 1.2497081913547626e-06,
+      "loss": 0.0558,
+      "step": 215
+    },
+    {
+      "epoch": 0.04909090909090909,
+      "grad_norm": 1.3858737024046333,
+      "learning_rate": 1.2497054580139642e-06,
+      "loss": 0.0773,
+      "step": 216
+    },
+    {
+      "epoch": 0.04931818181818182,
+      "grad_norm": 2.184973549763245,
+      "learning_rate": 1.2497027119343312e-06,
+      "loss": 0.0889,
+      "step": 217
+    },
+    {
+      "epoch": 0.049545454545454545,
+      "grad_norm": 2.0325673485953515,
+      "learning_rate": 1.2496999531159203e-06,
+      "loss": 0.0756,
+      "step": 218
+    },
+    {
+      "epoch": 0.049772727272727274,
+      "grad_norm": 2.8943515872478205,
+      "learning_rate": 1.2496971815587875e-06,
+      "loss": 0.1034,
+      "step": 219
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2961380345214608,
+      "learning_rate": 1.249694397262989e-06,
+      "loss": 0.0616,
+      "step": 220
+    },
+    {
+      "epoch": 0.050227272727272725,
+      "grad_norm": 2.064250131042833,
+      "learning_rate": 1.2496916002285823e-06,
+      "loss": 0.1022,
+      "step": 221
+    },
+    {
+      "epoch": 0.05045454545454545,
+      "grad_norm": 1.850902214887004,
+      "learning_rate": 1.249688790455624e-06,
+      "loss": 0.0863,
+      "step": 222
+    },
+    {
+      "epoch": 0.05068181818181818,
+      "grad_norm": 1.3601602707834528,
+      "learning_rate": 1.2496859679441712e-06,
+      "loss": 0.0917,
+      "step": 223
+    },
+    {
+      "epoch": 0.05090909090909091,
+      "grad_norm": 2.230653221740461,
+      "learning_rate": 1.2496831326942819e-06,
+      "loss": 0.0807,
+      "step": 224
+    },
+    {
+      "epoch": 0.05113636363636364,
+      "grad_norm": 2.2520045673700624,
+      "learning_rate": 1.2496802847060135e-06,
+      "loss": 0.1579,
+      "step": 225
+    },
+    {
+      "epoch": 0.05136363636363636,
+      "grad_norm": 1.7727065343386945,
+      "learning_rate": 1.2496774239794246e-06,
+      "loss": 0.0822,
+      "step": 226
+    },
+    {
+      "epoch": 0.05159090909090909,
+      "grad_norm": 1.7516674799106873,
+      "learning_rate": 1.249674550514573e-06,
+      "loss": 0.0647,
+      "step": 227
+    },
+    {
+      "epoch": 0.05181818181818182,
+      "grad_norm": 2.8655503636722295,
+      "learning_rate": 1.2496716643115175e-06,
+      "loss": 0.2004,
+      "step": 228
+    },
+    {
+      "epoch": 0.05204545454545455,
+      "grad_norm": 1.7737912421449948,
+      "learning_rate": 1.249668765370317e-06,
+      "loss": 0.073,
+      "step": 229
+    },
+    {
+      "epoch": 0.05227272727272727,
+      "grad_norm": 1.6990930798003503,
+      "learning_rate": 1.2496658536910306e-06,
+      "loss": 0.0511,
+      "step": 230
+    },
+    {
+      "epoch": 0.0525,
+      "grad_norm": 2.2407475052697343,
+      "learning_rate": 1.2496629292737176e-06,
+      "loss": 0.1562,
+      "step": 231
+    },
+    {
+      "epoch": 0.05272727272727273,
+      "grad_norm": 3.291267006622827,
+      "learning_rate": 1.2496599921184377e-06,
+      "loss": 0.202,
+      "step": 232
+    },
+    {
+      "epoch": 0.052954545454545456,
+      "grad_norm": 2.2439134739452107,
+      "learning_rate": 1.2496570422252507e-06,
+      "loss": 0.1301,
+      "step": 233
+    },
+    {
+      "epoch": 0.053181818181818184,
+      "grad_norm": 1.8166083071753045,
+      "learning_rate": 1.249654079594217e-06,
+      "loss": 0.1125,
+      "step": 234
+    },
+    {
+      "epoch": 0.053409090909090906,
+      "grad_norm": 2.7698145809319787,
+      "learning_rate": 1.249651104225397e-06,
+      "loss": 0.1622,
+      "step": 235
+    },
+    {
+      "epoch": 0.053636363636363635,
+      "grad_norm": 2.3797721932805755,
+      "learning_rate": 1.249648116118851e-06,
+      "loss": 0.1238,
+      "step": 236
+    },
+    {
+      "epoch": 0.053863636363636364,
+      "grad_norm": 3.518746249043222,
+      "learning_rate": 1.24964511527464e-06,
+      "loss": 0.1461,
+      "step": 237
+    },
+    {
+      "epoch": 0.05409090909090909,
+      "grad_norm": 4.3774666945579375,
+      "learning_rate": 1.2496421016928255e-06,
+      "loss": 0.2532,
+      "step": 238
+    },
+    {
+      "epoch": 0.05431818181818182,
+      "grad_norm": 2.621756361258652,
+      "learning_rate": 1.249639075373469e-06,
+      "loss": 0.1566,
+      "step": 239
+    },
+    {
+      "epoch": 0.05454545454545454,
+      "grad_norm": 2.2572357696726346,
+      "learning_rate": 1.2496360363166316e-06,
+      "loss": 0.1084,
+      "step": 240
+    },
+    {
+      "epoch": 0.05477272727272727,
+      "grad_norm": 1.4179540541089204,
+      "learning_rate": 1.2496329845223759e-06,
+      "loss": 0.0628,
+      "step": 241
+    },
+    {
+      "epoch": 0.055,
+      "grad_norm": 2.9698226750185763,
+      "learning_rate": 1.249629919990764e-06,
+      "loss": 0.093,
+      "step": 242
+    },
+    {
+      "epoch": 0.05522727272727273,
+      "grad_norm": 3.881399152203344,
+      "learning_rate": 1.2496268427218582e-06,
+      "loss": 0.1974,
+      "step": 243
+    },
+    {
+      "epoch": 0.05545454545454546,
+      "grad_norm": 2.3686058799447336,
+      "learning_rate": 1.2496237527157213e-06,
+      "loss": 0.1455,
+      "step": 244
+    },
+    {
+      "epoch": 0.05568181818181818,
+      "grad_norm": 1.7588162318161968,
+      "learning_rate": 1.2496206499724164e-06,
+      "loss": 0.0649,
+      "step": 245
+    },
+    {
+      "epoch": 0.05590909090909091,
+      "grad_norm": 1.3216028230327348,
+      "learning_rate": 1.2496175344920069e-06,
+      "loss": 0.0505,
+      "step": 246
+    },
+    {
+      "epoch": 0.05613636363636364,
+      "grad_norm": 2.6534613855097073,
+      "learning_rate": 1.249614406274556e-06,
+      "loss": 0.1076,
+      "step": 247
+    },
+    {
+      "epoch": 0.056363636363636366,
+      "grad_norm": 1.8415454162193927,
+      "learning_rate": 1.2496112653201276e-06,
+      "loss": 0.063,
+      "step": 248
+    },
+    {
+      "epoch": 0.05659090909090909,
+      "grad_norm": 1.8881944925712033,
+      "learning_rate": 1.249608111628786e-06,
+      "loss": 0.0639,
+      "step": 249
+    },
+    {
+      "epoch": 0.056818181818181816,
+      "grad_norm": 1.7770675986491005,
+      "learning_rate": 1.249604945200595e-06,
+      "loss": 0.1162,
+      "step": 250
+    },
+    {
+      "epoch": 0.057045454545454545,
+      "grad_norm": 2.468949611953055,
+      "learning_rate": 1.24960176603562e-06,
+      "loss": 0.1179,
+      "step": 251
+    },
+    {
+      "epoch": 0.057272727272727274,
+      "grad_norm": 1.5183145567460188,
+      "learning_rate": 1.249598574133925e-06,
+      "loss": 0.0823,
+      "step": 252
+    },
+    {
+      "epoch": 0.0575,
+      "grad_norm": 2.289293046356758,
+      "learning_rate": 1.2495953694955754e-06,
+      "loss": 0.1429,
+      "step": 253
+    },
+    {
+      "epoch": 0.057727272727272724,
+      "grad_norm": 1.4273193183639814,
+      "learning_rate": 1.2495921521206365e-06,
+      "loss": 0.0668,
+      "step": 254
+    },
+    {
+      "epoch": 0.05795454545454545,
+      "grad_norm": 2.185478688432408,
+      "learning_rate": 1.2495889220091741e-06,
+      "loss": 0.0818,
+      "step": 255
+    },
+    {
+      "epoch": 0.05818181818181818,
+      "grad_norm": 2.1628290443555547,
+      "learning_rate": 1.2495856791612538e-06,
+      "loss": 0.1045,
+      "step": 256
+    },
+    {
+      "epoch": 0.05840909090909091,
+      "grad_norm": 3.835421373612996,
+      "learning_rate": 1.2495824235769418e-06,
+      "loss": 0.1871,
+      "step": 257
+    },
+    {
+      "epoch": 0.05863636363636364,
+      "grad_norm": 3.02557031908882,
+      "learning_rate": 1.2495791552563047e-06,
+      "loss": 0.1459,
+      "step": 258
+    },
+    {
+      "epoch": 0.05886363636363636,
+      "grad_norm": 1.595722716155942,
+      "learning_rate": 1.249575874199409e-06,
+      "loss": 0.0814,
+      "step": 259
+    },
+    {
+      "epoch": 0.05909090909090909,
+      "grad_norm": 1.7519648568883437,
+      "learning_rate": 1.2495725804063214e-06,
+      "loss": 0.088,
+      "step": 260
+    },
+    {
+      "epoch": 0.05931818181818182,
+      "grad_norm": 2.2461777695706298,
+      "learning_rate": 1.2495692738771095e-06,
+      "loss": 0.0949,
+      "step": 261
+    },
+    {
+      "epoch": 0.05954545454545455,
+      "grad_norm": 2.1337077789558174,
+      "learning_rate": 1.2495659546118403e-06,
+      "loss": 0.0704,
+      "step": 262
+    },
+    {
+      "epoch": 0.059772727272727276,
+      "grad_norm": 2.7676472300076225,
+      "learning_rate": 1.2495626226105818e-06,
+      "loss": 0.1417,
+      "step": 263
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.03430481792065,
+      "learning_rate": 1.2495592778734018e-06,
+      "loss": 0.1544,
+      "step": 264
+    },
+    {
+      "epoch": 0.060227272727272727,
+      "grad_norm": 3.5722507549479112,
+      "learning_rate": 1.2495559204003684e-06,
+      "loss": 0.1448,
+      "step": 265
+    },
+    {
+      "epoch": 0.060454545454545455,
+      "grad_norm": 3.0034103142307305,
+      "learning_rate": 1.2495525501915503e-06,
+      "loss": 0.1265,
+      "step": 266
+    },
+    {
+      "epoch": 0.060681818181818184,
+      "grad_norm": 1.7429565189024705,
+      "learning_rate": 1.249549167247016e-06,
+      "loss": 0.07,
+      "step": 267
+    },
+    {
+      "epoch": 0.060909090909090906,
+      "grad_norm": 1.4823419588559634,
+      "learning_rate": 1.2495457715668346e-06,
+      "loss": 0.0548,
+      "step": 268
+    },
+    {
+      "epoch": 0.061136363636363635,
+      "grad_norm": 2.057498073957767,
+      "learning_rate": 1.2495423631510754e-06,
+      "loss": 0.0812,
+      "step": 269
+    },
+    {
+      "epoch": 0.06136363636363636,
+      "grad_norm": 1.62543310884278,
+      "learning_rate": 1.2495389419998078e-06,
+      "loss": 0.0693,
+      "step": 270
+    },
+    {
+      "epoch": 0.06159090909090909,
+      "grad_norm": 1.6029652911512395,
+      "learning_rate": 1.2495355081131017e-06,
+      "loss": 0.0676,
+      "step": 271
+    },
+    {
+      "epoch": 0.06181818181818182,
+      "grad_norm": 2.417231007500269,
+      "learning_rate": 1.249532061491027e-06,
+      "loss": 0.149,
+      "step": 272
+    },
+    {
+      "epoch": 0.06204545454545454,
+      "grad_norm": 1.6426200254134593,
+      "learning_rate": 1.249528602133654e-06,
+      "loss": 0.1005,
+      "step": 273
+    },
+    {
+      "epoch": 0.06227272727272727,
+      "grad_norm": 1.9082375656288766,
+      "learning_rate": 1.2495251300410533e-06,
+      "loss": 0.0775,
+      "step": 274
+    },
+    {
+      "epoch": 0.0625,
+      "grad_norm": 2.688489473925319,
+      "learning_rate": 1.2495216452132955e-06,
+      "loss": 0.1288,
+      "step": 275
+    },
+    {
+      "epoch": 0.06272727272727273,
+      "grad_norm": 1.6938834169152557,
+      "learning_rate": 1.249518147650452e-06,
+      "loss": 0.1176,
+      "step": 276
+    },
+    {
+      "epoch": 0.06295454545454546,
+      "grad_norm": 1.5452245399163091,
+      "learning_rate": 1.2495146373525937e-06,
+      "loss": 0.0643,
+      "step": 277
+    },
+    {
+      "epoch": 0.06318181818181819,
+      "grad_norm": 1.9180859307575704,
+      "learning_rate": 1.2495111143197928e-06,
+      "loss": 0.1233,
+      "step": 278
+    },
+    {
+      "epoch": 0.06340909090909091,
+      "grad_norm": 3.422030689922406,
+      "learning_rate": 1.2495075785521205e-06,
+      "loss": 0.1397,
+      "step": 279
+    },
+    {
+      "epoch": 0.06363636363636363,
+      "grad_norm": 3.531080114882214,
+      "learning_rate": 1.249504030049649e-06,
+      "loss": 0.2417,
+      "step": 280
+    },
+    {
+      "epoch": 0.06386363636363636,
+      "grad_norm": 1.9743055635290963,
+      "learning_rate": 1.249500468812451e-06,
+      "loss": 0.1225,
+      "step": 281
+    },
+    {
+      "epoch": 0.06409090909090909,
+      "grad_norm": 6.329996408994021,
+      "learning_rate": 1.2494968948405991e-06,
+      "loss": 0.1533,
+      "step": 282
+    },
+    {
+      "epoch": 0.06431818181818182,
+      "grad_norm": 1.5755788536122424,
+      "learning_rate": 1.2494933081341657e-06,
+      "loss": 0.0642,
+      "step": 283
+    },
+    {
+      "epoch": 0.06454545454545454,
+      "grad_norm": 1.3221463441180554,
+      "learning_rate": 1.2494897086932245e-06,
+      "loss": 0.0537,
+      "step": 284
+    },
+    {
+      "epoch": 0.06477272727272727,
+      "grad_norm": 1.603048514173173,
+      "learning_rate": 1.2494860965178484e-06,
+      "loss": 0.0635,
+      "step": 285
+    },
+    {
+      "epoch": 0.065,
+      "grad_norm": 1.8766919133194477,
+      "learning_rate": 1.2494824716081117e-06,
+      "loss": 0.1239,
+      "step": 286
+    },
+    {
+      "epoch": 0.06522727272727273,
+      "grad_norm": 1.721169250802494,
+      "learning_rate": 1.2494788339640877e-06,
+      "loss": 0.0992,
+      "step": 287
+    },
+    {
+      "epoch": 0.06545454545454546,
+      "grad_norm": 1.7399497704884956,
+      "learning_rate": 1.249475183585851e-06,
+      "loss": 0.078,
+      "step": 288
+    },
+    {
+      "epoch": 0.06568181818181819,
+      "grad_norm": 2.1086139954023975,
+      "learning_rate": 1.2494715204734756e-06,
+      "loss": 0.1419,
+      "step": 289
+    },
+    {
+      "epoch": 0.0659090909090909,
+      "grad_norm": 1.4689280794821082,
+      "learning_rate": 1.2494678446270364e-06,
+      "loss": 0.066,
+      "step": 290
+    },
+    {
+      "epoch": 0.06613636363636363,
+      "grad_norm": 1.325957734379718,
+      "learning_rate": 1.2494641560466087e-06,
+      "loss": 0.0566,
+      "step": 291
+    },
+    {
+      "epoch": 0.06636363636363636,
+      "grad_norm": 1.9330279629845104,
+      "learning_rate": 1.2494604547322672e-06,
+      "loss": 0.0868,
+      "step": 292
+    },
+    {
+      "epoch": 0.06659090909090909,
+      "grad_norm": 1.4241917606866943,
+      "learning_rate": 1.2494567406840876e-06,
+      "loss": 0.0743,
+      "step": 293
+    },
+    {
+      "epoch": 0.06681818181818182,
+      "grad_norm": 1.6497663657773214,
+      "learning_rate": 1.2494530139021458e-06,
+      "loss": 0.1211,
+      "step": 294
+    },
+    {
+      "epoch": 0.06704545454545455,
+      "grad_norm": 2.544547282033746,
+      "learning_rate": 1.2494492743865175e-06,
+      "loss": 0.1363,
+      "step": 295
+    },
+    {
+      "epoch": 0.06727272727272728,
+      "grad_norm": 1.8853504912496688,
+      "learning_rate": 1.249445522137279e-06,
+      "loss": 0.0693,
+      "step": 296
+    },
+    {
+      "epoch": 0.0675,
+      "grad_norm": 2.033522068542789,
+      "learning_rate": 1.249441757154507e-06,
+      "loss": 0.087,
+      "step": 297
+    },
+    {
+      "epoch": 0.06772727272727273,
+      "grad_norm": 2.511635736728221,
+      "learning_rate": 1.2494379794382782e-06,
+      "loss": 0.0785,
+      "step": 298
+    },
+    {
+      "epoch": 0.06795454545454545,
+      "grad_norm": 2.5268951051301385,
+      "learning_rate": 1.2494341889886695e-06,
+      "loss": 0.0839,
+      "step": 299
+    },
+    {
+      "epoch": 0.06818181818181818,
+      "grad_norm": 2.6520145374605635,
+      "learning_rate": 1.2494303858057586e-06,
+      "loss": 0.0872,
+      "step": 300
+    },
+    {
+      "epoch": 0.0684090909090909,
+      "grad_norm": 2.021469163968432,
+      "learning_rate": 1.2494265698896224e-06,
+      "loss": 0.0727,
+      "step": 301
+    },
+    {
+      "epoch": 0.06863636363636363,
+      "grad_norm": 4.526870661578714,
+      "learning_rate": 1.2494227412403393e-06,
+      "loss": 0.1981,
+      "step": 302
+    },
+    {
+      "epoch": 0.06886363636363636,
+      "grad_norm": 2.992078953455472,
+      "learning_rate": 1.249418899857987e-06,
+      "loss": 0.1463,
+      "step": 303
+    },
+    {
+      "epoch": 0.06909090909090909,
+      "grad_norm": 1.6114409274993262,
+      "learning_rate": 1.249415045742644e-06,
+      "loss": 0.0693,
+      "step": 304
+    },
+    {
+      "epoch": 0.06931818181818182,
+      "grad_norm": 3.112326641009545,
+      "learning_rate": 1.2494111788943888e-06,
+      "loss": 0.1525,
+      "step": 305
+    },
+    {
+      "epoch": 0.06954545454545455,
+      "grad_norm": 3.852673616390004,
+      "learning_rate": 1.2494072993133005e-06,
+      "loss": 0.1425,
+      "step": 306
+    },
+    {
+      "epoch": 0.06977272727272728,
+      "grad_norm": 3.363870796226044,
+      "learning_rate": 1.249403406999458e-06,
+      "loss": 0.1702,
+      "step": 307
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5549628930307637,
+      "learning_rate": 1.2493995019529407e-06,
+      "loss": 0.1104,
+      "step": 308
+    },
+    {
+      "epoch": 0.07022727272727272,
+      "grad_norm": 1.9630557632175043,
+      "learning_rate": 1.249395584173828e-06,
+      "loss": 0.1198,
+      "step": 309
+    },
+    {
+      "epoch": 0.07045454545454545,
+      "grad_norm": 2.9997206968488346,
+      "learning_rate": 1.2493916536622004e-06,
+      "loss": 0.2934,
+      "step": 310
+    },
+    {
+      "epoch": 0.07068181818181818,
+      "grad_norm": 2.1882411509428783,
+      "learning_rate": 1.2493877104181373e-06,
+      "loss": 0.11,
+      "step": 311
+    },
+    {
+      "epoch": 0.07090909090909091,
+      "grad_norm": 2.660023917098737,
+      "learning_rate": 1.2493837544417196e-06,
+      "loss": 0.0983,
+      "step": 312
+    },
+    {
+      "epoch": 0.07113636363636364,
+      "grad_norm": 2.0245673776777497,
+      "learning_rate": 1.249379785733028e-06,
+      "loss": 0.1159,
+      "step": 313
+    },
+    {
+      "epoch": 0.07136363636363637,
+      "grad_norm": 1.2395625498085538,
+      "learning_rate": 1.249375804292143e-06,
+      "loss": 0.0448,
+      "step": 314
+    },
+    {
+      "epoch": 0.0715909090909091,
+      "grad_norm": 1.6489119473071414,
+      "learning_rate": 1.2493718101191463e-06,
+      "loss": 0.0823,
+      "step": 315
+    },
+    {
+      "epoch": 0.07181818181818182,
+      "grad_norm": 2.067768279223006,
+      "learning_rate": 1.249367803214119e-06,
+      "loss": 0.1403,
+      "step": 316
+    },
+    {
+      "epoch": 0.07204545454545455,
+      "grad_norm": 1.4516326091445901,
+      "learning_rate": 1.249363783577143e-06,
+      "loss": 0.1027,
+      "step": 317
+    },
+    {
+      "epoch": 0.07227272727272727,
+      "grad_norm": 1.2447459964332626,
+      "learning_rate": 1.2493597512083003e-06,
+      "loss": 0.0596,
+      "step": 318
+    },
+    {
+      "epoch": 0.0725,
+      "grad_norm": 1.751177854137079,
+      "learning_rate": 1.2493557061076728e-06,
+      "loss": 0.1171,
+      "step": 319
+    },
+    {
+      "epoch": 0.07272727272727272,
+      "grad_norm": 1.6379916914670452,
+      "learning_rate": 1.2493516482753433e-06,
+      "loss": 0.091,
+      "step": 320
+    },
+    {
+      "epoch": 0.07295454545454545,
+      "grad_norm": 2.0179733919246323,
+      "learning_rate": 1.2493475777113945e-06,
+      "loss": 0.1143,
+      "step": 321
+    },
+    {
+      "epoch": 0.07318181818181818,
+      "grad_norm": 2.4536650793607655,
+      "learning_rate": 1.2493434944159091e-06,
+      "loss": 0.0919,
+      "step": 322
+    },
+    {
+      "epoch": 0.07340909090909091,
+      "grad_norm": 4.102261212867276,
+      "learning_rate": 1.249339398388971e-06,
+      "loss": 0.151,
+      "step": 323
+    },
+    {
+      "epoch": 0.07363636363636364,
+      "grad_norm": 2.261350484063778,
+      "learning_rate": 1.249335289630663e-06,
+      "loss": 0.1067,
+      "step": 324
+    },
+    {
+      "epoch": 0.07386363636363637,
+      "grad_norm": 2.305596537745846,
+      "learning_rate": 1.2493311681410693e-06,
+      "loss": 0.0992,
+      "step": 325
+    },
+    {
+      "epoch": 0.0740909090909091,
+      "grad_norm": 1.5740044945396432,
+      "learning_rate": 1.2493270339202742e-06,
+      "loss": 0.0736,
+      "step": 326
+    },
+    {
+      "epoch": 0.07431818181818182,
+      "grad_norm": 3.010931148003246,
+      "learning_rate": 1.2493228869683614e-06,
+      "loss": 0.087,
+      "step": 327
+    },
+    {
+      "epoch": 0.07454545454545454,
+      "grad_norm": 2.275280171329199,
+      "learning_rate": 1.2493187272854157e-06,
+      "loss": 0.1242,
+      "step": 328
+    },
+    {
+      "epoch": 0.07477272727272727,
+      "grad_norm": 1.3399237749250856,
+      "learning_rate": 1.249314554871522e-06,
+      "loss": 0.0745,
+      "step": 329
+    },
+    {
+      "epoch": 0.075,
+      "grad_norm": 2.004403419887356,
+      "learning_rate": 1.2493103697267656e-06,
+      "loss": 0.1145,
+      "step": 330
+    },
+    {
+      "epoch": 0.07522727272727273,
+      "grad_norm": 0.5652958377820397,
+      "learning_rate": 1.2493061718512314e-06,
+      "loss": 0.0309,
+      "step": 331
+    },
+    {
+      "epoch": 0.07545454545454545,
+      "grad_norm": 4.421367957310514,
+      "learning_rate": 1.2493019612450052e-06,
+      "loss": 0.2385,
+      "step": 332
+    },
+    {
+      "epoch": 0.07568181818181818,
+      "grad_norm": 1.1151229774385345,
+      "learning_rate": 1.2492977379081728e-06,
+      "loss": 0.0704,
+      "step": 333
+    },
+    {
+      "epoch": 0.07590909090909091,
+      "grad_norm": 1.6690937076560421,
+      "learning_rate": 1.2492935018408203e-06,
+      "loss": 0.0832,
+      "step": 334
+    },
+    {
+      "epoch": 0.07613636363636364,
+      "grad_norm": 2.408642072429063,
+      "learning_rate": 1.2492892530430343e-06,
+      "loss": 0.1231,
+      "step": 335
+    },
+    {
+      "epoch": 0.07636363636363637,
+      "grad_norm": 2.2150928876057567,
+      "learning_rate": 1.2492849915149013e-06,
+      "loss": 0.0871,
+      "step": 336
+    },
+    {
+      "epoch": 0.07659090909090908,
+      "grad_norm": 1.965635410961706,
+      "learning_rate": 1.2492807172565083e-06,
+      "loss": 0.0676,
+      "step": 337
+    },
+    {
+      "epoch": 0.07681818181818181,
+      "grad_norm": 1.228114166565457,
+      "learning_rate": 1.2492764302679422e-06,
+      "loss": 0.0683,
+      "step": 338
+    },
+    {
+      "epoch": 0.07704545454545454,
+      "grad_norm": 1.3795201506306647,
+      "learning_rate": 1.2492721305492905e-06,
+      "loss": 0.1095,
+      "step": 339
+    },
+    {
+      "epoch": 0.07727272727272727,
+      "grad_norm": 2.135634787327226,
+      "learning_rate": 1.2492678181006413e-06,
+      "loss": 0.1277,
+      "step": 340
+    },
+    {
+      "epoch": 0.0775,
+      "grad_norm": 1.5650111750467035,
+      "learning_rate": 1.2492634929220817e-06,
+      "loss": 0.068,
+      "step": 341
+    },
+    {
+      "epoch": 0.07772727272727273,
+      "grad_norm": 1.4958149919063035,
+      "learning_rate": 1.2492591550137008e-06,
+      "loss": 0.0752,
+      "step": 342
+    },
+    {
+      "epoch": 0.07795454545454546,
+      "grad_norm": 2.693458042828727,
+      "learning_rate": 1.2492548043755865e-06,
+      "loss": 0.1313,
+      "step": 343
+    },
+    {
+      "epoch": 0.07818181818181819,
+      "grad_norm": 2.21310548626628,
+      "learning_rate": 1.2492504410078276e-06,
+      "loss": 0.0795,
+      "step": 344
+    },
+    {
+      "epoch": 0.07840909090909091,
+      "grad_norm": 2.9116796541195082,
+      "learning_rate": 1.2492460649105133e-06,
+      "loss": 0.0816,
+      "step": 345
+    },
+    {
+      "epoch": 0.07863636363636364,
+      "grad_norm": 4.843344615772008,
+      "learning_rate": 1.2492416760837326e-06,
+      "loss": 0.1977,
+      "step": 346
+    },
+    {
+      "epoch": 0.07886363636363636,
+      "grad_norm": 3.2843856508241602,
+      "learning_rate": 1.249237274527575e-06,
+      "loss": 0.1879,
+      "step": 347
+    },
+    {
+      "epoch": 0.07909090909090909,
+      "grad_norm": 1.9860680300047149,
+      "learning_rate": 1.2492328602421305e-06,
+      "loss": 0.0758,
+      "step": 348
+    },
+    {
+      "epoch": 0.07931818181818182,
+      "grad_norm": 2.498991326391029,
+      "learning_rate": 1.2492284332274886e-06,
+      "loss": 0.0986,
+      "step": 349
+    },
+    {
+      "epoch": 0.07954545454545454,
+      "grad_norm": 2.929082591485183,
+      "learning_rate": 1.2492239934837405e-06,
+      "loss": 0.043,
+      "step": 350
+    },
+    {
+      "epoch": 0.07977272727272727,
+      "grad_norm": 1.5416351149362961,
+      "learning_rate": 1.2492195410109757e-06,
+      "loss": 0.0802,
+      "step": 351
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.192246458093708,
+      "learning_rate": 1.2492150758092855e-06,
+      "loss": 0.0844,
+      "step": 352
+    },
+    {
+      "epoch": 0.08022727272727273,
+      "grad_norm": 2.305333979771203,
+      "learning_rate": 1.2492105978787612e-06,
+      "loss": 0.1367,
+      "step": 353
+    },
+    {
+      "epoch": 0.08045454545454546,
+      "grad_norm": 1.15168439037637,
+      "learning_rate": 1.2492061072194934e-06,
+      "loss": 0.0818,
+      "step": 354
+    },
+    {
+      "epoch": 0.08068181818181819,
+      "grad_norm": 1.8348080472113009,
+      "learning_rate": 1.2492016038315744e-06,
+      "loss": 0.0929,
+      "step": 355
+    },
+    {
+      "epoch": 0.0809090909090909,
+      "grad_norm": 1.812954505649971,
+      "learning_rate": 1.2491970877150955e-06,
+      "loss": 0.0558,
+      "step": 356
+    },
+    {
+      "epoch": 0.08113636363636363,
+      "grad_norm": 2.0036065037903303,
+      "learning_rate": 1.249192558870149e-06,
+      "loss": 0.1782,
+      "step": 357
+    },
+    {
+      "epoch": 0.08136363636363636,
+      "grad_norm": 0.49665711616586355,
+      "learning_rate": 1.2491880172968273e-06,
+      "loss": 0.0273,
+      "step": 358
+    },
+    {
+      "epoch": 0.08159090909090909,
+      "grad_norm": 2.4198817632804266,
+      "learning_rate": 1.2491834629952232e-06,
+      "loss": 0.0853,
+      "step": 359
+    },
+    {
+      "epoch": 0.08181818181818182,
+      "grad_norm": 1.2744068138179512,
+      "learning_rate": 1.2491788959654291e-06,
+      "loss": 0.0717,
+      "step": 360
+    },
+    {
+      "epoch": 0.08204545454545455,
+      "grad_norm": 1.9693996198620667,
+      "learning_rate": 1.2491743162075384e-06,
+      "loss": 0.0876,
+      "step": 361
+    },
+    {
+      "epoch": 0.08227272727272728,
+      "grad_norm": 1.7490012322803337,
+      "learning_rate": 1.2491697237216446e-06,
+      "loss": 0.0944,
+      "step": 362
+    },
+    {
+      "epoch": 0.0825,
+      "grad_norm": 1.4396946996415372,
+      "learning_rate": 1.249165118507841e-06,
+      "loss": 0.0624,
+      "step": 363
+    },
+    {
+      "epoch": 0.08272727272727273,
+      "grad_norm": 2.428237901476937,
+      "learning_rate": 1.249160500566222e-06,
+      "loss": 0.1257,
+      "step": 364
+    },
+    {
+      "epoch": 0.08295454545454546,
+      "grad_norm": 3.027103196596242,
+      "learning_rate": 1.2491558698968812e-06,
+      "loss": 0.1036,
+      "step": 365
+    },
+    {
+      "epoch": 0.08318181818181818,
+      "grad_norm": 2.1011172447307893,
+      "learning_rate": 1.2491512264999135e-06,
+      "loss": 0.0827,
+      "step": 366
+    },
+    {
+      "epoch": 0.0834090909090909,
+      "grad_norm": 1.6700634379687471,
+      "learning_rate": 1.2491465703754132e-06,
+      "loss": 0.0853,
+      "step": 367
+    },
+    {
+      "epoch": 0.08363636363636363,
+      "grad_norm": 1.9102430881848167,
+      "learning_rate": 1.2491419015234758e-06,
+      "loss": 0.0666,
+      "step": 368
+    },
+    {
+      "epoch": 0.08386363636363636,
+      "grad_norm": 2.441496730306747,
+      "learning_rate": 1.249137219944196e-06,
+      "loss": 0.1171,
+      "step": 369
+    },
+    {
+      "epoch": 0.08409090909090909,
+      "grad_norm": 2.132920783325472,
+      "learning_rate": 1.2491325256376692e-06,
+      "loss": 0.0866,
+      "step": 370
+    },
+    {
+      "epoch": 0.08431818181818182,
+      "grad_norm": 2.678956053647621,
+      "learning_rate": 1.2491278186039916e-06,
+      "loss": 0.1209,
+      "step": 371
+    },
+    {
+      "epoch": 0.08454545454545455,
+      "grad_norm": 2.6043973428610814,
+      "learning_rate": 1.2491230988432587e-06,
+      "loss": 0.1453,
+      "step": 372
+    },
+    {
+      "epoch": 0.08477272727272728,
+      "grad_norm": 2.264048836729903,
+      "learning_rate": 1.2491183663555672e-06,
+      "loss": 0.1248,
+      "step": 373
+    },
+    {
+      "epoch": 0.085,
+      "grad_norm": 1.6554748542467863,
+      "learning_rate": 1.2491136211410132e-06,
+      "loss": 0.0675,
+      "step": 374
+    },
+    {
+      "epoch": 0.08522727272727272,
+      "grad_norm": 1.939357010974732,
+      "learning_rate": 1.2491088631996936e-06,
+      "loss": 0.0617,
+      "step": 375
+    },
+    {
+      "epoch": 0.08545454545454545,
+      "grad_norm": 0.9088713413667205,
+      "learning_rate": 1.2491040925317057e-06,
+      "loss": 0.0498,
+      "step": 376
+    },
+    {
+      "epoch": 0.08568181818181818,
+      "grad_norm": 2.3258356022361566,
+      "learning_rate": 1.2490993091371465e-06,
+      "loss": 0.1266,
+      "step": 377
+    },
+    {
+      "epoch": 0.08590909090909091,
+      "grad_norm": 2.3196919783367607,
+      "learning_rate": 1.2490945130161132e-06,
+      "loss": 0.1229,
+      "step": 378
+    },
+    {
+      "epoch": 0.08613636363636364,
+      "grad_norm": 2.301449624879497,
+      "learning_rate": 1.2490897041687044e-06,
+      "loss": 0.0805,
+      "step": 379
+    },
+    {
+      "epoch": 0.08636363636363636,
+      "grad_norm": 2.1068291179107654,
+      "learning_rate": 1.2490848825950178e-06,
+      "loss": 0.0679,
+      "step": 380
+    },
+    {
+      "epoch": 0.0865909090909091,
+      "grad_norm": 3.2855879297396533,
+      "learning_rate": 1.2490800482951515e-06,
+      "loss": 0.1691,
+      "step": 381
+    },
+    {
+      "epoch": 0.08681818181818182,
+      "grad_norm": 0.8307041198411539,
+      "learning_rate": 1.2490752012692042e-06,
+      "loss": 0.0486,
+      "step": 382
+    },
+    {
+      "epoch": 0.08704545454545455,
+      "grad_norm": 1.9504152700679762,
+      "learning_rate": 1.2490703415172747e-06,
+      "loss": 0.0877,
+      "step": 383
+    },
+    {
+      "epoch": 0.08727272727272728,
+      "grad_norm": 1.7503539406453896,
+      "learning_rate": 1.2490654690394624e-06,
+      "loss": 0.0485,
+      "step": 384
+    },
+    {
+      "epoch": 0.0875,
+      "grad_norm": 7.689462919603509,
+      "learning_rate": 1.2490605838358664e-06,
+      "loss": 0.2549,
+      "step": 385
+    },
+    {
+      "epoch": 0.08772727272727272,
+      "grad_norm": 1.7694890409967263,
+      "learning_rate": 1.2490556859065865e-06,
+      "loss": 0.0557,
+      "step": 386
+    },
+    {
+      "epoch": 0.08795454545454545,
+      "grad_norm": 1.8921383534682272,
+      "learning_rate": 1.2490507752517222e-06,
+      "loss": 0.0772,
+      "step": 387
+    },
+    {
+      "epoch": 0.08818181818181818,
+      "grad_norm": 1.4181747040167698,
+      "learning_rate": 1.249045851871374e-06,
+      "loss": 0.0603,
+      "step": 388
+    },
+    {
+      "epoch": 0.08840909090909091,
+      "grad_norm": 2.494181352909692,
+      "learning_rate": 1.2490409157656419e-06,
+      "loss": 0.1143,
+      "step": 389
+    },
+    {
+      "epoch": 0.08863636363636364,
+      "grad_norm": 1.4846798394521876,
+      "learning_rate": 1.2490359669346272e-06,
+      "loss": 0.0843,
+      "step": 390
+    },
+    {
+      "epoch": 0.08886363636363637,
+      "grad_norm": 2.8474414867631967,
+      "learning_rate": 1.2490310053784301e-06,
+      "loss": 0.1088,
+      "step": 391
+    },
+    {
+      "epoch": 0.0890909090909091,
+      "grad_norm": 1.6962328662105772,
+      "learning_rate": 1.2490260310971525e-06,
+      "loss": 0.0524,
+      "step": 392
+    },
+    {
+      "epoch": 0.08931818181818182,
+      "grad_norm": 2.3264152884475244,
+      "learning_rate": 1.249021044090895e-06,
+      "loss": 0.1216,
+      "step": 393
+    },
+    {
+      "epoch": 0.08954545454545454,
+      "grad_norm": 1.932945944497157,
+      "learning_rate": 1.24901604435976e-06,
+      "loss": 0.0808,
+      "step": 394
+    },
+    {
+      "epoch": 0.08977272727272727,
+      "grad_norm": 2.8474968957602877,
+      "learning_rate": 1.2490110319038495e-06,
+      "loss": 0.0836,
+      "step": 395
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8549668024983692,
+      "learning_rate": 1.249006006723265e-06,
+      "loss": 0.0736,
+      "step": 396
+    },
+    {
+      "epoch": 0.09022727272727273,
+      "grad_norm": 1.2646492907664195,
+      "learning_rate": 1.2490009688181095e-06,
+      "loss": 0.0845,
+      "step": 397
+    },
+    {
+      "epoch": 0.09045454545454545,
+      "grad_norm": 1.5603753080301555,
+      "learning_rate": 1.2489959181884856e-06,
+      "loss": 0.048,
+      "step": 398
+    },
+    {
+      "epoch": 0.09068181818181818,
+      "grad_norm": 2.235766401294709,
+      "learning_rate": 1.2489908548344965e-06,
+      "loss": 0.1544,
+      "step": 399
+    },
+    {
+      "epoch": 0.09090909090909091,
+      "grad_norm": 2.3647460204481288,
+      "learning_rate": 1.2489857787562449e-06,
+      "loss": 0.0635,
+      "step": 400
+    },
+    {
+      "epoch": 0.09113636363636364,
+      "grad_norm": 2.6570408996262365,
+      "learning_rate": 1.2489806899538349e-06,
+      "loss": 0.0755,
+      "step": 401
+    },
+    {
+      "epoch": 0.09136363636363637,
+      "grad_norm": 2.006062494494428,
+      "learning_rate": 1.24897558842737e-06,
+      "loss": 0.0807,
+      "step": 402
+    },
+    {
+      "epoch": 0.0915909090909091,
+      "grad_norm": 2.0185772553209707,
+      "learning_rate": 1.2489704741769544e-06,
+      "loss": 0.0828,
+      "step": 403
+    },
+    {
+      "epoch": 0.09181818181818181,
+      "grad_norm": 3.668694111134318,
+      "learning_rate": 1.2489653472026918e-06,
+      "loss": 0.1996,
+      "step": 404
+    },
+    {
+      "epoch": 0.09204545454545454,
+      "grad_norm": 2.1474795644397933,
+      "learning_rate": 1.2489602075046878e-06,
+      "loss": 0.1019,
+      "step": 405
+    },
+    {
+      "epoch": 0.09227272727272727,
+      "grad_norm": 2.0570867946879328,
+      "learning_rate": 1.2489550550830462e-06,
+      "loss": 0.1094,
+      "step": 406
+    },
+    {
+      "epoch": 0.0925,
+      "grad_norm": 2.999018427666295,
+      "learning_rate": 1.2489498899378726e-06,
+      "loss": 0.1635,
+      "step": 407
+    },
+    {
+      "epoch": 0.09272727272727273,
+      "grad_norm": 1.8644699222539551,
+      "learning_rate": 1.248944712069272e-06,
+      "loss": 0.0774,
+      "step": 408
+    },
+    {
+      "epoch": 0.09295454545454546,
+      "grad_norm": 2.0985351627750797,
+      "learning_rate": 1.2489395214773506e-06,
+      "loss": 0.0857,
+      "step": 409
+    },
+    {
+      "epoch": 0.09318181818181819,
+      "grad_norm": 2.111001580370495,
+      "learning_rate": 1.2489343181622138e-06,
+      "loss": 0.0839,
+      "step": 410
+    },
+    {
+      "epoch": 0.09340909090909091,
+      "grad_norm": 3.194644411242909,
+      "learning_rate": 1.2489291021239674e-06,
+      "loss": 0.1522,
+      "step": 411
+    },
+    {
+      "epoch": 0.09363636363636364,
+      "grad_norm": 1.7243339822687709,
+      "learning_rate": 1.2489238733627183e-06,
+      "loss": 0.0731,
+      "step": 412
+    },
+    {
+      "epoch": 0.09386363636363636,
+      "grad_norm": 1.8755445402497195,
+      "learning_rate": 1.2489186318785728e-06,
+      "loss": 0.0634,
+      "step": 413
+    },
+    {
+      "epoch": 0.09409090909090909,
+      "grad_norm": 2.5222348523581846,
+      "learning_rate": 1.248913377671638e-06,
+      "loss": 0.1243,
+      "step": 414
+    },
+    {
+      "epoch": 0.09431818181818181,
+      "grad_norm": 2.3693093268572056,
+      "learning_rate": 1.2489081107420212e-06,
+      "loss": 0.0868,
+      "step": 415
+    },
+    {
+      "epoch": 0.09454545454545454,
+      "grad_norm": 2.518360587827347,
+      "learning_rate": 1.2489028310898293e-06,
+      "loss": 0.1352,
+      "step": 416
+    },
+    {
+      "epoch": 0.09477272727272727,
+      "grad_norm": 2.778506603554396,
+      "learning_rate": 1.2488975387151702e-06,
+      "loss": 0.1047,
+      "step": 417
+    },
+    {
+      "epoch": 0.095,
+      "grad_norm": 2.1095524554701806,
+      "learning_rate": 1.2488922336181518e-06,
+      "loss": 0.1049,
+      "step": 418
+    },
+    {
+      "epoch": 0.09522727272727273,
+      "grad_norm": 2.6736629021342884,
+      "learning_rate": 1.2488869157988826e-06,
+      "loss": 0.0831,
+      "step": 419
+    },
+    {
+      "epoch": 0.09545454545454546,
+      "grad_norm": 2.9234631052793834,
+      "learning_rate": 1.2488815852574703e-06,
+      "loss": 0.1346,
+      "step": 420
+    },
+    {
+      "epoch": 0.09568181818181819,
+      "grad_norm": 1.9160296186544232,
+      "learning_rate": 1.2488762419940244e-06,
+      "loss": 0.1058,
+      "step": 421
+    },
+    {
+      "epoch": 0.0959090909090909,
+      "grad_norm": 1.7233838699574222,
+      "learning_rate": 1.2488708860086536e-06,
+      "loss": 0.0659,
+      "step": 422
+    },
+    {
+      "epoch": 0.09613636363636363,
+      "grad_norm": 2.4551197802456888,
+      "learning_rate": 1.2488655173014666e-06,
+      "loss": 0.079,
+      "step": 423
+    },
+    {
+      "epoch": 0.09636363636363636,
+      "grad_norm": 1.4606439864980938,
+      "learning_rate": 1.2488601358725737e-06,
+      "loss": 0.0795,
+      "step": 424
+    },
+    {
+      "epoch": 0.09659090909090909,
+      "grad_norm": 2.2199404794770397,
+      "learning_rate": 1.248854741722084e-06,
+      "loss": 0.0752,
+      "step": 425
+    },
+    {
+      "epoch": 0.09681818181818182,
+      "grad_norm": 2.448009652664545,
+      "learning_rate": 1.248849334850108e-06,
+      "loss": 0.0743,
+      "step": 426
+    },
+    {
+      "epoch": 0.09704545454545455,
+      "grad_norm": 2.5867177235825705,
+      "learning_rate": 1.2488439152567556e-06,
+      "loss": 0.0997,
+      "step": 427
+    },
+    {
+      "epoch": 0.09727272727272727,
+      "grad_norm": 1.2216800115304634,
+      "learning_rate": 1.248838482942137e-06,
+      "loss": 0.0742,
+      "step": 428
+    },
+    {
+      "epoch": 0.0975,
+      "grad_norm": 1.8029713380860561,
+      "learning_rate": 1.2488330379063641e-06,
+      "loss": 0.1177,
+      "step": 429
+    },
+    {
+      "epoch": 0.09772727272727273,
+      "grad_norm": 1.5451367085415704,
+      "learning_rate": 1.2488275801495467e-06,
+      "loss": 0.0777,
+      "step": 430
+    },
+    {
+      "epoch": 0.09795454545454546,
+      "grad_norm": 2.318821474524572,
+      "learning_rate": 1.2488221096717967e-06,
+      "loss": 0.0517,
+      "step": 431
+    },
+    {
+      "epoch": 0.09818181818181818,
+      "grad_norm": 2.52648183834825,
+      "learning_rate": 1.2488166264732256e-06,
+      "loss": 0.0908,
+      "step": 432
+    },
+    {
+      "epoch": 0.0984090909090909,
+      "grad_norm": 2.432215428359595,
+      "learning_rate": 1.2488111305539453e-06,
+      "loss": 0.1155,
+      "step": 433
+    },
+    {
+      "epoch": 0.09863636363636363,
+      "grad_norm": 1.7553227077151716,
+      "learning_rate": 1.2488056219140677e-06,
+      "loss": 0.0675,
+      "step": 434
+    },
+    {
+      "epoch": 0.09886363636363636,
+      "grad_norm": 1.9577753939611093,
+      "learning_rate": 1.2488001005537052e-06,
+      "loss": 0.0816,
+      "step": 435
+    },
+    {
+      "epoch": 0.09909090909090909,
+      "grad_norm": 2.652339238194842,
+      "learning_rate": 1.2487945664729703e-06,
+      "loss": 0.0996,
+      "step": 436
+    },
+    {
+      "epoch": 0.09931818181818182,
+      "grad_norm": 1.5594308671532067,
+      "learning_rate": 1.248789019671976e-06,
+      "loss": 0.0835,
+      "step": 437
+    },
+    {
+      "epoch": 0.09954545454545455,
+      "grad_norm": 2.344970890437003,
+      "learning_rate": 1.2487834601508353e-06,
+      "loss": 0.0924,
+      "step": 438
+    },
+    {
+      "epoch": 0.09977272727272728,
+      "grad_norm": 1.4792295361618915,
+      "learning_rate": 1.2487778879096615e-06,
+      "loss": 0.0512,
+      "step": 439
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4640564633292237,
+      "learning_rate": 1.2487723029485684e-06,
+      "loss": 0.0916,
+      "step": 440
+    },
+    {
+      "epoch": 0.10022727272727272,
+      "grad_norm": 2.6075538388387134,
+      "learning_rate": 1.2487667052676699e-06,
+      "loss": 0.1256,
+      "step": 441
+    },
+    {
+      "epoch": 0.10045454545454545,
+      "grad_norm": 3.4032335559769704,
+      "learning_rate": 1.2487610948670798e-06,
+      "loss": 0.1876,
+      "step": 442
+    },
+    {
+      "epoch": 0.10068181818181818,
+      "grad_norm": 1.5100029103510688,
+      "learning_rate": 1.248755471746913e-06,
+      "loss": 0.0708,
+      "step": 443
+    },
+    {
+      "epoch": 0.1009090909090909,
+      "grad_norm": 1.845332143296848,
+      "learning_rate": 1.248749835907284e-06,
+      "loss": 0.1386,
+      "step": 444
+    },
+    {
+      "epoch": 0.10113636363636364,
+      "grad_norm": 3.4888755359518693,
+      "learning_rate": 1.2487441873483075e-06,
+      "loss": 0.1146,
+      "step": 445
+    },
+    {
+      "epoch": 0.10136363636363636,
+      "grad_norm": 1.6111236780246716,
+      "learning_rate": 1.2487385260700987e-06,
+      "loss": 0.0721,
+      "step": 446
+    },
+    {
+      "epoch": 0.10159090909090909,
+      "grad_norm": 0.7765346492857228,
+      "learning_rate": 1.2487328520727734e-06,
+      "loss": 0.0309,
+      "step": 447
+    },
+    {
+      "epoch": 0.10181818181818182,
+      "grad_norm": 1.2214609099572236,
+      "learning_rate": 1.2487271653564471e-06,
+      "loss": 0.0458,
+      "step": 448
+    },
+    {
+      "epoch": 0.10204545454545455,
+      "grad_norm": 1.360973854703278,
+      "learning_rate": 1.2487214659212356e-06,
+      "loss": 0.0584,
+      "step": 449
+    },
+    {
+      "epoch": 0.10227272727272728,
+      "grad_norm": 1.196512716845981,
+      "learning_rate": 1.2487157537672554e-06,
+      "loss": 0.0612,
+      "step": 450
+    },
+    {
+      "epoch": 0.1025,
+      "grad_norm": 1.487496389988837,
+      "learning_rate": 1.2487100288946228e-06,
+      "loss": 0.0797,
+      "step": 451
+    },
+    {
+      "epoch": 0.10272727272727272,
+      "grad_norm": 2.1404575587075554,
+      "learning_rate": 1.2487042913034544e-06,
+      "loss": 0.0959,
+      "step": 452
+    },
+    {
+      "epoch": 0.10295454545454545,
+      "grad_norm": 1.9927997854090056,
+      "learning_rate": 1.2486985409938676e-06,
+      "loss": 0.0686,
+      "step": 453
+    },
+    {
+      "epoch": 0.10318181818181818,
+      "grad_norm": 1.8263677484737977,
+      "learning_rate": 1.2486927779659794e-06,
+      "loss": 0.1092,
+      "step": 454
+    },
+    {
+      "epoch": 0.10340909090909091,
+      "grad_norm": 4.416047806857413,
+      "learning_rate": 1.2486870022199074e-06,
+      "loss": 0.1116,
+      "step": 455
+    },
+    {
+      "epoch": 0.10363636363636364,
+      "grad_norm": 1.4422386304080221,
+      "learning_rate": 1.2486812137557693e-06,
+      "loss": 0.0578,
+      "step": 456
+    },
+    {
+      "epoch": 0.10386363636363637,
+      "grad_norm": 1.9266614706755454,
+      "learning_rate": 1.2486754125736833e-06,
+      "loss": 0.0571,
+      "step": 457
+    },
+    {
+      "epoch": 0.1040909090909091,
+      "grad_norm": 1.3244271825572036,
+      "learning_rate": 1.2486695986737673e-06,
+      "loss": 0.0397,
+      "step": 458
+    },
+    {
+      "epoch": 0.10431818181818182,
+      "grad_norm": 2.104707362823762,
+      "learning_rate": 1.2486637720561404e-06,
+      "loss": 0.0728,
+      "step": 459
+    },
+    {
+      "epoch": 0.10454545454545454,
+      "grad_norm": 1.7307804767375468,
+      "learning_rate": 1.248657932720921e-06,
+      "loss": 0.0616,
+      "step": 460
+    },
+    {
+      "epoch": 0.10477272727272727,
+      "grad_norm": 1.4739933784206947,
+      "learning_rate": 1.2486520806682283e-06,
+      "loss": 0.0563,
+      "step": 461
+    },
+    {
+      "epoch": 0.105,
+      "grad_norm": 1.7621351744687477,
+      "learning_rate": 1.248646215898182e-06,
+      "loss": 0.1016,
+      "step": 462
+    },
+    {
+      "epoch": 0.10522727272727272,
+      "grad_norm": 1.245244155972086,
+      "learning_rate": 1.2486403384109012e-06,
+      "loss": 0.0461,
+      "step": 463
+    },
+    {
+      "epoch": 0.10545454545454545,
+      "grad_norm": 1.6618991496732027,
+      "learning_rate": 1.2486344482065056e-06,
+      "loss": 0.0858,
+      "step": 464
+    },
+    {
+      "epoch": 0.10568181818181818,
+      "grad_norm": 2.5258028868808147,
+      "learning_rate": 1.2486285452851158e-06,
+      "loss": 0.0869,
+      "step": 465
+    },
+    {
+      "epoch": 0.10590909090909091,
+      "grad_norm": 2.2689283506887286,
+      "learning_rate": 1.248622629646852e-06,
+      "loss": 0.1068,
+      "step": 466
+    },
+    {
+      "epoch": 0.10613636363636364,
+      "grad_norm": 3.7401060871096874,
+      "learning_rate": 1.2486167012918347e-06,
+      "loss": 0.1326,
+      "step": 467
+    },
+    {
+      "epoch": 0.10636363636363637,
+      "grad_norm": 1.6774498136362916,
+      "learning_rate": 1.248610760220185e-06,
+      "loss": 0.0791,
+      "step": 468
+    },
+    {
+      "epoch": 0.1065909090909091,
+      "grad_norm": 0.5791858816991924,
+      "learning_rate": 1.248604806432024e-06,
+      "loss": 0.0449,
+      "step": 469
+    },
+    {
+      "epoch": 0.10681818181818181,
+      "grad_norm": 1.9521958750256818,
+      "learning_rate": 1.2485988399274728e-06,
+      "loss": 0.0504,
+      "step": 470
+    },
+    {
+      "epoch": 0.10704545454545454,
+      "grad_norm": 1.8539048276348093,
+      "learning_rate": 1.2485928607066537e-06,
+      "loss": 0.0898,
+      "step": 471
+    },
+    {
+      "epoch": 0.10727272727272727,
+      "grad_norm": 2.036799967782119,
+      "learning_rate": 1.248586868769688e-06,
+      "loss": 0.154,
+      "step": 472
+    },
+    {
+      "epoch": 0.1075,
+      "grad_norm": 2.8007102582051546,
+      "learning_rate": 1.2485808641166983e-06,
+      "loss": 0.176,
+      "step": 473
+    },
+    {
+      "epoch": 0.10772727272727273,
+      "grad_norm": 1.4772022890437055,
+      "learning_rate": 1.2485748467478069e-06,
+      "loss": 0.0879,
+      "step": 474
+    },
+    {
+      "epoch": 0.10795454545454546,
+      "grad_norm": 2.407450130929305,
+      "learning_rate": 1.2485688166631363e-06,
+      "loss": 0.0949,
+      "step": 475
+    },
+    {
+      "epoch": 0.10818181818181818,
+      "grad_norm": 2.061479521079572,
+      "learning_rate": 1.2485627738628097e-06,
+      "loss": 0.0762,
+      "step": 476
+    },
+    {
+      "epoch": 0.10840909090909091,
+      "grad_norm": 1.8548262700905604,
+      "learning_rate": 1.24855671834695e-06,
+      "loss": 0.14,
+      "step": 477
+    },
+    {
+      "epoch": 0.10863636363636364,
+      "grad_norm": 0.964476100134851,
+      "learning_rate": 1.2485506501156815e-06,
+      "loss": 0.0418,
+      "step": 478
+    },
+    {
+      "epoch": 0.10886363636363636,
+      "grad_norm": 1.1145266959461315,
+      "learning_rate": 1.2485445691691273e-06,
+      "loss": 0.0677,
+      "step": 479
+    },
+    {
+      "epoch": 0.10909090909090909,
+      "grad_norm": 2.330127881300682,
+      "learning_rate": 1.2485384755074112e-06,
+      "loss": 0.1005,
+      "step": 480
+    },
+    {
+      "epoch": 0.10931818181818181,
+      "grad_norm": 2.6983253247259427,
+      "learning_rate": 1.248532369130658e-06,
+      "loss": 0.0887,
+      "step": 481
+    },
+    {
+      "epoch": 0.10954545454545454,
+      "grad_norm": 2.4986750792970827,
+      "learning_rate": 1.2485262500389918e-06,
+      "loss": 0.1132,
+      "step": 482
+    },
+    {
+      "epoch": 0.10977272727272727,
+      "grad_norm": 2.3884347533881756,
+      "learning_rate": 1.2485201182325378e-06,
+      "loss": 0.0866,
+      "step": 483
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7236763475705132,
+      "learning_rate": 1.2485139737114208e-06,
+      "loss": 0.0732,
+      "step": 484
+    },
+    {
+      "epoch": 0.11022727272727273,
+      "grad_norm": 0.6594870797143155,
+      "learning_rate": 1.2485078164757659e-06,
+      "loss": 0.0405,
+      "step": 485
+    },
+    {
+      "epoch": 0.11045454545454546,
+      "grad_norm": 2.5371883392823706,
+      "learning_rate": 1.2485016465256987e-06,
+      "loss": 0.0762,
+      "step": 486
+    },
+    {
+      "epoch": 0.11068181818181819,
+      "grad_norm": 1.9519022387132825,
+      "learning_rate": 1.2484954638613456e-06,
+      "loss": 0.0678,
+      "step": 487
+    },
+    {
+      "epoch": 0.11090909090909092,
+      "grad_norm": 0.5386838297138675,
+      "learning_rate": 1.248489268482832e-06,
+      "loss": 0.0264,
+      "step": 488
+    },
+    {
+      "epoch": 0.11113636363636363,
+      "grad_norm": 2.200111981080962,
+      "learning_rate": 1.2484830603902846e-06,
+      "loss": 0.1782,
+      "step": 489
+    },
+    {
+      "epoch": 0.11136363636363636,
+      "grad_norm": 2.165457389848854,
+      "learning_rate": 1.2484768395838299e-06,
+      "loss": 0.1095,
+      "step": 490
+    },
+    {
+      "epoch": 0.11159090909090909,
+      "grad_norm": 1.6690565961021626,
+      "learning_rate": 1.2484706060635945e-06,
+      "loss": 0.0621,
+      "step": 491
+    },
+    {
+      "epoch": 0.11181818181818182,
+      "grad_norm": 1.0017787203741886,
+      "learning_rate": 1.2484643598297058e-06,
+      "loss": 0.0406,
+      "step": 492
+    },
+    {
+      "epoch": 0.11204545454545455,
+      "grad_norm": 0.8806988016458503,
+      "learning_rate": 1.2484581008822912e-06,
+      "loss": 0.0472,
+      "step": 493
+    },
+    {
+      "epoch": 0.11227272727272727,
+      "grad_norm": 1.9770058595085798,
+      "learning_rate": 1.2484518292214782e-06,
+      "loss": 0.1683,
+      "step": 494
+    },
+    {
+      "epoch": 0.1125,
+      "grad_norm": 1.7149135450486523,
+      "learning_rate": 1.2484455448473949e-06,
+      "loss": 0.0477,
+      "step": 495
+    },
+    {
+      "epoch": 0.11272727272727273,
+      "grad_norm": 2.5634460074003487,
+      "learning_rate": 1.248439247760169e-06,
+      "loss": 0.1804,
+      "step": 496
+    },
+    {
+      "epoch": 0.11295454545454546,
+      "grad_norm": 3.0275119245365727,
+      "learning_rate": 1.2484329379599292e-06,
+      "loss": 0.115,
+      "step": 497
+    },
+    {
+      "epoch": 0.11318181818181818,
+      "grad_norm": 1.6079090918603969,
+      "learning_rate": 1.2484266154468042e-06,
+      "loss": 0.0541,
+      "step": 498
+    },
+    {
+      "epoch": 0.1134090909090909,
+      "grad_norm": 1.698954435881134,
+      "learning_rate": 1.248420280220923e-06,
+      "loss": 0.0653,
+      "step": 499
+    },
+    {
+      "epoch": 0.11363636363636363,
+      "grad_norm": 2.32288025425098,
+      "learning_rate": 1.2484139322824147e-06,
+      "loss": 0.0883,
+      "step": 500
+    },
+    {
+      "epoch": 0.11386363636363636,
+      "grad_norm": 2.262741222608943,
+      "learning_rate": 1.2484075716314085e-06,
+      "loss": 0.1051,
+      "step": 501
+    },
+    {
+      "epoch": 0.11409090909090909,
+      "grad_norm": 2.3792798552294028,
+      "learning_rate": 1.2484011982680346e-06,
+      "loss": 0.1056,
+      "step": 502
+    },
+    {
+      "epoch": 0.11431818181818182,
+      "grad_norm": 2.673543843991085,
+      "learning_rate": 1.2483948121924224e-06,
+      "loss": 0.0978,
+      "step": 503
+    },
+    {
+      "epoch": 0.11454545454545455,
+      "grad_norm": 1.8961290857866375,
+      "learning_rate": 1.2483884134047026e-06,
+      "loss": 0.0754,
+      "step": 504
+    },
+    {
+      "epoch": 0.11477272727272728,
+      "grad_norm": 1.651727652691884,
+      "learning_rate": 1.2483820019050056e-06,
+      "loss": 0.095,
+      "step": 505
+    },
+    {
+      "epoch": 0.115,
+      "grad_norm": 3.0263085945449477,
+      "learning_rate": 1.2483755776934616e-06,
+      "loss": 0.1192,
+      "step": 506
+    },
+    {
+      "epoch": 0.11522727272727273,
+      "grad_norm": 2.244241504681122,
+      "learning_rate": 1.2483691407702025e-06,
+      "loss": 0.118,
+      "step": 507
+    },
+    {
+      "epoch": 0.11545454545454545,
+      "grad_norm": 2.310111668781045,
+      "learning_rate": 1.2483626911353589e-06,
+      "loss": 0.0752,
+      "step": 508
+    },
+    {
+      "epoch": 0.11568181818181818,
+      "grad_norm": 1.8436889026262906,
+      "learning_rate": 1.2483562287890625e-06,
+      "loss": 0.0691,
+      "step": 509
+    },
+    {
+      "epoch": 0.1159090909090909,
+      "grad_norm": 2.666846733942365,
+      "learning_rate": 1.248349753731445e-06,
+      "loss": 0.1043,
+      "step": 510
+    },
+    {
+      "epoch": 0.11613636363636363,
+      "grad_norm": 2.107687345505901,
+      "learning_rate": 1.2483432659626384e-06,
+      "loss": 0.1134,
+      "step": 511
+    },
+    {
+      "epoch": 0.11636363636363636,
+      "grad_norm": 3.5030619049624954,
+      "learning_rate": 1.2483367654827756e-06,
+      "loss": 0.0603,
+      "step": 512
+    },
+    {
+      "epoch": 0.11659090909090909,
+      "grad_norm": 4.213870361472748,
+      "learning_rate": 1.2483302522919884e-06,
+      "loss": 0.1389,
+      "step": 513
+    },
+    {
+      "epoch": 0.11681818181818182,
+      "grad_norm": 2.5454611079284937,
+      "learning_rate": 1.24832372639041e-06,
+      "loss": 0.1079,
+      "step": 514
+    },
+    {
+      "epoch": 0.11704545454545455,
+      "grad_norm": 1.3075425384287886,
+      "learning_rate": 1.2483171877781731e-06,
+      "loss": 0.0998,
+      "step": 515
+    },
+    {
+      "epoch": 0.11727272727272728,
+      "grad_norm": 1.5597885547847803,
+      "learning_rate": 1.2483106364554115e-06,
+      "loss": 0.0892,
+      "step": 516
+    },
+    {
+      "epoch": 0.1175,
+      "grad_norm": 2.6700601125230916,
+      "learning_rate": 1.2483040724222586e-06,
+      "loss": 0.1079,
+      "step": 517
+    },
+    {
+      "epoch": 0.11772727272727272,
+      "grad_norm": 3.511283195711642,
+      "learning_rate": 1.2482974956788483e-06,
+      "loss": 0.1664,
+      "step": 518
+    },
+    {
+      "epoch": 0.11795454545454545,
+      "grad_norm": 1.753249356483609,
+      "learning_rate": 1.2482909062253146e-06,
+      "loss": 0.0509,
+      "step": 519
+    },
+    {
+      "epoch": 0.11818181818181818,
+      "grad_norm": 1.2199463794601273,
+      "learning_rate": 1.2482843040617921e-06,
+      "loss": 0.0385,
+      "step": 520
+    },
+    {
+      "epoch": 0.11840909090909091,
+      "grad_norm": 4.774533358869784,
+      "learning_rate": 1.248277689188415e-06,
+      "loss": 0.1733,
+      "step": 521
+    },
+    {
+      "epoch": 0.11863636363636364,
+      "grad_norm": 1.9141585409726258,
+      "learning_rate": 1.2482710616053185e-06,
+      "loss": 0.0732,
+      "step": 522
+    },
+    {
+      "epoch": 0.11886363636363637,
+      "grad_norm": 2.718009089318278,
+      "learning_rate": 1.2482644213126377e-06,
+      "loss": 0.1042,
+      "step": 523
+    },
+    {
+      "epoch": 0.1190909090909091,
+      "grad_norm": 1.5291385060090936,
+      "learning_rate": 1.248257768310508e-06,
+      "loss": 0.0659,
+      "step": 524
+    },
+    {
+      "epoch": 0.11931818181818182,
+      "grad_norm": 1.5438844541718455,
+      "learning_rate": 1.2482511025990652e-06,
+      "loss": 0.0768,
+      "step": 525
+    },
+    {
+      "epoch": 0.11954545454545455,
+      "grad_norm": 2.0380468476624047,
+      "learning_rate": 1.248244424178445e-06,
+      "loss": 0.1205,
+      "step": 526
+    },
+    {
+      "epoch": 0.11977272727272727,
+      "grad_norm": 1.695751956727507,
+      "learning_rate": 1.2482377330487837e-06,
+      "loss": 0.089,
+      "step": 527
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7592189613511189,
+      "learning_rate": 1.2482310292102177e-06,
+      "loss": 0.1124,
+      "step": 528
+    },
+    {
+      "epoch": 0.12022727272727272,
+      "grad_norm": 4.995661129810231,
+      "learning_rate": 1.2482243126628838e-06,
+      "loss": 0.1282,
+      "step": 529
+    },
+    {
+      "epoch": 0.12045454545454545,
+      "grad_norm": 3.0408520694807644,
+      "learning_rate": 1.2482175834069188e-06,
+      "loss": 0.1119,
+      "step": 530
+    },
+    {
+      "epoch": 0.12068181818181818,
+      "grad_norm": 1.7647186023200758,
+      "learning_rate": 1.2482108414424602e-06,
+      "loss": 0.095,
+      "step": 531
+    },
+    {
+      "epoch": 0.12090909090909091,
+      "grad_norm": 3.6756871803312183,
+      "learning_rate": 1.2482040867696452e-06,
+      "loss": 0.0983,
+      "step": 532
+    },
+    {
+      "epoch": 0.12113636363636364,
+      "grad_norm": 2.3208636702458696,
+      "learning_rate": 1.2481973193886114e-06,
+      "loss": 0.1882,
+      "step": 533
+    },
+    {
+      "epoch": 0.12136363636363637,
+      "grad_norm": 2.626784420382633,
+      "learning_rate": 1.2481905392994971e-06,
+      "loss": 0.1097,
+      "step": 534
+    },
+    {
+      "epoch": 0.1215909090909091,
+      "grad_norm": 1.860774693118693,
+      "learning_rate": 1.2481837465024408e-06,
+      "loss": 0.0667,
+      "step": 535
+    },
+    {
+      "epoch": 0.12181818181818181,
+      "grad_norm": 1.446403593940181,
+      "learning_rate": 1.2481769409975805e-06,
+      "loss": 0.073,
+      "step": 536
+    },
+    {
+      "epoch": 0.12204545454545454,
+      "grad_norm": 2.173700581392551,
+      "learning_rate": 1.2481701227850551e-06,
+      "loss": 0.1161,
+      "step": 537
+    },
+    {
+      "epoch": 0.12227272727272727,
+      "grad_norm": 1.99797868342146,
+      "learning_rate": 1.248163291865004e-06,
+      "loss": 0.1031,
+      "step": 538
+    },
+    {
+      "epoch": 0.1225,
+      "grad_norm": 1.328622639093438,
+      "learning_rate": 1.2481564482375658e-06,
+      "loss": 0.0673,
+      "step": 539
+    },
+    {
+      "epoch": 0.12272727272727273,
+      "grad_norm": 0.9297667220989132,
+      "learning_rate": 1.2481495919028808e-06,
+      "loss": 0.0361,
+      "step": 540
+    },
+    {
+      "epoch": 0.12295454545454546,
+      "grad_norm": 7.973573472263443,
+      "learning_rate": 1.2481427228610881e-06,
+      "loss": 0.0575,
+      "step": 541
+    },
+    {
+      "epoch": 0.12318181818181818,
+      "grad_norm": 2.180417930341818,
+      "learning_rate": 1.2481358411123285e-06,
+      "loss": 0.098,
+      "step": 542
+    },
+    {
+      "epoch": 0.12340909090909091,
+      "grad_norm": 2.1498758546155874,
+      "learning_rate": 1.248128946656742e-06,
+      "loss": 0.1152,
+      "step": 543
+    },
+    {
+      "epoch": 0.12363636363636364,
+      "grad_norm": 1.6319464584953844,
+      "learning_rate": 1.248122039494469e-06,
+      "loss": 0.1062,
+      "step": 544
+    },
+    {
+      "epoch": 0.12386363636363637,
+      "grad_norm": 1.9248105451365445,
+      "learning_rate": 1.2481151196256506e-06,
+      "loss": 0.0799,
+      "step": 545
+    },
+    {
+      "epoch": 0.12409090909090909,
+      "grad_norm": 3.548093691433519,
+      "learning_rate": 1.2481081870504278e-06,
+      "loss": 0.1563,
+      "step": 546
+    },
+    {
+      "epoch": 0.12431818181818181,
+      "grad_norm": 2.145638512413841,
+      "learning_rate": 1.248101241768942e-06,
+      "loss": 0.0571,
+      "step": 547
+    },
+    {
+      "epoch": 0.12454545454545454,
+      "grad_norm": 1.922136955950023,
+      "learning_rate": 1.2480942837813348e-06,
+      "loss": 0.097,
+      "step": 548
+    },
+    {
+      "epoch": 0.12477272727272727,
+      "grad_norm": 0.6955005393551513,
+      "learning_rate": 1.248087313087748e-06,
+      "loss": 0.0451,
+      "step": 549
+    },
+    {
+      "epoch": 0.125,
+      "grad_norm": 1.7527975364628896,
+      "learning_rate": 1.248080329688324e-06,
+      "loss": 0.0817,
+      "step": 550
+    },
+    {
+      "epoch": 0.12522727272727271,
+      "grad_norm": 2.4150984769726085,
+      "learning_rate": 1.2480733335832052e-06,
+      "loss": 0.1106,
+      "step": 551
+    },
+    {
+      "epoch": 0.12545454545454546,
+      "grad_norm": 2.5048775933295024,
+      "learning_rate": 1.2480663247725338e-06,
+      "loss": 0.0727,
+      "step": 552
+    },
+    {
+      "epoch": 0.12568181818181817,
+      "grad_norm": 2.314610045698514,
+      "learning_rate": 1.2480593032564532e-06,
+      "loss": 0.1362,
+      "step": 553
+    },
+    {
+      "epoch": 0.12590909090909091,
+      "grad_norm": 3.2925046483379123,
+      "learning_rate": 1.2480522690351065e-06,
+      "loss": 0.1455,
+      "step": 554
+    },
+    {
+      "epoch": 0.12613636363636363,
+      "grad_norm": 1.3820687777050022,
+      "learning_rate": 1.248045222108637e-06,
+      "loss": 0.0597,
+      "step": 555
+    },
+    {
+      "epoch": 0.12636363636363637,
+      "grad_norm": 2.0729413776396637,
+      "learning_rate": 1.2480381624771882e-06,
+      "loss": 0.1771,
+      "step": 556
+    },
+    {
+      "epoch": 0.1265909090909091,
+      "grad_norm": 2.9901894706532772,
+      "learning_rate": 1.2480310901409046e-06,
+      "loss": 0.0855,
+      "step": 557
+    },
+    {
+      "epoch": 0.12681818181818183,
+      "grad_norm": 1.7018708570112,
+      "learning_rate": 1.24802400509993e-06,
+      "loss": 0.0642,
+      "step": 558
+    },
+    {
+      "epoch": 0.12704545454545454,
+      "grad_norm": 3.0554126375188426,
+      "learning_rate": 1.2480169073544092e-06,
+      "loss": 0.1334,
+      "step": 559
+    },
+    {
+      "epoch": 0.12727272727272726,
+      "grad_norm": 1.156208289095741,
+      "learning_rate": 1.2480097969044867e-06,
+      "loss": 0.0425,
+      "step": 560
+    },
+    {
+      "epoch": 0.1275,
+      "grad_norm": 2.09259199795833,
+      "learning_rate": 1.2480026737503073e-06,
+      "loss": 0.1024,
+      "step": 561
+    },
+    {
+      "epoch": 0.12772727272727272,
+      "grad_norm": 2.1077058051652937,
+      "learning_rate": 1.2479955378920167e-06,
+      "loss": 0.101,
+      "step": 562
+    },
+    {
+      "epoch": 0.12795454545454546,
+      "grad_norm": 2.508136779271039,
+      "learning_rate": 1.2479883893297602e-06,
+      "loss": 0.1143,
+      "step": 563
+    },
+    {
+      "epoch": 0.12818181818181817,
+      "grad_norm": 1.3873776835532647,
+      "learning_rate": 1.2479812280636836e-06,
+      "loss": 0.0536,
+      "step": 564
+    },
+    {
+      "epoch": 0.12840909090909092,
+      "grad_norm": 1.8747767005987064,
+      "learning_rate": 1.2479740540939328e-06,
+      "loss": 0.1205,
+      "step": 565
+    },
+    {
+      "epoch": 0.12863636363636363,
+      "grad_norm": 2.2559581352789406,
+      "learning_rate": 1.2479668674206543e-06,
+      "loss": 0.1377,
+      "step": 566
+    },
+    {
+      "epoch": 0.12886363636363637,
+      "grad_norm": 1.865382683864209,
+      "learning_rate": 1.2479596680439944e-06,
+      "loss": 0.0603,
+      "step": 567
+    },
+    {
+      "epoch": 0.1290909090909091,
+      "grad_norm": 3.1049373729212886,
+      "learning_rate": 1.2479524559641e-06,
+      "loss": 0.1468,
+      "step": 568
+    },
+    {
+      "epoch": 0.1293181818181818,
+      "grad_norm": 1.654156785974806,
+      "learning_rate": 1.2479452311811185e-06,
+      "loss": 0.0808,
+      "step": 569
+    },
+    {
+      "epoch": 0.12954545454545455,
+      "grad_norm": 2.3997893578872302,
+      "learning_rate": 1.2479379936951968e-06,
+      "loss": 0.1377,
+      "step": 570
+    },
+    {
+      "epoch": 0.12977272727272726,
+      "grad_norm": 1.667058229097542,
+      "learning_rate": 1.2479307435064827e-06,
+      "loss": 0.1071,
+      "step": 571
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0288361685365643,
+      "learning_rate": 1.2479234806151242e-06,
+      "loss": 0.1217,
+      "step": 572
+    },
+    {
+      "epoch": 0.13022727272727272,
+      "grad_norm": 1.7395690918734394,
+      "learning_rate": 1.2479162050212688e-06,
+      "loss": 0.0619,
+      "step": 573
+    },
+    {
+      "epoch": 0.13045454545454546,
+      "grad_norm": 1.5319554982801848,
+      "learning_rate": 1.2479089167250654e-06,
+      "loss": 0.0502,
+      "step": 574
+    },
+    {
+      "epoch": 0.13068181818181818,
+      "grad_norm": 1.8235299621848442,
+      "learning_rate": 1.2479016157266623e-06,
+      "loss": 0.0654,
+      "step": 575
+    },
+    {
+      "epoch": 0.13090909090909092,
+      "grad_norm": 1.777465934459937,
+      "learning_rate": 1.2478943020262087e-06,
+      "loss": 0.0895,
+      "step": 576
+    },
+    {
+      "epoch": 0.13113636363636363,
+      "grad_norm": 2.1564538052356608,
+      "learning_rate": 1.2478869756238536e-06,
+      "loss": 0.1044,
+      "step": 577
+    },
+    {
+      "epoch": 0.13136363636363638,
+      "grad_norm": 2.231105279853951,
+      "learning_rate": 1.2478796365197462e-06,
+      "loss": 0.1117,
+      "step": 578
+    },
+    {
+      "epoch": 0.1315909090909091,
+      "grad_norm": 2.093836133670983,
+      "learning_rate": 1.2478722847140367e-06,
+      "loss": 0.1363,
+      "step": 579
+    },
+    {
+      "epoch": 0.1318181818181818,
+      "grad_norm": 1.8193948285604047,
+      "learning_rate": 1.2478649202068744e-06,
+      "loss": 0.0671,
+      "step": 580
+    },
+    {
+      "epoch": 0.13204545454545455,
+      "grad_norm": 1.9443060761361328,
+      "learning_rate": 1.2478575429984097e-06,
+      "loss": 0.0904,
+      "step": 581
+    },
+    {
+      "epoch": 0.13227272727272726,
+      "grad_norm": 1.6346370023538672,
+      "learning_rate": 1.2478501530887933e-06,
+      "loss": 0.0871,
+      "step": 582
+    },
+    {
+      "epoch": 0.1325,
+      "grad_norm": 2.346913204468733,
+      "learning_rate": 1.2478427504781753e-06,
+      "loss": 0.1283,
+      "step": 583
+    },
+    {
+      "epoch": 0.13272727272727272,
+      "grad_norm": 1.9129761479995544,
+      "learning_rate": 1.2478353351667072e-06,
+      "loss": 0.0574,
+      "step": 584
+    },
+    {
+      "epoch": 0.13295454545454546,
+      "grad_norm": 1.8123643256486057,
+      "learning_rate": 1.24782790715454e-06,
+      "loss": 0.0716,
+      "step": 585
+    },
+    {
+      "epoch": 0.13318181818181818,
+      "grad_norm": 3.4725912075373526,
+      "learning_rate": 1.2478204664418254e-06,
+      "loss": 0.1327,
+      "step": 586
+    },
+    {
+      "epoch": 0.13340909090909092,
+      "grad_norm": 2.166070188664805,
+      "learning_rate": 1.2478130130287147e-06,
+      "loss": 0.1369,
+      "step": 587
+    },
+    {
+      "epoch": 0.13363636363636364,
+      "grad_norm": 1.1059426146403066,
+      "learning_rate": 1.2478055469153599e-06,
+      "loss": 0.0661,
+      "step": 588
+    },
+    {
+      "epoch": 0.13386363636363635,
+      "grad_norm": 2.847465595438576,
+      "learning_rate": 1.2477980681019139e-06,
+      "loss": 0.1207,
+      "step": 589
+    },
+    {
+      "epoch": 0.1340909090909091,
+      "grad_norm": 1.8370754566033425,
+      "learning_rate": 1.2477905765885285e-06,
+      "loss": 0.0517,
+      "step": 590
+    },
+    {
+      "epoch": 0.1343181818181818,
+      "grad_norm": 2.08230504589335,
+      "learning_rate": 1.2477830723753567e-06,
+      "loss": 0.0873,
+      "step": 591
+    },
+    {
+      "epoch": 0.13454545454545455,
+      "grad_norm": 1.8956305684762316,
+      "learning_rate": 1.2477755554625517e-06,
+      "loss": 0.0628,
+      "step": 592
+    },
+    {
+      "epoch": 0.13477272727272727,
+      "grad_norm": 1.404982841175674,
+      "learning_rate": 1.2477680258502664e-06,
+      "loss": 0.0497,
+      "step": 593
+    },
+    {
+      "epoch": 0.135,
+      "grad_norm": 2.95536935560541,
+      "learning_rate": 1.2477604835386548e-06,
+      "loss": 0.1285,
+      "step": 594
+    },
+    {
+      "epoch": 0.13522727272727272,
+      "grad_norm": 0.6865821105084395,
+      "learning_rate": 1.2477529285278704e-06,
+      "loss": 0.0422,
+      "step": 595
+    },
+    {
+      "epoch": 0.13545454545454547,
+      "grad_norm": 1.947781506196333,
+      "learning_rate": 1.2477453608180673e-06,
+      "loss": 0.1184,
+      "step": 596
+    },
+    {
+      "epoch": 0.13568181818181818,
+      "grad_norm": 2.176234177731866,
+      "learning_rate": 1.2477377804093998e-06,
+      "loss": 0.1628,
+      "step": 597
+    },
+    {
+      "epoch": 0.1359090909090909,
+      "grad_norm": 2.0282691282238505,
+      "learning_rate": 1.2477301873020228e-06,
+      "loss": 0.0865,
+      "step": 598
+    },
+    {
+      "epoch": 0.13613636363636364,
+      "grad_norm": 0.8578417291735129,
+      "learning_rate": 1.2477225814960905e-06,
+      "loss": 0.0492,
+      "step": 599
+    },
+    {
+      "epoch": 0.13636363636363635,
+      "grad_norm": 4.272836014951482,
+      "learning_rate": 1.2477149629917587e-06,
+      "loss": 0.1492,
+      "step": 600
+    },
+    {
+      "epoch": 0.1365909090909091,
+      "grad_norm": 1.2914964837931204,
+      "learning_rate": 1.2477073317891822e-06,
+      "loss": 0.0467,
+      "step": 601
+    },
+    {
+      "epoch": 0.1368181818181818,
+      "grad_norm": 1.942246100928008,
+      "learning_rate": 1.247699687888517e-06,
+      "loss": 0.0897,
+      "step": 602
+    },
+    {
+      "epoch": 0.13704545454545455,
+      "grad_norm": 2.7356604051165005,
+      "learning_rate": 1.2476920312899186e-06,
+      "loss": 0.188,
+      "step": 603
+    },
+    {
+      "epoch": 0.13727272727272727,
+      "grad_norm": 2.9175760757776428,
+      "learning_rate": 1.2476843619935434e-06,
+      "loss": 0.0839,
+      "step": 604
+    },
+    {
+      "epoch": 0.1375,
+      "grad_norm": 3.1103305433439608,
+      "learning_rate": 1.247676679999548e-06,
+      "loss": 0.0988,
+      "step": 605
+    },
+    {
+      "epoch": 0.13772727272727273,
+      "grad_norm": 1.8863501421794622,
+      "learning_rate": 1.2476689853080883e-06,
+      "loss": 0.0848,
+      "step": 606
+    },
+    {
+      "epoch": 0.13795454545454544,
+      "grad_norm": 3.7100914161754224,
+      "learning_rate": 1.247661277919322e-06,
+      "loss": 0.195,
+      "step": 607
+    },
+    {
+      "epoch": 0.13818181818181818,
+      "grad_norm": 0.49089918540637845,
+      "learning_rate": 1.247653557833406e-06,
+      "loss": 0.0195,
+      "step": 608
+    },
+    {
+      "epoch": 0.1384090909090909,
+      "grad_norm": 0.9651052176587062,
+      "learning_rate": 1.2476458250504974e-06,
+      "loss": 0.0426,
+      "step": 609
+    },
+    {
+      "epoch": 0.13863636363636364,
+      "grad_norm": 2.782401282877486,
+      "learning_rate": 1.2476380795707544e-06,
+      "loss": 0.0972,
+      "step": 610
+    },
+    {
+      "epoch": 0.13886363636363636,
+      "grad_norm": 3.0526176588811724,
+      "learning_rate": 1.2476303213943346e-06,
+      "loss": 0.1342,
+      "step": 611
+    },
+    {
+      "epoch": 0.1390909090909091,
+      "grad_norm": 1.294437488634513,
+      "learning_rate": 1.2476225505213964e-06,
+      "loss": 0.0593,
+      "step": 612
+    },
+    {
+      "epoch": 0.1393181818181818,
+      "grad_norm": 1.7596328031890145,
+      "learning_rate": 1.247614766952098e-06,
+      "loss": 0.0621,
+      "step": 613
+    },
+    {
+      "epoch": 0.13954545454545456,
+      "grad_norm": 2.382168684608339,
+      "learning_rate": 1.2476069706865983e-06,
+      "loss": 0.1307,
+      "step": 614
+    },
+    {
+      "epoch": 0.13977272727272727,
+      "grad_norm": 2.5661428625650493,
+      "learning_rate": 1.2475991617250566e-06,
+      "loss": 0.1001,
+      "step": 615
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.513631365561973,
+      "learning_rate": 1.2475913400676314e-06,
+      "loss": 0.2276,
+      "step": 616
+    },
+    {
+      "epoch": 0.14022727272727273,
+      "grad_norm": 2.6037909076878716,
+      "learning_rate": 1.247583505714483e-06,
+      "loss": 0.1378,
+      "step": 617
+    },
+    {
+      "epoch": 0.14045454545454544,
+      "grad_norm": 1.662682801656563,
+      "learning_rate": 1.2475756586657704e-06,
+      "loss": 0.0567,
+      "step": 618
+    },
+    {
+      "epoch": 0.14068181818181819,
+      "grad_norm": 1.6457027174437255,
+      "learning_rate": 1.2475677989216542e-06,
+      "loss": 0.0468,
+      "step": 619
+    },
+    {
+      "epoch": 0.1409090909090909,
+      "grad_norm": 1.557975778328669,
+      "learning_rate": 1.2475599264822945e-06,
+      "loss": 0.0782,
+      "step": 620
+    },
+    {
+      "epoch": 0.14113636363636364,
+      "grad_norm": 1.7665523674435315,
+      "learning_rate": 1.2475520413478516e-06,
+      "loss": 0.0913,
+      "step": 621
+    },
+    {
+      "epoch": 0.14136363636363636,
+      "grad_norm": 1.7280507776521359,
+      "learning_rate": 1.2475441435184865e-06,
+      "loss": 0.0796,
+      "step": 622
+    },
+    {
+      "epoch": 0.1415909090909091,
+      "grad_norm": 1.7887498462761924,
+      "learning_rate": 1.2475362329943602e-06,
+      "loss": 0.0914,
+      "step": 623
+    },
+    {
+      "epoch": 0.14181818181818182,
+      "grad_norm": 2.317140974254644,
+      "learning_rate": 1.247528309775634e-06,
+      "loss": 0.1202,
+      "step": 624
+    },
+    {
+      "epoch": 0.14204545454545456,
+      "grad_norm": 1.7507905648270672,
+      "learning_rate": 1.2475203738624697e-06,
+      "loss": 0.0494,
+      "step": 625
+    },
+    {
+      "epoch": 0.14227272727272727,
+      "grad_norm": 2.8870412055586887,
+      "learning_rate": 1.247512425255029e-06,
+      "loss": 0.0867,
+      "step": 626
+    },
+    {
+      "epoch": 0.1425,
+      "grad_norm": 1.6241049990484522,
+      "learning_rate": 1.2475044639534736e-06,
+      "loss": 0.0882,
+      "step": 627
+    },
+    {
+      "epoch": 0.14272727272727273,
+      "grad_norm": 2.9782737330409126,
+      "learning_rate": 1.2474964899579663e-06,
+      "loss": 0.1736,
+      "step": 628
+    },
+    {
+      "epoch": 0.14295454545454545,
+      "grad_norm": 3.390479922823892,
+      "learning_rate": 1.2474885032686696e-06,
+      "loss": 0.1184,
+      "step": 629
+    },
+    {
+      "epoch": 0.1431818181818182,
+      "grad_norm": 2.143155834589594,
+      "learning_rate": 1.2474805038857462e-06,
+      "loss": 0.1041,
+      "step": 630
+    },
+    {
+      "epoch": 0.1434090909090909,
+      "grad_norm": 2.446559275843494,
+      "learning_rate": 1.2474724918093594e-06,
+      "loss": 0.1067,
+      "step": 631
+    },
+    {
+      "epoch": 0.14363636363636365,
+      "grad_norm": 1.4094197463299392,
+      "learning_rate": 1.2474644670396726e-06,
+      "loss": 0.0875,
+      "step": 632
+    },
+    {
+      "epoch": 0.14386363636363636,
+      "grad_norm": 1.3971362026378875,
+      "learning_rate": 1.2474564295768493e-06,
+      "loss": 0.056,
+      "step": 633
+    },
+    {
+      "epoch": 0.1440909090909091,
+      "grad_norm": 2.4204105527276245,
+      "learning_rate": 1.2474483794210537e-06,
+      "loss": 0.0859,
+      "step": 634
+    },
+    {
+      "epoch": 0.14431818181818182,
+      "grad_norm": 1.6240037495426662,
+      "learning_rate": 1.2474403165724495e-06,
+      "loss": 0.0774,
+      "step": 635
+    },
+    {
+      "epoch": 0.14454545454545453,
+      "grad_norm": 1.382870120530831,
+      "learning_rate": 1.2474322410312012e-06,
+      "loss": 0.099,
+      "step": 636
+    },
+    {
+      "epoch": 0.14477272727272728,
+      "grad_norm": 1.9404530196216976,
+      "learning_rate": 1.2474241527974737e-06,
+      "loss": 0.0839,
+      "step": 637
+    },
+    {
+      "epoch": 0.145,
+      "grad_norm": 1.513674560400261,
+      "learning_rate": 1.2474160518714322e-06,
+      "loss": 0.0587,
+      "step": 638
+    },
+    {
+      "epoch": 0.14522727272727273,
+      "grad_norm": 2.6446611377722262,
+      "learning_rate": 1.2474079382532412e-06,
+      "loss": 0.0925,
+      "step": 639
+    },
+    {
+      "epoch": 0.14545454545454545,
+      "grad_norm": 1.353373006771438,
+      "learning_rate": 1.2473998119430665e-06,
+      "loss": 0.0636,
+      "step": 640
+    },
+    {
+      "epoch": 0.1456818181818182,
+      "grad_norm": 1.9775503441189939,
+      "learning_rate": 1.247391672941074e-06,
+      "loss": 0.0615,
+      "step": 641
+    },
+    {
+      "epoch": 0.1459090909090909,
+      "grad_norm": 2.446514998607641,
+      "learning_rate": 1.247383521247429e-06,
+      "loss": 0.1439,
+      "step": 642
+    },
+    {
+      "epoch": 0.14613636363636365,
+      "grad_norm": 1.7414956783291986,
+      "learning_rate": 1.2473753568622987e-06,
+      "loss": 0.1162,
+      "step": 643
+    },
+    {
+      "epoch": 0.14636363636363636,
+      "grad_norm": 1.3171430867182248,
+      "learning_rate": 1.247367179785849e-06,
+      "loss": 0.0921,
+      "step": 644
+    },
+    {
+      "epoch": 0.14659090909090908,
+      "grad_norm": 1.6875849679774773,
+      "learning_rate": 1.2473589900182464e-06,
+      "loss": 0.0801,
+      "step": 645
+    },
+    {
+      "epoch": 0.14681818181818182,
+      "grad_norm": 1.892863204360745,
+      "learning_rate": 1.2473507875596586e-06,
+      "loss": 0.112,
+      "step": 646
+    },
+    {
+      "epoch": 0.14704545454545453,
+      "grad_norm": 3.3694155738887894,
+      "learning_rate": 1.2473425724102523e-06,
+      "loss": 0.097,
+      "step": 647
+    },
+    {
+      "epoch": 0.14727272727272728,
+      "grad_norm": 1.0116034223713963,
+      "learning_rate": 1.2473343445701952e-06,
+      "loss": 0.0591,
+      "step": 648
+    },
+    {
+      "epoch": 0.1475,
+      "grad_norm": 2.499197822247887,
+      "learning_rate": 1.2473261040396553e-06,
+      "loss": 0.0838,
+      "step": 649
+    },
+    {
+      "epoch": 0.14772727272727273,
+      "grad_norm": 2.4076341458645873,
+      "learning_rate": 1.2473178508188003e-06,
+      "loss": 0.0833,
+      "step": 650
+    },
+    {
+      "epoch": 0.14795454545454545,
+      "grad_norm": 1.3200433927185546,
+      "learning_rate": 1.2473095849077984e-06,
+      "loss": 0.0631,
+      "step": 651
+    },
+    {
+      "epoch": 0.1481818181818182,
+      "grad_norm": 1.955963149327993,
+      "learning_rate": 1.2473013063068188e-06,
+      "loss": 0.0941,
+      "step": 652
+    },
+    {
+      "epoch": 0.1484090909090909,
+      "grad_norm": 1.3597272559971674,
+      "learning_rate": 1.2472930150160295e-06,
+      "loss": 0.092,
+      "step": 653
+    },
+    {
+      "epoch": 0.14863636363636365,
+      "grad_norm": 1.3273952368460749,
+      "learning_rate": 1.2472847110356002e-06,
+      "loss": 0.0802,
+      "step": 654
+    },
+    {
+      "epoch": 0.14886363636363636,
+      "grad_norm": 1.2283292368668424,
+      "learning_rate": 1.2472763943657e-06,
+      "loss": 0.051,
+      "step": 655
+    },
+    {
+      "epoch": 0.14909090909090908,
+      "grad_norm": 1.895207430235451,
+      "learning_rate": 1.2472680650064984e-06,
+      "loss": 0.0711,
+      "step": 656
+    },
+    {
+      "epoch": 0.14931818181818182,
+      "grad_norm": 2.2763988724900286,
+      "learning_rate": 1.2472597229581655e-06,
+      "loss": 0.0954,
+      "step": 657
+    },
+    {
+      "epoch": 0.14954545454545454,
+      "grad_norm": 3.868102685705611,
+      "learning_rate": 1.2472513682208711e-06,
+      "loss": 0.2411,
+      "step": 658
+    },
+    {
+      "epoch": 0.14977272727272728,
+      "grad_norm": 1.9315761977196009,
+      "learning_rate": 1.2472430007947859e-06,
+      "loss": 0.1027,
+      "step": 659
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.040258081849561,
+      "learning_rate": 1.2472346206800801e-06,
+      "loss": 0.0997,
+      "step": 660
+    },
+    {
+      "epoch": 0.15022727272727274,
+      "grad_norm": 1.1693035908168048,
+      "learning_rate": 1.247226227876925e-06,
+      "loss": 0.0533,
+      "step": 661
+    },
+    {
+      "epoch": 0.15045454545454545,
+      "grad_norm": 2.2078709491943576,
+      "learning_rate": 1.2472178223854916e-06,
+      "loss": 0.1263,
+      "step": 662
+    },
+    {
+      "epoch": 0.1506818181818182,
+      "grad_norm": 2.392885573283221,
+      "learning_rate": 1.2472094042059513e-06,
+      "loss": 0.1203,
+      "step": 663
+    },
+    {
+      "epoch": 0.1509090909090909,
+      "grad_norm": 2.6703758232590373,
+      "learning_rate": 1.2472009733384758e-06,
+      "loss": 0.1132,
+      "step": 664
+    },
+    {
+      "epoch": 0.15113636363636362,
+      "grad_norm": 2.6031131247591253,
+      "learning_rate": 1.2471925297832368e-06,
+      "loss": 0.1053,
+      "step": 665
+    },
+    {
+      "epoch": 0.15136363636363637,
+      "grad_norm": 1.5524701028486667,
+      "learning_rate": 1.2471840735404066e-06,
+      "loss": 0.0716,
+      "step": 666
+    },
+    {
+      "epoch": 0.15159090909090908,
+      "grad_norm": 3.249143499985677,
+      "learning_rate": 1.247175604610158e-06,
+      "loss": 0.1888,
+      "step": 667
+    },
+    {
+      "epoch": 0.15181818181818182,
+      "grad_norm": 1.7958799343302336,
+      "learning_rate": 1.247167122992663e-06,
+      "loss": 0.0789,
+      "step": 668
+    },
+    {
+      "epoch": 0.15204545454545454,
+      "grad_norm": 2.4023747200317325,
+      "learning_rate": 1.2471586286880952e-06,
+      "loss": 0.095,
+      "step": 669
+    },
+    {
+      "epoch": 0.15227272727272728,
+      "grad_norm": 2.2983028222314448,
+      "learning_rate": 1.2471501216966273e-06,
+      "loss": 0.1111,
+      "step": 670
+    },
+    {
+      "epoch": 0.1525,
+      "grad_norm": 2.2191005794886607,
+      "learning_rate": 1.2471416020184332e-06,
+      "loss": 0.0975,
+      "step": 671
+    },
+    {
+      "epoch": 0.15272727272727274,
+      "grad_norm": 3.012846122888053,
+      "learning_rate": 1.2471330696536865e-06,
+      "loss": 0.1184,
+      "step": 672
+    },
+    {
+      "epoch": 0.15295454545454545,
+      "grad_norm": 1.911663153816239,
+      "learning_rate": 1.247124524602561e-06,
+      "loss": 0.0929,
+      "step": 673
+    },
+    {
+      "epoch": 0.15318181818181817,
+      "grad_norm": 1.3331888331028157,
+      "learning_rate": 1.247115966865231e-06,
+      "loss": 0.0675,
+      "step": 674
+    },
+    {
+      "epoch": 0.1534090909090909,
+      "grad_norm": 2.2353017797821093,
+      "learning_rate": 1.2471073964418713e-06,
+      "loss": 0.1356,
+      "step": 675
+    },
+    {
+      "epoch": 0.15363636363636363,
+      "grad_norm": 2.6151316389217856,
+      "learning_rate": 1.2470988133326564e-06,
+      "loss": 0.0991,
+      "step": 676
+    },
+    {
+      "epoch": 0.15386363636363637,
+      "grad_norm": 1.6869842630512377,
+      "learning_rate": 1.2470902175377614e-06,
+      "loss": 0.08,
+      "step": 677
+    },
+    {
+      "epoch": 0.15409090909090908,
+      "grad_norm": 2.37201518523295,
+      "learning_rate": 1.2470816090573615e-06,
+      "loss": 0.1069,
+      "step": 678
+    },
+    {
+      "epoch": 0.15431818181818183,
+      "grad_norm": 1.1942903252980706,
+      "learning_rate": 1.2470729878916325e-06,
+      "loss": 0.0438,
+      "step": 679
+    },
+    {
+      "epoch": 0.15454545454545454,
+      "grad_norm": 2.596435232906105,
+      "learning_rate": 1.2470643540407499e-06,
+      "loss": 0.1517,
+      "step": 680
+    },
+    {
+      "epoch": 0.15477272727272728,
+      "grad_norm": 2.0573267514492874,
+      "learning_rate": 1.2470557075048897e-06,
+      "loss": 0.0988,
+      "step": 681
+    },
+    {
+      "epoch": 0.155,
+      "grad_norm": 1.853468244206894,
+      "learning_rate": 1.2470470482842285e-06,
+      "loss": 0.1619,
+      "step": 682
+    },
+    {
+      "epoch": 0.1552272727272727,
+      "grad_norm": 1.7327636164768434,
+      "learning_rate": 1.2470383763789428e-06,
+      "loss": 0.068,
+      "step": 683
+    },
+    {
+      "epoch": 0.15545454545454546,
+      "grad_norm": 2.3722883139266107,
+      "learning_rate": 1.2470296917892095e-06,
+      "loss": 0.0983,
+      "step": 684
+    },
+    {
+      "epoch": 0.15568181818181817,
+      "grad_norm": 1.1202170284199937,
+      "learning_rate": 1.2470209945152053e-06,
+      "loss": 0.0687,
+      "step": 685
+    },
+    {
+      "epoch": 0.1559090909090909,
+      "grad_norm": 1.8442094662230453,
+      "learning_rate": 1.2470122845571081e-06,
+      "loss": 0.0588,
+      "step": 686
+    },
+    {
+      "epoch": 0.15613636363636363,
+      "grad_norm": 2.360700094306487,
+      "learning_rate": 1.2470035619150952e-06,
+      "loss": 0.0922,
+      "step": 687
+    },
+    {
+      "epoch": 0.15636363636363637,
+      "grad_norm": 2.4571841665864476,
+      "learning_rate": 1.2469948265893445e-06,
+      "loss": 0.1343,
+      "step": 688
+    },
+    {
+      "epoch": 0.1565909090909091,
+      "grad_norm": 2.2244116299082064,
+      "learning_rate": 1.2469860785800342e-06,
+      "loss": 0.063,
+      "step": 689
+    },
+    {
+      "epoch": 0.15681818181818183,
+      "grad_norm": 2.8458504371522646,
+      "learning_rate": 1.2469773178873427e-06,
+      "loss": 0.1385,
+      "step": 690
+    },
+    {
+      "epoch": 0.15704545454545454,
+      "grad_norm": 2.1225914992930583,
+      "learning_rate": 1.2469685445114486e-06,
+      "loss": 0.1085,
+      "step": 691
+    },
+    {
+      "epoch": 0.1572727272727273,
+      "grad_norm": 1.6884250174810709,
+      "learning_rate": 1.2469597584525305e-06,
+      "loss": 0.0409,
+      "step": 692
+    },
+    {
+      "epoch": 0.1575,
+      "grad_norm": 2.1449937048619816,
+      "learning_rate": 1.246950959710768e-06,
+      "loss": 0.1136,
+      "step": 693
+    },
+    {
+      "epoch": 0.15772727272727272,
+      "grad_norm": 1.973158401452787,
+      "learning_rate": 1.2469421482863406e-06,
+      "loss": 0.0798,
+      "step": 694
+    },
+    {
+      "epoch": 0.15795454545454546,
+      "grad_norm": 2.7581864294647587,
+      "learning_rate": 1.2469333241794275e-06,
+      "loss": 0.1413,
+      "step": 695
+    },
+    {
+      "epoch": 0.15818181818181817,
+      "grad_norm": 1.8131709006177439,
+      "learning_rate": 1.2469244873902089e-06,
+      "loss": 0.0712,
+      "step": 696
+    },
+    {
+      "epoch": 0.15840909090909092,
+      "grad_norm": 1.9747534462388079,
+      "learning_rate": 1.246915637918865e-06,
+      "loss": 0.0863,
+      "step": 697
+    },
+    {
+      "epoch": 0.15863636363636363,
+      "grad_norm": 1.8364990024537176,
+      "learning_rate": 1.2469067757655765e-06,
+      "loss": 0.0783,
+      "step": 698
+    },
+    {
+      "epoch": 0.15886363636363637,
+      "grad_norm": 0.5444108510197619,
+      "learning_rate": 1.2468979009305238e-06,
+      "loss": 0.0233,
+      "step": 699
+    },
+    {
+      "epoch": 0.1590909090909091,
+      "grad_norm": 2.2242593303184957,
+      "learning_rate": 1.2468890134138878e-06,
+      "loss": 0.0711,
+      "step": 700
+    },
+    {
+      "epoch": 0.15931818181818183,
+      "grad_norm": 1.8183007003715497,
+      "learning_rate": 1.2468801132158499e-06,
+      "loss": 0.0758,
+      "step": 701
+    },
+    {
+      "epoch": 0.15954545454545455,
+      "grad_norm": 1.7379291682458051,
+      "learning_rate": 1.2468712003365916e-06,
+      "loss": 0.0621,
+      "step": 702
+    },
+    {
+      "epoch": 0.15977272727272726,
+      "grad_norm": 1.8413168189925682,
+      "learning_rate": 1.2468622747762945e-06,
+      "loss": 0.0609,
+      "step": 703
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.277500095608216,
+      "learning_rate": 1.2468533365351408e-06,
+      "loss": 0.0434,
+      "step": 704
+    },
+    {
+      "epoch": 0.16022727272727272,
+      "grad_norm": 1.6201568926129486,
+      "learning_rate": 1.2468443856133129e-06,
+      "loss": 0.0754,
+      "step": 705
+    },
+    {
+      "epoch": 0.16045454545454546,
+      "grad_norm": 1.3704095867486445,
+      "learning_rate": 1.2468354220109926e-06,
+      "loss": 0.0543,
+      "step": 706
+    },
+    {
+      "epoch": 0.16068181818181818,
+      "grad_norm": 1.6778138932296198,
+      "learning_rate": 1.2468264457283636e-06,
+      "loss": 0.0603,
+      "step": 707
+    },
+    {
+      "epoch": 0.16090909090909092,
+      "grad_norm": 1.7640770850950427,
+      "learning_rate": 1.2468174567656084e-06,
+      "loss": 0.0728,
+      "step": 708
+    },
+    {
+      "epoch": 0.16113636363636363,
+      "grad_norm": 1.818469561000814,
+      "learning_rate": 1.2468084551229106e-06,
+      "loss": 0.1046,
+      "step": 709
+    },
+    {
+      "epoch": 0.16136363636363638,
+      "grad_norm": 2.745686350898046,
+      "learning_rate": 1.2467994408004533e-06,
+      "loss": 0.0802,
+      "step": 710
+    },
+    {
+      "epoch": 0.1615909090909091,
+      "grad_norm": 3.5548511834444008,
+      "learning_rate": 1.2467904137984208e-06,
+      "loss": 0.1452,
+      "step": 711
+    },
+    {
+      "epoch": 0.1618181818181818,
+      "grad_norm": 2.085672767139168,
+      "learning_rate": 1.2467813741169971e-06,
+      "loss": 0.0729,
+      "step": 712
+    },
+    {
+      "epoch": 0.16204545454545455,
+      "grad_norm": 2.53532963285559,
+      "learning_rate": 1.2467723217563663e-06,
+      "loss": 0.1017,
+      "step": 713
+    },
+    {
+      "epoch": 0.16227272727272726,
+      "grad_norm": 1.5052081870977931,
+      "learning_rate": 1.246763256716713e-06,
+      "loss": 0.0702,
+      "step": 714
+    },
+    {
+      "epoch": 0.1625,
+      "grad_norm": 1.6940127695836484,
+      "learning_rate": 1.2467541789982223e-06,
+      "loss": 0.0561,
+      "step": 715
+    },
+    {
+      "epoch": 0.16272727272727272,
+      "grad_norm": 3.198035281802877,
+      "learning_rate": 1.246745088601079e-06,
+      "loss": 0.0906,
+      "step": 716
+    },
+    {
+      "epoch": 0.16295454545454546,
+      "grad_norm": 2.6660334187654686,
+      "learning_rate": 1.246735985525469e-06,
+      "loss": 0.1573,
+      "step": 717
+    },
+    {
+      "epoch": 0.16318181818181818,
+      "grad_norm": 1.3249472418058863,
+      "learning_rate": 1.2467268697715774e-06,
+      "loss": 0.0621,
+      "step": 718
+    },
+    {
+      "epoch": 0.16340909090909092,
+      "grad_norm": 1.8919433829896126,
+      "learning_rate": 1.2467177413395903e-06,
+      "loss": 0.1287,
+      "step": 719
+    },
+    {
+      "epoch": 0.16363636363636364,
+      "grad_norm": 1.7520889372081105,
+      "learning_rate": 1.2467086002296937e-06,
+      "loss": 0.0702,
+      "step": 720
+    },
+    {
+      "epoch": 0.16386363636363635,
+      "grad_norm": 2.6554377418784267,
+      "learning_rate": 1.246699446442074e-06,
+      "loss": 0.1673,
+      "step": 721
+    },
+    {
+      "epoch": 0.1640909090909091,
+      "grad_norm": 2.44412543576879,
+      "learning_rate": 1.246690279976918e-06,
+      "loss": 0.1407,
+      "step": 722
+    },
+    {
+      "epoch": 0.1643181818181818,
+      "grad_norm": 2.0026484930382047,
+      "learning_rate": 1.246681100834413e-06,
+      "loss": 0.0756,
+      "step": 723
+    },
+    {
+      "epoch": 0.16454545454545455,
+      "grad_norm": 2.887921224637937,
+      "learning_rate": 1.2466719090147455e-06,
+      "loss": 0.1512,
+      "step": 724
+    },
+    {
+      "epoch": 0.16477272727272727,
+      "grad_norm": 2.4967984224246385,
+      "learning_rate": 1.2466627045181032e-06,
+      "loss": 0.1368,
+      "step": 725
+    },
+    {
+      "epoch": 0.165,
+      "grad_norm": 1.353027925506024,
+      "learning_rate": 1.2466534873446738e-06,
+      "loss": 0.0815,
+      "step": 726
+    },
+    {
+      "epoch": 0.16522727272727272,
+      "grad_norm": 2.375041934717055,
+      "learning_rate": 1.2466442574946452e-06,
+      "loss": 0.1222,
+      "step": 727
+    },
+    {
+      "epoch": 0.16545454545454547,
+      "grad_norm": 1.6595233169614119,
+      "learning_rate": 1.246635014968206e-06,
+      "loss": 0.0829,
+      "step": 728
+    },
+    {
+      "epoch": 0.16568181818181818,
+      "grad_norm": 2.3247954780103344,
+      "learning_rate": 1.246625759765544e-06,
+      "loss": 0.0549,
+      "step": 729
+    },
+    {
+      "epoch": 0.16590909090909092,
+      "grad_norm": 1.7945460373131,
+      "learning_rate": 1.2466164918868484e-06,
+      "loss": 0.1079,
+      "step": 730
+    },
+    {
+      "epoch": 0.16613636363636364,
+      "grad_norm": 3.0146115872954438,
+      "learning_rate": 1.246607211332308e-06,
+      "loss": 0.1048,
+      "step": 731
+    },
+    {
+      "epoch": 0.16636363636363635,
+      "grad_norm": 1.4181800057813108,
+      "learning_rate": 1.246597918102112e-06,
+      "loss": 0.0516,
+      "step": 732
+    },
+    {
+      "epoch": 0.1665909090909091,
+      "grad_norm": 1.768582600208251,
+      "learning_rate": 1.24658861219645e-06,
+      "loss": 0.1111,
+      "step": 733
+    },
+    {
+      "epoch": 0.1668181818181818,
+      "grad_norm": 2.640778925979072,
+      "learning_rate": 1.2465792936155122e-06,
+      "loss": 0.1432,
+      "step": 734
+    },
+    {
+      "epoch": 0.16704545454545455,
+      "grad_norm": 1.7793634841874189,
+      "learning_rate": 1.246569962359488e-06,
+      "loss": 0.0838,
+      "step": 735
+    },
+    {
+      "epoch": 0.16727272727272727,
+      "grad_norm": 2.1248403252902577,
+      "learning_rate": 1.2465606184285679e-06,
+      "loss": 0.0935,
+      "step": 736
+    },
+    {
+      "epoch": 0.1675,
+      "grad_norm": 2.071060726156227,
+      "learning_rate": 1.2465512618229422e-06,
+      "loss": 0.1126,
+      "step": 737
+    },
+    {
+      "epoch": 0.16772727272727272,
+      "grad_norm": 1.8364221430759058,
+      "learning_rate": 1.246541892542802e-06,
+      "loss": 0.0536,
+      "step": 738
+    },
+    {
+      "epoch": 0.16795454545454547,
+      "grad_norm": 1.5803689231912126,
+      "learning_rate": 1.2465325105883383e-06,
+      "loss": 0.0577,
+      "step": 739
+    },
+    {
+      "epoch": 0.16818181818181818,
+      "grad_norm": 1.7260620078901308,
+      "learning_rate": 1.2465231159597425e-06,
+      "loss": 0.0702,
+      "step": 740
+    },
+    {
+      "epoch": 0.1684090909090909,
+      "grad_norm": 2.2316046562719554,
+      "learning_rate": 1.2465137086572057e-06,
+      "loss": 0.0918,
+      "step": 741
+    },
+    {
+      "epoch": 0.16863636363636364,
+      "grad_norm": 0.4518584357308151,
+      "learning_rate": 1.2465042886809204e-06,
+      "loss": 0.0204,
+      "step": 742
+    },
+    {
+      "epoch": 0.16886363636363635,
+      "grad_norm": 1.8363916049561753,
+      "learning_rate": 1.2464948560310785e-06,
+      "loss": 0.0554,
+      "step": 743
+    },
+    {
+      "epoch": 0.1690909090909091,
+      "grad_norm": 3.111049195780237,
+      "learning_rate": 1.2464854107078719e-06,
+      "loss": 0.1251,
+      "step": 744
+    },
+    {
+      "epoch": 0.1693181818181818,
+      "grad_norm": 2.068485881479512,
+      "learning_rate": 1.2464759527114935e-06,
+      "loss": 0.1075,
+      "step": 745
+    },
+    {
+      "epoch": 0.16954545454545455,
+      "grad_norm": 1.8036047004546578,
+      "learning_rate": 1.2464664820421365e-06,
+      "loss": 0.0936,
+      "step": 746
+    },
+    {
+      "epoch": 0.16977272727272727,
+      "grad_norm": 1.4962593603717844,
+      "learning_rate": 1.2464569986999937e-06,
+      "loss": 0.0979,
+      "step": 747
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.451365257941085,
+      "learning_rate": 1.2464475026852585e-06,
+      "loss": 0.0425,
+      "step": 748
+    },
+    {
+      "epoch": 0.17022727272727273,
+      "grad_norm": 2.457764827736832,
+      "learning_rate": 1.2464379939981245e-06,
+      "loss": 0.0829,
+      "step": 749
+    },
+    {
+      "epoch": 0.17045454545454544,
+      "grad_norm": 1.2633251317232443,
+      "learning_rate": 1.2464284726387854e-06,
+      "loss": 0.0764,
+      "step": 750
+    },
+    {
+      "epoch": 0.17068181818181818,
+      "grad_norm": 1.6158314754749314,
+      "learning_rate": 1.246418938607436e-06,
+      "loss": 0.0715,
+      "step": 751
+    },
+    {
+      "epoch": 0.1709090909090909,
+      "grad_norm": 3.5599187777713803,
+      "learning_rate": 1.24640939190427e-06,
+      "loss": 0.1685,
+      "step": 752
+    },
+    {
+      "epoch": 0.17113636363636364,
+      "grad_norm": 1.6031511316386666,
+      "learning_rate": 1.2463998325294827e-06,
+      "loss": 0.0551,
+      "step": 753
+    },
+    {
+      "epoch": 0.17136363636363636,
+      "grad_norm": 1.7927603989632424,
+      "learning_rate": 1.2463902604832684e-06,
+      "loss": 0.0657,
+      "step": 754
+    },
+    {
+      "epoch": 0.1715909090909091,
+      "grad_norm": 1.3426986611301193,
+      "learning_rate": 1.2463806757658227e-06,
+      "loss": 0.0624,
+      "step": 755
+    },
+    {
+      "epoch": 0.17181818181818181,
+      "grad_norm": 1.3975514303034429,
+      "learning_rate": 1.246371078377341e-06,
+      "loss": 0.0538,
+      "step": 756
+    },
+    {
+      "epoch": 0.17204545454545456,
+      "grad_norm": 2.427432809597591,
+      "learning_rate": 1.246361468318019e-06,
+      "loss": 0.0889,
+      "step": 757
+    },
+    {
+      "epoch": 0.17227272727272727,
+      "grad_norm": 1.7137040545028923,
+      "learning_rate": 1.2463518455880528e-06,
+      "loss": 0.0768,
+      "step": 758
+    },
+    {
+      "epoch": 0.1725,
+      "grad_norm": 1.7607680638751106,
+      "learning_rate": 1.2463422101876382e-06,
+      "loss": 0.0663,
+      "step": 759
+    },
+    {
+      "epoch": 0.17272727272727273,
+      "grad_norm": 3.089357065863354,
+      "learning_rate": 1.2463325621169719e-06,
+      "loss": 0.1501,
+      "step": 760
+    },
+    {
+      "epoch": 0.17295454545454544,
+      "grad_norm": 1.8273441386058986,
+      "learning_rate": 1.246322901376251e-06,
+      "loss": 0.0519,
+      "step": 761
+    },
+    {
+      "epoch": 0.1731818181818182,
+      "grad_norm": 2.665614758159654,
+      "learning_rate": 1.246313227965672e-06,
+      "loss": 0.1769,
+      "step": 762
+    },
+    {
+      "epoch": 0.1734090909090909,
+      "grad_norm": 1.14731792024219,
+      "learning_rate": 1.2463035418854322e-06,
+      "loss": 0.0592,
+      "step": 763
+    },
+    {
+      "epoch": 0.17363636363636364,
+      "grad_norm": 2.183586909347763,
+      "learning_rate": 1.2462938431357294e-06,
+      "loss": 0.1096,
+      "step": 764
+    },
+    {
+      "epoch": 0.17386363636363636,
+      "grad_norm": 2.2753623998424803,
+      "learning_rate": 1.2462841317167611e-06,
+      "loss": 0.0811,
+      "step": 765
+    },
+    {
+      "epoch": 0.1740909090909091,
+      "grad_norm": 2.1733020566620485,
+      "learning_rate": 1.2462744076287257e-06,
+      "loss": 0.165,
+      "step": 766
+    },
+    {
+      "epoch": 0.17431818181818182,
+      "grad_norm": 1.9498182364711891,
+      "learning_rate": 1.2462646708718211e-06,
+      "loss": 0.109,
+      "step": 767
+    },
+    {
+      "epoch": 0.17454545454545456,
+      "grad_norm": 2.0461055836402458,
+      "learning_rate": 1.2462549214462463e-06,
+      "loss": 0.0655,
+      "step": 768
+    },
+    {
+      "epoch": 0.17477272727272727,
+      "grad_norm": 2.0512469499211607,
+      "learning_rate": 1.2462451593521995e-06,
+      "loss": 0.089,
+      "step": 769
+    },
+    {
+      "epoch": 0.175,
+      "grad_norm": 2.645129156503398,
+      "learning_rate": 1.24623538458988e-06,
+      "loss": 0.09,
+      "step": 770
+    },
+    {
+      "epoch": 0.17522727272727273,
+      "grad_norm": 1.5613406608754885,
+      "learning_rate": 1.2462255971594874e-06,
+      "loss": 0.055,
+      "step": 771
+    },
+    {
+      "epoch": 0.17545454545454545,
+      "grad_norm": 1.4597419864094543,
+      "learning_rate": 1.246215797061221e-06,
+      "loss": 0.0519,
+      "step": 772
+    },
+    {
+      "epoch": 0.1756818181818182,
+      "grad_norm": 1.875588684332424,
+      "learning_rate": 1.246205984295281e-06,
+      "loss": 0.0488,
+      "step": 773
+    },
+    {
+      "epoch": 0.1759090909090909,
+      "grad_norm": 2.2920693702864794,
+      "learning_rate": 1.2461961588618669e-06,
+      "loss": 0.1015,
+      "step": 774
+    },
+    {
+      "epoch": 0.17613636363636365,
+      "grad_norm": 1.974571432280148,
+      "learning_rate": 1.2461863207611793e-06,
+      "loss": 0.0772,
+      "step": 775
+    },
+    {
+      "epoch": 0.17636363636363636,
+      "grad_norm": 3.3121151949301546,
+      "learning_rate": 1.2461764699934192e-06,
+      "loss": 0.1161,
+      "step": 776
+    },
+    {
+      "epoch": 0.1765909090909091,
+      "grad_norm": 2.2214476575414706,
+      "learning_rate": 1.246166606558787e-06,
+      "loss": 0.1432,
+      "step": 777
+    },
+    {
+      "epoch": 0.17681818181818182,
+      "grad_norm": 1.5497378987206079,
+      "learning_rate": 1.246156730457484e-06,
+      "loss": 0.0577,
+      "step": 778
+    },
+    {
+      "epoch": 0.17704545454545453,
+      "grad_norm": 1.8431792252298567,
+      "learning_rate": 1.2461468416897116e-06,
+      "loss": 0.0702,
+      "step": 779
+    },
+    {
+      "epoch": 0.17727272727272728,
+      "grad_norm": 1.5276395119205695,
+      "learning_rate": 1.2461369402556716e-06,
+      "loss": 0.056,
+      "step": 780
+    },
+    {
+      "epoch": 0.1775,
+      "grad_norm": 2.126104582881588,
+      "learning_rate": 1.2461270261555657e-06,
+      "loss": 0.0915,
+      "step": 781
+    },
+    {
+      "epoch": 0.17772727272727273,
+      "grad_norm": 1.84144161488585,
+      "learning_rate": 1.246117099389596e-06,
+      "loss": 0.0582,
+      "step": 782
+    },
+    {
+      "epoch": 0.17795454545454545,
+      "grad_norm": 1.9930193402593468,
+      "learning_rate": 1.2461071599579653e-06,
+      "loss": 0.135,
+      "step": 783
+    },
+    {
+      "epoch": 0.1781818181818182,
+      "grad_norm": 1.8489027687011008,
+      "learning_rate": 1.2460972078608757e-06,
+      "loss": 0.118,
+      "step": 784
+    },
+    {
+      "epoch": 0.1784090909090909,
+      "grad_norm": 1.9687403583926306,
+      "learning_rate": 1.2460872430985307e-06,
+      "loss": 0.0747,
+      "step": 785
+    },
+    {
+      "epoch": 0.17863636363636365,
+      "grad_norm": 2.032483601401086,
+      "learning_rate": 1.246077265671133e-06,
+      "loss": 0.1481,
+      "step": 786
+    },
+    {
+      "epoch": 0.17886363636363636,
+      "grad_norm": 2.194597891592595,
+      "learning_rate": 1.2460672755788867e-06,
+      "loss": 0.1258,
+      "step": 787
+    },
+    {
+      "epoch": 0.17909090909090908,
+      "grad_norm": 2.5790885170525337,
+      "learning_rate": 1.246057272821995e-06,
+      "loss": 0.1163,
+      "step": 788
+    },
+    {
+      "epoch": 0.17931818181818182,
+      "grad_norm": 1.538735297316767,
+      "learning_rate": 1.2460472574006619e-06,
+      "loss": 0.059,
+      "step": 789
+    },
+    {
+      "epoch": 0.17954545454545454,
+      "grad_norm": 1.531851811801296,
+      "learning_rate": 1.2460372293150919e-06,
+      "loss": 0.0459,
+      "step": 790
+    },
+    {
+      "epoch": 0.17977272727272728,
+      "grad_norm": 2.5391582770532066,
+      "learning_rate": 1.2460271885654891e-06,
+      "loss": 0.0606,
+      "step": 791
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.580247230663286,
+      "learning_rate": 1.2460171351520585e-06,
+      "loss": 0.2996,
+      "step": 792
+    },
+    {
+      "epoch": 0.18022727272727274,
+      "grad_norm": 3.2403422677842033,
+      "learning_rate": 1.2460070690750053e-06,
+      "loss": 0.2265,
+      "step": 793
+    },
+    {
+      "epoch": 0.18045454545454545,
+      "grad_norm": 1.9657007492903837,
+      "learning_rate": 1.2459969903345345e-06,
+      "loss": 0.0668,
+      "step": 794
+    },
+    {
+      "epoch": 0.1806818181818182,
+      "grad_norm": 1.4695504452415253,
+      "learning_rate": 1.2459868989308516e-06,
+      "loss": 0.0685,
+      "step": 795
+    },
+    {
+      "epoch": 0.1809090909090909,
+      "grad_norm": 2.4100621493395273,
+      "learning_rate": 1.2459767948641626e-06,
+      "loss": 0.1018,
+      "step": 796
+    },
+    {
+      "epoch": 0.18113636363636362,
+      "grad_norm": 2.487316454142042,
+      "learning_rate": 1.2459666781346732e-06,
+      "loss": 0.1463,
+      "step": 797
+    },
+    {
+      "epoch": 0.18136363636363637,
+      "grad_norm": 1.3787066354377613,
+      "learning_rate": 1.2459565487425901e-06,
+      "loss": 0.0783,
+      "step": 798
+    },
+    {
+      "epoch": 0.18159090909090908,
+      "grad_norm": 1.6240495266411359,
+      "learning_rate": 1.2459464066881196e-06,
+      "loss": 0.096,
+      "step": 799
+    },
+    {
+      "epoch": 0.18181818181818182,
+      "grad_norm": 3.270963622462829,
+      "learning_rate": 1.2459362519714686e-06,
+      "loss": 0.1159,
+      "step": 800
+    },
+    {
+      "epoch": 0.18204545454545454,
+      "grad_norm": 2.210646108623395,
+      "learning_rate": 1.2459260845928438e-06,
+      "loss": 0.1267,
+      "step": 801
+    },
+    {
+      "epoch": 0.18227272727272728,
+      "grad_norm": 1.9663225885729052,
+      "learning_rate": 1.2459159045524532e-06,
+      "loss": 0.0771,
+      "step": 802
+    },
+    {
+      "epoch": 0.1825,
+      "grad_norm": 2.334663987041024,
+      "learning_rate": 1.2459057118505042e-06,
+      "loss": 0.094,
+      "step": 803
+    },
+    {
+      "epoch": 0.18272727272727274,
+      "grad_norm": 2.3756707225045366,
+      "learning_rate": 1.245895506487204e-06,
+      "loss": 0.0905,
+      "step": 804
+    },
+    {
+      "epoch": 0.18295454545454545,
+      "grad_norm": 2.1627088224380566,
+      "learning_rate": 1.2458852884627618e-06,
+      "loss": 0.0801,
+      "step": 805
+    },
+    {
+      "epoch": 0.1831818181818182,
+      "grad_norm": 3.1482234778185947,
+      "learning_rate": 1.245875057777385e-06,
+      "loss": 0.1015,
+      "step": 806
+    },
+    {
+      "epoch": 0.1834090909090909,
+      "grad_norm": 2.5941248809020805,
+      "learning_rate": 1.2458648144312826e-06,
+      "loss": 0.1103,
+      "step": 807
+    },
+    {
+      "epoch": 0.18363636363636363,
+      "grad_norm": 1.9652008020516802,
+      "learning_rate": 1.2458545584246637e-06,
+      "loss": 0.0786,
+      "step": 808
+    },
+    {
+      "epoch": 0.18386363636363637,
+      "grad_norm": 1.2058349569686857,
+      "learning_rate": 1.245844289757737e-06,
+      "loss": 0.0766,
+      "step": 809
+    },
+    {
+      "epoch": 0.18409090909090908,
+      "grad_norm": 1.412079924321638,
+      "learning_rate": 1.2458340084307121e-06,
+      "loss": 0.0407,
+      "step": 810
+    },
+    {
+      "epoch": 0.18431818181818183,
+      "grad_norm": 2.133034298574448,
+      "learning_rate": 1.245823714443799e-06,
+      "loss": 0.1399,
+      "step": 811
+    },
+    {
+      "epoch": 0.18454545454545454,
+      "grad_norm": 1.8028274350146942,
+      "learning_rate": 1.245813407797207e-06,
+      "loss": 0.0947,
+      "step": 812
+    },
+    {
+      "epoch": 0.18477272727272728,
+      "grad_norm": 2.515995403059933,
+      "learning_rate": 1.2458030884911465e-06,
+      "loss": 0.096,
+      "step": 813
+    },
+    {
+      "epoch": 0.185,
+      "grad_norm": 1.788049060468717,
+      "learning_rate": 1.2457927565258279e-06,
+      "loss": 0.0762,
+      "step": 814
+    },
+    {
+      "epoch": 0.18522727272727274,
+      "grad_norm": 2.375670668838167,
+      "learning_rate": 1.2457824119014623e-06,
+      "loss": 0.1635,
+      "step": 815
+    },
+    {
+      "epoch": 0.18545454545454546,
+      "grad_norm": 1.5480533546565594,
+      "learning_rate": 1.24577205461826e-06,
+      "loss": 0.0837,
+      "step": 816
+    },
+    {
+      "epoch": 0.18568181818181817,
+      "grad_norm": 2.310392336980972,
+      "learning_rate": 1.2457616846764326e-06,
+      "loss": 0.09,
+      "step": 817
+    },
+    {
+      "epoch": 0.1859090909090909,
+      "grad_norm": 2.0748131612246614,
+      "learning_rate": 1.2457513020761916e-06,
+      "loss": 0.1098,
+      "step": 818
+    },
+    {
+      "epoch": 0.18613636363636363,
+      "grad_norm": 1.7275465770316718,
+      "learning_rate": 1.2457409068177484e-06,
+      "loss": 0.0872,
+      "step": 819
+    },
+    {
+      "epoch": 0.18636363636363637,
+      "grad_norm": 3.0592300729617747,
+      "learning_rate": 1.2457304989013153e-06,
+      "loss": 0.0967,
+      "step": 820
+    },
+    {
+      "epoch": 0.18659090909090909,
+      "grad_norm": 2.944692938824599,
+      "learning_rate": 1.2457200783271044e-06,
+      "loss": 0.1643,
+      "step": 821
+    },
+    {
+      "epoch": 0.18681818181818183,
+      "grad_norm": 1.547623000268646,
+      "learning_rate": 1.2457096450953282e-06,
+      "loss": 0.0845,
+      "step": 822
+    },
+    {
+      "epoch": 0.18704545454545454,
+      "grad_norm": 1.5108730293556807,
+      "learning_rate": 1.2456991992061993e-06,
+      "loss": 0.0432,
+      "step": 823
+    },
+    {
+      "epoch": 0.18727272727272729,
+      "grad_norm": 2.0771404552395785,
+      "learning_rate": 1.245688740659931e-06,
+      "loss": 0.0718,
+      "step": 824
+    },
+    {
+      "epoch": 0.1875,
+      "grad_norm": 2.0758513683189816,
+      "learning_rate": 1.2456782694567363e-06,
+      "loss": 0.0755,
+      "step": 825
+    },
+    {
+      "epoch": 0.18772727272727271,
+      "grad_norm": 2.051723322647701,
+      "learning_rate": 1.245667785596829e-06,
+      "loss": 0.0663,
+      "step": 826
+    },
+    {
+      "epoch": 0.18795454545454546,
+      "grad_norm": 1.440640444615283,
+      "learning_rate": 1.2456572890804228e-06,
+      "loss": 0.053,
+      "step": 827
+    },
+    {
+      "epoch": 0.18818181818181817,
+      "grad_norm": 1.993162031386313,
+      "learning_rate": 1.2456467799077315e-06,
+      "loss": 0.096,
+      "step": 828
+    },
+    {
+      "epoch": 0.18840909090909091,
+      "grad_norm": 2.415169459104793,
+      "learning_rate": 1.2456362580789697e-06,
+      "loss": 0.0799,
+      "step": 829
+    },
+    {
+      "epoch": 0.18863636363636363,
+      "grad_norm": 1.3404224302527155,
+      "learning_rate": 1.2456257235943516e-06,
+      "loss": 0.0514,
+      "step": 830
+    },
+    {
+      "epoch": 0.18886363636363637,
+      "grad_norm": 3.356720333121273,
+      "learning_rate": 1.2456151764540924e-06,
+      "loss": 0.1299,
+      "step": 831
+    },
+    {
+      "epoch": 0.1890909090909091,
+      "grad_norm": 2.0039472137487713,
+      "learning_rate": 1.2456046166584072e-06,
+      "loss": 0.1139,
+      "step": 832
+    },
+    {
+      "epoch": 0.18931818181818183,
+      "grad_norm": 2.6009341480246513,
+      "learning_rate": 1.245594044207511e-06,
+      "loss": 0.0829,
+      "step": 833
+    },
+    {
+      "epoch": 0.18954545454545454,
+      "grad_norm": 2.5820200751768514,
+      "learning_rate": 1.2455834591016196e-06,
+      "loss": 0.0953,
+      "step": 834
+    },
+    {
+      "epoch": 0.18977272727272726,
+      "grad_norm": 2.2113229131705086,
+      "learning_rate": 1.2455728613409487e-06,
+      "loss": 0.1461,
+      "step": 835
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0618043694022776,
+      "learning_rate": 1.2455622509257147e-06,
+      "loss": 0.0958,
+      "step": 836
+    },
+    {
+      "epoch": 0.19022727272727272,
+      "grad_norm": 0.6165205395481037,
+      "learning_rate": 1.2455516278561336e-06,
+      "loss": 0.0329,
+      "step": 837
+    },
+    {
+      "epoch": 0.19045454545454546,
+      "grad_norm": 2.0890932150423516,
+      "learning_rate": 1.2455409921324222e-06,
+      "loss": 0.0588,
+      "step": 838
+    },
+    {
+      "epoch": 0.19068181818181817,
+      "grad_norm": 2.1611832555008625,
+      "learning_rate": 1.2455303437547974e-06,
+      "loss": 0.0857,
+      "step": 839
+    },
+    {
+      "epoch": 0.19090909090909092,
+      "grad_norm": 1.5163575523147557,
+      "learning_rate": 1.2455196827234765e-06,
+      "loss": 0.0534,
+      "step": 840
+    },
+    {
+      "epoch": 0.19113636363636363,
+      "grad_norm": 2.57089747228702,
+      "learning_rate": 1.2455090090386765e-06,
+      "loss": 0.0863,
+      "step": 841
+    },
+    {
+      "epoch": 0.19136363636363637,
+      "grad_norm": 1.6540694522919728,
+      "learning_rate": 1.2454983227006153e-06,
+      "loss": 0.132,
+      "step": 842
+    },
+    {
+      "epoch": 0.1915909090909091,
+      "grad_norm": 1.9523118846949976,
+      "learning_rate": 1.2454876237095109e-06,
+      "loss": 0.0824,
+      "step": 843
+    },
+    {
+      "epoch": 0.1918181818181818,
+      "grad_norm": 1.8159616100245186,
+      "learning_rate": 1.2454769120655814e-06,
+      "loss": 0.0685,
+      "step": 844
+    },
+    {
+      "epoch": 0.19204545454545455,
+      "grad_norm": 1.6495482302059408,
+      "learning_rate": 1.245466187769045e-06,
+      "loss": 0.0827,
+      "step": 845
+    },
+    {
+      "epoch": 0.19227272727272726,
+      "grad_norm": 2.5919909761038764,
+      "learning_rate": 1.2454554508201205e-06,
+      "loss": 0.214,
+      "step": 846
+    },
+    {
+      "epoch": 0.1925,
+      "grad_norm": 2.022243414952102,
+      "learning_rate": 1.245444701219027e-06,
+      "loss": 0.0868,
+      "step": 847
+    },
+    {
+      "epoch": 0.19272727272727272,
+      "grad_norm": 2.22304175820554,
+      "learning_rate": 1.2454339389659837e-06,
+      "loss": 0.1009,
+      "step": 848
+    },
+    {
+      "epoch": 0.19295454545454546,
+      "grad_norm": 0.502552207230988,
+      "learning_rate": 1.24542316406121e-06,
+      "loss": 0.0309,
+      "step": 849
+    },
+    {
+      "epoch": 0.19318181818181818,
+      "grad_norm": 1.6375401716795237,
+      "learning_rate": 1.2454123765049256e-06,
+      "loss": 0.0825,
+      "step": 850
+    },
+    {
+      "epoch": 0.19340909090909092,
+      "grad_norm": 2.2630782882257763,
+      "learning_rate": 1.2454015762973505e-06,
+      "loss": 0.082,
+      "step": 851
+    },
+    {
+      "epoch": 0.19363636363636363,
+      "grad_norm": 2.9362466555203004,
+      "learning_rate": 1.2453907634387046e-06,
+      "loss": 0.1022,
+      "step": 852
+    },
+    {
+      "epoch": 0.19386363636363638,
+      "grad_norm": 1.8535158411327264,
+      "learning_rate": 1.2453799379292088e-06,
+      "loss": 0.0753,
+      "step": 853
+    },
+    {
+      "epoch": 0.1940909090909091,
+      "grad_norm": 1.137470685543181,
+      "learning_rate": 1.245369099769084e-06,
+      "loss": 0.0613,
+      "step": 854
+    },
+    {
+      "epoch": 0.1943181818181818,
+      "grad_norm": 2.238333322290331,
+      "learning_rate": 1.245358248958551e-06,
+      "loss": 0.0914,
+      "step": 855
+    },
+    {
+      "epoch": 0.19454545454545455,
+      "grad_norm": 1.4984942264509895,
+      "learning_rate": 1.2453473854978307e-06,
+      "loss": 0.1039,
+      "step": 856
+    },
+    {
+      "epoch": 0.19477272727272726,
+      "grad_norm": 2.9763586259406134,
+      "learning_rate": 1.245336509387145e-06,
+      "loss": 0.1003,
+      "step": 857
+    },
+    {
+      "epoch": 0.195,
+      "grad_norm": 3.2191775911726506,
+      "learning_rate": 1.2453256206267158e-06,
+      "loss": 0.1105,
+      "step": 858
+    },
+    {
+      "epoch": 0.19522727272727272,
+      "grad_norm": 1.1484381439991904,
+      "learning_rate": 1.245314719216765e-06,
+      "loss": 0.0408,
+      "step": 859
+    },
+    {
+      "epoch": 0.19545454545454546,
+      "grad_norm": 1.985380925147218,
+      "learning_rate": 1.2453038051575148e-06,
+      "loss": 0.0704,
+      "step": 860
+    },
+    {
+      "epoch": 0.19568181818181818,
+      "grad_norm": 2.514907725491543,
+      "learning_rate": 1.2452928784491877e-06,
+      "loss": 0.1124,
+      "step": 861
+    },
+    {
+      "epoch": 0.19590909090909092,
+      "grad_norm": 2.8703913354406843,
+      "learning_rate": 1.2452819390920068e-06,
+      "loss": 0.139,
+      "step": 862
+    },
+    {
+      "epoch": 0.19613636363636364,
+      "grad_norm": 2.292500813388667,
+      "learning_rate": 1.2452709870861952e-06,
+      "loss": 0.0671,
+      "step": 863
+    },
+    {
+      "epoch": 0.19636363636363635,
+      "grad_norm": 1.8648027783938061,
+      "learning_rate": 1.2452600224319758e-06,
+      "loss": 0.0662,
+      "step": 864
+    },
+    {
+      "epoch": 0.1965909090909091,
+      "grad_norm": 2.298323522002875,
+      "learning_rate": 1.2452490451295724e-06,
+      "loss": 0.0735,
+      "step": 865
+    },
+    {
+      "epoch": 0.1968181818181818,
+      "grad_norm": 2.832905047902425,
+      "learning_rate": 1.245238055179209e-06,
+      "loss": 0.1126,
+      "step": 866
+    },
+    {
+      "epoch": 0.19704545454545455,
+      "grad_norm": 1.800078111709904,
+      "learning_rate": 1.2452270525811095e-06,
+      "loss": 0.091,
+      "step": 867
+    },
+    {
+      "epoch": 0.19727272727272727,
+      "grad_norm": 2.023383671983363,
+      "learning_rate": 1.2452160373354985e-06,
+      "loss": 0.0817,
+      "step": 868
+    },
+    {
+      "epoch": 0.1975,
+      "grad_norm": 1.6950221941784396,
+      "learning_rate": 1.2452050094426006e-06,
+      "loss": 0.0629,
+      "step": 869
+    },
+    {
+      "epoch": 0.19772727272727272,
+      "grad_norm": 1.6903106587634438,
+      "learning_rate": 1.2451939689026401e-06,
+      "loss": 0.0732,
+      "step": 870
+    },
+    {
+      "epoch": 0.19795454545454547,
+      "grad_norm": 2.0700623262100226,
+      "learning_rate": 1.245182915715843e-06,
+      "loss": 0.0925,
+      "step": 871
+    },
+    {
+      "epoch": 0.19818181818181818,
+      "grad_norm": 3.2165682016635615,
+      "learning_rate": 1.2451718498824339e-06,
+      "loss": 0.1919,
+      "step": 872
+    },
+    {
+      "epoch": 0.1984090909090909,
+      "grad_norm": 2.1620700590122817,
+      "learning_rate": 1.245160771402639e-06,
+      "loss": 0.0851,
+      "step": 873
+    },
+    {
+      "epoch": 0.19863636363636364,
+      "grad_norm": 0.7664395568348693,
+      "learning_rate": 1.2451496802766843e-06,
+      "loss": 0.0385,
+      "step": 874
+    },
+    {
+      "epoch": 0.19886363636363635,
+      "grad_norm": 1.9374868963560667,
+      "learning_rate": 1.2451385765047955e-06,
+      "loss": 0.0488,
+      "step": 875
+    },
+    {
+      "epoch": 0.1990909090909091,
+      "grad_norm": 3.017359103777585,
+      "learning_rate": 1.2451274600871991e-06,
+      "loss": 0.1698,
+      "step": 876
+    },
+    {
+      "epoch": 0.1993181818181818,
+      "grad_norm": 1.7907471804311255,
+      "learning_rate": 1.245116331024122e-06,
+      "loss": 0.1037,
+      "step": 877
+    },
+    {
+      "epoch": 0.19954545454545455,
+      "grad_norm": 3.6572907738840126,
+      "learning_rate": 1.245105189315791e-06,
+      "loss": 0.1072,
+      "step": 878
+    },
+    {
+      "epoch": 0.19977272727272727,
+      "grad_norm": 2.41426903201482,
+      "learning_rate": 1.2450940349624337e-06,
+      "loss": 0.1125,
+      "step": 879
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4924254129068104,
+      "learning_rate": 1.2450828679642769e-06,
+      "loss": 0.092,
+      "step": 880
+    },
+    {
+      "epoch": 0.20022727272727273,
+      "grad_norm": 1.5945594482156806,
+      "learning_rate": 1.245071688321549e-06,
+      "loss": 0.1272,
+      "step": 881
+    },
+    {
+      "epoch": 0.20045454545454544,
+      "grad_norm": 1.4391837001441818,
+      "learning_rate": 1.245060496034477e-06,
+      "loss": 0.0387,
+      "step": 882
+    },
+    {
+      "epoch": 0.20068181818181818,
+      "grad_norm": 4.409140234512881,
+      "learning_rate": 1.24504929110329e-06,
+      "loss": 0.1808,
+      "step": 883
+    },
+    {
+      "epoch": 0.2009090909090909,
+      "grad_norm": 1.9932174980679984,
+      "learning_rate": 1.2450380735282164e-06,
+      "loss": 0.0949,
+      "step": 884
+    },
+    {
+      "epoch": 0.20113636363636364,
+      "grad_norm": 2.180469174564273,
+      "learning_rate": 1.2450268433094845e-06,
+      "loss": 0.0756,
+      "step": 885
+    },
+    {
+      "epoch": 0.20136363636363636,
+      "grad_norm": 3.5077492884511643,
+      "learning_rate": 1.2450156004473238e-06,
+      "loss": 0.1404,
+      "step": 886
+    },
+    {
+      "epoch": 0.2015909090909091,
+      "grad_norm": 0.6556441757304641,
+      "learning_rate": 1.2450043449419633e-06,
+      "loss": 0.0253,
+      "step": 887
+    },
+    {
+      "epoch": 0.2018181818181818,
+      "grad_norm": 0.7083434842408496,
+      "learning_rate": 1.2449930767936324e-06,
+      "loss": 0.0176,
+      "step": 888
+    },
+    {
+      "epoch": 0.20204545454545456,
+      "grad_norm": 2.624087022100737,
+      "learning_rate": 1.244981796002561e-06,
+      "loss": 0.1505,
+      "step": 889
+    },
+    {
+      "epoch": 0.20227272727272727,
+      "grad_norm": 2.8426146361578075,
+      "learning_rate": 1.2449705025689792e-06,
+      "loss": 0.1451,
+      "step": 890
+    },
+    {
+      "epoch": 0.2025,
+      "grad_norm": 1.3856140661607137,
+      "learning_rate": 1.2449591964931173e-06,
+      "loss": 0.0861,
+      "step": 891
+    },
+    {
+      "epoch": 0.20272727272727273,
+      "grad_norm": 1.3084756250114238,
+      "learning_rate": 1.2449478777752059e-06,
+      "loss": 0.0856,
+      "step": 892
+    },
+    {
+      "epoch": 0.20295454545454544,
+      "grad_norm": 1.4982932744143858,
+      "learning_rate": 1.2449365464154755e-06,
+      "loss": 0.0768,
+      "step": 893
+    },
+    {
+      "epoch": 0.20318181818181819,
+      "grad_norm": 2.851559285029384,
+      "learning_rate": 1.2449252024141577e-06,
+      "loss": 0.1001,
+      "step": 894
+    },
+    {
+      "epoch": 0.2034090909090909,
+      "grad_norm": 1.4918313399019223,
+      "learning_rate": 1.2449138457714832e-06,
+      "loss": 0.0585,
+      "step": 895
+    },
+    {
+      "epoch": 0.20363636363636364,
+      "grad_norm": 2.286157022220786,
+      "learning_rate": 1.2449024764876841e-06,
+      "loss": 0.1508,
+      "step": 896
+    },
+    {
+      "epoch": 0.20386363636363636,
+      "grad_norm": 1.9754658614353284,
+      "learning_rate": 1.244891094562992e-06,
+      "loss": 0.0749,
+      "step": 897
+    },
+    {
+      "epoch": 0.2040909090909091,
+      "grad_norm": 1.8122672681126089,
+      "learning_rate": 1.2448796999976389e-06,
+      "loss": 0.0906,
+      "step": 898
+    },
+    {
+      "epoch": 0.20431818181818182,
+      "grad_norm": 1.4397609632622894,
+      "learning_rate": 1.2448682927918573e-06,
+      "loss": 0.0613,
+      "step": 899
+    },
+    {
+      "epoch": 0.20454545454545456,
+      "grad_norm": 2.2608289380704756,
+      "learning_rate": 1.24485687294588e-06,
+      "loss": 0.0826,
+      "step": 900
+    },
+    {
+      "epoch": 0.20477272727272727,
+      "grad_norm": 1.4453175061468178,
+      "learning_rate": 1.2448454404599393e-06,
+      "loss": 0.0556,
+      "step": 901
+    },
+    {
+      "epoch": 0.205,
+      "grad_norm": 1.495371440754041,
+      "learning_rate": 1.244833995334269e-06,
+      "loss": 0.0905,
+      "step": 902
+    },
+    {
+      "epoch": 0.20522727272727273,
+      "grad_norm": 1.8483657167947465,
+      "learning_rate": 1.244822537569102e-06,
+      "loss": 0.0877,
+      "step": 903
+    },
+    {
+      "epoch": 0.20545454545454545,
+      "grad_norm": 1.8732553289689535,
+      "learning_rate": 1.244811067164672e-06,
+      "loss": 0.0693,
+      "step": 904
+    },
+    {
+      "epoch": 0.2056818181818182,
+      "grad_norm": 0.6418848127053353,
+      "learning_rate": 1.2447995841212133e-06,
+      "loss": 0.0295,
+      "step": 905
+    },
+    {
+      "epoch": 0.2059090909090909,
+      "grad_norm": 2.5683680553009274,
+      "learning_rate": 1.2447880884389597e-06,
+      "loss": 0.0842,
+      "step": 906
+    },
+    {
+      "epoch": 0.20613636363636365,
+      "grad_norm": 1.976100190197234,
+      "learning_rate": 1.2447765801181455e-06,
+      "loss": 0.1249,
+      "step": 907
+    },
+    {
+      "epoch": 0.20636363636363636,
+      "grad_norm": 1.6645560231134262,
+      "learning_rate": 1.2447650591590058e-06,
+      "loss": 0.1058,
+      "step": 908
+    },
+    {
+      "epoch": 0.2065909090909091,
+      "grad_norm": 1.9369753292589644,
+      "learning_rate": 1.2447535255617751e-06,
+      "loss": 0.1032,
+      "step": 909
+    },
+    {
+      "epoch": 0.20681818181818182,
+      "grad_norm": 2.189295860215505,
+      "learning_rate": 1.244741979326689e-06,
+      "loss": 0.0793,
+      "step": 910
+    },
+    {
+      "epoch": 0.20704545454545453,
+      "grad_norm": 2.110815371388909,
+      "learning_rate": 1.2447304204539827e-06,
+      "loss": 0.071,
+      "step": 911
+    },
+    {
+      "epoch": 0.20727272727272728,
+      "grad_norm": 1.3847597933653566,
+      "learning_rate": 1.244718848943892e-06,
+      "loss": 0.0482,
+      "step": 912
+    },
+    {
+      "epoch": 0.2075,
+      "grad_norm": 1.3676333921495836,
+      "learning_rate": 1.2447072647966527e-06,
+      "loss": 0.0573,
+      "step": 913
+    },
+    {
+      "epoch": 0.20772727272727273,
+      "grad_norm": 2.3010388599364706,
+      "learning_rate": 1.244695668012501e-06,
+      "loss": 0.0844,
+      "step": 914
+    },
+    {
+      "epoch": 0.20795454545454545,
+      "grad_norm": 1.1450481530586065,
+      "learning_rate": 1.244684058591674e-06,
+      "loss": 0.0727,
+      "step": 915
+    },
+    {
+      "epoch": 0.2081818181818182,
+      "grad_norm": 2.567231457354851,
+      "learning_rate": 1.2446724365344076e-06,
+      "loss": 0.0949,
+      "step": 916
+    },
+    {
+      "epoch": 0.2084090909090909,
+      "grad_norm": 2.3605592242361775,
+      "learning_rate": 1.2446608018409389e-06,
+      "loss": 0.0849,
+      "step": 917
+    },
+    {
+      "epoch": 0.20863636363636365,
+      "grad_norm": 2.373421703867287,
+      "learning_rate": 1.2446491545115057e-06,
+      "loss": 0.1011,
+      "step": 918
+    },
+    {
+      "epoch": 0.20886363636363636,
+      "grad_norm": 2.8253627602643006,
+      "learning_rate": 1.2446374945463451e-06,
+      "loss": 0.15,
+      "step": 919
+    },
+    {
+      "epoch": 0.20909090909090908,
+      "grad_norm": 1.7511811763512868,
+      "learning_rate": 1.2446258219456952e-06,
+      "loss": 0.0518,
+      "step": 920
+    },
+    {
+      "epoch": 0.20931818181818182,
+      "grad_norm": 2.1200669548599547,
+      "learning_rate": 1.2446141367097936e-06,
+      "loss": 0.1018,
+      "step": 921
+    },
+    {
+      "epoch": 0.20954545454545453,
+      "grad_norm": 1.6616499701829381,
+      "learning_rate": 1.2446024388388788e-06,
+      "loss": 0.0632,
+      "step": 922
+    },
+    {
+      "epoch": 0.20977272727272728,
+      "grad_norm": 2.1924064512604153,
+      "learning_rate": 1.2445907283331892e-06,
+      "loss": 0.131,
+      "step": 923
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6307638573120253,
+      "learning_rate": 1.244579005192964e-06,
+      "loss": 0.0683,
+      "step": 924
+    },
+    {
+      "epoch": 0.21022727272727273,
+      "grad_norm": 1.8365134908251521,
+      "learning_rate": 1.2445672694184418e-06,
+      "loss": 0.1225,
+      "step": 925
+    },
+    {
+      "epoch": 0.21045454545454545,
+      "grad_norm": 1.3367549499884048,
+      "learning_rate": 1.244555521009862e-06,
+      "loss": 0.0427,
+      "step": 926
+    },
+    {
+      "epoch": 0.2106818181818182,
+      "grad_norm": 2.0189607646415997,
+      "learning_rate": 1.2445437599674642e-06,
+      "loss": 0.1053,
+      "step": 927
+    },
+    {
+      "epoch": 0.2109090909090909,
+      "grad_norm": 1.7693950170887596,
+      "learning_rate": 1.2445319862914885e-06,
+      "loss": 0.1181,
+      "step": 928
+    },
+    {
+      "epoch": 0.21113636363636365,
+      "grad_norm": 1.0861541135419073,
+      "learning_rate": 1.2445201999821747e-06,
+      "loss": 0.0356,
+      "step": 929
+    },
+    {
+      "epoch": 0.21136363636363636,
+      "grad_norm": 2.3556419842546235,
+      "learning_rate": 1.244508401039763e-06,
+      "loss": 0.1439,
+      "step": 930
+    },
+    {
+      "epoch": 0.21159090909090908,
+      "grad_norm": 2.233656611439468,
+      "learning_rate": 1.2444965894644946e-06,
+      "loss": 0.0778,
+      "step": 931
+    },
+    {
+      "epoch": 0.21181818181818182,
+      "grad_norm": 2.2217036552199536,
+      "learning_rate": 1.2444847652566097e-06,
+      "loss": 0.079,
+      "step": 932
+    },
+    {
+      "epoch": 0.21204545454545454,
+      "grad_norm": 4.070513450062511,
+      "learning_rate": 1.2444729284163499e-06,
+      "loss": 0.1858,
+      "step": 933
+    },
+    {
+      "epoch": 0.21227272727272728,
+      "grad_norm": 1.59610353403429,
+      "learning_rate": 1.2444610789439563e-06,
+      "loss": 0.0836,
+      "step": 934
+    },
+    {
+      "epoch": 0.2125,
+      "grad_norm": 2.031808147201299,
+      "learning_rate": 1.2444492168396702e-06,
+      "loss": 0.0623,
+      "step": 935
+    },
+    {
+      "epoch": 0.21272727272727274,
+      "grad_norm": 0.9905555113177986,
+      "learning_rate": 1.2444373421037345e-06,
+      "loss": 0.047,
+      "step": 936
+    },
+    {
+      "epoch": 0.21295454545454545,
+      "grad_norm": 4.583150810929171,
+      "learning_rate": 1.2444254547363902e-06,
+      "loss": 0.2304,
+      "step": 937
+    },
+    {
+      "epoch": 0.2131818181818182,
+      "grad_norm": 1.7031249046960089,
+      "learning_rate": 1.2444135547378804e-06,
+      "loss": 0.0648,
+      "step": 938
+    },
+    {
+      "epoch": 0.2134090909090909,
+      "grad_norm": 1.8754469511859237,
+      "learning_rate": 1.2444016421084477e-06,
+      "loss": 0.1577,
+      "step": 939
+    },
+    {
+      "epoch": 0.21363636363636362,
+      "grad_norm": 2.155744662121728,
+      "learning_rate": 1.244389716848335e-06,
+      "loss": 0.0877,
+      "step": 940
+    },
+    {
+      "epoch": 0.21386363636363637,
+      "grad_norm": 1.4297533642103426,
+      "learning_rate": 1.2443777789577852e-06,
+      "loss": 0.0493,
+      "step": 941
+    },
+    {
+      "epoch": 0.21409090909090908,
+      "grad_norm": 2.2204578840268616,
+      "learning_rate": 1.244365828437042e-06,
+      "loss": 0.0949,
+      "step": 942
+    },
+    {
+      "epoch": 0.21431818181818182,
+      "grad_norm": 1.74532010122703,
+      "learning_rate": 1.2443538652863487e-06,
+      "loss": 0.0445,
+      "step": 943
+    },
+    {
+      "epoch": 0.21454545454545454,
+      "grad_norm": 3.0026539093406472,
+      "learning_rate": 1.24434188950595e-06,
+      "loss": 0.0971,
+      "step": 944
+    },
+    {
+      "epoch": 0.21477272727272728,
+      "grad_norm": 2.4587149088485525,
+      "learning_rate": 1.2443299010960896e-06,
+      "loss": 0.0826,
+      "step": 945
+    },
+    {
+      "epoch": 0.215,
+      "grad_norm": 1.7333048058418912,
+      "learning_rate": 1.244317900057012e-06,
+      "loss": 0.0852,
+      "step": 946
+    },
+    {
+      "epoch": 0.21522727272727274,
+      "grad_norm": 2.064117739394454,
+      "learning_rate": 1.2443058863889618e-06,
+      "loss": 0.1003,
+      "step": 947
+    },
+    {
+      "epoch": 0.21545454545454545,
+      "grad_norm": 2.2154213034151438,
+      "learning_rate": 1.2442938600921841e-06,
+      "loss": 0.0833,
+      "step": 948
+    },
+    {
+      "epoch": 0.21568181818181817,
+      "grad_norm": 1.8808894459732466,
+      "learning_rate": 1.2442818211669245e-06,
+      "loss": 0.0967,
+      "step": 949
+    },
+    {
+      "epoch": 0.2159090909090909,
+      "grad_norm": 1.7775742145051256,
+      "learning_rate": 1.2442697696134278e-06,
+      "loss": 0.0748,
+      "step": 950
+    },
+    {
+      "epoch": 0.21613636363636363,
+      "grad_norm": 1.4160366267350482,
+      "learning_rate": 1.2442577054319405e-06,
+      "loss": 0.0681,
+      "step": 951
+    },
+    {
+      "epoch": 0.21636363636363637,
+      "grad_norm": 1.8001022921706598,
+      "learning_rate": 1.2442456286227081e-06,
+      "loss": 0.0791,
+      "step": 952
+    },
+    {
+      "epoch": 0.21659090909090908,
+      "grad_norm": 2.681796594477653,
+      "learning_rate": 1.244233539185977e-06,
+      "loss": 0.1595,
+      "step": 953
+    },
+    {
+      "epoch": 0.21681818181818183,
+      "grad_norm": 1.5852774025417709,
+      "learning_rate": 1.2442214371219935e-06,
+      "loss": 0.0714,
+      "step": 954
+    },
+    {
+      "epoch": 0.21704545454545454,
+      "grad_norm": 2.5036043307921596,
+      "learning_rate": 1.2442093224310047e-06,
+      "loss": 0.1315,
+      "step": 955
+    },
+    {
+      "epoch": 0.21727272727272728,
+      "grad_norm": 1.5042350987246762,
+      "learning_rate": 1.2441971951132578e-06,
+      "loss": 0.0882,
+      "step": 956
+    },
+    {
+      "epoch": 0.2175,
+      "grad_norm": 2.4181780668959085,
+      "learning_rate": 1.2441850551689996e-06,
+      "loss": 0.1046,
+      "step": 957
+    },
+    {
+      "epoch": 0.2177272727272727,
+      "grad_norm": 2.5804342024474765,
+      "learning_rate": 1.244172902598478e-06,
+      "loss": 0.0819,
+      "step": 958
+    },
+    {
+      "epoch": 0.21795454545454546,
+      "grad_norm": 1.6927617530446317,
+      "learning_rate": 1.2441607374019407e-06,
+      "loss": 0.0894,
+      "step": 959
+    },
+    {
+      "epoch": 0.21818181818181817,
+      "grad_norm": 3.1485056093313064,
+      "learning_rate": 1.2441485595796358e-06,
+      "loss": 0.1001,
+      "step": 960
+    },
+    {
+      "epoch": 0.2184090909090909,
+      "grad_norm": 2.2145668987720257,
+      "learning_rate": 1.2441363691318114e-06,
+      "loss": 0.1026,
+      "step": 961
+    },
+    {
+      "epoch": 0.21863636363636363,
+      "grad_norm": 2.374231474652171,
+      "learning_rate": 1.2441241660587164e-06,
+      "loss": 0.1031,
+      "step": 962
+    },
+    {
+      "epoch": 0.21886363636363637,
+      "grad_norm": 27.81827256151305,
+      "learning_rate": 1.2441119503605997e-06,
+      "loss": 0.1051,
+      "step": 963
+    },
+    {
+      "epoch": 0.2190909090909091,
+      "grad_norm": 1.591802266640201,
+      "learning_rate": 1.24409972203771e-06,
+      "loss": 0.1008,
+      "step": 964
+    },
+    {
+      "epoch": 0.21931818181818183,
+      "grad_norm": 2.8597074764480768,
+      "learning_rate": 1.244087481090297e-06,
+      "loss": 0.1193,
+      "step": 965
+    },
+    {
+      "epoch": 0.21954545454545454,
+      "grad_norm": 2.448423353894707,
+      "learning_rate": 1.2440752275186102e-06,
+      "loss": 0.0878,
+      "step": 966
+    },
+    {
+      "epoch": 0.2197727272727273,
+      "grad_norm": 1.9456104566179393,
+      "learning_rate": 1.2440629613228996e-06,
+      "loss": 0.0853,
+      "step": 967
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3479105477149878,
+      "learning_rate": 1.244050682503415e-06,
+      "loss": 0.1771,
+      "step": 968
+    },
+    {
+      "epoch": 0.22022727272727272,
+      "grad_norm": 1.9058503128084854,
+      "learning_rate": 1.2440383910604071e-06,
+      "loss": 0.0907,
+      "step": 969
+    },
+    {
+      "epoch": 0.22045454545454546,
+      "grad_norm": 2.060967185911495,
+      "learning_rate": 1.2440260869941264e-06,
+      "loss": 0.0662,
+      "step": 970
+    },
+    {
+      "epoch": 0.22068181818181817,
+      "grad_norm": 2.0274985045026397,
+      "learning_rate": 1.244013770304824e-06,
+      "loss": 0.084,
+      "step": 971
+    },
+    {
+      "epoch": 0.22090909090909092,
+      "grad_norm": 2.040824795371165,
+      "learning_rate": 1.2440014409927506e-06,
+      "loss": 0.0749,
+      "step": 972
+    },
+    {
+      "epoch": 0.22113636363636363,
+      "grad_norm": 2.576360830983718,
+      "learning_rate": 1.243989099058158e-06,
+      "loss": 0.1475,
+      "step": 973
+    },
+    {
+      "epoch": 0.22136363636363637,
+      "grad_norm": 1.9017161332579158,
+      "learning_rate": 1.2439767445012978e-06,
+      "loss": 0.0808,
+      "step": 974
+    },
+    {
+      "epoch": 0.2215909090909091,
+      "grad_norm": 1.345311035791869,
+      "learning_rate": 1.2439643773224222e-06,
+      "loss": 0.0834,
+      "step": 975
+    },
+    {
+      "epoch": 0.22181818181818183,
+      "grad_norm": 4.09016271926954,
+      "learning_rate": 1.2439519975217828e-06,
+      "loss": 0.2172,
+      "step": 976
+    },
+    {
+      "epoch": 0.22204545454545455,
+      "grad_norm": 1.649397120707185,
+      "learning_rate": 1.2439396050996323e-06,
+      "loss": 0.0509,
+      "step": 977
+    },
+    {
+      "epoch": 0.22227272727272726,
+      "grad_norm": 2.7157471855206166,
+      "learning_rate": 1.2439272000562236e-06,
+      "loss": 0.1097,
+      "step": 978
+    },
+    {
+      "epoch": 0.2225,
+      "grad_norm": 1.685522220338484,
+      "learning_rate": 1.2439147823918094e-06,
+      "loss": 0.0785,
+      "step": 979
+    },
+    {
+      "epoch": 0.22272727272727272,
+      "grad_norm": 1.855765445402124,
+      "learning_rate": 1.2439023521066433e-06,
+      "loss": 0.0776,
+      "step": 980
+    },
+    {
+      "epoch": 0.22295454545454546,
+      "grad_norm": 2.036425862007418,
+      "learning_rate": 1.2438899092009783e-06,
+      "loss": 0.1633,
+      "step": 981
+    },
+    {
+      "epoch": 0.22318181818181818,
+      "grad_norm": 2.0862207585861645,
+      "learning_rate": 1.2438774536750682e-06,
+      "loss": 0.0833,
+      "step": 982
+    },
+    {
+      "epoch": 0.22340909090909092,
+      "grad_norm": 2.389926472838999,
+      "learning_rate": 1.2438649855291673e-06,
+      "loss": 0.1182,
+      "step": 983
+    },
+    {
+      "epoch": 0.22363636363636363,
+      "grad_norm": 3.146831679821472,
+      "learning_rate": 1.2438525047635296e-06,
+      "loss": 0.1071,
+      "step": 984
+    },
+    {
+      "epoch": 0.22386363636363638,
+      "grad_norm": 1.816559208098047,
+      "learning_rate": 1.2438400113784097e-06,
+      "loss": 0.0756,
+      "step": 985
+    },
+    {
+      "epoch": 0.2240909090909091,
+      "grad_norm": 2.003948586622291,
+      "learning_rate": 1.2438275053740624e-06,
+      "loss": 0.0815,
+      "step": 986
+    },
+    {
+      "epoch": 0.2243181818181818,
+      "grad_norm": 1.6405659484634025,
+      "learning_rate": 1.2438149867507427e-06,
+      "loss": 0.0529,
+      "step": 987
+    },
+    {
+      "epoch": 0.22454545454545455,
+      "grad_norm": 2.05629310395161,
+      "learning_rate": 1.2438024555087057e-06,
+      "loss": 0.122,
+      "step": 988
+    },
+    {
+      "epoch": 0.22477272727272726,
+      "grad_norm": 1.7559470347310315,
+      "learning_rate": 1.2437899116482072e-06,
+      "loss": 0.0739,
+      "step": 989
+    },
+    {
+      "epoch": 0.225,
+      "grad_norm": 3.1297881654081907,
+      "learning_rate": 1.2437773551695027e-06,
+      "loss": 0.1059,
+      "step": 990
+    },
+    {
+      "epoch": 0.22522727272727272,
+      "grad_norm": 1.8016255738332134,
+      "learning_rate": 1.2437647860728487e-06,
+      "loss": 0.0717,
+      "step": 991
+    },
+    {
+      "epoch": 0.22545454545454546,
+      "grad_norm": 3.076731247523777,
+      "learning_rate": 1.2437522043585009e-06,
+      "loss": 0.1408,
+      "step": 992
+    },
+    {
+      "epoch": 0.22568181818181818,
+      "grad_norm": 1.4535935422424164,
+      "learning_rate": 1.2437396100267162e-06,
+      "loss": 0.0663,
+      "step": 993
+    },
+    {
+      "epoch": 0.22590909090909092,
+      "grad_norm": 1.7976581805598622,
+      "learning_rate": 1.2437270030777516e-06,
+      "loss": 0.0946,
+      "step": 994
+    },
+    {
+      "epoch": 0.22613636363636364,
+      "grad_norm": 2.7778031437738107,
+      "learning_rate": 1.243714383511864e-06,
+      "loss": 0.154,
+      "step": 995
+    },
+    {
+      "epoch": 0.22636363636363635,
+      "grad_norm": 2.2433477607816825,
+      "learning_rate": 1.2437017513293107e-06,
+      "loss": 0.067,
+      "step": 996
+    },
+    {
+      "epoch": 0.2265909090909091,
+      "grad_norm": 2.177685018672325,
+      "learning_rate": 1.243689106530349e-06,
+      "loss": 0.0718,
+      "step": 997
+    },
+    {
+      "epoch": 0.2268181818181818,
+      "grad_norm": 2.263022319377964,
+      "learning_rate": 1.2436764491152372e-06,
+      "loss": 0.0861,
+      "step": 998
+    },
+    {
+      "epoch": 0.22704545454545455,
+      "grad_norm": 1.1439616250798952,
+      "learning_rate": 1.2436637790842335e-06,
+      "loss": 0.0398,
+      "step": 999
+    },
+    {
+      "epoch": 0.22727272727272727,
+      "grad_norm": 1.9249309425956758,
+      "learning_rate": 1.2436510964375958e-06,
+      "loss": 0.1006,
+      "step": 1000
+    },
+    {
+      "epoch": 0.2275,
+      "grad_norm": 2.4972685714532488,
+      "learning_rate": 1.243638401175583e-06,
+      "loss": 0.1082,
+      "step": 1001
+    },
+    {
+      "epoch": 0.22772727272727272,
+      "grad_norm": 2.2860358192026924,
+      "learning_rate": 1.243625693298454e-06,
+      "loss": 0.0794,
+      "step": 1002
+    },
+    {
+      "epoch": 0.22795454545454547,
+      "grad_norm": 0.6067297120767662,
+      "learning_rate": 1.2436129728064678e-06,
+      "loss": 0.0497,
+      "step": 1003
+    },
+    {
+      "epoch": 0.22818181818181818,
+      "grad_norm": 1.1098167723657022,
+      "learning_rate": 1.243600239699884e-06,
+      "loss": 0.0736,
+      "step": 1004
+    },
+    {
+      "epoch": 0.22840909090909092,
+      "grad_norm": 3.0171789078559237,
+      "learning_rate": 1.2435874939789618e-06,
+      "loss": 0.1488,
+      "step": 1005
+    },
+    {
+      "epoch": 0.22863636363636364,
+      "grad_norm": 3.302898043108915,
+      "learning_rate": 1.2435747356439614e-06,
+      "loss": 0.1319,
+      "step": 1006
+    },
+    {
+      "epoch": 0.22886363636363635,
+      "grad_norm": 1.6923239560882795,
+      "learning_rate": 1.2435619646951433e-06,
+      "loss": 0.0692,
+      "step": 1007
+    },
+    {
+      "epoch": 0.2290909090909091,
+      "grad_norm": 1.4467416528980033,
+      "learning_rate": 1.2435491811327674e-06,
+      "loss": 0.0414,
+      "step": 1008
+    },
+    {
+      "epoch": 0.2293181818181818,
+      "grad_norm": 2.3189518681418053,
+      "learning_rate": 1.2435363849570948e-06,
+      "loss": 0.1286,
+      "step": 1009
+    },
+    {
+      "epoch": 0.22954545454545455,
+      "grad_norm": 1.266680465408653,
+      "learning_rate": 1.243523576168386e-06,
+      "loss": 0.0398,
+      "step": 1010
+    },
+    {
+      "epoch": 0.22977272727272727,
+      "grad_norm": 3.0095614571869573,
+      "learning_rate": 1.2435107547669022e-06,
+      "loss": 0.1418,
+      "step": 1011
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6218631991421069,
+      "learning_rate": 1.2434979207529055e-06,
+      "loss": 0.0867,
+      "step": 1012
+    },
+    {
+      "epoch": 0.23022727272727272,
+      "grad_norm": 2.1105439584418604,
+      "learning_rate": 1.243485074126657e-06,
+      "loss": 0.0888,
+      "step": 1013
+    },
+    {
+      "epoch": 0.23045454545454547,
+      "grad_norm": 1.6731903593444284,
+      "learning_rate": 1.2434722148884188e-06,
+      "loss": 0.0864,
+      "step": 1014
+    },
+    {
+      "epoch": 0.23068181818181818,
+      "grad_norm": 2.249943688252698,
+      "learning_rate": 1.243459343038453e-06,
+      "loss": 0.0742,
+      "step": 1015
+    },
+    {
+      "epoch": 0.2309090909090909,
+      "grad_norm": 2.089603826245262,
+      "learning_rate": 1.2434464585770226e-06,
+      "loss": 0.0661,
+      "step": 1016
+    },
+    {
+      "epoch": 0.23113636363636364,
+      "grad_norm": 1.4152935582703243,
+      "learning_rate": 1.2434335615043896e-06,
+      "loss": 0.0667,
+      "step": 1017
+    },
+    {
+      "epoch": 0.23136363636363635,
+      "grad_norm": 2.1244725888070484,
+      "learning_rate": 1.2434206518208176e-06,
+      "loss": 0.0916,
+      "step": 1018
+    },
+    {
+      "epoch": 0.2315909090909091,
+      "grad_norm": 1.4377185909575454,
+      "learning_rate": 1.2434077295265695e-06,
+      "loss": 0.0869,
+      "step": 1019
+    },
+    {
+      "epoch": 0.2318181818181818,
+      "grad_norm": 1.5738153195979172,
+      "learning_rate": 1.2433947946219092e-06,
+      "loss": 0.0492,
+      "step": 1020
+    },
+    {
+      "epoch": 0.23204545454545455,
+      "grad_norm": 1.5217519959412786,
+      "learning_rate": 1.2433818471070998e-06,
+      "loss": 0.0706,
+      "step": 1021
+    },
+    {
+      "epoch": 0.23227272727272727,
+      "grad_norm": 2.72736435091497,
+      "learning_rate": 1.243368886982406e-06,
+      "loss": 0.1369,
+      "step": 1022
+    },
+    {
+      "epoch": 0.2325,
+      "grad_norm": 2.406173239989333,
+      "learning_rate": 1.2433559142480915e-06,
+      "loss": 0.1511,
+      "step": 1023
+    },
+    {
+      "epoch": 0.23272727272727273,
+      "grad_norm": 1.9810046474804894,
+      "learning_rate": 1.2433429289044212e-06,
+      "loss": 0.096,
+      "step": 1024
+    },
+    {
+      "epoch": 0.23295454545454544,
+      "grad_norm": 1.9048576854184494,
+      "learning_rate": 1.24332993095166e-06,
+      "loss": 0.0745,
+      "step": 1025
+    },
+    {
+      "epoch": 0.23318181818181818,
+      "grad_norm": 3.010305574553672,
+      "learning_rate": 1.2433169203900726e-06,
+      "loss": 0.138,
+      "step": 1026
+    },
+    {
+      "epoch": 0.2334090909090909,
+      "grad_norm": 1.4903137070053325,
+      "learning_rate": 1.2433038972199246e-06,
+      "loss": 0.0486,
+      "step": 1027
+    },
+    {
+      "epoch": 0.23363636363636364,
+      "grad_norm": 2.483338738925001,
+      "learning_rate": 1.2432908614414816e-06,
+      "loss": 0.1092,
+      "step": 1028
+    },
+    {
+      "epoch": 0.23386363636363636,
+      "grad_norm": 2.2431854289955027,
+      "learning_rate": 1.2432778130550088e-06,
+      "loss": 0.0942,
+      "step": 1029
+    },
+    {
+      "epoch": 0.2340909090909091,
+      "grad_norm": 1.2837481134984026,
+      "learning_rate": 1.2432647520607732e-06,
+      "loss": 0.0686,
+      "step": 1030
+    },
+    {
+      "epoch": 0.23431818181818181,
+      "grad_norm": 1.8307169411001316,
+      "learning_rate": 1.2432516784590405e-06,
+      "loss": 0.0572,
+      "step": 1031
+    },
+    {
+      "epoch": 0.23454545454545456,
+      "grad_norm": 2.4038189204345937,
+      "learning_rate": 1.2432385922500776e-06,
+      "loss": 0.1221,
+      "step": 1032
+    },
+    {
+      "epoch": 0.23477272727272727,
+      "grad_norm": 2.2444160247767266,
+      "learning_rate": 1.2432254934341508e-06,
+      "loss": 0.0894,
+      "step": 1033
+    },
+    {
+      "epoch": 0.235,
+      "grad_norm": 0.9265210031254871,
+      "learning_rate": 1.243212382011528e-06,
+      "loss": 0.0512,
+      "step": 1034
+    },
+    {
+      "epoch": 0.23522727272727273,
+      "grad_norm": 1.2031101715765897,
+      "learning_rate": 1.2431992579824762e-06,
+      "loss": 0.0585,
+      "step": 1035
+    },
+    {
+      "epoch": 0.23545454545454544,
+      "grad_norm": 0.49186429427950495,
+      "learning_rate": 1.243186121347263e-06,
+      "loss": 0.0284,
+      "step": 1036
+    },
+    {
+      "epoch": 0.2356818181818182,
+      "grad_norm": 2.885790681172995,
+      "learning_rate": 1.2431729721061563e-06,
+      "loss": 0.1563,
+      "step": 1037
+    },
+    {
+      "epoch": 0.2359090909090909,
+      "grad_norm": 2.4800422087638445,
+      "learning_rate": 1.243159810259424e-06,
+      "loss": 0.1049,
+      "step": 1038
+    },
+    {
+      "epoch": 0.23613636363636364,
+      "grad_norm": 1.1619963360981473,
+      "learning_rate": 1.2431466358073351e-06,
+      "loss": 0.0366,
+      "step": 1039
+    },
+    {
+      "epoch": 0.23636363636363636,
+      "grad_norm": 1.5509747062816737,
+      "learning_rate": 1.2431334487501578e-06,
+      "loss": 0.0977,
+      "step": 1040
+    },
+    {
+      "epoch": 0.2365909090909091,
+      "grad_norm": 2.996632936301279,
+      "learning_rate": 1.243120249088161e-06,
+      "loss": 0.1239,
+      "step": 1041
+    },
+    {
+      "epoch": 0.23681818181818182,
+      "grad_norm": 1.8053218176493258,
+      "learning_rate": 1.2431070368216138e-06,
+      "loss": 0.0665,
+      "step": 1042
+    },
+    {
+      "epoch": 0.23704545454545456,
+      "grad_norm": 1.3629430848525381,
+      "learning_rate": 1.243093811950786e-06,
+      "loss": 0.0567,
+      "step": 1043
+    },
+    {
+      "epoch": 0.23727272727272727,
+      "grad_norm": 2.9583085662885553,
+      "learning_rate": 1.243080574475947e-06,
+      "loss": 0.1404,
+      "step": 1044
+    },
+    {
+      "epoch": 0.2375,
+      "grad_norm": 2.067083078914586,
+      "learning_rate": 1.2430673243973667e-06,
+      "loss": 0.1014,
+      "step": 1045
+    },
+    {
+      "epoch": 0.23772727272727273,
+      "grad_norm": 2.155034157440008,
+      "learning_rate": 1.2430540617153156e-06,
+      "loss": 0.0961,
+      "step": 1046
+    },
+    {
+      "epoch": 0.23795454545454545,
+      "grad_norm": 1.6266352029511453,
+      "learning_rate": 1.2430407864300635e-06,
+      "loss": 0.0416,
+      "step": 1047
+    },
+    {
+      "epoch": 0.2381818181818182,
+      "grad_norm": 2.3451562028932775,
+      "learning_rate": 1.243027498541882e-06,
+      "loss": 0.0947,
+      "step": 1048
+    },
+    {
+      "epoch": 0.2384090909090909,
+      "grad_norm": 3.1999158244226313,
+      "learning_rate": 1.2430141980510412e-06,
+      "loss": 0.1627,
+      "step": 1049
+    },
+    {
+      "epoch": 0.23863636363636365,
+      "grad_norm": 1.5542894634630038,
+      "learning_rate": 1.243000884957813e-06,
+      "loss": 0.0769,
+      "step": 1050
+    },
+    {
+      "epoch": 0.23886363636363636,
+      "grad_norm": 3.116612358574622,
+      "learning_rate": 1.2429875592624685e-06,
+      "loss": 0.1412,
+      "step": 1051
+    },
+    {
+      "epoch": 0.2390909090909091,
+      "grad_norm": 1.7827707334530163,
+      "learning_rate": 1.2429742209652795e-06,
+      "loss": 0.0709,
+      "step": 1052
+    },
+    {
+      "epoch": 0.23931818181818182,
+      "grad_norm": 1.8717566302104287,
+      "learning_rate": 1.242960870066518e-06,
+      "loss": 0.0923,
+      "step": 1053
+    },
+    {
+      "epoch": 0.23954545454545453,
+      "grad_norm": 1.8058561688855062,
+      "learning_rate": 1.2429475065664564e-06,
+      "loss": 0.0748,
+      "step": 1054
+    },
+    {
+      "epoch": 0.23977272727272728,
+      "grad_norm": 1.8387901683698251,
+      "learning_rate": 1.242934130465367e-06,
+      "loss": 0.0522,
+      "step": 1055
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2960346550087407,
+      "learning_rate": 1.2429207417635226e-06,
+      "loss": 0.1104,
+      "step": 1056
+    },
+    {
+      "epoch": 0.24022727272727273,
+      "grad_norm": 2.1223994345524746,
+      "learning_rate": 1.2429073404611962e-06,
+      "loss": 0.0662,
+      "step": 1057
+    },
+    {
+      "epoch": 0.24045454545454545,
+      "grad_norm": 2.174752489159394,
+      "learning_rate": 1.2428939265586614e-06,
+      "loss": 0.0805,
+      "step": 1058
+    },
+    {
+      "epoch": 0.2406818181818182,
+      "grad_norm": 1.4103410195644692,
+      "learning_rate": 1.2428805000561913e-06,
+      "loss": 0.083,
+      "step": 1059
+    },
+    {
+      "epoch": 0.2409090909090909,
+      "grad_norm": 1.923532655864264,
+      "learning_rate": 1.2428670609540597e-06,
+      "loss": 0.1517,
+      "step": 1060
+    },
+    {
+      "epoch": 0.24113636363636365,
+      "grad_norm": 1.518909504850699,
+      "learning_rate": 1.242853609252541e-06,
+      "loss": 0.0459,
+      "step": 1061
+    },
+    {
+      "epoch": 0.24136363636363636,
+      "grad_norm": 2.0718844389443003,
+      "learning_rate": 1.2428401449519093e-06,
+      "loss": 0.0771,
+      "step": 1062
+    },
+    {
+      "epoch": 0.24159090909090908,
+      "grad_norm": 2.1096025150207733,
+      "learning_rate": 1.242826668052439e-06,
+      "loss": 0.072,
+      "step": 1063
+    },
+    {
+      "epoch": 0.24181818181818182,
+      "grad_norm": 1.8639655138178326,
+      "learning_rate": 1.242813178554405e-06,
+      "loss": 0.1056,
+      "step": 1064
+    },
+    {
+      "epoch": 0.24204545454545454,
+      "grad_norm": 1.8905828459193579,
+      "learning_rate": 1.242799676458083e-06,
+      "loss": 0.1153,
+      "step": 1065
+    },
+    {
+      "epoch": 0.24227272727272728,
+      "grad_norm": 3.412800540974206,
+      "learning_rate": 1.2427861617637472e-06,
+      "loss": 0.1403,
+      "step": 1066
+    },
+    {
+      "epoch": 0.2425,
+      "grad_norm": 1.6626627418763213,
+      "learning_rate": 1.242772634471674e-06,
+      "loss": 0.0562,
+      "step": 1067
+    },
+    {
+      "epoch": 0.24272727272727274,
+      "grad_norm": 1.3505221719342024,
+      "learning_rate": 1.242759094582139e-06,
+      "loss": 0.0756,
+      "step": 1068
+    },
+    {
+      "epoch": 0.24295454545454545,
+      "grad_norm": 1.997717373731209,
+      "learning_rate": 1.2427455420954183e-06,
+      "loss": 0.0729,
+      "step": 1069
+    },
+    {
+      "epoch": 0.2431818181818182,
+      "grad_norm": 2.509987961621086,
+      "learning_rate": 1.2427319770117884e-06,
+      "loss": 0.1552,
+      "step": 1070
+    },
+    {
+      "epoch": 0.2434090909090909,
+      "grad_norm": 2.2256355412875246,
+      "learning_rate": 1.2427183993315256e-06,
+      "loss": 0.1005,
+      "step": 1071
+    },
+    {
+      "epoch": 0.24363636363636362,
+      "grad_norm": 2.225597504507886,
+      "learning_rate": 1.2427048090549072e-06,
+      "loss": 0.0997,
+      "step": 1072
+    },
+    {
+      "epoch": 0.24386363636363637,
+      "grad_norm": 1.7649619058779473,
+      "learning_rate": 1.24269120618221e-06,
+      "loss": 0.065,
+      "step": 1073
+    },
+    {
+      "epoch": 0.24409090909090908,
+      "grad_norm": 0.5553132411485427,
+      "learning_rate": 1.2426775907137112e-06,
+      "loss": 0.0284,
+      "step": 1074
+    },
+    {
+      "epoch": 0.24431818181818182,
+      "grad_norm": 1.9139230013043653,
+      "learning_rate": 1.2426639626496892e-06,
+      "loss": 0.0705,
+      "step": 1075
+    },
+    {
+      "epoch": 0.24454545454545454,
+      "grad_norm": 2.094022935626849,
+      "learning_rate": 1.2426503219904213e-06,
+      "loss": 0.111,
+      "step": 1076
+    },
+    {
+      "epoch": 0.24477272727272728,
+      "grad_norm": 2.295418470612699,
+      "learning_rate": 1.2426366687361856e-06,
+      "loss": 0.0853,
+      "step": 1077
+    },
+    {
+      "epoch": 0.245,
+      "grad_norm": 1.7297873188495714,
+      "learning_rate": 1.2426230028872606e-06,
+      "loss": 0.0689,
+      "step": 1078
+    },
+    {
+      "epoch": 0.24522727272727274,
+      "grad_norm": 1.8017836546076818,
+      "learning_rate": 1.2426093244439254e-06,
+      "loss": 0.1458,
+      "step": 1079
+    },
+    {
+      "epoch": 0.24545454545454545,
+      "grad_norm": 2.5914568827154056,
+      "learning_rate": 1.2425956334064584e-06,
+      "loss": 0.0983,
+      "step": 1080
+    },
+    {
+      "epoch": 0.2456818181818182,
+      "grad_norm": 2.1757622517888193,
+      "learning_rate": 1.242581929775139e-06,
+      "loss": 0.1266,
+      "step": 1081
+    },
+    {
+      "epoch": 0.2459090909090909,
+      "grad_norm": 1.7447923167578314,
+      "learning_rate": 1.2425682135502468e-06,
+      "loss": 0.0686,
+      "step": 1082
+    },
+    {
+      "epoch": 0.24613636363636363,
+      "grad_norm": 2.5891563630082626,
+      "learning_rate": 1.2425544847320612e-06,
+      "loss": 0.1009,
+      "step": 1083
+    },
+    {
+      "epoch": 0.24636363636363637,
+      "grad_norm": 2.3278544105169336,
+      "learning_rate": 1.2425407433208623e-06,
+      "loss": 0.1388,
+      "step": 1084
+    },
+    {
+      "epoch": 0.24659090909090908,
+      "grad_norm": 1.5913360118574063,
+      "learning_rate": 1.24252698931693e-06,
+      "loss": 0.056,
+      "step": 1085
+    },
+    {
+      "epoch": 0.24681818181818183,
+      "grad_norm": 1.9980899173610092,
+      "learning_rate": 1.2425132227205456e-06,
+      "loss": 0.1009,
+      "step": 1086
+    },
+    {
+      "epoch": 0.24704545454545454,
+      "grad_norm": 1.363934438252317,
+      "learning_rate": 1.2424994435319887e-06,
+      "loss": 0.0469,
+      "step": 1087
+    },
+    {
+      "epoch": 0.24727272727272728,
+      "grad_norm": 1.843150856614852,
+      "learning_rate": 1.2424856517515413e-06,
+      "loss": 0.0574,
+      "step": 1088
+    },
+    {
+      "epoch": 0.2475,
+      "grad_norm": 1.3346312168415755,
+      "learning_rate": 1.242471847379484e-06,
+      "loss": 0.0584,
+      "step": 1089
+    },
+    {
+      "epoch": 0.24772727272727274,
+      "grad_norm": 2.794635993932223,
+      "learning_rate": 1.2424580304160986e-06,
+      "loss": 0.1138,
+      "step": 1090
+    },
+    {
+      "epoch": 0.24795454545454546,
+      "grad_norm": 1.6069208846041545,
+      "learning_rate": 1.2424442008616667e-06,
+      "loss": 0.089,
+      "step": 1091
+    },
+    {
+      "epoch": 0.24818181818181817,
+      "grad_norm": 3.025497909949808,
+      "learning_rate": 1.2424303587164702e-06,
+      "loss": 0.1354,
+      "step": 1092
+    },
+    {
+      "epoch": 0.2484090909090909,
+      "grad_norm": 1.0212102745305145,
+      "learning_rate": 1.2424165039807915e-06,
+      "loss": 0.0546,
+      "step": 1093
+    },
+    {
+      "epoch": 0.24863636363636363,
+      "grad_norm": 1.6409717398788335,
+      "learning_rate": 1.2424026366549134e-06,
+      "loss": 0.0767,
+      "step": 1094
+    },
+    {
+      "epoch": 0.24886363636363637,
+      "grad_norm": 2.432099944399322,
+      "learning_rate": 1.2423887567391182e-06,
+      "loss": 0.0976,
+      "step": 1095
+    },
+    {
+      "epoch": 0.24909090909090909,
+      "grad_norm": 2.4183689722389357,
+      "learning_rate": 1.2423748642336894e-06,
+      "loss": 0.1003,
+      "step": 1096
+    },
+    {
+      "epoch": 0.24931818181818183,
+      "grad_norm": 1.7962534751561647,
+      "learning_rate": 1.24236095913891e-06,
+      "loss": 0.0742,
+      "step": 1097
+    },
+    {
+      "epoch": 0.24954545454545454,
+      "grad_norm": 2.2729792247208063,
+      "learning_rate": 1.2423470414550634e-06,
+      "loss": 0.1048,
+      "step": 1098
+    },
+    {
+      "epoch": 0.24977272727272729,
+      "grad_norm": 1.5019820367803949,
+      "learning_rate": 1.2423331111824336e-06,
+      "loss": 0.1008,
+      "step": 1099
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8982871291289067,
+      "learning_rate": 1.2423191683213048e-06,
+      "loss": 0.0626,
+      "step": 1100
+    },
+    {
+      "epoch": 0.25022727272727274,
+      "grad_norm": 2.29170642889581,
+      "learning_rate": 1.2423052128719611e-06,
+      "loss": 0.1402,
+      "step": 1101
+    },
+    {
+      "epoch": 0.25045454545454543,
+      "grad_norm": 2.596693134128902,
+      "learning_rate": 1.2422912448346874e-06,
+      "loss": 0.1463,
+      "step": 1102
+    },
+    {
+      "epoch": 0.2506818181818182,
+      "grad_norm": 2.2669355480565336,
+      "learning_rate": 1.242277264209768e-06,
+      "loss": 0.1265,
+      "step": 1103
+    },
+    {
+      "epoch": 0.2509090909090909,
+      "grad_norm": 1.7017737891096427,
+      "learning_rate": 1.2422632709974884e-06,
+      "loss": 0.0699,
+      "step": 1104
+    },
+    {
+      "epoch": 0.25113636363636366,
+      "grad_norm": 1.450116713227821,
+      "learning_rate": 1.2422492651981337e-06,
+      "loss": 0.0855,
+      "step": 1105
+    },
+    {
+      "epoch": 0.25136363636363634,
+      "grad_norm": 1.630226029916931,
+      "learning_rate": 1.24223524681199e-06,
+      "loss": 0.0836,
+      "step": 1106
+    },
+    {
+      "epoch": 0.2515909090909091,
+      "grad_norm": 1.3381182224664045,
+      "learning_rate": 1.2422212158393424e-06,
+      "loss": 0.0849,
+      "step": 1107
+    },
+    {
+      "epoch": 0.25181818181818183,
+      "grad_norm": 1.6976878192174827,
+      "learning_rate": 1.2422071722804777e-06,
+      "loss": 0.0836,
+      "step": 1108
+    },
+    {
+      "epoch": 0.2520454545454546,
+      "grad_norm": 1.5644256968761323,
+      "learning_rate": 1.2421931161356817e-06,
+      "loss": 0.086,
+      "step": 1109
+    },
+    {
+      "epoch": 0.25227272727272726,
+      "grad_norm": 1.2005158854039986,
+      "learning_rate": 1.2421790474052414e-06,
+      "loss": 0.0738,
+      "step": 1110
+    },
+    {
+      "epoch": 0.2525,
+      "grad_norm": 1.8168765989958375,
+      "learning_rate": 1.2421649660894438e-06,
+      "loss": 0.0908,
+      "step": 1111
+    },
+    {
+      "epoch": 0.25272727272727274,
+      "grad_norm": 1.8574239730594904,
+      "learning_rate": 1.2421508721885756e-06,
+      "loss": 0.0741,
+      "step": 1112
+    },
+    {
+      "epoch": 0.25295454545454543,
+      "grad_norm": 1.3410023838943632,
+      "learning_rate": 1.2421367657029245e-06,
+      "loss": 0.0449,
+      "step": 1113
+    },
+    {
+      "epoch": 0.2531818181818182,
+      "grad_norm": 2.8520561623634406,
+      "learning_rate": 1.2421226466327782e-06,
+      "loss": 0.0853,
+      "step": 1114
+    },
+    {
+      "epoch": 0.2534090909090909,
+      "grad_norm": 2.52515440193051,
+      "learning_rate": 1.2421085149784242e-06,
+      "loss": 0.1445,
+      "step": 1115
+    },
+    {
+      "epoch": 0.25363636363636366,
+      "grad_norm": 2.0886910662596176,
+      "learning_rate": 1.2420943707401514e-06,
+      "loss": 0.0959,
+      "step": 1116
+    },
+    {
+      "epoch": 0.25386363636363635,
+      "grad_norm": 1.4629855901825601,
+      "learning_rate": 1.2420802139182475e-06,
+      "loss": 0.0654,
+      "step": 1117
+    },
+    {
+      "epoch": 0.2540909090909091,
+      "grad_norm": 1.6139773048996153,
+      "learning_rate": 1.2420660445130017e-06,
+      "loss": 0.062,
+      "step": 1118
+    },
+    {
+      "epoch": 0.25431818181818183,
+      "grad_norm": 3.088021822598228,
+      "learning_rate": 1.2420518625247024e-06,
+      "loss": 0.1836,
+      "step": 1119
+    },
+    {
+      "epoch": 0.2545454545454545,
+      "grad_norm": 2.5336312412147346,
+      "learning_rate": 1.2420376679536395e-06,
+      "loss": 0.1234,
+      "step": 1120
+    },
+    {
+      "epoch": 0.25477272727272726,
+      "grad_norm": 2.346533394511774,
+      "learning_rate": 1.2420234608001017e-06,
+      "loss": 0.1165,
+      "step": 1121
+    },
+    {
+      "epoch": 0.255,
+      "grad_norm": 2.266029056367262,
+      "learning_rate": 1.2420092410643795e-06,
+      "loss": 0.093,
+      "step": 1122
+    },
+    {
+      "epoch": 0.25522727272727275,
+      "grad_norm": 2.436224245061615,
+      "learning_rate": 1.241995008746762e-06,
+      "loss": 0.1217,
+      "step": 1123
+    },
+    {
+      "epoch": 0.25545454545454543,
+      "grad_norm": 2.5712910176787895,
+      "learning_rate": 1.24198076384754e-06,
+      "loss": 0.1156,
+      "step": 1124
+    },
+    {
+      "epoch": 0.2556818181818182,
+      "grad_norm": 4.650862562627461,
+      "learning_rate": 1.241966506367004e-06,
+      "loss": 0.1797,
+      "step": 1125
+    },
+    {
+      "epoch": 0.2559090909090909,
+      "grad_norm": 1.7247114446786174,
+      "learning_rate": 1.2419522363054446e-06,
+      "loss": 0.0715,
+      "step": 1126
+    },
+    {
+      "epoch": 0.25613636363636366,
+      "grad_norm": 2.76086364151186,
+      "learning_rate": 1.2419379536631525e-06,
+      "loss": 0.141,
+      "step": 1127
+    },
+    {
+      "epoch": 0.25636363636363635,
+      "grad_norm": 3.5486415149547472,
+      "learning_rate": 1.2419236584404196e-06,
+      "loss": 0.255,
+      "step": 1128
+    },
+    {
+      "epoch": 0.2565909090909091,
+      "grad_norm": 1.5441508941395343,
+      "learning_rate": 1.2419093506375367e-06,
+      "loss": 0.0681,
+      "step": 1129
+    },
+    {
+      "epoch": 0.25681818181818183,
+      "grad_norm": 1.979724952883299,
+      "learning_rate": 1.241895030254796e-06,
+      "loss": 0.1057,
+      "step": 1130
+    },
+    {
+      "epoch": 0.2570454545454545,
+      "grad_norm": 3.014312230194048,
+      "learning_rate": 1.2418806972924893e-06,
+      "loss": 0.166,
+      "step": 1131
+    },
+    {
+      "epoch": 0.25727272727272726,
+      "grad_norm": 2.6897406846349594,
+      "learning_rate": 1.241866351750909e-06,
+      "loss": 0.1049,
+      "step": 1132
+    },
+    {
+      "epoch": 0.2575,
+      "grad_norm": 1.713542765750914,
+      "learning_rate": 1.2418519936303477e-06,
+      "loss": 0.0584,
+      "step": 1133
+    },
+    {
+      "epoch": 0.25772727272727275,
+      "grad_norm": 2.5695844892802633,
+      "learning_rate": 1.2418376229310982e-06,
+      "loss": 0.1329,
+      "step": 1134
+    },
+    {
+      "epoch": 0.25795454545454544,
+      "grad_norm": 2.780772982061506,
+      "learning_rate": 1.2418232396534533e-06,
+      "loss": 0.1592,
+      "step": 1135
+    },
+    {
+      "epoch": 0.2581818181818182,
+      "grad_norm": 1.3890551362968824,
+      "learning_rate": 1.2418088437977063e-06,
+      "loss": 0.0664,
+      "step": 1136
+    },
+    {
+      "epoch": 0.2584090909090909,
+      "grad_norm": 1.6433698072838119,
+      "learning_rate": 1.2417944353641513e-06,
+      "loss": 0.0679,
+      "step": 1137
+    },
+    {
+      "epoch": 0.2586363636363636,
+      "grad_norm": 2.783225081947558,
+      "learning_rate": 1.241780014353081e-06,
+      "loss": 0.1172,
+      "step": 1138
+    },
+    {
+      "epoch": 0.25886363636363635,
+      "grad_norm": 0.9062182425535518,
+      "learning_rate": 1.2417655807647907e-06,
+      "loss": 0.0297,
+      "step": 1139
+    },
+    {
+      "epoch": 0.2590909090909091,
+      "grad_norm": 1.650082476518612,
+      "learning_rate": 1.2417511345995743e-06,
+      "loss": 0.0634,
+      "step": 1140
+    },
+    {
+      "epoch": 0.25931818181818184,
+      "grad_norm": 1.7680411763152404,
+      "learning_rate": 1.241736675857726e-06,
+      "loss": 0.0984,
+      "step": 1141
+    },
+    {
+      "epoch": 0.2595454545454545,
+      "grad_norm": 1.9551405493642489,
+      "learning_rate": 1.2417222045395411e-06,
+      "loss": 0.0707,
+      "step": 1142
+    },
+    {
+      "epoch": 0.25977272727272727,
+      "grad_norm": 2.132587844172924,
+      "learning_rate": 1.2417077206453143e-06,
+      "loss": 0.0669,
+      "step": 1143
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.73957938301342,
+      "learning_rate": 1.2416932241753412e-06,
+      "loss": 0.1732,
+      "step": 1144
+    },
+    {
+      "epoch": 0.26022727272727275,
+      "grad_norm": 3.159142677470461,
+      "learning_rate": 1.2416787151299176e-06,
+      "loss": 0.1298,
+      "step": 1145
+    },
+    {
+      "epoch": 0.26045454545454544,
+      "grad_norm": 4.931029032287271,
+      "learning_rate": 1.241664193509339e-06,
+      "loss": 0.2973,
+      "step": 1146
+    },
+    {
+      "epoch": 0.2606818181818182,
+      "grad_norm": 2.5593422204298983,
+      "learning_rate": 1.2416496593139019e-06,
+      "loss": 0.101,
+      "step": 1147
+    },
+    {
+      "epoch": 0.2609090909090909,
+      "grad_norm": 2.3824819524944574,
+      "learning_rate": 1.241635112543902e-06,
+      "loss": 0.1515,
+      "step": 1148
+    },
+    {
+      "epoch": 0.2611363636363636,
+      "grad_norm": 2.1267882182981532,
+      "learning_rate": 1.2416205531996368e-06,
+      "loss": 0.1111,
+      "step": 1149
+    },
+    {
+      "epoch": 0.26136363636363635,
+      "grad_norm": 1.872670705653664,
+      "learning_rate": 1.2416059812814024e-06,
+      "loss": 0.085,
+      "step": 1150
+    },
+    {
+      "epoch": 0.2615909090909091,
+      "grad_norm": 2.447922834726275,
+      "learning_rate": 1.2415913967894966e-06,
+      "loss": 0.0898,
+      "step": 1151
+    },
+    {
+      "epoch": 0.26181818181818184,
+      "grad_norm": 2.1133775223236038,
+      "learning_rate": 1.2415767997242163e-06,
+      "loss": 0.1138,
+      "step": 1152
+    },
+    {
+      "epoch": 0.2620454545454545,
+      "grad_norm": 1.7680857661148832,
+      "learning_rate": 1.2415621900858597e-06,
+      "loss": 0.1097,
+      "step": 1153
+    },
+    {
+      "epoch": 0.26227272727272727,
+      "grad_norm": 2.0495434360710663,
+      "learning_rate": 1.2415475678747241e-06,
+      "loss": 0.0815,
+      "step": 1154
+    },
+    {
+      "epoch": 0.2625,
+      "grad_norm": 1.5817369608384955,
+      "learning_rate": 1.241532933091108e-06,
+      "loss": 0.0711,
+      "step": 1155
+    },
+    {
+      "epoch": 0.26272727272727275,
+      "grad_norm": 2.077030006867295,
+      "learning_rate": 1.2415182857353098e-06,
+      "loss": 0.1021,
+      "step": 1156
+    },
+    {
+      "epoch": 0.26295454545454544,
+      "grad_norm": 2.0304327685878434,
+      "learning_rate": 1.2415036258076281e-06,
+      "loss": 0.098,
+      "step": 1157
+    },
+    {
+      "epoch": 0.2631818181818182,
+      "grad_norm": 2.0179819236674557,
+      "learning_rate": 1.241488953308362e-06,
+      "loss": 0.0832,
+      "step": 1158
+    },
+    {
+      "epoch": 0.2634090909090909,
+      "grad_norm": 2.3288500261271476,
+      "learning_rate": 1.2414742682378108e-06,
+      "loss": 0.0964,
+      "step": 1159
+    },
+    {
+      "epoch": 0.2636363636363636,
+      "grad_norm": 1.5148548073550192,
+      "learning_rate": 1.2414595705962735e-06,
+      "loss": 0.0693,
+      "step": 1160
+    },
+    {
+      "epoch": 0.26386363636363636,
+      "grad_norm": 2.515619267026407,
+      "learning_rate": 1.2414448603840504e-06,
+      "loss": 0.1309,
+      "step": 1161
+    },
+    {
+      "epoch": 0.2640909090909091,
+      "grad_norm": 1.6326659671070876,
+      "learning_rate": 1.2414301376014409e-06,
+      "loss": 0.1272,
+      "step": 1162
+    },
+    {
+      "epoch": 0.26431818181818184,
+      "grad_norm": 2.970613371600496,
+      "learning_rate": 1.2414154022487455e-06,
+      "loss": 0.1172,
+      "step": 1163
+    },
+    {
+      "epoch": 0.26454545454545453,
+      "grad_norm": 2.6251647812040892,
+      "learning_rate": 1.2414006543262647e-06,
+      "loss": 0.1516,
+      "step": 1164
+    },
+    {
+      "epoch": 0.26477272727272727,
+      "grad_norm": 2.7872106205480627,
+      "learning_rate": 1.241385893834299e-06,
+      "loss": 0.0816,
+      "step": 1165
+    },
+    {
+      "epoch": 0.265,
+      "grad_norm": 1.786101761693355,
+      "learning_rate": 1.24137112077315e-06,
+      "loss": 0.0608,
+      "step": 1166
+    },
+    {
+      "epoch": 0.2652272727272727,
+      "grad_norm": 1.831186762541302,
+      "learning_rate": 1.2413563351431183e-06,
+      "loss": 0.0575,
+      "step": 1167
+    },
+    {
+      "epoch": 0.26545454545454544,
+      "grad_norm": 1.5968775852586936,
+      "learning_rate": 1.2413415369445055e-06,
+      "loss": 0.0596,
+      "step": 1168
+    },
+    {
+      "epoch": 0.2656818181818182,
+      "grad_norm": 2.5283434230225925,
+      "learning_rate": 1.2413267261776135e-06,
+      "loss": 0.0983,
+      "step": 1169
+    },
+    {
+      "epoch": 0.26590909090909093,
+      "grad_norm": 1.9766043472417276,
+      "learning_rate": 1.2413119028427445e-06,
+      "loss": 0.0981,
+      "step": 1170
+    },
+    {
+      "epoch": 0.2661363636363636,
+      "grad_norm": 2.065356020610976,
+      "learning_rate": 1.2412970669402005e-06,
+      "loss": 0.1506,
+      "step": 1171
+    },
+    {
+      "epoch": 0.26636363636363636,
+      "grad_norm": 1.5003892463013624,
+      "learning_rate": 1.241282218470284e-06,
+      "loss": 0.0817,
+      "step": 1172
+    },
+    {
+      "epoch": 0.2665909090909091,
+      "grad_norm": 1.9697805643540396,
+      "learning_rate": 1.2412673574332982e-06,
+      "loss": 0.0949,
+      "step": 1173
+    },
+    {
+      "epoch": 0.26681818181818184,
+      "grad_norm": 1.8335048505238867,
+      "learning_rate": 1.2412524838295455e-06,
+      "loss": 0.0866,
+      "step": 1174
+    },
+    {
+      "epoch": 0.26704545454545453,
+      "grad_norm": 2.0191863323883705,
+      "learning_rate": 1.2412375976593298e-06,
+      "loss": 0.0908,
+      "step": 1175
+    },
+    {
+      "epoch": 0.2672727272727273,
+      "grad_norm": 2.4864989777084974,
+      "learning_rate": 1.2412226989229542e-06,
+      "loss": 0.0932,
+      "step": 1176
+    },
+    {
+      "epoch": 0.2675,
+      "grad_norm": 1.649407937206027,
+      "learning_rate": 1.241207787620723e-06,
+      "loss": 0.0648,
+      "step": 1177
+    },
+    {
+      "epoch": 0.2677272727272727,
+      "grad_norm": 1.7978698453008457,
+      "learning_rate": 1.2411928637529398e-06,
+      "loss": 0.1099,
+      "step": 1178
+    },
+    {
+      "epoch": 0.26795454545454545,
+      "grad_norm": 1.1880578380451114,
+      "learning_rate": 1.241177927319909e-06,
+      "loss": 0.0572,
+      "step": 1179
+    },
+    {
+      "epoch": 0.2681818181818182,
+      "grad_norm": 2.5218453363263804,
+      "learning_rate": 1.2411629783219352e-06,
+      "loss": 0.1173,
+      "step": 1180
+    },
+    {
+      "epoch": 0.26840909090909093,
+      "grad_norm": 2.7976484127966996,
+      "learning_rate": 1.2411480167593237e-06,
+      "loss": 0.1469,
+      "step": 1181
+    },
+    {
+      "epoch": 0.2686363636363636,
+      "grad_norm": 3.175973035532116,
+      "learning_rate": 1.241133042632379e-06,
+      "loss": 0.2072,
+      "step": 1182
+    },
+    {
+      "epoch": 0.26886363636363636,
+      "grad_norm": 2.3669209089885768,
+      "learning_rate": 1.2411180559414066e-06,
+      "loss": 0.1045,
+      "step": 1183
+    },
+    {
+      "epoch": 0.2690909090909091,
+      "grad_norm": 1.5072402255288602,
+      "learning_rate": 1.2411030566867122e-06,
+      "loss": 0.0486,
+      "step": 1184
+    },
+    {
+      "epoch": 0.26931818181818185,
+      "grad_norm": 2.179847893190514,
+      "learning_rate": 1.2410880448686017e-06,
+      "loss": 0.1161,
+      "step": 1185
+    },
+    {
+      "epoch": 0.26954545454545453,
+      "grad_norm": 3.257496974991969,
+      "learning_rate": 1.241073020487381e-06,
+      "loss": 0.1442,
+      "step": 1186
+    },
+    {
+      "epoch": 0.2697727272727273,
+      "grad_norm": 1.7959734323016614,
+      "learning_rate": 1.2410579835433566e-06,
+      "loss": 0.126,
+      "step": 1187
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2349978429174615,
+      "learning_rate": 1.2410429340368355e-06,
+      "loss": 0.088,
+      "step": 1188
+    },
+    {
+      "epoch": 0.2702272727272727,
+      "grad_norm": 1.388535586092169,
+      "learning_rate": 1.241027871968124e-06,
+      "loss": 0.0544,
+      "step": 1189
+    },
+    {
+      "epoch": 0.27045454545454545,
+      "grad_norm": 2.113972038690569,
+      "learning_rate": 1.2410127973375293e-06,
+      "loss": 0.1003,
+      "step": 1190
+    },
+    {
+      "epoch": 0.2706818181818182,
+      "grad_norm": 1.9046265783005238,
+      "learning_rate": 1.2409977101453591e-06,
+      "loss": 0.0654,
+      "step": 1191
+    },
+    {
+      "epoch": 0.27090909090909093,
+      "grad_norm": 2.229817943406569,
+      "learning_rate": 1.240982610391921e-06,
+      "loss": 0.1077,
+      "step": 1192
+    },
+    {
+      "epoch": 0.2711363636363636,
+      "grad_norm": 1.8854441793362515,
+      "learning_rate": 1.2409674980775227e-06,
+      "loss": 0.0922,
+      "step": 1193
+    },
+    {
+      "epoch": 0.27136363636363636,
+      "grad_norm": 2.187922823028847,
+      "learning_rate": 1.2409523732024726e-06,
+      "loss": 0.1009,
+      "step": 1194
+    },
+    {
+      "epoch": 0.2715909090909091,
+      "grad_norm": 2.616207583298245,
+      "learning_rate": 1.240937235767079e-06,
+      "loss": 0.1097,
+      "step": 1195
+    },
+    {
+      "epoch": 0.2718181818181818,
+      "grad_norm": 1.4257484163673362,
+      "learning_rate": 1.2409220857716506e-06,
+      "loss": 0.0468,
+      "step": 1196
+    },
+    {
+      "epoch": 0.27204545454545453,
+      "grad_norm": 2.369994412129498,
+      "learning_rate": 1.2409069232164964e-06,
+      "loss": 0.1297,
+      "step": 1197
+    },
+    {
+      "epoch": 0.2722727272727273,
+      "grad_norm": 1.8617877368684348,
+      "learning_rate": 1.2408917481019254e-06,
+      "loss": 0.0519,
+      "step": 1198
+    },
+    {
+      "epoch": 0.2725,
+      "grad_norm": 1.2039565522144606,
+      "learning_rate": 1.240876560428247e-06,
+      "loss": 0.0732,
+      "step": 1199
+    },
+    {
+      "epoch": 0.2727272727272727,
+      "grad_norm": 1.3768896989662787,
+      "learning_rate": 1.2408613601957714e-06,
+      "loss": 0.1022,
+      "step": 1200
+    },
+    {
+      "epoch": 0.27295454545454545,
+      "grad_norm": 2.8583554808760434,
+      "learning_rate": 1.2408461474048083e-06,
+      "loss": 0.1007,
+      "step": 1201
+    },
+    {
+      "epoch": 0.2731818181818182,
+      "grad_norm": 2.724157141035039,
+      "learning_rate": 1.2408309220556676e-06,
+      "loss": 0.1598,
+      "step": 1202
+    },
+    {
+      "epoch": 0.27340909090909093,
+      "grad_norm": 2.168844125083964,
+      "learning_rate": 1.2408156841486602e-06,
+      "loss": 0.0722,
+      "step": 1203
+    },
+    {
+      "epoch": 0.2736363636363636,
+      "grad_norm": 1.444920440815994,
+      "learning_rate": 1.2408004336840964e-06,
+      "loss": 0.0596,
+      "step": 1204
+    },
+    {
+      "epoch": 0.27386363636363636,
+      "grad_norm": 1.9996871382574168,
+      "learning_rate": 1.240785170662288e-06,
+      "loss": 0.103,
+      "step": 1205
+    },
+    {
+      "epoch": 0.2740909090909091,
+      "grad_norm": 2.338088410560865,
+      "learning_rate": 1.240769895083545e-06,
+      "loss": 0.1177,
+      "step": 1206
+    },
+    {
+      "epoch": 0.2743181818181818,
+      "grad_norm": 2.858735818863512,
+      "learning_rate": 1.2407546069481798e-06,
+      "loss": 0.1276,
+      "step": 1207
+    },
+    {
+      "epoch": 0.27454545454545454,
+      "grad_norm": 1.8447775180230737,
+      "learning_rate": 1.240739306256504e-06,
+      "loss": 0.1068,
+      "step": 1208
+    },
+    {
+      "epoch": 0.2747727272727273,
+      "grad_norm": 1.850061373380902,
+      "learning_rate": 1.2407239930088296e-06,
+      "loss": 0.0528,
+      "step": 1209
+    },
+    {
+      "epoch": 0.275,
+      "grad_norm": 3.6072839411216306,
+      "learning_rate": 1.2407086672054686e-06,
+      "loss": 0.1319,
+      "step": 1210
+    },
+    {
+      "epoch": 0.2752272727272727,
+      "grad_norm": 1.8083793289128827,
+      "learning_rate": 1.2406933288467337e-06,
+      "loss": 0.061,
+      "step": 1211
+    },
+    {
+      "epoch": 0.27545454545454545,
+      "grad_norm": 1.7201397260999347,
+      "learning_rate": 1.2406779779329378e-06,
+      "loss": 0.0575,
+      "step": 1212
+    },
+    {
+      "epoch": 0.2756818181818182,
+      "grad_norm": 2.9454765415902657,
+      "learning_rate": 1.240662614464394e-06,
+      "loss": 0.1449,
+      "step": 1213
+    },
+    {
+      "epoch": 0.2759090909090909,
+      "grad_norm": 1.6255670819663735,
+      "learning_rate": 1.240647238441415e-06,
+      "loss": 0.0563,
+      "step": 1214
+    },
+    {
+      "epoch": 0.2761363636363636,
+      "grad_norm": 1.894629187310187,
+      "learning_rate": 1.240631849864315e-06,
+      "loss": 0.0839,
+      "step": 1215
+    },
+    {
+      "epoch": 0.27636363636363637,
+      "grad_norm": 1.29983515225105,
+      "learning_rate": 1.2406164487334077e-06,
+      "loss": 0.0875,
+      "step": 1216
+    },
+    {
+      "epoch": 0.2765909090909091,
+      "grad_norm": 2.1901665993324744,
+      "learning_rate": 1.2406010350490069e-06,
+      "loss": 0.0628,
+      "step": 1217
+    },
+    {
+      "epoch": 0.2768181818181818,
+      "grad_norm": 2.0963230138472,
+      "learning_rate": 1.2405856088114268e-06,
+      "loss": 0.1047,
+      "step": 1218
+    },
+    {
+      "epoch": 0.27704545454545454,
+      "grad_norm": 1.7267740021905045,
+      "learning_rate": 1.2405701700209823e-06,
+      "loss": 0.0508,
+      "step": 1219
+    },
+    {
+      "epoch": 0.2772727272727273,
+      "grad_norm": 2.019619185116833,
+      "learning_rate": 1.2405547186779881e-06,
+      "loss": 0.0797,
+      "step": 1220
+    },
+    {
+      "epoch": 0.2775,
+      "grad_norm": 1.1096692397549046,
+      "learning_rate": 1.2405392547827594e-06,
+      "loss": 0.0506,
+      "step": 1221
+    },
+    {
+      "epoch": 0.2777272727272727,
+      "grad_norm": 2.2615511388416785,
+      "learning_rate": 1.2405237783356115e-06,
+      "loss": 0.0841,
+      "step": 1222
+    },
+    {
+      "epoch": 0.27795454545454545,
+      "grad_norm": 2.0918359387591945,
+      "learning_rate": 1.2405082893368598e-06,
+      "loss": 0.1008,
+      "step": 1223
+    },
+    {
+      "epoch": 0.2781818181818182,
+      "grad_norm": 2.328833683143935,
+      "learning_rate": 1.2404927877868204e-06,
+      "loss": 0.0999,
+      "step": 1224
+    },
+    {
+      "epoch": 0.2784090909090909,
+      "grad_norm": 1.3849526895112838,
+      "learning_rate": 1.240477273685809e-06,
+      "loss": 0.0451,
+      "step": 1225
+    },
+    {
+      "epoch": 0.2786363636363636,
+      "grad_norm": 2.053207876909769,
+      "learning_rate": 1.2404617470341423e-06,
+      "loss": 0.076,
+      "step": 1226
+    },
+    {
+      "epoch": 0.27886363636363637,
+      "grad_norm": 1.9168384089117194,
+      "learning_rate": 1.2404462078321368e-06,
+      "loss": 0.0787,
+      "step": 1227
+    },
+    {
+      "epoch": 0.2790909090909091,
+      "grad_norm": 1.8733743699442347,
+      "learning_rate": 1.2404306560801097e-06,
+      "loss": 0.0634,
+      "step": 1228
+    },
+    {
+      "epoch": 0.2793181818181818,
+      "grad_norm": 1.161031098884299,
+      "learning_rate": 1.2404150917783776e-06,
+      "loss": 0.0459,
+      "step": 1229
+    },
+    {
+      "epoch": 0.27954545454545454,
+      "grad_norm": 1.7938562055234542,
+      "learning_rate": 1.2403995149272581e-06,
+      "loss": 0.0503,
+      "step": 1230
+    },
+    {
+      "epoch": 0.2797727272727273,
+      "grad_norm": 4.701578477266401,
+      "learning_rate": 1.2403839255270693e-06,
+      "loss": 0.1175,
+      "step": 1231
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9405091958790215,
+      "learning_rate": 1.240368323578128e-06,
+      "loss": 0.0743,
+      "step": 1232
+    },
+    {
+      "epoch": 0.2802272727272727,
+      "grad_norm": 2.1300383368275053,
+      "learning_rate": 1.2403527090807535e-06,
+      "loss": 0.0997,
+      "step": 1233
+    },
+    {
+      "epoch": 0.28045454545454546,
+      "grad_norm": 1.6647495958077974,
+      "learning_rate": 1.2403370820352635e-06,
+      "loss": 0.0887,
+      "step": 1234
+    },
+    {
+      "epoch": 0.2806818181818182,
+      "grad_norm": 2.6147224451911195,
+      "learning_rate": 1.240321442441977e-06,
+      "loss": 0.0825,
+      "step": 1235
+    },
+    {
+      "epoch": 0.2809090909090909,
+      "grad_norm": 2.7612885507591827,
+      "learning_rate": 1.2403057903012128e-06,
+      "loss": 0.0909,
+      "step": 1236
+    },
+    {
+      "epoch": 0.28113636363636363,
+      "grad_norm": 1.0812134786546417,
+      "learning_rate": 1.2402901256132898e-06,
+      "loss": 0.0361,
+      "step": 1237
+    },
+    {
+      "epoch": 0.28136363636363637,
+      "grad_norm": 2.2690520240580976,
+      "learning_rate": 1.240274448378528e-06,
+      "loss": 0.0698,
+      "step": 1238
+    },
+    {
+      "epoch": 0.2815909090909091,
+      "grad_norm": 1.9328035164438566,
+      "learning_rate": 1.2402587585972466e-06,
+      "loss": 0.0787,
+      "step": 1239
+    },
+    {
+      "epoch": 0.2818181818181818,
+      "grad_norm": 1.53584596063646,
+      "learning_rate": 1.2402430562697661e-06,
+      "loss": 0.0597,
+      "step": 1240
+    },
+    {
+      "epoch": 0.28204545454545454,
+      "grad_norm": 1.8021391087814707,
+      "learning_rate": 1.240227341396406e-06,
+      "loss": 0.0674,
+      "step": 1241
+    },
+    {
+      "epoch": 0.2822727272727273,
+      "grad_norm": 2.7077474695474684,
+      "learning_rate": 1.2402116139774873e-06,
+      "loss": 0.0907,
+      "step": 1242
+    },
+    {
+      "epoch": 0.2825,
+      "grad_norm": 2.915762225546053,
+      "learning_rate": 1.2401958740133303e-06,
+      "loss": 0.1203,
+      "step": 1243
+    },
+    {
+      "epoch": 0.2827272727272727,
+      "grad_norm": 2.8321183666337344,
+      "learning_rate": 1.2401801215042563e-06,
+      "loss": 0.0737,
+      "step": 1244
+    },
+    {
+      "epoch": 0.28295454545454546,
+      "grad_norm": 1.6691973313862671,
+      "learning_rate": 1.2401643564505863e-06,
+      "loss": 0.0778,
+      "step": 1245
+    },
+    {
+      "epoch": 0.2831818181818182,
+      "grad_norm": 2.0742731435374306,
+      "learning_rate": 1.2401485788526418e-06,
+      "loss": 0.0722,
+      "step": 1246
+    },
+    {
+      "epoch": 0.2834090909090909,
+      "grad_norm": 2.6184662663801213,
+      "learning_rate": 1.2401327887107448e-06,
+      "loss": 0.0823,
+      "step": 1247
+    },
+    {
+      "epoch": 0.28363636363636363,
+      "grad_norm": 1.5745297368727198,
+      "learning_rate": 1.2401169860252168e-06,
+      "loss": 0.0918,
+      "step": 1248
+    },
+    {
+      "epoch": 0.2838636363636364,
+      "grad_norm": 2.080509261584933,
+      "learning_rate": 1.2401011707963806e-06,
+      "loss": 0.0971,
+      "step": 1249
+    },
+    {
+      "epoch": 0.2840909090909091,
+      "grad_norm": 1.0899802212649266,
+      "learning_rate": 1.2400853430245583e-06,
+      "loss": 0.0537,
+      "step": 1250
+    },
+    {
+      "epoch": 0.2843181818181818,
+      "grad_norm": 2.6147911841166236,
+      "learning_rate": 1.2400695027100725e-06,
+      "loss": 0.0811,
+      "step": 1251
+    },
+    {
+      "epoch": 0.28454545454545455,
+      "grad_norm": 2.0243573318290915,
+      "learning_rate": 1.240053649853247e-06,
+      "loss": 0.0626,
+      "step": 1252
+    },
+    {
+      "epoch": 0.2847727272727273,
+      "grad_norm": 2.2694739667224,
+      "learning_rate": 1.2400377844544042e-06,
+      "loss": 0.0744,
+      "step": 1253
+    },
+    {
+      "epoch": 0.285,
+      "grad_norm": 2.1645048766220887,
+      "learning_rate": 1.240021906513868e-06,
+      "loss": 0.0991,
+      "step": 1254
+    },
+    {
+      "epoch": 0.2852272727272727,
+      "grad_norm": 1.7620923011430991,
+      "learning_rate": 1.2400060160319624e-06,
+      "loss": 0.0714,
+      "step": 1255
+    },
+    {
+      "epoch": 0.28545454545454546,
+      "grad_norm": 2.7963710728975655,
+      "learning_rate": 1.2399901130090112e-06,
+      "loss": 0.1812,
+      "step": 1256
+    },
+    {
+      "epoch": 0.2856818181818182,
+      "grad_norm": 1.1327798925838666,
+      "learning_rate": 1.2399741974453385e-06,
+      "loss": 0.0382,
+      "step": 1257
+    },
+    {
+      "epoch": 0.2859090909090909,
+      "grad_norm": 2.382435624887536,
+      "learning_rate": 1.239958269341269e-06,
+      "loss": 0.0847,
+      "step": 1258
+    },
+    {
+      "epoch": 0.28613636363636363,
+      "grad_norm": 3.155060613463019,
+      "learning_rate": 1.2399423286971276e-06,
+      "loss": 0.146,
+      "step": 1259
+    },
+    {
+      "epoch": 0.2863636363636364,
+      "grad_norm": 2.666748197860957,
+      "learning_rate": 1.2399263755132393e-06,
+      "loss": 0.1256,
+      "step": 1260
+    },
+    {
+      "epoch": 0.2865909090909091,
+      "grad_norm": 2.3448548312892283,
+      "learning_rate": 1.2399104097899295e-06,
+      "loss": 0.0974,
+      "step": 1261
+    },
+    {
+      "epoch": 0.2868181818181818,
+      "grad_norm": 2.510445951195572,
+      "learning_rate": 1.2398944315275234e-06,
+      "loss": 0.1426,
+      "step": 1262
+    },
+    {
+      "epoch": 0.28704545454545455,
+      "grad_norm": 2.9064575156068084,
+      "learning_rate": 1.2398784407263476e-06,
+      "loss": 0.12,
+      "step": 1263
+    },
+    {
+      "epoch": 0.2872727272727273,
+      "grad_norm": 2.1100378095469456,
+      "learning_rate": 1.2398624373867273e-06,
+      "loss": 0.1033,
+      "step": 1264
+    },
+    {
+      "epoch": 0.2875,
+      "grad_norm": 1.722608995136586,
+      "learning_rate": 1.2398464215089892e-06,
+      "loss": 0.0503,
+      "step": 1265
+    },
+    {
+      "epoch": 0.2877272727272727,
+      "grad_norm": 1.8813369217033407,
+      "learning_rate": 1.2398303930934601e-06,
+      "loss": 0.078,
+      "step": 1266
+    },
+    {
+      "epoch": 0.28795454545454546,
+      "grad_norm": 2.448222579655358,
+      "learning_rate": 1.2398143521404666e-06,
+      "loss": 0.0952,
+      "step": 1267
+    },
+    {
+      "epoch": 0.2881818181818182,
+      "grad_norm": 3.7208668371852007,
+      "learning_rate": 1.2397982986503357e-06,
+      "loss": 0.1485,
+      "step": 1268
+    },
+    {
+      "epoch": 0.2884090909090909,
+      "grad_norm": 1.7463484585609228,
+      "learning_rate": 1.239782232623395e-06,
+      "loss": 0.0913,
+      "step": 1269
+    },
+    {
+      "epoch": 0.28863636363636364,
+      "grad_norm": 1.221832242084858,
+      "learning_rate": 1.239766154059972e-06,
+      "loss": 0.0451,
+      "step": 1270
+    },
+    {
+      "epoch": 0.2888636363636364,
+      "grad_norm": 2.5486306895583186,
+      "learning_rate": 1.2397500629603948e-06,
+      "loss": 0.0951,
+      "step": 1271
+    },
+    {
+      "epoch": 0.28909090909090907,
+      "grad_norm": 1.6648720279865687,
+      "learning_rate": 1.239733959324991e-06,
+      "loss": 0.1094,
+      "step": 1272
+    },
+    {
+      "epoch": 0.2893181818181818,
+      "grad_norm": 1.5018539333149457,
+      "learning_rate": 1.2397178431540897e-06,
+      "loss": 0.0875,
+      "step": 1273
+    },
+    {
+      "epoch": 0.28954545454545455,
+      "grad_norm": 1.97360524134053,
+      "learning_rate": 1.2397017144480188e-06,
+      "loss": 0.1527,
+      "step": 1274
+    },
+    {
+      "epoch": 0.2897727272727273,
+      "grad_norm": 1.9421809475967196,
+      "learning_rate": 1.2396855732071076e-06,
+      "loss": 0.0556,
+      "step": 1275
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8538880048050657,
+      "learning_rate": 1.2396694194316851e-06,
+      "loss": 0.0345,
+      "step": 1276
+    },
+    {
+      "epoch": 0.2902272727272727,
+      "grad_norm": 1.800995388693947,
+      "learning_rate": 1.2396532531220813e-06,
+      "loss": 0.0662,
+      "step": 1277
+    },
+    {
+      "epoch": 0.29045454545454547,
+      "grad_norm": 2.039566212975476,
+      "learning_rate": 1.239637074278625e-06,
+      "loss": 0.0754,
+      "step": 1278
+    },
+    {
+      "epoch": 0.2906818181818182,
+      "grad_norm": 2.5808249096556732,
+      "learning_rate": 1.2396208829016461e-06,
+      "loss": 0.1446,
+      "step": 1279
+    },
+    {
+      "epoch": 0.2909090909090909,
+      "grad_norm": 1.2474361598706714,
+      "learning_rate": 1.2396046789914756e-06,
+      "loss": 0.0487,
+      "step": 1280
+    },
+    {
+      "epoch": 0.29113636363636364,
+      "grad_norm": 2.0537277058114065,
+      "learning_rate": 1.2395884625484433e-06,
+      "loss": 0.1381,
+      "step": 1281
+    },
+    {
+      "epoch": 0.2913636363636364,
+      "grad_norm": 2.673771111617711,
+      "learning_rate": 1.2395722335728803e-06,
+      "loss": 0.0903,
+      "step": 1282
+    },
+    {
+      "epoch": 0.29159090909090907,
+      "grad_norm": 1.2169250008395864,
+      "learning_rate": 1.239555992065117e-06,
+      "loss": 0.0565,
+      "step": 1283
+    },
+    {
+      "epoch": 0.2918181818181818,
+      "grad_norm": 2.854515953494187,
+      "learning_rate": 1.2395397380254848e-06,
+      "loss": 0.1278,
+      "step": 1284
+    },
+    {
+      "epoch": 0.29204545454545455,
+      "grad_norm": 1.507376397758385,
+      "learning_rate": 1.2395234714543155e-06,
+      "loss": 0.0697,
+      "step": 1285
+    },
+    {
+      "epoch": 0.2922727272727273,
+      "grad_norm": 2.0768220986331443,
+      "learning_rate": 1.2395071923519403e-06,
+      "loss": 0.0862,
+      "step": 1286
+    },
+    {
+      "epoch": 0.2925,
+      "grad_norm": 1.4026693008466886,
+      "learning_rate": 1.2394909007186913e-06,
+      "loss": 0.046,
+      "step": 1287
+    },
+    {
+      "epoch": 0.2927272727272727,
+      "grad_norm": 2.228817655938715,
+      "learning_rate": 1.239474596554901e-06,
+      "loss": 0.0809,
+      "step": 1288
+    },
+    {
+      "epoch": 0.29295454545454547,
+      "grad_norm": 1.480309782773691,
+      "learning_rate": 1.2394582798609017e-06,
+      "loss": 0.0484,
+      "step": 1289
+    },
+    {
+      "epoch": 0.29318181818181815,
+      "grad_norm": 1.6280566424361982,
+      "learning_rate": 1.239441950637026e-06,
+      "loss": 0.0695,
+      "step": 1290
+    },
+    {
+      "epoch": 0.2934090909090909,
+      "grad_norm": 2.207144140111416,
+      "learning_rate": 1.2394256088836069e-06,
+      "loss": 0.1014,
+      "step": 1291
+    },
+    {
+      "epoch": 0.29363636363636364,
+      "grad_norm": 2.5021410890438025,
+      "learning_rate": 1.2394092546009778e-06,
+      "loss": 0.1061,
+      "step": 1292
+    },
+    {
+      "epoch": 0.2938636363636364,
+      "grad_norm": 2.2670624356621465,
+      "learning_rate": 1.2393928877894723e-06,
+      "loss": 0.1161,
+      "step": 1293
+    },
+    {
+      "epoch": 0.29409090909090907,
+      "grad_norm": 2.163277322378986,
+      "learning_rate": 1.2393765084494237e-06,
+      "loss": 0.1284,
+      "step": 1294
+    },
+    {
+      "epoch": 0.2943181818181818,
+      "grad_norm": 1.0285205884151147,
+      "learning_rate": 1.2393601165811661e-06,
+      "loss": 0.0462,
+      "step": 1295
+    },
+    {
+      "epoch": 0.29454545454545455,
+      "grad_norm": 2.12516170975103,
+      "learning_rate": 1.2393437121850342e-06,
+      "loss": 0.0771,
+      "step": 1296
+    },
+    {
+      "epoch": 0.2947727272727273,
+      "grad_norm": 1.522252459976902,
+      "learning_rate": 1.2393272952613621e-06,
+      "loss": 0.0537,
+      "step": 1297
+    },
+    {
+      "epoch": 0.295,
+      "grad_norm": 1.7061278408554315,
+      "learning_rate": 1.2393108658104846e-06,
+      "loss": 0.0901,
+      "step": 1298
+    },
+    {
+      "epoch": 0.2952272727272727,
+      "grad_norm": 1.244897879738489,
+      "learning_rate": 1.239294423832737e-06,
+      "loss": 0.0464,
+      "step": 1299
+    },
+    {
+      "epoch": 0.29545454545454547,
+      "grad_norm": 2.504970841797224,
+      "learning_rate": 1.2392779693284546e-06,
+      "loss": 0.1213,
+      "step": 1300
+    },
+    {
+      "epoch": 0.29568181818181816,
+      "grad_norm": 1.7659153169438087,
+      "learning_rate": 1.2392615022979723e-06,
+      "loss": 0.0659,
+      "step": 1301
+    },
+    {
+      "epoch": 0.2959090909090909,
+      "grad_norm": 2.511623907507154,
+      "learning_rate": 1.2392450227416267e-06,
+      "loss": 0.1315,
+      "step": 1302
+    },
+    {
+      "epoch": 0.29613636363636364,
+      "grad_norm": 1.4588780993812192,
+      "learning_rate": 1.2392285306597532e-06,
+      "loss": 0.0739,
+      "step": 1303
+    },
+    {
+      "epoch": 0.2963636363636364,
+      "grad_norm": 1.689421983947396,
+      "learning_rate": 1.2392120260526886e-06,
+      "loss": 0.07,
+      "step": 1304
+    },
+    {
+      "epoch": 0.29659090909090907,
+      "grad_norm": 2.4152975306497293,
+      "learning_rate": 1.239195508920769e-06,
+      "loss": 0.0949,
+      "step": 1305
+    },
+    {
+      "epoch": 0.2968181818181818,
+      "grad_norm": 1.263006864693506,
+      "learning_rate": 1.2391789792643317e-06,
+      "loss": 0.081,
+      "step": 1306
+    },
+    {
+      "epoch": 0.29704545454545456,
+      "grad_norm": 2.818794335818244,
+      "learning_rate": 1.2391624370837133e-06,
+      "loss": 0.1603,
+      "step": 1307
+    },
+    {
+      "epoch": 0.2972727272727273,
+      "grad_norm": 1.6163050779949044,
+      "learning_rate": 1.2391458823792515e-06,
+      "loss": 0.0989,
+      "step": 1308
+    },
+    {
+      "epoch": 0.2975,
+      "grad_norm": 1.55941530971334,
+      "learning_rate": 1.2391293151512835e-06,
+      "loss": 0.0903,
+      "step": 1309
+    },
+    {
+      "epoch": 0.29772727272727273,
+      "grad_norm": 1.479112887955286,
+      "learning_rate": 1.2391127354001475e-06,
+      "loss": 0.0695,
+      "step": 1310
+    },
+    {
+      "epoch": 0.29795454545454547,
+      "grad_norm": 2.2306555785535678,
+      "learning_rate": 1.2390961431261814e-06,
+      "loss": 0.1366,
+      "step": 1311
+    },
+    {
+      "epoch": 0.29818181818181816,
+      "grad_norm": 1.6413834419560913,
+      "learning_rate": 1.239079538329724e-06,
+      "loss": 0.0679,
+      "step": 1312
+    },
+    {
+      "epoch": 0.2984090909090909,
+      "grad_norm": 2.2081515671310785,
+      "learning_rate": 1.239062921011113e-06,
+      "loss": 0.0839,
+      "step": 1313
+    },
+    {
+      "epoch": 0.29863636363636364,
+      "grad_norm": 1.6250031109936125,
+      "learning_rate": 1.2390462911706877e-06,
+      "loss": 0.0745,
+      "step": 1314
+    },
+    {
+      "epoch": 0.2988636363636364,
+      "grad_norm": 1.2597375582479309,
+      "learning_rate": 1.2390296488087874e-06,
+      "loss": 0.0723,
+      "step": 1315
+    },
+    {
+      "epoch": 0.2990909090909091,
+      "grad_norm": 1.9611777017686367,
+      "learning_rate": 1.2390129939257515e-06,
+      "loss": 0.0828,
+      "step": 1316
+    },
+    {
+      "epoch": 0.2993181818181818,
+      "grad_norm": 1.8089848215539563,
+      "learning_rate": 1.2389963265219194e-06,
+      "loss": 0.0826,
+      "step": 1317
+    },
+    {
+      "epoch": 0.29954545454545456,
+      "grad_norm": 1.4991101051455944,
+      "learning_rate": 1.2389796465976309e-06,
+      "loss": 0.0539,
+      "step": 1318
+    },
+    {
+      "epoch": 0.29977272727272725,
+      "grad_norm": 3.2446077559145405,
+      "learning_rate": 1.2389629541532263e-06,
+      "loss": 0.1227,
+      "step": 1319
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0987104283726077,
+      "learning_rate": 1.238946249189046e-06,
+      "loss": 0.0885,
+      "step": 1320
+    },
+    {
+      "epoch": 0.30022727272727273,
+      "grad_norm": 2.5863813020378585,
+      "learning_rate": 1.2389295317054306e-06,
+      "loss": 0.1306,
+      "step": 1321
+    },
+    {
+      "epoch": 0.3004545454545455,
+      "grad_norm": 3.033306792730616,
+      "learning_rate": 1.2389128017027208e-06,
+      "loss": 0.1548,
+      "step": 1322
+    },
+    {
+      "epoch": 0.30068181818181816,
+      "grad_norm": 1.984134932574022,
+      "learning_rate": 1.2388960591812582e-06,
+      "loss": 0.0673,
+      "step": 1323
+    },
+    {
+      "epoch": 0.3009090909090909,
+      "grad_norm": 2.6833419192995374,
+      "learning_rate": 1.2388793041413838e-06,
+      "loss": 0.1554,
+      "step": 1324
+    },
+    {
+      "epoch": 0.30113636363636365,
+      "grad_norm": 3.4368075668350992,
+      "learning_rate": 1.2388625365834394e-06,
+      "loss": 0.2215,
+      "step": 1325
+    },
+    {
+      "epoch": 0.3013636363636364,
+      "grad_norm": 1.9317468716522008,
+      "learning_rate": 1.238845756507767e-06,
+      "loss": 0.1163,
+      "step": 1326
+    },
+    {
+      "epoch": 0.3015909090909091,
+      "grad_norm": 1.8689165736391575,
+      "learning_rate": 1.2388289639147088e-06,
+      "loss": 0.052,
+      "step": 1327
+    },
+    {
+      "epoch": 0.3018181818181818,
+      "grad_norm": 1.3675110756881914,
+      "learning_rate": 1.2388121588046071e-06,
+      "loss": 0.0777,
+      "step": 1328
+    },
+    {
+      "epoch": 0.30204545454545456,
+      "grad_norm": 2.2637530084850725,
+      "learning_rate": 1.2387953411778045e-06,
+      "loss": 0.1146,
+      "step": 1329
+    },
+    {
+      "epoch": 0.30227272727272725,
+      "grad_norm": 1.9247130682390652,
+      "learning_rate": 1.238778511034644e-06,
+      "loss": 0.0743,
+      "step": 1330
+    },
+    {
+      "epoch": 0.3025,
+      "grad_norm": 2.63862948281546,
+      "learning_rate": 1.2387616683754691e-06,
+      "loss": 0.1239,
+      "step": 1331
+    },
+    {
+      "epoch": 0.30272727272727273,
+      "grad_norm": 2.0076326115626157,
+      "learning_rate": 1.238744813200623e-06,
+      "loss": 0.0567,
+      "step": 1332
+    },
+    {
+      "epoch": 0.3029545454545455,
+      "grad_norm": 1.554211602845707,
+      "learning_rate": 1.2387279455104493e-06,
+      "loss": 0.0739,
+      "step": 1333
+    },
+    {
+      "epoch": 0.30318181818181816,
+      "grad_norm": 3.349058834588653,
+      "learning_rate": 1.2387110653052922e-06,
+      "loss": 0.1226,
+      "step": 1334
+    },
+    {
+      "epoch": 0.3034090909090909,
+      "grad_norm": 1.6049424892136057,
+      "learning_rate": 1.2386941725854956e-06,
+      "loss": 0.0717,
+      "step": 1335
+    },
+    {
+      "epoch": 0.30363636363636365,
+      "grad_norm": 2.132620729368777,
+      "learning_rate": 1.2386772673514044e-06,
+      "loss": 0.1059,
+      "step": 1336
+    },
+    {
+      "epoch": 0.3038636363636364,
+      "grad_norm": 1.8588838501372325,
+      "learning_rate": 1.2386603496033631e-06,
+      "loss": 0.0668,
+      "step": 1337
+    },
+    {
+      "epoch": 0.3040909090909091,
+      "grad_norm": 2.3505477867090634,
+      "learning_rate": 1.2386434193417165e-06,
+      "loss": 0.1159,
+      "step": 1338
+    },
+    {
+      "epoch": 0.3043181818181818,
+      "grad_norm": 1.2944787836007232,
+      "learning_rate": 1.23862647656681e-06,
+      "loss": 0.0497,
+      "step": 1339
+    },
+    {
+      "epoch": 0.30454545454545456,
+      "grad_norm": 2.4151675681732856,
+      "learning_rate": 1.2386095212789892e-06,
+      "loss": 0.0887,
+      "step": 1340
+    },
+    {
+      "epoch": 0.30477272727272725,
+      "grad_norm": 1.7667463018404936,
+      "learning_rate": 1.2385925534786e-06,
+      "loss": 0.0645,
+      "step": 1341
+    },
+    {
+      "epoch": 0.305,
+      "grad_norm": 2.0754039339687167,
+      "learning_rate": 1.2385755731659879e-06,
+      "loss": 0.059,
+      "step": 1342
+    },
+    {
+      "epoch": 0.30522727272727274,
+      "grad_norm": 1.6629693636815264,
+      "learning_rate": 1.2385585803414995e-06,
+      "loss": 0.0896,
+      "step": 1343
+    },
+    {
+      "epoch": 0.3054545454545455,
+      "grad_norm": 2.0540590430415753,
+      "learning_rate": 1.238541575005481e-06,
+      "loss": 0.1099,
+      "step": 1344
+    },
+    {
+      "epoch": 0.30568181818181817,
+      "grad_norm": 1.8113002554993654,
+      "learning_rate": 1.2385245571582797e-06,
+      "loss": 0.0902,
+      "step": 1345
+    },
+    {
+      "epoch": 0.3059090909090909,
+      "grad_norm": 2.533964910408494,
+      "learning_rate": 1.2385075268002423e-06,
+      "loss": 0.0895,
+      "step": 1346
+    },
+    {
+      "epoch": 0.30613636363636365,
+      "grad_norm": 1.0807929753467116,
+      "learning_rate": 1.238490483931716e-06,
+      "loss": 0.049,
+      "step": 1347
+    },
+    {
+      "epoch": 0.30636363636363634,
+      "grad_norm": 1.2927135996228083,
+      "learning_rate": 1.2384734285530484e-06,
+      "loss": 0.0537,
+      "step": 1348
+    },
+    {
+      "epoch": 0.3065909090909091,
+      "grad_norm": 2.2588593784223625,
+      "learning_rate": 1.2384563606645875e-06,
+      "loss": 0.0759,
+      "step": 1349
+    },
+    {
+      "epoch": 0.3068181818181818,
+      "grad_norm": 2.510567386109347,
+      "learning_rate": 1.238439280266681e-06,
+      "loss": 0.1177,
+      "step": 1350
+    },
+    {
+      "epoch": 0.30704545454545457,
+      "grad_norm": 2.9181222393386066,
+      "learning_rate": 1.2384221873596775e-06,
+      "loss": 0.1166,
+      "step": 1351
+    },
+    {
+      "epoch": 0.30727272727272725,
+      "grad_norm": 3.0103266056346767,
+      "learning_rate": 1.2384050819439252e-06,
+      "loss": 0.1379,
+      "step": 1352
+    },
+    {
+      "epoch": 0.3075,
+      "grad_norm": 1.9177396316497677,
+      "learning_rate": 1.2383879640197736e-06,
+      "loss": 0.0947,
+      "step": 1353
+    },
+    {
+      "epoch": 0.30772727272727274,
+      "grad_norm": 2.429556802355941,
+      "learning_rate": 1.238370833587571e-06,
+      "loss": 0.1012,
+      "step": 1354
+    },
+    {
+      "epoch": 0.3079545454545455,
+      "grad_norm": 1.3549988960782664,
+      "learning_rate": 1.2383536906476668e-06,
+      "loss": 0.0342,
+      "step": 1355
+    },
+    {
+      "epoch": 0.30818181818181817,
+      "grad_norm": 2.7515198593251604,
+      "learning_rate": 1.2383365352004111e-06,
+      "loss": 0.1166,
+      "step": 1356
+    },
+    {
+      "epoch": 0.3084090909090909,
+      "grad_norm": 1.0535861044290507,
+      "learning_rate": 1.2383193672461534e-06,
+      "loss": 0.0498,
+      "step": 1357
+    },
+    {
+      "epoch": 0.30863636363636365,
+      "grad_norm": 1.4364025876072166,
+      "learning_rate": 1.2383021867852438e-06,
+      "loss": 0.0862,
+      "step": 1358
+    },
+    {
+      "epoch": 0.30886363636363634,
+      "grad_norm": 2.313918281440171,
+      "learning_rate": 1.2382849938180324e-06,
+      "loss": 0.1274,
+      "step": 1359
+    },
+    {
+      "epoch": 0.3090909090909091,
+      "grad_norm": 1.912066895872404,
+      "learning_rate": 1.2382677883448704e-06,
+      "loss": 0.1404,
+      "step": 1360
+    },
+    {
+      "epoch": 0.3093181818181818,
+      "grad_norm": 2.040369813292881,
+      "learning_rate": 1.238250570366108e-06,
+      "loss": 0.0844,
+      "step": 1361
+    },
+    {
+      "epoch": 0.30954545454545457,
+      "grad_norm": 1.713904144474748,
+      "learning_rate": 1.238233339882097e-06,
+      "loss": 0.0908,
+      "step": 1362
+    },
+    {
+      "epoch": 0.30977272727272726,
+      "grad_norm": 1.2275948450897605,
+      "learning_rate": 1.238216096893188e-06,
+      "loss": 0.0438,
+      "step": 1363
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6618489020106064,
+      "learning_rate": 1.238198841399733e-06,
+      "loss": 0.0626,
+      "step": 1364
+    },
+    {
+      "epoch": 0.31022727272727274,
+      "grad_norm": 3.227270357917131,
+      "learning_rate": 1.2381815734020838e-06,
+      "loss": 0.1547,
+      "step": 1365
+    },
+    {
+      "epoch": 0.3104545454545454,
+      "grad_norm": 1.0136952181636993,
+      "learning_rate": 1.2381642929005927e-06,
+      "loss": 0.0595,
+      "step": 1366
+    },
+    {
+      "epoch": 0.31068181818181817,
+      "grad_norm": 1.93113274211823,
+      "learning_rate": 1.238146999895612e-06,
+      "loss": 0.082,
+      "step": 1367
+    },
+    {
+      "epoch": 0.3109090909090909,
+      "grad_norm": 1.588806503158415,
+      "learning_rate": 1.2381296943874943e-06,
+      "loss": 0.0768,
+      "step": 1368
+    },
+    {
+      "epoch": 0.31113636363636366,
+      "grad_norm": 2.821143223005106,
+      "learning_rate": 1.2381123763765922e-06,
+      "loss": 0.1251,
+      "step": 1369
+    },
+    {
+      "epoch": 0.31136363636363634,
+      "grad_norm": 2.398461523004137,
+      "learning_rate": 1.2380950458632593e-06,
+      "loss": 0.0928,
+      "step": 1370
+    },
+    {
+      "epoch": 0.3115909090909091,
+      "grad_norm": 2.314995380197198,
+      "learning_rate": 1.238077702847849e-06,
+      "loss": 0.1651,
+      "step": 1371
+    },
+    {
+      "epoch": 0.3118181818181818,
+      "grad_norm": 2.221367553911642,
+      "learning_rate": 1.2380603473307144e-06,
+      "loss": 0.07,
+      "step": 1372
+    },
+    {
+      "epoch": 0.31204545454545457,
+      "grad_norm": 1.6553449086603227,
+      "learning_rate": 1.2380429793122098e-06,
+      "loss": 0.0784,
+      "step": 1373
+    },
+    {
+      "epoch": 0.31227272727272726,
+      "grad_norm": 3.4255192158926615,
+      "learning_rate": 1.2380255987926896e-06,
+      "loss": 0.1199,
+      "step": 1374
+    },
+    {
+      "epoch": 0.3125,
+      "grad_norm": 1.4556770204636993,
+      "learning_rate": 1.2380082057725077e-06,
+      "loss": 0.0819,
+      "step": 1375
+    },
+    {
+      "epoch": 0.31272727272727274,
+      "grad_norm": 2.1167758823418197,
+      "learning_rate": 1.2379908002520191e-06,
+      "loss": 0.0886,
+      "step": 1376
+    },
+    {
+      "epoch": 0.31295454545454543,
+      "grad_norm": 2.60407649582862,
+      "learning_rate": 1.2379733822315786e-06,
+      "loss": 0.1672,
+      "step": 1377
+    },
+    {
+      "epoch": 0.3131818181818182,
+      "grad_norm": 1.5629969321384358,
+      "learning_rate": 1.2379559517115416e-06,
+      "loss": 0.0907,
+      "step": 1378
+    },
+    {
+      "epoch": 0.3134090909090909,
+      "grad_norm": 2.6416927595004256,
+      "learning_rate": 1.2379385086922635e-06,
+      "loss": 0.153,
+      "step": 1379
+    },
+    {
+      "epoch": 0.31363636363636366,
+      "grad_norm": 1.6352108310920914,
+      "learning_rate": 1.2379210531740996e-06,
+      "loss": 0.1017,
+      "step": 1380
+    },
+    {
+      "epoch": 0.31386363636363634,
+      "grad_norm": 1.8098549515246363,
+      "learning_rate": 1.2379035851574063e-06,
+      "loss": 0.1039,
+      "step": 1381
+    },
+    {
+      "epoch": 0.3140909090909091,
+      "grad_norm": 2.012273112503458,
+      "learning_rate": 1.2378861046425395e-06,
+      "loss": 0.0799,
+      "step": 1382
+    },
+    {
+      "epoch": 0.31431818181818183,
+      "grad_norm": 2.353014223725641,
+      "learning_rate": 1.2378686116298558e-06,
+      "loss": 0.0885,
+      "step": 1383
+    },
+    {
+      "epoch": 0.3145454545454546,
+      "grad_norm": 1.2905600962314943,
+      "learning_rate": 1.2378511061197121e-06,
+      "loss": 0.053,
+      "step": 1384
+    },
+    {
+      "epoch": 0.31477272727272726,
+      "grad_norm": 2.2776915706375283,
+      "learning_rate": 1.2378335881124651e-06,
+      "loss": 0.1219,
+      "step": 1385
+    },
+    {
+      "epoch": 0.315,
+      "grad_norm": 1.3826336906419114,
+      "learning_rate": 1.237816057608472e-06,
+      "loss": 0.0427,
+      "step": 1386
+    },
+    {
+      "epoch": 0.31522727272727274,
+      "grad_norm": 2.6421133925460665,
+      "learning_rate": 1.2377985146080904e-06,
+      "loss": 0.1335,
+      "step": 1387
+    },
+    {
+      "epoch": 0.31545454545454543,
+      "grad_norm": 2.1237594887121736,
+      "learning_rate": 1.237780959111678e-06,
+      "loss": 0.0769,
+      "step": 1388
+    },
+    {
+      "epoch": 0.3156818181818182,
+      "grad_norm": 1.4458374485607837,
+      "learning_rate": 1.2377633911195927e-06,
+      "loss": 0.1088,
+      "step": 1389
+    },
+    {
+      "epoch": 0.3159090909090909,
+      "grad_norm": 1.719106556417674,
+      "learning_rate": 1.237745810632193e-06,
+      "loss": 0.0623,
+      "step": 1390
+    },
+    {
+      "epoch": 0.31613636363636366,
+      "grad_norm": 1.589175111375019,
+      "learning_rate": 1.2377282176498371e-06,
+      "loss": 0.0637,
+      "step": 1391
+    },
+    {
+      "epoch": 0.31636363636363635,
+      "grad_norm": 2.7224177672329857,
+      "learning_rate": 1.2377106121728839e-06,
+      "loss": 0.1974,
+      "step": 1392
+    },
+    {
+      "epoch": 0.3165909090909091,
+      "grad_norm": 1.4673332012238611,
+      "learning_rate": 1.2376929942016923e-06,
+      "loss": 0.0701,
+      "step": 1393
+    },
+    {
+      "epoch": 0.31681818181818183,
+      "grad_norm": 1.2787936076073336,
+      "learning_rate": 1.2376753637366217e-06,
+      "loss": 0.0596,
+      "step": 1394
+    },
+    {
+      "epoch": 0.3170454545454545,
+      "grad_norm": 1.2646717405421106,
+      "learning_rate": 1.237657720778032e-06,
+      "loss": 0.0669,
+      "step": 1395
+    },
+    {
+      "epoch": 0.31727272727272726,
+      "grad_norm": 2.4949260226305605,
+      "learning_rate": 1.2376400653262817e-06,
+      "loss": 0.1023,
+      "step": 1396
+    },
+    {
+      "epoch": 0.3175,
+      "grad_norm": 1.6833671603112903,
+      "learning_rate": 1.2376223973817324e-06,
+      "loss": 0.0711,
+      "step": 1397
+    },
+    {
+      "epoch": 0.31772727272727275,
+      "grad_norm": 3.0825055528262792,
+      "learning_rate": 1.2376047169447433e-06,
+      "loss": 0.1191,
+      "step": 1398
+    },
+    {
+      "epoch": 0.31795454545454543,
+      "grad_norm": 2.483655456204137,
+      "learning_rate": 1.2375870240156754e-06,
+      "loss": 0.1176,
+      "step": 1399
+    },
+    {
+      "epoch": 0.3181818181818182,
+      "grad_norm": 1.8443325250287883,
+      "learning_rate": 1.2375693185948891e-06,
+      "loss": 0.1232,
+      "step": 1400
+    },
+    {
+      "epoch": 0.3184090909090909,
+      "grad_norm": 1.9232943537661107,
+      "learning_rate": 1.237551600682746e-06,
+      "loss": 0.1012,
+      "step": 1401
+    },
+    {
+      "epoch": 0.31863636363636366,
+      "grad_norm": 2.104756065489708,
+      "learning_rate": 1.237533870279607e-06,
+      "loss": 0.1081,
+      "step": 1402
+    },
+    {
+      "epoch": 0.31886363636363635,
+      "grad_norm": 3.141492397819032,
+      "learning_rate": 1.2375161273858338e-06,
+      "loss": 0.1033,
+      "step": 1403
+    },
+    {
+      "epoch": 0.3190909090909091,
+      "grad_norm": 1.5780329640570199,
+      "learning_rate": 1.2374983720017881e-06,
+      "loss": 0.0772,
+      "step": 1404
+    },
+    {
+      "epoch": 0.31931818181818183,
+      "grad_norm": 4.416999262177611,
+      "learning_rate": 1.237480604127832e-06,
+      "loss": 0.3898,
+      "step": 1405
+    },
+    {
+      "epoch": 0.3195454545454545,
+      "grad_norm": 1.638343245183661,
+      "learning_rate": 1.237462823764328e-06,
+      "loss": 0.0705,
+      "step": 1406
+    },
+    {
+      "epoch": 0.31977272727272726,
+      "grad_norm": 2.3035505239797263,
+      "learning_rate": 1.2374450309116384e-06,
+      "loss": 0.1055,
+      "step": 1407
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.48698871199634,
+      "learning_rate": 1.2374272255701262e-06,
+      "loss": 0.0831,
+      "step": 1408
+    },
+    {
+      "epoch": 0.32022727272727275,
+      "grad_norm": 1.750792535531329,
+      "learning_rate": 1.2374094077401545e-06,
+      "loss": 0.1007,
+      "step": 1409
+    },
+    {
+      "epoch": 0.32045454545454544,
+      "grad_norm": 3.5086717415998545,
+      "learning_rate": 1.2373915774220863e-06,
+      "loss": 0.1302,
+      "step": 1410
+    },
+    {
+      "epoch": 0.3206818181818182,
+      "grad_norm": 2.286312721348272,
+      "learning_rate": 1.2373737346162857e-06,
+      "loss": 0.1268,
+      "step": 1411
+    },
+    {
+      "epoch": 0.3209090909090909,
+      "grad_norm": 4.0452991589937835,
+      "learning_rate": 1.2373558793231163e-06,
+      "loss": 0.1037,
+      "step": 1412
+    },
+    {
+      "epoch": 0.3211363636363636,
+      "grad_norm": 2.133524892583949,
+      "learning_rate": 1.2373380115429421e-06,
+      "loss": 0.0921,
+      "step": 1413
+    },
+    {
+      "epoch": 0.32136363636363635,
+      "grad_norm": 1.9089442634393268,
+      "learning_rate": 1.2373201312761277e-06,
+      "loss": 0.0507,
+      "step": 1414
+    },
+    {
+      "epoch": 0.3215909090909091,
+      "grad_norm": 1.5059091394837045,
+      "learning_rate": 1.2373022385230374e-06,
+      "loss": 0.0889,
+      "step": 1415
+    },
+    {
+      "epoch": 0.32181818181818184,
+      "grad_norm": 2.039585134476412,
+      "learning_rate": 1.2372843332840364e-06,
+      "loss": 0.0629,
+      "step": 1416
+    },
+    {
+      "epoch": 0.3220454545454545,
+      "grad_norm": 1.8226421589383937,
+      "learning_rate": 1.2372664155594896e-06,
+      "loss": 0.0756,
+      "step": 1417
+    },
+    {
+      "epoch": 0.32227272727272727,
+      "grad_norm": 2.215458260725426,
+      "learning_rate": 1.2372484853497624e-06,
+      "loss": 0.1006,
+      "step": 1418
+    },
+    {
+      "epoch": 0.3225,
+      "grad_norm": 1.3627922961336536,
+      "learning_rate": 1.2372305426552207e-06,
+      "loss": 0.058,
+      "step": 1419
+    },
+    {
+      "epoch": 0.32272727272727275,
+      "grad_norm": 1.968842341432665,
+      "learning_rate": 1.2372125874762297e-06,
+      "loss": 0.0741,
+      "step": 1420
+    },
+    {
+      "epoch": 0.32295454545454544,
+      "grad_norm": 1.891018510949674,
+      "learning_rate": 1.2371946198131563e-06,
+      "loss": 0.0854,
+      "step": 1421
+    },
+    {
+      "epoch": 0.3231818181818182,
+      "grad_norm": 2.228256408971778,
+      "learning_rate": 1.2371766396663664e-06,
+      "loss": 0.136,
+      "step": 1422
+    },
+    {
+      "epoch": 0.3234090909090909,
+      "grad_norm": 2.370075926193462,
+      "learning_rate": 1.2371586470362268e-06,
+      "loss": 0.0998,
+      "step": 1423
+    },
+    {
+      "epoch": 0.3236363636363636,
+      "grad_norm": 1.728005860963192,
+      "learning_rate": 1.2371406419231045e-06,
+      "loss": 0.0994,
+      "step": 1424
+    },
+    {
+      "epoch": 0.32386363636363635,
+      "grad_norm": 1.9278264563597745,
+      "learning_rate": 1.2371226243273667e-06,
+      "loss": 0.0833,
+      "step": 1425
+    },
+    {
+      "epoch": 0.3240909090909091,
+      "grad_norm": 1.2944564277657056,
+      "learning_rate": 1.2371045942493804e-06,
+      "loss": 0.0648,
+      "step": 1426
+    },
+    {
+      "epoch": 0.32431818181818184,
+      "grad_norm": 1.4370451310826537,
+      "learning_rate": 1.2370865516895138e-06,
+      "loss": 0.0556,
+      "step": 1427
+    },
+    {
+      "epoch": 0.3245454545454545,
+      "grad_norm": 2.4448366206975427,
+      "learning_rate": 1.2370684966481344e-06,
+      "loss": 0.1144,
+      "step": 1428
+    },
+    {
+      "epoch": 0.32477272727272727,
+      "grad_norm": 1.8763596626197856,
+      "learning_rate": 1.2370504291256108e-06,
+      "loss": 0.0739,
+      "step": 1429
+    },
+    {
+      "epoch": 0.325,
+      "grad_norm": 2.0729993084379883,
+      "learning_rate": 1.2370323491223108e-06,
+      "loss": 0.0833,
+      "step": 1430
+    },
+    {
+      "epoch": 0.32522727272727275,
+      "grad_norm": 1.9668558999205996,
+      "learning_rate": 1.2370142566386038e-06,
+      "loss": 0.0759,
+      "step": 1431
+    },
+    {
+      "epoch": 0.32545454545454544,
+      "grad_norm": 1.7089744086620284,
+      "learning_rate": 1.2369961516748581e-06,
+      "loss": 0.0399,
+      "step": 1432
+    },
+    {
+      "epoch": 0.3256818181818182,
+      "grad_norm": 1.8551457553380208,
+      "learning_rate": 1.2369780342314434e-06,
+      "loss": 0.096,
+      "step": 1433
+    },
+    {
+      "epoch": 0.3259090909090909,
+      "grad_norm": 2.2395254052463507,
+      "learning_rate": 1.236959904308729e-06,
+      "loss": 0.0988,
+      "step": 1434
+    },
+    {
+      "epoch": 0.3261363636363636,
+      "grad_norm": 2.095575193591785,
+      "learning_rate": 1.2369417619070841e-06,
+      "loss": 0.0898,
+      "step": 1435
+    },
+    {
+      "epoch": 0.32636363636363636,
+      "grad_norm": 1.4288699427843854,
+      "learning_rate": 1.2369236070268795e-06,
+      "loss": 0.0593,
+      "step": 1436
+    },
+    {
+      "epoch": 0.3265909090909091,
+      "grad_norm": 1.6061391458434509,
+      "learning_rate": 1.2369054396684847e-06,
+      "loss": 0.0648,
+      "step": 1437
+    },
+    {
+      "epoch": 0.32681818181818184,
+      "grad_norm": 1.8357314627266657,
+      "learning_rate": 1.2368872598322707e-06,
+      "loss": 0.0739,
+      "step": 1438
+    },
+    {
+      "epoch": 0.32704545454545453,
+      "grad_norm": 1.8959224679952769,
+      "learning_rate": 1.2368690675186079e-06,
+      "loss": 0.0954,
+      "step": 1439
+    },
+    {
+      "epoch": 0.32727272727272727,
+      "grad_norm": 2.157378939052919,
+      "learning_rate": 1.2368508627278673e-06,
+      "loss": 0.0902,
+      "step": 1440
+    },
+    {
+      "epoch": 0.3275,
+      "grad_norm": 1.502984113413921,
+      "learning_rate": 1.2368326454604201e-06,
+      "loss": 0.0461,
+      "step": 1441
+    },
+    {
+      "epoch": 0.3277272727272727,
+      "grad_norm": 2.102461969903016,
+      "learning_rate": 1.236814415716638e-06,
+      "loss": 0.1039,
+      "step": 1442
+    },
+    {
+      "epoch": 0.32795454545454544,
+      "grad_norm": 3.994808339939499,
+      "learning_rate": 1.2367961734968926e-06,
+      "loss": 0.1572,
+      "step": 1443
+    },
+    {
+      "epoch": 0.3281818181818182,
+      "grad_norm": 1.9582538656434603,
+      "learning_rate": 1.2367779188015559e-06,
+      "loss": 0.0677,
+      "step": 1444
+    },
+    {
+      "epoch": 0.32840909090909093,
+      "grad_norm": 2.0059873979661846,
+      "learning_rate": 1.236759651631e-06,
+      "loss": 0.0892,
+      "step": 1445
+    },
+    {
+      "epoch": 0.3286363636363636,
+      "grad_norm": 1.6111491896250354,
+      "learning_rate": 1.2367413719855976e-06,
+      "loss": 0.0556,
+      "step": 1446
+    },
+    {
+      "epoch": 0.32886363636363636,
+      "grad_norm": 1.6709751875712553,
+      "learning_rate": 1.2367230798657214e-06,
+      "loss": 0.0859,
+      "step": 1447
+    },
+    {
+      "epoch": 0.3290909090909091,
+      "grad_norm": 1.6589722887971194,
+      "learning_rate": 1.2367047752717443e-06,
+      "loss": 0.0712,
+      "step": 1448
+    },
+    {
+      "epoch": 0.32931818181818184,
+      "grad_norm": 1.736227993875631,
+      "learning_rate": 1.2366864582040396e-06,
+      "loss": 0.0646,
+      "step": 1449
+    },
+    {
+      "epoch": 0.32954545454545453,
+      "grad_norm": 2.1489142060416273,
+      "learning_rate": 1.2366681286629811e-06,
+      "loss": 0.0721,
+      "step": 1450
+    },
+    {
+      "epoch": 0.3297727272727273,
+      "grad_norm": 1.2930262747371242,
+      "learning_rate": 1.2366497866489423e-06,
+      "loss": 0.0411,
+      "step": 1451
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2755834153855745,
+      "learning_rate": 1.236631432162297e-06,
+      "loss": 0.1189,
+      "step": 1452
+    },
+    {
+      "epoch": 0.3302272727272727,
+      "grad_norm": 0.4856784585304358,
+      "learning_rate": 1.23661306520342e-06,
+      "loss": 0.0293,
+      "step": 1453
+    },
+    {
+      "epoch": 0.33045454545454545,
+      "grad_norm": 2.5981183754144834,
+      "learning_rate": 1.2365946857726857e-06,
+      "loss": 0.0957,
+      "step": 1454
+    },
+    {
+      "epoch": 0.3306818181818182,
+      "grad_norm": 0.5319644453561735,
+      "learning_rate": 1.2365762938704687e-06,
+      "loss": 0.0292,
+      "step": 1455
+    },
+    {
+      "epoch": 0.33090909090909093,
+      "grad_norm": 2.352033509870476,
+      "learning_rate": 1.236557889497144e-06,
+      "loss": 0.086,
+      "step": 1456
+    },
+    {
+      "epoch": 0.3311363636363636,
+      "grad_norm": 1.5407525767834607,
+      "learning_rate": 1.236539472653087e-06,
+      "loss": 0.053,
+      "step": 1457
+    },
+    {
+      "epoch": 0.33136363636363636,
+      "grad_norm": 3.151539979759172,
+      "learning_rate": 1.2365210433386733e-06,
+      "loss": 0.11,
+      "step": 1458
+    },
+    {
+      "epoch": 0.3315909090909091,
+      "grad_norm": 1.5926509454663764,
+      "learning_rate": 1.2365026015542786e-06,
+      "loss": 0.099,
+      "step": 1459
+    },
+    {
+      "epoch": 0.33181818181818185,
+      "grad_norm": 1.8176368587498608,
+      "learning_rate": 1.2364841473002792e-06,
+      "loss": 0.0819,
+      "step": 1460
+    },
+    {
+      "epoch": 0.33204545454545453,
+      "grad_norm": 1.2864991254920919,
+      "learning_rate": 1.236465680577051e-06,
+      "loss": 0.0416,
+      "step": 1461
+    },
+    {
+      "epoch": 0.3322727272727273,
+      "grad_norm": 2.1805454805063427,
+      "learning_rate": 1.2364472013849711e-06,
+      "loss": 0.0744,
+      "step": 1462
+    },
+    {
+      "epoch": 0.3325,
+      "grad_norm": 1.3118861332755292,
+      "learning_rate": 1.236428709724416e-06,
+      "loss": 0.0977,
+      "step": 1463
+    },
+    {
+      "epoch": 0.3327272727272727,
+      "grad_norm": 2.1027403778508593,
+      "learning_rate": 1.2364102055957625e-06,
+      "loss": 0.0629,
+      "step": 1464
+    },
+    {
+      "epoch": 0.33295454545454545,
+      "grad_norm": 3.4676947138180845,
+      "learning_rate": 1.2363916889993884e-06,
+      "loss": 0.1298,
+      "step": 1465
+    },
+    {
+      "epoch": 0.3331818181818182,
+      "grad_norm": 2.093364664468466,
+      "learning_rate": 1.2363731599356712e-06,
+      "loss": 0.0907,
+      "step": 1466
+    },
+    {
+      "epoch": 0.33340909090909093,
+      "grad_norm": 2.3682045348140788,
+      "learning_rate": 1.2363546184049887e-06,
+      "loss": 0.0707,
+      "step": 1467
+    },
+    {
+      "epoch": 0.3336363636363636,
+      "grad_norm": 2.2132664595873646,
+      "learning_rate": 1.2363360644077192e-06,
+      "loss": 0.0896,
+      "step": 1468
+    },
+    {
+      "epoch": 0.33386363636363636,
+      "grad_norm": 2.716893539927445,
+      "learning_rate": 1.2363174979442403e-06,
+      "loss": 0.1416,
+      "step": 1469
+    },
+    {
+      "epoch": 0.3340909090909091,
+      "grad_norm": 1.4857401680488966,
+      "learning_rate": 1.2362989190149314e-06,
+      "loss": 0.0855,
+      "step": 1470
+    },
+    {
+      "epoch": 0.3343181818181818,
+      "grad_norm": 2.7016399668524103,
+      "learning_rate": 1.2362803276201709e-06,
+      "loss": 0.1199,
+      "step": 1471
+    },
+    {
+      "epoch": 0.33454545454545453,
+      "grad_norm": 2.159216470017884,
+      "learning_rate": 1.2362617237603384e-06,
+      "loss": 0.1086,
+      "step": 1472
+    },
+    {
+      "epoch": 0.3347727272727273,
+      "grad_norm": 1.4123583851049837,
+      "learning_rate": 1.2362431074358128e-06,
+      "loss": 0.0921,
+      "step": 1473
+    },
+    {
+      "epoch": 0.335,
+      "grad_norm": 1.87963112102474,
+      "learning_rate": 1.2362244786469737e-06,
+      "loss": 0.0736,
+      "step": 1474
+    },
+    {
+      "epoch": 0.3352272727272727,
+      "grad_norm": 1.6761129736886362,
+      "learning_rate": 1.2362058373942013e-06,
+      "loss": 0.1525,
+      "step": 1475
+    },
+    {
+      "epoch": 0.33545454545454545,
+      "grad_norm": 1.8634596612688648,
+      "learning_rate": 1.2361871836778755e-06,
+      "loss": 0.0924,
+      "step": 1476
+    },
+    {
+      "epoch": 0.3356818181818182,
+      "grad_norm": 1.6810807579733587,
+      "learning_rate": 1.2361685174983768e-06,
+      "loss": 0.0653,
+      "step": 1477
+    },
+    {
+      "epoch": 0.33590909090909093,
+      "grad_norm": 1.4089506431322627,
+      "learning_rate": 1.2361498388560859e-06,
+      "loss": 0.0504,
+      "step": 1478
+    },
+    {
+      "epoch": 0.3361363636363636,
+      "grad_norm": 3.6347885256291725,
+      "learning_rate": 1.2361311477513832e-06,
+      "loss": 0.2287,
+      "step": 1479
+    },
+    {
+      "epoch": 0.33636363636363636,
+      "grad_norm": 1.5046679441259465,
+      "learning_rate": 1.2361124441846506e-06,
+      "loss": 0.0574,
+      "step": 1480
+    },
+    {
+      "epoch": 0.3365909090909091,
+      "grad_norm": 2.927118052243882,
+      "learning_rate": 1.236093728156269e-06,
+      "loss": 0.1111,
+      "step": 1481
+    },
+    {
+      "epoch": 0.3368181818181818,
+      "grad_norm": 2.548451490519401,
+      "learning_rate": 1.23607499966662e-06,
+      "loss": 0.071,
+      "step": 1482
+    },
+    {
+      "epoch": 0.33704545454545454,
+      "grad_norm": 1.4721102366975014,
+      "learning_rate": 1.236056258716086e-06,
+      "loss": 0.0863,
+      "step": 1483
+    },
+    {
+      "epoch": 0.3372727272727273,
+      "grad_norm": 2.2936227555261417,
+      "learning_rate": 1.2360375053050486e-06,
+      "loss": 0.0764,
+      "step": 1484
+    },
+    {
+      "epoch": 0.3375,
+      "grad_norm": 1.6794639389114008,
+      "learning_rate": 1.2360187394338905e-06,
+      "loss": 0.1201,
+      "step": 1485
+    },
+    {
+      "epoch": 0.3377272727272727,
+      "grad_norm": 1.3184409132960824,
+      "learning_rate": 1.2359999611029944e-06,
+      "loss": 0.0797,
+      "step": 1486
+    },
+    {
+      "epoch": 0.33795454545454545,
+      "grad_norm": 1.58139173280543,
+      "learning_rate": 1.235981170312743e-06,
+      "loss": 0.0881,
+      "step": 1487
+    },
+    {
+      "epoch": 0.3381818181818182,
+      "grad_norm": 1.5216840168574433,
+      "learning_rate": 1.2359623670635195e-06,
+      "loss": 0.0756,
+      "step": 1488
+    },
+    {
+      "epoch": 0.3384090909090909,
+      "grad_norm": 1.2762699168318283,
+      "learning_rate": 1.2359435513557077e-06,
+      "loss": 0.0392,
+      "step": 1489
+    },
+    {
+      "epoch": 0.3386363636363636,
+      "grad_norm": 1.6081149931856848,
+      "learning_rate": 1.2359247231896912e-06,
+      "loss": 0.0772,
+      "step": 1490
+    },
+    {
+      "epoch": 0.33886363636363637,
+      "grad_norm": 1.256955636089201,
+      "learning_rate": 1.2359058825658534e-06,
+      "loss": 0.0641,
+      "step": 1491
+    },
+    {
+      "epoch": 0.3390909090909091,
+      "grad_norm": 2.5752194629060665,
+      "learning_rate": 1.2358870294845792e-06,
+      "loss": 0.1385,
+      "step": 1492
+    },
+    {
+      "epoch": 0.3393181818181818,
+      "grad_norm": 1.5491358792566432,
+      "learning_rate": 1.2358681639462524e-06,
+      "loss": 0.0839,
+      "step": 1493
+    },
+    {
+      "epoch": 0.33954545454545454,
+      "grad_norm": 2.407550690663252,
+      "learning_rate": 1.2358492859512583e-06,
+      "loss": 0.0975,
+      "step": 1494
+    },
+    {
+      "epoch": 0.3397727272727273,
+      "grad_norm": 1.5573829061437918,
+      "learning_rate": 1.2358303954999814e-06,
+      "loss": 0.1005,
+      "step": 1495
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3501566839549228,
+      "learning_rate": 1.2358114925928073e-06,
+      "loss": 0.0523,
+      "step": 1496
+    },
+    {
+      "epoch": 0.3402272727272727,
+      "grad_norm": 1.786935100934286,
+      "learning_rate": 1.235792577230121e-06,
+      "loss": 0.0479,
+      "step": 1497
+    },
+    {
+      "epoch": 0.34045454545454545,
+      "grad_norm": 2.9393581833217723,
+      "learning_rate": 1.2357736494123084e-06,
+      "loss": 0.1912,
+      "step": 1498
+    },
+    {
+      "epoch": 0.3406818181818182,
+      "grad_norm": 2.691010123156203,
+      "learning_rate": 1.235754709139756e-06,
+      "loss": 0.1356,
+      "step": 1499
+    },
+    {
+      "epoch": 0.3409090909090909,
+      "grad_norm": 0.3828531891963392,
+      "learning_rate": 1.2357357564128491e-06,
+      "loss": 0.0137,
+      "step": 1500
+    },
+    {
+      "epoch": 0.3411363636363636,
+      "grad_norm": 0.4901969639060127,
+      "learning_rate": 1.2357167912319747e-06,
+      "loss": 0.0278,
+      "step": 1501
+    },
+    {
+      "epoch": 0.34136363636363637,
+      "grad_norm": 1.976555091417265,
+      "learning_rate": 1.2356978135975196e-06,
+      "loss": 0.0985,
+      "step": 1502
+    },
+    {
+      "epoch": 0.3415909090909091,
+      "grad_norm": 1.8429541876773141,
+      "learning_rate": 1.2356788235098707e-06,
+      "loss": 0.0616,
+      "step": 1503
+    },
+    {
+      "epoch": 0.3418181818181818,
+      "grad_norm": 2.281600161531977,
+      "learning_rate": 1.235659820969415e-06,
+      "loss": 0.0811,
+      "step": 1504
+    },
+    {
+      "epoch": 0.34204545454545454,
+      "grad_norm": 1.2396449461248211,
+      "learning_rate": 1.2356408059765404e-06,
+      "loss": 0.0618,
+      "step": 1505
+    },
+    {
+      "epoch": 0.3422727272727273,
+      "grad_norm": 1.705787319904896,
+      "learning_rate": 1.2356217785316344e-06,
+      "loss": 0.0482,
+      "step": 1506
+    },
+    {
+      "epoch": 0.3425,
+      "grad_norm": 1.9670996505359966,
+      "learning_rate": 1.235602738635085e-06,
+      "loss": 0.1088,
+      "step": 1507
+    },
+    {
+      "epoch": 0.3427272727272727,
+      "grad_norm": 0.8975041046659857,
+      "learning_rate": 1.2355836862872805e-06,
+      "loss": 0.0406,
+      "step": 1508
+    },
+    {
+      "epoch": 0.34295454545454546,
+      "grad_norm": 1.293796439220399,
+      "learning_rate": 1.2355646214886094e-06,
+      "loss": 0.0519,
+      "step": 1509
+    },
+    {
+      "epoch": 0.3431818181818182,
+      "grad_norm": 1.8867798282734134,
+      "learning_rate": 1.2355455442394606e-06,
+      "loss": 0.1079,
+      "step": 1510
+    },
+    {
+      "epoch": 0.3434090909090909,
+      "grad_norm": 1.7822599828110839,
+      "learning_rate": 1.235526454540223e-06,
+      "loss": 0.0984,
+      "step": 1511
+    },
+    {
+      "epoch": 0.34363636363636363,
+      "grad_norm": 1.6709368977593508,
+      "learning_rate": 1.2355073523912857e-06,
+      "loss": 0.0696,
+      "step": 1512
+    },
+    {
+      "epoch": 0.34386363636363637,
+      "grad_norm": 2.7059454155437304,
+      "learning_rate": 1.2354882377930385e-06,
+      "loss": 0.0951,
+      "step": 1513
+    },
+    {
+      "epoch": 0.3440909090909091,
+      "grad_norm": 1.6577456623507638,
+      "learning_rate": 1.235469110745871e-06,
+      "loss": 0.0485,
+      "step": 1514
+    },
+    {
+      "epoch": 0.3443181818181818,
+      "grad_norm": 3.4233599763920943,
+      "learning_rate": 1.2354499712501733e-06,
+      "loss": 0.1458,
+      "step": 1515
+    },
+    {
+      "epoch": 0.34454545454545454,
+      "grad_norm": 1.437203281572912,
+      "learning_rate": 1.2354308193063358e-06,
+      "loss": 0.0836,
+      "step": 1516
+    },
+    {
+      "epoch": 0.3447727272727273,
+      "grad_norm": 2.4093422312627846,
+      "learning_rate": 1.235411654914749e-06,
+      "loss": 0.0902,
+      "step": 1517
+    },
+    {
+      "epoch": 0.345,
+      "grad_norm": 1.6125191292846897,
+      "learning_rate": 1.2353924780758034e-06,
+      "loss": 0.0516,
+      "step": 1518
+    },
+    {
+      "epoch": 0.3452272727272727,
+      "grad_norm": 1.7727560894835672,
+      "learning_rate": 1.2353732887898906e-06,
+      "loss": 0.0879,
+      "step": 1519
+    },
+    {
+      "epoch": 0.34545454545454546,
+      "grad_norm": 2.2131177070478563,
+      "learning_rate": 1.2353540870574014e-06,
+      "loss": 0.089,
+      "step": 1520
+    },
+    {
+      "epoch": 0.3456818181818182,
+      "grad_norm": 1.788097408866385,
+      "learning_rate": 1.2353348728787274e-06,
+      "loss": 0.0783,
+      "step": 1521
+    },
+    {
+      "epoch": 0.3459090909090909,
+      "grad_norm": 2.251462297712363,
+      "learning_rate": 1.2353156462542607e-06,
+      "loss": 0.0795,
+      "step": 1522
+    },
+    {
+      "epoch": 0.34613636363636363,
+      "grad_norm": 1.4821636144471846,
+      "learning_rate": 1.2352964071843932e-06,
+      "loss": 0.0592,
+      "step": 1523
+    },
+    {
+      "epoch": 0.3463636363636364,
+      "grad_norm": 2.400446497310218,
+      "learning_rate": 1.2352771556695174e-06,
+      "loss": 0.0987,
+      "step": 1524
+    },
+    {
+      "epoch": 0.3465909090909091,
+      "grad_norm": 2.8384047077280963,
+      "learning_rate": 1.2352578917100255e-06,
+      "loss": 0.1266,
+      "step": 1525
+    },
+    {
+      "epoch": 0.3468181818181818,
+      "grad_norm": 1.5827297669749045,
+      "learning_rate": 1.2352386153063107e-06,
+      "loss": 0.0864,
+      "step": 1526
+    },
+    {
+      "epoch": 0.34704545454545455,
+      "grad_norm": 1.412803637193178,
+      "learning_rate": 1.2352193264587658e-06,
+      "loss": 0.0367,
+      "step": 1527
+    },
+    {
+      "epoch": 0.3472727272727273,
+      "grad_norm": 2.136236011364923,
+      "learning_rate": 1.2352000251677843e-06,
+      "loss": 0.098,
+      "step": 1528
+    },
+    {
+      "epoch": 0.3475,
+      "grad_norm": 2.2506267859647613,
+      "learning_rate": 1.2351807114337595e-06,
+      "loss": 0.0639,
+      "step": 1529
+    },
+    {
+      "epoch": 0.3477272727272727,
+      "grad_norm": 2.1410721350166915,
+      "learning_rate": 1.2351613852570858e-06,
+      "loss": 0.0852,
+      "step": 1530
+    },
+    {
+      "epoch": 0.34795454545454546,
+      "grad_norm": 1.4980009330506854,
+      "learning_rate": 1.2351420466381566e-06,
+      "loss": 0.0833,
+      "step": 1531
+    },
+    {
+      "epoch": 0.3481818181818182,
+      "grad_norm": 1.5539135528843528,
+      "learning_rate": 1.235122695577367e-06,
+      "loss": 0.0886,
+      "step": 1532
+    },
+    {
+      "epoch": 0.3484090909090909,
+      "grad_norm": 2.9202716524292365,
+      "learning_rate": 1.235103332075111e-06,
+      "loss": 0.1108,
+      "step": 1533
+    },
+    {
+      "epoch": 0.34863636363636363,
+      "grad_norm": 2.429151581776052,
+      "learning_rate": 1.2350839561317838e-06,
+      "loss": 0.0868,
+      "step": 1534
+    },
+    {
+      "epoch": 0.3488636363636364,
+      "grad_norm": 2.1625593764047952,
+      "learning_rate": 1.2350645677477805e-06,
+      "loss": 0.0983,
+      "step": 1535
+    },
+    {
+      "epoch": 0.3490909090909091,
+      "grad_norm": 1.6649843739223296,
+      "learning_rate": 1.235045166923496e-06,
+      "loss": 0.0617,
+      "step": 1536
+    },
+    {
+      "epoch": 0.3493181818181818,
+      "grad_norm": 3.348656956671627,
+      "learning_rate": 1.2350257536593266e-06,
+      "loss": 0.0978,
+      "step": 1537
+    },
+    {
+      "epoch": 0.34954545454545455,
+      "grad_norm": 2.0026162180973333,
+      "learning_rate": 1.2350063279556677e-06,
+      "loss": 0.0721,
+      "step": 1538
+    },
+    {
+      "epoch": 0.3497727272727273,
+      "grad_norm": 2.097758435406637,
+      "learning_rate": 1.2349868898129155e-06,
+      "loss": 0.0787,
+      "step": 1539
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.968043077236136,
+      "learning_rate": 1.2349674392314666e-06,
+      "loss": 0.1234,
+      "step": 1540
+    },
+    {
+      "epoch": 0.3502272727272727,
+      "grad_norm": 1.3973203285321703,
+      "learning_rate": 1.2349479762117171e-06,
+      "loss": 0.0946,
+      "step": 1541
+    },
+    {
+      "epoch": 0.35045454545454546,
+      "grad_norm": 1.4283895433597191,
+      "learning_rate": 1.2349285007540646e-06,
+      "loss": 0.046,
+      "step": 1542
+    },
+    {
+      "epoch": 0.3506818181818182,
+      "grad_norm": 1.9572114362145603,
+      "learning_rate": 1.234909012858906e-06,
+      "loss": 0.098,
+      "step": 1543
+    },
+    {
+      "epoch": 0.3509090909090909,
+      "grad_norm": 2.439199933953352,
+      "learning_rate": 1.2348895125266384e-06,
+      "loss": 0.1015,
+      "step": 1544
+    },
+    {
+      "epoch": 0.35113636363636364,
+      "grad_norm": 1.3635786268942887,
+      "learning_rate": 1.2348699997576593e-06,
+      "loss": 0.076,
+      "step": 1545
+    },
+    {
+      "epoch": 0.3513636363636364,
+      "grad_norm": 1.7792702781144936,
+      "learning_rate": 1.2348504745523673e-06,
+      "loss": 0.1006,
+      "step": 1546
+    },
+    {
+      "epoch": 0.35159090909090907,
+      "grad_norm": 3.44408931136234,
+      "learning_rate": 1.2348309369111603e-06,
+      "loss": 0.1574,
+      "step": 1547
+    },
+    {
+      "epoch": 0.3518181818181818,
+      "grad_norm": 1.9334475482073272,
+      "learning_rate": 1.2348113868344363e-06,
+      "loss": 0.0535,
+      "step": 1548
+    },
+    {
+      "epoch": 0.35204545454545455,
+      "grad_norm": 2.198506914938487,
+      "learning_rate": 1.2347918243225942e-06,
+      "loss": 0.1183,
+      "step": 1549
+    },
+    {
+      "epoch": 0.3522727272727273,
+      "grad_norm": 2.54555493424984,
+      "learning_rate": 1.2347722493760331e-06,
+      "loss": 0.0946,
+      "step": 1550
+    },
+    {
+      "epoch": 0.3525,
+      "grad_norm": 1.6633456309607666,
+      "learning_rate": 1.2347526619951523e-06,
+      "loss": 0.0623,
+      "step": 1551
+    },
+    {
+      "epoch": 0.3527272727272727,
+      "grad_norm": 2.690164176462338,
+      "learning_rate": 1.2347330621803506e-06,
+      "loss": 0.1411,
+      "step": 1552
+    },
+    {
+      "epoch": 0.35295454545454547,
+      "grad_norm": 2.1073321027449,
+      "learning_rate": 1.234713449932028e-06,
+      "loss": 0.1181,
+      "step": 1553
+    },
+    {
+      "epoch": 0.3531818181818182,
+      "grad_norm": 2.730231311588262,
+      "learning_rate": 1.2346938252505847e-06,
+      "loss": 0.0965,
+      "step": 1554
+    },
+    {
+      "epoch": 0.3534090909090909,
+      "grad_norm": 2.362691138084924,
+      "learning_rate": 1.2346741881364206e-06,
+      "loss": 0.0806,
+      "step": 1555
+    },
+    {
+      "epoch": 0.35363636363636364,
+      "grad_norm": 1.8796434357689482,
+      "learning_rate": 1.2346545385899358e-06,
+      "loss": 0.0926,
+      "step": 1556
+    },
+    {
+      "epoch": 0.3538636363636364,
+      "grad_norm": 2.1900000300983637,
+      "learning_rate": 1.2346348766115318e-06,
+      "loss": 0.0927,
+      "step": 1557
+    },
+    {
+      "epoch": 0.35409090909090907,
+      "grad_norm": 1.773001036480335,
+      "learning_rate": 1.234615202201609e-06,
+      "loss": 0.1057,
+      "step": 1558
+    },
+    {
+      "epoch": 0.3543181818181818,
+      "grad_norm": 1.5793813912034906,
+      "learning_rate": 1.2345955153605685e-06,
+      "loss": 0.0809,
+      "step": 1559
+    },
+    {
+      "epoch": 0.35454545454545455,
+      "grad_norm": 1.0593616489604663,
+      "learning_rate": 1.2345758160888122e-06,
+      "loss": 0.0533,
+      "step": 1560
+    },
+    {
+      "epoch": 0.3547727272727273,
+      "grad_norm": 2.1576330234778363,
+      "learning_rate": 1.2345561043867413e-06,
+      "loss": 0.0867,
+      "step": 1561
+    },
+    {
+      "epoch": 0.355,
+      "grad_norm": 1.5365112788567976,
+      "learning_rate": 1.2345363802547581e-06,
+      "loss": 0.0591,
+      "step": 1562
+    },
+    {
+      "epoch": 0.3552272727272727,
+      "grad_norm": 0.8420687684452651,
+      "learning_rate": 1.2345166436932646e-06,
+      "loss": 0.0243,
+      "step": 1563
+    },
+    {
+      "epoch": 0.35545454545454547,
+      "grad_norm": 2.2010806869254322,
+      "learning_rate": 1.2344968947026635e-06,
+      "loss": 0.0835,
+      "step": 1564
+    },
+    {
+      "epoch": 0.35568181818181815,
+      "grad_norm": 2.1938423947725507,
+      "learning_rate": 1.2344771332833575e-06,
+      "loss": 0.1746,
+      "step": 1565
+    },
+    {
+      "epoch": 0.3559090909090909,
+      "grad_norm": 2.1197889309373323,
+      "learning_rate": 1.2344573594357493e-06,
+      "loss": 0.0979,
+      "step": 1566
+    },
+    {
+      "epoch": 0.35613636363636364,
+      "grad_norm": 2.184392072666271,
+      "learning_rate": 1.2344375731602426e-06,
+      "loss": 0.0976,
+      "step": 1567
+    },
+    {
+      "epoch": 0.3563636363636364,
+      "grad_norm": 1.2919427298618142,
+      "learning_rate": 1.2344177744572401e-06,
+      "loss": 0.0716,
+      "step": 1568
+    },
+    {
+      "epoch": 0.35659090909090907,
+      "grad_norm": 1.7778246564104034,
+      "learning_rate": 1.2343979633271463e-06,
+      "loss": 0.0502,
+      "step": 1569
+    },
+    {
+      "epoch": 0.3568181818181818,
+      "grad_norm": 1.9335153972006138,
+      "learning_rate": 1.2343781397703646e-06,
+      "loss": 0.0705,
+      "step": 1570
+    },
+    {
+      "epoch": 0.35704545454545455,
+      "grad_norm": 1.6853924116213475,
+      "learning_rate": 1.2343583037872998e-06,
+      "loss": 0.0625,
+      "step": 1571
+    },
+    {
+      "epoch": 0.3572727272727273,
+      "grad_norm": 2.091663508063597,
+      "learning_rate": 1.234338455378356e-06,
+      "loss": 0.1112,
+      "step": 1572
+    },
+    {
+      "epoch": 0.3575,
+      "grad_norm": 2.077620044349402,
+      "learning_rate": 1.234318594543938e-06,
+      "loss": 0.1254,
+      "step": 1573
+    },
+    {
+      "epoch": 0.3577272727272727,
+      "grad_norm": 2.2748395636685745,
+      "learning_rate": 1.234298721284451e-06,
+      "loss": 0.0764,
+      "step": 1574
+    },
+    {
+      "epoch": 0.35795454545454547,
+      "grad_norm": 2.185044681271537,
+      "learning_rate": 1.2342788356002998e-06,
+      "loss": 0.0937,
+      "step": 1575
+    },
+    {
+      "epoch": 0.35818181818181816,
+      "grad_norm": 2.466049016622862,
+      "learning_rate": 1.2342589374918905e-06,
+      "loss": 0.1089,
+      "step": 1576
+    },
+    {
+      "epoch": 0.3584090909090909,
+      "grad_norm": 5.593265308440061,
+      "learning_rate": 1.2342390269596283e-06,
+      "loss": 0.2187,
+      "step": 1577
+    },
+    {
+      "epoch": 0.35863636363636364,
+      "grad_norm": 1.570392173516243,
+      "learning_rate": 1.2342191040039197e-06,
+      "loss": 0.0786,
+      "step": 1578
+    },
+    {
+      "epoch": 0.3588636363636364,
+      "grad_norm": 1.332549086153333,
+      "learning_rate": 1.2341991686251706e-06,
+      "loss": 0.0489,
+      "step": 1579
+    },
+    {
+      "epoch": 0.35909090909090907,
+      "grad_norm": 1.00737360254056,
+      "learning_rate": 1.2341792208237876e-06,
+      "loss": 0.0727,
+      "step": 1580
+    },
+    {
+      "epoch": 0.3593181818181818,
+      "grad_norm": 2.5760629375112494,
+      "learning_rate": 1.2341592606001777e-06,
+      "loss": 0.0854,
+      "step": 1581
+    },
+    {
+      "epoch": 0.35954545454545456,
+      "grad_norm": 3.6279073840392315,
+      "learning_rate": 1.2341392879547474e-06,
+      "loss": 0.1788,
+      "step": 1582
+    },
+    {
+      "epoch": 0.3597727272727273,
+      "grad_norm": 1.4747862671984526,
+      "learning_rate": 1.2341193028879045e-06,
+      "loss": 0.0554,
+      "step": 1583
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6379645365662783,
+      "learning_rate": 1.2340993054000563e-06,
+      "loss": 0.1054,
+      "step": 1584
+    },
+    {
+      "epoch": 0.36022727272727273,
+      "grad_norm": 1.6765548557240038,
+      "learning_rate": 1.2340792954916109e-06,
+      "loss": 0.0956,
+      "step": 1585
+    },
+    {
+      "epoch": 0.36045454545454547,
+      "grad_norm": 1.3711301521934722,
+      "learning_rate": 1.2340592731629758e-06,
+      "loss": 0.1109,
+      "step": 1586
+    },
+    {
+      "epoch": 0.36068181818181816,
+      "grad_norm": 1.657596680702226,
+      "learning_rate": 1.2340392384145597e-06,
+      "loss": 0.0616,
+      "step": 1587
+    },
+    {
+      "epoch": 0.3609090909090909,
+      "grad_norm": 2.057298547365089,
+      "learning_rate": 1.2340191912467708e-06,
+      "loss": 0.0781,
+      "step": 1588
+    },
+    {
+      "epoch": 0.36113636363636364,
+      "grad_norm": 1.4258700201295547,
+      "learning_rate": 1.2339991316600183e-06,
+      "loss": 0.0689,
+      "step": 1589
+    },
+    {
+      "epoch": 0.3613636363636364,
+      "grad_norm": 1.7382596237999164,
+      "learning_rate": 1.233979059654711e-06,
+      "loss": 0.0826,
+      "step": 1590
+    },
+    {
+      "epoch": 0.3615909090909091,
+      "grad_norm": 1.5263352218325528,
+      "learning_rate": 1.2339589752312581e-06,
+      "loss": 0.0824,
+      "step": 1591
+    },
+    {
+      "epoch": 0.3618181818181818,
+      "grad_norm": 2.1444822885772092,
+      "learning_rate": 1.2339388783900694e-06,
+      "loss": 0.0825,
+      "step": 1592
+    },
+    {
+      "epoch": 0.36204545454545456,
+      "grad_norm": 1.1479906245701728,
+      "learning_rate": 1.2339187691315548e-06,
+      "loss": 0.0359,
+      "step": 1593
+    },
+    {
+      "epoch": 0.36227272727272725,
+      "grad_norm": 1.958124080122293,
+      "learning_rate": 1.233898647456124e-06,
+      "loss": 0.1088,
+      "step": 1594
+    },
+    {
+      "epoch": 0.3625,
+      "grad_norm": 2.326412823458403,
+      "learning_rate": 1.2338785133641876e-06,
+      "loss": 0.1015,
+      "step": 1595
+    },
+    {
+      "epoch": 0.36272727272727273,
+      "grad_norm": 2.231252330849337,
+      "learning_rate": 1.233858366856156e-06,
+      "loss": 0.1489,
+      "step": 1596
+    },
+    {
+      "epoch": 0.3629545454545455,
+      "grad_norm": 2.262765944360219,
+      "learning_rate": 1.23383820793244e-06,
+      "loss": 0.1247,
+      "step": 1597
+    },
+    {
+      "epoch": 0.36318181818181816,
+      "grad_norm": 2.7559783888132294,
+      "learning_rate": 1.233818036593451e-06,
+      "loss": 0.181,
+      "step": 1598
+    },
+    {
+      "epoch": 0.3634090909090909,
+      "grad_norm": 2.1276689611205226,
+      "learning_rate": 1.2337978528395998e-06,
+      "loss": 0.1224,
+      "step": 1599
+    },
+    {
+      "epoch": 0.36363636363636365,
+      "grad_norm": 1.2420986452633551,
+      "learning_rate": 1.2337776566712984e-06,
+      "loss": 0.0441,
+      "step": 1600
+    },
+    {
+      "epoch": 0.3638636363636364,
+      "grad_norm": 2.2205748138748347,
+      "learning_rate": 1.2337574480889585e-06,
+      "loss": 0.0848,
+      "step": 1601
+    },
+    {
+      "epoch": 0.3640909090909091,
+      "grad_norm": 2.110975545229239,
+      "learning_rate": 1.2337372270929923e-06,
+      "loss": 0.1374,
+      "step": 1602
+    },
+    {
+      "epoch": 0.3643181818181818,
+      "grad_norm": 2.1761659887985165,
+      "learning_rate": 1.2337169936838119e-06,
+      "loss": 0.0942,
+      "step": 1603
+    },
+    {
+      "epoch": 0.36454545454545456,
+      "grad_norm": 1.571662197097873,
+      "learning_rate": 1.23369674786183e-06,
+      "loss": 0.0763,
+      "step": 1604
+    },
+    {
+      "epoch": 0.36477272727272725,
+      "grad_norm": 2.1558186122062755,
+      "learning_rate": 1.2336764896274597e-06,
+      "loss": 0.0934,
+      "step": 1605
+    },
+    {
+      "epoch": 0.365,
+      "grad_norm": 1.8115612786120774,
+      "learning_rate": 1.2336562189811138e-06,
+      "loss": 0.1181,
+      "step": 1606
+    },
+    {
+      "epoch": 0.36522727272727273,
+      "grad_norm": 1.3125131551657305,
+      "learning_rate": 1.2336359359232055e-06,
+      "loss": 0.053,
+      "step": 1607
+    },
+    {
+      "epoch": 0.3654545454545455,
+      "grad_norm": 2.409012979551126,
+      "learning_rate": 1.233615640454149e-06,
+      "loss": 0.0684,
+      "step": 1608
+    },
+    {
+      "epoch": 0.36568181818181816,
+      "grad_norm": 2.5343212353122557,
+      "learning_rate": 1.2335953325743577e-06,
+      "loss": 0.1693,
+      "step": 1609
+    },
+    {
+      "epoch": 0.3659090909090909,
+      "grad_norm": 1.1674467453521455,
+      "learning_rate": 1.2335750122842458e-06,
+      "loss": 0.0571,
+      "step": 1610
+    },
+    {
+      "epoch": 0.36613636363636365,
+      "grad_norm": 3.1541085671465763,
+      "learning_rate": 1.2335546795842276e-06,
+      "loss": 0.1438,
+      "step": 1611
+    },
+    {
+      "epoch": 0.3663636363636364,
+      "grad_norm": 2.3211985288782797,
+      "learning_rate": 1.2335343344747177e-06,
+      "loss": 0.1488,
+      "step": 1612
+    },
+    {
+      "epoch": 0.3665909090909091,
+      "grad_norm": 2.400598012082991,
+      "learning_rate": 1.2335139769561314e-06,
+      "loss": 0.0958,
+      "step": 1613
+    },
+    {
+      "epoch": 0.3668181818181818,
+      "grad_norm": 1.1864507182973172,
+      "learning_rate": 1.2334936070288831e-06,
+      "loss": 0.063,
+      "step": 1614
+    },
+    {
+      "epoch": 0.36704545454545456,
+      "grad_norm": 1.4679678292469343,
+      "learning_rate": 1.233473224693389e-06,
+      "loss": 0.0776,
+      "step": 1615
+    },
+    {
+      "epoch": 0.36727272727272725,
+      "grad_norm": 1.7036552627204715,
+      "learning_rate": 1.233452829950064e-06,
+      "loss": 0.0455,
+      "step": 1616
+    },
+    {
+      "epoch": 0.3675,
+      "grad_norm": 1.9816630172793415,
+      "learning_rate": 1.2334324227993245e-06,
+      "loss": 0.0896,
+      "step": 1617
+    },
+    {
+      "epoch": 0.36772727272727274,
+      "grad_norm": 2.909570249786732,
+      "learning_rate": 1.233412003241586e-06,
+      "loss": 0.104,
+      "step": 1618
+    },
+    {
+      "epoch": 0.3679545454545455,
+      "grad_norm": 1.5257760912299119,
+      "learning_rate": 1.233391571277266e-06,
+      "loss": 0.0923,
+      "step": 1619
+    },
+    {
+      "epoch": 0.36818181818181817,
+      "grad_norm": 1.370181529675218,
+      "learning_rate": 1.2333711269067799e-06,
+      "loss": 0.0553,
+      "step": 1620
+    },
+    {
+      "epoch": 0.3684090909090909,
+      "grad_norm": 1.8105613815152266,
+      "learning_rate": 1.2333506701305453e-06,
+      "loss": 0.0789,
+      "step": 1621
+    },
+    {
+      "epoch": 0.36863636363636365,
+      "grad_norm": 2.1400549139602276,
+      "learning_rate": 1.2333302009489795e-06,
+      "loss": 0.0973,
+      "step": 1622
+    },
+    {
+      "epoch": 0.36886363636363634,
+      "grad_norm": 3.03758652011928,
+      "learning_rate": 1.2333097193624993e-06,
+      "loss": 0.1018,
+      "step": 1623
+    },
+    {
+      "epoch": 0.3690909090909091,
+      "grad_norm": 1.903064750258247,
+      "learning_rate": 1.233289225371523e-06,
+      "loss": 0.1043,
+      "step": 1624
+    },
+    {
+      "epoch": 0.3693181818181818,
+      "grad_norm": 2.4126646303869643,
+      "learning_rate": 1.2332687189764678e-06,
+      "loss": 0.091,
+      "step": 1625
+    },
+    {
+      "epoch": 0.36954545454545457,
+      "grad_norm": 1.5966076942637795,
+      "learning_rate": 1.2332482001777522e-06,
+      "loss": 0.1197,
+      "step": 1626
+    },
+    {
+      "epoch": 0.36977272727272725,
+      "grad_norm": 2.2707662820356953,
+      "learning_rate": 1.233227668975795e-06,
+      "loss": 0.1067,
+      "step": 1627
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.042977149010959,
+      "learning_rate": 1.2332071253710142e-06,
+      "loss": 0.0853,
+      "step": 1628
+    },
+    {
+      "epoch": 0.37022727272727274,
+      "grad_norm": 1.3099138121649299,
+      "learning_rate": 1.2331865693638292e-06,
+      "loss": 0.0821,
+      "step": 1629
+    },
+    {
+      "epoch": 0.3704545454545455,
+      "grad_norm": 1.8093111686923244,
+      "learning_rate": 1.2331660009546589e-06,
+      "loss": 0.1188,
+      "step": 1630
+    },
+    {
+      "epoch": 0.37068181818181817,
+      "grad_norm": 2.1179329031736143,
+      "learning_rate": 1.233145420143923e-06,
+      "loss": 0.0791,
+      "step": 1631
+    },
+    {
+      "epoch": 0.3709090909090909,
+      "grad_norm": 1.9796793707680245,
+      "learning_rate": 1.233124826932041e-06,
+      "loss": 0.0716,
+      "step": 1632
+    },
+    {
+      "epoch": 0.37113636363636365,
+      "grad_norm": 2.487705290817736,
+      "learning_rate": 1.2331042213194327e-06,
+      "loss": 0.091,
+      "step": 1633
+    },
+    {
+      "epoch": 0.37136363636363634,
+      "grad_norm": 1.8219412231735543,
+      "learning_rate": 1.2330836033065186e-06,
+      "loss": 0.0551,
+      "step": 1634
+    },
+    {
+      "epoch": 0.3715909090909091,
+      "grad_norm": 1.6590704434902679,
+      "learning_rate": 1.2330629728937187e-06,
+      "loss": 0.0898,
+      "step": 1635
+    },
+    {
+      "epoch": 0.3718181818181818,
+      "grad_norm": 2.267801530580309,
+      "learning_rate": 1.2330423300814542e-06,
+      "loss": 0.1066,
+      "step": 1636
+    },
+    {
+      "epoch": 0.37204545454545457,
+      "grad_norm": 2.4601246148035796,
+      "learning_rate": 1.2330216748701459e-06,
+      "loss": 0.1016,
+      "step": 1637
+    },
+    {
+      "epoch": 0.37227272727272726,
+      "grad_norm": 1.719434560567821,
+      "learning_rate": 1.2330010072602147e-06,
+      "loss": 0.0801,
+      "step": 1638
+    },
+    {
+      "epoch": 0.3725,
+      "grad_norm": 2.5923715115940733,
+      "learning_rate": 1.2329803272520823e-06,
+      "loss": 0.1074,
+      "step": 1639
+    },
+    {
+      "epoch": 0.37272727272727274,
+      "grad_norm": 2.5543370909695153,
+      "learning_rate": 1.2329596348461702e-06,
+      "loss": 0.1189,
+      "step": 1640
+    },
+    {
+      "epoch": 0.3729545454545454,
+      "grad_norm": 1.637408439199484,
+      "learning_rate": 1.2329389300429008e-06,
+      "loss": 0.0929,
+      "step": 1641
+    },
+    {
+      "epoch": 0.37318181818181817,
+      "grad_norm": 1.3732993851405224,
+      "learning_rate": 1.2329182128426958e-06,
+      "loss": 0.0596,
+      "step": 1642
+    },
+    {
+      "epoch": 0.3734090909090909,
+      "grad_norm": 2.87859584590602,
+      "learning_rate": 1.232897483245978e-06,
+      "loss": 0.1343,
+      "step": 1643
+    },
+    {
+      "epoch": 0.37363636363636366,
+      "grad_norm": 1.4195909706119643,
+      "learning_rate": 1.2328767412531698e-06,
+      "loss": 0.0712,
+      "step": 1644
+    },
+    {
+      "epoch": 0.37386363636363634,
+      "grad_norm": 1.5373828821559554,
+      "learning_rate": 1.2328559868646945e-06,
+      "loss": 0.0498,
+      "step": 1645
+    },
+    {
+      "epoch": 0.3740909090909091,
+      "grad_norm": 1.1782776069395948,
+      "learning_rate": 1.232835220080975e-06,
+      "loss": 0.0701,
+      "step": 1646
+    },
+    {
+      "epoch": 0.3743181818181818,
+      "grad_norm": 2.012889798281113,
+      "learning_rate": 1.232814440902435e-06,
+      "loss": 0.0898,
+      "step": 1647
+    },
+    {
+      "epoch": 0.37454545454545457,
+      "grad_norm": 3.0001550107430326,
+      "learning_rate": 1.2327936493294984e-06,
+      "loss": 0.0981,
+      "step": 1648
+    },
+    {
+      "epoch": 0.37477272727272726,
+      "grad_norm": 1.8987268442119174,
+      "learning_rate": 1.2327728453625887e-06,
+      "loss": 0.0667,
+      "step": 1649
+    },
+    {
+      "epoch": 0.375,
+      "grad_norm": 1.801799775116866,
+      "learning_rate": 1.2327520290021303e-06,
+      "loss": 0.0862,
+      "step": 1650
+    },
+    {
+      "epoch": 0.37522727272727274,
+      "grad_norm": 1.888236850759012,
+      "learning_rate": 1.232731200248548e-06,
+      "loss": 0.0788,
+      "step": 1651
+    },
+    {
+      "epoch": 0.37545454545454543,
+      "grad_norm": 2.2369131835856493,
+      "learning_rate": 1.2327103591022662e-06,
+      "loss": 0.0692,
+      "step": 1652
+    },
+    {
+      "epoch": 0.3756818181818182,
+      "grad_norm": 1.4625149520936138,
+      "learning_rate": 1.2326895055637098e-06,
+      "loss": 0.0717,
+      "step": 1653
+    },
+    {
+      "epoch": 0.3759090909090909,
+      "grad_norm": 1.4383899113667753,
+      "learning_rate": 1.2326686396333042e-06,
+      "loss": 0.084,
+      "step": 1654
+    },
+    {
+      "epoch": 0.37613636363636366,
+      "grad_norm": 1.3557193034970667,
+      "learning_rate": 1.2326477613114752e-06,
+      "loss": 0.0756,
+      "step": 1655
+    },
+    {
+      "epoch": 0.37636363636363634,
+      "grad_norm": 2.5289577513694295,
+      "learning_rate": 1.232626870598648e-06,
+      "loss": 0.1088,
+      "step": 1656
+    },
+    {
+      "epoch": 0.3765909090909091,
+      "grad_norm": 1.900911180519534,
+      "learning_rate": 1.232605967495249e-06,
+      "loss": 0.0747,
+      "step": 1657
+    },
+    {
+      "epoch": 0.37681818181818183,
+      "grad_norm": 2.3125506062645713,
+      "learning_rate": 1.232585052001704e-06,
+      "loss": 0.1651,
+      "step": 1658
+    },
+    {
+      "epoch": 0.3770454545454546,
+      "grad_norm": 1.2979612957454159,
+      "learning_rate": 1.2325641241184401e-06,
+      "loss": 0.0905,
+      "step": 1659
+    },
+    {
+      "epoch": 0.37727272727272726,
+      "grad_norm": 1.222397920073013,
+      "learning_rate": 1.2325431838458836e-06,
+      "loss": 0.0606,
+      "step": 1660
+    },
+    {
+      "epoch": 0.3775,
+      "grad_norm": 1.6903292554002185,
+      "learning_rate": 1.2325222311844617e-06,
+      "loss": 0.073,
+      "step": 1661
+    },
+    {
+      "epoch": 0.37772727272727274,
+      "grad_norm": 1.2552647377285655,
+      "learning_rate": 1.2325012661346018e-06,
+      "loss": 0.0458,
+      "step": 1662
+    },
+    {
+      "epoch": 0.37795454545454543,
+      "grad_norm": 1.3567541293851269,
+      "learning_rate": 1.232480288696731e-06,
+      "loss": 0.0505,
+      "step": 1663
+    },
+    {
+      "epoch": 0.3781818181818182,
+      "grad_norm": 1.8856650427709518,
+      "learning_rate": 1.2324592988712774e-06,
+      "loss": 0.0886,
+      "step": 1664
+    },
+    {
+      "epoch": 0.3784090909090909,
+      "grad_norm": 1.2997020009228193,
+      "learning_rate": 1.232438296658669e-06,
+      "loss": 0.0648,
+      "step": 1665
+    },
+    {
+      "epoch": 0.37863636363636366,
+      "grad_norm": 1.9594797042509664,
+      "learning_rate": 1.2324172820593339e-06,
+      "loss": 0.0927,
+      "step": 1666
+    },
+    {
+      "epoch": 0.37886363636363635,
+      "grad_norm": 1.6932859123287332,
+      "learning_rate": 1.2323962550737007e-06,
+      "loss": 0.0516,
+      "step": 1667
+    },
+    {
+      "epoch": 0.3790909090909091,
+      "grad_norm": 1.3818840893167152,
+      "learning_rate": 1.2323752157021982e-06,
+      "loss": 0.0793,
+      "step": 1668
+    },
+    {
+      "epoch": 0.37931818181818183,
+      "grad_norm": 1.6066660988616208,
+      "learning_rate": 1.2323541639452554e-06,
+      "loss": 0.0673,
+      "step": 1669
+    },
+    {
+      "epoch": 0.3795454545454545,
+      "grad_norm": 1.4637778283471876,
+      "learning_rate": 1.2323330998033016e-06,
+      "loss": 0.0759,
+      "step": 1670
+    },
+    {
+      "epoch": 0.37977272727272726,
+      "grad_norm": 2.748215600210496,
+      "learning_rate": 1.2323120232767667e-06,
+      "loss": 0.1166,
+      "step": 1671
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4495089059468085,
+      "learning_rate": 1.2322909343660798e-06,
+      "loss": 0.0757,
+      "step": 1672
+    },
+    {
+      "epoch": 0.38022727272727275,
+      "grad_norm": 2.492832078590292,
+      "learning_rate": 1.2322698330716713e-06,
+      "loss": 0.1225,
+      "step": 1673
+    },
+    {
+      "epoch": 0.38045454545454543,
+      "grad_norm": 1.644578846600651,
+      "learning_rate": 1.2322487193939716e-06,
+      "loss": 0.0647,
+      "step": 1674
+    },
+    {
+      "epoch": 0.3806818181818182,
+      "grad_norm": 2.2571236583598537,
+      "learning_rate": 1.2322275933334109e-06,
+      "loss": 0.0715,
+      "step": 1675
+    },
+    {
+      "epoch": 0.3809090909090909,
+      "grad_norm": 1.2472276576296193,
+      "learning_rate": 1.2322064548904202e-06,
+      "loss": 0.0587,
+      "step": 1676
+    },
+    {
+      "epoch": 0.38113636363636366,
+      "grad_norm": 4.161523599066941,
+      "learning_rate": 1.2321853040654308e-06,
+      "loss": 0.1298,
+      "step": 1677
+    },
+    {
+      "epoch": 0.38136363636363635,
+      "grad_norm": 2.952979132798579,
+      "learning_rate": 1.2321641408588736e-06,
+      "loss": 0.108,
+      "step": 1678
+    },
+    {
+      "epoch": 0.3815909090909091,
+      "grad_norm": 1.311970863517262,
+      "learning_rate": 1.2321429652711803e-06,
+      "loss": 0.0467,
+      "step": 1679
+    },
+    {
+      "epoch": 0.38181818181818183,
+      "grad_norm": 4.1063620379310075,
+      "learning_rate": 1.2321217773027827e-06,
+      "loss": 0.1297,
+      "step": 1680
+    },
+    {
+      "epoch": 0.3820454545454545,
+      "grad_norm": 2.0395677939741153,
+      "learning_rate": 1.232100576954113e-06,
+      "loss": 0.0943,
+      "step": 1681
+    },
+    {
+      "epoch": 0.38227272727272726,
+      "grad_norm": 2.84363115594454,
+      "learning_rate": 1.2320793642256031e-06,
+      "loss": 0.1386,
+      "step": 1682
+    },
+    {
+      "epoch": 0.3825,
+      "grad_norm": 1.3936183051099027,
+      "learning_rate": 1.232058139117686e-06,
+      "loss": 0.0645,
+      "step": 1683
+    },
+    {
+      "epoch": 0.38272727272727275,
+      "grad_norm": 2.865599282919753,
+      "learning_rate": 1.2320369016307947e-06,
+      "loss": 0.1297,
+      "step": 1684
+    },
+    {
+      "epoch": 0.38295454545454544,
+      "grad_norm": 1.8894366282652357,
+      "learning_rate": 1.2320156517653618e-06,
+      "loss": 0.1208,
+      "step": 1685
+    },
+    {
+      "epoch": 0.3831818181818182,
+      "grad_norm": 1.8148533452076563,
+      "learning_rate": 1.2319943895218205e-06,
+      "loss": 0.0574,
+      "step": 1686
+    },
+    {
+      "epoch": 0.3834090909090909,
+      "grad_norm": 1.2567034360748963,
+      "learning_rate": 1.2319731149006047e-06,
+      "loss": 0.053,
+      "step": 1687
+    },
+    {
+      "epoch": 0.3836363636363636,
+      "grad_norm": 1.4998239204376138,
+      "learning_rate": 1.2319518279021484e-06,
+      "loss": 0.0472,
+      "step": 1688
+    },
+    {
+      "epoch": 0.38386363636363635,
+      "grad_norm": 1.7343452545153306,
+      "learning_rate": 1.2319305285268852e-06,
+      "loss": 0.0634,
+      "step": 1689
+    },
+    {
+      "epoch": 0.3840909090909091,
+      "grad_norm": 1.5899330994089242,
+      "learning_rate": 1.2319092167752496e-06,
+      "loss": 0.077,
+      "step": 1690
+    },
+    {
+      "epoch": 0.38431818181818184,
+      "grad_norm": 1.8125021896984992,
+      "learning_rate": 1.2318878926476765e-06,
+      "loss": 0.062,
+      "step": 1691
+    },
+    {
+      "epoch": 0.3845454545454545,
+      "grad_norm": 2.6900029976200894,
+      "learning_rate": 1.2318665561446006e-06,
+      "loss": 0.123,
+      "step": 1692
+    },
+    {
+      "epoch": 0.38477272727272727,
+      "grad_norm": 1.3826876569403799,
+      "learning_rate": 1.2318452072664566e-06,
+      "loss": 0.0596,
+      "step": 1693
+    },
+    {
+      "epoch": 0.385,
+      "grad_norm": 2.4999039808840795,
+      "learning_rate": 1.2318238460136803e-06,
+      "loss": 0.1121,
+      "step": 1694
+    },
+    {
+      "epoch": 0.38522727272727275,
+      "grad_norm": 2.506446872392882,
+      "learning_rate": 1.231802472386707e-06,
+      "loss": 0.102,
+      "step": 1695
+    },
+    {
+      "epoch": 0.38545454545454544,
+      "grad_norm": 1.0983327523367696,
+      "learning_rate": 1.2317810863859728e-06,
+      "loss": 0.0434,
+      "step": 1696
+    },
+    {
+      "epoch": 0.3856818181818182,
+      "grad_norm": 1.7359347564961922,
+      "learning_rate": 1.2317596880119134e-06,
+      "loss": 0.0593,
+      "step": 1697
+    },
+    {
+      "epoch": 0.3859090909090909,
+      "grad_norm": 1.889806573952396,
+      "learning_rate": 1.2317382772649655e-06,
+      "loss": 0.107,
+      "step": 1698
+    },
+    {
+      "epoch": 0.3861363636363636,
+      "grad_norm": 0.5137350288571357,
+      "learning_rate": 1.2317168541455657e-06,
+      "loss": 0.028,
+      "step": 1699
+    },
+    {
+      "epoch": 0.38636363636363635,
+      "grad_norm": 0.912449056564532,
+      "learning_rate": 1.231695418654151e-06,
+      "loss": 0.0305,
+      "step": 1700
+    },
+    {
+      "epoch": 0.3865909090909091,
+      "grad_norm": 2.496079634571366,
+      "learning_rate": 1.231673970791158e-06,
+      "loss": 0.142,
+      "step": 1701
+    },
+    {
+      "epoch": 0.38681818181818184,
+      "grad_norm": 2.145334304066602,
+      "learning_rate": 1.2316525105570242e-06,
+      "loss": 0.077,
+      "step": 1702
+    },
+    {
+      "epoch": 0.3870454545454545,
+      "grad_norm": 1.2325099528283276,
+      "learning_rate": 1.2316310379521877e-06,
+      "loss": 0.063,
+      "step": 1703
+    },
+    {
+      "epoch": 0.38727272727272727,
+      "grad_norm": 1.3798470622848436,
+      "learning_rate": 1.2316095529770859e-06,
+      "loss": 0.0452,
+      "step": 1704
+    },
+    {
+      "epoch": 0.3875,
+      "grad_norm": 2.778234352860948,
+      "learning_rate": 1.231588055632157e-06,
+      "loss": 0.0947,
+      "step": 1705
+    },
+    {
+      "epoch": 0.38772727272727275,
+      "grad_norm": 3.9000705805788947,
+      "learning_rate": 1.231566545917839e-06,
+      "loss": 0.138,
+      "step": 1706
+    },
+    {
+      "epoch": 0.38795454545454544,
+      "grad_norm": 2.436109861550292,
+      "learning_rate": 1.2315450238345713e-06,
+      "loss": 0.1418,
+      "step": 1707
+    },
+    {
+      "epoch": 0.3881818181818182,
+      "grad_norm": 2.879681084618733,
+      "learning_rate": 1.2315234893827924e-06,
+      "loss": 0.1086,
+      "step": 1708
+    },
+    {
+      "epoch": 0.3884090909090909,
+      "grad_norm": 1.845239684888168,
+      "learning_rate": 1.2315019425629413e-06,
+      "loss": 0.0719,
+      "step": 1709
+    },
+    {
+      "epoch": 0.3886363636363636,
+      "grad_norm": 1.7338384742745092,
+      "learning_rate": 1.2314803833754576e-06,
+      "loss": 0.0856,
+      "step": 1710
+    },
+    {
+      "epoch": 0.38886363636363636,
+      "grad_norm": 3.059447499030947,
+      "learning_rate": 1.2314588118207808e-06,
+      "loss": 0.1318,
+      "step": 1711
+    },
+    {
+      "epoch": 0.3890909090909091,
+      "grad_norm": 1.3541009247729778,
+      "learning_rate": 1.2314372278993508e-06,
+      "loss": 0.0683,
+      "step": 1712
+    },
+    {
+      "epoch": 0.38931818181818184,
+      "grad_norm": 1.1109774538397592,
+      "learning_rate": 1.2314156316116078e-06,
+      "loss": 0.0567,
+      "step": 1713
+    },
+    {
+      "epoch": 0.38954545454545453,
+      "grad_norm": 1.8549933455733756,
+      "learning_rate": 1.231394022957992e-06,
+      "loss": 0.0866,
+      "step": 1714
+    },
+    {
+      "epoch": 0.38977272727272727,
+      "grad_norm": 2.2615710394079493,
+      "learning_rate": 1.2313724019389442e-06,
+      "loss": 0.075,
+      "step": 1715
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2618524397816246,
+      "learning_rate": 1.2313507685549054e-06,
+      "loss": 0.1086,
+      "step": 1716
+    },
+    {
+      "epoch": 0.3902272727272727,
+      "grad_norm": 2.7042938945826593,
+      "learning_rate": 1.2313291228063162e-06,
+      "loss": 0.1301,
+      "step": 1717
+    },
+    {
+      "epoch": 0.39045454545454544,
+      "grad_norm": 2.179979539509304,
+      "learning_rate": 1.2313074646936187e-06,
+      "loss": 0.1047,
+      "step": 1718
+    },
+    {
+      "epoch": 0.3906818181818182,
+      "grad_norm": 2.8299691960648956,
+      "learning_rate": 1.231285794217254e-06,
+      "loss": 0.1086,
+      "step": 1719
+    },
+    {
+      "epoch": 0.39090909090909093,
+      "grad_norm": 1.4318688520333198,
+      "learning_rate": 1.2312641113776646e-06,
+      "loss": 0.0803,
+      "step": 1720
+    },
+    {
+      "epoch": 0.3911363636363636,
+      "grad_norm": 1.563663217590128,
+      "learning_rate": 1.231242416175292e-06,
+      "loss": 0.0569,
+      "step": 1721
+    },
+    {
+      "epoch": 0.39136363636363636,
+      "grad_norm": 1.8728178378327789,
+      "learning_rate": 1.231220708610579e-06,
+      "loss": 0.0717,
+      "step": 1722
+    },
+    {
+      "epoch": 0.3915909090909091,
+      "grad_norm": 1.2710277012928968,
+      "learning_rate": 1.2311989886839682e-06,
+      "loss": 0.0583,
+      "step": 1723
+    },
+    {
+      "epoch": 0.39181818181818184,
+      "grad_norm": 3.226094375488176,
+      "learning_rate": 1.2311772563959022e-06,
+      "loss": 0.1432,
+      "step": 1724
+    },
+    {
+      "epoch": 0.39204545454545453,
+      "grad_norm": 2.2994765742793746,
+      "learning_rate": 1.2311555117468245e-06,
+      "loss": 0.1098,
+      "step": 1725
+    },
+    {
+      "epoch": 0.3922727272727273,
+      "grad_norm": 2.1765629186037145,
+      "learning_rate": 1.2311337547371785e-06,
+      "loss": 0.1004,
+      "step": 1726
+    },
+    {
+      "epoch": 0.3925,
+      "grad_norm": 3.7105548869853906,
+      "learning_rate": 1.2311119853674079e-06,
+      "loss": 0.1779,
+      "step": 1727
+    },
+    {
+      "epoch": 0.3927272727272727,
+      "grad_norm": 1.8899169485445602,
+      "learning_rate": 1.2310902036379563e-06,
+      "loss": 0.0721,
+      "step": 1728
+    },
+    {
+      "epoch": 0.39295454545454545,
+      "grad_norm": 1.844039177175954,
+      "learning_rate": 1.231068409549268e-06,
+      "loss": 0.0783,
+      "step": 1729
+    },
+    {
+      "epoch": 0.3931818181818182,
+      "grad_norm": 1.6586471317191023,
+      "learning_rate": 1.2310466031017876e-06,
+      "loss": 0.0664,
+      "step": 1730
+    },
+    {
+      "epoch": 0.39340909090909093,
+      "grad_norm": 1.364133628424039,
+      "learning_rate": 1.2310247842959597e-06,
+      "loss": 0.0544,
+      "step": 1731
+    },
+    {
+      "epoch": 0.3936363636363636,
+      "grad_norm": 1.1948505451987912,
+      "learning_rate": 1.2310029531322292e-06,
+      "loss": 0.0615,
+      "step": 1732
+    },
+    {
+      "epoch": 0.39386363636363636,
+      "grad_norm": 1.6121685110536068,
+      "learning_rate": 1.2309811096110412e-06,
+      "loss": 0.097,
+      "step": 1733
+    },
+    {
+      "epoch": 0.3940909090909091,
+      "grad_norm": 1.8001680731191108,
+      "learning_rate": 1.230959253732841e-06,
+      "loss": 0.0746,
+      "step": 1734
+    },
+    {
+      "epoch": 0.39431818181818185,
+      "grad_norm": 1.766264055522714,
+      "learning_rate": 1.2309373854980748e-06,
+      "loss": 0.1039,
+      "step": 1735
+    },
+    {
+      "epoch": 0.39454545454545453,
+      "grad_norm": 1.6331873520677913,
+      "learning_rate": 1.230915504907188e-06,
+      "loss": 0.0686,
+      "step": 1736
+    },
+    {
+      "epoch": 0.3947727272727273,
+      "grad_norm": 5.724033469505961,
+      "learning_rate": 1.2308936119606272e-06,
+      "loss": 0.2749,
+      "step": 1737
+    },
+    {
+      "epoch": 0.395,
+      "grad_norm": 1.4416117693860628,
+      "learning_rate": 1.2308717066588383e-06,
+      "loss": 0.065,
+      "step": 1738
+    },
+    {
+      "epoch": 0.3952272727272727,
+      "grad_norm": 2.135929443730567,
+      "learning_rate": 1.2308497890022684e-06,
+      "loss": 0.072,
+      "step": 1739
+    },
+    {
+      "epoch": 0.39545454545454545,
+      "grad_norm": 1.4882409008485717,
+      "learning_rate": 1.2308278589913644e-06,
+      "loss": 0.0599,
+      "step": 1740
+    },
+    {
+      "epoch": 0.3956818181818182,
+      "grad_norm": 1.2552505038442858,
+      "learning_rate": 1.2308059166265734e-06,
+      "loss": 0.042,
+      "step": 1741
+    },
+    {
+      "epoch": 0.39590909090909093,
+      "grad_norm": 1.7297656436455766,
+      "learning_rate": 1.2307839619083428e-06,
+      "loss": 0.0521,
+      "step": 1742
+    },
+    {
+      "epoch": 0.3961363636363636,
+      "grad_norm": 1.8099895333367777,
+      "learning_rate": 1.2307619948371205e-06,
+      "loss": 0.07,
+      "step": 1743
+    },
+    {
+      "epoch": 0.39636363636363636,
+      "grad_norm": 1.408784877097989,
+      "learning_rate": 1.2307400154133541e-06,
+      "loss": 0.0834,
+      "step": 1744
+    },
+    {
+      "epoch": 0.3965909090909091,
+      "grad_norm": 1.1836523057148116,
+      "learning_rate": 1.2307180236374922e-06,
+      "loss": 0.037,
+      "step": 1745
+    },
+    {
+      "epoch": 0.3968181818181818,
+      "grad_norm": 1.1564885292019196,
+      "learning_rate": 1.2306960195099833e-06,
+      "loss": 0.0637,
+      "step": 1746
+    },
+    {
+      "epoch": 0.39704545454545453,
+      "grad_norm": 1.5065137696748432,
+      "learning_rate": 1.2306740030312754e-06,
+      "loss": 0.0678,
+      "step": 1747
+    },
+    {
+      "epoch": 0.3972727272727273,
+      "grad_norm": 0.759563624355154,
+      "learning_rate": 1.230651974201818e-06,
+      "loss": 0.0526,
+      "step": 1748
+    },
+    {
+      "epoch": 0.3975,
+      "grad_norm": 1.8432947861573634,
+      "learning_rate": 1.2306299330220605e-06,
+      "loss": 0.118,
+      "step": 1749
+    },
+    {
+      "epoch": 0.3977272727272727,
+      "grad_norm": 2.7721313326015395,
+      "learning_rate": 1.230607879492452e-06,
+      "loss": 0.1231,
+      "step": 1750
+    },
+    {
+      "epoch": 0.39795454545454545,
+      "grad_norm": 1.248444513631358,
+      "learning_rate": 1.2305858136134422e-06,
+      "loss": 0.0484,
+      "step": 1751
+    },
+    {
+      "epoch": 0.3981818181818182,
+      "grad_norm": 1.5529937729055083,
+      "learning_rate": 1.230563735385481e-06,
+      "loss": 0.0649,
+      "step": 1752
+    },
+    {
+      "epoch": 0.39840909090909093,
+      "grad_norm": 1.9001309412805767,
+      "learning_rate": 1.2305416448090192e-06,
+      "loss": 0.118,
+      "step": 1753
+    },
+    {
+      "epoch": 0.3986363636363636,
+      "grad_norm": 1.7091841652096915,
+      "learning_rate": 1.2305195418845065e-06,
+      "loss": 0.1224,
+      "step": 1754
+    },
+    {
+      "epoch": 0.39886363636363636,
+      "grad_norm": 2.5129374510045586,
+      "learning_rate": 1.2304974266123941e-06,
+      "loss": 0.0858,
+      "step": 1755
+    },
+    {
+      "epoch": 0.3990909090909091,
+      "grad_norm": 2.1300245044473334,
+      "learning_rate": 1.2304752989931327e-06,
+      "loss": 0.0908,
+      "step": 1756
+    },
+    {
+      "epoch": 0.3993181818181818,
+      "grad_norm": 1.4867512126025368,
+      "learning_rate": 1.2304531590271738e-06,
+      "loss": 0.0591,
+      "step": 1757
+    },
+    {
+      "epoch": 0.39954545454545454,
+      "grad_norm": 2.434412962468815,
+      "learning_rate": 1.2304310067149686e-06,
+      "loss": 0.0974,
+      "step": 1758
+    },
+    {
+      "epoch": 0.3997727272727273,
+      "grad_norm": 2.7413669277563995,
+      "learning_rate": 1.230408842056969e-06,
+      "loss": 0.1074,
+      "step": 1759
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.474189431309057,
+      "learning_rate": 1.230386665053627e-06,
+      "loss": 0.0764,
+      "step": 1760
+    },
+    {
+      "epoch": 0.4002272727272727,
+      "grad_norm": 1.8698714536004617,
+      "learning_rate": 1.2303644757053945e-06,
+      "loss": 0.0737,
+      "step": 1761
+    },
+    {
+      "epoch": 0.40045454545454545,
+      "grad_norm": 1.6366038214456418,
+      "learning_rate": 1.2303422740127245e-06,
+      "loss": 0.0564,
+      "step": 1762
+    },
+    {
+      "epoch": 0.4006818181818182,
+      "grad_norm": 2.3906407195029784,
+      "learning_rate": 1.2303200599760692e-06,
+      "loss": 0.1006,
+      "step": 1763
+    },
+    {
+      "epoch": 0.4009090909090909,
+      "grad_norm": 2.5427194966293007,
+      "learning_rate": 1.2302978335958821e-06,
+      "loss": 0.1355,
+      "step": 1764
+    },
+    {
+      "epoch": 0.4011363636363636,
+      "grad_norm": 2.174718743842219,
+      "learning_rate": 1.2302755948726158e-06,
+      "loss": 0.121,
+      "step": 1765
+    },
+    {
+      "epoch": 0.40136363636363637,
+      "grad_norm": 2.631529757254184,
+      "learning_rate": 1.2302533438067247e-06,
+      "loss": 0.1256,
+      "step": 1766
+    },
+    {
+      "epoch": 0.4015909090909091,
+      "grad_norm": 1.8519852229846727,
+      "learning_rate": 1.2302310803986613e-06,
+      "loss": 0.1038,
+      "step": 1767
+    },
+    {
+      "epoch": 0.4018181818181818,
+      "grad_norm": 2.3198172534462236,
+      "learning_rate": 1.2302088046488809e-06,
+      "loss": 0.1423,
+      "step": 1768
+    },
+    {
+      "epoch": 0.40204545454545454,
+      "grad_norm": 2.437406517903106,
+      "learning_rate": 1.230186516557837e-06,
+      "loss": 0.1214,
+      "step": 1769
+    },
+    {
+      "epoch": 0.4022727272727273,
+      "grad_norm": 2.3863959464576716,
+      "learning_rate": 1.230164216125984e-06,
+      "loss": 0.1194,
+      "step": 1770
+    },
+    {
+      "epoch": 0.4025,
+      "grad_norm": 1.7261889768447511,
+      "learning_rate": 1.230141903353777e-06,
+      "loss": 0.0912,
+      "step": 1771
+    },
+    {
+      "epoch": 0.4027272727272727,
+      "grad_norm": 1.0563911532230497,
+      "learning_rate": 1.2301195782416707e-06,
+      "loss": 0.0648,
+      "step": 1772
+    },
+    {
+      "epoch": 0.40295454545454545,
+      "grad_norm": 1.4497492180202034,
+      "learning_rate": 1.2300972407901208e-06,
+      "loss": 0.092,
+      "step": 1773
+    },
+    {
+      "epoch": 0.4031818181818182,
+      "grad_norm": 1.554475995388426,
+      "learning_rate": 1.2300748909995824e-06,
+      "loss": 0.07,
+      "step": 1774
+    },
+    {
+      "epoch": 0.4034090909090909,
+      "grad_norm": 2.4139172125906403,
+      "learning_rate": 1.2300525288705113e-06,
+      "loss": 0.0978,
+      "step": 1775
+    },
+    {
+      "epoch": 0.4036363636363636,
+      "grad_norm": 2.3639679276972196,
+      "learning_rate": 1.2300301544033636e-06,
+      "loss": 0.0976,
+      "step": 1776
+    },
+    {
+      "epoch": 0.40386363636363637,
+      "grad_norm": 2.0532391891265744,
+      "learning_rate": 1.2300077675985956e-06,
+      "loss": 0.1083,
+      "step": 1777
+    },
+    {
+      "epoch": 0.4040909090909091,
+      "grad_norm": 2.3158859511372136,
+      "learning_rate": 1.2299853684566636e-06,
+      "loss": 0.1651,
+      "step": 1778
+    },
+    {
+      "epoch": 0.4043181818181818,
+      "grad_norm": 0.8167914613593289,
+      "learning_rate": 1.2299629569780246e-06,
+      "loss": 0.0253,
+      "step": 1779
+    },
+    {
+      "epoch": 0.40454545454545454,
+      "grad_norm": 1.1640005976916672,
+      "learning_rate": 1.2299405331631354e-06,
+      "loss": 0.0484,
+      "step": 1780
+    },
+    {
+      "epoch": 0.4047727272727273,
+      "grad_norm": 2.3863496262044523,
+      "learning_rate": 1.2299180970124533e-06,
+      "loss": 0.1385,
+      "step": 1781
+    },
+    {
+      "epoch": 0.405,
+      "grad_norm": 1.434064360593994,
+      "learning_rate": 1.229895648526436e-06,
+      "loss": 0.0786,
+      "step": 1782
+    },
+    {
+      "epoch": 0.4052272727272727,
+      "grad_norm": 1.6244601946944062,
+      "learning_rate": 1.229873187705541e-06,
+      "loss": 0.0684,
+      "step": 1783
+    },
+    {
+      "epoch": 0.40545454545454546,
+      "grad_norm": 1.4632609449180758,
+      "learning_rate": 1.2298507145502265e-06,
+      "loss": 0.0662,
+      "step": 1784
+    },
+    {
+      "epoch": 0.4056818181818182,
+      "grad_norm": 2.6444499756093744,
+      "learning_rate": 1.2298282290609507e-06,
+      "loss": 0.1372,
+      "step": 1785
+    },
+    {
+      "epoch": 0.4059090909090909,
+      "grad_norm": 1.6413941364835305,
+      "learning_rate": 1.2298057312381723e-06,
+      "loss": 0.09,
+      "step": 1786
+    },
+    {
+      "epoch": 0.40613636363636363,
+      "grad_norm": 1.8275043818517807,
+      "learning_rate": 1.2297832210823497e-06,
+      "loss": 0.0977,
+      "step": 1787
+    },
+    {
+      "epoch": 0.40636363636363637,
+      "grad_norm": 1.6912404423366334,
+      "learning_rate": 1.229760698593942e-06,
+      "loss": 0.0847,
+      "step": 1788
+    },
+    {
+      "epoch": 0.4065909090909091,
+      "grad_norm": 3.075673112883564,
+      "learning_rate": 1.2297381637734087e-06,
+      "loss": 0.151,
+      "step": 1789
+    },
+    {
+      "epoch": 0.4068181818181818,
+      "grad_norm": 2.4342340295664515,
+      "learning_rate": 1.2297156166212095e-06,
+      "loss": 0.1214,
+      "step": 1790
+    },
+    {
+      "epoch": 0.40704545454545454,
+      "grad_norm": 2.3047054670211686,
+      "learning_rate": 1.2296930571378035e-06,
+      "loss": 0.1339,
+      "step": 1791
+    },
+    {
+      "epoch": 0.4072727272727273,
+      "grad_norm": 2.0451700695710597,
+      "learning_rate": 1.2296704853236513e-06,
+      "loss": 0.0904,
+      "step": 1792
+    },
+    {
+      "epoch": 0.4075,
+      "grad_norm": 2.080789592318642,
+      "learning_rate": 1.2296479011792132e-06,
+      "loss": 0.1121,
+      "step": 1793
+    },
+    {
+      "epoch": 0.4077272727272727,
+      "grad_norm": 2.340047996080516,
+      "learning_rate": 1.2296253047049493e-06,
+      "loss": 0.1717,
+      "step": 1794
+    },
+    {
+      "epoch": 0.40795454545454546,
+      "grad_norm": 2.583885513054682,
+      "learning_rate": 1.2296026959013205e-06,
+      "loss": 0.1669,
+      "step": 1795
+    },
+    {
+      "epoch": 0.4081818181818182,
+      "grad_norm": 1.2592720216720525,
+      "learning_rate": 1.229580074768788e-06,
+      "loss": 0.0481,
+      "step": 1796
+    },
+    {
+      "epoch": 0.4084090909090909,
+      "grad_norm": 1.22803742847098,
+      "learning_rate": 1.2295574413078132e-06,
+      "loss": 0.0447,
+      "step": 1797
+    },
+    {
+      "epoch": 0.40863636363636363,
+      "grad_norm": 1.9299433470039473,
+      "learning_rate": 1.2295347955188574e-06,
+      "loss": 0.0581,
+      "step": 1798
+    },
+    {
+      "epoch": 0.4088636363636364,
+      "grad_norm": 2.0773783331431326,
+      "learning_rate": 1.2295121374023823e-06,
+      "loss": 0.0982,
+      "step": 1799
+    },
+    {
+      "epoch": 0.4090909090909091,
+      "grad_norm": 1.6264496411854865,
+      "learning_rate": 1.2294894669588504e-06,
+      "loss": 0.0548,
+      "step": 1800
+    },
+    {
+      "epoch": 0.4093181818181818,
+      "grad_norm": 1.2538536374881042,
+      "learning_rate": 1.2294667841887234e-06,
+      "loss": 0.0496,
+      "step": 1801
+    },
+    {
+      "epoch": 0.40954545454545455,
+      "grad_norm": 1.6345547780889969,
+      "learning_rate": 1.2294440890924642e-06,
+      "loss": 0.0533,
+      "step": 1802
+    },
+    {
+      "epoch": 0.4097727272727273,
+      "grad_norm": 2.0558249073212114,
+      "learning_rate": 1.2294213816705354e-06,
+      "loss": 0.0521,
+      "step": 1803
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.43164053235256,
+      "learning_rate": 1.2293986619234003e-06,
+      "loss": 0.1544,
+      "step": 1804
+    },
+    {
+      "epoch": 0.4102272727272727,
+      "grad_norm": 2.240369285344018,
+      "learning_rate": 1.229375929851522e-06,
+      "loss": 0.11,
+      "step": 1805
+    },
+    {
+      "epoch": 0.41045454545454546,
+      "grad_norm": 1.5776758790838217,
+      "learning_rate": 1.2293531854553642e-06,
+      "loss": 0.0578,
+      "step": 1806
+    },
+    {
+      "epoch": 0.4106818181818182,
+      "grad_norm": 2.045033034546197,
+      "learning_rate": 1.2293304287353905e-06,
+      "loss": 0.078,
+      "step": 1807
+    },
+    {
+      "epoch": 0.4109090909090909,
+      "grad_norm": 4.844774441556656,
+      "learning_rate": 1.229307659692065e-06,
+      "loss": 0.1669,
+      "step": 1808
+    },
+    {
+      "epoch": 0.41113636363636363,
+      "grad_norm": 1.7639269743354025,
+      "learning_rate": 1.2292848783258521e-06,
+      "loss": 0.0696,
+      "step": 1809
+    },
+    {
+      "epoch": 0.4113636363636364,
+      "grad_norm": 2.316882906225504,
+      "learning_rate": 1.2292620846372165e-06,
+      "loss": 0.1103,
+      "step": 1810
+    },
+    {
+      "epoch": 0.4115909090909091,
+      "grad_norm": 1.856432707288657,
+      "learning_rate": 1.2292392786266225e-06,
+      "loss": 0.1147,
+      "step": 1811
+    },
+    {
+      "epoch": 0.4118181818181818,
+      "grad_norm": 1.4309162165316784,
+      "learning_rate": 1.2292164602945356e-06,
+      "loss": 0.0581,
+      "step": 1812
+    },
+    {
+      "epoch": 0.41204545454545455,
+      "grad_norm": 2.401237200880494,
+      "learning_rate": 1.2291936296414209e-06,
+      "loss": 0.1109,
+      "step": 1813
+    },
+    {
+      "epoch": 0.4122727272727273,
+      "grad_norm": 2.1459494212843153,
+      "learning_rate": 1.229170786667744e-06,
+      "loss": 0.0825,
+      "step": 1814
+    },
+    {
+      "epoch": 0.4125,
+      "grad_norm": 1.6727624858933756,
+      "learning_rate": 1.2291479313739707e-06,
+      "loss": 0.0796,
+      "step": 1815
+    },
+    {
+      "epoch": 0.4127272727272727,
+      "grad_norm": 2.208720500026016,
+      "learning_rate": 1.2291250637605672e-06,
+      "loss": 0.0949,
+      "step": 1816
+    },
+    {
+      "epoch": 0.41295454545454546,
+      "grad_norm": 2.596337061719244,
+      "learning_rate": 1.2291021838279995e-06,
+      "loss": 0.08,
+      "step": 1817
+    },
+    {
+      "epoch": 0.4131818181818182,
+      "grad_norm": 1.7429583804897735,
+      "learning_rate": 1.2290792915767347e-06,
+      "loss": 0.1203,
+      "step": 1818
+    },
+    {
+      "epoch": 0.4134090909090909,
+      "grad_norm": 2.13023803252444,
+      "learning_rate": 1.229056387007239e-06,
+      "loss": 0.0838,
+      "step": 1819
+    },
+    {
+      "epoch": 0.41363636363636364,
+      "grad_norm": 2.2664228172517866,
+      "learning_rate": 1.22903347011998e-06,
+      "loss": 0.0711,
+      "step": 1820
+    },
+    {
+      "epoch": 0.4138636363636364,
+      "grad_norm": 1.3223832030524847,
+      "learning_rate": 1.2290105409154244e-06,
+      "loss": 0.071,
+      "step": 1821
+    },
+    {
+      "epoch": 0.41409090909090907,
+      "grad_norm": 2.1443327196676716,
+      "learning_rate": 1.2289875993940404e-06,
+      "loss": 0.0978,
+      "step": 1822
+    },
+    {
+      "epoch": 0.4143181818181818,
+      "grad_norm": 2.0278166978179546,
+      "learning_rate": 1.2289646455562954e-06,
+      "loss": 0.0704,
+      "step": 1823
+    },
+    {
+      "epoch": 0.41454545454545455,
+      "grad_norm": 1.4468455182508164,
+      "learning_rate": 1.2289416794026575e-06,
+      "loss": 0.06,
+      "step": 1824
+    },
+    {
+      "epoch": 0.4147727272727273,
+      "grad_norm": 2.2996866398013793,
+      "learning_rate": 1.2289187009335953e-06,
+      "loss": 0.1094,
+      "step": 1825
+    },
+    {
+      "epoch": 0.415,
+      "grad_norm": 1.5686672423041141,
+      "learning_rate": 1.2288957101495772e-06,
+      "loss": 0.0586,
+      "step": 1826
+    },
+    {
+      "epoch": 0.4152272727272727,
+      "grad_norm": 2.157844897447044,
+      "learning_rate": 1.2288727070510718e-06,
+      "loss": 0.1106,
+      "step": 1827
+    },
+    {
+      "epoch": 0.41545454545454547,
+      "grad_norm": 1.9110679208743553,
+      "learning_rate": 1.2288496916385484e-06,
+      "loss": 0.0949,
+      "step": 1828
+    },
+    {
+      "epoch": 0.4156818181818182,
+      "grad_norm": 3.075692171317193,
+      "learning_rate": 1.2288266639124767e-06,
+      "loss": 0.1287,
+      "step": 1829
+    },
+    {
+      "epoch": 0.4159090909090909,
+      "grad_norm": 2.5084199693722105,
+      "learning_rate": 1.2288036238733256e-06,
+      "loss": 0.1286,
+      "step": 1830
+    },
+    {
+      "epoch": 0.41613636363636364,
+      "grad_norm": 1.9957594286496898,
+      "learning_rate": 1.2287805715215651e-06,
+      "loss": 0.1177,
+      "step": 1831
+    },
+    {
+      "epoch": 0.4163636363636364,
+      "grad_norm": 1.4648259440091809,
+      "learning_rate": 1.2287575068576658e-06,
+      "loss": 0.0435,
+      "step": 1832
+    },
+    {
+      "epoch": 0.41659090909090907,
+      "grad_norm": 2.4127862690536395,
+      "learning_rate": 1.2287344298820972e-06,
+      "loss": 0.1023,
+      "step": 1833
+    },
+    {
+      "epoch": 0.4168181818181818,
+      "grad_norm": 2.89400644092995,
+      "learning_rate": 1.2287113405953306e-06,
+      "loss": 0.0941,
+      "step": 1834
+    },
+    {
+      "epoch": 0.41704545454545455,
+      "grad_norm": 2.7375643570367822,
+      "learning_rate": 1.2286882389978362e-06,
+      "loss": 0.1002,
+      "step": 1835
+    },
+    {
+      "epoch": 0.4172727272727273,
+      "grad_norm": 1.4069867829415468,
+      "learning_rate": 1.2286651250900858e-06,
+      "loss": 0.0679,
+      "step": 1836
+    },
+    {
+      "epoch": 0.4175,
+      "grad_norm": 1.7836971953403036,
+      "learning_rate": 1.22864199887255e-06,
+      "loss": 0.0803,
+      "step": 1837
+    },
+    {
+      "epoch": 0.4177272727272727,
+      "grad_norm": 2.3855115629855503,
+      "learning_rate": 1.228618860345701e-06,
+      "loss": 0.0581,
+      "step": 1838
+    },
+    {
+      "epoch": 0.41795454545454547,
+      "grad_norm": 1.604440776567854,
+      "learning_rate": 1.2285957095100102e-06,
+      "loss": 0.0713,
+      "step": 1839
+    },
+    {
+      "epoch": 0.41818181818181815,
+      "grad_norm": 3.235026953735444,
+      "learning_rate": 1.2285725463659501e-06,
+      "loss": 0.1626,
+      "step": 1840
+    },
+    {
+      "epoch": 0.4184090909090909,
+      "grad_norm": 2.71844905558046,
+      "learning_rate": 1.2285493709139925e-06,
+      "loss": 0.1226,
+      "step": 1841
+    },
+    {
+      "epoch": 0.41863636363636364,
+      "grad_norm": 1.6419425669013572,
+      "learning_rate": 1.2285261831546104e-06,
+      "loss": 0.0541,
+      "step": 1842
+    },
+    {
+      "epoch": 0.4188636363636364,
+      "grad_norm": 1.147444967865872,
+      "learning_rate": 1.2285029830882766e-06,
+      "loss": 0.0566,
+      "step": 1843
+    },
+    {
+      "epoch": 0.41909090909090907,
+      "grad_norm": 1.1429531622701397,
+      "learning_rate": 1.2284797707154638e-06,
+      "loss": 0.0324,
+      "step": 1844
+    },
+    {
+      "epoch": 0.4193181818181818,
+      "grad_norm": 1.4398502803521291,
+      "learning_rate": 1.228456546036646e-06,
+      "loss": 0.0573,
+      "step": 1845
+    },
+    {
+      "epoch": 0.41954545454545455,
+      "grad_norm": 1.7115208144378316,
+      "learning_rate": 1.2284333090522962e-06,
+      "loss": 0.0675,
+      "step": 1846
+    },
+    {
+      "epoch": 0.4197727272727273,
+      "grad_norm": 1.2971155569381565,
+      "learning_rate": 1.2284100597628886e-06,
+      "loss": 0.0481,
+      "step": 1847
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5299865816312909,
+      "learning_rate": 1.228386798168897e-06,
+      "loss": 0.0314,
+      "step": 1848
+    },
+    {
+      "epoch": 0.4202272727272727,
+      "grad_norm": 5.399356515943537,
+      "learning_rate": 1.2283635242707962e-06,
+      "loss": 0.1345,
+      "step": 1849
+    },
+    {
+      "epoch": 0.42045454545454547,
+      "grad_norm": 1.9573721525957488,
+      "learning_rate": 1.2283402380690603e-06,
+      "loss": 0.0711,
+      "step": 1850
+    },
+    {
+      "epoch": 0.42068181818181816,
+      "grad_norm": 1.5952641462753534,
+      "learning_rate": 1.2283169395641647e-06,
+      "loss": 0.0844,
+      "step": 1851
+    },
+    {
+      "epoch": 0.4209090909090909,
+      "grad_norm": 2.629971555287841,
+      "learning_rate": 1.228293628756584e-06,
+      "loss": 0.1025,
+      "step": 1852
+    },
+    {
+      "epoch": 0.42113636363636364,
+      "grad_norm": 2.048254950553639,
+      "learning_rate": 1.2282703056467934e-06,
+      "loss": 0.0959,
+      "step": 1853
+    },
+    {
+      "epoch": 0.4213636363636364,
+      "grad_norm": 1.876593969903028,
+      "learning_rate": 1.228246970235269e-06,
+      "loss": 0.1046,
+      "step": 1854
+    },
+    {
+      "epoch": 0.42159090909090907,
+      "grad_norm": 1.4337833008612024,
+      "learning_rate": 1.2282236225224864e-06,
+      "loss": 0.0639,
+      "step": 1855
+    },
+    {
+      "epoch": 0.4218181818181818,
+      "grad_norm": 0.9954283746560435,
+      "learning_rate": 1.228200262508922e-06,
+      "loss": 0.0544,
+      "step": 1856
+    },
+    {
+      "epoch": 0.42204545454545456,
+      "grad_norm": 2.479878410430168,
+      "learning_rate": 1.2281768901950518e-06,
+      "loss": 0.0931,
+      "step": 1857
+    },
+    {
+      "epoch": 0.4222727272727273,
+      "grad_norm": 1.8437014279960848,
+      "learning_rate": 1.2281535055813523e-06,
+      "loss": 0.0744,
+      "step": 1858
+    },
+    {
+      "epoch": 0.4225,
+      "grad_norm": 1.5831881134815051,
+      "learning_rate": 1.2281301086683008e-06,
+      "loss": 0.0528,
+      "step": 1859
+    },
+    {
+      "epoch": 0.42272727272727273,
+      "grad_norm": 2.114785691256069,
+      "learning_rate": 1.228106699456374e-06,
+      "loss": 0.1465,
+      "step": 1860
+    },
+    {
+      "epoch": 0.42295454545454547,
+      "grad_norm": 2.1940402601346913,
+      "learning_rate": 1.2280832779460494e-06,
+      "loss": 0.1356,
+      "step": 1861
+    },
+    {
+      "epoch": 0.42318181818181816,
+      "grad_norm": 2.372954487461523,
+      "learning_rate": 1.2280598441378048e-06,
+      "loss": 0.0907,
+      "step": 1862
+    },
+    {
+      "epoch": 0.4234090909090909,
+      "grad_norm": 1.4436058430379948,
+      "learning_rate": 1.2280363980321178e-06,
+      "loss": 0.0551,
+      "step": 1863
+    },
+    {
+      "epoch": 0.42363636363636364,
+      "grad_norm": 1.863767160035595,
+      "learning_rate": 1.2280129396294666e-06,
+      "loss": 0.1199,
+      "step": 1864
+    },
+    {
+      "epoch": 0.4238636363636364,
+      "grad_norm": 1.740161299945595,
+      "learning_rate": 1.2279894689303296e-06,
+      "loss": 0.0884,
+      "step": 1865
+    },
+    {
+      "epoch": 0.4240909090909091,
+      "grad_norm": 1.8401139352684723,
+      "learning_rate": 1.2279659859351853e-06,
+      "loss": 0.0732,
+      "step": 1866
+    },
+    {
+      "epoch": 0.4243181818181818,
+      "grad_norm": 1.7346652562659017,
+      "learning_rate": 1.2279424906445126e-06,
+      "loss": 0.0965,
+      "step": 1867
+    },
+    {
+      "epoch": 0.42454545454545456,
+      "grad_norm": 2.522518703288235,
+      "learning_rate": 1.2279189830587907e-06,
+      "loss": 0.0989,
+      "step": 1868
+    },
+    {
+      "epoch": 0.42477272727272725,
+      "grad_norm": 2.4612604090261856,
+      "learning_rate": 1.2278954631784987e-06,
+      "loss": 0.132,
+      "step": 1869
+    },
+    {
+      "epoch": 0.425,
+      "grad_norm": 2.1327155507914526,
+      "learning_rate": 1.2278719310041164e-06,
+      "loss": 0.0864,
+      "step": 1870
+    },
+    {
+      "epoch": 0.42522727272727273,
+      "grad_norm": 2.032708582509577,
+      "learning_rate": 1.2278483865361239e-06,
+      "loss": 0.0599,
+      "step": 1871
+    },
+    {
+      "epoch": 0.4254545454545455,
+      "grad_norm": 1.5030135209296547,
+      "learning_rate": 1.227824829775001e-06,
+      "loss": 0.061,
+      "step": 1872
+    },
+    {
+      "epoch": 0.42568181818181816,
+      "grad_norm": 1.6618030918516136,
+      "learning_rate": 1.227801260721228e-06,
+      "loss": 0.0345,
+      "step": 1873
+    },
+    {
+      "epoch": 0.4259090909090909,
+      "grad_norm": 2.432740325259115,
+      "learning_rate": 1.2277776793752857e-06,
+      "loss": 0.1626,
+      "step": 1874
+    },
+    {
+      "epoch": 0.42613636363636365,
+      "grad_norm": 1.1987308150359588,
+      "learning_rate": 1.2277540857376549e-06,
+      "loss": 0.0577,
+      "step": 1875
+    },
+    {
+      "epoch": 0.4263636363636364,
+      "grad_norm": 2.1550827951068827,
+      "learning_rate": 1.227730479808817e-06,
+      "loss": 0.1332,
+      "step": 1876
+    },
+    {
+      "epoch": 0.4265909090909091,
+      "grad_norm": 1.2476461703933936,
+      "learning_rate": 1.2277068615892528e-06,
+      "loss": 0.0771,
+      "step": 1877
+    },
+    {
+      "epoch": 0.4268181818181818,
+      "grad_norm": 1.825167402236287,
+      "learning_rate": 1.2276832310794442e-06,
+      "loss": 0.1007,
+      "step": 1878
+    },
+    {
+      "epoch": 0.42704545454545456,
+      "grad_norm": 3.005396590104723,
+      "learning_rate": 1.2276595882798733e-06,
+      "loss": 0.1938,
+      "step": 1879
+    },
+    {
+      "epoch": 0.42727272727272725,
+      "grad_norm": 2.3370263161403013,
+      "learning_rate": 1.2276359331910217e-06,
+      "loss": 0.1743,
+      "step": 1880
+    },
+    {
+      "epoch": 0.4275,
+      "grad_norm": 1.5765835047280845,
+      "learning_rate": 1.2276122658133723e-06,
+      "loss": 0.0487,
+      "step": 1881
+    },
+    {
+      "epoch": 0.42772727272727273,
+      "grad_norm": 2.059849592759223,
+      "learning_rate": 1.2275885861474075e-06,
+      "loss": 0.0839,
+      "step": 1882
+    },
+    {
+      "epoch": 0.4279545454545455,
+      "grad_norm": 3.6159883225193483,
+      "learning_rate": 1.22756489419361e-06,
+      "loss": 0.0978,
+      "step": 1883
+    },
+    {
+      "epoch": 0.42818181818181816,
+      "grad_norm": 1.028075751516862,
+      "learning_rate": 1.227541189952463e-06,
+      "loss": 0.05,
+      "step": 1884
+    },
+    {
+      "epoch": 0.4284090909090909,
+      "grad_norm": 1.5134200895516643,
+      "learning_rate": 1.2275174734244502e-06,
+      "loss": 0.0588,
+      "step": 1885
+    },
+    {
+      "epoch": 0.42863636363636365,
+      "grad_norm": 2.1229231491920784,
+      "learning_rate": 1.2274937446100548e-06,
+      "loss": 0.0822,
+      "step": 1886
+    },
+    {
+      "epoch": 0.4288636363636364,
+      "grad_norm": 2.796386073734362,
+      "learning_rate": 1.227470003509761e-06,
+      "loss": 0.0828,
+      "step": 1887
+    },
+    {
+      "epoch": 0.4290909090909091,
+      "grad_norm": 1.8096418706555726,
+      "learning_rate": 1.2274462501240526e-06,
+      "loss": 0.0638,
+      "step": 1888
+    },
+    {
+      "epoch": 0.4293181818181818,
+      "grad_norm": 1.70439311618849,
+      "learning_rate": 1.2274224844534143e-06,
+      "loss": 0.1004,
+      "step": 1889
+    },
+    {
+      "epoch": 0.42954545454545456,
+      "grad_norm": 1.7376372742402162,
+      "learning_rate": 1.2273987064983303e-06,
+      "loss": 0.0551,
+      "step": 1890
+    },
+    {
+      "epoch": 0.42977272727272725,
+      "grad_norm": 1.8988984403787164,
+      "learning_rate": 1.227374916259286e-06,
+      "loss": 0.0662,
+      "step": 1891
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2026596305504476,
+      "learning_rate": 1.227351113736766e-06,
+      "loss": 0.0968,
+      "step": 1892
+    },
+    {
+      "epoch": 0.43022727272727274,
+      "grad_norm": 2.6865404563368913,
+      "learning_rate": 1.2273272989312562e-06,
+      "loss": 0.1384,
+      "step": 1893
+    },
+    {
+      "epoch": 0.4304545454545455,
+      "grad_norm": 2.083376565953406,
+      "learning_rate": 1.2273034718432417e-06,
+      "loss": 0.0977,
+      "step": 1894
+    },
+    {
+      "epoch": 0.43068181818181817,
+      "grad_norm": 1.8484408161203076,
+      "learning_rate": 1.2272796324732088e-06,
+      "loss": 0.0921,
+      "step": 1895
+    },
+    {
+      "epoch": 0.4309090909090909,
+      "grad_norm": 1.4021429592366164,
+      "learning_rate": 1.2272557808216433e-06,
+      "loss": 0.0625,
+      "step": 1896
+    },
+    {
+      "epoch": 0.43113636363636365,
+      "grad_norm": 1.8938360780720347,
+      "learning_rate": 1.227231916889032e-06,
+      "loss": 0.086,
+      "step": 1897
+    },
+    {
+      "epoch": 0.43136363636363634,
+      "grad_norm": 2.629488181344048,
+      "learning_rate": 1.2272080406758612e-06,
+      "loss": 0.1913,
+      "step": 1898
+    },
+    {
+      "epoch": 0.4315909090909091,
+      "grad_norm": 1.613641401757471,
+      "learning_rate": 1.2271841521826177e-06,
+      "loss": 0.103,
+      "step": 1899
+    },
+    {
+      "epoch": 0.4318181818181818,
+      "grad_norm": 2.3338820987912388,
+      "learning_rate": 1.2271602514097888e-06,
+      "loss": 0.0657,
+      "step": 1900
+    },
+    {
+      "epoch": 0.43204545454545457,
+      "grad_norm": 1.3022987799474932,
+      "learning_rate": 1.2271363383578619e-06,
+      "loss": 0.0379,
+      "step": 1901
+    },
+    {
+      "epoch": 0.43227272727272725,
+      "grad_norm": 1.5643451957791583,
+      "learning_rate": 1.2271124130273245e-06,
+      "loss": 0.0532,
+      "step": 1902
+    },
+    {
+      "epoch": 0.4325,
+      "grad_norm": 2.639568556095883,
+      "learning_rate": 1.2270884754186645e-06,
+      "loss": 0.0919,
+      "step": 1903
+    },
+    {
+      "epoch": 0.43272727272727274,
+      "grad_norm": 1.495414529648914,
+      "learning_rate": 1.2270645255323704e-06,
+      "loss": 0.0464,
+      "step": 1904
+    },
+    {
+      "epoch": 0.4329545454545455,
+      "grad_norm": 1.7005431482155164,
+      "learning_rate": 1.22704056336893e-06,
+      "loss": 0.07,
+      "step": 1905
+    },
+    {
+      "epoch": 0.43318181818181817,
+      "grad_norm": 2.0927020451502227,
+      "learning_rate": 1.2270165889288325e-06,
+      "loss": 0.1295,
+      "step": 1906
+    },
+    {
+      "epoch": 0.4334090909090909,
+      "grad_norm": 1.852185538026283,
+      "learning_rate": 1.226992602212566e-06,
+      "loss": 0.0619,
+      "step": 1907
+    },
+    {
+      "epoch": 0.43363636363636365,
+      "grad_norm": 1.559915187819894,
+      "learning_rate": 1.2269686032206205e-06,
+      "loss": 0.0957,
+      "step": 1908
+    },
+    {
+      "epoch": 0.43386363636363634,
+      "grad_norm": 2.118265260299252,
+      "learning_rate": 1.2269445919534848e-06,
+      "loss": 0.0898,
+      "step": 1909
+    },
+    {
+      "epoch": 0.4340909090909091,
+      "grad_norm": 2.4600294081449317,
+      "learning_rate": 1.2269205684116486e-06,
+      "loss": 0.0988,
+      "step": 1910
+    },
+    {
+      "epoch": 0.4343181818181818,
+      "grad_norm": 2.23239658617845,
+      "learning_rate": 1.2268965325956022e-06,
+      "loss": 0.0959,
+      "step": 1911
+    },
+    {
+      "epoch": 0.43454545454545457,
+      "grad_norm": 1.9195266008564258,
+      "learning_rate": 1.2268724845058351e-06,
+      "loss": 0.1074,
+      "step": 1912
+    },
+    {
+      "epoch": 0.43477272727272726,
+      "grad_norm": 1.4912155967431884,
+      "learning_rate": 1.2268484241428383e-06,
+      "loss": 0.0747,
+      "step": 1913
+    },
+    {
+      "epoch": 0.435,
+      "grad_norm": 1.9312536655817063,
+      "learning_rate": 1.226824351507102e-06,
+      "loss": 0.1147,
+      "step": 1914
+    },
+    {
+      "epoch": 0.43522727272727274,
+      "grad_norm": 2.935348419483259,
+      "learning_rate": 1.2268002665991174e-06,
+      "loss": 0.1537,
+      "step": 1915
+    },
+    {
+      "epoch": 0.4354545454545454,
+      "grad_norm": 1.202068322193706,
+      "learning_rate": 1.2267761694193752e-06,
+      "loss": 0.0533,
+      "step": 1916
+    },
+    {
+      "epoch": 0.43568181818181817,
+      "grad_norm": 1.9644418724608632,
+      "learning_rate": 1.2267520599683673e-06,
+      "loss": 0.1,
+      "step": 1917
+    },
+    {
+      "epoch": 0.4359090909090909,
+      "grad_norm": 2.213572142511544,
+      "learning_rate": 1.226727938246585e-06,
+      "loss": 0.0993,
+      "step": 1918
+    },
+    {
+      "epoch": 0.43613636363636366,
+      "grad_norm": 2.0770023863787914,
+      "learning_rate": 1.2267038042545201e-06,
+      "loss": 0.1218,
+      "step": 1919
+    },
+    {
+      "epoch": 0.43636363636363634,
+      "grad_norm": 1.5557780171218951,
+      "learning_rate": 1.226679657992665e-06,
+      "loss": 0.0567,
+      "step": 1920
+    },
+    {
+      "epoch": 0.4365909090909091,
+      "grad_norm": 0.39985232979257096,
+      "learning_rate": 1.2266554994615121e-06,
+      "loss": 0.0209,
+      "step": 1921
+    },
+    {
+      "epoch": 0.4368181818181818,
+      "grad_norm": 1.3201214219196873,
+      "learning_rate": 1.2266313286615539e-06,
+      "loss": 0.0774,
+      "step": 1922
+    },
+    {
+      "epoch": 0.43704545454545457,
+      "grad_norm": 2.4828054733474567,
+      "learning_rate": 1.2266071455932833e-06,
+      "loss": 0.0918,
+      "step": 1923
+    },
+    {
+      "epoch": 0.43727272727272726,
+      "grad_norm": 2.6597581735004376,
+      "learning_rate": 1.2265829502571935e-06,
+      "loss": 0.0937,
+      "step": 1924
+    },
+    {
+      "epoch": 0.4375,
+      "grad_norm": 1.397267613792231,
+      "learning_rate": 1.2265587426537776e-06,
+      "loss": 0.0861,
+      "step": 1925
+    },
+    {
+      "epoch": 0.43772727272727274,
+      "grad_norm": 1.7002230058163235,
+      "learning_rate": 1.2265345227835295e-06,
+      "loss": 0.0866,
+      "step": 1926
+    },
+    {
+      "epoch": 0.43795454545454543,
+      "grad_norm": 2.050143924581109,
+      "learning_rate": 1.2265102906469433e-06,
+      "loss": 0.0912,
+      "step": 1927
+    },
+    {
+      "epoch": 0.4381818181818182,
+      "grad_norm": 3.0211337613716647,
+      "learning_rate": 1.2264860462445128e-06,
+      "loss": 0.1422,
+      "step": 1928
+    },
+    {
+      "epoch": 0.4384090909090909,
+      "grad_norm": 1.4452760602084282,
+      "learning_rate": 1.2264617895767323e-06,
+      "loss": 0.0428,
+      "step": 1929
+    },
+    {
+      "epoch": 0.43863636363636366,
+      "grad_norm": 2.955054720984963,
+      "learning_rate": 1.2264375206440967e-06,
+      "loss": 0.128,
+      "step": 1930
+    },
+    {
+      "epoch": 0.43886363636363634,
+      "grad_norm": 1.513598237043958,
+      "learning_rate": 1.2264132394471007e-06,
+      "loss": 0.0335,
+      "step": 1931
+    },
+    {
+      "epoch": 0.4390909090909091,
+      "grad_norm": 1.3511027067068113,
+      "learning_rate": 1.2263889459862397e-06,
+      "loss": 0.0629,
+      "step": 1932
+    },
+    {
+      "epoch": 0.43931818181818183,
+      "grad_norm": 1.3081623337022468,
+      "learning_rate": 1.2263646402620088e-06,
+      "loss": 0.0752,
+      "step": 1933
+    },
+    {
+      "epoch": 0.4395454545454546,
+      "grad_norm": 1.062681697830198,
+      "learning_rate": 1.2263403222749037e-06,
+      "loss": 0.0493,
+      "step": 1934
+    },
+    {
+      "epoch": 0.43977272727272726,
+      "grad_norm": 3.2584554786625,
+      "learning_rate": 1.2263159920254205e-06,
+      "loss": 0.0847,
+      "step": 1935
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5046342910963246,
+      "learning_rate": 1.226291649514055e-06,
+      "loss": 0.077,
+      "step": 1936
+    },
+    {
+      "epoch": 0.44022727272727274,
+      "grad_norm": 2.8131558481530456,
+      "learning_rate": 1.2262672947413037e-06,
+      "loss": 0.1157,
+      "step": 1937
+    },
+    {
+      "epoch": 0.44045454545454543,
+      "grad_norm": 2.321196306215072,
+      "learning_rate": 1.2262429277076634e-06,
+      "loss": 0.1042,
+      "step": 1938
+    },
+    {
+      "epoch": 0.4406818181818182,
+      "grad_norm": 1.9722249664443345,
+      "learning_rate": 1.2262185484136311e-06,
+      "loss": 0.1145,
+      "step": 1939
+    },
+    {
+      "epoch": 0.4409090909090909,
+      "grad_norm": 1.1162571948014945,
+      "learning_rate": 1.2261941568597033e-06,
+      "loss": 0.035,
+      "step": 1940
+    },
+    {
+      "epoch": 0.44113636363636366,
+      "grad_norm": 2.9158818993702336,
+      "learning_rate": 1.226169753046378e-06,
+      "loss": 0.0993,
+      "step": 1941
+    },
+    {
+      "epoch": 0.44136363636363635,
+      "grad_norm": 1.4954030530111027,
+      "learning_rate": 1.2261453369741528e-06,
+      "loss": 0.0445,
+      "step": 1942
+    },
+    {
+      "epoch": 0.4415909090909091,
+      "grad_norm": 1.4827297509182895,
+      "learning_rate": 1.2261209086435253e-06,
+      "loss": 0.0493,
+      "step": 1943
+    },
+    {
+      "epoch": 0.44181818181818183,
+      "grad_norm": 1.6079153193626776,
+      "learning_rate": 1.2260964680549935e-06,
+      "loss": 0.0816,
+      "step": 1944
+    },
+    {
+      "epoch": 0.4420454545454545,
+      "grad_norm": 1.7241303622727635,
+      "learning_rate": 1.2260720152090564e-06,
+      "loss": 0.0881,
+      "step": 1945
+    },
+    {
+      "epoch": 0.44227272727272726,
+      "grad_norm": 3.963796860580354,
+      "learning_rate": 1.2260475501062121e-06,
+      "loss": 0.1781,
+      "step": 1946
+    },
+    {
+      "epoch": 0.4425,
+      "grad_norm": 3.540831602882329,
+      "learning_rate": 1.2260230727469596e-06,
+      "loss": 0.1746,
+      "step": 1947
+    },
+    {
+      "epoch": 0.44272727272727275,
+      "grad_norm": 1.8852182755750877,
+      "learning_rate": 1.2259985831317984e-06,
+      "loss": 0.0831,
+      "step": 1948
+    },
+    {
+      "epoch": 0.44295454545454543,
+      "grad_norm": 2.9131811804495835,
+      "learning_rate": 1.2259740812612274e-06,
+      "loss": 0.1109,
+      "step": 1949
+    },
+    {
+      "epoch": 0.4431818181818182,
+      "grad_norm": 1.3046875615247273,
+      "learning_rate": 1.2259495671357462e-06,
+      "loss": 0.0497,
+      "step": 1950
+    },
+    {
+      "epoch": 0.4434090909090909,
+      "grad_norm": 2.1417569708659894,
+      "learning_rate": 1.2259250407558553e-06,
+      "loss": 0.0898,
+      "step": 1951
+    },
+    {
+      "epoch": 0.44363636363636366,
+      "grad_norm": 2.0384675383469557,
+      "learning_rate": 1.2259005021220544e-06,
+      "loss": 0.1491,
+      "step": 1952
+    },
+    {
+      "epoch": 0.44386363636363635,
+      "grad_norm": 3.294526276714794,
+      "learning_rate": 1.2258759512348438e-06,
+      "loss": 0.1256,
+      "step": 1953
+    },
+    {
+      "epoch": 0.4440909090909091,
+      "grad_norm": 1.5476269078532074,
+      "learning_rate": 1.2258513880947242e-06,
+      "loss": 0.0913,
+      "step": 1954
+    },
+    {
+      "epoch": 0.44431818181818183,
+      "grad_norm": 1.6989141995705577,
+      "learning_rate": 1.2258268127021967e-06,
+      "loss": 0.0938,
+      "step": 1955
+    },
+    {
+      "epoch": 0.4445454545454545,
+      "grad_norm": 1.5067716748219688,
+      "learning_rate": 1.2258022250577622e-06,
+      "loss": 0.0668,
+      "step": 1956
+    },
+    {
+      "epoch": 0.44477272727272726,
+      "grad_norm": 2.364851626457846,
+      "learning_rate": 1.2257776251619222e-06,
+      "loss": 0.0752,
+      "step": 1957
+    },
+    {
+      "epoch": 0.445,
+      "grad_norm": 2.646630727153874,
+      "learning_rate": 1.2257530130151785e-06,
+      "loss": 0.1385,
+      "step": 1958
+    },
+    {
+      "epoch": 0.44522727272727275,
+      "grad_norm": 2.504249641494026,
+      "learning_rate": 1.2257283886180325e-06,
+      "loss": 0.0954,
+      "step": 1959
+    },
+    {
+      "epoch": 0.44545454545454544,
+      "grad_norm": 2.536404885575026,
+      "learning_rate": 1.2257037519709868e-06,
+      "loss": 0.0777,
+      "step": 1960
+    },
+    {
+      "epoch": 0.4456818181818182,
+      "grad_norm": 3.9890948051056747,
+      "learning_rate": 1.2256791030745434e-06,
+      "loss": 0.167,
+      "step": 1961
+    },
+    {
+      "epoch": 0.4459090909090909,
+      "grad_norm": 1.2267810552056897,
+      "learning_rate": 1.2256544419292055e-06,
+      "loss": 0.0473,
+      "step": 1962
+    },
+    {
+      "epoch": 0.4461363636363636,
+      "grad_norm": 1.8096729454142217,
+      "learning_rate": 1.2256297685354753e-06,
+      "loss": 0.1017,
+      "step": 1963
+    },
+    {
+      "epoch": 0.44636363636363635,
+      "grad_norm": 1.6979617399760742,
+      "learning_rate": 1.2256050828938565e-06,
+      "loss": 0.084,
+      "step": 1964
+    },
+    {
+      "epoch": 0.4465909090909091,
+      "grad_norm": 1.4724725738890685,
+      "learning_rate": 1.225580385004852e-06,
+      "loss": 0.0589,
+      "step": 1965
+    },
+    {
+      "epoch": 0.44681818181818184,
+      "grad_norm": 1.353682068228786,
+      "learning_rate": 1.225555674868966e-06,
+      "loss": 0.0512,
+      "step": 1966
+    },
+    {
+      "epoch": 0.4470454545454545,
+      "grad_norm": 1.7496948383793156,
+      "learning_rate": 1.2255309524867017e-06,
+      "loss": 0.0691,
+      "step": 1967
+    },
+    {
+      "epoch": 0.44727272727272727,
+      "grad_norm": 1.9164986394504269,
+      "learning_rate": 1.225506217858564e-06,
+      "loss": 0.0936,
+      "step": 1968
+    },
+    {
+      "epoch": 0.4475,
+      "grad_norm": 2.380517690341414,
+      "learning_rate": 1.2254814709850566e-06,
+      "loss": 0.1499,
+      "step": 1969
+    },
+    {
+      "epoch": 0.44772727272727275,
+      "grad_norm": 1.325955312735287,
+      "learning_rate": 1.2254567118666845e-06,
+      "loss": 0.0805,
+      "step": 1970
+    },
+    {
+      "epoch": 0.44795454545454544,
+      "grad_norm": 2.130724757940079,
+      "learning_rate": 1.2254319405039524e-06,
+      "loss": 0.1223,
+      "step": 1971
+    },
+    {
+      "epoch": 0.4481818181818182,
+      "grad_norm": 2.23347112157379,
+      "learning_rate": 1.2254071568973656e-06,
+      "loss": 0.0862,
+      "step": 1972
+    },
+    {
+      "epoch": 0.4484090909090909,
+      "grad_norm": 1.8700883332933549,
+      "learning_rate": 1.2253823610474294e-06,
+      "loss": 0.0815,
+      "step": 1973
+    },
+    {
+      "epoch": 0.4486363636363636,
+      "grad_norm": 1.6679141223770326,
+      "learning_rate": 1.2253575529546495e-06,
+      "loss": 0.1228,
+      "step": 1974
+    },
+    {
+      "epoch": 0.44886363636363635,
+      "grad_norm": 1.6789094320463518,
+      "learning_rate": 1.2253327326195316e-06,
+      "loss": 0.0724,
+      "step": 1975
+    },
+    {
+      "epoch": 0.4490909090909091,
+      "grad_norm": 2.6143823113467235,
+      "learning_rate": 1.2253079000425818e-06,
+      "loss": 0.1185,
+      "step": 1976
+    },
+    {
+      "epoch": 0.44931818181818184,
+      "grad_norm": 2.04214728313808,
+      "learning_rate": 1.225283055224307e-06,
+      "loss": 0.0987,
+      "step": 1977
+    },
+    {
+      "epoch": 0.4495454545454545,
+      "grad_norm": 1.6689427218100592,
+      "learning_rate": 1.225258198165213e-06,
+      "loss": 0.0804,
+      "step": 1978
+    },
+    {
+      "epoch": 0.44977272727272727,
+      "grad_norm": 2.0717993473262566,
+      "learning_rate": 1.2252333288658076e-06,
+      "loss": 0.0941,
+      "step": 1979
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.8055755054346867,
+      "learning_rate": 1.2252084473265972e-06,
+      "loss": 0.0799,
+      "step": 1980
+    },
+    {
+      "epoch": 0.45022727272727275,
+      "grad_norm": 2.016009773088547,
+      "learning_rate": 1.2251835535480895e-06,
+      "loss": 0.1268,
+      "step": 1981
+    },
+    {
+      "epoch": 0.45045454545454544,
+      "grad_norm": 2.11345612598454,
+      "learning_rate": 1.225158647530792e-06,
+      "loss": 0.1077,
+      "step": 1982
+    },
+    {
+      "epoch": 0.4506818181818182,
+      "grad_norm": 2.0172975744184782,
+      "learning_rate": 1.2251337292752126e-06,
+      "loss": 0.1268,
+      "step": 1983
+    },
+    {
+      "epoch": 0.4509090909090909,
+      "grad_norm": 2.0488374820539295,
+      "learning_rate": 1.2251087987818598e-06,
+      "loss": 0.1176,
+      "step": 1984
+    },
+    {
+      "epoch": 0.4511363636363636,
+      "grad_norm": 0.8402312013739728,
+      "learning_rate": 1.2250838560512413e-06,
+      "loss": 0.0374,
+      "step": 1985
+    },
+    {
+      "epoch": 0.45136363636363636,
+      "grad_norm": 1.2886034939315518,
+      "learning_rate": 1.2250589010838662e-06,
+      "loss": 0.0691,
+      "step": 1986
+    },
+    {
+      "epoch": 0.4515909090909091,
+      "grad_norm": 2.4368479881205074,
+      "learning_rate": 1.2250339338802433e-06,
+      "loss": 0.119,
+      "step": 1987
+    },
+    {
+      "epoch": 0.45181818181818184,
+      "grad_norm": 2.0892749642039137,
+      "learning_rate": 1.2250089544408817e-06,
+      "loss": 0.084,
+      "step": 1988
+    },
+    {
+      "epoch": 0.45204545454545453,
+      "grad_norm": 1.9667466526731312,
+      "learning_rate": 1.2249839627662904e-06,
+      "loss": 0.0622,
+      "step": 1989
+    },
+    {
+      "epoch": 0.45227272727272727,
+      "grad_norm": 3.0616201055217616,
+      "learning_rate": 1.2249589588569796e-06,
+      "loss": 0.1259,
+      "step": 1990
+    },
+    {
+      "epoch": 0.4525,
+      "grad_norm": 0.7428798845220225,
+      "learning_rate": 1.224933942713459e-06,
+      "loss": 0.0541,
+      "step": 1991
+    },
+    {
+      "epoch": 0.4527272727272727,
+      "grad_norm": 1.1855765992872498,
+      "learning_rate": 1.2249089143362387e-06,
+      "loss": 0.0648,
+      "step": 1992
+    },
+    {
+      "epoch": 0.45295454545454544,
+      "grad_norm": 1.402472218190149,
+      "learning_rate": 1.224883873725829e-06,
+      "loss": 0.0682,
+      "step": 1993
+    },
+    {
+      "epoch": 0.4531818181818182,
+      "grad_norm": 2.244097600473901,
+      "learning_rate": 1.2248588208827405e-06,
+      "loss": 0.0933,
+      "step": 1994
+    },
+    {
+      "epoch": 0.45340909090909093,
+      "grad_norm": 2.183966654878923,
+      "learning_rate": 1.2248337558074842e-06,
+      "loss": 0.0774,
+      "step": 1995
+    },
+    {
+      "epoch": 0.4536363636363636,
+      "grad_norm": 2.0637443629337424,
+      "learning_rate": 1.2248086785005709e-06,
+      "loss": 0.0613,
+      "step": 1996
+    },
+    {
+      "epoch": 0.45386363636363636,
+      "grad_norm": 2.0523131924027482,
+      "learning_rate": 1.2247835889625126e-06,
+      "loss": 0.0926,
+      "step": 1997
+    },
+    {
+      "epoch": 0.4540909090909091,
+      "grad_norm": 1.7088531298576222,
+      "learning_rate": 1.2247584871938204e-06,
+      "loss": 0.0732,
+      "step": 1998
+    },
+    {
+      "epoch": 0.45431818181818184,
+      "grad_norm": 1.375656852792829,
+      "learning_rate": 1.2247333731950062e-06,
+      "loss": 0.0786,
+      "step": 1999
+    },
+    {
+      "epoch": 0.45454545454545453,
+      "grad_norm": 1.67019721713782,
+      "learning_rate": 1.2247082469665823e-06,
+      "loss": 0.1238,
+      "step": 2000
+    },
+    {
+      "epoch": 0.4547727272727273,
+      "grad_norm": 1.6132387143496802,
+      "learning_rate": 1.2246831085090611e-06,
+      "loss": 0.0808,
+      "step": 2001
+    },
+    {
+      "epoch": 0.455,
+      "grad_norm": 2.4323192322120955,
+      "learning_rate": 1.224657957822955e-06,
+      "loss": 0.1316,
+      "step": 2002
+    },
+    {
+      "epoch": 0.4552272727272727,
+      "grad_norm": 2.809654825507686,
+      "learning_rate": 1.2246327949087768e-06,
+      "loss": 0.0536,
+      "step": 2003
+    },
+    {
+      "epoch": 0.45545454545454545,
+      "grad_norm": 2.242241361432057,
+      "learning_rate": 1.2246076197670401e-06,
+      "loss": 0.1171,
+      "step": 2004
+    },
+    {
+      "epoch": 0.4556818181818182,
+      "grad_norm": 1.844084549745781,
+      "learning_rate": 1.2245824323982579e-06,
+      "loss": 0.0843,
+      "step": 2005
+    },
+    {
+      "epoch": 0.45590909090909093,
+      "grad_norm": 2.734228495716328,
+      "learning_rate": 1.2245572328029438e-06,
+      "loss": 0.1213,
+      "step": 2006
+    },
+    {
+      "epoch": 0.4561363636363636,
+      "grad_norm": 3.6961566475847962,
+      "learning_rate": 1.2245320209816116e-06,
+      "loss": 0.1535,
+      "step": 2007
+    },
+    {
+      "epoch": 0.45636363636363636,
+      "grad_norm": 2.0928859721519957,
+      "learning_rate": 1.2245067969347758e-06,
+      "loss": 0.0992,
+      "step": 2008
+    },
+    {
+      "epoch": 0.4565909090909091,
+      "grad_norm": 3.2960714904019857,
+      "learning_rate": 1.2244815606629504e-06,
+      "loss": 0.141,
+      "step": 2009
+    },
+    {
+      "epoch": 0.45681818181818185,
+      "grad_norm": 1.5646731274325274,
+      "learning_rate": 1.2244563121666501e-06,
+      "loss": 0.0708,
+      "step": 2010
+    },
+    {
+      "epoch": 0.45704545454545453,
+      "grad_norm": 1.882394205725334,
+      "learning_rate": 1.22443105144639e-06,
+      "loss": 0.0746,
+      "step": 2011
+    },
+    {
+      "epoch": 0.4572727272727273,
+      "grad_norm": 2.2974569053761287,
+      "learning_rate": 1.2244057785026847e-06,
+      "loss": 0.1126,
+      "step": 2012
+    },
+    {
+      "epoch": 0.4575,
+      "grad_norm": 3.4729498123147544,
+      "learning_rate": 1.22438049333605e-06,
+      "loss": 0.1713,
+      "step": 2013
+    },
+    {
+      "epoch": 0.4577272727272727,
+      "grad_norm": 1.6902338846648621,
+      "learning_rate": 1.2243551959470013e-06,
+      "loss": 0.0623,
+      "step": 2014
+    },
+    {
+      "epoch": 0.45795454545454545,
+      "grad_norm": 2.244685196416379,
+      "learning_rate": 1.2243298863360547e-06,
+      "loss": 0.1653,
+      "step": 2015
+    },
+    {
+      "epoch": 0.4581818181818182,
+      "grad_norm": 1.8027561589910048,
+      "learning_rate": 1.224304564503726e-06,
+      "loss": 0.1016,
+      "step": 2016
+    },
+    {
+      "epoch": 0.45840909090909093,
+      "grad_norm": 2.1087282680152533,
+      "learning_rate": 1.2242792304505317e-06,
+      "loss": 0.148,
+      "step": 2017
+    },
+    {
+      "epoch": 0.4586363636363636,
+      "grad_norm": 2.411572601517144,
+      "learning_rate": 1.2242538841769883e-06,
+      "loss": 0.1173,
+      "step": 2018
+    },
+    {
+      "epoch": 0.45886363636363636,
+      "grad_norm": 2.852733703338483,
+      "learning_rate": 1.2242285256836129e-06,
+      "loss": 0.0999,
+      "step": 2019
+    },
+    {
+      "epoch": 0.4590909090909091,
+      "grad_norm": 2.313803038153825,
+      "learning_rate": 1.2242031549709224e-06,
+      "loss": 0.1451,
+      "step": 2020
+    },
+    {
+      "epoch": 0.4593181818181818,
+      "grad_norm": 1.6978976201151466,
+      "learning_rate": 1.224177772039434e-06,
+      "loss": 0.0613,
+      "step": 2021
+    },
+    {
+      "epoch": 0.45954545454545453,
+      "grad_norm": 2.9067703662554765,
+      "learning_rate": 1.2241523768896657e-06,
+      "loss": 0.1491,
+      "step": 2022
+    },
+    {
+      "epoch": 0.4597727272727273,
+      "grad_norm": 2.608265937919946,
+      "learning_rate": 1.2241269695221352e-06,
+      "loss": 0.164,
+      "step": 2023
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4013611196179905,
+      "learning_rate": 1.2241015499373606e-06,
+      "loss": 0.0612,
+      "step": 2024
+    },
+    {
+      "epoch": 0.4602272727272727,
+      "grad_norm": 1.8112424681338959,
+      "learning_rate": 1.22407611813586e-06,
+      "loss": 0.1193,
+      "step": 2025
+    },
+    {
+      "epoch": 0.46045454545454545,
+      "grad_norm": 1.3681144895615605,
+      "learning_rate": 1.2240506741181523e-06,
+      "loss": 0.0554,
+      "step": 2026
+    },
+    {
+      "epoch": 0.4606818181818182,
+      "grad_norm": 0.8975635824616581,
+      "learning_rate": 1.2240252178847564e-06,
+      "loss": 0.0354,
+      "step": 2027
+    },
+    {
+      "epoch": 0.46090909090909093,
+      "grad_norm": 1.3516318818298965,
+      "learning_rate": 1.2239997494361911e-06,
+      "loss": 0.0769,
+      "step": 2028
+    },
+    {
+      "epoch": 0.4611363636363636,
+      "grad_norm": 1.5548119260295274,
+      "learning_rate": 1.223974268772976e-06,
+      "loss": 0.0846,
+      "step": 2029
+    },
+    {
+      "epoch": 0.46136363636363636,
+      "grad_norm": 2.2731770044205404,
+      "learning_rate": 1.2239487758956302e-06,
+      "loss": 0.1352,
+      "step": 2030
+    },
+    {
+      "epoch": 0.4615909090909091,
+      "grad_norm": 1.3908802784008327,
+      "learning_rate": 1.2239232708046745e-06,
+      "loss": 0.0623,
+      "step": 2031
+    },
+    {
+      "epoch": 0.4618181818181818,
+      "grad_norm": 1.5626752270256947,
+      "learning_rate": 1.223897753500628e-06,
+      "loss": 0.0699,
+      "step": 2032
+    },
+    {
+      "epoch": 0.46204545454545454,
+      "grad_norm": 1.8860600728992667,
+      "learning_rate": 1.2238722239840117e-06,
+      "loss": 0.0885,
+      "step": 2033
+    },
+    {
+      "epoch": 0.4622727272727273,
+      "grad_norm": 1.6662685798571717,
+      "learning_rate": 1.2238466822553459e-06,
+      "loss": 0.0999,
+      "step": 2034
+    },
+    {
+      "epoch": 0.4625,
+      "grad_norm": 1.4620257242510641,
+      "learning_rate": 1.2238211283151516e-06,
+      "loss": 0.0456,
+      "step": 2035
+    },
+    {
+      "epoch": 0.4627272727272727,
+      "grad_norm": 1.854044680465104,
+      "learning_rate": 1.2237955621639496e-06,
+      "loss": 0.0874,
+      "step": 2036
+    },
+    {
+      "epoch": 0.46295454545454545,
+      "grad_norm": 1.7449616718985754,
+      "learning_rate": 1.2237699838022614e-06,
+      "loss": 0.1151,
+      "step": 2037
+    },
+    {
+      "epoch": 0.4631818181818182,
+      "grad_norm": 1.6043654525770215,
+      "learning_rate": 1.223744393230609e-06,
+      "loss": 0.0655,
+      "step": 2038
+    },
+    {
+      "epoch": 0.4634090909090909,
+      "grad_norm": 2.156504486394697,
+      "learning_rate": 1.2237187904495134e-06,
+      "loss": 0.0768,
+      "step": 2039
+    },
+    {
+      "epoch": 0.4636363636363636,
+      "grad_norm": 1.3773757484755011,
+      "learning_rate": 1.2236931754594975e-06,
+      "loss": 0.0546,
+      "step": 2040
+    },
+    {
+      "epoch": 0.46386363636363637,
+      "grad_norm": 2.1423500396244086,
+      "learning_rate": 1.223667548261083e-06,
+      "loss": 0.095,
+      "step": 2041
+    },
+    {
+      "epoch": 0.4640909090909091,
+      "grad_norm": 0.5968540254331165,
+      "learning_rate": 1.223641908854793e-06,
+      "loss": 0.0325,
+      "step": 2042
+    },
+    {
+      "epoch": 0.4643181818181818,
+      "grad_norm": 1.6326657950649373,
+      "learning_rate": 1.2236162572411499e-06,
+      "loss": 0.0533,
+      "step": 2043
+    },
+    {
+      "epoch": 0.46454545454545454,
+      "grad_norm": 2.022565238549652,
+      "learning_rate": 1.2235905934206771e-06,
+      "loss": 0.0877,
+      "step": 2044
+    },
+    {
+      "epoch": 0.4647727272727273,
+      "grad_norm": 2.0291718560481278,
+      "learning_rate": 1.2235649173938976e-06,
+      "loss": 0.0834,
+      "step": 2045
+    },
+    {
+      "epoch": 0.465,
+      "grad_norm": 1.495107649387907,
+      "learning_rate": 1.2235392291613353e-06,
+      "loss": 0.0595,
+      "step": 2046
+    },
+    {
+      "epoch": 0.4652272727272727,
+      "grad_norm": 2.42604654900152,
+      "learning_rate": 1.223513528723514e-06,
+      "loss": 0.0773,
+      "step": 2047
+    },
+    {
+      "epoch": 0.46545454545454545,
+      "grad_norm": 2.3172413083839123,
+      "learning_rate": 1.2234878160809576e-06,
+      "loss": 0.065,
+      "step": 2048
+    },
+    {
+      "epoch": 0.4656818181818182,
+      "grad_norm": 2.057406539132907,
+      "learning_rate": 1.2234620912341905e-06,
+      "loss": 0.1141,
+      "step": 2049
+    },
+    {
+      "epoch": 0.4659090909090909,
+      "grad_norm": 1.9651325260633614,
+      "learning_rate": 1.2234363541837375e-06,
+      "loss": 0.0646,
+      "step": 2050
+    },
+    {
+      "epoch": 0.4661363636363636,
+      "grad_norm": 2.2648363224642023,
+      "learning_rate": 1.2234106049301228e-06,
+      "loss": 0.093,
+      "step": 2051
+    },
+    {
+      "epoch": 0.46636363636363637,
+      "grad_norm": 1.9673378058406235,
+      "learning_rate": 1.2233848434738721e-06,
+      "loss": 0.0444,
+      "step": 2052
+    },
+    {
+      "epoch": 0.4665909090909091,
+      "grad_norm": 3.0053562513766443,
+      "learning_rate": 1.2233590698155107e-06,
+      "loss": 0.1624,
+      "step": 2053
+    },
+    {
+      "epoch": 0.4668181818181818,
+      "grad_norm": 1.3950740379504807,
+      "learning_rate": 1.2233332839555636e-06,
+      "loss": 0.0512,
+      "step": 2054
+    },
+    {
+      "epoch": 0.46704545454545454,
+      "grad_norm": 1.7384846404939334,
+      "learning_rate": 1.2233074858945572e-06,
+      "loss": 0.1021,
+      "step": 2055
+    },
+    {
+      "epoch": 0.4672727272727273,
+      "grad_norm": 2.2761240829901133,
+      "learning_rate": 1.2232816756330173e-06,
+      "loss": 0.1359,
+      "step": 2056
+    },
+    {
+      "epoch": 0.4675,
+      "grad_norm": 3.0250757879962182,
+      "learning_rate": 1.2232558531714702e-06,
+      "loss": 0.0966,
+      "step": 2057
+    },
+    {
+      "epoch": 0.4677272727272727,
+      "grad_norm": 1.2774055419176162,
+      "learning_rate": 1.2232300185104428e-06,
+      "loss": 0.0602,
+      "step": 2058
+    },
+    {
+      "epoch": 0.46795454545454546,
+      "grad_norm": 1.9023951054909276,
+      "learning_rate": 1.2232041716504614e-06,
+      "loss": 0.1179,
+      "step": 2059
+    },
+    {
+      "epoch": 0.4681818181818182,
+      "grad_norm": 2.0573311903665297,
+      "learning_rate": 1.2231783125920535e-06,
+      "loss": 0.1076,
+      "step": 2060
+    },
+    {
+      "epoch": 0.4684090909090909,
+      "grad_norm": 2.301960510783131,
+      "learning_rate": 1.223152441335746e-06,
+      "loss": 0.1043,
+      "step": 2061
+    },
+    {
+      "epoch": 0.46863636363636363,
+      "grad_norm": 2.1393239097895873,
+      "learning_rate": 1.223126557882067e-06,
+      "loss": 0.0959,
+      "step": 2062
+    },
+    {
+      "epoch": 0.46886363636363637,
+      "grad_norm": 1.9004619794066873,
+      "learning_rate": 1.2231006622315437e-06,
+      "loss": 0.1218,
+      "step": 2063
+    },
+    {
+      "epoch": 0.4690909090909091,
+      "grad_norm": 1.6780884975636299,
+      "learning_rate": 1.2230747543847047e-06,
+      "loss": 0.0824,
+      "step": 2064
+    },
+    {
+      "epoch": 0.4693181818181818,
+      "grad_norm": 1.3781850315231052,
+      "learning_rate": 1.2230488343420777e-06,
+      "loss": 0.0667,
+      "step": 2065
+    },
+    {
+      "epoch": 0.46954545454545454,
+      "grad_norm": 2.00303124273341,
+      "learning_rate": 1.223022902104192e-06,
+      "loss": 0.128,
+      "step": 2066
+    },
+    {
+      "epoch": 0.4697727272727273,
+      "grad_norm": 1.6638584442200872,
+      "learning_rate": 1.2229969576715758e-06,
+      "loss": 0.0651,
+      "step": 2067
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.68787318448008,
+      "learning_rate": 1.2229710010447584e-06,
+      "loss": 0.1057,
+      "step": 2068
+    },
+    {
+      "epoch": 0.4702272727272727,
+      "grad_norm": 1.6597983893898522,
+      "learning_rate": 1.2229450322242692e-06,
+      "loss": 0.076,
+      "step": 2069
+    },
+    {
+      "epoch": 0.47045454545454546,
+      "grad_norm": 1.6573052051962283,
+      "learning_rate": 1.2229190512106376e-06,
+      "loss": 0.0747,
+      "step": 2070
+    },
+    {
+      "epoch": 0.4706818181818182,
+      "grad_norm": 1.8051192342459983,
+      "learning_rate": 1.2228930580043931e-06,
+      "loss": 0.0927,
+      "step": 2071
+    },
+    {
+      "epoch": 0.4709090909090909,
+      "grad_norm": 3.17248402590761,
+      "learning_rate": 1.2228670526060664e-06,
+      "loss": 0.1635,
+      "step": 2072
+    },
+    {
+      "epoch": 0.47113636363636363,
+      "grad_norm": 2.307842726947568,
+      "learning_rate": 1.2228410350161873e-06,
+      "loss": 0.136,
+      "step": 2073
+    },
+    {
+      "epoch": 0.4713636363636364,
+      "grad_norm": 2.5158964487801163,
+      "learning_rate": 1.2228150052352866e-06,
+      "loss": 0.1184,
+      "step": 2074
+    },
+    {
+      "epoch": 0.4715909090909091,
+      "grad_norm": 1.3614677435386413,
+      "learning_rate": 1.222788963263895e-06,
+      "loss": 0.0838,
+      "step": 2075
+    },
+    {
+      "epoch": 0.4718181818181818,
+      "grad_norm": 1.6931603811619722,
+      "learning_rate": 1.2227629091025437e-06,
+      "loss": 0.0478,
+      "step": 2076
+    },
+    {
+      "epoch": 0.47204545454545455,
+      "grad_norm": 1.2564113085156905,
+      "learning_rate": 1.2227368427517635e-06,
+      "loss": 0.0521,
+      "step": 2077
+    },
+    {
+      "epoch": 0.4722727272727273,
+      "grad_norm": 3.424383695651977,
+      "learning_rate": 1.2227107642120865e-06,
+      "loss": 0.1096,
+      "step": 2078
+    },
+    {
+      "epoch": 0.4725,
+      "grad_norm": 2.588060456136328,
+      "learning_rate": 1.2226846734840441e-06,
+      "loss": 0.0897,
+      "step": 2079
+    },
+    {
+      "epoch": 0.4727272727272727,
+      "grad_norm": 2.2107423258661707,
+      "learning_rate": 1.2226585705681685e-06,
+      "loss": 0.1354,
+      "step": 2080
+    },
+    {
+      "epoch": 0.47295454545454546,
+      "grad_norm": 2.0790615179890466,
+      "learning_rate": 1.2226324554649921e-06,
+      "loss": 0.1011,
+      "step": 2081
+    },
+    {
+      "epoch": 0.4731818181818182,
+      "grad_norm": 2.003608375374061,
+      "learning_rate": 1.2226063281750472e-06,
+      "loss": 0.1032,
+      "step": 2082
+    },
+    {
+      "epoch": 0.4734090909090909,
+      "grad_norm": 0.8979023237139657,
+      "learning_rate": 1.2225801886988666e-06,
+      "loss": 0.0417,
+      "step": 2083
+    },
+    {
+      "epoch": 0.47363636363636363,
+      "grad_norm": 1.4945785215359697,
+      "learning_rate": 1.2225540370369836e-06,
+      "loss": 0.0747,
+      "step": 2084
+    },
+    {
+      "epoch": 0.4738636363636364,
+      "grad_norm": 1.9889251061549393,
+      "learning_rate": 1.2225278731899312e-06,
+      "loss": 0.0676,
+      "step": 2085
+    },
+    {
+      "epoch": 0.4740909090909091,
+      "grad_norm": 1.7362551418134395,
+      "learning_rate": 1.222501697158243e-06,
+      "loss": 0.1001,
+      "step": 2086
+    },
+    {
+      "epoch": 0.4743181818181818,
+      "grad_norm": 2.08160772388357,
+      "learning_rate": 1.2224755089424527e-06,
+      "loss": 0.0721,
+      "step": 2087
+    },
+    {
+      "epoch": 0.47454545454545455,
+      "grad_norm": 2.3814187786880088,
+      "learning_rate": 1.2224493085430947e-06,
+      "loss": 0.162,
+      "step": 2088
+    },
+    {
+      "epoch": 0.4747727272727273,
+      "grad_norm": 1.7409998027116875,
+      "learning_rate": 1.222423095960703e-06,
+      "loss": 0.0687,
+      "step": 2089
+    },
+    {
+      "epoch": 0.475,
+      "grad_norm": 1.9142854431163119,
+      "learning_rate": 1.222396871195812e-06,
+      "loss": 0.0654,
+      "step": 2090
+    },
+    {
+      "epoch": 0.4752272727272727,
+      "grad_norm": 1.7622520113234186,
+      "learning_rate": 1.2223706342489565e-06,
+      "loss": 0.1135,
+      "step": 2091
+    },
+    {
+      "epoch": 0.47545454545454546,
+      "grad_norm": 2.1747961876585107,
+      "learning_rate": 1.2223443851206716e-06,
+      "loss": 0.1148,
+      "step": 2092
+    },
+    {
+      "epoch": 0.4756818181818182,
+      "grad_norm": 1.5840545078208916,
+      "learning_rate": 1.2223181238114926e-06,
+      "loss": 0.0927,
+      "step": 2093
+    },
+    {
+      "epoch": 0.4759090909090909,
+      "grad_norm": 1.8671979268660546,
+      "learning_rate": 1.222291850321955e-06,
+      "loss": 0.0962,
+      "step": 2094
+    },
+    {
+      "epoch": 0.47613636363636364,
+      "grad_norm": 1.2716676343422189,
+      "learning_rate": 1.2222655646525944e-06,
+      "loss": 0.0796,
+      "step": 2095
+    },
+    {
+      "epoch": 0.4763636363636364,
+      "grad_norm": 1.8957697226963222,
+      "learning_rate": 1.222239266803947e-06,
+      "loss": 0.0832,
+      "step": 2096
+    },
+    {
+      "epoch": 0.47659090909090907,
+      "grad_norm": 2.3985007558560234,
+      "learning_rate": 1.2222129567765491e-06,
+      "loss": 0.113,
+      "step": 2097
+    },
+    {
+      "epoch": 0.4768181818181818,
+      "grad_norm": 1.4507778868691785,
+      "learning_rate": 1.2221866345709372e-06,
+      "loss": 0.0446,
+      "step": 2098
+    },
+    {
+      "epoch": 0.47704545454545455,
+      "grad_norm": 2.0952734263470756,
+      "learning_rate": 1.2221603001876478e-06,
+      "loss": 0.1376,
+      "step": 2099
+    },
+    {
+      "epoch": 0.4772727272727273,
+      "grad_norm": 1.8824680239891596,
+      "learning_rate": 1.2221339536272183e-06,
+      "loss": 0.0836,
+      "step": 2100
+    },
+    {
+      "epoch": 0.4775,
+      "grad_norm": 1.266220169210467,
+      "learning_rate": 1.2221075948901856e-06,
+      "loss": 0.0465,
+      "step": 2101
+    },
+    {
+      "epoch": 0.4777272727272727,
+      "grad_norm": 1.5910249498822455,
+      "learning_rate": 1.2220812239770873e-06,
+      "loss": 0.0515,
+      "step": 2102
+    },
+    {
+      "epoch": 0.47795454545454547,
+      "grad_norm": 1.6920844567584827,
+      "learning_rate": 1.2220548408884611e-06,
+      "loss": 0.1077,
+      "step": 2103
+    },
+    {
+      "epoch": 0.4781818181818182,
+      "grad_norm": 2.1862866530040312,
+      "learning_rate": 1.2220284456248453e-06,
+      "loss": 0.0959,
+      "step": 2104
+    },
+    {
+      "epoch": 0.4784090909090909,
+      "grad_norm": 1.683300088993308,
+      "learning_rate": 1.2220020381867776e-06,
+      "loss": 0.0756,
+      "step": 2105
+    },
+    {
+      "epoch": 0.47863636363636364,
+      "grad_norm": 1.4765089705978804,
+      "learning_rate": 1.221975618574797e-06,
+      "loss": 0.0674,
+      "step": 2106
+    },
+    {
+      "epoch": 0.4788636363636364,
+      "grad_norm": 3.6644171564155728,
+      "learning_rate": 1.221949186789442e-06,
+      "loss": 0.1458,
+      "step": 2107
+    },
+    {
+      "epoch": 0.47909090909090907,
+      "grad_norm": 1.9209187086003014,
+      "learning_rate": 1.2219227428312517e-06,
+      "loss": 0.0961,
+      "step": 2108
+    },
+    {
+      "epoch": 0.4793181818181818,
+      "grad_norm": 1.507318405509234,
+      "learning_rate": 1.2218962867007654e-06,
+      "loss": 0.0772,
+      "step": 2109
+    },
+    {
+      "epoch": 0.47954545454545455,
+      "grad_norm": 1.842638354780919,
+      "learning_rate": 1.2218698183985223e-06,
+      "loss": 0.1062,
+      "step": 2110
+    },
+    {
+      "epoch": 0.4797727272727273,
+      "grad_norm": 1.9483572132621336,
+      "learning_rate": 1.2218433379250623e-06,
+      "loss": 0.0838,
+      "step": 2111
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.250842374452262,
+      "learning_rate": 1.2218168452809256e-06,
+      "loss": 0.0538,
+      "step": 2112
+    },
+    {
+      "epoch": 0.4802272727272727,
+      "grad_norm": 1.8885326168404062,
+      "learning_rate": 1.221790340466652e-06,
+      "loss": 0.0985,
+      "step": 2113
+    },
+    {
+      "epoch": 0.48045454545454547,
+      "grad_norm": 1.4030715215193765,
+      "learning_rate": 1.2217638234827822e-06,
+      "loss": 0.0633,
+      "step": 2114
+    },
+    {
+      "epoch": 0.48068181818181815,
+      "grad_norm": 1.7457219854690997,
+      "learning_rate": 1.221737294329857e-06,
+      "loss": 0.0974,
+      "step": 2115
+    },
+    {
+      "epoch": 0.4809090909090909,
+      "grad_norm": 1.3137777306600822,
+      "learning_rate": 1.2217107530084174e-06,
+      "loss": 0.0511,
+      "step": 2116
+    },
+    {
+      "epoch": 0.48113636363636364,
+      "grad_norm": 1.9641856816347008,
+      "learning_rate": 1.2216841995190042e-06,
+      "loss": 0.0832,
+      "step": 2117
+    },
+    {
+      "epoch": 0.4813636363636364,
+      "grad_norm": 1.6588180364800222,
+      "learning_rate": 1.2216576338621595e-06,
+      "loss": 0.0936,
+      "step": 2118
+    },
+    {
+      "epoch": 0.48159090909090907,
+      "grad_norm": 1.944918436359931,
+      "learning_rate": 1.2216310560384247e-06,
+      "loss": 0.0722,
+      "step": 2119
+    },
+    {
+      "epoch": 0.4818181818181818,
+      "grad_norm": 2.125656563867208,
+      "learning_rate": 1.2216044660483416e-06,
+      "loss": 0.1148,
+      "step": 2120
+    },
+    {
+      "epoch": 0.48204545454545455,
+      "grad_norm": 1.2872433041104554,
+      "learning_rate": 1.2215778638924527e-06,
+      "loss": 0.074,
+      "step": 2121
+    },
+    {
+      "epoch": 0.4822727272727273,
+      "grad_norm": 1.6198015725908101,
+      "learning_rate": 1.2215512495713006e-06,
+      "loss": 0.0694,
+      "step": 2122
+    },
+    {
+      "epoch": 0.4825,
+      "grad_norm": 1.6395877547842843,
+      "learning_rate": 1.2215246230854274e-06,
+      "loss": 0.0845,
+      "step": 2123
+    },
+    {
+      "epoch": 0.4827272727272727,
+      "grad_norm": 1.538367320349972,
+      "learning_rate": 1.2214979844353766e-06,
+      "loss": 0.0762,
+      "step": 2124
+    },
+    {
+      "epoch": 0.48295454545454547,
+      "grad_norm": 2.427329790261349,
+      "learning_rate": 1.2214713336216914e-06,
+      "loss": 0.0694,
+      "step": 2125
+    },
+    {
+      "epoch": 0.48318181818181816,
+      "grad_norm": 2.0959429817065565,
+      "learning_rate": 1.221444670644915e-06,
+      "loss": 0.1023,
+      "step": 2126
+    },
+    {
+      "epoch": 0.4834090909090909,
+      "grad_norm": 1.2073042931954405,
+      "learning_rate": 1.2214179955055912e-06,
+      "loss": 0.0537,
+      "step": 2127
+    },
+    {
+      "epoch": 0.48363636363636364,
+      "grad_norm": 1.3714255346617068,
+      "learning_rate": 1.2213913082042638e-06,
+      "loss": 0.0884,
+      "step": 2128
+    },
+    {
+      "epoch": 0.4838636363636364,
+      "grad_norm": 2.0196309897025078,
+      "learning_rate": 1.2213646087414774e-06,
+      "loss": 0.1049,
+      "step": 2129
+    },
+    {
+      "epoch": 0.48409090909090907,
+      "grad_norm": 1.5714041339765408,
+      "learning_rate": 1.2213378971177761e-06,
+      "loss": 0.0473,
+      "step": 2130
+    },
+    {
+      "epoch": 0.4843181818181818,
+      "grad_norm": 2.0535118963983754,
+      "learning_rate": 1.221311173333705e-06,
+      "loss": 0.0828,
+      "step": 2131
+    },
+    {
+      "epoch": 0.48454545454545456,
+      "grad_norm": 1.8557093618398435,
+      "learning_rate": 1.2212844373898084e-06,
+      "loss": 0.0656,
+      "step": 2132
+    },
+    {
+      "epoch": 0.4847727272727273,
+      "grad_norm": 2.293906468558044,
+      "learning_rate": 1.2212576892866317e-06,
+      "loss": 0.0654,
+      "step": 2133
+    },
+    {
+      "epoch": 0.485,
+      "grad_norm": 1.8319038470770035,
+      "learning_rate": 1.221230929024721e-06,
+      "loss": 0.0582,
+      "step": 2134
+    },
+    {
+      "epoch": 0.48522727272727273,
+      "grad_norm": 1.836663314598973,
+      "learning_rate": 1.2212041566046211e-06,
+      "loss": 0.0562,
+      "step": 2135
+    },
+    {
+      "epoch": 0.48545454545454547,
+      "grad_norm": 1.4260935788208016,
+      "learning_rate": 1.2211773720268784e-06,
+      "loss": 0.064,
+      "step": 2136
+    },
+    {
+      "epoch": 0.48568181818181816,
+      "grad_norm": 2.3866403797389557,
+      "learning_rate": 1.221150575292039e-06,
+      "loss": 0.093,
+      "step": 2137
+    },
+    {
+      "epoch": 0.4859090909090909,
+      "grad_norm": 1.5994725758155657,
+      "learning_rate": 1.2211237664006494e-06,
+      "loss": 0.1193,
+      "step": 2138
+    },
+    {
+      "epoch": 0.48613636363636364,
+      "grad_norm": 1.8436487788388003,
+      "learning_rate": 1.221096945353256e-06,
+      "loss": 0.0565,
+      "step": 2139
+    },
+    {
+      "epoch": 0.4863636363636364,
+      "grad_norm": 2.071351792841336,
+      "learning_rate": 1.2210701121504062e-06,
+      "loss": 0.0717,
+      "step": 2140
+    },
+    {
+      "epoch": 0.4865909090909091,
+      "grad_norm": 1.9841562353867281,
+      "learning_rate": 1.2210432667926467e-06,
+      "loss": 0.1112,
+      "step": 2141
+    },
+    {
+      "epoch": 0.4868181818181818,
+      "grad_norm": 2.8324828827087294,
+      "learning_rate": 1.2210164092805252e-06,
+      "loss": 0.1073,
+      "step": 2142
+    },
+    {
+      "epoch": 0.48704545454545456,
+      "grad_norm": 2.2672493946896717,
+      "learning_rate": 1.2209895396145894e-06,
+      "loss": 0.1046,
+      "step": 2143
+    },
+    {
+      "epoch": 0.48727272727272725,
+      "grad_norm": 1.5941676157307076,
+      "learning_rate": 1.220962657795387e-06,
+      "loss": 0.0976,
+      "step": 2144
+    },
+    {
+      "epoch": 0.4875,
+      "grad_norm": 1.7866225276196575,
+      "learning_rate": 1.2209357638234664e-06,
+      "loss": 0.0663,
+      "step": 2145
+    },
+    {
+      "epoch": 0.48772727272727273,
+      "grad_norm": 2.0198507951602784,
+      "learning_rate": 1.2209088576993757e-06,
+      "loss": 0.1385,
+      "step": 2146
+    },
+    {
+      "epoch": 0.4879545454545455,
+      "grad_norm": 2.6278090422971414,
+      "learning_rate": 1.220881939423664e-06,
+      "loss": 0.1625,
+      "step": 2147
+    },
+    {
+      "epoch": 0.48818181818181816,
+      "grad_norm": 1.8794441950819467,
+      "learning_rate": 1.22085500899688e-06,
+      "loss": 0.1098,
+      "step": 2148
+    },
+    {
+      "epoch": 0.4884090909090909,
+      "grad_norm": 1.5213706282397552,
+      "learning_rate": 1.2208280664195727e-06,
+      "loss": 0.0727,
+      "step": 2149
+    },
+    {
+      "epoch": 0.48863636363636365,
+      "grad_norm": 2.0038565825237593,
+      "learning_rate": 1.2208011116922918e-06,
+      "loss": 0.0865,
+      "step": 2150
+    },
+    {
+      "epoch": 0.4888636363636364,
+      "grad_norm": 1.9905472167120952,
+      "learning_rate": 1.2207741448155867e-06,
+      "loss": 0.0863,
+      "step": 2151
+    },
+    {
+      "epoch": 0.4890909090909091,
+      "grad_norm": 2.732978657948083,
+      "learning_rate": 1.2207471657900074e-06,
+      "loss": 0.1024,
+      "step": 2152
+    },
+    {
+      "epoch": 0.4893181818181818,
+      "grad_norm": 1.80716402703433,
+      "learning_rate": 1.220720174616104e-06,
+      "loss": 0.0799,
+      "step": 2153
+    },
+    {
+      "epoch": 0.48954545454545456,
+      "grad_norm": 2.3727737799554682,
+      "learning_rate": 1.220693171294427e-06,
+      "loss": 0.0943,
+      "step": 2154
+    },
+    {
+      "epoch": 0.48977272727272725,
+      "grad_norm": 1.6754346659362618,
+      "learning_rate": 1.220666155825527e-06,
+      "loss": 0.0812,
+      "step": 2155
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0104421967074875,
+      "learning_rate": 1.220639128209955e-06,
+      "loss": 0.102,
+      "step": 2156
+    },
+    {
+      "epoch": 0.49022727272727273,
+      "grad_norm": 2.139754496161981,
+      "learning_rate": 1.220612088448262e-06,
+      "loss": 0.1183,
+      "step": 2157
+    },
+    {
+      "epoch": 0.4904545454545455,
+      "grad_norm": 2.0339634866656007,
+      "learning_rate": 1.2205850365409992e-06,
+      "loss": 0.0962,
+      "step": 2158
+    },
+    {
+      "epoch": 0.49068181818181816,
+      "grad_norm": 1.4167854241263063,
+      "learning_rate": 1.2205579724887185e-06,
+      "loss": 0.0446,
+      "step": 2159
+    },
+    {
+      "epoch": 0.4909090909090909,
+      "grad_norm": 1.990265496354692,
+      "learning_rate": 1.2205308962919719e-06,
+      "loss": 0.1137,
+      "step": 2160
+    },
+    {
+      "epoch": 0.49113636363636365,
+      "grad_norm": 2.610733360469573,
+      "learning_rate": 1.2205038079513113e-06,
+      "loss": 0.0935,
+      "step": 2161
+    },
+    {
+      "epoch": 0.4913636363636364,
+      "grad_norm": 1.9608955402148542,
+      "learning_rate": 1.2204767074672891e-06,
+      "loss": 0.0734,
+      "step": 2162
+    },
+    {
+      "epoch": 0.4915909090909091,
+      "grad_norm": 2.650520953029873,
+      "learning_rate": 1.220449594840458e-06,
+      "loss": 0.1794,
+      "step": 2163
+    },
+    {
+      "epoch": 0.4918181818181818,
+      "grad_norm": 3.325736032318496,
+      "learning_rate": 1.2204224700713708e-06,
+      "loss": 0.1484,
+      "step": 2164
+    },
+    {
+      "epoch": 0.49204545454545456,
+      "grad_norm": 2.1390730213998523,
+      "learning_rate": 1.2203953331605806e-06,
+      "loss": 0.1055,
+      "step": 2165
+    },
+    {
+      "epoch": 0.49227272727272725,
+      "grad_norm": 2.198428640384916,
+      "learning_rate": 1.2203681841086409e-06,
+      "loss": 0.0933,
+      "step": 2166
+    },
+    {
+      "epoch": 0.4925,
+      "grad_norm": 2.871560218154371,
+      "learning_rate": 1.2203410229161052e-06,
+      "loss": 0.1115,
+      "step": 2167
+    },
+    {
+      "epoch": 0.49272727272727274,
+      "grad_norm": 1.9778534503922767,
+      "learning_rate": 1.2203138495835277e-06,
+      "loss": 0.081,
+      "step": 2168
+    },
+    {
+      "epoch": 0.4929545454545455,
+      "grad_norm": 1.4433774825330148,
+      "learning_rate": 1.2202866641114618e-06,
+      "loss": 0.0915,
+      "step": 2169
+    },
+    {
+      "epoch": 0.49318181818181817,
+      "grad_norm": 1.4031880535851808,
+      "learning_rate": 1.2202594665004628e-06,
+      "loss": 0.0669,
+      "step": 2170
+    },
+    {
+      "epoch": 0.4934090909090909,
+      "grad_norm": 2.1058098936363896,
+      "learning_rate": 1.2202322567510843e-06,
+      "loss": 0.0747,
+      "step": 2171
+    },
+    {
+      "epoch": 0.49363636363636365,
+      "grad_norm": 1.4507712762768656,
+      "learning_rate": 1.220205034863882e-06,
+      "loss": 0.0604,
+      "step": 2172
+    },
+    {
+      "epoch": 0.49386363636363634,
+      "grad_norm": 2.335914680646208,
+      "learning_rate": 1.2201778008394104e-06,
+      "loss": 0.1211,
+      "step": 2173
+    },
+    {
+      "epoch": 0.4940909090909091,
+      "grad_norm": 2.270599575918373,
+      "learning_rate": 1.2201505546782252e-06,
+      "loss": 0.0701,
+      "step": 2174
+    },
+    {
+      "epoch": 0.4943181818181818,
+      "grad_norm": 0.43798335186556875,
+      "learning_rate": 1.2201232963808817e-06,
+      "loss": 0.0235,
+      "step": 2175
+    },
+    {
+      "epoch": 0.49454545454545457,
+      "grad_norm": 2.2867611692596546,
+      "learning_rate": 1.2200960259479362e-06,
+      "loss": 0.0912,
+      "step": 2176
+    },
+    {
+      "epoch": 0.49477272727272725,
+      "grad_norm": 1.5077810085566947,
+      "learning_rate": 1.2200687433799444e-06,
+      "loss": 0.068,
+      "step": 2177
+    },
+    {
+      "epoch": 0.495,
+      "grad_norm": 2.128729953892156,
+      "learning_rate": 1.2200414486774629e-06,
+      "loss": 0.114,
+      "step": 2178
+    },
+    {
+      "epoch": 0.49522727272727274,
+      "grad_norm": 2.1996702538640664,
+      "learning_rate": 1.220014141841048e-06,
+      "loss": 0.0988,
+      "step": 2179
+    },
+    {
+      "epoch": 0.4954545454545455,
+      "grad_norm": 1.2409188086312404,
+      "learning_rate": 1.2199868228712568e-06,
+      "loss": 0.0694,
+      "step": 2180
+    },
+    {
+      "epoch": 0.49568181818181817,
+      "grad_norm": 1.7196825390126642,
+      "learning_rate": 1.219959491768646e-06,
+      "loss": 0.1041,
+      "step": 2181
+    },
+    {
+      "epoch": 0.4959090909090909,
+      "grad_norm": 2.4039885798950285,
+      "learning_rate": 1.2199321485337736e-06,
+      "loss": 0.1181,
+      "step": 2182
+    },
+    {
+      "epoch": 0.49613636363636365,
+      "grad_norm": 1.4896924194362018,
+      "learning_rate": 1.2199047931671966e-06,
+      "loss": 0.0523,
+      "step": 2183
+    },
+    {
+      "epoch": 0.49636363636363634,
+      "grad_norm": 3.1082916450642304,
+      "learning_rate": 1.219877425669473e-06,
+      "loss": 0.1415,
+      "step": 2184
+    },
+    {
+      "epoch": 0.4965909090909091,
+      "grad_norm": 2.5940817361798425,
+      "learning_rate": 1.2198500460411607e-06,
+      "loss": 0.11,
+      "step": 2185
+    },
+    {
+      "epoch": 0.4968181818181818,
+      "grad_norm": 2.8622562860448495,
+      "learning_rate": 1.2198226542828183e-06,
+      "loss": 0.1247,
+      "step": 2186
+    },
+    {
+      "epoch": 0.49704545454545457,
+      "grad_norm": 1.693316521689763,
+      "learning_rate": 1.2197952503950044e-06,
+      "loss": 0.0757,
+      "step": 2187
+    },
+    {
+      "epoch": 0.49727272727272726,
+      "grad_norm": 1.1134273916963404,
+      "learning_rate": 1.2197678343782777e-06,
+      "loss": 0.0518,
+      "step": 2188
+    },
+    {
+      "epoch": 0.4975,
+      "grad_norm": 2.255757579194932,
+      "learning_rate": 1.219740406233197e-06,
+      "loss": 0.0942,
+      "step": 2189
+    },
+    {
+      "epoch": 0.49772727272727274,
+      "grad_norm": 1.857361581350221,
+      "learning_rate": 1.2197129659603218e-06,
+      "loss": 0.0901,
+      "step": 2190
+    },
+    {
+      "epoch": 0.4979545454545454,
+      "grad_norm": 2.130050523355749,
+      "learning_rate": 1.219685513560212e-06,
+      "loss": 0.1156,
+      "step": 2191
+    },
+    {
+      "epoch": 0.49818181818181817,
+      "grad_norm": 1.9360384083426572,
+      "learning_rate": 1.219658049033427e-06,
+      "loss": 0.0767,
+      "step": 2192
+    },
+    {
+      "epoch": 0.4984090909090909,
+      "grad_norm": 2.409137967001759,
+      "learning_rate": 1.2196305723805269e-06,
+      "loss": 0.1254,
+      "step": 2193
+    },
+    {
+      "epoch": 0.49863636363636366,
+      "grad_norm": 2.5970130151715134,
+      "learning_rate": 1.2196030836020719e-06,
+      "loss": 0.0821,
+      "step": 2194
+    },
+    {
+      "epoch": 0.49886363636363634,
+      "grad_norm": 1.3662078034661904,
+      "learning_rate": 1.2195755826986227e-06,
+      "loss": 0.0498,
+      "step": 2195
+    },
+    {
+      "epoch": 0.4990909090909091,
+      "grad_norm": 2.225805066618288,
+      "learning_rate": 1.2195480696707401e-06,
+      "loss": 0.091,
+      "step": 2196
+    },
+    {
+      "epoch": 0.4993181818181818,
+      "grad_norm": 1.4721682406732306,
+      "learning_rate": 1.2195205445189853e-06,
+      "loss": 0.058,
+      "step": 2197
+    },
+    {
+      "epoch": 0.49954545454545457,
+      "grad_norm": 1.6106205175674284,
+      "learning_rate": 1.2194930072439193e-06,
+      "loss": 0.0644,
+      "step": 2198
+    },
+    {
+      "epoch": 0.49977272727272726,
+      "grad_norm": 1.666836834610145,
+      "learning_rate": 1.2194654578461036e-06,
+      "loss": 0.0732,
+      "step": 2199
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.907718004324924,
+      "learning_rate": 1.2194378963261002e-06,
+      "loss": 0.1333,
+      "step": 2200
+    },
+    {
+      "epoch": 0.5002272727272727,
+      "grad_norm": 1.9413416998033906,
+      "learning_rate": 1.2194103226844711e-06,
+      "loss": 0.1068,
+      "step": 2201
+    },
+    {
+      "epoch": 0.5004545454545455,
+      "grad_norm": 2.1805893997280994,
+      "learning_rate": 1.2193827369217783e-06,
+      "loss": 0.1566,
+      "step": 2202
+    },
+    {
+      "epoch": 0.5006818181818182,
+      "grad_norm": 2.7329315084136954,
+      "learning_rate": 1.219355139038585e-06,
+      "loss": 0.1262,
+      "step": 2203
+    },
+    {
+      "epoch": 0.5009090909090909,
+      "grad_norm": 1.7395593514728067,
+      "learning_rate": 1.2193275290354531e-06,
+      "loss": 0.0985,
+      "step": 2204
+    },
+    {
+      "epoch": 0.5011363636363636,
+      "grad_norm": 1.5135630718774011,
+      "learning_rate": 1.219299906912946e-06,
+      "loss": 0.0628,
+      "step": 2205
+    },
+    {
+      "epoch": 0.5013636363636363,
+      "grad_norm": 1.6650831951294658,
+      "learning_rate": 1.2192722726716272e-06,
+      "loss": 0.0774,
+      "step": 2206
+    },
+    {
+      "epoch": 0.5015909090909091,
+      "grad_norm": 1.6774301170101678,
+      "learning_rate": 1.21924462631206e-06,
+      "loss": 0.0501,
+      "step": 2207
+    },
+    {
+      "epoch": 0.5018181818181818,
+      "grad_norm": 2.216809926111673,
+      "learning_rate": 1.219216967834808e-06,
+      "loss": 0.1028,
+      "step": 2208
+    },
+    {
+      "epoch": 0.5020454545454546,
+      "grad_norm": 1.1100740134335398,
+      "learning_rate": 1.2191892972404353e-06,
+      "loss": 0.0436,
+      "step": 2209
+    },
+    {
+      "epoch": 0.5022727272727273,
+      "grad_norm": 1.3088981257450105,
+      "learning_rate": 1.2191616145295066e-06,
+      "loss": 0.0713,
+      "step": 2210
+    },
+    {
+      "epoch": 0.5025,
+      "grad_norm": 1.947637730855621,
+      "learning_rate": 1.2191339197025857e-06,
+      "loss": 0.1093,
+      "step": 2211
+    },
+    {
+      "epoch": 0.5027272727272727,
+      "grad_norm": 1.7752606878206498,
+      "learning_rate": 1.2191062127602377e-06,
+      "loss": 0.0786,
+      "step": 2212
+    },
+    {
+      "epoch": 0.5029545454545454,
+      "grad_norm": 1.513997645320129,
+      "learning_rate": 1.2190784937030275e-06,
+      "loss": 0.0596,
+      "step": 2213
+    },
+    {
+      "epoch": 0.5031818181818182,
+      "grad_norm": 1.9818214917311507,
+      "learning_rate": 1.2190507625315204e-06,
+      "loss": 0.0771,
+      "step": 2214
+    },
+    {
+      "epoch": 0.5034090909090909,
+      "grad_norm": 1.1747473738900922,
+      "learning_rate": 1.2190230192462819e-06,
+      "loss": 0.0593,
+      "step": 2215
+    },
+    {
+      "epoch": 0.5036363636363637,
+      "grad_norm": 1.7832938885085776,
+      "learning_rate": 1.2189952638478778e-06,
+      "loss": 0.0579,
+      "step": 2216
+    },
+    {
+      "epoch": 0.5038636363636364,
+      "grad_norm": 2.4445199054581797,
+      "learning_rate": 1.2189674963368738e-06,
+      "loss": 0.0748,
+      "step": 2217
+    },
+    {
+      "epoch": 0.5040909090909091,
+      "grad_norm": 1.1573219612201073,
+      "learning_rate": 1.2189397167138363e-06,
+      "loss": 0.0389,
+      "step": 2218
+    },
+    {
+      "epoch": 0.5043181818181818,
+      "grad_norm": 3.5642753223005665,
+      "learning_rate": 1.2189119249793318e-06,
+      "loss": 0.1657,
+      "step": 2219
+    },
+    {
+      "epoch": 0.5045454545454545,
+      "grad_norm": 2.343058014592198,
+      "learning_rate": 1.2188841211339273e-06,
+      "loss": 0.1188,
+      "step": 2220
+    },
+    {
+      "epoch": 0.5047727272727273,
+      "grad_norm": 2.170195245650058,
+      "learning_rate": 1.2188563051781894e-06,
+      "loss": 0.1167,
+      "step": 2221
+    },
+    {
+      "epoch": 0.505,
+      "grad_norm": 1.404039083813952,
+      "learning_rate": 1.2188284771126852e-06,
+      "loss": 0.0614,
+      "step": 2222
+    },
+    {
+      "epoch": 0.5052272727272727,
+      "grad_norm": 2.4104833587143406,
+      "learning_rate": 1.2188006369379823e-06,
+      "loss": 0.1147,
+      "step": 2223
+    },
+    {
+      "epoch": 0.5054545454545455,
+      "grad_norm": 2.446878487760744,
+      "learning_rate": 1.2187727846546487e-06,
+      "loss": 0.0945,
+      "step": 2224
+    },
+    {
+      "epoch": 0.5056818181818182,
+      "grad_norm": 1.7547506907944148,
+      "learning_rate": 1.218744920263252e-06,
+      "loss": 0.0775,
+      "step": 2225
+    },
+    {
+      "epoch": 0.5059090909090909,
+      "grad_norm": 1.8507356743078398,
+      "learning_rate": 1.2187170437643608e-06,
+      "loss": 0.0561,
+      "step": 2226
+    },
+    {
+      "epoch": 0.5061363636363636,
+      "grad_norm": 2.2333444007552856,
+      "learning_rate": 1.218689155158543e-06,
+      "loss": 0.0765,
+      "step": 2227
+    },
+    {
+      "epoch": 0.5063636363636363,
+      "grad_norm": 1.3051252599555019,
+      "learning_rate": 1.2186612544463677e-06,
+      "loss": 0.073,
+      "step": 2228
+    },
+    {
+      "epoch": 0.5065909090909091,
+      "grad_norm": 0.6324792663980657,
+      "learning_rate": 1.2186333416284036e-06,
+      "loss": 0.0313,
+      "step": 2229
+    },
+    {
+      "epoch": 0.5068181818181818,
+      "grad_norm": 1.8000188087662068,
+      "learning_rate": 1.21860541670522e-06,
+      "loss": 0.0606,
+      "step": 2230
+    },
+    {
+      "epoch": 0.5070454545454546,
+      "grad_norm": 1.3163052055293232,
+      "learning_rate": 1.2185774796773864e-06,
+      "loss": 0.0714,
+      "step": 2231
+    },
+    {
+      "epoch": 0.5072727272727273,
+      "grad_norm": 1.4951056676228704,
+      "learning_rate": 1.2185495305454724e-06,
+      "loss": 0.0583,
+      "step": 2232
+    },
+    {
+      "epoch": 0.5075,
+      "grad_norm": 1.5641461961689889,
+      "learning_rate": 1.2185215693100482e-06,
+      "loss": 0.0713,
+      "step": 2233
+    },
+    {
+      "epoch": 0.5077272727272727,
+      "grad_norm": 1.6459211098380448,
+      "learning_rate": 1.2184935959716836e-06,
+      "loss": 0.081,
+      "step": 2234
+    },
+    {
+      "epoch": 0.5079545454545454,
+      "grad_norm": 1.4647138364360204,
+      "learning_rate": 1.218465610530949e-06,
+      "loss": 0.0511,
+      "step": 2235
+    },
+    {
+      "epoch": 0.5081818181818182,
+      "grad_norm": 1.8120302130211619,
+      "learning_rate": 1.2184376129884156e-06,
+      "loss": 0.0931,
+      "step": 2236
+    },
+    {
+      "epoch": 0.5084090909090909,
+      "grad_norm": 1.5073449559297587,
+      "learning_rate": 1.2184096033446537e-06,
+      "loss": 0.0444,
+      "step": 2237
+    },
+    {
+      "epoch": 0.5086363636363637,
+      "grad_norm": 1.6847294530828452,
+      "learning_rate": 1.2183815816002347e-06,
+      "loss": 0.0838,
+      "step": 2238
+    },
+    {
+      "epoch": 0.5088636363636364,
+      "grad_norm": 2.3408785652904256,
+      "learning_rate": 1.2183535477557304e-06,
+      "loss": 0.1264,
+      "step": 2239
+    },
+    {
+      "epoch": 0.509090909090909,
+      "grad_norm": 1.8135266974209852,
+      "learning_rate": 1.2183255018117118e-06,
+      "loss": 0.0933,
+      "step": 2240
+    },
+    {
+      "epoch": 0.5093181818181818,
+      "grad_norm": 1.833193726751729,
+      "learning_rate": 1.2182974437687512e-06,
+      "loss": 0.0457,
+      "step": 2241
+    },
+    {
+      "epoch": 0.5095454545454545,
+      "grad_norm": 1.585629780358627,
+      "learning_rate": 1.2182693736274204e-06,
+      "loss": 0.0734,
+      "step": 2242
+    },
+    {
+      "epoch": 0.5097727272727273,
+      "grad_norm": 1.8745119005587123,
+      "learning_rate": 1.2182412913882924e-06,
+      "loss": 0.046,
+      "step": 2243
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.011358704094827,
+      "learning_rate": 1.2182131970519392e-06,
+      "loss": 0.0689,
+      "step": 2244
+    },
+    {
+      "epoch": 0.5102272727272728,
+      "grad_norm": 1.9646847328926522,
+      "learning_rate": 1.2181850906189341e-06,
+      "loss": 0.0772,
+      "step": 2245
+    },
+    {
+      "epoch": 0.5104545454545455,
+      "grad_norm": 1.671260543446508,
+      "learning_rate": 1.2181569720898503e-06,
+      "loss": 0.0981,
+      "step": 2246
+    },
+    {
+      "epoch": 0.5106818181818182,
+      "grad_norm": 1.9498890733375709,
+      "learning_rate": 1.218128841465261e-06,
+      "loss": 0.1082,
+      "step": 2247
+    },
+    {
+      "epoch": 0.5109090909090909,
+      "grad_norm": 1.4835323733995036,
+      "learning_rate": 1.2181006987457397e-06,
+      "loss": 0.0516,
+      "step": 2248
+    },
+    {
+      "epoch": 0.5111363636363636,
+      "grad_norm": 2.374329362162122,
+      "learning_rate": 1.2180725439318606e-06,
+      "loss": 0.0958,
+      "step": 2249
+    },
+    {
+      "epoch": 0.5113636363636364,
+      "grad_norm": 5.397695979915305,
+      "learning_rate": 1.2180443770241974e-06,
+      "loss": 0.2002,
+      "step": 2250
+    },
+    {
+      "epoch": 0.5115909090909091,
+      "grad_norm": 1.4560427679994012,
+      "learning_rate": 1.2180161980233249e-06,
+      "loss": 0.0557,
+      "step": 2251
+    },
+    {
+      "epoch": 0.5118181818181818,
+      "grad_norm": 2.5755074788993855,
+      "learning_rate": 1.2179880069298174e-06,
+      "loss": 0.1742,
+      "step": 2252
+    },
+    {
+      "epoch": 0.5120454545454546,
+      "grad_norm": 1.8589576110911314,
+      "learning_rate": 1.2179598037442502e-06,
+      "loss": 0.0902,
+      "step": 2253
+    },
+    {
+      "epoch": 0.5122727272727273,
+      "grad_norm": 1.8050547811291606,
+      "learning_rate": 1.217931588467198e-06,
+      "loss": 0.0798,
+      "step": 2254
+    },
+    {
+      "epoch": 0.5125,
+      "grad_norm": 2.02935305404429,
+      "learning_rate": 1.217903361099236e-06,
+      "loss": 0.0514,
+      "step": 2255
+    },
+    {
+      "epoch": 0.5127272727272727,
+      "grad_norm": 1.716127799255559,
+      "learning_rate": 1.2178751216409404e-06,
+      "loss": 0.0704,
+      "step": 2256
+    },
+    {
+      "epoch": 0.5129545454545454,
+      "grad_norm": 2.4866828568495554,
+      "learning_rate": 1.217846870092887e-06,
+      "loss": 0.1458,
+      "step": 2257
+    },
+    {
+      "epoch": 0.5131818181818182,
+      "grad_norm": 1.4075974437895173,
+      "learning_rate": 1.217818606455651e-06,
+      "loss": 0.076,
+      "step": 2258
+    },
+    {
+      "epoch": 0.5134090909090909,
+      "grad_norm": 1.9922715851075399,
+      "learning_rate": 1.2177903307298098e-06,
+      "loss": 0.0807,
+      "step": 2259
+    },
+    {
+      "epoch": 0.5136363636363637,
+      "grad_norm": 1.4619529475455997,
+      "learning_rate": 1.2177620429159395e-06,
+      "loss": 0.059,
+      "step": 2260
+    },
+    {
+      "epoch": 0.5138636363636364,
+      "grad_norm": 1.3567045418459571,
+      "learning_rate": 1.217733743014617e-06,
+      "loss": 0.0511,
+      "step": 2261
+    },
+    {
+      "epoch": 0.514090909090909,
+      "grad_norm": 2.530747668861107,
+      "learning_rate": 1.2177054310264195e-06,
+      "loss": 0.0975,
+      "step": 2262
+    },
+    {
+      "epoch": 0.5143181818181818,
+      "grad_norm": 1.1584139335609174,
+      "learning_rate": 1.217677106951924e-06,
+      "loss": 0.0574,
+      "step": 2263
+    },
+    {
+      "epoch": 0.5145454545454545,
+      "grad_norm": 1.952126393540434,
+      "learning_rate": 1.2176487707917084e-06,
+      "loss": 0.0787,
+      "step": 2264
+    },
+    {
+      "epoch": 0.5147727272727273,
+      "grad_norm": 1.710163466319092,
+      "learning_rate": 1.2176204225463506e-06,
+      "loss": 0.0744,
+      "step": 2265
+    },
+    {
+      "epoch": 0.515,
+      "grad_norm": 1.5812822773125854,
+      "learning_rate": 1.2175920622164284e-06,
+      "loss": 0.0801,
+      "step": 2266
+    },
+    {
+      "epoch": 0.5152272727272728,
+      "grad_norm": 1.9661177991254029,
+      "learning_rate": 1.21756368980252e-06,
+      "loss": 0.158,
+      "step": 2267
+    },
+    {
+      "epoch": 0.5154545454545455,
+      "grad_norm": 1.5913179432872733,
+      "learning_rate": 1.2175353053052044e-06,
+      "loss": 0.0451,
+      "step": 2268
+    },
+    {
+      "epoch": 0.5156818181818181,
+      "grad_norm": 1.9371148200939172,
+      "learning_rate": 1.2175069087250603e-06,
+      "loss": 0.0709,
+      "step": 2269
+    },
+    {
+      "epoch": 0.5159090909090909,
+      "grad_norm": 2.252331089065532,
+      "learning_rate": 1.2174785000626664e-06,
+      "loss": 0.0746,
+      "step": 2270
+    },
+    {
+      "epoch": 0.5161363636363636,
+      "grad_norm": 2.4662836413937854,
+      "learning_rate": 1.2174500793186024e-06,
+      "loss": 0.0793,
+      "step": 2271
+    },
+    {
+      "epoch": 0.5163636363636364,
+      "grad_norm": 1.3487870387861032,
+      "learning_rate": 1.2174216464934477e-06,
+      "loss": 0.0506,
+      "step": 2272
+    },
+    {
+      "epoch": 0.5165909090909091,
+      "grad_norm": 1.4023353148212503,
+      "learning_rate": 1.217393201587782e-06,
+      "loss": 0.0787,
+      "step": 2273
+    },
+    {
+      "epoch": 0.5168181818181818,
+      "grad_norm": 1.2839833667729184,
+      "learning_rate": 1.2173647446021856e-06,
+      "loss": 0.0422,
+      "step": 2274
+    },
+    {
+      "epoch": 0.5170454545454546,
+      "grad_norm": 2.4056574180706085,
+      "learning_rate": 1.2173362755372384e-06,
+      "loss": 0.0858,
+      "step": 2275
+    },
+    {
+      "epoch": 0.5172727272727272,
+      "grad_norm": 2.3625866705712903,
+      "learning_rate": 1.2173077943935212e-06,
+      "loss": 0.1462,
+      "step": 2276
+    },
+    {
+      "epoch": 0.5175,
+      "grad_norm": 3.0997327238674344,
+      "learning_rate": 1.2172793011716148e-06,
+      "loss": 0.1763,
+      "step": 2277
+    },
+    {
+      "epoch": 0.5177272727272727,
+      "grad_norm": 1.668246273098672,
+      "learning_rate": 1.2172507958721001e-06,
+      "loss": 0.0819,
+      "step": 2278
+    },
+    {
+      "epoch": 0.5179545454545454,
+      "grad_norm": 3.3009346690495924,
+      "learning_rate": 1.2172222784955586e-06,
+      "loss": 0.1423,
+      "step": 2279
+    },
+    {
+      "epoch": 0.5181818181818182,
+      "grad_norm": 1.790310668482315,
+      "learning_rate": 1.2171937490425715e-06,
+      "loss": 0.1061,
+      "step": 2280
+    },
+    {
+      "epoch": 0.5184090909090909,
+      "grad_norm": 1.6955747651879678,
+      "learning_rate": 1.2171652075137209e-06,
+      "loss": 0.0596,
+      "step": 2281
+    },
+    {
+      "epoch": 0.5186363636363637,
+      "grad_norm": 1.729132992331395,
+      "learning_rate": 1.2171366539095883e-06,
+      "loss": 0.0787,
+      "step": 2282
+    },
+    {
+      "epoch": 0.5188636363636364,
+      "grad_norm": 1.7727068017140468,
+      "learning_rate": 1.2171080882307564e-06,
+      "loss": 0.065,
+      "step": 2283
+    },
+    {
+      "epoch": 0.519090909090909,
+      "grad_norm": 1.3734908133864099,
+      "learning_rate": 1.2170795104778076e-06,
+      "loss": 0.053,
+      "step": 2284
+    },
+    {
+      "epoch": 0.5193181818181818,
+      "grad_norm": 2.9411795933300806,
+      "learning_rate": 1.2170509206513247e-06,
+      "loss": 0.1658,
+      "step": 2285
+    },
+    {
+      "epoch": 0.5195454545454545,
+      "grad_norm": 1.4838507014375277,
+      "learning_rate": 1.2170223187518908e-06,
+      "loss": 0.079,
+      "step": 2286
+    },
+    {
+      "epoch": 0.5197727272727273,
+      "grad_norm": 1.739290758012565,
+      "learning_rate": 1.2169937047800887e-06,
+      "loss": 0.0723,
+      "step": 2287
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4230985988825777,
+      "learning_rate": 1.2169650787365025e-06,
+      "loss": 0.126,
+      "step": 2288
+    },
+    {
+      "epoch": 0.5202272727272728,
+      "grad_norm": 1.498892067669245,
+      "learning_rate": 1.2169364406217154e-06,
+      "loss": 0.0698,
+      "step": 2289
+    },
+    {
+      "epoch": 0.5204545454545455,
+      "grad_norm": 1.9647393087981482,
+      "learning_rate": 1.2169077904363116e-06,
+      "loss": 0.0794,
+      "step": 2290
+    },
+    {
+      "epoch": 0.5206818181818181,
+      "grad_norm": 1.8691386057684936,
+      "learning_rate": 1.216879128180875e-06,
+      "loss": 0.0756,
+      "step": 2291
+    },
+    {
+      "epoch": 0.5209090909090909,
+      "grad_norm": 1.2647044581347964,
+      "learning_rate": 1.2168504538559908e-06,
+      "loss": 0.0548,
+      "step": 2292
+    },
+    {
+      "epoch": 0.5211363636363636,
+      "grad_norm": 2.6546166378428784,
+      "learning_rate": 1.2168217674622433e-06,
+      "loss": 0.1226,
+      "step": 2293
+    },
+    {
+      "epoch": 0.5213636363636364,
+      "grad_norm": 2.458916026380304,
+      "learning_rate": 1.2167930690002172e-06,
+      "loss": 0.1038,
+      "step": 2294
+    },
+    {
+      "epoch": 0.5215909090909091,
+      "grad_norm": 2.008621638526371,
+      "learning_rate": 1.2167643584704982e-06,
+      "loss": 0.116,
+      "step": 2295
+    },
+    {
+      "epoch": 0.5218181818181818,
+      "grad_norm": 1.808967975405842,
+      "learning_rate": 1.2167356358736714e-06,
+      "loss": 0.0751,
+      "step": 2296
+    },
+    {
+      "epoch": 0.5220454545454546,
+      "grad_norm": 1.7949294181516406,
+      "learning_rate": 1.2167069012103228e-06,
+      "loss": 0.1242,
+      "step": 2297
+    },
+    {
+      "epoch": 0.5222727272727272,
+      "grad_norm": 0.6436186804888646,
+      "learning_rate": 1.2166781544810384e-06,
+      "loss": 0.0342,
+      "step": 2298
+    },
+    {
+      "epoch": 0.5225,
+      "grad_norm": 1.2388788648361078,
+      "learning_rate": 1.2166493956864036e-06,
+      "loss": 0.0613,
+      "step": 2299
+    },
+    {
+      "epoch": 0.5227272727272727,
+      "grad_norm": 1.3441886486323542,
+      "learning_rate": 1.216620624827006e-06,
+      "loss": 0.0595,
+      "step": 2300
+    },
+    {
+      "epoch": 0.5229545454545454,
+      "grad_norm": 1.9413533773307794,
+      "learning_rate": 1.2165918419034312e-06,
+      "loss": 0.0712,
+      "step": 2301
+    },
+    {
+      "epoch": 0.5231818181818182,
+      "grad_norm": 2.5567000485491294,
+      "learning_rate": 1.216563046916267e-06,
+      "loss": 0.1364,
+      "step": 2302
+    },
+    {
+      "epoch": 0.5234090909090909,
+      "grad_norm": 1.4402190008872096,
+      "learning_rate": 1.2165342398661002e-06,
+      "loss": 0.076,
+      "step": 2303
+    },
+    {
+      "epoch": 0.5236363636363637,
+      "grad_norm": 1.2993121637133658,
+      "learning_rate": 1.2165054207535183e-06,
+      "loss": 0.0501,
+      "step": 2304
+    },
+    {
+      "epoch": 0.5238636363636363,
+      "grad_norm": 1.6932980256642107,
+      "learning_rate": 1.2164765895791088e-06,
+      "loss": 0.0788,
+      "step": 2305
+    },
+    {
+      "epoch": 0.524090909090909,
+      "grad_norm": 1.9102018993246501,
+      "learning_rate": 1.2164477463434599e-06,
+      "loss": 0.139,
+      "step": 2306
+    },
+    {
+      "epoch": 0.5243181818181818,
+      "grad_norm": 2.1519882374809036,
+      "learning_rate": 1.2164188910471593e-06,
+      "loss": 0.1185,
+      "step": 2307
+    },
+    {
+      "epoch": 0.5245454545454545,
+      "grad_norm": 1.359579357502171,
+      "learning_rate": 1.216390023690796e-06,
+      "loss": 0.0719,
+      "step": 2308
+    },
+    {
+      "epoch": 0.5247727272727273,
+      "grad_norm": 0.811018600239271,
+      "learning_rate": 1.2163611442749585e-06,
+      "loss": 0.0389,
+      "step": 2309
+    },
+    {
+      "epoch": 0.525,
+      "grad_norm": 2.595296359560413,
+      "learning_rate": 1.2163322528002352e-06,
+      "loss": 0.0756,
+      "step": 2310
+    },
+    {
+      "epoch": 0.5252272727272728,
+      "grad_norm": 1.5992700543742142,
+      "learning_rate": 1.216303349267216e-06,
+      "loss": 0.0559,
+      "step": 2311
+    },
+    {
+      "epoch": 0.5254545454545455,
+      "grad_norm": 1.8912855480440998,
+      "learning_rate": 1.2162744336764899e-06,
+      "loss": 0.0793,
+      "step": 2312
+    },
+    {
+      "epoch": 0.5256818181818181,
+      "grad_norm": 2.1781940043128714,
+      "learning_rate": 1.2162455060286462e-06,
+      "loss": 0.0649,
+      "step": 2313
+    },
+    {
+      "epoch": 0.5259090909090909,
+      "grad_norm": 2.953970419313081,
+      "learning_rate": 1.2162165663242754e-06,
+      "loss": 0.1699,
+      "step": 2314
+    },
+    {
+      "epoch": 0.5261363636363636,
+      "grad_norm": 1.6961734765510914,
+      "learning_rate": 1.2161876145639673e-06,
+      "loss": 0.0499,
+      "step": 2315
+    },
+    {
+      "epoch": 0.5263636363636364,
+      "grad_norm": 2.087466970250798,
+      "learning_rate": 1.2161586507483126e-06,
+      "loss": 0.1067,
+      "step": 2316
+    },
+    {
+      "epoch": 0.5265909090909091,
+      "grad_norm": 1.5903163266475338,
+      "learning_rate": 1.2161296748779013e-06,
+      "loss": 0.0797,
+      "step": 2317
+    },
+    {
+      "epoch": 0.5268181818181819,
+      "grad_norm": 3.2653792321387787,
+      "learning_rate": 1.2161006869533248e-06,
+      "loss": 0.1157,
+      "step": 2318
+    },
+    {
+      "epoch": 0.5270454545454546,
+      "grad_norm": 1.3491204525558678,
+      "learning_rate": 1.216071686975174e-06,
+      "loss": 0.0796,
+      "step": 2319
+    },
+    {
+      "epoch": 0.5272727272727272,
+      "grad_norm": 1.9963094639574306,
+      "learning_rate": 1.2160426749440405e-06,
+      "loss": 0.0876,
+      "step": 2320
+    },
+    {
+      "epoch": 0.5275,
+      "grad_norm": 2.049656523747854,
+      "learning_rate": 1.2160136508605156e-06,
+      "loss": 0.0969,
+      "step": 2321
+    },
+    {
+      "epoch": 0.5277272727272727,
+      "grad_norm": 2.1220790579946054,
+      "learning_rate": 1.2159846147251915e-06,
+      "loss": 0.081,
+      "step": 2322
+    },
+    {
+      "epoch": 0.5279545454545455,
+      "grad_norm": 2.08368812435842,
+      "learning_rate": 1.2159555665386599e-06,
+      "loss": 0.0944,
+      "step": 2323
+    },
+    {
+      "epoch": 0.5281818181818182,
+      "grad_norm": 2.0048398801132197,
+      "learning_rate": 1.2159265063015133e-06,
+      "loss": 0.111,
+      "step": 2324
+    },
+    {
+      "epoch": 0.5284090909090909,
+      "grad_norm": 1.754524594188052,
+      "learning_rate": 1.2158974340143442e-06,
+      "loss": 0.1303,
+      "step": 2325
+    },
+    {
+      "epoch": 0.5286363636363637,
+      "grad_norm": 1.912530896825054,
+      "learning_rate": 1.2158683496777457e-06,
+      "loss": 0.0836,
+      "step": 2326
+    },
+    {
+      "epoch": 0.5288636363636363,
+      "grad_norm": 1.231248917994634,
+      "learning_rate": 1.2158392532923107e-06,
+      "loss": 0.0677,
+      "step": 2327
+    },
+    {
+      "epoch": 0.5290909090909091,
+      "grad_norm": 1.337043499402379,
+      "learning_rate": 1.2158101448586327e-06,
+      "loss": 0.0747,
+      "step": 2328
+    },
+    {
+      "epoch": 0.5293181818181818,
+      "grad_norm": 1.9423354965250823,
+      "learning_rate": 1.2157810243773047e-06,
+      "loss": 0.0785,
+      "step": 2329
+    },
+    {
+      "epoch": 0.5295454545454545,
+      "grad_norm": 2.036119011040772,
+      "learning_rate": 1.2157518918489213e-06,
+      "loss": 0.0855,
+      "step": 2330
+    },
+    {
+      "epoch": 0.5297727272727273,
+      "grad_norm": 2.233825160286938,
+      "learning_rate": 1.215722747274076e-06,
+      "loss": 0.106,
+      "step": 2331
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.025611046725034,
+      "learning_rate": 1.2156935906533634e-06,
+      "loss": 0.1045,
+      "step": 2332
+    },
+    {
+      "epoch": 0.5302272727272728,
+      "grad_norm": 1.3243173734168672,
+      "learning_rate": 1.2156644219873781e-06,
+      "loss": 0.0646,
+      "step": 2333
+    },
+    {
+      "epoch": 0.5304545454545454,
+      "grad_norm": 2.8314489345092593,
+      "learning_rate": 1.2156352412767148e-06,
+      "loss": 0.1405,
+      "step": 2334
+    },
+    {
+      "epoch": 0.5306818181818181,
+      "grad_norm": 1.7013059116801548,
+      "learning_rate": 1.2156060485219681e-06,
+      "loss": 0.0644,
+      "step": 2335
+    },
+    {
+      "epoch": 0.5309090909090909,
+      "grad_norm": 1.183738304810414,
+      "learning_rate": 1.2155768437237342e-06,
+      "loss": 0.062,
+      "step": 2336
+    },
+    {
+      "epoch": 0.5311363636363636,
+      "grad_norm": 1.787953311281731,
+      "learning_rate": 1.2155476268826077e-06,
+      "loss": 0.095,
+      "step": 2337
+    },
+    {
+      "epoch": 0.5313636363636364,
+      "grad_norm": 3.0725766558387617,
+      "learning_rate": 1.215518397999185e-06,
+      "loss": 0.132,
+      "step": 2338
+    },
+    {
+      "epoch": 0.5315909090909091,
+      "grad_norm": 1.4312076090008239,
+      "learning_rate": 1.215489157074062e-06,
+      "loss": 0.0458,
+      "step": 2339
+    },
+    {
+      "epoch": 0.5318181818181819,
+      "grad_norm": 2.5927258063341525,
+      "learning_rate": 1.215459904107835e-06,
+      "loss": 0.1875,
+      "step": 2340
+    },
+    {
+      "epoch": 0.5320454545454546,
+      "grad_norm": 2.5216685093862865,
+      "learning_rate": 1.2154306391011003e-06,
+      "loss": 0.1628,
+      "step": 2341
+    },
+    {
+      "epoch": 0.5322727272727272,
+      "grad_norm": 2.0094549830055137,
+      "learning_rate": 1.2154013620544548e-06,
+      "loss": 0.0669,
+      "step": 2342
+    },
+    {
+      "epoch": 0.5325,
+      "grad_norm": 2.363681654077918,
+      "learning_rate": 1.2153720729684957e-06,
+      "loss": 0.0826,
+      "step": 2343
+    },
+    {
+      "epoch": 0.5327272727272727,
+      "grad_norm": 1.502767177229979,
+      "learning_rate": 1.2153427718438198e-06,
+      "loss": 0.0626,
+      "step": 2344
+    },
+    {
+      "epoch": 0.5329545454545455,
+      "grad_norm": 1.793183237904346,
+      "learning_rate": 1.215313458681025e-06,
+      "loss": 0.1001,
+      "step": 2345
+    },
+    {
+      "epoch": 0.5331818181818182,
+      "grad_norm": 1.9717233040588324,
+      "learning_rate": 1.215284133480709e-06,
+      "loss": 0.119,
+      "step": 2346
+    },
+    {
+      "epoch": 0.5334090909090909,
+      "grad_norm": 1.8348670598014714,
+      "learning_rate": 1.2152547962434694e-06,
+      "loss": 0.0651,
+      "step": 2347
+    },
+    {
+      "epoch": 0.5336363636363637,
+      "grad_norm": 2.715643976564033,
+      "learning_rate": 1.215225446969905e-06,
+      "loss": 0.0856,
+      "step": 2348
+    },
+    {
+      "epoch": 0.5338636363636363,
+      "grad_norm": 1.897729112158867,
+      "learning_rate": 1.215196085660614e-06,
+      "loss": 0.0941,
+      "step": 2349
+    },
+    {
+      "epoch": 0.5340909090909091,
+      "grad_norm": 1.5696563928645153,
+      "learning_rate": 1.2151667123161953e-06,
+      "loss": 0.0743,
+      "step": 2350
+    },
+    {
+      "epoch": 0.5343181818181818,
+      "grad_norm": 2.317787916978184,
+      "learning_rate": 1.2151373269372476e-06,
+      "loss": 0.1133,
+      "step": 2351
+    },
+    {
+      "epoch": 0.5345454545454545,
+      "grad_norm": 2.2350718122908733,
+      "learning_rate": 1.2151079295243702e-06,
+      "loss": 0.1422,
+      "step": 2352
+    },
+    {
+      "epoch": 0.5347727272727273,
+      "grad_norm": 1.1685377757183841,
+      "learning_rate": 1.2150785200781627e-06,
+      "loss": 0.0561,
+      "step": 2353
+    },
+    {
+      "epoch": 0.535,
+      "grad_norm": 2.032108978122722,
+      "learning_rate": 1.2150490985992248e-06,
+      "loss": 0.083,
+      "step": 2354
+    },
+    {
+      "epoch": 0.5352272727272728,
+      "grad_norm": 1.3360624520789264,
+      "learning_rate": 1.2150196650881563e-06,
+      "loss": 0.0403,
+      "step": 2355
+    },
+    {
+      "epoch": 0.5354545454545454,
+      "grad_norm": 2.3145350399566014,
+      "learning_rate": 1.2149902195455574e-06,
+      "loss": 0.0999,
+      "step": 2356
+    },
+    {
+      "epoch": 0.5356818181818181,
+      "grad_norm": 2.0843694020721357,
+      "learning_rate": 1.2149607619720286e-06,
+      "loss": 0.0824,
+      "step": 2357
+    },
+    {
+      "epoch": 0.5359090909090909,
+      "grad_norm": 1.718341705559437,
+      "learning_rate": 1.2149312923681707e-06,
+      "loss": 0.096,
+      "step": 2358
+    },
+    {
+      "epoch": 0.5361363636363636,
+      "grad_norm": 2.4041737883162586,
+      "learning_rate": 1.2149018107345847e-06,
+      "loss": 0.1292,
+      "step": 2359
+    },
+    {
+      "epoch": 0.5363636363636364,
+      "grad_norm": 1.4211699319278994,
+      "learning_rate": 1.2148723170718714e-06,
+      "loss": 0.0786,
+      "step": 2360
+    },
+    {
+      "epoch": 0.5365909090909091,
+      "grad_norm": 1.7477415886104959,
+      "learning_rate": 1.2148428113806326e-06,
+      "loss": 0.0945,
+      "step": 2361
+    },
+    {
+      "epoch": 0.5368181818181819,
+      "grad_norm": 1.3511234085516808,
+      "learning_rate": 1.2148132936614699e-06,
+      "loss": 0.0354,
+      "step": 2362
+    },
+    {
+      "epoch": 0.5370454545454545,
+      "grad_norm": 1.9062126143769242,
+      "learning_rate": 1.214783763914985e-06,
+      "loss": 0.0843,
+      "step": 2363
+    },
+    {
+      "epoch": 0.5372727272727272,
+      "grad_norm": 2.7368675528749025,
+      "learning_rate": 1.2147542221417801e-06,
+      "loss": 0.1107,
+      "step": 2364
+    },
+    {
+      "epoch": 0.5375,
+      "grad_norm": 1.8712850556448657,
+      "learning_rate": 1.214724668342458e-06,
+      "loss": 0.0654,
+      "step": 2365
+    },
+    {
+      "epoch": 0.5377272727272727,
+      "grad_norm": 1.6073936044098704,
+      "learning_rate": 1.214695102517621e-06,
+      "loss": 0.0854,
+      "step": 2366
+    },
+    {
+      "epoch": 0.5379545454545455,
+      "grad_norm": 1.3638312826197467,
+      "learning_rate": 1.214665524667872e-06,
+      "loss": 0.0764,
+      "step": 2367
+    },
+    {
+      "epoch": 0.5381818181818182,
+      "grad_norm": 1.9319506568130842,
+      "learning_rate": 1.2146359347938144e-06,
+      "loss": 0.1316,
+      "step": 2368
+    },
+    {
+      "epoch": 0.538409090909091,
+      "grad_norm": 1.2440711313218185,
+      "learning_rate": 1.2146063328960515e-06,
+      "loss": 0.0443,
+      "step": 2369
+    },
+    {
+      "epoch": 0.5386363636363637,
+      "grad_norm": 2.551641028593336,
+      "learning_rate": 1.2145767189751865e-06,
+      "loss": 0.1401,
+      "step": 2370
+    },
+    {
+      "epoch": 0.5388636363636363,
+      "grad_norm": 2.0799613136311956,
+      "learning_rate": 1.214547093031824e-06,
+      "loss": 0.1058,
+      "step": 2371
+    },
+    {
+      "epoch": 0.5390909090909091,
+      "grad_norm": 1.5639903135294775,
+      "learning_rate": 1.2145174550665673e-06,
+      "loss": 0.0472,
+      "step": 2372
+    },
+    {
+      "epoch": 0.5393181818181818,
+      "grad_norm": 4.5739976640472575,
+      "learning_rate": 1.2144878050800216e-06,
+      "loss": 0.0854,
+      "step": 2373
+    },
+    {
+      "epoch": 0.5395454545454546,
+      "grad_norm": 1.8137217133647905,
+      "learning_rate": 1.214458143072791e-06,
+      "loss": 0.1294,
+      "step": 2374
+    },
+    {
+      "epoch": 0.5397727272727273,
+      "grad_norm": 2.755623202473908,
+      "learning_rate": 1.2144284690454806e-06,
+      "loss": 0.1038,
+      "step": 2375
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.10446104571765,
+      "learning_rate": 1.2143987829986953e-06,
+      "loss": 0.0915,
+      "step": 2376
+    },
+    {
+      "epoch": 0.5402272727272728,
+      "grad_norm": 1.5676881717717075,
+      "learning_rate": 1.2143690849330405e-06,
+      "loss": 0.0464,
+      "step": 2377
+    },
+    {
+      "epoch": 0.5404545454545454,
+      "grad_norm": 3.359113182663159,
+      "learning_rate": 1.2143393748491218e-06,
+      "loss": 0.203,
+      "step": 2378
+    },
+    {
+      "epoch": 0.5406818181818182,
+      "grad_norm": 1.7995054960549952,
+      "learning_rate": 1.2143096527475451e-06,
+      "loss": 0.0643,
+      "step": 2379
+    },
+    {
+      "epoch": 0.5409090909090909,
+      "grad_norm": 1.685829795054111,
+      "learning_rate": 1.2142799186289166e-06,
+      "loss": 0.1027,
+      "step": 2380
+    },
+    {
+      "epoch": 0.5411363636363636,
+      "grad_norm": 2.672790941321297,
+      "learning_rate": 1.2142501724938425e-06,
+      "loss": 0.0816,
+      "step": 2381
+    },
+    {
+      "epoch": 0.5413636363636364,
+      "grad_norm": 1.7890249927542015,
+      "learning_rate": 1.2142204143429292e-06,
+      "loss": 0.0692,
+      "step": 2382
+    },
+    {
+      "epoch": 0.5415909090909091,
+      "grad_norm": 1.8259777590678228,
+      "learning_rate": 1.214190644176784e-06,
+      "loss": 0.0517,
+      "step": 2383
+    },
+    {
+      "epoch": 0.5418181818181819,
+      "grad_norm": 2.3963868075761825,
+      "learning_rate": 1.2141608619960135e-06,
+      "loss": 0.1675,
+      "step": 2384
+    },
+    {
+      "epoch": 0.5420454545454545,
+      "grad_norm": 1.618967506575223,
+      "learning_rate": 1.2141310678012253e-06,
+      "loss": 0.0916,
+      "step": 2385
+    },
+    {
+      "epoch": 0.5422727272727272,
+      "grad_norm": 1.4360787690114196,
+      "learning_rate": 1.2141012615930266e-06,
+      "loss": 0.0598,
+      "step": 2386
+    },
+    {
+      "epoch": 0.5425,
+      "grad_norm": 3.6154929062526913,
+      "learning_rate": 1.2140714433720256e-06,
+      "loss": 0.1758,
+      "step": 2387
+    },
+    {
+      "epoch": 0.5427272727272727,
+      "grad_norm": 1.3845833665838845,
+      "learning_rate": 1.2140416131388302e-06,
+      "loss": 0.078,
+      "step": 2388
+    },
+    {
+      "epoch": 0.5429545454545455,
+      "grad_norm": 2.928610601058639,
+      "learning_rate": 1.2140117708940486e-06,
+      "loss": 0.1268,
+      "step": 2389
+    },
+    {
+      "epoch": 0.5431818181818182,
+      "grad_norm": 2.319337705119755,
+      "learning_rate": 1.2139819166382893e-06,
+      "loss": 0.0829,
+      "step": 2390
+    },
+    {
+      "epoch": 0.543409090909091,
+      "grad_norm": 1.620304406242918,
+      "learning_rate": 1.2139520503721614e-06,
+      "loss": 0.0993,
+      "step": 2391
+    },
+    {
+      "epoch": 0.5436363636363636,
+      "grad_norm": 0.8278050186725971,
+      "learning_rate": 1.2139221720962734e-06,
+      "loss": 0.0341,
+      "step": 2392
+    },
+    {
+      "epoch": 0.5438636363636363,
+      "grad_norm": 1.3898291952646848,
+      "learning_rate": 1.2138922818112352e-06,
+      "loss": 0.0685,
+      "step": 2393
+    },
+    {
+      "epoch": 0.5440909090909091,
+      "grad_norm": 1.2614673278856372,
+      "learning_rate": 1.2138623795176557e-06,
+      "loss": 0.0679,
+      "step": 2394
+    },
+    {
+      "epoch": 0.5443181818181818,
+      "grad_norm": 1.9796993130443639,
+      "learning_rate": 1.2138324652161453e-06,
+      "loss": 0.0879,
+      "step": 2395
+    },
+    {
+      "epoch": 0.5445454545454546,
+      "grad_norm": 1.9282557049810003,
+      "learning_rate": 1.2138025389073133e-06,
+      "loss": 0.0563,
+      "step": 2396
+    },
+    {
+      "epoch": 0.5447727272727273,
+      "grad_norm": 1.6378294980113033,
+      "learning_rate": 1.2137726005917707e-06,
+      "loss": 0.0575,
+      "step": 2397
+    },
+    {
+      "epoch": 0.545,
+      "grad_norm": 1.4277976626159072,
+      "learning_rate": 1.2137426502701273e-06,
+      "loss": 0.0796,
+      "step": 2398
+    },
+    {
+      "epoch": 0.5452272727272728,
+      "grad_norm": 1.8542965730059344,
+      "learning_rate": 1.2137126879429942e-06,
+      "loss": 0.0957,
+      "step": 2399
+    },
+    {
+      "epoch": 0.5454545454545454,
+      "grad_norm": 1.2442541078224432,
+      "learning_rate": 1.2136827136109824e-06,
+      "loss": 0.0654,
+      "step": 2400
+    },
+    {
+      "epoch": 0.5456818181818182,
+      "grad_norm": 1.5101948707732744,
+      "learning_rate": 1.2136527272747032e-06,
+      "loss": 0.0794,
+      "step": 2401
+    },
+    {
+      "epoch": 0.5459090909090909,
+      "grad_norm": 1.7281333506167724,
+      "learning_rate": 1.2136227289347678e-06,
+      "loss": 0.0584,
+      "step": 2402
+    },
+    {
+      "epoch": 0.5461363636363636,
+      "grad_norm": 3.353012911312428,
+      "learning_rate": 1.2135927185917878e-06,
+      "loss": 0.195,
+      "step": 2403
+    },
+    {
+      "epoch": 0.5463636363636364,
+      "grad_norm": 1.9619936855526903,
+      "learning_rate": 1.2135626962463755e-06,
+      "loss": 0.0882,
+      "step": 2404
+    },
+    {
+      "epoch": 0.5465909090909091,
+      "grad_norm": 2.414248462257033,
+      "learning_rate": 1.2135326618991434e-06,
+      "loss": 0.1669,
+      "step": 2405
+    },
+    {
+      "epoch": 0.5468181818181819,
+      "grad_norm": 1.7544240754564406,
+      "learning_rate": 1.2135026155507033e-06,
+      "loss": 0.0654,
+      "step": 2406
+    },
+    {
+      "epoch": 0.5470454545454545,
+      "grad_norm": 2.4510496142978506,
+      "learning_rate": 1.2134725572016682e-06,
+      "loss": 0.0943,
+      "step": 2407
+    },
+    {
+      "epoch": 0.5472727272727272,
+      "grad_norm": 2.862411633162445,
+      "learning_rate": 1.2134424868526507e-06,
+      "loss": 0.1293,
+      "step": 2408
+    },
+    {
+      "epoch": 0.5475,
+      "grad_norm": 1.7843810190454428,
+      "learning_rate": 1.2134124045042648e-06,
+      "loss": 0.0786,
+      "step": 2409
+    },
+    {
+      "epoch": 0.5477272727272727,
+      "grad_norm": 2.8112518756326463,
+      "learning_rate": 1.2133823101571232e-06,
+      "loss": 0.0988,
+      "step": 2410
+    },
+    {
+      "epoch": 0.5479545454545455,
+      "grad_norm": 1.4935269273039218,
+      "learning_rate": 1.2133522038118398e-06,
+      "loss": 0.0434,
+      "step": 2411
+    },
+    {
+      "epoch": 0.5481818181818182,
+      "grad_norm": 1.6636377402747409,
+      "learning_rate": 1.2133220854690285e-06,
+      "loss": 0.0729,
+      "step": 2412
+    },
+    {
+      "epoch": 0.548409090909091,
+      "grad_norm": 1.976199235921085,
+      "learning_rate": 1.2132919551293034e-06,
+      "loss": 0.0663,
+      "step": 2413
+    },
+    {
+      "epoch": 0.5486363636363636,
+      "grad_norm": 1.5469602431015965,
+      "learning_rate": 1.2132618127932795e-06,
+      "loss": 0.0916,
+      "step": 2414
+    },
+    {
+      "epoch": 0.5488636363636363,
+      "grad_norm": 2.2886621287371036,
+      "learning_rate": 1.2132316584615704e-06,
+      "loss": 0.0737,
+      "step": 2415
+    },
+    {
+      "epoch": 0.5490909090909091,
+      "grad_norm": 1.8884672761235808,
+      "learning_rate": 1.2132014921347917e-06,
+      "loss": 0.0704,
+      "step": 2416
+    },
+    {
+      "epoch": 0.5493181818181818,
+      "grad_norm": 2.1617272255313145,
+      "learning_rate": 1.2131713138135585e-06,
+      "loss": 0.15,
+      "step": 2417
+    },
+    {
+      "epoch": 0.5495454545454546,
+      "grad_norm": 1.630854108523469,
+      "learning_rate": 1.2131411234984862e-06,
+      "loss": 0.0956,
+      "step": 2418
+    },
+    {
+      "epoch": 0.5497727272727273,
+      "grad_norm": 2.4005454122233725,
+      "learning_rate": 1.21311092119019e-06,
+      "loss": 0.0869,
+      "step": 2419
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.383691011733639,
+      "learning_rate": 1.2130807068892863e-06,
+      "loss": 0.1262,
+      "step": 2420
+    },
+    {
+      "epoch": 0.5502272727272727,
+      "grad_norm": 1.528818534910322,
+      "learning_rate": 1.213050480596391e-06,
+      "loss": 0.0822,
+      "step": 2421
+    },
+    {
+      "epoch": 0.5504545454545454,
+      "grad_norm": 1.7895530483577653,
+      "learning_rate": 1.2130202423121204e-06,
+      "loss": 0.0722,
+      "step": 2422
+    },
+    {
+      "epoch": 0.5506818181818182,
+      "grad_norm": 2.0416258322765546,
+      "learning_rate": 1.2129899920370913e-06,
+      "loss": 0.1038,
+      "step": 2423
+    },
+    {
+      "epoch": 0.5509090909090909,
+      "grad_norm": 1.8046954495923102,
+      "learning_rate": 1.2129597297719206e-06,
+      "loss": 0.064,
+      "step": 2424
+    },
+    {
+      "epoch": 0.5511363636363636,
+      "grad_norm": 2.3258782533884887,
+      "learning_rate": 1.212929455517225e-06,
+      "loss": 0.0658,
+      "step": 2425
+    },
+    {
+      "epoch": 0.5513636363636364,
+      "grad_norm": 1.7956899929095707,
+      "learning_rate": 1.2128991692736223e-06,
+      "loss": 0.047,
+      "step": 2426
+    },
+    {
+      "epoch": 0.5515909090909091,
+      "grad_norm": 2.56639721554455,
+      "learning_rate": 1.2128688710417298e-06,
+      "loss": 0.1313,
+      "step": 2427
+    },
+    {
+      "epoch": 0.5518181818181818,
+      "grad_norm": 2.1201706723235567,
+      "learning_rate": 1.2128385608221655e-06,
+      "loss": 0.1746,
+      "step": 2428
+    },
+    {
+      "epoch": 0.5520454545454545,
+      "grad_norm": 1.0493573915332868,
+      "learning_rate": 1.2128082386155472e-06,
+      "loss": 0.05,
+      "step": 2429
+    },
+    {
+      "epoch": 0.5522727272727272,
+      "grad_norm": 1.2528812404657423,
+      "learning_rate": 1.2127779044224936e-06,
+      "loss": 0.0638,
+      "step": 2430
+    },
+    {
+      "epoch": 0.5525,
+      "grad_norm": 2.748505495287872,
+      "learning_rate": 1.2127475582436232e-06,
+      "loss": 0.1016,
+      "step": 2431
+    },
+    {
+      "epoch": 0.5527272727272727,
+      "grad_norm": 1.5904672794460584,
+      "learning_rate": 1.2127172000795546e-06,
+      "loss": 0.0448,
+      "step": 2432
+    },
+    {
+      "epoch": 0.5529545454545455,
+      "grad_norm": 2.333914138249994,
+      "learning_rate": 1.2126868299309068e-06,
+      "loss": 0.0795,
+      "step": 2433
+    },
+    {
+      "epoch": 0.5531818181818182,
+      "grad_norm": 1.7709378641623688,
+      "learning_rate": 1.2126564477982995e-06,
+      "loss": 0.0856,
+      "step": 2434
+    },
+    {
+      "epoch": 0.553409090909091,
+      "grad_norm": 1.8800507636012862,
+      "learning_rate": 1.2126260536823517e-06,
+      "loss": 0.1519,
+      "step": 2435
+    },
+    {
+      "epoch": 0.5536363636363636,
+      "grad_norm": 1.8775391124724945,
+      "learning_rate": 1.2125956475836837e-06,
+      "loss": 0.0923,
+      "step": 2436
+    },
+    {
+      "epoch": 0.5538636363636363,
+      "grad_norm": 1.6307559200499402,
+      "learning_rate": 1.2125652295029152e-06,
+      "loss": 0.0868,
+      "step": 2437
+    },
+    {
+      "epoch": 0.5540909090909091,
+      "grad_norm": 2.231617299133525,
+      "learning_rate": 1.212534799440667e-06,
+      "loss": 0.0783,
+      "step": 2438
+    },
+    {
+      "epoch": 0.5543181818181818,
+      "grad_norm": 1.7387541123321502,
+      "learning_rate": 1.2125043573975588e-06,
+      "loss": 0.0524,
+      "step": 2439
+    },
+    {
+      "epoch": 0.5545454545454546,
+      "grad_norm": 1.296907892326737,
+      "learning_rate": 1.2124739033742119e-06,
+      "loss": 0.0318,
+      "step": 2440
+    },
+    {
+      "epoch": 0.5547727272727273,
+      "grad_norm": 1.8813886549744556,
+      "learning_rate": 1.2124434373712473e-06,
+      "loss": 0.0737,
+      "step": 2441
+    },
+    {
+      "epoch": 0.555,
+      "grad_norm": 1.3586520300668936,
+      "learning_rate": 1.212412959389286e-06,
+      "loss": 0.0664,
+      "step": 2442
+    },
+    {
+      "epoch": 0.5552272727272727,
+      "grad_norm": 4.270747737444282,
+      "learning_rate": 1.2123824694289498e-06,
+      "loss": 0.2246,
+      "step": 2443
+    },
+    {
+      "epoch": 0.5554545454545454,
+      "grad_norm": 0.4479603736239536,
+      "learning_rate": 1.2123519674908604e-06,
+      "loss": 0.0252,
+      "step": 2444
+    },
+    {
+      "epoch": 0.5556818181818182,
+      "grad_norm": 1.7724679020824081,
+      "learning_rate": 1.2123214535756394e-06,
+      "loss": 0.0993,
+      "step": 2445
+    },
+    {
+      "epoch": 0.5559090909090909,
+      "grad_norm": 1.9219307391836302,
+      "learning_rate": 1.2122909276839095e-06,
+      "loss": 0.0996,
+      "step": 2446
+    },
+    {
+      "epoch": 0.5561363636363637,
+      "grad_norm": 3.4739941592959704,
+      "learning_rate": 1.212260389816293e-06,
+      "loss": 0.1071,
+      "step": 2447
+    },
+    {
+      "epoch": 0.5563636363636364,
+      "grad_norm": 2.6480344676770486,
+      "learning_rate": 1.2122298399734127e-06,
+      "loss": 0.1316,
+      "step": 2448
+    },
+    {
+      "epoch": 0.5565909090909091,
+      "grad_norm": 1.8056085744663635,
+      "learning_rate": 1.2121992781558915e-06,
+      "loss": 0.0668,
+      "step": 2449
+    },
+    {
+      "epoch": 0.5568181818181818,
+      "grad_norm": 0.7329445465397755,
+      "learning_rate": 1.2121687043643523e-06,
+      "loss": 0.0408,
+      "step": 2450
+    },
+    {
+      "epoch": 0.5570454545454545,
+      "grad_norm": 1.5029999592498757,
+      "learning_rate": 1.2121381185994192e-06,
+      "loss": 0.0472,
+      "step": 2451
+    },
+    {
+      "epoch": 0.5572727272727273,
+      "grad_norm": 2.047321401855462,
+      "learning_rate": 1.2121075208617154e-06,
+      "loss": 0.0818,
+      "step": 2452
+    },
+    {
+      "epoch": 0.5575,
+      "grad_norm": 2.409846944460835,
+      "learning_rate": 1.212076911151865e-06,
+      "loss": 0.0931,
+      "step": 2453
+    },
+    {
+      "epoch": 0.5577272727272727,
+      "grad_norm": 2.843944141677551,
+      "learning_rate": 1.2120462894704922e-06,
+      "loss": 0.0951,
+      "step": 2454
+    },
+    {
+      "epoch": 0.5579545454545455,
+      "grad_norm": 1.615479346766341,
+      "learning_rate": 1.212015655818221e-06,
+      "loss": 0.0693,
+      "step": 2455
+    },
+    {
+      "epoch": 0.5581818181818182,
+      "grad_norm": 1.3277696740677072,
+      "learning_rate": 1.211985010195677e-06,
+      "loss": 0.0612,
+      "step": 2456
+    },
+    {
+      "epoch": 0.5584090909090909,
+      "grad_norm": 2.30851850501801,
+      "learning_rate": 1.2119543526034843e-06,
+      "loss": 0.0814,
+      "step": 2457
+    },
+    {
+      "epoch": 0.5586363636363636,
+      "grad_norm": 1.0037937240700514,
+      "learning_rate": 1.2119236830422685e-06,
+      "loss": 0.047,
+      "step": 2458
+    },
+    {
+      "epoch": 0.5588636363636363,
+      "grad_norm": 2.619840064116566,
+      "learning_rate": 1.2118930015126547e-06,
+      "loss": 0.1692,
+      "step": 2459
+    },
+    {
+      "epoch": 0.5590909090909091,
+      "grad_norm": 2.094282236060296,
+      "learning_rate": 1.2118623080152689e-06,
+      "loss": 0.1138,
+      "step": 2460
+    },
+    {
+      "epoch": 0.5593181818181818,
+      "grad_norm": 2.2682880240193777,
+      "learning_rate": 1.2118316025507369e-06,
+      "loss": 0.0872,
+      "step": 2461
+    },
+    {
+      "epoch": 0.5595454545454546,
+      "grad_norm": 2.0236387882546474,
+      "learning_rate": 1.2118008851196844e-06,
+      "loss": 0.1224,
+      "step": 2462
+    },
+    {
+      "epoch": 0.5597727272727273,
+      "grad_norm": 1.5684441558455617,
+      "learning_rate": 1.2117701557227383e-06,
+      "loss": 0.1044,
+      "step": 2463
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4651103230516145,
+      "learning_rate": 1.211739414360525e-06,
+      "loss": 0.0621,
+      "step": 2464
+    },
+    {
+      "epoch": 0.5602272727272727,
+      "grad_norm": 1.9406036369806878,
+      "learning_rate": 1.2117086610336714e-06,
+      "loss": 0.0873,
+      "step": 2465
+    },
+    {
+      "epoch": 0.5604545454545454,
+      "grad_norm": 1.329087064496546,
+      "learning_rate": 1.2116778957428046e-06,
+      "loss": 0.0576,
+      "step": 2466
+    },
+    {
+      "epoch": 0.5606818181818182,
+      "grad_norm": 3.1594193029350777,
+      "learning_rate": 1.211647118488552e-06,
+      "loss": 0.1082,
+      "step": 2467
+    },
+    {
+      "epoch": 0.5609090909090909,
+      "grad_norm": 1.516673551053669,
+      "learning_rate": 1.2116163292715414e-06,
+      "loss": 0.0513,
+      "step": 2468
+    },
+    {
+      "epoch": 0.5611363636363637,
+      "grad_norm": 1.949047010481494,
+      "learning_rate": 1.2115855280924003e-06,
+      "loss": 0.0757,
+      "step": 2469
+    },
+    {
+      "epoch": 0.5613636363636364,
+      "grad_norm": 1.935064997739515,
+      "learning_rate": 1.211554714951757e-06,
+      "loss": 0.0664,
+      "step": 2470
+    },
+    {
+      "epoch": 0.5615909090909091,
+      "grad_norm": 1.8538020868439247,
+      "learning_rate": 1.2115238898502395e-06,
+      "loss": 0.0684,
+      "step": 2471
+    },
+    {
+      "epoch": 0.5618181818181818,
+      "grad_norm": 2.354153575111514,
+      "learning_rate": 1.2114930527884769e-06,
+      "loss": 0.0914,
+      "step": 2472
+    },
+    {
+      "epoch": 0.5620454545454545,
+      "grad_norm": 1.4347417685741475,
+      "learning_rate": 1.2114622037670978e-06,
+      "loss": 0.0502,
+      "step": 2473
+    },
+    {
+      "epoch": 0.5622727272727273,
+      "grad_norm": 2.4324385831999766,
+      "learning_rate": 1.2114313427867311e-06,
+      "loss": 0.0899,
+      "step": 2474
+    },
+    {
+      "epoch": 0.5625,
+      "grad_norm": 1.4369620432290773,
+      "learning_rate": 1.211400469848006e-06,
+      "loss": 0.0495,
+      "step": 2475
+    },
+    {
+      "epoch": 0.5627272727272727,
+      "grad_norm": 2.4985968557169915,
+      "learning_rate": 1.2113695849515527e-06,
+      "loss": 0.1482,
+      "step": 2476
+    },
+    {
+      "epoch": 0.5629545454545455,
+      "grad_norm": 3.0776421302811747,
+      "learning_rate": 1.2113386880980005e-06,
+      "loss": 0.1185,
+      "step": 2477
+    },
+    {
+      "epoch": 0.5631818181818182,
+      "grad_norm": 2.6555101809144,
+      "learning_rate": 1.2113077792879792e-06,
+      "loss": 0.0717,
+      "step": 2478
+    },
+    {
+      "epoch": 0.5634090909090909,
+      "grad_norm": 1.04216134281261,
+      "learning_rate": 1.2112768585221199e-06,
+      "loss": 0.0714,
+      "step": 2479
+    },
+    {
+      "epoch": 0.5636363636363636,
+      "grad_norm": 1.3065222256207134,
+      "learning_rate": 1.2112459258010522e-06,
+      "loss": 0.0563,
+      "step": 2480
+    },
+    {
+      "epoch": 0.5638636363636363,
+      "grad_norm": 2.2206349279108464,
+      "learning_rate": 1.2112149811254076e-06,
+      "loss": 0.1205,
+      "step": 2481
+    },
+    {
+      "epoch": 0.5640909090909091,
+      "grad_norm": 2.0053115870155676,
+      "learning_rate": 1.2111840244958167e-06,
+      "loss": 0.0806,
+      "step": 2482
+    },
+    {
+      "epoch": 0.5643181818181818,
+      "grad_norm": 1.4475963720128207,
+      "learning_rate": 1.211153055912911e-06,
+      "loss": 0.0443,
+      "step": 2483
+    },
+    {
+      "epoch": 0.5645454545454546,
+      "grad_norm": 2.287330744984968,
+      "learning_rate": 1.2111220753773217e-06,
+      "loss": 0.1615,
+      "step": 2484
+    },
+    {
+      "epoch": 0.5647727272727273,
+      "grad_norm": 2.0416250743331976,
+      "learning_rate": 1.2110910828896808e-06,
+      "loss": 0.0883,
+      "step": 2485
+    },
+    {
+      "epoch": 0.565,
+      "grad_norm": 1.225199470061038,
+      "learning_rate": 1.2110600784506202e-06,
+      "loss": 0.0543,
+      "step": 2486
+    },
+    {
+      "epoch": 0.5652272727272727,
+      "grad_norm": 1.7751848152160221,
+      "learning_rate": 1.2110290620607724e-06,
+      "loss": 0.0894,
+      "step": 2487
+    },
+    {
+      "epoch": 0.5654545454545454,
+      "grad_norm": 0.9601529382004175,
+      "learning_rate": 1.2109980337207694e-06,
+      "loss": 0.0432,
+      "step": 2488
+    },
+    {
+      "epoch": 0.5656818181818182,
+      "grad_norm": 1.6406077729040545,
+      "learning_rate": 1.2109669934312443e-06,
+      "loss": 0.0872,
+      "step": 2489
+    },
+    {
+      "epoch": 0.5659090909090909,
+      "grad_norm": 2.7342554931017378,
+      "learning_rate": 1.2109359411928298e-06,
+      "loss": 0.0631,
+      "step": 2490
+    },
+    {
+      "epoch": 0.5661363636363637,
+      "grad_norm": 1.502650206613042,
+      "learning_rate": 1.2109048770061593e-06,
+      "loss": 0.076,
+      "step": 2491
+    },
+    {
+      "epoch": 0.5663636363636364,
+      "grad_norm": 1.805299369601701,
+      "learning_rate": 1.2108738008718662e-06,
+      "loss": 0.0927,
+      "step": 2492
+    },
+    {
+      "epoch": 0.5665909090909091,
+      "grad_norm": 3.1499302051227676,
+      "learning_rate": 1.2108427127905842e-06,
+      "loss": 0.1779,
+      "step": 2493
+    },
+    {
+      "epoch": 0.5668181818181818,
+      "grad_norm": 1.7764448781471702,
+      "learning_rate": 1.2108116127629472e-06,
+      "loss": 0.1217,
+      "step": 2494
+    },
+    {
+      "epoch": 0.5670454545454545,
+      "grad_norm": 3.3764188391646592,
+      "learning_rate": 1.2107805007895896e-06,
+      "loss": 0.1772,
+      "step": 2495
+    },
+    {
+      "epoch": 0.5672727272727273,
+      "grad_norm": 1.3089551578641163,
+      "learning_rate": 1.2107493768711453e-06,
+      "loss": 0.0925,
+      "step": 2496
+    },
+    {
+      "epoch": 0.5675,
+      "grad_norm": 2.144982417529584,
+      "learning_rate": 1.2107182410082496e-06,
+      "loss": 0.0708,
+      "step": 2497
+    },
+    {
+      "epoch": 0.5677272727272727,
+      "grad_norm": 3.758434911270266,
+      "learning_rate": 1.2106870932015369e-06,
+      "loss": 0.1014,
+      "step": 2498
+    },
+    {
+      "epoch": 0.5679545454545455,
+      "grad_norm": 1.7803879762758306,
+      "learning_rate": 1.2106559334516427e-06,
+      "loss": 0.077,
+      "step": 2499
+    },
+    {
+      "epoch": 0.5681818181818182,
+      "grad_norm": 1.6440139993932874,
+      "learning_rate": 1.2106247617592022e-06,
+      "loss": 0.0711,
+      "step": 2500
+    },
+    {
+      "epoch": 0.5684090909090909,
+      "grad_norm": 2.6121652863246174,
+      "learning_rate": 1.2105935781248513e-06,
+      "loss": 0.0996,
+      "step": 2501
+    },
+    {
+      "epoch": 0.5686363636363636,
+      "grad_norm": 1.4785380442034253,
+      "learning_rate": 1.2105623825492255e-06,
+      "loss": 0.0638,
+      "step": 2502
+    },
+    {
+      "epoch": 0.5688636363636363,
+      "grad_norm": 2.5545154933008702,
+      "learning_rate": 1.2105311750329612e-06,
+      "loss": 0.1067,
+      "step": 2503
+    },
+    {
+      "epoch": 0.5690909090909091,
+      "grad_norm": 1.8918886965672141,
+      "learning_rate": 1.2104999555766948e-06,
+      "loss": 0.1003,
+      "step": 2504
+    },
+    {
+      "epoch": 0.5693181818181818,
+      "grad_norm": 2.4086013945969498,
+      "learning_rate": 1.2104687241810628e-06,
+      "loss": 0.0849,
+      "step": 2505
+    },
+    {
+      "epoch": 0.5695454545454546,
+      "grad_norm": 1.6542608592360586,
+      "learning_rate": 1.2104374808467023e-06,
+      "loss": 0.0728,
+      "step": 2506
+    },
+    {
+      "epoch": 0.5697727272727273,
+      "grad_norm": 2.7223709567922705,
+      "learning_rate": 1.21040622557425e-06,
+      "loss": 0.1167,
+      "step": 2507
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4347139668525966,
+      "learning_rate": 1.2103749583643435e-06,
+      "loss": 0.0652,
+      "step": 2508
+    },
+    {
+      "epoch": 0.5702272727272727,
+      "grad_norm": 1.6214983494083488,
+      "learning_rate": 1.2103436792176204e-06,
+      "loss": 0.0808,
+      "step": 2509
+    },
+    {
+      "epoch": 0.5704545454545454,
+      "grad_norm": 2.3606490407491676,
+      "learning_rate": 1.2103123881347184e-06,
+      "loss": 0.1131,
+      "step": 2510
+    },
+    {
+      "epoch": 0.5706818181818182,
+      "grad_norm": 1.739613412064,
+      "learning_rate": 1.2102810851162758e-06,
+      "loss": 0.0856,
+      "step": 2511
+    },
+    {
+      "epoch": 0.5709090909090909,
+      "grad_norm": 2.0096344756147335,
+      "learning_rate": 1.2102497701629306e-06,
+      "loss": 0.0695,
+      "step": 2512
+    },
+    {
+      "epoch": 0.5711363636363637,
+      "grad_norm": 2.3276009230778825,
+      "learning_rate": 1.2102184432753217e-06,
+      "loss": 0.0739,
+      "step": 2513
+    },
+    {
+      "epoch": 0.5713636363636364,
+      "grad_norm": 1.0294625660186165,
+      "learning_rate": 1.2101871044540877e-06,
+      "loss": 0.0473,
+      "step": 2514
+    },
+    {
+      "epoch": 0.571590909090909,
+      "grad_norm": 1.4034659210354488,
+      "learning_rate": 1.210155753699868e-06,
+      "loss": 0.0789,
+      "step": 2515
+    },
+    {
+      "epoch": 0.5718181818181818,
+      "grad_norm": 2.3238123997584728,
+      "learning_rate": 1.2101243910133013e-06,
+      "loss": 0.1448,
+      "step": 2516
+    },
+    {
+      "epoch": 0.5720454545454545,
+      "grad_norm": 1.1010806441517047,
+      "learning_rate": 1.2100930163950275e-06,
+      "loss": 0.052,
+      "step": 2517
+    },
+    {
+      "epoch": 0.5722727272727273,
+      "grad_norm": 2.5201715180944007,
+      "learning_rate": 1.2100616298456865e-06,
+      "loss": 0.1124,
+      "step": 2518
+    },
+    {
+      "epoch": 0.5725,
+      "grad_norm": 2.6698702268379413,
+      "learning_rate": 1.210030231365918e-06,
+      "loss": 0.1153,
+      "step": 2519
+    },
+    {
+      "epoch": 0.5727272727272728,
+      "grad_norm": 1.5932079614003707,
+      "learning_rate": 1.2099988209563627e-06,
+      "loss": 0.1008,
+      "step": 2520
+    },
+    {
+      "epoch": 0.5729545454545455,
+      "grad_norm": 1.7436088984453948,
+      "learning_rate": 1.2099673986176604e-06,
+      "loss": 0.0695,
+      "step": 2521
+    },
+    {
+      "epoch": 0.5731818181818182,
+      "grad_norm": 2.2748936549939733,
+      "learning_rate": 1.2099359643504528e-06,
+      "loss": 0.1021,
+      "step": 2522
+    },
+    {
+      "epoch": 0.5734090909090909,
+      "grad_norm": 2.0201566348234175,
+      "learning_rate": 1.2099045181553803e-06,
+      "loss": 0.0587,
+      "step": 2523
+    },
+    {
+      "epoch": 0.5736363636363636,
+      "grad_norm": 1.647680787661945,
+      "learning_rate": 1.209873060033084e-06,
+      "loss": 0.0638,
+      "step": 2524
+    },
+    {
+      "epoch": 0.5738636363636364,
+      "grad_norm": 2.8637148105025814,
+      "learning_rate": 1.209841589984206e-06,
+      "loss": 0.1644,
+      "step": 2525
+    },
+    {
+      "epoch": 0.5740909090909091,
+      "grad_norm": 1.1940152727543474,
+      "learning_rate": 1.2098101080093873e-06,
+      "loss": 0.0432,
+      "step": 2526
+    },
+    {
+      "epoch": 0.5743181818181818,
+      "grad_norm": 2.942139732862501,
+      "learning_rate": 1.2097786141092704e-06,
+      "loss": 0.0784,
+      "step": 2527
+    },
+    {
+      "epoch": 0.5745454545454546,
+      "grad_norm": 1.5844323093611667,
+      "learning_rate": 1.2097471082844974e-06,
+      "loss": 0.0924,
+      "step": 2528
+    },
+    {
+      "epoch": 0.5747727272727273,
+      "grad_norm": 2.1166592850618184,
+      "learning_rate": 1.2097155905357107e-06,
+      "loss": 0.0985,
+      "step": 2529
+    },
+    {
+      "epoch": 0.575,
+      "grad_norm": 2.2228166621746714,
+      "learning_rate": 1.2096840608635529e-06,
+      "loss": 0.1047,
+      "step": 2530
+    },
+    {
+      "epoch": 0.5752272727272727,
+      "grad_norm": 2.082085821981918,
+      "learning_rate": 1.2096525192686673e-06,
+      "loss": 0.0886,
+      "step": 2531
+    },
+    {
+      "epoch": 0.5754545454545454,
+      "grad_norm": 1.5857056950020956,
+      "learning_rate": 1.2096209657516966e-06,
+      "loss": 0.1066,
+      "step": 2532
+    },
+    {
+      "epoch": 0.5756818181818182,
+      "grad_norm": 2.065555560586976,
+      "learning_rate": 1.2095894003132846e-06,
+      "loss": 0.0915,
+      "step": 2533
+    },
+    {
+      "epoch": 0.5759090909090909,
+      "grad_norm": 3.4129574506521845,
+      "learning_rate": 1.2095578229540748e-06,
+      "loss": 0.1414,
+      "step": 2534
+    },
+    {
+      "epoch": 0.5761363636363637,
+      "grad_norm": 2.1041755921694083,
+      "learning_rate": 1.2095262336747113e-06,
+      "loss": 0.0641,
+      "step": 2535
+    },
+    {
+      "epoch": 0.5763636363636364,
+      "grad_norm": 1.863682349326874,
+      "learning_rate": 1.209494632475838e-06,
+      "loss": 0.0712,
+      "step": 2536
+    },
+    {
+      "epoch": 0.576590909090909,
+      "grad_norm": 1.686593228826063,
+      "learning_rate": 1.2094630193580995e-06,
+      "loss": 0.0653,
+      "step": 2537
+    },
+    {
+      "epoch": 0.5768181818181818,
+      "grad_norm": 1.8208907526151081,
+      "learning_rate": 1.2094313943221404e-06,
+      "loss": 0.1189,
+      "step": 2538
+    },
+    {
+      "epoch": 0.5770454545454545,
+      "grad_norm": 1.4421370311704644,
+      "learning_rate": 1.2093997573686055e-06,
+      "loss": 0.0772,
+      "step": 2539
+    },
+    {
+      "epoch": 0.5772727272727273,
+      "grad_norm": 3.4162936458183686,
+      "learning_rate": 1.20936810849814e-06,
+      "loss": 0.119,
+      "step": 2540
+    },
+    {
+      "epoch": 0.5775,
+      "grad_norm": 2.476326863408967,
+      "learning_rate": 1.2093364477113893e-06,
+      "loss": 0.1353,
+      "step": 2541
+    },
+    {
+      "epoch": 0.5777272727272728,
+      "grad_norm": 1.5587685826096436,
+      "learning_rate": 1.209304775008999e-06,
+      "loss": 0.0697,
+      "step": 2542
+    },
+    {
+      "epoch": 0.5779545454545455,
+      "grad_norm": 2.0594594370448975,
+      "learning_rate": 1.209273090391615e-06,
+      "loss": 0.0901,
+      "step": 2543
+    },
+    {
+      "epoch": 0.5781818181818181,
+      "grad_norm": 1.6633178345173798,
+      "learning_rate": 1.2092413938598834e-06,
+      "loss": 0.0408,
+      "step": 2544
+    },
+    {
+      "epoch": 0.5784090909090909,
+      "grad_norm": 1.6370295412643938,
+      "learning_rate": 1.2092096854144506e-06,
+      "loss": 0.0657,
+      "step": 2545
+    },
+    {
+      "epoch": 0.5786363636363636,
+      "grad_norm": 1.8469554029838748,
+      "learning_rate": 1.2091779650559628e-06,
+      "loss": 0.0683,
+      "step": 2546
+    },
+    {
+      "epoch": 0.5788636363636364,
+      "grad_norm": 1.8588118614578548,
+      "learning_rate": 1.2091462327850674e-06,
+      "loss": 0.0895,
+      "step": 2547
+    },
+    {
+      "epoch": 0.5790909090909091,
+      "grad_norm": 3.0365685838848915,
+      "learning_rate": 1.2091144886024111e-06,
+      "loss": 0.1623,
+      "step": 2548
+    },
+    {
+      "epoch": 0.5793181818181818,
+      "grad_norm": 1.888362911798168,
+      "learning_rate": 1.2090827325086412e-06,
+      "loss": 0.0878,
+      "step": 2549
+    },
+    {
+      "epoch": 0.5795454545454546,
+      "grad_norm": 1.5094299544066772,
+      "learning_rate": 1.2090509645044058e-06,
+      "loss": 0.0543,
+      "step": 2550
+    },
+    {
+      "epoch": 0.5797727272727272,
+      "grad_norm": 1.787605894790316,
+      "learning_rate": 1.209019184590352e-06,
+      "loss": 0.1039,
+      "step": 2551
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3579181925736052,
+      "learning_rate": 1.2089873927671283e-06,
+      "loss": 0.0686,
+      "step": 2552
+    },
+    {
+      "epoch": 0.5802272727272727,
+      "grad_norm": 1.8013175116974323,
+      "learning_rate": 1.2089555890353826e-06,
+      "loss": 0.1086,
+      "step": 2553
+    },
+    {
+      "epoch": 0.5804545454545454,
+      "grad_norm": 2.7873266971186252,
+      "learning_rate": 1.2089237733957638e-06,
+      "loss": 0.1739,
+      "step": 2554
+    },
+    {
+      "epoch": 0.5806818181818182,
+      "grad_norm": 1.8689364474440187,
+      "learning_rate": 1.2088919458489206e-06,
+      "loss": 0.1152,
+      "step": 2555
+    },
+    {
+      "epoch": 0.5809090909090909,
+      "grad_norm": 1.3184064464740526,
+      "learning_rate": 1.2088601063955018e-06,
+      "loss": 0.0685,
+      "step": 2556
+    },
+    {
+      "epoch": 0.5811363636363637,
+      "grad_norm": 1.3082244174653535,
+      "learning_rate": 1.2088282550361572e-06,
+      "loss": 0.0542,
+      "step": 2557
+    },
+    {
+      "epoch": 0.5813636363636364,
+      "grad_norm": 1.0879073541089082,
+      "learning_rate": 1.2087963917715355e-06,
+      "loss": 0.0712,
+      "step": 2558
+    },
+    {
+      "epoch": 0.581590909090909,
+      "grad_norm": 1.2996397982537993,
+      "learning_rate": 1.2087645166022872e-06,
+      "loss": 0.0479,
+      "step": 2559
+    },
+    {
+      "epoch": 0.5818181818181818,
+      "grad_norm": 1.6422910547255567,
+      "learning_rate": 1.208732629529062e-06,
+      "loss": 0.0929,
+      "step": 2560
+    },
+    {
+      "epoch": 0.5820454545454545,
+      "grad_norm": 1.6281817791550055,
+      "learning_rate": 1.20870073055251e-06,
+      "loss": 0.0442,
+      "step": 2561
+    },
+    {
+      "epoch": 0.5822727272727273,
+      "grad_norm": 1.3929943708937942,
+      "learning_rate": 1.2086688196732815e-06,
+      "loss": 0.0793,
+      "step": 2562
+    },
+    {
+      "epoch": 0.5825,
+      "grad_norm": 1.8437700248703621,
+      "learning_rate": 1.208636896892028e-06,
+      "loss": 0.1076,
+      "step": 2563
+    },
+    {
+      "epoch": 0.5827272727272728,
+      "grad_norm": 1.80260030021939,
+      "learning_rate": 1.2086049622093996e-06,
+      "loss": 0.1179,
+      "step": 2564
+    },
+    {
+      "epoch": 0.5829545454545455,
+      "grad_norm": 1.0213330642717038,
+      "learning_rate": 1.208573015626048e-06,
+      "loss": 0.038,
+      "step": 2565
+    },
+    {
+      "epoch": 0.5831818181818181,
+      "grad_norm": 1.6742677210293815,
+      "learning_rate": 1.2085410571426244e-06,
+      "loss": 0.0881,
+      "step": 2566
+    },
+    {
+      "epoch": 0.5834090909090909,
+      "grad_norm": 2.180318333821991,
+      "learning_rate": 1.2085090867597806e-06,
+      "loss": 0.0803,
+      "step": 2567
+    },
+    {
+      "epoch": 0.5836363636363636,
+      "grad_norm": 1.9903413731813495,
+      "learning_rate": 1.2084771044781686e-06,
+      "loss": 0.0845,
+      "step": 2568
+    },
+    {
+      "epoch": 0.5838636363636364,
+      "grad_norm": 2.0873024021637474,
+      "learning_rate": 1.2084451102984407e-06,
+      "loss": 0.0813,
+      "step": 2569
+    },
+    {
+      "epoch": 0.5840909090909091,
+      "grad_norm": 2.0672111347124416,
+      "learning_rate": 1.2084131042212489e-06,
+      "loss": 0.1056,
+      "step": 2570
+    },
+    {
+      "epoch": 0.5843181818181818,
+      "grad_norm": 2.302657717994961,
+      "learning_rate": 1.208381086247246e-06,
+      "loss": 0.1602,
+      "step": 2571
+    },
+    {
+      "epoch": 0.5845454545454546,
+      "grad_norm": 1.9708573433901042,
+      "learning_rate": 1.2083490563770853e-06,
+      "loss": 0.089,
+      "step": 2572
+    },
+    {
+      "epoch": 0.5847727272727272,
+      "grad_norm": 1.9372771818282102,
+      "learning_rate": 1.2083170146114195e-06,
+      "loss": 0.0862,
+      "step": 2573
+    },
+    {
+      "epoch": 0.585,
+      "grad_norm": 1.5260810143806092,
+      "learning_rate": 1.208284960950902e-06,
+      "loss": 0.0601,
+      "step": 2574
+    },
+    {
+      "epoch": 0.5852272727272727,
+      "grad_norm": 2.1801528974382225,
+      "learning_rate": 1.2082528953961866e-06,
+      "loss": 0.0922,
+      "step": 2575
+    },
+    {
+      "epoch": 0.5854545454545454,
+      "grad_norm": 1.4864499664188107,
+      "learning_rate": 1.2082208179479272e-06,
+      "loss": 0.0851,
+      "step": 2576
+    },
+    {
+      "epoch": 0.5856818181818182,
+      "grad_norm": 1.9515791114707723,
+      "learning_rate": 1.2081887286067779e-06,
+      "loss": 0.1084,
+      "step": 2577
+    },
+    {
+      "epoch": 0.5859090909090909,
+      "grad_norm": 1.671397092978947,
+      "learning_rate": 1.208156627373393e-06,
+      "loss": 0.0557,
+      "step": 2578
+    },
+    {
+      "epoch": 0.5861363636363637,
+      "grad_norm": 2.656943786264855,
+      "learning_rate": 1.2081245142484268e-06,
+      "loss": 0.0975,
+      "step": 2579
+    },
+    {
+      "epoch": 0.5863636363636363,
+      "grad_norm": 1.9568696347664485,
+      "learning_rate": 1.2080923892325346e-06,
+      "loss": 0.0771,
+      "step": 2580
+    },
+    {
+      "epoch": 0.586590909090909,
+      "grad_norm": 2.258658753878424,
+      "learning_rate": 1.2080602523263715e-06,
+      "loss": 0.0995,
+      "step": 2581
+    },
+    {
+      "epoch": 0.5868181818181818,
+      "grad_norm": 1.9266524107264382,
+      "learning_rate": 1.2080281035305923e-06,
+      "loss": 0.066,
+      "step": 2582
+    },
+    {
+      "epoch": 0.5870454545454545,
+      "grad_norm": 1.685200233667084,
+      "learning_rate": 1.2079959428458533e-06,
+      "loss": 0.0774,
+      "step": 2583
+    },
+    {
+      "epoch": 0.5872727272727273,
+      "grad_norm": 1.8245648258265192,
+      "learning_rate": 1.2079637702728098e-06,
+      "loss": 0.0682,
+      "step": 2584
+    },
+    {
+      "epoch": 0.5875,
+      "grad_norm": 2.4861839730645707,
+      "learning_rate": 1.2079315858121178e-06,
+      "loss": 0.0727,
+      "step": 2585
+    },
+    {
+      "epoch": 0.5877272727272728,
+      "grad_norm": 1.6800529275199663,
+      "learning_rate": 1.207899389464434e-06,
+      "loss": 0.0624,
+      "step": 2586
+    },
+    {
+      "epoch": 0.5879545454545455,
+      "grad_norm": 2.517127292595301,
+      "learning_rate": 1.2078671812304145e-06,
+      "loss": 0.1606,
+      "step": 2587
+    },
+    {
+      "epoch": 0.5881818181818181,
+      "grad_norm": 1.9785298911628482,
+      "learning_rate": 1.2078349611107165e-06,
+      "loss": 0.05,
+      "step": 2588
+    },
+    {
+      "epoch": 0.5884090909090909,
+      "grad_norm": 1.9386057202899163,
+      "learning_rate": 1.2078027291059968e-06,
+      "loss": 0.1513,
+      "step": 2589
+    },
+    {
+      "epoch": 0.5886363636363636,
+      "grad_norm": 2.033082058800513,
+      "learning_rate": 1.2077704852169127e-06,
+      "loss": 0.1132,
+      "step": 2590
+    },
+    {
+      "epoch": 0.5888636363636364,
+      "grad_norm": 1.8799522313808212,
+      "learning_rate": 1.2077382294441218e-06,
+      "loss": 0.0826,
+      "step": 2591
+    },
+    {
+      "epoch": 0.5890909090909091,
+      "grad_norm": 2.8122815022465315,
+      "learning_rate": 1.2077059617882817e-06,
+      "loss": 0.1926,
+      "step": 2592
+    },
+    {
+      "epoch": 0.5893181818181819,
+      "grad_norm": 1.4511382103858386,
+      "learning_rate": 1.2076736822500502e-06,
+      "loss": 0.0663,
+      "step": 2593
+    },
+    {
+      "epoch": 0.5895454545454546,
+      "grad_norm": 2.4036698773262524,
+      "learning_rate": 1.207641390830086e-06,
+      "loss": 0.0899,
+      "step": 2594
+    },
+    {
+      "epoch": 0.5897727272727272,
+      "grad_norm": 1.505397864773569,
+      "learning_rate": 1.2076090875290476e-06,
+      "loss": 0.0611,
+      "step": 2595
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9414979110900152,
+      "learning_rate": 1.2075767723475932e-06,
+      "loss": 0.091,
+      "step": 2596
+    },
+    {
+      "epoch": 0.5902272727272727,
+      "grad_norm": 1.7841860541581893,
+      "learning_rate": 1.207544445286382e-06,
+      "loss": 0.0806,
+      "step": 2597
+    },
+    {
+      "epoch": 0.5904545454545455,
+      "grad_norm": 1.6167321002154313,
+      "learning_rate": 1.2075121063460733e-06,
+      "loss": 0.0549,
+      "step": 2598
+    },
+    {
+      "epoch": 0.5906818181818182,
+      "grad_norm": 1.859615432177371,
+      "learning_rate": 1.2074797555273265e-06,
+      "loss": 0.11,
+      "step": 2599
+    },
+    {
+      "epoch": 0.5909090909090909,
+      "grad_norm": 1.9486346356116189,
+      "learning_rate": 1.2074473928308015e-06,
+      "loss": 0.1134,
+      "step": 2600
+    },
+    {
+      "epoch": 0.5911363636363637,
+      "grad_norm": 1.609098435942125,
+      "learning_rate": 1.2074150182571579e-06,
+      "loss": 0.0823,
+      "step": 2601
+    },
+    {
+      "epoch": 0.5913636363636363,
+      "grad_norm": 1.3758084115469416,
+      "learning_rate": 1.2073826318070558e-06,
+      "loss": 0.0568,
+      "step": 2602
+    },
+    {
+      "epoch": 0.5915909090909091,
+      "grad_norm": 1.6828076333099042,
+      "learning_rate": 1.207350233481156e-06,
+      "loss": 0.079,
+      "step": 2603
+    },
+    {
+      "epoch": 0.5918181818181818,
+      "grad_norm": 1.6511246835279099,
+      "learning_rate": 1.207317823280119e-06,
+      "loss": 0.0767,
+      "step": 2604
+    },
+    {
+      "epoch": 0.5920454545454545,
+      "grad_norm": 1.6415951754995992,
+      "learning_rate": 1.2072854012046056e-06,
+      "loss": 0.0806,
+      "step": 2605
+    },
+    {
+      "epoch": 0.5922727272727273,
+      "grad_norm": 1.2136471021447068,
+      "learning_rate": 1.2072529672552771e-06,
+      "loss": 0.0469,
+      "step": 2606
+    },
+    {
+      "epoch": 0.5925,
+      "grad_norm": 1.7983177386626361,
+      "learning_rate": 1.2072205214327947e-06,
+      "loss": 0.1045,
+      "step": 2607
+    },
+    {
+      "epoch": 0.5927272727272728,
+      "grad_norm": 2.1392333891161743,
+      "learning_rate": 1.2071880637378202e-06,
+      "loss": 0.0859,
+      "step": 2608
+    },
+    {
+      "epoch": 0.5929545454545454,
+      "grad_norm": 1.442023914698665,
+      "learning_rate": 1.2071555941710154e-06,
+      "loss": 0.0837,
+      "step": 2609
+    },
+    {
+      "epoch": 0.5931818181818181,
+      "grad_norm": 1.8155510720546082,
+      "learning_rate": 1.2071231127330425e-06,
+      "loss": 0.0587,
+      "step": 2610
+    },
+    {
+      "epoch": 0.5934090909090909,
+      "grad_norm": 2.1502458710786008,
+      "learning_rate": 1.2070906194245634e-06,
+      "loss": 0.155,
+      "step": 2611
+    },
+    {
+      "epoch": 0.5936363636363636,
+      "grad_norm": 2.936975997367636,
+      "learning_rate": 1.2070581142462415e-06,
+      "loss": 0.1532,
+      "step": 2612
+    },
+    {
+      "epoch": 0.5938636363636364,
+      "grad_norm": 1.602759578874033,
+      "learning_rate": 1.2070255971987392e-06,
+      "loss": 0.0541,
+      "step": 2613
+    },
+    {
+      "epoch": 0.5940909090909091,
+      "grad_norm": 1.2683463229327416,
+      "learning_rate": 1.2069930682827192e-06,
+      "loss": 0.0517,
+      "step": 2614
+    },
+    {
+      "epoch": 0.5943181818181819,
+      "grad_norm": 1.456749607213772,
+      "learning_rate": 1.2069605274988453e-06,
+      "loss": 0.0518,
+      "step": 2615
+    },
+    {
+      "epoch": 0.5945454545454546,
+      "grad_norm": 1.1235114795184564,
+      "learning_rate": 1.2069279748477812e-06,
+      "loss": 0.0413,
+      "step": 2616
+    },
+    {
+      "epoch": 0.5947727272727272,
+      "grad_norm": 1.9425598105038089,
+      "learning_rate": 1.2068954103301903e-06,
+      "loss": 0.0891,
+      "step": 2617
+    },
+    {
+      "epoch": 0.595,
+      "grad_norm": 3.107879015602975,
+      "learning_rate": 1.2068628339467368e-06,
+      "loss": 0.1746,
+      "step": 2618
+    },
+    {
+      "epoch": 0.5952272727272727,
+      "grad_norm": 1.3506068068006425,
+      "learning_rate": 1.206830245698085e-06,
+      "loss": 0.0689,
+      "step": 2619
+    },
+    {
+      "epoch": 0.5954545454545455,
+      "grad_norm": 2.2987323844382357,
+      "learning_rate": 1.2067976455848992e-06,
+      "loss": 0.0995,
+      "step": 2620
+    },
+    {
+      "epoch": 0.5956818181818182,
+      "grad_norm": 1.9599829457817857,
+      "learning_rate": 1.206765033607845e-06,
+      "loss": 0.0956,
+      "step": 2621
+    },
+    {
+      "epoch": 0.5959090909090909,
+      "grad_norm": 1.4966295478837615,
+      "learning_rate": 1.2067324097675863e-06,
+      "loss": 0.0885,
+      "step": 2622
+    },
+    {
+      "epoch": 0.5961363636363637,
+      "grad_norm": 1.8031185319924365,
+      "learning_rate": 1.206699774064789e-06,
+      "loss": 0.0714,
+      "step": 2623
+    },
+    {
+      "epoch": 0.5963636363636363,
+      "grad_norm": 2.0339941541478095,
+      "learning_rate": 1.2066671265001185e-06,
+      "loss": 0.0894,
+      "step": 2624
+    },
+    {
+      "epoch": 0.5965909090909091,
+      "grad_norm": 1.0890826422637505,
+      "learning_rate": 1.2066344670742405e-06,
+      "loss": 0.0611,
+      "step": 2625
+    },
+    {
+      "epoch": 0.5968181818181818,
+      "grad_norm": 1.343749142024692,
+      "learning_rate": 1.2066017957878212e-06,
+      "loss": 0.0898,
+      "step": 2626
+    },
+    {
+      "epoch": 0.5970454545454545,
+      "grad_norm": 2.0396564871190663,
+      "learning_rate": 1.2065691126415262e-06,
+      "loss": 0.0688,
+      "step": 2627
+    },
+    {
+      "epoch": 0.5972727272727273,
+      "grad_norm": 2.097504617899022,
+      "learning_rate": 1.2065364176360227e-06,
+      "loss": 0.085,
+      "step": 2628
+    },
+    {
+      "epoch": 0.5975,
+      "grad_norm": 2.550211501291697,
+      "learning_rate": 1.206503710771977e-06,
+      "loss": 0.0654,
+      "step": 2629
+    },
+    {
+      "epoch": 0.5977272727272728,
+      "grad_norm": 1.747758826539172,
+      "learning_rate": 1.2064709920500565e-06,
+      "loss": 0.0721,
+      "step": 2630
+    },
+    {
+      "epoch": 0.5979545454545454,
+      "grad_norm": 1.8704775708182058,
+      "learning_rate": 1.2064382614709276e-06,
+      "loss": 0.1024,
+      "step": 2631
+    },
+    {
+      "epoch": 0.5981818181818181,
+      "grad_norm": 1.7309841765622334,
+      "learning_rate": 1.2064055190352584e-06,
+      "loss": 0.0787,
+      "step": 2632
+    },
+    {
+      "epoch": 0.5984090909090909,
+      "grad_norm": 2.5032506675652595,
+      "learning_rate": 1.2063727647437162e-06,
+      "loss": 0.0874,
+      "step": 2633
+    },
+    {
+      "epoch": 0.5986363636363636,
+      "grad_norm": 1.731722347167956,
+      "learning_rate": 1.2063399985969692e-06,
+      "loss": 0.0889,
+      "step": 2634
+    },
+    {
+      "epoch": 0.5988636363636364,
+      "grad_norm": 1.854372110194003,
+      "learning_rate": 1.2063072205956852e-06,
+      "loss": 0.0726,
+      "step": 2635
+    },
+    {
+      "epoch": 0.5990909090909091,
+      "grad_norm": 2.2904042236645714,
+      "learning_rate": 1.206274430740533e-06,
+      "loss": 0.0811,
+      "step": 2636
+    },
+    {
+      "epoch": 0.5993181818181819,
+      "grad_norm": 1.0926937165208397,
+      "learning_rate": 1.206241629032181e-06,
+      "loss": 0.0578,
+      "step": 2637
+    },
+    {
+      "epoch": 0.5995454545454545,
+      "grad_norm": 2.109496709479513,
+      "learning_rate": 1.2062088154712983e-06,
+      "loss": 0.1587,
+      "step": 2638
+    },
+    {
+      "epoch": 0.5997727272727272,
+      "grad_norm": 2.113557699436018,
+      "learning_rate": 1.2061759900585538e-06,
+      "loss": 0.0784,
+      "step": 2639
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3977691964103396,
+      "learning_rate": 1.2061431527946169e-06,
+      "loss": 0.0663,
+      "step": 2640
+    },
+    {
+      "epoch": 0.6002272727272727,
+      "grad_norm": 1.8977099520047114,
+      "learning_rate": 1.2061103036801573e-06,
+      "loss": 0.046,
+      "step": 2641
+    },
+    {
+      "epoch": 0.6004545454545455,
+      "grad_norm": 2.9590484044634486,
+      "learning_rate": 1.2060774427158448e-06,
+      "loss": 0.1009,
+      "step": 2642
+    },
+    {
+      "epoch": 0.6006818181818182,
+      "grad_norm": 1.368896860845168,
+      "learning_rate": 1.2060445699023493e-06,
+      "loss": 0.0569,
+      "step": 2643
+    },
+    {
+      "epoch": 0.600909090909091,
+      "grad_norm": 0.6250400128196438,
+      "learning_rate": 1.2060116852403417e-06,
+      "loss": 0.0406,
+      "step": 2644
+    },
+    {
+      "epoch": 0.6011363636363637,
+      "grad_norm": 1.1166985300000885,
+      "learning_rate": 1.205978788730492e-06,
+      "loss": 0.0613,
+      "step": 2645
+    },
+    {
+      "epoch": 0.6013636363636363,
+      "grad_norm": 3.1119072555025316,
+      "learning_rate": 1.2059458803734712e-06,
+      "loss": 0.1552,
+      "step": 2646
+    },
+    {
+      "epoch": 0.6015909090909091,
+      "grad_norm": 1.3625427452156238,
+      "learning_rate": 1.2059129601699504e-06,
+      "loss": 0.0566,
+      "step": 2647
+    },
+    {
+      "epoch": 0.6018181818181818,
+      "grad_norm": 2.4705397050335263,
+      "learning_rate": 1.205880028120601e-06,
+      "loss": 0.1016,
+      "step": 2648
+    },
+    {
+      "epoch": 0.6020454545454546,
+      "grad_norm": 1.7603648265984961,
+      "learning_rate": 1.2058470842260941e-06,
+      "loss": 0.0916,
+      "step": 2649
+    },
+    {
+      "epoch": 0.6022727272727273,
+      "grad_norm": 1.4840058497848245,
+      "learning_rate": 1.2058141284871021e-06,
+      "loss": 0.0352,
+      "step": 2650
+    },
+    {
+      "epoch": 0.6025,
+      "grad_norm": 1.4206816917095877,
+      "learning_rate": 1.2057811609042968e-06,
+      "loss": 0.0598,
+      "step": 2651
+    },
+    {
+      "epoch": 0.6027272727272728,
+      "grad_norm": 1.3814475589682744,
+      "learning_rate": 1.2057481814783503e-06,
+      "loss": 0.0705,
+      "step": 2652
+    },
+    {
+      "epoch": 0.6029545454545454,
+      "grad_norm": 2.535282662612228,
+      "learning_rate": 1.2057151902099352e-06,
+      "loss": 0.1069,
+      "step": 2653
+    },
+    {
+      "epoch": 0.6031818181818182,
+      "grad_norm": 1.7463413841552615,
+      "learning_rate": 1.2056821870997242e-06,
+      "loss": 0.0801,
+      "step": 2654
+    },
+    {
+      "epoch": 0.6034090909090909,
+      "grad_norm": 2.0683386086925073,
+      "learning_rate": 1.2056491721483904e-06,
+      "loss": 0.1007,
+      "step": 2655
+    },
+    {
+      "epoch": 0.6036363636363636,
+      "grad_norm": 1.7184713285634416,
+      "learning_rate": 1.205616145356607e-06,
+      "loss": 0.0493,
+      "step": 2656
+    },
+    {
+      "epoch": 0.6038636363636364,
+      "grad_norm": 1.875177300618362,
+      "learning_rate": 1.2055831067250474e-06,
+      "loss": 0.0557,
+      "step": 2657
+    },
+    {
+      "epoch": 0.6040909090909091,
+      "grad_norm": 5.307901042340458,
+      "learning_rate": 1.2055500562543854e-06,
+      "loss": 0.145,
+      "step": 2658
+    },
+    {
+      "epoch": 0.6043181818181819,
+      "grad_norm": 1.6997194963935633,
+      "learning_rate": 1.205516993945295e-06,
+      "loss": 0.1053,
+      "step": 2659
+    },
+    {
+      "epoch": 0.6045454545454545,
+      "grad_norm": 1.5966905458293268,
+      "learning_rate": 1.2054839197984503e-06,
+      "loss": 0.0788,
+      "step": 2660
+    },
+    {
+      "epoch": 0.6047727272727272,
+      "grad_norm": 1.6569414577650399,
+      "learning_rate": 1.2054508338145257e-06,
+      "loss": 0.0676,
+      "step": 2661
+    },
+    {
+      "epoch": 0.605,
+      "grad_norm": 1.444977738687939,
+      "learning_rate": 1.205417735994196e-06,
+      "loss": 0.0817,
+      "step": 2662
+    },
+    {
+      "epoch": 0.6052272727272727,
+      "grad_norm": 2.064752495544887,
+      "learning_rate": 1.205384626338136e-06,
+      "loss": 0.1374,
+      "step": 2663
+    },
+    {
+      "epoch": 0.6054545454545455,
+      "grad_norm": 2.3432892558637914,
+      "learning_rate": 1.205351504847021e-06,
+      "loss": 0.0927,
+      "step": 2664
+    },
+    {
+      "epoch": 0.6056818181818182,
+      "grad_norm": 1.3201336064024416,
+      "learning_rate": 1.2053183715215262e-06,
+      "loss": 0.0768,
+      "step": 2665
+    },
+    {
+      "epoch": 0.605909090909091,
+      "grad_norm": 1.4623064388105913,
+      "learning_rate": 1.2052852263623274e-06,
+      "loss": 0.0769,
+      "step": 2666
+    },
+    {
+      "epoch": 0.6061363636363636,
+      "grad_norm": 2.237618856328588,
+      "learning_rate": 1.2052520693701007e-06,
+      "loss": 0.1384,
+      "step": 2667
+    },
+    {
+      "epoch": 0.6063636363636363,
+      "grad_norm": 2.9296589260703367,
+      "learning_rate": 1.205218900545522e-06,
+      "loss": 0.1723,
+      "step": 2668
+    },
+    {
+      "epoch": 0.6065909090909091,
+      "grad_norm": 2.836566288382695,
+      "learning_rate": 1.2051857198892674e-06,
+      "loss": 0.1659,
+      "step": 2669
+    },
+    {
+      "epoch": 0.6068181818181818,
+      "grad_norm": 1.5215709526336856,
+      "learning_rate": 1.205152527402014e-06,
+      "loss": 0.0794,
+      "step": 2670
+    },
+    {
+      "epoch": 0.6070454545454546,
+      "grad_norm": 2.8346958331913275,
+      "learning_rate": 1.2051193230844382e-06,
+      "loss": 0.1283,
+      "step": 2671
+    },
+    {
+      "epoch": 0.6072727272727273,
+      "grad_norm": 2.2148653055306,
+      "learning_rate": 1.2050861069372175e-06,
+      "loss": 0.069,
+      "step": 2672
+    },
+    {
+      "epoch": 0.6075,
+      "grad_norm": 4.003274871883982,
+      "learning_rate": 1.205052878961029e-06,
+      "loss": 0.1177,
+      "step": 2673
+    },
+    {
+      "epoch": 0.6077272727272728,
+      "grad_norm": 1.4815104824400815,
+      "learning_rate": 1.2050196391565504e-06,
+      "loss": 0.1068,
+      "step": 2674
+    },
+    {
+      "epoch": 0.6079545454545454,
+      "grad_norm": 2.3417498328060677,
+      "learning_rate": 1.2049863875244595e-06,
+      "loss": 0.1395,
+      "step": 2675
+    },
+    {
+      "epoch": 0.6081818181818182,
+      "grad_norm": 1.4966387487509107,
+      "learning_rate": 1.2049531240654343e-06,
+      "loss": 0.0627,
+      "step": 2676
+    },
+    {
+      "epoch": 0.6084090909090909,
+      "grad_norm": 2.5286938476144307,
+      "learning_rate": 1.204919848780153e-06,
+      "loss": 0.1279,
+      "step": 2677
+    },
+    {
+      "epoch": 0.6086363636363636,
+      "grad_norm": 3.170950871095497,
+      "learning_rate": 1.2048865616692945e-06,
+      "loss": 0.2204,
+      "step": 2678
+    },
+    {
+      "epoch": 0.6088636363636364,
+      "grad_norm": 1.7580453831948917,
+      "learning_rate": 1.204853262733537e-06,
+      "loss": 0.074,
+      "step": 2679
+    },
+    {
+      "epoch": 0.6090909090909091,
+      "grad_norm": 1.584565150542556,
+      "learning_rate": 1.2048199519735601e-06,
+      "loss": 0.0655,
+      "step": 2680
+    },
+    {
+      "epoch": 0.6093181818181819,
+      "grad_norm": 2.185521772962692,
+      "learning_rate": 1.2047866293900428e-06,
+      "loss": 0.0846,
+      "step": 2681
+    },
+    {
+      "epoch": 0.6095454545454545,
+      "grad_norm": 1.2382844516450735,
+      "learning_rate": 1.2047532949836645e-06,
+      "loss": 0.0545,
+      "step": 2682
+    },
+    {
+      "epoch": 0.6097727272727272,
+      "grad_norm": 1.1853206746623002,
+      "learning_rate": 1.2047199487551051e-06,
+      "loss": 0.0444,
+      "step": 2683
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.381398385179321,
+      "learning_rate": 1.2046865907050447e-06,
+      "loss": 0.047,
+      "step": 2684
+    },
+    {
+      "epoch": 0.6102272727272727,
+      "grad_norm": 1.2622720895312578,
+      "learning_rate": 1.2046532208341636e-06,
+      "loss": 0.0426,
+      "step": 2685
+    },
+    {
+      "epoch": 0.6104545454545455,
+      "grad_norm": 1.802134722523953,
+      "learning_rate": 1.2046198391431415e-06,
+      "loss": 0.0762,
+      "step": 2686
+    },
+    {
+      "epoch": 0.6106818181818182,
+      "grad_norm": 1.9582726110956412,
+      "learning_rate": 1.20458644563266e-06,
+      "loss": 0.0772,
+      "step": 2687
+    },
+    {
+      "epoch": 0.610909090909091,
+      "grad_norm": 1.87358457596608,
+      "learning_rate": 1.2045530403033998e-06,
+      "loss": 0.1047,
+      "step": 2688
+    },
+    {
+      "epoch": 0.6111363636363636,
+      "grad_norm": 2.00156902692273,
+      "learning_rate": 1.204519623156042e-06,
+      "loss": 0.0833,
+      "step": 2689
+    },
+    {
+      "epoch": 0.6113636363636363,
+      "grad_norm": 1.6025352677036835,
+      "learning_rate": 1.204486194191268e-06,
+      "loss": 0.064,
+      "step": 2690
+    },
+    {
+      "epoch": 0.6115909090909091,
+      "grad_norm": 2.3494461622033627,
+      "learning_rate": 1.2044527534097595e-06,
+      "loss": 0.1129,
+      "step": 2691
+    },
+    {
+      "epoch": 0.6118181818181818,
+      "grad_norm": 1.0553866387635873,
+      "learning_rate": 1.2044193008121985e-06,
+      "loss": 0.0438,
+      "step": 2692
+    },
+    {
+      "epoch": 0.6120454545454546,
+      "grad_norm": 3.767037210076653,
+      "learning_rate": 1.2043858363992671e-06,
+      "loss": 0.1847,
+      "step": 2693
+    },
+    {
+      "epoch": 0.6122727272727273,
+      "grad_norm": 2.312919984855907,
+      "learning_rate": 1.2043523601716477e-06,
+      "loss": 0.0776,
+      "step": 2694
+    },
+    {
+      "epoch": 0.6125,
+      "grad_norm": 1.7117253550023581,
+      "learning_rate": 1.2043188721300229e-06,
+      "loss": 0.0731,
+      "step": 2695
+    },
+    {
+      "epoch": 0.6127272727272727,
+      "grad_norm": 2.35912804523733,
+      "learning_rate": 1.2042853722750756e-06,
+      "loss": 0.0636,
+      "step": 2696
+    },
+    {
+      "epoch": 0.6129545454545454,
+      "grad_norm": 1.6826495415237341,
+      "learning_rate": 1.204251860607489e-06,
+      "loss": 0.0635,
+      "step": 2697
+    },
+    {
+      "epoch": 0.6131818181818182,
+      "grad_norm": 1.6708913124861995,
+      "learning_rate": 1.2042183371279464e-06,
+      "loss": 0.0974,
+      "step": 2698
+    },
+    {
+      "epoch": 0.6134090909090909,
+      "grad_norm": 2.5318698554251022,
+      "learning_rate": 1.2041848018371314e-06,
+      "loss": 0.1191,
+      "step": 2699
+    },
+    {
+      "epoch": 0.6136363636363636,
+      "grad_norm": 1.9681209701767246,
+      "learning_rate": 1.2041512547357278e-06,
+      "loss": 0.0772,
+      "step": 2700
+    },
+    {
+      "epoch": 0.6138636363636364,
+      "grad_norm": 2.3841010530683193,
+      "learning_rate": 1.2041176958244197e-06,
+      "loss": 0.0891,
+      "step": 2701
+    },
+    {
+      "epoch": 0.6140909090909091,
+      "grad_norm": 1.730788512863966,
+      "learning_rate": 1.2040841251038915e-06,
+      "loss": 0.0958,
+      "step": 2702
+    },
+    {
+      "epoch": 0.6143181818181818,
+      "grad_norm": 1.87099828695,
+      "learning_rate": 1.2040505425748279e-06,
+      "loss": 0.1174,
+      "step": 2703
+    },
+    {
+      "epoch": 0.6145454545454545,
+      "grad_norm": 1.8334472098760894,
+      "learning_rate": 1.2040169482379132e-06,
+      "loss": 0.0783,
+      "step": 2704
+    },
+    {
+      "epoch": 0.6147727272727272,
+      "grad_norm": 1.4521534586898392,
+      "learning_rate": 1.203983342093833e-06,
+      "loss": 0.0771,
+      "step": 2705
+    },
+    {
+      "epoch": 0.615,
+      "grad_norm": 0.8157470198695895,
+      "learning_rate": 1.2039497241432724e-06,
+      "loss": 0.0272,
+      "step": 2706
+    },
+    {
+      "epoch": 0.6152272727272727,
+      "grad_norm": 1.5712454110611502,
+      "learning_rate": 1.2039160943869168e-06,
+      "loss": 0.084,
+      "step": 2707
+    },
+    {
+      "epoch": 0.6154545454545455,
+      "grad_norm": 2.2838647066361815,
+      "learning_rate": 1.203882452825452e-06,
+      "loss": 0.0966,
+      "step": 2708
+    },
+    {
+      "epoch": 0.6156818181818182,
+      "grad_norm": 1.644035208022435,
+      "learning_rate": 1.203848799459564e-06,
+      "loss": 0.0821,
+      "step": 2709
+    },
+    {
+      "epoch": 0.615909090909091,
+      "grad_norm": 2.29282779780281,
+      "learning_rate": 1.2038151342899395e-06,
+      "loss": 0.1092,
+      "step": 2710
+    },
+    {
+      "epoch": 0.6161363636363636,
+      "grad_norm": 2.8013303717523015,
+      "learning_rate": 1.2037814573172642e-06,
+      "loss": 0.0993,
+      "step": 2711
+    },
+    {
+      "epoch": 0.6163636363636363,
+      "grad_norm": 1.3843147826017947,
+      "learning_rate": 1.2037477685422255e-06,
+      "loss": 0.0749,
+      "step": 2712
+    },
+    {
+      "epoch": 0.6165909090909091,
+      "grad_norm": 2.209901099423209,
+      "learning_rate": 1.20371406796551e-06,
+      "loss": 0.0847,
+      "step": 2713
+    },
+    {
+      "epoch": 0.6168181818181818,
+      "grad_norm": 1.7326747243095593,
+      "learning_rate": 1.203680355587805e-06,
+      "loss": 0.0615,
+      "step": 2714
+    },
+    {
+      "epoch": 0.6170454545454546,
+      "grad_norm": 2.1873682941671775,
+      "learning_rate": 1.203646631409798e-06,
+      "loss": 0.1293,
+      "step": 2715
+    },
+    {
+      "epoch": 0.6172727272727273,
+      "grad_norm": 1.4692986773790602,
+      "learning_rate": 1.2036128954321768e-06,
+      "loss": 0.0323,
+      "step": 2716
+    },
+    {
+      "epoch": 0.6175,
+      "grad_norm": 2.5999745195027075,
+      "learning_rate": 1.2035791476556292e-06,
+      "loss": 0.1158,
+      "step": 2717
+    },
+    {
+      "epoch": 0.6177272727272727,
+      "grad_norm": 2.1876574558455717,
+      "learning_rate": 1.2035453880808435e-06,
+      "loss": 0.095,
+      "step": 2718
+    },
+    {
+      "epoch": 0.6179545454545454,
+      "grad_norm": 2.3542582169693254,
+      "learning_rate": 1.2035116167085077e-06,
+      "loss": 0.092,
+      "step": 2719
+    },
+    {
+      "epoch": 0.6181818181818182,
+      "grad_norm": 1.580915500606124,
+      "learning_rate": 1.2034778335393111e-06,
+      "loss": 0.1307,
+      "step": 2720
+    },
+    {
+      "epoch": 0.6184090909090909,
+      "grad_norm": 2.1984794245281467,
+      "learning_rate": 1.2034440385739418e-06,
+      "loss": 0.0847,
+      "step": 2721
+    },
+    {
+      "epoch": 0.6186363636363637,
+      "grad_norm": 2.1863835800943354,
+      "learning_rate": 1.2034102318130897e-06,
+      "loss": 0.0707,
+      "step": 2722
+    },
+    {
+      "epoch": 0.6188636363636364,
+      "grad_norm": 1.0303335246082843,
+      "learning_rate": 1.2033764132574439e-06,
+      "loss": 0.0685,
+      "step": 2723
+    },
+    {
+      "epoch": 0.6190909090909091,
+      "grad_norm": 1.6699123871285466,
+      "learning_rate": 1.2033425829076941e-06,
+      "loss": 0.0823,
+      "step": 2724
+    },
+    {
+      "epoch": 0.6193181818181818,
+      "grad_norm": 1.2622847507110209,
+      "learning_rate": 1.2033087407645297e-06,
+      "loss": 0.0549,
+      "step": 2725
+    },
+    {
+      "epoch": 0.6195454545454545,
+      "grad_norm": 3.002377277332965,
+      "learning_rate": 1.2032748868286415e-06,
+      "loss": 0.1049,
+      "step": 2726
+    },
+    {
+      "epoch": 0.6197727272727273,
+      "grad_norm": 1.9361186022331929,
+      "learning_rate": 1.2032410211007194e-06,
+      "loss": 0.0679,
+      "step": 2727
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4210759820500183,
+      "learning_rate": 1.203207143581454e-06,
+      "loss": 0.0712,
+      "step": 2728
+    },
+    {
+      "epoch": 0.6202272727272727,
+      "grad_norm": 2.430000769857462,
+      "learning_rate": 1.2031732542715363e-06,
+      "loss": 0.0926,
+      "step": 2729
+    },
+    {
+      "epoch": 0.6204545454545455,
+      "grad_norm": 1.315371886470588,
+      "learning_rate": 1.2031393531716573e-06,
+      "loss": 0.0619,
+      "step": 2730
+    },
+    {
+      "epoch": 0.6206818181818182,
+      "grad_norm": 2.4164038698383186,
+      "learning_rate": 1.2031054402825082e-06,
+      "loss": 0.1118,
+      "step": 2731
+    },
+    {
+      "epoch": 0.6209090909090909,
+      "grad_norm": 2.373794567592861,
+      "learning_rate": 1.2030715156047805e-06,
+      "loss": 0.0626,
+      "step": 2732
+    },
+    {
+      "epoch": 0.6211363636363636,
+      "grad_norm": 1.2890737615187022,
+      "learning_rate": 1.2030375791391663e-06,
+      "loss": 0.0463,
+      "step": 2733
+    },
+    {
+      "epoch": 0.6213636363636363,
+      "grad_norm": 1.9986408003909022,
+      "learning_rate": 1.2030036308863575e-06,
+      "loss": 0.0778,
+      "step": 2734
+    },
+    {
+      "epoch": 0.6215909090909091,
+      "grad_norm": 1.8089512650392563,
+      "learning_rate": 1.2029696708470461e-06,
+      "loss": 0.0699,
+      "step": 2735
+    },
+    {
+      "epoch": 0.6218181818181818,
+      "grad_norm": 1.2803169423627847,
+      "learning_rate": 1.2029356990219248e-06,
+      "loss": 0.0448,
+      "step": 2736
+    },
+    {
+      "epoch": 0.6220454545454546,
+      "grad_norm": 1.3369807174345374,
+      "learning_rate": 1.2029017154116863e-06,
+      "loss": 0.0736,
+      "step": 2737
+    },
+    {
+      "epoch": 0.6222727272727273,
+      "grad_norm": 1.658053354345458,
+      "learning_rate": 1.2028677200170236e-06,
+      "loss": 0.0783,
+      "step": 2738
+    },
+    {
+      "epoch": 0.6225,
+      "grad_norm": 1.8228597417702626,
+      "learning_rate": 1.20283371283863e-06,
+      "loss": 0.0801,
+      "step": 2739
+    },
+    {
+      "epoch": 0.6227272727272727,
+      "grad_norm": 0.8906116622819438,
+      "learning_rate": 1.2027996938771992e-06,
+      "loss": 0.0352,
+      "step": 2740
+    },
+    {
+      "epoch": 0.6229545454545454,
+      "grad_norm": 1.2869902520349028,
+      "learning_rate": 1.2027656631334242e-06,
+      "loss": 0.0535,
+      "step": 2741
+    },
+    {
+      "epoch": 0.6231818181818182,
+      "grad_norm": 1.2946496681783033,
+      "learning_rate": 1.2027316206079995e-06,
+      "loss": 0.0419,
+      "step": 2742
+    },
+    {
+      "epoch": 0.6234090909090909,
+      "grad_norm": 2.0714040448147344,
+      "learning_rate": 1.2026975663016191e-06,
+      "loss": 0.0871,
+      "step": 2743
+    },
+    {
+      "epoch": 0.6236363636363637,
+      "grad_norm": 2.2619270225484334,
+      "learning_rate": 1.2026635002149775e-06,
+      "loss": 0.1008,
+      "step": 2744
+    },
+    {
+      "epoch": 0.6238636363636364,
+      "grad_norm": 1.2728375696661645,
+      "learning_rate": 1.2026294223487693e-06,
+      "loss": 0.0723,
+      "step": 2745
+    },
+    {
+      "epoch": 0.6240909090909091,
+      "grad_norm": 1.8332702063454163,
+      "learning_rate": 1.2025953327036897e-06,
+      "loss": 0.0568,
+      "step": 2746
+    },
+    {
+      "epoch": 0.6243181818181818,
+      "grad_norm": 2.0295270503683467,
+      "learning_rate": 1.2025612312804333e-06,
+      "loss": 0.117,
+      "step": 2747
+    },
+    {
+      "epoch": 0.6245454545454545,
+      "grad_norm": 2.231411496996103,
+      "learning_rate": 1.2025271180796957e-06,
+      "loss": 0.0794,
+      "step": 2748
+    },
+    {
+      "epoch": 0.6247727272727273,
+      "grad_norm": 1.2317995677184084,
+      "learning_rate": 1.2024929931021727e-06,
+      "loss": 0.0653,
+      "step": 2749
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 1.5979753996952715,
+      "learning_rate": 1.2024588563485604e-06,
+      "loss": 0.1137,
+      "step": 2750
+    },
+    {
+      "epoch": 0.6252272727272727,
+      "grad_norm": 1.6609242015957866,
+      "learning_rate": 1.2024247078195542e-06,
+      "loss": 0.0472,
+      "step": 2751
+    },
+    {
+      "epoch": 0.6254545454545455,
+      "grad_norm": 2.039357151644558,
+      "learning_rate": 1.2023905475158511e-06,
+      "loss": 0.0771,
+      "step": 2752
+    },
+    {
+      "epoch": 0.6256818181818182,
+      "grad_norm": 2.415210514126609,
+      "learning_rate": 1.2023563754381472e-06,
+      "loss": 0.0827,
+      "step": 2753
+    },
+    {
+      "epoch": 0.6259090909090909,
+      "grad_norm": 1.334175039365404,
+      "learning_rate": 1.2023221915871398e-06,
+      "loss": 0.0338,
+      "step": 2754
+    },
+    {
+      "epoch": 0.6261363636363636,
+      "grad_norm": 1.542262717515143,
+      "learning_rate": 1.2022879959635255e-06,
+      "loss": 0.0556,
+      "step": 2755
+    },
+    {
+      "epoch": 0.6263636363636363,
+      "grad_norm": 3.0249656563234373,
+      "learning_rate": 1.202253788568002e-06,
+      "loss": 0.1119,
+      "step": 2756
+    },
+    {
+      "epoch": 0.6265909090909091,
+      "grad_norm": 1.5993767867148962,
+      "learning_rate": 1.2022195694012667e-06,
+      "loss": 0.0699,
+      "step": 2757
+    },
+    {
+      "epoch": 0.6268181818181818,
+      "grad_norm": 2.1427788537644714,
+      "learning_rate": 1.2021853384640172e-06,
+      "loss": 0.1176,
+      "step": 2758
+    },
+    {
+      "epoch": 0.6270454545454546,
+      "grad_norm": 1.6013286911817204,
+      "learning_rate": 1.2021510957569517e-06,
+      "loss": 0.0777,
+      "step": 2759
+    },
+    {
+      "epoch": 0.6272727272727273,
+      "grad_norm": 2.4682677727652282,
+      "learning_rate": 1.2021168412807684e-06,
+      "loss": 0.1137,
+      "step": 2760
+    },
+    {
+      "epoch": 0.6275,
+      "grad_norm": 6.5295865276318406,
+      "learning_rate": 1.202082575036166e-06,
+      "loss": 0.1785,
+      "step": 2761
+    },
+    {
+      "epoch": 0.6277272727272727,
+      "grad_norm": 2.8273228115522535,
+      "learning_rate": 1.2020482970238432e-06,
+      "loss": 0.1181,
+      "step": 2762
+    },
+    {
+      "epoch": 0.6279545454545454,
+      "grad_norm": 2.2441544291647686,
+      "learning_rate": 1.2020140072444988e-06,
+      "loss": 0.0777,
+      "step": 2763
+    },
+    {
+      "epoch": 0.6281818181818182,
+      "grad_norm": 1.8004465872268522,
+      "learning_rate": 1.2019797056988323e-06,
+      "loss": 0.0832,
+      "step": 2764
+    },
+    {
+      "epoch": 0.6284090909090909,
+      "grad_norm": 1.676200356065674,
+      "learning_rate": 1.2019453923875426e-06,
+      "loss": 0.0792,
+      "step": 2765
+    },
+    {
+      "epoch": 0.6286363636363637,
+      "grad_norm": 1.0751769718108544,
+      "learning_rate": 1.2019110673113302e-06,
+      "loss": 0.0533,
+      "step": 2766
+    },
+    {
+      "epoch": 0.6288636363636364,
+      "grad_norm": 1.8873340660493554,
+      "learning_rate": 1.2018767304708946e-06,
+      "loss": 0.0838,
+      "step": 2767
+    },
+    {
+      "epoch": 0.6290909090909091,
+      "grad_norm": 0.3223713620382924,
+      "learning_rate": 1.2018423818669359e-06,
+      "loss": 0.0151,
+      "step": 2768
+    },
+    {
+      "epoch": 0.6293181818181818,
+      "grad_norm": 1.3467711367239803,
+      "learning_rate": 1.2018080215001548e-06,
+      "loss": 0.0395,
+      "step": 2769
+    },
+    {
+      "epoch": 0.6295454545454545,
+      "grad_norm": 1.4471415668237795,
+      "learning_rate": 1.2017736493712517e-06,
+      "loss": 0.0511,
+      "step": 2770
+    },
+    {
+      "epoch": 0.6297727272727273,
+      "grad_norm": 1.905115742665941,
+      "learning_rate": 1.2017392654809278e-06,
+      "loss": 0.0961,
+      "step": 2771
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1446589276461987,
+      "learning_rate": 1.201704869829884e-06,
+      "loss": 0.0739,
+      "step": 2772
+    },
+    {
+      "epoch": 0.6302272727272727,
+      "grad_norm": 2.1946472282080283,
+      "learning_rate": 1.2016704624188219e-06,
+      "loss": 0.0822,
+      "step": 2773
+    },
+    {
+      "epoch": 0.6304545454545455,
+      "grad_norm": 1.2195937075278085,
+      "learning_rate": 1.2016360432484429e-06,
+      "loss": 0.0432,
+      "step": 2774
+    },
+    {
+      "epoch": 0.6306818181818182,
+      "grad_norm": 3.10817555572515,
+      "learning_rate": 1.2016016123194491e-06,
+      "loss": 0.1292,
+      "step": 2775
+    },
+    {
+      "epoch": 0.6309090909090909,
+      "grad_norm": 2.171983408710466,
+      "learning_rate": 1.2015671696325423e-06,
+      "loss": 0.0844,
+      "step": 2776
+    },
+    {
+      "epoch": 0.6311363636363636,
+      "grad_norm": 1.692512170704687,
+      "learning_rate": 1.2015327151884253e-06,
+      "loss": 0.0601,
+      "step": 2777
+    },
+    {
+      "epoch": 0.6313636363636363,
+      "grad_norm": 1.9514986311884983,
+      "learning_rate": 1.2014982489878001e-06,
+      "loss": 0.1071,
+      "step": 2778
+    },
+    {
+      "epoch": 0.6315909090909091,
+      "grad_norm": 2.5096169333022984,
+      "learning_rate": 1.2014637710313701e-06,
+      "loss": 0.0673,
+      "step": 2779
+    },
+    {
+      "epoch": 0.6318181818181818,
+      "grad_norm": 1.8070740227338022,
+      "learning_rate": 1.201429281319838e-06,
+      "loss": 0.106,
+      "step": 2780
+    },
+    {
+      "epoch": 0.6320454545454546,
+      "grad_norm": 2.7505241206066136,
+      "learning_rate": 1.2013947798539073e-06,
+      "loss": 0.1437,
+      "step": 2781
+    },
+    {
+      "epoch": 0.6322727272727273,
+      "grad_norm": 2.8009250766193388,
+      "learning_rate": 1.2013602666342813e-06,
+      "loss": 0.1395,
+      "step": 2782
+    },
+    {
+      "epoch": 0.6325,
+      "grad_norm": 1.6416417935035441,
+      "learning_rate": 1.2013257416616642e-06,
+      "loss": 0.0743,
+      "step": 2783
+    },
+    {
+      "epoch": 0.6327272727272727,
+      "grad_norm": 1.8880519034718402,
+      "learning_rate": 1.2012912049367597e-06,
+      "loss": 0.0885,
+      "step": 2784
+    },
+    {
+      "epoch": 0.6329545454545454,
+      "grad_norm": 1.0589396270445648,
+      "learning_rate": 1.2012566564602721e-06,
+      "loss": 0.0449,
+      "step": 2785
+    },
+    {
+      "epoch": 0.6331818181818182,
+      "grad_norm": 1.851462030787762,
+      "learning_rate": 1.2012220962329058e-06,
+      "loss": 0.082,
+      "step": 2786
+    },
+    {
+      "epoch": 0.6334090909090909,
+      "grad_norm": 2.191264245400896,
+      "learning_rate": 1.201187524255366e-06,
+      "loss": 0.0835,
+      "step": 2787
+    },
+    {
+      "epoch": 0.6336363636363637,
+      "grad_norm": 2.3554071345494254,
+      "learning_rate": 1.201152940528357e-06,
+      "loss": 0.1176,
+      "step": 2788
+    },
+    {
+      "epoch": 0.6338636363636364,
+      "grad_norm": 1.7108400792642384,
+      "learning_rate": 1.2011183450525846e-06,
+      "loss": 0.0647,
+      "step": 2789
+    },
+    {
+      "epoch": 0.634090909090909,
+      "grad_norm": 2.536159731465754,
+      "learning_rate": 1.201083737828754e-06,
+      "loss": 0.1311,
+      "step": 2790
+    },
+    {
+      "epoch": 0.6343181818181818,
+      "grad_norm": 2.6591233842300297,
+      "learning_rate": 1.201049118857571e-06,
+      "loss": 0.0637,
+      "step": 2791
+    },
+    {
+      "epoch": 0.6345454545454545,
+      "grad_norm": 1.4366521088963762,
+      "learning_rate": 1.2010144881397414e-06,
+      "loss": 0.0841,
+      "step": 2792
+    },
+    {
+      "epoch": 0.6347727272727273,
+      "grad_norm": 1.628525099546028,
+      "learning_rate": 1.2009798456759713e-06,
+      "loss": 0.0707,
+      "step": 2793
+    },
+    {
+      "epoch": 0.635,
+      "grad_norm": 1.427738883174935,
+      "learning_rate": 1.2009451914669675e-06,
+      "loss": 0.0681,
+      "step": 2794
+    },
+    {
+      "epoch": 0.6352272727272728,
+      "grad_norm": 2.320548575619766,
+      "learning_rate": 1.2009105255134366e-06,
+      "loss": 0.1216,
+      "step": 2795
+    },
+    {
+      "epoch": 0.6354545454545455,
+      "grad_norm": 2.6089131165954744,
+      "learning_rate": 1.2008758478160853e-06,
+      "loss": 0.1702,
+      "step": 2796
+    },
+    {
+      "epoch": 0.6356818181818182,
+      "grad_norm": 2.5333253521528976,
+      "learning_rate": 1.2008411583756208e-06,
+      "loss": 0.1047,
+      "step": 2797
+    },
+    {
+      "epoch": 0.6359090909090909,
+      "grad_norm": 1.8763741756020003,
+      "learning_rate": 1.2008064571927502e-06,
+      "loss": 0.0713,
+      "step": 2798
+    },
+    {
+      "epoch": 0.6361363636363636,
+      "grad_norm": 1.715857523829253,
+      "learning_rate": 1.2007717442681816e-06,
+      "loss": 0.0742,
+      "step": 2799
+    },
+    {
+      "epoch": 0.6363636363636364,
+      "grad_norm": 2.54288566596264,
+      "learning_rate": 1.2007370196026228e-06,
+      "loss": 0.158,
+      "step": 2800
+    },
+    {
+      "epoch": 0.6365909090909091,
+      "grad_norm": 1.4588155298893413,
+      "learning_rate": 1.2007022831967813e-06,
+      "loss": 0.0771,
+      "step": 2801
+    },
+    {
+      "epoch": 0.6368181818181818,
+      "grad_norm": 3.0610492513208327,
+      "learning_rate": 1.2006675350513663e-06,
+      "loss": 0.0969,
+      "step": 2802
+    },
+    {
+      "epoch": 0.6370454545454546,
+      "grad_norm": 2.137847940435915,
+      "learning_rate": 1.2006327751670857e-06,
+      "loss": 0.0998,
+      "step": 2803
+    },
+    {
+      "epoch": 0.6372727272727273,
+      "grad_norm": 1.2697158666316657,
+      "learning_rate": 1.2005980035446488e-06,
+      "loss": 0.0594,
+      "step": 2804
+    },
+    {
+      "epoch": 0.6375,
+      "grad_norm": 1.536341611570717,
+      "learning_rate": 1.2005632201847641e-06,
+      "loss": 0.0488,
+      "step": 2805
+    },
+    {
+      "epoch": 0.6377272727272727,
+      "grad_norm": 1.6921028499050814,
+      "learning_rate": 1.2005284250881417e-06,
+      "loss": 0.0813,
+      "step": 2806
+    },
+    {
+      "epoch": 0.6379545454545454,
+      "grad_norm": 1.159402272943863,
+      "learning_rate": 1.2004936182554903e-06,
+      "loss": 0.0448,
+      "step": 2807
+    },
+    {
+      "epoch": 0.6381818181818182,
+      "grad_norm": 1.9178702706174064,
+      "learning_rate": 1.20045879968752e-06,
+      "loss": 0.0884,
+      "step": 2808
+    },
+    {
+      "epoch": 0.6384090909090909,
+      "grad_norm": 1.2522610506351823,
+      "learning_rate": 1.200423969384941e-06,
+      "loss": 0.0484,
+      "step": 2809
+    },
+    {
+      "epoch": 0.6386363636363637,
+      "grad_norm": 1.9584787387664584,
+      "learning_rate": 1.200389127348463e-06,
+      "loss": 0.1358,
+      "step": 2810
+    },
+    {
+      "epoch": 0.6388636363636364,
+      "grad_norm": 0.4397944343172517,
+      "learning_rate": 1.2003542735787973e-06,
+      "loss": 0.0248,
+      "step": 2811
+    },
+    {
+      "epoch": 0.639090909090909,
+      "grad_norm": 2.599053256000105,
+      "learning_rate": 1.200319408076654e-06,
+      "loss": 0.1412,
+      "step": 2812
+    },
+    {
+      "epoch": 0.6393181818181818,
+      "grad_norm": 2.974619888325304,
+      "learning_rate": 1.2002845308427444e-06,
+      "loss": 0.1204,
+      "step": 2813
+    },
+    {
+      "epoch": 0.6395454545454545,
+      "grad_norm": 1.4882595938566046,
+      "learning_rate": 1.2002496418777793e-06,
+      "loss": 0.0765,
+      "step": 2814
+    },
+    {
+      "epoch": 0.6397727272727273,
+      "grad_norm": 2.0651435478825455,
+      "learning_rate": 1.2002147411824708e-06,
+      "loss": 0.0375,
+      "step": 2815
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.8552653033229616,
+      "learning_rate": 1.20017982875753e-06,
+      "loss": 0.1503,
+      "step": 2816
+    },
+    {
+      "epoch": 0.6402272727272728,
+      "grad_norm": 2.835382321798183,
+      "learning_rate": 1.2001449046036692e-06,
+      "loss": 0.0683,
+      "step": 2817
+    },
+    {
+      "epoch": 0.6404545454545455,
+      "grad_norm": 1.499863055733183,
+      "learning_rate": 1.2001099687216002e-06,
+      "loss": 0.0395,
+      "step": 2818
+    },
+    {
+      "epoch": 0.6406818181818181,
+      "grad_norm": 1.8343309397148817,
+      "learning_rate": 1.2000750211120356e-06,
+      "loss": 0.0999,
+      "step": 2819
+    },
+    {
+      "epoch": 0.6409090909090909,
+      "grad_norm": 2.2426929587556743,
+      "learning_rate": 1.200040061775688e-06,
+      "loss": 0.0936,
+      "step": 2820
+    },
+    {
+      "epoch": 0.6411363636363636,
+      "grad_norm": 1.7022797653902806,
+      "learning_rate": 1.2000050907132705e-06,
+      "loss": 0.1112,
+      "step": 2821
+    },
+    {
+      "epoch": 0.6413636363636364,
+      "grad_norm": 1.3063828520993694,
+      "learning_rate": 1.199970107925496e-06,
+      "loss": 0.0467,
+      "step": 2822
+    },
+    {
+      "epoch": 0.6415909090909091,
+      "grad_norm": 1.7537647434540795,
+      "learning_rate": 1.199935113413078e-06,
+      "loss": 0.0903,
+      "step": 2823
+    },
+    {
+      "epoch": 0.6418181818181818,
+      "grad_norm": 2.321294044295073,
+      "learning_rate": 1.19990010717673e-06,
+      "loss": 0.1119,
+      "step": 2824
+    },
+    {
+      "epoch": 0.6420454545454546,
+      "grad_norm": 2.301121831073146,
+      "learning_rate": 1.1998650892171657e-06,
+      "loss": 0.1242,
+      "step": 2825
+    },
+    {
+      "epoch": 0.6422727272727272,
+      "grad_norm": 2.0967762703686734,
+      "learning_rate": 1.1998300595350993e-06,
+      "loss": 0.0999,
+      "step": 2826
+    },
+    {
+      "epoch": 0.6425,
+      "grad_norm": 2.807455699109957,
+      "learning_rate": 1.1997950181312453e-06,
+      "loss": 0.0865,
+      "step": 2827
+    },
+    {
+      "epoch": 0.6427272727272727,
+      "grad_norm": 1.2342356288665468,
+      "learning_rate": 1.1997599650063179e-06,
+      "loss": 0.0812,
+      "step": 2828
+    },
+    {
+      "epoch": 0.6429545454545454,
+      "grad_norm": 1.6784646835275798,
+      "learning_rate": 1.1997249001610322e-06,
+      "loss": 0.0749,
+      "step": 2829
+    },
+    {
+      "epoch": 0.6431818181818182,
+      "grad_norm": 1.1949960878360781,
+      "learning_rate": 1.1996898235961032e-06,
+      "loss": 0.034,
+      "step": 2830
+    },
+    {
+      "epoch": 0.6434090909090909,
+      "grad_norm": 1.141960410353299,
+      "learning_rate": 1.1996547353122461e-06,
+      "loss": 0.0406,
+      "step": 2831
+    },
+    {
+      "epoch": 0.6436363636363637,
+      "grad_norm": 2.2126969968254135,
+      "learning_rate": 1.1996196353101764e-06,
+      "loss": 0.0914,
+      "step": 2832
+    },
+    {
+      "epoch": 0.6438636363636364,
+      "grad_norm": 2.0281621517959145,
+      "learning_rate": 1.1995845235906098e-06,
+      "loss": 0.0834,
+      "step": 2833
+    },
+    {
+      "epoch": 0.644090909090909,
+      "grad_norm": 1.149834077643026,
+      "learning_rate": 1.1995494001542622e-06,
+      "loss": 0.0515,
+      "step": 2834
+    },
+    {
+      "epoch": 0.6443181818181818,
+      "grad_norm": 1.722588812903194,
+      "learning_rate": 1.1995142650018502e-06,
+      "loss": 0.0806,
+      "step": 2835
+    },
+    {
+      "epoch": 0.6445454545454545,
+      "grad_norm": 1.6110292623568108,
+      "learning_rate": 1.1994791181340897e-06,
+      "loss": 0.0647,
+      "step": 2836
+    },
+    {
+      "epoch": 0.6447727272727273,
+      "grad_norm": 1.663462164837718,
+      "learning_rate": 1.1994439595516982e-06,
+      "loss": 0.1169,
+      "step": 2837
+    },
+    {
+      "epoch": 0.645,
+      "grad_norm": 2.3035066836206957,
+      "learning_rate": 1.199408789255392e-06,
+      "loss": 0.1183,
+      "step": 2838
+    },
+    {
+      "epoch": 0.6452272727272728,
+      "grad_norm": 3.524071967574788,
+      "learning_rate": 1.1993736072458884e-06,
+      "loss": 0.1769,
+      "step": 2839
+    },
+    {
+      "epoch": 0.6454545454545455,
+      "grad_norm": 1.9974021974976148,
+      "learning_rate": 1.199338413523905e-06,
+      "loss": 0.1277,
+      "step": 2840
+    },
+    {
+      "epoch": 0.6456818181818181,
+      "grad_norm": 2.582577993665508,
+      "learning_rate": 1.1993032080901593e-06,
+      "loss": 0.1015,
+      "step": 2841
+    },
+    {
+      "epoch": 0.6459090909090909,
+      "grad_norm": 1.8270630568147168,
+      "learning_rate": 1.1992679909453693e-06,
+      "loss": 0.1103,
+      "step": 2842
+    },
+    {
+      "epoch": 0.6461363636363636,
+      "grad_norm": 0.5433756920680516,
+      "learning_rate": 1.199232762090253e-06,
+      "loss": 0.0357,
+      "step": 2843
+    },
+    {
+      "epoch": 0.6463636363636364,
+      "grad_norm": 2.3222202716145914,
+      "learning_rate": 1.1991975215255292e-06,
+      "loss": 0.1122,
+      "step": 2844
+    },
+    {
+      "epoch": 0.6465909090909091,
+      "grad_norm": 1.181373482864075,
+      "learning_rate": 1.1991622692519158e-06,
+      "loss": 0.0583,
+      "step": 2845
+    },
+    {
+      "epoch": 0.6468181818181818,
+      "grad_norm": 1.7977704100350211,
+      "learning_rate": 1.1991270052701323e-06,
+      "loss": 0.0592,
+      "step": 2846
+    },
+    {
+      "epoch": 0.6470454545454546,
+      "grad_norm": 2.456168634776501,
+      "learning_rate": 1.1990917295808974e-06,
+      "loss": 0.0967,
+      "step": 2847
+    },
+    {
+      "epoch": 0.6472727272727272,
+      "grad_norm": 1.7800750099024378,
+      "learning_rate": 1.1990564421849306e-06,
+      "loss": 0.0698,
+      "step": 2848
+    },
+    {
+      "epoch": 0.6475,
+      "grad_norm": 3.111296463485175,
+      "learning_rate": 1.1990211430829514e-06,
+      "loss": 0.1511,
+      "step": 2849
+    },
+    {
+      "epoch": 0.6477272727272727,
+      "grad_norm": 1.310484564508483,
+      "learning_rate": 1.1989858322756797e-06,
+      "loss": 0.0473,
+      "step": 2850
+    },
+    {
+      "epoch": 0.6479545454545454,
+      "grad_norm": 1.534480468698642,
+      "learning_rate": 1.1989505097638357e-06,
+      "loss": 0.0502,
+      "step": 2851
+    },
+    {
+      "epoch": 0.6481818181818182,
+      "grad_norm": 1.7060341551022182,
+      "learning_rate": 1.1989151755481393e-06,
+      "loss": 0.066,
+      "step": 2852
+    },
+    {
+      "epoch": 0.6484090909090909,
+      "grad_norm": 2.1774547975142875,
+      "learning_rate": 1.1988798296293112e-06,
+      "loss": 0.138,
+      "step": 2853
+    },
+    {
+      "epoch": 0.6486363636363637,
+      "grad_norm": 1.7523721896139135,
+      "learning_rate": 1.1988444720080723e-06,
+      "loss": 0.1057,
+      "step": 2854
+    },
+    {
+      "epoch": 0.6488636363636363,
+      "grad_norm": 1.3910509612086217,
+      "learning_rate": 1.1988091026851434e-06,
+      "loss": 0.0713,
+      "step": 2855
+    },
+    {
+      "epoch": 0.649090909090909,
+      "grad_norm": 2.3494030818374774,
+      "learning_rate": 1.198773721661246e-06,
+      "loss": 0.131,
+      "step": 2856
+    },
+    {
+      "epoch": 0.6493181818181818,
+      "grad_norm": 1.591454238579702,
+      "learning_rate": 1.1987383289371011e-06,
+      "loss": 0.0691,
+      "step": 2857
+    },
+    {
+      "epoch": 0.6495454545454545,
+      "grad_norm": 1.8732716996872931,
+      "learning_rate": 1.198702924513431e-06,
+      "loss": 0.0705,
+      "step": 2858
+    },
+    {
+      "epoch": 0.6497727272727273,
+      "grad_norm": 0.6713321645525313,
+      "learning_rate": 1.198667508390957e-06,
+      "loss": 0.0409,
+      "step": 2859
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.658773772307118,
+      "learning_rate": 1.1986320805704021e-06,
+      "loss": 0.0929,
+      "step": 2860
+    },
+    {
+      "epoch": 0.6502272727272728,
+      "grad_norm": 1.1233753740793804,
+      "learning_rate": 1.1985966410524883e-06,
+      "loss": 0.0762,
+      "step": 2861
+    },
+    {
+      "epoch": 0.6504545454545455,
+      "grad_norm": 1.77073955188842,
+      "learning_rate": 1.1985611898379381e-06,
+      "loss": 0.0925,
+      "step": 2862
+    },
+    {
+      "epoch": 0.6506818181818181,
+      "grad_norm": 1.4610166579765138,
+      "learning_rate": 1.1985257269274748e-06,
+      "loss": 0.0899,
+      "step": 2863
+    },
+    {
+      "epoch": 0.6509090909090909,
+      "grad_norm": 2.232674960518595,
+      "learning_rate": 1.1984902523218213e-06,
+      "loss": 0.1053,
+      "step": 2864
+    },
+    {
+      "epoch": 0.6511363636363636,
+      "grad_norm": 1.9002817417837703,
+      "learning_rate": 1.198454766021701e-06,
+      "loss": 0.0508,
+      "step": 2865
+    },
+    {
+      "epoch": 0.6513636363636364,
+      "grad_norm": 1.6905639873830671,
+      "learning_rate": 1.1984192680278376e-06,
+      "loss": 0.0998,
+      "step": 2866
+    },
+    {
+      "epoch": 0.6515909090909091,
+      "grad_norm": 1.4065434859818637,
+      "learning_rate": 1.1983837583409549e-06,
+      "loss": 0.0575,
+      "step": 2867
+    },
+    {
+      "epoch": 0.6518181818181819,
+      "grad_norm": 2.5514765042162386,
+      "learning_rate": 1.198348236961777e-06,
+      "loss": 0.1507,
+      "step": 2868
+    },
+    {
+      "epoch": 0.6520454545454546,
+      "grad_norm": 2.1901110420468255,
+      "learning_rate": 1.1983127038910283e-06,
+      "loss": 0.103,
+      "step": 2869
+    },
+    {
+      "epoch": 0.6522727272727272,
+      "grad_norm": 2.5229362316997612,
+      "learning_rate": 1.1982771591294335e-06,
+      "loss": 0.1119,
+      "step": 2870
+    },
+    {
+      "epoch": 0.6525,
+      "grad_norm": 2.0788293020826685,
+      "learning_rate": 1.1982416026777172e-06,
+      "loss": 0.0699,
+      "step": 2871
+    },
+    {
+      "epoch": 0.6527272727272727,
+      "grad_norm": 1.6612777300889554,
+      "learning_rate": 1.1982060345366046e-06,
+      "loss": 0.0602,
+      "step": 2872
+    },
+    {
+      "epoch": 0.6529545454545455,
+      "grad_norm": 2.216286099702645,
+      "learning_rate": 1.1981704547068208e-06,
+      "loss": 0.1067,
+      "step": 2873
+    },
+    {
+      "epoch": 0.6531818181818182,
+      "grad_norm": 1.912556695920962,
+      "learning_rate": 1.1981348631890916e-06,
+      "loss": 0.1209,
+      "step": 2874
+    },
+    {
+      "epoch": 0.6534090909090909,
+      "grad_norm": 3.1010135215772965,
+      "learning_rate": 1.1980992599841426e-06,
+      "loss": 0.1068,
+      "step": 2875
+    },
+    {
+      "epoch": 0.6536363636363637,
+      "grad_norm": 2.046252197369452,
+      "learning_rate": 1.1980636450926999e-06,
+      "loss": 0.1532,
+      "step": 2876
+    },
+    {
+      "epoch": 0.6538636363636363,
+      "grad_norm": 2.790766644600982,
+      "learning_rate": 1.1980280185154896e-06,
+      "loss": 0.091,
+      "step": 2877
+    },
+    {
+      "epoch": 0.6540909090909091,
+      "grad_norm": 2.3470041491996967,
+      "learning_rate": 1.1979923802532386e-06,
+      "loss": 0.07,
+      "step": 2878
+    },
+    {
+      "epoch": 0.6543181818181818,
+      "grad_norm": 1.8958451397701002,
+      "learning_rate": 1.1979567303066731e-06,
+      "loss": 0.0585,
+      "step": 2879
+    },
+    {
+      "epoch": 0.6545454545454545,
+      "grad_norm": 1.4664226992627907,
+      "learning_rate": 1.1979210686765203e-06,
+      "loss": 0.0588,
+      "step": 2880
+    },
+    {
+      "epoch": 0.6547727272727273,
+      "grad_norm": 2.2251094910439453,
+      "learning_rate": 1.1978853953635074e-06,
+      "loss": 0.0892,
+      "step": 2881
+    },
+    {
+      "epoch": 0.655,
+      "grad_norm": 1.6382855235735432,
+      "learning_rate": 1.1978497103683618e-06,
+      "loss": 0.052,
+      "step": 2882
+    },
+    {
+      "epoch": 0.6552272727272728,
+      "grad_norm": 1.2905375842785058,
+      "learning_rate": 1.1978140136918113e-06,
+      "loss": 0.0458,
+      "step": 2883
+    },
+    {
+      "epoch": 0.6554545454545454,
+      "grad_norm": 1.323755451134127,
+      "learning_rate": 1.1977783053345837e-06,
+      "loss": 0.0481,
+      "step": 2884
+    },
+    {
+      "epoch": 0.6556818181818181,
+      "grad_norm": 1.3043381182041067,
+      "learning_rate": 1.1977425852974072e-06,
+      "loss": 0.0678,
+      "step": 2885
+    },
+    {
+      "epoch": 0.6559090909090909,
+      "grad_norm": 3.081442694870073,
+      "learning_rate": 1.1977068535810101e-06,
+      "loss": 0.0872,
+      "step": 2886
+    },
+    {
+      "epoch": 0.6561363636363636,
+      "grad_norm": 1.8724875676716217,
+      "learning_rate": 1.1976711101861213e-06,
+      "loss": 0.0975,
+      "step": 2887
+    },
+    {
+      "epoch": 0.6563636363636364,
+      "grad_norm": 1.8632687375017623,
+      "learning_rate": 1.1976353551134692e-06,
+      "loss": 0.1348,
+      "step": 2888
+    },
+    {
+      "epoch": 0.6565909090909091,
+      "grad_norm": 2.042032135613588,
+      "learning_rate": 1.1975995883637835e-06,
+      "loss": 0.0758,
+      "step": 2889
+    },
+    {
+      "epoch": 0.6568181818181819,
+      "grad_norm": 1.8420145708651958,
+      "learning_rate": 1.197563809937793e-06,
+      "loss": 0.0572,
+      "step": 2890
+    },
+    {
+      "epoch": 0.6570454545454546,
+      "grad_norm": 3.082324578017795,
+      "learning_rate": 1.1975280198362276e-06,
+      "loss": 0.2249,
+      "step": 2891
+    },
+    {
+      "epoch": 0.6572727272727272,
+      "grad_norm": 1.778237991242713,
+      "learning_rate": 1.1974922180598172e-06,
+      "loss": 0.0569,
+      "step": 2892
+    },
+    {
+      "epoch": 0.6575,
+      "grad_norm": 1.5183686390690982,
+      "learning_rate": 1.1974564046092916e-06,
+      "loss": 0.0411,
+      "step": 2893
+    },
+    {
+      "epoch": 0.6577272727272727,
+      "grad_norm": 1.7949964917333234,
+      "learning_rate": 1.197420579485381e-06,
+      "loss": 0.0992,
+      "step": 2894
+    },
+    {
+      "epoch": 0.6579545454545455,
+      "grad_norm": 0.9064339773129967,
+      "learning_rate": 1.1973847426888164e-06,
+      "loss": 0.0261,
+      "step": 2895
+    },
+    {
+      "epoch": 0.6581818181818182,
+      "grad_norm": 1.7352741836242742,
+      "learning_rate": 1.1973488942203282e-06,
+      "loss": 0.0507,
+      "step": 2896
+    },
+    {
+      "epoch": 0.6584090909090909,
+      "grad_norm": 1.864562840498135,
+      "learning_rate": 1.1973130340806475e-06,
+      "loss": 0.0586,
+      "step": 2897
+    },
+    {
+      "epoch": 0.6586363636363637,
+      "grad_norm": 2.4346622268708624,
+      "learning_rate": 1.1972771622705056e-06,
+      "loss": 0.1017,
+      "step": 2898
+    },
+    {
+      "epoch": 0.6588636363636363,
+      "grad_norm": 1.9939545185521317,
+      "learning_rate": 1.197241278790634e-06,
+      "loss": 0.0974,
+      "step": 2899
+    },
+    {
+      "epoch": 0.6590909090909091,
+      "grad_norm": 2.706078090329856,
+      "learning_rate": 1.1972053836417646e-06,
+      "loss": 0.1265,
+      "step": 2900
+    },
+    {
+      "epoch": 0.6593181818181818,
+      "grad_norm": 1.7931594556301123,
+      "learning_rate": 1.197169476824629e-06,
+      "loss": 0.0719,
+      "step": 2901
+    },
+    {
+      "epoch": 0.6595454545454545,
+      "grad_norm": 1.7226188845349641,
+      "learning_rate": 1.1971335583399593e-06,
+      "loss": 0.0686,
+      "step": 2902
+    },
+    {
+      "epoch": 0.6597727272727273,
+      "grad_norm": 1.7350144009976884,
+      "learning_rate": 1.1970976281884886e-06,
+      "loss": 0.1134,
+      "step": 2903
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.026465341046622,
+      "learning_rate": 1.1970616863709488e-06,
+      "loss": 0.1428,
+      "step": 2904
+    },
+    {
+      "epoch": 0.6602272727272728,
+      "grad_norm": 1.314434744359462,
+      "learning_rate": 1.1970257328880735e-06,
+      "loss": 0.0588,
+      "step": 2905
+    },
+    {
+      "epoch": 0.6604545454545454,
+      "grad_norm": 2.5238867242522125,
+      "learning_rate": 1.1969897677405956e-06,
+      "loss": 0.0871,
+      "step": 2906
+    },
+    {
+      "epoch": 0.6606818181818181,
+      "grad_norm": 2.649953534648723,
+      "learning_rate": 1.1969537909292483e-06,
+      "loss": 0.0773,
+      "step": 2907
+    },
+    {
+      "epoch": 0.6609090909090909,
+      "grad_norm": 2.6649852386689807,
+      "learning_rate": 1.1969178024547653e-06,
+      "loss": 0.0949,
+      "step": 2908
+    },
+    {
+      "epoch": 0.6611363636363636,
+      "grad_norm": 1.5991047776661518,
+      "learning_rate": 1.1968818023178809e-06,
+      "loss": 0.0605,
+      "step": 2909
+    },
+    {
+      "epoch": 0.6613636363636364,
+      "grad_norm": 2.047546998475509,
+      "learning_rate": 1.1968457905193284e-06,
+      "loss": 0.0473,
+      "step": 2910
+    },
+    {
+      "epoch": 0.6615909090909091,
+      "grad_norm": 1.6100690159543594,
+      "learning_rate": 1.1968097670598428e-06,
+      "loss": 0.0643,
+      "step": 2911
+    },
+    {
+      "epoch": 0.6618181818181819,
+      "grad_norm": 2.0043399948432277,
+      "learning_rate": 1.1967737319401585e-06,
+      "loss": 0.0857,
+      "step": 2912
+    },
+    {
+      "epoch": 0.6620454545454545,
+      "grad_norm": 2.213298490605103,
+      "learning_rate": 1.19673768516101e-06,
+      "loss": 0.1185,
+      "step": 2913
+    },
+    {
+      "epoch": 0.6622727272727272,
+      "grad_norm": 2.0778849859535606,
+      "learning_rate": 1.1967016267231329e-06,
+      "loss": 0.0861,
+      "step": 2914
+    },
+    {
+      "epoch": 0.6625,
+      "grad_norm": 1.9812488161689261,
+      "learning_rate": 1.1966655566272623e-06,
+      "loss": 0.1011,
+      "step": 2915
+    },
+    {
+      "epoch": 0.6627272727272727,
+      "grad_norm": 1.9255650563055717,
+      "learning_rate": 1.1966294748741336e-06,
+      "loss": 0.112,
+      "step": 2916
+    },
+    {
+      "epoch": 0.6629545454545455,
+      "grad_norm": 1.935240452123203,
+      "learning_rate": 1.1965933814644825e-06,
+      "loss": 0.0932,
+      "step": 2917
+    },
+    {
+      "epoch": 0.6631818181818182,
+      "grad_norm": 2.0522718991955546,
+      "learning_rate": 1.1965572763990454e-06,
+      "loss": 0.1212,
+      "step": 2918
+    },
+    {
+      "epoch": 0.663409090909091,
+      "grad_norm": 2.8512150820153193,
+      "learning_rate": 1.1965211596785582e-06,
+      "loss": 0.1645,
+      "step": 2919
+    },
+    {
+      "epoch": 0.6636363636363637,
+      "grad_norm": 1.4699348858154087,
+      "learning_rate": 1.1964850313037576e-06,
+      "loss": 0.0619,
+      "step": 2920
+    },
+    {
+      "epoch": 0.6638636363636363,
+      "grad_norm": 1.538920632304403,
+      "learning_rate": 1.19644889127538e-06,
+      "loss": 0.0835,
+      "step": 2921
+    },
+    {
+      "epoch": 0.6640909090909091,
+      "grad_norm": 1.7199030775695106,
+      "learning_rate": 1.1964127395941627e-06,
+      "loss": 0.073,
+      "step": 2922
+    },
+    {
+      "epoch": 0.6643181818181818,
+      "grad_norm": 1.8573873613955478,
+      "learning_rate": 1.1963765762608426e-06,
+      "loss": 0.1091,
+      "step": 2923
+    },
+    {
+      "epoch": 0.6645454545454546,
+      "grad_norm": 2.5886273049650006,
+      "learning_rate": 1.1963404012761573e-06,
+      "loss": 0.1018,
+      "step": 2924
+    },
+    {
+      "epoch": 0.6647727272727273,
+      "grad_norm": 1.0760830756804765,
+      "learning_rate": 1.1963042146408447e-06,
+      "loss": 0.0271,
+      "step": 2925
+    },
+    {
+      "epoch": 0.665,
+      "grad_norm": 0.6073579693748098,
+      "learning_rate": 1.1962680163556424e-06,
+      "loss": 0.0359,
+      "step": 2926
+    },
+    {
+      "epoch": 0.6652272727272728,
+      "grad_norm": 1.4420583061707182,
+      "learning_rate": 1.1962318064212882e-06,
+      "loss": 0.1033,
+      "step": 2927
+    },
+    {
+      "epoch": 0.6654545454545454,
+      "grad_norm": 1.738089842229266,
+      "learning_rate": 1.1961955848385213e-06,
+      "loss": 0.0695,
+      "step": 2928
+    },
+    {
+      "epoch": 0.6656818181818182,
+      "grad_norm": 2.0381496745300236,
+      "learning_rate": 1.19615935160808e-06,
+      "loss": 0.0852,
+      "step": 2929
+    },
+    {
+      "epoch": 0.6659090909090909,
+      "grad_norm": 1.6913825989545193,
+      "learning_rate": 1.1961231067307027e-06,
+      "loss": 0.0691,
+      "step": 2930
+    },
+    {
+      "epoch": 0.6661363636363636,
+      "grad_norm": 1.8027815840687806,
+      "learning_rate": 1.1960868502071294e-06,
+      "loss": 0.0855,
+      "step": 2931
+    },
+    {
+      "epoch": 0.6663636363636364,
+      "grad_norm": 2.8050100099917623,
+      "learning_rate": 1.1960505820380984e-06,
+      "loss": 0.1656,
+      "step": 2932
+    },
+    {
+      "epoch": 0.6665909090909091,
+      "grad_norm": 3.5082355035397774,
+      "learning_rate": 1.19601430222435e-06,
+      "loss": 0.1617,
+      "step": 2933
+    },
+    {
+      "epoch": 0.6668181818181819,
+      "grad_norm": 2.0453596686360904,
+      "learning_rate": 1.1959780107666238e-06,
+      "loss": 0.0806,
+      "step": 2934
+    },
+    {
+      "epoch": 0.6670454545454545,
+      "grad_norm": 1.5559763655986945,
+      "learning_rate": 1.19594170766566e-06,
+      "loss": 0.0494,
+      "step": 2935
+    },
+    {
+      "epoch": 0.6672727272727272,
+      "grad_norm": 1.5756019162301427,
+      "learning_rate": 1.1959053929221984e-06,
+      "loss": 0.0627,
+      "step": 2936
+    },
+    {
+      "epoch": 0.6675,
+      "grad_norm": 1.709317245310985,
+      "learning_rate": 1.1958690665369798e-06,
+      "loss": 0.0752,
+      "step": 2937
+    },
+    {
+      "epoch": 0.6677272727272727,
+      "grad_norm": 2.2298597009269328,
+      "learning_rate": 1.1958327285107452e-06,
+      "loss": 0.119,
+      "step": 2938
+    },
+    {
+      "epoch": 0.6679545454545455,
+      "grad_norm": 1.5745088137054357,
+      "learning_rate": 1.1957963788442352e-06,
+      "loss": 0.0517,
+      "step": 2939
+    },
+    {
+      "epoch": 0.6681818181818182,
+      "grad_norm": 1.6590111427368455,
+      "learning_rate": 1.1957600175381916e-06,
+      "loss": 0.1119,
+      "step": 2940
+    },
+    {
+      "epoch": 0.668409090909091,
+      "grad_norm": 1.9782165090232047,
+      "learning_rate": 1.1957236445933553e-06,
+      "loss": 0.0911,
+      "step": 2941
+    },
+    {
+      "epoch": 0.6686363636363636,
+      "grad_norm": 2.106775147788912,
+      "learning_rate": 1.195687260010468e-06,
+      "loss": 0.1458,
+      "step": 2942
+    },
+    {
+      "epoch": 0.6688636363636363,
+      "grad_norm": 2.323074138493056,
+      "learning_rate": 1.1956508637902722e-06,
+      "loss": 0.0878,
+      "step": 2943
+    },
+    {
+      "epoch": 0.6690909090909091,
+      "grad_norm": 1.995163505525096,
+      "learning_rate": 1.1956144559335096e-06,
+      "loss": 0.1267,
+      "step": 2944
+    },
+    {
+      "epoch": 0.6693181818181818,
+      "grad_norm": 1.6965073330326321,
+      "learning_rate": 1.1955780364409226e-06,
+      "loss": 0.131,
+      "step": 2945
+    },
+    {
+      "epoch": 0.6695454545454546,
+      "grad_norm": 2.0501389348432406,
+      "learning_rate": 1.1955416053132542e-06,
+      "loss": 0.0816,
+      "step": 2946
+    },
+    {
+      "epoch": 0.6697727272727273,
+      "grad_norm": 2.308967065930361,
+      "learning_rate": 1.195505162551247e-06,
+      "loss": 0.0777,
+      "step": 2947
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1423362966617665,
+      "learning_rate": 1.1954687081556442e-06,
+      "loss": 0.0617,
+      "step": 2948
+    },
+    {
+      "epoch": 0.6702272727272728,
+      "grad_norm": 2.539068949831609,
+      "learning_rate": 1.1954322421271893e-06,
+      "loss": 0.1469,
+      "step": 2949
+    },
+    {
+      "epoch": 0.6704545454545454,
+      "grad_norm": 2.2052459997246157,
+      "learning_rate": 1.195395764466626e-06,
+      "loss": 0.1341,
+      "step": 2950
+    },
+    {
+      "epoch": 0.6706818181818182,
+      "grad_norm": 1.4623074599411574,
+      "learning_rate": 1.1953592751746976e-06,
+      "loss": 0.0604,
+      "step": 2951
+    },
+    {
+      "epoch": 0.6709090909090909,
+      "grad_norm": 1.9391731818630806,
+      "learning_rate": 1.1953227742521486e-06,
+      "loss": 0.0626,
+      "step": 2952
+    },
+    {
+      "epoch": 0.6711363636363636,
+      "grad_norm": 2.80060336011388,
+      "learning_rate": 1.1952862616997234e-06,
+      "loss": 0.2079,
+      "step": 2953
+    },
+    {
+      "epoch": 0.6713636363636364,
+      "grad_norm": 3.1851333698982214,
+      "learning_rate": 1.1952497375181662e-06,
+      "loss": 0.2119,
+      "step": 2954
+    },
+    {
+      "epoch": 0.6715909090909091,
+      "grad_norm": 1.2850085687293442,
+      "learning_rate": 1.1952132017082222e-06,
+      "loss": 0.0569,
+      "step": 2955
+    },
+    {
+      "epoch": 0.6718181818181819,
+      "grad_norm": 2.0732894631450622,
+      "learning_rate": 1.1951766542706362e-06,
+      "loss": 0.1437,
+      "step": 2956
+    },
+    {
+      "epoch": 0.6720454545454545,
+      "grad_norm": 2.0297557798629087,
+      "learning_rate": 1.1951400952061535e-06,
+      "loss": 0.119,
+      "step": 2957
+    },
+    {
+      "epoch": 0.6722727272727272,
+      "grad_norm": 2.1958433065097926,
+      "learning_rate": 1.1951035245155194e-06,
+      "loss": 0.1161,
+      "step": 2958
+    },
+    {
+      "epoch": 0.6725,
+      "grad_norm": 2.7138863301421416,
+      "learning_rate": 1.19506694219948e-06,
+      "loss": 0.0968,
+      "step": 2959
+    },
+    {
+      "epoch": 0.6727272727272727,
+      "grad_norm": 1.3725670469051048,
+      "learning_rate": 1.195030348258781e-06,
+      "loss": 0.0748,
+      "step": 2960
+    },
+    {
+      "epoch": 0.6729545454545455,
+      "grad_norm": 2.205976625393546,
+      "learning_rate": 1.1949937426941689e-06,
+      "loss": 0.1099,
+      "step": 2961
+    },
+    {
+      "epoch": 0.6731818181818182,
+      "grad_norm": 1.6595949584474334,
+      "learning_rate": 1.1949571255063898e-06,
+      "loss": 0.0584,
+      "step": 2962
+    },
+    {
+      "epoch": 0.673409090909091,
+      "grad_norm": 2.1710829976624315,
+      "learning_rate": 1.1949204966961906e-06,
+      "loss": 0.066,
+      "step": 2963
+    },
+    {
+      "epoch": 0.6736363636363636,
+      "grad_norm": 0.5363980111328321,
+      "learning_rate": 1.1948838562643184e-06,
+      "loss": 0.0348,
+      "step": 2964
+    },
+    {
+      "epoch": 0.6738636363636363,
+      "grad_norm": 2.71499269804727,
+      "learning_rate": 1.19484720421152e-06,
+      "loss": 0.1283,
+      "step": 2965
+    },
+    {
+      "epoch": 0.6740909090909091,
+      "grad_norm": 2.6154385386177585,
+      "learning_rate": 1.1948105405385428e-06,
+      "loss": 0.1109,
+      "step": 2966
+    },
+    {
+      "epoch": 0.6743181818181818,
+      "grad_norm": 1.6556329649300991,
+      "learning_rate": 1.1947738652461346e-06,
+      "loss": 0.1175,
+      "step": 2967
+    },
+    {
+      "epoch": 0.6745454545454546,
+      "grad_norm": 1.9058120179208964,
+      "learning_rate": 1.1947371783350436e-06,
+      "loss": 0.0704,
+      "step": 2968
+    },
+    {
+      "epoch": 0.6747727272727273,
+      "grad_norm": 2.4704817966154238,
+      "learning_rate": 1.1947004798060174e-06,
+      "loss": 0.0891,
+      "step": 2969
+    },
+    {
+      "epoch": 0.675,
+      "grad_norm": 1.645638518871364,
+      "learning_rate": 1.1946637696598043e-06,
+      "loss": 0.1229,
+      "step": 2970
+    },
+    {
+      "epoch": 0.6752272727272727,
+      "grad_norm": 1.982703695287054,
+      "learning_rate": 1.1946270478971533e-06,
+      "loss": 0.1161,
+      "step": 2971
+    },
+    {
+      "epoch": 0.6754545454545454,
+      "grad_norm": 1.434473158843332,
+      "learning_rate": 1.1945903145188132e-06,
+      "loss": 0.0963,
+      "step": 2972
+    },
+    {
+      "epoch": 0.6756818181818182,
+      "grad_norm": 2.276748581195608,
+      "learning_rate": 1.1945535695255325e-06,
+      "loss": 0.0731,
+      "step": 2973
+    },
+    {
+      "epoch": 0.6759090909090909,
+      "grad_norm": 2.266424125144421,
+      "learning_rate": 1.1945168129180613e-06,
+      "loss": 0.108,
+      "step": 2974
+    },
+    {
+      "epoch": 0.6761363636363636,
+      "grad_norm": 1.1291420619614252,
+      "learning_rate": 1.1944800446971483e-06,
+      "loss": 0.0525,
+      "step": 2975
+    },
+    {
+      "epoch": 0.6763636363636364,
+      "grad_norm": 3.7115940065030246,
+      "learning_rate": 1.194443264863544e-06,
+      "loss": 0.1855,
+      "step": 2976
+    },
+    {
+      "epoch": 0.6765909090909091,
+      "grad_norm": 1.1134638607594658,
+      "learning_rate": 1.1944064734179977e-06,
+      "loss": 0.0476,
+      "step": 2977
+    },
+    {
+      "epoch": 0.6768181818181818,
+      "grad_norm": 1.4203901896146813,
+      "learning_rate": 1.1943696703612604e-06,
+      "loss": 0.052,
+      "step": 2978
+    },
+    {
+      "epoch": 0.6770454545454545,
+      "grad_norm": 2.465302498547221,
+      "learning_rate": 1.194332855694082e-06,
+      "loss": 0.1743,
+      "step": 2979
+    },
+    {
+      "epoch": 0.6772727272727272,
+      "grad_norm": 2.934077700393342,
+      "learning_rate": 1.1942960294172135e-06,
+      "loss": 0.1982,
+      "step": 2980
+    },
+    {
+      "epoch": 0.6775,
+      "grad_norm": 2.0120279022356096,
+      "learning_rate": 1.1942591915314058e-06,
+      "loss": 0.0651,
+      "step": 2981
+    },
+    {
+      "epoch": 0.6777272727272727,
+      "grad_norm": 1.1989022058583192,
+      "learning_rate": 1.19422234203741e-06,
+      "loss": 0.0647,
+      "step": 2982
+    },
+    {
+      "epoch": 0.6779545454545455,
+      "grad_norm": 1.2610226449089499,
+      "learning_rate": 1.1941854809359775e-06,
+      "loss": 0.0413,
+      "step": 2983
+    },
+    {
+      "epoch": 0.6781818181818182,
+      "grad_norm": 1.8182479291629559,
+      "learning_rate": 1.1941486082278602e-06,
+      "loss": 0.0738,
+      "step": 2984
+    },
+    {
+      "epoch": 0.678409090909091,
+      "grad_norm": 2.320972545665281,
+      "learning_rate": 1.1941117239138098e-06,
+      "loss": 0.134,
+      "step": 2985
+    },
+    {
+      "epoch": 0.6786363636363636,
+      "grad_norm": 2.1320124526042386,
+      "learning_rate": 1.1940748279945784e-06,
+      "loss": 0.0999,
+      "step": 2986
+    },
+    {
+      "epoch": 0.6788636363636363,
+      "grad_norm": 2.1155750241959512,
+      "learning_rate": 1.1940379204709186e-06,
+      "loss": 0.0729,
+      "step": 2987
+    },
+    {
+      "epoch": 0.6790909090909091,
+      "grad_norm": 2.512289437722847,
+      "learning_rate": 1.1940010013435829e-06,
+      "loss": 0.0989,
+      "step": 2988
+    },
+    {
+      "epoch": 0.6793181818181818,
+      "grad_norm": 1.8604426745130982,
+      "learning_rate": 1.1939640706133237e-06,
+      "loss": 0.0784,
+      "step": 2989
+    },
+    {
+      "epoch": 0.6795454545454546,
+      "grad_norm": 0.75360891249419,
+      "learning_rate": 1.193927128280895e-06,
+      "loss": 0.0349,
+      "step": 2990
+    },
+    {
+      "epoch": 0.6797727272727273,
+      "grad_norm": 0.73855068160215,
+      "learning_rate": 1.1938901743470494e-06,
+      "loss": 0.0312,
+      "step": 2991
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4566567103112065,
+      "learning_rate": 1.1938532088125405e-06,
+      "loss": 0.0824,
+      "step": 2992
+    },
+    {
+      "epoch": 0.6802272727272727,
+      "grad_norm": 3.5963037263263833,
+      "learning_rate": 1.1938162316781223e-06,
+      "loss": 0.0902,
+      "step": 2993
+    },
+    {
+      "epoch": 0.6804545454545454,
+      "grad_norm": 1.3813558817173497,
+      "learning_rate": 1.1937792429445489e-06,
+      "loss": 0.0435,
+      "step": 2994
+    },
+    {
+      "epoch": 0.6806818181818182,
+      "grad_norm": 1.9212648313362939,
+      "learning_rate": 1.1937422426125745e-06,
+      "loss": 0.0586,
+      "step": 2995
+    },
+    {
+      "epoch": 0.6809090909090909,
+      "grad_norm": 1.884837512305607,
+      "learning_rate": 1.1937052306829534e-06,
+      "loss": 0.079,
+      "step": 2996
+    },
+    {
+      "epoch": 0.6811363636363637,
+      "grad_norm": 2.097766321356603,
+      "learning_rate": 1.1936682071564407e-06,
+      "loss": 0.071,
+      "step": 2997
+    },
+    {
+      "epoch": 0.6813636363636364,
+      "grad_norm": 1.5069673653706013,
+      "learning_rate": 1.193631172033791e-06,
+      "loss": 0.0749,
+      "step": 2998
+    },
+    {
+      "epoch": 0.6815909090909091,
+      "grad_norm": 1.317010807342748,
+      "learning_rate": 1.1935941253157596e-06,
+      "loss": 0.0684,
+      "step": 2999
+    },
+    {
+      "epoch": 0.6818181818181818,
+      "grad_norm": 1.2911145495313106,
+      "learning_rate": 1.1935570670031022e-06,
+      "loss": 0.0557,
+      "step": 3000
+    },
+    {
+      "epoch": 0.6820454545454545,
+      "grad_norm": 2.1064924829256393,
+      "learning_rate": 1.1935199970965741e-06,
+      "loss": 0.1085,
+      "step": 3001
+    },
+    {
+      "epoch": 0.6822727272727273,
+      "grad_norm": 1.311011695102738,
+      "learning_rate": 1.1934829155969315e-06,
+      "loss": 0.0574,
+      "step": 3002
+    },
+    {
+      "epoch": 0.6825,
+      "grad_norm": 2.1193766179577085,
+      "learning_rate": 1.1934458225049306e-06,
+      "loss": 0.1211,
+      "step": 3003
+    },
+    {
+      "epoch": 0.6827272727272727,
+      "grad_norm": 1.2931060461491533,
+      "learning_rate": 1.1934087178213275e-06,
+      "loss": 0.0564,
+      "step": 3004
+    },
+    {
+      "epoch": 0.6829545454545455,
+      "grad_norm": 2.249954335926216,
+      "learning_rate": 1.193371601546879e-06,
+      "loss": 0.0904,
+      "step": 3005
+    },
+    {
+      "epoch": 0.6831818181818182,
+      "grad_norm": 2.0471415720374293,
+      "learning_rate": 1.193334473682342e-06,
+      "loss": 0.1031,
+      "step": 3006
+    },
+    {
+      "epoch": 0.6834090909090909,
+      "grad_norm": 1.9975830914192465,
+      "learning_rate": 1.1932973342284737e-06,
+      "loss": 0.0967,
+      "step": 3007
+    },
+    {
+      "epoch": 0.6836363636363636,
+      "grad_norm": 1.9221816396111169,
+      "learning_rate": 1.1932601831860313e-06,
+      "loss": 0.1159,
+      "step": 3008
+    },
+    {
+      "epoch": 0.6838636363636363,
+      "grad_norm": 1.7311311765078239,
+      "learning_rate": 1.1932230205557721e-06,
+      "loss": 0.1083,
+      "step": 3009
+    },
+    {
+      "epoch": 0.6840909090909091,
+      "grad_norm": 0.5631760870556558,
+      "learning_rate": 1.1931858463384544e-06,
+      "loss": 0.0159,
+      "step": 3010
+    },
+    {
+      "epoch": 0.6843181818181818,
+      "grad_norm": 0.9838468359781708,
+      "learning_rate": 1.193148660534836e-06,
+      "loss": 0.0418,
+      "step": 3011
+    },
+    {
+      "epoch": 0.6845454545454546,
+      "grad_norm": 1.7603987231253975,
+      "learning_rate": 1.1931114631456753e-06,
+      "loss": 0.07,
+      "step": 3012
+    },
+    {
+      "epoch": 0.6847727272727273,
+      "grad_norm": 1.8999791195340185,
+      "learning_rate": 1.1930742541717307e-06,
+      "loss": 0.0861,
+      "step": 3013
+    },
+    {
+      "epoch": 0.685,
+      "grad_norm": 1.2088327298237413,
+      "learning_rate": 1.193037033613761e-06,
+      "loss": 0.0461,
+      "step": 3014
+    },
+    {
+      "epoch": 0.6852272727272727,
+      "grad_norm": 1.6946915198421277,
+      "learning_rate": 1.1929998014725252e-06,
+      "loss": 0.0681,
+      "step": 3015
+    },
+    {
+      "epoch": 0.6854545454545454,
+      "grad_norm": 2.3048856648862177,
+      "learning_rate": 1.1929625577487825e-06,
+      "loss": 0.1277,
+      "step": 3016
+    },
+    {
+      "epoch": 0.6856818181818182,
+      "grad_norm": 1.8818158920301469,
+      "learning_rate": 1.1929253024432925e-06,
+      "loss": 0.111,
+      "step": 3017
+    },
+    {
+      "epoch": 0.6859090909090909,
+      "grad_norm": 1.6214756090662403,
+      "learning_rate": 1.1928880355568146e-06,
+      "loss": 0.0576,
+      "step": 3018
+    },
+    {
+      "epoch": 0.6861363636363637,
+      "grad_norm": 2.238972484084478,
+      "learning_rate": 1.1928507570901089e-06,
+      "loss": 0.1366,
+      "step": 3019
+    },
+    {
+      "epoch": 0.6863636363636364,
+      "grad_norm": 1.890351173018935,
+      "learning_rate": 1.1928134670439357e-06,
+      "loss": 0.1342,
+      "step": 3020
+    },
+    {
+      "epoch": 0.6865909090909091,
+      "grad_norm": 3.21499886167954,
+      "learning_rate": 1.1927761654190552e-06,
+      "loss": 0.1049,
+      "step": 3021
+    },
+    {
+      "epoch": 0.6868181818181818,
+      "grad_norm": 3.9267729044225024,
+      "learning_rate": 1.1927388522162282e-06,
+      "loss": 0.1055,
+      "step": 3022
+    },
+    {
+      "epoch": 0.6870454545454545,
+      "grad_norm": 2.2526762414041612,
+      "learning_rate": 1.1927015274362156e-06,
+      "loss": 0.0893,
+      "step": 3023
+    },
+    {
+      "epoch": 0.6872727272727273,
+      "grad_norm": 1.8326126505580012,
+      "learning_rate": 1.1926641910797785e-06,
+      "loss": 0.0706,
+      "step": 3024
+    },
+    {
+      "epoch": 0.6875,
+      "grad_norm": 1.8255334874688318,
+      "learning_rate": 1.1926268431476777e-06,
+      "loss": 0.0747,
+      "step": 3025
+    },
+    {
+      "epoch": 0.6877272727272727,
+      "grad_norm": 2.358498424294261,
+      "learning_rate": 1.192589483640676e-06,
+      "loss": 0.0603,
+      "step": 3026
+    },
+    {
+      "epoch": 0.6879545454545455,
+      "grad_norm": 2.9988160089673124,
+      "learning_rate": 1.192552112559534e-06,
+      "loss": 0.2,
+      "step": 3027
+    },
+    {
+      "epoch": 0.6881818181818182,
+      "grad_norm": 2.0208266645529767,
+      "learning_rate": 1.1925147299050147e-06,
+      "loss": 0.0716,
+      "step": 3028
+    },
+    {
+      "epoch": 0.6884090909090909,
+      "grad_norm": 1.5983505270766545,
+      "learning_rate": 1.1924773356778797e-06,
+      "loss": 0.0723,
+      "step": 3029
+    },
+    {
+      "epoch": 0.6886363636363636,
+      "grad_norm": 0.8971175289344515,
+      "learning_rate": 1.1924399298788917e-06,
+      "loss": 0.0234,
+      "step": 3030
+    },
+    {
+      "epoch": 0.6888636363636363,
+      "grad_norm": 3.352833592715963,
+      "learning_rate": 1.1924025125088138e-06,
+      "loss": 0.2062,
+      "step": 3031
+    },
+    {
+      "epoch": 0.6890909090909091,
+      "grad_norm": 1.803217920660286,
+      "learning_rate": 1.1923650835684085e-06,
+      "loss": 0.0897,
+      "step": 3032
+    },
+    {
+      "epoch": 0.6893181818181818,
+      "grad_norm": 1.930289386769726,
+      "learning_rate": 1.1923276430584398e-06,
+      "loss": 0.0517,
+      "step": 3033
+    },
+    {
+      "epoch": 0.6895454545454546,
+      "grad_norm": 1.6676748458640007,
+      "learning_rate": 1.1922901909796705e-06,
+      "loss": 0.0984,
+      "step": 3034
+    },
+    {
+      "epoch": 0.6897727272727273,
+      "grad_norm": 1.2362281541013442,
+      "learning_rate": 1.1922527273328644e-06,
+      "loss": 0.0477,
+      "step": 3035
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5251772668475776,
+      "learning_rate": 1.1922152521187854e-06,
+      "loss": 0.0579,
+      "step": 3036
+    },
+    {
+      "epoch": 0.6902272727272727,
+      "grad_norm": 3.399835305702526,
+      "learning_rate": 1.1921777653381981e-06,
+      "loss": 0.1011,
+      "step": 3037
+    },
+    {
+      "epoch": 0.6904545454545454,
+      "grad_norm": 3.15057044695853,
+      "learning_rate": 1.1921402669918666e-06,
+      "loss": 0.0904,
+      "step": 3038
+    },
+    {
+      "epoch": 0.6906818181818182,
+      "grad_norm": 1.6199699590228094,
+      "learning_rate": 1.1921027570805557e-06,
+      "loss": 0.0871,
+      "step": 3039
+    },
+    {
+      "epoch": 0.6909090909090909,
+      "grad_norm": 1.4839557623145854,
+      "learning_rate": 1.1920652356050303e-06,
+      "loss": 0.0649,
+      "step": 3040
+    },
+    {
+      "epoch": 0.6911363636363637,
+      "grad_norm": 2.489378895550056,
+      "learning_rate": 1.1920277025660553e-06,
+      "loss": 0.118,
+      "step": 3041
+    },
+    {
+      "epoch": 0.6913636363636364,
+      "grad_norm": 1.773232278358991,
+      "learning_rate": 1.191990157964396e-06,
+      "loss": 0.0763,
+      "step": 3042
+    },
+    {
+      "epoch": 0.6915909090909091,
+      "grad_norm": 1.9078040734239965,
+      "learning_rate": 1.1919526018008185e-06,
+      "loss": 0.0737,
+      "step": 3043
+    },
+    {
+      "epoch": 0.6918181818181818,
+      "grad_norm": 1.8742402874469573,
+      "learning_rate": 1.1919150340760884e-06,
+      "loss": 0.0982,
+      "step": 3044
+    },
+    {
+      "epoch": 0.6920454545454545,
+      "grad_norm": 1.467595920804719,
+      "learning_rate": 1.1918774547909716e-06,
+      "loss": 0.0873,
+      "step": 3045
+    },
+    {
+      "epoch": 0.6922727272727273,
+      "grad_norm": 2.0597149319327372,
+      "learning_rate": 1.1918398639462345e-06,
+      "loss": 0.1074,
+      "step": 3046
+    },
+    {
+      "epoch": 0.6925,
+      "grad_norm": 1.992550223959394,
+      "learning_rate": 1.1918022615426438e-06,
+      "loss": 0.0954,
+      "step": 3047
+    },
+    {
+      "epoch": 0.6927272727272727,
+      "grad_norm": 1.6695673054070386,
+      "learning_rate": 1.191764647580966e-06,
+      "loss": 0.0581,
+      "step": 3048
+    },
+    {
+      "epoch": 0.6929545454545455,
+      "grad_norm": 2.3011408764946366,
+      "learning_rate": 1.1917270220619682e-06,
+      "loss": 0.1592,
+      "step": 3049
+    },
+    {
+      "epoch": 0.6931818181818182,
+      "grad_norm": 2.6449243924525416,
+      "learning_rate": 1.1916893849864177e-06,
+      "loss": 0.1476,
+      "step": 3050
+    },
+    {
+      "epoch": 0.6934090909090909,
+      "grad_norm": 1.475653079479369,
+      "learning_rate": 1.1916517363550821e-06,
+      "loss": 0.0662,
+      "step": 3051
+    },
+    {
+      "epoch": 0.6936363636363636,
+      "grad_norm": 1.385434997468204,
+      "learning_rate": 1.1916140761687292e-06,
+      "loss": 0.0403,
+      "step": 3052
+    },
+    {
+      "epoch": 0.6938636363636363,
+      "grad_norm": 1.768879144871322,
+      "learning_rate": 1.1915764044281266e-06,
+      "loss": 0.0993,
+      "step": 3053
+    },
+    {
+      "epoch": 0.6940909090909091,
+      "grad_norm": 2.2140514309903847,
+      "learning_rate": 1.1915387211340427e-06,
+      "loss": 0.0798,
+      "step": 3054
+    },
+    {
+      "epoch": 0.6943181818181818,
+      "grad_norm": 1.3978027992232365,
+      "learning_rate": 1.1915010262872459e-06,
+      "loss": 0.1019,
+      "step": 3055
+    },
+    {
+      "epoch": 0.6945454545454546,
+      "grad_norm": 2.1235867359364833,
+      "learning_rate": 1.1914633198885047e-06,
+      "loss": 0.0543,
+      "step": 3056
+    },
+    {
+      "epoch": 0.6947727272727273,
+      "grad_norm": 1.4618118965357383,
+      "learning_rate": 1.1914256019385883e-06,
+      "loss": 0.075,
+      "step": 3057
+    },
+    {
+      "epoch": 0.695,
+      "grad_norm": 2.1103331698999703,
+      "learning_rate": 1.1913878724382658e-06,
+      "loss": 0.067,
+      "step": 3058
+    },
+    {
+      "epoch": 0.6952272727272727,
+      "grad_norm": 2.4666884007098417,
+      "learning_rate": 1.1913501313883061e-06,
+      "loss": 0.0926,
+      "step": 3059
+    },
+    {
+      "epoch": 0.6954545454545454,
+      "grad_norm": 1.7550996344916832,
+      "learning_rate": 1.1913123787894795e-06,
+      "loss": 0.1014,
+      "step": 3060
+    },
+    {
+      "epoch": 0.6956818181818182,
+      "grad_norm": 1.7727575440975376,
+      "learning_rate": 1.1912746146425555e-06,
+      "loss": 0.1062,
+      "step": 3061
+    },
+    {
+      "epoch": 0.6959090909090909,
+      "grad_norm": 2.1707071888252805,
+      "learning_rate": 1.1912368389483041e-06,
+      "loss": 0.0918,
+      "step": 3062
+    },
+    {
+      "epoch": 0.6961363636363637,
+      "grad_norm": 1.1671559095504347,
+      "learning_rate": 1.1911990517074957e-06,
+      "loss": 0.0681,
+      "step": 3063
+    },
+    {
+      "epoch": 0.6963636363636364,
+      "grad_norm": 2.1622259969049415,
+      "learning_rate": 1.1911612529209008e-06,
+      "loss": 0.1241,
+      "step": 3064
+    },
+    {
+      "epoch": 0.696590909090909,
+      "grad_norm": 3.138451810170633,
+      "learning_rate": 1.1911234425892904e-06,
+      "loss": 0.1996,
+      "step": 3065
+    },
+    {
+      "epoch": 0.6968181818181818,
+      "grad_norm": 1.8833450243345458,
+      "learning_rate": 1.1910856207134352e-06,
+      "loss": 0.1118,
+      "step": 3066
+    },
+    {
+      "epoch": 0.6970454545454545,
+      "grad_norm": 1.140702085325538,
+      "learning_rate": 1.1910477872941065e-06,
+      "loss": 0.0593,
+      "step": 3067
+    },
+    {
+      "epoch": 0.6972727272727273,
+      "grad_norm": 1.1567595680746212,
+      "learning_rate": 1.1910099423320763e-06,
+      "loss": 0.0518,
+      "step": 3068
+    },
+    {
+      "epoch": 0.6975,
+      "grad_norm": 1.307510162998218,
+      "learning_rate": 1.1909720858281155e-06,
+      "loss": 0.0407,
+      "step": 3069
+    },
+    {
+      "epoch": 0.6977272727272728,
+      "grad_norm": 1.9931274825921341,
+      "learning_rate": 1.1909342177829967e-06,
+      "loss": 0.0807,
+      "step": 3070
+    },
+    {
+      "epoch": 0.6979545454545455,
+      "grad_norm": 1.6173269099107999,
+      "learning_rate": 1.1908963381974916e-06,
+      "loss": 0.0805,
+      "step": 3071
+    },
+    {
+      "epoch": 0.6981818181818182,
+      "grad_norm": 1.745374775636812,
+      "learning_rate": 1.190858447072373e-06,
+      "loss": 0.0403,
+      "step": 3072
+    },
+    {
+      "epoch": 0.6984090909090909,
+      "grad_norm": 2.4938443254493565,
+      "learning_rate": 1.1908205444084138e-06,
+      "loss": 0.1133,
+      "step": 3073
+    },
+    {
+      "epoch": 0.6986363636363636,
+      "grad_norm": 1.9187883583062475,
+      "learning_rate": 1.1907826302063863e-06,
+      "loss": 0.0904,
+      "step": 3074
+    },
+    {
+      "epoch": 0.6988636363636364,
+      "grad_norm": 2.155179232853511,
+      "learning_rate": 1.190744704467064e-06,
+      "loss": 0.0902,
+      "step": 3075
+    },
+    {
+      "epoch": 0.6990909090909091,
+      "grad_norm": 1.250638449578486,
+      "learning_rate": 1.19070676719122e-06,
+      "loss": 0.0737,
+      "step": 3076
+    },
+    {
+      "epoch": 0.6993181818181818,
+      "grad_norm": 1.933560007410491,
+      "learning_rate": 1.1906688183796282e-06,
+      "loss": 0.1049,
+      "step": 3077
+    },
+    {
+      "epoch": 0.6995454545454546,
+      "grad_norm": 1.2730288231078695,
+      "learning_rate": 1.1906308580330625e-06,
+      "loss": 0.0607,
+      "step": 3078
+    },
+    {
+      "epoch": 0.6997727272727273,
+      "grad_norm": 1.8858318109550323,
+      "learning_rate": 1.1905928861522964e-06,
+      "loss": 0.101,
+      "step": 3079
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.418767742026248,
+      "learning_rate": 1.190554902738105e-06,
+      "loss": 0.0438,
+      "step": 3080
+    },
+    {
+      "epoch": 0.7002272727272727,
+      "grad_norm": 1.9908610339182096,
+      "learning_rate": 1.1905169077912623e-06,
+      "loss": 0.0841,
+      "step": 3081
+    },
+    {
+      "epoch": 0.7004545454545454,
+      "grad_norm": 1.8417952031408806,
+      "learning_rate": 1.1904789013125432e-06,
+      "loss": 0.067,
+      "step": 3082
+    },
+    {
+      "epoch": 0.7006818181818182,
+      "grad_norm": 2.0357087313607365,
+      "learning_rate": 1.190440883302723e-06,
+      "loss": 0.1195,
+      "step": 3083
+    },
+    {
+      "epoch": 0.7009090909090909,
+      "grad_norm": 1.8936857554016093,
+      "learning_rate": 1.1904028537625765e-06,
+      "loss": 0.0477,
+      "step": 3084
+    },
+    {
+      "epoch": 0.7011363636363637,
+      "grad_norm": 1.9611984204233002,
+      "learning_rate": 1.1903648126928796e-06,
+      "loss": 0.0892,
+      "step": 3085
+    },
+    {
+      "epoch": 0.7013636363636364,
+      "grad_norm": 1.5868121487972757,
+      "learning_rate": 1.1903267600944077e-06,
+      "loss": 0.0646,
+      "step": 3086
+    },
+    {
+      "epoch": 0.701590909090909,
+      "grad_norm": 1.3380520384841221,
+      "learning_rate": 1.190288695967937e-06,
+      "loss": 0.0637,
+      "step": 3087
+    },
+    {
+      "epoch": 0.7018181818181818,
+      "grad_norm": 1.3132438325342677,
+      "learning_rate": 1.1902506203142434e-06,
+      "loss": 0.0404,
+      "step": 3088
+    },
+    {
+      "epoch": 0.7020454545454545,
+      "grad_norm": 2.335660621690764,
+      "learning_rate": 1.1902125331341036e-06,
+      "loss": 0.1458,
+      "step": 3089
+    },
+    {
+      "epoch": 0.7022727272727273,
+      "grad_norm": 1.423425186336246,
+      "learning_rate": 1.1901744344282942e-06,
+      "loss": 0.0997,
+      "step": 3090
+    },
+    {
+      "epoch": 0.7025,
+      "grad_norm": 1.782565541664032,
+      "learning_rate": 1.1901363241975921e-06,
+      "loss": 0.0605,
+      "step": 3091
+    },
+    {
+      "epoch": 0.7027272727272728,
+      "grad_norm": 2.1830949939273228,
+      "learning_rate": 1.1900982024427745e-06,
+      "loss": 0.102,
+      "step": 3092
+    },
+    {
+      "epoch": 0.7029545454545455,
+      "grad_norm": 1.6752858590070172,
+      "learning_rate": 1.1900600691646187e-06,
+      "loss": 0.1188,
+      "step": 3093
+    },
+    {
+      "epoch": 0.7031818181818181,
+      "grad_norm": 1.5691859691410739,
+      "learning_rate": 1.190021924363902e-06,
+      "loss": 0.052,
+      "step": 3094
+    },
+    {
+      "epoch": 0.7034090909090909,
+      "grad_norm": 2.362297723123779,
+      "learning_rate": 1.1899837680414028e-06,
+      "loss": 0.1356,
+      "step": 3095
+    },
+    {
+      "epoch": 0.7036363636363636,
+      "grad_norm": 1.5386712947575067,
+      "learning_rate": 1.1899456001978987e-06,
+      "loss": 0.0633,
+      "step": 3096
+    },
+    {
+      "epoch": 0.7038636363636364,
+      "grad_norm": 2.2405559444129803,
+      "learning_rate": 1.1899074208341685e-06,
+      "loss": 0.1225,
+      "step": 3097
+    },
+    {
+      "epoch": 0.7040909090909091,
+      "grad_norm": 2.8567154513375086,
+      "learning_rate": 1.1898692299509903e-06,
+      "loss": 0.1255,
+      "step": 3098
+    },
+    {
+      "epoch": 0.7043181818181818,
+      "grad_norm": 1.8530166415114095,
+      "learning_rate": 1.189831027549143e-06,
+      "loss": 0.068,
+      "step": 3099
+    },
+    {
+      "epoch": 0.7045454545454546,
+      "grad_norm": 2.7779654501447566,
+      "learning_rate": 1.1897928136294055e-06,
+      "loss": 0.1163,
+      "step": 3100
+    },
+    {
+      "epoch": 0.7047727272727272,
+      "grad_norm": 1.698128468533611,
+      "learning_rate": 1.1897545881925573e-06,
+      "loss": 0.0569,
+      "step": 3101
+    },
+    {
+      "epoch": 0.705,
+      "grad_norm": 1.5086950193392006,
+      "learning_rate": 1.189716351239378e-06,
+      "loss": 0.0449,
+      "step": 3102
+    },
+    {
+      "epoch": 0.7052272727272727,
+      "grad_norm": 1.2506386756660945,
+      "learning_rate": 1.1896781027706469e-06,
+      "loss": 0.0634,
+      "step": 3103
+    },
+    {
+      "epoch": 0.7054545454545454,
+      "grad_norm": 1.570987467500789,
+      "learning_rate": 1.1896398427871442e-06,
+      "loss": 0.0543,
+      "step": 3104
+    },
+    {
+      "epoch": 0.7056818181818182,
+      "grad_norm": 1.13444366028076,
+      "learning_rate": 1.18960157128965e-06,
+      "loss": 0.07,
+      "step": 3105
+    },
+    {
+      "epoch": 0.7059090909090909,
+      "grad_norm": 2.492136130696584,
+      "learning_rate": 1.1895632882789447e-06,
+      "loss": 0.139,
+      "step": 3106
+    },
+    {
+      "epoch": 0.7061363636363637,
+      "grad_norm": 1.8195294415365988,
+      "learning_rate": 1.1895249937558091e-06,
+      "loss": 0.0678,
+      "step": 3107
+    },
+    {
+      "epoch": 0.7063636363636364,
+      "grad_norm": 1.800764395706198,
+      "learning_rate": 1.189486687721024e-06,
+      "loss": 0.0389,
+      "step": 3108
+    },
+    {
+      "epoch": 0.706590909090909,
+      "grad_norm": 2.0532856931160453,
+      "learning_rate": 1.1894483701753707e-06,
+      "loss": 0.121,
+      "step": 3109
+    },
+    {
+      "epoch": 0.7068181818181818,
+      "grad_norm": 1.6523093119032983,
+      "learning_rate": 1.1894100411196302e-06,
+      "loss": 0.0592,
+      "step": 3110
+    },
+    {
+      "epoch": 0.7070454545454545,
+      "grad_norm": 4.007257255438708,
+      "learning_rate": 1.1893717005545843e-06,
+      "loss": 0.1588,
+      "step": 3111
+    },
+    {
+      "epoch": 0.7072727272727273,
+      "grad_norm": 2.3952345664247643,
+      "learning_rate": 1.1893333484810148e-06,
+      "loss": 0.0925,
+      "step": 3112
+    },
+    {
+      "epoch": 0.7075,
+      "grad_norm": 1.6198372804502819,
+      "learning_rate": 1.1892949848997038e-06,
+      "loss": 0.0759,
+      "step": 3113
+    },
+    {
+      "epoch": 0.7077272727272728,
+      "grad_norm": 1.0694168502121828,
+      "learning_rate": 1.1892566098114337e-06,
+      "loss": 0.0555,
+      "step": 3114
+    },
+    {
+      "epoch": 0.7079545454545455,
+      "grad_norm": 2.60051757829192,
+      "learning_rate": 1.1892182232169869e-06,
+      "loss": 0.118,
+      "step": 3115
+    },
+    {
+      "epoch": 0.7081818181818181,
+      "grad_norm": 1.7061433547073823,
+      "learning_rate": 1.189179825117146e-06,
+      "loss": 0.0657,
+      "step": 3116
+    },
+    {
+      "epoch": 0.7084090909090909,
+      "grad_norm": 1.7695071036158467,
+      "learning_rate": 1.1891414155126944e-06,
+      "loss": 0.0639,
+      "step": 3117
+    },
+    {
+      "epoch": 0.7086363636363636,
+      "grad_norm": 1.8019425861155092,
+      "learning_rate": 1.1891029944044153e-06,
+      "loss": 0.0961,
+      "step": 3118
+    },
+    {
+      "epoch": 0.7088636363636364,
+      "grad_norm": 2.152712056609558,
+      "learning_rate": 1.1890645617930916e-06,
+      "loss": 0.0991,
+      "step": 3119
+    },
+    {
+      "epoch": 0.7090909090909091,
+      "grad_norm": 1.5028572929589992,
+      "learning_rate": 1.1890261176795078e-06,
+      "loss": 0.0853,
+      "step": 3120
+    },
+    {
+      "epoch": 0.7093181818181818,
+      "grad_norm": 1.8431437970613613,
+      "learning_rate": 1.1889876620644472e-06,
+      "loss": 0.0512,
+      "step": 3121
+    },
+    {
+      "epoch": 0.7095454545454546,
+      "grad_norm": 2.1799668951870137,
+      "learning_rate": 1.1889491949486944e-06,
+      "loss": 0.1197,
+      "step": 3122
+    },
+    {
+      "epoch": 0.7097727272727272,
+      "grad_norm": 2.0167070986183884,
+      "learning_rate": 1.1889107163330335e-06,
+      "loss": 0.0682,
+      "step": 3123
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6895003304711165,
+      "learning_rate": 1.1888722262182495e-06,
+      "loss": 0.0842,
+      "step": 3124
+    },
+    {
+      "epoch": 0.7102272727272727,
+      "grad_norm": 2.11517832132824,
+      "learning_rate": 1.1888337246051268e-06,
+      "loss": 0.0558,
+      "step": 3125
+    },
+    {
+      "epoch": 0.7104545454545454,
+      "grad_norm": 1.743553228219526,
+      "learning_rate": 1.1887952114944509e-06,
+      "loss": 0.0865,
+      "step": 3126
+    },
+    {
+      "epoch": 0.7106818181818182,
+      "grad_norm": 2.865426207768221,
+      "learning_rate": 1.1887566868870071e-06,
+      "loss": 0.1259,
+      "step": 3127
+    },
+    {
+      "epoch": 0.7109090909090909,
+      "grad_norm": 1.457252716372001,
+      "learning_rate": 1.188718150783581e-06,
+      "loss": 0.0924,
+      "step": 3128
+    },
+    {
+      "epoch": 0.7111363636363637,
+      "grad_norm": 0.9672955605105978,
+      "learning_rate": 1.1886796031849583e-06,
+      "loss": 0.0466,
+      "step": 3129
+    },
+    {
+      "epoch": 0.7113636363636363,
+      "grad_norm": 1.746359789168319,
+      "learning_rate": 1.188641044091925e-06,
+      "loss": 0.0901,
+      "step": 3130
+    },
+    {
+      "epoch": 0.711590909090909,
+      "grad_norm": 2.073490483756267,
+      "learning_rate": 1.1886024735052676e-06,
+      "loss": 0.0676,
+      "step": 3131
+    },
+    {
+      "epoch": 0.7118181818181818,
+      "grad_norm": 2.19664342914708,
+      "learning_rate": 1.1885638914257725e-06,
+      "loss": 0.0892,
+      "step": 3132
+    },
+    {
+      "epoch": 0.7120454545454545,
+      "grad_norm": 1.6238169803661224,
+      "learning_rate": 1.1885252978542265e-06,
+      "loss": 0.0932,
+      "step": 3133
+    },
+    {
+      "epoch": 0.7122727272727273,
+      "grad_norm": 1.349881117360508,
+      "learning_rate": 1.1884866927914165e-06,
+      "loss": 0.086,
+      "step": 3134
+    },
+    {
+      "epoch": 0.7125,
+      "grad_norm": 1.2375809922658874,
+      "learning_rate": 1.1884480762381296e-06,
+      "loss": 0.038,
+      "step": 3135
+    },
+    {
+      "epoch": 0.7127272727272728,
+      "grad_norm": 1.5549359821313384,
+      "learning_rate": 1.1884094481951535e-06,
+      "loss": 0.0928,
+      "step": 3136
+    },
+    {
+      "epoch": 0.7129545454545455,
+      "grad_norm": 1.8459272474836266,
+      "learning_rate": 1.188370808663276e-06,
+      "loss": 0.0561,
+      "step": 3137
+    },
+    {
+      "epoch": 0.7131818181818181,
+      "grad_norm": 1.621071440588219,
+      "learning_rate": 1.1883321576432849e-06,
+      "loss": 0.0532,
+      "step": 3138
+    },
+    {
+      "epoch": 0.7134090909090909,
+      "grad_norm": 2.011403902289039,
+      "learning_rate": 1.1882934951359682e-06,
+      "loss": 0.0754,
+      "step": 3139
+    },
+    {
+      "epoch": 0.7136363636363636,
+      "grad_norm": 1.3237407500876721,
+      "learning_rate": 1.1882548211421144e-06,
+      "loss": 0.0791,
+      "step": 3140
+    },
+    {
+      "epoch": 0.7138636363636364,
+      "grad_norm": 1.5982329543989025,
+      "learning_rate": 1.1882161356625122e-06,
+      "loss": 0.0912,
+      "step": 3141
+    },
+    {
+      "epoch": 0.7140909090909091,
+      "grad_norm": 2.595289644745982,
+      "learning_rate": 1.1881774386979503e-06,
+      "loss": 0.0921,
+      "step": 3142
+    },
+    {
+      "epoch": 0.7143181818181819,
+      "grad_norm": 2.2985600349311164,
+      "learning_rate": 1.188138730249218e-06,
+      "loss": 0.1309,
+      "step": 3143
+    },
+    {
+      "epoch": 0.7145454545454546,
+      "grad_norm": 2.3185116829095267,
+      "learning_rate": 1.1881000103171045e-06,
+      "loss": 0.097,
+      "step": 3144
+    },
+    {
+      "epoch": 0.7147727272727272,
+      "grad_norm": 3.4562396787067344,
+      "learning_rate": 1.1880612789023993e-06,
+      "loss": 0.1508,
+      "step": 3145
+    },
+    {
+      "epoch": 0.715,
+      "grad_norm": 2.1649105893732026,
+      "learning_rate": 1.1880225360058925e-06,
+      "loss": 0.0556,
+      "step": 3146
+    },
+    {
+      "epoch": 0.7152272727272727,
+      "grad_norm": 2.08679084699186,
+      "learning_rate": 1.187983781628374e-06,
+      "loss": 0.0998,
+      "step": 3147
+    },
+    {
+      "epoch": 0.7154545454545455,
+      "grad_norm": 2.1285697751336428,
+      "learning_rate": 1.1879450157706338e-06,
+      "loss": 0.1125,
+      "step": 3148
+    },
+    {
+      "epoch": 0.7156818181818182,
+      "grad_norm": 2.436148391891243,
+      "learning_rate": 1.1879062384334628e-06,
+      "loss": 0.1093,
+      "step": 3149
+    },
+    {
+      "epoch": 0.7159090909090909,
+      "grad_norm": 1.4794121004470495,
+      "learning_rate": 1.1878674496176514e-06,
+      "loss": 0.0408,
+      "step": 3150
+    },
+    {
+      "epoch": 0.7161363636363637,
+      "grad_norm": 1.5166443813701576,
+      "learning_rate": 1.1878286493239907e-06,
+      "loss": 0.0377,
+      "step": 3151
+    },
+    {
+      "epoch": 0.7163636363636363,
+      "grad_norm": 1.5714107507587272,
+      "learning_rate": 1.1877898375532721e-06,
+      "loss": 0.0637,
+      "step": 3152
+    },
+    {
+      "epoch": 0.7165909090909091,
+      "grad_norm": 1.9645946917950303,
+      "learning_rate": 1.1877510143062867e-06,
+      "loss": 0.0595,
+      "step": 3153
+    },
+    {
+      "epoch": 0.7168181818181818,
+      "grad_norm": 2.3699215454069162,
+      "learning_rate": 1.1877121795838265e-06,
+      "loss": 0.0913,
+      "step": 3154
+    },
+    {
+      "epoch": 0.7170454545454545,
+      "grad_norm": 1.454902466349191,
+      "learning_rate": 1.187673333386683e-06,
+      "loss": 0.0932,
+      "step": 3155
+    },
+    {
+      "epoch": 0.7172727272727273,
+      "grad_norm": 1.5051810496325053,
+      "learning_rate": 1.187634475715649e-06,
+      "loss": 0.0752,
+      "step": 3156
+    },
+    {
+      "epoch": 0.7175,
+      "grad_norm": 2.2670369821312706,
+      "learning_rate": 1.1875956065715162e-06,
+      "loss": 0.1342,
+      "step": 3157
+    },
+    {
+      "epoch": 0.7177272727272728,
+      "grad_norm": 2.300548919317992,
+      "learning_rate": 1.1875567259550776e-06,
+      "loss": 0.0677,
+      "step": 3158
+    },
+    {
+      "epoch": 0.7179545454545454,
+      "grad_norm": 2.1385843604749555,
+      "learning_rate": 1.187517833867126e-06,
+      "loss": 0.0845,
+      "step": 3159
+    },
+    {
+      "epoch": 0.7181818181818181,
+      "grad_norm": 1.5974455151677092,
+      "learning_rate": 1.1874789303084541e-06,
+      "loss": 0.0598,
+      "step": 3160
+    },
+    {
+      "epoch": 0.7184090909090909,
+      "grad_norm": 1.7591596873018964,
+      "learning_rate": 1.1874400152798557e-06,
+      "loss": 0.0695,
+      "step": 3161
+    },
+    {
+      "epoch": 0.7186363636363636,
+      "grad_norm": 0.5225351826960436,
+      "learning_rate": 1.1874010887821242e-06,
+      "loss": 0.0269,
+      "step": 3162
+    },
+    {
+      "epoch": 0.7188636363636364,
+      "grad_norm": 1.8902865956766421,
+      "learning_rate": 1.1873621508160531e-06,
+      "loss": 0.0794,
+      "step": 3163
+    },
+    {
+      "epoch": 0.7190909090909091,
+      "grad_norm": 0.8216304532796697,
+      "learning_rate": 1.1873232013824367e-06,
+      "loss": 0.0142,
+      "step": 3164
+    },
+    {
+      "epoch": 0.7193181818181819,
+      "grad_norm": 0.40351440517212,
+      "learning_rate": 1.1872842404820694e-06,
+      "loss": 0.0161,
+      "step": 3165
+    },
+    {
+      "epoch": 0.7195454545454546,
+      "grad_norm": 2.5073103542189172,
+      "learning_rate": 1.1872452681157453e-06,
+      "loss": 0.1092,
+      "step": 3166
+    },
+    {
+      "epoch": 0.7197727272727272,
+      "grad_norm": 1.497743914853747,
+      "learning_rate": 1.1872062842842596e-06,
+      "loss": 0.0969,
+      "step": 3167
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.139654269379934,
+      "learning_rate": 1.1871672889884067e-06,
+      "loss": 0.0846,
+      "step": 3168
+    },
+    {
+      "epoch": 0.7202272727272727,
+      "grad_norm": 2.117772071683374,
+      "learning_rate": 1.187128282228982e-06,
+      "loss": 0.0868,
+      "step": 3169
+    },
+    {
+      "epoch": 0.7204545454545455,
+      "grad_norm": 2.485002910873948,
+      "learning_rate": 1.187089264006781e-06,
+      "loss": 0.0981,
+      "step": 3170
+    },
+    {
+      "epoch": 0.7206818181818182,
+      "grad_norm": 2.91004722766395,
+      "learning_rate": 1.1870502343225992e-06,
+      "loss": 0.1149,
+      "step": 3171
+    },
+    {
+      "epoch": 0.7209090909090909,
+      "grad_norm": 1.396443583765936,
+      "learning_rate": 1.1870111931772326e-06,
+      "loss": 0.0691,
+      "step": 3172
+    },
+    {
+      "epoch": 0.7211363636363637,
+      "grad_norm": 2.129562705166514,
+      "learning_rate": 1.1869721405714774e-06,
+      "loss": 0.1264,
+      "step": 3173
+    },
+    {
+      "epoch": 0.7213636363636363,
+      "grad_norm": 1.5833857330120678,
+      "learning_rate": 1.1869330765061299e-06,
+      "loss": 0.0649,
+      "step": 3174
+    },
+    {
+      "epoch": 0.7215909090909091,
+      "grad_norm": 3.2619414007443255,
+      "learning_rate": 1.1868940009819863e-06,
+      "loss": 0.1261,
+      "step": 3175
+    },
+    {
+      "epoch": 0.7218181818181818,
+      "grad_norm": 1.351918142342257,
+      "learning_rate": 1.186854913999844e-06,
+      "loss": 0.0549,
+      "step": 3176
+    },
+    {
+      "epoch": 0.7220454545454545,
+      "grad_norm": 2.366897879675844,
+      "learning_rate": 1.1868158155604998e-06,
+      "loss": 0.1112,
+      "step": 3177
+    },
+    {
+      "epoch": 0.7222727272727273,
+      "grad_norm": 1.6818860872916088,
+      "learning_rate": 1.186776705664751e-06,
+      "loss": 0.0679,
+      "step": 3178
+    },
+    {
+      "epoch": 0.7225,
+      "grad_norm": 1.7503965463503908,
+      "learning_rate": 1.186737584313395e-06,
+      "loss": 0.0847,
+      "step": 3179
+    },
+    {
+      "epoch": 0.7227272727272728,
+      "grad_norm": 0.5019435450071682,
+      "learning_rate": 1.1866984515072296e-06,
+      "loss": 0.0235,
+      "step": 3180
+    },
+    {
+      "epoch": 0.7229545454545454,
+      "grad_norm": 1.435128444912289,
+      "learning_rate": 1.1866593072470527e-06,
+      "loss": 0.0444,
+      "step": 3181
+    },
+    {
+      "epoch": 0.7231818181818181,
+      "grad_norm": 1.5250625366569928,
+      "learning_rate": 1.186620151533663e-06,
+      "loss": 0.0687,
+      "step": 3182
+    },
+    {
+      "epoch": 0.7234090909090909,
+      "grad_norm": 1.9177491323858178,
+      "learning_rate": 1.1865809843678586e-06,
+      "loss": 0.0708,
+      "step": 3183
+    },
+    {
+      "epoch": 0.7236363636363636,
+      "grad_norm": 2.212758907598171,
+      "learning_rate": 1.186541805750438e-06,
+      "loss": 0.0769,
+      "step": 3184
+    },
+    {
+      "epoch": 0.7238636363636364,
+      "grad_norm": 2.94842335147326,
+      "learning_rate": 1.1865026156822004e-06,
+      "loss": 0.2131,
+      "step": 3185
+    },
+    {
+      "epoch": 0.7240909090909091,
+      "grad_norm": 3.385259848757481,
+      "learning_rate": 1.1864634141639448e-06,
+      "loss": 0.1357,
+      "step": 3186
+    },
+    {
+      "epoch": 0.7243181818181819,
+      "grad_norm": 1.6809327785054005,
+      "learning_rate": 1.1864242011964708e-06,
+      "loss": 0.0962,
+      "step": 3187
+    },
+    {
+      "epoch": 0.7245454545454545,
+      "grad_norm": 2.2559511161667993,
+      "learning_rate": 1.1863849767805776e-06,
+      "loss": 0.1399,
+      "step": 3188
+    },
+    {
+      "epoch": 0.7247727272727272,
+      "grad_norm": 1.3449604012001577,
+      "learning_rate": 1.1863457409170655e-06,
+      "loss": 0.0893,
+      "step": 3189
+    },
+    {
+      "epoch": 0.725,
+      "grad_norm": 2.278972184831493,
+      "learning_rate": 1.1863064936067345e-06,
+      "loss": 0.0792,
+      "step": 3190
+    },
+    {
+      "epoch": 0.7252272727272727,
+      "grad_norm": 2.4107250789089156,
+      "learning_rate": 1.1862672348503848e-06,
+      "loss": 0.1169,
+      "step": 3191
+    },
+    {
+      "epoch": 0.7254545454545455,
+      "grad_norm": 2.0610989793184338,
+      "learning_rate": 1.186227964648817e-06,
+      "loss": 0.1092,
+      "step": 3192
+    },
+    {
+      "epoch": 0.7256818181818182,
+      "grad_norm": 1.8617024317461783,
+      "learning_rate": 1.186188683002832e-06,
+      "loss": 0.0723,
+      "step": 3193
+    },
+    {
+      "epoch": 0.725909090909091,
+      "grad_norm": 2.733632033028653,
+      "learning_rate": 1.1861493899132306e-06,
+      "loss": 0.1463,
+      "step": 3194
+    },
+    {
+      "epoch": 0.7261363636363637,
+      "grad_norm": 1.5271419655125813,
+      "learning_rate": 1.1861100853808142e-06,
+      "loss": 0.0855,
+      "step": 3195
+    },
+    {
+      "epoch": 0.7263636363636363,
+      "grad_norm": 2.6723111362646175,
+      "learning_rate": 1.186070769406384e-06,
+      "loss": 0.1031,
+      "step": 3196
+    },
+    {
+      "epoch": 0.7265909090909091,
+      "grad_norm": 1.1272659153242577,
+      "learning_rate": 1.1860314419907422e-06,
+      "loss": 0.0682,
+      "step": 3197
+    },
+    {
+      "epoch": 0.7268181818181818,
+      "grad_norm": 1.8472736925843287,
+      "learning_rate": 1.1859921031346905e-06,
+      "loss": 0.1092,
+      "step": 3198
+    },
+    {
+      "epoch": 0.7270454545454546,
+      "grad_norm": 2.0828662799018844,
+      "learning_rate": 1.1859527528390311e-06,
+      "loss": 0.104,
+      "step": 3199
+    },
+    {
+      "epoch": 0.7272727272727273,
+      "grad_norm": 2.432341708226813,
+      "learning_rate": 1.1859133911045664e-06,
+      "loss": 0.0781,
+      "step": 3200
+    },
+    {
+      "epoch": 0.7275,
+      "grad_norm": 1.3937014970530908,
+      "learning_rate": 1.185874017932099e-06,
+      "loss": 0.0525,
+      "step": 3201
+    },
+    {
+      "epoch": 0.7277272727272728,
+      "grad_norm": 2.490244304188187,
+      "learning_rate": 1.185834633322432e-06,
+      "loss": 0.1294,
+      "step": 3202
+    },
+    {
+      "epoch": 0.7279545454545454,
+      "grad_norm": 3.104183133119907,
+      "learning_rate": 1.1857952372763683e-06,
+      "loss": 0.0972,
+      "step": 3203
+    },
+    {
+      "epoch": 0.7281818181818182,
+      "grad_norm": 1.7604463008338758,
+      "learning_rate": 1.1857558297947116e-06,
+      "loss": 0.0683,
+      "step": 3204
+    },
+    {
+      "epoch": 0.7284090909090909,
+      "grad_norm": 1.5287338673494095,
+      "learning_rate": 1.1857164108782651e-06,
+      "loss": 0.0724,
+      "step": 3205
+    },
+    {
+      "epoch": 0.7286363636363636,
+      "grad_norm": 2.06471300175637,
+      "learning_rate": 1.1856769805278327e-06,
+      "loss": 0.0727,
+      "step": 3206
+    },
+    {
+      "epoch": 0.7288636363636364,
+      "grad_norm": 0.950340616734179,
+      "learning_rate": 1.1856375387442185e-06,
+      "loss": 0.037,
+      "step": 3207
+    },
+    {
+      "epoch": 0.7290909090909091,
+      "grad_norm": 2.481422076770132,
+      "learning_rate": 1.185598085528227e-06,
+      "loss": 0.1345,
+      "step": 3208
+    },
+    {
+      "epoch": 0.7293181818181819,
+      "grad_norm": 2.6220763826771556,
+      "learning_rate": 1.1855586208806622e-06,
+      "loss": 0.1216,
+      "step": 3209
+    },
+    {
+      "epoch": 0.7295454545454545,
+      "grad_norm": 2.2095731312324016,
+      "learning_rate": 1.1855191448023294e-06,
+      "loss": 0.0912,
+      "step": 3210
+    },
+    {
+      "epoch": 0.7297727272727272,
+      "grad_norm": 3.6546617091760063,
+      "learning_rate": 1.1854796572940332e-06,
+      "loss": 0.2045,
+      "step": 3211
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.158278437132916,
+      "learning_rate": 1.1854401583565792e-06,
+      "loss": 0.11,
+      "step": 3212
+    },
+    {
+      "epoch": 0.7302272727272727,
+      "grad_norm": 1.496847874911391,
+      "learning_rate": 1.1854006479907726e-06,
+      "loss": 0.0481,
+      "step": 3213
+    },
+    {
+      "epoch": 0.7304545454545455,
+      "grad_norm": 1.5707416717177456,
+      "learning_rate": 1.185361126197419e-06,
+      "loss": 0.0812,
+      "step": 3214
+    },
+    {
+      "epoch": 0.7306818181818182,
+      "grad_norm": 1.9107480880954688,
+      "learning_rate": 1.1853215929773243e-06,
+      "loss": 0.1021,
+      "step": 3215
+    },
+    {
+      "epoch": 0.730909090909091,
+      "grad_norm": 2.8738878147094757,
+      "learning_rate": 1.1852820483312951e-06,
+      "loss": 0.0683,
+      "step": 3216
+    },
+    {
+      "epoch": 0.7311363636363636,
+      "grad_norm": 2.9333047217365253,
+      "learning_rate": 1.1852424922601372e-06,
+      "loss": 0.0829,
+      "step": 3217
+    },
+    {
+      "epoch": 0.7313636363636363,
+      "grad_norm": 1.901951418329287,
+      "learning_rate": 1.1852029247646576e-06,
+      "loss": 0.079,
+      "step": 3218
+    },
+    {
+      "epoch": 0.7315909090909091,
+      "grad_norm": 1.9062607414109665,
+      "learning_rate": 1.185163345845663e-06,
+      "loss": 0.0944,
+      "step": 3219
+    },
+    {
+      "epoch": 0.7318181818181818,
+      "grad_norm": 1.6828342850850389,
+      "learning_rate": 1.1851237555039607e-06,
+      "loss": 0.0789,
+      "step": 3220
+    },
+    {
+      "epoch": 0.7320454545454546,
+      "grad_norm": 1.174404884768972,
+      "learning_rate": 1.1850841537403577e-06,
+      "loss": 0.0771,
+      "step": 3221
+    },
+    {
+      "epoch": 0.7322727272727273,
+      "grad_norm": 1.774141267503811,
+      "learning_rate": 1.1850445405556615e-06,
+      "loss": 0.0763,
+      "step": 3222
+    },
+    {
+      "epoch": 0.7325,
+      "grad_norm": 3.4857435964102987,
+      "learning_rate": 1.1850049159506804e-06,
+      "loss": 0.1433,
+      "step": 3223
+    },
+    {
+      "epoch": 0.7327272727272728,
+      "grad_norm": 1.6557428863044095,
+      "learning_rate": 1.1849652799262219e-06,
+      "loss": 0.0871,
+      "step": 3224
+    },
+    {
+      "epoch": 0.7329545454545454,
+      "grad_norm": 1.5516715241016872,
+      "learning_rate": 1.1849256324830943e-06,
+      "loss": 0.0761,
+      "step": 3225
+    },
+    {
+      "epoch": 0.7331818181818182,
+      "grad_norm": 1.96342263801315,
+      "learning_rate": 1.1848859736221062e-06,
+      "loss": 0.0994,
+      "step": 3226
+    },
+    {
+      "epoch": 0.7334090909090909,
+      "grad_norm": 1.9618139517406206,
+      "learning_rate": 1.1848463033440667e-06,
+      "loss": 0.1035,
+      "step": 3227
+    },
+    {
+      "epoch": 0.7336363636363636,
+      "grad_norm": 2.2287714535389167,
+      "learning_rate": 1.1848066216497842e-06,
+      "loss": 0.1008,
+      "step": 3228
+    },
+    {
+      "epoch": 0.7338636363636364,
+      "grad_norm": 1.1351221917498473,
+      "learning_rate": 1.184766928540068e-06,
+      "loss": 0.0627,
+      "step": 3229
+    },
+    {
+      "epoch": 0.7340909090909091,
+      "grad_norm": 1.4725679259164604,
+      "learning_rate": 1.1847272240157276e-06,
+      "loss": 0.0612,
+      "step": 3230
+    },
+    {
+      "epoch": 0.7343181818181819,
+      "grad_norm": 1.45709355980599,
+      "learning_rate": 1.1846875080775724e-06,
+      "loss": 0.063,
+      "step": 3231
+    },
+    {
+      "epoch": 0.7345454545454545,
+      "grad_norm": 1.1231677776489322,
+      "learning_rate": 1.1846477807264127e-06,
+      "loss": 0.0571,
+      "step": 3232
+    },
+    {
+      "epoch": 0.7347727272727272,
+      "grad_norm": 3.210313293433556,
+      "learning_rate": 1.1846080419630584e-06,
+      "loss": 0.1458,
+      "step": 3233
+    },
+    {
+      "epoch": 0.735,
+      "grad_norm": 2.0545511473211238,
+      "learning_rate": 1.1845682917883197e-06,
+      "loss": 0.0791,
+      "step": 3234
+    },
+    {
+      "epoch": 0.7352272727272727,
+      "grad_norm": 2.0848501046904255,
+      "learning_rate": 1.1845285302030073e-06,
+      "loss": 0.0741,
+      "step": 3235
+    },
+    {
+      "epoch": 0.7354545454545455,
+      "grad_norm": 1.317834306135167,
+      "learning_rate": 1.1844887572079322e-06,
+      "loss": 0.0702,
+      "step": 3236
+    },
+    {
+      "epoch": 0.7356818181818182,
+      "grad_norm": 2.4484661653830413,
+      "learning_rate": 1.1844489728039052e-06,
+      "loss": 0.0806,
+      "step": 3237
+    },
+    {
+      "epoch": 0.735909090909091,
+      "grad_norm": 3.0039839516264815,
+      "learning_rate": 1.1844091769917376e-06,
+      "loss": 0.0949,
+      "step": 3238
+    },
+    {
+      "epoch": 0.7361363636363636,
+      "grad_norm": 0.687336011150035,
+      "learning_rate": 1.1843693697722408e-06,
+      "loss": 0.0319,
+      "step": 3239
+    },
+    {
+      "epoch": 0.7363636363636363,
+      "grad_norm": 1.6157568432622522,
+      "learning_rate": 1.1843295511462269e-06,
+      "loss": 0.0423,
+      "step": 3240
+    },
+    {
+      "epoch": 0.7365909090909091,
+      "grad_norm": 1.831903963630847,
+      "learning_rate": 1.1842897211145075e-06,
+      "loss": 0.0746,
+      "step": 3241
+    },
+    {
+      "epoch": 0.7368181818181818,
+      "grad_norm": 1.2258028537584886,
+      "learning_rate": 1.184249879677895e-06,
+      "loss": 0.0411,
+      "step": 3242
+    },
+    {
+      "epoch": 0.7370454545454546,
+      "grad_norm": 2.380015229557806,
+      "learning_rate": 1.1842100268372018e-06,
+      "loss": 0.0596,
+      "step": 3243
+    },
+    {
+      "epoch": 0.7372727272727273,
+      "grad_norm": 1.649667395948596,
+      "learning_rate": 1.1841701625932405e-06,
+      "loss": 0.1129,
+      "step": 3244
+    },
+    {
+      "epoch": 0.7375,
+      "grad_norm": 1.4179784036217702,
+      "learning_rate": 1.1841302869468242e-06,
+      "loss": 0.0485,
+      "step": 3245
+    },
+    {
+      "epoch": 0.7377272727272727,
+      "grad_norm": 1.6187745419785824,
+      "learning_rate": 1.1840903998987657e-06,
+      "loss": 0.0666,
+      "step": 3246
+    },
+    {
+      "epoch": 0.7379545454545454,
+      "grad_norm": 2.240181317638905,
+      "learning_rate": 1.1840505014498788e-06,
+      "loss": 0.089,
+      "step": 3247
+    },
+    {
+      "epoch": 0.7381818181818182,
+      "grad_norm": 2.4024689352249213,
+      "learning_rate": 1.1840105916009767e-06,
+      "loss": 0.1032,
+      "step": 3248
+    },
+    {
+      "epoch": 0.7384090909090909,
+      "grad_norm": 1.147256062979827,
+      "learning_rate": 1.1839706703528734e-06,
+      "loss": 0.0344,
+      "step": 3249
+    },
+    {
+      "epoch": 0.7386363636363636,
+      "grad_norm": 4.511983511507155,
+      "learning_rate": 1.183930737706383e-06,
+      "loss": 0.2493,
+      "step": 3250
+    },
+    {
+      "epoch": 0.7388636363636364,
+      "grad_norm": 1.5968451676567126,
+      "learning_rate": 1.1838907936623196e-06,
+      "loss": 0.0717,
+      "step": 3251
+    },
+    {
+      "epoch": 0.7390909090909091,
+      "grad_norm": 1.6073553834104273,
+      "learning_rate": 1.183850838221498e-06,
+      "loss": 0.074,
+      "step": 3252
+    },
+    {
+      "epoch": 0.7393181818181818,
+      "grad_norm": 2.183917496233476,
+      "learning_rate": 1.1838108713847329e-06,
+      "loss": 0.0913,
+      "step": 3253
+    },
+    {
+      "epoch": 0.7395454545454545,
+      "grad_norm": 1.8747982507553513,
+      "learning_rate": 1.183770893152839e-06,
+      "loss": 0.105,
+      "step": 3254
+    },
+    {
+      "epoch": 0.7397727272727272,
+      "grad_norm": 1.2815578803412901,
+      "learning_rate": 1.183730903526632e-06,
+      "loss": 0.0594,
+      "step": 3255
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.357343925212435,
+      "learning_rate": 1.183690902506927e-06,
+      "loss": 0.0885,
+      "step": 3256
+    },
+    {
+      "epoch": 0.7402272727272727,
+      "grad_norm": 1.4103081470632541,
+      "learning_rate": 1.18365089009454e-06,
+      "loss": 0.0675,
+      "step": 3257
+    },
+    {
+      "epoch": 0.7404545454545455,
+      "grad_norm": 2.016873168940554,
+      "learning_rate": 1.1836108662902865e-06,
+      "loss": 0.1197,
+      "step": 3258
+    },
+    {
+      "epoch": 0.7406818181818182,
+      "grad_norm": 2.3671256716421087,
+      "learning_rate": 1.183570831094983e-06,
+      "loss": 0.1135,
+      "step": 3259
+    },
+    {
+      "epoch": 0.740909090909091,
+      "grad_norm": 2.014540007575336,
+      "learning_rate": 1.1835307845094457e-06,
+      "loss": 0.1127,
+      "step": 3260
+    },
+    {
+      "epoch": 0.7411363636363636,
+      "grad_norm": 0.9885102573689944,
+      "learning_rate": 1.1834907265344913e-06,
+      "loss": 0.0567,
+      "step": 3261
+    },
+    {
+      "epoch": 0.7413636363636363,
+      "grad_norm": 2.3830323187964537,
+      "learning_rate": 1.1834506571709368e-06,
+      "loss": 0.1053,
+      "step": 3262
+    },
+    {
+      "epoch": 0.7415909090909091,
+      "grad_norm": 2.193964553053558,
+      "learning_rate": 1.183410576419599e-06,
+      "loss": 0.0681,
+      "step": 3263
+    },
+    {
+      "epoch": 0.7418181818181818,
+      "grad_norm": 2.6673717849790672,
+      "learning_rate": 1.1833704842812955e-06,
+      "loss": 0.1454,
+      "step": 3264
+    },
+    {
+      "epoch": 0.7420454545454546,
+      "grad_norm": 1.6345198748515106,
+      "learning_rate": 1.1833303807568433e-06,
+      "loss": 0.0748,
+      "step": 3265
+    },
+    {
+      "epoch": 0.7422727272727273,
+      "grad_norm": 3.968089561074626,
+      "learning_rate": 1.1832902658470608e-06,
+      "loss": 0.1623,
+      "step": 3266
+    },
+    {
+      "epoch": 0.7425,
+      "grad_norm": 1.8225018749891344,
+      "learning_rate": 1.183250139552766e-06,
+      "loss": 0.0915,
+      "step": 3267
+    },
+    {
+      "epoch": 0.7427272727272727,
+      "grad_norm": 1.4141595050600955,
+      "learning_rate": 1.1832100018747766e-06,
+      "loss": 0.0566,
+      "step": 3268
+    },
+    {
+      "epoch": 0.7429545454545454,
+      "grad_norm": 1.919292322235338,
+      "learning_rate": 1.1831698528139115e-06,
+      "loss": 0.0732,
+      "step": 3269
+    },
+    {
+      "epoch": 0.7431818181818182,
+      "grad_norm": 2.8931410117353584,
+      "learning_rate": 1.1831296923709895e-06,
+      "loss": 0.1162,
+      "step": 3270
+    },
+    {
+      "epoch": 0.7434090909090909,
+      "grad_norm": 1.9135779168832838,
+      "learning_rate": 1.1830895205468293e-06,
+      "loss": 0.1066,
+      "step": 3271
+    },
+    {
+      "epoch": 0.7436363636363637,
+      "grad_norm": 1.8792280999261253,
+      "learning_rate": 1.1830493373422499e-06,
+      "loss": 0.1314,
+      "step": 3272
+    },
+    {
+      "epoch": 0.7438636363636364,
+      "grad_norm": 1.8218932656372298,
+      "learning_rate": 1.1830091427580713e-06,
+      "loss": 0.1352,
+      "step": 3273
+    },
+    {
+      "epoch": 0.7440909090909091,
+      "grad_norm": 1.2987797527715346,
+      "learning_rate": 1.1829689367951125e-06,
+      "loss": 0.0463,
+      "step": 3274
+    },
+    {
+      "epoch": 0.7443181818181818,
+      "grad_norm": 2.5721891414070024,
+      "learning_rate": 1.182928719454194e-06,
+      "loss": 0.1776,
+      "step": 3275
+    },
+    {
+      "epoch": 0.7445454545454545,
+      "grad_norm": 1.6233839024791856,
+      "learning_rate": 1.182888490736135e-06,
+      "loss": 0.0742,
+      "step": 3276
+    },
+    {
+      "epoch": 0.7447727272727273,
+      "grad_norm": 2.896390602106791,
+      "learning_rate": 1.1828482506417569e-06,
+      "loss": 0.0796,
+      "step": 3277
+    },
+    {
+      "epoch": 0.745,
+      "grad_norm": 1.623488463700975,
+      "learning_rate": 1.1828079991718793e-06,
+      "loss": 0.0628,
+      "step": 3278
+    },
+    {
+      "epoch": 0.7452272727272727,
+      "grad_norm": 1.3014648179713673,
+      "learning_rate": 1.1827677363273238e-06,
+      "loss": 0.0611,
+      "step": 3279
+    },
+    {
+      "epoch": 0.7454545454545455,
+      "grad_norm": 2.5095720061910805,
+      "learning_rate": 1.182727462108911e-06,
+      "loss": 0.1494,
+      "step": 3280
+    },
+    {
+      "epoch": 0.7456818181818182,
+      "grad_norm": 2.178770518131945,
+      "learning_rate": 1.1826871765174622e-06,
+      "loss": 0.1029,
+      "step": 3281
+    },
+    {
+      "epoch": 0.7459090909090909,
+      "grad_norm": 2.362217877126384,
+      "learning_rate": 1.1826468795537988e-06,
+      "loss": 0.0889,
+      "step": 3282
+    },
+    {
+      "epoch": 0.7461363636363636,
+      "grad_norm": 1.1757622093133553,
+      "learning_rate": 1.1826065712187428e-06,
+      "loss": 0.0629,
+      "step": 3283
+    },
+    {
+      "epoch": 0.7463636363636363,
+      "grad_norm": 1.3468500090266693,
+      "learning_rate": 1.182566251513116e-06,
+      "loss": 0.0719,
+      "step": 3284
+    },
+    {
+      "epoch": 0.7465909090909091,
+      "grad_norm": 1.9654351822233602,
+      "learning_rate": 1.1825259204377406e-06,
+      "loss": 0.1341,
+      "step": 3285
+    },
+    {
+      "epoch": 0.7468181818181818,
+      "grad_norm": 2.1532896752213584,
+      "learning_rate": 1.1824855779934392e-06,
+      "loss": 0.0638,
+      "step": 3286
+    },
+    {
+      "epoch": 0.7470454545454546,
+      "grad_norm": 2.0604917748295923,
+      "learning_rate": 1.182445224181034e-06,
+      "loss": 0.0745,
+      "step": 3287
+    },
+    {
+      "epoch": 0.7472727272727273,
+      "grad_norm": 1.996407718380684,
+      "learning_rate": 1.182404859001348e-06,
+      "loss": 0.076,
+      "step": 3288
+    },
+    {
+      "epoch": 0.7475,
+      "grad_norm": 2.863017937299549,
+      "learning_rate": 1.182364482455205e-06,
+      "loss": 0.0943,
+      "step": 3289
+    },
+    {
+      "epoch": 0.7477272727272727,
+      "grad_norm": 1.3918360803274215,
+      "learning_rate": 1.1823240945434276e-06,
+      "loss": 0.0547,
+      "step": 3290
+    },
+    {
+      "epoch": 0.7479545454545454,
+      "grad_norm": 2.445793399673117,
+      "learning_rate": 1.1822836952668397e-06,
+      "loss": 0.1237,
+      "step": 3291
+    },
+    {
+      "epoch": 0.7481818181818182,
+      "grad_norm": 1.6064402814243726,
+      "learning_rate": 1.1822432846262648e-06,
+      "loss": 0.0746,
+      "step": 3292
+    },
+    {
+      "epoch": 0.7484090909090909,
+      "grad_norm": 2.0244437113878515,
+      "learning_rate": 1.1822028626225272e-06,
+      "loss": 0.1429,
+      "step": 3293
+    },
+    {
+      "epoch": 0.7486363636363637,
+      "grad_norm": 2.0832940981451413,
+      "learning_rate": 1.1821624292564512e-06,
+      "loss": 0.1229,
+      "step": 3294
+    },
+    {
+      "epoch": 0.7488636363636364,
+      "grad_norm": 2.7796158841103082,
+      "learning_rate": 1.1821219845288611e-06,
+      "loss": 0.0805,
+      "step": 3295
+    },
+    {
+      "epoch": 0.7490909090909091,
+      "grad_norm": 1.4675993985829239,
+      "learning_rate": 1.182081528440582e-06,
+      "loss": 0.0696,
+      "step": 3296
+    },
+    {
+      "epoch": 0.7493181818181818,
+      "grad_norm": 3.0289247069586365,
+      "learning_rate": 1.1820410609924388e-06,
+      "loss": 0.1692,
+      "step": 3297
+    },
+    {
+      "epoch": 0.7495454545454545,
+      "grad_norm": 1.5002653238061479,
+      "learning_rate": 1.1820005821852561e-06,
+      "loss": 0.0846,
+      "step": 3298
+    },
+    {
+      "epoch": 0.7497727272727273,
+      "grad_norm": 1.7319434468593593,
+      "learning_rate": 1.1819600920198601e-06,
+      "loss": 0.0525,
+      "step": 3299
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2018515748754734,
+      "learning_rate": 1.1819195904970762e-06,
+      "loss": 0.0497,
+      "step": 3300
+    },
+    {
+      "epoch": 0.7502272727272727,
+      "grad_norm": 1.3002477695049235,
+      "learning_rate": 1.18187907761773e-06,
+      "loss": 0.0322,
+      "step": 3301
+    },
+    {
+      "epoch": 0.7504545454545455,
+      "grad_norm": 1.8211852312667853,
+      "learning_rate": 1.181838553382648e-06,
+      "loss": 0.0975,
+      "step": 3302
+    },
+    {
+      "epoch": 0.7506818181818182,
+      "grad_norm": 2.4302741166103994,
+      "learning_rate": 1.1817980177926566e-06,
+      "loss": 0.0856,
+      "step": 3303
+    },
+    {
+      "epoch": 0.7509090909090909,
+      "grad_norm": 1.6276678636430295,
+      "learning_rate": 1.181757470848582e-06,
+      "loss": 0.0533,
+      "step": 3304
+    },
+    {
+      "epoch": 0.7511363636363636,
+      "grad_norm": 2.216820951126393,
+      "learning_rate": 1.1817169125512512e-06,
+      "loss": 0.1731,
+      "step": 3305
+    },
+    {
+      "epoch": 0.7513636363636363,
+      "grad_norm": 1.9886139996613732,
+      "learning_rate": 1.1816763429014917e-06,
+      "loss": 0.1597,
+      "step": 3306
+    },
+    {
+      "epoch": 0.7515909090909091,
+      "grad_norm": 1.515068811697469,
+      "learning_rate": 1.18163576190013e-06,
+      "loss": 0.0586,
+      "step": 3307
+    },
+    {
+      "epoch": 0.7518181818181818,
+      "grad_norm": 2.593510209720562,
+      "learning_rate": 1.181595169547994e-06,
+      "loss": 0.0999,
+      "step": 3308
+    },
+    {
+      "epoch": 0.7520454545454546,
+      "grad_norm": 2.155435662999111,
+      "learning_rate": 1.1815545658459116e-06,
+      "loss": 0.0667,
+      "step": 3309
+    },
+    {
+      "epoch": 0.7522727272727273,
+      "grad_norm": 1.6139150698829854,
+      "learning_rate": 1.1815139507947107e-06,
+      "loss": 0.0354,
+      "step": 3310
+    },
+    {
+      "epoch": 0.7525,
+      "grad_norm": 2.31010693687784,
+      "learning_rate": 1.1814733243952193e-06,
+      "loss": 0.1126,
+      "step": 3311
+    },
+    {
+      "epoch": 0.7527272727272727,
+      "grad_norm": 1.9167830859349608,
+      "learning_rate": 1.181432686648266e-06,
+      "loss": 0.089,
+      "step": 3312
+    },
+    {
+      "epoch": 0.7529545454545454,
+      "grad_norm": 2.7932409610263402,
+      "learning_rate": 1.1813920375546798e-06,
+      "loss": 0.1521,
+      "step": 3313
+    },
+    {
+      "epoch": 0.7531818181818182,
+      "grad_norm": 2.447145827497416,
+      "learning_rate": 1.181351377115289e-06,
+      "loss": 0.1009,
+      "step": 3314
+    },
+    {
+      "epoch": 0.7534090909090909,
+      "grad_norm": 1.4494323560611377,
+      "learning_rate": 1.181310705330923e-06,
+      "loss": 0.0568,
+      "step": 3315
+    },
+    {
+      "epoch": 0.7536363636363637,
+      "grad_norm": 1.504319947296005,
+      "learning_rate": 1.1812700222024111e-06,
+      "loss": 0.0577,
+      "step": 3316
+    },
+    {
+      "epoch": 0.7538636363636364,
+      "grad_norm": 2.6338885287535136,
+      "learning_rate": 1.181229327730583e-06,
+      "loss": 0.1144,
+      "step": 3317
+    },
+    {
+      "epoch": 0.7540909090909091,
+      "grad_norm": 1.8880351541226525,
+      "learning_rate": 1.181188621916269e-06,
+      "loss": 0.1007,
+      "step": 3318
+    },
+    {
+      "epoch": 0.7543181818181818,
+      "grad_norm": 1.7339479308796308,
+      "learning_rate": 1.181147904760298e-06,
+      "loss": 0.0805,
+      "step": 3319
+    },
+    {
+      "epoch": 0.7545454545454545,
+      "grad_norm": 2.121457200549142,
+      "learning_rate": 1.1811071762635011e-06,
+      "loss": 0.0694,
+      "step": 3320
+    },
+    {
+      "epoch": 0.7547727272727273,
+      "grad_norm": 1.6159838997829656,
+      "learning_rate": 1.1810664364267092e-06,
+      "loss": 0.0779,
+      "step": 3321
+    },
+    {
+      "epoch": 0.755,
+      "grad_norm": 1.7501479639894753,
+      "learning_rate": 1.1810256852507521e-06,
+      "loss": 0.1092,
+      "step": 3322
+    },
+    {
+      "epoch": 0.7552272727272727,
+      "grad_norm": 1.180847240214158,
+      "learning_rate": 1.1809849227364614e-06,
+      "loss": 0.096,
+      "step": 3323
+    },
+    {
+      "epoch": 0.7554545454545455,
+      "grad_norm": 2.8862001661109407,
+      "learning_rate": 1.180944148884668e-06,
+      "loss": 0.0678,
+      "step": 3324
+    },
+    {
+      "epoch": 0.7556818181818182,
+      "grad_norm": 1.365332893248773,
+      "learning_rate": 1.1809033636962037e-06,
+      "loss": 0.0882,
+      "step": 3325
+    },
+    {
+      "epoch": 0.7559090909090909,
+      "grad_norm": 1.6619423932568613,
+      "learning_rate": 1.1808625671718999e-06,
+      "loss": 0.0988,
+      "step": 3326
+    },
+    {
+      "epoch": 0.7561363636363636,
+      "grad_norm": 1.7926391524490464,
+      "learning_rate": 1.1808217593125884e-06,
+      "loss": 0.0939,
+      "step": 3327
+    },
+    {
+      "epoch": 0.7563636363636363,
+      "grad_norm": 1.5590575902634451,
+      "learning_rate": 1.1807809401191018e-06,
+      "loss": 0.1005,
+      "step": 3328
+    },
+    {
+      "epoch": 0.7565909090909091,
+      "grad_norm": 1.213768842859606,
+      "learning_rate": 1.1807401095922722e-06,
+      "loss": 0.0585,
+      "step": 3329
+    },
+    {
+      "epoch": 0.7568181818181818,
+      "grad_norm": 0.8268735113939348,
+      "learning_rate": 1.180699267732932e-06,
+      "loss": 0.0466,
+      "step": 3330
+    },
+    {
+      "epoch": 0.7570454545454546,
+      "grad_norm": 2.468760940442858,
+      "learning_rate": 1.1806584145419144e-06,
+      "loss": 0.0876,
+      "step": 3331
+    },
+    {
+      "epoch": 0.7572727272727273,
+      "grad_norm": 1.7197597289927355,
+      "learning_rate": 1.1806175500200526e-06,
+      "loss": 0.0856,
+      "step": 3332
+    },
+    {
+      "epoch": 0.7575,
+      "grad_norm": 2.0041450401459726,
+      "learning_rate": 1.1805766741681793e-06,
+      "loss": 0.0637,
+      "step": 3333
+    },
+    {
+      "epoch": 0.7577272727272727,
+      "grad_norm": 1.3596367071540918,
+      "learning_rate": 1.1805357869871284e-06,
+      "loss": 0.0393,
+      "step": 3334
+    },
+    {
+      "epoch": 0.7579545454545454,
+      "grad_norm": 2.0649743990212985,
+      "learning_rate": 1.1804948884777336e-06,
+      "loss": 0.0893,
+      "step": 3335
+    },
+    {
+      "epoch": 0.7581818181818182,
+      "grad_norm": 2.142096698417246,
+      "learning_rate": 1.1804539786408292e-06,
+      "loss": 0.0923,
+      "step": 3336
+    },
+    {
+      "epoch": 0.7584090909090909,
+      "grad_norm": 2.290347034374846,
+      "learning_rate": 1.1804130574772488e-06,
+      "loss": 0.1046,
+      "step": 3337
+    },
+    {
+      "epoch": 0.7586363636363637,
+      "grad_norm": 1.9028368198854437,
+      "learning_rate": 1.1803721249878272e-06,
+      "loss": 0.1585,
+      "step": 3338
+    },
+    {
+      "epoch": 0.7588636363636364,
+      "grad_norm": 3.2058924884707416,
+      "learning_rate": 1.1803311811733994e-06,
+      "loss": 0.0831,
+      "step": 3339
+    },
+    {
+      "epoch": 0.759090909090909,
+      "grad_norm": 1.4751778814378802,
+      "learning_rate": 1.1802902260347996e-06,
+      "loss": 0.0476,
+      "step": 3340
+    },
+    {
+      "epoch": 0.7593181818181818,
+      "grad_norm": 3.537018421862872,
+      "learning_rate": 1.1802492595728634e-06,
+      "loss": 0.2044,
+      "step": 3341
+    },
+    {
+      "epoch": 0.7595454545454545,
+      "grad_norm": 1.933743955542011,
+      "learning_rate": 1.1802082817884263e-06,
+      "loss": 0.0685,
+      "step": 3342
+    },
+    {
+      "epoch": 0.7597727272727273,
+      "grad_norm": 1.553572762805268,
+      "learning_rate": 1.1801672926823237e-06,
+      "loss": 0.0654,
+      "step": 3343
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.074285279945886,
+      "learning_rate": 1.1801262922553912e-06,
+      "loss": 0.0871,
+      "step": 3344
+    },
+    {
+      "epoch": 0.7602272727272728,
+      "grad_norm": 1.7264075529555258,
+      "learning_rate": 1.1800852805084655e-06,
+      "loss": 0.0655,
+      "step": 3345
+    },
+    {
+      "epoch": 0.7604545454545455,
+      "grad_norm": 2.2519801308307463,
+      "learning_rate": 1.1800442574423823e-06,
+      "loss": 0.1124,
+      "step": 3346
+    },
+    {
+      "epoch": 0.7606818181818182,
+      "grad_norm": 1.602917066016388,
+      "learning_rate": 1.1800032230579784e-06,
+      "loss": 0.0799,
+      "step": 3347
+    },
+    {
+      "epoch": 0.7609090909090909,
+      "grad_norm": 1.6459386690029847,
+      "learning_rate": 1.1799621773560904e-06,
+      "loss": 0.0791,
+      "step": 3348
+    },
+    {
+      "epoch": 0.7611363636363636,
+      "grad_norm": 1.8548605695108307,
+      "learning_rate": 1.1799211203375556e-06,
+      "loss": 0.0725,
+      "step": 3349
+    },
+    {
+      "epoch": 0.7613636363636364,
+      "grad_norm": 2.290031892432448,
+      "learning_rate": 1.179880052003211e-06,
+      "loss": 0.0893,
+      "step": 3350
+    },
+    {
+      "epoch": 0.7615909090909091,
+      "grad_norm": 2.0692758727203016,
+      "learning_rate": 1.1798389723538942e-06,
+      "loss": 0.1046,
+      "step": 3351
+    },
+    {
+      "epoch": 0.7618181818181818,
+      "grad_norm": 2.9175904275030295,
+      "learning_rate": 1.1797978813904425e-06,
+      "loss": 0.1052,
+      "step": 3352
+    },
+    {
+      "epoch": 0.7620454545454546,
+      "grad_norm": 2.3551895606697277,
+      "learning_rate": 1.1797567791136944e-06,
+      "loss": 0.1302,
+      "step": 3353
+    },
+    {
+      "epoch": 0.7622727272727273,
+      "grad_norm": 1.347866881913114,
+      "learning_rate": 1.1797156655244875e-06,
+      "loss": 0.0664,
+      "step": 3354
+    },
+    {
+      "epoch": 0.7625,
+      "grad_norm": 1.503176986766777,
+      "learning_rate": 1.1796745406236607e-06,
+      "loss": 0.0739,
+      "step": 3355
+    },
+    {
+      "epoch": 0.7627272727272727,
+      "grad_norm": 2.9896871184247567,
+      "learning_rate": 1.1796334044120522e-06,
+      "loss": 0.1707,
+      "step": 3356
+    },
+    {
+      "epoch": 0.7629545454545454,
+      "grad_norm": 2.234268251802013,
+      "learning_rate": 1.179592256890501e-06,
+      "loss": 0.0709,
+      "step": 3357
+    },
+    {
+      "epoch": 0.7631818181818182,
+      "grad_norm": 4.0404951606209325,
+      "learning_rate": 1.179551098059846e-06,
+      "loss": 0.138,
+      "step": 3358
+    },
+    {
+      "epoch": 0.7634090909090909,
+      "grad_norm": 2.5509216126670085,
+      "learning_rate": 1.1795099279209266e-06,
+      "loss": 0.1071,
+      "step": 3359
+    },
+    {
+      "epoch": 0.7636363636363637,
+      "grad_norm": 1.8935370095631916,
+      "learning_rate": 1.1794687464745826e-06,
+      "loss": 0.088,
+      "step": 3360
+    },
+    {
+      "epoch": 0.7638636363636364,
+      "grad_norm": 2.480861985670505,
+      "learning_rate": 1.1794275537216534e-06,
+      "loss": 0.146,
+      "step": 3361
+    },
+    {
+      "epoch": 0.764090909090909,
+      "grad_norm": 1.2609856415324576,
+      "learning_rate": 1.1793863496629793e-06,
+      "loss": 0.0705,
+      "step": 3362
+    },
+    {
+      "epoch": 0.7643181818181818,
+      "grad_norm": 3.0220432754115922,
+      "learning_rate": 1.1793451342994002e-06,
+      "loss": 0.1628,
+      "step": 3363
+    },
+    {
+      "epoch": 0.7645454545454545,
+      "grad_norm": 1.9224428320242208,
+      "learning_rate": 1.1793039076317567e-06,
+      "loss": 0.0695,
+      "step": 3364
+    },
+    {
+      "epoch": 0.7647727272727273,
+      "grad_norm": 1.6903938730840111,
+      "learning_rate": 1.1792626696608895e-06,
+      "loss": 0.073,
+      "step": 3365
+    },
+    {
+      "epoch": 0.765,
+      "grad_norm": 1.4981797875679106,
+      "learning_rate": 1.1792214203876396e-06,
+      "loss": 0.0513,
+      "step": 3366
+    },
+    {
+      "epoch": 0.7652272727272728,
+      "grad_norm": 2.504222686226482,
+      "learning_rate": 1.179180159812848e-06,
+      "loss": 0.1155,
+      "step": 3367
+    },
+    {
+      "epoch": 0.7654545454545455,
+      "grad_norm": 1.1971333380458538,
+      "learning_rate": 1.179138887937356e-06,
+      "loss": 0.0628,
+      "step": 3368
+    },
+    {
+      "epoch": 0.7656818181818181,
+      "grad_norm": 2.314837455035228,
+      "learning_rate": 1.1790976047620053e-06,
+      "loss": 0.0871,
+      "step": 3369
+    },
+    {
+      "epoch": 0.7659090909090909,
+      "grad_norm": 1.6844740449515379,
+      "learning_rate": 1.179056310287638e-06,
+      "loss": 0.1056,
+      "step": 3370
+    },
+    {
+      "epoch": 0.7661363636363636,
+      "grad_norm": 1.0860350876050247,
+      "learning_rate": 1.1790150045150958e-06,
+      "loss": 0.0821,
+      "step": 3371
+    },
+    {
+      "epoch": 0.7663636363636364,
+      "grad_norm": 2.662696535040499,
+      "learning_rate": 1.178973687445221e-06,
+      "loss": 0.0699,
+      "step": 3372
+    },
+    {
+      "epoch": 0.7665909090909091,
+      "grad_norm": 2.615296006937713,
+      "learning_rate": 1.1789323590788565e-06,
+      "loss": 0.0863,
+      "step": 3373
+    },
+    {
+      "epoch": 0.7668181818181818,
+      "grad_norm": 2.3695336503138202,
+      "learning_rate": 1.1788910194168446e-06,
+      "loss": 0.113,
+      "step": 3374
+    },
+    {
+      "epoch": 0.7670454545454546,
+      "grad_norm": 1.7596080778550796,
+      "learning_rate": 1.1788496684600287e-06,
+      "loss": 0.1024,
+      "step": 3375
+    },
+    {
+      "epoch": 0.7672727272727272,
+      "grad_norm": 1.3698380663306018,
+      "learning_rate": 1.1788083062092518e-06,
+      "loss": 0.0893,
+      "step": 3376
+    },
+    {
+      "epoch": 0.7675,
+      "grad_norm": 1.4822161369088767,
+      "learning_rate": 1.1787669326653572e-06,
+      "loss": 0.0645,
+      "step": 3377
+    },
+    {
+      "epoch": 0.7677272727272727,
+      "grad_norm": 0.5830928102049023,
+      "learning_rate": 1.1787255478291888e-06,
+      "loss": 0.0354,
+      "step": 3378
+    },
+    {
+      "epoch": 0.7679545454545454,
+      "grad_norm": 2.125505439726047,
+      "learning_rate": 1.1786841517015904e-06,
+      "loss": 0.0907,
+      "step": 3379
+    },
+    {
+      "epoch": 0.7681818181818182,
+      "grad_norm": 1.8042299191394489,
+      "learning_rate": 1.1786427442834065e-06,
+      "loss": 0.0932,
+      "step": 3380
+    },
+    {
+      "epoch": 0.7684090909090909,
+      "grad_norm": 1.7473263017335916,
+      "learning_rate": 1.1786013255754808e-06,
+      "loss": 0.0544,
+      "step": 3381
+    },
+    {
+      "epoch": 0.7686363636363637,
+      "grad_norm": 2.7269418407993964,
+      "learning_rate": 1.1785598955786583e-06,
+      "loss": 0.1311,
+      "step": 3382
+    },
+    {
+      "epoch": 0.7688636363636364,
+      "grad_norm": 2.358638412692771,
+      "learning_rate": 1.178518454293784e-06,
+      "loss": 0.1298,
+      "step": 3383
+    },
+    {
+      "epoch": 0.769090909090909,
+      "grad_norm": 2.0397268692535553,
+      "learning_rate": 1.1784770017217027e-06,
+      "loss": 0.1489,
+      "step": 3384
+    },
+    {
+      "epoch": 0.7693181818181818,
+      "grad_norm": 1.533382519059826,
+      "learning_rate": 1.1784355378632594e-06,
+      "loss": 0.0488,
+      "step": 3385
+    },
+    {
+      "epoch": 0.7695454545454545,
+      "grad_norm": 1.850563481511943,
+      "learning_rate": 1.1783940627193002e-06,
+      "loss": 0.0461,
+      "step": 3386
+    },
+    {
+      "epoch": 0.7697727272727273,
+      "grad_norm": 1.9365422837104185,
+      "learning_rate": 1.1783525762906707e-06,
+      "loss": 0.06,
+      "step": 3387
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.249703136345413,
+      "learning_rate": 1.1783110785782168e-06,
+      "loss": 0.0741,
+      "step": 3388
+    },
+    {
+      "epoch": 0.7702272727272728,
+      "grad_norm": 1.3315163669213756,
+      "learning_rate": 1.1782695695827846e-06,
+      "loss": 0.0343,
+      "step": 3389
+    },
+    {
+      "epoch": 0.7704545454545455,
+      "grad_norm": 1.4126341056651222,
+      "learning_rate": 1.1782280493052207e-06,
+      "loss": 0.084,
+      "step": 3390
+    },
+    {
+      "epoch": 0.7706818181818181,
+      "grad_norm": 3.577943308604818,
+      "learning_rate": 1.1781865177463717e-06,
+      "loss": 0.1602,
+      "step": 3391
+    },
+    {
+      "epoch": 0.7709090909090909,
+      "grad_norm": 1.763447962012633,
+      "learning_rate": 1.1781449749070845e-06,
+      "loss": 0.0689,
+      "step": 3392
+    },
+    {
+      "epoch": 0.7711363636363636,
+      "grad_norm": 1.6485968269438755,
+      "learning_rate": 1.1781034207882062e-06,
+      "loss": 0.1082,
+      "step": 3393
+    },
+    {
+      "epoch": 0.7713636363636364,
+      "grad_norm": 1.5245467692120065,
+      "learning_rate": 1.1780618553905843e-06,
+      "loss": 0.0792,
+      "step": 3394
+    },
+    {
+      "epoch": 0.7715909090909091,
+      "grad_norm": 1.4537189605953076,
+      "learning_rate": 1.1780202787150663e-06,
+      "loss": 0.0446,
+      "step": 3395
+    },
+    {
+      "epoch": 0.7718181818181818,
+      "grad_norm": 2.9376444320683706,
+      "learning_rate": 1.1779786907625e-06,
+      "loss": 0.1413,
+      "step": 3396
+    },
+    {
+      "epoch": 0.7720454545454546,
+      "grad_norm": 1.031798674184616,
+      "learning_rate": 1.1779370915337334e-06,
+      "loss": 0.032,
+      "step": 3397
+    },
+    {
+      "epoch": 0.7722727272727272,
+      "grad_norm": 1.6301698928522537,
+      "learning_rate": 1.177895481029615e-06,
+      "loss": 0.0603,
+      "step": 3398
+    },
+    {
+      "epoch": 0.7725,
+      "grad_norm": 2.6940305639213085,
+      "learning_rate": 1.177853859250993e-06,
+      "loss": 0.119,
+      "step": 3399
+    },
+    {
+      "epoch": 0.7727272727272727,
+      "grad_norm": 2.276009870271611,
+      "learning_rate": 1.1778122261987163e-06,
+      "loss": 0.0736,
+      "step": 3400
+    },
+    {
+      "epoch": 0.7729545454545454,
+      "grad_norm": 2.198800222806949,
+      "learning_rate": 1.177770581873634e-06,
+      "loss": 0.0825,
+      "step": 3401
+    },
+    {
+      "epoch": 0.7731818181818182,
+      "grad_norm": 2.2629291850710804,
+      "learning_rate": 1.1777289262765953e-06,
+      "loss": 0.0935,
+      "step": 3402
+    },
+    {
+      "epoch": 0.7734090909090909,
+      "grad_norm": 1.4096062614969453,
+      "learning_rate": 1.1776872594084493e-06,
+      "loss": 0.0682,
+      "step": 3403
+    },
+    {
+      "epoch": 0.7736363636363637,
+      "grad_norm": 1.6381465637127088,
+      "learning_rate": 1.1776455812700458e-06,
+      "loss": 0.0688,
+      "step": 3404
+    },
+    {
+      "epoch": 0.7738636363636363,
+      "grad_norm": 1.1157480835412887,
+      "learning_rate": 1.1776038918622348e-06,
+      "loss": 0.0362,
+      "step": 3405
+    },
+    {
+      "epoch": 0.774090909090909,
+      "grad_norm": 1.5595342337819464,
+      "learning_rate": 1.1775621911858665e-06,
+      "loss": 0.067,
+      "step": 3406
+    },
+    {
+      "epoch": 0.7743181818181818,
+      "grad_norm": 2.3513052133279926,
+      "learning_rate": 1.177520479241791e-06,
+      "loss": 0.0814,
+      "step": 3407
+    },
+    {
+      "epoch": 0.7745454545454545,
+      "grad_norm": 2.4196426912157363,
+      "learning_rate": 1.177478756030859e-06,
+      "loss": 0.0884,
+      "step": 3408
+    },
+    {
+      "epoch": 0.7747727272727273,
+      "grad_norm": 2.924970325060619,
+      "learning_rate": 1.1774370215539215e-06,
+      "loss": 0.1165,
+      "step": 3409
+    },
+    {
+      "epoch": 0.775,
+      "grad_norm": 1.7113955346241214,
+      "learning_rate": 1.177395275811829e-06,
+      "loss": 0.0654,
+      "step": 3410
+    },
+    {
+      "epoch": 0.7752272727272728,
+      "grad_norm": 2.0533270931253886,
+      "learning_rate": 1.1773535188054336e-06,
+      "loss": 0.1191,
+      "step": 3411
+    },
+    {
+      "epoch": 0.7754545454545455,
+      "grad_norm": 1.2860160966192278,
+      "learning_rate": 1.1773117505355861e-06,
+      "loss": 0.0688,
+      "step": 3412
+    },
+    {
+      "epoch": 0.7756818181818181,
+      "grad_norm": 0.9611524728296987,
+      "learning_rate": 1.1772699710031382e-06,
+      "loss": 0.0604,
+      "step": 3413
+    },
+    {
+      "epoch": 0.7759090909090909,
+      "grad_norm": 2.655400306086309,
+      "learning_rate": 1.1772281802089424e-06,
+      "loss": 0.0847,
+      "step": 3414
+    },
+    {
+      "epoch": 0.7761363636363636,
+      "grad_norm": 2.841922010227259,
+      "learning_rate": 1.1771863781538505e-06,
+      "loss": 0.1217,
+      "step": 3415
+    },
+    {
+      "epoch": 0.7763636363636364,
+      "grad_norm": 1.47278958511318,
+      "learning_rate": 1.177144564838715e-06,
+      "loss": 0.0715,
+      "step": 3416
+    },
+    {
+      "epoch": 0.7765909090909091,
+      "grad_norm": 2.003463705239972,
+      "learning_rate": 1.1771027402643885e-06,
+      "loss": 0.0676,
+      "step": 3417
+    },
+    {
+      "epoch": 0.7768181818181819,
+      "grad_norm": 2.1259776719398933,
+      "learning_rate": 1.177060904431724e-06,
+      "loss": 0.156,
+      "step": 3418
+    },
+    {
+      "epoch": 0.7770454545454546,
+      "grad_norm": 2.043442440560488,
+      "learning_rate": 1.1770190573415744e-06,
+      "loss": 0.0912,
+      "step": 3419
+    },
+    {
+      "epoch": 0.7772727272727272,
+      "grad_norm": 1.536357144429238,
+      "learning_rate": 1.1769771989947934e-06,
+      "loss": 0.0924,
+      "step": 3420
+    },
+    {
+      "epoch": 0.7775,
+      "grad_norm": 2.294575235888452,
+      "learning_rate": 1.1769353293922341e-06,
+      "loss": 0.0971,
+      "step": 3421
+    },
+    {
+      "epoch": 0.7777272727272727,
+      "grad_norm": 2.0021104683094983,
+      "learning_rate": 1.1768934485347508e-06,
+      "loss": 0.0936,
+      "step": 3422
+    },
+    {
+      "epoch": 0.7779545454545455,
+      "grad_norm": 1.7301217230132726,
+      "learning_rate": 1.176851556423197e-06,
+      "loss": 0.058,
+      "step": 3423
+    },
+    {
+      "epoch": 0.7781818181818182,
+      "grad_norm": 1.6305618515321931,
+      "learning_rate": 1.1768096530584276e-06,
+      "loss": 0.0617,
+      "step": 3424
+    },
+    {
+      "epoch": 0.7784090909090909,
+      "grad_norm": 1.7359580482744565,
+      "learning_rate": 1.1767677384412963e-06,
+      "loss": 0.079,
+      "step": 3425
+    },
+    {
+      "epoch": 0.7786363636363637,
+      "grad_norm": 2.097106689252018,
+      "learning_rate": 1.1767258125726584e-06,
+      "loss": 0.1342,
+      "step": 3426
+    },
+    {
+      "epoch": 0.7788636363636363,
+      "grad_norm": 2.031401239454995,
+      "learning_rate": 1.1766838754533684e-06,
+      "loss": 0.0794,
+      "step": 3427
+    },
+    {
+      "epoch": 0.7790909090909091,
+      "grad_norm": 2.0544637114892255,
+      "learning_rate": 1.176641927084282e-06,
+      "loss": 0.1301,
+      "step": 3428
+    },
+    {
+      "epoch": 0.7793181818181818,
+      "grad_norm": 1.6725013276194722,
+      "learning_rate": 1.176599967466254e-06,
+      "loss": 0.0847,
+      "step": 3429
+    },
+    {
+      "epoch": 0.7795454545454545,
+      "grad_norm": 2.228918379242946,
+      "learning_rate": 1.1765579966001406e-06,
+      "loss": 0.0899,
+      "step": 3430
+    },
+    {
+      "epoch": 0.7797727272727273,
+      "grad_norm": 1.2795788532339094,
+      "learning_rate": 1.1765160144867972e-06,
+      "loss": 0.0722,
+      "step": 3431
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.009327480794945,
+      "learning_rate": 1.1764740211270803e-06,
+      "loss": 0.1465,
+      "step": 3432
+    },
+    {
+      "epoch": 0.7802272727272728,
+      "grad_norm": 1.591146986855067,
+      "learning_rate": 1.176432016521846e-06,
+      "loss": 0.0843,
+      "step": 3433
+    },
+    {
+      "epoch": 0.7804545454545454,
+      "grad_norm": 4.0103579010841655,
+      "learning_rate": 1.1763900006719506e-06,
+      "loss": 0.184,
+      "step": 3434
+    },
+    {
+      "epoch": 0.7806818181818181,
+      "grad_norm": 1.8713460121415848,
+      "learning_rate": 1.1763479735782513e-06,
+      "loss": 0.0838,
+      "step": 3435
+    },
+    {
+      "epoch": 0.7809090909090909,
+      "grad_norm": 2.7507398879067435,
+      "learning_rate": 1.176305935241605e-06,
+      "loss": 0.082,
+      "step": 3436
+    },
+    {
+      "epoch": 0.7811363636363636,
+      "grad_norm": 1.647689723314221,
+      "learning_rate": 1.176263885662869e-06,
+      "loss": 0.0906,
+      "step": 3437
+    },
+    {
+      "epoch": 0.7813636363636364,
+      "grad_norm": 1.1610470710770897,
+      "learning_rate": 1.1762218248429003e-06,
+      "loss": 0.0512,
+      "step": 3438
+    },
+    {
+      "epoch": 0.7815909090909091,
+      "grad_norm": 1.4338986912727294,
+      "learning_rate": 1.176179752782557e-06,
+      "loss": 0.0881,
+      "step": 3439
+    },
+    {
+      "epoch": 0.7818181818181819,
+      "grad_norm": 2.0183638947088496,
+      "learning_rate": 1.1761376694826974e-06,
+      "loss": 0.1469,
+      "step": 3440
+    },
+    {
+      "epoch": 0.7820454545454546,
+      "grad_norm": 1.779540783346273,
+      "learning_rate": 1.1760955749441786e-06,
+      "loss": 0.0612,
+      "step": 3441
+    },
+    {
+      "epoch": 0.7822727272727272,
+      "grad_norm": 1.4637899499961542,
+      "learning_rate": 1.1760534691678598e-06,
+      "loss": 0.0775,
+      "step": 3442
+    },
+    {
+      "epoch": 0.7825,
+      "grad_norm": 1.266397881351739,
+      "learning_rate": 1.1760113521545995e-06,
+      "loss": 0.0599,
+      "step": 3443
+    },
+    {
+      "epoch": 0.7827272727272727,
+      "grad_norm": 1.3209434208520028,
+      "learning_rate": 1.1759692239052564e-06,
+      "loss": 0.0384,
+      "step": 3444
+    },
+    {
+      "epoch": 0.7829545454545455,
+      "grad_norm": 1.3498468490032287,
+      "learning_rate": 1.1759270844206897e-06,
+      "loss": 0.0505,
+      "step": 3445
+    },
+    {
+      "epoch": 0.7831818181818182,
+      "grad_norm": 1.9729258786087218,
+      "learning_rate": 1.1758849337017587e-06,
+      "loss": 0.0831,
+      "step": 3446
+    },
+    {
+      "epoch": 0.7834090909090909,
+      "grad_norm": 2.4028580658426453,
+      "learning_rate": 1.1758427717493225e-06,
+      "loss": 0.0986,
+      "step": 3447
+    },
+    {
+      "epoch": 0.7836363636363637,
+      "grad_norm": 2.216279579348642,
+      "learning_rate": 1.1758005985642411e-06,
+      "loss": 0.1584,
+      "step": 3448
+    },
+    {
+      "epoch": 0.7838636363636363,
+      "grad_norm": 1.459060850814956,
+      "learning_rate": 1.175758414147375e-06,
+      "loss": 0.0641,
+      "step": 3449
+    },
+    {
+      "epoch": 0.7840909090909091,
+      "grad_norm": 1.6234863554436845,
+      "learning_rate": 1.1757162184995838e-06,
+      "loss": 0.0699,
+      "step": 3450
+    },
+    {
+      "epoch": 0.7843181818181818,
+      "grad_norm": 1.345642109050307,
+      "learning_rate": 1.175674011621728e-06,
+      "loss": 0.0408,
+      "step": 3451
+    },
+    {
+      "epoch": 0.7845454545454545,
+      "grad_norm": 3.0373111002242354,
+      "learning_rate": 1.1756317935146685e-06,
+      "loss": 0.109,
+      "step": 3452
+    },
+    {
+      "epoch": 0.7847727272727273,
+      "grad_norm": 1.7750396393462928,
+      "learning_rate": 1.1755895641792661e-06,
+      "loss": 0.0661,
+      "step": 3453
+    },
+    {
+      "epoch": 0.785,
+      "grad_norm": 1.2197596223619043,
+      "learning_rate": 1.175547323616382e-06,
+      "loss": 0.0454,
+      "step": 3454
+    },
+    {
+      "epoch": 0.7852272727272728,
+      "grad_norm": 1.7348046965381685,
+      "learning_rate": 1.1755050718268773e-06,
+      "loss": 0.0835,
+      "step": 3455
+    },
+    {
+      "epoch": 0.7854545454545454,
+      "grad_norm": 2.08383492972899,
+      "learning_rate": 1.1754628088116138e-06,
+      "loss": 0.0687,
+      "step": 3456
+    },
+    {
+      "epoch": 0.7856818181818181,
+      "grad_norm": 1.7791768213351544,
+      "learning_rate": 1.1754205345714535e-06,
+      "loss": 0.0813,
+      "step": 3457
+    },
+    {
+      "epoch": 0.7859090909090909,
+      "grad_norm": 1.6209951337909214,
+      "learning_rate": 1.1753782491072583e-06,
+      "loss": 0.083,
+      "step": 3458
+    },
+    {
+      "epoch": 0.7861363636363636,
+      "grad_norm": 2.1125261971659564,
+      "learning_rate": 1.17533595241989e-06,
+      "loss": 0.0553,
+      "step": 3459
+    },
+    {
+      "epoch": 0.7863636363636364,
+      "grad_norm": 1.5276248217402943,
+      "learning_rate": 1.1752936445102118e-06,
+      "loss": 0.0353,
+      "step": 3460
+    },
+    {
+      "epoch": 0.7865909090909091,
+      "grad_norm": 3.215804309092215,
+      "learning_rate": 1.1752513253790861e-06,
+      "loss": 0.1224,
+      "step": 3461
+    },
+    {
+      "epoch": 0.7868181818181819,
+      "grad_norm": 1.2939797442170993,
+      "learning_rate": 1.175208995027376e-06,
+      "loss": 0.0528,
+      "step": 3462
+    },
+    {
+      "epoch": 0.7870454545454545,
+      "grad_norm": 1.7469193962551406,
+      "learning_rate": 1.1751666534559445e-06,
+      "loss": 0.0932,
+      "step": 3463
+    },
+    {
+      "epoch": 0.7872727272727272,
+      "grad_norm": 2.6856647280519734,
+      "learning_rate": 1.175124300665655e-06,
+      "loss": 0.1555,
+      "step": 3464
+    },
+    {
+      "epoch": 0.7875,
+      "grad_norm": 1.830276254051702,
+      "learning_rate": 1.1750819366573716e-06,
+      "loss": 0.0778,
+      "step": 3465
+    },
+    {
+      "epoch": 0.7877272727272727,
+      "grad_norm": 1.9980463403268816,
+      "learning_rate": 1.1750395614319576e-06,
+      "loss": 0.0983,
+      "step": 3466
+    },
+    {
+      "epoch": 0.7879545454545455,
+      "grad_norm": 1.9434628405231293,
+      "learning_rate": 1.1749971749902773e-06,
+      "loss": 0.1171,
+      "step": 3467
+    },
+    {
+      "epoch": 0.7881818181818182,
+      "grad_norm": 2.4751182042434325,
+      "learning_rate": 1.1749547773331952e-06,
+      "loss": 0.1283,
+      "step": 3468
+    },
+    {
+      "epoch": 0.788409090909091,
+      "grad_norm": 2.4696288104985142,
+      "learning_rate": 1.1749123684615757e-06,
+      "loss": 0.1607,
+      "step": 3469
+    },
+    {
+      "epoch": 0.7886363636363637,
+      "grad_norm": 1.5787568706392552,
+      "learning_rate": 1.1748699483762836e-06,
+      "loss": 0.0697,
+      "step": 3470
+    },
+    {
+      "epoch": 0.7888636363636363,
+      "grad_norm": 2.0696492582750365,
+      "learning_rate": 1.174827517078184e-06,
+      "loss": 0.0498,
+      "step": 3471
+    },
+    {
+      "epoch": 0.7890909090909091,
+      "grad_norm": 1.2638206371919458,
+      "learning_rate": 1.174785074568142e-06,
+      "loss": 0.0742,
+      "step": 3472
+    },
+    {
+      "epoch": 0.7893181818181818,
+      "grad_norm": 2.723449584930399,
+      "learning_rate": 1.1747426208470234e-06,
+      "loss": 0.1262,
+      "step": 3473
+    },
+    {
+      "epoch": 0.7895454545454546,
+      "grad_norm": 2.414992107111454,
+      "learning_rate": 1.1747001559156936e-06,
+      "loss": 0.0661,
+      "step": 3474
+    },
+    {
+      "epoch": 0.7897727272727273,
+      "grad_norm": 1.7820558645524207,
+      "learning_rate": 1.1746576797750185e-06,
+      "loss": 0.0727,
+      "step": 3475
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9989152203643796,
+      "learning_rate": 1.1746151924258644e-06,
+      "loss": 0.0841,
+      "step": 3476
+    },
+    {
+      "epoch": 0.7902272727272728,
+      "grad_norm": 1.9636499016629976,
+      "learning_rate": 1.1745726938690976e-06,
+      "loss": 0.0687,
+      "step": 3477
+    },
+    {
+      "epoch": 0.7904545454545454,
+      "grad_norm": 1.7938037053137452,
+      "learning_rate": 1.1745301841055851e-06,
+      "loss": 0.0617,
+      "step": 3478
+    },
+    {
+      "epoch": 0.7906818181818182,
+      "grad_norm": 2.810210101723205,
+      "learning_rate": 1.1744876631361931e-06,
+      "loss": 0.0989,
+      "step": 3479
+    },
+    {
+      "epoch": 0.7909090909090909,
+      "grad_norm": 1.445732600731667,
+      "learning_rate": 1.1744451309617893e-06,
+      "loss": 0.0492,
+      "step": 3480
+    },
+    {
+      "epoch": 0.7911363636363636,
+      "grad_norm": 1.9155480872455986,
+      "learning_rate": 1.1744025875832405e-06,
+      "loss": 0.0899,
+      "step": 3481
+    },
+    {
+      "epoch": 0.7913636363636364,
+      "grad_norm": 2.579539331134932,
+      "learning_rate": 1.1743600330014147e-06,
+      "loss": 0.0793,
+      "step": 3482
+    },
+    {
+      "epoch": 0.7915909090909091,
+      "grad_norm": 1.9112923242537943,
+      "learning_rate": 1.1743174672171791e-06,
+      "loss": 0.073,
+      "step": 3483
+    },
+    {
+      "epoch": 0.7918181818181819,
+      "grad_norm": 2.8904819114288323,
+      "learning_rate": 1.174274890231402e-06,
+      "loss": 0.0751,
+      "step": 3484
+    },
+    {
+      "epoch": 0.7920454545454545,
+      "grad_norm": 1.2326226727922684,
+      "learning_rate": 1.174232302044952e-06,
+      "loss": 0.0366,
+      "step": 3485
+    },
+    {
+      "epoch": 0.7922727272727272,
+      "grad_norm": 1.7766681293470674,
+      "learning_rate": 1.174189702658697e-06,
+      "loss": 0.0608,
+      "step": 3486
+    },
+    {
+      "epoch": 0.7925,
+      "grad_norm": 2.0918798778034544,
+      "learning_rate": 1.1741470920735058e-06,
+      "loss": 0.0772,
+      "step": 3487
+    },
+    {
+      "epoch": 0.7927272727272727,
+      "grad_norm": 1.5677400043071592,
+      "learning_rate": 1.1741044702902473e-06,
+      "loss": 0.0428,
+      "step": 3488
+    },
+    {
+      "epoch": 0.7929545454545455,
+      "grad_norm": 1.2118804413443642,
+      "learning_rate": 1.1740618373097907e-06,
+      "loss": 0.0729,
+      "step": 3489
+    },
+    {
+      "epoch": 0.7931818181818182,
+      "grad_norm": 1.8495695214250223,
+      "learning_rate": 1.1740191931330053e-06,
+      "loss": 0.0712,
+      "step": 3490
+    },
+    {
+      "epoch": 0.793409090909091,
+      "grad_norm": 1.756361523892247,
+      "learning_rate": 1.173976537760761e-06,
+      "loss": 0.116,
+      "step": 3491
+    },
+    {
+      "epoch": 0.7936363636363636,
+      "grad_norm": 2.004327134281911,
+      "learning_rate": 1.1739338711939272e-06,
+      "loss": 0.0732,
+      "step": 3492
+    },
+    {
+      "epoch": 0.7938636363636363,
+      "grad_norm": 2.1934256174855986,
+      "learning_rate": 1.173891193433374e-06,
+      "loss": 0.081,
+      "step": 3493
+    },
+    {
+      "epoch": 0.7940909090909091,
+      "grad_norm": 2.460116050851841,
+      "learning_rate": 1.173848504479972e-06,
+      "loss": 0.1268,
+      "step": 3494
+    },
+    {
+      "epoch": 0.7943181818181818,
+      "grad_norm": 2.583925205591781,
+      "learning_rate": 1.1738058043345911e-06,
+      "loss": 0.1072,
+      "step": 3495
+    },
+    {
+      "epoch": 0.7945454545454546,
+      "grad_norm": 1.491296758471164,
+      "learning_rate": 1.1737630929981026e-06,
+      "loss": 0.0593,
+      "step": 3496
+    },
+    {
+      "epoch": 0.7947727272727273,
+      "grad_norm": 2.7855019496495226,
+      "learning_rate": 1.1737203704713773e-06,
+      "loss": 0.1935,
+      "step": 3497
+    },
+    {
+      "epoch": 0.795,
+      "grad_norm": 1.4981743637877658,
+      "learning_rate": 1.1736776367552862e-06,
+      "loss": 0.0648,
+      "step": 3498
+    },
+    {
+      "epoch": 0.7952272727272728,
+      "grad_norm": 2.194093312364921,
+      "learning_rate": 1.1736348918507011e-06,
+      "loss": 0.1096,
+      "step": 3499
+    },
+    {
+      "epoch": 0.7954545454545454,
+      "grad_norm": 1.817298104546263,
+      "learning_rate": 1.1735921357584935e-06,
+      "loss": 0.0807,
+      "step": 3500
+    },
+    {
+      "epoch": 0.7956818181818182,
+      "grad_norm": 1.4973439822244594,
+      "learning_rate": 1.1735493684795348e-06,
+      "loss": 0.0605,
+      "step": 3501
+    },
+    {
+      "epoch": 0.7959090909090909,
+      "grad_norm": 1.568939246498925,
+      "learning_rate": 1.1735065900146978e-06,
+      "loss": 0.0865,
+      "step": 3502
+    },
+    {
+      "epoch": 0.7961363636363636,
+      "grad_norm": 1.9969149701958637,
+      "learning_rate": 1.1734638003648544e-06,
+      "loss": 0.069,
+      "step": 3503
+    },
+    {
+      "epoch": 0.7963636363636364,
+      "grad_norm": 1.9946820116503106,
+      "learning_rate": 1.1734209995308772e-06,
+      "loss": 0.1361,
+      "step": 3504
+    },
+    {
+      "epoch": 0.7965909090909091,
+      "grad_norm": 2.1491283475522374,
+      "learning_rate": 1.1733781875136391e-06,
+      "loss": 0.0652,
+      "step": 3505
+    },
+    {
+      "epoch": 0.7968181818181819,
+      "grad_norm": 1.827286733048145,
+      "learning_rate": 1.173335364314013e-06,
+      "loss": 0.0656,
+      "step": 3506
+    },
+    {
+      "epoch": 0.7970454545454545,
+      "grad_norm": 1.778378846523844,
+      "learning_rate": 1.1732925299328723e-06,
+      "loss": 0.0952,
+      "step": 3507
+    },
+    {
+      "epoch": 0.7972727272727272,
+      "grad_norm": 1.6355588555465883,
+      "learning_rate": 1.1732496843710903e-06,
+      "loss": 0.0848,
+      "step": 3508
+    },
+    {
+      "epoch": 0.7975,
+      "grad_norm": 1.757631610367313,
+      "learning_rate": 1.1732068276295409e-06,
+      "loss": 0.0753,
+      "step": 3509
+    },
+    {
+      "epoch": 0.7977272727272727,
+      "grad_norm": 1.8358146067326448,
+      "learning_rate": 1.1731639597090977e-06,
+      "loss": 0.0495,
+      "step": 3510
+    },
+    {
+      "epoch": 0.7979545454545455,
+      "grad_norm": 2.325449147050743,
+      "learning_rate": 1.173121080610635e-06,
+      "loss": 0.094,
+      "step": 3511
+    },
+    {
+      "epoch": 0.7981818181818182,
+      "grad_norm": 2.700938562375825,
+      "learning_rate": 1.1730781903350273e-06,
+      "loss": 0.1115,
+      "step": 3512
+    },
+    {
+      "epoch": 0.798409090909091,
+      "grad_norm": 1.6383677375601764,
+      "learning_rate": 1.1730352888831493e-06,
+      "loss": 0.069,
+      "step": 3513
+    },
+    {
+      "epoch": 0.7986363636363636,
+      "grad_norm": 1.613734878123903,
+      "learning_rate": 1.1729923762558754e-06,
+      "loss": 0.0606,
+      "step": 3514
+    },
+    {
+      "epoch": 0.7988636363636363,
+      "grad_norm": 2.1219951034804345,
+      "learning_rate": 1.172949452454081e-06,
+      "loss": 0.0689,
+      "step": 3515
+    },
+    {
+      "epoch": 0.7990909090909091,
+      "grad_norm": 2.1918243692100727,
+      "learning_rate": 1.1729065174786414e-06,
+      "loss": 0.1286,
+      "step": 3516
+    },
+    {
+      "epoch": 0.7993181818181818,
+      "grad_norm": 1.4127001843965743,
+      "learning_rate": 1.1728635713304317e-06,
+      "loss": 0.0835,
+      "step": 3517
+    },
+    {
+      "epoch": 0.7995454545454546,
+      "grad_norm": 1.9557449646644307,
+      "learning_rate": 1.1728206140103286e-06,
+      "loss": 0.0685,
+      "step": 3518
+    },
+    {
+      "epoch": 0.7997727272727273,
+      "grad_norm": 2.4898937194075974,
+      "learning_rate": 1.172777645519207e-06,
+      "loss": 0.0652,
+      "step": 3519
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.778293480988369,
+      "learning_rate": 1.1727346658579437e-06,
+      "loss": 0.0561,
+      "step": 3520
+    },
+    {
+      "epoch": 0.8002272727272727,
+      "grad_norm": 1.9863662575469052,
+      "learning_rate": 1.1726916750274148e-06,
+      "loss": 0.0993,
+      "step": 3521
+    },
+    {
+      "epoch": 0.8004545454545454,
+      "grad_norm": 1.5473881765298039,
+      "learning_rate": 1.1726486730284974e-06,
+      "loss": 0.0492,
+      "step": 3522
+    },
+    {
+      "epoch": 0.8006818181818182,
+      "grad_norm": 3.8203427132156746,
+      "learning_rate": 1.172605659862068e-06,
+      "loss": 0.0745,
+      "step": 3523
+    },
+    {
+      "epoch": 0.8009090909090909,
+      "grad_norm": 1.749194153645411,
+      "learning_rate": 1.1725626355290037e-06,
+      "loss": 0.0692,
+      "step": 3524
+    },
+    {
+      "epoch": 0.8011363636363636,
+      "grad_norm": 2.6835277038398337,
+      "learning_rate": 1.1725196000301823e-06,
+      "loss": 0.127,
+      "step": 3525
+    },
+    {
+      "epoch": 0.8013636363636364,
+      "grad_norm": 2.8410293057735183,
+      "learning_rate": 1.1724765533664808e-06,
+      "loss": 0.1694,
+      "step": 3526
+    },
+    {
+      "epoch": 0.8015909090909091,
+      "grad_norm": 1.8773776397186959,
+      "learning_rate": 1.1724334955387776e-06,
+      "loss": 0.079,
+      "step": 3527
+    },
+    {
+      "epoch": 0.8018181818181818,
+      "grad_norm": 2.0765640552693188,
+      "learning_rate": 1.1723904265479498e-06,
+      "loss": 0.1006,
+      "step": 3528
+    },
+    {
+      "epoch": 0.8020454545454545,
+      "grad_norm": 1.212812245859335,
+      "learning_rate": 1.1723473463948768e-06,
+      "loss": 0.0516,
+      "step": 3529
+    },
+    {
+      "epoch": 0.8022727272727272,
+      "grad_norm": 1.8252815457920362,
+      "learning_rate": 1.1723042550804362e-06,
+      "loss": 0.082,
+      "step": 3530
+    },
+    {
+      "epoch": 0.8025,
+      "grad_norm": 1.1129411732592984,
+      "learning_rate": 1.1722611526055073e-06,
+      "loss": 0.0632,
+      "step": 3531
+    },
+    {
+      "epoch": 0.8027272727272727,
+      "grad_norm": 3.0652455876988673,
+      "learning_rate": 1.1722180389709685e-06,
+      "loss": 0.1127,
+      "step": 3532
+    },
+    {
+      "epoch": 0.8029545454545455,
+      "grad_norm": 3.130303760196722,
+      "learning_rate": 1.1721749141776993e-06,
+      "loss": 0.1105,
+      "step": 3533
+    },
+    {
+      "epoch": 0.8031818181818182,
+      "grad_norm": 1.3788846538455606,
+      "learning_rate": 1.1721317782265786e-06,
+      "loss": 0.0586,
+      "step": 3534
+    },
+    {
+      "epoch": 0.803409090909091,
+      "grad_norm": 2.4086925252735427,
+      "learning_rate": 1.1720886311184871e-06,
+      "loss": 0.1117,
+      "step": 3535
+    },
+    {
+      "epoch": 0.8036363636363636,
+      "grad_norm": 1.9950603043144288,
+      "learning_rate": 1.1720454728543034e-06,
+      "loss": 0.1115,
+      "step": 3536
+    },
+    {
+      "epoch": 0.8038636363636363,
+      "grad_norm": 1.6331907827364318,
+      "learning_rate": 1.1720023034349084e-06,
+      "loss": 0.0732,
+      "step": 3537
+    },
+    {
+      "epoch": 0.8040909090909091,
+      "grad_norm": 2.740312185115448,
+      "learning_rate": 1.1719591228611821e-06,
+      "loss": 0.0846,
+      "step": 3538
+    },
+    {
+      "epoch": 0.8043181818181818,
+      "grad_norm": 2.651788965616502,
+      "learning_rate": 1.171915931134005e-06,
+      "loss": 0.1087,
+      "step": 3539
+    },
+    {
+      "epoch": 0.8045454545454546,
+      "grad_norm": 2.7475124244893423,
+      "learning_rate": 1.171872728254258e-06,
+      "loss": 0.098,
+      "step": 3540
+    },
+    {
+      "epoch": 0.8047727272727273,
+      "grad_norm": 2.523185252525073,
+      "learning_rate": 1.171829514222822e-06,
+      "loss": 0.1665,
+      "step": 3541
+    },
+    {
+      "epoch": 0.805,
+      "grad_norm": 2.5016791614616394,
+      "learning_rate": 1.1717862890405782e-06,
+      "loss": 0.0973,
+      "step": 3542
+    },
+    {
+      "epoch": 0.8052272727272727,
+      "grad_norm": 1.2200237895888395,
+      "learning_rate": 1.171743052708408e-06,
+      "loss": 0.0603,
+      "step": 3543
+    },
+    {
+      "epoch": 0.8054545454545454,
+      "grad_norm": 1.1997971493805508,
+      "learning_rate": 1.1716998052271933e-06,
+      "loss": 0.0757,
+      "step": 3544
+    },
+    {
+      "epoch": 0.8056818181818182,
+      "grad_norm": 1.8947205036658537,
+      "learning_rate": 1.1716565465978156e-06,
+      "loss": 0.093,
+      "step": 3545
+    },
+    {
+      "epoch": 0.8059090909090909,
+      "grad_norm": 1.464781373618728,
+      "learning_rate": 1.1716132768211572e-06,
+      "loss": 0.0603,
+      "step": 3546
+    },
+    {
+      "epoch": 0.8061363636363637,
+      "grad_norm": 1.9631789584104535,
+      "learning_rate": 1.1715699958981006e-06,
+      "loss": 0.0573,
+      "step": 3547
+    },
+    {
+      "epoch": 0.8063636363636364,
+      "grad_norm": 1.3392510628422465,
+      "learning_rate": 1.1715267038295283e-06,
+      "loss": 0.0697,
+      "step": 3548
+    },
+    {
+      "epoch": 0.8065909090909091,
+      "grad_norm": 2.2830003798380956,
+      "learning_rate": 1.1714834006163228e-06,
+      "loss": 0.0794,
+      "step": 3549
+    },
+    {
+      "epoch": 0.8068181818181818,
+      "grad_norm": 1.2787646958297776,
+      "learning_rate": 1.1714400862593675e-06,
+      "loss": 0.0477,
+      "step": 3550
+    },
+    {
+      "epoch": 0.8070454545454545,
+      "grad_norm": 2.076714407702622,
+      "learning_rate": 1.1713967607595455e-06,
+      "loss": 0.0723,
+      "step": 3551
+    },
+    {
+      "epoch": 0.8072727272727273,
+      "grad_norm": 2.715852662493627,
+      "learning_rate": 1.17135342411774e-06,
+      "loss": 0.071,
+      "step": 3552
+    },
+    {
+      "epoch": 0.8075,
+      "grad_norm": 1.0676564370624806,
+      "learning_rate": 1.1713100763348355e-06,
+      "loss": 0.0684,
+      "step": 3553
+    },
+    {
+      "epoch": 0.8077272727272727,
+      "grad_norm": 1.4315501987970651,
+      "learning_rate": 1.1712667174117154e-06,
+      "loss": 0.052,
+      "step": 3554
+    },
+    {
+      "epoch": 0.8079545454545455,
+      "grad_norm": 1.602016634893764,
+      "learning_rate": 1.1712233473492637e-06,
+      "loss": 0.1008,
+      "step": 3555
+    },
+    {
+      "epoch": 0.8081818181818182,
+      "grad_norm": 1.310479104119806,
+      "learning_rate": 1.1711799661483653e-06,
+      "loss": 0.0878,
+      "step": 3556
+    },
+    {
+      "epoch": 0.8084090909090909,
+      "grad_norm": 3.2550080159361707,
+      "learning_rate": 1.1711365738099042e-06,
+      "loss": 0.2172,
+      "step": 3557
+    },
+    {
+      "epoch": 0.8086363636363636,
+      "grad_norm": 1.3931991205037246,
+      "learning_rate": 1.1710931703347657e-06,
+      "loss": 0.0429,
+      "step": 3558
+    },
+    {
+      "epoch": 0.8088636363636363,
+      "grad_norm": 2.026693337129852,
+      "learning_rate": 1.1710497557238348e-06,
+      "loss": 0.1082,
+      "step": 3559
+    },
+    {
+      "epoch": 0.8090909090909091,
+      "grad_norm": 1.0604696753032001,
+      "learning_rate": 1.1710063299779966e-06,
+      "loss": 0.034,
+      "step": 3560
+    },
+    {
+      "epoch": 0.8093181818181818,
+      "grad_norm": 1.7436396803642356,
+      "learning_rate": 1.170962893098137e-06,
+      "loss": 0.0609,
+      "step": 3561
+    },
+    {
+      "epoch": 0.8095454545454546,
+      "grad_norm": 1.6447767136781473,
+      "learning_rate": 1.1709194450851414e-06,
+      "loss": 0.0744,
+      "step": 3562
+    },
+    {
+      "epoch": 0.8097727272727273,
+      "grad_norm": 0.9270005186312174,
+      "learning_rate": 1.1708759859398958e-06,
+      "loss": 0.0561,
+      "step": 3563
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.732913072574565,
+      "learning_rate": 1.1708325156632867e-06,
+      "loss": 0.101,
+      "step": 3564
+    },
+    {
+      "epoch": 0.8102272727272727,
+      "grad_norm": 2.3638090908853187,
+      "learning_rate": 1.1707890342562003e-06,
+      "loss": 0.0791,
+      "step": 3565
+    },
+    {
+      "epoch": 0.8104545454545454,
+      "grad_norm": 1.8612362360388737,
+      "learning_rate": 1.1707455417195231e-06,
+      "loss": 0.071,
+      "step": 3566
+    },
+    {
+      "epoch": 0.8106818181818182,
+      "grad_norm": 1.3310980714306575,
+      "learning_rate": 1.1707020380541424e-06,
+      "loss": 0.0571,
+      "step": 3567
+    },
+    {
+      "epoch": 0.8109090909090909,
+      "grad_norm": 1.9545779843372855,
+      "learning_rate": 1.1706585232609453e-06,
+      "loss": 0.0709,
+      "step": 3568
+    },
+    {
+      "epoch": 0.8111363636363637,
+      "grad_norm": 1.3143324179694822,
+      "learning_rate": 1.1706149973408187e-06,
+      "loss": 0.0539,
+      "step": 3569
+    },
+    {
+      "epoch": 0.8113636363636364,
+      "grad_norm": 1.606087923362984,
+      "learning_rate": 1.1705714602946502e-06,
+      "loss": 0.0816,
+      "step": 3570
+    },
+    {
+      "epoch": 0.8115909090909091,
+      "grad_norm": 1.078483904442067,
+      "learning_rate": 1.170527912123328e-06,
+      "loss": 0.0644,
+      "step": 3571
+    },
+    {
+      "epoch": 0.8118181818181818,
+      "grad_norm": 1.9614558292698876,
+      "learning_rate": 1.1704843528277403e-06,
+      "loss": 0.0572,
+      "step": 3572
+    },
+    {
+      "epoch": 0.8120454545454545,
+      "grad_norm": 2.379993131890751,
+      "learning_rate": 1.1704407824087747e-06,
+      "loss": 0.1139,
+      "step": 3573
+    },
+    {
+      "epoch": 0.8122727272727273,
+      "grad_norm": 1.5632938079630414,
+      "learning_rate": 1.17039720086732e-06,
+      "loss": 0.0812,
+      "step": 3574
+    },
+    {
+      "epoch": 0.8125,
+      "grad_norm": 1.3610646083548785,
+      "learning_rate": 1.1703536082042649e-06,
+      "loss": 0.0705,
+      "step": 3575
+    },
+    {
+      "epoch": 0.8127272727272727,
+      "grad_norm": 2.297614235450191,
+      "learning_rate": 1.1703100044204984e-06,
+      "loss": 0.1606,
+      "step": 3576
+    },
+    {
+      "epoch": 0.8129545454545455,
+      "grad_norm": 1.6941771319929875,
+      "learning_rate": 1.1702663895169095e-06,
+      "loss": 0.0542,
+      "step": 3577
+    },
+    {
+      "epoch": 0.8131818181818182,
+      "grad_norm": 1.747208370010215,
+      "learning_rate": 1.1702227634943874e-06,
+      "loss": 0.0633,
+      "step": 3578
+    },
+    {
+      "epoch": 0.8134090909090909,
+      "grad_norm": 2.245779448707662,
+      "learning_rate": 1.1701791263538223e-06,
+      "loss": 0.0885,
+      "step": 3579
+    },
+    {
+      "epoch": 0.8136363636363636,
+      "grad_norm": 2.482639519934004,
+      "learning_rate": 1.1701354780961035e-06,
+      "loss": 0.1613,
+      "step": 3580
+    },
+    {
+      "epoch": 0.8138636363636363,
+      "grad_norm": 1.4282655648151736,
+      "learning_rate": 1.1700918187221214e-06,
+      "loss": 0.0885,
+      "step": 3581
+    },
+    {
+      "epoch": 0.8140909090909091,
+      "grad_norm": 1.69896432083361,
+      "learning_rate": 1.1700481482327661e-06,
+      "loss": 0.066,
+      "step": 3582
+    },
+    {
+      "epoch": 0.8143181818181818,
+      "grad_norm": 1.9595861791222973,
+      "learning_rate": 1.1700044666289282e-06,
+      "loss": 0.1242,
+      "step": 3583
+    },
+    {
+      "epoch": 0.8145454545454546,
+      "grad_norm": 1.6689659457702426,
+      "learning_rate": 1.1699607739114982e-06,
+      "loss": 0.0724,
+      "step": 3584
+    },
+    {
+      "epoch": 0.8147727272727273,
+      "grad_norm": 1.5645169005603807,
+      "learning_rate": 1.1699170700813676e-06,
+      "loss": 0.0754,
+      "step": 3585
+    },
+    {
+      "epoch": 0.815,
+      "grad_norm": 3.5443787781412706,
+      "learning_rate": 1.169873355139427e-06,
+      "loss": 0.1829,
+      "step": 3586
+    },
+    {
+      "epoch": 0.8152272727272727,
+      "grad_norm": 1.7030758710249223,
+      "learning_rate": 1.1698296290865683e-06,
+      "loss": 0.0604,
+      "step": 3587
+    },
+    {
+      "epoch": 0.8154545454545454,
+      "grad_norm": 1.6339689190919238,
+      "learning_rate": 1.1697858919236828e-06,
+      "loss": 0.0819,
+      "step": 3588
+    },
+    {
+      "epoch": 0.8156818181818182,
+      "grad_norm": 3.145099295958885,
+      "learning_rate": 1.1697421436516626e-06,
+      "loss": 0.2269,
+      "step": 3589
+    },
+    {
+      "epoch": 0.8159090909090909,
+      "grad_norm": 2.568864961326701,
+      "learning_rate": 1.1696983842713996e-06,
+      "loss": 0.0878,
+      "step": 3590
+    },
+    {
+      "epoch": 0.8161363636363637,
+      "grad_norm": 1.4639334080996294,
+      "learning_rate": 1.1696546137837865e-06,
+      "loss": 0.0785,
+      "step": 3591
+    },
+    {
+      "epoch": 0.8163636363636364,
+      "grad_norm": 1.0913945617146397,
+      "learning_rate": 1.1696108321897156e-06,
+      "loss": 0.0556,
+      "step": 3592
+    },
+    {
+      "epoch": 0.8165909090909091,
+      "grad_norm": 1.4704722576426374,
+      "learning_rate": 1.1695670394900797e-06,
+      "loss": 0.0661,
+      "step": 3593
+    },
+    {
+      "epoch": 0.8168181818181818,
+      "grad_norm": 1.740123048203194,
+      "learning_rate": 1.1695232356857717e-06,
+      "loss": 0.0439,
+      "step": 3594
+    },
+    {
+      "epoch": 0.8170454545454545,
+      "grad_norm": 1.4625421676025163,
+      "learning_rate": 1.169479420777685e-06,
+      "loss": 0.1039,
+      "step": 3595
+    },
+    {
+      "epoch": 0.8172727272727273,
+      "grad_norm": 1.3461633617316193,
+      "learning_rate": 1.169435594766713e-06,
+      "loss": 0.0674,
+      "step": 3596
+    },
+    {
+      "epoch": 0.8175,
+      "grad_norm": 2.336494958434851,
+      "learning_rate": 1.1693917576537494e-06,
+      "loss": 0.1165,
+      "step": 3597
+    },
+    {
+      "epoch": 0.8177272727272727,
+      "grad_norm": 2.0368623141463273,
+      "learning_rate": 1.169347909439688e-06,
+      "loss": 0.0782,
+      "step": 3598
+    },
+    {
+      "epoch": 0.8179545454545455,
+      "grad_norm": 2.7208696406844486,
+      "learning_rate": 1.1693040501254234e-06,
+      "loss": 0.1603,
+      "step": 3599
+    },
+    {
+      "epoch": 0.8181818181818182,
+      "grad_norm": 1.2999185494402925,
+      "learning_rate": 1.1692601797118495e-06,
+      "loss": 0.0585,
+      "step": 3600
+    },
+    {
+      "epoch": 0.8184090909090909,
+      "grad_norm": 1.9222549527917865,
+      "learning_rate": 1.1692162981998608e-06,
+      "loss": 0.0691,
+      "step": 3601
+    },
+    {
+      "epoch": 0.8186363636363636,
+      "grad_norm": 1.330561932322522,
+      "learning_rate": 1.1691724055903524e-06,
+      "loss": 0.071,
+      "step": 3602
+    },
+    {
+      "epoch": 0.8188636363636363,
+      "grad_norm": 1.358178659962048,
+      "learning_rate": 1.1691285018842195e-06,
+      "loss": 0.0567,
+      "step": 3603
+    },
+    {
+      "epoch": 0.8190909090909091,
+      "grad_norm": 1.8183031578891857,
+      "learning_rate": 1.169084587082357e-06,
+      "loss": 0.0929,
+      "step": 3604
+    },
+    {
+      "epoch": 0.8193181818181818,
+      "grad_norm": 1.3016324212003028,
+      "learning_rate": 1.1690406611856607e-06,
+      "loss": 0.0663,
+      "step": 3605
+    },
+    {
+      "epoch": 0.8195454545454546,
+      "grad_norm": 2.380838460176052,
+      "learning_rate": 1.1689967241950263e-06,
+      "loss": 0.0948,
+      "step": 3606
+    },
+    {
+      "epoch": 0.8197727272727273,
+      "grad_norm": 1.3284079795559904,
+      "learning_rate": 1.1689527761113492e-06,
+      "loss": 0.0499,
+      "step": 3607
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9688054841661031,
+      "learning_rate": 1.1689088169355262e-06,
+      "loss": 0.1248,
+      "step": 3608
+    },
+    {
+      "epoch": 0.8202272727272727,
+      "grad_norm": 2.4637020862061916,
+      "learning_rate": 1.1688648466684537e-06,
+      "loss": 0.0771,
+      "step": 3609
+    },
+    {
+      "epoch": 0.8204545454545454,
+      "grad_norm": 1.6649235272402232,
+      "learning_rate": 1.168820865311028e-06,
+      "loss": 0.0711,
+      "step": 3610
+    },
+    {
+      "epoch": 0.8206818181818182,
+      "grad_norm": 1.747797656932806,
+      "learning_rate": 1.168776872864146e-06,
+      "loss": 0.0687,
+      "step": 3611
+    },
+    {
+      "epoch": 0.8209090909090909,
+      "grad_norm": 1.7483980875556524,
+      "learning_rate": 1.1687328693287052e-06,
+      "loss": 0.0767,
+      "step": 3612
+    },
+    {
+      "epoch": 0.8211363636363637,
+      "grad_norm": 1.7196514080367133,
+      "learning_rate": 1.1686888547056025e-06,
+      "loss": 0.0561,
+      "step": 3613
+    },
+    {
+      "epoch": 0.8213636363636364,
+      "grad_norm": 2.1893324539474537,
+      "learning_rate": 1.1686448289957352e-06,
+      "loss": 0.1272,
+      "step": 3614
+    },
+    {
+      "epoch": 0.821590909090909,
+      "grad_norm": 1.565043419435565,
+      "learning_rate": 1.1686007922000017e-06,
+      "loss": 0.0821,
+      "step": 3615
+    },
+    {
+      "epoch": 0.8218181818181818,
+      "grad_norm": 2.470456574722027,
+      "learning_rate": 1.1685567443192996e-06,
+      "loss": 0.0954,
+      "step": 3616
+    },
+    {
+      "epoch": 0.8220454545454545,
+      "grad_norm": 2.99504415245268,
+      "learning_rate": 1.1685126853545272e-06,
+      "loss": 0.1471,
+      "step": 3617
+    },
+    {
+      "epoch": 0.8222727272727273,
+      "grad_norm": 1.3889627744556392,
+      "learning_rate": 1.1684686153065832e-06,
+      "loss": 0.0597,
+      "step": 3618
+    },
+    {
+      "epoch": 0.8225,
+      "grad_norm": 1.7650673575772635,
+      "learning_rate": 1.1684245341763658e-06,
+      "loss": 0.0798,
+      "step": 3619
+    },
+    {
+      "epoch": 0.8227272727272728,
+      "grad_norm": 1.1506833684466944,
+      "learning_rate": 1.168380441964774e-06,
+      "loss": 0.052,
+      "step": 3620
+    },
+    {
+      "epoch": 0.8229545454545455,
+      "grad_norm": 1.671578551275523,
+      "learning_rate": 1.168336338672707e-06,
+      "loss": 0.0641,
+      "step": 3621
+    },
+    {
+      "epoch": 0.8231818181818182,
+      "grad_norm": 2.109786700545356,
+      "learning_rate": 1.1682922243010644e-06,
+      "loss": 0.0683,
+      "step": 3622
+    },
+    {
+      "epoch": 0.8234090909090909,
+      "grad_norm": 1.67364813487367,
+      "learning_rate": 1.1682480988507452e-06,
+      "loss": 0.0714,
+      "step": 3623
+    },
+    {
+      "epoch": 0.8236363636363636,
+      "grad_norm": 1.5360287363164207,
+      "learning_rate": 1.1682039623226497e-06,
+      "loss": 0.0495,
+      "step": 3624
+    },
+    {
+      "epoch": 0.8238636363636364,
+      "grad_norm": 1.4527582932036625,
+      "learning_rate": 1.1681598147176777e-06,
+      "loss": 0.0579,
+      "step": 3625
+    },
+    {
+      "epoch": 0.8240909090909091,
+      "grad_norm": 2.0652286600034158,
+      "learning_rate": 1.1681156560367296e-06,
+      "loss": 0.0875,
+      "step": 3626
+    },
+    {
+      "epoch": 0.8243181818181818,
+      "grad_norm": 1.3645674659724296,
+      "learning_rate": 1.1680714862807054e-06,
+      "loss": 0.0359,
+      "step": 3627
+    },
+    {
+      "epoch": 0.8245454545454546,
+      "grad_norm": 2.8565385426686105,
+      "learning_rate": 1.1680273054505065e-06,
+      "loss": 0.1331,
+      "step": 3628
+    },
+    {
+      "epoch": 0.8247727272727273,
+      "grad_norm": 1.6252782963289103,
+      "learning_rate": 1.1679831135470335e-06,
+      "loss": 0.1039,
+      "step": 3629
+    },
+    {
+      "epoch": 0.825,
+      "grad_norm": 1.8439075607134685,
+      "learning_rate": 1.1679389105711873e-06,
+      "loss": 0.0644,
+      "step": 3630
+    },
+    {
+      "epoch": 0.8252272727272727,
+      "grad_norm": 1.0291839117634192,
+      "learning_rate": 1.1678946965238697e-06,
+      "loss": 0.0489,
+      "step": 3631
+    },
+    {
+      "epoch": 0.8254545454545454,
+      "grad_norm": 2.273889668788591,
+      "learning_rate": 1.167850471405982e-06,
+      "loss": 0.1611,
+      "step": 3632
+    },
+    {
+      "epoch": 0.8256818181818182,
+      "grad_norm": 2.853208064394543,
+      "learning_rate": 1.167806235218426e-06,
+      "loss": 0.0912,
+      "step": 3633
+    },
+    {
+      "epoch": 0.8259090909090909,
+      "grad_norm": 1.128829753895101,
+      "learning_rate": 1.1677619879621042e-06,
+      "loss": 0.0325,
+      "step": 3634
+    },
+    {
+      "epoch": 0.8261363636363637,
+      "grad_norm": 2.4013559072687554,
+      "learning_rate": 1.1677177296379182e-06,
+      "loss": 0.1022,
+      "step": 3635
+    },
+    {
+      "epoch": 0.8263636363636364,
+      "grad_norm": 2.1500789984987834,
+      "learning_rate": 1.1676734602467713e-06,
+      "loss": 0.0949,
+      "step": 3636
+    },
+    {
+      "epoch": 0.826590909090909,
+      "grad_norm": 1.5174421941221496,
+      "learning_rate": 1.1676291797895655e-06,
+      "loss": 0.0852,
+      "step": 3637
+    },
+    {
+      "epoch": 0.8268181818181818,
+      "grad_norm": 2.5157314068227823,
+      "learning_rate": 1.1675848882672042e-06,
+      "loss": 0.0751,
+      "step": 3638
+    },
+    {
+      "epoch": 0.8270454545454545,
+      "grad_norm": 1.382477907592964,
+      "learning_rate": 1.1675405856805901e-06,
+      "loss": 0.0555,
+      "step": 3639
+    },
+    {
+      "epoch": 0.8272727272727273,
+      "grad_norm": 1.6850809449589774,
+      "learning_rate": 1.1674962720306273e-06,
+      "loss": 0.0622,
+      "step": 3640
+    },
+    {
+      "epoch": 0.8275,
+      "grad_norm": 1.6323955245410309,
+      "learning_rate": 1.1674519473182192e-06,
+      "loss": 0.0883,
+      "step": 3641
+    },
+    {
+      "epoch": 0.8277272727272728,
+      "grad_norm": 1.8406010775571398,
+      "learning_rate": 1.1674076115442693e-06,
+      "loss": 0.064,
+      "step": 3642
+    },
+    {
+      "epoch": 0.8279545454545455,
+      "grad_norm": 2.174815555849599,
+      "learning_rate": 1.1673632647096819e-06,
+      "loss": 0.1285,
+      "step": 3643
+    },
+    {
+      "epoch": 0.8281818181818181,
+      "grad_norm": 4.2001996646217385,
+      "learning_rate": 1.1673189068153611e-06,
+      "loss": 0.1831,
+      "step": 3644
+    },
+    {
+      "epoch": 0.8284090909090909,
+      "grad_norm": 1.5268671033202517,
+      "learning_rate": 1.167274537862212e-06,
+      "loss": 0.0522,
+      "step": 3645
+    },
+    {
+      "epoch": 0.8286363636363636,
+      "grad_norm": 1.4747397504708597,
+      "learning_rate": 1.1672301578511392e-06,
+      "loss": 0.0633,
+      "step": 3646
+    },
+    {
+      "epoch": 0.8288636363636364,
+      "grad_norm": 1.0590282503545403,
+      "learning_rate": 1.167185766783047e-06,
+      "loss": 0.0281,
+      "step": 3647
+    },
+    {
+      "epoch": 0.8290909090909091,
+      "grad_norm": 1.9026730835985461,
+      "learning_rate": 1.1671413646588414e-06,
+      "loss": 0.0679,
+      "step": 3648
+    },
+    {
+      "epoch": 0.8293181818181818,
+      "grad_norm": 2.1128238942195914,
+      "learning_rate": 1.1670969514794277e-06,
+      "loss": 0.1023,
+      "step": 3649
+    },
+    {
+      "epoch": 0.8295454545454546,
+      "grad_norm": 1.4581548976473662,
+      "learning_rate": 1.1670525272457112e-06,
+      "loss": 0.077,
+      "step": 3650
+    },
+    {
+      "epoch": 0.8297727272727272,
+      "grad_norm": 3.9363734412192963,
+      "learning_rate": 1.167008091958598e-06,
+      "loss": 0.0713,
+      "step": 3651
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4448725413230668,
+      "learning_rate": 1.1669636456189943e-06,
+      "loss": 0.1022,
+      "step": 3652
+    },
+    {
+      "epoch": 0.8302272727272727,
+      "grad_norm": 1.8180463069462873,
+      "learning_rate": 1.1669191882278063e-06,
+      "loss": 0.0862,
+      "step": 3653
+    },
+    {
+      "epoch": 0.8304545454545454,
+      "grad_norm": 1.097051486840231,
+      "learning_rate": 1.1668747197859406e-06,
+      "loss": 0.0509,
+      "step": 3654
+    },
+    {
+      "epoch": 0.8306818181818182,
+      "grad_norm": 3.4142737812858988,
+      "learning_rate": 1.1668302402943041e-06,
+      "loss": 0.2062,
+      "step": 3655
+    },
+    {
+      "epoch": 0.8309090909090909,
+      "grad_norm": 2.0868602086793198,
+      "learning_rate": 1.1667857497538037e-06,
+      "loss": 0.0682,
+      "step": 3656
+    },
+    {
+      "epoch": 0.8311363636363637,
+      "grad_norm": 3.015540953274811,
+      "learning_rate": 1.1667412481653467e-06,
+      "loss": 0.1139,
+      "step": 3657
+    },
+    {
+      "epoch": 0.8313636363636364,
+      "grad_norm": 2.28777189759415,
+      "learning_rate": 1.1666967355298406e-06,
+      "loss": 0.0883,
+      "step": 3658
+    },
+    {
+      "epoch": 0.831590909090909,
+      "grad_norm": 0.9540412948149161,
+      "learning_rate": 1.166652211848193e-06,
+      "loss": 0.0314,
+      "step": 3659
+    },
+    {
+      "epoch": 0.8318181818181818,
+      "grad_norm": 1.8432475210126984,
+      "learning_rate": 1.1666076771213116e-06,
+      "loss": 0.079,
+      "step": 3660
+    },
+    {
+      "epoch": 0.8320454545454545,
+      "grad_norm": 1.5000310294131094,
+      "learning_rate": 1.166563131350105e-06,
+      "loss": 0.0663,
+      "step": 3661
+    },
+    {
+      "epoch": 0.8322727272727273,
+      "grad_norm": 2.78554964934304,
+      "learning_rate": 1.1665185745354813e-06,
+      "loss": 0.1892,
+      "step": 3662
+    },
+    {
+      "epoch": 0.8325,
+      "grad_norm": 2.15714319354658,
+      "learning_rate": 1.166474006678349e-06,
+      "loss": 0.1331,
+      "step": 3663
+    },
+    {
+      "epoch": 0.8327272727272728,
+      "grad_norm": 1.3664857834067135,
+      "learning_rate": 1.1664294277796173e-06,
+      "loss": 0.0651,
+      "step": 3664
+    },
+    {
+      "epoch": 0.8329545454545455,
+      "grad_norm": 0.9251647686747521,
+      "learning_rate": 1.1663848378401948e-06,
+      "loss": 0.0327,
+      "step": 3665
+    },
+    {
+      "epoch": 0.8331818181818181,
+      "grad_norm": 2.0289974442281067,
+      "learning_rate": 1.166340236860991e-06,
+      "loss": 0.0796,
+      "step": 3666
+    },
+    {
+      "epoch": 0.8334090909090909,
+      "grad_norm": 2.7297421081683453,
+      "learning_rate": 1.1662956248429155e-06,
+      "loss": 0.1256,
+      "step": 3667
+    },
+    {
+      "epoch": 0.8336363636363636,
+      "grad_norm": 1.7609987293024179,
+      "learning_rate": 1.1662510017868777e-06,
+      "loss": 0.0947,
+      "step": 3668
+    },
+    {
+      "epoch": 0.8338636363636364,
+      "grad_norm": 1.0988774382826345,
+      "learning_rate": 1.1662063676937876e-06,
+      "loss": 0.0585,
+      "step": 3669
+    },
+    {
+      "epoch": 0.8340909090909091,
+      "grad_norm": 1.1204087898176016,
+      "learning_rate": 1.1661617225645557e-06,
+      "loss": 0.0598,
+      "step": 3670
+    },
+    {
+      "epoch": 0.8343181818181818,
+      "grad_norm": 1.8875956219116163,
+      "learning_rate": 1.166117066400092e-06,
+      "loss": 0.0892,
+      "step": 3671
+    },
+    {
+      "epoch": 0.8345454545454546,
+      "grad_norm": 1.9329057183818676,
+      "learning_rate": 1.1660723992013074e-06,
+      "loss": 0.0572,
+      "step": 3672
+    },
+    {
+      "epoch": 0.8347727272727272,
+      "grad_norm": 1.5489744523205755,
+      "learning_rate": 1.1660277209691125e-06,
+      "loss": 0.0858,
+      "step": 3673
+    },
+    {
+      "epoch": 0.835,
+      "grad_norm": 2.391675747536509,
+      "learning_rate": 1.1659830317044188e-06,
+      "loss": 0.1164,
+      "step": 3674
+    },
+    {
+      "epoch": 0.8352272727272727,
+      "grad_norm": 1.530411485676199,
+      "learning_rate": 1.165938331408137e-06,
+      "loss": 0.0554,
+      "step": 3675
+    },
+    {
+      "epoch": 0.8354545454545454,
+      "grad_norm": 1.9758789306408389,
+      "learning_rate": 1.1658936200811789e-06,
+      "loss": 0.0603,
+      "step": 3676
+    },
+    {
+      "epoch": 0.8356818181818182,
+      "grad_norm": 2.0603582400860194,
+      "learning_rate": 1.1658488977244564e-06,
+      "loss": 0.086,
+      "step": 3677
+    },
+    {
+      "epoch": 0.8359090909090909,
+      "grad_norm": 1.5468977787792555,
+      "learning_rate": 1.1658041643388811e-06,
+      "loss": 0.0822,
+      "step": 3678
+    },
+    {
+      "epoch": 0.8361363636363637,
+      "grad_norm": 1.5211534135712619,
+      "learning_rate": 1.1657594199253657e-06,
+      "loss": 0.1049,
+      "step": 3679
+    },
+    {
+      "epoch": 0.8363636363636363,
+      "grad_norm": 1.4745458139570355,
+      "learning_rate": 1.1657146644848221e-06,
+      "loss": 0.0559,
+      "step": 3680
+    },
+    {
+      "epoch": 0.836590909090909,
+      "grad_norm": 1.7790056135780778,
+      "learning_rate": 1.1656698980181633e-06,
+      "loss": 0.119,
+      "step": 3681
+    },
+    {
+      "epoch": 0.8368181818181818,
+      "grad_norm": 2.582513854594154,
+      "learning_rate": 1.165625120526302e-06,
+      "loss": 0.1206,
+      "step": 3682
+    },
+    {
+      "epoch": 0.8370454545454545,
+      "grad_norm": 1.2543480594406957,
+      "learning_rate": 1.1655803320101512e-06,
+      "loss": 0.0403,
+      "step": 3683
+    },
+    {
+      "epoch": 0.8372727272727273,
+      "grad_norm": 1.5992608199142426,
+      "learning_rate": 1.1655355324706245e-06,
+      "loss": 0.0881,
+      "step": 3684
+    },
+    {
+      "epoch": 0.8375,
+      "grad_norm": 2.435838991627666,
+      "learning_rate": 1.1654907219086351e-06,
+      "loss": 0.0742,
+      "step": 3685
+    },
+    {
+      "epoch": 0.8377272727272728,
+      "grad_norm": 1.4013126706679857,
+      "learning_rate": 1.1654459003250971e-06,
+      "loss": 0.0494,
+      "step": 3686
+    },
+    {
+      "epoch": 0.8379545454545455,
+      "grad_norm": 1.0530913075434298,
+      "learning_rate": 1.1654010677209242e-06,
+      "loss": 0.0504,
+      "step": 3687
+    },
+    {
+      "epoch": 0.8381818181818181,
+      "grad_norm": 0.9705634631714875,
+      "learning_rate": 1.1653562240970307e-06,
+      "loss": 0.0549,
+      "step": 3688
+    },
+    {
+      "epoch": 0.8384090909090909,
+      "grad_norm": 1.0434261636708255,
+      "learning_rate": 1.1653113694543315e-06,
+      "loss": 0.0323,
+      "step": 3689
+    },
+    {
+      "epoch": 0.8386363636363636,
+      "grad_norm": 2.84087868678207,
+      "learning_rate": 1.1652665037937404e-06,
+      "loss": 0.1471,
+      "step": 3690
+    },
+    {
+      "epoch": 0.8388636363636364,
+      "grad_norm": 3.7201009638386777,
+      "learning_rate": 1.1652216271161728e-06,
+      "loss": 0.2188,
+      "step": 3691
+    },
+    {
+      "epoch": 0.8390909090909091,
+      "grad_norm": 1.206371557529756,
+      "learning_rate": 1.1651767394225437e-06,
+      "loss": 0.0435,
+      "step": 3692
+    },
+    {
+      "epoch": 0.8393181818181819,
+      "grad_norm": 1.7475542860112911,
+      "learning_rate": 1.1651318407137688e-06,
+      "loss": 0.101,
+      "step": 3693
+    },
+    {
+      "epoch": 0.8395454545454546,
+      "grad_norm": 1.3844976250862548,
+      "learning_rate": 1.1650869309907633e-06,
+      "loss": 0.052,
+      "step": 3694
+    },
+    {
+      "epoch": 0.8397727272727272,
+      "grad_norm": 2.47917199626916,
+      "learning_rate": 1.165042010254443e-06,
+      "loss": 0.0955,
+      "step": 3695
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.8310882591723177,
+      "learning_rate": 1.1649970785057238e-06,
+      "loss": 0.1474,
+      "step": 3696
+    },
+    {
+      "epoch": 0.8402272727272727,
+      "grad_norm": 1.7784701207113174,
+      "learning_rate": 1.1649521357455222e-06,
+      "loss": 0.0691,
+      "step": 3697
+    },
+    {
+      "epoch": 0.8404545454545455,
+      "grad_norm": 2.3706424498170358,
+      "learning_rate": 1.1649071819747545e-06,
+      "loss": 0.0849,
+      "step": 3698
+    },
+    {
+      "epoch": 0.8406818181818182,
+      "grad_norm": 2.091859752392473,
+      "learning_rate": 1.1648622171943374e-06,
+      "loss": 0.0895,
+      "step": 3699
+    },
+    {
+      "epoch": 0.8409090909090909,
+      "grad_norm": 2.5844758847564653,
+      "learning_rate": 1.1648172414051882e-06,
+      "loss": 0.1456,
+      "step": 3700
+    },
+    {
+      "epoch": 0.8411363636363637,
+      "grad_norm": 1.4378995767718907,
+      "learning_rate": 1.1647722546082232e-06,
+      "loss": 0.052,
+      "step": 3701
+    },
+    {
+      "epoch": 0.8413636363636363,
+      "grad_norm": 0.9955401500772467,
+      "learning_rate": 1.1647272568043605e-06,
+      "loss": 0.0426,
+      "step": 3702
+    },
+    {
+      "epoch": 0.8415909090909091,
+      "grad_norm": 2.885918432484691,
+      "learning_rate": 1.1646822479945174e-06,
+      "loss": 0.1055,
+      "step": 3703
+    },
+    {
+      "epoch": 0.8418181818181818,
+      "grad_norm": 1.6649279385819231,
+      "learning_rate": 1.1646372281796116e-06,
+      "loss": 0.0796,
+      "step": 3704
+    },
+    {
+      "epoch": 0.8420454545454545,
+      "grad_norm": 1.8732797752466608,
+      "learning_rate": 1.1645921973605616e-06,
+      "loss": 0.0813,
+      "step": 3705
+    },
+    {
+      "epoch": 0.8422727272727273,
+      "grad_norm": 1.9616669644042317,
+      "learning_rate": 1.164547155538285e-06,
+      "loss": 0.0884,
+      "step": 3706
+    },
+    {
+      "epoch": 0.8425,
+      "grad_norm": 0.6052210231641323,
+      "learning_rate": 1.1645021027137007e-06,
+      "loss": 0.0389,
+      "step": 3707
+    },
+    {
+      "epoch": 0.8427272727272728,
+      "grad_norm": 2.0086691082031822,
+      "learning_rate": 1.1644570388877274e-06,
+      "loss": 0.0977,
+      "step": 3708
+    },
+    {
+      "epoch": 0.8429545454545454,
+      "grad_norm": 2.3046531646127697,
+      "learning_rate": 1.164411964061284e-06,
+      "loss": 0.174,
+      "step": 3709
+    },
+    {
+      "epoch": 0.8431818181818181,
+      "grad_norm": 2.172343933018854,
+      "learning_rate": 1.1643668782352894e-06,
+      "loss": 0.0838,
+      "step": 3710
+    },
+    {
+      "epoch": 0.8434090909090909,
+      "grad_norm": 2.150221645458971,
+      "learning_rate": 1.1643217814106633e-06,
+      "loss": 0.0623,
+      "step": 3711
+    },
+    {
+      "epoch": 0.8436363636363636,
+      "grad_norm": 1.0025191330686263,
+      "learning_rate": 1.1642766735883252e-06,
+      "loss": 0.0521,
+      "step": 3712
+    },
+    {
+      "epoch": 0.8438636363636364,
+      "grad_norm": 1.6021490205065236,
+      "learning_rate": 1.1642315547691948e-06,
+      "loss": 0.0798,
+      "step": 3713
+    },
+    {
+      "epoch": 0.8440909090909091,
+      "grad_norm": 2.108196658613105,
+      "learning_rate": 1.1641864249541922e-06,
+      "loss": 0.1252,
+      "step": 3714
+    },
+    {
+      "epoch": 0.8443181818181819,
+      "grad_norm": 2.221078874169877,
+      "learning_rate": 1.164141284144238e-06,
+      "loss": 0.0787,
+      "step": 3715
+    },
+    {
+      "epoch": 0.8445454545454546,
+      "grad_norm": 0.9796909451611757,
+      "learning_rate": 1.1640961323402522e-06,
+      "loss": 0.0364,
+      "step": 3716
+    },
+    {
+      "epoch": 0.8447727272727272,
+      "grad_norm": 1.9856165418250564,
+      "learning_rate": 1.1640509695431558e-06,
+      "loss": 0.0698,
+      "step": 3717
+    },
+    {
+      "epoch": 0.845,
+      "grad_norm": 4.581886054459653,
+      "learning_rate": 1.1640057957538696e-06,
+      "loss": 0.0993,
+      "step": 3718
+    },
+    {
+      "epoch": 0.8452272727272727,
+      "grad_norm": 1.7746254795068783,
+      "learning_rate": 1.163960610973315e-06,
+      "loss": 0.1318,
+      "step": 3719
+    },
+    {
+      "epoch": 0.8454545454545455,
+      "grad_norm": 1.912924803687727,
+      "learning_rate": 1.1639154152024132e-06,
+      "loss": 0.1,
+      "step": 3720
+    },
+    {
+      "epoch": 0.8456818181818182,
+      "grad_norm": 1.6378503454212314,
+      "learning_rate": 1.163870208442086e-06,
+      "loss": 0.0923,
+      "step": 3721
+    },
+    {
+      "epoch": 0.8459090909090909,
+      "grad_norm": 2.3185840096026133,
+      "learning_rate": 1.1638249906932552e-06,
+      "loss": 0.0822,
+      "step": 3722
+    },
+    {
+      "epoch": 0.8461363636363637,
+      "grad_norm": 1.6806695562492917,
+      "learning_rate": 1.1637797619568429e-06,
+      "loss": 0.0573,
+      "step": 3723
+    },
+    {
+      "epoch": 0.8463636363636363,
+      "grad_norm": 0.9760494681670611,
+      "learning_rate": 1.163734522233771e-06,
+      "loss": 0.0298,
+      "step": 3724
+    },
+    {
+      "epoch": 0.8465909090909091,
+      "grad_norm": 3.0313516516436105,
+      "learning_rate": 1.1636892715249623e-06,
+      "loss": 0.1046,
+      "step": 3725
+    },
+    {
+      "epoch": 0.8468181818181818,
+      "grad_norm": 2.7308280146542745,
+      "learning_rate": 1.1636440098313398e-06,
+      "loss": 0.0845,
+      "step": 3726
+    },
+    {
+      "epoch": 0.8470454545454545,
+      "grad_norm": 1.9005049494850415,
+      "learning_rate": 1.1635987371538262e-06,
+      "loss": 0.1185,
+      "step": 3727
+    },
+    {
+      "epoch": 0.8472727272727273,
+      "grad_norm": 2.2901196579011254,
+      "learning_rate": 1.1635534534933446e-06,
+      "loss": 0.1382,
+      "step": 3728
+    },
+    {
+      "epoch": 0.8475,
+      "grad_norm": 1.8615931280857345,
+      "learning_rate": 1.1635081588508185e-06,
+      "loss": 0.1008,
+      "step": 3729
+    },
+    {
+      "epoch": 0.8477272727272728,
+      "grad_norm": 1.8603516604381467,
+      "learning_rate": 1.1634628532271717e-06,
+      "loss": 0.12,
+      "step": 3730
+    },
+    {
+      "epoch": 0.8479545454545454,
+      "grad_norm": 2.321514658007446,
+      "learning_rate": 1.1634175366233278e-06,
+      "loss": 0.1143,
+      "step": 3731
+    },
+    {
+      "epoch": 0.8481818181818181,
+      "grad_norm": 2.693419999388435,
+      "learning_rate": 1.163372209040211e-06,
+      "loss": 0.1064,
+      "step": 3732
+    },
+    {
+      "epoch": 0.8484090909090909,
+      "grad_norm": 1.7849846938062608,
+      "learning_rate": 1.1633268704787456e-06,
+      "loss": 0.0739,
+      "step": 3733
+    },
+    {
+      "epoch": 0.8486363636363636,
+      "grad_norm": 1.6441965525694489,
+      "learning_rate": 1.1632815209398561e-06,
+      "loss": 0.0487,
+      "step": 3734
+    },
+    {
+      "epoch": 0.8488636363636364,
+      "grad_norm": 2.0178159246393292,
+      "learning_rate": 1.1632361604244677e-06,
+      "loss": 0.1302,
+      "step": 3735
+    },
+    {
+      "epoch": 0.8490909090909091,
+      "grad_norm": 1.815811160208809,
+      "learning_rate": 1.1631907889335046e-06,
+      "loss": 0.0706,
+      "step": 3736
+    },
+    {
+      "epoch": 0.8493181818181819,
+      "grad_norm": 0.8855409610796527,
+      "learning_rate": 1.1631454064678924e-06,
+      "loss": 0.0439,
+      "step": 3737
+    },
+    {
+      "epoch": 0.8495454545454545,
+      "grad_norm": 2.4705826647517966,
+      "learning_rate": 1.1631000130285565e-06,
+      "loss": 0.1213,
+      "step": 3738
+    },
+    {
+      "epoch": 0.8497727272727272,
+      "grad_norm": 2.1269474726791766,
+      "learning_rate": 1.163054608616423e-06,
+      "loss": 0.0549,
+      "step": 3739
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1561246805441057,
+      "learning_rate": 1.1630091932324169e-06,
+      "loss": 0.1141,
+      "step": 3740
+    },
+    {
+      "epoch": 0.8502272727272727,
+      "grad_norm": 2.2201489952374884,
+      "learning_rate": 1.162963766877465e-06,
+      "loss": 0.1192,
+      "step": 3741
+    },
+    {
+      "epoch": 0.8504545454545455,
+      "grad_norm": 2.6687633031085984,
+      "learning_rate": 1.1629183295524931e-06,
+      "loss": 0.1271,
+      "step": 3742
+    },
+    {
+      "epoch": 0.8506818181818182,
+      "grad_norm": 4.02004698837629,
+      "learning_rate": 1.1628728812584282e-06,
+      "loss": 0.1586,
+      "step": 3743
+    },
+    {
+      "epoch": 0.850909090909091,
+      "grad_norm": 2.5638353133544567,
+      "learning_rate": 1.1628274219961967e-06,
+      "loss": 0.1247,
+      "step": 3744
+    },
+    {
+      "epoch": 0.8511363636363637,
+      "grad_norm": 1.3406007767590298,
+      "learning_rate": 1.162781951766726e-06,
+      "loss": 0.0633,
+      "step": 3745
+    },
+    {
+      "epoch": 0.8513636363636363,
+      "grad_norm": 1.707668089822094,
+      "learning_rate": 1.1627364705709428e-06,
+      "loss": 0.0545,
+      "step": 3746
+    },
+    {
+      "epoch": 0.8515909090909091,
+      "grad_norm": 1.3391828404106119,
+      "learning_rate": 1.162690978409775e-06,
+      "loss": 0.0463,
+      "step": 3747
+    },
+    {
+      "epoch": 0.8518181818181818,
+      "grad_norm": 2.3464088651372315,
+      "learning_rate": 1.1626454752841501e-06,
+      "loss": 0.0985,
+      "step": 3748
+    },
+    {
+      "epoch": 0.8520454545454546,
+      "grad_norm": 1.5782442819844713,
+      "learning_rate": 1.1625999611949962e-06,
+      "loss": 0.1044,
+      "step": 3749
+    },
+    {
+      "epoch": 0.8522727272727273,
+      "grad_norm": 1.7075830528317257,
+      "learning_rate": 1.1625544361432409e-06,
+      "loss": 0.0735,
+      "step": 3750
+    },
+    {
+      "epoch": 0.8525,
+      "grad_norm": 1.5092829779211863,
+      "learning_rate": 1.1625089001298129e-06,
+      "loss": 0.0686,
+      "step": 3751
+    },
+    {
+      "epoch": 0.8527272727272728,
+      "grad_norm": 0.7055849252400382,
+      "learning_rate": 1.1624633531556409e-06,
+      "loss": 0.0357,
+      "step": 3752
+    },
+    {
+      "epoch": 0.8529545454545454,
+      "grad_norm": 1.563446160250751,
+      "learning_rate": 1.162417795221653e-06,
+      "loss": 0.0672,
+      "step": 3753
+    },
+    {
+      "epoch": 0.8531818181818182,
+      "grad_norm": 2.027263581028918,
+      "learning_rate": 1.162372226328779e-06,
+      "loss": 0.0683,
+      "step": 3754
+    },
+    {
+      "epoch": 0.8534090909090909,
+      "grad_norm": 1.7052912762164785,
+      "learning_rate": 1.1623266464779478e-06,
+      "loss": 0.0561,
+      "step": 3755
+    },
+    {
+      "epoch": 0.8536363636363636,
+      "grad_norm": 1.4315406617225686,
+      "learning_rate": 1.1622810556700889e-06,
+      "loss": 0.0732,
+      "step": 3756
+    },
+    {
+      "epoch": 0.8538636363636364,
+      "grad_norm": 2.276876768113902,
+      "learning_rate": 1.162235453906132e-06,
+      "loss": 0.0978,
+      "step": 3757
+    },
+    {
+      "epoch": 0.8540909090909091,
+      "grad_norm": 1.5214305104494021,
+      "learning_rate": 1.1621898411870066e-06,
+      "loss": 0.0652,
+      "step": 3758
+    },
+    {
+      "epoch": 0.8543181818181819,
+      "grad_norm": 1.7388589511092751,
+      "learning_rate": 1.1621442175136436e-06,
+      "loss": 0.0773,
+      "step": 3759
+    },
+    {
+      "epoch": 0.8545454545454545,
+      "grad_norm": 1.874088305130784,
+      "learning_rate": 1.1620985828869725e-06,
+      "loss": 0.0976,
+      "step": 3760
+    },
+    {
+      "epoch": 0.8547727272727272,
+      "grad_norm": 2.9459847313259604,
+      "learning_rate": 1.1620529373079246e-06,
+      "loss": 0.1101,
+      "step": 3761
+    },
+    {
+      "epoch": 0.855,
+      "grad_norm": 1.0970042166897314,
+      "learning_rate": 1.1620072807774301e-06,
+      "loss": 0.0718,
+      "step": 3762
+    },
+    {
+      "epoch": 0.8552272727272727,
+      "grad_norm": 2.236121593345304,
+      "learning_rate": 1.1619616132964203e-06,
+      "loss": 0.0853,
+      "step": 3763
+    },
+    {
+      "epoch": 0.8554545454545455,
+      "grad_norm": 2.1148682551983136,
+      "learning_rate": 1.1619159348658265e-06,
+      "loss": 0.0917,
+      "step": 3764
+    },
+    {
+      "epoch": 0.8556818181818182,
+      "grad_norm": 1.2277901561449394,
+      "learning_rate": 1.1618702454865802e-06,
+      "loss": 0.0521,
+      "step": 3765
+    },
+    {
+      "epoch": 0.855909090909091,
+      "grad_norm": 0.7967407547081585,
+      "learning_rate": 1.1618245451596128e-06,
+      "loss": 0.0312,
+      "step": 3766
+    },
+    {
+      "epoch": 0.8561363636363636,
+      "grad_norm": 2.2346529341058887,
+      "learning_rate": 1.1617788338858565e-06,
+      "loss": 0.099,
+      "step": 3767
+    },
+    {
+      "epoch": 0.8563636363636363,
+      "grad_norm": 1.8086948914258147,
+      "learning_rate": 1.1617331116662433e-06,
+      "loss": 0.1126,
+      "step": 3768
+    },
+    {
+      "epoch": 0.8565909090909091,
+      "grad_norm": 1.5784344231740588,
+      "learning_rate": 1.1616873785017054e-06,
+      "loss": 0.0524,
+      "step": 3769
+    },
+    {
+      "epoch": 0.8568181818181818,
+      "grad_norm": 1.2136958168974634,
+      "learning_rate": 1.1616416343931758e-06,
+      "loss": 0.071,
+      "step": 3770
+    },
+    {
+      "epoch": 0.8570454545454546,
+      "grad_norm": 1.3535380648569775,
+      "learning_rate": 1.161595879341587e-06,
+      "loss": 0.0656,
+      "step": 3771
+    },
+    {
+      "epoch": 0.8572727272727273,
+      "grad_norm": 1.4523331459935278,
+      "learning_rate": 1.1615501133478722e-06,
+      "loss": 0.081,
+      "step": 3772
+    },
+    {
+      "epoch": 0.8575,
+      "grad_norm": 2.0500149945599855,
+      "learning_rate": 1.1615043364129642e-06,
+      "loss": 0.1034,
+      "step": 3773
+    },
+    {
+      "epoch": 0.8577272727272728,
+      "grad_norm": 1.5157065688756708,
+      "learning_rate": 1.1614585485377969e-06,
+      "loss": 0.0662,
+      "step": 3774
+    },
+    {
+      "epoch": 0.8579545454545454,
+      "grad_norm": 1.887007201004279,
+      "learning_rate": 1.1614127497233039e-06,
+      "loss": 0.1042,
+      "step": 3775
+    },
+    {
+      "epoch": 0.8581818181818182,
+      "grad_norm": 1.8969942050738087,
+      "learning_rate": 1.1613669399704192e-06,
+      "loss": 0.069,
+      "step": 3776
+    },
+    {
+      "epoch": 0.8584090909090909,
+      "grad_norm": 2.780043148198906,
+      "learning_rate": 1.1613211192800768e-06,
+      "loss": 0.1072,
+      "step": 3777
+    },
+    {
+      "epoch": 0.8586363636363636,
+      "grad_norm": 1.8657576565878335,
+      "learning_rate": 1.161275287653211e-06,
+      "loss": 0.1178,
+      "step": 3778
+    },
+    {
+      "epoch": 0.8588636363636364,
+      "grad_norm": 1.5641878851733673,
+      "learning_rate": 1.1612294450907568e-06,
+      "loss": 0.0632,
+      "step": 3779
+    },
+    {
+      "epoch": 0.8590909090909091,
+      "grad_norm": 1.4288463035128731,
+      "learning_rate": 1.1611835915936484e-06,
+      "loss": 0.071,
+      "step": 3780
+    },
+    {
+      "epoch": 0.8593181818181819,
+      "grad_norm": 2.2294642537839784,
+      "learning_rate": 1.1611377271628213e-06,
+      "loss": 0.1207,
+      "step": 3781
+    },
+    {
+      "epoch": 0.8595454545454545,
+      "grad_norm": 2.6996398459348723,
+      "learning_rate": 1.1610918517992107e-06,
+      "loss": 0.0908,
+      "step": 3782
+    },
+    {
+      "epoch": 0.8597727272727272,
+      "grad_norm": 1.608232806653256,
+      "learning_rate": 1.1610459655037518e-06,
+      "loss": 0.0436,
+      "step": 3783
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7556871402010557,
+      "learning_rate": 1.1610000682773806e-06,
+      "loss": 0.0899,
+      "step": 3784
+    },
+    {
+      "epoch": 0.8602272727272727,
+      "grad_norm": 2.8597275382963634,
+      "learning_rate": 1.1609541601210329e-06,
+      "loss": 0.1202,
+      "step": 3785
+    },
+    {
+      "epoch": 0.8604545454545455,
+      "grad_norm": 1.8676630434261714,
+      "learning_rate": 1.1609082410356447e-06,
+      "loss": 0.0575,
+      "step": 3786
+    },
+    {
+      "epoch": 0.8606818181818182,
+      "grad_norm": 1.9125171487681878,
+      "learning_rate": 1.1608623110221526e-06,
+      "loss": 0.0739,
+      "step": 3787
+    },
+    {
+      "epoch": 0.860909090909091,
+      "grad_norm": 2.5850797531194143,
+      "learning_rate": 1.160816370081493e-06,
+      "loss": 0.1016,
+      "step": 3788
+    },
+    {
+      "epoch": 0.8611363636363636,
+      "grad_norm": 2.207419889410306,
+      "learning_rate": 1.160770418214603e-06,
+      "loss": 0.1102,
+      "step": 3789
+    },
+    {
+      "epoch": 0.8613636363636363,
+      "grad_norm": 1.8875052187780155,
+      "learning_rate": 1.1607244554224196e-06,
+      "loss": 0.0792,
+      "step": 3790
+    },
+    {
+      "epoch": 0.8615909090909091,
+      "grad_norm": 1.8049444198059654,
+      "learning_rate": 1.1606784817058797e-06,
+      "loss": 0.0813,
+      "step": 3791
+    },
+    {
+      "epoch": 0.8618181818181818,
+      "grad_norm": 1.731283248463127,
+      "learning_rate": 1.1606324970659212e-06,
+      "loss": 0.0726,
+      "step": 3792
+    },
+    {
+      "epoch": 0.8620454545454546,
+      "grad_norm": 1.6753196557113461,
+      "learning_rate": 1.1605865015034813e-06,
+      "loss": 0.045,
+      "step": 3793
+    },
+    {
+      "epoch": 0.8622727272727273,
+      "grad_norm": 1.7092061436760069,
+      "learning_rate": 1.1605404950194985e-06,
+      "loss": 0.1098,
+      "step": 3794
+    },
+    {
+      "epoch": 0.8625,
+      "grad_norm": 1.179088770378713,
+      "learning_rate": 1.1604944776149107e-06,
+      "loss": 0.0871,
+      "step": 3795
+    },
+    {
+      "epoch": 0.8627272727272727,
+      "grad_norm": 2.122837111022549,
+      "learning_rate": 1.1604484492906562e-06,
+      "loss": 0.1072,
+      "step": 3796
+    },
+    {
+      "epoch": 0.8629545454545454,
+      "grad_norm": 1.0980393851902674,
+      "learning_rate": 1.1604024100476736e-06,
+      "loss": 0.0419,
+      "step": 3797
+    },
+    {
+      "epoch": 0.8631818181818182,
+      "grad_norm": 1.8879681179064947,
+      "learning_rate": 1.1603563598869019e-06,
+      "loss": 0.0993,
+      "step": 3798
+    },
+    {
+      "epoch": 0.8634090909090909,
+      "grad_norm": 1.8023167498765775,
+      "learning_rate": 1.16031029880928e-06,
+      "loss": 0.0721,
+      "step": 3799
+    },
+    {
+      "epoch": 0.8636363636363636,
+      "grad_norm": 1.9532427861005313,
+      "learning_rate": 1.1602642268157473e-06,
+      "loss": 0.0624,
+      "step": 3800
+    },
+    {
+      "epoch": 0.8638636363636364,
+      "grad_norm": 1.29512717056828,
+      "learning_rate": 1.1602181439072432e-06,
+      "loss": 0.0711,
+      "step": 3801
+    },
+    {
+      "epoch": 0.8640909090909091,
+      "grad_norm": 1.6463202563878068,
+      "learning_rate": 1.1601720500847075e-06,
+      "loss": 0.0734,
+      "step": 3802
+    },
+    {
+      "epoch": 0.8643181818181818,
+      "grad_norm": 2.2128052745873443,
+      "learning_rate": 1.1601259453490798e-06,
+      "loss": 0.1026,
+      "step": 3803
+    },
+    {
+      "epoch": 0.8645454545454545,
+      "grad_norm": 1.5915594742494008,
+      "learning_rate": 1.1600798297013004e-06,
+      "loss": 0.0659,
+      "step": 3804
+    },
+    {
+      "epoch": 0.8647727272727272,
+      "grad_norm": 2.1522881917428336,
+      "learning_rate": 1.16003370314231e-06,
+      "loss": 0.0856,
+      "step": 3805
+    },
+    {
+      "epoch": 0.865,
+      "grad_norm": 1.5820600498243733,
+      "learning_rate": 1.1599875656730492e-06,
+      "loss": 0.0568,
+      "step": 3806
+    },
+    {
+      "epoch": 0.8652272727272727,
+      "grad_norm": 2.326087117148058,
+      "learning_rate": 1.1599414172944582e-06,
+      "loss": 0.0773,
+      "step": 3807
+    },
+    {
+      "epoch": 0.8654545454545455,
+      "grad_norm": 1.4873605884782242,
+      "learning_rate": 1.1598952580074787e-06,
+      "loss": 0.0747,
+      "step": 3808
+    },
+    {
+      "epoch": 0.8656818181818182,
+      "grad_norm": 2.2733943557149536,
+      "learning_rate": 1.1598490878130514e-06,
+      "loss": 0.1003,
+      "step": 3809
+    },
+    {
+      "epoch": 0.865909090909091,
+      "grad_norm": 1.5108305917075844,
+      "learning_rate": 1.1598029067121182e-06,
+      "loss": 0.0402,
+      "step": 3810
+    },
+    {
+      "epoch": 0.8661363636363636,
+      "grad_norm": 1.6245874913487663,
+      "learning_rate": 1.159756714705621e-06,
+      "loss": 0.0883,
+      "step": 3811
+    },
+    {
+      "epoch": 0.8663636363636363,
+      "grad_norm": 2.3071165021075153,
+      "learning_rate": 1.1597105117945014e-06,
+      "loss": 0.0946,
+      "step": 3812
+    },
+    {
+      "epoch": 0.8665909090909091,
+      "grad_norm": 1.2486012524831787,
+      "learning_rate": 1.1596642979797015e-06,
+      "loss": 0.0573,
+      "step": 3813
+    },
+    {
+      "epoch": 0.8668181818181818,
+      "grad_norm": 1.2217193087926537,
+      "learning_rate": 1.1596180732621638e-06,
+      "loss": 0.0396,
+      "step": 3814
+    },
+    {
+      "epoch": 0.8670454545454546,
+      "grad_norm": 1.562718869672984,
+      "learning_rate": 1.1595718376428308e-06,
+      "loss": 0.0705,
+      "step": 3815
+    },
+    {
+      "epoch": 0.8672727272727273,
+      "grad_norm": 1.677131670099555,
+      "learning_rate": 1.1595255911226456e-06,
+      "loss": 0.0961,
+      "step": 3816
+    },
+    {
+      "epoch": 0.8675,
+      "grad_norm": 1.2344500320863865,
+      "learning_rate": 1.159479333702551e-06,
+      "loss": 0.045,
+      "step": 3817
+    },
+    {
+      "epoch": 0.8677272727272727,
+      "grad_norm": 1.6627481418309438,
+      "learning_rate": 1.1594330653834902e-06,
+      "loss": 0.0853,
+      "step": 3818
+    },
+    {
+      "epoch": 0.8679545454545454,
+      "grad_norm": 1.5639644789501952,
+      "learning_rate": 1.1593867861664068e-06,
+      "loss": 0.0555,
+      "step": 3819
+    },
+    {
+      "epoch": 0.8681818181818182,
+      "grad_norm": 1.6525106119980297,
+      "learning_rate": 1.1593404960522448e-06,
+      "loss": 0.0688,
+      "step": 3820
+    },
+    {
+      "epoch": 0.8684090909090909,
+      "grad_norm": 1.6118040028901992,
+      "learning_rate": 1.1592941950419475e-06,
+      "loss": 0.0522,
+      "step": 3821
+    },
+    {
+      "epoch": 0.8686363636363637,
+      "grad_norm": 1.9535665745204844,
+      "learning_rate": 1.1592478831364598e-06,
+      "loss": 0.1209,
+      "step": 3822
+    },
+    {
+      "epoch": 0.8688636363636364,
+      "grad_norm": 0.9854462369432426,
+      "learning_rate": 1.1592015603367257e-06,
+      "loss": 0.0669,
+      "step": 3823
+    },
+    {
+      "epoch": 0.8690909090909091,
+      "grad_norm": 1.4073751507856134,
+      "learning_rate": 1.1591552266436896e-06,
+      "loss": 0.0947,
+      "step": 3824
+    },
+    {
+      "epoch": 0.8693181818181818,
+      "grad_norm": 1.7348769653944212,
+      "learning_rate": 1.1591088820582965e-06,
+      "loss": 0.0836,
+      "step": 3825
+    },
+    {
+      "epoch": 0.8695454545454545,
+      "grad_norm": 2.492631619796814,
+      "learning_rate": 1.1590625265814918e-06,
+      "loss": 0.1125,
+      "step": 3826
+    },
+    {
+      "epoch": 0.8697727272727273,
+      "grad_norm": 1.8576275997121587,
+      "learning_rate": 1.1590161602142204e-06,
+      "loss": 0.0779,
+      "step": 3827
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.517609415085955,
+      "learning_rate": 1.1589697829574275e-06,
+      "loss": 0.1052,
+      "step": 3828
+    },
+    {
+      "epoch": 0.8702272727272727,
+      "grad_norm": 1.747003101895003,
+      "learning_rate": 1.1589233948120596e-06,
+      "loss": 0.1001,
+      "step": 3829
+    },
+    {
+      "epoch": 0.8704545454545455,
+      "grad_norm": 1.7181798506335069,
+      "learning_rate": 1.158876995779062e-06,
+      "loss": 0.0832,
+      "step": 3830
+    },
+    {
+      "epoch": 0.8706818181818182,
+      "grad_norm": 0.8522328192056194,
+      "learning_rate": 1.1588305858593811e-06,
+      "loss": 0.0455,
+      "step": 3831
+    },
+    {
+      "epoch": 0.8709090909090909,
+      "grad_norm": 2.6697307997506514,
+      "learning_rate": 1.158784165053963e-06,
+      "loss": 0.0946,
+      "step": 3832
+    },
+    {
+      "epoch": 0.8711363636363636,
+      "grad_norm": 1.726787305172552,
+      "learning_rate": 1.158737733363755e-06,
+      "loss": 0.0698,
+      "step": 3833
+    },
+    {
+      "epoch": 0.8713636363636363,
+      "grad_norm": 1.5878363595632754,
+      "learning_rate": 1.1586912907897032e-06,
+      "loss": 0.0774,
+      "step": 3834
+    },
+    {
+      "epoch": 0.8715909090909091,
+      "grad_norm": 1.635294242650449,
+      "learning_rate": 1.158644837332755e-06,
+      "loss": 0.115,
+      "step": 3835
+    },
+    {
+      "epoch": 0.8718181818181818,
+      "grad_norm": 1.2546172783340133,
+      "learning_rate": 1.1585983729938575e-06,
+      "loss": 0.0655,
+      "step": 3836
+    },
+    {
+      "epoch": 0.8720454545454546,
+      "grad_norm": 1.6334916039284195,
+      "learning_rate": 1.158551897773958e-06,
+      "loss": 0.0633,
+      "step": 3837
+    },
+    {
+      "epoch": 0.8722727272727273,
+      "grad_norm": 1.7531124202980284,
+      "learning_rate": 1.1585054116740047e-06,
+      "loss": 0.0773,
+      "step": 3838
+    },
+    {
+      "epoch": 0.8725,
+      "grad_norm": 1.8111474570826078,
+      "learning_rate": 1.1584589146949455e-06,
+      "loss": 0.0664,
+      "step": 3839
+    },
+    {
+      "epoch": 0.8727272727272727,
+      "grad_norm": 1.3847047434321522,
+      "learning_rate": 1.1584124068377282e-06,
+      "loss": 0.0596,
+      "step": 3840
+    },
+    {
+      "epoch": 0.8729545454545454,
+      "grad_norm": 1.789737102825547,
+      "learning_rate": 1.1583658881033013e-06,
+      "loss": 0.0785,
+      "step": 3841
+    },
+    {
+      "epoch": 0.8731818181818182,
+      "grad_norm": 1.7271800723991413,
+      "learning_rate": 1.1583193584926133e-06,
+      "loss": 0.0712,
+      "step": 3842
+    },
+    {
+      "epoch": 0.8734090909090909,
+      "grad_norm": 1.4123204406886343,
+      "learning_rate": 1.1582728180066135e-06,
+      "loss": 0.06,
+      "step": 3843
+    },
+    {
+      "epoch": 0.8736363636363637,
+      "grad_norm": 0.7623817373232129,
+      "learning_rate": 1.1582262666462503e-06,
+      "loss": 0.0376,
+      "step": 3844
+    },
+    {
+      "epoch": 0.8738636363636364,
+      "grad_norm": 1.8435511236810238,
+      "learning_rate": 1.1581797044124733e-06,
+      "loss": 0.064,
+      "step": 3845
+    },
+    {
+      "epoch": 0.8740909090909091,
+      "grad_norm": 1.9982238238227406,
+      "learning_rate": 1.158133131306232e-06,
+      "loss": 0.0982,
+      "step": 3846
+    },
+    {
+      "epoch": 0.8743181818181818,
+      "grad_norm": 1.3702479499287277,
+      "learning_rate": 1.1580865473284762e-06,
+      "loss": 0.063,
+      "step": 3847
+    },
+    {
+      "epoch": 0.8745454545454545,
+      "grad_norm": 2.8084076148650245,
+      "learning_rate": 1.1580399524801555e-06,
+      "loss": 0.1453,
+      "step": 3848
+    },
+    {
+      "epoch": 0.8747727272727273,
+      "grad_norm": 1.915851180502446,
+      "learning_rate": 1.1579933467622201e-06,
+      "loss": 0.1018,
+      "step": 3849
+    },
+    {
+      "epoch": 0.875,
+      "grad_norm": 1.7580485049710362,
+      "learning_rate": 1.157946730175621e-06,
+      "loss": 0.0912,
+      "step": 3850
+    },
+    {
+      "epoch": 0.8752272727272727,
+      "grad_norm": 1.870265140460761,
+      "learning_rate": 1.1579001027213078e-06,
+      "loss": 0.0798,
+      "step": 3851
+    },
+    {
+      "epoch": 0.8754545454545455,
+      "grad_norm": 1.6254912182081138,
+      "learning_rate": 1.1578534644002319e-06,
+      "loss": 0.0383,
+      "step": 3852
+    },
+    {
+      "epoch": 0.8756818181818182,
+      "grad_norm": 2.141231557844524,
+      "learning_rate": 1.1578068152133443e-06,
+      "loss": 0.0864,
+      "step": 3853
+    },
+    {
+      "epoch": 0.8759090909090909,
+      "grad_norm": 2.067220884807508,
+      "learning_rate": 1.1577601551615962e-06,
+      "loss": 0.1271,
+      "step": 3854
+    },
+    {
+      "epoch": 0.8761363636363636,
+      "grad_norm": 2.190292266697076,
+      "learning_rate": 1.1577134842459391e-06,
+      "loss": 0.0872,
+      "step": 3855
+    },
+    {
+      "epoch": 0.8763636363636363,
+      "grad_norm": 1.4065098554338686,
+      "learning_rate": 1.1576668024673248e-06,
+      "loss": 0.0406,
+      "step": 3856
+    },
+    {
+      "epoch": 0.8765909090909091,
+      "grad_norm": 1.6543624064921079,
+      "learning_rate": 1.157620109826705e-06,
+      "loss": 0.0765,
+      "step": 3857
+    },
+    {
+      "epoch": 0.8768181818181818,
+      "grad_norm": 1.9047704052082228,
+      "learning_rate": 1.1575734063250318e-06,
+      "loss": 0.0949,
+      "step": 3858
+    },
+    {
+      "epoch": 0.8770454545454546,
+      "grad_norm": 2.110083074290689,
+      "learning_rate": 1.157526691963258e-06,
+      "loss": 0.099,
+      "step": 3859
+    },
+    {
+      "epoch": 0.8772727272727273,
+      "grad_norm": 2.758019264189553,
+      "learning_rate": 1.157479966742336e-06,
+      "loss": 0.1162,
+      "step": 3860
+    },
+    {
+      "epoch": 0.8775,
+      "grad_norm": 2.212766522143702,
+      "learning_rate": 1.157433230663218e-06,
+      "loss": 0.0929,
+      "step": 3861
+    },
+    {
+      "epoch": 0.8777272727272727,
+      "grad_norm": 0.901961384026111,
+      "learning_rate": 1.157386483726858e-06,
+      "loss": 0.0538,
+      "step": 3862
+    },
+    {
+      "epoch": 0.8779545454545454,
+      "grad_norm": 1.7931277746228507,
+      "learning_rate": 1.1573397259342086e-06,
+      "loss": 0.0683,
+      "step": 3863
+    },
+    {
+      "epoch": 0.8781818181818182,
+      "grad_norm": 2.6803649514069723,
+      "learning_rate": 1.1572929572862235e-06,
+      "loss": 0.1701,
+      "step": 3864
+    },
+    {
+      "epoch": 0.8784090909090909,
+      "grad_norm": 1.7005726048352199,
+      "learning_rate": 1.1572461777838564e-06,
+      "loss": 0.1059,
+      "step": 3865
+    },
+    {
+      "epoch": 0.8786363636363637,
+      "grad_norm": 2.0095332752006834,
+      "learning_rate": 1.1571993874280611e-06,
+      "loss": 0.0648,
+      "step": 3866
+    },
+    {
+      "epoch": 0.8788636363636364,
+      "grad_norm": 2.0648597793796926,
+      "learning_rate": 1.1571525862197918e-06,
+      "loss": 0.1203,
+      "step": 3867
+    },
+    {
+      "epoch": 0.8790909090909091,
+      "grad_norm": 1.9525048960953386,
+      "learning_rate": 1.1571057741600031e-06,
+      "loss": 0.0886,
+      "step": 3868
+    },
+    {
+      "epoch": 0.8793181818181818,
+      "grad_norm": 1.0042862007111508,
+      "learning_rate": 1.1570589512496492e-06,
+      "loss": 0.0463,
+      "step": 3869
+    },
+    {
+      "epoch": 0.8795454545454545,
+      "grad_norm": 2.839755655611553,
+      "learning_rate": 1.1570121174896851e-06,
+      "loss": 0.1863,
+      "step": 3870
+    },
+    {
+      "epoch": 0.8797727272727273,
+      "grad_norm": 2.0950280976830644,
+      "learning_rate": 1.1569652728810658e-06,
+      "loss": 0.1055,
+      "step": 3871
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4581057892594993,
+      "learning_rate": 1.1569184174247465e-06,
+      "loss": 0.0645,
+      "step": 3872
+    },
+    {
+      "epoch": 0.8802272727272727,
+      "grad_norm": 1.93168189409377,
+      "learning_rate": 1.1568715511216826e-06,
+      "loss": 0.0785,
+      "step": 3873
+    },
+    {
+      "epoch": 0.8804545454545455,
+      "grad_norm": 1.3729048292484265,
+      "learning_rate": 1.1568246739728297e-06,
+      "loss": 0.0703,
+      "step": 3874
+    },
+    {
+      "epoch": 0.8806818181818182,
+      "grad_norm": 3.1819301774993223,
+      "learning_rate": 1.1567777859791443e-06,
+      "loss": 0.1742,
+      "step": 3875
+    },
+    {
+      "epoch": 0.8809090909090909,
+      "grad_norm": 1.7551225806033077,
+      "learning_rate": 1.156730887141582e-06,
+      "loss": 0.1267,
+      "step": 3876
+    },
+    {
+      "epoch": 0.8811363636363636,
+      "grad_norm": 2.707297882974657,
+      "learning_rate": 1.1566839774610992e-06,
+      "loss": 0.1082,
+      "step": 3877
+    },
+    {
+      "epoch": 0.8813636363636363,
+      "grad_norm": 1.8682929393616259,
+      "learning_rate": 1.1566370569386523e-06,
+      "loss": 0.1045,
+      "step": 3878
+    },
+    {
+      "epoch": 0.8815909090909091,
+      "grad_norm": 1.6045190696335685,
+      "learning_rate": 1.1565901255751984e-06,
+      "loss": 0.0748,
+      "step": 3879
+    },
+    {
+      "epoch": 0.8818181818181818,
+      "grad_norm": 2.6142069551019103,
+      "learning_rate": 1.1565431833716947e-06,
+      "loss": 0.1305,
+      "step": 3880
+    },
+    {
+      "epoch": 0.8820454545454546,
+      "grad_norm": 1.5982697452587504,
+      "learning_rate": 1.156496230329098e-06,
+      "loss": 0.0752,
+      "step": 3881
+    },
+    {
+      "epoch": 0.8822727272727273,
+      "grad_norm": 2.2716531265870477,
+      "learning_rate": 1.156449266448366e-06,
+      "loss": 0.1144,
+      "step": 3882
+    },
+    {
+      "epoch": 0.8825,
+      "grad_norm": 1.9897427717929292,
+      "learning_rate": 1.156402291730456e-06,
+      "loss": 0.1325,
+      "step": 3883
+    },
+    {
+      "epoch": 0.8827272727272727,
+      "grad_norm": 1.2623679738226494,
+      "learning_rate": 1.1563553061763266e-06,
+      "loss": 0.0593,
+      "step": 3884
+    },
+    {
+      "epoch": 0.8829545454545454,
+      "grad_norm": 2.053751934922433,
+      "learning_rate": 1.1563083097869351e-06,
+      "loss": 0.074,
+      "step": 3885
+    },
+    {
+      "epoch": 0.8831818181818182,
+      "grad_norm": 2.7122650791978176,
+      "learning_rate": 1.1562613025632406e-06,
+      "loss": 0.168,
+      "step": 3886
+    },
+    {
+      "epoch": 0.8834090909090909,
+      "grad_norm": 1.310958808072135,
+      "learning_rate": 1.156214284506201e-06,
+      "loss": 0.0671,
+      "step": 3887
+    },
+    {
+      "epoch": 0.8836363636363637,
+      "grad_norm": 3.1750029905449866,
+      "learning_rate": 1.1561672556167757e-06,
+      "loss": 0.152,
+      "step": 3888
+    },
+    {
+      "epoch": 0.8838636363636364,
+      "grad_norm": 2.58043955884006,
+      "learning_rate": 1.156120215895923e-06,
+      "loss": 0.1022,
+      "step": 3889
+    },
+    {
+      "epoch": 0.884090909090909,
+      "grad_norm": 2.0103742948130385,
+      "learning_rate": 1.1560731653446029e-06,
+      "loss": 0.0991,
+      "step": 3890
+    },
+    {
+      "epoch": 0.8843181818181818,
+      "grad_norm": 2.3180293142528705,
+      "learning_rate": 1.1560261039637744e-06,
+      "loss": 0.1327,
+      "step": 3891
+    },
+    {
+      "epoch": 0.8845454545454545,
+      "grad_norm": 1.5925718561233075,
+      "learning_rate": 1.155979031754397e-06,
+      "loss": 0.0541,
+      "step": 3892
+    },
+    {
+      "epoch": 0.8847727272727273,
+      "grad_norm": 1.5523268614089774,
+      "learning_rate": 1.1559319487174308e-06,
+      "loss": 0.0662,
+      "step": 3893
+    },
+    {
+      "epoch": 0.885,
+      "grad_norm": 1.4815071480394388,
+      "learning_rate": 1.155884854853836e-06,
+      "loss": 0.0545,
+      "step": 3894
+    },
+    {
+      "epoch": 0.8852272727272728,
+      "grad_norm": 1.9819795662948663,
+      "learning_rate": 1.155837750164573e-06,
+      "loss": 0.1004,
+      "step": 3895
+    },
+    {
+      "epoch": 0.8854545454545455,
+      "grad_norm": 2.5713229209408617,
+      "learning_rate": 1.1557906346506021e-06,
+      "loss": 0.1421,
+      "step": 3896
+    },
+    {
+      "epoch": 0.8856818181818182,
+      "grad_norm": 1.1404729800830042,
+      "learning_rate": 1.1557435083128839e-06,
+      "loss": 0.0374,
+      "step": 3897
+    },
+    {
+      "epoch": 0.8859090909090909,
+      "grad_norm": 1.0571987612656373,
+      "learning_rate": 1.1556963711523796e-06,
+      "loss": 0.0466,
+      "step": 3898
+    },
+    {
+      "epoch": 0.8861363636363636,
+      "grad_norm": 1.5454922495344094,
+      "learning_rate": 1.155649223170051e-06,
+      "loss": 0.0734,
+      "step": 3899
+    },
+    {
+      "epoch": 0.8863636363636364,
+      "grad_norm": 2.042803507786921,
+      "learning_rate": 1.1556020643668583e-06,
+      "loss": 0.1002,
+      "step": 3900
+    },
+    {
+      "epoch": 0.8865909090909091,
+      "grad_norm": 1.964220770076743,
+      "learning_rate": 1.1555548947437642e-06,
+      "loss": 0.065,
+      "step": 3901
+    },
+    {
+      "epoch": 0.8868181818181818,
+      "grad_norm": 1.6568024236484018,
+      "learning_rate": 1.15550771430173e-06,
+      "loss": 0.0653,
+      "step": 3902
+    },
+    {
+      "epoch": 0.8870454545454546,
+      "grad_norm": 2.082550539195056,
+      "learning_rate": 1.1554605230417181e-06,
+      "loss": 0.0673,
+      "step": 3903
+    },
+    {
+      "epoch": 0.8872727272727273,
+      "grad_norm": 1.4977922918039945,
+      "learning_rate": 1.155413320964691e-06,
+      "loss": 0.0413,
+      "step": 3904
+    },
+    {
+      "epoch": 0.8875,
+      "grad_norm": 3.4622915705976327,
+      "learning_rate": 1.1553661080716106e-06,
+      "loss": 0.1452,
+      "step": 3905
+    },
+    {
+      "epoch": 0.8877272727272727,
+      "grad_norm": 1.7828971479995832,
+      "learning_rate": 1.1553188843634399e-06,
+      "loss": 0.0903,
+      "step": 3906
+    },
+    {
+      "epoch": 0.8879545454545454,
+      "grad_norm": 2.230762807998579,
+      "learning_rate": 1.155271649841142e-06,
+      "loss": 0.0946,
+      "step": 3907
+    },
+    {
+      "epoch": 0.8881818181818182,
+      "grad_norm": 1.6634773653032022,
+      "learning_rate": 1.1552244045056801e-06,
+      "loss": 0.0725,
+      "step": 3908
+    },
+    {
+      "epoch": 0.8884090909090909,
+      "grad_norm": 3.009021681734323,
+      "learning_rate": 1.1551771483580176e-06,
+      "loss": 0.0778,
+      "step": 3909
+    },
+    {
+      "epoch": 0.8886363636363637,
+      "grad_norm": 1.7731340712186003,
+      "learning_rate": 1.1551298813991182e-06,
+      "loss": 0.0774,
+      "step": 3910
+    },
+    {
+      "epoch": 0.8888636363636364,
+      "grad_norm": 1.7232108823646044,
+      "learning_rate": 1.1550826036299455e-06,
+      "loss": 0.081,
+      "step": 3911
+    },
+    {
+      "epoch": 0.889090909090909,
+      "grad_norm": 2.2909854033611112,
+      "learning_rate": 1.1550353150514637e-06,
+      "loss": 0.0779,
+      "step": 3912
+    },
+    {
+      "epoch": 0.8893181818181818,
+      "grad_norm": 1.4036755683769517,
+      "learning_rate": 1.1549880156646373e-06,
+      "loss": 0.0663,
+      "step": 3913
+    },
+    {
+      "epoch": 0.8895454545454545,
+      "grad_norm": 2.305293827951395,
+      "learning_rate": 1.1549407054704304e-06,
+      "loss": 0.0791,
+      "step": 3914
+    },
+    {
+      "epoch": 0.8897727272727273,
+      "grad_norm": 1.2890671061010743,
+      "learning_rate": 1.1548933844698082e-06,
+      "loss": 0.0805,
+      "step": 3915
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4723754760997125,
+      "learning_rate": 1.1548460526637354e-06,
+      "loss": 0.078,
+      "step": 3916
+    },
+    {
+      "epoch": 0.8902272727272728,
+      "grad_norm": 1.2042273564842731,
+      "learning_rate": 1.154798710053177e-06,
+      "loss": 0.061,
+      "step": 3917
+    },
+    {
+      "epoch": 0.8904545454545455,
+      "grad_norm": 1.3211554833108732,
+      "learning_rate": 1.154751356639099e-06,
+      "loss": 0.0462,
+      "step": 3918
+    },
+    {
+      "epoch": 0.8906818181818181,
+      "grad_norm": 1.6480479752797892,
+      "learning_rate": 1.1547039924224664e-06,
+      "loss": 0.0571,
+      "step": 3919
+    },
+    {
+      "epoch": 0.8909090909090909,
+      "grad_norm": 1.9964039903083493,
+      "learning_rate": 1.1546566174042454e-06,
+      "loss": 0.1169,
+      "step": 3920
+    },
+    {
+      "epoch": 0.8911363636363636,
+      "grad_norm": 0.8027502823009605,
+      "learning_rate": 1.1546092315854017e-06,
+      "loss": 0.0435,
+      "step": 3921
+    },
+    {
+      "epoch": 0.8913636363636364,
+      "grad_norm": 1.5742520234224997,
+      "learning_rate": 1.1545618349669018e-06,
+      "loss": 0.077,
+      "step": 3922
+    },
+    {
+      "epoch": 0.8915909090909091,
+      "grad_norm": 2.0165233779470477,
+      "learning_rate": 1.1545144275497126e-06,
+      "loss": 0.082,
+      "step": 3923
+    },
+    {
+      "epoch": 0.8918181818181818,
+      "grad_norm": 2.39060237144151,
+      "learning_rate": 1.1544670093348001e-06,
+      "loss": 0.155,
+      "step": 3924
+    },
+    {
+      "epoch": 0.8920454545454546,
+      "grad_norm": 1.2585260025451235,
+      "learning_rate": 1.1544195803231316e-06,
+      "loss": 0.0509,
+      "step": 3925
+    },
+    {
+      "epoch": 0.8922727272727272,
+      "grad_norm": 2.196931185382406,
+      "learning_rate": 1.1543721405156744e-06,
+      "loss": 0.115,
+      "step": 3926
+    },
+    {
+      "epoch": 0.8925,
+      "grad_norm": 1.7989864475795871,
+      "learning_rate": 1.1543246899133956e-06,
+      "loss": 0.0977,
+      "step": 3927
+    },
+    {
+      "epoch": 0.8927272727272727,
+      "grad_norm": 2.015185166131887,
+      "learning_rate": 1.154277228517263e-06,
+      "loss": 0.0895,
+      "step": 3928
+    },
+    {
+      "epoch": 0.8929545454545454,
+      "grad_norm": 2.1540832670099093,
+      "learning_rate": 1.1542297563282442e-06,
+      "loss": 0.0882,
+      "step": 3929
+    },
+    {
+      "epoch": 0.8931818181818182,
+      "grad_norm": 1.5742730846381199,
+      "learning_rate": 1.1541822733473077e-06,
+      "loss": 0.0953,
+      "step": 3930
+    },
+    {
+      "epoch": 0.8934090909090909,
+      "grad_norm": 2.1427907481337005,
+      "learning_rate": 1.154134779575421e-06,
+      "loss": 0.1055,
+      "step": 3931
+    },
+    {
+      "epoch": 0.8936363636363637,
+      "grad_norm": 1.29644662851884,
+      "learning_rate": 1.1540872750135532e-06,
+      "loss": 0.0452,
+      "step": 3932
+    },
+    {
+      "epoch": 0.8938636363636364,
+      "grad_norm": 2.4799015871545076,
+      "learning_rate": 1.1540397596626728e-06,
+      "loss": 0.1236,
+      "step": 3933
+    },
+    {
+      "epoch": 0.894090909090909,
+      "grad_norm": 1.5387654182985584,
+      "learning_rate": 1.1539922335237488e-06,
+      "loss": 0.0894,
+      "step": 3934
+    },
+    {
+      "epoch": 0.8943181818181818,
+      "grad_norm": 1.884381813423836,
+      "learning_rate": 1.1539446965977503e-06,
+      "loss": 0.131,
+      "step": 3935
+    },
+    {
+      "epoch": 0.8945454545454545,
+      "grad_norm": 1.8472883034403247,
+      "learning_rate": 1.1538971488856465e-06,
+      "loss": 0.1069,
+      "step": 3936
+    },
+    {
+      "epoch": 0.8947727272727273,
+      "grad_norm": 2.3752666365936372,
+      "learning_rate": 1.1538495903884073e-06,
+      "loss": 0.1455,
+      "step": 3937
+    },
+    {
+      "epoch": 0.895,
+      "grad_norm": 2.233617871889245,
+      "learning_rate": 1.1538020211070022e-06,
+      "loss": 0.1086,
+      "step": 3938
+    },
+    {
+      "epoch": 0.8952272727272728,
+      "grad_norm": 2.536470310379701,
+      "learning_rate": 1.1537544410424015e-06,
+      "loss": 0.1399,
+      "step": 3939
+    },
+    {
+      "epoch": 0.8954545454545455,
+      "grad_norm": 1.743881122733296,
+      "learning_rate": 1.153706850195575e-06,
+      "loss": 0.11,
+      "step": 3940
+    },
+    {
+      "epoch": 0.8956818181818181,
+      "grad_norm": 2.539487126015612,
+      "learning_rate": 1.153659248567494e-06,
+      "loss": 0.1003,
+      "step": 3941
+    },
+    {
+      "epoch": 0.8959090909090909,
+      "grad_norm": 1.5293945956373043,
+      "learning_rate": 1.153611636159128e-06,
+      "loss": 0.0774,
+      "step": 3942
+    },
+    {
+      "epoch": 0.8961363636363636,
+      "grad_norm": 2.7090442770586303,
+      "learning_rate": 1.1535640129714487e-06,
+      "loss": 0.1314,
+      "step": 3943
+    },
+    {
+      "epoch": 0.8963636363636364,
+      "grad_norm": 2.0848193854864525,
+      "learning_rate": 1.1535163790054272e-06,
+      "loss": 0.0727,
+      "step": 3944
+    },
+    {
+      "epoch": 0.8965909090909091,
+      "grad_norm": 1.4554727271278807,
+      "learning_rate": 1.1534687342620346e-06,
+      "loss": 0.0795,
+      "step": 3945
+    },
+    {
+      "epoch": 0.8968181818181818,
+      "grad_norm": 1.4748427988547006,
+      "learning_rate": 1.1534210787422425e-06,
+      "loss": 0.0489,
+      "step": 3946
+    },
+    {
+      "epoch": 0.8970454545454546,
+      "grad_norm": 2.070802206415807,
+      "learning_rate": 1.1533734124470229e-06,
+      "loss": 0.0984,
+      "step": 3947
+    },
+    {
+      "epoch": 0.8972727272727272,
+      "grad_norm": 2.012394436426914,
+      "learning_rate": 1.1533257353773474e-06,
+      "loss": 0.1118,
+      "step": 3948
+    },
+    {
+      "epoch": 0.8975,
+      "grad_norm": 1.6140823055604816,
+      "learning_rate": 1.1532780475341884e-06,
+      "loss": 0.0619,
+      "step": 3949
+    },
+    {
+      "epoch": 0.8977272727272727,
+      "grad_norm": 1.5748755102912533,
+      "learning_rate": 1.1532303489185185e-06,
+      "loss": 0.0951,
+      "step": 3950
+    },
+    {
+      "epoch": 0.8979545454545454,
+      "grad_norm": 1.9303877904190574,
+      "learning_rate": 1.1531826395313104e-06,
+      "loss": 0.0644,
+      "step": 3951
+    },
+    {
+      "epoch": 0.8981818181818182,
+      "grad_norm": 1.1211123147706952,
+      "learning_rate": 1.1531349193735366e-06,
+      "loss": 0.0461,
+      "step": 3952
+    },
+    {
+      "epoch": 0.8984090909090909,
+      "grad_norm": 2.439657692287954,
+      "learning_rate": 1.1530871884461705e-06,
+      "loss": 0.1241,
+      "step": 3953
+    },
+    {
+      "epoch": 0.8986363636363637,
+      "grad_norm": 1.5996303857396836,
+      "learning_rate": 1.153039446750185e-06,
+      "loss": 0.0846,
+      "step": 3954
+    },
+    {
+      "epoch": 0.8988636363636363,
+      "grad_norm": 1.6368065075146792,
+      "learning_rate": 1.1529916942865544e-06,
+      "loss": 0.0884,
+      "step": 3955
+    },
+    {
+      "epoch": 0.899090909090909,
+      "grad_norm": 2.134155105621591,
+      "learning_rate": 1.152943931056252e-06,
+      "loss": 0.0618,
+      "step": 3956
+    },
+    {
+      "epoch": 0.8993181818181818,
+      "grad_norm": 1.8615327782503983,
+      "learning_rate": 1.1528961570602517e-06,
+      "loss": 0.1049,
+      "step": 3957
+    },
+    {
+      "epoch": 0.8995454545454545,
+      "grad_norm": 2.340354878751805,
+      "learning_rate": 1.1528483722995277e-06,
+      "loss": 0.07,
+      "step": 3958
+    },
+    {
+      "epoch": 0.8997727272727273,
+      "grad_norm": 1.4132626904408265,
+      "learning_rate": 1.1528005767750546e-06,
+      "loss": 0.0464,
+      "step": 3959
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9720468261073771,
+      "learning_rate": 1.152752770487807e-06,
+      "loss": 0.0873,
+      "step": 3960
+    },
+    {
+      "epoch": 0.9002272727272728,
+      "grad_norm": 2.016101961197667,
+      "learning_rate": 1.1527049534387595e-06,
+      "loss": 0.0817,
+      "step": 3961
+    },
+    {
+      "epoch": 0.9004545454545455,
+      "grad_norm": 1.2709490877257785,
+      "learning_rate": 1.1526571256288874e-06,
+      "loss": 0.0536,
+      "step": 3962
+    },
+    {
+      "epoch": 0.9006818181818181,
+      "grad_norm": 1.5562121678277405,
+      "learning_rate": 1.1526092870591662e-06,
+      "loss": 0.0667,
+      "step": 3963
+    },
+    {
+      "epoch": 0.9009090909090909,
+      "grad_norm": 1.6732637371370518,
+      "learning_rate": 1.152561437730571e-06,
+      "loss": 0.0747,
+      "step": 3964
+    },
+    {
+      "epoch": 0.9011363636363636,
+      "grad_norm": 1.7089502473214286,
+      "learning_rate": 1.1525135776440777e-06,
+      "loss": 0.0778,
+      "step": 3965
+    },
+    {
+      "epoch": 0.9013636363636364,
+      "grad_norm": 1.7840855883910682,
+      "learning_rate": 1.1524657068006622e-06,
+      "loss": 0.0677,
+      "step": 3966
+    },
+    {
+      "epoch": 0.9015909090909091,
+      "grad_norm": 1.808503333042701,
+      "learning_rate": 1.1524178252013007e-06,
+      "loss": 0.0944,
+      "step": 3967
+    },
+    {
+      "epoch": 0.9018181818181819,
+      "grad_norm": 3.0514566954886746,
+      "learning_rate": 1.1523699328469698e-06,
+      "loss": 0.1303,
+      "step": 3968
+    },
+    {
+      "epoch": 0.9020454545454546,
+      "grad_norm": 1.233520035472673,
+      "learning_rate": 1.152322029738646e-06,
+      "loss": 0.0664,
+      "step": 3969
+    },
+    {
+      "epoch": 0.9022727272727272,
+      "grad_norm": 1.6221868949365736,
+      "learning_rate": 1.1522741158773057e-06,
+      "loss": 0.0454,
+      "step": 3970
+    },
+    {
+      "epoch": 0.9025,
+      "grad_norm": 1.9974890104761802,
+      "learning_rate": 1.1522261912639266e-06,
+      "loss": 0.0947,
+      "step": 3971
+    },
+    {
+      "epoch": 0.9027272727272727,
+      "grad_norm": 1.5783890529021098,
+      "learning_rate": 1.1521782558994854e-06,
+      "loss": 0.1067,
+      "step": 3972
+    },
+    {
+      "epoch": 0.9029545454545455,
+      "grad_norm": 1.305608054418615,
+      "learning_rate": 1.1521303097849601e-06,
+      "loss": 0.0534,
+      "step": 3973
+    },
+    {
+      "epoch": 0.9031818181818182,
+      "grad_norm": 3.2629007263717815,
+      "learning_rate": 1.152082352921328e-06,
+      "loss": 0.1663,
+      "step": 3974
+    },
+    {
+      "epoch": 0.9034090909090909,
+      "grad_norm": 2.0555624666555263,
+      "learning_rate": 1.1520343853095673e-06,
+      "loss": 0.1044,
+      "step": 3975
+    },
+    {
+      "epoch": 0.9036363636363637,
+      "grad_norm": 2.7094558155589565,
+      "learning_rate": 1.151986406950656e-06,
+      "loss": 0.1222,
+      "step": 3976
+    },
+    {
+      "epoch": 0.9038636363636363,
+      "grad_norm": 2.554967103271389,
+      "learning_rate": 1.1519384178455723e-06,
+      "loss": 0.1251,
+      "step": 3977
+    },
+    {
+      "epoch": 0.9040909090909091,
+      "grad_norm": 2.875767277934412,
+      "learning_rate": 1.151890417995295e-06,
+      "loss": 0.1348,
+      "step": 3978
+    },
+    {
+      "epoch": 0.9043181818181818,
+      "grad_norm": 2.3915482189942012,
+      "learning_rate": 1.151842407400803e-06,
+      "loss": 0.1205,
+      "step": 3979
+    },
+    {
+      "epoch": 0.9045454545454545,
+      "grad_norm": 2.0219736123414758,
+      "learning_rate": 1.1517943860630753e-06,
+      "loss": 0.1151,
+      "step": 3980
+    },
+    {
+      "epoch": 0.9047727272727273,
+      "grad_norm": 2.246068762070019,
+      "learning_rate": 1.1517463539830908e-06,
+      "loss": 0.1222,
+      "step": 3981
+    },
+    {
+      "epoch": 0.905,
+      "grad_norm": 2.0827092984597213,
+      "learning_rate": 1.1516983111618293e-06,
+      "loss": 0.0575,
+      "step": 3982
+    },
+    {
+      "epoch": 0.9052272727272728,
+      "grad_norm": 2.2558998443492575,
+      "learning_rate": 1.1516502576002702e-06,
+      "loss": 0.1523,
+      "step": 3983
+    },
+    {
+      "epoch": 0.9054545454545454,
+      "grad_norm": 1.8679218521263514,
+      "learning_rate": 1.1516021932993937e-06,
+      "loss": 0.0821,
+      "step": 3984
+    },
+    {
+      "epoch": 0.9056818181818181,
+      "grad_norm": 2.2129252470934855,
+      "learning_rate": 1.1515541182601796e-06,
+      "loss": 0.0739,
+      "step": 3985
+    },
+    {
+      "epoch": 0.9059090909090909,
+      "grad_norm": 1.4034095927313608,
+      "learning_rate": 1.1515060324836088e-06,
+      "loss": 0.084,
+      "step": 3986
+    },
+    {
+      "epoch": 0.9061363636363636,
+      "grad_norm": 2.6395906222059646,
+      "learning_rate": 1.151457935970661e-06,
+      "loss": 0.0953,
+      "step": 3987
+    },
+    {
+      "epoch": 0.9063636363636364,
+      "grad_norm": 1.792210053049276,
+      "learning_rate": 1.1514098287223177e-06,
+      "loss": 0.0925,
+      "step": 3988
+    },
+    {
+      "epoch": 0.9065909090909091,
+      "grad_norm": 3.1412752246478988,
+      "learning_rate": 1.1513617107395596e-06,
+      "loss": 0.1271,
+      "step": 3989
+    },
+    {
+      "epoch": 0.9068181818181819,
+      "grad_norm": 1.8439858605862962,
+      "learning_rate": 1.1513135820233678e-06,
+      "loss": 0.103,
+      "step": 3990
+    },
+    {
+      "epoch": 0.9070454545454546,
+      "grad_norm": 1.3306836712519547,
+      "learning_rate": 1.151265442574724e-06,
+      "loss": 0.0498,
+      "step": 3991
+    },
+    {
+      "epoch": 0.9072727272727272,
+      "grad_norm": 2.0270324703615907,
+      "learning_rate": 1.1512172923946097e-06,
+      "loss": 0.098,
+      "step": 3992
+    },
+    {
+      "epoch": 0.9075,
+      "grad_norm": 1.57038125111986,
+      "learning_rate": 1.1511691314840067e-06,
+      "loss": 0.0609,
+      "step": 3993
+    },
+    {
+      "epoch": 0.9077272727272727,
+      "grad_norm": 2.680210976753738,
+      "learning_rate": 1.1511209598438972e-06,
+      "loss": 0.1028,
+      "step": 3994
+    },
+    {
+      "epoch": 0.9079545454545455,
+      "grad_norm": 1.80519669930538,
+      "learning_rate": 1.1510727774752633e-06,
+      "loss": 0.0614,
+      "step": 3995
+    },
+    {
+      "epoch": 0.9081818181818182,
+      "grad_norm": 1.834507803426322,
+      "learning_rate": 1.151024584379088e-06,
+      "loss": 0.1032,
+      "step": 3996
+    },
+    {
+      "epoch": 0.9084090909090909,
+      "grad_norm": 1.3354662741507817,
+      "learning_rate": 1.1509763805563534e-06,
+      "loss": 0.0492,
+      "step": 3997
+    },
+    {
+      "epoch": 0.9086363636363637,
+      "grad_norm": 2.1555250998613658,
+      "learning_rate": 1.150928166008043e-06,
+      "loss": 0.0858,
+      "step": 3998
+    },
+    {
+      "epoch": 0.9088636363636363,
+      "grad_norm": 1.9578771713325538,
+      "learning_rate": 1.1508799407351396e-06,
+      "loss": 0.0884,
+      "step": 3999
+    },
+    {
+      "epoch": 0.9090909090909091,
+      "grad_norm": 1.3990561217280268,
+      "learning_rate": 1.150831704738627e-06,
+      "loss": 0.0531,
+      "step": 4000
+    },
+    {
+      "epoch": 0.9093181818181818,
+      "grad_norm": 3.054935311597894,
+      "learning_rate": 1.1507834580194883e-06,
+      "loss": 0.1427,
+      "step": 4001
+    },
+    {
+      "epoch": 0.9095454545454545,
+      "grad_norm": 2.825211864210614,
+      "learning_rate": 1.1507352005787078e-06,
+      "loss": 0.1047,
+      "step": 4002
+    },
+    {
+      "epoch": 0.9097727272727273,
+      "grad_norm": 1.698989230473553,
+      "learning_rate": 1.1506869324172693e-06,
+      "loss": 0.0953,
+      "step": 4003
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9809273665585636,
+      "learning_rate": 1.1506386535361572e-06,
+      "loss": 0.061,
+      "step": 4004
+    },
+    {
+      "epoch": 0.9102272727272728,
+      "grad_norm": 1.238134025191778,
+      "learning_rate": 1.1505903639363559e-06,
+      "loss": 0.0352,
+      "step": 4005
+    },
+    {
+      "epoch": 0.9104545454545454,
+      "grad_norm": 3.13073451555956,
+      "learning_rate": 1.15054206361885e-06,
+      "loss": 0.1245,
+      "step": 4006
+    },
+    {
+      "epoch": 0.9106818181818181,
+      "grad_norm": 1.4219619563904515,
+      "learning_rate": 1.1504937525846247e-06,
+      "loss": 0.0684,
+      "step": 4007
+    },
+    {
+      "epoch": 0.9109090909090909,
+      "grad_norm": 1.5454849127760617,
+      "learning_rate": 1.1504454308346649e-06,
+      "loss": 0.0651,
+      "step": 4008
+    },
+    {
+      "epoch": 0.9111363636363636,
+      "grad_norm": 2.522993617835452,
+      "learning_rate": 1.1503970983699561e-06,
+      "loss": 0.1289,
+      "step": 4009
+    },
+    {
+      "epoch": 0.9113636363636364,
+      "grad_norm": 2.1964763693503446,
+      "learning_rate": 1.150348755191484e-06,
+      "loss": 0.1283,
+      "step": 4010
+    },
+    {
+      "epoch": 0.9115909090909091,
+      "grad_norm": 3.788701000029646,
+      "learning_rate": 1.1503004013002341e-06,
+      "loss": 0.1497,
+      "step": 4011
+    },
+    {
+      "epoch": 0.9118181818181819,
+      "grad_norm": 2.849902989221603,
+      "learning_rate": 1.1502520366971927e-06,
+      "loss": 0.0907,
+      "step": 4012
+    },
+    {
+      "epoch": 0.9120454545454545,
+      "grad_norm": 2.0627563104745184,
+      "learning_rate": 1.1502036613833458e-06,
+      "loss": 0.0662,
+      "step": 4013
+    },
+    {
+      "epoch": 0.9122727272727272,
+      "grad_norm": 0.8895195454037521,
+      "learning_rate": 1.1501552753596801e-06,
+      "loss": 0.0443,
+      "step": 4014
+    },
+    {
+      "epoch": 0.9125,
+      "grad_norm": 2.0016902184921332,
+      "learning_rate": 1.1501068786271821e-06,
+      "loss": 0.1233,
+      "step": 4015
+    },
+    {
+      "epoch": 0.9127272727272727,
+      "grad_norm": 0.8939360843309885,
+      "learning_rate": 1.1500584711868388e-06,
+      "loss": 0.0391,
+      "step": 4016
+    },
+    {
+      "epoch": 0.9129545454545455,
+      "grad_norm": 1.7193569285912085,
+      "learning_rate": 1.150010053039637e-06,
+      "loss": 0.0844,
+      "step": 4017
+    },
+    {
+      "epoch": 0.9131818181818182,
+      "grad_norm": 2.42833674979026,
+      "learning_rate": 1.1499616241865646e-06,
+      "loss": 0.1009,
+      "step": 4018
+    },
+    {
+      "epoch": 0.913409090909091,
+      "grad_norm": 1.9331247665215425,
+      "learning_rate": 1.1499131846286087e-06,
+      "loss": 0.1252,
+      "step": 4019
+    },
+    {
+      "epoch": 0.9136363636363637,
+      "grad_norm": 2.845347604367503,
+      "learning_rate": 1.1498647343667574e-06,
+      "loss": 0.0685,
+      "step": 4020
+    },
+    {
+      "epoch": 0.9138636363636363,
+      "grad_norm": 2.361095999814292,
+      "learning_rate": 1.1498162734019983e-06,
+      "loss": 0.1553,
+      "step": 4021
+    },
+    {
+      "epoch": 0.9140909090909091,
+      "grad_norm": 1.633949090072858,
+      "learning_rate": 1.1497678017353199e-06,
+      "loss": 0.041,
+      "step": 4022
+    },
+    {
+      "epoch": 0.9143181818181818,
+      "grad_norm": 2.1728388540386794,
+      "learning_rate": 1.1497193193677102e-06,
+      "loss": 0.1581,
+      "step": 4023
+    },
+    {
+      "epoch": 0.9145454545454546,
+      "grad_norm": 1.8670825356985663,
+      "learning_rate": 1.1496708263001584e-06,
+      "loss": 0.0606,
+      "step": 4024
+    },
+    {
+      "epoch": 0.9147727272727273,
+      "grad_norm": 1.9369362016636977,
+      "learning_rate": 1.1496223225336533e-06,
+      "loss": 0.0729,
+      "step": 4025
+    },
+    {
+      "epoch": 0.915,
+      "grad_norm": 2.7467485503203934,
+      "learning_rate": 1.1495738080691834e-06,
+      "loss": 0.138,
+      "step": 4026
+    },
+    {
+      "epoch": 0.9152272727272728,
+      "grad_norm": 2.0898876171459246,
+      "learning_rate": 1.1495252829077385e-06,
+      "loss": 0.0582,
+      "step": 4027
+    },
+    {
+      "epoch": 0.9154545454545454,
+      "grad_norm": 2.0288412302631733,
+      "learning_rate": 1.1494767470503083e-06,
+      "loss": 0.0711,
+      "step": 4028
+    },
+    {
+      "epoch": 0.9156818181818182,
+      "grad_norm": 2.3753020103747553,
+      "learning_rate": 1.149428200497882e-06,
+      "loss": 0.1463,
+      "step": 4029
+    },
+    {
+      "epoch": 0.9159090909090909,
+      "grad_norm": 1.9123629716168327,
+      "learning_rate": 1.14937964325145e-06,
+      "loss": 0.0923,
+      "step": 4030
+    },
+    {
+      "epoch": 0.9161363636363636,
+      "grad_norm": 1.5291141129367314,
+      "learning_rate": 1.1493310753120017e-06,
+      "loss": 0.0608,
+      "step": 4031
+    },
+    {
+      "epoch": 0.9163636363636364,
+      "grad_norm": 1.3292163413016709,
+      "learning_rate": 1.1492824966805283e-06,
+      "loss": 0.0517,
+      "step": 4032
+    },
+    {
+      "epoch": 0.9165909090909091,
+      "grad_norm": 1.2484228298876365,
+      "learning_rate": 1.1492339073580203e-06,
+      "loss": 0.0533,
+      "step": 4033
+    },
+    {
+      "epoch": 0.9168181818181819,
+      "grad_norm": 1.1859858953632234,
+      "learning_rate": 1.1491853073454682e-06,
+      "loss": 0.0366,
+      "step": 4034
+    },
+    {
+      "epoch": 0.9170454545454545,
+      "grad_norm": 1.4706764904658964,
+      "learning_rate": 1.1491366966438633e-06,
+      "loss": 0.0878,
+      "step": 4035
+    },
+    {
+      "epoch": 0.9172727272727272,
+      "grad_norm": 1.1779792699695288,
+      "learning_rate": 1.1490880752541967e-06,
+      "loss": 0.049,
+      "step": 4036
+    },
+    {
+      "epoch": 0.9175,
+      "grad_norm": 1.123609744589482,
+      "learning_rate": 1.14903944317746e-06,
+      "loss": 0.0543,
+      "step": 4037
+    },
+    {
+      "epoch": 0.9177272727272727,
+      "grad_norm": 1.8312013116735653,
+      "learning_rate": 1.1489908004146444e-06,
+      "loss": 0.0707,
+      "step": 4038
+    },
+    {
+      "epoch": 0.9179545454545455,
+      "grad_norm": 2.618530409350869,
+      "learning_rate": 1.1489421469667425e-06,
+      "loss": 0.1343,
+      "step": 4039
+    },
+    {
+      "epoch": 0.9181818181818182,
+      "grad_norm": 2.019872133965422,
+      "learning_rate": 1.1488934828347465e-06,
+      "loss": 0.1416,
+      "step": 4040
+    },
+    {
+      "epoch": 0.918409090909091,
+      "grad_norm": 1.5055973053434788,
+      "learning_rate": 1.148844808019648e-06,
+      "loss": 0.0667,
+      "step": 4041
+    },
+    {
+      "epoch": 0.9186363636363636,
+      "grad_norm": 1.852530951844095,
+      "learning_rate": 1.14879612252244e-06,
+      "loss": 0.0773,
+      "step": 4042
+    },
+    {
+      "epoch": 0.9188636363636363,
+      "grad_norm": 1.6215103073547816,
+      "learning_rate": 1.1487474263441154e-06,
+      "loss": 0.0548,
+      "step": 4043
+    },
+    {
+      "epoch": 0.9190909090909091,
+      "grad_norm": 1.4539647604536925,
+      "learning_rate": 1.148698719485667e-06,
+      "loss": 0.0679,
+      "step": 4044
+    },
+    {
+      "epoch": 0.9193181818181818,
+      "grad_norm": 1.326780893715488,
+      "learning_rate": 1.148650001948088e-06,
+      "loss": 0.0594,
+      "step": 4045
+    },
+    {
+      "epoch": 0.9195454545454546,
+      "grad_norm": 2.6395660379670685,
+      "learning_rate": 1.1486012737323718e-06,
+      "loss": 0.1337,
+      "step": 4046
+    },
+    {
+      "epoch": 0.9197727272727273,
+      "grad_norm": 1.6890440942467548,
+      "learning_rate": 1.1485525348395124e-06,
+      "loss": 0.0781,
+      "step": 4047
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1930388243055137,
+      "learning_rate": 1.1485037852705033e-06,
+      "loss": 0.0958,
+      "step": 4048
+    },
+    {
+      "epoch": 0.9202272727272728,
+      "grad_norm": 2.699578355351962,
+      "learning_rate": 1.1484550250263388e-06,
+      "loss": 0.1602,
+      "step": 4049
+    },
+    {
+      "epoch": 0.9204545454545454,
+      "grad_norm": 2.1361497260321407,
+      "learning_rate": 1.148406254108013e-06,
+      "loss": 0.1163,
+      "step": 4050
+    },
+    {
+      "epoch": 0.9206818181818182,
+      "grad_norm": 1.9521749281246998,
+      "learning_rate": 1.1483574725165207e-06,
+      "loss": 0.0703,
+      "step": 4051
+    },
+    {
+      "epoch": 0.9209090909090909,
+      "grad_norm": 1.5764417483136657,
+      "learning_rate": 1.1483086802528563e-06,
+      "loss": 0.1005,
+      "step": 4052
+    },
+    {
+      "epoch": 0.9211363636363636,
+      "grad_norm": 1.7279927573795442,
+      "learning_rate": 1.148259877318015e-06,
+      "loss": 0.0584,
+      "step": 4053
+    },
+    {
+      "epoch": 0.9213636363636364,
+      "grad_norm": 1.369719668320114,
+      "learning_rate": 1.148211063712992e-06,
+      "loss": 0.0762,
+      "step": 4054
+    },
+    {
+      "epoch": 0.9215909090909091,
+      "grad_norm": 1.7738342394684923,
+      "learning_rate": 1.1481622394387825e-06,
+      "loss": 0.1101,
+      "step": 4055
+    },
+    {
+      "epoch": 0.9218181818181819,
+      "grad_norm": 2.0631662602265344,
+      "learning_rate": 1.1481134044963825e-06,
+      "loss": 0.1007,
+      "step": 4056
+    },
+    {
+      "epoch": 0.9220454545454545,
+      "grad_norm": 1.6050238534480612,
+      "learning_rate": 1.1480645588867872e-06,
+      "loss": 0.0791,
+      "step": 4057
+    },
+    {
+      "epoch": 0.9222727272727272,
+      "grad_norm": 1.187524746974321,
+      "learning_rate": 1.148015702610993e-06,
+      "loss": 0.0391,
+      "step": 4058
+    },
+    {
+      "epoch": 0.9225,
+      "grad_norm": 2.5447563045956776,
+      "learning_rate": 1.1479668356699966e-06,
+      "loss": 0.1702,
+      "step": 4059
+    },
+    {
+      "epoch": 0.9227272727272727,
+      "grad_norm": 2.572624853347682,
+      "learning_rate": 1.1479179580647938e-06,
+      "loss": 0.1403,
+      "step": 4060
+    },
+    {
+      "epoch": 0.9229545454545455,
+      "grad_norm": 3.0293607390586046,
+      "learning_rate": 1.1478690697963814e-06,
+      "loss": 0.1339,
+      "step": 4061
+    },
+    {
+      "epoch": 0.9231818181818182,
+      "grad_norm": 2.612886838273605,
+      "learning_rate": 1.1478201708657568e-06,
+      "loss": 0.1367,
+      "step": 4062
+    },
+    {
+      "epoch": 0.923409090909091,
+      "grad_norm": 1.853748027110018,
+      "learning_rate": 1.1477712612739168e-06,
+      "loss": 0.1093,
+      "step": 4063
+    },
+    {
+      "epoch": 0.9236363636363636,
+      "grad_norm": 1.2432416758859535,
+      "learning_rate": 1.1477223410218584e-06,
+      "loss": 0.0484,
+      "step": 4064
+    },
+    {
+      "epoch": 0.9238636363636363,
+      "grad_norm": 1.7323414250977922,
+      "learning_rate": 1.1476734101105797e-06,
+      "loss": 0.0679,
+      "step": 4065
+    },
+    {
+      "epoch": 0.9240909090909091,
+      "grad_norm": 1.7732614569130716,
+      "learning_rate": 1.1476244685410784e-06,
+      "loss": 0.0656,
+      "step": 4066
+    },
+    {
+      "epoch": 0.9243181818181818,
+      "grad_norm": 2.140936400436928,
+      "learning_rate": 1.1475755163143523e-06,
+      "loss": 0.0961,
+      "step": 4067
+    },
+    {
+      "epoch": 0.9245454545454546,
+      "grad_norm": 1.5423619950602365,
+      "learning_rate": 1.1475265534313997e-06,
+      "loss": 0.0595,
+      "step": 4068
+    },
+    {
+      "epoch": 0.9247727272727273,
+      "grad_norm": 1.53447117202004,
+      "learning_rate": 1.1474775798932192e-06,
+      "loss": 0.0807,
+      "step": 4069
+    },
+    {
+      "epoch": 0.925,
+      "grad_norm": 2.3857173231802373,
+      "learning_rate": 1.1474285957008093e-06,
+      "loss": 0.0807,
+      "step": 4070
+    },
+    {
+      "epoch": 0.9252272727272727,
+      "grad_norm": 2.037134732230061,
+      "learning_rate": 1.147379600855169e-06,
+      "loss": 0.1069,
+      "step": 4071
+    },
+    {
+      "epoch": 0.9254545454545454,
+      "grad_norm": 1.899682737008444,
+      "learning_rate": 1.1473305953572973e-06,
+      "loss": 0.0722,
+      "step": 4072
+    },
+    {
+      "epoch": 0.9256818181818182,
+      "grad_norm": 1.8062644779192272,
+      "learning_rate": 1.1472815792081935e-06,
+      "loss": 0.1004,
+      "step": 4073
+    },
+    {
+      "epoch": 0.9259090909090909,
+      "grad_norm": 1.2743613565986185,
+      "learning_rate": 1.147232552408857e-06,
+      "loss": 0.067,
+      "step": 4074
+    },
+    {
+      "epoch": 0.9261363636363636,
+      "grad_norm": 1.0098641624164855,
+      "learning_rate": 1.1471835149602877e-06,
+      "loss": 0.033,
+      "step": 4075
+    },
+    {
+      "epoch": 0.9263636363636364,
+      "grad_norm": 2.6870591352352227,
+      "learning_rate": 1.1471344668634854e-06,
+      "loss": 0.1266,
+      "step": 4076
+    },
+    {
+      "epoch": 0.9265909090909091,
+      "grad_norm": 1.2084288891980501,
+      "learning_rate": 1.1470854081194505e-06,
+      "loss": 0.0684,
+      "step": 4077
+    },
+    {
+      "epoch": 0.9268181818181818,
+      "grad_norm": 2.512536540491046,
+      "learning_rate": 1.1470363387291836e-06,
+      "loss": 0.1238,
+      "step": 4078
+    },
+    {
+      "epoch": 0.9270454545454545,
+      "grad_norm": 1.9858140489009224,
+      "learning_rate": 1.1469872586936845e-06,
+      "loss": 0.0741,
+      "step": 4079
+    },
+    {
+      "epoch": 0.9272727272727272,
+      "grad_norm": 2.3294938386131965,
+      "learning_rate": 1.1469381680139548e-06,
+      "loss": 0.0941,
+      "step": 4080
+    },
+    {
+      "epoch": 0.9275,
+      "grad_norm": 1.4846132259576668,
+      "learning_rate": 1.1468890666909954e-06,
+      "loss": 0.0602,
+      "step": 4081
+    },
+    {
+      "epoch": 0.9277272727272727,
+      "grad_norm": 1.3656795883727402,
+      "learning_rate": 1.1468399547258073e-06,
+      "loss": 0.0511,
+      "step": 4082
+    },
+    {
+      "epoch": 0.9279545454545455,
+      "grad_norm": 3.1016744100720555,
+      "learning_rate": 1.1467908321193922e-06,
+      "loss": 0.1316,
+      "step": 4083
+    },
+    {
+      "epoch": 0.9281818181818182,
+      "grad_norm": 2.2643738367221893,
+      "learning_rate": 1.1467416988727516e-06,
+      "loss": 0.061,
+      "step": 4084
+    },
+    {
+      "epoch": 0.928409090909091,
+      "grad_norm": 2.460283616212408,
+      "learning_rate": 1.1466925549868877e-06,
+      "loss": 0.1756,
+      "step": 4085
+    },
+    {
+      "epoch": 0.9286363636363636,
+      "grad_norm": 1.3820976544074317,
+      "learning_rate": 1.1466434004628023e-06,
+      "loss": 0.0535,
+      "step": 4086
+    },
+    {
+      "epoch": 0.9288636363636363,
+      "grad_norm": 2.5208756533619945,
+      "learning_rate": 1.146594235301498e-06,
+      "loss": 0.1101,
+      "step": 4087
+    },
+    {
+      "epoch": 0.9290909090909091,
+      "grad_norm": 2.577452957466939,
+      "learning_rate": 1.1465450595039772e-06,
+      "loss": 0.0898,
+      "step": 4088
+    },
+    {
+      "epoch": 0.9293181818181818,
+      "grad_norm": 1.6448002413982894,
+      "learning_rate": 1.146495873071243e-06,
+      "loss": 0.0729,
+      "step": 4089
+    },
+    {
+      "epoch": 0.9295454545454546,
+      "grad_norm": 2.108549918752881,
+      "learning_rate": 1.146446676004298e-06,
+      "loss": 0.0792,
+      "step": 4090
+    },
+    {
+      "epoch": 0.9297727272727273,
+      "grad_norm": 1.828036392914514,
+      "learning_rate": 1.1463974683041455e-06,
+      "loss": 0.1061,
+      "step": 4091
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3111532341628684,
+      "learning_rate": 1.146348249971789e-06,
+      "loss": 0.0811,
+      "step": 4092
+    },
+    {
+      "epoch": 0.9302272727272727,
+      "grad_norm": 1.7362366202428745,
+      "learning_rate": 1.1462990210082322e-06,
+      "loss": 0.0937,
+      "step": 4093
+    },
+    {
+      "epoch": 0.9304545454545454,
+      "grad_norm": 2.6140652911750246,
+      "learning_rate": 1.1462497814144788e-06,
+      "loss": 0.1281,
+      "step": 4094
+    },
+    {
+      "epoch": 0.9306818181818182,
+      "grad_norm": 2.2321852296743487,
+      "learning_rate": 1.146200531191533e-06,
+      "loss": 0.1289,
+      "step": 4095
+    },
+    {
+      "epoch": 0.9309090909090909,
+      "grad_norm": 2.358180195175633,
+      "learning_rate": 1.1461512703403992e-06,
+      "loss": 0.118,
+      "step": 4096
+    },
+    {
+      "epoch": 0.9311363636363637,
+      "grad_norm": 2.7467335640391703,
+      "learning_rate": 1.1461019988620817e-06,
+      "loss": 0.1127,
+      "step": 4097
+    },
+    {
+      "epoch": 0.9313636363636364,
+      "grad_norm": 3.024326355376512,
+      "learning_rate": 1.1460527167575851e-06,
+      "loss": 0.0757,
+      "step": 4098
+    },
+    {
+      "epoch": 0.9315909090909091,
+      "grad_norm": 1.3314264715671604,
+      "learning_rate": 1.146003424027915e-06,
+      "loss": 0.088,
+      "step": 4099
+    },
+    {
+      "epoch": 0.9318181818181818,
+      "grad_norm": 1.7268944782363447,
+      "learning_rate": 1.145954120674076e-06,
+      "loss": 0.0793,
+      "step": 4100
+    },
+    {
+      "epoch": 0.9320454545454545,
+      "grad_norm": 1.5843348517621019,
+      "learning_rate": 1.1459048066970736e-06,
+      "loss": 0.0903,
+      "step": 4101
+    },
+    {
+      "epoch": 0.9322727272727273,
+      "grad_norm": 0.8940932600784471,
+      "learning_rate": 1.1458554820979134e-06,
+      "loss": 0.0464,
+      "step": 4102
+    },
+    {
+      "epoch": 0.9325,
+      "grad_norm": 1.3903029511227578,
+      "learning_rate": 1.1458061468776013e-06,
+      "loss": 0.031,
+      "step": 4103
+    },
+    {
+      "epoch": 0.9327272727272727,
+      "grad_norm": 2.1377798092462177,
+      "learning_rate": 1.1457568010371432e-06,
+      "loss": 0.0561,
+      "step": 4104
+    },
+    {
+      "epoch": 0.9329545454545455,
+      "grad_norm": 1.7843887974515982,
+      "learning_rate": 1.1457074445775456e-06,
+      "loss": 0.0747,
+      "step": 4105
+    },
+    {
+      "epoch": 0.9331818181818182,
+      "grad_norm": 1.4857286610205904,
+      "learning_rate": 1.1456580774998146e-06,
+      "loss": 0.0793,
+      "step": 4106
+    },
+    {
+      "epoch": 0.9334090909090909,
+      "grad_norm": 1.4561791965151842,
+      "learning_rate": 1.145608699804957e-06,
+      "loss": 0.0704,
+      "step": 4107
+    },
+    {
+      "epoch": 0.9336363636363636,
+      "grad_norm": 0.962868979099717,
+      "learning_rate": 1.1455593114939799e-06,
+      "loss": 0.0641,
+      "step": 4108
+    },
+    {
+      "epoch": 0.9338636363636363,
+      "grad_norm": 1.6547394822460417,
+      "learning_rate": 1.1455099125678901e-06,
+      "loss": 0.0838,
+      "step": 4109
+    },
+    {
+      "epoch": 0.9340909090909091,
+      "grad_norm": 1.8404631122211592,
+      "learning_rate": 1.1454605030276953e-06,
+      "loss": 0.0634,
+      "step": 4110
+    },
+    {
+      "epoch": 0.9343181818181818,
+      "grad_norm": 1.7625533372673097,
+      "learning_rate": 1.1454110828744027e-06,
+      "loss": 0.0549,
+      "step": 4111
+    },
+    {
+      "epoch": 0.9345454545454546,
+      "grad_norm": 2.1643060904302045,
+      "learning_rate": 1.1453616521090204e-06,
+      "loss": 0.0853,
+      "step": 4112
+    },
+    {
+      "epoch": 0.9347727272727273,
+      "grad_norm": 1.9943707168463451,
+      "learning_rate": 1.1453122107325561e-06,
+      "loss": 0.0763,
+      "step": 4113
+    },
+    {
+      "epoch": 0.935,
+      "grad_norm": 1.0127405492344288,
+      "learning_rate": 1.145262758746018e-06,
+      "loss": 0.0689,
+      "step": 4114
+    },
+    {
+      "epoch": 0.9352272727272727,
+      "grad_norm": 1.292601824100203,
+      "learning_rate": 1.1452132961504149e-06,
+      "loss": 0.0475,
+      "step": 4115
+    },
+    {
+      "epoch": 0.9354545454545454,
+      "grad_norm": 2.213023025196244,
+      "learning_rate": 1.1451638229467547e-06,
+      "loss": 0.1043,
+      "step": 4116
+    },
+    {
+      "epoch": 0.9356818181818182,
+      "grad_norm": 1.9491957175456023,
+      "learning_rate": 1.145114339136047e-06,
+      "loss": 0.0988,
+      "step": 4117
+    },
+    {
+      "epoch": 0.9359090909090909,
+      "grad_norm": 1.2473946701475687,
+      "learning_rate": 1.1450648447193004e-06,
+      "loss": 0.049,
+      "step": 4118
+    },
+    {
+      "epoch": 0.9361363636363637,
+      "grad_norm": 1.8637987908997338,
+      "learning_rate": 1.1450153396975243e-06,
+      "loss": 0.0682,
+      "step": 4119
+    },
+    {
+      "epoch": 0.9363636363636364,
+      "grad_norm": 2.694403759161847,
+      "learning_rate": 1.1449658240717282e-06,
+      "loss": 0.1401,
+      "step": 4120
+    },
+    {
+      "epoch": 0.9365909090909091,
+      "grad_norm": 2.3344169453731327,
+      "learning_rate": 1.1449162978429218e-06,
+      "loss": 0.0665,
+      "step": 4121
+    },
+    {
+      "epoch": 0.9368181818181818,
+      "grad_norm": 2.218707023326159,
+      "learning_rate": 1.1448667610121152e-06,
+      "loss": 0.1658,
+      "step": 4122
+    },
+    {
+      "epoch": 0.9370454545454545,
+      "grad_norm": 3.080877041580417,
+      "learning_rate": 1.1448172135803183e-06,
+      "loss": 0.1897,
+      "step": 4123
+    },
+    {
+      "epoch": 0.9372727272727273,
+      "grad_norm": 1.3259664798672035,
+      "learning_rate": 1.1447676555485416e-06,
+      "loss": 0.078,
+      "step": 4124
+    },
+    {
+      "epoch": 0.9375,
+      "grad_norm": 0.5150745167657237,
+      "learning_rate": 1.1447180869177953e-06,
+      "loss": 0.0279,
+      "step": 4125
+    },
+    {
+      "epoch": 0.9377272727272727,
+      "grad_norm": 1.0366635025263082,
+      "learning_rate": 1.144668507689091e-06,
+      "loss": 0.0525,
+      "step": 4126
+    },
+    {
+      "epoch": 0.9379545454545455,
+      "grad_norm": 2.1734326869464553,
+      "learning_rate": 1.1446189178634388e-06,
+      "loss": 0.1375,
+      "step": 4127
+    },
+    {
+      "epoch": 0.9381818181818182,
+      "grad_norm": 1.878813663790923,
+      "learning_rate": 1.1445693174418505e-06,
+      "loss": 0.0811,
+      "step": 4128
+    },
+    {
+      "epoch": 0.9384090909090909,
+      "grad_norm": 2.037563982688254,
+      "learning_rate": 1.1445197064253375e-06,
+      "loss": 0.0802,
+      "step": 4129
+    },
+    {
+      "epoch": 0.9386363636363636,
+      "grad_norm": 2.220929172969581,
+      "learning_rate": 1.1444700848149113e-06,
+      "loss": 0.1106,
+      "step": 4130
+    },
+    {
+      "epoch": 0.9388636363636363,
+      "grad_norm": 1.6314558830127448,
+      "learning_rate": 1.1444204526115837e-06,
+      "loss": 0.0746,
+      "step": 4131
+    },
+    {
+      "epoch": 0.9390909090909091,
+      "grad_norm": 2.353523888044355,
+      "learning_rate": 1.1443708098163669e-06,
+      "loss": 0.0809,
+      "step": 4132
+    },
+    {
+      "epoch": 0.9393181818181818,
+      "grad_norm": 1.2480610890212498,
+      "learning_rate": 1.1443211564302731e-06,
+      "loss": 0.088,
+      "step": 4133
+    },
+    {
+      "epoch": 0.9395454545454546,
+      "grad_norm": 1.4984614156711462,
+      "learning_rate": 1.1442714924543151e-06,
+      "loss": 0.0611,
+      "step": 4134
+    },
+    {
+      "epoch": 0.9397727272727273,
+      "grad_norm": 1.3884908276684538,
+      "learning_rate": 1.1442218178895053e-06,
+      "loss": 0.0627,
+      "step": 4135
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9566750798329622,
+      "learning_rate": 1.144172132736857e-06,
+      "loss": 0.0539,
+      "step": 4136
+    },
+    {
+      "epoch": 0.9402272727272727,
+      "grad_norm": 2.2472194581860965,
+      "learning_rate": 1.144122436997383e-06,
+      "loss": 0.0941,
+      "step": 4137
+    },
+    {
+      "epoch": 0.9404545454545454,
+      "grad_norm": 0.9695086439387153,
+      "learning_rate": 1.1440727306720969e-06,
+      "loss": 0.0491,
+      "step": 4138
+    },
+    {
+      "epoch": 0.9406818181818182,
+      "grad_norm": 2.8069787012682683,
+      "learning_rate": 1.1440230137620124e-06,
+      "loss": 0.1144,
+      "step": 4139
+    },
+    {
+      "epoch": 0.9409090909090909,
+      "grad_norm": 1.6857675982388427,
+      "learning_rate": 1.143973286268143e-06,
+      "loss": 0.0969,
+      "step": 4140
+    },
+    {
+      "epoch": 0.9411363636363637,
+      "grad_norm": 1.9326304937312258,
+      "learning_rate": 1.1439235481915028e-06,
+      "loss": 0.0777,
+      "step": 4141
+    },
+    {
+      "epoch": 0.9413636363636364,
+      "grad_norm": 1.4326907419249764,
+      "learning_rate": 1.1438737995331061e-06,
+      "loss": 0.0781,
+      "step": 4142
+    },
+    {
+      "epoch": 0.9415909090909091,
+      "grad_norm": 3.073828077761356,
+      "learning_rate": 1.1438240402939676e-06,
+      "loss": 0.1277,
+      "step": 4143
+    },
+    {
+      "epoch": 0.9418181818181818,
+      "grad_norm": 2.498765581555338,
+      "learning_rate": 1.1437742704751018e-06,
+      "loss": 0.1295,
+      "step": 4144
+    },
+    {
+      "epoch": 0.9420454545454545,
+      "grad_norm": 2.9984741942634754,
+      "learning_rate": 1.1437244900775235e-06,
+      "loss": 0.1621,
+      "step": 4145
+    },
+    {
+      "epoch": 0.9422727272727273,
+      "grad_norm": 1.5863717158753057,
+      "learning_rate": 1.1436746991022479e-06,
+      "loss": 0.0577,
+      "step": 4146
+    },
+    {
+      "epoch": 0.9425,
+      "grad_norm": 1.784607262244383,
+      "learning_rate": 1.1436248975502904e-06,
+      "loss": 0.0398,
+      "step": 4147
+    },
+    {
+      "epoch": 0.9427272727272727,
+      "grad_norm": 2.9801408197896304,
+      "learning_rate": 1.1435750854226664e-06,
+      "loss": 0.1302,
+      "step": 4148
+    },
+    {
+      "epoch": 0.9429545454545455,
+      "grad_norm": 1.54369443297821,
+      "learning_rate": 1.1435252627203917e-06,
+      "loss": 0.0705,
+      "step": 4149
+    },
+    {
+      "epoch": 0.9431818181818182,
+      "grad_norm": 1.485993399262697,
+      "learning_rate": 1.143475429444482e-06,
+      "loss": 0.0805,
+      "step": 4150
+    },
+    {
+      "epoch": 0.9434090909090909,
+      "grad_norm": 2.139722848284652,
+      "learning_rate": 1.143425585595954e-06,
+      "loss": 0.1272,
+      "step": 4151
+    },
+    {
+      "epoch": 0.9436363636363636,
+      "grad_norm": 0.910636087124845,
+      "learning_rate": 1.143375731175824e-06,
+      "loss": 0.0398,
+      "step": 4152
+    },
+    {
+      "epoch": 0.9438636363636363,
+      "grad_norm": 1.5294010261514184,
+      "learning_rate": 1.1433258661851082e-06,
+      "loss": 0.0412,
+      "step": 4153
+    },
+    {
+      "epoch": 0.9440909090909091,
+      "grad_norm": 2.1973013496379643,
+      "learning_rate": 1.1432759906248239e-06,
+      "loss": 0.0987,
+      "step": 4154
+    },
+    {
+      "epoch": 0.9443181818181818,
+      "grad_norm": 1.8704097147771948,
+      "learning_rate": 1.143226104495988e-06,
+      "loss": 0.0548,
+      "step": 4155
+    },
+    {
+      "epoch": 0.9445454545454546,
+      "grad_norm": 1.639513461913086,
+      "learning_rate": 1.1431762077996174e-06,
+      "loss": 0.0984,
+      "step": 4156
+    },
+    {
+      "epoch": 0.9447727272727273,
+      "grad_norm": 2.921361744082297,
+      "learning_rate": 1.1431263005367305e-06,
+      "loss": 0.1292,
+      "step": 4157
+    },
+    {
+      "epoch": 0.945,
+      "grad_norm": 1.6466575793725589,
+      "learning_rate": 1.143076382708344e-06,
+      "loss": 0.0581,
+      "step": 4158
+    },
+    {
+      "epoch": 0.9452272727272727,
+      "grad_norm": 1.7850271580748038,
+      "learning_rate": 1.1430264543154763e-06,
+      "loss": 0.0724,
+      "step": 4159
+    },
+    {
+      "epoch": 0.9454545454545454,
+      "grad_norm": 1.6772527017433567,
+      "learning_rate": 1.1429765153591455e-06,
+      "loss": 0.0585,
+      "step": 4160
+    },
+    {
+      "epoch": 0.9456818181818182,
+      "grad_norm": 2.2926202195785628,
+      "learning_rate": 1.1429265658403698e-06,
+      "loss": 0.087,
+      "step": 4161
+    },
+    {
+      "epoch": 0.9459090909090909,
+      "grad_norm": 1.6591188505732637,
+      "learning_rate": 1.1428766057601679e-06,
+      "loss": 0.0712,
+      "step": 4162
+    },
+    {
+      "epoch": 0.9461363636363637,
+      "grad_norm": 1.544962006475865,
+      "learning_rate": 1.1428266351195585e-06,
+      "loss": 0.0532,
+      "step": 4163
+    },
+    {
+      "epoch": 0.9463636363636364,
+      "grad_norm": 2.6498145018805572,
+      "learning_rate": 1.1427766539195608e-06,
+      "loss": 0.1197,
+      "step": 4164
+    },
+    {
+      "epoch": 0.946590909090909,
+      "grad_norm": 1.6627095691399898,
+      "learning_rate": 1.1427266621611936e-06,
+      "loss": 0.0496,
+      "step": 4165
+    },
+    {
+      "epoch": 0.9468181818181818,
+      "grad_norm": 1.3554188784814045,
+      "learning_rate": 1.1426766598454768e-06,
+      "loss": 0.0639,
+      "step": 4166
+    },
+    {
+      "epoch": 0.9470454545454545,
+      "grad_norm": 1.9437442982490087,
+      "learning_rate": 1.1426266469734293e-06,
+      "loss": 0.0944,
+      "step": 4167
+    },
+    {
+      "epoch": 0.9472727272727273,
+      "grad_norm": 2.3755819892665615,
+      "learning_rate": 1.1425766235460716e-06,
+      "loss": 0.1203,
+      "step": 4168
+    },
+    {
+      "epoch": 0.9475,
+      "grad_norm": 1.5831374564082885,
+      "learning_rate": 1.1425265895644239e-06,
+      "loss": 0.0831,
+      "step": 4169
+    },
+    {
+      "epoch": 0.9477272727272728,
+      "grad_norm": 2.398317312326405,
+      "learning_rate": 1.1424765450295058e-06,
+      "loss": 0.1114,
+      "step": 4170
+    },
+    {
+      "epoch": 0.9479545454545455,
+      "grad_norm": 1.2995137674176975,
+      "learning_rate": 1.1424264899423383e-06,
+      "loss": 0.0815,
+      "step": 4171
+    },
+    {
+      "epoch": 0.9481818181818182,
+      "grad_norm": 2.0274122937438857,
+      "learning_rate": 1.142376424303942e-06,
+      "loss": 0.1135,
+      "step": 4172
+    },
+    {
+      "epoch": 0.9484090909090909,
+      "grad_norm": 2.5101193951374174,
+      "learning_rate": 1.1423263481153376e-06,
+      "loss": 0.1402,
+      "step": 4173
+    },
+    {
+      "epoch": 0.9486363636363636,
+      "grad_norm": 3.0017520765861083,
+      "learning_rate": 1.1422762613775467e-06,
+      "loss": 0.1314,
+      "step": 4174
+    },
+    {
+      "epoch": 0.9488636363636364,
+      "grad_norm": 1.7317402244161635,
+      "learning_rate": 1.1422261640915902e-06,
+      "loss": 0.076,
+      "step": 4175
+    },
+    {
+      "epoch": 0.9490909090909091,
+      "grad_norm": 1.148093125080391,
+      "learning_rate": 1.1421760562584901e-06,
+      "loss": 0.0763,
+      "step": 4176
+    },
+    {
+      "epoch": 0.9493181818181818,
+      "grad_norm": 2.4449316822833227,
+      "learning_rate": 1.1421259378792676e-06,
+      "loss": 0.1041,
+      "step": 4177
+    },
+    {
+      "epoch": 0.9495454545454546,
+      "grad_norm": 1.8530967389882742,
+      "learning_rate": 1.1420758089549453e-06,
+      "loss": 0.0726,
+      "step": 4178
+    },
+    {
+      "epoch": 0.9497727272727273,
+      "grad_norm": 1.2671094633757,
+      "learning_rate": 1.1420256694865451e-06,
+      "loss": 0.0577,
+      "step": 4179
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3984993766797098,
+      "learning_rate": 1.1419755194750894e-06,
+      "loss": 0.0591,
+      "step": 4180
+    },
+    {
+      "epoch": 0.9502272727272727,
+      "grad_norm": 1.1746934037309682,
+      "learning_rate": 1.1419253589216012e-06,
+      "loss": 0.0494,
+      "step": 4181
+    },
+    {
+      "epoch": 0.9504545454545454,
+      "grad_norm": 1.6451767981559486,
+      "learning_rate": 1.141875187827103e-06,
+      "loss": 0.0533,
+      "step": 4182
+    },
+    {
+      "epoch": 0.9506818181818182,
+      "grad_norm": 2.0769237483799445,
+      "learning_rate": 1.141825006192618e-06,
+      "loss": 0.1,
+      "step": 4183
+    },
+    {
+      "epoch": 0.9509090909090909,
+      "grad_norm": 2.4081317841523093,
+      "learning_rate": 1.1417748140191693e-06,
+      "loss": 0.0644,
+      "step": 4184
+    },
+    {
+      "epoch": 0.9511363636363637,
+      "grad_norm": 2.714102404867192,
+      "learning_rate": 1.1417246113077807e-06,
+      "loss": 0.1097,
+      "step": 4185
+    },
+    {
+      "epoch": 0.9513636363636364,
+      "grad_norm": 1.2662468188157625,
+      "learning_rate": 1.1416743980594758e-06,
+      "loss": 0.0535,
+      "step": 4186
+    },
+    {
+      "epoch": 0.951590909090909,
+      "grad_norm": 2.167343070591957,
+      "learning_rate": 1.1416241742752787e-06,
+      "loss": 0.1261,
+      "step": 4187
+    },
+    {
+      "epoch": 0.9518181818181818,
+      "grad_norm": 2.6019436779968967,
+      "learning_rate": 1.1415739399562132e-06,
+      "loss": 0.1298,
+      "step": 4188
+    },
+    {
+      "epoch": 0.9520454545454545,
+      "grad_norm": 1.1718473582162454,
+      "learning_rate": 1.1415236951033038e-06,
+      "loss": 0.0499,
+      "step": 4189
+    },
+    {
+      "epoch": 0.9522727272727273,
+      "grad_norm": 1.7889350248696054,
+      "learning_rate": 1.141473439717575e-06,
+      "loss": 0.0887,
+      "step": 4190
+    },
+    {
+      "epoch": 0.9525,
+      "grad_norm": 2.137283770334028,
+      "learning_rate": 1.1414231738000521e-06,
+      "loss": 0.1292,
+      "step": 4191
+    },
+    {
+      "epoch": 0.9527272727272728,
+      "grad_norm": 1.463730097615998,
+      "learning_rate": 1.1413728973517597e-06,
+      "loss": 0.0643,
+      "step": 4192
+    },
+    {
+      "epoch": 0.9529545454545455,
+      "grad_norm": 1.7675952635444532,
+      "learning_rate": 1.1413226103737228e-06,
+      "loss": 0.0641,
+      "step": 4193
+    },
+    {
+      "epoch": 0.9531818181818181,
+      "grad_norm": 1.28762055628509,
+      "learning_rate": 1.1412723128669674e-06,
+      "loss": 0.0458,
+      "step": 4194
+    },
+    {
+      "epoch": 0.9534090909090909,
+      "grad_norm": 2.4134655283114603,
+      "learning_rate": 1.1412220048325186e-06,
+      "loss": 0.0759,
+      "step": 4195
+    },
+    {
+      "epoch": 0.9536363636363636,
+      "grad_norm": 1.5383396750570955,
+      "learning_rate": 1.1411716862714027e-06,
+      "loss": 0.0709,
+      "step": 4196
+    },
+    {
+      "epoch": 0.9538636363636364,
+      "grad_norm": 1.4474936665981037,
+      "learning_rate": 1.1411213571846454e-06,
+      "loss": 0.0619,
+      "step": 4197
+    },
+    {
+      "epoch": 0.9540909090909091,
+      "grad_norm": 2.6025192373483654,
+      "learning_rate": 1.1410710175732734e-06,
+      "loss": 0.1059,
+      "step": 4198
+    },
+    {
+      "epoch": 0.9543181818181818,
+      "grad_norm": 1.6667338703130723,
+      "learning_rate": 1.141020667438313e-06,
+      "loss": 0.1005,
+      "step": 4199
+    },
+    {
+      "epoch": 0.9545454545454546,
+      "grad_norm": 1.5214217497711937,
+      "learning_rate": 1.1409703067807906e-06,
+      "loss": 0.0679,
+      "step": 4200
+    },
+    {
+      "epoch": 0.9547727272727272,
+      "grad_norm": 1.6422106211104552,
+      "learning_rate": 1.1409199356017339e-06,
+      "loss": 0.0656,
+      "step": 4201
+    },
+    {
+      "epoch": 0.955,
+      "grad_norm": 2.1288139599561697,
+      "learning_rate": 1.1408695539021694e-06,
+      "loss": 0.1279,
+      "step": 4202
+    },
+    {
+      "epoch": 0.9552272727272727,
+      "grad_norm": 2.401710766729204,
+      "learning_rate": 1.140819161683125e-06,
+      "loss": 0.0795,
+      "step": 4203
+    },
+    {
+      "epoch": 0.9554545454545454,
+      "grad_norm": 2.0805395102975925,
+      "learning_rate": 1.1407687589456277e-06,
+      "loss": 0.0976,
+      "step": 4204
+    },
+    {
+      "epoch": 0.9556818181818182,
+      "grad_norm": 1.7581374331144912,
+      "learning_rate": 1.1407183456907057e-06,
+      "loss": 0.1178,
+      "step": 4205
+    },
+    {
+      "epoch": 0.9559090909090909,
+      "grad_norm": 1.7923042056544218,
+      "learning_rate": 1.140667921919387e-06,
+      "loss": 0.0674,
+      "step": 4206
+    },
+    {
+      "epoch": 0.9561363636363637,
+      "grad_norm": 2.9041366578583037,
+      "learning_rate": 1.1406174876326995e-06,
+      "loss": 0.0871,
+      "step": 4207
+    },
+    {
+      "epoch": 0.9563636363636364,
+      "grad_norm": 1.4272828943331408,
+      "learning_rate": 1.140567042831672e-06,
+      "loss": 0.0998,
+      "step": 4208
+    },
+    {
+      "epoch": 0.956590909090909,
+      "grad_norm": 2.0540799816088304,
+      "learning_rate": 1.1405165875173333e-06,
+      "loss": 0.0757,
+      "step": 4209
+    },
+    {
+      "epoch": 0.9568181818181818,
+      "grad_norm": 2.035416900420297,
+      "learning_rate": 1.1404661216907117e-06,
+      "loss": 0.1306,
+      "step": 4210
+    },
+    {
+      "epoch": 0.9570454545454545,
+      "grad_norm": 1.85329820637599,
+      "learning_rate": 1.1404156453528367e-06,
+      "loss": 0.1041,
+      "step": 4211
+    },
+    {
+      "epoch": 0.9572727272727273,
+      "grad_norm": 2.411956384518529,
+      "learning_rate": 1.1403651585047373e-06,
+      "loss": 0.0758,
+      "step": 4212
+    },
+    {
+      "epoch": 0.9575,
+      "grad_norm": 1.069620866002623,
+      "learning_rate": 1.1403146611474437e-06,
+      "loss": 0.045,
+      "step": 4213
+    },
+    {
+      "epoch": 0.9577272727272728,
+      "grad_norm": 1.7241717405550674,
+      "learning_rate": 1.1402641532819847e-06,
+      "loss": 0.0876,
+      "step": 4214
+    },
+    {
+      "epoch": 0.9579545454545455,
+      "grad_norm": 1.7597464347280758,
+      "learning_rate": 1.140213634909391e-06,
+      "loss": 0.0715,
+      "step": 4215
+    },
+    {
+      "epoch": 0.9581818181818181,
+      "grad_norm": 1.2620310663093226,
+      "learning_rate": 1.1401631060306921e-06,
+      "loss": 0.0454,
+      "step": 4216
+    },
+    {
+      "epoch": 0.9584090909090909,
+      "grad_norm": 1.3273345900215745,
+      "learning_rate": 1.140112566646919e-06,
+      "loss": 0.0458,
+      "step": 4217
+    },
+    {
+      "epoch": 0.9586363636363636,
+      "grad_norm": 2.248816924790617,
+      "learning_rate": 1.1400620167591018e-06,
+      "loss": 0.0759,
+      "step": 4218
+    },
+    {
+      "epoch": 0.9588636363636364,
+      "grad_norm": 2.1379480124563712,
+      "learning_rate": 1.1400114563682714e-06,
+      "loss": 0.0917,
+      "step": 4219
+    },
+    {
+      "epoch": 0.9590909090909091,
+      "grad_norm": 1.58142617239529,
+      "learning_rate": 1.1399608854754593e-06,
+      "loss": 0.0983,
+      "step": 4220
+    },
+    {
+      "epoch": 0.9593181818181818,
+      "grad_norm": 2.3161509690539464,
+      "learning_rate": 1.1399103040816963e-06,
+      "loss": 0.0829,
+      "step": 4221
+    },
+    {
+      "epoch": 0.9595454545454546,
+      "grad_norm": 0.9870135394889189,
+      "learning_rate": 1.1398597121880136e-06,
+      "loss": 0.0493,
+      "step": 4222
+    },
+    {
+      "epoch": 0.9597727272727272,
+      "grad_norm": 1.5903512170252132,
+      "learning_rate": 1.1398091097954433e-06,
+      "loss": 0.0549,
+      "step": 4223
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.811624339924134,
+      "learning_rate": 1.139758496905017e-06,
+      "loss": 0.0551,
+      "step": 4224
+    },
+    {
+      "epoch": 0.9602272727272727,
+      "grad_norm": 2.119880197737849,
+      "learning_rate": 1.1397078735177673e-06,
+      "loss": 0.0647,
+      "step": 4225
+    },
+    {
+      "epoch": 0.9604545454545454,
+      "grad_norm": 1.812998372280847,
+      "learning_rate": 1.1396572396347257e-06,
+      "loss": 0.0645,
+      "step": 4226
+    },
+    {
+      "epoch": 0.9606818181818182,
+      "grad_norm": 3.3033364845623545,
+      "learning_rate": 1.1396065952569254e-06,
+      "loss": 0.1572,
+      "step": 4227
+    },
+    {
+      "epoch": 0.9609090909090909,
+      "grad_norm": 2.3913336956433513,
+      "learning_rate": 1.1395559403853986e-06,
+      "loss": 0.0764,
+      "step": 4228
+    },
+    {
+      "epoch": 0.9611363636363637,
+      "grad_norm": 1.5221679903661862,
+      "learning_rate": 1.1395052750211786e-06,
+      "loss": 0.067,
+      "step": 4229
+    },
+    {
+      "epoch": 0.9613636363636363,
+      "grad_norm": 1.8783922926951706,
+      "learning_rate": 1.1394545991652983e-06,
+      "loss": 0.1428,
+      "step": 4230
+    },
+    {
+      "epoch": 0.961590909090909,
+      "grad_norm": 1.4406203247741542,
+      "learning_rate": 1.1394039128187914e-06,
+      "loss": 0.0627,
+      "step": 4231
+    },
+    {
+      "epoch": 0.9618181818181818,
+      "grad_norm": 1.3426432525410104,
+      "learning_rate": 1.1393532159826913e-06,
+      "loss": 0.0765,
+      "step": 4232
+    },
+    {
+      "epoch": 0.9620454545454545,
+      "grad_norm": 1.1837097111134178,
+      "learning_rate": 1.1393025086580316e-06,
+      "loss": 0.0641,
+      "step": 4233
+    },
+    {
+      "epoch": 0.9622727272727273,
+      "grad_norm": 2.021812935094423,
+      "learning_rate": 1.1392517908458465e-06,
+      "loss": 0.1165,
+      "step": 4234
+    },
+    {
+      "epoch": 0.9625,
+      "grad_norm": 3.43362606500914,
+      "learning_rate": 1.1392010625471702e-06,
+      "loss": 0.1453,
+      "step": 4235
+    },
+    {
+      "epoch": 0.9627272727272728,
+      "grad_norm": 2.414494441447142,
+      "learning_rate": 1.1391503237630375e-06,
+      "loss": 0.08,
+      "step": 4236
+    },
+    {
+      "epoch": 0.9629545454545455,
+      "grad_norm": 1.630639215849712,
+      "learning_rate": 1.1390995744944823e-06,
+      "loss": 0.054,
+      "step": 4237
+    },
+    {
+      "epoch": 0.9631818181818181,
+      "grad_norm": 2.0738371825332247,
+      "learning_rate": 1.1390488147425403e-06,
+      "loss": 0.1173,
+      "step": 4238
+    },
+    {
+      "epoch": 0.9634090909090909,
+      "grad_norm": 1.8996042973813274,
+      "learning_rate": 1.138998044508246e-06,
+      "loss": 0.0671,
+      "step": 4239
+    },
+    {
+      "epoch": 0.9636363636363636,
+      "grad_norm": 1.076906919876832,
+      "learning_rate": 1.1389472637926346e-06,
+      "loss": 0.0381,
+      "step": 4240
+    },
+    {
+      "epoch": 0.9638636363636364,
+      "grad_norm": 1.8734124032072546,
+      "learning_rate": 1.1388964725967423e-06,
+      "loss": 0.0713,
+      "step": 4241
+    },
+    {
+      "epoch": 0.9640909090909091,
+      "grad_norm": 1.745340937370693,
+      "learning_rate": 1.138845670921604e-06,
+      "loss": 0.1194,
+      "step": 4242
+    },
+    {
+      "epoch": 0.9643181818181819,
+      "grad_norm": 2.0841283901214633,
+      "learning_rate": 1.1387948587682561e-06,
+      "loss": 0.0637,
+      "step": 4243
+    },
+    {
+      "epoch": 0.9645454545454546,
+      "grad_norm": 1.0977852849534613,
+      "learning_rate": 1.1387440361377346e-06,
+      "loss": 0.0593,
+      "step": 4244
+    },
+    {
+      "epoch": 0.9647727272727272,
+      "grad_norm": 2.0809123357871053,
+      "learning_rate": 1.1386932030310762e-06,
+      "loss": 0.0772,
+      "step": 4245
+    },
+    {
+      "epoch": 0.965,
+      "grad_norm": 1.5700371953967343,
+      "learning_rate": 1.138642359449317e-06,
+      "loss": 0.0752,
+      "step": 4246
+    },
+    {
+      "epoch": 0.9652272727272727,
+      "grad_norm": 2.5120750104219196,
+      "learning_rate": 1.138591505393494e-06,
+      "loss": 0.1289,
+      "step": 4247
+    },
+    {
+      "epoch": 0.9654545454545455,
+      "grad_norm": 2.0916168342723416,
+      "learning_rate": 1.1385406408646443e-06,
+      "loss": 0.0722,
+      "step": 4248
+    },
+    {
+      "epoch": 0.9656818181818182,
+      "grad_norm": 1.194015398919941,
+      "learning_rate": 1.1384897658638049e-06,
+      "loss": 0.0576,
+      "step": 4249
+    },
+    {
+      "epoch": 0.9659090909090909,
+      "grad_norm": 2.027044293610933,
+      "learning_rate": 1.1384388803920134e-06,
+      "loss": 0.0929,
+      "step": 4250
+    },
+    {
+      "epoch": 0.9661363636363637,
+      "grad_norm": 2.558394754173468,
+      "learning_rate": 1.1383879844503073e-06,
+      "loss": 0.1209,
+      "step": 4251
+    },
+    {
+      "epoch": 0.9663636363636363,
+      "grad_norm": 1.8637436247305066,
+      "learning_rate": 1.1383370780397246e-06,
+      "loss": 0.1465,
+      "step": 4252
+    },
+    {
+      "epoch": 0.9665909090909091,
+      "grad_norm": 2.3988232024436464,
+      "learning_rate": 1.1382861611613033e-06,
+      "loss": 0.1252,
+      "step": 4253
+    },
+    {
+      "epoch": 0.9668181818181818,
+      "grad_norm": 2.214143957042044,
+      "learning_rate": 1.138235233816082e-06,
+      "loss": 0.0807,
+      "step": 4254
+    },
+    {
+      "epoch": 0.9670454545454545,
+      "grad_norm": 1.0144965136248965,
+      "learning_rate": 1.1381842960050985e-06,
+      "loss": 0.038,
+      "step": 4255
+    },
+    {
+      "epoch": 0.9672727272727273,
+      "grad_norm": 1.5705124319523491,
+      "learning_rate": 1.1381333477293918e-06,
+      "loss": 0.0883,
+      "step": 4256
+    },
+    {
+      "epoch": 0.9675,
+      "grad_norm": 1.6623455578601383,
+      "learning_rate": 1.1380823889900013e-06,
+      "loss": 0.0999,
+      "step": 4257
+    },
+    {
+      "epoch": 0.9677272727272728,
+      "grad_norm": 0.4978114157472213,
+      "learning_rate": 1.1380314197879654e-06,
+      "loss": 0.0339,
+      "step": 4258
+    },
+    {
+      "epoch": 0.9679545454545454,
+      "grad_norm": 1.575778800121303,
+      "learning_rate": 1.137980440124324e-06,
+      "loss": 0.0615,
+      "step": 4259
+    },
+    {
+      "epoch": 0.9681818181818181,
+      "grad_norm": 2.085077233662171,
+      "learning_rate": 1.1379294500001167e-06,
+      "loss": 0.1061,
+      "step": 4260
+    },
+    {
+      "epoch": 0.9684090909090909,
+      "grad_norm": 2.914987323441988,
+      "learning_rate": 1.137878449416383e-06,
+      "loss": 0.1715,
+      "step": 4261
+    },
+    {
+      "epoch": 0.9686363636363636,
+      "grad_norm": 2.6008369670041023,
+      "learning_rate": 1.1378274383741628e-06,
+      "loss": 0.0908,
+      "step": 4262
+    },
+    {
+      "epoch": 0.9688636363636364,
+      "grad_norm": 2.2225011995335513,
+      "learning_rate": 1.1377764168744964e-06,
+      "loss": 0.1073,
+      "step": 4263
+    },
+    {
+      "epoch": 0.9690909090909091,
+      "grad_norm": 1.8426143769231054,
+      "learning_rate": 1.1377253849184244e-06,
+      "loss": 0.0783,
+      "step": 4264
+    },
+    {
+      "epoch": 0.9693181818181819,
+      "grad_norm": 2.3344463521478476,
+      "learning_rate": 1.1376743425069871e-06,
+      "loss": 0.1462,
+      "step": 4265
+    },
+    {
+      "epoch": 0.9695454545454546,
+      "grad_norm": 1.446851287385785,
+      "learning_rate": 1.137623289641226e-06,
+      "loss": 0.046,
+      "step": 4266
+    },
+    {
+      "epoch": 0.9697727272727272,
+      "grad_norm": 1.5106473728110372,
+      "learning_rate": 1.1375722263221812e-06,
+      "loss": 0.0766,
+      "step": 4267
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6571546926794984,
+      "learning_rate": 1.1375211525508948e-06,
+      "loss": 0.0507,
+      "step": 4268
+    },
+    {
+      "epoch": 0.9702272727272727,
+      "grad_norm": 2.451109622759049,
+      "learning_rate": 1.1374700683284078e-06,
+      "loss": 0.0662,
+      "step": 4269
+    },
+    {
+      "epoch": 0.9704545454545455,
+      "grad_norm": 1.9402692610052594,
+      "learning_rate": 1.1374189736557621e-06,
+      "loss": 0.1083,
+      "step": 4270
+    },
+    {
+      "epoch": 0.9706818181818182,
+      "grad_norm": 2.039219148899409,
+      "learning_rate": 1.1373678685339994e-06,
+      "loss": 0.0464,
+      "step": 4271
+    },
+    {
+      "epoch": 0.9709090909090909,
+      "grad_norm": 1.2145488619857463,
+      "learning_rate": 1.1373167529641625e-06,
+      "loss": 0.0396,
+      "step": 4272
+    },
+    {
+      "epoch": 0.9711363636363637,
+      "grad_norm": 2.0144367247111075,
+      "learning_rate": 1.1372656269472926e-06,
+      "loss": 0.0554,
+      "step": 4273
+    },
+    {
+      "epoch": 0.9713636363636363,
+      "grad_norm": 1.8755521848418497,
+      "learning_rate": 1.1372144904844331e-06,
+      "loss": 0.0888,
+      "step": 4274
+    },
+    {
+      "epoch": 0.9715909090909091,
+      "grad_norm": 2.3536933269702565,
+      "learning_rate": 1.1371633435766266e-06,
+      "loss": 0.1138,
+      "step": 4275
+    },
+    {
+      "epoch": 0.9718181818181818,
+      "grad_norm": 1.7609097302571808,
+      "learning_rate": 1.137112186224916e-06,
+      "loss": 0.0697,
+      "step": 4276
+    },
+    {
+      "epoch": 0.9720454545454545,
+      "grad_norm": 2.1116399964552093,
+      "learning_rate": 1.1370610184303445e-06,
+      "loss": 0.0929,
+      "step": 4277
+    },
+    {
+      "epoch": 0.9722727272727273,
+      "grad_norm": 1.331438526807452,
+      "learning_rate": 1.1370098401939555e-06,
+      "loss": 0.0753,
+      "step": 4278
+    },
+    {
+      "epoch": 0.9725,
+      "grad_norm": 1.614029344362771,
+      "learning_rate": 1.1369586515167925e-06,
+      "loss": 0.0919,
+      "step": 4279
+    },
+    {
+      "epoch": 0.9727272727272728,
+      "grad_norm": 1.7565570755831104,
+      "learning_rate": 1.1369074523998997e-06,
+      "loss": 0.1201,
+      "step": 4280
+    },
+    {
+      "epoch": 0.9729545454545454,
+      "grad_norm": 1.3972135532857344,
+      "learning_rate": 1.1368562428443205e-06,
+      "loss": 0.0594,
+      "step": 4281
+    },
+    {
+      "epoch": 0.9731818181818181,
+      "grad_norm": 1.817259074025937,
+      "learning_rate": 1.1368050228510998e-06,
+      "loss": 0.0992,
+      "step": 4282
+    },
+    {
+      "epoch": 0.9734090909090909,
+      "grad_norm": 1.3940930439052885,
+      "learning_rate": 1.1367537924212817e-06,
+      "loss": 0.0719,
+      "step": 4283
+    },
+    {
+      "epoch": 0.9736363636363636,
+      "grad_norm": 1.4284729072621676,
+      "learning_rate": 1.1367025515559108e-06,
+      "loss": 0.0825,
+      "step": 4284
+    },
+    {
+      "epoch": 0.9738636363636364,
+      "grad_norm": 1.6321989994842947,
+      "learning_rate": 1.1366513002560324e-06,
+      "loss": 0.0757,
+      "step": 4285
+    },
+    {
+      "epoch": 0.9740909090909091,
+      "grad_norm": 2.7215773549188946,
+      "learning_rate": 1.1366000385226913e-06,
+      "loss": 0.1364,
+      "step": 4286
+    },
+    {
+      "epoch": 0.9743181818181819,
+      "grad_norm": 2.1105843992473243,
+      "learning_rate": 1.136548766356933e-06,
+      "loss": 0.1206,
+      "step": 4287
+    },
+    {
+      "epoch": 0.9745454545454545,
+      "grad_norm": 1.1634435300258883,
+      "learning_rate": 1.1364974837598029e-06,
+      "loss": 0.0462,
+      "step": 4288
+    },
+    {
+      "epoch": 0.9747727272727272,
+      "grad_norm": 1.2099534428116623,
+      "learning_rate": 1.1364461907323465e-06,
+      "loss": 0.0611,
+      "step": 4289
+    },
+    {
+      "epoch": 0.975,
+      "grad_norm": 1.5920023462523307,
+      "learning_rate": 1.1363948872756101e-06,
+      "loss": 0.0594,
+      "step": 4290
+    },
+    {
+      "epoch": 0.9752272727272727,
+      "grad_norm": 1.7120153873328574,
+      "learning_rate": 1.1363435733906398e-06,
+      "loss": 0.1176,
+      "step": 4291
+    },
+    {
+      "epoch": 0.9754545454545455,
+      "grad_norm": 1.0831410269186577,
+      "learning_rate": 1.136292249078482e-06,
+      "loss": 0.034,
+      "step": 4292
+    },
+    {
+      "epoch": 0.9756818181818182,
+      "grad_norm": 1.4261614696930718,
+      "learning_rate": 1.136240914340183e-06,
+      "loss": 0.0468,
+      "step": 4293
+    },
+    {
+      "epoch": 0.975909090909091,
+      "grad_norm": 1.6593865465573472,
+      "learning_rate": 1.1361895691767902e-06,
+      "loss": 0.0872,
+      "step": 4294
+    },
+    {
+      "epoch": 0.9761363636363637,
+      "grad_norm": 2.4141948187678586,
+      "learning_rate": 1.13613821358935e-06,
+      "loss": 0.1344,
+      "step": 4295
+    },
+    {
+      "epoch": 0.9763636363636363,
+      "grad_norm": 1.649159828912975,
+      "learning_rate": 1.13608684757891e-06,
+      "loss": 0.0749,
+      "step": 4296
+    },
+    {
+      "epoch": 0.9765909090909091,
+      "grad_norm": 2.357881369665485,
+      "learning_rate": 1.1360354711465177e-06,
+      "loss": 0.0853,
+      "step": 4297
+    },
+    {
+      "epoch": 0.9768181818181818,
+      "grad_norm": 1.2641122692742646,
+      "learning_rate": 1.1359840842932204e-06,
+      "loss": 0.0464,
+      "step": 4298
+    },
+    {
+      "epoch": 0.9770454545454546,
+      "grad_norm": 1.7156367919564661,
+      "learning_rate": 1.135932687020066e-06,
+      "loss": 0.0804,
+      "step": 4299
+    },
+    {
+      "epoch": 0.9772727272727273,
+      "grad_norm": 1.4603663304481989,
+      "learning_rate": 1.135881279328103e-06,
+      "loss": 0.0569,
+      "step": 4300
+    },
+    {
+      "epoch": 0.9775,
+      "grad_norm": 2.1269177464321367,
+      "learning_rate": 1.1358298612183793e-06,
+      "loss": 0.064,
+      "step": 4301
+    },
+    {
+      "epoch": 0.9777272727272728,
+      "grad_norm": 1.7931050554151522,
+      "learning_rate": 1.1357784326919439e-06,
+      "loss": 0.0757,
+      "step": 4302
+    },
+    {
+      "epoch": 0.9779545454545454,
+      "grad_norm": 1.5175713953766488,
+      "learning_rate": 1.1357269937498448e-06,
+      "loss": 0.0638,
+      "step": 4303
+    },
+    {
+      "epoch": 0.9781818181818182,
+      "grad_norm": 2.128476060151315,
+      "learning_rate": 1.1356755443931313e-06,
+      "loss": 0.1023,
+      "step": 4304
+    },
+    {
+      "epoch": 0.9784090909090909,
+      "grad_norm": 2.219715495324755,
+      "learning_rate": 1.1356240846228524e-06,
+      "loss": 0.1523,
+      "step": 4305
+    },
+    {
+      "epoch": 0.9786363636363636,
+      "grad_norm": 1.2024570771265481,
+      "learning_rate": 1.135572614440058e-06,
+      "loss": 0.0727,
+      "step": 4306
+    },
+    {
+      "epoch": 0.9788636363636364,
+      "grad_norm": 0.6058770607943758,
+      "learning_rate": 1.135521133845797e-06,
+      "loss": 0.0343,
+      "step": 4307
+    },
+    {
+      "epoch": 0.9790909090909091,
+      "grad_norm": 2.108632415706878,
+      "learning_rate": 1.1354696428411193e-06,
+      "loss": 0.1176,
+      "step": 4308
+    },
+    {
+      "epoch": 0.9793181818181819,
+      "grad_norm": 1.3782004131483923,
+      "learning_rate": 1.135418141427075e-06,
+      "loss": 0.0375,
+      "step": 4309
+    },
+    {
+      "epoch": 0.9795454545454545,
+      "grad_norm": 1.6893872723082506,
+      "learning_rate": 1.1353666296047148e-06,
+      "loss": 0.0539,
+      "step": 4310
+    },
+    {
+      "epoch": 0.9797727272727272,
+      "grad_norm": 1.4606354698451325,
+      "learning_rate": 1.1353151073750882e-06,
+      "loss": 0.0691,
+      "step": 4311
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.497208513934938,
+      "learning_rate": 1.1352635747392465e-06,
+      "loss": 0.0972,
+      "step": 4312
+    },
+    {
+      "epoch": 0.9802272727272727,
+      "grad_norm": 2.3473856031255247,
+      "learning_rate": 1.13521203169824e-06,
+      "loss": 0.098,
+      "step": 4313
+    },
+    {
+      "epoch": 0.9804545454545455,
+      "grad_norm": 1.2849247602886635,
+      "learning_rate": 1.1351604782531203e-06,
+      "loss": 0.0423,
+      "step": 4314
+    },
+    {
+      "epoch": 0.9806818181818182,
+      "grad_norm": 1.5438396179864842,
+      "learning_rate": 1.1351089144049384e-06,
+      "loss": 0.0814,
+      "step": 4315
+    },
+    {
+      "epoch": 0.980909090909091,
+      "grad_norm": 1.7111590923646158,
+      "learning_rate": 1.1350573401547457e-06,
+      "loss": 0.0672,
+      "step": 4316
+    },
+    {
+      "epoch": 0.9811363636363636,
+      "grad_norm": 1.9928245889786709,
+      "learning_rate": 1.1350057555035943e-06,
+      "loss": 0.1098,
+      "step": 4317
+    },
+    {
+      "epoch": 0.9813636363636363,
+      "grad_norm": 0.5336042618357226,
+      "learning_rate": 1.1349541604525355e-06,
+      "loss": 0.0318,
+      "step": 4318
+    },
+    {
+      "epoch": 0.9815909090909091,
+      "grad_norm": 2.236681079991919,
+      "learning_rate": 1.1349025550026215e-06,
+      "loss": 0.1353,
+      "step": 4319
+    },
+    {
+      "epoch": 0.9818181818181818,
+      "grad_norm": 1.4611129047086024,
+      "learning_rate": 1.1348509391549053e-06,
+      "loss": 0.0678,
+      "step": 4320
+    },
+    {
+      "epoch": 0.9820454545454546,
+      "grad_norm": 1.8300624607621967,
+      "learning_rate": 1.1347993129104386e-06,
+      "loss": 0.1038,
+      "step": 4321
+    },
+    {
+      "epoch": 0.9822727272727273,
+      "grad_norm": 1.5575697361221585,
+      "learning_rate": 1.1347476762702747e-06,
+      "loss": 0.0745,
+      "step": 4322
+    },
+    {
+      "epoch": 0.9825,
+      "grad_norm": 1.3449653638022425,
+      "learning_rate": 1.1346960292354665e-06,
+      "loss": 0.0463,
+      "step": 4323
+    },
+    {
+      "epoch": 0.9827272727272728,
+      "grad_norm": 1.910971985040669,
+      "learning_rate": 1.1346443718070668e-06,
+      "loss": 0.0851,
+      "step": 4324
+    },
+    {
+      "epoch": 0.9829545454545454,
+      "grad_norm": 1.4219845394472848,
+      "learning_rate": 1.1345927039861293e-06,
+      "loss": 0.0436,
+      "step": 4325
+    },
+    {
+      "epoch": 0.9831818181818182,
+      "grad_norm": 1.7020360153394452,
+      "learning_rate": 1.1345410257737078e-06,
+      "loss": 0.0752,
+      "step": 4326
+    },
+    {
+      "epoch": 0.9834090909090909,
+      "grad_norm": 1.3774716736815111,
+      "learning_rate": 1.1344893371708555e-06,
+      "loss": 0.0514,
+      "step": 4327
+    },
+    {
+      "epoch": 0.9836363636363636,
+      "grad_norm": 1.5607267034352892,
+      "learning_rate": 1.1344376381786268e-06,
+      "loss": 0.0736,
+      "step": 4328
+    },
+    {
+      "epoch": 0.9838636363636364,
+      "grad_norm": 2.1410094640162054,
+      "learning_rate": 1.134385928798076e-06,
+      "loss": 0.0675,
+      "step": 4329
+    },
+    {
+      "epoch": 0.9840909090909091,
+      "grad_norm": 1.9246316000546906,
+      "learning_rate": 1.1343342090302575e-06,
+      "loss": 0.0915,
+      "step": 4330
+    },
+    {
+      "epoch": 0.9843181818181819,
+      "grad_norm": 3.0301716501830995,
+      "learning_rate": 1.1342824788762258e-06,
+      "loss": 0.1077,
+      "step": 4331
+    },
+    {
+      "epoch": 0.9845454545454545,
+      "grad_norm": 1.469821233021225,
+      "learning_rate": 1.1342307383370359e-06,
+      "loss": 0.0979,
+      "step": 4332
+    },
+    {
+      "epoch": 0.9847727272727272,
+      "grad_norm": 1.1460330604595226,
+      "learning_rate": 1.1341789874137428e-06,
+      "loss": 0.0477,
+      "step": 4333
+    },
+    {
+      "epoch": 0.985,
+      "grad_norm": 2.306325783748454,
+      "learning_rate": 1.1341272261074018e-06,
+      "loss": 0.1297,
+      "step": 4334
+    },
+    {
+      "epoch": 0.9852272727272727,
+      "grad_norm": 1.319536000244395,
+      "learning_rate": 1.1340754544190685e-06,
+      "loss": 0.0807,
+      "step": 4335
+    },
+    {
+      "epoch": 0.9854545454545455,
+      "grad_norm": 1.5136499455444101,
+      "learning_rate": 1.1340236723497985e-06,
+      "loss": 0.0893,
+      "step": 4336
+    },
+    {
+      "epoch": 0.9856818181818182,
+      "grad_norm": 1.8754320606221733,
+      "learning_rate": 1.1339718799006478e-06,
+      "loss": 0.089,
+      "step": 4337
+    },
+    {
+      "epoch": 0.985909090909091,
+      "grad_norm": 1.7822785028735206,
+      "learning_rate": 1.1339200770726726e-06,
+      "loss": 0.0624,
+      "step": 4338
+    },
+    {
+      "epoch": 0.9861363636363636,
+      "grad_norm": 1.6194957496226232,
+      "learning_rate": 1.1338682638669291e-06,
+      "loss": 0.0938,
+      "step": 4339
+    },
+    {
+      "epoch": 0.9863636363636363,
+      "grad_norm": 2.5634667996366645,
+      "learning_rate": 1.1338164402844739e-06,
+      "loss": 0.1056,
+      "step": 4340
+    },
+    {
+      "epoch": 0.9865909090909091,
+      "grad_norm": 1.814064596024776,
+      "learning_rate": 1.133764606326364e-06,
+      "loss": 0.1023,
+      "step": 4341
+    },
+    {
+      "epoch": 0.9868181818181818,
+      "grad_norm": 1.7952384887443167,
+      "learning_rate": 1.133712761993656e-06,
+      "loss": 0.0643,
+      "step": 4342
+    },
+    {
+      "epoch": 0.9870454545454546,
+      "grad_norm": 1.6655656741536045,
+      "learning_rate": 1.1336609072874072e-06,
+      "loss": 0.0635,
+      "step": 4343
+    },
+    {
+      "epoch": 0.9872727272727273,
+      "grad_norm": 1.2114931960996036,
+      "learning_rate": 1.133609042208675e-06,
+      "loss": 0.055,
+      "step": 4344
+    },
+    {
+      "epoch": 0.9875,
+      "grad_norm": 1.7867390454776044,
+      "learning_rate": 1.1335571667585175e-06,
+      "loss": 0.069,
+      "step": 4345
+    },
+    {
+      "epoch": 0.9877272727272727,
+      "grad_norm": 1.8451366289833822,
+      "learning_rate": 1.1335052809379921e-06,
+      "loss": 0.064,
+      "step": 4346
+    },
+    {
+      "epoch": 0.9879545454545454,
+      "grad_norm": 2.1444247512042867,
+      "learning_rate": 1.1334533847481566e-06,
+      "loss": 0.0932,
+      "step": 4347
+    },
+    {
+      "epoch": 0.9881818181818182,
+      "grad_norm": 1.6779813046360383,
+      "learning_rate": 1.1334014781900696e-06,
+      "loss": 0.1114,
+      "step": 4348
+    },
+    {
+      "epoch": 0.9884090909090909,
+      "grad_norm": 1.6532985861679539,
+      "learning_rate": 1.1333495612647896e-06,
+      "loss": 0.044,
+      "step": 4349
+    },
+    {
+      "epoch": 0.9886363636363636,
+      "grad_norm": 1.5090893171441755,
+      "learning_rate": 1.1332976339733753e-06,
+      "loss": 0.091,
+      "step": 4350
+    },
+    {
+      "epoch": 0.9888636363636364,
+      "grad_norm": 2.0940410656334505,
+      "learning_rate": 1.1332456963168854e-06,
+      "loss": 0.0784,
+      "step": 4351
+    },
+    {
+      "epoch": 0.9890909090909091,
+      "grad_norm": 2.491248740340114,
+      "learning_rate": 1.133193748296379e-06,
+      "loss": 0.1093,
+      "step": 4352
+    },
+    {
+      "epoch": 0.9893181818181818,
+      "grad_norm": 1.7243680617375707,
+      "learning_rate": 1.1331417899129155e-06,
+      "loss": 0.107,
+      "step": 4353
+    },
+    {
+      "epoch": 0.9895454545454545,
+      "grad_norm": 2.999777617114768,
+      "learning_rate": 1.1330898211675544e-06,
+      "loss": 0.1609,
+      "step": 4354
+    },
+    {
+      "epoch": 0.9897727272727272,
+      "grad_norm": 1.2271535275893963,
+      "learning_rate": 1.1330378420613553e-06,
+      "loss": 0.0728,
+      "step": 4355
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7775840532219367,
+      "learning_rate": 1.1329858525953785e-06,
+      "loss": 0.1154,
+      "step": 4356
+    },
+    {
+      "epoch": 0.9902272727272727,
+      "grad_norm": 2.782665107953299,
+      "learning_rate": 1.1329338527706837e-06,
+      "loss": 0.1402,
+      "step": 4357
+    },
+    {
+      "epoch": 0.9904545454545455,
+      "grad_norm": 2.0649260185530345,
+      "learning_rate": 1.1328818425883317e-06,
+      "loss": 0.0952,
+      "step": 4358
+    },
+    {
+      "epoch": 0.9906818181818182,
+      "grad_norm": 1.931992128077991,
+      "learning_rate": 1.1328298220493828e-06,
+      "loss": 0.1,
+      "step": 4359
+    },
+    {
+      "epoch": 0.990909090909091,
+      "grad_norm": 1.8072363772703106,
+      "learning_rate": 1.132777791154898e-06,
+      "loss": 0.074,
+      "step": 4360
+    },
+    {
+      "epoch": 0.9911363636363636,
+      "grad_norm": 0.990220970317101,
+      "learning_rate": 1.132725749905938e-06,
+      "loss": 0.0401,
+      "step": 4361
+    },
+    {
+      "epoch": 0.9913636363636363,
+      "grad_norm": 1.7102784121857937,
+      "learning_rate": 1.1326736983035644e-06,
+      "loss": 0.0663,
+      "step": 4362
+    },
+    {
+      "epoch": 0.9915909090909091,
+      "grad_norm": 2.228420401265974,
+      "learning_rate": 1.1326216363488381e-06,
+      "loss": 0.1103,
+      "step": 4363
+    },
+    {
+      "epoch": 0.9918181818181818,
+      "grad_norm": 2.3863178367410063,
+      "learning_rate": 1.1325695640428214e-06,
+      "loss": 0.1138,
+      "step": 4364
+    },
+    {
+      "epoch": 0.9920454545454546,
+      "grad_norm": 2.3407525445084874,
+      "learning_rate": 1.1325174813865756e-06,
+      "loss": 0.1053,
+      "step": 4365
+    },
+    {
+      "epoch": 0.9922727272727273,
+      "grad_norm": 1.5317367976945508,
+      "learning_rate": 1.132465388381163e-06,
+      "loss": 0.0924,
+      "step": 4366
+    },
+    {
+      "epoch": 0.9925,
+      "grad_norm": 2.671177014060101,
+      "learning_rate": 1.1324132850276454e-06,
+      "loss": 0.1771,
+      "step": 4367
+    },
+    {
+      "epoch": 0.9927272727272727,
+      "grad_norm": 3.1686244575604716,
+      "learning_rate": 1.1323611713270862e-06,
+      "loss": 0.121,
+      "step": 4368
+    },
+    {
+      "epoch": 0.9929545454545454,
+      "grad_norm": 1.870931564256806,
+      "learning_rate": 1.1323090472805473e-06,
+      "loss": 0.0717,
+      "step": 4369
+    },
+    {
+      "epoch": 0.9931818181818182,
+      "grad_norm": 2.623536867080072,
+      "learning_rate": 1.132256912889092e-06,
+      "loss": 0.1336,
+      "step": 4370
+    },
+    {
+      "epoch": 0.9934090909090909,
+      "grad_norm": 1.834660373466206,
+      "learning_rate": 1.1322047681537834e-06,
+      "loss": 0.062,
+      "step": 4371
+    },
+    {
+      "epoch": 0.9936363636363637,
+      "grad_norm": 1.5491338887973025,
+      "learning_rate": 1.1321526130756844e-06,
+      "loss": 0.0785,
+      "step": 4372
+    },
+    {
+      "epoch": 0.9938636363636364,
+      "grad_norm": 2.432604677887877,
+      "learning_rate": 1.1321004476558593e-06,
+      "loss": 0.0966,
+      "step": 4373
+    },
+    {
+      "epoch": 0.9940909090909091,
+      "grad_norm": 2.3905989077575054,
+      "learning_rate": 1.132048271895371e-06,
+      "loss": 0.0879,
+      "step": 4374
+    },
+    {
+      "epoch": 0.9943181818181818,
+      "grad_norm": 2.067213586305291,
+      "learning_rate": 1.131996085795284e-06,
+      "loss": 0.0847,
+      "step": 4375
+    },
+    {
+      "epoch": 0.9945454545454545,
+      "grad_norm": 1.303483758217693,
+      "learning_rate": 1.131943889356662e-06,
+      "loss": 0.049,
+      "step": 4376
+    },
+    {
+      "epoch": 0.9947727272727273,
+      "grad_norm": 2.0864156006382326,
+      "learning_rate": 1.1318916825805702e-06,
+      "loss": 0.1267,
+      "step": 4377
+    },
+    {
+      "epoch": 0.995,
+      "grad_norm": 2.416811507822547,
+      "learning_rate": 1.1318394654680723e-06,
+      "loss": 0.092,
+      "step": 4378
+    },
+    {
+      "epoch": 0.9952272727272727,
+      "grad_norm": 1.48194817368532,
+      "learning_rate": 1.1317872380202335e-06,
+      "loss": 0.0593,
+      "step": 4379
+    },
+    {
+      "epoch": 0.9954545454545455,
+      "grad_norm": 2.3307114845077046,
+      "learning_rate": 1.131735000238119e-06,
+      "loss": 0.0892,
+      "step": 4380
+    },
+    {
+      "epoch": 0.9956818181818182,
+      "grad_norm": 0.577643539957353,
+      "learning_rate": 1.1316827521227935e-06,
+      "loss": 0.0365,
+      "step": 4381
+    },
+    {
+      "epoch": 0.9959090909090909,
+      "grad_norm": 1.756862652471502,
+      "learning_rate": 1.131630493675323e-06,
+      "loss": 0.1019,
+      "step": 4382
+    },
+    {
+      "epoch": 0.9961363636363636,
+      "grad_norm": 2.283798732254271,
+      "learning_rate": 1.1315782248967727e-06,
+      "loss": 0.0737,
+      "step": 4383
+    },
+    {
+      "epoch": 0.9963636363636363,
+      "grad_norm": 1.4432216870497228,
+      "learning_rate": 1.1315259457882085e-06,
+      "loss": 0.068,
+      "step": 4384
+    },
+    {
+      "epoch": 0.9965909090909091,
+      "grad_norm": 1.9207065941938133,
+      "learning_rate": 1.1314736563506969e-06,
+      "loss": 0.095,
+      "step": 4385
+    },
+    {
+      "epoch": 0.9968181818181818,
+      "grad_norm": 0.7971348239093567,
+      "learning_rate": 1.1314213565853036e-06,
+      "loss": 0.0479,
+      "step": 4386
+    },
+    {
+      "epoch": 0.9970454545454546,
+      "grad_norm": 2.5498190214482546,
+      "learning_rate": 1.1313690464930956e-06,
+      "loss": 0.1036,
+      "step": 4387
+    },
+    {
+      "epoch": 0.9972727272727273,
+      "grad_norm": 1.8963598637281185,
+      "learning_rate": 1.1313167260751392e-06,
+      "loss": 0.0779,
+      "step": 4388
+    },
+    {
+      "epoch": 0.9975,
+      "grad_norm": 1.8741121967767678,
+      "learning_rate": 1.1312643953325013e-06,
+      "loss": 0.0803,
+      "step": 4389
+    },
+    {
+      "epoch": 0.9977272727272727,
+      "grad_norm": 1.6427343228699114,
+      "learning_rate": 1.1312120542662493e-06,
+      "loss": 0.0513,
+      "step": 4390
+    },
+    {
+      "epoch": 0.9979545454545454,
+      "grad_norm": 1.2549116489545877,
+      "learning_rate": 1.1311597028774503e-06,
+      "loss": 0.0778,
+      "step": 4391
+    },
+    {
+      "epoch": 0.9981818181818182,
+      "grad_norm": 2.26811625417695,
+      "learning_rate": 1.1311073411671718e-06,
+      "loss": 0.1097,
+      "step": 4392
+    },
+    {
+      "epoch": 0.9984090909090909,
+      "grad_norm": 1.8836464776794812,
+      "learning_rate": 1.1310549691364819e-06,
+      "loss": 0.0889,
+      "step": 4393
+    },
+    {
+      "epoch": 0.9986363636363637,
+      "grad_norm": 2.7786558956477427,
+      "learning_rate": 1.1310025867864482e-06,
+      "loss": 0.1026,
+      "step": 4394
+    },
+    {
+      "epoch": 0.9988636363636364,
+      "grad_norm": 1.6062383970993732,
+      "learning_rate": 1.130950194118139e-06,
+      "loss": 0.0761,
+      "step": 4395
+    },
+    {
+      "epoch": 0.9990909090909091,
+      "grad_norm": 1.4583447171714556,
+      "learning_rate": 1.1308977911326229e-06,
+      "loss": 0.0433,
+      "step": 4396
+    },
+    {
+      "epoch": 0.9993181818181818,
+      "grad_norm": 2.125013607693135,
+      "learning_rate": 1.1308453778309681e-06,
+      "loss": 0.1091,
+      "step": 4397
+    },
+    {
+      "epoch": 0.9995454545454545,
+      "grad_norm": 2.1141086693331763,
+      "learning_rate": 1.1307929542142435e-06,
+      "loss": 0.1059,
+      "step": 4398
+    },
+    {
+      "epoch": 0.9997727272727273,
+      "grad_norm": 1.0452569154174431,
+      "learning_rate": 1.130740520283518e-06,
+      "loss": 0.0575,
+      "step": 4399
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9502283023578147,
+      "learning_rate": 1.1306880760398611e-06,
+      "loss": 0.0547,
+      "step": 4400
+    },
+    {
+      "epoch": 1.0002272727272727,
+      "grad_norm": 0.9425820283709417,
+      "learning_rate": 1.1306356214843423e-06,
+      "loss": 0.0302,
+      "step": 4401
+    },
+    {
+      "epoch": 1.0004545454545455,
+      "grad_norm": 1.1419421721852552,
+      "learning_rate": 1.130583156618031e-06,
+      "loss": 0.0355,
+      "step": 4402
+    },
+    {
+      "epoch": 1.0006818181818182,
+      "grad_norm": 0.7399963190873251,
+      "learning_rate": 1.1305306814419969e-06,
+      "loss": 0.0311,
+      "step": 4403
+    },
+    {
+      "epoch": 1.000909090909091,
+      "grad_norm": 1.0763134620694756,
+      "learning_rate": 1.1304781959573103e-06,
+      "loss": 0.0527,
+      "step": 4404
+    },
+    {
+      "epoch": 1.0011363636363637,
+      "grad_norm": 2.3540248565430093,
+      "learning_rate": 1.1304257001650415e-06,
+      "loss": 0.0713,
+      "step": 4405
+    },
+    {
+      "epoch": 1.0013636363636365,
+      "grad_norm": 1.5768777803046357,
+      "learning_rate": 1.1303731940662608e-06,
+      "loss": 0.0762,
+      "step": 4406
+    },
+    {
+      "epoch": 1.0015909090909092,
+      "grad_norm": 1.28321837107857,
+      "learning_rate": 1.130320677662039e-06,
+      "loss": 0.0358,
+      "step": 4407
+    },
+    {
+      "epoch": 1.0018181818181817,
+      "grad_norm": 1.1642494624362285,
+      "learning_rate": 1.1302681509534473e-06,
+      "loss": 0.0542,
+      "step": 4408
+    },
+    {
+      "epoch": 1.0020454545454545,
+      "grad_norm": 1.1146365953353923,
+      "learning_rate": 1.1302156139415563e-06,
+      "loss": 0.0314,
+      "step": 4409
+    },
+    {
+      "epoch": 1.0022727272727272,
+      "grad_norm": 1.4883636948809214,
+      "learning_rate": 1.1301630666274374e-06,
+      "loss": 0.0965,
+      "step": 4410
+    },
+    {
+      "epoch": 1.0025,
+      "grad_norm": 1.7548677800547257,
+      "learning_rate": 1.1301105090121624e-06,
+      "loss": 0.0521,
+      "step": 4411
+    },
+    {
+      "epoch": 1.0027272727272727,
+      "grad_norm": 1.437784059806614,
+      "learning_rate": 1.130057941096803e-06,
+      "loss": 0.049,
+      "step": 4412
+    },
+    {
+      "epoch": 1.0029545454545454,
+      "grad_norm": 1.4551877123322565,
+      "learning_rate": 1.1300053628824308e-06,
+      "loss": 0.0484,
+      "step": 4413
+    },
+    {
+      "epoch": 1.0031818181818182,
+      "grad_norm": 1.1912945173392573,
+      "learning_rate": 1.1299527743701184e-06,
+      "loss": 0.0604,
+      "step": 4414
+    },
+    {
+      "epoch": 1.003409090909091,
+      "grad_norm": 1.6583949130016453,
+      "learning_rate": 1.1299001755609381e-06,
+      "loss": 0.0829,
+      "step": 4415
+    },
+    {
+      "epoch": 1.0036363636363637,
+      "grad_norm": 2.114355085425396,
+      "learning_rate": 1.1298475664559622e-06,
+      "loss": 0.0978,
+      "step": 4416
+    },
+    {
+      "epoch": 1.0038636363636364,
+      "grad_norm": 1.6350764474439796,
+      "learning_rate": 1.1297949470562637e-06,
+      "loss": 0.0639,
+      "step": 4417
+    },
+    {
+      "epoch": 1.0040909090909091,
+      "grad_norm": 1.203332608422028,
+      "learning_rate": 1.1297423173629155e-06,
+      "loss": 0.057,
+      "step": 4418
+    },
+    {
+      "epoch": 1.0043181818181819,
+      "grad_norm": 1.4803870221865691,
+      "learning_rate": 1.1296896773769912e-06,
+      "loss": 0.0694,
+      "step": 4419
+    },
+    {
+      "epoch": 1.0045454545454546,
+      "grad_norm": 1.6559790163379575,
+      "learning_rate": 1.1296370270995635e-06,
+      "loss": 0.0781,
+      "step": 4420
+    },
+    {
+      "epoch": 1.0047727272727274,
+      "grad_norm": 1.3031382023760363,
+      "learning_rate": 1.1295843665317067e-06,
+      "loss": 0.0582,
+      "step": 4421
+    },
+    {
+      "epoch": 1.005,
+      "grad_norm": 0.5533701101020213,
+      "learning_rate": 1.1295316956744943e-06,
+      "loss": 0.0101,
+      "step": 4422
+    },
+    {
+      "epoch": 1.0052272727272726,
+      "grad_norm": 0.720011260623196,
+      "learning_rate": 1.1294790145290003e-06,
+      "loss": 0.0247,
+      "step": 4423
+    },
+    {
+      "epoch": 1.0054545454545454,
+      "grad_norm": 1.5886945733108813,
+      "learning_rate": 1.1294263230962992e-06,
+      "loss": 0.0328,
+      "step": 4424
+    },
+    {
+      "epoch": 1.0056818181818181,
+      "grad_norm": 1.0829454537871013,
+      "learning_rate": 1.1293736213774653e-06,
+      "loss": 0.0367,
+      "step": 4425
+    },
+    {
+      "epoch": 1.0059090909090909,
+      "grad_norm": 1.2793603225848462,
+      "learning_rate": 1.1293209093735732e-06,
+      "loss": 0.0158,
+      "step": 4426
+    },
+    {
+      "epoch": 1.0061363636363636,
+      "grad_norm": 1.1444714119422865,
+      "learning_rate": 1.1292681870856982e-06,
+      "loss": 0.0466,
+      "step": 4427
+    },
+    {
+      "epoch": 1.0063636363636363,
+      "grad_norm": 1.631167607301591,
+      "learning_rate": 1.129215454514915e-06,
+      "loss": 0.0625,
+      "step": 4428
+    },
+    {
+      "epoch": 1.006590909090909,
+      "grad_norm": 2.728998972396216,
+      "learning_rate": 1.129162711662299e-06,
+      "loss": 0.093,
+      "step": 4429
+    },
+    {
+      "epoch": 1.0068181818181818,
+      "grad_norm": 2.0133273446918727,
+      "learning_rate": 1.129109958528926e-06,
+      "loss": 0.045,
+      "step": 4430
+    },
+    {
+      "epoch": 1.0070454545454546,
+      "grad_norm": 1.8210916234392738,
+      "learning_rate": 1.129057195115871e-06,
+      "loss": 0.0527,
+      "step": 4431
+    },
+    {
+      "epoch": 1.0072727272727273,
+      "grad_norm": 1.2503655043801538,
+      "learning_rate": 1.1290044214242108e-06,
+      "loss": 0.0486,
+      "step": 4432
+    },
+    {
+      "epoch": 1.0075,
+      "grad_norm": 1.6967181682756363,
+      "learning_rate": 1.128951637455021e-06,
+      "loss": 0.0624,
+      "step": 4433
+    },
+    {
+      "epoch": 1.0077272727272728,
+      "grad_norm": 1.6215477095298543,
+      "learning_rate": 1.1288988432093781e-06,
+      "loss": 0.0467,
+      "step": 4434
+    },
+    {
+      "epoch": 1.0079545454545455,
+      "grad_norm": 1.4363233532748438,
+      "learning_rate": 1.1288460386883587e-06,
+      "loss": 0.0868,
+      "step": 4435
+    },
+    {
+      "epoch": 1.0081818181818183,
+      "grad_norm": 1.681227724645006,
+      "learning_rate": 1.1287932238930397e-06,
+      "loss": 0.0631,
+      "step": 4436
+    },
+    {
+      "epoch": 1.0084090909090908,
+      "grad_norm": 2.470727977368941,
+      "learning_rate": 1.1287403988244977e-06,
+      "loss": 0.0597,
+      "step": 4437
+    },
+    {
+      "epoch": 1.0086363636363636,
+      "grad_norm": 1.625112323938301,
+      "learning_rate": 1.1286875634838104e-06,
+      "loss": 0.0646,
+      "step": 4438
+    },
+    {
+      "epoch": 1.0088636363636363,
+      "grad_norm": 1.4756499716919682,
+      "learning_rate": 1.1286347178720546e-06,
+      "loss": 0.0348,
+      "step": 4439
+    },
+    {
+      "epoch": 1.009090909090909,
+      "grad_norm": 2.0338549054721593,
+      "learning_rate": 1.1285818619903086e-06,
+      "loss": 0.0317,
+      "step": 4440
+    },
+    {
+      "epoch": 1.0093181818181818,
+      "grad_norm": 2.0284314695359233,
+      "learning_rate": 1.12852899583965e-06,
+      "loss": 0.0964,
+      "step": 4441
+    },
+    {
+      "epoch": 1.0095454545454545,
+      "grad_norm": 1.9015441919023626,
+      "learning_rate": 1.1284761194211564e-06,
+      "loss": 0.0637,
+      "step": 4442
+    },
+    {
+      "epoch": 1.0097727272727273,
+      "grad_norm": 1.973004734028038,
+      "learning_rate": 1.1284232327359065e-06,
+      "loss": 0.0421,
+      "step": 4443
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.871656009933019,
+      "learning_rate": 1.1283703357849786e-06,
+      "loss": 0.041,
+      "step": 4444
+    },
+    {
+      "epoch": 1.0102272727272728,
+      "grad_norm": 1.7146551963601457,
+      "learning_rate": 1.1283174285694511e-06,
+      "loss": 0.0615,
+      "step": 4445
+    },
+    {
+      "epoch": 1.0104545454545455,
+      "grad_norm": 1.0638255063997892,
+      "learning_rate": 1.1282645110904036e-06,
+      "loss": 0.032,
+      "step": 4446
+    },
+    {
+      "epoch": 1.0106818181818182,
+      "grad_norm": 1.4421280074459613,
+      "learning_rate": 1.1282115833489147e-06,
+      "loss": 0.0332,
+      "step": 4447
+    },
+    {
+      "epoch": 1.010909090909091,
+      "grad_norm": 0.938559029273199,
+      "learning_rate": 1.1281586453460634e-06,
+      "loss": 0.0243,
+      "step": 4448
+    },
+    {
+      "epoch": 1.0111363636363637,
+      "grad_norm": 2.4168919267329785,
+      "learning_rate": 1.1281056970829297e-06,
+      "loss": 0.1162,
+      "step": 4449
+    },
+    {
+      "epoch": 1.0113636363636365,
+      "grad_norm": 1.2289906360410185,
+      "learning_rate": 1.128052738560593e-06,
+      "loss": 0.0587,
+      "step": 4450
+    },
+    {
+      "epoch": 1.011590909090909,
+      "grad_norm": 1.7439119243712977,
+      "learning_rate": 1.1279997697801334e-06,
+      "loss": 0.0789,
+      "step": 4451
+    },
+    {
+      "epoch": 1.0118181818181817,
+      "grad_norm": 1.4492910180996101,
+      "learning_rate": 1.1279467907426312e-06,
+      "loss": 0.0388,
+      "step": 4452
+    },
+    {
+      "epoch": 1.0120454545454545,
+      "grad_norm": 1.3050576065056476,
+      "learning_rate": 1.127893801449166e-06,
+      "loss": 0.0338,
+      "step": 4453
+    },
+    {
+      "epoch": 1.0122727272727272,
+      "grad_norm": 1.8961911527609079,
+      "learning_rate": 1.1278408019008194e-06,
+      "loss": 0.0857,
+      "step": 4454
+    },
+    {
+      "epoch": 1.0125,
+      "grad_norm": 1.5440012795079177,
+      "learning_rate": 1.1277877920986713e-06,
+      "loss": 0.0491,
+      "step": 4455
+    },
+    {
+      "epoch": 1.0127272727272727,
+      "grad_norm": 1.4322298243901592,
+      "learning_rate": 1.1277347720438028e-06,
+      "loss": 0.0241,
+      "step": 4456
+    },
+    {
+      "epoch": 1.0129545454545454,
+      "grad_norm": 1.7536980747916535,
+      "learning_rate": 1.1276817417372956e-06,
+      "loss": 0.0314,
+      "step": 4457
+    },
+    {
+      "epoch": 1.0131818181818182,
+      "grad_norm": 1.5336294772814205,
+      "learning_rate": 1.1276287011802304e-06,
+      "loss": 0.0478,
+      "step": 4458
+    },
+    {
+      "epoch": 1.013409090909091,
+      "grad_norm": 1.3370363974423305,
+      "learning_rate": 1.1275756503736893e-06,
+      "loss": 0.0314,
+      "step": 4459
+    },
+    {
+      "epoch": 1.0136363636363637,
+      "grad_norm": 1.5762463726628273,
+      "learning_rate": 1.1275225893187538e-06,
+      "loss": 0.0279,
+      "step": 4460
+    },
+    {
+      "epoch": 1.0138636363636364,
+      "grad_norm": 1.1983497101253744,
+      "learning_rate": 1.127469518016506e-06,
+      "loss": 0.0366,
+      "step": 4461
+    },
+    {
+      "epoch": 1.0140909090909092,
+      "grad_norm": 1.864303500116511,
+      "learning_rate": 1.1274164364680282e-06,
+      "loss": 0.0371,
+      "step": 4462
+    },
+    {
+      "epoch": 1.014318181818182,
+      "grad_norm": 1.5216001366394558,
+      "learning_rate": 1.1273633446744026e-06,
+      "loss": 0.0366,
+      "step": 4463
+    },
+    {
+      "epoch": 1.0145454545454546,
+      "grad_norm": 2.473776723680216,
+      "learning_rate": 1.1273102426367122e-06,
+      "loss": 0.0799,
+      "step": 4464
+    },
+    {
+      "epoch": 1.0147727272727274,
+      "grad_norm": 1.6802951714081034,
+      "learning_rate": 1.1272571303560394e-06,
+      "loss": 0.053,
+      "step": 4465
+    },
+    {
+      "epoch": 1.015,
+      "grad_norm": 2.2191996868463035,
+      "learning_rate": 1.1272040078334675e-06,
+      "loss": 0.037,
+      "step": 4466
+    },
+    {
+      "epoch": 1.0152272727272726,
+      "grad_norm": 2.296192880852401,
+      "learning_rate": 1.1271508750700799e-06,
+      "loss": 0.0766,
+      "step": 4467
+    },
+    {
+      "epoch": 1.0154545454545454,
+      "grad_norm": 1.7715031130735963,
+      "learning_rate": 1.1270977320669597e-06,
+      "loss": 0.0398,
+      "step": 4468
+    },
+    {
+      "epoch": 1.0156818181818181,
+      "grad_norm": 1.615967289018618,
+      "learning_rate": 1.127044578825191e-06,
+      "loss": 0.0558,
+      "step": 4469
+    },
+    {
+      "epoch": 1.0159090909090909,
+      "grad_norm": 1.196756511917262,
+      "learning_rate": 1.1269914153458574e-06,
+      "loss": 0.028,
+      "step": 4470
+    },
+    {
+      "epoch": 1.0161363636363636,
+      "grad_norm": 2.0055334607547617,
+      "learning_rate": 1.1269382416300429e-06,
+      "loss": 0.0676,
+      "step": 4471
+    },
+    {
+      "epoch": 1.0163636363636364,
+      "grad_norm": 1.9036597804036681,
+      "learning_rate": 1.1268850576788322e-06,
+      "loss": 0.049,
+      "step": 4472
+    },
+    {
+      "epoch": 1.016590909090909,
+      "grad_norm": 1.1298642650738135,
+      "learning_rate": 1.1268318634933095e-06,
+      "loss": 0.0439,
+      "step": 4473
+    },
+    {
+      "epoch": 1.0168181818181818,
+      "grad_norm": 1.1347724327571096,
+      "learning_rate": 1.1267786590745596e-06,
+      "loss": 0.0416,
+      "step": 4474
+    },
+    {
+      "epoch": 1.0170454545454546,
+      "grad_norm": 2.101669207780539,
+      "learning_rate": 1.1267254444236676e-06,
+      "loss": 0.0728,
+      "step": 4475
+    },
+    {
+      "epoch": 1.0172727272727273,
+      "grad_norm": 2.099737125191835,
+      "learning_rate": 1.126672219541718e-06,
+      "loss": 0.0413,
+      "step": 4476
+    },
+    {
+      "epoch": 1.0175,
+      "grad_norm": 1.6119207009533494,
+      "learning_rate": 1.1266189844297972e-06,
+      "loss": 0.0411,
+      "step": 4477
+    },
+    {
+      "epoch": 1.0177272727272728,
+      "grad_norm": 1.5100996949842131,
+      "learning_rate": 1.12656573908899e-06,
+      "loss": 0.0447,
+      "step": 4478
+    },
+    {
+      "epoch": 1.0179545454545456,
+      "grad_norm": 2.266908850414485,
+      "learning_rate": 1.1265124835203823e-06,
+      "loss": 0.0328,
+      "step": 4479
+    },
+    {
+      "epoch": 1.018181818181818,
+      "grad_norm": 1.6117353800571463,
+      "learning_rate": 1.1264592177250599e-06,
+      "loss": 0.0466,
+      "step": 4480
+    },
+    {
+      "epoch": 1.0184090909090908,
+      "grad_norm": 1.9149979744782872,
+      "learning_rate": 1.1264059417041095e-06,
+      "loss": 0.056,
+      "step": 4481
+    },
+    {
+      "epoch": 1.0186363636363636,
+      "grad_norm": 2.4807812244806646,
+      "learning_rate": 1.1263526554586171e-06,
+      "loss": 0.1245,
+      "step": 4482
+    },
+    {
+      "epoch": 1.0188636363636363,
+      "grad_norm": 1.4240254148766252,
+      "learning_rate": 1.1262993589896696e-06,
+      "loss": 0.0444,
+      "step": 4483
+    },
+    {
+      "epoch": 1.019090909090909,
+      "grad_norm": 1.1744813679773853,
+      "learning_rate": 1.1262460522983535e-06,
+      "loss": 0.0339,
+      "step": 4484
+    },
+    {
+      "epoch": 1.0193181818181818,
+      "grad_norm": 3.483338887158205,
+      "learning_rate": 1.1261927353857557e-06,
+      "loss": 0.0832,
+      "step": 4485
+    },
+    {
+      "epoch": 1.0195454545454545,
+      "grad_norm": 2.3982544506665477,
+      "learning_rate": 1.1261394082529638e-06,
+      "loss": 0.0873,
+      "step": 4486
+    },
+    {
+      "epoch": 1.0197727272727273,
+      "grad_norm": 1.6948813213348688,
+      "learning_rate": 1.1260860709010652e-06,
+      "loss": 0.0777,
+      "step": 4487
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.425671764595094,
+      "learning_rate": 1.1260327233311472e-06,
+      "loss": 0.0232,
+      "step": 4488
+    },
+    {
+      "epoch": 1.0202272727272728,
+      "grad_norm": 1.9357162477306928,
+      "learning_rate": 1.125979365544298e-06,
+      "loss": 0.0411,
+      "step": 4489
+    },
+    {
+      "epoch": 1.0204545454545455,
+      "grad_norm": 1.3853575031204737,
+      "learning_rate": 1.1259259975416054e-06,
+      "loss": 0.0449,
+      "step": 4490
+    },
+    {
+      "epoch": 1.0206818181818182,
+      "grad_norm": 1.8196960516666034,
+      "learning_rate": 1.1258726193241578e-06,
+      "loss": 0.0198,
+      "step": 4491
+    },
+    {
+      "epoch": 1.020909090909091,
+      "grad_norm": 1.2875968625481067,
+      "learning_rate": 1.125819230893044e-06,
+      "loss": 0.0352,
+      "step": 4492
+    },
+    {
+      "epoch": 1.0211363636363637,
+      "grad_norm": 0.8497025408765546,
+      "learning_rate": 1.125765832249352e-06,
+      "loss": 0.0221,
+      "step": 4493
+    },
+    {
+      "epoch": 1.0213636363636365,
+      "grad_norm": 1.969531029334638,
+      "learning_rate": 1.1257124233941714e-06,
+      "loss": 0.0838,
+      "step": 4494
+    },
+    {
+      "epoch": 1.021590909090909,
+      "grad_norm": 2.177236654402906,
+      "learning_rate": 1.1256590043285906e-06,
+      "loss": 0.0718,
+      "step": 4495
+    },
+    {
+      "epoch": 1.0218181818181817,
+      "grad_norm": 1.2571302652546494,
+      "learning_rate": 1.1256055750536992e-06,
+      "loss": 0.027,
+      "step": 4496
+    },
+    {
+      "epoch": 1.0220454545454545,
+      "grad_norm": 1.605346979421165,
+      "learning_rate": 1.125552135570587e-06,
+      "loss": 0.0303,
+      "step": 4497
+    },
+    {
+      "epoch": 1.0222727272727272,
+      "grad_norm": 1.6249952801444105,
+      "learning_rate": 1.1254986858803434e-06,
+      "loss": 0.07,
+      "step": 4498
+    },
+    {
+      "epoch": 1.0225,
+      "grad_norm": 1.8630591245768675,
+      "learning_rate": 1.1254452259840586e-06,
+      "loss": 0.0901,
+      "step": 4499
+    },
+    {
+      "epoch": 1.0227272727272727,
+      "grad_norm": 1.242407476574553,
+      "learning_rate": 1.1253917558828225e-06,
+      "loss": 0.0276,
+      "step": 4500
+    },
+    {
+      "epoch": 1.0229545454545454,
+      "grad_norm": 2.689196423894623,
+      "learning_rate": 1.1253382755777253e-06,
+      "loss": 0.0467,
+      "step": 4501
+    },
+    {
+      "epoch": 1.0231818181818182,
+      "grad_norm": 2.174166180595865,
+      "learning_rate": 1.125284785069858e-06,
+      "loss": 0.0368,
+      "step": 4502
+    },
+    {
+      "epoch": 1.023409090909091,
+      "grad_norm": 1.438411566051344,
+      "learning_rate": 1.1252312843603111e-06,
+      "loss": 0.0508,
+      "step": 4503
+    },
+    {
+      "epoch": 1.0236363636363637,
+      "grad_norm": 1.629649169048908,
+      "learning_rate": 1.1251777734501756e-06,
+      "loss": 0.0365,
+      "step": 4504
+    },
+    {
+      "epoch": 1.0238636363636364,
+      "grad_norm": 2.278514960070012,
+      "learning_rate": 1.1251242523405425e-06,
+      "loss": 0.0565,
+      "step": 4505
+    },
+    {
+      "epoch": 1.0240909090909092,
+      "grad_norm": 2.181095220208998,
+      "learning_rate": 1.1250707210325036e-06,
+      "loss": 0.0389,
+      "step": 4506
+    },
+    {
+      "epoch": 1.024318181818182,
+      "grad_norm": 1.2590105088155732,
+      "learning_rate": 1.1250171795271503e-06,
+      "loss": 0.0216,
+      "step": 4507
+    },
+    {
+      "epoch": 1.0245454545454546,
+      "grad_norm": 1.669065259438309,
+      "learning_rate": 1.1249636278255741e-06,
+      "loss": 0.0645,
+      "step": 4508
+    },
+    {
+      "epoch": 1.0247727272727272,
+      "grad_norm": 7.70366275508918,
+      "learning_rate": 1.1249100659288676e-06,
+      "loss": 0.0464,
+      "step": 4509
+    },
+    {
+      "epoch": 1.025,
+      "grad_norm": 1.4005712838262998,
+      "learning_rate": 1.1248564938381226e-06,
+      "loss": 0.0324,
+      "step": 4510
+    },
+    {
+      "epoch": 1.0252272727272727,
+      "grad_norm": 2.020101919476888,
+      "learning_rate": 1.1248029115544319e-06,
+      "loss": 0.0668,
+      "step": 4511
+    },
+    {
+      "epoch": 1.0254545454545454,
+      "grad_norm": 0.7518584373027258,
+      "learning_rate": 1.1247493190788874e-06,
+      "loss": 0.0238,
+      "step": 4512
+    },
+    {
+      "epoch": 1.0256818181818181,
+      "grad_norm": 1.164745762440263,
+      "learning_rate": 1.1246957164125827e-06,
+      "loss": 0.0404,
+      "step": 4513
+    },
+    {
+      "epoch": 1.0259090909090909,
+      "grad_norm": 1.4589308202140303,
+      "learning_rate": 1.1246421035566105e-06,
+      "loss": 0.0747,
+      "step": 4514
+    },
+    {
+      "epoch": 1.0261363636363636,
+      "grad_norm": 3.4015869147149527,
+      "learning_rate": 1.1245884805120642e-06,
+      "loss": 0.1107,
+      "step": 4515
+    },
+    {
+      "epoch": 1.0263636363636364,
+      "grad_norm": 1.5164314746929222,
+      "learning_rate": 1.1245348472800372e-06,
+      "loss": 0.048,
+      "step": 4516
+    },
+    {
+      "epoch": 1.026590909090909,
+      "grad_norm": 2.277254573853051,
+      "learning_rate": 1.1244812038616232e-06,
+      "loss": 0.046,
+      "step": 4517
+    },
+    {
+      "epoch": 1.0268181818181819,
+      "grad_norm": 1.6230630476010843,
+      "learning_rate": 1.124427550257916e-06,
+      "loss": 0.0828,
+      "step": 4518
+    },
+    {
+      "epoch": 1.0270454545454546,
+      "grad_norm": 1.9465867411209372,
+      "learning_rate": 1.1243738864700096e-06,
+      "loss": 0.0755,
+      "step": 4519
+    },
+    {
+      "epoch": 1.0272727272727273,
+      "grad_norm": 1.737128803239455,
+      "learning_rate": 1.1243202124989986e-06,
+      "loss": 0.057,
+      "step": 4520
+    },
+    {
+      "epoch": 1.0275,
+      "grad_norm": 2.62885786290785,
+      "learning_rate": 1.1242665283459775e-06,
+      "loss": 0.1165,
+      "step": 4521
+    },
+    {
+      "epoch": 1.0277272727272728,
+      "grad_norm": 1.966515469464289,
+      "learning_rate": 1.1242128340120407e-06,
+      "loss": 0.0645,
+      "step": 4522
+    },
+    {
+      "epoch": 1.0279545454545453,
+      "grad_norm": 0.9969880762049771,
+      "learning_rate": 1.1241591294982833e-06,
+      "loss": 0.0334,
+      "step": 4523
+    },
+    {
+      "epoch": 1.028181818181818,
+      "grad_norm": 1.8464070766893315,
+      "learning_rate": 1.1241054148058003e-06,
+      "loss": 0.0642,
+      "step": 4524
+    },
+    {
+      "epoch": 1.0284090909090908,
+      "grad_norm": 1.2069957226924493,
+      "learning_rate": 1.1240516899356874e-06,
+      "loss": 0.0226,
+      "step": 4525
+    },
+    {
+      "epoch": 1.0286363636363636,
+      "grad_norm": 0.9788137927747425,
+      "learning_rate": 1.1239979548890398e-06,
+      "loss": 0.0372,
+      "step": 4526
+    },
+    {
+      "epoch": 1.0288636363636363,
+      "grad_norm": 0.8783126186564265,
+      "learning_rate": 1.1239442096669534e-06,
+      "loss": 0.0346,
+      "step": 4527
+    },
+    {
+      "epoch": 1.029090909090909,
+      "grad_norm": 2.8587280229303613,
+      "learning_rate": 1.123890454270524e-06,
+      "loss": 0.0963,
+      "step": 4528
+    },
+    {
+      "epoch": 1.0293181818181818,
+      "grad_norm": 1.7673297627489406,
+      "learning_rate": 1.1238366887008478e-06,
+      "loss": 0.0282,
+      "step": 4529
+    },
+    {
+      "epoch": 1.0295454545454545,
+      "grad_norm": 1.6294672195204603,
+      "learning_rate": 1.1237829129590215e-06,
+      "loss": 0.033,
+      "step": 4530
+    },
+    {
+      "epoch": 1.0297727272727273,
+      "grad_norm": 1.7068974579712866,
+      "learning_rate": 1.1237291270461413e-06,
+      "loss": 0.048,
+      "step": 4531
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.42897692996257575,
+      "learning_rate": 1.1236753309633041e-06,
+      "loss": 0.0157,
+      "step": 4532
+    },
+    {
+      "epoch": 1.0302272727272728,
+      "grad_norm": 1.7103071757344743,
+      "learning_rate": 1.123621524711607e-06,
+      "loss": 0.0473,
+      "step": 4533
+    },
+    {
+      "epoch": 1.0304545454545455,
+      "grad_norm": 1.87179079916954,
+      "learning_rate": 1.123567708292147e-06,
+      "loss": 0.0279,
+      "step": 4534
+    },
+    {
+      "epoch": 1.0306818181818183,
+      "grad_norm": 1.6682196417399746,
+      "learning_rate": 1.1235138817060218e-06,
+      "loss": 0.0348,
+      "step": 4535
+    },
+    {
+      "epoch": 1.030909090909091,
+      "grad_norm": 1.2123820335677686,
+      "learning_rate": 1.1234600449543288e-06,
+      "loss": 0.0429,
+      "step": 4536
+    },
+    {
+      "epoch": 1.0311363636363637,
+      "grad_norm": 3.460660789064435,
+      "learning_rate": 1.1234061980381658e-06,
+      "loss": 0.1014,
+      "step": 4537
+    },
+    {
+      "epoch": 1.0313636363636363,
+      "grad_norm": 1.4819507928001907,
+      "learning_rate": 1.1233523409586308e-06,
+      "loss": 0.0409,
+      "step": 4538
+    },
+    {
+      "epoch": 1.031590909090909,
+      "grad_norm": 2.0431418805789034,
+      "learning_rate": 1.1232984737168221e-06,
+      "loss": 0.0427,
+      "step": 4539
+    },
+    {
+      "epoch": 1.0318181818181817,
+      "grad_norm": 2.0678918892546534,
+      "learning_rate": 1.1232445963138385e-06,
+      "loss": 0.051,
+      "step": 4540
+    },
+    {
+      "epoch": 1.0320454545454545,
+      "grad_norm": 2.135187981135868,
+      "learning_rate": 1.123190708750778e-06,
+      "loss": 0.0807,
+      "step": 4541
+    },
+    {
+      "epoch": 1.0322727272727272,
+      "grad_norm": 2.281991909126744,
+      "learning_rate": 1.1231368110287402e-06,
+      "loss": 0.06,
+      "step": 4542
+    },
+    {
+      "epoch": 1.0325,
+      "grad_norm": 2.553319830611825,
+      "learning_rate": 1.1230829031488235e-06,
+      "loss": 0.0627,
+      "step": 4543
+    },
+    {
+      "epoch": 1.0327272727272727,
+      "grad_norm": 1.0886894708043413,
+      "learning_rate": 1.1230289851121276e-06,
+      "loss": 0.0225,
+      "step": 4544
+    },
+    {
+      "epoch": 1.0329545454545455,
+      "grad_norm": 1.3924686710889085,
+      "learning_rate": 1.1229750569197518e-06,
+      "loss": 0.0434,
+      "step": 4545
+    },
+    {
+      "epoch": 1.0331818181818182,
+      "grad_norm": 1.3594026296086008,
+      "learning_rate": 1.1229211185727957e-06,
+      "loss": 0.0252,
+      "step": 4546
+    },
+    {
+      "epoch": 1.033409090909091,
+      "grad_norm": 2.216012984551213,
+      "learning_rate": 1.1228671700723595e-06,
+      "loss": 0.0452,
+      "step": 4547
+    },
+    {
+      "epoch": 1.0336363636363637,
+      "grad_norm": 2.730484589778948,
+      "learning_rate": 1.1228132114195429e-06,
+      "loss": 0.0482,
+      "step": 4548
+    },
+    {
+      "epoch": 1.0338636363636364,
+      "grad_norm": 1.4236113905307006,
+      "learning_rate": 1.1227592426154465e-06,
+      "loss": 0.0295,
+      "step": 4549
+    },
+    {
+      "epoch": 1.0340909090909092,
+      "grad_norm": 2.553567497288934,
+      "learning_rate": 1.122705263661171e-06,
+      "loss": 0.0415,
+      "step": 4550
+    },
+    {
+      "epoch": 1.034318181818182,
+      "grad_norm": 2.3783996871322355,
+      "learning_rate": 1.1226512745578166e-06,
+      "loss": 0.075,
+      "step": 4551
+    },
+    {
+      "epoch": 1.0345454545454547,
+      "grad_norm": 1.0160084721655762,
+      "learning_rate": 1.1225972753064847e-06,
+      "loss": 0.0185,
+      "step": 4552
+    },
+    {
+      "epoch": 1.0347727272727272,
+      "grad_norm": 2.0477187526541254,
+      "learning_rate": 1.1225432659082762e-06,
+      "loss": 0.0334,
+      "step": 4553
+    },
+    {
+      "epoch": 1.035,
+      "grad_norm": 1.6440847411892014,
+      "learning_rate": 1.1224892463642924e-06,
+      "loss": 0.0311,
+      "step": 4554
+    },
+    {
+      "epoch": 1.0352272727272727,
+      "grad_norm": 2.677032600238327,
+      "learning_rate": 1.122435216675635e-06,
+      "loss": 0.11,
+      "step": 4555
+    },
+    {
+      "epoch": 1.0354545454545454,
+      "grad_norm": 2.202525366588216,
+      "learning_rate": 1.1223811768434057e-06,
+      "loss": 0.0715,
+      "step": 4556
+    },
+    {
+      "epoch": 1.0356818181818181,
+      "grad_norm": 3.0682870894776584,
+      "learning_rate": 1.1223271268687066e-06,
+      "loss": 0.1229,
+      "step": 4557
+    },
+    {
+      "epoch": 1.035909090909091,
+      "grad_norm": 1.4668266470510136,
+      "learning_rate": 1.1222730667526396e-06,
+      "loss": 0.033,
+      "step": 4558
+    },
+    {
+      "epoch": 1.0361363636363636,
+      "grad_norm": 2.3226002144756905,
+      "learning_rate": 1.1222189964963072e-06,
+      "loss": 0.0662,
+      "step": 4559
+    },
+    {
+      "epoch": 1.0363636363636364,
+      "grad_norm": 1.5593070857325566,
+      "learning_rate": 1.1221649161008122e-06,
+      "loss": 0.0436,
+      "step": 4560
+    },
+    {
+      "epoch": 1.0365909090909091,
+      "grad_norm": 1.7470009607513028,
+      "learning_rate": 1.122110825567257e-06,
+      "loss": 0.0369,
+      "step": 4561
+    },
+    {
+      "epoch": 1.0368181818181819,
+      "grad_norm": 2.5706682548672513,
+      "learning_rate": 1.1220567248967448e-06,
+      "loss": 0.0815,
+      "step": 4562
+    },
+    {
+      "epoch": 1.0370454545454546,
+      "grad_norm": 1.7545158540741426,
+      "learning_rate": 1.122002614090379e-06,
+      "loss": 0.0527,
+      "step": 4563
+    },
+    {
+      "epoch": 1.0372727272727273,
+      "grad_norm": 1.4696083479376076,
+      "learning_rate": 1.1219484931492627e-06,
+      "loss": 0.0268,
+      "step": 4564
+    },
+    {
+      "epoch": 1.0375,
+      "grad_norm": 1.7250212511931629,
+      "learning_rate": 1.1218943620744998e-06,
+      "loss": 0.0429,
+      "step": 4565
+    },
+    {
+      "epoch": 1.0377272727272728,
+      "grad_norm": 1.66228962633756,
+      "learning_rate": 1.1218402208671938e-06,
+      "loss": 0.0484,
+      "step": 4566
+    },
+    {
+      "epoch": 1.0379545454545454,
+      "grad_norm": 1.8954134833667535,
+      "learning_rate": 1.121786069528449e-06,
+      "loss": 0.0681,
+      "step": 4567
+    },
+    {
+      "epoch": 1.038181818181818,
+      "grad_norm": 2.3437731280497407,
+      "learning_rate": 1.1217319080593695e-06,
+      "loss": 0.0756,
+      "step": 4568
+    },
+    {
+      "epoch": 1.0384090909090908,
+      "grad_norm": 3.0787191714839053,
+      "learning_rate": 1.1216777364610598e-06,
+      "loss": 0.0562,
+      "step": 4569
+    },
+    {
+      "epoch": 1.0386363636363636,
+      "grad_norm": 2.372096607613791,
+      "learning_rate": 1.1216235547346246e-06,
+      "loss": 0.0736,
+      "step": 4570
+    },
+    {
+      "epoch": 1.0388636363636363,
+      "grad_norm": 0.6963299269909989,
+      "learning_rate": 1.1215693628811688e-06,
+      "loss": 0.0144,
+      "step": 4571
+    },
+    {
+      "epoch": 1.039090909090909,
+      "grad_norm": 1.500568303033402,
+      "learning_rate": 1.1215151609017974e-06,
+      "loss": 0.0478,
+      "step": 4572
+    },
+    {
+      "epoch": 1.0393181818181818,
+      "grad_norm": 1.7504631667079678,
+      "learning_rate": 1.1214609487976155e-06,
+      "loss": 0.0317,
+      "step": 4573
+    },
+    {
+      "epoch": 1.0395454545454546,
+      "grad_norm": 3.44608831842479,
+      "learning_rate": 1.1214067265697288e-06,
+      "loss": 0.0807,
+      "step": 4574
+    },
+    {
+      "epoch": 1.0397727272727273,
+      "grad_norm": 3.322683418305735,
+      "learning_rate": 1.1213524942192428e-06,
+      "loss": 0.127,
+      "step": 4575
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.8185716559514873,
+      "learning_rate": 1.1212982517472636e-06,
+      "loss": 0.0522,
+      "step": 4576
+    },
+    {
+      "epoch": 1.0402272727272728,
+      "grad_norm": 2.609230724858619,
+      "learning_rate": 1.1212439991548971e-06,
+      "loss": 0.0547,
+      "step": 4577
+    },
+    {
+      "epoch": 1.0404545454545455,
+      "grad_norm": 1.6898886057213525,
+      "learning_rate": 1.1211897364432498e-06,
+      "loss": 0.0247,
+      "step": 4578
+    },
+    {
+      "epoch": 1.0406818181818183,
+      "grad_norm": 2.0357321256541634,
+      "learning_rate": 1.1211354636134282e-06,
+      "loss": 0.0614,
+      "step": 4579
+    },
+    {
+      "epoch": 1.040909090909091,
+      "grad_norm": 1.484433753561575,
+      "learning_rate": 1.121081180666539e-06,
+      "loss": 0.0382,
+      "step": 4580
+    },
+    {
+      "epoch": 1.0411363636363635,
+      "grad_norm": 1.3613809903867997,
+      "learning_rate": 1.1210268876036888e-06,
+      "loss": 0.0504,
+      "step": 4581
+    },
+    {
+      "epoch": 1.0413636363636363,
+      "grad_norm": 1.2936082643486164,
+      "learning_rate": 1.1209725844259852e-06,
+      "loss": 0.0414,
+      "step": 4582
+    },
+    {
+      "epoch": 1.041590909090909,
+      "grad_norm": 1.2203372199383005,
+      "learning_rate": 1.1209182711345353e-06,
+      "loss": 0.0326,
+      "step": 4583
+    },
+    {
+      "epoch": 1.0418181818181818,
+      "grad_norm": 1.2362932969198732,
+      "learning_rate": 1.1208639477304466e-06,
+      "loss": 0.0269,
+      "step": 4584
+    },
+    {
+      "epoch": 1.0420454545454545,
+      "grad_norm": 2.9729039635591987,
+      "learning_rate": 1.120809614214827e-06,
+      "loss": 0.0798,
+      "step": 4585
+    },
+    {
+      "epoch": 1.0422727272727272,
+      "grad_norm": 2.807342310295373,
+      "learning_rate": 1.1207552705887841e-06,
+      "loss": 0.1385,
+      "step": 4586
+    },
+    {
+      "epoch": 1.0425,
+      "grad_norm": 1.9573116541606324,
+      "learning_rate": 1.1207009168534268e-06,
+      "loss": 0.0616,
+      "step": 4587
+    },
+    {
+      "epoch": 1.0427272727272727,
+      "grad_norm": 1.9335958823518944,
+      "learning_rate": 1.1206465530098628e-06,
+      "loss": 0.0728,
+      "step": 4588
+    },
+    {
+      "epoch": 1.0429545454545455,
+      "grad_norm": 2.301701995369291,
+      "learning_rate": 1.1205921790592008e-06,
+      "loss": 0.0787,
+      "step": 4589
+    },
+    {
+      "epoch": 1.0431818181818182,
+      "grad_norm": 1.4059283144980412,
+      "learning_rate": 1.1205377950025499e-06,
+      "loss": 0.0532,
+      "step": 4590
+    },
+    {
+      "epoch": 1.043409090909091,
+      "grad_norm": 2.3082128062019285,
+      "learning_rate": 1.1204834008410184e-06,
+      "loss": 0.073,
+      "step": 4591
+    },
+    {
+      "epoch": 1.0436363636363637,
+      "grad_norm": 1.3561529503013328,
+      "learning_rate": 1.1204289965757163e-06,
+      "loss": 0.0353,
+      "step": 4592
+    },
+    {
+      "epoch": 1.0438636363636364,
+      "grad_norm": 2.3875262341467645,
+      "learning_rate": 1.1203745822077524e-06,
+      "loss": 0.0598,
+      "step": 4593
+    },
+    {
+      "epoch": 1.0440909090909092,
+      "grad_norm": 2.281002211397296,
+      "learning_rate": 1.120320157738237e-06,
+      "loss": 0.046,
+      "step": 4594
+    },
+    {
+      "epoch": 1.044318181818182,
+      "grad_norm": 1.5287759689489324,
+      "learning_rate": 1.120265723168279e-06,
+      "loss": 0.0404,
+      "step": 4595
+    },
+    {
+      "epoch": 1.0445454545454544,
+      "grad_norm": 1.273619582080558,
+      "learning_rate": 1.1202112784989891e-06,
+      "loss": 0.0336,
+      "step": 4596
+    },
+    {
+      "epoch": 1.0447727272727272,
+      "grad_norm": 3.16112779807842,
+      "learning_rate": 1.1201568237314772e-06,
+      "loss": 0.0939,
+      "step": 4597
+    },
+    {
+      "epoch": 1.045,
+      "grad_norm": 1.921011048450419,
+      "learning_rate": 1.1201023588668537e-06,
+      "loss": 0.0517,
+      "step": 4598
+    },
+    {
+      "epoch": 1.0452272727272727,
+      "grad_norm": 1.494456728665966,
+      "learning_rate": 1.1200478839062296e-06,
+      "loss": 0.0561,
+      "step": 4599
+    },
+    {
+      "epoch": 1.0454545454545454,
+      "grad_norm": 2.151524008006665,
+      "learning_rate": 1.1199933988507153e-06,
+      "loss": 0.0357,
+      "step": 4600
+    },
+    {
+      "epoch": 1.0456818181818182,
+      "grad_norm": 2.280020981791357,
+      "learning_rate": 1.1199389037014221e-06,
+      "loss": 0.0732,
+      "step": 4601
+    },
+    {
+      "epoch": 1.045909090909091,
+      "grad_norm": 3.1707224549900532,
+      "learning_rate": 1.1198843984594612e-06,
+      "loss": 0.074,
+      "step": 4602
+    },
+    {
+      "epoch": 1.0461363636363636,
+      "grad_norm": 1.925678229731674,
+      "learning_rate": 1.1198298831259438e-06,
+      "loss": 0.0338,
+      "step": 4603
+    },
+    {
+      "epoch": 1.0463636363636364,
+      "grad_norm": 1.3217354273255484,
+      "learning_rate": 1.119775357701982e-06,
+      "loss": 0.0305,
+      "step": 4604
+    },
+    {
+      "epoch": 1.0465909090909091,
+      "grad_norm": 1.7616391801936857,
+      "learning_rate": 1.1197208221886876e-06,
+      "loss": 0.0458,
+      "step": 4605
+    },
+    {
+      "epoch": 1.0468181818181819,
+      "grad_norm": 1.9131534663949283,
+      "learning_rate": 1.1196662765871725e-06,
+      "loss": 0.044,
+      "step": 4606
+    },
+    {
+      "epoch": 1.0470454545454546,
+      "grad_norm": 1.5090874432621642,
+      "learning_rate": 1.1196117208985489e-06,
+      "loss": 0.0478,
+      "step": 4607
+    },
+    {
+      "epoch": 1.0472727272727274,
+      "grad_norm": 1.0102396654725028,
+      "learning_rate": 1.1195571551239296e-06,
+      "loss": 0.0174,
+      "step": 4608
+    },
+    {
+      "epoch": 1.0475,
+      "grad_norm": 1.133572472406678,
+      "learning_rate": 1.119502579264427e-06,
+      "loss": 0.0259,
+      "step": 4609
+    },
+    {
+      "epoch": 1.0477272727272728,
+      "grad_norm": 2.0018841795274898,
+      "learning_rate": 1.1194479933211542e-06,
+      "loss": 0.0597,
+      "step": 4610
+    },
+    {
+      "epoch": 1.0479545454545454,
+      "grad_norm": 1.777539584227708,
+      "learning_rate": 1.119393397295224e-06,
+      "loss": 0.0778,
+      "step": 4611
+    },
+    {
+      "epoch": 1.048181818181818,
+      "grad_norm": 1.483352102411854,
+      "learning_rate": 1.1193387911877497e-06,
+      "loss": 0.0424,
+      "step": 4612
+    },
+    {
+      "epoch": 1.0484090909090908,
+      "grad_norm": 1.2471739710560439,
+      "learning_rate": 1.1192841749998455e-06,
+      "loss": 0.0416,
+      "step": 4613
+    },
+    {
+      "epoch": 1.0486363636363636,
+      "grad_norm": 1.0090228684225833,
+      "learning_rate": 1.1192295487326245e-06,
+      "loss": 0.0294,
+      "step": 4614
+    },
+    {
+      "epoch": 1.0488636363636363,
+      "grad_norm": 1.6752841741319333,
+      "learning_rate": 1.1191749123872008e-06,
+      "loss": 0.0393,
+      "step": 4615
+    },
+    {
+      "epoch": 1.049090909090909,
+      "grad_norm": 1.8947525471078184,
+      "learning_rate": 1.1191202659646883e-06,
+      "loss": 0.0417,
+      "step": 4616
+    },
+    {
+      "epoch": 1.0493181818181818,
+      "grad_norm": 2.106395158364556,
+      "learning_rate": 1.1190656094662017e-06,
+      "loss": 0.0537,
+      "step": 4617
+    },
+    {
+      "epoch": 1.0495454545454546,
+      "grad_norm": 2.9971401242633355,
+      "learning_rate": 1.1190109428928554e-06,
+      "loss": 0.1047,
+      "step": 4618
+    },
+    {
+      "epoch": 1.0497727272727273,
+      "grad_norm": 2.47138129678393,
+      "learning_rate": 1.118956266245764e-06,
+      "loss": 0.0481,
+      "step": 4619
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.2699812261750643,
+      "learning_rate": 1.1189015795260427e-06,
+      "loss": 0.0232,
+      "step": 4620
+    },
+    {
+      "epoch": 1.0502272727272728,
+      "grad_norm": 1.2445309368268458,
+      "learning_rate": 1.1188468827348066e-06,
+      "loss": 0.0391,
+      "step": 4621
+    },
+    {
+      "epoch": 1.0504545454545455,
+      "grad_norm": 2.6519593037382863,
+      "learning_rate": 1.118792175873171e-06,
+      "loss": 0.1115,
+      "step": 4622
+    },
+    {
+      "epoch": 1.0506818181818183,
+      "grad_norm": 1.9250105135192146,
+      "learning_rate": 1.1187374589422512e-06,
+      "loss": 0.0347,
+      "step": 4623
+    },
+    {
+      "epoch": 1.050909090909091,
+      "grad_norm": 1.5855175704379072,
+      "learning_rate": 1.1186827319431635e-06,
+      "loss": 0.0438,
+      "step": 4624
+    },
+    {
+      "epoch": 1.0511363636363635,
+      "grad_norm": 2.311002353977136,
+      "learning_rate": 1.1186279948770235e-06,
+      "loss": 0.0375,
+      "step": 4625
+    },
+    {
+      "epoch": 1.0513636363636363,
+      "grad_norm": 3.1795580748847923,
+      "learning_rate": 1.1185732477449475e-06,
+      "loss": 0.1075,
+      "step": 4626
+    },
+    {
+      "epoch": 1.051590909090909,
+      "grad_norm": 1.5797040242492333,
+      "learning_rate": 1.1185184905480518e-06,
+      "loss": 0.0531,
+      "step": 4627
+    },
+    {
+      "epoch": 1.0518181818181818,
+      "grad_norm": 2.193078530233923,
+      "learning_rate": 1.1184637232874533e-06,
+      "loss": 0.0547,
+      "step": 4628
+    },
+    {
+      "epoch": 1.0520454545454545,
+      "grad_norm": 1.6611197467271455,
+      "learning_rate": 1.1184089459642686e-06,
+      "loss": 0.0461,
+      "step": 4629
+    },
+    {
+      "epoch": 1.0522727272727272,
+      "grad_norm": 1.370280803910132,
+      "learning_rate": 1.1183541585796146e-06,
+      "loss": 0.0599,
+      "step": 4630
+    },
+    {
+      "epoch": 1.0525,
+      "grad_norm": 1.947550714966326,
+      "learning_rate": 1.1182993611346084e-06,
+      "loss": 0.0741,
+      "step": 4631
+    },
+    {
+      "epoch": 1.0527272727272727,
+      "grad_norm": 2.026262568086528,
+      "learning_rate": 1.118244553630368e-06,
+      "loss": 0.0433,
+      "step": 4632
+    },
+    {
+      "epoch": 1.0529545454545455,
+      "grad_norm": 1.4620383696975703,
+      "learning_rate": 1.1181897360680102e-06,
+      "loss": 0.0583,
+      "step": 4633
+    },
+    {
+      "epoch": 1.0531818181818182,
+      "grad_norm": 2.2279057565177456,
+      "learning_rate": 1.1181349084486536e-06,
+      "loss": 0.0793,
+      "step": 4634
+    },
+    {
+      "epoch": 1.053409090909091,
+      "grad_norm": 1.4857221865370456,
+      "learning_rate": 1.1180800707734157e-06,
+      "loss": 0.0261,
+      "step": 4635
+    },
+    {
+      "epoch": 1.0536363636363637,
+      "grad_norm": 1.360240209800206,
+      "learning_rate": 1.1180252230434151e-06,
+      "loss": 0.0389,
+      "step": 4636
+    },
+    {
+      "epoch": 1.0538636363636364,
+      "grad_norm": 1.6676391500376944,
+      "learning_rate": 1.11797036525977e-06,
+      "loss": 0.0266,
+      "step": 4637
+    },
+    {
+      "epoch": 1.0540909090909092,
+      "grad_norm": 1.9741218472390614,
+      "learning_rate": 1.1179154974235992e-06,
+      "loss": 0.0842,
+      "step": 4638
+    },
+    {
+      "epoch": 1.0543181818181817,
+      "grad_norm": 0.37707255162816933,
+      "learning_rate": 1.1178606195360214e-06,
+      "loss": 0.0112,
+      "step": 4639
+    },
+    {
+      "epoch": 1.0545454545454545,
+      "grad_norm": 1.5256673008327686,
+      "learning_rate": 1.1178057315981559e-06,
+      "loss": 0.0334,
+      "step": 4640
+    },
+    {
+      "epoch": 1.0547727272727272,
+      "grad_norm": 1.2241000799689084,
+      "learning_rate": 1.1177508336111215e-06,
+      "loss": 0.0332,
+      "step": 4641
+    },
+    {
+      "epoch": 1.055,
+      "grad_norm": 0.6259055578126628,
+      "learning_rate": 1.1176959255760381e-06,
+      "loss": 0.0239,
+      "step": 4642
+    },
+    {
+      "epoch": 1.0552272727272727,
+      "grad_norm": 1.3875951391600954,
+      "learning_rate": 1.117641007494025e-06,
+      "loss": 0.0247,
+      "step": 4643
+    },
+    {
+      "epoch": 1.0554545454545454,
+      "grad_norm": 2.1398019981014027,
+      "learning_rate": 1.1175860793662027e-06,
+      "loss": 0.0521,
+      "step": 4644
+    },
+    {
+      "epoch": 1.0556818181818182,
+      "grad_norm": 1.7428297081188309,
+      "learning_rate": 1.1175311411936907e-06,
+      "loss": 0.0489,
+      "step": 4645
+    },
+    {
+      "epoch": 1.055909090909091,
+      "grad_norm": 1.8585800139597195,
+      "learning_rate": 1.1174761929776094e-06,
+      "loss": 0.0477,
+      "step": 4646
+    },
+    {
+      "epoch": 1.0561363636363637,
+      "grad_norm": 1.4523018599061597,
+      "learning_rate": 1.1174212347190791e-06,
+      "loss": 0.025,
+      "step": 4647
+    },
+    {
+      "epoch": 1.0563636363636364,
+      "grad_norm": 1.0944893900291153,
+      "learning_rate": 1.117366266419221e-06,
+      "loss": 0.0202,
+      "step": 4648
+    },
+    {
+      "epoch": 1.0565909090909091,
+      "grad_norm": 1.3523967340978638,
+      "learning_rate": 1.1173112880791558e-06,
+      "loss": 0.0249,
+      "step": 4649
+    },
+    {
+      "epoch": 1.0568181818181819,
+      "grad_norm": 0.7146044843029394,
+      "learning_rate": 1.1172562997000046e-06,
+      "loss": 0.036,
+      "step": 4650
+    },
+    {
+      "epoch": 1.0570454545454546,
+      "grad_norm": 1.7323861471628588,
+      "learning_rate": 1.1172013012828887e-06,
+      "loss": 0.0506,
+      "step": 4651
+    },
+    {
+      "epoch": 1.0572727272727274,
+      "grad_norm": 1.4507777607434256,
+      "learning_rate": 1.1171462928289296e-06,
+      "loss": 0.0631,
+      "step": 4652
+    },
+    {
+      "epoch": 1.0575,
+      "grad_norm": 1.4860156213602762,
+      "learning_rate": 1.1170912743392487e-06,
+      "loss": 0.0201,
+      "step": 4653
+    },
+    {
+      "epoch": 1.0577272727272726,
+      "grad_norm": 1.3918603636510827,
+      "learning_rate": 1.1170362458149682e-06,
+      "loss": 0.0368,
+      "step": 4654
+    },
+    {
+      "epoch": 1.0579545454545454,
+      "grad_norm": 2.163461333750783,
+      "learning_rate": 1.1169812072572106e-06,
+      "loss": 0.0695,
+      "step": 4655
+    },
+    {
+      "epoch": 1.0581818181818181,
+      "grad_norm": 1.7710438427497461,
+      "learning_rate": 1.1169261586670976e-06,
+      "loss": 0.0503,
+      "step": 4656
+    },
+    {
+      "epoch": 1.0584090909090909,
+      "grad_norm": 1.912616088951064,
+      "learning_rate": 1.116871100045752e-06,
+      "loss": 0.0442,
+      "step": 4657
+    },
+    {
+      "epoch": 1.0586363636363636,
+      "grad_norm": 3.4605897657142757,
+      "learning_rate": 1.1168160313942967e-06,
+      "loss": 0.0598,
+      "step": 4658
+    },
+    {
+      "epoch": 1.0588636363636363,
+      "grad_norm": 2.1785461141022378,
+      "learning_rate": 1.1167609527138544e-06,
+      "loss": 0.0421,
+      "step": 4659
+    },
+    {
+      "epoch": 1.059090909090909,
+      "grad_norm": 2.2277261345875115,
+      "learning_rate": 1.1167058640055484e-06,
+      "loss": 0.03,
+      "step": 4660
+    },
+    {
+      "epoch": 1.0593181818181818,
+      "grad_norm": 1.4587341314062245,
+      "learning_rate": 1.1166507652705021e-06,
+      "loss": 0.038,
+      "step": 4661
+    },
+    {
+      "epoch": 1.0595454545454546,
+      "grad_norm": 1.2673212161753997,
+      "learning_rate": 1.116595656509839e-06,
+      "loss": 0.0174,
+      "step": 4662
+    },
+    {
+      "epoch": 1.0597727272727273,
+      "grad_norm": 2.1840367635829216,
+      "learning_rate": 1.1165405377246826e-06,
+      "loss": 0.0464,
+      "step": 4663
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.0095752368809947,
+      "learning_rate": 1.1164854089161572e-06,
+      "loss": 0.0443,
+      "step": 4664
+    },
+    {
+      "epoch": 1.0602272727272728,
+      "grad_norm": 1.852148575539562,
+      "learning_rate": 1.1164302700853868e-06,
+      "loss": 0.0242,
+      "step": 4665
+    },
+    {
+      "epoch": 1.0604545454545455,
+      "grad_norm": 3.148802287702087,
+      "learning_rate": 1.1163751212334962e-06,
+      "loss": 0.1476,
+      "step": 4666
+    },
+    {
+      "epoch": 1.0606818181818183,
+      "grad_norm": 1.705953643140737,
+      "learning_rate": 1.1163199623616092e-06,
+      "loss": 0.0293,
+      "step": 4667
+    },
+    {
+      "epoch": 1.0609090909090908,
+      "grad_norm": 1.6425825028043914,
+      "learning_rate": 1.1162647934708513e-06,
+      "loss": 0.0277,
+      "step": 4668
+    },
+    {
+      "epoch": 1.0611363636363635,
+      "grad_norm": 1.5920433402965968,
+      "learning_rate": 1.1162096145623468e-06,
+      "loss": 0.0421,
+      "step": 4669
+    },
+    {
+      "epoch": 1.0613636363636363,
+      "grad_norm": 1.6144613706374977,
+      "learning_rate": 1.1161544256372218e-06,
+      "loss": 0.0599,
+      "step": 4670
+    },
+    {
+      "epoch": 1.061590909090909,
+      "grad_norm": 2.1686572957463133,
+      "learning_rate": 1.116099226696601e-06,
+      "loss": 0.0766,
+      "step": 4671
+    },
+    {
+      "epoch": 1.0618181818181818,
+      "grad_norm": 1.4455313472587796,
+      "learning_rate": 1.1160440177416098e-06,
+      "loss": 0.023,
+      "step": 4672
+    },
+    {
+      "epoch": 1.0620454545454545,
+      "grad_norm": 1.8947813719293993,
+      "learning_rate": 1.115988798773375e-06,
+      "loss": 0.0307,
+      "step": 4673
+    },
+    {
+      "epoch": 1.0622727272727273,
+      "grad_norm": 2.27552539112849,
+      "learning_rate": 1.1159335697930218e-06,
+      "loss": 0.0715,
+      "step": 4674
+    },
+    {
+      "epoch": 1.0625,
+      "grad_norm": 2.1341550347837757,
+      "learning_rate": 1.1158783308016764e-06,
+      "loss": 0.0505,
+      "step": 4675
+    },
+    {
+      "epoch": 1.0627272727272727,
+      "grad_norm": 1.5307905531589279,
+      "learning_rate": 1.1158230818004656e-06,
+      "loss": 0.0443,
+      "step": 4676
+    },
+    {
+      "epoch": 1.0629545454545455,
+      "grad_norm": 1.0308160769219068,
+      "learning_rate": 1.115767822790516e-06,
+      "loss": 0.0368,
+      "step": 4677
+    },
+    {
+      "epoch": 1.0631818181818182,
+      "grad_norm": 1.7909089663697881,
+      "learning_rate": 1.1157125537729543e-06,
+      "loss": 0.0543,
+      "step": 4678
+    },
+    {
+      "epoch": 1.063409090909091,
+      "grad_norm": 1.5568486818068032,
+      "learning_rate": 1.1156572747489073e-06,
+      "loss": 0.0583,
+      "step": 4679
+    },
+    {
+      "epoch": 1.0636363636363637,
+      "grad_norm": 1.7888431066930088,
+      "learning_rate": 1.1156019857195027e-06,
+      "loss": 0.0587,
+      "step": 4680
+    },
+    {
+      "epoch": 1.0638636363636365,
+      "grad_norm": 1.8689545938961292,
+      "learning_rate": 1.1155466866858677e-06,
+      "loss": 0.047,
+      "step": 4681
+    },
+    {
+      "epoch": 1.064090909090909,
+      "grad_norm": 2.023697175201692,
+      "learning_rate": 1.1154913776491297e-06,
+      "loss": 0.081,
+      "step": 4682
+    },
+    {
+      "epoch": 1.0643181818181817,
+      "grad_norm": 1.8483187052559895,
+      "learning_rate": 1.1154360586104171e-06,
+      "loss": 0.029,
+      "step": 4683
+    },
+    {
+      "epoch": 1.0645454545454545,
+      "grad_norm": 1.278619498388582,
+      "learning_rate": 1.1153807295708575e-06,
+      "loss": 0.0293,
+      "step": 4684
+    },
+    {
+      "epoch": 1.0647727272727272,
+      "grad_norm": 1.5778097179325143,
+      "learning_rate": 1.1153253905315794e-06,
+      "loss": 0.0454,
+      "step": 4685
+    },
+    {
+      "epoch": 1.065,
+      "grad_norm": 1.8204833214745673,
+      "learning_rate": 1.1152700414937111e-06,
+      "loss": 0.0309,
+      "step": 4686
+    },
+    {
+      "epoch": 1.0652272727272727,
+      "grad_norm": 1.7218945056745676,
+      "learning_rate": 1.1152146824583814e-06,
+      "loss": 0.0469,
+      "step": 4687
+    },
+    {
+      "epoch": 1.0654545454545454,
+      "grad_norm": 1.8625410485370588,
+      "learning_rate": 1.1151593134267191e-06,
+      "loss": 0.0398,
+      "step": 4688
+    },
+    {
+      "epoch": 1.0656818181818182,
+      "grad_norm": 1.64804286208676,
+      "learning_rate": 1.1151039343998531e-06,
+      "loss": 0.0473,
+      "step": 4689
+    },
+    {
+      "epoch": 1.065909090909091,
+      "grad_norm": 2.235576414462196,
+      "learning_rate": 1.1150485453789131e-06,
+      "loss": 0.0425,
+      "step": 4690
+    },
+    {
+      "epoch": 1.0661363636363637,
+      "grad_norm": 1.283897352508624,
+      "learning_rate": 1.114993146365028e-06,
+      "loss": 0.0262,
+      "step": 4691
+    },
+    {
+      "epoch": 1.0663636363636364,
+      "grad_norm": 2.0326757930395045,
+      "learning_rate": 1.1149377373593282e-06,
+      "loss": 0.0541,
+      "step": 4692
+    },
+    {
+      "epoch": 1.0665909090909091,
+      "grad_norm": 1.961389484439353,
+      "learning_rate": 1.114882318362943e-06,
+      "loss": 0.0534,
+      "step": 4693
+    },
+    {
+      "epoch": 1.0668181818181819,
+      "grad_norm": 1.4279916314363637,
+      "learning_rate": 1.1148268893770026e-06,
+      "loss": 0.0203,
+      "step": 4694
+    },
+    {
+      "epoch": 1.0670454545454546,
+      "grad_norm": 1.7245347432311546,
+      "learning_rate": 1.1147714504026374e-06,
+      "loss": 0.045,
+      "step": 4695
+    },
+    {
+      "epoch": 1.0672727272727274,
+      "grad_norm": 1.412418023217374,
+      "learning_rate": 1.1147160014409779e-06,
+      "loss": 0.0543,
+      "step": 4696
+    },
+    {
+      "epoch": 1.0675,
+      "grad_norm": 1.8653589010422076,
+      "learning_rate": 1.1146605424931548e-06,
+      "loss": 0.0496,
+      "step": 4697
+    },
+    {
+      "epoch": 1.0677272727272726,
+      "grad_norm": 2.022268039445269,
+      "learning_rate": 1.114605073560299e-06,
+      "loss": 0.0489,
+      "step": 4698
+    },
+    {
+      "epoch": 1.0679545454545454,
+      "grad_norm": 1.8348574049048878,
+      "learning_rate": 1.1145495946435413e-06,
+      "loss": 0.0529,
+      "step": 4699
+    },
+    {
+      "epoch": 1.0681818181818181,
+      "grad_norm": 1.7205180875183934,
+      "learning_rate": 1.1144941057440136e-06,
+      "loss": 0.0411,
+      "step": 4700
+    },
+    {
+      "epoch": 1.0684090909090909,
+      "grad_norm": 2.540770327171062,
+      "learning_rate": 1.1144386068628472e-06,
+      "loss": 0.0848,
+      "step": 4701
+    },
+    {
+      "epoch": 1.0686363636363636,
+      "grad_norm": 1.6870134441238867,
+      "learning_rate": 1.1143830980011735e-06,
+      "loss": 0.0723,
+      "step": 4702
+    },
+    {
+      "epoch": 1.0688636363636363,
+      "grad_norm": 1.3689670079371763,
+      "learning_rate": 1.1143275791601249e-06,
+      "loss": 0.0515,
+      "step": 4703
+    },
+    {
+      "epoch": 1.069090909090909,
+      "grad_norm": 2.2996620389093,
+      "learning_rate": 1.114272050340833e-06,
+      "loss": 0.0562,
+      "step": 4704
+    },
+    {
+      "epoch": 1.0693181818181818,
+      "grad_norm": 2.0572508886933143,
+      "learning_rate": 1.1142165115444306e-06,
+      "loss": 0.0466,
+      "step": 4705
+    },
+    {
+      "epoch": 1.0695454545454546,
+      "grad_norm": 2.0304436501047634,
+      "learning_rate": 1.1141609627720501e-06,
+      "loss": 0.0638,
+      "step": 4706
+    },
+    {
+      "epoch": 1.0697727272727273,
+      "grad_norm": 1.362976302948961,
+      "learning_rate": 1.114105404024824e-06,
+      "loss": 0.0466,
+      "step": 4707
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.9656391728365048,
+      "learning_rate": 1.1140498353038857e-06,
+      "loss": 0.0578,
+      "step": 4708
+    },
+    {
+      "epoch": 1.0702272727272728,
+      "grad_norm": 1.961271393936768,
+      "learning_rate": 1.1139942566103677e-06,
+      "loss": 0.046,
+      "step": 4709
+    },
+    {
+      "epoch": 1.0704545454545455,
+      "grad_norm": 2.5072584583902082,
+      "learning_rate": 1.1139386679454038e-06,
+      "loss": 0.0753,
+      "step": 4710
+    },
+    {
+      "epoch": 1.0706818181818183,
+      "grad_norm": 2.003611002706506,
+      "learning_rate": 1.1138830693101277e-06,
+      "loss": 0.0575,
+      "step": 4711
+    },
+    {
+      "epoch": 1.0709090909090908,
+      "grad_norm": 1.369542802464638,
+      "learning_rate": 1.1138274607056726e-06,
+      "loss": 0.045,
+      "step": 4712
+    },
+    {
+      "epoch": 1.0711363636363636,
+      "grad_norm": 1.8923950696502545,
+      "learning_rate": 1.113771842133173e-06,
+      "loss": 0.0648,
+      "step": 4713
+    },
+    {
+      "epoch": 1.0713636363636363,
+      "grad_norm": 1.0732689383310612,
+      "learning_rate": 1.1137162135937625e-06,
+      "loss": 0.0295,
+      "step": 4714
+    },
+    {
+      "epoch": 1.071590909090909,
+      "grad_norm": 1.7173386593328757,
+      "learning_rate": 1.1136605750885762e-06,
+      "loss": 0.0465,
+      "step": 4715
+    },
+    {
+      "epoch": 1.0718181818181818,
+      "grad_norm": 2.803162719557196,
+      "learning_rate": 1.1136049266187481e-06,
+      "loss": 0.0676,
+      "step": 4716
+    },
+    {
+      "epoch": 1.0720454545454545,
+      "grad_norm": 1.7901109132721604,
+      "learning_rate": 1.1135492681854132e-06,
+      "loss": 0.0424,
+      "step": 4717
+    },
+    {
+      "epoch": 1.0722727272727273,
+      "grad_norm": 2.239657059868241,
+      "learning_rate": 1.1134935997897062e-06,
+      "loss": 0.0596,
+      "step": 4718
+    },
+    {
+      "epoch": 1.0725,
+      "grad_norm": 3.3557263729711426,
+      "learning_rate": 1.1134379214327625e-06,
+      "loss": 0.0666,
+      "step": 4719
+    },
+    {
+      "epoch": 1.0727272727272728,
+      "grad_norm": 1.4655874791310097,
+      "learning_rate": 1.1133822331157177e-06,
+      "loss": 0.0836,
+      "step": 4720
+    },
+    {
+      "epoch": 1.0729545454545455,
+      "grad_norm": 2.273467086567405,
+      "learning_rate": 1.113326534839707e-06,
+      "loss": 0.0655,
+      "step": 4721
+    },
+    {
+      "epoch": 1.0731818181818182,
+      "grad_norm": 1.9613884417977132,
+      "learning_rate": 1.1132708266058662e-06,
+      "loss": 0.0536,
+      "step": 4722
+    },
+    {
+      "epoch": 1.073409090909091,
+      "grad_norm": 1.7001548338543722,
+      "learning_rate": 1.1132151084153318e-06,
+      "loss": 0.0496,
+      "step": 4723
+    },
+    {
+      "epoch": 1.0736363636363637,
+      "grad_norm": 1.208260948976263,
+      "learning_rate": 1.1131593802692392e-06,
+      "loss": 0.0198,
+      "step": 4724
+    },
+    {
+      "epoch": 1.0738636363636365,
+      "grad_norm": 1.0412870681437465,
+      "learning_rate": 1.1131036421687252e-06,
+      "loss": 0.0235,
+      "step": 4725
+    },
+    {
+      "epoch": 1.0740909090909092,
+      "grad_norm": 1.7494532802317748,
+      "learning_rate": 1.1130478941149268e-06,
+      "loss": 0.031,
+      "step": 4726
+    },
+    {
+      "epoch": 1.0743181818181817,
+      "grad_norm": 1.9561676584677523,
+      "learning_rate": 1.11299213610898e-06,
+      "loss": 0.0505,
+      "step": 4727
+    },
+    {
+      "epoch": 1.0745454545454545,
+      "grad_norm": 1.465520447066641,
+      "learning_rate": 1.1129363681520223e-06,
+      "loss": 0.0277,
+      "step": 4728
+    },
+    {
+      "epoch": 1.0747727272727272,
+      "grad_norm": 1.554219482921872,
+      "learning_rate": 1.1128805902451907e-06,
+      "loss": 0.0375,
+      "step": 4729
+    },
+    {
+      "epoch": 1.075,
+      "grad_norm": 1.8978883653708278,
+      "learning_rate": 1.112824802389623e-06,
+      "loss": 0.0593,
+      "step": 4730
+    },
+    {
+      "epoch": 1.0752272727272727,
+      "grad_norm": 1.360582881722943,
+      "learning_rate": 1.1127690045864561e-06,
+      "loss": 0.0231,
+      "step": 4731
+    },
+    {
+      "epoch": 1.0754545454545454,
+      "grad_norm": 1.3425022343698356,
+      "learning_rate": 1.1127131968368285e-06,
+      "loss": 0.0323,
+      "step": 4732
+    },
+    {
+      "epoch": 1.0756818181818182,
+      "grad_norm": 1.4517211159673424,
+      "learning_rate": 1.1126573791418778e-06,
+      "loss": 0.0562,
+      "step": 4733
+    },
+    {
+      "epoch": 1.075909090909091,
+      "grad_norm": 2.3301104882588626,
+      "learning_rate": 1.1126015515027428e-06,
+      "loss": 0.0466,
+      "step": 4734
+    },
+    {
+      "epoch": 1.0761363636363637,
+      "grad_norm": 2.2883523320843797,
+      "learning_rate": 1.1125457139205612e-06,
+      "loss": 0.0833,
+      "step": 4735
+    },
+    {
+      "epoch": 1.0763636363636364,
+      "grad_norm": 1.9700552520953654,
+      "learning_rate": 1.1124898663964718e-06,
+      "loss": 0.0456,
+      "step": 4736
+    },
+    {
+      "epoch": 1.0765909090909092,
+      "grad_norm": 1.4799583497461888,
+      "learning_rate": 1.1124340089316137e-06,
+      "loss": 0.035,
+      "step": 4737
+    },
+    {
+      "epoch": 1.076818181818182,
+      "grad_norm": 1.9507268383093115,
+      "learning_rate": 1.1123781415271257e-06,
+      "loss": 0.0262,
+      "step": 4738
+    },
+    {
+      "epoch": 1.0770454545454546,
+      "grad_norm": 1.9829648068735974,
+      "learning_rate": 1.112322264184147e-06,
+      "loss": 0.0414,
+      "step": 4739
+    },
+    {
+      "epoch": 1.0772727272727272,
+      "grad_norm": 1.6090149554639368,
+      "learning_rate": 1.1122663769038175e-06,
+      "loss": 0.0434,
+      "step": 4740
+    },
+    {
+      "epoch": 1.0775,
+      "grad_norm": 1.3149064428948851,
+      "learning_rate": 1.1122104796872763e-06,
+      "loss": 0.0421,
+      "step": 4741
+    },
+    {
+      "epoch": 1.0777272727272726,
+      "grad_norm": 2.0202831767149023,
+      "learning_rate": 1.1121545725356635e-06,
+      "loss": 0.0274,
+      "step": 4742
+    },
+    {
+      "epoch": 1.0779545454545454,
+      "grad_norm": 1.9942843757105446,
+      "learning_rate": 1.112098655450119e-06,
+      "loss": 0.0664,
+      "step": 4743
+    },
+    {
+      "epoch": 1.0781818181818181,
+      "grad_norm": 1.7838401384777247,
+      "learning_rate": 1.1120427284317832e-06,
+      "loss": 0.0608,
+      "step": 4744
+    },
+    {
+      "epoch": 1.0784090909090909,
+      "grad_norm": 2.014946685603903,
+      "learning_rate": 1.1119867914817963e-06,
+      "loss": 0.0464,
+      "step": 4745
+    },
+    {
+      "epoch": 1.0786363636363636,
+      "grad_norm": 1.8693332569980727,
+      "learning_rate": 1.1119308446012993e-06,
+      "loss": 0.0553,
+      "step": 4746
+    },
+    {
+      "epoch": 1.0788636363636364,
+      "grad_norm": 1.716298141984726,
+      "learning_rate": 1.1118748877914328e-06,
+      "loss": 0.062,
+      "step": 4747
+    },
+    {
+      "epoch": 1.079090909090909,
+      "grad_norm": 1.4278043162750018,
+      "learning_rate": 1.111818921053338e-06,
+      "loss": 0.0367,
+      "step": 4748
+    },
+    {
+      "epoch": 1.0793181818181818,
+      "grad_norm": 2.054263397649353,
+      "learning_rate": 1.1117629443881562e-06,
+      "loss": 0.0757,
+      "step": 4749
+    },
+    {
+      "epoch": 1.0795454545454546,
+      "grad_norm": 1.613177747174749,
+      "learning_rate": 1.1117069577970286e-06,
+      "loss": 0.0339,
+      "step": 4750
+    },
+    {
+      "epoch": 1.0797727272727273,
+      "grad_norm": 2.082804348446984,
+      "learning_rate": 1.111650961281097e-06,
+      "loss": 0.0543,
+      "step": 4751
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.3775110273001903,
+      "learning_rate": 1.1115949548415033e-06,
+      "loss": 0.0446,
+      "step": 4752
+    },
+    {
+      "epoch": 1.0802272727272728,
+      "grad_norm": 1.560661331022803,
+      "learning_rate": 1.1115389384793898e-06,
+      "loss": 0.0206,
+      "step": 4753
+    },
+    {
+      "epoch": 1.0804545454545456,
+      "grad_norm": 1.383407141522372,
+      "learning_rate": 1.1114829121958983e-06,
+      "loss": 0.027,
+      "step": 4754
+    },
+    {
+      "epoch": 1.080681818181818,
+      "grad_norm": 1.2674718600704338,
+      "learning_rate": 1.1114268759921714e-06,
+      "loss": 0.0325,
+      "step": 4755
+    },
+    {
+      "epoch": 1.0809090909090908,
+      "grad_norm": 3.016468962761092,
+      "learning_rate": 1.111370829869352e-06,
+      "loss": 0.0396,
+      "step": 4756
+    },
+    {
+      "epoch": 1.0811363636363636,
+      "grad_norm": 1.3003260300528643,
+      "learning_rate": 1.111314773828583e-06,
+      "loss": 0.0419,
+      "step": 4757
+    },
+    {
+      "epoch": 1.0813636363636363,
+      "grad_norm": 1.7869085517635384,
+      "learning_rate": 1.1112587078710072e-06,
+      "loss": 0.0458,
+      "step": 4758
+    },
+    {
+      "epoch": 1.081590909090909,
+      "grad_norm": 1.654192007098248,
+      "learning_rate": 1.1112026319977681e-06,
+      "loss": 0.0643,
+      "step": 4759
+    },
+    {
+      "epoch": 1.0818181818181818,
+      "grad_norm": 0.944777168728353,
+      "learning_rate": 1.111146546210009e-06,
+      "loss": 0.0309,
+      "step": 4760
+    },
+    {
+      "epoch": 1.0820454545454545,
+      "grad_norm": 1.6252255903570694,
+      "learning_rate": 1.1110904505088738e-06,
+      "loss": 0.0396,
+      "step": 4761
+    },
+    {
+      "epoch": 1.0822727272727273,
+      "grad_norm": 1.5298453840621196,
+      "learning_rate": 1.1110343448955062e-06,
+      "loss": 0.0387,
+      "step": 4762
+    },
+    {
+      "epoch": 1.0825,
+      "grad_norm": 2.2571554542063597,
+      "learning_rate": 1.1109782293710503e-06,
+      "loss": 0.083,
+      "step": 4763
+    },
+    {
+      "epoch": 1.0827272727272728,
+      "grad_norm": 2.6818950767028413,
+      "learning_rate": 1.1109221039366506e-06,
+      "loss": 0.045,
+      "step": 4764
+    },
+    {
+      "epoch": 1.0829545454545455,
+      "grad_norm": 1.0675713606519137,
+      "learning_rate": 1.1108659685934514e-06,
+      "loss": 0.0314,
+      "step": 4765
+    },
+    {
+      "epoch": 1.0831818181818182,
+      "grad_norm": 2.845769639120061,
+      "learning_rate": 1.1108098233425977e-06,
+      "loss": 0.0687,
+      "step": 4766
+    },
+    {
+      "epoch": 1.083409090909091,
+      "grad_norm": 1.7558261172794873,
+      "learning_rate": 1.1107536681852337e-06,
+      "loss": 0.0844,
+      "step": 4767
+    },
+    {
+      "epoch": 1.0836363636363637,
+      "grad_norm": 1.168290042914909,
+      "learning_rate": 1.1106975031225053e-06,
+      "loss": 0.0252,
+      "step": 4768
+    },
+    {
+      "epoch": 1.0838636363636365,
+      "grad_norm": 3.098093182792178,
+      "learning_rate": 1.1106413281555574e-06,
+      "loss": 0.0549,
+      "step": 4769
+    },
+    {
+      "epoch": 1.084090909090909,
+      "grad_norm": 2.9474150459955175,
+      "learning_rate": 1.1105851432855354e-06,
+      "loss": 0.081,
+      "step": 4770
+    },
+    {
+      "epoch": 1.0843181818181817,
+      "grad_norm": 2.1943982807858404,
+      "learning_rate": 1.1105289485135855e-06,
+      "loss": 0.0797,
+      "step": 4771
+    },
+    {
+      "epoch": 1.0845454545454545,
+      "grad_norm": 1.3094909893095585,
+      "learning_rate": 1.1104727438408531e-06,
+      "loss": 0.0444,
+      "step": 4772
+    },
+    {
+      "epoch": 1.0847727272727272,
+      "grad_norm": 2.141554249470172,
+      "learning_rate": 1.1104165292684846e-06,
+      "loss": 0.057,
+      "step": 4773
+    },
+    {
+      "epoch": 1.085,
+      "grad_norm": 1.234850995034782,
+      "learning_rate": 1.1103603047976261e-06,
+      "loss": 0.03,
+      "step": 4774
+    },
+    {
+      "epoch": 1.0852272727272727,
+      "grad_norm": 1.550804453491793,
+      "learning_rate": 1.1103040704294242e-06,
+      "loss": 0.062,
+      "step": 4775
+    },
+    {
+      "epoch": 1.0854545454545454,
+      "grad_norm": 1.43646009275343,
+      "learning_rate": 1.1102478261650258e-06,
+      "loss": 0.0515,
+      "step": 4776
+    },
+    {
+      "epoch": 1.0856818181818182,
+      "grad_norm": 1.5416902020207772,
+      "learning_rate": 1.1101915720055776e-06,
+      "loss": 0.0206,
+      "step": 4777
+    },
+    {
+      "epoch": 1.085909090909091,
+      "grad_norm": 1.8227652070812788,
+      "learning_rate": 1.1101353079522266e-06,
+      "loss": 0.0454,
+      "step": 4778
+    },
+    {
+      "epoch": 1.0861363636363637,
+      "grad_norm": 1.2708367677606058,
+      "learning_rate": 1.1100790340061204e-06,
+      "loss": 0.0205,
+      "step": 4779
+    },
+    {
+      "epoch": 1.0863636363636364,
+      "grad_norm": 1.5742561240805495,
+      "learning_rate": 1.1100227501684066e-06,
+      "loss": 0.0401,
+      "step": 4780
+    },
+    {
+      "epoch": 1.0865909090909092,
+      "grad_norm": 2.887242135671193,
+      "learning_rate": 1.1099664564402327e-06,
+      "loss": 0.055,
+      "step": 4781
+    },
+    {
+      "epoch": 1.086818181818182,
+      "grad_norm": 0.9971198791688094,
+      "learning_rate": 1.1099101528227465e-06,
+      "loss": 0.0145,
+      "step": 4782
+    },
+    {
+      "epoch": 1.0870454545454546,
+      "grad_norm": 1.24584141707147,
+      "learning_rate": 1.1098538393170964e-06,
+      "loss": 0.0255,
+      "step": 4783
+    },
+    {
+      "epoch": 1.0872727272727274,
+      "grad_norm": 2.1851999326859426,
+      "learning_rate": 1.1097975159244307e-06,
+      "loss": 0.0605,
+      "step": 4784
+    },
+    {
+      "epoch": 1.0875,
+      "grad_norm": 1.7590211213436533,
+      "learning_rate": 1.1097411826458978e-06,
+      "loss": 0.0335,
+      "step": 4785
+    },
+    {
+      "epoch": 1.0877272727272727,
+      "grad_norm": 2.3141829439648913,
+      "learning_rate": 1.1096848394826467e-06,
+      "loss": 0.0732,
+      "step": 4786
+    },
+    {
+      "epoch": 1.0879545454545454,
+      "grad_norm": 2.207513174680549,
+      "learning_rate": 1.109628486435826e-06,
+      "loss": 0.0367,
+      "step": 4787
+    },
+    {
+      "epoch": 1.0881818181818181,
+      "grad_norm": 1.4358521694489041,
+      "learning_rate": 1.109572123506585e-06,
+      "loss": 0.0627,
+      "step": 4788
+    },
+    {
+      "epoch": 1.0884090909090909,
+      "grad_norm": 2.315328308235063,
+      "learning_rate": 1.109515750696073e-06,
+      "loss": 0.1074,
+      "step": 4789
+    },
+    {
+      "epoch": 1.0886363636363636,
+      "grad_norm": 0.6082851241008974,
+      "learning_rate": 1.1094593680054393e-06,
+      "loss": 0.0151,
+      "step": 4790
+    },
+    {
+      "epoch": 1.0888636363636364,
+      "grad_norm": 2.5556755444598624,
+      "learning_rate": 1.1094029754358343e-06,
+      "loss": 0.0444,
+      "step": 4791
+    },
+    {
+      "epoch": 1.089090909090909,
+      "grad_norm": 1.8898717352642442,
+      "learning_rate": 1.1093465729884075e-06,
+      "loss": 0.044,
+      "step": 4792
+    },
+    {
+      "epoch": 1.0893181818181819,
+      "grad_norm": 1.8336992994731915,
+      "learning_rate": 1.1092901606643089e-06,
+      "loss": 0.0523,
+      "step": 4793
+    },
+    {
+      "epoch": 1.0895454545454546,
+      "grad_norm": 1.6157566547432225,
+      "learning_rate": 1.1092337384646892e-06,
+      "loss": 0.0304,
+      "step": 4794
+    },
+    {
+      "epoch": 1.0897727272727273,
+      "grad_norm": 1.5340153119597624,
+      "learning_rate": 1.1091773063906988e-06,
+      "loss": 0.057,
+      "step": 4795
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.8172765854397301,
+      "learning_rate": 1.1091208644434882e-06,
+      "loss": 0.0677,
+      "step": 4796
+    },
+    {
+      "epoch": 1.0902272727272728,
+      "grad_norm": 1.936791710971849,
+      "learning_rate": 1.1090644126242086e-06,
+      "loss": 0.0531,
+      "step": 4797
+    },
+    {
+      "epoch": 1.0904545454545453,
+      "grad_norm": 1.8488165838722133,
+      "learning_rate": 1.1090079509340113e-06,
+      "loss": 0.0387,
+      "step": 4798
+    },
+    {
+      "epoch": 1.090681818181818,
+      "grad_norm": 2.0492124005052412,
+      "learning_rate": 1.1089514793740474e-06,
+      "loss": 0.0473,
+      "step": 4799
+    },
+    {
+      "epoch": 1.0909090909090908,
+      "grad_norm": 3.943733667465746,
+      "learning_rate": 1.1088949979454686e-06,
+      "loss": 0.0724,
+      "step": 4800
+    },
+    {
+      "epoch": 1.0911363636363636,
+      "grad_norm": 1.6737383078226007,
+      "learning_rate": 1.1088385066494267e-06,
+      "loss": 0.0312,
+      "step": 4801
+    },
+    {
+      "epoch": 1.0913636363636363,
+      "grad_norm": 2.061673771925797,
+      "learning_rate": 1.1087820054870733e-06,
+      "loss": 0.0549,
+      "step": 4802
+    },
+    {
+      "epoch": 1.091590909090909,
+      "grad_norm": 1.8425950268784845,
+      "learning_rate": 1.108725494459561e-06,
+      "loss": 0.057,
+      "step": 4803
+    },
+    {
+      "epoch": 1.0918181818181818,
+      "grad_norm": 2.0739892711298014,
+      "learning_rate": 1.1086689735680418e-06,
+      "loss": 0.097,
+      "step": 4804
+    },
+    {
+      "epoch": 1.0920454545454545,
+      "grad_norm": 3.247895911719444,
+      "learning_rate": 1.1086124428136686e-06,
+      "loss": 0.0883,
+      "step": 4805
+    },
+    {
+      "epoch": 1.0922727272727273,
+      "grad_norm": 1.8432992215305162,
+      "learning_rate": 1.1085559021975937e-06,
+      "loss": 0.0354,
+      "step": 4806
+    },
+    {
+      "epoch": 1.0925,
+      "grad_norm": 1.728727553570927,
+      "learning_rate": 1.1084993517209704e-06,
+      "loss": 0.0559,
+      "step": 4807
+    },
+    {
+      "epoch": 1.0927272727272728,
+      "grad_norm": 2.1385925408420876,
+      "learning_rate": 1.108442791384952e-06,
+      "loss": 0.0376,
+      "step": 4808
+    },
+    {
+      "epoch": 1.0929545454545455,
+      "grad_norm": 1.6718735099508213,
+      "learning_rate": 1.1083862211906918e-06,
+      "loss": 0.0565,
+      "step": 4809
+    },
+    {
+      "epoch": 1.0931818181818183,
+      "grad_norm": 1.388161809391227,
+      "learning_rate": 1.108329641139343e-06,
+      "loss": 0.066,
+      "step": 4810
+    },
+    {
+      "epoch": 1.093409090909091,
+      "grad_norm": 1.5983781183741037,
+      "learning_rate": 1.1082730512320597e-06,
+      "loss": 0.0388,
+      "step": 4811
+    },
+    {
+      "epoch": 1.0936363636363637,
+      "grad_norm": 1.528650310925019,
+      "learning_rate": 1.1082164514699956e-06,
+      "loss": 0.0503,
+      "step": 4812
+    },
+    {
+      "epoch": 1.0938636363636363,
+      "grad_norm": 2.9076445745624526,
+      "learning_rate": 1.108159841854305e-06,
+      "loss": 0.0867,
+      "step": 4813
+    },
+    {
+      "epoch": 1.094090909090909,
+      "grad_norm": 1.0108446398124642,
+      "learning_rate": 1.1081032223861427e-06,
+      "loss": 0.0218,
+      "step": 4814
+    },
+    {
+      "epoch": 1.0943181818181817,
+      "grad_norm": 1.7459742078256846,
+      "learning_rate": 1.1080465930666625e-06,
+      "loss": 0.071,
+      "step": 4815
+    },
+    {
+      "epoch": 1.0945454545454545,
+      "grad_norm": 2.549474268620618,
+      "learning_rate": 1.1079899538970196e-06,
+      "loss": 0.0548,
+      "step": 4816
+    },
+    {
+      "epoch": 1.0947727272727272,
+      "grad_norm": 1.5527088510181024,
+      "learning_rate": 1.107933304878369e-06,
+      "loss": 0.0345,
+      "step": 4817
+    },
+    {
+      "epoch": 1.095,
+      "grad_norm": 2.092768036104715,
+      "learning_rate": 1.1078766460118659e-06,
+      "loss": 0.0467,
+      "step": 4818
+    },
+    {
+      "epoch": 1.0952272727272727,
+      "grad_norm": 1.2128449988743568,
+      "learning_rate": 1.1078199772986652e-06,
+      "loss": 0.0443,
+      "step": 4819
+    },
+    {
+      "epoch": 1.0954545454545455,
+      "grad_norm": 2.410026791669714,
+      "learning_rate": 1.1077632987399232e-06,
+      "loss": 0.0868,
+      "step": 4820
+    },
+    {
+      "epoch": 1.0956818181818182,
+      "grad_norm": 2.2939559447804183,
+      "learning_rate": 1.1077066103367954e-06,
+      "loss": 0.0553,
+      "step": 4821
+    },
+    {
+      "epoch": 1.095909090909091,
+      "grad_norm": 2.2204561253294206,
+      "learning_rate": 1.1076499120904375e-06,
+      "loss": 0.0517,
+      "step": 4822
+    },
+    {
+      "epoch": 1.0961363636363637,
+      "grad_norm": 1.1693291074335297,
+      "learning_rate": 1.1075932040020057e-06,
+      "loss": 0.0372,
+      "step": 4823
+    },
+    {
+      "epoch": 1.0963636363636364,
+      "grad_norm": 1.7475479872891104,
+      "learning_rate": 1.1075364860726568e-06,
+      "loss": 0.027,
+      "step": 4824
+    },
+    {
+      "epoch": 1.0965909090909092,
+      "grad_norm": 1.5770048506672252,
+      "learning_rate": 1.1074797583035468e-06,
+      "loss": 0.0538,
+      "step": 4825
+    },
+    {
+      "epoch": 1.096818181818182,
+      "grad_norm": 1.9708444603475774,
+      "learning_rate": 1.1074230206958332e-06,
+      "loss": 0.0457,
+      "step": 4826
+    },
+    {
+      "epoch": 1.0970454545454547,
+      "grad_norm": 1.883277679280038,
+      "learning_rate": 1.1073662732506721e-06,
+      "loss": 0.0326,
+      "step": 4827
+    },
+    {
+      "epoch": 1.0972727272727272,
+      "grad_norm": 0.8558381512938085,
+      "learning_rate": 1.1073095159692215e-06,
+      "loss": 0.0157,
+      "step": 4828
+    },
+    {
+      "epoch": 1.0975,
+      "grad_norm": 1.862079598524587,
+      "learning_rate": 1.1072527488526383e-06,
+      "loss": 0.0308,
+      "step": 4829
+    },
+    {
+      "epoch": 1.0977272727272727,
+      "grad_norm": 1.7856660623717684,
+      "learning_rate": 1.1071959719020802e-06,
+      "loss": 0.0478,
+      "step": 4830
+    },
+    {
+      "epoch": 1.0979545454545454,
+      "grad_norm": 1.4166243486714976,
+      "learning_rate": 1.1071391851187052e-06,
+      "loss": 0.0406,
+      "step": 4831
+    },
+    {
+      "epoch": 1.0981818181818181,
+      "grad_norm": 1.733439568742794,
+      "learning_rate": 1.107082388503671e-06,
+      "loss": 0.0703,
+      "step": 4832
+    },
+    {
+      "epoch": 1.098409090909091,
+      "grad_norm": 1.9025836591002105,
+      "learning_rate": 1.1070255820581357e-06,
+      "loss": 0.0335,
+      "step": 4833
+    },
+    {
+      "epoch": 1.0986363636363636,
+      "grad_norm": 2.5483889229139307,
+      "learning_rate": 1.1069687657832582e-06,
+      "loss": 0.0626,
+      "step": 4834
+    },
+    {
+      "epoch": 1.0988636363636364,
+      "grad_norm": 1.5143896639626062,
+      "learning_rate": 1.1069119396801964e-06,
+      "loss": 0.0347,
+      "step": 4835
+    },
+    {
+      "epoch": 1.0990909090909091,
+      "grad_norm": 1.430611865407937,
+      "learning_rate": 1.1068551037501093e-06,
+      "loss": 0.0394,
+      "step": 4836
+    },
+    {
+      "epoch": 1.0993181818181819,
+      "grad_norm": 1.4586088375432649,
+      "learning_rate": 1.1067982579941563e-06,
+      "loss": 0.0311,
+      "step": 4837
+    },
+    {
+      "epoch": 1.0995454545454546,
+      "grad_norm": 1.447808836832053,
+      "learning_rate": 1.1067414024134963e-06,
+      "loss": 0.0286,
+      "step": 4838
+    },
+    {
+      "epoch": 1.0997727272727273,
+      "grad_norm": 2.5829327672692433,
+      "learning_rate": 1.1066845370092882e-06,
+      "loss": 0.0965,
+      "step": 4839
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.8547962732394403,
+      "learning_rate": 1.1066276617826923e-06,
+      "loss": 0.0797,
+      "step": 4840
+    },
+    {
+      "epoch": 1.1002272727272726,
+      "grad_norm": 1.3396503666646729,
+      "learning_rate": 1.1065707767348684e-06,
+      "loss": 0.0419,
+      "step": 4841
+    },
+    {
+      "epoch": 1.1004545454545454,
+      "grad_norm": 1.7799584026934774,
+      "learning_rate": 1.106513881866976e-06,
+      "loss": 0.0412,
+      "step": 4842
+    },
+    {
+      "epoch": 1.100681818181818,
+      "grad_norm": 1.3828220697366593,
+      "learning_rate": 1.1064569771801754e-06,
+      "loss": 0.0636,
+      "step": 4843
+    },
+    {
+      "epoch": 1.1009090909090908,
+      "grad_norm": 1.7080577729095203,
+      "learning_rate": 1.1064000626756273e-06,
+      "loss": 0.04,
+      "step": 4844
+    },
+    {
+      "epoch": 1.1011363636363636,
+      "grad_norm": 1.3738802528756324,
+      "learning_rate": 1.106343138354492e-06,
+      "loss": 0.0367,
+      "step": 4845
+    },
+    {
+      "epoch": 1.1013636363636363,
+      "grad_norm": 1.7881002926924678,
+      "learning_rate": 1.1062862042179302e-06,
+      "loss": 0.0337,
+      "step": 4846
+    },
+    {
+      "epoch": 1.101590909090909,
+      "grad_norm": 1.41944467134885,
+      "learning_rate": 1.106229260267103e-06,
+      "loss": 0.033,
+      "step": 4847
+    },
+    {
+      "epoch": 1.1018181818181818,
+      "grad_norm": 1.2853960647388392,
+      "learning_rate": 1.106172306503172e-06,
+      "loss": 0.0304,
+      "step": 4848
+    },
+    {
+      "epoch": 1.1020454545454546,
+      "grad_norm": 2.7691216509589403,
+      "learning_rate": 1.106115342927298e-06,
+      "loss": 0.0707,
+      "step": 4849
+    },
+    {
+      "epoch": 1.1022727272727273,
+      "grad_norm": 2.5880262144869195,
+      "learning_rate": 1.1060583695406429e-06,
+      "loss": 0.123,
+      "step": 4850
+    },
+    {
+      "epoch": 1.1025,
+      "grad_norm": 1.1148114248837189,
+      "learning_rate": 1.1060013863443683e-06,
+      "loss": 0.0401,
+      "step": 4851
+    },
+    {
+      "epoch": 1.1027272727272728,
+      "grad_norm": 1.7780700890775754,
+      "learning_rate": 1.1059443933396364e-06,
+      "loss": 0.0673,
+      "step": 4852
+    },
+    {
+      "epoch": 1.1029545454545455,
+      "grad_norm": 1.3204589570398344,
+      "learning_rate": 1.105887390527609e-06,
+      "loss": 0.0323,
+      "step": 4853
+    },
+    {
+      "epoch": 1.1031818181818183,
+      "grad_norm": 1.8696299309727866,
+      "learning_rate": 1.105830377909449e-06,
+      "loss": 0.0529,
+      "step": 4854
+    },
+    {
+      "epoch": 1.103409090909091,
+      "grad_norm": 1.4744936648750753,
+      "learning_rate": 1.1057733554863188e-06,
+      "loss": 0.0541,
+      "step": 4855
+    },
+    {
+      "epoch": 1.1036363636363635,
+      "grad_norm": 1.6478396985531218,
+      "learning_rate": 1.1057163232593808e-06,
+      "loss": 0.0329,
+      "step": 4856
+    },
+    {
+      "epoch": 1.1038636363636363,
+      "grad_norm": 1.9050161659543776,
+      "learning_rate": 1.1056592812297986e-06,
+      "loss": 0.0373,
+      "step": 4857
+    },
+    {
+      "epoch": 1.104090909090909,
+      "grad_norm": 1.4698841404765637,
+      "learning_rate": 1.105602229398735e-06,
+      "loss": 0.0344,
+      "step": 4858
+    },
+    {
+      "epoch": 1.1043181818181818,
+      "grad_norm": 1.784928465370008,
+      "learning_rate": 1.1055451677673535e-06,
+      "loss": 0.0508,
+      "step": 4859
+    },
+    {
+      "epoch": 1.1045454545454545,
+      "grad_norm": 1.1545738218675543,
+      "learning_rate": 1.1054880963368176e-06,
+      "loss": 0.0422,
+      "step": 4860
+    },
+    {
+      "epoch": 1.1047727272727272,
+      "grad_norm": 1.4647907317850974,
+      "learning_rate": 1.1054310151082913e-06,
+      "loss": 0.0423,
+      "step": 4861
+    },
+    {
+      "epoch": 1.105,
+      "grad_norm": 1.846775909457833,
+      "learning_rate": 1.1053739240829384e-06,
+      "loss": 0.0779,
+      "step": 4862
+    },
+    {
+      "epoch": 1.1052272727272727,
+      "grad_norm": 1.2129115189744755,
+      "learning_rate": 1.1053168232619229e-06,
+      "loss": 0.0371,
+      "step": 4863
+    },
+    {
+      "epoch": 1.1054545454545455,
+      "grad_norm": 2.28064583664225,
+      "learning_rate": 1.1052597126464098e-06,
+      "loss": 0.0691,
+      "step": 4864
+    },
+    {
+      "epoch": 1.1056818181818182,
+      "grad_norm": 1.5120819693366516,
+      "learning_rate": 1.105202592237563e-06,
+      "loss": 0.0398,
+      "step": 4865
+    },
+    {
+      "epoch": 1.105909090909091,
+      "grad_norm": 1.6544420876072015,
+      "learning_rate": 1.1051454620365475e-06,
+      "loss": 0.0299,
+      "step": 4866
+    },
+    {
+      "epoch": 1.1061363636363637,
+      "grad_norm": 1.2181366610701414,
+      "learning_rate": 1.1050883220445287e-06,
+      "loss": 0.0379,
+      "step": 4867
+    },
+    {
+      "epoch": 1.1063636363636364,
+      "grad_norm": 2.3999558812946873,
+      "learning_rate": 1.1050311722626713e-06,
+      "loss": 0.0671,
+      "step": 4868
+    },
+    {
+      "epoch": 1.1065909090909092,
+      "grad_norm": 2.0181565472669596,
+      "learning_rate": 1.104974012692141e-06,
+      "loss": 0.0456,
+      "step": 4869
+    },
+    {
+      "epoch": 1.106818181818182,
+      "grad_norm": 3.1113124935696757,
+      "learning_rate": 1.1049168433341028e-06,
+      "loss": 0.0459,
+      "step": 4870
+    },
+    {
+      "epoch": 1.1070454545454544,
+      "grad_norm": 1.8353724900447563,
+      "learning_rate": 1.1048596641897233e-06,
+      "loss": 0.0321,
+      "step": 4871
+    },
+    {
+      "epoch": 1.1072727272727272,
+      "grad_norm": 2.0081636606015527,
+      "learning_rate": 1.104802475260168e-06,
+      "loss": 0.0329,
+      "step": 4872
+    },
+    {
+      "epoch": 1.1075,
+      "grad_norm": 0.910597290722714,
+      "learning_rate": 1.1047452765466032e-06,
+      "loss": 0.0243,
+      "step": 4873
+    },
+    {
+      "epoch": 1.1077272727272727,
+      "grad_norm": 2.442529059797623,
+      "learning_rate": 1.104688068050195e-06,
+      "loss": 0.0577,
+      "step": 4874
+    },
+    {
+      "epoch": 1.1079545454545454,
+      "grad_norm": 1.4488617726321709,
+      "learning_rate": 1.1046308497721105e-06,
+      "loss": 0.0453,
+      "step": 4875
+    },
+    {
+      "epoch": 1.1081818181818182,
+      "grad_norm": 1.872236796957777,
+      "learning_rate": 1.104573621713516e-06,
+      "loss": 0.0874,
+      "step": 4876
+    },
+    {
+      "epoch": 1.108409090909091,
+      "grad_norm": 2.288279537895583,
+      "learning_rate": 1.104516383875579e-06,
+      "loss": 0.0759,
+      "step": 4877
+    },
+    {
+      "epoch": 1.1086363636363636,
+      "grad_norm": 1.76031318394151,
+      "learning_rate": 1.104459136259466e-06,
+      "loss": 0.0241,
+      "step": 4878
+    },
+    {
+      "epoch": 1.1088636363636364,
+      "grad_norm": 2.2144522069008956,
+      "learning_rate": 1.1044018788663453e-06,
+      "loss": 0.0634,
+      "step": 4879
+    },
+    {
+      "epoch": 1.1090909090909091,
+      "grad_norm": 1.8330622364816345,
+      "learning_rate": 1.1043446116973833e-06,
+      "loss": 0.0745,
+      "step": 4880
+    },
+    {
+      "epoch": 1.1093181818181819,
+      "grad_norm": 1.8671793494911033,
+      "learning_rate": 1.1042873347537485e-06,
+      "loss": 0.0447,
+      "step": 4881
+    },
+    {
+      "epoch": 1.1095454545454546,
+      "grad_norm": 1.5347348877280815,
+      "learning_rate": 1.104230048036609e-06,
+      "loss": 0.0361,
+      "step": 4882
+    },
+    {
+      "epoch": 1.1097727272727274,
+      "grad_norm": 7.475316396516895,
+      "learning_rate": 1.104172751547133e-06,
+      "loss": 0.0287,
+      "step": 4883
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.0786464514684018,
+      "learning_rate": 1.104115445286488e-06,
+      "loss": 0.0361,
+      "step": 4884
+    },
+    {
+      "epoch": 1.1102272727272728,
+      "grad_norm": 1.9737475397098612,
+      "learning_rate": 1.1040581292558438e-06,
+      "loss": 0.0394,
+      "step": 4885
+    },
+    {
+      "epoch": 1.1104545454545454,
+      "grad_norm": 1.6369079195262495,
+      "learning_rate": 1.1040008034563682e-06,
+      "loss": 0.0402,
+      "step": 4886
+    },
+    {
+      "epoch": 1.110681818181818,
+      "grad_norm": 2.4421724178701916,
+      "learning_rate": 1.1039434678892308e-06,
+      "loss": 0.0727,
+      "step": 4887
+    },
+    {
+      "epoch": 1.1109090909090908,
+      "grad_norm": 1.7954429025979513,
+      "learning_rate": 1.1038861225556004e-06,
+      "loss": 0.0406,
+      "step": 4888
+    },
+    {
+      "epoch": 1.1111363636363636,
+      "grad_norm": 0.9902204826231455,
+      "learning_rate": 1.1038287674566465e-06,
+      "loss": 0.0416,
+      "step": 4889
+    },
+    {
+      "epoch": 1.1113636363636363,
+      "grad_norm": 1.5329916531885142,
+      "learning_rate": 1.1037714025935386e-06,
+      "loss": 0.0203,
+      "step": 4890
+    },
+    {
+      "epoch": 1.111590909090909,
+      "grad_norm": 1.8430529243836746,
+      "learning_rate": 1.1037140279674468e-06,
+      "loss": 0.0269,
+      "step": 4891
+    },
+    {
+      "epoch": 1.1118181818181818,
+      "grad_norm": 1.5151445293638608,
+      "learning_rate": 1.1036566435795404e-06,
+      "loss": 0.0612,
+      "step": 4892
+    },
+    {
+      "epoch": 1.1120454545454546,
+      "grad_norm": 2.8258197268137066,
+      "learning_rate": 1.1035992494309903e-06,
+      "loss": 0.0455,
+      "step": 4893
+    },
+    {
+      "epoch": 1.1122727272727273,
+      "grad_norm": 1.7675152975337152,
+      "learning_rate": 1.1035418455229664e-06,
+      "loss": 0.0363,
+      "step": 4894
+    },
+    {
+      "epoch": 1.1125,
+      "grad_norm": 1.6576002950426816,
+      "learning_rate": 1.1034844318566396e-06,
+      "loss": 0.0261,
+      "step": 4895
+    },
+    {
+      "epoch": 1.1127272727272728,
+      "grad_norm": 1.770623117428466,
+      "learning_rate": 1.1034270084331803e-06,
+      "loss": 0.0396,
+      "step": 4896
+    },
+    {
+      "epoch": 1.1129545454545455,
+      "grad_norm": 1.5282569430867814,
+      "learning_rate": 1.1033695752537596e-06,
+      "loss": 0.0505,
+      "step": 4897
+    },
+    {
+      "epoch": 1.1131818181818183,
+      "grad_norm": 2.577721828436324,
+      "learning_rate": 1.1033121323195487e-06,
+      "loss": 0.05,
+      "step": 4898
+    },
+    {
+      "epoch": 1.1134090909090908,
+      "grad_norm": 0.7125537672882208,
+      "learning_rate": 1.103254679631719e-06,
+      "loss": 0.04,
+      "step": 4899
+    },
+    {
+      "epoch": 1.1136363636363635,
+      "grad_norm": 1.516361077357395,
+      "learning_rate": 1.103197217191442e-06,
+      "loss": 0.0278,
+      "step": 4900
+    },
+    {
+      "epoch": 1.1138636363636363,
+      "grad_norm": 1.8579980384939787,
+      "learning_rate": 1.1031397449998896e-06,
+      "loss": 0.0719,
+      "step": 4901
+    },
+    {
+      "epoch": 1.114090909090909,
+      "grad_norm": 1.805583692340183,
+      "learning_rate": 1.1030822630582335e-06,
+      "loss": 0.0456,
+      "step": 4902
+    },
+    {
+      "epoch": 1.1143181818181818,
+      "grad_norm": 2.2102404701816183,
+      "learning_rate": 1.1030247713676462e-06,
+      "loss": 0.0571,
+      "step": 4903
+    },
+    {
+      "epoch": 1.1145454545454545,
+      "grad_norm": 3.0706498171879377,
+      "learning_rate": 1.1029672699292996e-06,
+      "loss": 0.0742,
+      "step": 4904
+    },
+    {
+      "epoch": 1.1147727272727272,
+      "grad_norm": 1.8454225383200098,
+      "learning_rate": 1.1029097587443665e-06,
+      "loss": 0.0671,
+      "step": 4905
+    },
+    {
+      "epoch": 1.115,
+      "grad_norm": 2.672357894142236,
+      "learning_rate": 1.10285223781402e-06,
+      "loss": 0.0601,
+      "step": 4906
+    },
+    {
+      "epoch": 1.1152272727272727,
+      "grad_norm": 3.6590417576888523,
+      "learning_rate": 1.1027947071394324e-06,
+      "loss": 0.1055,
+      "step": 4907
+    },
+    {
+      "epoch": 1.1154545454545455,
+      "grad_norm": 1.5893102208835928,
+      "learning_rate": 1.1027371667217773e-06,
+      "loss": 0.0562,
+      "step": 4908
+    },
+    {
+      "epoch": 1.1156818181818182,
+      "grad_norm": 2.261591653040477,
+      "learning_rate": 1.1026796165622279e-06,
+      "loss": 0.0964,
+      "step": 4909
+    },
+    {
+      "epoch": 1.115909090909091,
+      "grad_norm": 1.9021111721439226,
+      "learning_rate": 1.1026220566619575e-06,
+      "loss": 0.0482,
+      "step": 4910
+    },
+    {
+      "epoch": 1.1161363636363637,
+      "grad_norm": 1.4990123113765952,
+      "learning_rate": 1.1025644870221405e-06,
+      "loss": 0.0316,
+      "step": 4911
+    },
+    {
+      "epoch": 1.1163636363636364,
+      "grad_norm": 1.9594980062652603,
+      "learning_rate": 1.1025069076439502e-06,
+      "loss": 0.0389,
+      "step": 4912
+    },
+    {
+      "epoch": 1.1165909090909092,
+      "grad_norm": 2.3021411672037946,
+      "learning_rate": 1.1024493185285613e-06,
+      "loss": 0.0407,
+      "step": 4913
+    },
+    {
+      "epoch": 1.1168181818181817,
+      "grad_norm": 2.1160739843686627,
+      "learning_rate": 1.1023917196771473e-06,
+      "loss": 0.037,
+      "step": 4914
+    },
+    {
+      "epoch": 1.1170454545454545,
+      "grad_norm": 1.794453509822364,
+      "learning_rate": 1.1023341110908837e-06,
+      "loss": 0.0398,
+      "step": 4915
+    },
+    {
+      "epoch": 1.1172727272727272,
+      "grad_norm": 1.8063785300369348,
+      "learning_rate": 1.1022764927709447e-06,
+      "loss": 0.0592,
+      "step": 4916
+    },
+    {
+      "epoch": 1.1175,
+      "grad_norm": 2.2290038496662707,
+      "learning_rate": 1.1022188647185053e-06,
+      "loss": 0.1372,
+      "step": 4917
+    },
+    {
+      "epoch": 1.1177272727272727,
+      "grad_norm": 1.9584188231837434,
+      "learning_rate": 1.1021612269347408e-06,
+      "loss": 0.0852,
+      "step": 4918
+    },
+    {
+      "epoch": 1.1179545454545454,
+      "grad_norm": 1.9992798236731009,
+      "learning_rate": 1.1021035794208265e-06,
+      "loss": 0.0502,
+      "step": 4919
+    },
+    {
+      "epoch": 1.1181818181818182,
+      "grad_norm": 1.2676760380994871,
+      "learning_rate": 1.1020459221779376e-06,
+      "loss": 0.0168,
+      "step": 4920
+    },
+    {
+      "epoch": 1.118409090909091,
+      "grad_norm": 3.1932974589545093,
+      "learning_rate": 1.1019882552072502e-06,
+      "loss": 0.0671,
+      "step": 4921
+    },
+    {
+      "epoch": 1.1186363636363637,
+      "grad_norm": 1.7858223136796074,
+      "learning_rate": 1.1019305785099403e-06,
+      "loss": 0.0612,
+      "step": 4922
+    },
+    {
+      "epoch": 1.1188636363636364,
+      "grad_norm": 1.809690210855179,
+      "learning_rate": 1.1018728920871837e-06,
+      "loss": 0.0352,
+      "step": 4923
+    },
+    {
+      "epoch": 1.1190909090909091,
+      "grad_norm": 1.498448314229979,
+      "learning_rate": 1.101815195940157e-06,
+      "loss": 0.0419,
+      "step": 4924
+    },
+    {
+      "epoch": 1.1193181818181819,
+      "grad_norm": 2.1258651972808464,
+      "learning_rate": 1.1017574900700364e-06,
+      "loss": 0.0353,
+      "step": 4925
+    },
+    {
+      "epoch": 1.1195454545454546,
+      "grad_norm": 1.7631283448812385,
+      "learning_rate": 1.101699774477999e-06,
+      "loss": 0.0678,
+      "step": 4926
+    },
+    {
+      "epoch": 1.1197727272727274,
+      "grad_norm": 1.7489993063544393,
+      "learning_rate": 1.1016420491652215e-06,
+      "loss": 0.0366,
+      "step": 4927
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.4436400047059152,
+      "learning_rate": 1.101584314132881e-06,
+      "loss": 0.0205,
+      "step": 4928
+    },
+    {
+      "epoch": 1.1202272727272726,
+      "grad_norm": 1.3003582975312107,
+      "learning_rate": 1.101526569382155e-06,
+      "loss": 0.0425,
+      "step": 4929
+    },
+    {
+      "epoch": 1.1204545454545454,
+      "grad_norm": 0.7784766416929302,
+      "learning_rate": 1.1014688149142208e-06,
+      "loss": 0.0174,
+      "step": 4930
+    },
+    {
+      "epoch": 1.1206818181818181,
+      "grad_norm": 3.2573415575527513,
+      "learning_rate": 1.1014110507302563e-06,
+      "loss": 0.0903,
+      "step": 4931
+    },
+    {
+      "epoch": 1.1209090909090909,
+      "grad_norm": 2.5449622758608066,
+      "learning_rate": 1.1013532768314394e-06,
+      "loss": 0.0308,
+      "step": 4932
+    },
+    {
+      "epoch": 1.1211363636363636,
+      "grad_norm": 5.63382410372314,
+      "learning_rate": 1.1012954932189479e-06,
+      "loss": 0.0289,
+      "step": 4933
+    },
+    {
+      "epoch": 1.1213636363636363,
+      "grad_norm": 1.8425138909924708,
+      "learning_rate": 1.1012376998939605e-06,
+      "loss": 0.0172,
+      "step": 4934
+    },
+    {
+      "epoch": 1.121590909090909,
+      "grad_norm": 2.236848089164351,
+      "learning_rate": 1.1011798968576557e-06,
+      "loss": 0.0679,
+      "step": 4935
+    },
+    {
+      "epoch": 1.1218181818181818,
+      "grad_norm": 1.8316659297013718,
+      "learning_rate": 1.1011220841112118e-06,
+      "loss": 0.0539,
+      "step": 4936
+    },
+    {
+      "epoch": 1.1220454545454546,
+      "grad_norm": 1.7903231340215837,
+      "learning_rate": 1.1010642616558083e-06,
+      "loss": 0.059,
+      "step": 4937
+    },
+    {
+      "epoch": 1.1222727272727273,
+      "grad_norm": 2.0720250607148007,
+      "learning_rate": 1.1010064294926236e-06,
+      "loss": 0.0945,
+      "step": 4938
+    },
+    {
+      "epoch": 1.1225,
+      "grad_norm": 1.9282583852636839,
+      "learning_rate": 1.1009485876228374e-06,
+      "loss": 0.0329,
+      "step": 4939
+    },
+    {
+      "epoch": 1.1227272727272728,
+      "grad_norm": 1.9285030418785103,
+      "learning_rate": 1.1008907360476295e-06,
+      "loss": 0.0663,
+      "step": 4940
+    },
+    {
+      "epoch": 1.1229545454545455,
+      "grad_norm": 1.651477465722451,
+      "learning_rate": 1.1008328747681788e-06,
+      "loss": 0.0653,
+      "step": 4941
+    },
+    {
+      "epoch": 1.1231818181818183,
+      "grad_norm": 1.890184923032407,
+      "learning_rate": 1.100775003785666e-06,
+      "loss": 0.0815,
+      "step": 4942
+    },
+    {
+      "epoch": 1.123409090909091,
+      "grad_norm": 1.5472772306804972,
+      "learning_rate": 1.1007171231012707e-06,
+      "loss": 0.0195,
+      "step": 4943
+    },
+    {
+      "epoch": 1.1236363636363635,
+      "grad_norm": 1.937157921538579,
+      "learning_rate": 1.1006592327161734e-06,
+      "loss": 0.0403,
+      "step": 4944
+    },
+    {
+      "epoch": 1.1238636363636363,
+      "grad_norm": 1.576370506908117,
+      "learning_rate": 1.1006013326315542e-06,
+      "loss": 0.0455,
+      "step": 4945
+    },
+    {
+      "epoch": 1.124090909090909,
+      "grad_norm": 2.556186205075861,
+      "learning_rate": 1.1005434228485945e-06,
+      "loss": 0.0634,
+      "step": 4946
+    },
+    {
+      "epoch": 1.1243181818181818,
+      "grad_norm": 2.0986470443340712,
+      "learning_rate": 1.1004855033684747e-06,
+      "loss": 0.0647,
+      "step": 4947
+    },
+    {
+      "epoch": 1.1245454545454545,
+      "grad_norm": 2.0842690125209318,
+      "learning_rate": 1.100427574192376e-06,
+      "loss": 0.0642,
+      "step": 4948
+    },
+    {
+      "epoch": 1.1247727272727273,
+      "grad_norm": 2.7347632882243094,
+      "learning_rate": 1.1003696353214792e-06,
+      "loss": 0.067,
+      "step": 4949
+    },
+    {
+      "epoch": 1.125,
+      "grad_norm": 1.3107681011345376,
+      "learning_rate": 1.1003116867569666e-06,
+      "loss": 0.0288,
+      "step": 4950
+    },
+    {
+      "epoch": 1.1252272727272727,
+      "grad_norm": 1.1679187397020871,
+      "learning_rate": 1.1002537285000196e-06,
+      "loss": 0.0342,
+      "step": 4951
+    },
+    {
+      "epoch": 1.1254545454545455,
+      "grad_norm": 1.2197695871792982,
+      "learning_rate": 1.1001957605518197e-06,
+      "loss": 0.0386,
+      "step": 4952
+    },
+    {
+      "epoch": 1.1256818181818182,
+      "grad_norm": 2.030797562322339,
+      "learning_rate": 1.100137782913549e-06,
+      "loss": 0.051,
+      "step": 4953
+    },
+    {
+      "epoch": 1.125909090909091,
+      "grad_norm": 1.5668120653474955,
+      "learning_rate": 1.1000797955863905e-06,
+      "loss": 0.0619,
+      "step": 4954
+    },
+    {
+      "epoch": 1.1261363636363637,
+      "grad_norm": 1.8757861422066466,
+      "learning_rate": 1.1000217985715259e-06,
+      "loss": 0.0426,
+      "step": 4955
+    },
+    {
+      "epoch": 1.1263636363636365,
+      "grad_norm": 2.4091256627777295,
+      "learning_rate": 1.099963791870138e-06,
+      "loss": 0.0897,
+      "step": 4956
+    },
+    {
+      "epoch": 1.126590909090909,
+      "grad_norm": 1.413279421127508,
+      "learning_rate": 1.0999057754834099e-06,
+      "loss": 0.0332,
+      "step": 4957
+    },
+    {
+      "epoch": 1.126818181818182,
+      "grad_norm": 1.8358188890570435,
+      "learning_rate": 1.0998477494125245e-06,
+      "loss": 0.0596,
+      "step": 4958
+    },
+    {
+      "epoch": 1.1270454545454545,
+      "grad_norm": 1.6947972498813093,
+      "learning_rate": 1.099789713658665e-06,
+      "loss": 0.0325,
+      "step": 4959
+    },
+    {
+      "epoch": 1.1272727272727272,
+      "grad_norm": 1.8353870227615545,
+      "learning_rate": 1.099731668223015e-06,
+      "loss": 0.0688,
+      "step": 4960
+    },
+    {
+      "epoch": 1.1275,
+      "grad_norm": 1.8339067164301777,
+      "learning_rate": 1.099673613106758e-06,
+      "loss": 0.0354,
+      "step": 4961
+    },
+    {
+      "epoch": 1.1277272727272727,
+      "grad_norm": 1.3731642093385958,
+      "learning_rate": 1.0996155483110778e-06,
+      "loss": 0.04,
+      "step": 4962
+    },
+    {
+      "epoch": 1.1279545454545454,
+      "grad_norm": 1.7476833080042646,
+      "learning_rate": 1.099557473837159e-06,
+      "loss": 0.0494,
+      "step": 4963
+    },
+    {
+      "epoch": 1.1281818181818182,
+      "grad_norm": 1.7839956239208856,
+      "learning_rate": 1.099499389686185e-06,
+      "loss": 0.0463,
+      "step": 4964
+    },
+    {
+      "epoch": 1.128409090909091,
+      "grad_norm": 1.5934108219892096,
+      "learning_rate": 1.099441295859341e-06,
+      "loss": 0.084,
+      "step": 4965
+    },
+    {
+      "epoch": 1.1286363636363637,
+      "grad_norm": 1.5492024846330916,
+      "learning_rate": 1.0993831923578107e-06,
+      "loss": 0.0373,
+      "step": 4966
+    },
+    {
+      "epoch": 1.1288636363636364,
+      "grad_norm": 1.7311689638196666,
+      "learning_rate": 1.09932507918278e-06,
+      "loss": 0.0645,
+      "step": 4967
+    },
+    {
+      "epoch": 1.1290909090909091,
+      "grad_norm": 3.1390453210028104,
+      "learning_rate": 1.0992669563354333e-06,
+      "loss": 0.083,
+      "step": 4968
+    },
+    {
+      "epoch": 1.1293181818181819,
+      "grad_norm": 1.8450719132500204,
+      "learning_rate": 1.0992088238169558e-06,
+      "loss": 0.0551,
+      "step": 4969
+    },
+    {
+      "epoch": 1.1295454545454546,
+      "grad_norm": 1.9381061502522718,
+      "learning_rate": 1.0991506816285333e-06,
+      "loss": 0.0513,
+      "step": 4970
+    },
+    {
+      "epoch": 1.1297727272727274,
+      "grad_norm": 0.7905316767891853,
+      "learning_rate": 1.099092529771351e-06,
+      "loss": 0.0193,
+      "step": 4971
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.006519018809741,
+      "learning_rate": 1.099034368246595e-06,
+      "loss": 0.0299,
+      "step": 4972
+    },
+    {
+      "epoch": 1.1302272727272726,
+      "grad_norm": 2.3849873441411087,
+      "learning_rate": 1.0989761970554514e-06,
+      "loss": 0.0751,
+      "step": 4973
+    },
+    {
+      "epoch": 1.1304545454545454,
+      "grad_norm": 2.309900623285752,
+      "learning_rate": 1.0989180161991058e-06,
+      "loss": 0.0642,
+      "step": 4974
+    },
+    {
+      "epoch": 1.1306818181818181,
+      "grad_norm": 2.153871285446705,
+      "learning_rate": 1.0988598256787455e-06,
+      "loss": 0.0748,
+      "step": 4975
+    },
+    {
+      "epoch": 1.1309090909090909,
+      "grad_norm": 2.7452661654259014,
+      "learning_rate": 1.0988016254955565e-06,
+      "loss": 0.0552,
+      "step": 4976
+    },
+    {
+      "epoch": 1.1311363636363636,
+      "grad_norm": 0.9629758773654579,
+      "learning_rate": 1.0987434156507257e-06,
+      "loss": 0.0418,
+      "step": 4977
+    },
+    {
+      "epoch": 1.1313636363636363,
+      "grad_norm": 1.2932023215435706,
+      "learning_rate": 1.0986851961454403e-06,
+      "loss": 0.0159,
+      "step": 4978
+    },
+    {
+      "epoch": 1.131590909090909,
+      "grad_norm": 2.095606092064554,
+      "learning_rate": 1.0986269669808874e-06,
+      "loss": 0.0778,
+      "step": 4979
+    },
+    {
+      "epoch": 1.1318181818181818,
+      "grad_norm": 1.6999163835087923,
+      "learning_rate": 1.0985687281582541e-06,
+      "loss": 0.0658,
+      "step": 4980
+    },
+    {
+      "epoch": 1.1320454545454546,
+      "grad_norm": 5.5761124129218675,
+      "learning_rate": 1.0985104796787285e-06,
+      "loss": 0.096,
+      "step": 4981
+    },
+    {
+      "epoch": 1.1322727272727273,
+      "grad_norm": 2.5089215846951833,
+      "learning_rate": 1.098452221543498e-06,
+      "loss": 0.0555,
+      "step": 4982
+    },
+    {
+      "epoch": 1.1325,
+      "grad_norm": 2.101665485551473,
+      "learning_rate": 1.0983939537537506e-06,
+      "loss": 0.0799,
+      "step": 4983
+    },
+    {
+      "epoch": 1.1327272727272728,
+      "grad_norm": 3.450061930592721,
+      "learning_rate": 1.0983356763106747e-06,
+      "loss": 0.0488,
+      "step": 4984
+    },
+    {
+      "epoch": 1.1329545454545455,
+      "grad_norm": 1.7853724807135813,
+      "learning_rate": 1.0982773892154589e-06,
+      "loss": 0.0384,
+      "step": 4985
+    },
+    {
+      "epoch": 1.1331818181818183,
+      "grad_norm": 2.0369901595722752,
+      "learning_rate": 1.0982190924692911e-06,
+      "loss": 0.0503,
+      "step": 4986
+    },
+    {
+      "epoch": 1.1334090909090908,
+      "grad_norm": 1.6323125407949572,
+      "learning_rate": 1.0981607860733604e-06,
+      "loss": 0.0273,
+      "step": 4987
+    },
+    {
+      "epoch": 1.1336363636363636,
+      "grad_norm": 2.876640786179609,
+      "learning_rate": 1.098102470028856e-06,
+      "loss": 0.0741,
+      "step": 4988
+    },
+    {
+      "epoch": 1.1338636363636363,
+      "grad_norm": 1.4699422520327001,
+      "learning_rate": 1.0980441443369668e-06,
+      "loss": 0.0345,
+      "step": 4989
+    },
+    {
+      "epoch": 1.134090909090909,
+      "grad_norm": 1.314856564473248,
+      "learning_rate": 1.0979858089988821e-06,
+      "loss": 0.0337,
+      "step": 4990
+    },
+    {
+      "epoch": 1.1343181818181818,
+      "grad_norm": 1.5270066749526119,
+      "learning_rate": 1.0979274640157919e-06,
+      "loss": 0.0377,
+      "step": 4991
+    },
+    {
+      "epoch": 1.1345454545454545,
+      "grad_norm": 1.4431715039326163,
+      "learning_rate": 1.0978691093888853e-06,
+      "loss": 0.0228,
+      "step": 4992
+    },
+    {
+      "epoch": 1.1347727272727273,
+      "grad_norm": 2.294879753091361,
+      "learning_rate": 1.0978107451193527e-06,
+      "loss": 0.0565,
+      "step": 4993
+    },
+    {
+      "epoch": 1.135,
+      "grad_norm": 1.76242645078388,
+      "learning_rate": 1.0977523712083843e-06,
+      "loss": 0.0458,
+      "step": 4994
+    },
+    {
+      "epoch": 1.1352272727272728,
+      "grad_norm": 2.225695959286449,
+      "learning_rate": 1.09769398765717e-06,
+      "loss": 0.0743,
+      "step": 4995
+    },
+    {
+      "epoch": 1.1354545454545455,
+      "grad_norm": 1.3210374417232083,
+      "learning_rate": 1.097635594466901e-06,
+      "loss": 0.0281,
+      "step": 4996
+    },
+    {
+      "epoch": 1.1356818181818182,
+      "grad_norm": 3.79259691474038,
+      "learning_rate": 1.0975771916387673e-06,
+      "loss": 0.1418,
+      "step": 4997
+    },
+    {
+      "epoch": 1.135909090909091,
+      "grad_norm": 1.529704929101522,
+      "learning_rate": 1.0975187791739604e-06,
+      "loss": 0.0333,
+      "step": 4998
+    },
+    {
+      "epoch": 1.1361363636363637,
+      "grad_norm": 1.8524321478644188,
+      "learning_rate": 1.097460357073671e-06,
+      "loss": 0.0516,
+      "step": 4999
+    },
+    {
+      "epoch": 1.1363636363636362,
+      "grad_norm": 2.333913497597312,
+      "learning_rate": 1.097401925339091e-06,
+      "loss": 0.064,
+      "step": 5000
+    },
+    {
+      "epoch": 1.1365909090909092,
+      "grad_norm": 1.4417591836019805,
+      "learning_rate": 1.0973434839714116e-06,
+      "loss": 0.0431,
+      "step": 5001
+    },
+    {
+      "epoch": 1.1368181818181817,
+      "grad_norm": 2.218266503530642,
+      "learning_rate": 1.0972850329718244e-06,
+      "loss": 0.0609,
+      "step": 5002
+    },
+    {
+      "epoch": 1.1370454545454545,
+      "grad_norm": 1.9431198286561804,
+      "learning_rate": 1.0972265723415212e-06,
+      "loss": 0.0519,
+      "step": 5003
+    },
+    {
+      "epoch": 1.1372727272727272,
+      "grad_norm": 1.4833804692216126,
+      "learning_rate": 1.0971681020816946e-06,
+      "loss": 0.04,
+      "step": 5004
+    },
+    {
+      "epoch": 1.1375,
+      "grad_norm": 1.4804697807666454,
+      "learning_rate": 1.0971096221935369e-06,
+      "loss": 0.0653,
+      "step": 5005
+    },
+    {
+      "epoch": 1.1377272727272727,
+      "grad_norm": 1.6019070025929252,
+      "learning_rate": 1.0970511326782398e-06,
+      "loss": 0.0532,
+      "step": 5006
+    },
+    {
+      "epoch": 1.1379545454545454,
+      "grad_norm": 1.5534587357719767,
+      "learning_rate": 1.0969926335369968e-06,
+      "loss": 0.0542,
+      "step": 5007
+    },
+    {
+      "epoch": 1.1381818181818182,
+      "grad_norm": 1.0326463906117043,
+      "learning_rate": 1.0969341247710009e-06,
+      "loss": 0.0179,
+      "step": 5008
+    },
+    {
+      "epoch": 1.138409090909091,
+      "grad_norm": 1.302797067859672,
+      "learning_rate": 1.0968756063814444e-06,
+      "loss": 0.041,
+      "step": 5009
+    },
+    {
+      "epoch": 1.1386363636363637,
+      "grad_norm": 1.3624001232840413,
+      "learning_rate": 1.0968170783695212e-06,
+      "loss": 0.0341,
+      "step": 5010
+    },
+    {
+      "epoch": 1.1388636363636364,
+      "grad_norm": 3.0037968640850883,
+      "learning_rate": 1.0967585407364246e-06,
+      "loss": 0.0659,
+      "step": 5011
+    },
+    {
+      "epoch": 1.1390909090909092,
+      "grad_norm": 1.9999683627747737,
+      "learning_rate": 1.0966999934833484e-06,
+      "loss": 0.042,
+      "step": 5012
+    },
+    {
+      "epoch": 1.139318181818182,
+      "grad_norm": 2.973633016991446,
+      "learning_rate": 1.0966414366114864e-06,
+      "loss": 0.0838,
+      "step": 5013
+    },
+    {
+      "epoch": 1.1395454545454546,
+      "grad_norm": 1.6803260148295083,
+      "learning_rate": 1.0965828701220326e-06,
+      "loss": 0.0313,
+      "step": 5014
+    },
+    {
+      "epoch": 1.1397727272727272,
+      "grad_norm": 1.7425380195311566,
+      "learning_rate": 1.0965242940161815e-06,
+      "loss": 0.0543,
+      "step": 5015
+    },
+    {
+      "epoch": 1.1400000000000001,
+      "grad_norm": 2.4630446830648953,
+      "learning_rate": 1.0964657082951274e-06,
+      "loss": 0.0675,
+      "step": 5016
+    },
+    {
+      "epoch": 1.1402272727272726,
+      "grad_norm": 2.3125501636279484,
+      "learning_rate": 1.096407112960065e-06,
+      "loss": 0.0657,
+      "step": 5017
+    },
+    {
+      "epoch": 1.1404545454545454,
+      "grad_norm": 1.9546781370565112,
+      "learning_rate": 1.0963485080121892e-06,
+      "loss": 0.0946,
+      "step": 5018
+    },
+    {
+      "epoch": 1.1406818181818181,
+      "grad_norm": 1.841810016209379,
+      "learning_rate": 1.0962898934526947e-06,
+      "loss": 0.043,
+      "step": 5019
+    },
+    {
+      "epoch": 1.1409090909090909,
+      "grad_norm": 3.303023967189705,
+      "learning_rate": 1.0962312692827775e-06,
+      "loss": 0.0967,
+      "step": 5020
+    },
+    {
+      "epoch": 1.1411363636363636,
+      "grad_norm": 1.0223695099020857,
+      "learning_rate": 1.0961726355036324e-06,
+      "loss": 0.0215,
+      "step": 5021
+    },
+    {
+      "epoch": 1.1413636363636364,
+      "grad_norm": 2.249208682576478,
+      "learning_rate": 1.096113992116455e-06,
+      "loss": 0.0642,
+      "step": 5022
+    },
+    {
+      "epoch": 1.141590909090909,
+      "grad_norm": 4.552071441878801,
+      "learning_rate": 1.0960553391224419e-06,
+      "loss": 0.0532,
+      "step": 5023
+    },
+    {
+      "epoch": 1.1418181818181818,
+      "grad_norm": 1.5817965413870598,
+      "learning_rate": 1.0959966765227883e-06,
+      "loss": 0.0437,
+      "step": 5024
+    },
+    {
+      "epoch": 1.1420454545454546,
+      "grad_norm": 1.6314490449148602,
+      "learning_rate": 1.0959380043186906e-06,
+      "loss": 0.0392,
+      "step": 5025
+    },
+    {
+      "epoch": 1.1422727272727273,
+      "grad_norm": 2.765758425918672,
+      "learning_rate": 1.0958793225113459e-06,
+      "loss": 0.122,
+      "step": 5026
+    },
+    {
+      "epoch": 1.1425,
+      "grad_norm": 1.931900264354739,
+      "learning_rate": 1.0958206311019499e-06,
+      "loss": 0.0773,
+      "step": 5027
+    },
+    {
+      "epoch": 1.1427272727272728,
+      "grad_norm": 1.3182762286701026,
+      "learning_rate": 1.0957619300917001e-06,
+      "loss": 0.0408,
+      "step": 5028
+    },
+    {
+      "epoch": 1.1429545454545456,
+      "grad_norm": 2.3861897605484734,
+      "learning_rate": 1.0957032194817933e-06,
+      "loss": 0.0721,
+      "step": 5029
+    },
+    {
+      "epoch": 1.143181818181818,
+      "grad_norm": 2.38490355509836,
+      "learning_rate": 1.0956444992734264e-06,
+      "loss": 0.0463,
+      "step": 5030
+    },
+    {
+      "epoch": 1.1434090909090908,
+      "grad_norm": 1.5449383842802626,
+      "learning_rate": 1.0955857694677971e-06,
+      "loss": 0.0358,
+      "step": 5031
+    },
+    {
+      "epoch": 1.1436363636363636,
+      "grad_norm": 2.614518863318884,
+      "learning_rate": 1.095527030066103e-06,
+      "loss": 0.0327,
+      "step": 5032
+    },
+    {
+      "epoch": 1.1438636363636363,
+      "grad_norm": 1.7145031101756034,
+      "learning_rate": 1.0954682810695422e-06,
+      "loss": 0.0343,
+      "step": 5033
+    },
+    {
+      "epoch": 1.144090909090909,
+      "grad_norm": 1.5783169041259,
+      "learning_rate": 1.0954095224793122e-06,
+      "loss": 0.0457,
+      "step": 5034
+    },
+    {
+      "epoch": 1.1443181818181818,
+      "grad_norm": 1.8955232967398636,
+      "learning_rate": 1.0953507542966113e-06,
+      "loss": 0.0779,
+      "step": 5035
+    },
+    {
+      "epoch": 1.1445454545454545,
+      "grad_norm": 1.2760556038338748,
+      "learning_rate": 1.0952919765226378e-06,
+      "loss": 0.0655,
+      "step": 5036
+    },
+    {
+      "epoch": 1.1447727272727273,
+      "grad_norm": 1.6953733851912518,
+      "learning_rate": 1.0952331891585905e-06,
+      "loss": 0.0262,
+      "step": 5037
+    },
+    {
+      "epoch": 1.145,
+      "grad_norm": 1.4210063854881803,
+      "learning_rate": 1.0951743922056683e-06,
+      "loss": 0.0482,
+      "step": 5038
+    },
+    {
+      "epoch": 1.1452272727272728,
+      "grad_norm": 0.9474888911851631,
+      "learning_rate": 1.0951155856650697e-06,
+      "loss": 0.0225,
+      "step": 5039
+    },
+    {
+      "epoch": 1.1454545454545455,
+      "grad_norm": 0.5484659916881125,
+      "learning_rate": 1.0950567695379943e-06,
+      "loss": 0.0212,
+      "step": 5040
+    },
+    {
+      "epoch": 1.1456818181818182,
+      "grad_norm": 2.1529091226078116,
+      "learning_rate": 1.0949979438256415e-06,
+      "loss": 0.0528,
+      "step": 5041
+    },
+    {
+      "epoch": 1.145909090909091,
+      "grad_norm": 1.9142152776515304,
+      "learning_rate": 1.0949391085292104e-06,
+      "loss": 0.054,
+      "step": 5042
+    },
+    {
+      "epoch": 1.1461363636363637,
+      "grad_norm": 2.1442157700752573,
+      "learning_rate": 1.094880263649901e-06,
+      "loss": 0.0557,
+      "step": 5043
+    },
+    {
+      "epoch": 1.1463636363636365,
+      "grad_norm": 1.5789467624302242,
+      "learning_rate": 1.0948214091889135e-06,
+      "loss": 0.043,
+      "step": 5044
+    },
+    {
+      "epoch": 1.146590909090909,
+      "grad_norm": 2.4821983619552763,
+      "learning_rate": 1.0947625451474475e-06,
+      "loss": 0.0486,
+      "step": 5045
+    },
+    {
+      "epoch": 1.1468181818181817,
+      "grad_norm": 1.9730118251319657,
+      "learning_rate": 1.0947036715267039e-06,
+      "loss": 0.0818,
+      "step": 5046
+    },
+    {
+      "epoch": 1.1470454545454545,
+      "grad_norm": 1.470780826758543,
+      "learning_rate": 1.094644788327883e-06,
+      "loss": 0.0308,
+      "step": 5047
+    },
+    {
+      "epoch": 1.1472727272727272,
+      "grad_norm": 1.592270431994418,
+      "learning_rate": 1.0945858955521855e-06,
+      "loss": 0.0442,
+      "step": 5048
+    },
+    {
+      "epoch": 1.1475,
+      "grad_norm": 3.5543362668137215,
+      "learning_rate": 1.0945269932008123e-06,
+      "loss": 0.0852,
+      "step": 5049
+    },
+    {
+      "epoch": 1.1477272727272727,
+      "grad_norm": 2.3219793307736127,
+      "learning_rate": 1.0944680812749647e-06,
+      "loss": 0.0767,
+      "step": 5050
+    },
+    {
+      "epoch": 1.1479545454545454,
+      "grad_norm": 2.2599223082821314,
+      "learning_rate": 1.0944091597758438e-06,
+      "loss": 0.0481,
+      "step": 5051
+    },
+    {
+      "epoch": 1.1481818181818182,
+      "grad_norm": 2.368632461511502,
+      "learning_rate": 1.094350228704651e-06,
+      "loss": 0.0614,
+      "step": 5052
+    },
+    {
+      "epoch": 1.148409090909091,
+      "grad_norm": 1.476499783147443,
+      "learning_rate": 1.0942912880625885e-06,
+      "loss": 0.0392,
+      "step": 5053
+    },
+    {
+      "epoch": 1.1486363636363637,
+      "grad_norm": 0.9326263174304582,
+      "learning_rate": 1.0942323378508579e-06,
+      "loss": 0.0213,
+      "step": 5054
+    },
+    {
+      "epoch": 1.1488636363636364,
+      "grad_norm": 2.4275563448351667,
+      "learning_rate": 1.094173378070661e-06,
+      "loss": 0.0697,
+      "step": 5055
+    },
+    {
+      "epoch": 1.1490909090909092,
+      "grad_norm": 1.920403777349689,
+      "learning_rate": 1.0941144087232008e-06,
+      "loss": 0.0594,
+      "step": 5056
+    },
+    {
+      "epoch": 1.149318181818182,
+      "grad_norm": 2.98238775909552,
+      "learning_rate": 1.0940554298096793e-06,
+      "loss": 0.0813,
+      "step": 5057
+    },
+    {
+      "epoch": 1.1495454545454544,
+      "grad_norm": 1.357761869042494,
+      "learning_rate": 1.093996441331299e-06,
+      "loss": 0.0309,
+      "step": 5058
+    },
+    {
+      "epoch": 1.1497727272727274,
+      "grad_norm": 2.169005354776055,
+      "learning_rate": 1.0939374432892632e-06,
+      "loss": 0.0638,
+      "step": 5059
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.859159697443024,
+      "learning_rate": 1.093878435684775e-06,
+      "loss": 0.0763,
+      "step": 5060
+    },
+    {
+      "epoch": 1.1502272727272727,
+      "grad_norm": 1.8808317040772302,
+      "learning_rate": 1.0938194185190374e-06,
+      "loss": 0.0449,
+      "step": 5061
+    },
+    {
+      "epoch": 1.1504545454545454,
+      "grad_norm": 1.6635105089444204,
+      "learning_rate": 1.0937603917932538e-06,
+      "loss": 0.033,
+      "step": 5062
+    },
+    {
+      "epoch": 1.1506818181818181,
+      "grad_norm": 1.9141034641353405,
+      "learning_rate": 1.0937013555086282e-06,
+      "loss": 0.1068,
+      "step": 5063
+    },
+    {
+      "epoch": 1.1509090909090909,
+      "grad_norm": 1.8460101347411895,
+      "learning_rate": 1.0936423096663643e-06,
+      "loss": 0.0554,
+      "step": 5064
+    },
+    {
+      "epoch": 1.1511363636363636,
+      "grad_norm": 2.6584243579862394,
+      "learning_rate": 1.0935832542676658e-06,
+      "loss": 0.0733,
+      "step": 5065
+    },
+    {
+      "epoch": 1.1513636363636364,
+      "grad_norm": 1.6141573718756972,
+      "learning_rate": 1.0935241893137376e-06,
+      "loss": 0.0259,
+      "step": 5066
+    },
+    {
+      "epoch": 1.151590909090909,
+      "grad_norm": 0.5949382462215499,
+      "learning_rate": 1.0934651148057838e-06,
+      "loss": 0.0239,
+      "step": 5067
+    },
+    {
+      "epoch": 1.1518181818181819,
+      "grad_norm": 1.5084942490274709,
+      "learning_rate": 1.0934060307450087e-06,
+      "loss": 0.0366,
+      "step": 5068
+    },
+    {
+      "epoch": 1.1520454545454546,
+      "grad_norm": 1.5751208332265745,
+      "learning_rate": 1.0933469371326177e-06,
+      "loss": 0.0659,
+      "step": 5069
+    },
+    {
+      "epoch": 1.1522727272727273,
+      "grad_norm": 3.427577265736409,
+      "learning_rate": 1.0932878339698154e-06,
+      "loss": 0.0691,
+      "step": 5070
+    },
+    {
+      "epoch": 1.1525,
+      "grad_norm": 2.220562507659086,
+      "learning_rate": 1.0932287212578075e-06,
+      "loss": 0.0796,
+      "step": 5071
+    },
+    {
+      "epoch": 1.1527272727272728,
+      "grad_norm": 2.343984081429403,
+      "learning_rate": 1.093169598997799e-06,
+      "loss": 0.0546,
+      "step": 5072
+    },
+    {
+      "epoch": 1.1529545454545453,
+      "grad_norm": 1.619143783313097,
+      "learning_rate": 1.0931104671909953e-06,
+      "loss": 0.0236,
+      "step": 5073
+    },
+    {
+      "epoch": 1.153181818181818,
+      "grad_norm": 1.6304009901081276,
+      "learning_rate": 1.0930513258386027e-06,
+      "loss": 0.0515,
+      "step": 5074
+    },
+    {
+      "epoch": 1.1534090909090908,
+      "grad_norm": 2.0849452691646744,
+      "learning_rate": 1.092992174941827e-06,
+      "loss": 0.0355,
+      "step": 5075
+    },
+    {
+      "epoch": 1.1536363636363636,
+      "grad_norm": 2.531389327482295,
+      "learning_rate": 1.0929330145018747e-06,
+      "loss": 0.0967,
+      "step": 5076
+    },
+    {
+      "epoch": 1.1538636363636363,
+      "grad_norm": 1.867241601400514,
+      "learning_rate": 1.0928738445199515e-06,
+      "loss": 0.0713,
+      "step": 5077
+    },
+    {
+      "epoch": 1.154090909090909,
+      "grad_norm": 1.8795741391670542,
+      "learning_rate": 1.0928146649972646e-06,
+      "loss": 0.0661,
+      "step": 5078
+    },
+    {
+      "epoch": 1.1543181818181818,
+      "grad_norm": 0.7849916522008997,
+      "learning_rate": 1.0927554759350203e-06,
+      "loss": 0.0213,
+      "step": 5079
+    },
+    {
+      "epoch": 1.1545454545454545,
+      "grad_norm": 2.496493504965869,
+      "learning_rate": 1.092696277334426e-06,
+      "loss": 0.0463,
+      "step": 5080
+    },
+    {
+      "epoch": 1.1547727272727273,
+      "grad_norm": 1.222493799162277,
+      "learning_rate": 1.0926370691966883e-06,
+      "loss": 0.0363,
+      "step": 5081
+    },
+    {
+      "epoch": 1.155,
+      "grad_norm": 2.1491022529084,
+      "learning_rate": 1.0925778515230155e-06,
+      "loss": 0.0555,
+      "step": 5082
+    },
+    {
+      "epoch": 1.1552272727272728,
+      "grad_norm": 1.7653427474938752,
+      "learning_rate": 1.092518624314614e-06,
+      "loss": 0.0816,
+      "step": 5083
+    },
+    {
+      "epoch": 1.1554545454545455,
+      "grad_norm": 1.2717074287084165,
+      "learning_rate": 1.0924593875726926e-06,
+      "loss": 0.056,
+      "step": 5084
+    },
+    {
+      "epoch": 1.1556818181818183,
+      "grad_norm": 1.443954594361313,
+      "learning_rate": 1.0924001412984585e-06,
+      "loss": 0.0488,
+      "step": 5085
+    },
+    {
+      "epoch": 1.155909090909091,
+      "grad_norm": 1.6807410354088965,
+      "learning_rate": 1.0923408854931202e-06,
+      "loss": 0.047,
+      "step": 5086
+    },
+    {
+      "epoch": 1.1561363636363637,
+      "grad_norm": 2.4282032207571524,
+      "learning_rate": 1.0922816201578858e-06,
+      "loss": 0.0499,
+      "step": 5087
+    },
+    {
+      "epoch": 1.1563636363636363,
+      "grad_norm": 1.9123997833112043,
+      "learning_rate": 1.092222345293964e-06,
+      "loss": 0.0467,
+      "step": 5088
+    },
+    {
+      "epoch": 1.156590909090909,
+      "grad_norm": 1.7163404190498956,
+      "learning_rate": 1.0921630609025635e-06,
+      "loss": 0.0785,
+      "step": 5089
+    },
+    {
+      "epoch": 1.1568181818181817,
+      "grad_norm": 2.184612411709692,
+      "learning_rate": 1.0921037669848933e-06,
+      "loss": 0.0795,
+      "step": 5090
+    },
+    {
+      "epoch": 1.1570454545454545,
+      "grad_norm": 1.507590504362358,
+      "learning_rate": 1.0920444635421622e-06,
+      "loss": 0.0286,
+      "step": 5091
+    },
+    {
+      "epoch": 1.1572727272727272,
+      "grad_norm": 1.66138110269742,
+      "learning_rate": 1.0919851505755796e-06,
+      "loss": 0.0446,
+      "step": 5092
+    },
+    {
+      "epoch": 1.1575,
+      "grad_norm": 1.3217382600551768,
+      "learning_rate": 1.0919258280863554e-06,
+      "loss": 0.0228,
+      "step": 5093
+    },
+    {
+      "epoch": 1.1577272727272727,
+      "grad_norm": 1.594697984673048,
+      "learning_rate": 1.0918664960756986e-06,
+      "loss": 0.0481,
+      "step": 5094
+    },
+    {
+      "epoch": 1.1579545454545455,
+      "grad_norm": 2.5565335983247386,
+      "learning_rate": 1.0918071545448198e-06,
+      "loss": 0.0865,
+      "step": 5095
+    },
+    {
+      "epoch": 1.1581818181818182,
+      "grad_norm": 2.102035205837115,
+      "learning_rate": 1.0917478034949285e-06,
+      "loss": 0.0338,
+      "step": 5096
+    },
+    {
+      "epoch": 1.158409090909091,
+      "grad_norm": 2.062104171364163,
+      "learning_rate": 1.0916884429272352e-06,
+      "loss": 0.0693,
+      "step": 5097
+    },
+    {
+      "epoch": 1.1586363636363637,
+      "grad_norm": 2.0324742577819537,
+      "learning_rate": 1.0916290728429505e-06,
+      "loss": 0.1016,
+      "step": 5098
+    },
+    {
+      "epoch": 1.1588636363636364,
+      "grad_norm": 1.9182190278994002,
+      "learning_rate": 1.0915696932432847e-06,
+      "loss": 0.0332,
+      "step": 5099
+    },
+    {
+      "epoch": 1.1590909090909092,
+      "grad_norm": 1.347313399131043,
+      "learning_rate": 1.0915103041294492e-06,
+      "loss": 0.024,
+      "step": 5100
+    },
+    {
+      "epoch": 1.159318181818182,
+      "grad_norm": 2.935136177700199,
+      "learning_rate": 1.0914509055026545e-06,
+      "loss": 0.1346,
+      "step": 5101
+    },
+    {
+      "epoch": 1.1595454545454547,
+      "grad_norm": 1.5142236724047984,
+      "learning_rate": 1.0913914973641122e-06,
+      "loss": 0.0533,
+      "step": 5102
+    },
+    {
+      "epoch": 1.1597727272727272,
+      "grad_norm": 2.137328777779087,
+      "learning_rate": 1.0913320797150335e-06,
+      "loss": 0.0596,
+      "step": 5103
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.3191266717152796,
+      "learning_rate": 1.0912726525566302e-06,
+      "loss": 0.0601,
+      "step": 5104
+    },
+    {
+      "epoch": 1.1602272727272727,
+      "grad_norm": 1.5510932322529758,
+      "learning_rate": 1.091213215890114e-06,
+      "loss": 0.0269,
+      "step": 5105
+    },
+    {
+      "epoch": 1.1604545454545454,
+      "grad_norm": 1.9753187696772005,
+      "learning_rate": 1.0911537697166967e-06,
+      "loss": 0.0358,
+      "step": 5106
+    },
+    {
+      "epoch": 1.1606818181818181,
+      "grad_norm": 1.2718986423834242,
+      "learning_rate": 1.0910943140375913e-06,
+      "loss": 0.0476,
+      "step": 5107
+    },
+    {
+      "epoch": 1.160909090909091,
+      "grad_norm": 2.6297670026821756,
+      "learning_rate": 1.0910348488540093e-06,
+      "loss": 0.0888,
+      "step": 5108
+    },
+    {
+      "epoch": 1.1611363636363636,
+      "grad_norm": 1.841786792840535,
+      "learning_rate": 1.090975374167164e-06,
+      "loss": 0.0359,
+      "step": 5109
+    },
+    {
+      "epoch": 1.1613636363636364,
+      "grad_norm": 0.9845039279983854,
+      "learning_rate": 1.0909158899782677e-06,
+      "loss": 0.0195,
+      "step": 5110
+    },
+    {
+      "epoch": 1.1615909090909091,
+      "grad_norm": 1.6017887852364958,
+      "learning_rate": 1.0908563962885337e-06,
+      "loss": 0.0364,
+      "step": 5111
+    },
+    {
+      "epoch": 1.1618181818181819,
+      "grad_norm": 1.5872145985387398,
+      "learning_rate": 1.090796893099175e-06,
+      "loss": 0.0392,
+      "step": 5112
+    },
+    {
+      "epoch": 1.1620454545454546,
+      "grad_norm": 1.7865134931307511,
+      "learning_rate": 1.090737380411405e-06,
+      "loss": 0.0352,
+      "step": 5113
+    },
+    {
+      "epoch": 1.1622727272727273,
+      "grad_norm": 2.4108078723659174,
+      "learning_rate": 1.0906778582264372e-06,
+      "loss": 0.0582,
+      "step": 5114
+    },
+    {
+      "epoch": 1.1625,
+      "grad_norm": 2.630760583610465,
+      "learning_rate": 1.0906183265454856e-06,
+      "loss": 0.0713,
+      "step": 5115
+    },
+    {
+      "epoch": 1.1627272727272726,
+      "grad_norm": 1.9078905376216406,
+      "learning_rate": 1.0905587853697638e-06,
+      "loss": 0.0753,
+      "step": 5116
+    },
+    {
+      "epoch": 1.1629545454545456,
+      "grad_norm": 2.33887922590744,
+      "learning_rate": 1.0904992347004865e-06,
+      "loss": 0.0538,
+      "step": 5117
+    },
+    {
+      "epoch": 1.163181818181818,
+      "grad_norm": 3.1048224504099613,
+      "learning_rate": 1.0904396745388674e-06,
+      "loss": 0.0545,
+      "step": 5118
+    },
+    {
+      "epoch": 1.1634090909090908,
+      "grad_norm": 1.4756379635009635,
+      "learning_rate": 1.0903801048861214e-06,
+      "loss": 0.0408,
+      "step": 5119
+    },
+    {
+      "epoch": 1.1636363636363636,
+      "grad_norm": 1.2129763851204427,
+      "learning_rate": 1.0903205257434632e-06,
+      "loss": 0.0317,
+      "step": 5120
+    },
+    {
+      "epoch": 1.1638636363636363,
+      "grad_norm": 1.6714875215204756,
+      "learning_rate": 1.0902609371121076e-06,
+      "loss": 0.0247,
+      "step": 5121
+    },
+    {
+      "epoch": 1.164090909090909,
+      "grad_norm": 2.013380422893122,
+      "learning_rate": 1.09020133899327e-06,
+      "loss": 0.0358,
+      "step": 5122
+    },
+    {
+      "epoch": 1.1643181818181818,
+      "grad_norm": 1.996161014171097,
+      "learning_rate": 1.0901417313881652e-06,
+      "loss": 0.0973,
+      "step": 5123
+    },
+    {
+      "epoch": 1.1645454545454546,
+      "grad_norm": 1.8586367477589767,
+      "learning_rate": 1.0900821142980091e-06,
+      "loss": 0.049,
+      "step": 5124
+    },
+    {
+      "epoch": 1.1647727272727273,
+      "grad_norm": 0.8572275435470702,
+      "learning_rate": 1.0900224877240176e-06,
+      "loss": 0.0205,
+      "step": 5125
+    },
+    {
+      "epoch": 1.165,
+      "grad_norm": 1.4062904059948114,
+      "learning_rate": 1.089962851667406e-06,
+      "loss": 0.0221,
+      "step": 5126
+    },
+    {
+      "epoch": 1.1652272727272728,
+      "grad_norm": 1.9367360569701295,
+      "learning_rate": 1.0899032061293908e-06,
+      "loss": 0.0602,
+      "step": 5127
+    },
+    {
+      "epoch": 1.1654545454545455,
+      "grad_norm": 1.4409472171746869,
+      "learning_rate": 1.0898435511111881e-06,
+      "loss": 0.023,
+      "step": 5128
+    },
+    {
+      "epoch": 1.1656818181818183,
+      "grad_norm": 1.9976534031293283,
+      "learning_rate": 1.0897838866140145e-06,
+      "loss": 0.0799,
+      "step": 5129
+    },
+    {
+      "epoch": 1.165909090909091,
+      "grad_norm": 1.454304098329238,
+      "learning_rate": 1.0897242126390864e-06,
+      "loss": 0.0416,
+      "step": 5130
+    },
+    {
+      "epoch": 1.1661363636363635,
+      "grad_norm": 2.6928723447363265,
+      "learning_rate": 1.089664529187621e-06,
+      "loss": 0.0964,
+      "step": 5131
+    },
+    {
+      "epoch": 1.1663636363636363,
+      "grad_norm": 1.5223210639525748,
+      "learning_rate": 1.0896048362608352e-06,
+      "loss": 0.0336,
+      "step": 5132
+    },
+    {
+      "epoch": 1.166590909090909,
+      "grad_norm": 1.3594045476011392,
+      "learning_rate": 1.0895451338599464e-06,
+      "loss": 0.0315,
+      "step": 5133
+    },
+    {
+      "epoch": 1.1668181818181818,
+      "grad_norm": 3.9853636779804558,
+      "learning_rate": 1.0894854219861717e-06,
+      "loss": 0.0917,
+      "step": 5134
+    },
+    {
+      "epoch": 1.1670454545454545,
+      "grad_norm": 2.0043623039851926,
+      "learning_rate": 1.089425700640729e-06,
+      "loss": 0.0523,
+      "step": 5135
+    },
+    {
+      "epoch": 1.1672727272727272,
+      "grad_norm": 3.1860646556112315,
+      "learning_rate": 1.0893659698248358e-06,
+      "loss": 0.0631,
+      "step": 5136
+    },
+    {
+      "epoch": 1.1675,
+      "grad_norm": 1.507265373918131,
+      "learning_rate": 1.0893062295397105e-06,
+      "loss": 0.0247,
+      "step": 5137
+    },
+    {
+      "epoch": 1.1677272727272727,
+      "grad_norm": 1.7206897120722071,
+      "learning_rate": 1.0892464797865711e-06,
+      "loss": 0.0343,
+      "step": 5138
+    },
+    {
+      "epoch": 1.1679545454545455,
+      "grad_norm": 2.1748626750964397,
+      "learning_rate": 1.0891867205666362e-06,
+      "loss": 0.0502,
+      "step": 5139
+    },
+    {
+      "epoch": 1.1681818181818182,
+      "grad_norm": 2.052116912680236,
+      "learning_rate": 1.089126951881124e-06,
+      "loss": 0.0459,
+      "step": 5140
+    },
+    {
+      "epoch": 1.168409090909091,
+      "grad_norm": 2.406320782044234,
+      "learning_rate": 1.0890671737312538e-06,
+      "loss": 0.1051,
+      "step": 5141
+    },
+    {
+      "epoch": 1.1686363636363637,
+      "grad_norm": 1.0748113964945492,
+      "learning_rate": 1.0890073861182443e-06,
+      "loss": 0.0244,
+      "step": 5142
+    },
+    {
+      "epoch": 1.1688636363636364,
+      "grad_norm": 1.4857506270664127,
+      "learning_rate": 1.0889475890433145e-06,
+      "loss": 0.0786,
+      "step": 5143
+    },
+    {
+      "epoch": 1.1690909090909092,
+      "grad_norm": 2.6039602869170695,
+      "learning_rate": 1.088887782507684e-06,
+      "loss": 0.043,
+      "step": 5144
+    },
+    {
+      "epoch": 1.169318181818182,
+      "grad_norm": 1.5916269688822686,
+      "learning_rate": 1.0888279665125725e-06,
+      "loss": 0.0617,
+      "step": 5145
+    },
+    {
+      "epoch": 1.1695454545454544,
+      "grad_norm": 1.882786419787967,
+      "learning_rate": 1.0887681410591994e-06,
+      "loss": 0.0611,
+      "step": 5146
+    },
+    {
+      "epoch": 1.1697727272727272,
+      "grad_norm": 1.9953882526597417,
+      "learning_rate": 1.0887083061487848e-06,
+      "loss": 0.074,
+      "step": 5147
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.720715860742873,
+      "learning_rate": 1.0886484617825491e-06,
+      "loss": 0.0736,
+      "step": 5148
+    },
+    {
+      "epoch": 1.1702272727272727,
+      "grad_norm": 1.723736921215915,
+      "learning_rate": 1.088588607961712e-06,
+      "loss": 0.0346,
+      "step": 5149
+    },
+    {
+      "epoch": 1.1704545454545454,
+      "grad_norm": 1.0905410010838446,
+      "learning_rate": 1.0885287446874946e-06,
+      "loss": 0.0328,
+      "step": 5150
+    },
+    {
+      "epoch": 1.1706818181818182,
+      "grad_norm": 2.739052904266121,
+      "learning_rate": 1.0884688719611176e-06,
+      "loss": 0.0694,
+      "step": 5151
+    },
+    {
+      "epoch": 1.170909090909091,
+      "grad_norm": 1.5980696893052533,
+      "learning_rate": 1.0884089897838016e-06,
+      "loss": 0.0698,
+      "step": 5152
+    },
+    {
+      "epoch": 1.1711363636363636,
+      "grad_norm": 1.4398844317725827,
+      "learning_rate": 1.0883490981567676e-06,
+      "loss": 0.071,
+      "step": 5153
+    },
+    {
+      "epoch": 1.1713636363636364,
+      "grad_norm": 1.5371208039836883,
+      "learning_rate": 1.0882891970812374e-06,
+      "loss": 0.032,
+      "step": 5154
+    },
+    {
+      "epoch": 1.1715909090909091,
+      "grad_norm": 2.5141219612102046,
+      "learning_rate": 1.0882292865584323e-06,
+      "loss": 0.0596,
+      "step": 5155
+    },
+    {
+      "epoch": 1.1718181818181819,
+      "grad_norm": 2.657163041031824,
+      "learning_rate": 1.0881693665895737e-06,
+      "loss": 0.063,
+      "step": 5156
+    },
+    {
+      "epoch": 1.1720454545454546,
+      "grad_norm": 1.8894123734351775,
+      "learning_rate": 1.0881094371758839e-06,
+      "loss": 0.0361,
+      "step": 5157
+    },
+    {
+      "epoch": 1.1722727272727274,
+      "grad_norm": 1.713992365370292,
+      "learning_rate": 1.0880494983185843e-06,
+      "loss": 0.0344,
+      "step": 5158
+    },
+    {
+      "epoch": 1.1724999999999999,
+      "grad_norm": 1.5411044654134718,
+      "learning_rate": 1.0879895500188982e-06,
+      "loss": 0.0712,
+      "step": 5159
+    },
+    {
+      "epoch": 1.1727272727272728,
+      "grad_norm": 1.6681387291662777,
+      "learning_rate": 1.087929592278047e-06,
+      "loss": 0.0343,
+      "step": 5160
+    },
+    {
+      "epoch": 1.1729545454545454,
+      "grad_norm": 2.067918461679956,
+      "learning_rate": 1.0878696250972536e-06,
+      "loss": 0.0579,
+      "step": 5161
+    },
+    {
+      "epoch": 1.173181818181818,
+      "grad_norm": 2.1133543332542235,
+      "learning_rate": 1.0878096484777414e-06,
+      "loss": 0.0802,
+      "step": 5162
+    },
+    {
+      "epoch": 1.1734090909090908,
+      "grad_norm": 2.2936069426201993,
+      "learning_rate": 1.0877496624207333e-06,
+      "loss": 0.0651,
+      "step": 5163
+    },
+    {
+      "epoch": 1.1736363636363636,
+      "grad_norm": 2.595634321241766,
+      "learning_rate": 1.0876896669274517e-06,
+      "loss": 0.0822,
+      "step": 5164
+    },
+    {
+      "epoch": 1.1738636363636363,
+      "grad_norm": 2.310441731990821,
+      "learning_rate": 1.0876296619991208e-06,
+      "loss": 0.0658,
+      "step": 5165
+    },
+    {
+      "epoch": 1.174090909090909,
+      "grad_norm": 0.7594378023440516,
+      "learning_rate": 1.087569647636964e-06,
+      "loss": 0.0156,
+      "step": 5166
+    },
+    {
+      "epoch": 1.1743181818181818,
+      "grad_norm": 2.6434575918988994,
+      "learning_rate": 1.0875096238422051e-06,
+      "loss": 0.0729,
+      "step": 5167
+    },
+    {
+      "epoch": 1.1745454545454546,
+      "grad_norm": 2.3001634887078093,
+      "learning_rate": 1.087449590616068e-06,
+      "loss": 0.0604,
+      "step": 5168
+    },
+    {
+      "epoch": 1.1747727272727273,
+      "grad_norm": 1.2919050396200233,
+      "learning_rate": 1.0873895479597773e-06,
+      "loss": 0.033,
+      "step": 5169
+    },
+    {
+      "epoch": 1.175,
+      "grad_norm": 3.273048982929147,
+      "learning_rate": 1.0873294958745568e-06,
+      "loss": 0.0585,
+      "step": 5170
+    },
+    {
+      "epoch": 1.1752272727272728,
+      "grad_norm": 3.577808092121959,
+      "learning_rate": 1.0872694343616312e-06,
+      "loss": 0.0666,
+      "step": 5171
+    },
+    {
+      "epoch": 1.1754545454545455,
+      "grad_norm": 2.0706532944470237,
+      "learning_rate": 1.0872093634222255e-06,
+      "loss": 0.0463,
+      "step": 5172
+    },
+    {
+      "epoch": 1.1756818181818183,
+      "grad_norm": 1.776522978103439,
+      "learning_rate": 1.0871492830575642e-06,
+      "loss": 0.048,
+      "step": 5173
+    },
+    {
+      "epoch": 1.1759090909090908,
+      "grad_norm": 1.8254932044672114,
+      "learning_rate": 1.0870891932688733e-06,
+      "loss": 0.0431,
+      "step": 5174
+    },
+    {
+      "epoch": 1.1761363636363638,
+      "grad_norm": 2.9848322735159605,
+      "learning_rate": 1.0870290940573772e-06,
+      "loss": 0.069,
+      "step": 5175
+    },
+    {
+      "epoch": 1.1763636363636363,
+      "grad_norm": 1.877703236947816,
+      "learning_rate": 1.0869689854243019e-06,
+      "loss": 0.0462,
+      "step": 5176
+    },
+    {
+      "epoch": 1.176590909090909,
+      "grad_norm": 2.0337181515987273,
+      "learning_rate": 1.086908867370873e-06,
+      "loss": 0.0487,
+      "step": 5177
+    },
+    {
+      "epoch": 1.1768181818181818,
+      "grad_norm": 2.117821341036171,
+      "learning_rate": 1.0868487398983166e-06,
+      "loss": 0.0639,
+      "step": 5178
+    },
+    {
+      "epoch": 1.1770454545454545,
+      "grad_norm": 2.24384406535897,
+      "learning_rate": 1.0867886030078585e-06,
+      "loss": 0.0806,
+      "step": 5179
+    },
+    {
+      "epoch": 1.1772727272727272,
+      "grad_norm": 1.4653014530637805,
+      "learning_rate": 1.0867284567007252e-06,
+      "loss": 0.0382,
+      "step": 5180
+    },
+    {
+      "epoch": 1.1775,
+      "grad_norm": 2.2111068381652035,
+      "learning_rate": 1.0866683009781432e-06,
+      "loss": 0.0523,
+      "step": 5181
+    },
+    {
+      "epoch": 1.1777272727272727,
+      "grad_norm": 3.1333480649035526,
+      "learning_rate": 1.086608135841339e-06,
+      "loss": 0.0697,
+      "step": 5182
+    },
+    {
+      "epoch": 1.1779545454545455,
+      "grad_norm": 1.8588313038623643,
+      "learning_rate": 1.0865479612915398e-06,
+      "loss": 0.0467,
+      "step": 5183
+    },
+    {
+      "epoch": 1.1781818181818182,
+      "grad_norm": 2.7078592879232244,
+      "learning_rate": 1.0864877773299724e-06,
+      "loss": 0.0413,
+      "step": 5184
+    },
+    {
+      "epoch": 1.178409090909091,
+      "grad_norm": 2.593670689690114,
+      "learning_rate": 1.0864275839578637e-06,
+      "loss": 0.0477,
+      "step": 5185
+    },
+    {
+      "epoch": 1.1786363636363637,
+      "grad_norm": 1.8694518361156613,
+      "learning_rate": 1.0863673811764419e-06,
+      "loss": 0.0628,
+      "step": 5186
+    },
+    {
+      "epoch": 1.1788636363636364,
+      "grad_norm": 1.2783744234860879,
+      "learning_rate": 1.0863071689869345e-06,
+      "loss": 0.03,
+      "step": 5187
+    },
+    {
+      "epoch": 1.1790909090909092,
+      "grad_norm": 1.6639897803611134,
+      "learning_rate": 1.0862469473905688e-06,
+      "loss": 0.0343,
+      "step": 5188
+    },
+    {
+      "epoch": 1.1793181818181817,
+      "grad_norm": 1.5979979251741798,
+      "learning_rate": 1.0861867163885731e-06,
+      "loss": 0.0283,
+      "step": 5189
+    },
+    {
+      "epoch": 1.1795454545454545,
+      "grad_norm": 3.0223354560934426,
+      "learning_rate": 1.086126475982176e-06,
+      "loss": 0.0551,
+      "step": 5190
+    },
+    {
+      "epoch": 1.1797727272727272,
+      "grad_norm": 1.7642438993429952,
+      "learning_rate": 1.0860662261726054e-06,
+      "loss": 0.036,
+      "step": 5191
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.1878582185562188,
+      "learning_rate": 1.08600596696109e-06,
+      "loss": 0.0407,
+      "step": 5192
+    },
+    {
+      "epoch": 1.1802272727272727,
+      "grad_norm": 2.161671356428624,
+      "learning_rate": 1.0859456983488588e-06,
+      "loss": 0.0597,
+      "step": 5193
+    },
+    {
+      "epoch": 1.1804545454545454,
+      "grad_norm": 1.6695546049284322,
+      "learning_rate": 1.0858854203371406e-06,
+      "loss": 0.058,
+      "step": 5194
+    },
+    {
+      "epoch": 1.1806818181818182,
+      "grad_norm": 2.4986795834890407,
+      "learning_rate": 1.0858251329271645e-06,
+      "loss": 0.0581,
+      "step": 5195
+    },
+    {
+      "epoch": 1.180909090909091,
+      "grad_norm": 1.151531270161502,
+      "learning_rate": 1.0857648361201603e-06,
+      "loss": 0.0314,
+      "step": 5196
+    },
+    {
+      "epoch": 1.1811363636363637,
+      "grad_norm": 2.4205689924151046,
+      "learning_rate": 1.0857045299173569e-06,
+      "loss": 0.0539,
+      "step": 5197
+    },
+    {
+      "epoch": 1.1813636363636364,
+      "grad_norm": 2.055541178979146,
+      "learning_rate": 1.0856442143199846e-06,
+      "loss": 0.0557,
+      "step": 5198
+    },
+    {
+      "epoch": 1.1815909090909091,
+      "grad_norm": 2.0483552597330084,
+      "learning_rate": 1.085583889329273e-06,
+      "loss": 0.0682,
+      "step": 5199
+    },
+    {
+      "epoch": 1.1818181818181819,
+      "grad_norm": 2.0934149661197003,
+      "learning_rate": 1.0855235549464523e-06,
+      "loss": 0.0443,
+      "step": 5200
+    },
+    {
+      "epoch": 1.1820454545454546,
+      "grad_norm": 1.3838177028418355,
+      "learning_rate": 1.085463211172753e-06,
+      "loss": 0.0338,
+      "step": 5201
+    },
+    {
+      "epoch": 1.1822727272727274,
+      "grad_norm": 2.52853617140007,
+      "learning_rate": 1.0854028580094057e-06,
+      "loss": 0.037,
+      "step": 5202
+    },
+    {
+      "epoch": 1.1825,
+      "grad_norm": 2.376962323304628,
+      "learning_rate": 1.0853424954576408e-06,
+      "loss": 0.0821,
+      "step": 5203
+    },
+    {
+      "epoch": 1.1827272727272726,
+      "grad_norm": 2.2300503322944563,
+      "learning_rate": 1.0852821235186892e-06,
+      "loss": 0.0902,
+      "step": 5204
+    },
+    {
+      "epoch": 1.1829545454545454,
+      "grad_norm": 2.398765833796079,
+      "learning_rate": 1.085221742193782e-06,
+      "loss": 0.0553,
+      "step": 5205
+    },
+    {
+      "epoch": 1.1831818181818181,
+      "grad_norm": 1.4047468768989952,
+      "learning_rate": 1.0851613514841508e-06,
+      "loss": 0.0317,
+      "step": 5206
+    },
+    {
+      "epoch": 1.1834090909090909,
+      "grad_norm": 1.9915487394379,
+      "learning_rate": 1.0851009513910268e-06,
+      "loss": 0.0764,
+      "step": 5207
+    },
+    {
+      "epoch": 1.1836363636363636,
+      "grad_norm": 1.9792102179925553,
+      "learning_rate": 1.0850405419156415e-06,
+      "loss": 0.0444,
+      "step": 5208
+    },
+    {
+      "epoch": 1.1838636363636363,
+      "grad_norm": 2.2709880750319194,
+      "learning_rate": 1.084980123059227e-06,
+      "loss": 0.0627,
+      "step": 5209
+    },
+    {
+      "epoch": 1.184090909090909,
+      "grad_norm": 1.3174584953004969,
+      "learning_rate": 1.0849196948230156e-06,
+      "loss": 0.0308,
+      "step": 5210
+    },
+    {
+      "epoch": 1.1843181818181818,
+      "grad_norm": 1.535270313236034,
+      "learning_rate": 1.0848592572082391e-06,
+      "loss": 0.0502,
+      "step": 5211
+    },
+    {
+      "epoch": 1.1845454545454546,
+      "grad_norm": 1.7402049807455255,
+      "learning_rate": 1.08479881021613e-06,
+      "loss": 0.0286,
+      "step": 5212
+    },
+    {
+      "epoch": 1.1847727272727273,
+      "grad_norm": 2.6811573472492056,
+      "learning_rate": 1.0847383538479211e-06,
+      "loss": 0.0456,
+      "step": 5213
+    },
+    {
+      "epoch": 1.185,
+      "grad_norm": 1.8797181801308762,
+      "learning_rate": 1.0846778881048451e-06,
+      "loss": 0.0539,
+      "step": 5214
+    },
+    {
+      "epoch": 1.1852272727272728,
+      "grad_norm": 1.7085266182905627,
+      "learning_rate": 1.084617412988135e-06,
+      "loss": 0.059,
+      "step": 5215
+    },
+    {
+      "epoch": 1.1854545454545455,
+      "grad_norm": 2.671054145090461,
+      "learning_rate": 1.084556928499024e-06,
+      "loss": 0.0655,
+      "step": 5216
+    },
+    {
+      "epoch": 1.185681818181818,
+      "grad_norm": 2.5930032715496107,
+      "learning_rate": 1.0844964346387454e-06,
+      "loss": 0.0458,
+      "step": 5217
+    },
+    {
+      "epoch": 1.185909090909091,
+      "grad_norm": 1.306561884435171,
+      "learning_rate": 1.084435931408533e-06,
+      "loss": 0.0379,
+      "step": 5218
+    },
+    {
+      "epoch": 1.1861363636363635,
+      "grad_norm": 1.6016294393679504,
+      "learning_rate": 1.0843754188096203e-06,
+      "loss": 0.0234,
+      "step": 5219
+    },
+    {
+      "epoch": 1.1863636363636363,
+      "grad_norm": 2.0516335680437114,
+      "learning_rate": 1.0843148968432416e-06,
+      "loss": 0.0958,
+      "step": 5220
+    },
+    {
+      "epoch": 1.186590909090909,
+      "grad_norm": 2.0519308217211805,
+      "learning_rate": 1.0842543655106305e-06,
+      "loss": 0.0916,
+      "step": 5221
+    },
+    {
+      "epoch": 1.1868181818181818,
+      "grad_norm": 1.5782476978317301,
+      "learning_rate": 1.0841938248130218e-06,
+      "loss": 0.0354,
+      "step": 5222
+    },
+    {
+      "epoch": 1.1870454545454545,
+      "grad_norm": 1.0780711453686451,
+      "learning_rate": 1.08413327475165e-06,
+      "loss": 0.0274,
+      "step": 5223
+    },
+    {
+      "epoch": 1.1872727272727273,
+      "grad_norm": 1.0914986243435105,
+      "learning_rate": 1.0840727153277498e-06,
+      "loss": 0.0258,
+      "step": 5224
+    },
+    {
+      "epoch": 1.1875,
+      "grad_norm": 1.4588000538695622,
+      "learning_rate": 1.084012146542556e-06,
+      "loss": 0.0315,
+      "step": 5225
+    },
+    {
+      "epoch": 1.1877272727272727,
+      "grad_norm": 2.8606419277146684,
+      "learning_rate": 1.0839515683973035e-06,
+      "loss": 0.0755,
+      "step": 5226
+    },
+    {
+      "epoch": 1.1879545454545455,
+      "grad_norm": 2.528481828560485,
+      "learning_rate": 1.083890980893228e-06,
+      "loss": 0.0483,
+      "step": 5227
+    },
+    {
+      "epoch": 1.1881818181818182,
+      "grad_norm": 2.1222624507452372,
+      "learning_rate": 1.0838303840315647e-06,
+      "loss": 0.052,
+      "step": 5228
+    },
+    {
+      "epoch": 1.188409090909091,
+      "grad_norm": 1.5196467810151457,
+      "learning_rate": 1.0837697778135498e-06,
+      "loss": 0.0631,
+      "step": 5229
+    },
+    {
+      "epoch": 1.1886363636363637,
+      "grad_norm": 1.977839217973498,
+      "learning_rate": 1.0837091622404183e-06,
+      "loss": 0.0635,
+      "step": 5230
+    },
+    {
+      "epoch": 1.1888636363636365,
+      "grad_norm": 1.951407866964508,
+      "learning_rate": 1.083648537313407e-06,
+      "loss": 0.0659,
+      "step": 5231
+    },
+    {
+      "epoch": 1.189090909090909,
+      "grad_norm": 1.858069863032203,
+      "learning_rate": 1.0835879030337519e-06,
+      "loss": 0.0508,
+      "step": 5232
+    },
+    {
+      "epoch": 1.189318181818182,
+      "grad_norm": 2.9249323498095743,
+      "learning_rate": 1.0835272594026892e-06,
+      "loss": 0.0397,
+      "step": 5233
+    },
+    {
+      "epoch": 1.1895454545454545,
+      "grad_norm": 1.4428089846622973,
+      "learning_rate": 1.0834666064214557e-06,
+      "loss": 0.0516,
+      "step": 5234
+    },
+    {
+      "epoch": 1.1897727272727272,
+      "grad_norm": 2.2001161907857494,
+      "learning_rate": 1.0834059440912884e-06,
+      "loss": 0.0387,
+      "step": 5235
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.124928573270036,
+      "learning_rate": 1.083345272413424e-06,
+      "loss": 0.0424,
+      "step": 5236
+    },
+    {
+      "epoch": 1.1902272727272727,
+      "grad_norm": 2.2141984874275176,
+      "learning_rate": 1.0832845913891002e-06,
+      "loss": 0.0932,
+      "step": 5237
+    },
+    {
+      "epoch": 1.1904545454545454,
+      "grad_norm": 2.022488979813883,
+      "learning_rate": 1.0832239010195538e-06,
+      "loss": 0.0513,
+      "step": 5238
+    },
+    {
+      "epoch": 1.1906818181818182,
+      "grad_norm": 0.9234929410812226,
+      "learning_rate": 1.0831632013060227e-06,
+      "loss": 0.0219,
+      "step": 5239
+    },
+    {
+      "epoch": 1.190909090909091,
+      "grad_norm": 1.9604296335833677,
+      "learning_rate": 1.0831024922497447e-06,
+      "loss": 0.0562,
+      "step": 5240
+    },
+    {
+      "epoch": 1.1911363636363637,
+      "grad_norm": 4.1921984238377705,
+      "learning_rate": 1.0830417738519575e-06,
+      "loss": 0.0982,
+      "step": 5241
+    },
+    {
+      "epoch": 1.1913636363636364,
+      "grad_norm": 2.183714236316245,
+      "learning_rate": 1.0829810461138996e-06,
+      "loss": 0.0535,
+      "step": 5242
+    },
+    {
+      "epoch": 1.1915909090909091,
+      "grad_norm": 1.5657848143629447,
+      "learning_rate": 1.082920309036809e-06,
+      "loss": 0.0343,
+      "step": 5243
+    },
+    {
+      "epoch": 1.1918181818181819,
+      "grad_norm": 1.1921638740061917,
+      "learning_rate": 1.0828595626219245e-06,
+      "loss": 0.0304,
+      "step": 5244
+    },
+    {
+      "epoch": 1.1920454545454546,
+      "grad_norm": 2.6434710285451914,
+      "learning_rate": 1.0827988068704847e-06,
+      "loss": 0.0825,
+      "step": 5245
+    },
+    {
+      "epoch": 1.1922727272727274,
+      "grad_norm": 2.254141010863687,
+      "learning_rate": 1.0827380417837287e-06,
+      "loss": 0.0642,
+      "step": 5246
+    },
+    {
+      "epoch": 1.1925,
+      "grad_norm": 2.8076082619764002,
+      "learning_rate": 1.0826772673628954e-06,
+      "loss": 0.0436,
+      "step": 5247
+    },
+    {
+      "epoch": 1.1927272727272726,
+      "grad_norm": 2.7212242077890254,
+      "learning_rate": 1.0826164836092242e-06,
+      "loss": 0.0638,
+      "step": 5248
+    },
+    {
+      "epoch": 1.1929545454545454,
+      "grad_norm": 2.1327307408811387,
+      "learning_rate": 1.0825556905239545e-06,
+      "loss": 0.0607,
+      "step": 5249
+    },
+    {
+      "epoch": 1.1931818181818181,
+      "grad_norm": 1.6787494979236237,
+      "learning_rate": 1.0824948881083259e-06,
+      "loss": 0.0849,
+      "step": 5250
+    },
+    {
+      "epoch": 1.1934090909090909,
+      "grad_norm": 1.4144099761135183,
+      "learning_rate": 1.0824340763635785e-06,
+      "loss": 0.0268,
+      "step": 5251
+    },
+    {
+      "epoch": 1.1936363636363636,
+      "grad_norm": 1.508916846664373,
+      "learning_rate": 1.082373255290952e-06,
+      "loss": 0.038,
+      "step": 5252
+    },
+    {
+      "epoch": 1.1938636363636363,
+      "grad_norm": 2.027087966460394,
+      "learning_rate": 1.0823124248916874e-06,
+      "loss": 0.1235,
+      "step": 5253
+    },
+    {
+      "epoch": 1.194090909090909,
+      "grad_norm": 1.3820423295486974,
+      "learning_rate": 1.0822515851670243e-06,
+      "loss": 0.0293,
+      "step": 5254
+    },
+    {
+      "epoch": 1.1943181818181818,
+      "grad_norm": 1.4767703794338802,
+      "learning_rate": 1.0821907361182037e-06,
+      "loss": 0.029,
+      "step": 5255
+    },
+    {
+      "epoch": 1.1945454545454546,
+      "grad_norm": 1.9731758573958356,
+      "learning_rate": 1.0821298777464665e-06,
+      "loss": 0.0212,
+      "step": 5256
+    },
+    {
+      "epoch": 1.1947727272727273,
+      "grad_norm": 1.4911126912463306,
+      "learning_rate": 1.0820690100530535e-06,
+      "loss": 0.0317,
+      "step": 5257
+    },
+    {
+      "epoch": 1.195,
+      "grad_norm": 1.0996524273119277,
+      "learning_rate": 1.082008133039206e-06,
+      "loss": 0.0178,
+      "step": 5258
+    },
+    {
+      "epoch": 1.1952272727272728,
+      "grad_norm": 1.4999889981867667,
+      "learning_rate": 1.0819472467061653e-06,
+      "loss": 0.0228,
+      "step": 5259
+    },
+    {
+      "epoch": 1.1954545454545455,
+      "grad_norm": 3.063977567320875,
+      "learning_rate": 1.081886351055173e-06,
+      "loss": 0.0601,
+      "step": 5260
+    },
+    {
+      "epoch": 1.1956818181818183,
+      "grad_norm": 1.4164705949684557,
+      "learning_rate": 1.081825446087471e-06,
+      "loss": 0.0474,
+      "step": 5261
+    },
+    {
+      "epoch": 1.1959090909090908,
+      "grad_norm": 1.6705884257510695,
+      "learning_rate": 1.0817645318043011e-06,
+      "loss": 0.0445,
+      "step": 5262
+    },
+    {
+      "epoch": 1.1961363636363636,
+      "grad_norm": 2.9841993948730043,
+      "learning_rate": 1.0817036082069057e-06,
+      "loss": 0.0754,
+      "step": 5263
+    },
+    {
+      "epoch": 1.1963636363636363,
+      "grad_norm": 0.48460408222373524,
+      "learning_rate": 1.0816426752965268e-06,
+      "loss": 0.0179,
+      "step": 5264
+    },
+    {
+      "epoch": 1.196590909090909,
+      "grad_norm": 1.5876131401326055,
+      "learning_rate": 1.0815817330744073e-06,
+      "loss": 0.0341,
+      "step": 5265
+    },
+    {
+      "epoch": 1.1968181818181818,
+      "grad_norm": 1.188492075543643,
+      "learning_rate": 1.0815207815417894e-06,
+      "loss": 0.0219,
+      "step": 5266
+    },
+    {
+      "epoch": 1.1970454545454545,
+      "grad_norm": 2.427279248611199,
+      "learning_rate": 1.0814598206999165e-06,
+      "loss": 0.0519,
+      "step": 5267
+    },
+    {
+      "epoch": 1.1972727272727273,
+      "grad_norm": 2.151839050715557,
+      "learning_rate": 1.0813988505500316e-06,
+      "loss": 0.0318,
+      "step": 5268
+    },
+    {
+      "epoch": 1.1975,
+      "grad_norm": 2.4498353181678207,
+      "learning_rate": 1.0813378710933777e-06,
+      "loss": 0.0742,
+      "step": 5269
+    },
+    {
+      "epoch": 1.1977272727272728,
+      "grad_norm": 3.213795937502913,
+      "learning_rate": 1.0812768823311987e-06,
+      "loss": 0.0416,
+      "step": 5270
+    },
+    {
+      "epoch": 1.1979545454545455,
+      "grad_norm": 3.6984565449931366,
+      "learning_rate": 1.081215884264738e-06,
+      "loss": 0.1295,
+      "step": 5271
+    },
+    {
+      "epoch": 1.1981818181818182,
+      "grad_norm": 2.4197278920363168,
+      "learning_rate": 1.0811548768952394e-06,
+      "loss": 0.0633,
+      "step": 5272
+    },
+    {
+      "epoch": 1.198409090909091,
+      "grad_norm": 1.4456653173110139,
+      "learning_rate": 1.0810938602239471e-06,
+      "loss": 0.0209,
+      "step": 5273
+    },
+    {
+      "epoch": 1.1986363636363637,
+      "grad_norm": 1.7684631526922778,
+      "learning_rate": 1.0810328342521054e-06,
+      "loss": 0.0361,
+      "step": 5274
+    },
+    {
+      "epoch": 1.1988636363636362,
+      "grad_norm": 0.5974046792563614,
+      "learning_rate": 1.0809717989809585e-06,
+      "loss": 0.0112,
+      "step": 5275
+    },
+    {
+      "epoch": 1.1990909090909092,
+      "grad_norm": 2.4638451062565285,
+      "learning_rate": 1.0809107544117511e-06,
+      "loss": 0.0867,
+      "step": 5276
+    },
+    {
+      "epoch": 1.1993181818181817,
+      "grad_norm": 1.2776879776250487,
+      "learning_rate": 1.0808497005457283e-06,
+      "loss": 0.0341,
+      "step": 5277
+    },
+    {
+      "epoch": 1.1995454545454545,
+      "grad_norm": 1.4898278964880984,
+      "learning_rate": 1.0807886373841344e-06,
+      "loss": 0.0367,
+      "step": 5278
+    },
+    {
+      "epoch": 1.1997727272727272,
+      "grad_norm": 1.330738412095984,
+      "learning_rate": 1.0807275649282153e-06,
+      "loss": 0.0218,
+      "step": 5279
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.9977029136850857,
+      "learning_rate": 1.080666483179216e-06,
+      "loss": 0.0437,
+      "step": 5280
+    },
+    {
+      "epoch": 1.2002272727272727,
+      "grad_norm": 4.330048794883879,
+      "learning_rate": 1.0806053921383823e-06,
+      "loss": 0.1129,
+      "step": 5281
+    },
+    {
+      "epoch": 1.2004545454545454,
+      "grad_norm": 2.983749885852835,
+      "learning_rate": 1.0805442918069597e-06,
+      "loss": 0.0435,
+      "step": 5282
+    },
+    {
+      "epoch": 1.2006818181818182,
+      "grad_norm": 2.3942007318276346,
+      "learning_rate": 1.0804831821861942e-06,
+      "loss": 0.1095,
+      "step": 5283
+    },
+    {
+      "epoch": 1.200909090909091,
+      "grad_norm": 2.6204221427439363,
+      "learning_rate": 1.080422063277332e-06,
+      "loss": 0.0526,
+      "step": 5284
+    },
+    {
+      "epoch": 1.2011363636363637,
+      "grad_norm": 2.876363022849003,
+      "learning_rate": 1.0803609350816196e-06,
+      "loss": 0.0642,
+      "step": 5285
+    },
+    {
+      "epoch": 1.2013636363636364,
+      "grad_norm": 1.4427434949824094,
+      "learning_rate": 1.0802997976003031e-06,
+      "loss": 0.0237,
+      "step": 5286
+    },
+    {
+      "epoch": 1.2015909090909092,
+      "grad_norm": 2.3084960014372076,
+      "learning_rate": 1.0802386508346293e-06,
+      "loss": 0.0619,
+      "step": 5287
+    },
+    {
+      "epoch": 1.201818181818182,
+      "grad_norm": 2.5372855889966046,
+      "learning_rate": 1.0801774947858453e-06,
+      "loss": 0.0732,
+      "step": 5288
+    },
+    {
+      "epoch": 1.2020454545454546,
+      "grad_norm": 1.4432804603850256,
+      "learning_rate": 1.0801163294551983e-06,
+      "loss": 0.029,
+      "step": 5289
+    },
+    {
+      "epoch": 1.2022727272727272,
+      "grad_norm": 2.137495312554455,
+      "learning_rate": 1.0800551548439353e-06,
+      "loss": 0.0459,
+      "step": 5290
+    },
+    {
+      "epoch": 1.2025000000000001,
+      "grad_norm": 1.2666315629366403,
+      "learning_rate": 1.0799939709533036e-06,
+      "loss": 0.0391,
+      "step": 5291
+    },
+    {
+      "epoch": 1.2027272727272726,
+      "grad_norm": 2.060078695511436,
+      "learning_rate": 1.0799327777845512e-06,
+      "loss": 0.0461,
+      "step": 5292
+    },
+    {
+      "epoch": 1.2029545454545454,
+      "grad_norm": 2.410135873977101,
+      "learning_rate": 1.0798715753389258e-06,
+      "loss": 0.1022,
+      "step": 5293
+    },
+    {
+      "epoch": 1.2031818181818181,
+      "grad_norm": 0.665386416942162,
+      "learning_rate": 1.079810363617675e-06,
+      "loss": 0.0116,
+      "step": 5294
+    },
+    {
+      "epoch": 1.2034090909090909,
+      "grad_norm": 1.7061198582800272,
+      "learning_rate": 1.079749142622048e-06,
+      "loss": 0.0297,
+      "step": 5295
+    },
+    {
+      "epoch": 1.2036363636363636,
+      "grad_norm": 1.7393620213081726,
+      "learning_rate": 1.0796879123532924e-06,
+      "loss": 0.0833,
+      "step": 5296
+    },
+    {
+      "epoch": 1.2038636363636364,
+      "grad_norm": 1.4321692667210129,
+      "learning_rate": 1.079626672812657e-06,
+      "loss": 0.0511,
+      "step": 5297
+    },
+    {
+      "epoch": 1.204090909090909,
+      "grad_norm": 1.3289771215893684,
+      "learning_rate": 1.0795654240013906e-06,
+      "loss": 0.0391,
+      "step": 5298
+    },
+    {
+      "epoch": 1.2043181818181818,
+      "grad_norm": 1.7918517538169159,
+      "learning_rate": 1.0795041659207424e-06,
+      "loss": 0.0732,
+      "step": 5299
+    },
+    {
+      "epoch": 1.2045454545454546,
+      "grad_norm": 1.9492856824686968,
+      "learning_rate": 1.079442898571961e-06,
+      "loss": 0.042,
+      "step": 5300
+    },
+    {
+      "epoch": 1.2047727272727273,
+      "grad_norm": 2.2047010477617763,
+      "learning_rate": 1.0793816219562963e-06,
+      "loss": 0.0641,
+      "step": 5301
+    },
+    {
+      "epoch": 1.205,
+      "grad_norm": 2.0599500592527824,
+      "learning_rate": 1.0793203360749974e-06,
+      "loss": 0.0448,
+      "step": 5302
+    },
+    {
+      "epoch": 1.2052272727272728,
+      "grad_norm": 2.222983229085742,
+      "learning_rate": 1.0792590409293144e-06,
+      "loss": 0.0553,
+      "step": 5303
+    },
+    {
+      "epoch": 1.2054545454545456,
+      "grad_norm": 2.032389365138482,
+      "learning_rate": 1.079197736520497e-06,
+      "loss": 0.0399,
+      "step": 5304
+    },
+    {
+      "epoch": 1.205681818181818,
+      "grad_norm": 1.1304761687730498,
+      "learning_rate": 1.0791364228497953e-06,
+      "loss": 0.0228,
+      "step": 5305
+    },
+    {
+      "epoch": 1.2059090909090908,
+      "grad_norm": 1.5270744047773357,
+      "learning_rate": 1.0790750999184598e-06,
+      "loss": 0.0416,
+      "step": 5306
+    },
+    {
+      "epoch": 1.2061363636363636,
+      "grad_norm": 1.2473200090962286,
+      "learning_rate": 1.0790137677277407e-06,
+      "loss": 0.0444,
+      "step": 5307
+    },
+    {
+      "epoch": 1.2063636363636363,
+      "grad_norm": 1.9462924584388555,
+      "learning_rate": 1.078952426278889e-06,
+      "loss": 0.0372,
+      "step": 5308
+    },
+    {
+      "epoch": 1.206590909090909,
+      "grad_norm": 2.928442939752278,
+      "learning_rate": 1.0788910755731553e-06,
+      "loss": 0.0956,
+      "step": 5309
+    },
+    {
+      "epoch": 1.2068181818181818,
+      "grad_norm": 2.4124485439728263,
+      "learning_rate": 1.0788297156117906e-06,
+      "loss": 0.061,
+      "step": 5310
+    },
+    {
+      "epoch": 1.2070454545454545,
+      "grad_norm": 2.3222764378328407,
+      "learning_rate": 1.0787683463960462e-06,
+      "loss": 0.0632,
+      "step": 5311
+    },
+    {
+      "epoch": 1.2072727272727273,
+      "grad_norm": 7.235900020306009,
+      "learning_rate": 1.0787069679271738e-06,
+      "loss": 0.0469,
+      "step": 5312
+    },
+    {
+      "epoch": 1.2075,
+      "grad_norm": 1.3990534830894166,
+      "learning_rate": 1.0786455802064246e-06,
+      "loss": 0.0532,
+      "step": 5313
+    },
+    {
+      "epoch": 1.2077272727272728,
+      "grad_norm": 1.1507618423635084,
+      "learning_rate": 1.0785841832350506e-06,
+      "loss": 0.0209,
+      "step": 5314
+    },
+    {
+      "epoch": 1.2079545454545455,
+      "grad_norm": 1.6383932837303263,
+      "learning_rate": 1.078522777014304e-06,
+      "loss": 0.0724,
+      "step": 5315
+    },
+    {
+      "epoch": 1.2081818181818182,
+      "grad_norm": 2.9557558204131222,
+      "learning_rate": 1.0784613615454365e-06,
+      "loss": 0.0674,
+      "step": 5316
+    },
+    {
+      "epoch": 1.208409090909091,
+      "grad_norm": 2.7304364503171676,
+      "learning_rate": 1.0783999368297006e-06,
+      "loss": 0.0738,
+      "step": 5317
+    },
+    {
+      "epoch": 1.2086363636363637,
+      "grad_norm": 3.1225083320804234,
+      "learning_rate": 1.0783385028683494e-06,
+      "loss": 0.0801,
+      "step": 5318
+    },
+    {
+      "epoch": 1.2088636363636365,
+      "grad_norm": 2.041235320399034,
+      "learning_rate": 1.0782770596626352e-06,
+      "loss": 0.0269,
+      "step": 5319
+    },
+    {
+      "epoch": 1.209090909090909,
+      "grad_norm": 2.3210026784820683,
+      "learning_rate": 1.0782156072138107e-06,
+      "loss": 0.0482,
+      "step": 5320
+    },
+    {
+      "epoch": 1.2093181818181817,
+      "grad_norm": 1.8092726941859452,
+      "learning_rate": 1.0781541455231294e-06,
+      "loss": 0.0376,
+      "step": 5321
+    },
+    {
+      "epoch": 1.2095454545454545,
+      "grad_norm": 2.4976784408528228,
+      "learning_rate": 1.0780926745918446e-06,
+      "loss": 0.0501,
+      "step": 5322
+    },
+    {
+      "epoch": 1.2097727272727272,
+      "grad_norm": 1.8970582591979632,
+      "learning_rate": 1.0780311944212095e-06,
+      "loss": 0.0416,
+      "step": 5323
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.087554302725568,
+      "learning_rate": 1.0779697050124783e-06,
+      "loss": 0.0891,
+      "step": 5324
+    },
+    {
+      "epoch": 1.2102272727272727,
+      "grad_norm": 1.2437652583465568,
+      "learning_rate": 1.0779082063669042e-06,
+      "loss": 0.042,
+      "step": 5325
+    },
+    {
+      "epoch": 1.2104545454545454,
+      "grad_norm": 3.0286591621966323,
+      "learning_rate": 1.077846698485742e-06,
+      "loss": 0.0876,
+      "step": 5326
+    },
+    {
+      "epoch": 1.2106818181818182,
+      "grad_norm": 1.8403369434118733,
+      "learning_rate": 1.0777851813702453e-06,
+      "loss": 0.0406,
+      "step": 5327
+    },
+    {
+      "epoch": 1.210909090909091,
+      "grad_norm": 2.130219321092055,
+      "learning_rate": 1.0777236550216688e-06,
+      "loss": 0.0632,
+      "step": 5328
+    },
+    {
+      "epoch": 1.2111363636363637,
+      "grad_norm": 3.260945386247149,
+      "learning_rate": 1.0776621194412673e-06,
+      "loss": 0.0927,
+      "step": 5329
+    },
+    {
+      "epoch": 1.2113636363636364,
+      "grad_norm": 1.3932839324115354,
+      "learning_rate": 1.0776005746302955e-06,
+      "loss": 0.0261,
+      "step": 5330
+    },
+    {
+      "epoch": 1.2115909090909092,
+      "grad_norm": 1.7876730389674256,
+      "learning_rate": 1.0775390205900084e-06,
+      "loss": 0.0232,
+      "step": 5331
+    },
+    {
+      "epoch": 1.211818181818182,
+      "grad_norm": 1.9791987962199296,
+      "learning_rate": 1.0774774573216609e-06,
+      "loss": 0.0411,
+      "step": 5332
+    },
+    {
+      "epoch": 1.2120454545454544,
+      "grad_norm": 2.172522973267638,
+      "learning_rate": 1.077415884826509e-06,
+      "loss": 0.043,
+      "step": 5333
+    },
+    {
+      "epoch": 1.2122727272727274,
+      "grad_norm": 2.692214327930651,
+      "learning_rate": 1.0773543031058077e-06,
+      "loss": 0.0479,
+      "step": 5334
+    },
+    {
+      "epoch": 1.2125,
+      "grad_norm": 1.5151516362235038,
+      "learning_rate": 1.077292712160813e-06,
+      "loss": 0.0472,
+      "step": 5335
+    },
+    {
+      "epoch": 1.2127272727272727,
+      "grad_norm": 2.6719287589213216,
+      "learning_rate": 1.0772311119927808e-06,
+      "loss": 0.0807,
+      "step": 5336
+    },
+    {
+      "epoch": 1.2129545454545454,
+      "grad_norm": 1.3370030347514332,
+      "learning_rate": 1.0771695026029673e-06,
+      "loss": 0.0322,
+      "step": 5337
+    },
+    {
+      "epoch": 1.2131818181818181,
+      "grad_norm": 1.794354286853342,
+      "learning_rate": 1.0771078839926292e-06,
+      "loss": 0.0316,
+      "step": 5338
+    },
+    {
+      "epoch": 1.2134090909090909,
+      "grad_norm": 2.049749695699831,
+      "learning_rate": 1.0770462561630222e-06,
+      "loss": 0.0335,
+      "step": 5339
+    },
+    {
+      "epoch": 1.2136363636363636,
+      "grad_norm": 3.707661558645731,
+      "learning_rate": 1.0769846191154032e-06,
+      "loss": 0.0633,
+      "step": 5340
+    },
+    {
+      "epoch": 1.2138636363636364,
+      "grad_norm": 3.8363049482229594,
+      "learning_rate": 1.0769229728510298e-06,
+      "loss": 0.0915,
+      "step": 5341
+    },
+    {
+      "epoch": 1.214090909090909,
+      "grad_norm": 2.02573094187132,
+      "learning_rate": 1.0768613173711582e-06,
+      "loss": 0.0271,
+      "step": 5342
+    },
+    {
+      "epoch": 1.2143181818181819,
+      "grad_norm": 1.5998203055994782,
+      "learning_rate": 1.0767996526770462e-06,
+      "loss": 0.0448,
+      "step": 5343
+    },
+    {
+      "epoch": 1.2145454545454546,
+      "grad_norm": 1.4164015279223179,
+      "learning_rate": 1.0767379787699507e-06,
+      "loss": 0.0306,
+      "step": 5344
+    },
+    {
+      "epoch": 1.2147727272727273,
+      "grad_norm": 2.70822760805152,
+      "learning_rate": 1.07667629565113e-06,
+      "loss": 0.0501,
+      "step": 5345
+    },
+    {
+      "epoch": 1.215,
+      "grad_norm": 2.2822865405959853,
+      "learning_rate": 1.0766146033218417e-06,
+      "loss": 0.0408,
+      "step": 5346
+    },
+    {
+      "epoch": 1.2152272727272728,
+      "grad_norm": 3.0513243014183744,
+      "learning_rate": 1.0765529017833436e-06,
+      "loss": 0.0522,
+      "step": 5347
+    },
+    {
+      "epoch": 1.2154545454545453,
+      "grad_norm": 1.3683024576722072,
+      "learning_rate": 1.0764911910368943e-06,
+      "loss": 0.0275,
+      "step": 5348
+    },
+    {
+      "epoch": 1.215681818181818,
+      "grad_norm": 1.5044299224995652,
+      "learning_rate": 1.076429471083752e-06,
+      "loss": 0.0388,
+      "step": 5349
+    },
+    {
+      "epoch": 1.2159090909090908,
+      "grad_norm": 2.428352267078033,
+      "learning_rate": 1.0763677419251747e-06,
+      "loss": 0.0727,
+      "step": 5350
+    },
+    {
+      "epoch": 1.2161363636363636,
+      "grad_norm": 1.7260882439640153,
+      "learning_rate": 1.076306003562422e-06,
+      "loss": 0.0289,
+      "step": 5351
+    },
+    {
+      "epoch": 1.2163636363636363,
+      "grad_norm": 1.796502033054265,
+      "learning_rate": 1.0762442559967527e-06,
+      "loss": 0.0299,
+      "step": 5352
+    },
+    {
+      "epoch": 1.216590909090909,
+      "grad_norm": 2.3079696732009953,
+      "learning_rate": 1.0761824992294255e-06,
+      "loss": 0.0618,
+      "step": 5353
+    },
+    {
+      "epoch": 1.2168181818181818,
+      "grad_norm": 1.570138900278559,
+      "learning_rate": 1.0761207332617002e-06,
+      "loss": 0.0259,
+      "step": 5354
+    },
+    {
+      "epoch": 1.2170454545454545,
+      "grad_norm": 1.8374761623831064,
+      "learning_rate": 1.076058958094836e-06,
+      "loss": 0.0206,
+      "step": 5355
+    },
+    {
+      "epoch": 1.2172727272727273,
+      "grad_norm": 1.3804360839846175,
+      "learning_rate": 1.0759971737300928e-06,
+      "loss": 0.0197,
+      "step": 5356
+    },
+    {
+      "epoch": 1.2175,
+      "grad_norm": 1.7778326829204434,
+      "learning_rate": 1.0759353801687304e-06,
+      "loss": 0.0437,
+      "step": 5357
+    },
+    {
+      "epoch": 1.2177272727272728,
+      "grad_norm": 1.3926597881922738,
+      "learning_rate": 1.075873577412009e-06,
+      "loss": 0.038,
+      "step": 5358
+    },
+    {
+      "epoch": 1.2179545454545455,
+      "grad_norm": 1.6838162021105392,
+      "learning_rate": 1.0758117654611886e-06,
+      "loss": 0.0466,
+      "step": 5359
+    },
+    {
+      "epoch": 1.2181818181818183,
+      "grad_norm": 2.119455844398616,
+      "learning_rate": 1.0757499443175298e-06,
+      "loss": 0.0354,
+      "step": 5360
+    },
+    {
+      "epoch": 1.218409090909091,
+      "grad_norm": 1.121271400928087,
+      "learning_rate": 1.0756881139822934e-06,
+      "loss": 0.0277,
+      "step": 5361
+    },
+    {
+      "epoch": 1.2186363636363637,
+      "grad_norm": 2.266068562740744,
+      "learning_rate": 1.07562627445674e-06,
+      "loss": 0.0468,
+      "step": 5362
+    },
+    {
+      "epoch": 1.2188636363636363,
+      "grad_norm": 2.8907368605381274,
+      "learning_rate": 1.0755644257421308e-06,
+      "loss": 0.0639,
+      "step": 5363
+    },
+    {
+      "epoch": 1.219090909090909,
+      "grad_norm": 1.5229253435836179,
+      "learning_rate": 1.0755025678397269e-06,
+      "loss": 0.0641,
+      "step": 5364
+    },
+    {
+      "epoch": 1.2193181818181817,
+      "grad_norm": 1.4875712053726673,
+      "learning_rate": 1.0754407007507896e-06,
+      "loss": 0.0269,
+      "step": 5365
+    },
+    {
+      "epoch": 1.2195454545454545,
+      "grad_norm": 2.1794350472306183,
+      "learning_rate": 1.0753788244765805e-06,
+      "loss": 0.0336,
+      "step": 5366
+    },
+    {
+      "epoch": 1.2197727272727272,
+      "grad_norm": 1.9496680848964596,
+      "learning_rate": 1.0753169390183616e-06,
+      "loss": 0.072,
+      "step": 5367
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.9200551074592207,
+      "learning_rate": 1.0752550443773948e-06,
+      "loss": 0.056,
+      "step": 5368
+    },
+    {
+      "epoch": 1.2202272727272727,
+      "grad_norm": 2.4416343264745763,
+      "learning_rate": 1.0751931405549419e-06,
+      "loss": 0.0642,
+      "step": 5369
+    },
+    {
+      "epoch": 1.2204545454545455,
+      "grad_norm": 2.821351561588249,
+      "learning_rate": 1.0751312275522655e-06,
+      "loss": 0.0516,
+      "step": 5370
+    },
+    {
+      "epoch": 1.2206818181818182,
+      "grad_norm": 1.9097597706139426,
+      "learning_rate": 1.0750693053706282e-06,
+      "loss": 0.0426,
+      "step": 5371
+    },
+    {
+      "epoch": 1.220909090909091,
+      "grad_norm": 1.9594436938959916,
+      "learning_rate": 1.0750073740112926e-06,
+      "loss": 0.0555,
+      "step": 5372
+    },
+    {
+      "epoch": 1.2211363636363637,
+      "grad_norm": 2.081513325662645,
+      "learning_rate": 1.0749454334755214e-06,
+      "loss": 0.0519,
+      "step": 5373
+    },
+    {
+      "epoch": 1.2213636363636364,
+      "grad_norm": 1.6153479429957252,
+      "learning_rate": 1.0748834837645777e-06,
+      "loss": 0.0476,
+      "step": 5374
+    },
+    {
+      "epoch": 1.2215909090909092,
+      "grad_norm": 1.3571254888474302,
+      "learning_rate": 1.0748215248797253e-06,
+      "loss": 0.0401,
+      "step": 5375
+    },
+    {
+      "epoch": 1.221818181818182,
+      "grad_norm": 2.8976204626771316,
+      "learning_rate": 1.0747595568222268e-06,
+      "loss": 0.0792,
+      "step": 5376
+    },
+    {
+      "epoch": 1.2220454545454547,
+      "grad_norm": 2.237567522930413,
+      "learning_rate": 1.0746975795933466e-06,
+      "loss": 0.0515,
+      "step": 5377
+    },
+    {
+      "epoch": 1.2222727272727272,
+      "grad_norm": 2.0334664243831497,
+      "learning_rate": 1.074635593194348e-06,
+      "loss": 0.0485,
+      "step": 5378
+    },
+    {
+      "epoch": 1.2225,
+      "grad_norm": 1.5043473978819477,
+      "learning_rate": 1.0745735976264955e-06,
+      "loss": 0.0436,
+      "step": 5379
+    },
+    {
+      "epoch": 1.2227272727272727,
+      "grad_norm": 3.16515296484789,
+      "learning_rate": 1.074511592891053e-06,
+      "loss": 0.027,
+      "step": 5380
+    },
+    {
+      "epoch": 1.2229545454545454,
+      "grad_norm": 1.717881977330667,
+      "learning_rate": 1.0744495789892848e-06,
+      "loss": 0.0456,
+      "step": 5381
+    },
+    {
+      "epoch": 1.2231818181818181,
+      "grad_norm": 2.7938899600500555,
+      "learning_rate": 1.0743875559224554e-06,
+      "loss": 0.0782,
+      "step": 5382
+    },
+    {
+      "epoch": 1.223409090909091,
+      "grad_norm": 1.9019498759634357,
+      "learning_rate": 1.07432552369183e-06,
+      "loss": 0.084,
+      "step": 5383
+    },
+    {
+      "epoch": 1.2236363636363636,
+      "grad_norm": 1.7395917391554507,
+      "learning_rate": 1.074263482298673e-06,
+      "loss": 0.0272,
+      "step": 5384
+    },
+    {
+      "epoch": 1.2238636363636364,
+      "grad_norm": 1.2316315631060288,
+      "learning_rate": 1.07420143174425e-06,
+      "loss": 0.0228,
+      "step": 5385
+    },
+    {
+      "epoch": 1.2240909090909091,
+      "grad_norm": 1.763003889744533,
+      "learning_rate": 1.0741393720298263e-06,
+      "loss": 0.0425,
+      "step": 5386
+    },
+    {
+      "epoch": 1.2243181818181819,
+      "grad_norm": 1.7896972082522633,
+      "learning_rate": 1.074077303156667e-06,
+      "loss": 0.0287,
+      "step": 5387
+    },
+    {
+      "epoch": 1.2245454545454546,
+      "grad_norm": 1.5376383728942098,
+      "learning_rate": 1.0740152251260382e-06,
+      "loss": 0.0384,
+      "step": 5388
+    },
+    {
+      "epoch": 1.2247727272727273,
+      "grad_norm": 2.1122891307608906,
+      "learning_rate": 1.0739531379392053e-06,
+      "loss": 0.1096,
+      "step": 5389
+    },
+    {
+      "epoch": 1.225,
+      "grad_norm": 1.1585645038431518,
+      "learning_rate": 1.073891041597435e-06,
+      "loss": 0.0306,
+      "step": 5390
+    },
+    {
+      "epoch": 1.2252272727272726,
+      "grad_norm": 2.1410743728633723,
+      "learning_rate": 1.073828936101993e-06,
+      "loss": 0.056,
+      "step": 5391
+    },
+    {
+      "epoch": 1.2254545454545456,
+      "grad_norm": 2.010424399331064,
+      "learning_rate": 1.0737668214541463e-06,
+      "loss": 0.0449,
+      "step": 5392
+    },
+    {
+      "epoch": 1.225681818181818,
+      "grad_norm": 2.50971635505183,
+      "learning_rate": 1.073704697655161e-06,
+      "loss": 0.047,
+      "step": 5393
+    },
+    {
+      "epoch": 1.2259090909090908,
+      "grad_norm": 1.2599759355849673,
+      "learning_rate": 1.073642564706304e-06,
+      "loss": 0.0383,
+      "step": 5394
+    },
+    {
+      "epoch": 1.2261363636363636,
+      "grad_norm": 2.1292531880189123,
+      "learning_rate": 1.0735804226088423e-06,
+      "loss": 0.0532,
+      "step": 5395
+    },
+    {
+      "epoch": 1.2263636363636363,
+      "grad_norm": 0.8733451582737255,
+      "learning_rate": 1.0735182713640436e-06,
+      "loss": 0.0174,
+      "step": 5396
+    },
+    {
+      "epoch": 1.226590909090909,
+      "grad_norm": 1.004368066426233,
+      "learning_rate": 1.0734561109731746e-06,
+      "loss": 0.0153,
+      "step": 5397
+    },
+    {
+      "epoch": 1.2268181818181818,
+      "grad_norm": 2.345401122866967,
+      "learning_rate": 1.073393941437503e-06,
+      "loss": 0.056,
+      "step": 5398
+    },
+    {
+      "epoch": 1.2270454545454546,
+      "grad_norm": 1.4714318415755372,
+      "learning_rate": 1.073331762758297e-06,
+      "loss": 0.0462,
+      "step": 5399
+    },
+    {
+      "epoch": 1.2272727272727273,
+      "grad_norm": 1.4238300804268775,
+      "learning_rate": 1.0732695749368239e-06,
+      "loss": 0.0336,
+      "step": 5400
+    },
+    {
+      "epoch": 1.2275,
+      "grad_norm": 2.591248695742394,
+      "learning_rate": 1.0732073779743523e-06,
+      "loss": 0.1014,
+      "step": 5401
+    },
+    {
+      "epoch": 1.2277272727272728,
+      "grad_norm": 1.9088120203056949,
+      "learning_rate": 1.0731451718721502e-06,
+      "loss": 0.0272,
+      "step": 5402
+    },
+    {
+      "epoch": 1.2279545454545455,
+      "grad_norm": 1.502648280303289,
+      "learning_rate": 1.0730829566314863e-06,
+      "loss": 0.0212,
+      "step": 5403
+    },
+    {
+      "epoch": 1.2281818181818183,
+      "grad_norm": 1.5024699253265572,
+      "learning_rate": 1.0730207322536292e-06,
+      "loss": 0.035,
+      "step": 5404
+    },
+    {
+      "epoch": 1.228409090909091,
+      "grad_norm": 2.3686440925274144,
+      "learning_rate": 1.0729584987398479e-06,
+      "loss": 0.0401,
+      "step": 5405
+    },
+    {
+      "epoch": 1.2286363636363635,
+      "grad_norm": 1.9610175612455618,
+      "learning_rate": 1.0728962560914108e-06,
+      "loss": 0.0368,
+      "step": 5406
+    },
+    {
+      "epoch": 1.2288636363636363,
+      "grad_norm": 2.102584960984363,
+      "learning_rate": 1.072834004309588e-06,
+      "loss": 0.0302,
+      "step": 5407
+    },
+    {
+      "epoch": 1.229090909090909,
+      "grad_norm": 1.6183458889518192,
+      "learning_rate": 1.0727717433956487e-06,
+      "loss": 0.0351,
+      "step": 5408
+    },
+    {
+      "epoch": 1.2293181818181818,
+      "grad_norm": 1.3784511884568462,
+      "learning_rate": 1.0727094733508623e-06,
+      "loss": 0.0375,
+      "step": 5409
+    },
+    {
+      "epoch": 1.2295454545454545,
+      "grad_norm": 2.092147076918062,
+      "learning_rate": 1.0726471941764986e-06,
+      "loss": 0.0544,
+      "step": 5410
+    },
+    {
+      "epoch": 1.2297727272727272,
+      "grad_norm": 0.8788325965322696,
+      "learning_rate": 1.0725849058738274e-06,
+      "loss": 0.0221,
+      "step": 5411
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.4985547211266255,
+      "learning_rate": 1.0725226084441192e-06,
+      "loss": 0.0557,
+      "step": 5412
+    },
+    {
+      "epoch": 1.2302272727272727,
+      "grad_norm": 2.3937617729451546,
+      "learning_rate": 1.0724603018886445e-06,
+      "loss": 0.0685,
+      "step": 5413
+    },
+    {
+      "epoch": 1.2304545454545455,
+      "grad_norm": 1.7652659216760813,
+      "learning_rate": 1.0723979862086735e-06,
+      "loss": 0.0781,
+      "step": 5414
+    },
+    {
+      "epoch": 1.2306818181818182,
+      "grad_norm": 2.796197278806462,
+      "learning_rate": 1.0723356614054772e-06,
+      "loss": 0.0759,
+      "step": 5415
+    },
+    {
+      "epoch": 1.230909090909091,
+      "grad_norm": 2.0054225063032787,
+      "learning_rate": 1.0722733274803261e-06,
+      "loss": 0.0604,
+      "step": 5416
+    },
+    {
+      "epoch": 1.2311363636363637,
+      "grad_norm": 1.4856035051354308,
+      "learning_rate": 1.0722109844344916e-06,
+      "loss": 0.0584,
+      "step": 5417
+    },
+    {
+      "epoch": 1.2313636363636364,
+      "grad_norm": 1.5638623513018448,
+      "learning_rate": 1.072148632269245e-06,
+      "loss": 0.0453,
+      "step": 5418
+    },
+    {
+      "epoch": 1.2315909090909092,
+      "grad_norm": 0.6505516062198353,
+      "learning_rate": 1.0720862709858577e-06,
+      "loss": 0.0183,
+      "step": 5419
+    },
+    {
+      "epoch": 1.231818181818182,
+      "grad_norm": 1.8632305116238028,
+      "learning_rate": 1.0720239005856013e-06,
+      "loss": 0.0296,
+      "step": 5420
+    },
+    {
+      "epoch": 1.2320454545454544,
+      "grad_norm": 1.6920730146989291,
+      "learning_rate": 1.0719615210697476e-06,
+      "loss": 0.0293,
+      "step": 5421
+    },
+    {
+      "epoch": 1.2322727272727272,
+      "grad_norm": 1.5876330168879198,
+      "learning_rate": 1.071899132439569e-06,
+      "loss": 0.0619,
+      "step": 5422
+    },
+    {
+      "epoch": 1.2325,
+      "grad_norm": 2.658982781948774,
+      "learning_rate": 1.071836734696337e-06,
+      "loss": 0.0698,
+      "step": 5423
+    },
+    {
+      "epoch": 1.2327272727272727,
+      "grad_norm": 2.5505810675963323,
+      "learning_rate": 1.0717743278413248e-06,
+      "loss": 0.078,
+      "step": 5424
+    },
+    {
+      "epoch": 1.2329545454545454,
+      "grad_norm": 2.0095689745797247,
+      "learning_rate": 1.0717119118758043e-06,
+      "loss": 0.0704,
+      "step": 5425
+    },
+    {
+      "epoch": 1.2331818181818182,
+      "grad_norm": 1.6962684355116315,
+      "learning_rate": 1.0716494868010488e-06,
+      "loss": 0.037,
+      "step": 5426
+    },
+    {
+      "epoch": 1.233409090909091,
+      "grad_norm": 3.1640638345202388,
+      "learning_rate": 1.0715870526183312e-06,
+      "loss": 0.1077,
+      "step": 5427
+    },
+    {
+      "epoch": 1.2336363636363636,
+      "grad_norm": 1.7485038596984206,
+      "learning_rate": 1.0715246093289241e-06,
+      "loss": 0.0538,
+      "step": 5428
+    },
+    {
+      "epoch": 1.2338636363636364,
+      "grad_norm": 1.2640986497086053,
+      "learning_rate": 1.0714621569341016e-06,
+      "loss": 0.0362,
+      "step": 5429
+    },
+    {
+      "epoch": 1.2340909090909091,
+      "grad_norm": 1.5256620550087732,
+      "learning_rate": 1.0713996954351364e-06,
+      "loss": 0.0304,
+      "step": 5430
+    },
+    {
+      "epoch": 1.2343181818181819,
+      "grad_norm": 2.9420490272027275,
+      "learning_rate": 1.071337224833303e-06,
+      "loss": 0.0709,
+      "step": 5431
+    },
+    {
+      "epoch": 1.2345454545454546,
+      "grad_norm": 1.8154271365347503,
+      "learning_rate": 1.0712747451298747e-06,
+      "loss": 0.0401,
+      "step": 5432
+    },
+    {
+      "epoch": 1.2347727272727274,
+      "grad_norm": 2.156331657025114,
+      "learning_rate": 1.071212256326126e-06,
+      "loss": 0.0394,
+      "step": 5433
+    },
+    {
+      "epoch": 1.2349999999999999,
+      "grad_norm": 2.2940456706551773,
+      "learning_rate": 1.0711497584233306e-06,
+      "loss": 0.0424,
+      "step": 5434
+    },
+    {
+      "epoch": 1.2352272727272728,
+      "grad_norm": 2.5683868537093635,
+      "learning_rate": 1.0710872514227635e-06,
+      "loss": 0.0514,
+      "step": 5435
+    },
+    {
+      "epoch": 1.2354545454545454,
+      "grad_norm": 1.79928223898622,
+      "learning_rate": 1.0710247353256988e-06,
+      "loss": 0.0317,
+      "step": 5436
+    },
+    {
+      "epoch": 1.235681818181818,
+      "grad_norm": 1.4663678517579,
+      "learning_rate": 1.070962210133412e-06,
+      "loss": 0.0532,
+      "step": 5437
+    },
+    {
+      "epoch": 1.2359090909090908,
+      "grad_norm": 1.941541924072884,
+      "learning_rate": 1.0708996758471776e-06,
+      "loss": 0.0358,
+      "step": 5438
+    },
+    {
+      "epoch": 1.2361363636363636,
+      "grad_norm": 1.3936187087957523,
+      "learning_rate": 1.070837132468271e-06,
+      "loss": 0.0518,
+      "step": 5439
+    },
+    {
+      "epoch": 1.2363636363636363,
+      "grad_norm": 1.5734088339955632,
+      "learning_rate": 1.0707745799979672e-06,
+      "loss": 0.0227,
+      "step": 5440
+    },
+    {
+      "epoch": 1.236590909090909,
+      "grad_norm": 2.8343768188498575,
+      "learning_rate": 1.0707120184375422e-06,
+      "loss": 0.0705,
+      "step": 5441
+    },
+    {
+      "epoch": 1.2368181818181818,
+      "grad_norm": 1.2094812786849611,
+      "learning_rate": 1.0706494477882714e-06,
+      "loss": 0.02,
+      "step": 5442
+    },
+    {
+      "epoch": 1.2370454545454546,
+      "grad_norm": 1.0434242637372761,
+      "learning_rate": 1.070586868051431e-06,
+      "loss": 0.0213,
+      "step": 5443
+    },
+    {
+      "epoch": 1.2372727272727273,
+      "grad_norm": 3.4012564485859706,
+      "learning_rate": 1.0705242792282968e-06,
+      "loss": 0.0833,
+      "step": 5444
+    },
+    {
+      "epoch": 1.2375,
+      "grad_norm": 1.4415592186969706,
+      "learning_rate": 1.0704616813201458e-06,
+      "loss": 0.0467,
+      "step": 5445
+    },
+    {
+      "epoch": 1.2377272727272728,
+      "grad_norm": 1.5591936822944976,
+      "learning_rate": 1.0703990743282534e-06,
+      "loss": 0.0593,
+      "step": 5446
+    },
+    {
+      "epoch": 1.2379545454545455,
+      "grad_norm": 1.3876412425785902,
+      "learning_rate": 1.0703364582538971e-06,
+      "loss": 0.0361,
+      "step": 5447
+    },
+    {
+      "epoch": 1.2381818181818183,
+      "grad_norm": 1.5656607742685487,
+      "learning_rate": 1.0702738330983534e-06,
+      "loss": 0.0388,
+      "step": 5448
+    },
+    {
+      "epoch": 1.2384090909090908,
+      "grad_norm": 1.8766150867365687,
+      "learning_rate": 1.0702111988628997e-06,
+      "loss": 0.0596,
+      "step": 5449
+    },
+    {
+      "epoch": 1.2386363636363638,
+      "grad_norm": 2.27076366608418,
+      "learning_rate": 1.0701485555488125e-06,
+      "loss": 0.0528,
+      "step": 5450
+    },
+    {
+      "epoch": 1.2388636363636363,
+      "grad_norm": 2.3937812857707375,
+      "learning_rate": 1.07008590315737e-06,
+      "loss": 0.049,
+      "step": 5451
+    },
+    {
+      "epoch": 1.239090909090909,
+      "grad_norm": 1.949825335663674,
+      "learning_rate": 1.0700232416898493e-06,
+      "loss": 0.0759,
+      "step": 5452
+    },
+    {
+      "epoch": 1.2393181818181818,
+      "grad_norm": 1.514932207099186,
+      "learning_rate": 1.0699605711475283e-06,
+      "loss": 0.0473,
+      "step": 5453
+    },
+    {
+      "epoch": 1.2395454545454545,
+      "grad_norm": 1.0961170500804767,
+      "learning_rate": 1.069897891531685e-06,
+      "loss": 0.0241,
+      "step": 5454
+    },
+    {
+      "epoch": 1.2397727272727272,
+      "grad_norm": 1.9687040063101748,
+      "learning_rate": 1.0698352028435973e-06,
+      "loss": 0.0609,
+      "step": 5455
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.578615102097434,
+      "learning_rate": 1.069772505084544e-06,
+      "loss": 0.0454,
+      "step": 5456
+    },
+    {
+      "epoch": 1.2402272727272727,
+      "grad_norm": 2.9493846526415606,
+      "learning_rate": 1.0697097982558036e-06,
+      "loss": 0.0328,
+      "step": 5457
+    },
+    {
+      "epoch": 1.2404545454545455,
+      "grad_norm": 1.7916871178680576,
+      "learning_rate": 1.069647082358654e-06,
+      "loss": 0.0523,
+      "step": 5458
+    },
+    {
+      "epoch": 1.2406818181818182,
+      "grad_norm": 2.191976417978301,
+      "learning_rate": 1.0695843573943752e-06,
+      "loss": 0.0594,
+      "step": 5459
+    },
+    {
+      "epoch": 1.240909090909091,
+      "grad_norm": 2.1415536834446893,
+      "learning_rate": 1.0695216233642456e-06,
+      "loss": 0.0516,
+      "step": 5460
+    },
+    {
+      "epoch": 1.2411363636363637,
+      "grad_norm": 1.4405913247086002,
+      "learning_rate": 1.0694588802695443e-06,
+      "loss": 0.0242,
+      "step": 5461
+    },
+    {
+      "epoch": 1.2413636363636364,
+      "grad_norm": 1.9306256251355607,
+      "learning_rate": 1.0693961281115513e-06,
+      "loss": 0.0255,
+      "step": 5462
+    },
+    {
+      "epoch": 1.2415909090909092,
+      "grad_norm": 1.4824173046137992,
+      "learning_rate": 1.069333366891546e-06,
+      "loss": 0.0623,
+      "step": 5463
+    },
+    {
+      "epoch": 1.2418181818181817,
+      "grad_norm": 2.325423189706263,
+      "learning_rate": 1.069270596610808e-06,
+      "loss": 0.0436,
+      "step": 5464
+    },
+    {
+      "epoch": 1.2420454545454545,
+      "grad_norm": 2.6689054962778136,
+      "learning_rate": 1.0692078172706174e-06,
+      "loss": 0.0611,
+      "step": 5465
+    },
+    {
+      "epoch": 1.2422727272727272,
+      "grad_norm": 1.9983876157799607,
+      "learning_rate": 1.0691450288722545e-06,
+      "loss": 0.0631,
+      "step": 5466
+    },
+    {
+      "epoch": 1.2425,
+      "grad_norm": 2.140023600622837,
+      "learning_rate": 1.0690822314169997e-06,
+      "loss": 0.071,
+      "step": 5467
+    },
+    {
+      "epoch": 1.2427272727272727,
+      "grad_norm": 1.2766665102178953,
+      "learning_rate": 1.0690194249061334e-06,
+      "loss": 0.0282,
+      "step": 5468
+    },
+    {
+      "epoch": 1.2429545454545454,
+      "grad_norm": 1.9411797232298165,
+      "learning_rate": 1.0689566093409366e-06,
+      "loss": 0.0653,
+      "step": 5469
+    },
+    {
+      "epoch": 1.2431818181818182,
+      "grad_norm": 1.5225310937070995,
+      "learning_rate": 1.0688937847226897e-06,
+      "loss": 0.0406,
+      "step": 5470
+    },
+    {
+      "epoch": 1.243409090909091,
+      "grad_norm": 1.8677288220667447,
+      "learning_rate": 1.0688309510526742e-06,
+      "loss": 0.0614,
+      "step": 5471
+    },
+    {
+      "epoch": 1.2436363636363637,
+      "grad_norm": 2.3632226347225354,
+      "learning_rate": 1.0687681083321714e-06,
+      "loss": 0.0466,
+      "step": 5472
+    },
+    {
+      "epoch": 1.2438636363636364,
+      "grad_norm": 2.152467639807551,
+      "learning_rate": 1.0687052565624627e-06,
+      "loss": 0.038,
+      "step": 5473
+    },
+    {
+      "epoch": 1.2440909090909091,
+      "grad_norm": 2.5635333247777994,
+      "learning_rate": 1.0686423957448295e-06,
+      "loss": 0.0504,
+      "step": 5474
+    },
+    {
+      "epoch": 1.2443181818181819,
+      "grad_norm": 0.6929031132205887,
+      "learning_rate": 1.068579525880554e-06,
+      "loss": 0.0284,
+      "step": 5475
+    },
+    {
+      "epoch": 1.2445454545454546,
+      "grad_norm": 2.3259786643247424,
+      "learning_rate": 1.0685166469709181e-06,
+      "loss": 0.0404,
+      "step": 5476
+    },
+    {
+      "epoch": 1.2447727272727274,
+      "grad_norm": 2.2460566034243983,
+      "learning_rate": 1.0684537590172042e-06,
+      "loss": 0.0456,
+      "step": 5477
+    },
+    {
+      "epoch": 1.245,
+      "grad_norm": 2.4302393950912062,
+      "learning_rate": 1.0683908620206942e-06,
+      "loss": 0.0709,
+      "step": 5478
+    },
+    {
+      "epoch": 1.2452272727272726,
+      "grad_norm": 6.881886267633361,
+      "learning_rate": 1.0683279559826712e-06,
+      "loss": 0.0764,
+      "step": 5479
+    },
+    {
+      "epoch": 1.2454545454545454,
+      "grad_norm": 1.2741479503878248,
+      "learning_rate": 1.0682650409044178e-06,
+      "loss": 0.0392,
+      "step": 5480
+    },
+    {
+      "epoch": 1.2456818181818181,
+      "grad_norm": 1.8936353869811233,
+      "learning_rate": 1.0682021167872166e-06,
+      "loss": 0.0485,
+      "step": 5481
+    },
+    {
+      "epoch": 1.2459090909090909,
+      "grad_norm": 1.4303850876691886,
+      "learning_rate": 1.0681391836323513e-06,
+      "loss": 0.0217,
+      "step": 5482
+    },
+    {
+      "epoch": 1.2461363636363636,
+      "grad_norm": 2.4985036954785165,
+      "learning_rate": 1.068076241441105e-06,
+      "loss": 0.0408,
+      "step": 5483
+    },
+    {
+      "epoch": 1.2463636363636363,
+      "grad_norm": 0.6964119660009599,
+      "learning_rate": 1.0680132902147609e-06,
+      "loss": 0.0212,
+      "step": 5484
+    },
+    {
+      "epoch": 1.246590909090909,
+      "grad_norm": 1.9683426184350912,
+      "learning_rate": 1.0679503299546029e-06,
+      "loss": 0.0437,
+      "step": 5485
+    },
+    {
+      "epoch": 1.2468181818181818,
+      "grad_norm": 0.4091489801212166,
+      "learning_rate": 1.0678873606619152e-06,
+      "loss": 0.0103,
+      "step": 5486
+    },
+    {
+      "epoch": 1.2470454545454546,
+      "grad_norm": 1.3136565927396147,
+      "learning_rate": 1.0678243823379813e-06,
+      "loss": 0.0313,
+      "step": 5487
+    },
+    {
+      "epoch": 1.2472727272727273,
+      "grad_norm": 1.9307317961069879,
+      "learning_rate": 1.067761394984086e-06,
+      "loss": 0.0698,
+      "step": 5488
+    },
+    {
+      "epoch": 1.2475,
+      "grad_norm": 2.5345068862208384,
+      "learning_rate": 1.067698398601513e-06,
+      "loss": 0.0626,
+      "step": 5489
+    },
+    {
+      "epoch": 1.2477272727272728,
+      "grad_norm": 1.35683760642962,
+      "learning_rate": 1.0676353931915475e-06,
+      "loss": 0.0396,
+      "step": 5490
+    },
+    {
+      "epoch": 1.2479545454545455,
+      "grad_norm": 2.34558968049307,
+      "learning_rate": 1.0675723787554743e-06,
+      "loss": 0.0426,
+      "step": 5491
+    },
+    {
+      "epoch": 1.248181818181818,
+      "grad_norm": 2.146328246707034,
+      "learning_rate": 1.067509355294578e-06,
+      "loss": 0.0588,
+      "step": 5492
+    },
+    {
+      "epoch": 1.248409090909091,
+      "grad_norm": 2.5118467360070498,
+      "learning_rate": 1.0674463228101439e-06,
+      "loss": 0.0676,
+      "step": 5493
+    },
+    {
+      "epoch": 1.2486363636363635,
+      "grad_norm": 2.2905480012747184,
+      "learning_rate": 1.0673832813034576e-06,
+      "loss": 0.0504,
+      "step": 5494
+    },
+    {
+      "epoch": 1.2488636363636363,
+      "grad_norm": 2.452890622117196,
+      "learning_rate": 1.067320230775804e-06,
+      "loss": 0.039,
+      "step": 5495
+    },
+    {
+      "epoch": 1.249090909090909,
+      "grad_norm": 1.6239620348494257,
+      "learning_rate": 1.0672571712284697e-06,
+      "loss": 0.0389,
+      "step": 5496
+    },
+    {
+      "epoch": 1.2493181818181818,
+      "grad_norm": 2.4380786000888723,
+      "learning_rate": 1.0671941026627399e-06,
+      "loss": 0.0571,
+      "step": 5497
+    },
+    {
+      "epoch": 1.2495454545454545,
+      "grad_norm": 2.0610843526047202,
+      "learning_rate": 1.067131025079901e-06,
+      "loss": 0.0395,
+      "step": 5498
+    },
+    {
+      "epoch": 1.2497727272727273,
+      "grad_norm": 1.7105852872445155,
+      "learning_rate": 1.067067938481239e-06,
+      "loss": 0.0494,
+      "step": 5499
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.5700724779180417,
+      "learning_rate": 1.0670048428680405e-06,
+      "loss": 0.051,
+      "step": 5500
+    },
+    {
+      "epoch": 1.2502272727272727,
+      "grad_norm": 1.8011929438116228,
+      "learning_rate": 1.0669417382415923e-06,
+      "loss": 0.0756,
+      "step": 5501
+    },
+    {
+      "epoch": 1.2504545454545455,
+      "grad_norm": 0.5990242278414336,
+      "learning_rate": 1.066878624603181e-06,
+      "loss": 0.0363,
+      "step": 5502
+    },
+    {
+      "epoch": 1.2506818181818182,
+      "grad_norm": 1.402068100947407,
+      "learning_rate": 1.0668155019540936e-06,
+      "loss": 0.0356,
+      "step": 5503
+    },
+    {
+      "epoch": 1.250909090909091,
+      "grad_norm": 1.9464128467592472,
+      "learning_rate": 1.0667523702956174e-06,
+      "loss": 0.0477,
+      "step": 5504
+    },
+    {
+      "epoch": 1.2511363636363637,
+      "grad_norm": 1.5403423696925505,
+      "learning_rate": 1.0666892296290396e-06,
+      "loss": 0.0364,
+      "step": 5505
+    },
+    {
+      "epoch": 1.2513636363636365,
+      "grad_norm": 2.7505728656977313,
+      "learning_rate": 1.0666260799556477e-06,
+      "loss": 0.088,
+      "step": 5506
+    },
+    {
+      "epoch": 1.251590909090909,
+      "grad_norm": 1.1666032654596976,
+      "learning_rate": 1.0665629212767298e-06,
+      "loss": 0.0389,
+      "step": 5507
+    },
+    {
+      "epoch": 1.251818181818182,
+      "grad_norm": 1.251071726703666,
+      "learning_rate": 1.0664997535935735e-06,
+      "loss": 0.0397,
+      "step": 5508
+    },
+    {
+      "epoch": 1.2520454545454545,
+      "grad_norm": 1.8300401004696574,
+      "learning_rate": 1.066436576907467e-06,
+      "loss": 0.0366,
+      "step": 5509
+    },
+    {
+      "epoch": 1.2522727272727272,
+      "grad_norm": 2.0351460190684705,
+      "learning_rate": 1.0663733912196985e-06,
+      "loss": 0.0477,
+      "step": 5510
+    },
+    {
+      "epoch": 1.2525,
+      "grad_norm": 1.3432421640810701,
+      "learning_rate": 1.0663101965315568e-06,
+      "loss": 0.0517,
+      "step": 5511
+    },
+    {
+      "epoch": 1.2527272727272727,
+      "grad_norm": 1.812523982864153,
+      "learning_rate": 1.0662469928443301e-06,
+      "loss": 0.0387,
+      "step": 5512
+    },
+    {
+      "epoch": 1.2529545454545454,
+      "grad_norm": 2.615103686584389,
+      "learning_rate": 1.0661837801593073e-06,
+      "loss": 0.057,
+      "step": 5513
+    },
+    {
+      "epoch": 1.2531818181818182,
+      "grad_norm": 1.70601326154419,
+      "learning_rate": 1.0661205584777777e-06,
+      "loss": 0.0612,
+      "step": 5514
+    },
+    {
+      "epoch": 1.253409090909091,
+      "grad_norm": 2.252090841351334,
+      "learning_rate": 1.0660573278010303e-06,
+      "loss": 0.0432,
+      "step": 5515
+    },
+    {
+      "epoch": 1.2536363636363637,
+      "grad_norm": 1.8889787826084883,
+      "learning_rate": 1.0659940881303545e-06,
+      "loss": 0.0387,
+      "step": 5516
+    },
+    {
+      "epoch": 1.2538636363636364,
+      "grad_norm": 1.619194315627806,
+      "learning_rate": 1.0659308394670398e-06,
+      "loss": 0.0559,
+      "step": 5517
+    },
+    {
+      "epoch": 1.2540909090909091,
+      "grad_norm": 1.4435399068780987,
+      "learning_rate": 1.0658675818123762e-06,
+      "loss": 0.0528,
+      "step": 5518
+    },
+    {
+      "epoch": 1.2543181818181819,
+      "grad_norm": 2.1501256324985927,
+      "learning_rate": 1.0658043151676533e-06,
+      "loss": 0.0429,
+      "step": 5519
+    },
+    {
+      "epoch": 1.2545454545454544,
+      "grad_norm": 2.1005875088606407,
+      "learning_rate": 1.0657410395341616e-06,
+      "loss": 0.0485,
+      "step": 5520
+    },
+    {
+      "epoch": 1.2547727272727274,
+      "grad_norm": 1.8831043082256378,
+      "learning_rate": 1.065677754913191e-06,
+      "loss": 0.0414,
+      "step": 5521
+    },
+    {
+      "epoch": 1.255,
+      "grad_norm": 2.017814565081032,
+      "learning_rate": 1.0656144613060324e-06,
+      "loss": 0.0615,
+      "step": 5522
+    },
+    {
+      "epoch": 1.2552272727272729,
+      "grad_norm": 1.1598109314809522,
+      "learning_rate": 1.065551158713976e-06,
+      "loss": 0.0326,
+      "step": 5523
+    },
+    {
+      "epoch": 1.2554545454545454,
+      "grad_norm": 1.4430240169562198,
+      "learning_rate": 1.065487847138313e-06,
+      "loss": 0.031,
+      "step": 5524
+    },
+    {
+      "epoch": 1.2556818181818181,
+      "grad_norm": 1.931609108347081,
+      "learning_rate": 1.0654245265803343e-06,
+      "loss": 0.0729,
+      "step": 5525
+    },
+    {
+      "epoch": 1.2559090909090909,
+      "grad_norm": 1.7283199868899752,
+      "learning_rate": 1.0653611970413311e-06,
+      "loss": 0.0307,
+      "step": 5526
+    },
+    {
+      "epoch": 1.2561363636363636,
+      "grad_norm": 1.728710145617959,
+      "learning_rate": 1.0652978585225948e-06,
+      "loss": 0.0449,
+      "step": 5527
+    },
+    {
+      "epoch": 1.2563636363636363,
+      "grad_norm": 1.3336631514800863,
+      "learning_rate": 1.0652345110254173e-06,
+      "loss": 0.0216,
+      "step": 5528
+    },
+    {
+      "epoch": 1.256590909090909,
+      "grad_norm": 1.471263498134401,
+      "learning_rate": 1.0651711545510898e-06,
+      "loss": 0.0427,
+      "step": 5529
+    },
+    {
+      "epoch": 1.2568181818181818,
+      "grad_norm": 1.7391748427653422,
+      "learning_rate": 1.0651077891009047e-06,
+      "loss": 0.0515,
+      "step": 5530
+    },
+    {
+      "epoch": 1.2570454545454546,
+      "grad_norm": 1.1299732918286165,
+      "learning_rate": 1.0650444146761537e-06,
+      "loss": 0.0191,
+      "step": 5531
+    },
+    {
+      "epoch": 1.2572727272727273,
+      "grad_norm": 1.4178143190786752,
+      "learning_rate": 1.0649810312781297e-06,
+      "loss": 0.0481,
+      "step": 5532
+    },
+    {
+      "epoch": 1.2575,
+      "grad_norm": 1.8898184971393313,
+      "learning_rate": 1.0649176389081248e-06,
+      "loss": 0.0497,
+      "step": 5533
+    },
+    {
+      "epoch": 1.2577272727272728,
+      "grad_norm": 1.7319119582463967,
+      "learning_rate": 1.0648542375674317e-06,
+      "loss": 0.0239,
+      "step": 5534
+    },
+    {
+      "epoch": 1.2579545454545453,
+      "grad_norm": 3.2396287778516633,
+      "learning_rate": 1.0647908272573433e-06,
+      "loss": 0.0711,
+      "step": 5535
+    },
+    {
+      "epoch": 1.2581818181818183,
+      "grad_norm": 1.4676574277002499,
+      "learning_rate": 1.0647274079791529e-06,
+      "loss": 0.0253,
+      "step": 5536
+    },
+    {
+      "epoch": 1.2584090909090908,
+      "grad_norm": 0.8750526700126028,
+      "learning_rate": 1.0646639797341531e-06,
+      "loss": 0.0187,
+      "step": 5537
+    },
+    {
+      "epoch": 1.2586363636363636,
+      "grad_norm": 1.634489396442035,
+      "learning_rate": 1.064600542523638e-06,
+      "loss": 0.0402,
+      "step": 5538
+    },
+    {
+      "epoch": 1.2588636363636363,
+      "grad_norm": 2.1604080557355037,
+      "learning_rate": 1.064537096348901e-06,
+      "loss": 0.0627,
+      "step": 5539
+    },
+    {
+      "epoch": 1.259090909090909,
+      "grad_norm": 2.7237064101144908,
+      "learning_rate": 1.0644736412112358e-06,
+      "loss": 0.0561,
+      "step": 5540
+    },
+    {
+      "epoch": 1.2593181818181818,
+      "grad_norm": 2.061744966574245,
+      "learning_rate": 1.0644101771119361e-06,
+      "loss": 0.0461,
+      "step": 5541
+    },
+    {
+      "epoch": 1.2595454545454545,
+      "grad_norm": 2.721038192876396,
+      "learning_rate": 1.0643467040522964e-06,
+      "loss": 0.0507,
+      "step": 5542
+    },
+    {
+      "epoch": 1.2597727272727273,
+      "grad_norm": 2.995685565788023,
+      "learning_rate": 1.064283222033611e-06,
+      "loss": 0.0822,
+      "step": 5543
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.2753410505857152,
+      "learning_rate": 1.064219731057174e-06,
+      "loss": 0.0333,
+      "step": 5544
+    },
+    {
+      "epoch": 1.2602272727272728,
+      "grad_norm": 1.501550612317906,
+      "learning_rate": 1.064156231124281e-06,
+      "loss": 0.036,
+      "step": 5545
+    },
+    {
+      "epoch": 1.2604545454545455,
+      "grad_norm": 2.2635388974079174,
+      "learning_rate": 1.0640927222362258e-06,
+      "loss": 0.0466,
+      "step": 5546
+    },
+    {
+      "epoch": 1.2606818181818182,
+      "grad_norm": 1.2680254239491842,
+      "learning_rate": 1.0640292043943042e-06,
+      "loss": 0.0575,
+      "step": 5547
+    },
+    {
+      "epoch": 1.260909090909091,
+      "grad_norm": 2.634991854315612,
+      "learning_rate": 1.0639656775998111e-06,
+      "loss": 0.1075,
+      "step": 5548
+    },
+    {
+      "epoch": 1.2611363636363637,
+      "grad_norm": 1.5931653057600696,
+      "learning_rate": 1.063902141854042e-06,
+      "loss": 0.0415,
+      "step": 5549
+    },
+    {
+      "epoch": 1.2613636363636362,
+      "grad_norm": 3.1095218706748904,
+      "learning_rate": 1.0638385971582927e-06,
+      "loss": 0.0584,
+      "step": 5550
+    },
+    {
+      "epoch": 1.2615909090909092,
+      "grad_norm": 1.641007785866066,
+      "learning_rate": 1.0637750435138586e-06,
+      "loss": 0.0669,
+      "step": 5551
+    },
+    {
+      "epoch": 1.2618181818181817,
+      "grad_norm": 2.6877653258235568,
+      "learning_rate": 1.0637114809220358e-06,
+      "loss": 0.0447,
+      "step": 5552
+    },
+    {
+      "epoch": 1.2620454545454545,
+      "grad_norm": 1.4500236732301777,
+      "learning_rate": 1.0636479093841209e-06,
+      "loss": 0.024,
+      "step": 5553
+    },
+    {
+      "epoch": 1.2622727272727272,
+      "grad_norm": 3.6943894199836285,
+      "learning_rate": 1.0635843289014097e-06,
+      "loss": 0.0813,
+      "step": 5554
+    },
+    {
+      "epoch": 1.2625,
+      "grad_norm": 2.4304956397181976,
+      "learning_rate": 1.0635207394751986e-06,
+      "loss": 0.0932,
+      "step": 5555
+    },
+    {
+      "epoch": 1.2627272727272727,
+      "grad_norm": 2.2069950577657877,
+      "learning_rate": 1.063457141106785e-06,
+      "loss": 0.0961,
+      "step": 5556
+    },
+    {
+      "epoch": 1.2629545454545454,
+      "grad_norm": 2.194933787463352,
+      "learning_rate": 1.063393533797465e-06,
+      "loss": 0.0352,
+      "step": 5557
+    },
+    {
+      "epoch": 1.2631818181818182,
+      "grad_norm": 3.2821742594743735,
+      "learning_rate": 1.0633299175485363e-06,
+      "loss": 0.0685,
+      "step": 5558
+    },
+    {
+      "epoch": 1.263409090909091,
+      "grad_norm": 2.253875534530101,
+      "learning_rate": 1.0632662923612957e-06,
+      "loss": 0.0663,
+      "step": 5559
+    },
+    {
+      "epoch": 1.2636363636363637,
+      "grad_norm": 2.8465926587271375,
+      "learning_rate": 1.063202658237041e-06,
+      "loss": 0.0598,
+      "step": 5560
+    },
+    {
+      "epoch": 1.2638636363636364,
+      "grad_norm": 2.1191305438819414,
+      "learning_rate": 1.0631390151770693e-06,
+      "loss": 0.0817,
+      "step": 5561
+    },
+    {
+      "epoch": 1.2640909090909092,
+      "grad_norm": 1.043260183879419,
+      "learning_rate": 1.063075363182679e-06,
+      "loss": 0.0495,
+      "step": 5562
+    },
+    {
+      "epoch": 1.264318181818182,
+      "grad_norm": 1.9175150520825361,
+      "learning_rate": 1.0630117022551677e-06,
+      "loss": 0.0336,
+      "step": 5563
+    },
+    {
+      "epoch": 1.2645454545454546,
+      "grad_norm": 1.7252409431800535,
+      "learning_rate": 1.0629480323958335e-06,
+      "loss": 0.0391,
+      "step": 5564
+    },
+    {
+      "epoch": 1.2647727272727272,
+      "grad_norm": 2.039389532760748,
+      "learning_rate": 1.0628843536059751e-06,
+      "loss": 0.0562,
+      "step": 5565
+    },
+    {
+      "epoch": 1.2650000000000001,
+      "grad_norm": 5.787788622778891,
+      "learning_rate": 1.0628206658868906e-06,
+      "loss": 0.0949,
+      "step": 5566
+    },
+    {
+      "epoch": 1.2652272727272726,
+      "grad_norm": 2.107608481401556,
+      "learning_rate": 1.0627569692398792e-06,
+      "loss": 0.0377,
+      "step": 5567
+    },
+    {
+      "epoch": 1.2654545454545454,
+      "grad_norm": 2.1588623305695767,
+      "learning_rate": 1.062693263666239e-06,
+      "loss": 0.0504,
+      "step": 5568
+    },
+    {
+      "epoch": 1.2656818181818181,
+      "grad_norm": 1.6192590140356906,
+      "learning_rate": 1.06262954916727e-06,
+      "loss": 0.0529,
+      "step": 5569
+    },
+    {
+      "epoch": 1.2659090909090909,
+      "grad_norm": 2.29657681726111,
+      "learning_rate": 1.062565825744271e-06,
+      "loss": 0.0966,
+      "step": 5570
+    },
+    {
+      "epoch": 1.2661363636363636,
+      "grad_norm": 1.4938337578913778,
+      "learning_rate": 1.0625020933985411e-06,
+      "loss": 0.0566,
+      "step": 5571
+    },
+    {
+      "epoch": 1.2663636363636364,
+      "grad_norm": 1.238829210565282,
+      "learning_rate": 1.0624383521313803e-06,
+      "loss": 0.0237,
+      "step": 5572
+    },
+    {
+      "epoch": 1.266590909090909,
+      "grad_norm": 1.3670958068272823,
+      "learning_rate": 1.0623746019440886e-06,
+      "loss": 0.0225,
+      "step": 5573
+    },
+    {
+      "epoch": 1.2668181818181818,
+      "grad_norm": 1.8127436420395129,
+      "learning_rate": 1.0623108428379656e-06,
+      "loss": 0.0516,
+      "step": 5574
+    },
+    {
+      "epoch": 1.2670454545454546,
+      "grad_norm": 2.8379922489850893,
+      "learning_rate": 1.0622470748143115e-06,
+      "loss": 0.0548,
+      "step": 5575
+    },
+    {
+      "epoch": 1.2672727272727273,
+      "grad_norm": 1.7197015046910904,
+      "learning_rate": 1.0621832978744267e-06,
+      "loss": 0.0524,
+      "step": 5576
+    },
+    {
+      "epoch": 1.2675,
+      "grad_norm": 1.3370806560924258,
+      "learning_rate": 1.062119512019612e-06,
+      "loss": 0.0382,
+      "step": 5577
+    },
+    {
+      "epoch": 1.2677272727272726,
+      "grad_norm": 1.339475409503897,
+      "learning_rate": 1.0620557172511679e-06,
+      "loss": 0.0396,
+      "step": 5578
+    },
+    {
+      "epoch": 1.2679545454545456,
+      "grad_norm": 2.1027547400036606,
+      "learning_rate": 1.061991913570395e-06,
+      "loss": 0.0805,
+      "step": 5579
+    },
+    {
+      "epoch": 1.268181818181818,
+      "grad_norm": 2.246429191001638,
+      "learning_rate": 1.0619281009785946e-06,
+      "loss": 0.0484,
+      "step": 5580
+    },
+    {
+      "epoch": 1.268409090909091,
+      "grad_norm": 3.088088967893037,
+      "learning_rate": 1.0618642794770679e-06,
+      "loss": 0.074,
+      "step": 5581
+    },
+    {
+      "epoch": 1.2686363636363636,
+      "grad_norm": 2.04936485498736,
+      "learning_rate": 1.0618004490671165e-06,
+      "loss": 0.0505,
+      "step": 5582
+    },
+    {
+      "epoch": 1.2688636363636363,
+      "grad_norm": 0.875979479659346,
+      "learning_rate": 1.061736609750042e-06,
+      "loss": 0.0272,
+      "step": 5583
+    },
+    {
+      "epoch": 1.269090909090909,
+      "grad_norm": 1.8093837507150505,
+      "learning_rate": 1.061672761527146e-06,
+      "loss": 0.0346,
+      "step": 5584
+    },
+    {
+      "epoch": 1.2693181818181818,
+      "grad_norm": 2.418285251440226,
+      "learning_rate": 1.0616089043997306e-06,
+      "loss": 0.0543,
+      "step": 5585
+    },
+    {
+      "epoch": 1.2695454545454545,
+      "grad_norm": 2.225673429184962,
+      "learning_rate": 1.0615450383690978e-06,
+      "loss": 0.0502,
+      "step": 5586
+    },
+    {
+      "epoch": 1.2697727272727273,
+      "grad_norm": 1.7510860792003948,
+      "learning_rate": 1.06148116343655e-06,
+      "loss": 0.0467,
+      "step": 5587
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.8401687160163367,
+      "learning_rate": 1.0614172796033903e-06,
+      "loss": 0.0727,
+      "step": 5588
+    },
+    {
+      "epoch": 1.2702272727272728,
+      "grad_norm": 1.0938429855174314,
+      "learning_rate": 1.0613533868709206e-06,
+      "loss": 0.0489,
+      "step": 5589
+    },
+    {
+      "epoch": 1.2704545454545455,
+      "grad_norm": 1.4640949576860556,
+      "learning_rate": 1.061289485240444e-06,
+      "loss": 0.0278,
+      "step": 5590
+    },
+    {
+      "epoch": 1.2706818181818182,
+      "grad_norm": 2.6674771506060733,
+      "learning_rate": 1.0612255747132637e-06,
+      "loss": 0.063,
+      "step": 5591
+    },
+    {
+      "epoch": 1.270909090909091,
+      "grad_norm": 1.7057496265764376,
+      "learning_rate": 1.0611616552906829e-06,
+      "loss": 0.0439,
+      "step": 5592
+    },
+    {
+      "epoch": 1.2711363636363635,
+      "grad_norm": 2.9763309786688157,
+      "learning_rate": 1.061097726974005e-06,
+      "loss": 0.0734,
+      "step": 5593
+    },
+    {
+      "epoch": 1.2713636363636365,
+      "grad_norm": 1.1747324445012333,
+      "learning_rate": 1.0610337897645338e-06,
+      "loss": 0.0251,
+      "step": 5594
+    },
+    {
+      "epoch": 1.271590909090909,
+      "grad_norm": 1.6550342466087353,
+      "learning_rate": 1.0609698436635727e-06,
+      "loss": 0.0491,
+      "step": 5595
+    },
+    {
+      "epoch": 1.2718181818181817,
+      "grad_norm": 1.828025665814938,
+      "learning_rate": 1.0609058886724258e-06,
+      "loss": 0.0272,
+      "step": 5596
+    },
+    {
+      "epoch": 1.2720454545454545,
+      "grad_norm": 2.173439847830357,
+      "learning_rate": 1.0608419247923975e-06,
+      "loss": 0.0369,
+      "step": 5597
+    },
+    {
+      "epoch": 1.2722727272727272,
+      "grad_norm": 1.998055895199448,
+      "learning_rate": 1.0607779520247922e-06,
+      "loss": 0.0361,
+      "step": 5598
+    },
+    {
+      "epoch": 1.2725,
+      "grad_norm": 1.2347560151704742,
+      "learning_rate": 1.060713970370914e-06,
+      "loss": 0.0244,
+      "step": 5599
+    },
+    {
+      "epoch": 1.2727272727272727,
+      "grad_norm": 1.9253672332665652,
+      "learning_rate": 1.0606499798320678e-06,
+      "loss": 0.0513,
+      "step": 5600
+    },
+    {
+      "epoch": 1.2729545454545454,
+      "grad_norm": 2.6297570280129543,
+      "learning_rate": 1.0605859804095587e-06,
+      "loss": 0.051,
+      "step": 5601
+    },
+    {
+      "epoch": 1.2731818181818182,
+      "grad_norm": 2.8648349665496515,
+      "learning_rate": 1.060521972104691e-06,
+      "loss": 0.0578,
+      "step": 5602
+    },
+    {
+      "epoch": 1.273409090909091,
+      "grad_norm": 2.173815036350961,
+      "learning_rate": 1.060457954918771e-06,
+      "loss": 0.0413,
+      "step": 5603
+    },
+    {
+      "epoch": 1.2736363636363637,
+      "grad_norm": 2.0102473299038985,
+      "learning_rate": 1.0603939288531035e-06,
+      "loss": 0.0468,
+      "step": 5604
+    },
+    {
+      "epoch": 1.2738636363636364,
+      "grad_norm": 3.2030509049269074,
+      "learning_rate": 1.060329893908994e-06,
+      "loss": 0.0732,
+      "step": 5605
+    },
+    {
+      "epoch": 1.2740909090909092,
+      "grad_norm": 1.5336648819436391,
+      "learning_rate": 1.0602658500877486e-06,
+      "loss": 0.0263,
+      "step": 5606
+    },
+    {
+      "epoch": 1.274318181818182,
+      "grad_norm": 1.8725801008574783,
+      "learning_rate": 1.060201797390673e-06,
+      "loss": 0.0489,
+      "step": 5607
+    },
+    {
+      "epoch": 1.2745454545454544,
+      "grad_norm": 1.4942132728304909,
+      "learning_rate": 1.0601377358190739e-06,
+      "loss": 0.0421,
+      "step": 5608
+    },
+    {
+      "epoch": 1.2747727272727274,
+      "grad_norm": 1.8570027383123724,
+      "learning_rate": 1.060073665374257e-06,
+      "loss": 0.0652,
+      "step": 5609
+    },
+    {
+      "epoch": 1.275,
+      "grad_norm": 2.1913635993782794,
+      "learning_rate": 1.0600095860575288e-06,
+      "loss": 0.0343,
+      "step": 5610
+    },
+    {
+      "epoch": 1.2752272727272727,
+      "grad_norm": 1.6348539063360767,
+      "learning_rate": 1.0599454978701965e-06,
+      "loss": 0.0473,
+      "step": 5611
+    },
+    {
+      "epoch": 1.2754545454545454,
+      "grad_norm": 2.8996671068174327,
+      "learning_rate": 1.0598814008135668e-06,
+      "loss": 0.0981,
+      "step": 5612
+    },
+    {
+      "epoch": 1.2756818181818181,
+      "grad_norm": 1.1293125496184229,
+      "learning_rate": 1.0598172948889463e-06,
+      "loss": 0.0359,
+      "step": 5613
+    },
+    {
+      "epoch": 1.2759090909090909,
+      "grad_norm": 2.046717425161776,
+      "learning_rate": 1.0597531800976428e-06,
+      "loss": 0.042,
+      "step": 5614
+    },
+    {
+      "epoch": 1.2761363636363636,
+      "grad_norm": 1.3814163845970813,
+      "learning_rate": 1.0596890564409633e-06,
+      "loss": 0.0287,
+      "step": 5615
+    },
+    {
+      "epoch": 1.2763636363636364,
+      "grad_norm": 1.4440539947210602,
+      "learning_rate": 1.0596249239202158e-06,
+      "loss": 0.0369,
+      "step": 5616
+    },
+    {
+      "epoch": 1.276590909090909,
+      "grad_norm": 2.5582775357465164,
+      "learning_rate": 1.059560782536708e-06,
+      "loss": 0.0815,
+      "step": 5617
+    },
+    {
+      "epoch": 1.2768181818181819,
+      "grad_norm": 1.3244117254766996,
+      "learning_rate": 1.0594966322917474e-06,
+      "loss": 0.0354,
+      "step": 5618
+    },
+    {
+      "epoch": 1.2770454545454546,
+      "grad_norm": 2.258384997079986,
+      "learning_rate": 1.0594324731866426e-06,
+      "loss": 0.0502,
+      "step": 5619
+    },
+    {
+      "epoch": 1.2772727272727273,
+      "grad_norm": 1.0775126588223614,
+      "learning_rate": 1.0593683052227016e-06,
+      "loss": 0.0189,
+      "step": 5620
+    },
+    {
+      "epoch": 1.2775,
+      "grad_norm": 2.077795402350193,
+      "learning_rate": 1.0593041284012333e-06,
+      "loss": 0.0484,
+      "step": 5621
+    },
+    {
+      "epoch": 1.2777272727272728,
+      "grad_norm": 1.6559687291931415,
+      "learning_rate": 1.059239942723546e-06,
+      "loss": 0.0322,
+      "step": 5622
+    },
+    {
+      "epoch": 1.2779545454545453,
+      "grad_norm": 1.8167171918372653,
+      "learning_rate": 1.0591757481909486e-06,
+      "loss": 0.0581,
+      "step": 5623
+    },
+    {
+      "epoch": 1.2781818181818183,
+      "grad_norm": 1.8888879347474403,
+      "learning_rate": 1.0591115448047504e-06,
+      "loss": 0.0333,
+      "step": 5624
+    },
+    {
+      "epoch": 1.2784090909090908,
+      "grad_norm": 1.553620901770995,
+      "learning_rate": 1.0590473325662603e-06,
+      "loss": 0.0439,
+      "step": 5625
+    },
+    {
+      "epoch": 1.2786363636363636,
+      "grad_norm": 1.5759124452756084,
+      "learning_rate": 1.058983111476788e-06,
+      "loss": 0.0491,
+      "step": 5626
+    },
+    {
+      "epoch": 1.2788636363636363,
+      "grad_norm": 1.9452091264324922,
+      "learning_rate": 1.0589188815376429e-06,
+      "loss": 0.0613,
+      "step": 5627
+    },
+    {
+      "epoch": 1.279090909090909,
+      "grad_norm": 2.619120583760919,
+      "learning_rate": 1.0588546427501349e-06,
+      "loss": 0.0559,
+      "step": 5628
+    },
+    {
+      "epoch": 1.2793181818181818,
+      "grad_norm": 2.0612512387256956,
+      "learning_rate": 1.0587903951155733e-06,
+      "loss": 0.0718,
+      "step": 5629
+    },
+    {
+      "epoch": 1.2795454545454545,
+      "grad_norm": 2.0819585681045303,
+      "learning_rate": 1.0587261386352695e-06,
+      "loss": 0.0464,
+      "step": 5630
+    },
+    {
+      "epoch": 1.2797727272727273,
+      "grad_norm": 2.8341831707999225,
+      "learning_rate": 1.0586618733105327e-06,
+      "loss": 0.108,
+      "step": 5631
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.5564502103409648,
+      "learning_rate": 1.0585975991426736e-06,
+      "loss": 0.06,
+      "step": 5632
+    },
+    {
+      "epoch": 1.2802272727272728,
+      "grad_norm": 2.466613360596151,
+      "learning_rate": 1.058533316133003e-06,
+      "loss": 0.0554,
+      "step": 5633
+    },
+    {
+      "epoch": 1.2804545454545455,
+      "grad_norm": 1.2662220429663975,
+      "learning_rate": 1.0584690242828321e-06,
+      "loss": 0.0734,
+      "step": 5634
+    },
+    {
+      "epoch": 1.2806818181818183,
+      "grad_norm": 1.1083857120852436,
+      "learning_rate": 1.0584047235934713e-06,
+      "loss": 0.0233,
+      "step": 5635
+    },
+    {
+      "epoch": 1.2809090909090908,
+      "grad_norm": 1.9835988000719624,
+      "learning_rate": 1.0583404140662322e-06,
+      "loss": 0.0551,
+      "step": 5636
+    },
+    {
+      "epoch": 1.2811363636363637,
+      "grad_norm": 2.4129451681198026,
+      "learning_rate": 1.058276095702426e-06,
+      "loss": 0.0344,
+      "step": 5637
+    },
+    {
+      "epoch": 1.2813636363636363,
+      "grad_norm": 2.098913813302621,
+      "learning_rate": 1.0582117685033642e-06,
+      "loss": 0.0942,
+      "step": 5638
+    },
+    {
+      "epoch": 1.2815909090909092,
+      "grad_norm": 1.6877932362035897,
+      "learning_rate": 1.0581474324703589e-06,
+      "loss": 0.0332,
+      "step": 5639
+    },
+    {
+      "epoch": 1.2818181818181817,
+      "grad_norm": 1.7375988492533376,
+      "learning_rate": 1.0580830876047216e-06,
+      "loss": 0.0571,
+      "step": 5640
+    },
+    {
+      "epoch": 1.2820454545454545,
+      "grad_norm": 2.1283518984958074,
+      "learning_rate": 1.0580187339077647e-06,
+      "loss": 0.0455,
+      "step": 5641
+    },
+    {
+      "epoch": 1.2822727272727272,
+      "grad_norm": 0.5577290420583969,
+      "learning_rate": 1.0579543713808002e-06,
+      "loss": 0.0183,
+      "step": 5642
+    },
+    {
+      "epoch": 1.2825,
+      "grad_norm": 2.0661196816979897,
+      "learning_rate": 1.0578900000251408e-06,
+      "loss": 0.0555,
+      "step": 5643
+    },
+    {
+      "epoch": 1.2827272727272727,
+      "grad_norm": 2.446532727349597,
+      "learning_rate": 1.0578256198420991e-06,
+      "loss": 0.0466,
+      "step": 5644
+    },
+    {
+      "epoch": 1.2829545454545455,
+      "grad_norm": 2.5444758942769745,
+      "learning_rate": 1.0577612308329881e-06,
+      "loss": 0.0873,
+      "step": 5645
+    },
+    {
+      "epoch": 1.2831818181818182,
+      "grad_norm": 0.8579521870107516,
+      "learning_rate": 1.0576968329991205e-06,
+      "loss": 0.0223,
+      "step": 5646
+    },
+    {
+      "epoch": 1.283409090909091,
+      "grad_norm": 3.3585615962041118,
+      "learning_rate": 1.0576324263418093e-06,
+      "loss": 0.1471,
+      "step": 5647
+    },
+    {
+      "epoch": 1.2836363636363637,
+      "grad_norm": 1.9700027568100038,
+      "learning_rate": 1.0575680108623685e-06,
+      "loss": 0.048,
+      "step": 5648
+    },
+    {
+      "epoch": 1.2838636363636364,
+      "grad_norm": 3.202237392616603,
+      "learning_rate": 1.057503586562111e-06,
+      "loss": 0.0913,
+      "step": 5649
+    },
+    {
+      "epoch": 1.2840909090909092,
+      "grad_norm": 2.9943901243279867,
+      "learning_rate": 1.0574391534423512e-06,
+      "loss": 0.0599,
+      "step": 5650
+    },
+    {
+      "epoch": 1.2843181818181817,
+      "grad_norm": 2.524721635886412,
+      "learning_rate": 1.0573747115044025e-06,
+      "loss": 0.0432,
+      "step": 5651
+    },
+    {
+      "epoch": 1.2845454545454547,
+      "grad_norm": 1.490728442637951,
+      "learning_rate": 1.057310260749579e-06,
+      "loss": 0.0343,
+      "step": 5652
+    },
+    {
+      "epoch": 1.2847727272727272,
+      "grad_norm": 1.7279144236147763,
+      "learning_rate": 1.0572458011791951e-06,
+      "loss": 0.0582,
+      "step": 5653
+    },
+    {
+      "epoch": 1.285,
+      "grad_norm": 1.8613185521101585,
+      "learning_rate": 1.0571813327945655e-06,
+      "loss": 0.0356,
+      "step": 5654
+    },
+    {
+      "epoch": 1.2852272727272727,
+      "grad_norm": 1.8365770320179706,
+      "learning_rate": 1.0571168555970041e-06,
+      "loss": 0.046,
+      "step": 5655
+    },
+    {
+      "epoch": 1.2854545454545454,
+      "grad_norm": 2.4412419109196284,
+      "learning_rate": 1.0570523695878264e-06,
+      "loss": 0.0483,
+      "step": 5656
+    },
+    {
+      "epoch": 1.2856818181818181,
+      "grad_norm": 1.87838319648248,
+      "learning_rate": 1.056987874768347e-06,
+      "loss": 0.0707,
+      "step": 5657
+    },
+    {
+      "epoch": 1.285909090909091,
+      "grad_norm": 1.2733641074078905,
+      "learning_rate": 1.0569233711398813e-06,
+      "loss": 0.0386,
+      "step": 5658
+    },
+    {
+      "epoch": 1.2861363636363636,
+      "grad_norm": 1.073397919747364,
+      "learning_rate": 1.0568588587037444e-06,
+      "loss": 0.0375,
+      "step": 5659
+    },
+    {
+      "epoch": 1.2863636363636364,
+      "grad_norm": 2.5953606808450207,
+      "learning_rate": 1.056794337461252e-06,
+      "loss": 0.0645,
+      "step": 5660
+    },
+    {
+      "epoch": 1.2865909090909091,
+      "grad_norm": 2.8178255204173466,
+      "learning_rate": 1.05672980741372e-06,
+      "loss": 0.0706,
+      "step": 5661
+    },
+    {
+      "epoch": 1.2868181818181819,
+      "grad_norm": 1.5193220645121037,
+      "learning_rate": 1.0566652685624637e-06,
+      "loss": 0.0276,
+      "step": 5662
+    },
+    {
+      "epoch": 1.2870454545454546,
+      "grad_norm": 1.357416151366874,
+      "learning_rate": 1.0566007209087997e-06,
+      "loss": 0.0288,
+      "step": 5663
+    },
+    {
+      "epoch": 1.2872727272727273,
+      "grad_norm": 1.9700199097111466,
+      "learning_rate": 1.056536164454044e-06,
+      "loss": 0.0391,
+      "step": 5664
+    },
+    {
+      "epoch": 1.2875,
+      "grad_norm": 1.6884885995289407,
+      "learning_rate": 1.056471599199513e-06,
+      "loss": 0.07,
+      "step": 5665
+    },
+    {
+      "epoch": 1.2877272727272726,
+      "grad_norm": 2.426558466639006,
+      "learning_rate": 1.0564070251465233e-06,
+      "loss": 0.0548,
+      "step": 5666
+    },
+    {
+      "epoch": 1.2879545454545456,
+      "grad_norm": 2.4561234996695043,
+      "learning_rate": 1.056342442296392e-06,
+      "loss": 0.0539,
+      "step": 5667
+    },
+    {
+      "epoch": 1.288181818181818,
+      "grad_norm": 4.62068853469211,
+      "learning_rate": 1.0562778506504354e-06,
+      "loss": 0.0958,
+      "step": 5668
+    },
+    {
+      "epoch": 1.2884090909090908,
+      "grad_norm": 1.8408155180990782,
+      "learning_rate": 1.0562132502099711e-06,
+      "loss": 0.0457,
+      "step": 5669
+    },
+    {
+      "epoch": 1.2886363636363636,
+      "grad_norm": 1.3662076327706965,
+      "learning_rate": 1.0561486409763164e-06,
+      "loss": 0.0557,
+      "step": 5670
+    },
+    {
+      "epoch": 1.2888636363636363,
+      "grad_norm": 1.6539840645655008,
+      "learning_rate": 1.0560840229507888e-06,
+      "loss": 0.0713,
+      "step": 5671
+    },
+    {
+      "epoch": 1.289090909090909,
+      "grad_norm": 1.2620567522710022,
+      "learning_rate": 1.056019396134706e-06,
+      "loss": 0.0189,
+      "step": 5672
+    },
+    {
+      "epoch": 1.2893181818181818,
+      "grad_norm": 1.3040130110929482,
+      "learning_rate": 1.0559547605293854e-06,
+      "loss": 0.0223,
+      "step": 5673
+    },
+    {
+      "epoch": 1.2895454545454546,
+      "grad_norm": 2.4020763872382096,
+      "learning_rate": 1.0558901161361456e-06,
+      "loss": 0.0537,
+      "step": 5674
+    },
+    {
+      "epoch": 1.2897727272727273,
+      "grad_norm": 2.258262909804577,
+      "learning_rate": 1.0558254629563048e-06,
+      "loss": 0.045,
+      "step": 5675
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.8009351157148303,
+      "learning_rate": 1.055760800991181e-06,
+      "loss": 0.081,
+      "step": 5676
+    },
+    {
+      "epoch": 1.2902272727272728,
+      "grad_norm": 1.638858186625124,
+      "learning_rate": 1.0556961302420929e-06,
+      "loss": 0.0333,
+      "step": 5677
+    },
+    {
+      "epoch": 1.2904545454545455,
+      "grad_norm": 1.8963061041531926,
+      "learning_rate": 1.0556314507103594e-06,
+      "loss": 0.0417,
+      "step": 5678
+    },
+    {
+      "epoch": 1.2906818181818183,
+      "grad_norm": 2.8112918053029263,
+      "learning_rate": 1.0555667623972993e-06,
+      "loss": 0.0433,
+      "step": 5679
+    },
+    {
+      "epoch": 1.290909090909091,
+      "grad_norm": 1.937840914362235,
+      "learning_rate": 1.0555020653042317e-06,
+      "loss": 0.0504,
+      "step": 5680
+    },
+    {
+      "epoch": 1.2911363636363635,
+      "grad_norm": 2.383144282995582,
+      "learning_rate": 1.0554373594324762e-06,
+      "loss": 0.0434,
+      "step": 5681
+    },
+    {
+      "epoch": 1.2913636363636365,
+      "grad_norm": 1.5469052808147288,
+      "learning_rate": 1.0553726447833517e-06,
+      "loss": 0.0406,
+      "step": 5682
+    },
+    {
+      "epoch": 1.291590909090909,
+      "grad_norm": 2.5130192455087776,
+      "learning_rate": 1.0553079213581782e-06,
+      "loss": 0.0616,
+      "step": 5683
+    },
+    {
+      "epoch": 1.2918181818181818,
+      "grad_norm": 1.715126607860654,
+      "learning_rate": 1.0552431891582755e-06,
+      "loss": 0.0428,
+      "step": 5684
+    },
+    {
+      "epoch": 1.2920454545454545,
+      "grad_norm": 1.2258562040762135,
+      "learning_rate": 1.0551784481849637e-06,
+      "loss": 0.0321,
+      "step": 5685
+    },
+    {
+      "epoch": 1.2922727272727272,
+      "grad_norm": 2.5979591607584727,
+      "learning_rate": 1.0551136984395627e-06,
+      "loss": 0.0374,
+      "step": 5686
+    },
+    {
+      "epoch": 1.2925,
+      "grad_norm": 1.6665784389530447,
+      "learning_rate": 1.0550489399233932e-06,
+      "loss": 0.0356,
+      "step": 5687
+    },
+    {
+      "epoch": 1.2927272727272727,
+      "grad_norm": 1.8886111941318735,
+      "learning_rate": 1.0549841726377754e-06,
+      "loss": 0.0633,
+      "step": 5688
+    },
+    {
+      "epoch": 1.2929545454545455,
+      "grad_norm": 1.618205636362081,
+      "learning_rate": 1.0549193965840302e-06,
+      "loss": 0.0227,
+      "step": 5689
+    },
+    {
+      "epoch": 1.2931818181818182,
+      "grad_norm": 1.935366520954372,
+      "learning_rate": 1.0548546117634783e-06,
+      "loss": 0.0659,
+      "step": 5690
+    },
+    {
+      "epoch": 1.293409090909091,
+      "grad_norm": 1.311486934358705,
+      "learning_rate": 1.0547898181774413e-06,
+      "loss": 0.0604,
+      "step": 5691
+    },
+    {
+      "epoch": 1.2936363636363637,
+      "grad_norm": 2.54937880245274,
+      "learning_rate": 1.05472501582724e-06,
+      "loss": 0.0471,
+      "step": 5692
+    },
+    {
+      "epoch": 1.2938636363636364,
+      "grad_norm": 2.696998293416156,
+      "learning_rate": 1.0546602047141956e-06,
+      "loss": 0.0421,
+      "step": 5693
+    },
+    {
+      "epoch": 1.294090909090909,
+      "grad_norm": 2.698055371526489,
+      "learning_rate": 1.0545953848396303e-06,
+      "loss": 0.0429,
+      "step": 5694
+    },
+    {
+      "epoch": 1.294318181818182,
+      "grad_norm": 1.204312360747689,
+      "learning_rate": 1.0545305562048654e-06,
+      "loss": 0.0412,
+      "step": 5695
+    },
+    {
+      "epoch": 1.2945454545454544,
+      "grad_norm": 1.5021387484372215,
+      "learning_rate": 1.0544657188112238e-06,
+      "loss": 0.0492,
+      "step": 5696
+    },
+    {
+      "epoch": 1.2947727272727274,
+      "grad_norm": 1.1029702754561421,
+      "learning_rate": 1.054400872660026e-06,
+      "loss": 0.0224,
+      "step": 5697
+    },
+    {
+      "epoch": 1.295,
+      "grad_norm": 2.0447613479111473,
+      "learning_rate": 1.054336017752596e-06,
+      "loss": 0.0674,
+      "step": 5698
+    },
+    {
+      "epoch": 1.2952272727272727,
+      "grad_norm": 3.7238802597886456,
+      "learning_rate": 1.0542711540902554e-06,
+      "loss": 0.1129,
+      "step": 5699
+    },
+    {
+      "epoch": 1.2954545454545454,
+      "grad_norm": 1.9827730590528019,
+      "learning_rate": 1.0542062816743268e-06,
+      "loss": 0.0605,
+      "step": 5700
+    },
+    {
+      "epoch": 1.2956818181818182,
+      "grad_norm": 1.8458351895907612,
+      "learning_rate": 1.0541414005061334e-06,
+      "loss": 0.0558,
+      "step": 5701
+    },
+    {
+      "epoch": 1.295909090909091,
+      "grad_norm": 2.198940456997998,
+      "learning_rate": 1.0540765105869984e-06,
+      "loss": 0.077,
+      "step": 5702
+    },
+    {
+      "epoch": 1.2961363636363636,
+      "grad_norm": 2.2805491580404595,
+      "learning_rate": 1.0540116119182448e-06,
+      "loss": 0.0533,
+      "step": 5703
+    },
+    {
+      "epoch": 1.2963636363636364,
+      "grad_norm": 2.8927159684308523,
+      "learning_rate": 1.0539467045011957e-06,
+      "loss": 0.0727,
+      "step": 5704
+    },
+    {
+      "epoch": 1.2965909090909091,
+      "grad_norm": 1.51691601929214,
+      "learning_rate": 1.0538817883371752e-06,
+      "loss": 0.0425,
+      "step": 5705
+    },
+    {
+      "epoch": 1.2968181818181819,
+      "grad_norm": 1.5599714249302488,
+      "learning_rate": 1.0538168634275067e-06,
+      "loss": 0.0299,
+      "step": 5706
+    },
+    {
+      "epoch": 1.2970454545454546,
+      "grad_norm": 2.223894001758107,
+      "learning_rate": 1.053751929773514e-06,
+      "loss": 0.0983,
+      "step": 5707
+    },
+    {
+      "epoch": 1.2972727272727274,
+      "grad_norm": 1.2296250752334503,
+      "learning_rate": 1.053686987376522e-06,
+      "loss": 0.0302,
+      "step": 5708
+    },
+    {
+      "epoch": 1.2974999999999999,
+      "grad_norm": 2.125395296883189,
+      "learning_rate": 1.0536220362378541e-06,
+      "loss": 0.0534,
+      "step": 5709
+    },
+    {
+      "epoch": 1.2977272727272728,
+      "grad_norm": 1.8698610513246763,
+      "learning_rate": 1.053557076358835e-06,
+      "loss": 0.0533,
+      "step": 5710
+    },
+    {
+      "epoch": 1.2979545454545454,
+      "grad_norm": 1.2734342222264392,
+      "learning_rate": 1.0534921077407898e-06,
+      "loss": 0.036,
+      "step": 5711
+    },
+    {
+      "epoch": 1.298181818181818,
+      "grad_norm": 1.9521968326779104,
+      "learning_rate": 1.0534271303850426e-06,
+      "loss": 0.0607,
+      "step": 5712
+    },
+    {
+      "epoch": 1.2984090909090908,
+      "grad_norm": 3.207173278623383,
+      "learning_rate": 1.053362144292919e-06,
+      "loss": 0.0509,
+      "step": 5713
+    },
+    {
+      "epoch": 1.2986363636363636,
+      "grad_norm": 1.2755946111365526,
+      "learning_rate": 1.0532971494657441e-06,
+      "loss": 0.0334,
+      "step": 5714
+    },
+    {
+      "epoch": 1.2988636363636363,
+      "grad_norm": 1.738299309555183,
+      "learning_rate": 1.0532321459048427e-06,
+      "loss": 0.024,
+      "step": 5715
+    },
+    {
+      "epoch": 1.299090909090909,
+      "grad_norm": 3.6758776636955584,
+      "learning_rate": 1.0531671336115409e-06,
+      "loss": 0.1664,
+      "step": 5716
+    },
+    {
+      "epoch": 1.2993181818181818,
+      "grad_norm": 2.346496608599762,
+      "learning_rate": 1.0531021125871645e-06,
+      "loss": 0.0885,
+      "step": 5717
+    },
+    {
+      "epoch": 1.2995454545454546,
+      "grad_norm": 1.73667525717087,
+      "learning_rate": 1.053037082833039e-06,
+      "loss": 0.065,
+      "step": 5718
+    },
+    {
+      "epoch": 1.2997727272727273,
+      "grad_norm": 2.661409420732672,
+      "learning_rate": 1.0529720443504907e-06,
+      "loss": 0.0635,
+      "step": 5719
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.8751641086325483,
+      "learning_rate": 1.0529069971408457e-06,
+      "loss": 0.0573,
+      "step": 5720
+    },
+    {
+      "epoch": 1.3002272727272728,
+      "grad_norm": 1.4453189098320416,
+      "learning_rate": 1.0528419412054306e-06,
+      "loss": 0.0214,
+      "step": 5721
+    },
+    {
+      "epoch": 1.3004545454545455,
+      "grad_norm": 1.6955055150130638,
+      "learning_rate": 1.0527768765455717e-06,
+      "loss": 0.0357,
+      "step": 5722
+    },
+    {
+      "epoch": 1.3006818181818183,
+      "grad_norm": 3.1376087222965476,
+      "learning_rate": 1.0527118031625963e-06,
+      "loss": 0.0807,
+      "step": 5723
+    },
+    {
+      "epoch": 1.3009090909090908,
+      "grad_norm": 2.0505470677000646,
+      "learning_rate": 1.0526467210578308e-06,
+      "loss": 0.0255,
+      "step": 5724
+    },
+    {
+      "epoch": 1.3011363636363638,
+      "grad_norm": 1.9041781257543249,
+      "learning_rate": 1.0525816302326026e-06,
+      "loss": 0.0389,
+      "step": 5725
+    },
+    {
+      "epoch": 1.3013636363636363,
+      "grad_norm": 2.0552890654349096,
+      "learning_rate": 1.052516530688239e-06,
+      "loss": 0.0474,
+      "step": 5726
+    },
+    {
+      "epoch": 1.301590909090909,
+      "grad_norm": 2.0181426620044727,
+      "learning_rate": 1.0524514224260676e-06,
+      "loss": 0.0285,
+      "step": 5727
+    },
+    {
+      "epoch": 1.3018181818181818,
+      "grad_norm": 1.1056388401872261,
+      "learning_rate": 1.052386305447416e-06,
+      "loss": 0.0205,
+      "step": 5728
+    },
+    {
+      "epoch": 1.3020454545454545,
+      "grad_norm": 1.1483687156872373,
+      "learning_rate": 1.0523211797536121e-06,
+      "loss": 0.0268,
+      "step": 5729
+    },
+    {
+      "epoch": 1.3022727272727272,
+      "grad_norm": 1.6695402331098228,
+      "learning_rate": 1.0522560453459835e-06,
+      "loss": 0.0332,
+      "step": 5730
+    },
+    {
+      "epoch": 1.3025,
+      "grad_norm": 2.3826004754067145,
+      "learning_rate": 1.0521909022258589e-06,
+      "loss": 0.0739,
+      "step": 5731
+    },
+    {
+      "epoch": 1.3027272727272727,
+      "grad_norm": 2.210837459220217,
+      "learning_rate": 1.0521257503945667e-06,
+      "loss": 0.0758,
+      "step": 5732
+    },
+    {
+      "epoch": 1.3029545454545455,
+      "grad_norm": 2.2293882840362573,
+      "learning_rate": 1.052060589853435e-06,
+      "loss": 0.0637,
+      "step": 5733
+    },
+    {
+      "epoch": 1.3031818181818182,
+      "grad_norm": 0.8224423419728871,
+      "learning_rate": 1.051995420603793e-06,
+      "loss": 0.0107,
+      "step": 5734
+    },
+    {
+      "epoch": 1.303409090909091,
+      "grad_norm": 1.1576388900763916,
+      "learning_rate": 1.0519302426469693e-06,
+      "loss": 0.0176,
+      "step": 5735
+    },
+    {
+      "epoch": 1.3036363636363637,
+      "grad_norm": 2.8001741808787806,
+      "learning_rate": 1.0518650559842934e-06,
+      "loss": 0.0521,
+      "step": 5736
+    },
+    {
+      "epoch": 1.3038636363636364,
+      "grad_norm": 1.5709873715565927,
+      "learning_rate": 1.051799860617094e-06,
+      "loss": 0.0368,
+      "step": 5737
+    },
+    {
+      "epoch": 1.3040909090909092,
+      "grad_norm": 2.9286576465751235,
+      "learning_rate": 1.051734656546701e-06,
+      "loss": 0.053,
+      "step": 5738
+    },
+    {
+      "epoch": 1.3043181818181817,
+      "grad_norm": 1.663779735820643,
+      "learning_rate": 1.051669443774444e-06,
+      "loss": 0.0562,
+      "step": 5739
+    },
+    {
+      "epoch": 1.3045454545454547,
+      "grad_norm": 1.9015979509626213,
+      "learning_rate": 1.0516042223016527e-06,
+      "loss": 0.0595,
+      "step": 5740
+    },
+    {
+      "epoch": 1.3047727272727272,
+      "grad_norm": 1.4401591166560777,
+      "learning_rate": 1.0515389921296568e-06,
+      "loss": 0.0451,
+      "step": 5741
+    },
+    {
+      "epoch": 1.305,
+      "grad_norm": 1.9367699903367825,
+      "learning_rate": 1.0514737532597867e-06,
+      "loss": 0.0323,
+      "step": 5742
+    },
+    {
+      "epoch": 1.3052272727272727,
+      "grad_norm": 1.810547305799554,
+      "learning_rate": 1.051408505693373e-06,
+      "loss": 0.0698,
+      "step": 5743
+    },
+    {
+      "epoch": 1.3054545454545454,
+      "grad_norm": 1.7400751459733117,
+      "learning_rate": 1.0513432494317457e-06,
+      "loss": 0.039,
+      "step": 5744
+    },
+    {
+      "epoch": 1.3056818181818182,
+      "grad_norm": 2.4577391452694233,
+      "learning_rate": 1.0512779844762358e-06,
+      "loss": 0.0735,
+      "step": 5745
+    },
+    {
+      "epoch": 1.305909090909091,
+      "grad_norm": 1.5285746248173266,
+      "learning_rate": 1.0512127108281741e-06,
+      "loss": 0.0242,
+      "step": 5746
+    },
+    {
+      "epoch": 1.3061363636363637,
+      "grad_norm": 2.054916067340606,
+      "learning_rate": 1.0511474284888918e-06,
+      "loss": 0.0373,
+      "step": 5747
+    },
+    {
+      "epoch": 1.3063636363636364,
+      "grad_norm": 1.70104407662197,
+      "learning_rate": 1.05108213745972e-06,
+      "loss": 0.0516,
+      "step": 5748
+    },
+    {
+      "epoch": 1.3065909090909091,
+      "grad_norm": 1.6568463180333592,
+      "learning_rate": 1.05101683774199e-06,
+      "loss": 0.0367,
+      "step": 5749
+    },
+    {
+      "epoch": 1.3068181818181819,
+      "grad_norm": 1.673161217718892,
+      "learning_rate": 1.0509515293370334e-06,
+      "loss": 0.0409,
+      "step": 5750
+    },
+    {
+      "epoch": 1.3070454545454546,
+      "grad_norm": 2.2146594322204582,
+      "learning_rate": 1.0508862122461822e-06,
+      "loss": 0.0753,
+      "step": 5751
+    },
+    {
+      "epoch": 1.3072727272727271,
+      "grad_norm": 1.4414743208358212,
+      "learning_rate": 1.0508208864707677e-06,
+      "loss": 0.0419,
+      "step": 5752
+    },
+    {
+      "epoch": 1.3075,
+      "grad_norm": 1.6920706501923468,
+      "learning_rate": 1.0507555520121228e-06,
+      "loss": 0.0346,
+      "step": 5753
+    },
+    {
+      "epoch": 1.3077272727272726,
+      "grad_norm": 3.159589466629026,
+      "learning_rate": 1.0506902088715793e-06,
+      "loss": 0.0884,
+      "step": 5754
+    },
+    {
+      "epoch": 1.3079545454545456,
+      "grad_norm": 3.522281831090253,
+      "learning_rate": 1.0506248570504698e-06,
+      "loss": 0.0228,
+      "step": 5755
+    },
+    {
+      "epoch": 1.3081818181818181,
+      "grad_norm": 0.7175730672700847,
+      "learning_rate": 1.050559496550127e-06,
+      "loss": 0.0287,
+      "step": 5756
+    },
+    {
+      "epoch": 1.3084090909090909,
+      "grad_norm": 2.0286305448785225,
+      "learning_rate": 1.0504941273718836e-06,
+      "loss": 0.0434,
+      "step": 5757
+    },
+    {
+      "epoch": 1.3086363636363636,
+      "grad_norm": 1.5304133303100405,
+      "learning_rate": 1.0504287495170727e-06,
+      "loss": 0.0377,
+      "step": 5758
+    },
+    {
+      "epoch": 1.3088636363636363,
+      "grad_norm": 3.3620835820777506,
+      "learning_rate": 1.0503633629870272e-06,
+      "loss": 0.0894,
+      "step": 5759
+    },
+    {
+      "epoch": 1.309090909090909,
+      "grad_norm": 1.514579912629456,
+      "learning_rate": 1.0502979677830808e-06,
+      "loss": 0.0477,
+      "step": 5760
+    },
+    {
+      "epoch": 1.3093181818181818,
+      "grad_norm": 2.1403102634842868,
+      "learning_rate": 1.0502325639065669e-06,
+      "loss": 0.044,
+      "step": 5761
+    },
+    {
+      "epoch": 1.3095454545454546,
+      "grad_norm": 1.4281275411666172,
+      "learning_rate": 1.050167151358819e-06,
+      "loss": 0.0267,
+      "step": 5762
+    },
+    {
+      "epoch": 1.3097727272727273,
+      "grad_norm": 2.1645907639397133,
+      "learning_rate": 1.0501017301411712e-06,
+      "loss": 0.0568,
+      "step": 5763
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.2517907143535747,
+      "learning_rate": 1.0500363002549576e-06,
+      "loss": 0.0516,
+      "step": 5764
+    },
+    {
+      "epoch": 1.3102272727272728,
+      "grad_norm": 1.9320945719608364,
+      "learning_rate": 1.049970861701512e-06,
+      "loss": 0.0303,
+      "step": 5765
+    },
+    {
+      "epoch": 1.3104545454545455,
+      "grad_norm": 2.526743789425754,
+      "learning_rate": 1.0499054144821695e-06,
+      "loss": 0.053,
+      "step": 5766
+    },
+    {
+      "epoch": 1.310681818181818,
+      "grad_norm": 2.8884881048849964,
+      "learning_rate": 1.049839958598264e-06,
+      "loss": 0.0791,
+      "step": 5767
+    },
+    {
+      "epoch": 1.310909090909091,
+      "grad_norm": 2.347720779893212,
+      "learning_rate": 1.0497744940511308e-06,
+      "loss": 0.0601,
+      "step": 5768
+    },
+    {
+      "epoch": 1.3111363636363635,
+      "grad_norm": 1.3493160023181348,
+      "learning_rate": 1.0497090208421045e-06,
+      "loss": 0.0374,
+      "step": 5769
+    },
+    {
+      "epoch": 1.3113636363636363,
+      "grad_norm": 0.5550743434926473,
+      "learning_rate": 1.0496435389725205e-06,
+      "loss": 0.0214,
+      "step": 5770
+    },
+    {
+      "epoch": 1.311590909090909,
+      "grad_norm": 1.7517672684002712,
+      "learning_rate": 1.0495780484437138e-06,
+      "loss": 0.0371,
+      "step": 5771
+    },
+    {
+      "epoch": 1.3118181818181818,
+      "grad_norm": 1.4406899684319283,
+      "learning_rate": 1.0495125492570198e-06,
+      "loss": 0.0488,
+      "step": 5772
+    },
+    {
+      "epoch": 1.3120454545454545,
+      "grad_norm": 2.1382190321549195,
+      "learning_rate": 1.0494470414137745e-06,
+      "loss": 0.0535,
+      "step": 5773
+    },
+    {
+      "epoch": 1.3122727272727273,
+      "grad_norm": 2.3520459028260317,
+      "learning_rate": 1.0493815249153135e-06,
+      "loss": 0.0569,
+      "step": 5774
+    },
+    {
+      "epoch": 1.3125,
+      "grad_norm": 3.9220004023109776,
+      "learning_rate": 1.049315999762973e-06,
+      "loss": 0.1188,
+      "step": 5775
+    },
+    {
+      "epoch": 1.3127272727272727,
+      "grad_norm": 1.3712347136494794,
+      "learning_rate": 1.0492504659580888e-06,
+      "loss": 0.0325,
+      "step": 5776
+    },
+    {
+      "epoch": 1.3129545454545455,
+      "grad_norm": 0.9865107589545173,
+      "learning_rate": 1.0491849235019974e-06,
+      "loss": 0.0215,
+      "step": 5777
+    },
+    {
+      "epoch": 1.3131818181818182,
+      "grad_norm": 1.9217118797721169,
+      "learning_rate": 1.0491193723960356e-06,
+      "loss": 0.049,
+      "step": 5778
+    },
+    {
+      "epoch": 1.313409090909091,
+      "grad_norm": 2.0037063171497445,
+      "learning_rate": 1.0490538126415396e-06,
+      "loss": 0.0349,
+      "step": 5779
+    },
+    {
+      "epoch": 1.3136363636363637,
+      "grad_norm": 1.7069978694323018,
+      "learning_rate": 1.0489882442398467e-06,
+      "loss": 0.0359,
+      "step": 5780
+    },
+    {
+      "epoch": 1.3138636363636365,
+      "grad_norm": 1.913659456586191,
+      "learning_rate": 1.0489226671922938e-06,
+      "loss": 0.0377,
+      "step": 5781
+    },
+    {
+      "epoch": 1.314090909090909,
+      "grad_norm": 1.48789342817106,
+      "learning_rate": 1.0488570815002181e-06,
+      "loss": 0.0408,
+      "step": 5782
+    },
+    {
+      "epoch": 1.314318181818182,
+      "grad_norm": 2.5431162580061217,
+      "learning_rate": 1.048791487164957e-06,
+      "loss": 0.0983,
+      "step": 5783
+    },
+    {
+      "epoch": 1.3145454545454545,
+      "grad_norm": 1.5346238011691686,
+      "learning_rate": 1.0487258841878483e-06,
+      "loss": 0.0618,
+      "step": 5784
+    },
+    {
+      "epoch": 1.3147727272727272,
+      "grad_norm": 1.0892329741792026,
+      "learning_rate": 1.0486602725702297e-06,
+      "loss": 0.0331,
+      "step": 5785
+    },
+    {
+      "epoch": 1.315,
+      "grad_norm": 1.463810296958906,
+      "learning_rate": 1.048594652313439e-06,
+      "loss": 0.0448,
+      "step": 5786
+    },
+    {
+      "epoch": 1.3152272727272727,
+      "grad_norm": 1.565495843633412,
+      "learning_rate": 1.048529023418814e-06,
+      "loss": 0.0326,
+      "step": 5787
+    },
+    {
+      "epoch": 1.3154545454545454,
+      "grad_norm": 2.500114563921178,
+      "learning_rate": 1.0484633858876935e-06,
+      "loss": 0.08,
+      "step": 5788
+    },
+    {
+      "epoch": 1.3156818181818182,
+      "grad_norm": 1.1954186354053733,
+      "learning_rate": 1.0483977397214158e-06,
+      "loss": 0.0227,
+      "step": 5789
+    },
+    {
+      "epoch": 1.315909090909091,
+      "grad_norm": 1.4357032247299892,
+      "learning_rate": 1.0483320849213198e-06,
+      "loss": 0.0259,
+      "step": 5790
+    },
+    {
+      "epoch": 1.3161363636363637,
+      "grad_norm": 2.497538924709891,
+      "learning_rate": 1.0482664214887436e-06,
+      "loss": 0.0852,
+      "step": 5791
+    },
+    {
+      "epoch": 1.3163636363636364,
+      "grad_norm": 1.4797453320959413,
+      "learning_rate": 1.0482007494250268e-06,
+      "loss": 0.0516,
+      "step": 5792
+    },
+    {
+      "epoch": 1.3165909090909091,
+      "grad_norm": 2.46500291474213,
+      "learning_rate": 1.0481350687315084e-06,
+      "loss": 0.0387,
+      "step": 5793
+    },
+    {
+      "epoch": 1.3168181818181819,
+      "grad_norm": 3.11275080685461,
+      "learning_rate": 1.048069379409528e-06,
+      "loss": 0.0781,
+      "step": 5794
+    },
+    {
+      "epoch": 1.3170454545454544,
+      "grad_norm": 2.6445030442726196,
+      "learning_rate": 1.0480036814604243e-06,
+      "loss": 0.0808,
+      "step": 5795
+    },
+    {
+      "epoch": 1.3172727272727274,
+      "grad_norm": 1.4159385550571317,
+      "learning_rate": 1.047937974885538e-06,
+      "loss": 0.0318,
+      "step": 5796
+    },
+    {
+      "epoch": 1.3175,
+      "grad_norm": 3.1475565366252183,
+      "learning_rate": 1.0478722596862085e-06,
+      "loss": 0.1209,
+      "step": 5797
+    },
+    {
+      "epoch": 1.3177272727272729,
+      "grad_norm": 1.6020916182499287,
+      "learning_rate": 1.0478065358637759e-06,
+      "loss": 0.0483,
+      "step": 5798
+    },
+    {
+      "epoch": 1.3179545454545454,
+      "grad_norm": 2.1602135323327136,
+      "learning_rate": 1.0477408034195803e-06,
+      "loss": 0.0288,
+      "step": 5799
+    },
+    {
+      "epoch": 1.3181818181818181,
+      "grad_norm": 1.5218848683359965,
+      "learning_rate": 1.0476750623549622e-06,
+      "loss": 0.0425,
+      "step": 5800
+    },
+    {
+      "epoch": 1.3184090909090909,
+      "grad_norm": 0.8458927539194409,
+      "learning_rate": 1.0476093126712622e-06,
+      "loss": 0.0213,
+      "step": 5801
+    },
+    {
+      "epoch": 1.3186363636363636,
+      "grad_norm": 3.8266406953781553,
+      "learning_rate": 1.047543554369821e-06,
+      "loss": 0.0895,
+      "step": 5802
+    },
+    {
+      "epoch": 1.3188636363636363,
+      "grad_norm": 1.9430334133819582,
+      "learning_rate": 1.0474777874519797e-06,
+      "loss": 0.0562,
+      "step": 5803
+    },
+    {
+      "epoch": 1.319090909090909,
+      "grad_norm": 3.0739678289947947,
+      "learning_rate": 1.0474120119190792e-06,
+      "loss": 0.1162,
+      "step": 5804
+    },
+    {
+      "epoch": 1.3193181818181818,
+      "grad_norm": 2.6349216913793576,
+      "learning_rate": 1.0473462277724608e-06,
+      "loss": 0.0715,
+      "step": 5805
+    },
+    {
+      "epoch": 1.3195454545454546,
+      "grad_norm": 1.611460440257656,
+      "learning_rate": 1.0472804350134661e-06,
+      "loss": 0.0667,
+      "step": 5806
+    },
+    {
+      "epoch": 1.3197727272727273,
+      "grad_norm": 1.937644349058035,
+      "learning_rate": 1.0472146336434364e-06,
+      "loss": 0.0692,
+      "step": 5807
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.30963439609036,
+      "learning_rate": 1.047148823663714e-06,
+      "loss": 0.0408,
+      "step": 5808
+    },
+    {
+      "epoch": 1.3202272727272728,
+      "grad_norm": 1.622916349658791,
+      "learning_rate": 1.0470830050756402e-06,
+      "loss": 0.0305,
+      "step": 5809
+    },
+    {
+      "epoch": 1.3204545454545453,
+      "grad_norm": 1.3608238553313583,
+      "learning_rate": 1.047017177880558e-06,
+      "loss": 0.0307,
+      "step": 5810
+    },
+    {
+      "epoch": 1.3206818181818183,
+      "grad_norm": 1.0900334588274883,
+      "learning_rate": 1.0469513420798091e-06,
+      "loss": 0.0384,
+      "step": 5811
+    },
+    {
+      "epoch": 1.3209090909090908,
+      "grad_norm": 2.1185903077409063,
+      "learning_rate": 1.0468854976747362e-06,
+      "loss": 0.0755,
+      "step": 5812
+    },
+    {
+      "epoch": 1.3211363636363636,
+      "grad_norm": 1.158936393059536,
+      "learning_rate": 1.046819644666682e-06,
+      "loss": 0.021,
+      "step": 5813
+    },
+    {
+      "epoch": 1.3213636363636363,
+      "grad_norm": 2.7498997913500243,
+      "learning_rate": 1.0467537830569893e-06,
+      "loss": 0.0645,
+      "step": 5814
+    },
+    {
+      "epoch": 1.321590909090909,
+      "grad_norm": 2.3248882070924215,
+      "learning_rate": 1.0466879128470012e-06,
+      "loss": 0.0569,
+      "step": 5815
+    },
+    {
+      "epoch": 1.3218181818181818,
+      "grad_norm": 2.5098137012915034,
+      "learning_rate": 1.046622034038061e-06,
+      "loss": 0.06,
+      "step": 5816
+    },
+    {
+      "epoch": 1.3220454545454545,
+      "grad_norm": 2.119276593340401,
+      "learning_rate": 1.046556146631512e-06,
+      "loss": 0.0478,
+      "step": 5817
+    },
+    {
+      "epoch": 1.3222727272727273,
+      "grad_norm": 1.9539806278906247,
+      "learning_rate": 1.0464902506286976e-06,
+      "loss": 0.0393,
+      "step": 5818
+    },
+    {
+      "epoch": 1.3225,
+      "grad_norm": 1.8049848102548873,
+      "learning_rate": 1.0464243460309614e-06,
+      "loss": 0.0655,
+      "step": 5819
+    },
+    {
+      "epoch": 1.3227272727272728,
+      "grad_norm": 1.3507606810268873,
+      "learning_rate": 1.046358432839648e-06,
+      "loss": 0.058,
+      "step": 5820
+    },
+    {
+      "epoch": 1.3229545454545455,
+      "grad_norm": 2.486784328583078,
+      "learning_rate": 1.046292511056101e-06,
+      "loss": 0.0797,
+      "step": 5821
+    },
+    {
+      "epoch": 1.3231818181818182,
+      "grad_norm": 2.0048722734447826,
+      "learning_rate": 1.0462265806816643e-06,
+      "loss": 0.0627,
+      "step": 5822
+    },
+    {
+      "epoch": 1.323409090909091,
+      "grad_norm": 2.922856319036663,
+      "learning_rate": 1.046160641717683e-06,
+      "loss": 0.0695,
+      "step": 5823
+    },
+    {
+      "epoch": 1.3236363636363637,
+      "grad_norm": 0.945320575112193,
+      "learning_rate": 1.0460946941655016e-06,
+      "loss": 0.0205,
+      "step": 5824
+    },
+    {
+      "epoch": 1.3238636363636362,
+      "grad_norm": 1.332520232186204,
+      "learning_rate": 1.0460287380264648e-06,
+      "loss": 0.0283,
+      "step": 5825
+    },
+    {
+      "epoch": 1.3240909090909092,
+      "grad_norm": 2.3304483435164793,
+      "learning_rate": 1.0459627733019172e-06,
+      "loss": 0.0494,
+      "step": 5826
+    },
+    {
+      "epoch": 1.3243181818181817,
+      "grad_norm": 1.7906481517544943,
+      "learning_rate": 1.0458967999932043e-06,
+      "loss": 0.0662,
+      "step": 5827
+    },
+    {
+      "epoch": 1.3245454545454545,
+      "grad_norm": 2.216317333624357,
+      "learning_rate": 1.0458308181016717e-06,
+      "loss": 0.059,
+      "step": 5828
+    },
+    {
+      "epoch": 1.3247727272727272,
+      "grad_norm": 2.2531118995214223,
+      "learning_rate": 1.0457648276286643e-06,
+      "loss": 0.0898,
+      "step": 5829
+    },
+    {
+      "epoch": 1.325,
+      "grad_norm": 1.4157564651424128,
+      "learning_rate": 1.045698828575528e-06,
+      "loss": 0.03,
+      "step": 5830
+    },
+    {
+      "epoch": 1.3252272727272727,
+      "grad_norm": 2.2244950802936714,
+      "learning_rate": 1.0456328209436085e-06,
+      "loss": 0.0691,
+      "step": 5831
+    },
+    {
+      "epoch": 1.3254545454545454,
+      "grad_norm": 2.10911714439626,
+      "learning_rate": 1.045566804734252e-06,
+      "loss": 0.0649,
+      "step": 5832
+    },
+    {
+      "epoch": 1.3256818181818182,
+      "grad_norm": 2.2823183328616636,
+      "learning_rate": 1.0455007799488048e-06,
+      "loss": 0.0585,
+      "step": 5833
+    },
+    {
+      "epoch": 1.325909090909091,
+      "grad_norm": 1.8491356694715912,
+      "learning_rate": 1.045434746588613e-06,
+      "loss": 0.0337,
+      "step": 5834
+    },
+    {
+      "epoch": 1.3261363636363637,
+      "grad_norm": 2.9098935152978145,
+      "learning_rate": 1.0453687046550232e-06,
+      "loss": 0.0948,
+      "step": 5835
+    },
+    {
+      "epoch": 1.3263636363636364,
+      "grad_norm": 1.7520585572012446,
+      "learning_rate": 1.0453026541493822e-06,
+      "loss": 0.0378,
+      "step": 5836
+    },
+    {
+      "epoch": 1.3265909090909092,
+      "grad_norm": 1.8140724900392504,
+      "learning_rate": 1.0452365950730367e-06,
+      "loss": 0.0433,
+      "step": 5837
+    },
+    {
+      "epoch": 1.326818181818182,
+      "grad_norm": 1.0547893848062193,
+      "learning_rate": 1.0451705274273337e-06,
+      "loss": 0.0158,
+      "step": 5838
+    },
+    {
+      "epoch": 1.3270454545454546,
+      "grad_norm": 1.8377926887228162,
+      "learning_rate": 1.045104451213621e-06,
+      "loss": 0.0548,
+      "step": 5839
+    },
+    {
+      "epoch": 1.3272727272727272,
+      "grad_norm": 2.086072172208237,
+      "learning_rate": 1.0450383664332455e-06,
+      "loss": 0.0431,
+      "step": 5840
+    },
+    {
+      "epoch": 1.3275000000000001,
+      "grad_norm": 2.0001499966457157,
+      "learning_rate": 1.044972273087555e-06,
+      "loss": 0.0667,
+      "step": 5841
+    },
+    {
+      "epoch": 1.3277272727272726,
+      "grad_norm": 2.627301408906523,
+      "learning_rate": 1.044906171177897e-06,
+      "loss": 0.0645,
+      "step": 5842
+    },
+    {
+      "epoch": 1.3279545454545454,
+      "grad_norm": 2.128922410496881,
+      "learning_rate": 1.04484006070562e-06,
+      "loss": 0.0445,
+      "step": 5843
+    },
+    {
+      "epoch": 1.3281818181818181,
+      "grad_norm": 2.350527836089654,
+      "learning_rate": 1.0447739416720713e-06,
+      "loss": 0.0355,
+      "step": 5844
+    },
+    {
+      "epoch": 1.3284090909090909,
+      "grad_norm": 2.078721842712213,
+      "learning_rate": 1.0447078140785997e-06,
+      "loss": 0.0363,
+      "step": 5845
+    },
+    {
+      "epoch": 1.3286363636363636,
+      "grad_norm": 1.7635899774544552,
+      "learning_rate": 1.0446416779265535e-06,
+      "loss": 0.0649,
+      "step": 5846
+    },
+    {
+      "epoch": 1.3288636363636364,
+      "grad_norm": 1.4864437599800973,
+      "learning_rate": 1.0445755332172815e-06,
+      "loss": 0.0249,
+      "step": 5847
+    },
+    {
+      "epoch": 1.329090909090909,
+      "grad_norm": 2.159007824001488,
+      "learning_rate": 1.044509379952132e-06,
+      "loss": 0.0869,
+      "step": 5848
+    },
+    {
+      "epoch": 1.3293181818181818,
+      "grad_norm": 1.5373434039858398,
+      "learning_rate": 1.044443218132455e-06,
+      "loss": 0.0307,
+      "step": 5849
+    },
+    {
+      "epoch": 1.3295454545454546,
+      "grad_norm": 3.6274806596183695,
+      "learning_rate": 1.0443770477595986e-06,
+      "loss": 0.1105,
+      "step": 5850
+    },
+    {
+      "epoch": 1.3297727272727273,
+      "grad_norm": 1.453191236655391,
+      "learning_rate": 1.0443108688349127e-06,
+      "loss": 0.0496,
+      "step": 5851
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.0511920860577988,
+      "learning_rate": 1.0442446813597468e-06,
+      "loss": 0.033,
+      "step": 5852
+    },
+    {
+      "epoch": 1.3302272727272726,
+      "grad_norm": 1.9064999406264067,
+      "learning_rate": 1.0441784853354501e-06,
+      "loss": 0.0391,
+      "step": 5853
+    },
+    {
+      "epoch": 1.3304545454545456,
+      "grad_norm": 1.0923648228189506,
+      "learning_rate": 1.044112280763373e-06,
+      "loss": 0.0203,
+      "step": 5854
+    },
+    {
+      "epoch": 1.330681818181818,
+      "grad_norm": 2.707896505680224,
+      "learning_rate": 1.0440460676448652e-06,
+      "loss": 0.0474,
+      "step": 5855
+    },
+    {
+      "epoch": 1.330909090909091,
+      "grad_norm": 3.160486117875778,
+      "learning_rate": 1.0439798459812772e-06,
+      "loss": 0.0495,
+      "step": 5856
+    },
+    {
+      "epoch": 1.3311363636363636,
+      "grad_norm": 3.0050505799826372,
+      "learning_rate": 1.043913615773959e-06,
+      "loss": 0.0622,
+      "step": 5857
+    },
+    {
+      "epoch": 1.3313636363636363,
+      "grad_norm": 2.546252424472989,
+      "learning_rate": 1.0438473770242614e-06,
+      "loss": 0.0676,
+      "step": 5858
+    },
+    {
+      "epoch": 1.331590909090909,
+      "grad_norm": 1.788355849231007,
+      "learning_rate": 1.043781129733535e-06,
+      "loss": 0.027,
+      "step": 5859
+    },
+    {
+      "epoch": 1.3318181818181818,
+      "grad_norm": 4.237798864681473,
+      "learning_rate": 1.0437148739031308e-06,
+      "loss": 0.0829,
+      "step": 5860
+    },
+    {
+      "epoch": 1.3320454545454545,
+      "grad_norm": 2.4101603752964262,
+      "learning_rate": 1.0436486095344e-06,
+      "loss": 0.0689,
+      "step": 5861
+    },
+    {
+      "epoch": 1.3322727272727273,
+      "grad_norm": 3.0941647746399426,
+      "learning_rate": 1.0435823366286934e-06,
+      "loss": 0.0687,
+      "step": 5862
+    },
+    {
+      "epoch": 1.3325,
+      "grad_norm": 1.729580309911627,
+      "learning_rate": 1.043516055187363e-06,
+      "loss": 0.0435,
+      "step": 5863
+    },
+    {
+      "epoch": 1.3327272727272728,
+      "grad_norm": 1.400933655500334,
+      "learning_rate": 1.0434497652117598e-06,
+      "loss": 0.0275,
+      "step": 5864
+    },
+    {
+      "epoch": 1.3329545454545455,
+      "grad_norm": 3.1778972807361265,
+      "learning_rate": 1.0433834667032362e-06,
+      "loss": 0.0664,
+      "step": 5865
+    },
+    {
+      "epoch": 1.3331818181818182,
+      "grad_norm": 1.2943635167067695,
+      "learning_rate": 1.0433171596631433e-06,
+      "loss": 0.0317,
+      "step": 5866
+    },
+    {
+      "epoch": 1.333409090909091,
+      "grad_norm": 2.7627085228063497,
+      "learning_rate": 1.043250844092834e-06,
+      "loss": 0.0605,
+      "step": 5867
+    },
+    {
+      "epoch": 1.3336363636363635,
+      "grad_norm": 2.209643574621457,
+      "learning_rate": 1.0431845199936606e-06,
+      "loss": 0.0603,
+      "step": 5868
+    },
+    {
+      "epoch": 1.3338636363636365,
+      "grad_norm": 1.599684056260015,
+      "learning_rate": 1.043118187366975e-06,
+      "loss": 0.0433,
+      "step": 5869
+    },
+    {
+      "epoch": 1.334090909090909,
+      "grad_norm": 1.544112589113023,
+      "learning_rate": 1.0430518462141302e-06,
+      "loss": 0.0236,
+      "step": 5870
+    },
+    {
+      "epoch": 1.3343181818181817,
+      "grad_norm": 1.8013367927520214,
+      "learning_rate": 1.0429854965364786e-06,
+      "loss": 0.0365,
+      "step": 5871
+    },
+    {
+      "epoch": 1.3345454545454545,
+      "grad_norm": 3.059315316640698,
+      "learning_rate": 1.0429191383353738e-06,
+      "loss": 0.0768,
+      "step": 5872
+    },
+    {
+      "epoch": 1.3347727272727272,
+      "grad_norm": 2.287422639318977,
+      "learning_rate": 1.0428527716121686e-06,
+      "loss": 0.0433,
+      "step": 5873
+    },
+    {
+      "epoch": 1.335,
+      "grad_norm": 2.121257972371813,
+      "learning_rate": 1.0427863963682164e-06,
+      "loss": 0.0571,
+      "step": 5874
+    },
+    {
+      "epoch": 1.3352272727272727,
+      "grad_norm": 1.7702165077115837,
+      "learning_rate": 1.0427200126048708e-06,
+      "loss": 0.036,
+      "step": 5875
+    },
+    {
+      "epoch": 1.3354545454545454,
+      "grad_norm": 3.089161870148214,
+      "learning_rate": 1.0426536203234851e-06,
+      "loss": 0.0528,
+      "step": 5876
+    },
+    {
+      "epoch": 1.3356818181818182,
+      "grad_norm": 3.9116796247796026,
+      "learning_rate": 1.0425872195254138e-06,
+      "loss": 0.1046,
+      "step": 5877
+    },
+    {
+      "epoch": 1.335909090909091,
+      "grad_norm": 1.8570542965154597,
+      "learning_rate": 1.0425208102120105e-06,
+      "loss": 0.0467,
+      "step": 5878
+    },
+    {
+      "epoch": 1.3361363636363637,
+      "grad_norm": 1.7672278299991155,
+      "learning_rate": 1.0424543923846295e-06,
+      "loss": 0.0496,
+      "step": 5879
+    },
+    {
+      "epoch": 1.3363636363636364,
+      "grad_norm": 2.488818327347778,
+      "learning_rate": 1.042387966044625e-06,
+      "loss": 0.0854,
+      "step": 5880
+    },
+    {
+      "epoch": 1.3365909090909092,
+      "grad_norm": 1.5991330672226733,
+      "learning_rate": 1.0423215311933516e-06,
+      "loss": 0.0382,
+      "step": 5881
+    },
+    {
+      "epoch": 1.336818181818182,
+      "grad_norm": 1.7560132301460083,
+      "learning_rate": 1.0422550878321643e-06,
+      "loss": 0.0473,
+      "step": 5882
+    },
+    {
+      "epoch": 1.3370454545454544,
+      "grad_norm": 1.5903539392061654,
+      "learning_rate": 1.0421886359624178e-06,
+      "loss": 0.0542,
+      "step": 5883
+    },
+    {
+      "epoch": 1.3372727272727274,
+      "grad_norm": 2.380692599006304,
+      "learning_rate": 1.0421221755854673e-06,
+      "loss": 0.0249,
+      "step": 5884
+    },
+    {
+      "epoch": 1.3375,
+      "grad_norm": 0.5280268260471602,
+      "learning_rate": 1.042055706702668e-06,
+      "loss": 0.0141,
+      "step": 5885
+    },
+    {
+      "epoch": 1.3377272727272727,
+      "grad_norm": 1.974372770643765,
+      "learning_rate": 1.0419892293153749e-06,
+      "loss": 0.0365,
+      "step": 5886
+    },
+    {
+      "epoch": 1.3379545454545454,
+      "grad_norm": 2.2349430814757585,
+      "learning_rate": 1.041922743424944e-06,
+      "loss": 0.0649,
+      "step": 5887
+    },
+    {
+      "epoch": 1.3381818181818181,
+      "grad_norm": 2.301773179732204,
+      "learning_rate": 1.041856249032731e-06,
+      "loss": 0.0755,
+      "step": 5888
+    },
+    {
+      "epoch": 1.3384090909090909,
+      "grad_norm": 2.763025310305679,
+      "learning_rate": 1.0417897461400923e-06,
+      "loss": 0.0546,
+      "step": 5889
+    },
+    {
+      "epoch": 1.3386363636363636,
+      "grad_norm": 1.7499657517616938,
+      "learning_rate": 1.041723234748383e-06,
+      "loss": 0.0596,
+      "step": 5890
+    },
+    {
+      "epoch": 1.3388636363636364,
+      "grad_norm": 3.841479047882086,
+      "learning_rate": 1.04165671485896e-06,
+      "loss": 0.0993,
+      "step": 5891
+    },
+    {
+      "epoch": 1.339090909090909,
+      "grad_norm": 1.2828714016874656,
+      "learning_rate": 1.04159018647318e-06,
+      "loss": 0.0245,
+      "step": 5892
+    },
+    {
+      "epoch": 1.3393181818181819,
+      "grad_norm": 2.4892850027687587,
+      "learning_rate": 1.041523649592399e-06,
+      "loss": 0.0445,
+      "step": 5893
+    },
+    {
+      "epoch": 1.3395454545454546,
+      "grad_norm": 1.2382884053222338,
+      "learning_rate": 1.0414571042179743e-06,
+      "loss": 0.0528,
+      "step": 5894
+    },
+    {
+      "epoch": 1.3397727272727273,
+      "grad_norm": 1.4240250045365652,
+      "learning_rate": 1.0413905503512626e-06,
+      "loss": 0.0347,
+      "step": 5895
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.4797377191183856,
+      "learning_rate": 1.0413239879936213e-06,
+      "loss": 0.0378,
+      "step": 5896
+    },
+    {
+      "epoch": 1.3402272727272728,
+      "grad_norm": 3.387702513048427,
+      "learning_rate": 1.0412574171464074e-06,
+      "loss": 0.0787,
+      "step": 5897
+    },
+    {
+      "epoch": 1.3404545454545453,
+      "grad_norm": 2.070543856045891,
+      "learning_rate": 1.0411908378109786e-06,
+      "loss": 0.038,
+      "step": 5898
+    },
+    {
+      "epoch": 1.3406818181818183,
+      "grad_norm": 0.992403559340145,
+      "learning_rate": 1.0411242499886925e-06,
+      "loss": 0.0436,
+      "step": 5899
+    },
+    {
+      "epoch": 1.3409090909090908,
+      "grad_norm": 2.2546961472680587,
+      "learning_rate": 1.041057653680907e-06,
+      "loss": 0.079,
+      "step": 5900
+    },
+    {
+      "epoch": 1.3411363636363636,
+      "grad_norm": 1.112394475591204,
+      "learning_rate": 1.0409910488889801e-06,
+      "loss": 0.033,
+      "step": 5901
+    },
+    {
+      "epoch": 1.3413636363636363,
+      "grad_norm": 1.6790685657247355,
+      "learning_rate": 1.04092443561427e-06,
+      "loss": 0.0622,
+      "step": 5902
+    },
+    {
+      "epoch": 1.341590909090909,
+      "grad_norm": 1.9833000824815776,
+      "learning_rate": 1.040857813858135e-06,
+      "loss": 0.0596,
+      "step": 5903
+    },
+    {
+      "epoch": 1.3418181818181818,
+      "grad_norm": 1.2868101637721396,
+      "learning_rate": 1.0407911836219337e-06,
+      "loss": 0.0285,
+      "step": 5904
+    },
+    {
+      "epoch": 1.3420454545454545,
+      "grad_norm": 1.7827838310756388,
+      "learning_rate": 1.0407245449070248e-06,
+      "loss": 0.0477,
+      "step": 5905
+    },
+    {
+      "epoch": 1.3422727272727273,
+      "grad_norm": 1.9752412942783695,
+      "learning_rate": 1.0406578977147672e-06,
+      "loss": 0.0337,
+      "step": 5906
+    },
+    {
+      "epoch": 1.3425,
+      "grad_norm": 0.7829025898368536,
+      "learning_rate": 1.0405912420465197e-06,
+      "loss": 0.025,
+      "step": 5907
+    },
+    {
+      "epoch": 1.3427272727272728,
+      "grad_norm": 1.7611715649838977,
+      "learning_rate": 1.0405245779036418e-06,
+      "loss": 0.0391,
+      "step": 5908
+    },
+    {
+      "epoch": 1.3429545454545455,
+      "grad_norm": 4.323374098044104,
+      "learning_rate": 1.040457905287493e-06,
+      "loss": 0.0579,
+      "step": 5909
+    },
+    {
+      "epoch": 1.3431818181818183,
+      "grad_norm": 1.3575505597779043,
+      "learning_rate": 1.0403912241994326e-06,
+      "loss": 0.0511,
+      "step": 5910
+    },
+    {
+      "epoch": 1.3434090909090908,
+      "grad_norm": 3.1555978391827915,
+      "learning_rate": 1.0403245346408205e-06,
+      "loss": 0.0626,
+      "step": 5911
+    },
+    {
+      "epoch": 1.3436363636363637,
+      "grad_norm": 1.850844105240432,
+      "learning_rate": 1.0402578366130165e-06,
+      "loss": 0.0366,
+      "step": 5912
+    },
+    {
+      "epoch": 1.3438636363636363,
+      "grad_norm": 1.526608328254569,
+      "learning_rate": 1.0401911301173808e-06,
+      "loss": 0.03,
+      "step": 5913
+    },
+    {
+      "epoch": 1.3440909090909092,
+      "grad_norm": 1.5940106315926283,
+      "learning_rate": 1.0401244151552736e-06,
+      "loss": 0.0466,
+      "step": 5914
+    },
+    {
+      "epoch": 1.3443181818181817,
+      "grad_norm": 1.198972496099865,
+      "learning_rate": 1.0400576917280553e-06,
+      "loss": 0.0316,
+      "step": 5915
+    },
+    {
+      "epoch": 1.3445454545454545,
+      "grad_norm": 1.2811071613738076,
+      "learning_rate": 1.0399909598370866e-06,
+      "loss": 0.0313,
+      "step": 5916
+    },
+    {
+      "epoch": 1.3447727272727272,
+      "grad_norm": 1.2633593144412767,
+      "learning_rate": 1.0399242194837283e-06,
+      "loss": 0.0299,
+      "step": 5917
+    },
+    {
+      "epoch": 1.345,
+      "grad_norm": 1.5429277426971215,
+      "learning_rate": 1.0398574706693412e-06,
+      "loss": 0.0398,
+      "step": 5918
+    },
+    {
+      "epoch": 1.3452272727272727,
+      "grad_norm": 1.5189901206541445,
+      "learning_rate": 1.0397907133952866e-06,
+      "loss": 0.0332,
+      "step": 5919
+    },
+    {
+      "epoch": 1.3454545454545455,
+      "grad_norm": 2.2167590059947457,
+      "learning_rate": 1.0397239476629258e-06,
+      "loss": 0.0669,
+      "step": 5920
+    },
+    {
+      "epoch": 1.3456818181818182,
+      "grad_norm": 1.1005502277578845,
+      "learning_rate": 1.03965717347362e-06,
+      "loss": 0.0251,
+      "step": 5921
+    },
+    {
+      "epoch": 1.345909090909091,
+      "grad_norm": 1.2639133915002743,
+      "learning_rate": 1.0395903908287312e-06,
+      "loss": 0.0291,
+      "step": 5922
+    },
+    {
+      "epoch": 1.3461363636363637,
+      "grad_norm": 2.4066573343427433,
+      "learning_rate": 1.0395235997296207e-06,
+      "loss": 0.0776,
+      "step": 5923
+    },
+    {
+      "epoch": 1.3463636363636364,
+      "grad_norm": 1.8273351964001583,
+      "learning_rate": 1.0394568001776511e-06,
+      "loss": 0.0655,
+      "step": 5924
+    },
+    {
+      "epoch": 1.3465909090909092,
+      "grad_norm": 1.5334925999098066,
+      "learning_rate": 1.0393899921741841e-06,
+      "loss": 0.0319,
+      "step": 5925
+    },
+    {
+      "epoch": 1.3468181818181817,
+      "grad_norm": 2.3027924429414988,
+      "learning_rate": 1.0393231757205824e-06,
+      "loss": 0.0487,
+      "step": 5926
+    },
+    {
+      "epoch": 1.3470454545454547,
+      "grad_norm": 2.030598708776666,
+      "learning_rate": 1.0392563508182082e-06,
+      "loss": 0.0832,
+      "step": 5927
+    },
+    {
+      "epoch": 1.3472727272727272,
+      "grad_norm": 2.165473638962749,
+      "learning_rate": 1.0391895174684244e-06,
+      "loss": 0.0548,
+      "step": 5928
+    },
+    {
+      "epoch": 1.3475,
+      "grad_norm": 1.6089703536875661,
+      "learning_rate": 1.0391226756725936e-06,
+      "loss": 0.0514,
+      "step": 5929
+    },
+    {
+      "epoch": 1.3477272727272727,
+      "grad_norm": 1.5371892090950725,
+      "learning_rate": 1.0390558254320792e-06,
+      "loss": 0.0369,
+      "step": 5930
+    },
+    {
+      "epoch": 1.3479545454545454,
+      "grad_norm": 1.5444947382773546,
+      "learning_rate": 1.038988966748244e-06,
+      "loss": 0.0405,
+      "step": 5931
+    },
+    {
+      "epoch": 1.3481818181818181,
+      "grad_norm": 1.3494754921198482,
+      "learning_rate": 1.0389220996224516e-06,
+      "loss": 0.0252,
+      "step": 5932
+    },
+    {
+      "epoch": 1.348409090909091,
+      "grad_norm": 1.6922149891021008,
+      "learning_rate": 1.0388552240560654e-06,
+      "loss": 0.033,
+      "step": 5933
+    },
+    {
+      "epoch": 1.3486363636363636,
+      "grad_norm": 1.4098154210459641,
+      "learning_rate": 1.0387883400504492e-06,
+      "loss": 0.0273,
+      "step": 5934
+    },
+    {
+      "epoch": 1.3488636363636364,
+      "grad_norm": 1.0904282728332966,
+      "learning_rate": 1.0387214476069667e-06,
+      "loss": 0.0367,
+      "step": 5935
+    },
+    {
+      "epoch": 1.3490909090909091,
+      "grad_norm": 1.2813324706005378,
+      "learning_rate": 1.0386545467269823e-06,
+      "loss": 0.0469,
+      "step": 5936
+    },
+    {
+      "epoch": 1.3493181818181819,
+      "grad_norm": 1.668494891688939,
+      "learning_rate": 1.0385876374118599e-06,
+      "loss": 0.0422,
+      "step": 5937
+    },
+    {
+      "epoch": 1.3495454545454546,
+      "grad_norm": 3.0871559662844925,
+      "learning_rate": 1.0385207196629642e-06,
+      "loss": 0.0577,
+      "step": 5938
+    },
+    {
+      "epoch": 1.3497727272727273,
+      "grad_norm": 2.842064113888904,
+      "learning_rate": 1.0384537934816594e-06,
+      "loss": 0.0985,
+      "step": 5939
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.821631822393064,
+      "learning_rate": 1.0383868588693106e-06,
+      "loss": 0.0667,
+      "step": 5940
+    },
+    {
+      "epoch": 1.3502272727272726,
+      "grad_norm": 1.5183556062708006,
+      "learning_rate": 1.0383199158272825e-06,
+      "loss": 0.0478,
+      "step": 5941
+    },
+    {
+      "epoch": 1.3504545454545456,
+      "grad_norm": 1.1077932880017614,
+      "learning_rate": 1.0382529643569402e-06,
+      "loss": 0.0272,
+      "step": 5942
+    },
+    {
+      "epoch": 1.350681818181818,
+      "grad_norm": 1.7026904082930177,
+      "learning_rate": 1.0381860044596491e-06,
+      "loss": 0.0652,
+      "step": 5943
+    },
+    {
+      "epoch": 1.3509090909090908,
+      "grad_norm": 3.166279133019957,
+      "learning_rate": 1.0381190361367745e-06,
+      "loss": 0.0739,
+      "step": 5944
+    },
+    {
+      "epoch": 1.3511363636363636,
+      "grad_norm": 1.594175265484251,
+      "learning_rate": 1.038052059389682e-06,
+      "loss": 0.055,
+      "step": 5945
+    },
+    {
+      "epoch": 1.3513636363636363,
+      "grad_norm": 1.4135074493999866,
+      "learning_rate": 1.0379850742197374e-06,
+      "loss": 0.0407,
+      "step": 5946
+    },
+    {
+      "epoch": 1.351590909090909,
+      "grad_norm": 2.4915923160497946,
+      "learning_rate": 1.0379180806283066e-06,
+      "loss": 0.0647,
+      "step": 5947
+    },
+    {
+      "epoch": 1.3518181818181818,
+      "grad_norm": 1.4611710317154516,
+      "learning_rate": 1.0378510786167559e-06,
+      "loss": 0.0689,
+      "step": 5948
+    },
+    {
+      "epoch": 1.3520454545454546,
+      "grad_norm": 1.8656462903275892,
+      "learning_rate": 1.0377840681864513e-06,
+      "loss": 0.0525,
+      "step": 5949
+    },
+    {
+      "epoch": 1.3522727272727273,
+      "grad_norm": 2.230336642507931,
+      "learning_rate": 1.0377170493387593e-06,
+      "loss": 0.0771,
+      "step": 5950
+    },
+    {
+      "epoch": 1.3525,
+      "grad_norm": 1.9091220364267307,
+      "learning_rate": 1.037650022075047e-06,
+      "loss": 0.0369,
+      "step": 5951
+    },
+    {
+      "epoch": 1.3527272727272728,
+      "grad_norm": 1.4648149158075157,
+      "learning_rate": 1.0375829863966805e-06,
+      "loss": 0.0665,
+      "step": 5952
+    },
+    {
+      "epoch": 1.3529545454545455,
+      "grad_norm": 1.0558413844087298,
+      "learning_rate": 1.0375159423050272e-06,
+      "loss": 0.02,
+      "step": 5953
+    },
+    {
+      "epoch": 1.3531818181818183,
+      "grad_norm": 2.608300470231648,
+      "learning_rate": 1.0374488898014543e-06,
+      "loss": 0.0892,
+      "step": 5954
+    },
+    {
+      "epoch": 1.353409090909091,
+      "grad_norm": 1.815891425384508,
+      "learning_rate": 1.0373818288873288e-06,
+      "loss": 0.0357,
+      "step": 5955
+    },
+    {
+      "epoch": 1.3536363636363635,
+      "grad_norm": 1.6038288665541909,
+      "learning_rate": 1.0373147595640183e-06,
+      "loss": 0.0416,
+      "step": 5956
+    },
+    {
+      "epoch": 1.3538636363636365,
+      "grad_norm": 1.3432893562481214,
+      "learning_rate": 1.037247681832891e-06,
+      "loss": 0.0239,
+      "step": 5957
+    },
+    {
+      "epoch": 1.354090909090909,
+      "grad_norm": 2.768503799732782,
+      "learning_rate": 1.0371805956953137e-06,
+      "loss": 0.0757,
+      "step": 5958
+    },
+    {
+      "epoch": 1.3543181818181818,
+      "grad_norm": 2.1318114866062188,
+      "learning_rate": 1.0371135011526554e-06,
+      "loss": 0.0267,
+      "step": 5959
+    },
+    {
+      "epoch": 1.3545454545454545,
+      "grad_norm": 1.1607697264526429,
+      "learning_rate": 1.0370463982062836e-06,
+      "loss": 0.023,
+      "step": 5960
+    },
+    {
+      "epoch": 1.3547727272727272,
+      "grad_norm": 2.134410418437462,
+      "learning_rate": 1.0369792868575669e-06,
+      "loss": 0.0605,
+      "step": 5961
+    },
+    {
+      "epoch": 1.355,
+      "grad_norm": 1.5264637427211103,
+      "learning_rate": 1.036912167107874e-06,
+      "loss": 0.0411,
+      "step": 5962
+    },
+    {
+      "epoch": 1.3552272727272727,
+      "grad_norm": 2.5474230645529468,
+      "learning_rate": 1.0368450389585733e-06,
+      "loss": 0.0982,
+      "step": 5963
+    },
+    {
+      "epoch": 1.3554545454545455,
+      "grad_norm": 1.8040893971654277,
+      "learning_rate": 1.0367779024110336e-06,
+      "loss": 0.0423,
+      "step": 5964
+    },
+    {
+      "epoch": 1.3556818181818182,
+      "grad_norm": 1.939986999846257,
+      "learning_rate": 1.036710757466624e-06,
+      "loss": 0.0634,
+      "step": 5965
+    },
+    {
+      "epoch": 1.355909090909091,
+      "grad_norm": 3.0550772944573366,
+      "learning_rate": 1.0366436041267142e-06,
+      "loss": 0.0616,
+      "step": 5966
+    },
+    {
+      "epoch": 1.3561363636363637,
+      "grad_norm": 2.041508891987108,
+      "learning_rate": 1.0365764423926727e-06,
+      "loss": 0.0616,
+      "step": 5967
+    },
+    {
+      "epoch": 1.3563636363636364,
+      "grad_norm": 1.8665079006260703,
+      "learning_rate": 1.0365092722658698e-06,
+      "loss": 0.0676,
+      "step": 5968
+    },
+    {
+      "epoch": 1.356590909090909,
+      "grad_norm": 1.271776469051963,
+      "learning_rate": 1.036442093747675e-06,
+      "loss": 0.0339,
+      "step": 5969
+    },
+    {
+      "epoch": 1.356818181818182,
+      "grad_norm": 1.6732456998028444,
+      "learning_rate": 1.036374906839458e-06,
+      "loss": 0.0291,
+      "step": 5970
+    },
+    {
+      "epoch": 1.3570454545454544,
+      "grad_norm": 3.716638849160032,
+      "learning_rate": 1.0363077115425888e-06,
+      "loss": 0.0722,
+      "step": 5971
+    },
+    {
+      "epoch": 1.3572727272727274,
+      "grad_norm": 1.3066328002867371,
+      "learning_rate": 1.0362405078584378e-06,
+      "loss": 0.057,
+      "step": 5972
+    },
+    {
+      "epoch": 1.3575,
+      "grad_norm": 2.0368840601106712,
+      "learning_rate": 1.0361732957883758e-06,
+      "loss": 0.0364,
+      "step": 5973
+    },
+    {
+      "epoch": 1.3577272727272727,
+      "grad_norm": 2.333789576095563,
+      "learning_rate": 1.0361060753337725e-06,
+      "loss": 0.0579,
+      "step": 5974
+    },
+    {
+      "epoch": 1.3579545454545454,
+      "grad_norm": 1.6469206957592908,
+      "learning_rate": 1.0360388464959994e-06,
+      "loss": 0.0205,
+      "step": 5975
+    },
+    {
+      "epoch": 1.3581818181818182,
+      "grad_norm": 1.7876961321583458,
+      "learning_rate": 1.0359716092764268e-06,
+      "loss": 0.0438,
+      "step": 5976
+    },
+    {
+      "epoch": 1.358409090909091,
+      "grad_norm": 2.720580947385,
+      "learning_rate": 1.0359043636764265e-06,
+      "loss": 0.0951,
+      "step": 5977
+    },
+    {
+      "epoch": 1.3586363636363636,
+      "grad_norm": 1.9929815834182179,
+      "learning_rate": 1.0358371096973691e-06,
+      "loss": 0.0685,
+      "step": 5978
+    },
+    {
+      "epoch": 1.3588636363636364,
+      "grad_norm": 1.3937331813842555,
+      "learning_rate": 1.0357698473406266e-06,
+      "loss": 0.0297,
+      "step": 5979
+    },
+    {
+      "epoch": 1.3590909090909091,
+      "grad_norm": 1.034015093467045,
+      "learning_rate": 1.0357025766075699e-06,
+      "loss": 0.0194,
+      "step": 5980
+    },
+    {
+      "epoch": 1.3593181818181819,
+      "grad_norm": 1.4417098929256444,
+      "learning_rate": 1.0356352974995713e-06,
+      "loss": 0.0224,
+      "step": 5981
+    },
+    {
+      "epoch": 1.3595454545454546,
+      "grad_norm": 1.2328105665742755,
+      "learning_rate": 1.0355680100180027e-06,
+      "loss": 0.0458,
+      "step": 5982
+    },
+    {
+      "epoch": 1.3597727272727274,
+      "grad_norm": 1.3059948263183156,
+      "learning_rate": 1.035500714164236e-06,
+      "loss": 0.0174,
+      "step": 5983
+    },
+    {
+      "epoch": 1.3599999999999999,
+      "grad_norm": 3.0744769746650524,
+      "learning_rate": 1.0354334099396435e-06,
+      "loss": 0.0722,
+      "step": 5984
+    },
+    {
+      "epoch": 1.3602272727272728,
+      "grad_norm": 1.755199570341227,
+      "learning_rate": 1.035366097345598e-06,
+      "loss": 0.0668,
+      "step": 5985
+    },
+    {
+      "epoch": 1.3604545454545454,
+      "grad_norm": 1.9081867602455036,
+      "learning_rate": 1.0352987763834716e-06,
+      "loss": 0.0611,
+      "step": 5986
+    },
+    {
+      "epoch": 1.360681818181818,
+      "grad_norm": 1.0982530562484025,
+      "learning_rate": 1.0352314470546373e-06,
+      "loss": 0.0167,
+      "step": 5987
+    },
+    {
+      "epoch": 1.3609090909090908,
+      "grad_norm": 1.896125053458228,
+      "learning_rate": 1.0351641093604684e-06,
+      "loss": 0.0396,
+      "step": 5988
+    },
+    {
+      "epoch": 1.3611363636363636,
+      "grad_norm": 3.0903160557564577,
+      "learning_rate": 1.0350967633023375e-06,
+      "loss": 0.1178,
+      "step": 5989
+    },
+    {
+      "epoch": 1.3613636363636363,
+      "grad_norm": 2.1925760406514248,
+      "learning_rate": 1.0350294088816181e-06,
+      "loss": 0.055,
+      "step": 5990
+    },
+    {
+      "epoch": 1.361590909090909,
+      "grad_norm": 2.4148678659448897,
+      "learning_rate": 1.034962046099684e-06,
+      "loss": 0.0872,
+      "step": 5991
+    },
+    {
+      "epoch": 1.3618181818181818,
+      "grad_norm": 1.8083166340577037,
+      "learning_rate": 1.0348946749579083e-06,
+      "loss": 0.0295,
+      "step": 5992
+    },
+    {
+      "epoch": 1.3620454545454546,
+      "grad_norm": 1.3988170199125154,
+      "learning_rate": 1.0348272954576653e-06,
+      "loss": 0.0194,
+      "step": 5993
+    },
+    {
+      "epoch": 1.3622727272727273,
+      "grad_norm": 2.0826427862777317,
+      "learning_rate": 1.0347599076003284e-06,
+      "loss": 0.0561,
+      "step": 5994
+    },
+    {
+      "epoch": 1.3625,
+      "grad_norm": 3.3801473281158634,
+      "learning_rate": 1.0346925113872724e-06,
+      "loss": 0.0687,
+      "step": 5995
+    },
+    {
+      "epoch": 1.3627272727272728,
+      "grad_norm": 1.0703365597299879,
+      "learning_rate": 1.0346251068198712e-06,
+      "loss": 0.0435,
+      "step": 5996
+    },
+    {
+      "epoch": 1.3629545454545455,
+      "grad_norm": 2.735981585844725,
+      "learning_rate": 1.0345576938994998e-06,
+      "loss": 0.0828,
+      "step": 5997
+    },
+    {
+      "epoch": 1.3631818181818183,
+      "grad_norm": 1.2286389783354679,
+      "learning_rate": 1.034490272627532e-06,
+      "loss": 0.0265,
+      "step": 5998
+    },
+    {
+      "epoch": 1.3634090909090908,
+      "grad_norm": 1.8420945013234222,
+      "learning_rate": 1.0344228430053435e-06,
+      "loss": 0.0689,
+      "step": 5999
+    },
+    {
+      "epoch": 1.3636363636363638,
+      "grad_norm": 1.714016468418457,
+      "learning_rate": 1.0343554050343088e-06,
+      "loss": 0.0262,
+      "step": 6000
+    },
+    {
+      "epoch": 1.3638636363636363,
+      "grad_norm": 1.3232264480511176,
+      "learning_rate": 1.0342879587158033e-06,
+      "loss": 0.0361,
+      "step": 6001
+    },
+    {
+      "epoch": 1.364090909090909,
+      "grad_norm": 2.585239182006051,
+      "learning_rate": 1.0342205040512021e-06,
+      "loss": 0.0344,
+      "step": 6002
+    },
+    {
+      "epoch": 1.3643181818181818,
+      "grad_norm": 1.873062844306329,
+      "learning_rate": 1.0341530410418813e-06,
+      "loss": 0.0445,
+      "step": 6003
+    },
+    {
+      "epoch": 1.3645454545454545,
+      "grad_norm": 3.426771320637717,
+      "learning_rate": 1.0340855696892157e-06,
+      "loss": 0.1258,
+      "step": 6004
+    },
+    {
+      "epoch": 1.3647727272727272,
+      "grad_norm": 1.12101502008358,
+      "learning_rate": 1.0340180899945821e-06,
+      "loss": 0.0254,
+      "step": 6005
+    },
+    {
+      "epoch": 1.365,
+      "grad_norm": 2.089070519912733,
+      "learning_rate": 1.0339506019593557e-06,
+      "loss": 0.0576,
+      "step": 6006
+    },
+    {
+      "epoch": 1.3652272727272727,
+      "grad_norm": 2.1602426910446177,
+      "learning_rate": 1.0338831055849133e-06,
+      "loss": 0.0424,
+      "step": 6007
+    },
+    {
+      "epoch": 1.3654545454545455,
+      "grad_norm": 1.0477912011144992,
+      "learning_rate": 1.033815600872631e-06,
+      "loss": 0.0259,
+      "step": 6008
+    },
+    {
+      "epoch": 1.3656818181818182,
+      "grad_norm": 1.0914390898338273,
+      "learning_rate": 1.0337480878238852e-06,
+      "loss": 0.0224,
+      "step": 6009
+    },
+    {
+      "epoch": 1.365909090909091,
+      "grad_norm": 2.545229371001049,
+      "learning_rate": 1.033680566440053e-06,
+      "loss": 0.072,
+      "step": 6010
+    },
+    {
+      "epoch": 1.3661363636363637,
+      "grad_norm": 1.7924747360058921,
+      "learning_rate": 1.033613036722511e-06,
+      "loss": 0.0569,
+      "step": 6011
+    },
+    {
+      "epoch": 1.3663636363636364,
+      "grad_norm": 2.9629439085465967,
+      "learning_rate": 1.0335454986726363e-06,
+      "loss": 0.0813,
+      "step": 6012
+    },
+    {
+      "epoch": 1.3665909090909092,
+      "grad_norm": 2.4000095035522437,
+      "learning_rate": 1.0334779522918062e-06,
+      "loss": 0.0499,
+      "step": 6013
+    },
+    {
+      "epoch": 1.3668181818181817,
+      "grad_norm": 3.8820811443030716,
+      "learning_rate": 1.033410397581398e-06,
+      "loss": 0.0774,
+      "step": 6014
+    },
+    {
+      "epoch": 1.3670454545454547,
+      "grad_norm": 3.93358780849887,
+      "learning_rate": 1.0333428345427894e-06,
+      "loss": 0.0769,
+      "step": 6015
+    },
+    {
+      "epoch": 1.3672727272727272,
+      "grad_norm": 1.2516314638466965,
+      "learning_rate": 1.0332752631773578e-06,
+      "loss": 0.0354,
+      "step": 6016
+    },
+    {
+      "epoch": 1.3675,
+      "grad_norm": 2.3957295873033675,
+      "learning_rate": 1.0332076834864812e-06,
+      "loss": 0.093,
+      "step": 6017
+    },
+    {
+      "epoch": 1.3677272727272727,
+      "grad_norm": 1.7971392829059276,
+      "learning_rate": 1.033140095471538e-06,
+      "loss": 0.056,
+      "step": 6018
+    },
+    {
+      "epoch": 1.3679545454545454,
+      "grad_norm": 0.9540107183010806,
+      "learning_rate": 1.0330724991339064e-06,
+      "loss": 0.0171,
+      "step": 6019
+    },
+    {
+      "epoch": 1.3681818181818182,
+      "grad_norm": 2.2678510743528113,
+      "learning_rate": 1.0330048944749643e-06,
+      "loss": 0.0601,
+      "step": 6020
+    },
+    {
+      "epoch": 1.368409090909091,
+      "grad_norm": 2.3396796693901916,
+      "learning_rate": 1.0329372814960906e-06,
+      "loss": 0.0527,
+      "step": 6021
+    },
+    {
+      "epoch": 1.3686363636363637,
+      "grad_norm": 2.231017139254196,
+      "learning_rate": 1.0328696601986641e-06,
+      "loss": 0.0461,
+      "step": 6022
+    },
+    {
+      "epoch": 1.3688636363636364,
+      "grad_norm": 2.0880650077557505,
+      "learning_rate": 1.0328020305840638e-06,
+      "loss": 0.0334,
+      "step": 6023
+    },
+    {
+      "epoch": 1.3690909090909091,
+      "grad_norm": 1.4294861712863618,
+      "learning_rate": 1.0327343926536684e-06,
+      "loss": 0.0497,
+      "step": 6024
+    },
+    {
+      "epoch": 1.3693181818181819,
+      "grad_norm": 1.1149563284380632,
+      "learning_rate": 1.0326667464088577e-06,
+      "loss": 0.0204,
+      "step": 6025
+    },
+    {
+      "epoch": 1.3695454545454546,
+      "grad_norm": 1.8079064931504012,
+      "learning_rate": 1.0325990918510107e-06,
+      "loss": 0.0343,
+      "step": 6026
+    },
+    {
+      "epoch": 1.3697727272727271,
+      "grad_norm": 1.9432520348134596,
+      "learning_rate": 1.032531428981507e-06,
+      "loss": 0.0365,
+      "step": 6027
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.3204040194926567,
+      "learning_rate": 1.0324637578017266e-06,
+      "loss": 0.035,
+      "step": 6028
+    },
+    {
+      "epoch": 1.3702272727272726,
+      "grad_norm": 1.891658625697901,
+      "learning_rate": 1.0323960783130494e-06,
+      "loss": 0.0573,
+      "step": 6029
+    },
+    {
+      "epoch": 1.3704545454545456,
+      "grad_norm": 2.6239880084486997,
+      "learning_rate": 1.0323283905168552e-06,
+      "loss": 0.1052,
+      "step": 6030
+    },
+    {
+      "epoch": 1.3706818181818181,
+      "grad_norm": 2.1355623692538965,
+      "learning_rate": 1.0322606944145248e-06,
+      "loss": 0.0673,
+      "step": 6031
+    },
+    {
+      "epoch": 1.3709090909090909,
+      "grad_norm": 1.7383116147873612,
+      "learning_rate": 1.032192990007438e-06,
+      "loss": 0.0588,
+      "step": 6032
+    },
+    {
+      "epoch": 1.3711363636363636,
+      "grad_norm": 1.3146546521530664,
+      "learning_rate": 1.0321252772969758e-06,
+      "loss": 0.0255,
+      "step": 6033
+    },
+    {
+      "epoch": 1.3713636363636363,
+      "grad_norm": 2.8097800315051353,
+      "learning_rate": 1.0320575562845192e-06,
+      "loss": 0.0513,
+      "step": 6034
+    },
+    {
+      "epoch": 1.371590909090909,
+      "grad_norm": 1.6733478120476895,
+      "learning_rate": 1.0319898269714485e-06,
+      "loss": 0.0272,
+      "step": 6035
+    },
+    {
+      "epoch": 1.3718181818181818,
+      "grad_norm": 2.2657129555998483,
+      "learning_rate": 1.0319220893591456e-06,
+      "loss": 0.0668,
+      "step": 6036
+    },
+    {
+      "epoch": 1.3720454545454546,
+      "grad_norm": 2.2218561451307695,
+      "learning_rate": 1.031854343448991e-06,
+      "loss": 0.0515,
+      "step": 6037
+    },
+    {
+      "epoch": 1.3722727272727273,
+      "grad_norm": 1.5852477189659253,
+      "learning_rate": 1.0317865892423668e-06,
+      "loss": 0.0423,
+      "step": 6038
+    },
+    {
+      "epoch": 1.3725,
+      "grad_norm": 3.019217598178217,
+      "learning_rate": 1.0317188267406542e-06,
+      "loss": 0.0654,
+      "step": 6039
+    },
+    {
+      "epoch": 1.3727272727272728,
+      "grad_norm": 1.7542015529022565,
+      "learning_rate": 1.0316510559452352e-06,
+      "loss": 0.0531,
+      "step": 6040
+    },
+    {
+      "epoch": 1.3729545454545455,
+      "grad_norm": 1.0792091721210464,
+      "learning_rate": 1.031583276857492e-06,
+      "loss": 0.0415,
+      "step": 6041
+    },
+    {
+      "epoch": 1.373181818181818,
+      "grad_norm": 2.5786081500725753,
+      "learning_rate": 1.0315154894788063e-06,
+      "loss": 0.0508,
+      "step": 6042
+    },
+    {
+      "epoch": 1.373409090909091,
+      "grad_norm": 1.319851045107395,
+      "learning_rate": 1.0314476938105604e-06,
+      "loss": 0.0249,
+      "step": 6043
+    },
+    {
+      "epoch": 1.3736363636363635,
+      "grad_norm": 2.354812138824377,
+      "learning_rate": 1.031379889854137e-06,
+      "loss": 0.0412,
+      "step": 6044
+    },
+    {
+      "epoch": 1.3738636363636363,
+      "grad_norm": 1.4003548606733776,
+      "learning_rate": 1.0313120776109188e-06,
+      "loss": 0.0464,
+      "step": 6045
+    },
+    {
+      "epoch": 1.374090909090909,
+      "grad_norm": 2.3674681835320253,
+      "learning_rate": 1.0312442570822883e-06,
+      "loss": 0.0424,
+      "step": 6046
+    },
+    {
+      "epoch": 1.3743181818181818,
+      "grad_norm": 1.4143608177874358,
+      "learning_rate": 1.0311764282696288e-06,
+      "loss": 0.0313,
+      "step": 6047
+    },
+    {
+      "epoch": 1.3745454545454545,
+      "grad_norm": 2.348588579547304,
+      "learning_rate": 1.0311085911743233e-06,
+      "loss": 0.0466,
+      "step": 6048
+    },
+    {
+      "epoch": 1.3747727272727273,
+      "grad_norm": 2.110276160834307,
+      "learning_rate": 1.031040745797755e-06,
+      "loss": 0.0341,
+      "step": 6049
+    },
+    {
+      "epoch": 1.375,
+      "grad_norm": 3.18145773808921,
+      "learning_rate": 1.0309728921413078e-06,
+      "loss": 0.1048,
+      "step": 6050
+    },
+    {
+      "epoch": 1.3752272727272727,
+      "grad_norm": 1.8153134625175578,
+      "learning_rate": 1.0309050302063647e-06,
+      "loss": 0.079,
+      "step": 6051
+    },
+    {
+      "epoch": 1.3754545454545455,
+      "grad_norm": 1.828221378399437,
+      "learning_rate": 1.0308371599943103e-06,
+      "loss": 0.051,
+      "step": 6052
+    },
+    {
+      "epoch": 1.3756818181818182,
+      "grad_norm": 3.3481913735105153,
+      "learning_rate": 1.0307692815065278e-06,
+      "loss": 0.137,
+      "step": 6053
+    },
+    {
+      "epoch": 1.375909090909091,
+      "grad_norm": 2.405746174238826,
+      "learning_rate": 1.0307013947444021e-06,
+      "loss": 0.1029,
+      "step": 6054
+    },
+    {
+      "epoch": 1.3761363636363637,
+      "grad_norm": 2.429074761776705,
+      "learning_rate": 1.030633499709317e-06,
+      "loss": 0.0487,
+      "step": 6055
+    },
+    {
+      "epoch": 1.3763636363636365,
+      "grad_norm": 1.7842971322374512,
+      "learning_rate": 1.0305655964026573e-06,
+      "loss": 0.0552,
+      "step": 6056
+    },
+    {
+      "epoch": 1.376590909090909,
+      "grad_norm": 1.981654071219296,
+      "learning_rate": 1.0304976848258073e-06,
+      "loss": 0.0644,
+      "step": 6057
+    },
+    {
+      "epoch": 1.376818181818182,
+      "grad_norm": 2.0009713720693774,
+      "learning_rate": 1.0304297649801523e-06,
+      "loss": 0.05,
+      "step": 6058
+    },
+    {
+      "epoch": 1.3770454545454545,
+      "grad_norm": 2.178177461457059,
+      "learning_rate": 1.0303618368670769e-06,
+      "loss": 0.082,
+      "step": 6059
+    },
+    {
+      "epoch": 1.3772727272727272,
+      "grad_norm": 2.924664394232642,
+      "learning_rate": 1.0302939004879664e-06,
+      "loss": 0.0851,
+      "step": 6060
+    },
+    {
+      "epoch": 1.3775,
+      "grad_norm": 0.6243132179254567,
+      "learning_rate": 1.0302259558442065e-06,
+      "loss": 0.0246,
+      "step": 6061
+    },
+    {
+      "epoch": 1.3777272727272727,
+      "grad_norm": 1.835688104489339,
+      "learning_rate": 1.0301580029371823e-06,
+      "loss": 0.0408,
+      "step": 6062
+    },
+    {
+      "epoch": 1.3779545454545454,
+      "grad_norm": 1.6029573226232519,
+      "learning_rate": 1.0300900417682793e-06,
+      "loss": 0.0348,
+      "step": 6063
+    },
+    {
+      "epoch": 1.3781818181818182,
+      "grad_norm": 1.8201609448117153,
+      "learning_rate": 1.030022072338884e-06,
+      "loss": 0.043,
+      "step": 6064
+    },
+    {
+      "epoch": 1.378409090909091,
+      "grad_norm": 1.8054747625922765,
+      "learning_rate": 1.029954094650382e-06,
+      "loss": 0.0387,
+      "step": 6065
+    },
+    {
+      "epoch": 1.3786363636363637,
+      "grad_norm": 2.530629583011827,
+      "learning_rate": 1.0298861087041593e-06,
+      "loss": 0.0541,
+      "step": 6066
+    },
+    {
+      "epoch": 1.3788636363636364,
+      "grad_norm": 1.3384229964587346,
+      "learning_rate": 1.0298181145016027e-06,
+      "loss": 0.0285,
+      "step": 6067
+    },
+    {
+      "epoch": 1.3790909090909091,
+      "grad_norm": 2.1974469071660017,
+      "learning_rate": 1.0297501120440984e-06,
+      "loss": 0.0505,
+      "step": 6068
+    },
+    {
+      "epoch": 1.3793181818181819,
+      "grad_norm": 1.4501967836103347,
+      "learning_rate": 1.029682101333033e-06,
+      "loss": 0.0289,
+      "step": 6069
+    },
+    {
+      "epoch": 1.3795454545454544,
+      "grad_norm": 1.7861368891362222,
+      "learning_rate": 1.0296140823697937e-06,
+      "loss": 0.0461,
+      "step": 6070
+    },
+    {
+      "epoch": 1.3797727272727274,
+      "grad_norm": 2.532526911653803,
+      "learning_rate": 1.0295460551557676e-06,
+      "loss": 0.0518,
+      "step": 6071
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.1067786620016664,
+      "learning_rate": 1.0294780196923415e-06,
+      "loss": 0.019,
+      "step": 6072
+    },
+    {
+      "epoch": 1.3802272727272729,
+      "grad_norm": 1.4315092342209796,
+      "learning_rate": 1.0294099759809028e-06,
+      "loss": 0.0329,
+      "step": 6073
+    },
+    {
+      "epoch": 1.3804545454545454,
+      "grad_norm": 2.3604854236422304,
+      "learning_rate": 1.0293419240228393e-06,
+      "loss": 0.1055,
+      "step": 6074
+    },
+    {
+      "epoch": 1.3806818181818181,
+      "grad_norm": 1.9424544280310132,
+      "learning_rate": 1.0292738638195385e-06,
+      "loss": 0.04,
+      "step": 6075
+    },
+    {
+      "epoch": 1.3809090909090909,
+      "grad_norm": 1.533023566790882,
+      "learning_rate": 1.0292057953723883e-06,
+      "loss": 0.0297,
+      "step": 6076
+    },
+    {
+      "epoch": 1.3811363636363636,
+      "grad_norm": 1.7479795253345218,
+      "learning_rate": 1.0291377186827768e-06,
+      "loss": 0.0256,
+      "step": 6077
+    },
+    {
+      "epoch": 1.3813636363636363,
+      "grad_norm": 1.6048905743945805,
+      "learning_rate": 1.0290696337520923e-06,
+      "loss": 0.0487,
+      "step": 6078
+    },
+    {
+      "epoch": 1.381590909090909,
+      "grad_norm": 1.6771166055470363,
+      "learning_rate": 1.0290015405817228e-06,
+      "loss": 0.0493,
+      "step": 6079
+    },
+    {
+      "epoch": 1.3818181818181818,
+      "grad_norm": 2.6755517118693173,
+      "learning_rate": 1.0289334391730572e-06,
+      "loss": 0.0974,
+      "step": 6080
+    },
+    {
+      "epoch": 1.3820454545454546,
+      "grad_norm": 2.998152395766971,
+      "learning_rate": 1.028865329527484e-06,
+      "loss": 0.0879,
+      "step": 6081
+    },
+    {
+      "epoch": 1.3822727272727273,
+      "grad_norm": 1.6607832339656288,
+      "learning_rate": 1.0287972116463922e-06,
+      "loss": 0.0597,
+      "step": 6082
+    },
+    {
+      "epoch": 1.3825,
+      "grad_norm": 1.2908356479031304,
+      "learning_rate": 1.0287290855311707e-06,
+      "loss": 0.031,
+      "step": 6083
+    },
+    {
+      "epoch": 1.3827272727272728,
+      "grad_norm": 1.9154662785056142,
+      "learning_rate": 1.028660951183209e-06,
+      "loss": 0.0468,
+      "step": 6084
+    },
+    {
+      "epoch": 1.3829545454545453,
+      "grad_norm": 1.6941839189444157,
+      "learning_rate": 1.028592808603896e-06,
+      "loss": 0.0263,
+      "step": 6085
+    },
+    {
+      "epoch": 1.3831818181818183,
+      "grad_norm": 3.065137078812516,
+      "learning_rate": 1.0285246577946217e-06,
+      "loss": 0.0857,
+      "step": 6086
+    },
+    {
+      "epoch": 1.3834090909090908,
+      "grad_norm": 2.4000968036143577,
+      "learning_rate": 1.0284564987567758e-06,
+      "loss": 0.0446,
+      "step": 6087
+    },
+    {
+      "epoch": 1.3836363636363636,
+      "grad_norm": 1.9893670677543338,
+      "learning_rate": 1.0283883314917476e-06,
+      "loss": 0.0424,
+      "step": 6088
+    },
+    {
+      "epoch": 1.3838636363636363,
+      "grad_norm": 2.0958834741503316,
+      "learning_rate": 1.028320156000928e-06,
+      "loss": 0.0545,
+      "step": 6089
+    },
+    {
+      "epoch": 1.384090909090909,
+      "grad_norm": 2.299217405089398,
+      "learning_rate": 1.0282519722857066e-06,
+      "loss": 0.0498,
+      "step": 6090
+    },
+    {
+      "epoch": 1.3843181818181818,
+      "grad_norm": 1.9740573787684388,
+      "learning_rate": 1.0281837803474739e-06,
+      "loss": 0.0544,
+      "step": 6091
+    },
+    {
+      "epoch": 1.3845454545454545,
+      "grad_norm": 3.6894590622242704,
+      "learning_rate": 1.0281155801876205e-06,
+      "loss": 0.095,
+      "step": 6092
+    },
+    {
+      "epoch": 1.3847727272727273,
+      "grad_norm": 2.333400313451479,
+      "learning_rate": 1.0280473718075374e-06,
+      "loss": 0.0771,
+      "step": 6093
+    },
+    {
+      "epoch": 1.385,
+      "grad_norm": 2.2414971179847507,
+      "learning_rate": 1.0279791552086151e-06,
+      "loss": 0.0903,
+      "step": 6094
+    },
+    {
+      "epoch": 1.3852272727272728,
+      "grad_norm": 1.5939604325166685,
+      "learning_rate": 1.0279109303922446e-06,
+      "loss": 0.0768,
+      "step": 6095
+    },
+    {
+      "epoch": 1.3854545454545455,
+      "grad_norm": 1.1715464655133843,
+      "learning_rate": 1.0278426973598176e-06,
+      "loss": 0.0387,
+      "step": 6096
+    },
+    {
+      "epoch": 1.3856818181818182,
+      "grad_norm": 1.1135274800338246,
+      "learning_rate": 1.027774456112725e-06,
+      "loss": 0.0165,
+      "step": 6097
+    },
+    {
+      "epoch": 1.385909090909091,
+      "grad_norm": 1.5096272509154933,
+      "learning_rate": 1.0277062066523587e-06,
+      "loss": 0.058,
+      "step": 6098
+    },
+    {
+      "epoch": 1.3861363636363637,
+      "grad_norm": 1.1494054278376482,
+      "learning_rate": 1.0276379489801102e-06,
+      "loss": 0.0291,
+      "step": 6099
+    },
+    {
+      "epoch": 1.3863636363636362,
+      "grad_norm": 1.075443397247181,
+      "learning_rate": 1.0275696830973717e-06,
+      "loss": 0.0206,
+      "step": 6100
+    },
+    {
+      "epoch": 1.3865909090909092,
+      "grad_norm": 0.5361956354733862,
+      "learning_rate": 1.0275014090055346e-06,
+      "loss": 0.027,
+      "step": 6101
+    },
+    {
+      "epoch": 1.3868181818181817,
+      "grad_norm": 1.7016557499597316,
+      "learning_rate": 1.027433126705992e-06,
+      "loss": 0.0541,
+      "step": 6102
+    },
+    {
+      "epoch": 1.3870454545454545,
+      "grad_norm": 2.1669306408930598,
+      "learning_rate": 1.0273648362001357e-06,
+      "loss": 0.098,
+      "step": 6103
+    },
+    {
+      "epoch": 1.3872727272727272,
+      "grad_norm": 2.160533865378526,
+      "learning_rate": 1.0272965374893584e-06,
+      "loss": 0.0426,
+      "step": 6104
+    },
+    {
+      "epoch": 1.3875,
+      "grad_norm": 2.1010119539824768,
+      "learning_rate": 1.0272282305750528e-06,
+      "loss": 0.0493,
+      "step": 6105
+    },
+    {
+      "epoch": 1.3877272727272727,
+      "grad_norm": 1.5542342021245366,
+      "learning_rate": 1.0271599154586119e-06,
+      "loss": 0.0415,
+      "step": 6106
+    },
+    {
+      "epoch": 1.3879545454545454,
+      "grad_norm": 1.3344812369883112,
+      "learning_rate": 1.0270915921414287e-06,
+      "loss": 0.0315,
+      "step": 6107
+    },
+    {
+      "epoch": 1.3881818181818182,
+      "grad_norm": 1.8225757302802708,
+      "learning_rate": 1.0270232606248964e-06,
+      "loss": 0.0501,
+      "step": 6108
+    },
+    {
+      "epoch": 1.388409090909091,
+      "grad_norm": 1.3062833053859115,
+      "learning_rate": 1.0269549209104088e-06,
+      "loss": 0.0509,
+      "step": 6109
+    },
+    {
+      "epoch": 1.3886363636363637,
+      "grad_norm": 2.1154733911674684,
+      "learning_rate": 1.0268865729993588e-06,
+      "loss": 0.0905,
+      "step": 6110
+    },
+    {
+      "epoch": 1.3888636363636364,
+      "grad_norm": 3.290197968923886,
+      "learning_rate": 1.0268182168931404e-06,
+      "loss": 0.0517,
+      "step": 6111
+    },
+    {
+      "epoch": 1.3890909090909092,
+      "grad_norm": 2.8064116488335022,
+      "learning_rate": 1.0267498525931478e-06,
+      "loss": 0.0546,
+      "step": 6112
+    },
+    {
+      "epoch": 1.389318181818182,
+      "grad_norm": 2.473460535470631,
+      "learning_rate": 1.0266814801007747e-06,
+      "loss": 0.0531,
+      "step": 6113
+    },
+    {
+      "epoch": 1.3895454545454546,
+      "grad_norm": 1.838605052105167,
+      "learning_rate": 1.0266130994174153e-06,
+      "loss": 0.0393,
+      "step": 6114
+    },
+    {
+      "epoch": 1.3897727272727272,
+      "grad_norm": 1.6337017597771801,
+      "learning_rate": 1.0265447105444643e-06,
+      "loss": 0.0408,
+      "step": 6115
+    },
+    {
+      "epoch": 1.3900000000000001,
+      "grad_norm": 2.965869599073169,
+      "learning_rate": 1.026476313483316e-06,
+      "loss": 0.0874,
+      "step": 6116
+    },
+    {
+      "epoch": 1.3902272727272726,
+      "grad_norm": 1.5772727143571752,
+      "learning_rate": 1.0264079082353654e-06,
+      "loss": 0.0405,
+      "step": 6117
+    },
+    {
+      "epoch": 1.3904545454545454,
+      "grad_norm": 1.853703306761269,
+      "learning_rate": 1.0263394948020071e-06,
+      "loss": 0.0429,
+      "step": 6118
+    },
+    {
+      "epoch": 1.3906818181818181,
+      "grad_norm": 1.937355435683476,
+      "learning_rate": 1.0262710731846365e-06,
+      "loss": 0.0462,
+      "step": 6119
+    },
+    {
+      "epoch": 1.3909090909090909,
+      "grad_norm": 1.3369136065879834,
+      "learning_rate": 1.0262026433846484e-06,
+      "loss": 0.0278,
+      "step": 6120
+    },
+    {
+      "epoch": 1.3911363636363636,
+      "grad_norm": 1.9486566637239195,
+      "learning_rate": 1.0261342054034387e-06,
+      "loss": 0.0621,
+      "step": 6121
+    },
+    {
+      "epoch": 1.3913636363636364,
+      "grad_norm": 2.191137008592484,
+      "learning_rate": 1.0260657592424028e-06,
+      "loss": 0.0352,
+      "step": 6122
+    },
+    {
+      "epoch": 1.391590909090909,
+      "grad_norm": 2.698667860902118,
+      "learning_rate": 1.0259973049029362e-06,
+      "loss": 0.0612,
+      "step": 6123
+    },
+    {
+      "epoch": 1.3918181818181818,
+      "grad_norm": 2.137679760086553,
+      "learning_rate": 1.025928842386435e-06,
+      "loss": 0.0402,
+      "step": 6124
+    },
+    {
+      "epoch": 1.3920454545454546,
+      "grad_norm": 1.3695512818792643,
+      "learning_rate": 1.0258603716942952e-06,
+      "loss": 0.0283,
+      "step": 6125
+    },
+    {
+      "epoch": 1.3922727272727273,
+      "grad_norm": 2.217280173974909,
+      "learning_rate": 1.025791892827913e-06,
+      "loss": 0.0366,
+      "step": 6126
+    },
+    {
+      "epoch": 1.3925,
+      "grad_norm": 1.3043955288837583,
+      "learning_rate": 1.0257234057886854e-06,
+      "loss": 0.0355,
+      "step": 6127
+    },
+    {
+      "epoch": 1.3927272727272726,
+      "grad_norm": 1.8402065390719715,
+      "learning_rate": 1.0256549105780078e-06,
+      "loss": 0.0424,
+      "step": 6128
+    },
+    {
+      "epoch": 1.3929545454545456,
+      "grad_norm": 2.207061169759894,
+      "learning_rate": 1.0255864071972781e-06,
+      "loss": 0.059,
+      "step": 6129
+    },
+    {
+      "epoch": 1.393181818181818,
+      "grad_norm": 2.669438748665288,
+      "learning_rate": 1.0255178956478925e-06,
+      "loss": 0.0679,
+      "step": 6130
+    },
+    {
+      "epoch": 1.393409090909091,
+      "grad_norm": 1.6691340603501479,
+      "learning_rate": 1.0254493759312482e-06,
+      "loss": 0.0477,
+      "step": 6131
+    },
+    {
+      "epoch": 1.3936363636363636,
+      "grad_norm": 2.5605714879615036,
+      "learning_rate": 1.0253808480487425e-06,
+      "loss": 0.0493,
+      "step": 6132
+    },
+    {
+      "epoch": 1.3938636363636363,
+      "grad_norm": 0.4430386073835019,
+      "learning_rate": 1.0253123120017728e-06,
+      "loss": 0.0146,
+      "step": 6133
+    },
+    {
+      "epoch": 1.394090909090909,
+      "grad_norm": 3.1696198510860953,
+      "learning_rate": 1.0252437677917367e-06,
+      "loss": 0.0469,
+      "step": 6134
+    },
+    {
+      "epoch": 1.3943181818181818,
+      "grad_norm": 2.3156015316568634,
+      "learning_rate": 1.025175215420032e-06,
+      "loss": 0.0416,
+      "step": 6135
+    },
+    {
+      "epoch": 1.3945454545454545,
+      "grad_norm": 2.00758114097602,
+      "learning_rate": 1.0251066548880564e-06,
+      "loss": 0.051,
+      "step": 6136
+    },
+    {
+      "epoch": 1.3947727272727273,
+      "grad_norm": 1.674801066752738,
+      "learning_rate": 1.0250380861972081e-06,
+      "loss": 0.0247,
+      "step": 6137
+    },
+    {
+      "epoch": 1.395,
+      "grad_norm": 1.5475608381588666,
+      "learning_rate": 1.0249695093488854e-06,
+      "loss": 0.0519,
+      "step": 6138
+    },
+    {
+      "epoch": 1.3952272727272728,
+      "grad_norm": 1.6845508132309344,
+      "learning_rate": 1.0249009243444865e-06,
+      "loss": 0.0655,
+      "step": 6139
+    },
+    {
+      "epoch": 1.3954545454545455,
+      "grad_norm": 3.8225553749191072,
+      "learning_rate": 1.0248323311854102e-06,
+      "loss": 0.0929,
+      "step": 6140
+    },
+    {
+      "epoch": 1.3956818181818182,
+      "grad_norm": 2.2935446954829435,
+      "learning_rate": 1.024763729873055e-06,
+      "loss": 0.0579,
+      "step": 6141
+    },
+    {
+      "epoch": 1.395909090909091,
+      "grad_norm": 1.2780537891162016,
+      "learning_rate": 1.0246951204088202e-06,
+      "loss": 0.0486,
+      "step": 6142
+    },
+    {
+      "epoch": 1.3961363636363635,
+      "grad_norm": 2.202223995492104,
+      "learning_rate": 1.0246265027941043e-06,
+      "loss": 0.0626,
+      "step": 6143
+    },
+    {
+      "epoch": 1.3963636363636365,
+      "grad_norm": 1.6344748675812861,
+      "learning_rate": 1.0245578770303071e-06,
+      "loss": 0.0538,
+      "step": 6144
+    },
+    {
+      "epoch": 1.396590909090909,
+      "grad_norm": 1.9092718642861908,
+      "learning_rate": 1.0244892431188274e-06,
+      "loss": 0.0342,
+      "step": 6145
+    },
+    {
+      "epoch": 1.3968181818181817,
+      "grad_norm": 1.5986700776576825,
+      "learning_rate": 1.0244206010610653e-06,
+      "loss": 0.0578,
+      "step": 6146
+    },
+    {
+      "epoch": 1.3970454545454545,
+      "grad_norm": 1.855932467580437,
+      "learning_rate": 1.02435195085842e-06,
+      "loss": 0.0442,
+      "step": 6147
+    },
+    {
+      "epoch": 1.3972727272727272,
+      "grad_norm": 2.6044850883041044,
+      "learning_rate": 1.0242832925122921e-06,
+      "loss": 0.0683,
+      "step": 6148
+    },
+    {
+      "epoch": 1.3975,
+      "grad_norm": 1.2373316245572046,
+      "learning_rate": 1.0242146260240811e-06,
+      "loss": 0.0457,
+      "step": 6149
+    },
+    {
+      "epoch": 1.3977272727272727,
+      "grad_norm": 1.5608772384520868,
+      "learning_rate": 1.0241459513951875e-06,
+      "loss": 0.0309,
+      "step": 6150
+    },
+    {
+      "epoch": 1.3979545454545454,
+      "grad_norm": 3.500176576989866,
+      "learning_rate": 1.0240772686270115e-06,
+      "loss": 0.1149,
+      "step": 6151
+    },
+    {
+      "epoch": 1.3981818181818182,
+      "grad_norm": 1.544853173513096,
+      "learning_rate": 1.0240085777209537e-06,
+      "loss": 0.0553,
+      "step": 6152
+    },
+    {
+      "epoch": 1.398409090909091,
+      "grad_norm": 2.152367919295889,
+      "learning_rate": 1.023939878678415e-06,
+      "loss": 0.0671,
+      "step": 6153
+    },
+    {
+      "epoch": 1.3986363636363637,
+      "grad_norm": 2.4108481049502495,
+      "learning_rate": 1.0238711715007961e-06,
+      "loss": 0.1015,
+      "step": 6154
+    },
+    {
+      "epoch": 1.3988636363636364,
+      "grad_norm": 1.8502036026214632,
+      "learning_rate": 1.0238024561894985e-06,
+      "loss": 0.0526,
+      "step": 6155
+    },
+    {
+      "epoch": 1.3990909090909092,
+      "grad_norm": 1.4176175337174548,
+      "learning_rate": 1.0237337327459226e-06,
+      "loss": 0.0375,
+      "step": 6156
+    },
+    {
+      "epoch": 1.399318181818182,
+      "grad_norm": 2.167351956349145,
+      "learning_rate": 1.0236650011714704e-06,
+      "loss": 0.0718,
+      "step": 6157
+    },
+    {
+      "epoch": 1.3995454545454544,
+      "grad_norm": 2.879168123816235,
+      "learning_rate": 1.0235962614675435e-06,
+      "loss": 0.078,
+      "step": 6158
+    },
+    {
+      "epoch": 1.3997727272727274,
+      "grad_norm": 1.7058691470429217,
+      "learning_rate": 1.0235275136355434e-06,
+      "loss": 0.0513,
+      "step": 6159
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.2413573225660826,
+      "learning_rate": 1.0234587576768719e-06,
+      "loss": 0.0374,
+      "step": 6160
+    },
+    {
+      "epoch": 1.4002272727272727,
+      "grad_norm": 1.0935666420959869,
+      "learning_rate": 1.0233899935929311e-06,
+      "loss": 0.0251,
+      "step": 6161
+    },
+    {
+      "epoch": 1.4004545454545454,
+      "grad_norm": 1.2912230475729147,
+      "learning_rate": 1.0233212213851235e-06,
+      "loss": 0.0257,
+      "step": 6162
+    },
+    {
+      "epoch": 1.4006818181818181,
+      "grad_norm": 2.4854250092386825,
+      "learning_rate": 1.023252441054851e-06,
+      "loss": 0.0974,
+      "step": 6163
+    },
+    {
+      "epoch": 1.4009090909090909,
+      "grad_norm": 1.755963526001346,
+      "learning_rate": 1.023183652603517e-06,
+      "loss": 0.0309,
+      "step": 6164
+    },
+    {
+      "epoch": 1.4011363636363636,
+      "grad_norm": 1.57332895229544,
+      "learning_rate": 1.0231148560325232e-06,
+      "loss": 0.0642,
+      "step": 6165
+    },
+    {
+      "epoch": 1.4013636363636364,
+      "grad_norm": 2.3957896418537548,
+      "learning_rate": 1.023046051343273e-06,
+      "loss": 0.0498,
+      "step": 6166
+    },
+    {
+      "epoch": 1.401590909090909,
+      "grad_norm": 2.4679944773329368,
+      "learning_rate": 1.0229772385371695e-06,
+      "loss": 0.0919,
+      "step": 6167
+    },
+    {
+      "epoch": 1.4018181818181819,
+      "grad_norm": 1.8087308260106476,
+      "learning_rate": 1.0229084176156156e-06,
+      "loss": 0.0424,
+      "step": 6168
+    },
+    {
+      "epoch": 1.4020454545454546,
+      "grad_norm": 0.7245737251515612,
+      "learning_rate": 1.022839588580015e-06,
+      "loss": 0.0135,
+      "step": 6169
+    },
+    {
+      "epoch": 1.4022727272727273,
+      "grad_norm": 2.340886719436232,
+      "learning_rate": 1.022770751431771e-06,
+      "loss": 0.0682,
+      "step": 6170
+    },
+    {
+      "epoch": 1.4025,
+      "grad_norm": 2.292171571798208,
+      "learning_rate": 1.0227019061722877e-06,
+      "loss": 0.0739,
+      "step": 6171
+    },
+    {
+      "epoch": 1.4027272727272728,
+      "grad_norm": 2.1651821353761718,
+      "learning_rate": 1.0226330528029684e-06,
+      "loss": 0.0446,
+      "step": 6172
+    },
+    {
+      "epoch": 1.4029545454545453,
+      "grad_norm": 1.5702818515145056,
+      "learning_rate": 1.0225641913252176e-06,
+      "loss": 0.0459,
+      "step": 6173
+    },
+    {
+      "epoch": 1.4031818181818183,
+      "grad_norm": 1.1509846243402606,
+      "learning_rate": 1.0224953217404394e-06,
+      "loss": 0.0276,
+      "step": 6174
+    },
+    {
+      "epoch": 1.4034090909090908,
+      "grad_norm": 2.113019993274709,
+      "learning_rate": 1.0224264440500382e-06,
+      "loss": 0.0595,
+      "step": 6175
+    },
+    {
+      "epoch": 1.4036363636363636,
+      "grad_norm": 0.9751984740381435,
+      "learning_rate": 1.022357558255418e-06,
+      "loss": 0.0388,
+      "step": 6176
+    },
+    {
+      "epoch": 1.4038636363636363,
+      "grad_norm": 2.263245935446486,
+      "learning_rate": 1.0222886643579844e-06,
+      "loss": 0.0422,
+      "step": 6177
+    },
+    {
+      "epoch": 1.404090909090909,
+      "grad_norm": 2.3447223599657927,
+      "learning_rate": 1.0222197623591418e-06,
+      "loss": 0.0383,
+      "step": 6178
+    },
+    {
+      "epoch": 1.4043181818181818,
+      "grad_norm": 1.9665887429048219,
+      "learning_rate": 1.022150852260295e-06,
+      "loss": 0.0788,
+      "step": 6179
+    },
+    {
+      "epoch": 1.4045454545454545,
+      "grad_norm": 1.8275708584158676,
+      "learning_rate": 1.0220819340628497e-06,
+      "loss": 0.0756,
+      "step": 6180
+    },
+    {
+      "epoch": 1.4047727272727273,
+      "grad_norm": 1.9564688709403895,
+      "learning_rate": 1.022013007768211e-06,
+      "loss": 0.0437,
+      "step": 6181
+    },
+    {
+      "epoch": 1.405,
+      "grad_norm": 1.4307666892547968,
+      "learning_rate": 1.0219440733777843e-06,
+      "loss": 0.0263,
+      "step": 6182
+    },
+    {
+      "epoch": 1.4052272727272728,
+      "grad_norm": 1.128973763588542,
+      "learning_rate": 1.0218751308929755e-06,
+      "loss": 0.0244,
+      "step": 6183
+    },
+    {
+      "epoch": 1.4054545454545455,
+      "grad_norm": 1.698985504043439,
+      "learning_rate": 1.0218061803151906e-06,
+      "loss": 0.0402,
+      "step": 6184
+    },
+    {
+      "epoch": 1.4056818181818183,
+      "grad_norm": 1.430665093207686,
+      "learning_rate": 1.0217372216458348e-06,
+      "loss": 0.0273,
+      "step": 6185
+    },
+    {
+      "epoch": 1.4059090909090908,
+      "grad_norm": 2.3756220521317757,
+      "learning_rate": 1.0216682548863155e-06,
+      "loss": 0.0853,
+      "step": 6186
+    },
+    {
+      "epoch": 1.4061363636363637,
+      "grad_norm": 3.159920128367306,
+      "learning_rate": 1.0215992800380384e-06,
+      "loss": 0.0815,
+      "step": 6187
+    },
+    {
+      "epoch": 1.4063636363636363,
+      "grad_norm": 1.997184745635932,
+      "learning_rate": 1.0215302971024099e-06,
+      "loss": 0.0546,
+      "step": 6188
+    },
+    {
+      "epoch": 1.4065909090909092,
+      "grad_norm": 1.6920784647496654,
+      "learning_rate": 1.021461306080837e-06,
+      "loss": 0.0254,
+      "step": 6189
+    },
+    {
+      "epoch": 1.4068181818181817,
+      "grad_norm": 1.8417236848042329,
+      "learning_rate": 1.021392306974726e-06,
+      "loss": 0.0646,
+      "step": 6190
+    },
+    {
+      "epoch": 1.4070454545454545,
+      "grad_norm": 2.0136652439295384,
+      "learning_rate": 1.0213232997854848e-06,
+      "loss": 0.0372,
+      "step": 6191
+    },
+    {
+      "epoch": 1.4072727272727272,
+      "grad_norm": 2.253351359296166,
+      "learning_rate": 1.02125428451452e-06,
+      "loss": 0.0711,
+      "step": 6192
+    },
+    {
+      "epoch": 1.4075,
+      "grad_norm": 2.20712092412831,
+      "learning_rate": 1.021185261163239e-06,
+      "loss": 0.0404,
+      "step": 6193
+    },
+    {
+      "epoch": 1.4077272727272727,
+      "grad_norm": 2.182650844568821,
+      "learning_rate": 1.0211162297330494e-06,
+      "loss": 0.0522,
+      "step": 6194
+    },
+    {
+      "epoch": 1.4079545454545455,
+      "grad_norm": 1.9596632561426677,
+      "learning_rate": 1.0210471902253585e-06,
+      "loss": 0.07,
+      "step": 6195
+    },
+    {
+      "epoch": 1.4081818181818182,
+      "grad_norm": 1.7191018567894356,
+      "learning_rate": 1.0209781426415747e-06,
+      "loss": 0.0321,
+      "step": 6196
+    },
+    {
+      "epoch": 1.408409090909091,
+      "grad_norm": 4.939484112747691,
+      "learning_rate": 1.0209090869831058e-06,
+      "loss": 0.096,
+      "step": 6197
+    },
+    {
+      "epoch": 1.4086363636363637,
+      "grad_norm": 1.58319253718961,
+      "learning_rate": 1.0208400232513598e-06,
+      "loss": 0.0667,
+      "step": 6198
+    },
+    {
+      "epoch": 1.4088636363636364,
+      "grad_norm": 3.146351482645228,
+      "learning_rate": 1.0207709514477452e-06,
+      "loss": 0.1502,
+      "step": 6199
+    },
+    {
+      "epoch": 1.4090909090909092,
+      "grad_norm": 2.7417911865708273,
+      "learning_rate": 1.0207018715736704e-06,
+      "loss": 0.0605,
+      "step": 6200
+    },
+    {
+      "epoch": 1.4093181818181817,
+      "grad_norm": 1.5558288960911015,
+      "learning_rate": 1.020632783630544e-06,
+      "loss": 0.0301,
+      "step": 6201
+    },
+    {
+      "epoch": 1.4095454545454547,
+      "grad_norm": 2.1624674804610424,
+      "learning_rate": 1.0205636876197752e-06,
+      "loss": 0.0973,
+      "step": 6202
+    },
+    {
+      "epoch": 1.4097727272727272,
+      "grad_norm": 1.9679174554890828,
+      "learning_rate": 1.0204945835427725e-06,
+      "loss": 0.046,
+      "step": 6203
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.2021636547842176,
+      "learning_rate": 1.0204254714009453e-06,
+      "loss": 0.0855,
+      "step": 6204
+    },
+    {
+      "epoch": 1.4102272727272727,
+      "grad_norm": 2.9885640290882916,
+      "learning_rate": 1.020356351195703e-06,
+      "loss": 0.0749,
+      "step": 6205
+    },
+    {
+      "epoch": 1.4104545454545454,
+      "grad_norm": 1.767071835041449,
+      "learning_rate": 1.0202872229284546e-06,
+      "loss": 0.054,
+      "step": 6206
+    },
+    {
+      "epoch": 1.4106818181818181,
+      "grad_norm": 2.638750106896759,
+      "learning_rate": 1.0202180866006105e-06,
+      "loss": 0.0874,
+      "step": 6207
+    },
+    {
+      "epoch": 1.410909090909091,
+      "grad_norm": 1.928977191423767,
+      "learning_rate": 1.0201489422135798e-06,
+      "loss": 0.0629,
+      "step": 6208
+    },
+    {
+      "epoch": 1.4111363636363636,
+      "grad_norm": 1.1489762194521922,
+      "learning_rate": 1.020079789768773e-06,
+      "loss": 0.0262,
+      "step": 6209
+    },
+    {
+      "epoch": 1.4113636363636364,
+      "grad_norm": 1.0705599188362818,
+      "learning_rate": 1.0200106292676e-06,
+      "loss": 0.0359,
+      "step": 6210
+    },
+    {
+      "epoch": 1.4115909090909091,
+      "grad_norm": 1.4944103933008261,
+      "learning_rate": 1.0199414607114713e-06,
+      "loss": 0.0423,
+      "step": 6211
+    },
+    {
+      "epoch": 1.4118181818181819,
+      "grad_norm": 1.6617957785692394,
+      "learning_rate": 1.019872284101797e-06,
+      "loss": 0.0428,
+      "step": 6212
+    },
+    {
+      "epoch": 1.4120454545454546,
+      "grad_norm": 1.1556713147150746,
+      "learning_rate": 1.019803099439988e-06,
+      "loss": 0.0165,
+      "step": 6213
+    },
+    {
+      "epoch": 1.4122727272727273,
+      "grad_norm": 2.6024306931985164,
+      "learning_rate": 1.019733906727455e-06,
+      "loss": 0.0847,
+      "step": 6214
+    },
+    {
+      "epoch": 1.4125,
+      "grad_norm": 1.7825669229940564,
+      "learning_rate": 1.019664705965609e-06,
+      "loss": 0.0546,
+      "step": 6215
+    },
+    {
+      "epoch": 1.4127272727272726,
+      "grad_norm": 1.537815171357164,
+      "learning_rate": 1.0195954971558613e-06,
+      "loss": 0.0249,
+      "step": 6216
+    },
+    {
+      "epoch": 1.4129545454545456,
+      "grad_norm": 1.1789172537417076,
+      "learning_rate": 1.0195262802996226e-06,
+      "loss": 0.035,
+      "step": 6217
+    },
+    {
+      "epoch": 1.413181818181818,
+      "grad_norm": 2.730586746583894,
+      "learning_rate": 1.0194570553983049e-06,
+      "loss": 0.0788,
+      "step": 6218
+    },
+    {
+      "epoch": 1.4134090909090908,
+      "grad_norm": 1.1952770239369728,
+      "learning_rate": 1.01938782245332e-06,
+      "loss": 0.0389,
+      "step": 6219
+    },
+    {
+      "epoch": 1.4136363636363636,
+      "grad_norm": 1.674275165326814,
+      "learning_rate": 1.019318581466079e-06,
+      "loss": 0.0335,
+      "step": 6220
+    },
+    {
+      "epoch": 1.4138636363636363,
+      "grad_norm": 2.0166205707653866,
+      "learning_rate": 1.0192493324379943e-06,
+      "loss": 0.0597,
+      "step": 6221
+    },
+    {
+      "epoch": 1.414090909090909,
+      "grad_norm": 2.406642790204674,
+      "learning_rate": 1.0191800753704778e-06,
+      "loss": 0.0815,
+      "step": 6222
+    },
+    {
+      "epoch": 1.4143181818181818,
+      "grad_norm": 2.515181328667471,
+      "learning_rate": 1.0191108102649421e-06,
+      "loss": 0.0475,
+      "step": 6223
+    },
+    {
+      "epoch": 1.4145454545454546,
+      "grad_norm": 1.628077934871075,
+      "learning_rate": 1.0190415371227994e-06,
+      "loss": 0.03,
+      "step": 6224
+    },
+    {
+      "epoch": 1.4147727272727273,
+      "grad_norm": 2.6526265588190228,
+      "learning_rate": 1.0189722559454621e-06,
+      "loss": 0.0399,
+      "step": 6225
+    },
+    {
+      "epoch": 1.415,
+      "grad_norm": 2.5670775424439634,
+      "learning_rate": 1.0189029667343433e-06,
+      "loss": 0.06,
+      "step": 6226
+    },
+    {
+      "epoch": 1.4152272727272728,
+      "grad_norm": 3.278098576240127,
+      "learning_rate": 1.018833669490856e-06,
+      "loss": 0.1036,
+      "step": 6227
+    },
+    {
+      "epoch": 1.4154545454545455,
+      "grad_norm": 2.4698810915304636,
+      "learning_rate": 1.0187643642164127e-06,
+      "loss": 0.0316,
+      "step": 6228
+    },
+    {
+      "epoch": 1.4156818181818183,
+      "grad_norm": 1.5698858276064769,
+      "learning_rate": 1.0186950509124272e-06,
+      "loss": 0.0495,
+      "step": 6229
+    },
+    {
+      "epoch": 1.415909090909091,
+      "grad_norm": 2.0523482036442933,
+      "learning_rate": 1.0186257295803128e-06,
+      "loss": 0.0475,
+      "step": 6230
+    },
+    {
+      "epoch": 1.4161363636363635,
+      "grad_norm": 2.0749845910369373,
+      "learning_rate": 1.0185564002214831e-06,
+      "loss": 0.0755,
+      "step": 6231
+    },
+    {
+      "epoch": 1.4163636363636365,
+      "grad_norm": 1.384240995248522,
+      "learning_rate": 1.0184870628373517e-06,
+      "loss": 0.027,
+      "step": 6232
+    },
+    {
+      "epoch": 1.416590909090909,
+      "grad_norm": 2.882248261374469,
+      "learning_rate": 1.0184177174293327e-06,
+      "loss": 0.077,
+      "step": 6233
+    },
+    {
+      "epoch": 1.4168181818181818,
+      "grad_norm": 2.224208410543184,
+      "learning_rate": 1.0183483639988402e-06,
+      "loss": 0.0665,
+      "step": 6234
+    },
+    {
+      "epoch": 1.4170454545454545,
+      "grad_norm": 1.2285481339794369,
+      "learning_rate": 1.0182790025472882e-06,
+      "loss": 0.0244,
+      "step": 6235
+    },
+    {
+      "epoch": 1.4172727272727272,
+      "grad_norm": 3.1979562148679856,
+      "learning_rate": 1.0182096330760912e-06,
+      "loss": 0.1269,
+      "step": 6236
+    },
+    {
+      "epoch": 1.4175,
+      "grad_norm": 3.381358918110057,
+      "learning_rate": 1.0181402555866638e-06,
+      "loss": 0.0651,
+      "step": 6237
+    },
+    {
+      "epoch": 1.4177272727272727,
+      "grad_norm": 1.3335449978514418,
+      "learning_rate": 1.018070870080421e-06,
+      "loss": 0.0484,
+      "step": 6238
+    },
+    {
+      "epoch": 1.4179545454545455,
+      "grad_norm": 1.194884624546463,
+      "learning_rate": 1.018001476558777e-06,
+      "loss": 0.0253,
+      "step": 6239
+    },
+    {
+      "epoch": 1.4181818181818182,
+      "grad_norm": 2.2071849335337945,
+      "learning_rate": 1.0179320750231473e-06,
+      "loss": 0.0613,
+      "step": 6240
+    },
+    {
+      "epoch": 1.418409090909091,
+      "grad_norm": 2.0970275179064792,
+      "learning_rate": 1.0178626654749473e-06,
+      "loss": 0.0682,
+      "step": 6241
+    },
+    {
+      "epoch": 1.4186363636363637,
+      "grad_norm": 1.3100116266039097,
+      "learning_rate": 1.017793247915592e-06,
+      "loss": 0.0242,
+      "step": 6242
+    },
+    {
+      "epoch": 1.4188636363636364,
+      "grad_norm": 1.7972153655632137,
+      "learning_rate": 1.0177238223464973e-06,
+      "loss": 0.0499,
+      "step": 6243
+    },
+    {
+      "epoch": 1.419090909090909,
+      "grad_norm": 2.2761143579437415,
+      "learning_rate": 1.0176543887690785e-06,
+      "loss": 0.035,
+      "step": 6244
+    },
+    {
+      "epoch": 1.419318181818182,
+      "grad_norm": 1.5274887413411984,
+      "learning_rate": 1.0175849471847518e-06,
+      "loss": 0.0289,
+      "step": 6245
+    },
+    {
+      "epoch": 1.4195454545454544,
+      "grad_norm": 2.408305600556025,
+      "learning_rate": 1.017515497594933e-06,
+      "loss": 0.0464,
+      "step": 6246
+    },
+    {
+      "epoch": 1.4197727272727274,
+      "grad_norm": 1.0849649061589346,
+      "learning_rate": 1.0174460400010386e-06,
+      "loss": 0.0241,
+      "step": 6247
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.9978161803679924,
+      "learning_rate": 1.0173765744044846e-06,
+      "loss": 0.0526,
+      "step": 6248
+    },
+    {
+      "epoch": 1.4202272727272727,
+      "grad_norm": 0.8219377032127845,
+      "learning_rate": 1.0173071008066878e-06,
+      "loss": 0.019,
+      "step": 6249
+    },
+    {
+      "epoch": 1.4204545454545454,
+      "grad_norm": 1.4515367432995625,
+      "learning_rate": 1.0172376192090648e-06,
+      "loss": 0.0303,
+      "step": 6250
+    },
+    {
+      "epoch": 1.4206818181818182,
+      "grad_norm": 2.511070878734959,
+      "learning_rate": 1.0171681296130326e-06,
+      "loss": 0.111,
+      "step": 6251
+    },
+    {
+      "epoch": 1.420909090909091,
+      "grad_norm": 2.044081682654695,
+      "learning_rate": 1.017098632020008e-06,
+      "loss": 0.0927,
+      "step": 6252
+    },
+    {
+      "epoch": 1.4211363636363636,
+      "grad_norm": 1.3412493834132027,
+      "learning_rate": 1.017029126431408e-06,
+      "loss": 0.0465,
+      "step": 6253
+    },
+    {
+      "epoch": 1.4213636363636364,
+      "grad_norm": 2.1342958224551696,
+      "learning_rate": 1.0169596128486503e-06,
+      "loss": 0.043,
+      "step": 6254
+    },
+    {
+      "epoch": 1.4215909090909091,
+      "grad_norm": 0.9957308239944688,
+      "learning_rate": 1.0168900912731526e-06,
+      "loss": 0.0261,
+      "step": 6255
+    },
+    {
+      "epoch": 1.4218181818181819,
+      "grad_norm": 1.4875746647869024,
+      "learning_rate": 1.0168205617063318e-06,
+      "loss": 0.0264,
+      "step": 6256
+    },
+    {
+      "epoch": 1.4220454545454546,
+      "grad_norm": 1.242642069068493,
+      "learning_rate": 1.0167510241496065e-06,
+      "loss": 0.0285,
+      "step": 6257
+    },
+    {
+      "epoch": 1.4222727272727274,
+      "grad_norm": 1.8316659073218056,
+      "learning_rate": 1.016681478604394e-06,
+      "loss": 0.0374,
+      "step": 6258
+    },
+    {
+      "epoch": 1.4224999999999999,
+      "grad_norm": 2.2362950746465393,
+      "learning_rate": 1.0166119250721133e-06,
+      "loss": 0.0421,
+      "step": 6259
+    },
+    {
+      "epoch": 1.4227272727272728,
+      "grad_norm": 2.8565600838638776,
+      "learning_rate": 1.0165423635541819e-06,
+      "loss": 0.1086,
+      "step": 6260
+    },
+    {
+      "epoch": 1.4229545454545454,
+      "grad_norm": 2.288699097340909,
+      "learning_rate": 1.0164727940520185e-06,
+      "loss": 0.0694,
+      "step": 6261
+    },
+    {
+      "epoch": 1.423181818181818,
+      "grad_norm": 1.445081886674453,
+      "learning_rate": 1.0164032165670423e-06,
+      "loss": 0.0362,
+      "step": 6262
+    },
+    {
+      "epoch": 1.4234090909090908,
+      "grad_norm": 2.0017693912469534,
+      "learning_rate": 1.0163336311006714e-06,
+      "loss": 0.0613,
+      "step": 6263
+    },
+    {
+      "epoch": 1.4236363636363636,
+      "grad_norm": 1.600735229717396,
+      "learning_rate": 1.0162640376543251e-06,
+      "loss": 0.0436,
+      "step": 6264
+    },
+    {
+      "epoch": 1.4238636363636363,
+      "grad_norm": 1.593316998726824,
+      "learning_rate": 1.0161944362294225e-06,
+      "loss": 0.0201,
+      "step": 6265
+    },
+    {
+      "epoch": 1.424090909090909,
+      "grad_norm": 2.1284007400803495,
+      "learning_rate": 1.016124826827383e-06,
+      "loss": 0.0634,
+      "step": 6266
+    },
+    {
+      "epoch": 1.4243181818181818,
+      "grad_norm": 2.075051864065432,
+      "learning_rate": 1.0160552094496256e-06,
+      "loss": 0.0482,
+      "step": 6267
+    },
+    {
+      "epoch": 1.4245454545454546,
+      "grad_norm": 1.1832179067036461,
+      "learning_rate": 1.0159855840975705e-06,
+      "loss": 0.0288,
+      "step": 6268
+    },
+    {
+      "epoch": 1.4247727272727273,
+      "grad_norm": 2.360764922145612,
+      "learning_rate": 1.0159159507726371e-06,
+      "loss": 0.0503,
+      "step": 6269
+    },
+    {
+      "epoch": 1.425,
+      "grad_norm": 1.1177907530668978,
+      "learning_rate": 1.0158463094762456e-06,
+      "loss": 0.0335,
+      "step": 6270
+    },
+    {
+      "epoch": 1.4252272727272728,
+      "grad_norm": 2.4079973603552496,
+      "learning_rate": 1.015776660209816e-06,
+      "loss": 0.0558,
+      "step": 6271
+    },
+    {
+      "epoch": 1.4254545454545455,
+      "grad_norm": 1.3144050986858768,
+      "learning_rate": 1.0157070029747683e-06,
+      "loss": 0.0299,
+      "step": 6272
+    },
+    {
+      "epoch": 1.4256818181818183,
+      "grad_norm": 2.6987177705198837,
+      "learning_rate": 1.0156373377725234e-06,
+      "loss": 0.0789,
+      "step": 6273
+    },
+    {
+      "epoch": 1.4259090909090908,
+      "grad_norm": 1.5949162602238702,
+      "learning_rate": 1.0155676646045015e-06,
+      "loss": 0.0405,
+      "step": 6274
+    },
+    {
+      "epoch": 1.4261363636363638,
+      "grad_norm": 1.4639703694283597,
+      "learning_rate": 1.0154979834721235e-06,
+      "loss": 0.0522,
+      "step": 6275
+    },
+    {
+      "epoch": 1.4263636363636363,
+      "grad_norm": 2.517068802401391,
+      "learning_rate": 1.0154282943768106e-06,
+      "loss": 0.0619,
+      "step": 6276
+    },
+    {
+      "epoch": 1.426590909090909,
+      "grad_norm": 1.9493460635048738,
+      "learning_rate": 1.0153585973199837e-06,
+      "loss": 0.0626,
+      "step": 6277
+    },
+    {
+      "epoch": 1.4268181818181818,
+      "grad_norm": 1.3958052416380633,
+      "learning_rate": 1.0152888923030637e-06,
+      "loss": 0.0393,
+      "step": 6278
+    },
+    {
+      "epoch": 1.4270454545454545,
+      "grad_norm": 2.0319350661938542,
+      "learning_rate": 1.0152191793274723e-06,
+      "loss": 0.0314,
+      "step": 6279
+    },
+    {
+      "epoch": 1.4272727272727272,
+      "grad_norm": 2.734387470350427,
+      "learning_rate": 1.0151494583946315e-06,
+      "loss": 0.0641,
+      "step": 6280
+    },
+    {
+      "epoch": 1.4275,
+      "grad_norm": 2.3055348208111113,
+      "learning_rate": 1.0150797295059623e-06,
+      "loss": 0.0535,
+      "step": 6281
+    },
+    {
+      "epoch": 1.4277272727272727,
+      "grad_norm": 2.574427521123117,
+      "learning_rate": 1.0150099926628868e-06,
+      "loss": 0.0498,
+      "step": 6282
+    },
+    {
+      "epoch": 1.4279545454545455,
+      "grad_norm": 2.107305181996583,
+      "learning_rate": 1.014940247866827e-06,
+      "loss": 0.0396,
+      "step": 6283
+    },
+    {
+      "epoch": 1.4281818181818182,
+      "grad_norm": 1.6100985910897674,
+      "learning_rate": 1.0148704951192055e-06,
+      "loss": 0.0435,
+      "step": 6284
+    },
+    {
+      "epoch": 1.428409090909091,
+      "grad_norm": 3.8541259465940034,
+      "learning_rate": 1.0148007344214445e-06,
+      "loss": 0.0912,
+      "step": 6285
+    },
+    {
+      "epoch": 1.4286363636363637,
+      "grad_norm": 1.655223187396699,
+      "learning_rate": 1.0147309657749663e-06,
+      "loss": 0.0731,
+      "step": 6286
+    },
+    {
+      "epoch": 1.4288636363636364,
+      "grad_norm": 2.088528568754076,
+      "learning_rate": 1.0146611891811938e-06,
+      "loss": 0.0356,
+      "step": 6287
+    },
+    {
+      "epoch": 1.4290909090909092,
+      "grad_norm": 1.8977024354717393,
+      "learning_rate": 1.0145914046415499e-06,
+      "loss": 0.0497,
+      "step": 6288
+    },
+    {
+      "epoch": 1.4293181818181817,
+      "grad_norm": 1.510543981545033,
+      "learning_rate": 1.0145216121574575e-06,
+      "loss": 0.0451,
+      "step": 6289
+    },
+    {
+      "epoch": 1.4295454545454547,
+      "grad_norm": 1.8895150219878092,
+      "learning_rate": 1.0144518117303402e-06,
+      "loss": 0.0437,
+      "step": 6290
+    },
+    {
+      "epoch": 1.4297727272727272,
+      "grad_norm": 2.1759751216122005,
+      "learning_rate": 1.0143820033616206e-06,
+      "loss": 0.0629,
+      "step": 6291
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.7296396614160543,
+      "learning_rate": 1.0143121870527227e-06,
+      "loss": 0.052,
+      "step": 6292
+    },
+    {
+      "epoch": 1.4302272727272727,
+      "grad_norm": 1.416731814986604,
+      "learning_rate": 1.0142423628050704e-06,
+      "loss": 0.0342,
+      "step": 6293
+    },
+    {
+      "epoch": 1.4304545454545454,
+      "grad_norm": 2.1847972328743723,
+      "learning_rate": 1.014172530620087e-06,
+      "loss": 0.0572,
+      "step": 6294
+    },
+    {
+      "epoch": 1.4306818181818182,
+      "grad_norm": 2.984150072755174,
+      "learning_rate": 1.014102690499197e-06,
+      "loss": 0.0521,
+      "step": 6295
+    },
+    {
+      "epoch": 1.430909090909091,
+      "grad_norm": 1.111696038952851,
+      "learning_rate": 1.0140328424438242e-06,
+      "loss": 0.0379,
+      "step": 6296
+    },
+    {
+      "epoch": 1.4311363636363637,
+      "grad_norm": 3.153415606735708,
+      "learning_rate": 1.013962986455393e-06,
+      "loss": 0.0601,
+      "step": 6297
+    },
+    {
+      "epoch": 1.4313636363636364,
+      "grad_norm": 1.384163906831715,
+      "learning_rate": 1.0138931225353278e-06,
+      "loss": 0.0493,
+      "step": 6298
+    },
+    {
+      "epoch": 1.4315909090909091,
+      "grad_norm": 2.527321386269838,
+      "learning_rate": 1.0138232506850533e-06,
+      "loss": 0.0393,
+      "step": 6299
+    },
+    {
+      "epoch": 1.4318181818181819,
+      "grad_norm": 1.7158440564296347,
+      "learning_rate": 1.0137533709059947e-06,
+      "loss": 0.05,
+      "step": 6300
+    },
+    {
+      "epoch": 1.4320454545454546,
+      "grad_norm": 1.9382856176506886,
+      "learning_rate": 1.0136834831995765e-06,
+      "loss": 0.0354,
+      "step": 6301
+    },
+    {
+      "epoch": 1.4322727272727271,
+      "grad_norm": 1.8539117852909803,
+      "learning_rate": 1.013613587567224e-06,
+      "loss": 0.0608,
+      "step": 6302
+    },
+    {
+      "epoch": 1.4325,
+      "grad_norm": 2.1173995133244894,
+      "learning_rate": 1.0135436840103624e-06,
+      "loss": 0.0454,
+      "step": 6303
+    },
+    {
+      "epoch": 1.4327272727272726,
+      "grad_norm": 0.5418750712115254,
+      "learning_rate": 1.0134737725304174e-06,
+      "loss": 0.0226,
+      "step": 6304
+    },
+    {
+      "epoch": 1.4329545454545456,
+      "grad_norm": 2.0216804880691255,
+      "learning_rate": 1.0134038531288142e-06,
+      "loss": 0.067,
+      "step": 6305
+    },
+    {
+      "epoch": 1.4331818181818181,
+      "grad_norm": 2.048861875538986,
+      "learning_rate": 1.0133339258069791e-06,
+      "loss": 0.0555,
+      "step": 6306
+    },
+    {
+      "epoch": 1.4334090909090909,
+      "grad_norm": 2.406028217214649,
+      "learning_rate": 1.0132639905663376e-06,
+      "loss": 0.0683,
+      "step": 6307
+    },
+    {
+      "epoch": 1.4336363636363636,
+      "grad_norm": 2.1359501577118594,
+      "learning_rate": 1.013194047408316e-06,
+      "loss": 0.0632,
+      "step": 6308
+    },
+    {
+      "epoch": 1.4338636363636363,
+      "grad_norm": 1.211567038689289,
+      "learning_rate": 1.0131240963343405e-06,
+      "loss": 0.0434,
+      "step": 6309
+    },
+    {
+      "epoch": 1.434090909090909,
+      "grad_norm": 0.8802892411789037,
+      "learning_rate": 1.0130541373458377e-06,
+      "loss": 0.0227,
+      "step": 6310
+    },
+    {
+      "epoch": 1.4343181818181818,
+      "grad_norm": 3.918131565120923,
+      "learning_rate": 1.0129841704442339e-06,
+      "loss": 0.1199,
+      "step": 6311
+    },
+    {
+      "epoch": 1.4345454545454546,
+      "grad_norm": 2.297944940258336,
+      "learning_rate": 1.012914195630956e-06,
+      "loss": 0.0737,
+      "step": 6312
+    },
+    {
+      "epoch": 1.4347727272727273,
+      "grad_norm": 2.1829452062688612,
+      "learning_rate": 1.012844212907431e-06,
+      "loss": 0.0718,
+      "step": 6313
+    },
+    {
+      "epoch": 1.435,
+      "grad_norm": 2.0698608489487675,
+      "learning_rate": 1.0127742222750858e-06,
+      "loss": 0.0726,
+      "step": 6314
+    },
+    {
+      "epoch": 1.4352272727272728,
+      "grad_norm": 1.9709660987547581,
+      "learning_rate": 1.0127042237353477e-06,
+      "loss": 0.0551,
+      "step": 6315
+    },
+    {
+      "epoch": 1.4354545454545455,
+      "grad_norm": 1.1285441092853792,
+      "learning_rate": 1.012634217289644e-06,
+      "loss": 0.024,
+      "step": 6316
+    },
+    {
+      "epoch": 1.435681818181818,
+      "grad_norm": 2.3917449926449135,
+      "learning_rate": 1.0125642029394025e-06,
+      "loss": 0.0686,
+      "step": 6317
+    },
+    {
+      "epoch": 1.435909090909091,
+      "grad_norm": 1.382793472239119,
+      "learning_rate": 1.0124941806860506e-06,
+      "loss": 0.0281,
+      "step": 6318
+    },
+    {
+      "epoch": 1.4361363636363635,
+      "grad_norm": 2.4077667172502335,
+      "learning_rate": 1.0124241505310164e-06,
+      "loss": 0.1088,
+      "step": 6319
+    },
+    {
+      "epoch": 1.4363636363636363,
+      "grad_norm": 2.2402806871599594,
+      "learning_rate": 1.0123541124757278e-06,
+      "loss": 0.0478,
+      "step": 6320
+    },
+    {
+      "epoch": 1.436590909090909,
+      "grad_norm": 1.748334089361085,
+      "learning_rate": 1.0122840665216133e-06,
+      "loss": 0.0608,
+      "step": 6321
+    },
+    {
+      "epoch": 1.4368181818181818,
+      "grad_norm": 1.3710244033949548,
+      "learning_rate": 1.012214012670101e-06,
+      "loss": 0.0307,
+      "step": 6322
+    },
+    {
+      "epoch": 1.4370454545454545,
+      "grad_norm": 1.0878918832979907,
+      "learning_rate": 1.0121439509226192e-06,
+      "loss": 0.0169,
+      "step": 6323
+    },
+    {
+      "epoch": 1.4372727272727273,
+      "grad_norm": 2.2269205062919037,
+      "learning_rate": 1.0120738812805968e-06,
+      "loss": 0.0723,
+      "step": 6324
+    },
+    {
+      "epoch": 1.4375,
+      "grad_norm": 1.8511981232630557,
+      "learning_rate": 1.012003803745463e-06,
+      "loss": 0.0311,
+      "step": 6325
+    },
+    {
+      "epoch": 1.4377272727272727,
+      "grad_norm": 1.3617235013159903,
+      "learning_rate": 1.0119337183186465e-06,
+      "loss": 0.0463,
+      "step": 6326
+    },
+    {
+      "epoch": 1.4379545454545455,
+      "grad_norm": 1.3653824896372255,
+      "learning_rate": 1.0118636250015762e-06,
+      "loss": 0.0221,
+      "step": 6327
+    },
+    {
+      "epoch": 1.4381818181818182,
+      "grad_norm": 1.5844355911542374,
+      "learning_rate": 1.0117935237956817e-06,
+      "loss": 0.0406,
+      "step": 6328
+    },
+    {
+      "epoch": 1.438409090909091,
+      "grad_norm": 1.846258874453668,
+      "learning_rate": 1.0117234147023923e-06,
+      "loss": 0.0734,
+      "step": 6329
+    },
+    {
+      "epoch": 1.4386363636363637,
+      "grad_norm": 1.9697906324583152,
+      "learning_rate": 1.011653297723138e-06,
+      "loss": 0.0425,
+      "step": 6330
+    },
+    {
+      "epoch": 1.4388636363636365,
+      "grad_norm": 1.4127352173867909,
+      "learning_rate": 1.0115831728593483e-06,
+      "loss": 0.0495,
+      "step": 6331
+    },
+    {
+      "epoch": 1.439090909090909,
+      "grad_norm": 1.4082495500831291,
+      "learning_rate": 1.0115130401124534e-06,
+      "loss": 0.0353,
+      "step": 6332
+    },
+    {
+      "epoch": 1.439318181818182,
+      "grad_norm": 2.9904259416779913,
+      "learning_rate": 1.0114428994838835e-06,
+      "loss": 0.0773,
+      "step": 6333
+    },
+    {
+      "epoch": 1.4395454545454545,
+      "grad_norm": 1.3285289213699503,
+      "learning_rate": 1.0113727509750683e-06,
+      "loss": 0.0301,
+      "step": 6334
+    },
+    {
+      "epoch": 1.4397727272727272,
+      "grad_norm": 4.040021486045353,
+      "learning_rate": 1.0113025945874388e-06,
+      "loss": 0.1269,
+      "step": 6335
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.9274729337795006,
+      "learning_rate": 1.0112324303224255e-06,
+      "loss": 0.0125,
+      "step": 6336
+    },
+    {
+      "epoch": 1.4402272727272727,
+      "grad_norm": 2.5472879742505556,
+      "learning_rate": 1.011162258181459e-06,
+      "loss": 0.0704,
+      "step": 6337
+    },
+    {
+      "epoch": 1.4404545454545454,
+      "grad_norm": 1.8239379624626917,
+      "learning_rate": 1.0110920781659706e-06,
+      "loss": 0.0471,
+      "step": 6338
+    },
+    {
+      "epoch": 1.4406818181818182,
+      "grad_norm": 2.5960302702914295,
+      "learning_rate": 1.011021890277391e-06,
+      "loss": 0.0427,
+      "step": 6339
+    },
+    {
+      "epoch": 1.440909090909091,
+      "grad_norm": 1.901666539581469,
+      "learning_rate": 1.0109516945171516e-06,
+      "loss": 0.0379,
+      "step": 6340
+    },
+    {
+      "epoch": 1.4411363636363637,
+      "grad_norm": 1.821504827740914,
+      "learning_rate": 1.0108814908866837e-06,
+      "loss": 0.0486,
+      "step": 6341
+    },
+    {
+      "epoch": 1.4413636363636364,
+      "grad_norm": 2.0190687754751826,
+      "learning_rate": 1.0108112793874192e-06,
+      "loss": 0.0458,
+      "step": 6342
+    },
+    {
+      "epoch": 1.4415909090909091,
+      "grad_norm": 0.9377369632693148,
+      "learning_rate": 1.0107410600207894e-06,
+      "loss": 0.0257,
+      "step": 6343
+    },
+    {
+      "epoch": 1.4418181818181819,
+      "grad_norm": 1.6815507067814794,
+      "learning_rate": 1.0106708327882266e-06,
+      "loss": 0.0754,
+      "step": 6344
+    },
+    {
+      "epoch": 1.4420454545454544,
+      "grad_norm": 2.5964587576916855,
+      "learning_rate": 1.0106005976911626e-06,
+      "loss": 0.0322,
+      "step": 6345
+    },
+    {
+      "epoch": 1.4422727272727274,
+      "grad_norm": 2.715828517609748,
+      "learning_rate": 1.0105303547310298e-06,
+      "loss": 0.0824,
+      "step": 6346
+    },
+    {
+      "epoch": 1.4425,
+      "grad_norm": 1.3633734401697726,
+      "learning_rate": 1.0104601039092603e-06,
+      "loss": 0.0297,
+      "step": 6347
+    },
+    {
+      "epoch": 1.4427272727272729,
+      "grad_norm": 2.2196343211377525,
+      "learning_rate": 1.0103898452272871e-06,
+      "loss": 0.0456,
+      "step": 6348
+    },
+    {
+      "epoch": 1.4429545454545454,
+      "grad_norm": 2.5203586694738656,
+      "learning_rate": 1.0103195786865424e-06,
+      "loss": 0.0506,
+      "step": 6349
+    },
+    {
+      "epoch": 1.4431818181818181,
+      "grad_norm": 2.069827429955396,
+      "learning_rate": 1.0102493042884592e-06,
+      "loss": 0.0491,
+      "step": 6350
+    },
+    {
+      "epoch": 1.4434090909090909,
+      "grad_norm": 1.3276044073671076,
+      "learning_rate": 1.0101790220344707e-06,
+      "loss": 0.0287,
+      "step": 6351
+    },
+    {
+      "epoch": 1.4436363636363636,
+      "grad_norm": 4.614059146669947,
+      "learning_rate": 1.0101087319260098e-06,
+      "loss": 0.1035,
+      "step": 6352
+    },
+    {
+      "epoch": 1.4438636363636363,
+      "grad_norm": 2.556601040355897,
+      "learning_rate": 1.0100384339645103e-06,
+      "loss": 0.0405,
+      "step": 6353
+    },
+    {
+      "epoch": 1.444090909090909,
+      "grad_norm": 1.4403747959059388,
+      "learning_rate": 1.009968128151405e-06,
+      "loss": 0.0164,
+      "step": 6354
+    },
+    {
+      "epoch": 1.4443181818181818,
+      "grad_norm": 1.2408904017780624,
+      "learning_rate": 1.0098978144881283e-06,
+      "loss": 0.0271,
+      "step": 6355
+    },
+    {
+      "epoch": 1.4445454545454546,
+      "grad_norm": 1.6255591488506014,
+      "learning_rate": 1.0098274929761134e-06,
+      "loss": 0.0301,
+      "step": 6356
+    },
+    {
+      "epoch": 1.4447727272727273,
+      "grad_norm": 1.800085324567748,
+      "learning_rate": 1.009757163616795e-06,
+      "loss": 0.0309,
+      "step": 6357
+    },
+    {
+      "epoch": 1.445,
+      "grad_norm": 1.467922238332108,
+      "learning_rate": 1.0096868264116065e-06,
+      "loss": 0.0387,
+      "step": 6358
+    },
+    {
+      "epoch": 1.4452272727272728,
+      "grad_norm": 2.2429316418928074,
+      "learning_rate": 1.0096164813619823e-06,
+      "loss": 0.0398,
+      "step": 6359
+    },
+    {
+      "epoch": 1.4454545454545453,
+      "grad_norm": 1.5778394466890282,
+      "learning_rate": 1.0095461284693573e-06,
+      "loss": 0.0444,
+      "step": 6360
+    },
+    {
+      "epoch": 1.4456818181818183,
+      "grad_norm": 1.9777040724913126,
+      "learning_rate": 1.0094757677351658e-06,
+      "loss": 0.0412,
+      "step": 6361
+    },
+    {
+      "epoch": 1.4459090909090908,
+      "grad_norm": 2.2992166459442145,
+      "learning_rate": 1.0094053991608425e-06,
+      "loss": 0.0417,
+      "step": 6362
+    },
+    {
+      "epoch": 1.4461363636363636,
+      "grad_norm": 2.367668995994971,
+      "learning_rate": 1.0093350227478225e-06,
+      "loss": 0.0426,
+      "step": 6363
+    },
+    {
+      "epoch": 1.4463636363636363,
+      "grad_norm": 2.447919722998443,
+      "learning_rate": 1.009264638497541e-06,
+      "loss": 0.0502,
+      "step": 6364
+    },
+    {
+      "epoch": 1.446590909090909,
+      "grad_norm": 2.4587189623569548,
+      "learning_rate": 1.0091942464114332e-06,
+      "loss": 0.0406,
+      "step": 6365
+    },
+    {
+      "epoch": 1.4468181818181818,
+      "grad_norm": 1.2314543874349417,
+      "learning_rate": 1.0091238464909343e-06,
+      "loss": 0.0348,
+      "step": 6366
+    },
+    {
+      "epoch": 1.4470454545454545,
+      "grad_norm": 2.333934107634729,
+      "learning_rate": 1.0090534387374801e-06,
+      "loss": 0.0656,
+      "step": 6367
+    },
+    {
+      "epoch": 1.4472727272727273,
+      "grad_norm": 3.165476131381416,
+      "learning_rate": 1.0089830231525063e-06,
+      "loss": 0.1075,
+      "step": 6368
+    },
+    {
+      "epoch": 1.4475,
+      "grad_norm": 1.7280449262056026,
+      "learning_rate": 1.0089125997374488e-06,
+      "loss": 0.0743,
+      "step": 6369
+    },
+    {
+      "epoch": 1.4477272727272728,
+      "grad_norm": 1.076164248491903,
+      "learning_rate": 1.0088421684937434e-06,
+      "loss": 0.0179,
+      "step": 6370
+    },
+    {
+      "epoch": 1.4479545454545455,
+      "grad_norm": 1.8420352614436393,
+      "learning_rate": 1.0087717294228266e-06,
+      "loss": 0.0439,
+      "step": 6371
+    },
+    {
+      "epoch": 1.4481818181818182,
+      "grad_norm": 1.9677601589137401,
+      "learning_rate": 1.008701282526135e-06,
+      "loss": 0.0487,
+      "step": 6372
+    },
+    {
+      "epoch": 1.448409090909091,
+      "grad_norm": 1.76208865916273,
+      "learning_rate": 1.0086308278051045e-06,
+      "loss": 0.0461,
+      "step": 6373
+    },
+    {
+      "epoch": 1.4486363636363637,
+      "grad_norm": 1.393149618398718,
+      "learning_rate": 1.0085603652611723e-06,
+      "loss": 0.0314,
+      "step": 6374
+    },
+    {
+      "epoch": 1.4488636363636362,
+      "grad_norm": 1.2713531011931198,
+      "learning_rate": 1.008489894895775e-06,
+      "loss": 0.0241,
+      "step": 6375
+    },
+    {
+      "epoch": 1.4490909090909092,
+      "grad_norm": 1.1682825374549832,
+      "learning_rate": 1.0084194167103499e-06,
+      "loss": 0.0218,
+      "step": 6376
+    },
+    {
+      "epoch": 1.4493181818181817,
+      "grad_norm": 1.3454282664821646,
+      "learning_rate": 1.0083489307063338e-06,
+      "loss": 0.0449,
+      "step": 6377
+    },
+    {
+      "epoch": 1.4495454545454545,
+      "grad_norm": 1.857823890252562,
+      "learning_rate": 1.0082784368851641e-06,
+      "loss": 0.0321,
+      "step": 6378
+    },
+    {
+      "epoch": 1.4497727272727272,
+      "grad_norm": 1.82879798502903,
+      "learning_rate": 1.0082079352482787e-06,
+      "loss": 0.0567,
+      "step": 6379
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.1524491286860874,
+      "learning_rate": 1.0081374257971148e-06,
+      "loss": 0.0651,
+      "step": 6380
+    },
+    {
+      "epoch": 1.4502272727272727,
+      "grad_norm": 1.0540551132064586,
+      "learning_rate": 1.0080669085331104e-06,
+      "loss": 0.0176,
+      "step": 6381
+    },
+    {
+      "epoch": 1.4504545454545454,
+      "grad_norm": 1.5007874772008847,
+      "learning_rate": 1.0079963834577034e-06,
+      "loss": 0.0288,
+      "step": 6382
+    },
+    {
+      "epoch": 1.4506818181818182,
+      "grad_norm": 1.581750253647596,
+      "learning_rate": 1.0079258505723322e-06,
+      "loss": 0.0343,
+      "step": 6383
+    },
+    {
+      "epoch": 1.450909090909091,
+      "grad_norm": 5.123612482053277,
+      "learning_rate": 1.0078553098784345e-06,
+      "loss": 0.0957,
+      "step": 6384
+    },
+    {
+      "epoch": 1.4511363636363637,
+      "grad_norm": 2.2710636291746806,
+      "learning_rate": 1.007784761377449e-06,
+      "loss": 0.0751,
+      "step": 6385
+    },
+    {
+      "epoch": 1.4513636363636364,
+      "grad_norm": 1.4221832583294105,
+      "learning_rate": 1.007714205070815e-06,
+      "loss": 0.0505,
+      "step": 6386
+    },
+    {
+      "epoch": 1.4515909090909092,
+      "grad_norm": 1.4281442438018763,
+      "learning_rate": 1.0076436409599702e-06,
+      "loss": 0.0388,
+      "step": 6387
+    },
+    {
+      "epoch": 1.451818181818182,
+      "grad_norm": 2.7884875333498655,
+      "learning_rate": 1.0075730690463542e-06,
+      "loss": 0.0433,
+      "step": 6388
+    },
+    {
+      "epoch": 1.4520454545454546,
+      "grad_norm": 1.0829483817704513,
+      "learning_rate": 1.0075024893314058e-06,
+      "loss": 0.0278,
+      "step": 6389
+    },
+    {
+      "epoch": 1.4522727272727272,
+      "grad_norm": 1.1537700847547614,
+      "learning_rate": 1.0074319018165646e-06,
+      "loss": 0.0341,
+      "step": 6390
+    },
+    {
+      "epoch": 1.4525000000000001,
+      "grad_norm": 2.4744134078147084,
+      "learning_rate": 1.0073613065032694e-06,
+      "loss": 0.0685,
+      "step": 6391
+    },
+    {
+      "epoch": 1.4527272727272726,
+      "grad_norm": 1.955178161858656,
+      "learning_rate": 1.00729070339296e-06,
+      "loss": 0.0464,
+      "step": 6392
+    },
+    {
+      "epoch": 1.4529545454545454,
+      "grad_norm": 2.2919687501302275,
+      "learning_rate": 1.0072200924870765e-06,
+      "loss": 0.0504,
+      "step": 6393
+    },
+    {
+      "epoch": 1.4531818181818181,
+      "grad_norm": 1.5052706756703638,
+      "learning_rate": 1.0071494737870584e-06,
+      "loss": 0.0283,
+      "step": 6394
+    },
+    {
+      "epoch": 1.4534090909090909,
+      "grad_norm": 2.3618255624548206,
+      "learning_rate": 1.007078847294346e-06,
+      "loss": 0.0426,
+      "step": 6395
+    },
+    {
+      "epoch": 1.4536363636363636,
+      "grad_norm": 1.7396826884912506,
+      "learning_rate": 1.007008213010379e-06,
+      "loss": 0.0344,
+      "step": 6396
+    },
+    {
+      "epoch": 1.4538636363636364,
+      "grad_norm": 3.0452003062968567,
+      "learning_rate": 1.0069375709365984e-06,
+      "loss": 0.0473,
+      "step": 6397
+    },
+    {
+      "epoch": 1.454090909090909,
+      "grad_norm": 2.4305157855839368,
+      "learning_rate": 1.0068669210744443e-06,
+      "loss": 0.0482,
+      "step": 6398
+    },
+    {
+      "epoch": 1.4543181818181818,
+      "grad_norm": 1.9108694659055117,
+      "learning_rate": 1.0067962634253574e-06,
+      "loss": 0.0782,
+      "step": 6399
+    },
+    {
+      "epoch": 1.4545454545454546,
+      "grad_norm": 2.2009851599204824,
+      "learning_rate": 1.0067255979907787e-06,
+      "loss": 0.0597,
+      "step": 6400
+    },
+    {
+      "epoch": 1.4547727272727273,
+      "grad_norm": 1.5075623907424796,
+      "learning_rate": 1.0066549247721489e-06,
+      "loss": 0.0587,
+      "step": 6401
+    },
+    {
+      "epoch": 1.455,
+      "grad_norm": 1.5792849495095576,
+      "learning_rate": 1.0065842437709094e-06,
+      "loss": 0.0358,
+      "step": 6402
+    },
+    {
+      "epoch": 1.4552272727272726,
+      "grad_norm": 1.7687714801401462,
+      "learning_rate": 1.0065135549885017e-06,
+      "loss": 0.0485,
+      "step": 6403
+    },
+    {
+      "epoch": 1.4554545454545456,
+      "grad_norm": 2.003506902086739,
+      "learning_rate": 1.006442858426367e-06,
+      "loss": 0.0861,
+      "step": 6404
+    },
+    {
+      "epoch": 1.455681818181818,
+      "grad_norm": 1.0322870470915375,
+      "learning_rate": 1.0063721540859467e-06,
+      "loss": 0.0326,
+      "step": 6405
+    },
+    {
+      "epoch": 1.455909090909091,
+      "grad_norm": 2.1771924031443777,
+      "learning_rate": 1.006301441968683e-06,
+      "loss": 0.0919,
+      "step": 6406
+    },
+    {
+      "epoch": 1.4561363636363636,
+      "grad_norm": 1.4979487938468306,
+      "learning_rate": 1.0062307220760177e-06,
+      "loss": 0.0346,
+      "step": 6407
+    },
+    {
+      "epoch": 1.4563636363636363,
+      "grad_norm": 1.8864157928697511,
+      "learning_rate": 1.0061599944093928e-06,
+      "loss": 0.0331,
+      "step": 6408
+    },
+    {
+      "epoch": 1.456590909090909,
+      "grad_norm": 3.1506434004792485,
+      "learning_rate": 1.0060892589702506e-06,
+      "loss": 0.0871,
+      "step": 6409
+    },
+    {
+      "epoch": 1.4568181818181818,
+      "grad_norm": 2.4991603303607386,
+      "learning_rate": 1.0060185157600337e-06,
+      "loss": 0.064,
+      "step": 6410
+    },
+    {
+      "epoch": 1.4570454545454545,
+      "grad_norm": 2.6940043752145013,
+      "learning_rate": 1.0059477647801845e-06,
+      "loss": 0.075,
+      "step": 6411
+    },
+    {
+      "epoch": 1.4572727272727273,
+      "grad_norm": 3.400618876969209,
+      "learning_rate": 1.0058770060321457e-06,
+      "loss": 0.1032,
+      "step": 6412
+    },
+    {
+      "epoch": 1.4575,
+      "grad_norm": 1.8348461829014868,
+      "learning_rate": 1.0058062395173604e-06,
+      "loss": 0.0552,
+      "step": 6413
+    },
+    {
+      "epoch": 1.4577272727272728,
+      "grad_norm": 1.7459255498800592,
+      "learning_rate": 1.0057354652372715e-06,
+      "loss": 0.0338,
+      "step": 6414
+    },
+    {
+      "epoch": 1.4579545454545455,
+      "grad_norm": 2.9960765369041735,
+      "learning_rate": 1.005664683193322e-06,
+      "loss": 0.0864,
+      "step": 6415
+    },
+    {
+      "epoch": 1.4581818181818182,
+      "grad_norm": 2.1233540663369284,
+      "learning_rate": 1.0055938933869557e-06,
+      "loss": 0.0481,
+      "step": 6416
+    },
+    {
+      "epoch": 1.458409090909091,
+      "grad_norm": 2.205182153641412,
+      "learning_rate": 1.005523095819616e-06,
+      "loss": 0.0778,
+      "step": 6417
+    },
+    {
+      "epoch": 1.4586363636363635,
+      "grad_norm": 1.2476026406439347,
+      "learning_rate": 1.0054522904927463e-06,
+      "loss": 0.0314,
+      "step": 6418
+    },
+    {
+      "epoch": 1.4588636363636365,
+      "grad_norm": 2.173714448031302,
+      "learning_rate": 1.0053814774077909e-06,
+      "loss": 0.0561,
+      "step": 6419
+    },
+    {
+      "epoch": 1.459090909090909,
+      "grad_norm": 1.5369522759758425,
+      "learning_rate": 1.0053106565661935e-06,
+      "loss": 0.0331,
+      "step": 6420
+    },
+    {
+      "epoch": 1.4593181818181817,
+      "grad_norm": 1.4952005521007345,
+      "learning_rate": 1.0052398279693984e-06,
+      "loss": 0.0298,
+      "step": 6421
+    },
+    {
+      "epoch": 1.4595454545454545,
+      "grad_norm": 2.464738458694198,
+      "learning_rate": 1.0051689916188497e-06,
+      "loss": 0.1078,
+      "step": 6422
+    },
+    {
+      "epoch": 1.4597727272727272,
+      "grad_norm": 2.692557296371626,
+      "learning_rate": 1.005098147515992e-06,
+      "loss": 0.0885,
+      "step": 6423
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.893741206133349,
+      "learning_rate": 1.00502729566227e-06,
+      "loss": 0.0404,
+      "step": 6424
+    },
+    {
+      "epoch": 1.4602272727272727,
+      "grad_norm": 1.734119045172255,
+      "learning_rate": 1.0049564360591285e-06,
+      "loss": 0.0544,
+      "step": 6425
+    },
+    {
+      "epoch": 1.4604545454545454,
+      "grad_norm": 2.6086618092099934,
+      "learning_rate": 1.0048855687080122e-06,
+      "loss": 0.0406,
+      "step": 6426
+    },
+    {
+      "epoch": 1.4606818181818182,
+      "grad_norm": 1.3304806672410248,
+      "learning_rate": 1.0048146936103668e-06,
+      "loss": 0.0226,
+      "step": 6427
+    },
+    {
+      "epoch": 1.460909090909091,
+      "grad_norm": 1.414890569973089,
+      "learning_rate": 1.004743810767637e-06,
+      "loss": 0.0336,
+      "step": 6428
+    },
+    {
+      "epoch": 1.4611363636363637,
+      "grad_norm": 2.5409008708476,
+      "learning_rate": 1.0046729201812681e-06,
+      "loss": 0.0672,
+      "step": 6429
+    },
+    {
+      "epoch": 1.4613636363636364,
+      "grad_norm": 1.3859075795177764,
+      "learning_rate": 1.0046020218527063e-06,
+      "loss": 0.0336,
+      "step": 6430
+    },
+    {
+      "epoch": 1.4615909090909092,
+      "grad_norm": 1.1781413650364054,
+      "learning_rate": 1.0045311157833968e-06,
+      "loss": 0.0215,
+      "step": 6431
+    },
+    {
+      "epoch": 1.461818181818182,
+      "grad_norm": 1.8584510368889886,
+      "learning_rate": 1.0044602019747859e-06,
+      "loss": 0.0379,
+      "step": 6432
+    },
+    {
+      "epoch": 1.4620454545454544,
+      "grad_norm": 2.0498455819651933,
+      "learning_rate": 1.0043892804283195e-06,
+      "loss": 0.0636,
+      "step": 6433
+    },
+    {
+      "epoch": 1.4622727272727274,
+      "grad_norm": 1.6351628507308538,
+      "learning_rate": 1.0043183511454439e-06,
+      "loss": 0.0527,
+      "step": 6434
+    },
+    {
+      "epoch": 1.4625,
+      "grad_norm": 1.4035587239944722,
+      "learning_rate": 1.004247414127605e-06,
+      "loss": 0.0253,
+      "step": 6435
+    },
+    {
+      "epoch": 1.4627272727272727,
+      "grad_norm": 1.597077663003012,
+      "learning_rate": 1.0041764693762499e-06,
+      "loss": 0.0237,
+      "step": 6436
+    },
+    {
+      "epoch": 1.4629545454545454,
+      "grad_norm": 2.3757862644193612,
+      "learning_rate": 1.0041055168928252e-06,
+      "loss": 0.0525,
+      "step": 6437
+    },
+    {
+      "epoch": 1.4631818181818181,
+      "grad_norm": 2.432912390780326,
+      "learning_rate": 1.0040345566787773e-06,
+      "loss": 0.0531,
+      "step": 6438
+    },
+    {
+      "epoch": 1.4634090909090909,
+      "grad_norm": 1.7895530530389683,
+      "learning_rate": 1.0039635887355538e-06,
+      "loss": 0.0534,
+      "step": 6439
+    },
+    {
+      "epoch": 1.4636363636363636,
+      "grad_norm": 2.3522061131316296,
+      "learning_rate": 1.0038926130646015e-06,
+      "loss": 0.0719,
+      "step": 6440
+    },
+    {
+      "epoch": 1.4638636363636364,
+      "grad_norm": 0.7281323875085393,
+      "learning_rate": 1.0038216296673678e-06,
+      "loss": 0.0325,
+      "step": 6441
+    },
+    {
+      "epoch": 1.464090909090909,
+      "grad_norm": 1.8528409007283853,
+      "learning_rate": 1.0037506385453e-06,
+      "loss": 0.0453,
+      "step": 6442
+    },
+    {
+      "epoch": 1.4643181818181819,
+      "grad_norm": 1.217755613121193,
+      "learning_rate": 1.0036796396998463e-06,
+      "loss": 0.0434,
+      "step": 6443
+    },
+    {
+      "epoch": 1.4645454545454546,
+      "grad_norm": 1.18239754662164,
+      "learning_rate": 1.0036086331324538e-06,
+      "loss": 0.0306,
+      "step": 6444
+    },
+    {
+      "epoch": 1.4647727272727273,
+      "grad_norm": 2.2886036932436746,
+      "learning_rate": 1.003537618844571e-06,
+      "loss": 0.083,
+      "step": 6445
+    },
+    {
+      "epoch": 1.465,
+      "grad_norm": 1.248644252250673,
+      "learning_rate": 1.0034665968376457e-06,
+      "loss": 0.0331,
+      "step": 6446
+    },
+    {
+      "epoch": 1.4652272727272728,
+      "grad_norm": 2.2341162082132366,
+      "learning_rate": 1.003395567113126e-06,
+      "loss": 0.0255,
+      "step": 6447
+    },
+    {
+      "epoch": 1.4654545454545453,
+      "grad_norm": 2.303722496186659,
+      "learning_rate": 1.0033245296724607e-06,
+      "loss": 0.0755,
+      "step": 6448
+    },
+    {
+      "epoch": 1.4656818181818183,
+      "grad_norm": 2.772545418552375,
+      "learning_rate": 1.0032534845170982e-06,
+      "loss": 0.0571,
+      "step": 6449
+    },
+    {
+      "epoch": 1.4659090909090908,
+      "grad_norm": 1.2990212180120995,
+      "learning_rate": 1.003182431648487e-06,
+      "loss": 0.0305,
+      "step": 6450
+    },
+    {
+      "epoch": 1.4661363636363636,
+      "grad_norm": 2.369344649762382,
+      "learning_rate": 1.0031113710680767e-06,
+      "loss": 0.0821,
+      "step": 6451
+    },
+    {
+      "epoch": 1.4663636363636363,
+      "grad_norm": 2.9474697259164544,
+      "learning_rate": 1.0030403027773156e-06,
+      "loss": 0.0386,
+      "step": 6452
+    },
+    {
+      "epoch": 1.466590909090909,
+      "grad_norm": 1.4737302209558387,
+      "learning_rate": 1.0029692267776533e-06,
+      "loss": 0.0376,
+      "step": 6453
+    },
+    {
+      "epoch": 1.4668181818181818,
+      "grad_norm": 1.6223687234387505,
+      "learning_rate": 1.0028981430705389e-06,
+      "loss": 0.0455,
+      "step": 6454
+    },
+    {
+      "epoch": 1.4670454545454545,
+      "grad_norm": 1.1582551823843719,
+      "learning_rate": 1.0028270516574224e-06,
+      "loss": 0.0164,
+      "step": 6455
+    },
+    {
+      "epoch": 1.4672727272727273,
+      "grad_norm": 1.9614856745766123,
+      "learning_rate": 1.002755952539753e-06,
+      "loss": 0.0374,
+      "step": 6456
+    },
+    {
+      "epoch": 1.4675,
+      "grad_norm": 3.501032261160683,
+      "learning_rate": 1.0026848457189805e-06,
+      "loss": 0.0714,
+      "step": 6457
+    },
+    {
+      "epoch": 1.4677272727272728,
+      "grad_norm": 0.6449062886566868,
+      "learning_rate": 1.0026137311965554e-06,
+      "loss": 0.0228,
+      "step": 6458
+    },
+    {
+      "epoch": 1.4679545454545455,
+      "grad_norm": 1.8905696018145564,
+      "learning_rate": 1.0025426089739273e-06,
+      "loss": 0.0403,
+      "step": 6459
+    },
+    {
+      "epoch": 1.4681818181818183,
+      "grad_norm": 1.5978645464950156,
+      "learning_rate": 1.0024714790525469e-06,
+      "loss": 0.0634,
+      "step": 6460
+    },
+    {
+      "epoch": 1.4684090909090908,
+      "grad_norm": 2.076665503170499,
+      "learning_rate": 1.0024003414338645e-06,
+      "loss": 0.0643,
+      "step": 6461
+    },
+    {
+      "epoch": 1.4686363636363637,
+      "grad_norm": 2.301950907133932,
+      "learning_rate": 1.0023291961193308e-06,
+      "loss": 0.0909,
+      "step": 6462
+    },
+    {
+      "epoch": 1.4688636363636363,
+      "grad_norm": 2.8497012275632447,
+      "learning_rate": 1.0022580431103962e-06,
+      "loss": 0.0795,
+      "step": 6463
+    },
+    {
+      "epoch": 1.4690909090909092,
+      "grad_norm": 1.7464864860665767,
+      "learning_rate": 1.002186882408512e-06,
+      "loss": 0.0743,
+      "step": 6464
+    },
+    {
+      "epoch": 1.4693181818181817,
+      "grad_norm": 1.9358920529899404,
+      "learning_rate": 1.0021157140151296e-06,
+      "loss": 0.0549,
+      "step": 6465
+    },
+    {
+      "epoch": 1.4695454545454545,
+      "grad_norm": 1.4938451787516729,
+      "learning_rate": 1.0020445379316996e-06,
+      "loss": 0.0303,
+      "step": 6466
+    },
+    {
+      "epoch": 1.4697727272727272,
+      "grad_norm": 1.7919836234782096,
+      "learning_rate": 1.0019733541596738e-06,
+      "loss": 0.0477,
+      "step": 6467
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.300573545212087,
+      "learning_rate": 1.0019021627005036e-06,
+      "loss": 0.0379,
+      "step": 6468
+    },
+    {
+      "epoch": 1.4702272727272727,
+      "grad_norm": 1.5286004070302437,
+      "learning_rate": 1.0018309635556406e-06,
+      "loss": 0.0431,
+      "step": 6469
+    },
+    {
+      "epoch": 1.4704545454545455,
+      "grad_norm": 2.451816070539352,
+      "learning_rate": 1.001759756726537e-06,
+      "loss": 0.0605,
+      "step": 6470
+    },
+    {
+      "epoch": 1.4706818181818182,
+      "grad_norm": 1.4824796444184447,
+      "learning_rate": 1.0016885422146445e-06,
+      "loss": 0.0377,
+      "step": 6471
+    },
+    {
+      "epoch": 1.470909090909091,
+      "grad_norm": 1.8219110867151567,
+      "learning_rate": 1.0016173200214155e-06,
+      "loss": 0.0297,
+      "step": 6472
+    },
+    {
+      "epoch": 1.4711363636363637,
+      "grad_norm": 1.6018129029298336,
+      "learning_rate": 1.0015460901483025e-06,
+      "loss": 0.0456,
+      "step": 6473
+    },
+    {
+      "epoch": 1.4713636363636364,
+      "grad_norm": 1.4670094463757926,
+      "learning_rate": 1.0014748525967577e-06,
+      "loss": 0.0229,
+      "step": 6474
+    },
+    {
+      "epoch": 1.4715909090909092,
+      "grad_norm": 1.9719249163296253,
+      "learning_rate": 1.0014036073682337e-06,
+      "loss": 0.0512,
+      "step": 6475
+    },
+    {
+      "epoch": 1.4718181818181817,
+      "grad_norm": 3.144858040445848,
+      "learning_rate": 1.0013323544641837e-06,
+      "loss": 0.0764,
+      "step": 6476
+    },
+    {
+      "epoch": 1.4720454545454547,
+      "grad_norm": 1.7446444678620994,
+      "learning_rate": 1.0012610938860603e-06,
+      "loss": 0.0472,
+      "step": 6477
+    },
+    {
+      "epoch": 1.4722727272727272,
+      "grad_norm": 1.681896179526469,
+      "learning_rate": 1.001189825635317e-06,
+      "loss": 0.0277,
+      "step": 6478
+    },
+    {
+      "epoch": 1.4725,
+      "grad_norm": 1.5735000701431439,
+      "learning_rate": 1.0011185497134066e-06,
+      "loss": 0.0288,
+      "step": 6479
+    },
+    {
+      "epoch": 1.4727272727272727,
+      "grad_norm": 1.9807440128683147,
+      "learning_rate": 1.0010472661217828e-06,
+      "loss": 0.049,
+      "step": 6480
+    },
+    {
+      "epoch": 1.4729545454545454,
+      "grad_norm": 2.8019354708497035,
+      "learning_rate": 1.0009759748618993e-06,
+      "loss": 0.0841,
+      "step": 6481
+    },
+    {
+      "epoch": 1.4731818181818181,
+      "grad_norm": 1.3141462012641578,
+      "learning_rate": 1.0009046759352099e-06,
+      "loss": 0.0359,
+      "step": 6482
+    },
+    {
+      "epoch": 1.473409090909091,
+      "grad_norm": 1.8099518864671829,
+      "learning_rate": 1.000833369343168e-06,
+      "loss": 0.0638,
+      "step": 6483
+    },
+    {
+      "epoch": 1.4736363636363636,
+      "grad_norm": 1.7928747147717299,
+      "learning_rate": 1.0007620550872282e-06,
+      "loss": 0.0299,
+      "step": 6484
+    },
+    {
+      "epoch": 1.4738636363636364,
+      "grad_norm": 0.9084807330373733,
+      "learning_rate": 1.0006907331688445e-06,
+      "loss": 0.0161,
+      "step": 6485
+    },
+    {
+      "epoch": 1.4740909090909091,
+      "grad_norm": 1.7492076973586586,
+      "learning_rate": 1.0006194035894714e-06,
+      "loss": 0.0536,
+      "step": 6486
+    },
+    {
+      "epoch": 1.4743181818181819,
+      "grad_norm": 1.6401618932878868,
+      "learning_rate": 1.000548066350563e-06,
+      "loss": 0.0336,
+      "step": 6487
+    },
+    {
+      "epoch": 1.4745454545454546,
+      "grad_norm": 2.1237557803350673,
+      "learning_rate": 1.0004767214535749e-06,
+      "loss": 0.0819,
+      "step": 6488
+    },
+    {
+      "epoch": 1.4747727272727273,
+      "grad_norm": 3.0234266451569214,
+      "learning_rate": 1.0004053688999613e-06,
+      "loss": 0.1082,
+      "step": 6489
+    },
+    {
+      "epoch": 1.475,
+      "grad_norm": 1.6624811436506461,
+      "learning_rate": 1.000334008691177e-06,
+      "loss": 0.0463,
+      "step": 6490
+    },
+    {
+      "epoch": 1.4752272727272726,
+      "grad_norm": 2.8324656831104127,
+      "learning_rate": 1.0002626408286777e-06,
+      "loss": 0.1034,
+      "step": 6491
+    },
+    {
+      "epoch": 1.4754545454545456,
+      "grad_norm": 2.9367717500298562,
+      "learning_rate": 1.0001912653139182e-06,
+      "loss": 0.0677,
+      "step": 6492
+    },
+    {
+      "epoch": 1.475681818181818,
+      "grad_norm": 1.549604658085629,
+      "learning_rate": 1.0001198821483545e-06,
+      "loss": 0.0379,
+      "step": 6493
+    },
+    {
+      "epoch": 1.4759090909090908,
+      "grad_norm": 1.7312949189933577,
+      "learning_rate": 1.000048491333442e-06,
+      "loss": 0.0606,
+      "step": 6494
+    },
+    {
+      "epoch": 1.4761363636363636,
+      "grad_norm": 0.7045728291151284,
+      "learning_rate": 9.999770928706364e-07,
+      "loss": 0.0242,
+      "step": 6495
+    },
+    {
+      "epoch": 1.4763636363636363,
+      "grad_norm": 2.518501684205732,
+      "learning_rate": 9.999056867613935e-07,
+      "loss": 0.065,
+      "step": 6496
+    },
+    {
+      "epoch": 1.476590909090909,
+      "grad_norm": 1.1160542032812777,
+      "learning_rate": 9.998342730071698e-07,
+      "loss": 0.0251,
+      "step": 6497
+    },
+    {
+      "epoch": 1.4768181818181818,
+      "grad_norm": 1.7517132477168245,
+      "learning_rate": 9.997628516094213e-07,
+      "loss": 0.0376,
+      "step": 6498
+    },
+    {
+      "epoch": 1.4770454545454546,
+      "grad_norm": 1.2291051238794855,
+      "learning_rate": 9.996914225696044e-07,
+      "loss": 0.0427,
+      "step": 6499
+    },
+    {
+      "epoch": 1.4772727272727273,
+      "grad_norm": 1.5036364585758628,
+      "learning_rate": 9.996199858891757e-07,
+      "loss": 0.0274,
+      "step": 6500
+    },
+    {
+      "epoch": 1.4775,
+      "grad_norm": 1.7036474020302164,
+      "learning_rate": 9.995485415695919e-07,
+      "loss": 0.0396,
+      "step": 6501
+    },
+    {
+      "epoch": 1.4777272727272728,
+      "grad_norm": 1.5595659874079888,
+      "learning_rate": 9.9947708961231e-07,
+      "loss": 0.0251,
+      "step": 6502
+    },
+    {
+      "epoch": 1.4779545454545455,
+      "grad_norm": 1.8474452364712497,
+      "learning_rate": 9.994056300187868e-07,
+      "loss": 0.0329,
+      "step": 6503
+    },
+    {
+      "epoch": 1.4781818181818183,
+      "grad_norm": 1.562892744521227,
+      "learning_rate": 9.993341627904799e-07,
+      "loss": 0.0328,
+      "step": 6504
+    },
+    {
+      "epoch": 1.478409090909091,
+      "grad_norm": 1.037840631430209,
+      "learning_rate": 9.992626879288461e-07,
+      "loss": 0.0288,
+      "step": 6505
+    },
+    {
+      "epoch": 1.4786363636363635,
+      "grad_norm": 1.6087571583707625,
+      "learning_rate": 9.99191205435343e-07,
+      "loss": 0.0381,
+      "step": 6506
+    },
+    {
+      "epoch": 1.4788636363636365,
+      "grad_norm": 1.7643342492245495,
+      "learning_rate": 9.991197153114288e-07,
+      "loss": 0.0599,
+      "step": 6507
+    },
+    {
+      "epoch": 1.479090909090909,
+      "grad_norm": 0.3258987791918649,
+      "learning_rate": 9.990482175585606e-07,
+      "loss": 0.0092,
+      "step": 6508
+    },
+    {
+      "epoch": 1.4793181818181818,
+      "grad_norm": 1.0451107645988214,
+      "learning_rate": 9.98976712178197e-07,
+      "loss": 0.0229,
+      "step": 6509
+    },
+    {
+      "epoch": 1.4795454545454545,
+      "grad_norm": 1.791551021819946,
+      "learning_rate": 9.989051991717956e-07,
+      "loss": 0.0383,
+      "step": 6510
+    },
+    {
+      "epoch": 1.4797727272727272,
+      "grad_norm": 2.2130294778801,
+      "learning_rate": 9.988336785408147e-07,
+      "loss": 0.0873,
+      "step": 6511
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.287135963548202,
+      "learning_rate": 9.987621502867132e-07,
+      "loss": 0.034,
+      "step": 6512
+    },
+    {
+      "epoch": 1.4802272727272727,
+      "grad_norm": 1.9810031598120177,
+      "learning_rate": 9.986906144109494e-07,
+      "loss": 0.0657,
+      "step": 6513
+    },
+    {
+      "epoch": 1.4804545454545455,
+      "grad_norm": 1.6798273106163748,
+      "learning_rate": 9.98619070914982e-07,
+      "loss": 0.0492,
+      "step": 6514
+    },
+    {
+      "epoch": 1.4806818181818182,
+      "grad_norm": 1.518315408723825,
+      "learning_rate": 9.985475198002697e-07,
+      "loss": 0.0341,
+      "step": 6515
+    },
+    {
+      "epoch": 1.480909090909091,
+      "grad_norm": 1.863929313176273,
+      "learning_rate": 9.984759610682721e-07,
+      "loss": 0.0307,
+      "step": 6516
+    },
+    {
+      "epoch": 1.4811363636363637,
+      "grad_norm": 2.276051984922568,
+      "learning_rate": 9.98404394720448e-07,
+      "loss": 0.0508,
+      "step": 6517
+    },
+    {
+      "epoch": 1.4813636363636364,
+      "grad_norm": 1.922673671116797,
+      "learning_rate": 9.983328207582566e-07,
+      "loss": 0.0578,
+      "step": 6518
+    },
+    {
+      "epoch": 1.481590909090909,
+      "grad_norm": 2.2840132596481033,
+      "learning_rate": 9.982612391831579e-07,
+      "loss": 0.0565,
+      "step": 6519
+    },
+    {
+      "epoch": 1.481818181818182,
+      "grad_norm": 1.741349289831192,
+      "learning_rate": 9.981896499966114e-07,
+      "loss": 0.0493,
+      "step": 6520
+    },
+    {
+      "epoch": 1.4820454545454544,
+      "grad_norm": 2.0564818777505613,
+      "learning_rate": 9.981180532000768e-07,
+      "loss": 0.0358,
+      "step": 6521
+    },
+    {
+      "epoch": 1.4822727272727274,
+      "grad_norm": 2.471218321009787,
+      "learning_rate": 9.980464487950138e-07,
+      "loss": 0.0584,
+      "step": 6522
+    },
+    {
+      "epoch": 1.4825,
+      "grad_norm": 1.2926271239070892,
+      "learning_rate": 9.979748367828833e-07,
+      "loss": 0.0261,
+      "step": 6523
+    },
+    {
+      "epoch": 1.4827272727272727,
+      "grad_norm": 2.1052778091316524,
+      "learning_rate": 9.979032171651453e-07,
+      "loss": 0.0659,
+      "step": 6524
+    },
+    {
+      "epoch": 1.4829545454545454,
+      "grad_norm": 1.077833648371343,
+      "learning_rate": 9.978315899432598e-07,
+      "loss": 0.0378,
+      "step": 6525
+    },
+    {
+      "epoch": 1.4831818181818182,
+      "grad_norm": 2.5647186452982855,
+      "learning_rate": 9.97759955118688e-07,
+      "loss": 0.0803,
+      "step": 6526
+    },
+    {
+      "epoch": 1.483409090909091,
+      "grad_norm": 2.012607105773318,
+      "learning_rate": 9.976883126928903e-07,
+      "loss": 0.0625,
+      "step": 6527
+    },
+    {
+      "epoch": 1.4836363636363636,
+      "grad_norm": 2.1871079203579233,
+      "learning_rate": 9.976166626673278e-07,
+      "loss": 0.0516,
+      "step": 6528
+    },
+    {
+      "epoch": 1.4838636363636364,
+      "grad_norm": 1.5676719613990289,
+      "learning_rate": 9.975450050434613e-07,
+      "loss": 0.0392,
+      "step": 6529
+    },
+    {
+      "epoch": 1.4840909090909091,
+      "grad_norm": 2.297928929053252,
+      "learning_rate": 9.97473339822752e-07,
+      "loss": 0.0604,
+      "step": 6530
+    },
+    {
+      "epoch": 1.4843181818181819,
+      "grad_norm": 1.9558656899247615,
+      "learning_rate": 9.97401667006662e-07,
+      "loss": 0.0217,
+      "step": 6531
+    },
+    {
+      "epoch": 1.4845454545454546,
+      "grad_norm": 2.8488470671970054,
+      "learning_rate": 9.97329986596652e-07,
+      "loss": 0.1054,
+      "step": 6532
+    },
+    {
+      "epoch": 1.4847727272727274,
+      "grad_norm": 1.682800802942668,
+      "learning_rate": 9.972582985941842e-07,
+      "loss": 0.0574,
+      "step": 6533
+    },
+    {
+      "epoch": 1.4849999999999999,
+      "grad_norm": 2.2412195419135563,
+      "learning_rate": 9.971866030007201e-07,
+      "loss": 0.0484,
+      "step": 6534
+    },
+    {
+      "epoch": 1.4852272727272728,
+      "grad_norm": 2.445211719342115,
+      "learning_rate": 9.97114899817722e-07,
+      "loss": 0.0833,
+      "step": 6535
+    },
+    {
+      "epoch": 1.4854545454545454,
+      "grad_norm": 1.690177412523068,
+      "learning_rate": 9.970431890466516e-07,
+      "loss": 0.0611,
+      "step": 6536
+    },
+    {
+      "epoch": 1.485681818181818,
+      "grad_norm": 2.187370803303269,
+      "learning_rate": 9.96971470688972e-07,
+      "loss": 0.0678,
+      "step": 6537
+    },
+    {
+      "epoch": 1.4859090909090908,
+      "grad_norm": 2.920787283244549,
+      "learning_rate": 9.968997447461448e-07,
+      "loss": 0.0523,
+      "step": 6538
+    },
+    {
+      "epoch": 1.4861363636363636,
+      "grad_norm": 3.1682804322358225,
+      "learning_rate": 9.96828011219633e-07,
+      "loss": 0.0824,
+      "step": 6539
+    },
+    {
+      "epoch": 1.4863636363636363,
+      "grad_norm": 1.3339242040681767,
+      "learning_rate": 9.967562701108994e-07,
+      "loss": 0.0369,
+      "step": 6540
+    },
+    {
+      "epoch": 1.486590909090909,
+      "grad_norm": 2.635823413546444,
+      "learning_rate": 9.96684521421407e-07,
+      "loss": 0.04,
+      "step": 6541
+    },
+    {
+      "epoch": 1.4868181818181818,
+      "grad_norm": 2.891038203915287,
+      "learning_rate": 9.966127651526186e-07,
+      "loss": 0.098,
+      "step": 6542
+    },
+    {
+      "epoch": 1.4870454545454546,
+      "grad_norm": 1.5473372660790663,
+      "learning_rate": 9.96541001305998e-07,
+      "loss": 0.0646,
+      "step": 6543
+    },
+    {
+      "epoch": 1.4872727272727273,
+      "grad_norm": 1.670056165674067,
+      "learning_rate": 9.964692298830077e-07,
+      "loss": 0.0332,
+      "step": 6544
+    },
+    {
+      "epoch": 1.4875,
+      "grad_norm": 1.847073472092438,
+      "learning_rate": 9.963974508851117e-07,
+      "loss": 0.0332,
+      "step": 6545
+    },
+    {
+      "epoch": 1.4877272727272728,
+      "grad_norm": 1.005747812739735,
+      "learning_rate": 9.963256643137743e-07,
+      "loss": 0.0301,
+      "step": 6546
+    },
+    {
+      "epoch": 1.4879545454545455,
+      "grad_norm": 1.9141397310249786,
+      "learning_rate": 9.962538701704582e-07,
+      "loss": 0.0415,
+      "step": 6547
+    },
+    {
+      "epoch": 1.4881818181818183,
+      "grad_norm": 1.4617793685723888,
+      "learning_rate": 9.961820684566283e-07,
+      "loss": 0.0206,
+      "step": 6548
+    },
+    {
+      "epoch": 1.4884090909090908,
+      "grad_norm": 1.3009819885903207,
+      "learning_rate": 9.961102591737484e-07,
+      "loss": 0.0282,
+      "step": 6549
+    },
+    {
+      "epoch": 1.4886363636363638,
+      "grad_norm": 1.5555495132167816,
+      "learning_rate": 9.96038442323283e-07,
+      "loss": 0.0393,
+      "step": 6550
+    },
+    {
+      "epoch": 1.4888636363636363,
+      "grad_norm": 1.4452620299142844,
+      "learning_rate": 9.959666179066965e-07,
+      "loss": 0.0323,
+      "step": 6551
+    },
+    {
+      "epoch": 1.489090909090909,
+      "grad_norm": 2.9664664773067804,
+      "learning_rate": 9.958947859254533e-07,
+      "loss": 0.0923,
+      "step": 6552
+    },
+    {
+      "epoch": 1.4893181818181818,
+      "grad_norm": 1.3616203862747707,
+      "learning_rate": 9.958229463810184e-07,
+      "loss": 0.035,
+      "step": 6553
+    },
+    {
+      "epoch": 1.4895454545454545,
+      "grad_norm": 1.7010559932006986,
+      "learning_rate": 9.957510992748565e-07,
+      "loss": 0.0438,
+      "step": 6554
+    },
+    {
+      "epoch": 1.4897727272727272,
+      "grad_norm": 1.5338385437916502,
+      "learning_rate": 9.956792446084332e-07,
+      "loss": 0.0349,
+      "step": 6555
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 2.946423364462156,
+      "learning_rate": 9.95607382383213e-07,
+      "loss": 0.0813,
+      "step": 6556
+    },
+    {
+      "epoch": 1.4902272727272727,
+      "grad_norm": 2.0548951778142412,
+      "learning_rate": 9.95535512600662e-07,
+      "loss": 0.0453,
+      "step": 6557
+    },
+    {
+      "epoch": 1.4904545454545455,
+      "grad_norm": 2.128400268762302,
+      "learning_rate": 9.954636352622456e-07,
+      "loss": 0.0407,
+      "step": 6558
+    },
+    {
+      "epoch": 1.4906818181818182,
+      "grad_norm": 1.56307745594216,
+      "learning_rate": 9.953917503694291e-07,
+      "loss": 0.0222,
+      "step": 6559
+    },
+    {
+      "epoch": 1.490909090909091,
+      "grad_norm": 1.7005560009422307,
+      "learning_rate": 9.953198579236788e-07,
+      "loss": 0.0347,
+      "step": 6560
+    },
+    {
+      "epoch": 1.4911363636363637,
+      "grad_norm": 2.115870008741916,
+      "learning_rate": 9.952479579264605e-07,
+      "loss": 0.0599,
+      "step": 6561
+    },
+    {
+      "epoch": 1.4913636363636364,
+      "grad_norm": 2.420441214813657,
+      "learning_rate": 9.951760503792403e-07,
+      "loss": 0.0628,
+      "step": 6562
+    },
+    {
+      "epoch": 1.4915909090909092,
+      "grad_norm": 1.9418115530149749,
+      "learning_rate": 9.951041352834849e-07,
+      "loss": 0.0485,
+      "step": 6563
+    },
+    {
+      "epoch": 1.4918181818181817,
+      "grad_norm": 2.7883452388707797,
+      "learning_rate": 9.950322126406601e-07,
+      "loss": 0.1236,
+      "step": 6564
+    },
+    {
+      "epoch": 1.4920454545454547,
+      "grad_norm": 2.162365744426155,
+      "learning_rate": 9.949602824522331e-07,
+      "loss": 0.042,
+      "step": 6565
+    },
+    {
+      "epoch": 1.4922727272727272,
+      "grad_norm": 2.3856751836432144,
+      "learning_rate": 9.948883447196706e-07,
+      "loss": 0.0524,
+      "step": 6566
+    },
+    {
+      "epoch": 1.4925,
+      "grad_norm": 2.602329683200007,
+      "learning_rate": 9.948163994444393e-07,
+      "loss": 0.0649,
+      "step": 6567
+    },
+    {
+      "epoch": 1.4927272727272727,
+      "grad_norm": 1.937078436099697,
+      "learning_rate": 9.947444466280066e-07,
+      "loss": 0.0286,
+      "step": 6568
+    },
+    {
+      "epoch": 1.4929545454545454,
+      "grad_norm": 1.367135501671428,
+      "learning_rate": 9.946724862718395e-07,
+      "loss": 0.0281,
+      "step": 6569
+    },
+    {
+      "epoch": 1.4931818181818182,
+      "grad_norm": 1.6910299778095137,
+      "learning_rate": 9.946005183774054e-07,
+      "loss": 0.0477,
+      "step": 6570
+    },
+    {
+      "epoch": 1.493409090909091,
+      "grad_norm": 2.182095568247313,
+      "learning_rate": 9.945285429461722e-07,
+      "loss": 0.0748,
+      "step": 6571
+    },
+    {
+      "epoch": 1.4936363636363637,
+      "grad_norm": 2.2490082178103763,
+      "learning_rate": 9.94456559979607e-07,
+      "loss": 0.1153,
+      "step": 6572
+    },
+    {
+      "epoch": 1.4938636363636364,
+      "grad_norm": 2.0101644604030344,
+      "learning_rate": 9.943845694791782e-07,
+      "loss": 0.054,
+      "step": 6573
+    },
+    {
+      "epoch": 1.4940909090909091,
+      "grad_norm": 2.496389860358231,
+      "learning_rate": 9.943125714463532e-07,
+      "loss": 0.0302,
+      "step": 6574
+    },
+    {
+      "epoch": 1.4943181818181819,
+      "grad_norm": 1.3259820587414055,
+      "learning_rate": 9.94240565882601e-07,
+      "loss": 0.0705,
+      "step": 6575
+    },
+    {
+      "epoch": 1.4945454545454546,
+      "grad_norm": 3.2107717196420333,
+      "learning_rate": 9.941685527893894e-07,
+      "loss": 0.0613,
+      "step": 6576
+    },
+    {
+      "epoch": 1.4947727272727271,
+      "grad_norm": 1.6679951678339786,
+      "learning_rate": 9.94096532168187e-07,
+      "loss": 0.0488,
+      "step": 6577
+    },
+    {
+      "epoch": 1.495,
+      "grad_norm": 1.7577693139306658,
+      "learning_rate": 9.940245040204625e-07,
+      "loss": 0.0723,
+      "step": 6578
+    },
+    {
+      "epoch": 1.4952272727272726,
+      "grad_norm": 2.351816483400507,
+      "learning_rate": 9.939524683476842e-07,
+      "loss": 0.0503,
+      "step": 6579
+    },
+    {
+      "epoch": 1.4954545454545456,
+      "grad_norm": 2.716166248567904,
+      "learning_rate": 9.938804251513215e-07,
+      "loss": 0.0562,
+      "step": 6580
+    },
+    {
+      "epoch": 1.4956818181818181,
+      "grad_norm": 1.5072126007550888,
+      "learning_rate": 9.938083744328436e-07,
+      "loss": 0.0442,
+      "step": 6581
+    },
+    {
+      "epoch": 1.4959090909090909,
+      "grad_norm": 1.5730440886460517,
+      "learning_rate": 9.937363161937193e-07,
+      "loss": 0.0234,
+      "step": 6582
+    },
+    {
+      "epoch": 1.4961363636363636,
+      "grad_norm": 2.1193402438481583,
+      "learning_rate": 9.936642504354184e-07,
+      "loss": 0.0777,
+      "step": 6583
+    },
+    {
+      "epoch": 1.4963636363636363,
+      "grad_norm": 1.7596625976177753,
+      "learning_rate": 9.935921771594101e-07,
+      "loss": 0.0285,
+      "step": 6584
+    },
+    {
+      "epoch": 1.496590909090909,
+      "grad_norm": 1.8836154694453486,
+      "learning_rate": 9.935200963671644e-07,
+      "loss": 0.0509,
+      "step": 6585
+    },
+    {
+      "epoch": 1.4968181818181818,
+      "grad_norm": 1.4459430708633927,
+      "learning_rate": 9.934480080601507e-07,
+      "loss": 0.0651,
+      "step": 6586
+    },
+    {
+      "epoch": 1.4970454545454546,
+      "grad_norm": 1.8867530431205157,
+      "learning_rate": 9.933759122398397e-07,
+      "loss": 0.0338,
+      "step": 6587
+    },
+    {
+      "epoch": 1.4972727272727273,
+      "grad_norm": 2.4752614902230072,
+      "learning_rate": 9.93303808907701e-07,
+      "loss": 0.0727,
+      "step": 6588
+    },
+    {
+      "epoch": 1.4975,
+      "grad_norm": 2.561480096734537,
+      "learning_rate": 9.93231698065205e-07,
+      "loss": 0.0486,
+      "step": 6589
+    },
+    {
+      "epoch": 1.4977272727272728,
+      "grad_norm": 1.80270078136266,
+      "learning_rate": 9.931595797138227e-07,
+      "loss": 0.0248,
+      "step": 6590
+    },
+    {
+      "epoch": 1.4979545454545455,
+      "grad_norm": 1.7313584666570403,
+      "learning_rate": 9.930874538550238e-07,
+      "loss": 0.0591,
+      "step": 6591
+    },
+    {
+      "epoch": 1.498181818181818,
+      "grad_norm": 1.8823880712644492,
+      "learning_rate": 9.930153204902798e-07,
+      "loss": 0.0772,
+      "step": 6592
+    },
+    {
+      "epoch": 1.498409090909091,
+      "grad_norm": 2.3520793462563554,
+      "learning_rate": 9.929431796210614e-07,
+      "loss": 0.0603,
+      "step": 6593
+    },
+    {
+      "epoch": 1.4986363636363635,
+      "grad_norm": 1.8382200858958435,
+      "learning_rate": 9.928710312488395e-07,
+      "loss": 0.044,
+      "step": 6594
+    },
+    {
+      "epoch": 1.4988636363636363,
+      "grad_norm": 1.244293094470016,
+      "learning_rate": 9.927988753750856e-07,
+      "loss": 0.028,
+      "step": 6595
+    },
+    {
+      "epoch": 1.499090909090909,
+      "grad_norm": 1.4590445235715856,
+      "learning_rate": 9.92726712001271e-07,
+      "loss": 0.0359,
+      "step": 6596
+    },
+    {
+      "epoch": 1.4993181818181818,
+      "grad_norm": 1.66237864092323,
+      "learning_rate": 9.926545411288672e-07,
+      "loss": 0.0375,
+      "step": 6597
+    },
+    {
+      "epoch": 1.4995454545454545,
+      "grad_norm": 1.7217744929490182,
+      "learning_rate": 9.925823627593456e-07,
+      "loss": 0.0225,
+      "step": 6598
+    },
+    {
+      "epoch": 1.4997727272727273,
+      "grad_norm": 2.740343577341118,
+      "learning_rate": 9.925101768941787e-07,
+      "loss": 0.0508,
+      "step": 6599
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.0129000119327585,
+      "learning_rate": 9.92437983534838e-07,
+      "loss": 0.0309,
+      "step": 6600
+    },
+    {
+      "epoch": 1.5002272727272727,
+      "grad_norm": 2.4159561580989344,
+      "learning_rate": 9.923657826827958e-07,
+      "loss": 0.0636,
+      "step": 6601
+    },
+    {
+      "epoch": 1.5004545454545455,
+      "grad_norm": 2.9023923977599884,
+      "learning_rate": 9.922935743395243e-07,
+      "loss": 0.0933,
+      "step": 6602
+    },
+    {
+      "epoch": 1.5006818181818182,
+      "grad_norm": 1.6983483585834407,
+      "learning_rate": 9.922213585064962e-07,
+      "loss": 0.0515,
+      "step": 6603
+    },
+    {
+      "epoch": 1.500909090909091,
+      "grad_norm": 2.9106752152990047,
+      "learning_rate": 9.921491351851839e-07,
+      "loss": 0.0557,
+      "step": 6604
+    },
+    {
+      "epoch": 1.5011363636363635,
+      "grad_norm": 2.1625458630857013,
+      "learning_rate": 9.920769043770602e-07,
+      "loss": 0.0628,
+      "step": 6605
+    },
+    {
+      "epoch": 1.5013636363636365,
+      "grad_norm": 1.2168091297813395,
+      "learning_rate": 9.92004666083598e-07,
+      "loss": 0.029,
+      "step": 6606
+    },
+    {
+      "epoch": 1.501590909090909,
+      "grad_norm": 1.7538977975385164,
+      "learning_rate": 9.919324203062704e-07,
+      "loss": 0.0418,
+      "step": 6607
+    },
+    {
+      "epoch": 1.501818181818182,
+      "grad_norm": 1.9044952514884834,
+      "learning_rate": 9.918601670465506e-07,
+      "loss": 0.0524,
+      "step": 6608
+    },
+    {
+      "epoch": 1.5020454545454545,
+      "grad_norm": 2.3032763388533084,
+      "learning_rate": 9.917879063059118e-07,
+      "loss": 0.038,
+      "step": 6609
+    },
+    {
+      "epoch": 1.5022727272727274,
+      "grad_norm": 1.1532897586288675,
+      "learning_rate": 9.917156380858279e-07,
+      "loss": 0.0288,
+      "step": 6610
+    },
+    {
+      "epoch": 1.5025,
+      "grad_norm": 2.1348094484342326,
+      "learning_rate": 9.916433623877722e-07,
+      "loss": 0.049,
+      "step": 6611
+    },
+    {
+      "epoch": 1.5027272727272727,
+      "grad_norm": 2.371059489026449,
+      "learning_rate": 9.91571079213219e-07,
+      "loss": 0.0436,
+      "step": 6612
+    },
+    {
+      "epoch": 1.5029545454545454,
+      "grad_norm": 2.023291412928825,
+      "learning_rate": 9.914987885636418e-07,
+      "loss": 0.0644,
+      "step": 6613
+    },
+    {
+      "epoch": 1.5031818181818182,
+      "grad_norm": 2.141256033061052,
+      "learning_rate": 9.914264904405148e-07,
+      "loss": 0.0488,
+      "step": 6614
+    },
+    {
+      "epoch": 1.503409090909091,
+      "grad_norm": 1.7093943029064382,
+      "learning_rate": 9.913541848453126e-07,
+      "loss": 0.0331,
+      "step": 6615
+    },
+    {
+      "epoch": 1.5036363636363637,
+      "grad_norm": 1.9342212927772962,
+      "learning_rate": 9.91281871779509e-07,
+      "loss": 0.0444,
+      "step": 6616
+    },
+    {
+      "epoch": 1.5038636363636364,
+      "grad_norm": 1.619302448168268,
+      "learning_rate": 9.912095512445797e-07,
+      "loss": 0.0367,
+      "step": 6617
+    },
+    {
+      "epoch": 1.5040909090909091,
+      "grad_norm": 1.34185241708385,
+      "learning_rate": 9.911372232419983e-07,
+      "loss": 0.0323,
+      "step": 6618
+    },
+    {
+      "epoch": 1.5043181818181819,
+      "grad_norm": 2.182873040551185,
+      "learning_rate": 9.910648877732404e-07,
+      "loss": 0.0481,
+      "step": 6619
+    },
+    {
+      "epoch": 1.5045454545454544,
+      "grad_norm": 2.9577541061551593,
+      "learning_rate": 9.909925448397806e-07,
+      "loss": 0.0615,
+      "step": 6620
+    },
+    {
+      "epoch": 1.5047727272727274,
+      "grad_norm": 1.8247739409819554,
+      "learning_rate": 9.909201944430943e-07,
+      "loss": 0.0399,
+      "step": 6621
+    },
+    {
+      "epoch": 1.505,
+      "grad_norm": 1.8312817741277252,
+      "learning_rate": 9.908478365846569e-07,
+      "loss": 0.059,
+      "step": 6622
+    },
+    {
+      "epoch": 1.5052272727272729,
+      "grad_norm": 2.264416753228814,
+      "learning_rate": 9.90775471265944e-07,
+      "loss": 0.0482,
+      "step": 6623
+    },
+    {
+      "epoch": 1.5054545454545454,
+      "grad_norm": 2.6440040990676446,
+      "learning_rate": 9.907030984884309e-07,
+      "loss": 0.0426,
+      "step": 6624
+    },
+    {
+      "epoch": 1.5056818181818183,
+      "grad_norm": 1.8542599417105803,
+      "learning_rate": 9.906307182535938e-07,
+      "loss": 0.0472,
+      "step": 6625
+    },
+    {
+      "epoch": 1.5059090909090909,
+      "grad_norm": 1.7873782551517983,
+      "learning_rate": 9.905583305629082e-07,
+      "loss": 0.0455,
+      "step": 6626
+    },
+    {
+      "epoch": 1.5061363636363636,
+      "grad_norm": 1.9297159114297056,
+      "learning_rate": 9.904859354178507e-07,
+      "loss": 0.0305,
+      "step": 6627
+    },
+    {
+      "epoch": 1.5063636363636363,
+      "grad_norm": 1.1905482582650695,
+      "learning_rate": 9.904135328198974e-07,
+      "loss": 0.0322,
+      "step": 6628
+    },
+    {
+      "epoch": 1.506590909090909,
+      "grad_norm": 2.209206355673789,
+      "learning_rate": 9.903411227705244e-07,
+      "loss": 0.0555,
+      "step": 6629
+    },
+    {
+      "epoch": 1.5068181818181818,
+      "grad_norm": 1.6959847823236476,
+      "learning_rate": 9.902687052712089e-07,
+      "loss": 0.0227,
+      "step": 6630
+    },
+    {
+      "epoch": 1.5070454545454546,
+      "grad_norm": 2.0489052941818082,
+      "learning_rate": 9.90196280323427e-07,
+      "loss": 0.0551,
+      "step": 6631
+    },
+    {
+      "epoch": 1.5072727272727273,
+      "grad_norm": 2.5293718255401414,
+      "learning_rate": 9.901238479286559e-07,
+      "loss": 0.0595,
+      "step": 6632
+    },
+    {
+      "epoch": 1.5074999999999998,
+      "grad_norm": 1.9796403712514603,
+      "learning_rate": 9.900514080883724e-07,
+      "loss": 0.0289,
+      "step": 6633
+    },
+    {
+      "epoch": 1.5077272727272728,
+      "grad_norm": 1.2695931167190475,
+      "learning_rate": 9.899789608040542e-07,
+      "loss": 0.0273,
+      "step": 6634
+    },
+    {
+      "epoch": 1.5079545454545453,
+      "grad_norm": 2.1063855693630695,
+      "learning_rate": 9.89906506077178e-07,
+      "loss": 0.0362,
+      "step": 6635
+    },
+    {
+      "epoch": 1.5081818181818183,
+      "grad_norm": 2.2193950376202514,
+      "learning_rate": 9.898340439092216e-07,
+      "loss": 0.0481,
+      "step": 6636
+    },
+    {
+      "epoch": 1.5084090909090908,
+      "grad_norm": 1.768341539304703,
+      "learning_rate": 9.897615743016623e-07,
+      "loss": 0.0387,
+      "step": 6637
+    },
+    {
+      "epoch": 1.5086363636363638,
+      "grad_norm": 1.8648379693059878,
+      "learning_rate": 9.896890972559786e-07,
+      "loss": 0.0809,
+      "step": 6638
+    },
+    {
+      "epoch": 1.5088636363636363,
+      "grad_norm": 2.030144559656272,
+      "learning_rate": 9.896166127736476e-07,
+      "loss": 0.0334,
+      "step": 6639
+    },
+    {
+      "epoch": 1.509090909090909,
+      "grad_norm": 2.2910314249358112,
+      "learning_rate": 9.895441208561478e-07,
+      "loss": 0.0418,
+      "step": 6640
+    },
+    {
+      "epoch": 1.5093181818181818,
+      "grad_norm": 1.884500966355803,
+      "learning_rate": 9.894716215049575e-07,
+      "loss": 0.0475,
+      "step": 6641
+    },
+    {
+      "epoch": 1.5095454545454545,
+      "grad_norm": 1.8545069480244365,
+      "learning_rate": 9.89399114721555e-07,
+      "loss": 0.0215,
+      "step": 6642
+    },
+    {
+      "epoch": 1.5097727272727273,
+      "grad_norm": 2.4785797176208235,
+      "learning_rate": 9.893266005074186e-07,
+      "loss": 0.0538,
+      "step": 6643
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.7499181366043006,
+      "learning_rate": 9.892540788640275e-07,
+      "loss": 0.0566,
+      "step": 6644
+    },
+    {
+      "epoch": 1.5102272727272728,
+      "grad_norm": 5.722979957150903,
+      "learning_rate": 9.891815497928602e-07,
+      "loss": 0.1061,
+      "step": 6645
+    },
+    {
+      "epoch": 1.5104545454545455,
+      "grad_norm": 1.6446382774172605,
+      "learning_rate": 9.891090132953957e-07,
+      "loss": 0.0556,
+      "step": 6646
+    },
+    {
+      "epoch": 1.5106818181818182,
+      "grad_norm": 1.8882415623314464,
+      "learning_rate": 9.89036469373113e-07,
+      "loss": 0.0422,
+      "step": 6647
+    },
+    {
+      "epoch": 1.5109090909090908,
+      "grad_norm": 2.398471457008044,
+      "learning_rate": 9.889639180274919e-07,
+      "loss": 0.0403,
+      "step": 6648
+    },
+    {
+      "epoch": 1.5111363636363637,
+      "grad_norm": 2.34756378132755,
+      "learning_rate": 9.888913592600114e-07,
+      "loss": 0.0654,
+      "step": 6649
+    },
+    {
+      "epoch": 1.5113636363636362,
+      "grad_norm": 1.9674117729888982,
+      "learning_rate": 9.888187930721514e-07,
+      "loss": 0.0368,
+      "step": 6650
+    },
+    {
+      "epoch": 1.5115909090909092,
+      "grad_norm": 1.4716057412873393,
+      "learning_rate": 9.887462194653912e-07,
+      "loss": 0.0384,
+      "step": 6651
+    },
+    {
+      "epoch": 1.5118181818181817,
+      "grad_norm": 1.5743338077794378,
+      "learning_rate": 9.886736384412112e-07,
+      "loss": 0.0477,
+      "step": 6652
+    },
+    {
+      "epoch": 1.5120454545454547,
+      "grad_norm": 1.2808418344865269,
+      "learning_rate": 9.886010500010911e-07,
+      "loss": 0.0417,
+      "step": 6653
+    },
+    {
+      "epoch": 1.5122727272727272,
+      "grad_norm": 2.8799889855353484,
+      "learning_rate": 9.885284541465113e-07,
+      "loss": 0.044,
+      "step": 6654
+    },
+    {
+      "epoch": 1.5125,
+      "grad_norm": 1.633796647462169,
+      "learning_rate": 9.884558508789522e-07,
+      "loss": 0.0349,
+      "step": 6655
+    },
+    {
+      "epoch": 1.5127272727272727,
+      "grad_norm": 0.5886035004201491,
+      "learning_rate": 9.883832401998942e-07,
+      "loss": 0.0129,
+      "step": 6656
+    },
+    {
+      "epoch": 1.5129545454545454,
+      "grad_norm": 1.8358633881625528,
+      "learning_rate": 9.883106221108179e-07,
+      "loss": 0.063,
+      "step": 6657
+    },
+    {
+      "epoch": 1.5131818181818182,
+      "grad_norm": 1.698154292866491,
+      "learning_rate": 9.882379966132043e-07,
+      "loss": 0.0566,
+      "step": 6658
+    },
+    {
+      "epoch": 1.513409090909091,
+      "grad_norm": 1.2944470256235217,
+      "learning_rate": 9.881653637085341e-07,
+      "loss": 0.0581,
+      "step": 6659
+    },
+    {
+      "epoch": 1.5136363636363637,
+      "grad_norm": 2.2901706704273423,
+      "learning_rate": 9.880927233982887e-07,
+      "loss": 0.094,
+      "step": 6660
+    },
+    {
+      "epoch": 1.5138636363636364,
+      "grad_norm": 1.5877349976276107,
+      "learning_rate": 9.880200756839492e-07,
+      "loss": 0.039,
+      "step": 6661
+    },
+    {
+      "epoch": 1.5140909090909092,
+      "grad_norm": 2.2049235918361996,
+      "learning_rate": 9.87947420566997e-07,
+      "loss": 0.0652,
+      "step": 6662
+    },
+    {
+      "epoch": 1.5143181818181817,
+      "grad_norm": 1.6704563201757021,
+      "learning_rate": 9.878747580489136e-07,
+      "loss": 0.0497,
+      "step": 6663
+    },
+    {
+      "epoch": 1.5145454545454546,
+      "grad_norm": 1.8883765305663291,
+      "learning_rate": 9.878020881311809e-07,
+      "loss": 0.0488,
+      "step": 6664
+    },
+    {
+      "epoch": 1.5147727272727272,
+      "grad_norm": 2.0798929939054775,
+      "learning_rate": 9.877294108152807e-07,
+      "loss": 0.0505,
+      "step": 6665
+    },
+    {
+      "epoch": 1.5150000000000001,
+      "grad_norm": 1.9711322250157581,
+      "learning_rate": 9.87656726102695e-07,
+      "loss": 0.0276,
+      "step": 6666
+    },
+    {
+      "epoch": 1.5152272727272726,
+      "grad_norm": 1.3563638910184237,
+      "learning_rate": 9.87584033994906e-07,
+      "loss": 0.0293,
+      "step": 6667
+    },
+    {
+      "epoch": 1.5154545454545456,
+      "grad_norm": 1.0617940134086636,
+      "learning_rate": 9.87511334493396e-07,
+      "loss": 0.0268,
+      "step": 6668
+    },
+    {
+      "epoch": 1.5156818181818181,
+      "grad_norm": 1.7740161801308283,
+      "learning_rate": 9.874386275996473e-07,
+      "loss": 0.0734,
+      "step": 6669
+    },
+    {
+      "epoch": 1.5159090909090909,
+      "grad_norm": 1.4972878454786418,
+      "learning_rate": 9.873659133151426e-07,
+      "loss": 0.0552,
+      "step": 6670
+    },
+    {
+      "epoch": 1.5161363636363636,
+      "grad_norm": 2.4775660817992677,
+      "learning_rate": 9.87293191641365e-07,
+      "loss": 0.0729,
+      "step": 6671
+    },
+    {
+      "epoch": 1.5163636363636364,
+      "grad_norm": 1.3911198144674335,
+      "learning_rate": 9.872204625797972e-07,
+      "loss": 0.0355,
+      "step": 6672
+    },
+    {
+      "epoch": 1.516590909090909,
+      "grad_norm": 3.487369633336006,
+      "learning_rate": 9.871477261319222e-07,
+      "loss": 0.0772,
+      "step": 6673
+    },
+    {
+      "epoch": 1.5168181818181818,
+      "grad_norm": 11.161220626005917,
+      "learning_rate": 9.870749822992228e-07,
+      "loss": 0.0422,
+      "step": 6674
+    },
+    {
+      "epoch": 1.5170454545454546,
+      "grad_norm": 1.9740932850627055,
+      "learning_rate": 9.870022310831835e-07,
+      "loss": 0.038,
+      "step": 6675
+    },
+    {
+      "epoch": 1.517272727272727,
+      "grad_norm": 2.0101835376231048,
+      "learning_rate": 9.869294724852868e-07,
+      "loss": 0.0405,
+      "step": 6676
+    },
+    {
+      "epoch": 1.5175,
+      "grad_norm": 2.9141032669312903,
+      "learning_rate": 9.868567065070167e-07,
+      "loss": 0.06,
+      "step": 6677
+    },
+    {
+      "epoch": 1.5177272727272726,
+      "grad_norm": 1.2816639552900657,
+      "learning_rate": 9.867839331498571e-07,
+      "loss": 0.046,
+      "step": 6678
+    },
+    {
+      "epoch": 1.5179545454545456,
+      "grad_norm": 2.3346434242465697,
+      "learning_rate": 9.86711152415292e-07,
+      "loss": 0.0644,
+      "step": 6679
+    },
+    {
+      "epoch": 1.518181818181818,
+      "grad_norm": 2.266167129167941,
+      "learning_rate": 9.866383643048057e-07,
+      "loss": 0.0831,
+      "step": 6680
+    },
+    {
+      "epoch": 1.518409090909091,
+      "grad_norm": 2.8239198248803112,
+      "learning_rate": 9.86565568819882e-07,
+      "loss": 0.0664,
+      "step": 6681
+    },
+    {
+      "epoch": 1.5186363636363636,
+      "grad_norm": 1.4108559495551638,
+      "learning_rate": 9.864927659620055e-07,
+      "loss": 0.041,
+      "step": 6682
+    },
+    {
+      "epoch": 1.5188636363636365,
+      "grad_norm": 1.8351193418256948,
+      "learning_rate": 9.86419955732661e-07,
+      "loss": 0.0586,
+      "step": 6683
+    },
+    {
+      "epoch": 1.519090909090909,
+      "grad_norm": 1.9809466678117487,
+      "learning_rate": 9.863471381333328e-07,
+      "loss": 0.0447,
+      "step": 6684
+    },
+    {
+      "epoch": 1.5193181818181818,
+      "grad_norm": 1.6216202334027818,
+      "learning_rate": 9.862743131655063e-07,
+      "loss": 0.0425,
+      "step": 6685
+    },
+    {
+      "epoch": 1.5195454545454545,
+      "grad_norm": 2.310947864149758,
+      "learning_rate": 9.86201480830666e-07,
+      "loss": 0.0916,
+      "step": 6686
+    },
+    {
+      "epoch": 1.5197727272727273,
+      "grad_norm": 1.4000737586705951,
+      "learning_rate": 9.861286411302978e-07,
+      "loss": 0.038,
+      "step": 6687
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 2.2384277589826755,
+      "learning_rate": 9.860557940658864e-07,
+      "loss": 0.0606,
+      "step": 6688
+    },
+    {
+      "epoch": 1.5202272727272728,
+      "grad_norm": 1.7329984193121752,
+      "learning_rate": 9.859829396389172e-07,
+      "loss": 0.0271,
+      "step": 6689
+    },
+    {
+      "epoch": 1.5204545454545455,
+      "grad_norm": 2.221239104342009,
+      "learning_rate": 9.859100778508766e-07,
+      "loss": 0.0327,
+      "step": 6690
+    },
+    {
+      "epoch": 1.520681818181818,
+      "grad_norm": 1.7794481283853263,
+      "learning_rate": 9.858372087032493e-07,
+      "loss": 0.0241,
+      "step": 6691
+    },
+    {
+      "epoch": 1.520909090909091,
+      "grad_norm": 1.7383193527121834,
+      "learning_rate": 9.85764332197522e-07,
+      "loss": 0.044,
+      "step": 6692
+    },
+    {
+      "epoch": 1.5211363636363635,
+      "grad_norm": 2.290082512425131,
+      "learning_rate": 9.85691448335181e-07,
+      "loss": 0.0958,
+      "step": 6693
+    },
+    {
+      "epoch": 1.5213636363636365,
+      "grad_norm": 2.2808955940090594,
+      "learning_rate": 9.856185571177115e-07,
+      "loss": 0.0408,
+      "step": 6694
+    },
+    {
+      "epoch": 1.521590909090909,
+      "grad_norm": 1.9278406193624633,
+      "learning_rate": 9.85545658546601e-07,
+      "loss": 0.065,
+      "step": 6695
+    },
+    {
+      "epoch": 1.521818181818182,
+      "grad_norm": 1.802768773545635,
+      "learning_rate": 9.854727526233351e-07,
+      "loss": 0.0403,
+      "step": 6696
+    },
+    {
+      "epoch": 1.5220454545454545,
+      "grad_norm": 1.6948779650199108,
+      "learning_rate": 9.85399839349401e-07,
+      "loss": 0.0458,
+      "step": 6697
+    },
+    {
+      "epoch": 1.5222727272727272,
+      "grad_norm": 1.9882626093204763,
+      "learning_rate": 9.853269187262856e-07,
+      "loss": 0.0334,
+      "step": 6698
+    },
+    {
+      "epoch": 1.5225,
+      "grad_norm": 2.4368260787872815,
+      "learning_rate": 9.852539907554755e-07,
+      "loss": 0.0689,
+      "step": 6699
+    },
+    {
+      "epoch": 1.5227272727272727,
+      "grad_norm": 1.8577798352383617,
+      "learning_rate": 9.85181055438458e-07,
+      "loss": 0.0629,
+      "step": 6700
+    },
+    {
+      "epoch": 1.5229545454545454,
+      "grad_norm": 1.6749472337972824,
+      "learning_rate": 9.851081127767206e-07,
+      "loss": 0.0372,
+      "step": 6701
+    },
+    {
+      "epoch": 1.5231818181818182,
+      "grad_norm": 1.753568101982999,
+      "learning_rate": 9.850351627717505e-07,
+      "loss": 0.0291,
+      "step": 6702
+    },
+    {
+      "epoch": 1.523409090909091,
+      "grad_norm": 1.8104500356923812,
+      "learning_rate": 9.849622054250353e-07,
+      "loss": 0.0809,
+      "step": 6703
+    },
+    {
+      "epoch": 1.5236363636363637,
+      "grad_norm": 1.9326733311176403,
+      "learning_rate": 9.848892407380628e-07,
+      "loss": 0.0516,
+      "step": 6704
+    },
+    {
+      "epoch": 1.5238636363636364,
+      "grad_norm": 1.9208526214872532,
+      "learning_rate": 9.848162687123206e-07,
+      "loss": 0.037,
+      "step": 6705
+    },
+    {
+      "epoch": 1.524090909090909,
+      "grad_norm": 1.4619680518669451,
+      "learning_rate": 9.847432893492972e-07,
+      "loss": 0.0517,
+      "step": 6706
+    },
+    {
+      "epoch": 1.524318181818182,
+      "grad_norm": 2.844940482966167,
+      "learning_rate": 9.846703026504804e-07,
+      "loss": 0.0961,
+      "step": 6707
+    },
+    {
+      "epoch": 1.5245454545454544,
+      "grad_norm": 2.916369027625539,
+      "learning_rate": 9.845973086173587e-07,
+      "loss": 0.0803,
+      "step": 6708
+    },
+    {
+      "epoch": 1.5247727272727274,
+      "grad_norm": 1.2213353513217777,
+      "learning_rate": 9.845243072514204e-07,
+      "loss": 0.0157,
+      "step": 6709
+    },
+    {
+      "epoch": 1.525,
+      "grad_norm": 2.2694293031997024,
+      "learning_rate": 9.844512985541544e-07,
+      "loss": 0.0518,
+      "step": 6710
+    },
+    {
+      "epoch": 1.5252272727272729,
+      "grad_norm": 1.9066761732104662,
+      "learning_rate": 9.843782825270494e-07,
+      "loss": 0.0456,
+      "step": 6711
+    },
+    {
+      "epoch": 1.5254545454545454,
+      "grad_norm": 1.4251936412424198,
+      "learning_rate": 9.843052591715938e-07,
+      "loss": 0.0385,
+      "step": 6712
+    },
+    {
+      "epoch": 1.5256818181818181,
+      "grad_norm": 1.527342249164405,
+      "learning_rate": 9.842322284892776e-07,
+      "loss": 0.0393,
+      "step": 6713
+    },
+    {
+      "epoch": 1.5259090909090909,
+      "grad_norm": 3.264670111563198,
+      "learning_rate": 9.841591904815893e-07,
+      "loss": 0.0813,
+      "step": 6714
+    },
+    {
+      "epoch": 1.5261363636363636,
+      "grad_norm": 1.4698731021498699,
+      "learning_rate": 9.840861451500187e-07,
+      "loss": 0.0432,
+      "step": 6715
+    },
+    {
+      "epoch": 1.5263636363636364,
+      "grad_norm": 1.3978391898822724,
+      "learning_rate": 9.84013092496055e-07,
+      "loss": 0.0382,
+      "step": 6716
+    },
+    {
+      "epoch": 1.526590909090909,
+      "grad_norm": 2.023765187076473,
+      "learning_rate": 9.839400325211882e-07,
+      "loss": 0.0544,
+      "step": 6717
+    },
+    {
+      "epoch": 1.5268181818181819,
+      "grad_norm": 3.765601660360754,
+      "learning_rate": 9.838669652269076e-07,
+      "loss": 0.1438,
+      "step": 6718
+    },
+    {
+      "epoch": 1.5270454545454546,
+      "grad_norm": 1.6042768105358107,
+      "learning_rate": 9.837938906147037e-07,
+      "loss": 0.0311,
+      "step": 6719
+    },
+    {
+      "epoch": 1.5272727272727273,
+      "grad_norm": 1.5579458971185538,
+      "learning_rate": 9.837208086860665e-07,
+      "loss": 0.0579,
+      "step": 6720
+    },
+    {
+      "epoch": 1.5274999999999999,
+      "grad_norm": 2.5440178292388396,
+      "learning_rate": 9.83647719442486e-07,
+      "loss": 0.0437,
+      "step": 6721
+    },
+    {
+      "epoch": 1.5277272727272728,
+      "grad_norm": 2.573828954180803,
+      "learning_rate": 9.83574622885453e-07,
+      "loss": 0.087,
+      "step": 6722
+    },
+    {
+      "epoch": 1.5279545454545453,
+      "grad_norm": 1.4310637612558077,
+      "learning_rate": 9.835015190164577e-07,
+      "loss": 0.0278,
+      "step": 6723
+    },
+    {
+      "epoch": 1.5281818181818183,
+      "grad_norm": 1.1133619375677042,
+      "learning_rate": 9.83428407836991e-07,
+      "loss": 0.0222,
+      "step": 6724
+    },
+    {
+      "epoch": 1.5284090909090908,
+      "grad_norm": 1.0782833949807211,
+      "learning_rate": 9.833552893485438e-07,
+      "loss": 0.0173,
+      "step": 6725
+    },
+    {
+      "epoch": 1.5286363636363638,
+      "grad_norm": 2.223050179060035,
+      "learning_rate": 9.832821635526072e-07,
+      "loss": 0.0405,
+      "step": 6726
+    },
+    {
+      "epoch": 1.5288636363636363,
+      "grad_norm": 2.790163609999724,
+      "learning_rate": 9.83209030450672e-07,
+      "loss": 0.0674,
+      "step": 6727
+    },
+    {
+      "epoch": 1.529090909090909,
+      "grad_norm": 2.010152338868296,
+      "learning_rate": 9.831358900442299e-07,
+      "loss": 0.0771,
+      "step": 6728
+    },
+    {
+      "epoch": 1.5293181818181818,
+      "grad_norm": 2.062354082865681,
+      "learning_rate": 9.83062742334772e-07,
+      "loss": 0.0302,
+      "step": 6729
+    },
+    {
+      "epoch": 1.5295454545454545,
+      "grad_norm": 1.2447537257793293,
+      "learning_rate": 9.829895873237904e-07,
+      "loss": 0.0322,
+      "step": 6730
+    },
+    {
+      "epoch": 1.5297727272727273,
+      "grad_norm": 1.9115628967435787,
+      "learning_rate": 9.829164250127761e-07,
+      "loss": 0.0363,
+      "step": 6731
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.9533719234881497,
+      "learning_rate": 9.828432554032218e-07,
+      "loss": 0.0276,
+      "step": 6732
+    },
+    {
+      "epoch": 1.5302272727272728,
+      "grad_norm": 1.4943079344158874,
+      "learning_rate": 9.827700784966195e-07,
+      "loss": 0.0427,
+      "step": 6733
+    },
+    {
+      "epoch": 1.5304545454545453,
+      "grad_norm": 2.1051202046811777,
+      "learning_rate": 9.826968942944607e-07,
+      "loss": 0.0947,
+      "step": 6734
+    },
+    {
+      "epoch": 1.5306818181818183,
+      "grad_norm": 1.4936079233728665,
+      "learning_rate": 9.826237027982385e-07,
+      "loss": 0.0493,
+      "step": 6735
+    },
+    {
+      "epoch": 1.5309090909090908,
+      "grad_norm": 1.819256739282101,
+      "learning_rate": 9.82550504009445e-07,
+      "loss": 0.0267,
+      "step": 6736
+    },
+    {
+      "epoch": 1.5311363636363637,
+      "grad_norm": 2.684302491751897,
+      "learning_rate": 9.824772979295727e-07,
+      "loss": 0.1753,
+      "step": 6737
+    },
+    {
+      "epoch": 1.5313636363636363,
+      "grad_norm": 2.806086079793147,
+      "learning_rate": 9.824040845601148e-07,
+      "loss": 0.1025,
+      "step": 6738
+    },
+    {
+      "epoch": 1.5315909090909092,
+      "grad_norm": 1.4034928861485916,
+      "learning_rate": 9.823308639025642e-07,
+      "loss": 0.0434,
+      "step": 6739
+    },
+    {
+      "epoch": 1.5318181818181817,
+      "grad_norm": 2.0713315486984247,
+      "learning_rate": 9.822576359584138e-07,
+      "loss": 0.0436,
+      "step": 6740
+    },
+    {
+      "epoch": 1.5320454545454547,
+      "grad_norm": 2.416907156695531,
+      "learning_rate": 9.82184400729157e-07,
+      "loss": 0.0701,
+      "step": 6741
+    },
+    {
+      "epoch": 1.5322727272727272,
+      "grad_norm": 2.8045032639980487,
+      "learning_rate": 9.821111582162868e-07,
+      "loss": 0.0755,
+      "step": 6742
+    },
+    {
+      "epoch": 1.5325,
+      "grad_norm": 1.2049176993244477,
+      "learning_rate": 9.820379084212974e-07,
+      "loss": 0.0334,
+      "step": 6743
+    },
+    {
+      "epoch": 1.5327272727272727,
+      "grad_norm": 2.415253207097337,
+      "learning_rate": 9.81964651345682e-07,
+      "loss": 0.0741,
+      "step": 6744
+    },
+    {
+      "epoch": 1.5329545454545455,
+      "grad_norm": 1.1710568761522047,
+      "learning_rate": 9.81891386990935e-07,
+      "loss": 0.0425,
+      "step": 6745
+    },
+    {
+      "epoch": 1.5331818181818182,
+      "grad_norm": 2.74059426183991,
+      "learning_rate": 9.818181153585495e-07,
+      "loss": 0.0543,
+      "step": 6746
+    },
+    {
+      "epoch": 1.533409090909091,
+      "grad_norm": 1.76686890109403,
+      "learning_rate": 9.817448364500202e-07,
+      "loss": 0.0369,
+      "step": 6747
+    },
+    {
+      "epoch": 1.5336363636363637,
+      "grad_norm": 2.362608355128909,
+      "learning_rate": 9.816715502668416e-07,
+      "loss": 0.0652,
+      "step": 6748
+    },
+    {
+      "epoch": 1.5338636363636362,
+      "grad_norm": 1.8130549641642744,
+      "learning_rate": 9.815982568105075e-07,
+      "loss": 0.0488,
+      "step": 6749
+    },
+    {
+      "epoch": 1.5340909090909092,
+      "grad_norm": 2.032717907072296,
+      "learning_rate": 9.81524956082513e-07,
+      "loss": 0.0739,
+      "step": 6750
+    },
+    {
+      "epoch": 1.5343181818181817,
+      "grad_norm": 1.8498309893278984,
+      "learning_rate": 9.814516480843526e-07,
+      "loss": 0.0423,
+      "step": 6751
+    },
+    {
+      "epoch": 1.5345454545454547,
+      "grad_norm": 3.320145555621259,
+      "learning_rate": 9.813783328175215e-07,
+      "loss": 0.0825,
+      "step": 6752
+    },
+    {
+      "epoch": 1.5347727272727272,
+      "grad_norm": 1.8268015569761031,
+      "learning_rate": 9.81305010283514e-07,
+      "loss": 0.0344,
+      "step": 6753
+    },
+    {
+      "epoch": 1.5350000000000001,
+      "grad_norm": 0.9899957934938497,
+      "learning_rate": 9.81231680483826e-07,
+      "loss": 0.0202,
+      "step": 6754
+    },
+    {
+      "epoch": 1.5352272727272727,
+      "grad_norm": 3.0637387014921886,
+      "learning_rate": 9.811583434199524e-07,
+      "loss": 0.062,
+      "step": 6755
+    },
+    {
+      "epoch": 1.5354545454545454,
+      "grad_norm": 1.6057557905593962,
+      "learning_rate": 9.810849990933891e-07,
+      "loss": 0.0312,
+      "step": 6756
+    },
+    {
+      "epoch": 1.5356818181818181,
+      "grad_norm": 1.9273913281550794,
+      "learning_rate": 9.81011647505631e-07,
+      "loss": 0.0941,
+      "step": 6757
+    },
+    {
+      "epoch": 1.535909090909091,
+      "grad_norm": 1.6499550407087165,
+      "learning_rate": 9.809382886581746e-07,
+      "loss": 0.0279,
+      "step": 6758
+    },
+    {
+      "epoch": 1.5361363636363636,
+      "grad_norm": 2.6285061979322353,
+      "learning_rate": 9.808649225525158e-07,
+      "loss": 0.0492,
+      "step": 6759
+    },
+    {
+      "epoch": 1.5363636363636364,
+      "grad_norm": 1.414679304810263,
+      "learning_rate": 9.8079154919015e-07,
+      "loss": 0.0337,
+      "step": 6760
+    },
+    {
+      "epoch": 1.5365909090909091,
+      "grad_norm": 2.1063869112045595,
+      "learning_rate": 9.80718168572574e-07,
+      "loss": 0.0391,
+      "step": 6761
+    },
+    {
+      "epoch": 1.5368181818181819,
+      "grad_norm": 2.66723717512687,
+      "learning_rate": 9.806447807012836e-07,
+      "loss": 0.0513,
+      "step": 6762
+    },
+    {
+      "epoch": 1.5370454545454546,
+      "grad_norm": 1.9045880703373552,
+      "learning_rate": 9.80571385577776e-07,
+      "loss": 0.0551,
+      "step": 6763
+    },
+    {
+      "epoch": 1.5372727272727271,
+      "grad_norm": 1.039650689710681,
+      "learning_rate": 9.804979832035475e-07,
+      "loss": 0.0262,
+      "step": 6764
+    },
+    {
+      "epoch": 1.5375,
+      "grad_norm": 1.309429269940395,
+      "learning_rate": 9.804245735800948e-07,
+      "loss": 0.034,
+      "step": 6765
+    },
+    {
+      "epoch": 1.5377272727272726,
+      "grad_norm": 1.1547316784600563,
+      "learning_rate": 9.803511567089151e-07,
+      "loss": 0.0218,
+      "step": 6766
+    },
+    {
+      "epoch": 1.5379545454545456,
+      "grad_norm": 3.2501138356536963,
+      "learning_rate": 9.802777325915052e-07,
+      "loss": 0.0673,
+      "step": 6767
+    },
+    {
+      "epoch": 1.538181818181818,
+      "grad_norm": 1.965777302830094,
+      "learning_rate": 9.802043012293627e-07,
+      "loss": 0.0658,
+      "step": 6768
+    },
+    {
+      "epoch": 1.538409090909091,
+      "grad_norm": 1.4488983392244423,
+      "learning_rate": 9.801308626239847e-07,
+      "loss": 0.0284,
+      "step": 6769
+    },
+    {
+      "epoch": 1.5386363636363636,
+      "grad_norm": 2.015877442928182,
+      "learning_rate": 9.80057416776869e-07,
+      "loss": 0.0356,
+      "step": 6770
+    },
+    {
+      "epoch": 1.5388636363636363,
+      "grad_norm": 2.6439043318115694,
+      "learning_rate": 9.799839636895129e-07,
+      "loss": 0.0386,
+      "step": 6771
+    },
+    {
+      "epoch": 1.539090909090909,
+      "grad_norm": 2.5780463571950585,
+      "learning_rate": 9.799105033634146e-07,
+      "loss": 0.1073,
+      "step": 6772
+    },
+    {
+      "epoch": 1.5393181818181818,
+      "grad_norm": 2.049189108178466,
+      "learning_rate": 9.798370358000719e-07,
+      "loss": 0.0321,
+      "step": 6773
+    },
+    {
+      "epoch": 1.5395454545454546,
+      "grad_norm": 1.830656046194352,
+      "learning_rate": 9.797635610009832e-07,
+      "loss": 0.0691,
+      "step": 6774
+    },
+    {
+      "epoch": 1.5397727272727273,
+      "grad_norm": 3.1627512824350132,
+      "learning_rate": 9.796900789676463e-07,
+      "loss": 0.0444,
+      "step": 6775
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.8041719508165215,
+      "learning_rate": 9.796165897015602e-07,
+      "loss": 0.0322,
+      "step": 6776
+    },
+    {
+      "epoch": 1.5402272727272728,
+      "grad_norm": 3.4300956966308638,
+      "learning_rate": 9.79543093204223e-07,
+      "loss": 0.0972,
+      "step": 6777
+    },
+    {
+      "epoch": 1.5404545454545455,
+      "grad_norm": 1.6556642682997482,
+      "learning_rate": 9.794695894771335e-07,
+      "loss": 0.025,
+      "step": 6778
+    },
+    {
+      "epoch": 1.540681818181818,
+      "grad_norm": 2.501274121318188,
+      "learning_rate": 9.793960785217908e-07,
+      "loss": 0.0644,
+      "step": 6779
+    },
+    {
+      "epoch": 1.540909090909091,
+      "grad_norm": 1.8248040337881233,
+      "learning_rate": 9.793225603396936e-07,
+      "loss": 0.0399,
+      "step": 6780
+    },
+    {
+      "epoch": 1.5411363636363635,
+      "grad_norm": 1.7020597122524117,
+      "learning_rate": 9.792490349323415e-07,
+      "loss": 0.0737,
+      "step": 6781
+    },
+    {
+      "epoch": 1.5413636363636365,
+      "grad_norm": 1.928411380175593,
+      "learning_rate": 9.791755023012334e-07,
+      "loss": 0.0512,
+      "step": 6782
+    },
+    {
+      "epoch": 1.541590909090909,
+      "grad_norm": 2.3326305166457937,
+      "learning_rate": 9.791019624478689e-07,
+      "loss": 0.0658,
+      "step": 6783
+    },
+    {
+      "epoch": 1.541818181818182,
+      "grad_norm": 0.6347043695292657,
+      "learning_rate": 9.79028415373748e-07,
+      "loss": 0.0143,
+      "step": 6784
+    },
+    {
+      "epoch": 1.5420454545454545,
+      "grad_norm": 4.1893033669068585,
+      "learning_rate": 9.789548610803696e-07,
+      "loss": 0.1086,
+      "step": 6785
+    },
+    {
+      "epoch": 1.5422727272727272,
+      "grad_norm": 2.0801022732174097,
+      "learning_rate": 9.788812995692342e-07,
+      "loss": 0.0815,
+      "step": 6786
+    },
+    {
+      "epoch": 1.5425,
+      "grad_norm": 1.167509248770417,
+      "learning_rate": 9.788077308418416e-07,
+      "loss": 0.0552,
+      "step": 6787
+    },
+    {
+      "epoch": 1.5427272727272727,
+      "grad_norm": 2.7892363333900687,
+      "learning_rate": 9.787341548996924e-07,
+      "loss": 0.0363,
+      "step": 6788
+    },
+    {
+      "epoch": 1.5429545454545455,
+      "grad_norm": 1.8149955927822858,
+      "learning_rate": 9.786605717442864e-07,
+      "loss": 0.0416,
+      "step": 6789
+    },
+    {
+      "epoch": 1.5431818181818182,
+      "grad_norm": 1.8488300123406378,
+      "learning_rate": 9.785869813771246e-07,
+      "loss": 0.0473,
+      "step": 6790
+    },
+    {
+      "epoch": 1.543409090909091,
+      "grad_norm": 2.6380167831799524,
+      "learning_rate": 9.78513383799707e-07,
+      "loss": 0.0905,
+      "step": 6791
+    },
+    {
+      "epoch": 1.5436363636363635,
+      "grad_norm": 1.5333176011586538,
+      "learning_rate": 9.784397790135352e-07,
+      "loss": 0.0325,
+      "step": 6792
+    },
+    {
+      "epoch": 1.5438636363636364,
+      "grad_norm": 2.28957656903153,
+      "learning_rate": 9.783661670201094e-07,
+      "loss": 0.1048,
+      "step": 6793
+    },
+    {
+      "epoch": 1.544090909090909,
+      "grad_norm": 2.439834815651016,
+      "learning_rate": 9.782925478209312e-07,
+      "loss": 0.0702,
+      "step": 6794
+    },
+    {
+      "epoch": 1.544318181818182,
+      "grad_norm": 1.4738272599126518,
+      "learning_rate": 9.782189214175014e-07,
+      "loss": 0.0375,
+      "step": 6795
+    },
+    {
+      "epoch": 1.5445454545454544,
+      "grad_norm": 2.26886949273962,
+      "learning_rate": 9.781452878113219e-07,
+      "loss": 0.0646,
+      "step": 6796
+    },
+    {
+      "epoch": 1.5447727272727274,
+      "grad_norm": 1.9502187222938947,
+      "learning_rate": 9.780716470038936e-07,
+      "loss": 0.0503,
+      "step": 6797
+    },
+    {
+      "epoch": 1.545,
+      "grad_norm": 2.14398561094954,
+      "learning_rate": 9.779979989967184e-07,
+      "loss": 0.0616,
+      "step": 6798
+    },
+    {
+      "epoch": 1.545227272727273,
+      "grad_norm": 3.545993293693219,
+      "learning_rate": 9.779243437912983e-07,
+      "loss": 0.1182,
+      "step": 6799
+    },
+    {
+      "epoch": 1.5454545454545454,
+      "grad_norm": 1.5872311920373674,
+      "learning_rate": 9.778506813891351e-07,
+      "loss": 0.0396,
+      "step": 6800
+    },
+    {
+      "epoch": 1.5456818181818182,
+      "grad_norm": 1.489134583462971,
+      "learning_rate": 9.77777011791731e-07,
+      "loss": 0.0251,
+      "step": 6801
+    },
+    {
+      "epoch": 1.545909090909091,
+      "grad_norm": 1.1195606983770476,
+      "learning_rate": 9.77703335000588e-07,
+      "loss": 0.0263,
+      "step": 6802
+    },
+    {
+      "epoch": 1.5461363636363636,
+      "grad_norm": 2.021931586100243,
+      "learning_rate": 9.776296510172086e-07,
+      "loss": 0.04,
+      "step": 6803
+    },
+    {
+      "epoch": 1.5463636363636364,
+      "grad_norm": 1.854011085791509,
+      "learning_rate": 9.775559598430955e-07,
+      "loss": 0.0432,
+      "step": 6804
+    },
+    {
+      "epoch": 1.5465909090909091,
+      "grad_norm": 2.507816092517634,
+      "learning_rate": 9.774822614797517e-07,
+      "loss": 0.0729,
+      "step": 6805
+    },
+    {
+      "epoch": 1.5468181818181819,
+      "grad_norm": 2.224348915770611,
+      "learning_rate": 9.774085559286793e-07,
+      "loss": 0.0444,
+      "step": 6806
+    },
+    {
+      "epoch": 1.5470454545454544,
+      "grad_norm": 2.4042242935562563,
+      "learning_rate": 9.773348431913814e-07,
+      "loss": 0.0475,
+      "step": 6807
+    },
+    {
+      "epoch": 1.5472727272727274,
+      "grad_norm": 1.7186554054102094,
+      "learning_rate": 9.772611232693616e-07,
+      "loss": 0.0465,
+      "step": 6808
+    },
+    {
+      "epoch": 1.5474999999999999,
+      "grad_norm": 2.7315312115665047,
+      "learning_rate": 9.771873961641232e-07,
+      "loss": 0.07,
+      "step": 6809
+    },
+    {
+      "epoch": 1.5477272727272728,
+      "grad_norm": 1.386138038285807,
+      "learning_rate": 9.77113661877169e-07,
+      "loss": 0.0424,
+      "step": 6810
+    },
+    {
+      "epoch": 1.5479545454545454,
+      "grad_norm": 1.1911799323957195,
+      "learning_rate": 9.770399204100032e-07,
+      "loss": 0.0353,
+      "step": 6811
+    },
+    {
+      "epoch": 1.5481818181818183,
+      "grad_norm": 2.302129163707838,
+      "learning_rate": 9.76966171764129e-07,
+      "loss": 0.0447,
+      "step": 6812
+    },
+    {
+      "epoch": 1.5484090909090908,
+      "grad_norm": 2.602524231295,
+      "learning_rate": 9.768924159410507e-07,
+      "loss": 0.0497,
+      "step": 6813
+    },
+    {
+      "epoch": 1.5486363636363636,
+      "grad_norm": 2.3242292070180177,
+      "learning_rate": 9.76818652942272e-07,
+      "loss": 0.0391,
+      "step": 6814
+    },
+    {
+      "epoch": 1.5488636363636363,
+      "grad_norm": 1.3689850954333216,
+      "learning_rate": 9.767448827692974e-07,
+      "loss": 0.0302,
+      "step": 6815
+    },
+    {
+      "epoch": 1.549090909090909,
+      "grad_norm": 1.4059725406976444,
+      "learning_rate": 9.766711054236308e-07,
+      "loss": 0.0718,
+      "step": 6816
+    },
+    {
+      "epoch": 1.5493181818181818,
+      "grad_norm": 2.127157392926515,
+      "learning_rate": 9.765973209067768e-07,
+      "loss": 0.0935,
+      "step": 6817
+    },
+    {
+      "epoch": 1.5495454545454546,
+      "grad_norm": 2.216739969446934,
+      "learning_rate": 9.765235292202402e-07,
+      "loss": 0.0485,
+      "step": 6818
+    },
+    {
+      "epoch": 1.5497727272727273,
+      "grad_norm": 1.2815133690924512,
+      "learning_rate": 9.764497303655255e-07,
+      "loss": 0.0245,
+      "step": 6819
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 3.5306909375959825,
+      "learning_rate": 9.763759243441376e-07,
+      "loss": 0.071,
+      "step": 6820
+    },
+    {
+      "epoch": 1.5502272727272728,
+      "grad_norm": 1.6950655149210414,
+      "learning_rate": 9.763021111575816e-07,
+      "loss": 0.0555,
+      "step": 6821
+    },
+    {
+      "epoch": 1.5504545454545453,
+      "grad_norm": 2.720875245136892,
+      "learning_rate": 9.762282908073628e-07,
+      "loss": 0.0496,
+      "step": 6822
+    },
+    {
+      "epoch": 1.5506818181818183,
+      "grad_norm": 1.7693435216934754,
+      "learning_rate": 9.761544632949864e-07,
+      "loss": 0.0491,
+      "step": 6823
+    },
+    {
+      "epoch": 1.5509090909090908,
+      "grad_norm": 2.2432257264953144,
+      "learning_rate": 9.760806286219578e-07,
+      "loss": 0.0756,
+      "step": 6824
+    },
+    {
+      "epoch": 1.5511363636363638,
+      "grad_norm": 3.416269211328338,
+      "learning_rate": 9.760067867897827e-07,
+      "loss": 0.0817,
+      "step": 6825
+    },
+    {
+      "epoch": 1.5513636363636363,
+      "grad_norm": 1.1619790064849471,
+      "learning_rate": 9.75932937799967e-07,
+      "loss": 0.026,
+      "step": 6826
+    },
+    {
+      "epoch": 1.5515909090909092,
+      "grad_norm": 1.4268986738912546,
+      "learning_rate": 9.758590816540161e-07,
+      "loss": 0.053,
+      "step": 6827
+    },
+    {
+      "epoch": 1.5518181818181818,
+      "grad_norm": 1.490143889812267,
+      "learning_rate": 9.757852183534366e-07,
+      "loss": 0.0527,
+      "step": 6828
+    },
+    {
+      "epoch": 1.5520454545454545,
+      "grad_norm": 2.2787856827509025,
+      "learning_rate": 9.757113478997345e-07,
+      "loss": 0.0484,
+      "step": 6829
+    },
+    {
+      "epoch": 1.5522727272727272,
+      "grad_norm": 1.7603921489872771,
+      "learning_rate": 9.756374702944164e-07,
+      "loss": 0.0495,
+      "step": 6830
+    },
+    {
+      "epoch": 1.5525,
+      "grad_norm": 1.295838587810353,
+      "learning_rate": 9.755635855389884e-07,
+      "loss": 0.0261,
+      "step": 6831
+    },
+    {
+      "epoch": 1.5527272727272727,
+      "grad_norm": 1.0884979695348804,
+      "learning_rate": 9.75489693634957e-07,
+      "loss": 0.0233,
+      "step": 6832
+    },
+    {
+      "epoch": 1.5529545454545455,
+      "grad_norm": 1.4434745317686974,
+      "learning_rate": 9.754157945838294e-07,
+      "loss": 0.0341,
+      "step": 6833
+    },
+    {
+      "epoch": 1.5531818181818182,
+      "grad_norm": 1.9509534931035277,
+      "learning_rate": 9.753418883871126e-07,
+      "loss": 0.0546,
+      "step": 6834
+    },
+    {
+      "epoch": 1.553409090909091,
+      "grad_norm": 2.213272857043201,
+      "learning_rate": 9.752679750463136e-07,
+      "loss": 0.0642,
+      "step": 6835
+    },
+    {
+      "epoch": 1.5536363636363637,
+      "grad_norm": 2.3563297096032687,
+      "learning_rate": 9.751940545629394e-07,
+      "loss": 0.0526,
+      "step": 6836
+    },
+    {
+      "epoch": 1.5538636363636362,
+      "grad_norm": 1.5567507951324977,
+      "learning_rate": 9.751201269384974e-07,
+      "loss": 0.0357,
+      "step": 6837
+    },
+    {
+      "epoch": 1.5540909090909092,
+      "grad_norm": 2.309327848925876,
+      "learning_rate": 9.750461921744952e-07,
+      "loss": 0.0416,
+      "step": 6838
+    },
+    {
+      "epoch": 1.5543181818181817,
+      "grad_norm": 1.8296725855396314,
+      "learning_rate": 9.749722502724404e-07,
+      "loss": 0.0406,
+      "step": 6839
+    },
+    {
+      "epoch": 1.5545454545454547,
+      "grad_norm": 2.0319340948554108,
+      "learning_rate": 9.74898301233841e-07,
+      "loss": 0.0471,
+      "step": 6840
+    },
+    {
+      "epoch": 1.5547727272727272,
+      "grad_norm": 2.036579111179166,
+      "learning_rate": 9.748243450602049e-07,
+      "loss": 0.0556,
+      "step": 6841
+    },
+    {
+      "epoch": 1.5550000000000002,
+      "grad_norm": 2.165811111363408,
+      "learning_rate": 9.747503817530399e-07,
+      "loss": 0.0626,
+      "step": 6842
+    },
+    {
+      "epoch": 1.5552272727272727,
+      "grad_norm": 1.957608431854787,
+      "learning_rate": 9.746764113138546e-07,
+      "loss": 0.05,
+      "step": 6843
+    },
+    {
+      "epoch": 1.5554545454545454,
+      "grad_norm": 2.0084055027681873,
+      "learning_rate": 9.74602433744157e-07,
+      "loss": 0.04,
+      "step": 6844
+    },
+    {
+      "epoch": 1.5556818181818182,
+      "grad_norm": 1.4262768844486917,
+      "learning_rate": 9.74528449045456e-07,
+      "loss": 0.0304,
+      "step": 6845
+    },
+    {
+      "epoch": 1.555909090909091,
+      "grad_norm": 0.9890203237283013,
+      "learning_rate": 9.7445445721926e-07,
+      "loss": 0.04,
+      "step": 6846
+    },
+    {
+      "epoch": 1.5561363636363637,
+      "grad_norm": 2.660266279808844,
+      "learning_rate": 9.743804582670782e-07,
+      "loss": 0.1073,
+      "step": 6847
+    },
+    {
+      "epoch": 1.5563636363636364,
+      "grad_norm": 1.7373082712079313,
+      "learning_rate": 9.743064521904192e-07,
+      "loss": 0.0527,
+      "step": 6848
+    },
+    {
+      "epoch": 1.5565909090909091,
+      "grad_norm": 1.6146704346301228,
+      "learning_rate": 9.742324389907923e-07,
+      "loss": 0.0603,
+      "step": 6849
+    },
+    {
+      "epoch": 1.5568181818181817,
+      "grad_norm": 1.9802780002209959,
+      "learning_rate": 9.741584186697067e-07,
+      "loss": 0.0581,
+      "step": 6850
+    },
+    {
+      "epoch": 1.5570454545454546,
+      "grad_norm": 2.0599631151535087,
+      "learning_rate": 9.740843912286717e-07,
+      "loss": 0.0391,
+      "step": 6851
+    },
+    {
+      "epoch": 1.5572727272727271,
+      "grad_norm": 1.3136422917171706,
+      "learning_rate": 9.740103566691969e-07,
+      "loss": 0.0282,
+      "step": 6852
+    },
+    {
+      "epoch": 1.5575,
+      "grad_norm": 1.522341361488454,
+      "learning_rate": 9.739363149927922e-07,
+      "loss": 0.0349,
+      "step": 6853
+    },
+    {
+      "epoch": 1.5577272727272726,
+      "grad_norm": 2.7606277559034442,
+      "learning_rate": 9.738622662009673e-07,
+      "loss": 0.0992,
+      "step": 6854
+    },
+    {
+      "epoch": 1.5579545454545456,
+      "grad_norm": 2.2393822858898647,
+      "learning_rate": 9.73788210295232e-07,
+      "loss": 0.0858,
+      "step": 6855
+    },
+    {
+      "epoch": 1.5581818181818181,
+      "grad_norm": 2.5446768525583,
+      "learning_rate": 9.737141472770967e-07,
+      "loss": 0.0935,
+      "step": 6856
+    },
+    {
+      "epoch": 1.5584090909090909,
+      "grad_norm": 0.5603075352713585,
+      "learning_rate": 9.736400771480716e-07,
+      "loss": 0.0211,
+      "step": 6857
+    },
+    {
+      "epoch": 1.5586363636363636,
+      "grad_norm": 1.6452145856323497,
+      "learning_rate": 9.735659999096671e-07,
+      "loss": 0.0397,
+      "step": 6858
+    },
+    {
+      "epoch": 1.5588636363636363,
+      "grad_norm": 2.5419501787088787,
+      "learning_rate": 9.734919155633937e-07,
+      "loss": 0.047,
+      "step": 6859
+    },
+    {
+      "epoch": 1.559090909090909,
+      "grad_norm": 0.9095067788687364,
+      "learning_rate": 9.734178241107622e-07,
+      "loss": 0.0246,
+      "step": 6860
+    },
+    {
+      "epoch": 1.5593181818181818,
+      "grad_norm": 2.186532130779292,
+      "learning_rate": 9.733437255532836e-07,
+      "loss": 0.0518,
+      "step": 6861
+    },
+    {
+      "epoch": 1.5595454545454546,
+      "grad_norm": 1.877374317241115,
+      "learning_rate": 9.732696198924685e-07,
+      "loss": 0.0609,
+      "step": 6862
+    },
+    {
+      "epoch": 1.5597727272727273,
+      "grad_norm": 1.726351565847404,
+      "learning_rate": 9.73195507129828e-07,
+      "loss": 0.0343,
+      "step": 6863
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 2.5491996096210934,
+      "learning_rate": 9.73121387266874e-07,
+      "loss": 0.0383,
+      "step": 6864
+    },
+    {
+      "epoch": 1.5602272727272726,
+      "grad_norm": 1.5932429948159958,
+      "learning_rate": 9.730472603051175e-07,
+      "loss": 0.0359,
+      "step": 6865
+    },
+    {
+      "epoch": 1.5604545454545455,
+      "grad_norm": 1.5598615415870212,
+      "learning_rate": 9.729731262460702e-07,
+      "loss": 0.0305,
+      "step": 6866
+    },
+    {
+      "epoch": 1.560681818181818,
+      "grad_norm": 2.2495628688743605,
+      "learning_rate": 9.728989850912437e-07,
+      "loss": 0.0473,
+      "step": 6867
+    },
+    {
+      "epoch": 1.560909090909091,
+      "grad_norm": 2.3592452580955214,
+      "learning_rate": 9.7282483684215e-07,
+      "loss": 0.0862,
+      "step": 6868
+    },
+    {
+      "epoch": 1.5611363636363635,
+      "grad_norm": 1.951131867434589,
+      "learning_rate": 9.727506815003012e-07,
+      "loss": 0.0392,
+      "step": 6869
+    },
+    {
+      "epoch": 1.5613636363636365,
+      "grad_norm": 0.8928245111091362,
+      "learning_rate": 9.726765190672092e-07,
+      "loss": 0.0295,
+      "step": 6870
+    },
+    {
+      "epoch": 1.561590909090909,
+      "grad_norm": 1.9080449949788156,
+      "learning_rate": 9.726023495443866e-07,
+      "loss": 0.0568,
+      "step": 6871
+    },
+    {
+      "epoch": 1.5618181818181818,
+      "grad_norm": 1.783202777655329,
+      "learning_rate": 9.725281729333453e-07,
+      "loss": 0.0313,
+      "step": 6872
+    },
+    {
+      "epoch": 1.5620454545454545,
+      "grad_norm": 1.7374638304685375,
+      "learning_rate": 9.724539892355987e-07,
+      "loss": 0.0357,
+      "step": 6873
+    },
+    {
+      "epoch": 1.5622727272727273,
+      "grad_norm": 1.8466640610486271,
+      "learning_rate": 9.723797984526586e-07,
+      "loss": 0.0405,
+      "step": 6874
+    },
+    {
+      "epoch": 1.5625,
+      "grad_norm": 0.9350372780129971,
+      "learning_rate": 9.723056005860386e-07,
+      "loss": 0.0174,
+      "step": 6875
+    },
+    {
+      "epoch": 1.5627272727272727,
+      "grad_norm": 1.7407683061123609,
+      "learning_rate": 9.722313956372515e-07,
+      "loss": 0.0623,
+      "step": 6876
+    },
+    {
+      "epoch": 1.5629545454545455,
+      "grad_norm": 1.5507333605637434,
+      "learning_rate": 9.721571836078105e-07,
+      "loss": 0.0433,
+      "step": 6877
+    },
+    {
+      "epoch": 1.5631818181818182,
+      "grad_norm": 1.4444142708835948,
+      "learning_rate": 9.720829644992287e-07,
+      "loss": 0.0351,
+      "step": 6878
+    },
+    {
+      "epoch": 1.563409090909091,
+      "grad_norm": 1.8707624444582083,
+      "learning_rate": 9.7200873831302e-07,
+      "loss": 0.0406,
+      "step": 6879
+    },
+    {
+      "epoch": 1.5636363636363635,
+      "grad_norm": 2.2900844139372807,
+      "learning_rate": 9.719345050506975e-07,
+      "loss": 0.0423,
+      "step": 6880
+    },
+    {
+      "epoch": 1.5638636363636365,
+      "grad_norm": 1.3148758786340635,
+      "learning_rate": 9.718602647137754e-07,
+      "loss": 0.0353,
+      "step": 6881
+    },
+    {
+      "epoch": 1.564090909090909,
+      "grad_norm": 2.8339272980608237,
+      "learning_rate": 9.717860173037673e-07,
+      "loss": 0.0646,
+      "step": 6882
+    },
+    {
+      "epoch": 1.564318181818182,
+      "grad_norm": 1.049241599470918,
+      "learning_rate": 9.717117628221872e-07,
+      "loss": 0.026,
+      "step": 6883
+    },
+    {
+      "epoch": 1.5645454545454545,
+      "grad_norm": 2.7357745891543392,
+      "learning_rate": 9.716375012705494e-07,
+      "loss": 0.062,
+      "step": 6884
+    },
+    {
+      "epoch": 1.5647727272727274,
+      "grad_norm": 2.0757787615654935,
+      "learning_rate": 9.715632326503684e-07,
+      "loss": 0.0668,
+      "step": 6885
+    },
+    {
+      "epoch": 1.565,
+      "grad_norm": 2.276087127899769,
+      "learning_rate": 9.714889569631583e-07,
+      "loss": 0.0885,
+      "step": 6886
+    },
+    {
+      "epoch": 1.5652272727272727,
+      "grad_norm": 2.847193363445694,
+      "learning_rate": 9.71414674210434e-07,
+      "loss": 0.0571,
+      "step": 6887
+    },
+    {
+      "epoch": 1.5654545454545454,
+      "grad_norm": 4.854448906867819,
+      "learning_rate": 9.713403843937099e-07,
+      "loss": 0.1478,
+      "step": 6888
+    },
+    {
+      "epoch": 1.5656818181818182,
+      "grad_norm": 1.7066857254219476,
+      "learning_rate": 9.712660875145014e-07,
+      "loss": 0.025,
+      "step": 6889
+    },
+    {
+      "epoch": 1.565909090909091,
+      "grad_norm": 2.2619974838446915,
+      "learning_rate": 9.711917835743231e-07,
+      "loss": 0.0726,
+      "step": 6890
+    },
+    {
+      "epoch": 1.5661363636363637,
+      "grad_norm": 1.8257438507583097,
+      "learning_rate": 9.711174725746904e-07,
+      "loss": 0.0406,
+      "step": 6891
+    },
+    {
+      "epoch": 1.5663636363636364,
+      "grad_norm": 1.7972237126988377,
+      "learning_rate": 9.710431545171186e-07,
+      "loss": 0.0289,
+      "step": 6892
+    },
+    {
+      "epoch": 1.5665909090909091,
+      "grad_norm": 2.3904070826655945,
+      "learning_rate": 9.709688294031233e-07,
+      "loss": 0.0842,
+      "step": 6893
+    },
+    {
+      "epoch": 1.5668181818181819,
+      "grad_norm": 1.93656782165886,
+      "learning_rate": 9.7089449723422e-07,
+      "loss": 0.0608,
+      "step": 6894
+    },
+    {
+      "epoch": 1.5670454545454544,
+      "grad_norm": 3.1270781990541634,
+      "learning_rate": 9.708201580119245e-07,
+      "loss": 0.061,
+      "step": 6895
+    },
+    {
+      "epoch": 1.5672727272727274,
+      "grad_norm": 1.774640295180161,
+      "learning_rate": 9.707458117377525e-07,
+      "loss": 0.0738,
+      "step": 6896
+    },
+    {
+      "epoch": 1.5675,
+      "grad_norm": 1.9895583445785936,
+      "learning_rate": 9.7067145841322e-07,
+      "loss": 0.0573,
+      "step": 6897
+    },
+    {
+      "epoch": 1.5677272727272729,
+      "grad_norm": 1.3572578941876061,
+      "learning_rate": 9.705970980398437e-07,
+      "loss": 0.029,
+      "step": 6898
+    },
+    {
+      "epoch": 1.5679545454545454,
+      "grad_norm": 2.0776905568320854,
+      "learning_rate": 9.705227306191397e-07,
+      "loss": 0.0432,
+      "step": 6899
+    },
+    {
+      "epoch": 1.5681818181818183,
+      "grad_norm": 2.5524901541585074,
+      "learning_rate": 9.704483561526242e-07,
+      "loss": 0.0538,
+      "step": 6900
+    },
+    {
+      "epoch": 1.5684090909090909,
+      "grad_norm": 2.4906405675924272,
+      "learning_rate": 9.703739746418141e-07,
+      "loss": 0.0568,
+      "step": 6901
+    },
+    {
+      "epoch": 1.5686363636363636,
+      "grad_norm": 1.698201527763677,
+      "learning_rate": 9.702995860882263e-07,
+      "loss": 0.0325,
+      "step": 6902
+    },
+    {
+      "epoch": 1.5688636363636363,
+      "grad_norm": 1.588692305123336,
+      "learning_rate": 9.702251904933774e-07,
+      "loss": 0.0621,
+      "step": 6903
+    },
+    {
+      "epoch": 1.569090909090909,
+      "grad_norm": 1.4609379268658331,
+      "learning_rate": 9.701507878587848e-07,
+      "loss": 0.0313,
+      "step": 6904
+    },
+    {
+      "epoch": 1.5693181818181818,
+      "grad_norm": 3.449951838540384,
+      "learning_rate": 9.700763781859651e-07,
+      "loss": 0.0692,
+      "step": 6905
+    },
+    {
+      "epoch": 1.5695454545454546,
+      "grad_norm": 1.8258110399602714,
+      "learning_rate": 9.700019614764364e-07,
+      "loss": 0.0511,
+      "step": 6906
+    },
+    {
+      "epoch": 1.5697727272727273,
+      "grad_norm": 1.545994914695442,
+      "learning_rate": 9.699275377317154e-07,
+      "loss": 0.0331,
+      "step": 6907
+    },
+    {
+      "epoch": 1.5699999999999998,
+      "grad_norm": 1.8603553063298737,
+      "learning_rate": 9.698531069533205e-07,
+      "loss": 0.0818,
+      "step": 6908
+    },
+    {
+      "epoch": 1.5702272727272728,
+      "grad_norm": 1.6917031970757013,
+      "learning_rate": 9.697786691427691e-07,
+      "loss": 0.0616,
+      "step": 6909
+    },
+    {
+      "epoch": 1.5704545454545453,
+      "grad_norm": 2.056217851685307,
+      "learning_rate": 9.69704224301579e-07,
+      "loss": 0.0355,
+      "step": 6910
+    },
+    {
+      "epoch": 1.5706818181818183,
+      "grad_norm": 1.254617764030147,
+      "learning_rate": 9.696297724312683e-07,
+      "loss": 0.0389,
+      "step": 6911
+    },
+    {
+      "epoch": 1.5709090909090908,
+      "grad_norm": 1.6529903928227603,
+      "learning_rate": 9.695553135333555e-07,
+      "loss": 0.0459,
+      "step": 6912
+    },
+    {
+      "epoch": 1.5711363636363638,
+      "grad_norm": 1.698381905577737,
+      "learning_rate": 9.694808476093588e-07,
+      "loss": 0.0357,
+      "step": 6913
+    },
+    {
+      "epoch": 1.5713636363636363,
+      "grad_norm": 1.3259179693634504,
+      "learning_rate": 9.694063746607965e-07,
+      "loss": 0.0318,
+      "step": 6914
+    },
+    {
+      "epoch": 1.571590909090909,
+      "grad_norm": 3.393915332595363,
+      "learning_rate": 9.693318946891874e-07,
+      "loss": 0.1796,
+      "step": 6915
+    },
+    {
+      "epoch": 1.5718181818181818,
+      "grad_norm": 2.5369773258615065,
+      "learning_rate": 9.692574076960501e-07,
+      "loss": 0.0286,
+      "step": 6916
+    },
+    {
+      "epoch": 1.5720454545454545,
+      "grad_norm": 2.037257852791563,
+      "learning_rate": 9.69182913682904e-07,
+      "loss": 0.0588,
+      "step": 6917
+    },
+    {
+      "epoch": 1.5722727272727273,
+      "grad_norm": 2.175903364174516,
+      "learning_rate": 9.691084126512675e-07,
+      "loss": 0.0427,
+      "step": 6918
+    },
+    {
+      "epoch": 1.5725,
+      "grad_norm": 2.567252182073996,
+      "learning_rate": 9.690339046026603e-07,
+      "loss": 0.034,
+      "step": 6919
+    },
+    {
+      "epoch": 1.5727272727272728,
+      "grad_norm": 1.6547186894320518,
+      "learning_rate": 9.689593895386015e-07,
+      "loss": 0.072,
+      "step": 6920
+    },
+    {
+      "epoch": 1.5729545454545455,
+      "grad_norm": 1.7398523742272685,
+      "learning_rate": 9.688848674606108e-07,
+      "loss": 0.0489,
+      "step": 6921
+    },
+    {
+      "epoch": 1.5731818181818182,
+      "grad_norm": 1.9745769305846579,
+      "learning_rate": 9.688103383702078e-07,
+      "loss": 0.0516,
+      "step": 6922
+    },
+    {
+      "epoch": 1.5734090909090908,
+      "grad_norm": 1.6559447285764568,
+      "learning_rate": 9.68735802268912e-07,
+      "loss": 0.0344,
+      "step": 6923
+    },
+    {
+      "epoch": 1.5736363636363637,
+      "grad_norm": 1.3120250999950993,
+      "learning_rate": 9.686612591582434e-07,
+      "loss": 0.0215,
+      "step": 6924
+    },
+    {
+      "epoch": 1.5738636363636362,
+      "grad_norm": 1.3999468945689102,
+      "learning_rate": 9.685867090397223e-07,
+      "loss": 0.0337,
+      "step": 6925
+    },
+    {
+      "epoch": 1.5740909090909092,
+      "grad_norm": 1.3095111023168315,
+      "learning_rate": 9.68512151914869e-07,
+      "loss": 0.0317,
+      "step": 6926
+    },
+    {
+      "epoch": 1.5743181818181817,
+      "grad_norm": 1.875630047009875,
+      "learning_rate": 9.684375877852033e-07,
+      "loss": 0.0437,
+      "step": 6927
+    },
+    {
+      "epoch": 1.5745454545454547,
+      "grad_norm": 1.0980610513579543,
+      "learning_rate": 9.683630166522463e-07,
+      "loss": 0.0196,
+      "step": 6928
+    },
+    {
+      "epoch": 1.5747727272727272,
+      "grad_norm": 1.8352103351952316,
+      "learning_rate": 9.682884385175182e-07,
+      "loss": 0.0291,
+      "step": 6929
+    },
+    {
+      "epoch": 1.575,
+      "grad_norm": 0.9977617462415428,
+      "learning_rate": 9.6821385338254e-07,
+      "loss": 0.022,
+      "step": 6930
+    },
+    {
+      "epoch": 1.5752272727272727,
+      "grad_norm": 1.6656507099114792,
+      "learning_rate": 9.681392612488326e-07,
+      "loss": 0.0368,
+      "step": 6931
+    },
+    {
+      "epoch": 1.5754545454545454,
+      "grad_norm": 2.152441297640792,
+      "learning_rate": 9.680646621179168e-07,
+      "loss": 0.0625,
+      "step": 6932
+    },
+    {
+      "epoch": 1.5756818181818182,
+      "grad_norm": 2.945162760340573,
+      "learning_rate": 9.679900559913141e-07,
+      "loss": 0.0778,
+      "step": 6933
+    },
+    {
+      "epoch": 1.575909090909091,
+      "grad_norm": 0.9872293323183857,
+      "learning_rate": 9.67915442870546e-07,
+      "loss": 0.0262,
+      "step": 6934
+    },
+    {
+      "epoch": 1.5761363636363637,
+      "grad_norm": 1.2681449054368013,
+      "learning_rate": 9.678408227571337e-07,
+      "loss": 0.0233,
+      "step": 6935
+    },
+    {
+      "epoch": 1.5763636363636364,
+      "grad_norm": 2.162602229489527,
+      "learning_rate": 9.677661956525986e-07,
+      "loss": 0.0811,
+      "step": 6936
+    },
+    {
+      "epoch": 1.5765909090909092,
+      "grad_norm": 2.1953628053032315,
+      "learning_rate": 9.67691561558463e-07,
+      "loss": 0.0534,
+      "step": 6937
+    },
+    {
+      "epoch": 1.5768181818181817,
+      "grad_norm": 2.0119966486789793,
+      "learning_rate": 9.676169204762488e-07,
+      "loss": 0.0346,
+      "step": 6938
+    },
+    {
+      "epoch": 1.5770454545454546,
+      "grad_norm": 3.041587774098328,
+      "learning_rate": 9.675422724074776e-07,
+      "loss": 0.07,
+      "step": 6939
+    },
+    {
+      "epoch": 1.5772727272727272,
+      "grad_norm": 1.8579825661495712,
+      "learning_rate": 9.67467617353672e-07,
+      "loss": 0.0474,
+      "step": 6940
+    },
+    {
+      "epoch": 1.5775000000000001,
+      "grad_norm": 1.5388648343687132,
+      "learning_rate": 9.673929553163541e-07,
+      "loss": 0.0424,
+      "step": 6941
+    },
+    {
+      "epoch": 1.5777272727272726,
+      "grad_norm": 1.4669255986283598,
+      "learning_rate": 9.673182862970467e-07,
+      "loss": 0.0349,
+      "step": 6942
+    },
+    {
+      "epoch": 1.5779545454545456,
+      "grad_norm": 4.155992258677547,
+      "learning_rate": 9.672436102972722e-07,
+      "loss": 0.0968,
+      "step": 6943
+    },
+    {
+      "epoch": 1.5781818181818181,
+      "grad_norm": 2.736680327220514,
+      "learning_rate": 9.671689273185533e-07,
+      "loss": 0.0534,
+      "step": 6944
+    },
+    {
+      "epoch": 1.5784090909090909,
+      "grad_norm": 2.641233920696683,
+      "learning_rate": 9.670942373624128e-07,
+      "loss": 0.0477,
+      "step": 6945
+    },
+    {
+      "epoch": 1.5786363636363636,
+      "grad_norm": 2.2469985701517636,
+      "learning_rate": 9.670195404303744e-07,
+      "loss": 0.0408,
+      "step": 6946
+    },
+    {
+      "epoch": 1.5788636363636364,
+      "grad_norm": 2.620827419508843,
+      "learning_rate": 9.669448365239606e-07,
+      "loss": 0.0696,
+      "step": 6947
+    },
+    {
+      "epoch": 1.579090909090909,
+      "grad_norm": 2.380620216352754,
+      "learning_rate": 9.668701256446952e-07,
+      "loss": 0.0571,
+      "step": 6948
+    },
+    {
+      "epoch": 1.5793181818181818,
+      "grad_norm": 1.78995851346502,
+      "learning_rate": 9.667954077941014e-07,
+      "loss": 0.0586,
+      "step": 6949
+    },
+    {
+      "epoch": 1.5795454545454546,
+      "grad_norm": 2.236469114947499,
+      "learning_rate": 9.66720682973703e-07,
+      "loss": 0.1009,
+      "step": 6950
+    },
+    {
+      "epoch": 1.579772727272727,
+      "grad_norm": 1.9743860998427731,
+      "learning_rate": 9.666459511850237e-07,
+      "loss": 0.034,
+      "step": 6951
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 3.6739795685053815,
+      "learning_rate": 9.665712124295875e-07,
+      "loss": 0.0735,
+      "step": 6952
+    },
+    {
+      "epoch": 1.5802272727272726,
+      "grad_norm": 0.6136312845015321,
+      "learning_rate": 9.664964667089184e-07,
+      "loss": 0.0179,
+      "step": 6953
+    },
+    {
+      "epoch": 1.5804545454545456,
+      "grad_norm": 2.286578978128098,
+      "learning_rate": 9.664217140245404e-07,
+      "loss": 0.0393,
+      "step": 6954
+    },
+    {
+      "epoch": 1.580681818181818,
+      "grad_norm": 3.130296918270931,
+      "learning_rate": 9.66346954377978e-07,
+      "loss": 0.0709,
+      "step": 6955
+    },
+    {
+      "epoch": 1.580909090909091,
+      "grad_norm": 1.5022355519196218,
+      "learning_rate": 9.662721877707557e-07,
+      "loss": 0.0438,
+      "step": 6956
+    },
+    {
+      "epoch": 1.5811363636363636,
+      "grad_norm": 2.0209027546596348,
+      "learning_rate": 9.661974142043985e-07,
+      "loss": 0.0561,
+      "step": 6957
+    },
+    {
+      "epoch": 1.5813636363636365,
+      "grad_norm": 2.8414193399838417,
+      "learning_rate": 9.661226336804302e-07,
+      "loss": 0.0879,
+      "step": 6958
+    },
+    {
+      "epoch": 1.581590909090909,
+      "grad_norm": 2.160233105569862,
+      "learning_rate": 9.660478462003765e-07,
+      "loss": 0.0385,
+      "step": 6959
+    },
+    {
+      "epoch": 1.5818181818181818,
+      "grad_norm": 1.3218626666229685,
+      "learning_rate": 9.659730517657624e-07,
+      "loss": 0.0342,
+      "step": 6960
+    },
+    {
+      "epoch": 1.5820454545454545,
+      "grad_norm": 1.2263304924344687,
+      "learning_rate": 9.658982503781127e-07,
+      "loss": 0.0219,
+      "step": 6961
+    },
+    {
+      "epoch": 1.5822727272727273,
+      "grad_norm": 1.6171109474718837,
+      "learning_rate": 9.658234420389528e-07,
+      "loss": 0.0424,
+      "step": 6962
+    },
+    {
+      "epoch": 1.5825,
+      "grad_norm": 1.7477592270350997,
+      "learning_rate": 9.657486267498087e-07,
+      "loss": 0.0304,
+      "step": 6963
+    },
+    {
+      "epoch": 1.5827272727272728,
+      "grad_norm": 1.3518610590090532,
+      "learning_rate": 9.656738045122055e-07,
+      "loss": 0.0374,
+      "step": 6964
+    },
+    {
+      "epoch": 1.5829545454545455,
+      "grad_norm": 1.8790926516782127,
+      "learning_rate": 9.65598975327669e-07,
+      "loss": 0.0793,
+      "step": 6965
+    },
+    {
+      "epoch": 1.583181818181818,
+      "grad_norm": 2.3033247886927066,
+      "learning_rate": 9.655241391977251e-07,
+      "loss": 0.0689,
+      "step": 6966
+    },
+    {
+      "epoch": 1.583409090909091,
+      "grad_norm": 2.298707348686805,
+      "learning_rate": 9.654492961239002e-07,
+      "loss": 0.0576,
+      "step": 6967
+    },
+    {
+      "epoch": 1.5836363636363635,
+      "grad_norm": 1.6308583401038033,
+      "learning_rate": 9.6537444610772e-07,
+      "loss": 0.0407,
+      "step": 6968
+    },
+    {
+      "epoch": 1.5838636363636365,
+      "grad_norm": 2.2730074053614,
+      "learning_rate": 9.65299589150711e-07,
+      "loss": 0.0412,
+      "step": 6969
+    },
+    {
+      "epoch": 1.584090909090909,
+      "grad_norm": 1.3445166282891845,
+      "learning_rate": 9.652247252544e-07,
+      "loss": 0.0462,
+      "step": 6970
+    },
+    {
+      "epoch": 1.584318181818182,
+      "grad_norm": 1.6025213926787112,
+      "learning_rate": 9.65149854420313e-07,
+      "loss": 0.037,
+      "step": 6971
+    },
+    {
+      "epoch": 1.5845454545454545,
+      "grad_norm": 2.6352548536738563,
+      "learning_rate": 9.650749766499772e-07,
+      "loss": 0.0659,
+      "step": 6972
+    },
+    {
+      "epoch": 1.5847727272727272,
+      "grad_norm": 1.0359184012948683,
+      "learning_rate": 9.650000919449193e-07,
+      "loss": 0.0234,
+      "step": 6973
+    },
+    {
+      "epoch": 1.585,
+      "grad_norm": 1.816992364253317,
+      "learning_rate": 9.649252003066664e-07,
+      "loss": 0.0425,
+      "step": 6974
+    },
+    {
+      "epoch": 1.5852272727272727,
+      "grad_norm": 1.2498924311941122,
+      "learning_rate": 9.648503017367453e-07,
+      "loss": 0.0458,
+      "step": 6975
+    },
+    {
+      "epoch": 1.5854545454545454,
+      "grad_norm": 1.683250598529712,
+      "learning_rate": 9.64775396236684e-07,
+      "loss": 0.0449,
+      "step": 6976
+    },
+    {
+      "epoch": 1.5856818181818182,
+      "grad_norm": 1.7141344495214583,
+      "learning_rate": 9.647004838080095e-07,
+      "loss": 0.0475,
+      "step": 6977
+    },
+    {
+      "epoch": 1.585909090909091,
+      "grad_norm": 1.606190964184522,
+      "learning_rate": 9.646255644522496e-07,
+      "loss": 0.0469,
+      "step": 6978
+    },
+    {
+      "epoch": 1.5861363636363637,
+      "grad_norm": 0.7994875390188432,
+      "learning_rate": 9.645506381709318e-07,
+      "loss": 0.0135,
+      "step": 6979
+    },
+    {
+      "epoch": 1.5863636363636364,
+      "grad_norm": 1.9816501133691953,
+      "learning_rate": 9.644757049655839e-07,
+      "loss": 0.0632,
+      "step": 6980
+    },
+    {
+      "epoch": 1.586590909090909,
+      "grad_norm": 2.0746633721314938,
+      "learning_rate": 9.644007648377347e-07,
+      "loss": 0.0657,
+      "step": 6981
+    },
+    {
+      "epoch": 1.586818181818182,
+      "grad_norm": 1.3175063907202123,
+      "learning_rate": 9.643258177889112e-07,
+      "loss": 0.0358,
+      "step": 6982
+    },
+    {
+      "epoch": 1.5870454545454544,
+      "grad_norm": 1.6686373790259794,
+      "learning_rate": 9.642508638206428e-07,
+      "loss": 0.0306,
+      "step": 6983
+    },
+    {
+      "epoch": 1.5872727272727274,
+      "grad_norm": 2.4973552328115933,
+      "learning_rate": 9.64175902934457e-07,
+      "loss": 0.0892,
+      "step": 6984
+    },
+    {
+      "epoch": 1.5875,
+      "grad_norm": 1.7516445482216865,
+      "learning_rate": 9.641009351318831e-07,
+      "loss": 0.0364,
+      "step": 6985
+    },
+    {
+      "epoch": 1.5877272727272729,
+      "grad_norm": 0.9904852926147393,
+      "learning_rate": 9.640259604144493e-07,
+      "loss": 0.0279,
+      "step": 6986
+    },
+    {
+      "epoch": 1.5879545454545454,
+      "grad_norm": 2.3489546895312357,
+      "learning_rate": 9.639509787836849e-07,
+      "loss": 0.0781,
+      "step": 6987
+    },
+    {
+      "epoch": 1.5881818181818181,
+      "grad_norm": 1.3489711535907318,
+      "learning_rate": 9.638759902411187e-07,
+      "loss": 0.0451,
+      "step": 6988
+    },
+    {
+      "epoch": 1.5884090909090909,
+      "grad_norm": 2.4038198614716255,
+      "learning_rate": 9.638009947882796e-07,
+      "loss": 0.0572,
+      "step": 6989
+    },
+    {
+      "epoch": 1.5886363636363636,
+      "grad_norm": 1.0791561943473056,
+      "learning_rate": 9.637259924266973e-07,
+      "loss": 0.0253,
+      "step": 6990
+    },
+    {
+      "epoch": 1.5888636363636364,
+      "grad_norm": 2.0126921771842743,
+      "learning_rate": 9.636509831579013e-07,
+      "loss": 0.0372,
+      "step": 6991
+    },
+    {
+      "epoch": 1.589090909090909,
+      "grad_norm": 1.9951612553954299,
+      "learning_rate": 9.635759669834207e-07,
+      "loss": 0.0593,
+      "step": 6992
+    },
+    {
+      "epoch": 1.5893181818181819,
+      "grad_norm": 1.3126177770415957,
+      "learning_rate": 9.635009439047855e-07,
+      "loss": 0.035,
+      "step": 6993
+    },
+    {
+      "epoch": 1.5895454545454546,
+      "grad_norm": 1.6244753674629762,
+      "learning_rate": 9.634259139235256e-07,
+      "loss": 0.0478,
+      "step": 6994
+    },
+    {
+      "epoch": 1.5897727272727273,
+      "grad_norm": 2.4728954774345833,
+      "learning_rate": 9.633508770411708e-07,
+      "loss": 0.0675,
+      "step": 6995
+    },
+    {
+      "epoch": 1.5899999999999999,
+      "grad_norm": 1.804519391603642,
+      "learning_rate": 9.632758332592514e-07,
+      "loss": 0.0893,
+      "step": 6996
+    },
+    {
+      "epoch": 1.5902272727272728,
+      "grad_norm": 2.0386329079028385,
+      "learning_rate": 9.632007825792975e-07,
+      "loss": 0.0484,
+      "step": 6997
+    },
+    {
+      "epoch": 1.5904545454545453,
+      "grad_norm": 2.914096446576712,
+      "learning_rate": 9.631257250028397e-07,
+      "loss": 0.0662,
+      "step": 6998
+    },
+    {
+      "epoch": 1.5906818181818183,
+      "grad_norm": 1.9615460789841885,
+      "learning_rate": 9.630506605314085e-07,
+      "loss": 0.039,
+      "step": 6999
+    },
+    {
+      "epoch": 1.5909090909090908,
+      "grad_norm": 1.7760634004975622,
+      "learning_rate": 9.629755891665345e-07,
+      "loss": 0.0442,
+      "step": 7000
+    },
+    {
+      "epoch": 1.5911363636363638,
+      "grad_norm": 2.031171305204827,
+      "learning_rate": 9.629005109097486e-07,
+      "loss": 0.0383,
+      "step": 7001
+    },
+    {
+      "epoch": 1.5913636363636363,
+      "grad_norm": 2.4586950827726377,
+      "learning_rate": 9.628254257625818e-07,
+      "loss": 0.0532,
+      "step": 7002
+    },
+    {
+      "epoch": 1.591590909090909,
+      "grad_norm": 1.259798546698914,
+      "learning_rate": 9.627503337265652e-07,
+      "loss": 0.0338,
+      "step": 7003
+    },
+    {
+      "epoch": 1.5918181818181818,
+      "grad_norm": 2.6713605052277427,
+      "learning_rate": 9.6267523480323e-07,
+      "loss": 0.1048,
+      "step": 7004
+    },
+    {
+      "epoch": 1.5920454545454545,
+      "grad_norm": 1.79374862511291,
+      "learning_rate": 9.626001289941079e-07,
+      "loss": 0.0486,
+      "step": 7005
+    },
+    {
+      "epoch": 1.5922727272727273,
+      "grad_norm": 2.0087856888057036,
+      "learning_rate": 9.6252501630073e-07,
+      "loss": 0.062,
+      "step": 7006
+    },
+    {
+      "epoch": 1.5925,
+      "grad_norm": 1.58493408672267,
+      "learning_rate": 9.624498967246283e-07,
+      "loss": 0.0314,
+      "step": 7007
+    },
+    {
+      "epoch": 1.5927272727272728,
+      "grad_norm": 2.1365771017626387,
+      "learning_rate": 9.623747702673342e-07,
+      "loss": 0.0528,
+      "step": 7008
+    },
+    {
+      "epoch": 1.5929545454545453,
+      "grad_norm": 2.717144097963536,
+      "learning_rate": 9.622996369303802e-07,
+      "loss": 0.0395,
+      "step": 7009
+    },
+    {
+      "epoch": 1.5931818181818183,
+      "grad_norm": 1.192280676489987,
+      "learning_rate": 9.62224496715298e-07,
+      "loss": 0.0214,
+      "step": 7010
+    },
+    {
+      "epoch": 1.5934090909090908,
+      "grad_norm": 1.5134949619702012,
+      "learning_rate": 9.6214934962362e-07,
+      "loss": 0.045,
+      "step": 7011
+    },
+    {
+      "epoch": 1.5936363636363637,
+      "grad_norm": 1.8572829993190476,
+      "learning_rate": 9.620741956568789e-07,
+      "loss": 0.0495,
+      "step": 7012
+    },
+    {
+      "epoch": 1.5938636363636363,
+      "grad_norm": 2.5334050805407626,
+      "learning_rate": 9.619990348166064e-07,
+      "loss": 0.0367,
+      "step": 7013
+    },
+    {
+      "epoch": 1.5940909090909092,
+      "grad_norm": 1.795330625436214,
+      "learning_rate": 9.61923867104336e-07,
+      "loss": 0.0641,
+      "step": 7014
+    },
+    {
+      "epoch": 1.5943181818181817,
+      "grad_norm": 2.28337936740196,
+      "learning_rate": 9.618486925216e-07,
+      "loss": 0.0481,
+      "step": 7015
+    },
+    {
+      "epoch": 1.5945454545454547,
+      "grad_norm": 1.938619571187411,
+      "learning_rate": 9.617735110699317e-07,
+      "loss": 0.0516,
+      "step": 7016
+    },
+    {
+      "epoch": 1.5947727272727272,
+      "grad_norm": 1.3998363669503953,
+      "learning_rate": 9.616983227508637e-07,
+      "loss": 0.0512,
+      "step": 7017
+    },
+    {
+      "epoch": 1.595,
+      "grad_norm": 1.3458371739157742,
+      "learning_rate": 9.616231275659296e-07,
+      "loss": 0.0456,
+      "step": 7018
+    },
+    {
+      "epoch": 1.5952272727272727,
+      "grad_norm": 1.2325722199960043,
+      "learning_rate": 9.615479255166627e-07,
+      "loss": 0.0254,
+      "step": 7019
+    },
+    {
+      "epoch": 1.5954545454545455,
+      "grad_norm": 1.7193539563415663,
+      "learning_rate": 9.614727166045964e-07,
+      "loss": 0.06,
+      "step": 7020
+    },
+    {
+      "epoch": 1.5956818181818182,
+      "grad_norm": 1.2972416801291846,
+      "learning_rate": 9.613975008312646e-07,
+      "loss": 0.0348,
+      "step": 7021
+    },
+    {
+      "epoch": 1.595909090909091,
+      "grad_norm": 2.1335239669641437,
+      "learning_rate": 9.613222781982008e-07,
+      "loss": 0.0436,
+      "step": 7022
+    },
+    {
+      "epoch": 1.5961363636363637,
+      "grad_norm": 1.9329416171522265,
+      "learning_rate": 9.612470487069388e-07,
+      "loss": 0.083,
+      "step": 7023
+    },
+    {
+      "epoch": 1.5963636363636362,
+      "grad_norm": 3.0167900510372205,
+      "learning_rate": 9.611718123590128e-07,
+      "loss": 0.0682,
+      "step": 7024
+    },
+    {
+      "epoch": 1.5965909090909092,
+      "grad_norm": 1.1528564041879104,
+      "learning_rate": 9.610965691559573e-07,
+      "loss": 0.0316,
+      "step": 7025
+    },
+    {
+      "epoch": 1.5968181818181817,
+      "grad_norm": 3.230423179215356,
+      "learning_rate": 9.610213190993063e-07,
+      "loss": 0.078,
+      "step": 7026
+    },
+    {
+      "epoch": 1.5970454545454547,
+      "grad_norm": 1.3575015985904624,
+      "learning_rate": 9.609460621905945e-07,
+      "loss": 0.0507,
+      "step": 7027
+    },
+    {
+      "epoch": 1.5972727272727272,
+      "grad_norm": 1.4368606726367485,
+      "learning_rate": 9.608707984313563e-07,
+      "loss": 0.0475,
+      "step": 7028
+    },
+    {
+      "epoch": 1.5975000000000001,
+      "grad_norm": 2.4291756531926114,
+      "learning_rate": 9.607955278231267e-07,
+      "loss": 0.0881,
+      "step": 7029
+    },
+    {
+      "epoch": 1.5977272727272727,
+      "grad_norm": 3.3365356268801705,
+      "learning_rate": 9.607202503674403e-07,
+      "loss": 0.1667,
+      "step": 7030
+    },
+    {
+      "epoch": 1.5979545454545454,
+      "grad_norm": 1.3199171660387716,
+      "learning_rate": 9.606449660658323e-07,
+      "loss": 0.0255,
+      "step": 7031
+    },
+    {
+      "epoch": 1.5981818181818181,
+      "grad_norm": 2.7631316592936033,
+      "learning_rate": 9.605696749198378e-07,
+      "loss": 0.0511,
+      "step": 7032
+    },
+    {
+      "epoch": 1.598409090909091,
+      "grad_norm": 1.4266383056624796,
+      "learning_rate": 9.604943769309924e-07,
+      "loss": 0.0138,
+      "step": 7033
+    },
+    {
+      "epoch": 1.5986363636363636,
+      "grad_norm": 1.7035816352049808,
+      "learning_rate": 9.604190721008313e-07,
+      "loss": 0.0557,
+      "step": 7034
+    },
+    {
+      "epoch": 1.5988636363636364,
+      "grad_norm": 1.759646885815036,
+      "learning_rate": 9.603437604308902e-07,
+      "loss": 0.0576,
+      "step": 7035
+    },
+    {
+      "epoch": 1.5990909090909091,
+      "grad_norm": 1.6333513836723723,
+      "learning_rate": 9.602684419227047e-07,
+      "loss": 0.0327,
+      "step": 7036
+    },
+    {
+      "epoch": 1.5993181818181819,
+      "grad_norm": 1.9512115534472827,
+      "learning_rate": 9.601931165778107e-07,
+      "loss": 0.0435,
+      "step": 7037
+    },
+    {
+      "epoch": 1.5995454545454546,
+      "grad_norm": 1.4582255155558084,
+      "learning_rate": 9.601177843977445e-07,
+      "loss": 0.0201,
+      "step": 7038
+    },
+    {
+      "epoch": 1.5997727272727271,
+      "grad_norm": 1.3848663216075132,
+      "learning_rate": 9.600424453840418e-07,
+      "loss": 0.029,
+      "step": 7039
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.490272658098656,
+      "learning_rate": 9.59967099538239e-07,
+      "loss": 0.0617,
+      "step": 7040
+    },
+    {
+      "epoch": 1.6002272727272726,
+      "grad_norm": 1.713695620743472,
+      "learning_rate": 9.59891746861873e-07,
+      "loss": 0.0262,
+      "step": 7041
+    },
+    {
+      "epoch": 1.6004545454545456,
+      "grad_norm": 2.313745030640142,
+      "learning_rate": 9.5981638735648e-07,
+      "loss": 0.0438,
+      "step": 7042
+    },
+    {
+      "epoch": 1.600681818181818,
+      "grad_norm": 1.575446230983195,
+      "learning_rate": 9.597410210235966e-07,
+      "loss": 0.0423,
+      "step": 7043
+    },
+    {
+      "epoch": 1.600909090909091,
+      "grad_norm": 3.6844295543942613,
+      "learning_rate": 9.596656478647598e-07,
+      "loss": 0.1064,
+      "step": 7044
+    },
+    {
+      "epoch": 1.6011363636363636,
+      "grad_norm": 2.054687963846532,
+      "learning_rate": 9.595902678815066e-07,
+      "loss": 0.0458,
+      "step": 7045
+    },
+    {
+      "epoch": 1.6013636363636363,
+      "grad_norm": 3.019214391522692,
+      "learning_rate": 9.59514881075374e-07,
+      "loss": 0.0913,
+      "step": 7046
+    },
+    {
+      "epoch": 1.601590909090909,
+      "grad_norm": 1.038159144954777,
+      "learning_rate": 9.594394874478997e-07,
+      "loss": 0.0167,
+      "step": 7047
+    },
+    {
+      "epoch": 1.6018181818181818,
+      "grad_norm": 1.9583055332544943,
+      "learning_rate": 9.593640870006204e-07,
+      "loss": 0.0529,
+      "step": 7048
+    },
+    {
+      "epoch": 1.6020454545454546,
+      "grad_norm": 2.2763357056977704,
+      "learning_rate": 9.592886797350742e-07,
+      "loss": 0.0329,
+      "step": 7049
+    },
+    {
+      "epoch": 1.6022727272727273,
+      "grad_norm": 2.5851791852995794,
+      "learning_rate": 9.592132656527986e-07,
+      "loss": 0.0576,
+      "step": 7050
+    },
+    {
+      "epoch": 1.6025,
+      "grad_norm": 1.997376822158563,
+      "learning_rate": 9.591378447553316e-07,
+      "loss": 0.0423,
+      "step": 7051
+    },
+    {
+      "epoch": 1.6027272727272728,
+      "grad_norm": 2.1517510460755123,
+      "learning_rate": 9.590624170442109e-07,
+      "loss": 0.0434,
+      "step": 7052
+    },
+    {
+      "epoch": 1.6029545454545455,
+      "grad_norm": 2.3871334389037977,
+      "learning_rate": 9.589869825209749e-07,
+      "loss": 0.0635,
+      "step": 7053
+    },
+    {
+      "epoch": 1.603181818181818,
+      "grad_norm": 1.2022877390766304,
+      "learning_rate": 9.589115411871613e-07,
+      "loss": 0.0249,
+      "step": 7054
+    },
+    {
+      "epoch": 1.603409090909091,
+      "grad_norm": 2.8809610734904543,
+      "learning_rate": 9.588360930443091e-07,
+      "loss": 0.0519,
+      "step": 7055
+    },
+    {
+      "epoch": 1.6036363636363635,
+      "grad_norm": 1.6945857197411167,
+      "learning_rate": 9.587606380939565e-07,
+      "loss": 0.054,
+      "step": 7056
+    },
+    {
+      "epoch": 1.6038636363636365,
+      "grad_norm": 3.324927832129725,
+      "learning_rate": 9.58685176337642e-07,
+      "loss": 0.0499,
+      "step": 7057
+    },
+    {
+      "epoch": 1.604090909090909,
+      "grad_norm": 3.728284643048274,
+      "learning_rate": 9.58609707776905e-07,
+      "loss": 0.064,
+      "step": 7058
+    },
+    {
+      "epoch": 1.604318181818182,
+      "grad_norm": 2.2019345670617483,
+      "learning_rate": 9.585342324132838e-07,
+      "loss": 0.0604,
+      "step": 7059
+    },
+    {
+      "epoch": 1.6045454545454545,
+      "grad_norm": 1.1678034634033678,
+      "learning_rate": 9.584587502483178e-07,
+      "loss": 0.0227,
+      "step": 7060
+    },
+    {
+      "epoch": 1.6047727272727272,
+      "grad_norm": 2.1901462163040977,
+      "learning_rate": 9.58383261283546e-07,
+      "loss": 0.0364,
+      "step": 7061
+    },
+    {
+      "epoch": 1.605,
+      "grad_norm": 2.5099968386668223,
+      "learning_rate": 9.583077655205078e-07,
+      "loss": 0.0959,
+      "step": 7062
+    },
+    {
+      "epoch": 1.6052272727272727,
+      "grad_norm": 1.607741385534304,
+      "learning_rate": 9.58232262960743e-07,
+      "loss": 0.0289,
+      "step": 7063
+    },
+    {
+      "epoch": 1.6054545454545455,
+      "grad_norm": 2.3235093450132402,
+      "learning_rate": 9.58156753605791e-07,
+      "loss": 0.061,
+      "step": 7064
+    },
+    {
+      "epoch": 1.6056818181818182,
+      "grad_norm": 1.3146222556217584,
+      "learning_rate": 9.580812374571913e-07,
+      "loss": 0.0338,
+      "step": 7065
+    },
+    {
+      "epoch": 1.605909090909091,
+      "grad_norm": 1.47402193767477,
+      "learning_rate": 9.580057145164844e-07,
+      "loss": 0.065,
+      "step": 7066
+    },
+    {
+      "epoch": 1.6061363636363635,
+      "grad_norm": 3.21289033360165,
+      "learning_rate": 9.5793018478521e-07,
+      "loss": 0.1128,
+      "step": 7067
+    },
+    {
+      "epoch": 1.6063636363636364,
+      "grad_norm": 3.086715163317407,
+      "learning_rate": 9.57854648264908e-07,
+      "loss": 0.1009,
+      "step": 7068
+    },
+    {
+      "epoch": 1.606590909090909,
+      "grad_norm": 1.0302566953546661,
+      "learning_rate": 9.577791049571195e-07,
+      "loss": 0.0347,
+      "step": 7069
+    },
+    {
+      "epoch": 1.606818181818182,
+      "grad_norm": 1.4580140766863894,
+      "learning_rate": 9.57703554863384e-07,
+      "loss": 0.0281,
+      "step": 7070
+    },
+    {
+      "epoch": 1.6070454545454544,
+      "grad_norm": 1.5532290618391158,
+      "learning_rate": 9.576279979852429e-07,
+      "loss": 0.0285,
+      "step": 7071
+    },
+    {
+      "epoch": 1.6072727272727274,
+      "grad_norm": 1.4978904303057445,
+      "learning_rate": 9.575524343242366e-07,
+      "loss": 0.0189,
+      "step": 7072
+    },
+    {
+      "epoch": 1.6075,
+      "grad_norm": 3.773305010662702,
+      "learning_rate": 9.57476863881906e-07,
+      "loss": 0.0945,
+      "step": 7073
+    },
+    {
+      "epoch": 1.607727272727273,
+      "grad_norm": 2.1929523928640386,
+      "learning_rate": 9.574012866597918e-07,
+      "loss": 0.0408,
+      "step": 7074
+    },
+    {
+      "epoch": 1.6079545454545454,
+      "grad_norm": 1.7068290809584246,
+      "learning_rate": 9.573257026594357e-07,
+      "loss": 0.0626,
+      "step": 7075
+    },
+    {
+      "epoch": 1.6081818181818182,
+      "grad_norm": 1.3863283146199201,
+      "learning_rate": 9.572501118823787e-07,
+      "loss": 0.0455,
+      "step": 7076
+    },
+    {
+      "epoch": 1.608409090909091,
+      "grad_norm": 1.9416534999415143,
+      "learning_rate": 9.571745143301624e-07,
+      "loss": 0.0268,
+      "step": 7077
+    },
+    {
+      "epoch": 1.6086363636363636,
+      "grad_norm": 1.13386244792801,
+      "learning_rate": 9.57098910004328e-07,
+      "loss": 0.0249,
+      "step": 7078
+    },
+    {
+      "epoch": 1.6088636363636364,
+      "grad_norm": 1.6276349437403683,
+      "learning_rate": 9.570232989064177e-07,
+      "loss": 0.0297,
+      "step": 7079
+    },
+    {
+      "epoch": 1.6090909090909091,
+      "grad_norm": 0.9700036354237241,
+      "learning_rate": 9.56947681037973e-07,
+      "loss": 0.0205,
+      "step": 7080
+    },
+    {
+      "epoch": 1.6093181818181819,
+      "grad_norm": 2.794337639658697,
+      "learning_rate": 9.56872056400536e-07,
+      "loss": 0.0869,
+      "step": 7081
+    },
+    {
+      "epoch": 1.6095454545454544,
+      "grad_norm": 2.1304972703937057,
+      "learning_rate": 9.567964249956487e-07,
+      "loss": 0.0562,
+      "step": 7082
+    },
+    {
+      "epoch": 1.6097727272727274,
+      "grad_norm": 1.9330617879515541,
+      "learning_rate": 9.567207868248535e-07,
+      "loss": 0.0296,
+      "step": 7083
+    },
+    {
+      "epoch": 1.6099999999999999,
+      "grad_norm": 2.223389576271924,
+      "learning_rate": 9.566451418896927e-07,
+      "loss": 0.0428,
+      "step": 7084
+    },
+    {
+      "epoch": 1.6102272727272728,
+      "grad_norm": 1.9653537824910907,
+      "learning_rate": 9.56569490191709e-07,
+      "loss": 0.0285,
+      "step": 7085
+    },
+    {
+      "epoch": 1.6104545454545454,
+      "grad_norm": 2.076981197734736,
+      "learning_rate": 9.564938317324447e-07,
+      "loss": 0.0476,
+      "step": 7086
+    },
+    {
+      "epoch": 1.6106818181818183,
+      "grad_norm": 1.5265316376468734,
+      "learning_rate": 9.564181665134429e-07,
+      "loss": 0.0379,
+      "step": 7087
+    },
+    {
+      "epoch": 1.6109090909090908,
+      "grad_norm": 1.4101576226844277,
+      "learning_rate": 9.563424945362469e-07,
+      "loss": 0.0518,
+      "step": 7088
+    },
+    {
+      "epoch": 1.6111363636363636,
+      "grad_norm": 2.196462204193946,
+      "learning_rate": 9.562668158023988e-07,
+      "loss": 0.0721,
+      "step": 7089
+    },
+    {
+      "epoch": 1.6113636363636363,
+      "grad_norm": 2.6058021104269424,
+      "learning_rate": 9.561911303134427e-07,
+      "loss": 0.0915,
+      "step": 7090
+    },
+    {
+      "epoch": 1.611590909090909,
+      "grad_norm": 2.284706319088539,
+      "learning_rate": 9.561154380709218e-07,
+      "loss": 0.0476,
+      "step": 7091
+    },
+    {
+      "epoch": 1.6118181818181818,
+      "grad_norm": 2.4595789942493362,
+      "learning_rate": 9.56039739076379e-07,
+      "loss": 0.0496,
+      "step": 7092
+    },
+    {
+      "epoch": 1.6120454545454546,
+      "grad_norm": 2.271106493860753,
+      "learning_rate": 9.559640333313588e-07,
+      "loss": 0.037,
+      "step": 7093
+    },
+    {
+      "epoch": 1.6122727272727273,
+      "grad_norm": 3.030593227171128,
+      "learning_rate": 9.558883208374043e-07,
+      "loss": 0.1203,
+      "step": 7094
+    },
+    {
+      "epoch": 1.6125,
+      "grad_norm": 0.9509165589321237,
+      "learning_rate": 9.558126015960597e-07,
+      "loss": 0.0293,
+      "step": 7095
+    },
+    {
+      "epoch": 1.6127272727272728,
+      "grad_norm": 2.2490172759103078,
+      "learning_rate": 9.557368756088691e-07,
+      "loss": 0.0758,
+      "step": 7096
+    },
+    {
+      "epoch": 1.6129545454545453,
+      "grad_norm": 0.4763980114947887,
+      "learning_rate": 9.556611428773767e-07,
+      "loss": 0.0136,
+      "step": 7097
+    },
+    {
+      "epoch": 1.6131818181818183,
+      "grad_norm": 1.6013891350868983,
+      "learning_rate": 9.555854034031268e-07,
+      "loss": 0.0599,
+      "step": 7098
+    },
+    {
+      "epoch": 1.6134090909090908,
+      "grad_norm": 1.932484836077564,
+      "learning_rate": 9.555096571876635e-07,
+      "loss": 0.0743,
+      "step": 7099
+    },
+    {
+      "epoch": 1.6136363636363638,
+      "grad_norm": 1.5593310584365039,
+      "learning_rate": 9.554339042325319e-07,
+      "loss": 0.042,
+      "step": 7100
+    },
+    {
+      "epoch": 1.6138636363636363,
+      "grad_norm": 1.9725781596620136,
+      "learning_rate": 9.553581445392762e-07,
+      "loss": 0.0397,
+      "step": 7101
+    },
+    {
+      "epoch": 1.6140909090909092,
+      "grad_norm": 1.908423470656671,
+      "learning_rate": 9.55282378109442e-07,
+      "loss": 0.0471,
+      "step": 7102
+    },
+    {
+      "epoch": 1.6143181818181818,
+      "grad_norm": 0.7215840291926634,
+      "learning_rate": 9.552066049445737e-07,
+      "loss": 0.0188,
+      "step": 7103
+    },
+    {
+      "epoch": 1.6145454545454545,
+      "grad_norm": 1.9899909634813406,
+      "learning_rate": 9.551308250462167e-07,
+      "loss": 0.0487,
+      "step": 7104
+    },
+    {
+      "epoch": 1.6147727272727272,
+      "grad_norm": 1.8662922521907108,
+      "learning_rate": 9.550550384159162e-07,
+      "loss": 0.0639,
+      "step": 7105
+    },
+    {
+      "epoch": 1.615,
+      "grad_norm": 2.18120356279244,
+      "learning_rate": 9.549792450552176e-07,
+      "loss": 0.0459,
+      "step": 7106
+    },
+    {
+      "epoch": 1.6152272727272727,
+      "grad_norm": 1.8006767487244186,
+      "learning_rate": 9.549034449656666e-07,
+      "loss": 0.0499,
+      "step": 7107
+    },
+    {
+      "epoch": 1.6154545454545455,
+      "grad_norm": 3.614170523661449,
+      "learning_rate": 9.548276381488087e-07,
+      "loss": 0.1109,
+      "step": 7108
+    },
+    {
+      "epoch": 1.6156818181818182,
+      "grad_norm": 2.5272890491780258,
+      "learning_rate": 9.547518246061898e-07,
+      "loss": 0.0617,
+      "step": 7109
+    },
+    {
+      "epoch": 1.615909090909091,
+      "grad_norm": 1.8171473765207873,
+      "learning_rate": 9.546760043393563e-07,
+      "loss": 0.0285,
+      "step": 7110
+    },
+    {
+      "epoch": 1.6161363636363637,
+      "grad_norm": 1.6245202875307403,
+      "learning_rate": 9.546001773498537e-07,
+      "loss": 0.049,
+      "step": 7111
+    },
+    {
+      "epoch": 1.6163636363636362,
+      "grad_norm": 1.5473075113533221,
+      "learning_rate": 9.545243436392283e-07,
+      "loss": 0.0436,
+      "step": 7112
+    },
+    {
+      "epoch": 1.6165909090909092,
+      "grad_norm": 1.7074114814360732,
+      "learning_rate": 9.544485032090268e-07,
+      "loss": 0.041,
+      "step": 7113
+    },
+    {
+      "epoch": 1.6168181818181817,
+      "grad_norm": 2.070948052437633,
+      "learning_rate": 9.543726560607954e-07,
+      "loss": 0.0656,
+      "step": 7114
+    },
+    {
+      "epoch": 1.6170454545454547,
+      "grad_norm": 1.269688317967587,
+      "learning_rate": 9.54296802196081e-07,
+      "loss": 0.0353,
+      "step": 7115
+    },
+    {
+      "epoch": 1.6172727272727272,
+      "grad_norm": 2.4266855303774446,
+      "learning_rate": 9.542209416164305e-07,
+      "loss": 0.0823,
+      "step": 7116
+    },
+    {
+      "epoch": 1.6175000000000002,
+      "grad_norm": 1.7536186685300652,
+      "learning_rate": 9.541450743233903e-07,
+      "loss": 0.055,
+      "step": 7117
+    },
+    {
+      "epoch": 1.6177272727272727,
+      "grad_norm": 1.3826227447528292,
+      "learning_rate": 9.54069200318508e-07,
+      "loss": 0.0352,
+      "step": 7118
+    },
+    {
+      "epoch": 1.6179545454545454,
+      "grad_norm": 2.516382901734116,
+      "learning_rate": 9.539933196033309e-07,
+      "loss": 0.0594,
+      "step": 7119
+    },
+    {
+      "epoch": 1.6181818181818182,
+      "grad_norm": 2.2326139927927104,
+      "learning_rate": 9.539174321794058e-07,
+      "loss": 0.064,
+      "step": 7120
+    },
+    {
+      "epoch": 1.618409090909091,
+      "grad_norm": 1.4872874077105194,
+      "learning_rate": 9.538415380482804e-07,
+      "loss": 0.0304,
+      "step": 7121
+    },
+    {
+      "epoch": 1.6186363636363637,
+      "grad_norm": 1.368579183547115,
+      "learning_rate": 9.537656372115023e-07,
+      "loss": 0.0149,
+      "step": 7122
+    },
+    {
+      "epoch": 1.6188636363636364,
+      "grad_norm": 1.4316263407301535,
+      "learning_rate": 9.536897296706193e-07,
+      "loss": 0.044,
+      "step": 7123
+    },
+    {
+      "epoch": 1.6190909090909091,
+      "grad_norm": 1.6294037091149727,
+      "learning_rate": 9.536138154271793e-07,
+      "loss": 0.0816,
+      "step": 7124
+    },
+    {
+      "epoch": 1.6193181818181817,
+      "grad_norm": 1.512060574134086,
+      "learning_rate": 9.535378944827303e-07,
+      "loss": 0.0445,
+      "step": 7125
+    },
+    {
+      "epoch": 1.6195454545454546,
+      "grad_norm": 1.1411928143482328,
+      "learning_rate": 9.534619668388206e-07,
+      "loss": 0.0536,
+      "step": 7126
+    },
+    {
+      "epoch": 1.6197727272727271,
+      "grad_norm": 1.1721626391944626,
+      "learning_rate": 9.533860324969983e-07,
+      "loss": 0.0184,
+      "step": 7127
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 2.8134247055389965,
+      "learning_rate": 9.533100914588119e-07,
+      "loss": 0.0527,
+      "step": 7128
+    },
+    {
+      "epoch": 1.6202272727272726,
+      "grad_norm": 1.5972751682943231,
+      "learning_rate": 9.532341437258099e-07,
+      "loss": 0.031,
+      "step": 7129
+    },
+    {
+      "epoch": 1.6204545454545456,
+      "grad_norm": 1.7252859535158258,
+      "learning_rate": 9.531581892995412e-07,
+      "loss": 0.0682,
+      "step": 7130
+    },
+    {
+      "epoch": 1.6206818181818181,
+      "grad_norm": 1.614779007245336,
+      "learning_rate": 9.530822281815543e-07,
+      "loss": 0.0457,
+      "step": 7131
+    },
+    {
+      "epoch": 1.6209090909090909,
+      "grad_norm": 2.7165889350056047,
+      "learning_rate": 9.530062603733986e-07,
+      "loss": 0.0452,
+      "step": 7132
+    },
+    {
+      "epoch": 1.6211363636363636,
+      "grad_norm": 2.824926880478656,
+      "learning_rate": 9.529302858766229e-07,
+      "loss": 0.0765,
+      "step": 7133
+    },
+    {
+      "epoch": 1.6213636363636363,
+      "grad_norm": 2.3417338879584966,
+      "learning_rate": 9.528543046927766e-07,
+      "loss": 0.0767,
+      "step": 7134
+    },
+    {
+      "epoch": 1.621590909090909,
+      "grad_norm": 2.706442257716524,
+      "learning_rate": 9.527783168234091e-07,
+      "loss": 0.0855,
+      "step": 7135
+    },
+    {
+      "epoch": 1.6218181818181818,
+      "grad_norm": 2.951804571308263,
+      "learning_rate": 9.527023222700697e-07,
+      "loss": 0.0793,
+      "step": 7136
+    },
+    {
+      "epoch": 1.6220454545454546,
+      "grad_norm": 1.8665996096420145,
+      "learning_rate": 9.526263210343084e-07,
+      "loss": 0.0353,
+      "step": 7137
+    },
+    {
+      "epoch": 1.6222727272727273,
+      "grad_norm": 1.4068936820901912,
+      "learning_rate": 9.525503131176749e-07,
+      "loss": 0.0362,
+      "step": 7138
+    },
+    {
+      "epoch": 1.6225,
+      "grad_norm": 0.7956410806587328,
+      "learning_rate": 9.524742985217189e-07,
+      "loss": 0.0156,
+      "step": 7139
+    },
+    {
+      "epoch": 1.6227272727272726,
+      "grad_norm": 1.74676579356523,
+      "learning_rate": 9.523982772479906e-07,
+      "loss": 0.0494,
+      "step": 7140
+    },
+    {
+      "epoch": 1.6229545454545455,
+      "grad_norm": 1.6783142090321146,
+      "learning_rate": 9.523222492980405e-07,
+      "loss": 0.0505,
+      "step": 7141
+    },
+    {
+      "epoch": 1.623181818181818,
+      "grad_norm": 1.3948939461353989,
+      "learning_rate": 9.522462146734185e-07,
+      "loss": 0.043,
+      "step": 7142
+    },
+    {
+      "epoch": 1.623409090909091,
+      "grad_norm": 2.527044895787894,
+      "learning_rate": 9.521701733756755e-07,
+      "loss": 0.0816,
+      "step": 7143
+    },
+    {
+      "epoch": 1.6236363636363635,
+      "grad_norm": 2.182537026436156,
+      "learning_rate": 9.520941254063617e-07,
+      "loss": 0.0799,
+      "step": 7144
+    },
+    {
+      "epoch": 1.6238636363636365,
+      "grad_norm": 1.3454451334828743,
+      "learning_rate": 9.520180707670279e-07,
+      "loss": 0.0251,
+      "step": 7145
+    },
+    {
+      "epoch": 1.624090909090909,
+      "grad_norm": 3.5471408108666775,
+      "learning_rate": 9.519420094592252e-07,
+      "loss": 0.1171,
+      "step": 7146
+    },
+    {
+      "epoch": 1.6243181818181818,
+      "grad_norm": 1.150860613185303,
+      "learning_rate": 9.518659414845047e-07,
+      "loss": 0.0305,
+      "step": 7147
+    },
+    {
+      "epoch": 1.6245454545454545,
+      "grad_norm": 1.1077637217991052,
+      "learning_rate": 9.517898668444174e-07,
+      "loss": 0.0246,
+      "step": 7148
+    },
+    {
+      "epoch": 1.6247727272727273,
+      "grad_norm": 2.1916332360149227,
+      "learning_rate": 9.517137855405145e-07,
+      "loss": 0.0562,
+      "step": 7149
+    },
+    {
+      "epoch": 1.625,
+      "grad_norm": 2.5475189109119705,
+      "learning_rate": 9.516376975743477e-07,
+      "loss": 0.0588,
+      "step": 7150
+    },
+    {
+      "epoch": 1.6252272727272727,
+      "grad_norm": 1.8953505414434642,
+      "learning_rate": 9.51561602947468e-07,
+      "loss": 0.0515,
+      "step": 7151
+    },
+    {
+      "epoch": 1.6254545454545455,
+      "grad_norm": 1.9009725327545872,
+      "learning_rate": 9.51485501661428e-07,
+      "loss": 0.0348,
+      "step": 7152
+    },
+    {
+      "epoch": 1.6256818181818182,
+      "grad_norm": 1.6500851308176623,
+      "learning_rate": 9.514093937177788e-07,
+      "loss": 0.0297,
+      "step": 7153
+    },
+    {
+      "epoch": 1.625909090909091,
+      "grad_norm": 2.4379085957691564,
+      "learning_rate": 9.513332791180724e-07,
+      "loss": 0.0794,
+      "step": 7154
+    },
+    {
+      "epoch": 1.6261363636363635,
+      "grad_norm": 3.7246125212047527,
+      "learning_rate": 9.512571578638613e-07,
+      "loss": 0.081,
+      "step": 7155
+    },
+    {
+      "epoch": 1.6263636363636365,
+      "grad_norm": 1.6313460145750327,
+      "learning_rate": 9.511810299566976e-07,
+      "loss": 0.0225,
+      "step": 7156
+    },
+    {
+      "epoch": 1.626590909090909,
+      "grad_norm": 1.6635739052475564,
+      "learning_rate": 9.511048953981336e-07,
+      "loss": 0.0317,
+      "step": 7157
+    },
+    {
+      "epoch": 1.626818181818182,
+      "grad_norm": 2.389377655453193,
+      "learning_rate": 9.51028754189722e-07,
+      "loss": 0.0846,
+      "step": 7158
+    },
+    {
+      "epoch": 1.6270454545454545,
+      "grad_norm": 2.262419834421432,
+      "learning_rate": 9.509526063330151e-07,
+      "loss": 0.0719,
+      "step": 7159
+    },
+    {
+      "epoch": 1.6272727272727274,
+      "grad_norm": 2.4878809074183104,
+      "learning_rate": 9.50876451829566e-07,
+      "loss": 0.0604,
+      "step": 7160
+    },
+    {
+      "epoch": 1.6275,
+      "grad_norm": 1.945703827777146,
+      "learning_rate": 9.508002906809276e-07,
+      "loss": 0.0489,
+      "step": 7161
+    },
+    {
+      "epoch": 1.6277272727272727,
+      "grad_norm": 1.0597307599067247,
+      "learning_rate": 9.507241228886527e-07,
+      "loss": 0.0263,
+      "step": 7162
+    },
+    {
+      "epoch": 1.6279545454545454,
+      "grad_norm": 0.7096898394656725,
+      "learning_rate": 9.506479484542949e-07,
+      "loss": 0.0135,
+      "step": 7163
+    },
+    {
+      "epoch": 1.6281818181818182,
+      "grad_norm": 2.106885735321502,
+      "learning_rate": 9.50571767379407e-07,
+      "loss": 0.0529,
+      "step": 7164
+    },
+    {
+      "epoch": 1.628409090909091,
+      "grad_norm": 2.583720960670451,
+      "learning_rate": 9.504955796655432e-07,
+      "loss": 0.0552,
+      "step": 7165
+    },
+    {
+      "epoch": 1.6286363636363637,
+      "grad_norm": 1.994585686938468,
+      "learning_rate": 9.504193853142564e-07,
+      "loss": 0.0377,
+      "step": 7166
+    },
+    {
+      "epoch": 1.6288636363636364,
+      "grad_norm": 1.5345357008791765,
+      "learning_rate": 9.503431843271006e-07,
+      "loss": 0.0346,
+      "step": 7167
+    },
+    {
+      "epoch": 1.6290909090909091,
+      "grad_norm": 2.0289172265537605,
+      "learning_rate": 9.502669767056298e-07,
+      "loss": 0.0643,
+      "step": 7168
+    },
+    {
+      "epoch": 1.6293181818181819,
+      "grad_norm": 0.9808396428633941,
+      "learning_rate": 9.501907624513978e-07,
+      "loss": 0.0195,
+      "step": 7169
+    },
+    {
+      "epoch": 1.6295454545454544,
+      "grad_norm": 3.2739146996827633,
+      "learning_rate": 9.501145415659588e-07,
+      "loss": 0.1171,
+      "step": 7170
+    },
+    {
+      "epoch": 1.6297727272727274,
+      "grad_norm": 2.260029350671522,
+      "learning_rate": 9.500383140508671e-07,
+      "loss": 0.0571,
+      "step": 7171
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.7216669381868175,
+      "learning_rate": 9.499620799076772e-07,
+      "loss": 0.0537,
+      "step": 7172
+    },
+    {
+      "epoch": 1.6302272727272729,
+      "grad_norm": 1.5865187315568803,
+      "learning_rate": 9.498858391379434e-07,
+      "loss": 0.0201,
+      "step": 7173
+    },
+    {
+      "epoch": 1.6304545454545454,
+      "grad_norm": 2.490976735218659,
+      "learning_rate": 9.49809591743221e-07,
+      "loss": 0.0639,
+      "step": 7174
+    },
+    {
+      "epoch": 1.6306818181818183,
+      "grad_norm": 2.539811767411571,
+      "learning_rate": 9.497333377250639e-07,
+      "loss": 0.1016,
+      "step": 7175
+    },
+    {
+      "epoch": 1.6309090909090909,
+      "grad_norm": 2.6434212243532755,
+      "learning_rate": 9.496570770850274e-07,
+      "loss": 0.0774,
+      "step": 7176
+    },
+    {
+      "epoch": 1.6311363636363636,
+      "grad_norm": 1.3534638353084532,
+      "learning_rate": 9.49580809824667e-07,
+      "loss": 0.0361,
+      "step": 7177
+    },
+    {
+      "epoch": 1.6313636363636363,
+      "grad_norm": 1.580308710016567,
+      "learning_rate": 9.495045359455377e-07,
+      "loss": 0.0502,
+      "step": 7178
+    },
+    {
+      "epoch": 1.631590909090909,
+      "grad_norm": 1.2061293167356977,
+      "learning_rate": 9.494282554491946e-07,
+      "loss": 0.0333,
+      "step": 7179
+    },
+    {
+      "epoch": 1.6318181818181818,
+      "grad_norm": 2.1533355249749535,
+      "learning_rate": 9.493519683371936e-07,
+      "loss": 0.0338,
+      "step": 7180
+    },
+    {
+      "epoch": 1.6320454545454546,
+      "grad_norm": 1.6848768934793539,
+      "learning_rate": 9.492756746110898e-07,
+      "loss": 0.0359,
+      "step": 7181
+    },
+    {
+      "epoch": 1.6322727272727273,
+      "grad_norm": 1.0623337528161858,
+      "learning_rate": 9.491993742724394e-07,
+      "loss": 0.0325,
+      "step": 7182
+    },
+    {
+      "epoch": 1.6324999999999998,
+      "grad_norm": 1.8600267189282638,
+      "learning_rate": 9.491230673227983e-07,
+      "loss": 0.0388,
+      "step": 7183
+    },
+    {
+      "epoch": 1.6327272727272728,
+      "grad_norm": 2.2966040383059165,
+      "learning_rate": 9.490467537637221e-07,
+      "loss": 0.0541,
+      "step": 7184
+    },
+    {
+      "epoch": 1.6329545454545453,
+      "grad_norm": 1.7733533597106637,
+      "learning_rate": 9.489704335967674e-07,
+      "loss": 0.0535,
+      "step": 7185
+    },
+    {
+      "epoch": 1.6331818181818183,
+      "grad_norm": 1.2080877019687903,
+      "learning_rate": 9.488941068234904e-07,
+      "loss": 0.0299,
+      "step": 7186
+    },
+    {
+      "epoch": 1.6334090909090908,
+      "grad_norm": 1.7972144716676748,
+      "learning_rate": 9.488177734454477e-07,
+      "loss": 0.0337,
+      "step": 7187
+    },
+    {
+      "epoch": 1.6336363636363638,
+      "grad_norm": 2.139717699918852,
+      "learning_rate": 9.487414334641952e-07,
+      "loss": 0.0815,
+      "step": 7188
+    },
+    {
+      "epoch": 1.6338636363636363,
+      "grad_norm": 1.078454532036465,
+      "learning_rate": 9.486650868812905e-07,
+      "loss": 0.024,
+      "step": 7189
+    },
+    {
+      "epoch": 1.634090909090909,
+      "grad_norm": 2.1254986087835155,
+      "learning_rate": 9.485887336982896e-07,
+      "loss": 0.0571,
+      "step": 7190
+    },
+    {
+      "epoch": 1.6343181818181818,
+      "grad_norm": 2.2268546219554057,
+      "learning_rate": 9.485123739167501e-07,
+      "loss": 0.0976,
+      "step": 7191
+    },
+    {
+      "epoch": 1.6345454545454545,
+      "grad_norm": 1.929303130770742,
+      "learning_rate": 9.48436007538229e-07,
+      "loss": 0.0487,
+      "step": 7192
+    },
+    {
+      "epoch": 1.6347727272727273,
+      "grad_norm": 2.596779251457908,
+      "learning_rate": 9.483596345642833e-07,
+      "loss": 0.0743,
+      "step": 7193
+    },
+    {
+      "epoch": 1.635,
+      "grad_norm": 2.185906624930221,
+      "learning_rate": 9.482832549964708e-07,
+      "loss": 0.0706,
+      "step": 7194
+    },
+    {
+      "epoch": 1.6352272727272728,
+      "grad_norm": 3.7845646356337497,
+      "learning_rate": 9.482068688363483e-07,
+      "loss": 0.1289,
+      "step": 7195
+    },
+    {
+      "epoch": 1.6354545454545455,
+      "grad_norm": 2.8452349126116263,
+      "learning_rate": 9.481304760854743e-07,
+      "loss": 0.049,
+      "step": 7196
+    },
+    {
+      "epoch": 1.6356818181818182,
+      "grad_norm": 2.1940721545718915,
+      "learning_rate": 9.480540767454058e-07,
+      "loss": 0.0587,
+      "step": 7197
+    },
+    {
+      "epoch": 1.6359090909090908,
+      "grad_norm": 2.4339992671509294,
+      "learning_rate": 9.479776708177013e-07,
+      "loss": 0.0559,
+      "step": 7198
+    },
+    {
+      "epoch": 1.6361363636363637,
+      "grad_norm": 1.3659779117656714,
+      "learning_rate": 9.479012583039184e-07,
+      "loss": 0.0377,
+      "step": 7199
+    },
+    {
+      "epoch": 1.6363636363636362,
+      "grad_norm": 1.3960309007404041,
+      "learning_rate": 9.478248392056158e-07,
+      "loss": 0.0539,
+      "step": 7200
+    },
+    {
+      "epoch": 1.6365909090909092,
+      "grad_norm": 3.5267451218435237,
+      "learning_rate": 9.477484135243513e-07,
+      "loss": 0.1077,
+      "step": 7201
+    },
+    {
+      "epoch": 1.6368181818181817,
+      "grad_norm": 1.895467471769694,
+      "learning_rate": 9.476719812616836e-07,
+      "loss": 0.0396,
+      "step": 7202
+    },
+    {
+      "epoch": 1.6370454545454547,
+      "grad_norm": 2.1097544881579346,
+      "learning_rate": 9.475955424191715e-07,
+      "loss": 0.0582,
+      "step": 7203
+    },
+    {
+      "epoch": 1.6372727272727272,
+      "grad_norm": 1.6731034448491104,
+      "learning_rate": 9.475190969983733e-07,
+      "loss": 0.0662,
+      "step": 7204
+    },
+    {
+      "epoch": 1.6375,
+      "grad_norm": 2.7502840723435775,
+      "learning_rate": 9.474426450008481e-07,
+      "loss": 0.074,
+      "step": 7205
+    },
+    {
+      "epoch": 1.6377272727272727,
+      "grad_norm": 2.351984372037657,
+      "learning_rate": 9.473661864281547e-07,
+      "loss": 0.0712,
+      "step": 7206
+    },
+    {
+      "epoch": 1.6379545454545454,
+      "grad_norm": 2.335913013585984,
+      "learning_rate": 9.472897212818524e-07,
+      "loss": 0.0312,
+      "step": 7207
+    },
+    {
+      "epoch": 1.6381818181818182,
+      "grad_norm": 3.0165726855444785,
+      "learning_rate": 9.472132495635005e-07,
+      "loss": 0.0735,
+      "step": 7208
+    },
+    {
+      "epoch": 1.638409090909091,
+      "grad_norm": 2.407482195169821,
+      "learning_rate": 9.471367712746585e-07,
+      "loss": 0.0603,
+      "step": 7209
+    },
+    {
+      "epoch": 1.6386363636363637,
+      "grad_norm": 1.3307317904049354,
+      "learning_rate": 9.470602864168854e-07,
+      "loss": 0.0613,
+      "step": 7210
+    },
+    {
+      "epoch": 1.6388636363636364,
+      "grad_norm": 1.3267256878157243,
+      "learning_rate": 9.469837949917415e-07,
+      "loss": 0.02,
+      "step": 7211
+    },
+    {
+      "epoch": 1.6390909090909092,
+      "grad_norm": 1.365346217036175,
+      "learning_rate": 9.469072970007861e-07,
+      "loss": 0.0421,
+      "step": 7212
+    },
+    {
+      "epoch": 1.6393181818181817,
+      "grad_norm": 2.576923899702594,
+      "learning_rate": 9.468307924455792e-07,
+      "loss": 0.0414,
+      "step": 7213
+    },
+    {
+      "epoch": 1.6395454545454546,
+      "grad_norm": 1.9513000383484858,
+      "learning_rate": 9.467542813276812e-07,
+      "loss": 0.0714,
+      "step": 7214
+    },
+    {
+      "epoch": 1.6397727272727272,
+      "grad_norm": 1.376551718712285,
+      "learning_rate": 9.466777636486518e-07,
+      "loss": 0.0503,
+      "step": 7215
+    },
+    {
+      "epoch": 1.6400000000000001,
+      "grad_norm": 2.0807689977346673,
+      "learning_rate": 9.466012394100519e-07,
+      "loss": 0.0388,
+      "step": 7216
+    },
+    {
+      "epoch": 1.6402272727272726,
+      "grad_norm": 1.884782066503792,
+      "learning_rate": 9.465247086134416e-07,
+      "loss": 0.0475,
+      "step": 7217
+    },
+    {
+      "epoch": 1.6404545454545456,
+      "grad_norm": 2.543168961955036,
+      "learning_rate": 9.464481712603815e-07,
+      "loss": 0.0815,
+      "step": 7218
+    },
+    {
+      "epoch": 1.6406818181818181,
+      "grad_norm": 1.567536672146352,
+      "learning_rate": 9.463716273524325e-07,
+      "loss": 0.0502,
+      "step": 7219
+    },
+    {
+      "epoch": 1.6409090909090909,
+      "grad_norm": 1.3828377899299424,
+      "learning_rate": 9.462950768911553e-07,
+      "loss": 0.0496,
+      "step": 7220
+    },
+    {
+      "epoch": 1.6411363636363636,
+      "grad_norm": 3.806034866547095,
+      "learning_rate": 9.46218519878111e-07,
+      "loss": 0.0485,
+      "step": 7221
+    },
+    {
+      "epoch": 1.6413636363636364,
+      "grad_norm": 1.1218200152875641,
+      "learning_rate": 9.461419563148605e-07,
+      "loss": 0.0163,
+      "step": 7222
+    },
+    {
+      "epoch": 1.641590909090909,
+      "grad_norm": 1.2193575853543897,
+      "learning_rate": 9.460653862029653e-07,
+      "loss": 0.0353,
+      "step": 7223
+    },
+    {
+      "epoch": 1.6418181818181818,
+      "grad_norm": 0.9193324641746788,
+      "learning_rate": 9.459888095439868e-07,
+      "loss": 0.0148,
+      "step": 7224
+    },
+    {
+      "epoch": 1.6420454545454546,
+      "grad_norm": 2.1449536350657192,
+      "learning_rate": 9.459122263394866e-07,
+      "loss": 0.0557,
+      "step": 7225
+    },
+    {
+      "epoch": 1.642272727272727,
+      "grad_norm": 3.1324562857013687,
+      "learning_rate": 9.458356365910259e-07,
+      "loss": 0.062,
+      "step": 7226
+    },
+    {
+      "epoch": 1.6425,
+      "grad_norm": 2.0322816965439237,
+      "learning_rate": 9.457590403001671e-07,
+      "loss": 0.0761,
+      "step": 7227
+    },
+    {
+      "epoch": 1.6427272727272726,
+      "grad_norm": 2.086380068936033,
+      "learning_rate": 9.456824374684718e-07,
+      "loss": 0.0531,
+      "step": 7228
+    },
+    {
+      "epoch": 1.6429545454545456,
+      "grad_norm": 1.8150417861457833,
+      "learning_rate": 9.456058280975022e-07,
+      "loss": 0.0388,
+      "step": 7229
+    },
+    {
+      "epoch": 1.643181818181818,
+      "grad_norm": 2.3550465543847983,
+      "learning_rate": 9.455292121888204e-07,
+      "loss": 0.0302,
+      "step": 7230
+    },
+    {
+      "epoch": 1.643409090909091,
+      "grad_norm": 2.2632538555042,
+      "learning_rate": 9.454525897439888e-07,
+      "loss": 0.0697,
+      "step": 7231
+    },
+    {
+      "epoch": 1.6436363636363636,
+      "grad_norm": 2.8698628311551895,
+      "learning_rate": 9.453759607645698e-07,
+      "loss": 0.0824,
+      "step": 7232
+    },
+    {
+      "epoch": 1.6438636363636365,
+      "grad_norm": 2.3530629201717956,
+      "learning_rate": 9.452993252521261e-07,
+      "loss": 0.0409,
+      "step": 7233
+    },
+    {
+      "epoch": 1.644090909090909,
+      "grad_norm": 1.0649979908813227,
+      "learning_rate": 9.452226832082203e-07,
+      "loss": 0.0301,
+      "step": 7234
+    },
+    {
+      "epoch": 1.6443181818181818,
+      "grad_norm": 1.7231541808411115,
+      "learning_rate": 9.451460346344152e-07,
+      "loss": 0.0309,
+      "step": 7235
+    },
+    {
+      "epoch": 1.6445454545454545,
+      "grad_norm": 1.5313982803385777,
+      "learning_rate": 9.450693795322742e-07,
+      "loss": 0.0404,
+      "step": 7236
+    },
+    {
+      "epoch": 1.6447727272727273,
+      "grad_norm": 2.304214874577828,
+      "learning_rate": 9.449927179033599e-07,
+      "loss": 0.0844,
+      "step": 7237
+    },
+    {
+      "epoch": 1.645,
+      "grad_norm": 1.6614132341495451,
+      "learning_rate": 9.44916049749236e-07,
+      "loss": 0.0585,
+      "step": 7238
+    },
+    {
+      "epoch": 1.6452272727272728,
+      "grad_norm": 2.9582080344038637,
+      "learning_rate": 9.448393750714657e-07,
+      "loss": 0.0885,
+      "step": 7239
+    },
+    {
+      "epoch": 1.6454545454545455,
+      "grad_norm": 1.9386916482313785,
+      "learning_rate": 9.447626938716126e-07,
+      "loss": 0.0414,
+      "step": 7240
+    },
+    {
+      "epoch": 1.645681818181818,
+      "grad_norm": 1.7832802647702939,
+      "learning_rate": 9.446860061512402e-07,
+      "loss": 0.039,
+      "step": 7241
+    },
+    {
+      "epoch": 1.645909090909091,
+      "grad_norm": 0.8837959559014347,
+      "learning_rate": 9.446093119119125e-07,
+      "loss": 0.0269,
+      "step": 7242
+    },
+    {
+      "epoch": 1.6461363636363635,
+      "grad_norm": 2.302796751164841,
+      "learning_rate": 9.445326111551932e-07,
+      "loss": 0.0417,
+      "step": 7243
+    },
+    {
+      "epoch": 1.6463636363636365,
+      "grad_norm": 1.6347693962138448,
+      "learning_rate": 9.444559038826466e-07,
+      "loss": 0.0325,
+      "step": 7244
+    },
+    {
+      "epoch": 1.646590909090909,
+      "grad_norm": 0.7713199466281704,
+      "learning_rate": 9.443791900958367e-07,
+      "loss": 0.017,
+      "step": 7245
+    },
+    {
+      "epoch": 1.646818181818182,
+      "grad_norm": 1.9188341521320469,
+      "learning_rate": 9.443024697963279e-07,
+      "loss": 0.0588,
+      "step": 7246
+    },
+    {
+      "epoch": 1.6470454545454545,
+      "grad_norm": 2.1097762854915283,
+      "learning_rate": 9.442257429856849e-07,
+      "loss": 0.0606,
+      "step": 7247
+    },
+    {
+      "epoch": 1.6472727272727272,
+      "grad_norm": 1.2873885024059522,
+      "learning_rate": 9.441490096654721e-07,
+      "loss": 0.0277,
+      "step": 7248
+    },
+    {
+      "epoch": 1.6475,
+      "grad_norm": 1.310896480870737,
+      "learning_rate": 9.440722698372541e-07,
+      "loss": 0.0327,
+      "step": 7249
+    },
+    {
+      "epoch": 1.6477272727272727,
+      "grad_norm": 1.4251103046919469,
+      "learning_rate": 9.439955235025957e-07,
+      "loss": 0.0337,
+      "step": 7250
+    },
+    {
+      "epoch": 1.6479545454545454,
+      "grad_norm": 2.5284424867921222,
+      "learning_rate": 9.439187706630625e-07,
+      "loss": 0.0625,
+      "step": 7251
+    },
+    {
+      "epoch": 1.6481818181818182,
+      "grad_norm": 2.015012034575042,
+      "learning_rate": 9.438420113202187e-07,
+      "loss": 0.0588,
+      "step": 7252
+    },
+    {
+      "epoch": 1.648409090909091,
+      "grad_norm": 1.7658050966749288,
+      "learning_rate": 9.437652454756302e-07,
+      "loss": 0.0451,
+      "step": 7253
+    },
+    {
+      "epoch": 1.6486363636363637,
+      "grad_norm": 2.824568822105477,
+      "learning_rate": 9.436884731308623e-07,
+      "loss": 0.0592,
+      "step": 7254
+    },
+    {
+      "epoch": 1.6488636363636364,
+      "grad_norm": 3.2368511446692354,
+      "learning_rate": 9.436116942874806e-07,
+      "loss": 0.0803,
+      "step": 7255
+    },
+    {
+      "epoch": 1.649090909090909,
+      "grad_norm": 1.773893387017909,
+      "learning_rate": 9.435349089470502e-07,
+      "loss": 0.032,
+      "step": 7256
+    },
+    {
+      "epoch": 1.649318181818182,
+      "grad_norm": 2.119255211940223,
+      "learning_rate": 9.434581171111376e-07,
+      "loss": 0.0295,
+      "step": 7257
+    },
+    {
+      "epoch": 1.6495454545454544,
+      "grad_norm": 1.3003972781391355,
+      "learning_rate": 9.433813187813084e-07,
+      "loss": 0.0349,
+      "step": 7258
+    },
+    {
+      "epoch": 1.6497727272727274,
+      "grad_norm": 2.9968111925498553,
+      "learning_rate": 9.433045139591284e-07,
+      "loss": 0.0762,
+      "step": 7259
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 2.415733627336174,
+      "learning_rate": 9.432277026461643e-07,
+      "loss": 0.0655,
+      "step": 7260
+    },
+    {
+      "epoch": 1.6502272727272729,
+      "grad_norm": 2.109039454150422,
+      "learning_rate": 9.431508848439822e-07,
+      "loss": 0.0591,
+      "step": 7261
+    },
+    {
+      "epoch": 1.6504545454545454,
+      "grad_norm": 1.931937732095629,
+      "learning_rate": 9.430740605541483e-07,
+      "loss": 0.0765,
+      "step": 7262
+    },
+    {
+      "epoch": 1.6506818181818181,
+      "grad_norm": 1.0001025868795805,
+      "learning_rate": 9.429972297782295e-07,
+      "loss": 0.0252,
+      "step": 7263
+    },
+    {
+      "epoch": 1.6509090909090909,
+      "grad_norm": 0.8962415099643066,
+      "learning_rate": 9.429203925177925e-07,
+      "loss": 0.028,
+      "step": 7264
+    },
+    {
+      "epoch": 1.6511363636363636,
+      "grad_norm": 1.5046254479560917,
+      "learning_rate": 9.428435487744039e-07,
+      "loss": 0.0399,
+      "step": 7265
+    },
+    {
+      "epoch": 1.6513636363636364,
+      "grad_norm": 1.7477512106594773,
+      "learning_rate": 9.427666985496311e-07,
+      "loss": 0.0402,
+      "step": 7266
+    },
+    {
+      "epoch": 1.651590909090909,
+      "grad_norm": 2.1362698404163893,
+      "learning_rate": 9.426898418450406e-07,
+      "loss": 0.0363,
+      "step": 7267
+    },
+    {
+      "epoch": 1.6518181818181819,
+      "grad_norm": 2.3142525415294086,
+      "learning_rate": 9.426129786622001e-07,
+      "loss": 0.0844,
+      "step": 7268
+    },
+    {
+      "epoch": 1.6520454545454546,
+      "grad_norm": 1.0878452667660172,
+      "learning_rate": 9.425361090026769e-07,
+      "loss": 0.0303,
+      "step": 7269
+    },
+    {
+      "epoch": 1.6522727272727273,
+      "grad_norm": 2.078069193846706,
+      "learning_rate": 9.424592328680385e-07,
+      "loss": 0.0733,
+      "step": 7270
+    },
+    {
+      "epoch": 1.6524999999999999,
+      "grad_norm": 0.6867592776339865,
+      "learning_rate": 9.423823502598526e-07,
+      "loss": 0.0227,
+      "step": 7271
+    },
+    {
+      "epoch": 1.6527272727272728,
+      "grad_norm": 1.4707455821517383,
+      "learning_rate": 9.423054611796866e-07,
+      "loss": 0.0223,
+      "step": 7272
+    },
+    {
+      "epoch": 1.6529545454545453,
+      "grad_norm": 2.625182000449637,
+      "learning_rate": 9.422285656291089e-07,
+      "loss": 0.1171,
+      "step": 7273
+    },
+    {
+      "epoch": 1.6531818181818183,
+      "grad_norm": 1.4375571320020601,
+      "learning_rate": 9.421516636096871e-07,
+      "loss": 0.0294,
+      "step": 7274
+    },
+    {
+      "epoch": 1.6534090909090908,
+      "grad_norm": 2.0134178985318596,
+      "learning_rate": 9.420747551229898e-07,
+      "loss": 0.0298,
+      "step": 7275
+    },
+    {
+      "epoch": 1.6536363636363638,
+      "grad_norm": 1.7167472638446124,
+      "learning_rate": 9.419978401705849e-07,
+      "loss": 0.054,
+      "step": 7276
+    },
+    {
+      "epoch": 1.6538636363636363,
+      "grad_norm": 2.1684321878602453,
+      "learning_rate": 9.419209187540411e-07,
+      "loss": 0.0401,
+      "step": 7277
+    },
+    {
+      "epoch": 1.654090909090909,
+      "grad_norm": 2.0779126086116975,
+      "learning_rate": 9.418439908749267e-07,
+      "loss": 0.035,
+      "step": 7278
+    },
+    {
+      "epoch": 1.6543181818181818,
+      "grad_norm": 1.661883279202413,
+      "learning_rate": 9.417670565348108e-07,
+      "loss": 0.0601,
+      "step": 7279
+    },
+    {
+      "epoch": 1.6545454545454545,
+      "grad_norm": 3.2993182731087267,
+      "learning_rate": 9.41690115735262e-07,
+      "loss": 0.0801,
+      "step": 7280
+    },
+    {
+      "epoch": 1.6547727272727273,
+      "grad_norm": 1.0697810847518392,
+      "learning_rate": 9.41613168477849e-07,
+      "loss": 0.0168,
+      "step": 7281
+    },
+    {
+      "epoch": 1.655,
+      "grad_norm": 3.077717575039108,
+      "learning_rate": 9.415362147641413e-07,
+      "loss": 0.0739,
+      "step": 7282
+    },
+    {
+      "epoch": 1.6552272727272728,
+      "grad_norm": 2.0911823896347586,
+      "learning_rate": 9.414592545957078e-07,
+      "loss": 0.0323,
+      "step": 7283
+    },
+    {
+      "epoch": 1.6554545454545453,
+      "grad_norm": 1.693228647034296,
+      "learning_rate": 9.413822879741183e-07,
+      "loss": 0.0445,
+      "step": 7284
+    },
+    {
+      "epoch": 1.6556818181818183,
+      "grad_norm": 1.457673780345477,
+      "learning_rate": 9.413053149009417e-07,
+      "loss": 0.0363,
+      "step": 7285
+    },
+    {
+      "epoch": 1.6559090909090908,
+      "grad_norm": 1.993447927830899,
+      "learning_rate": 9.412283353777482e-07,
+      "loss": 0.0463,
+      "step": 7286
+    },
+    {
+      "epoch": 1.6561363636363637,
+      "grad_norm": 2.4323478074820097,
+      "learning_rate": 9.41151349406107e-07,
+      "loss": 0.0276,
+      "step": 7287
+    },
+    {
+      "epoch": 1.6563636363636363,
+      "grad_norm": 1.3239356537010734,
+      "learning_rate": 9.410743569875883e-07,
+      "loss": 0.0506,
+      "step": 7288
+    },
+    {
+      "epoch": 1.6565909090909092,
+      "grad_norm": 1.1886444116758035,
+      "learning_rate": 9.40997358123762e-07,
+      "loss": 0.0234,
+      "step": 7289
+    },
+    {
+      "epoch": 1.6568181818181817,
+      "grad_norm": 3.1854288825537695,
+      "learning_rate": 9.409203528161984e-07,
+      "loss": 0.061,
+      "step": 7290
+    },
+    {
+      "epoch": 1.6570454545454547,
+      "grad_norm": 1.9732775185845277,
+      "learning_rate": 9.408433410664675e-07,
+      "loss": 0.0486,
+      "step": 7291
+    },
+    {
+      "epoch": 1.6572727272727272,
+      "grad_norm": 1.2765794967481123,
+      "learning_rate": 9.407663228761399e-07,
+      "loss": 0.0375,
+      "step": 7292
+    },
+    {
+      "epoch": 1.6575,
+      "grad_norm": 2.1199018109856924,
+      "learning_rate": 9.406892982467862e-07,
+      "loss": 0.0415,
+      "step": 7293
+    },
+    {
+      "epoch": 1.6577272727272727,
+      "grad_norm": 1.7767989085204314,
+      "learning_rate": 9.406122671799768e-07,
+      "loss": 0.0561,
+      "step": 7294
+    },
+    {
+      "epoch": 1.6579545454545455,
+      "grad_norm": 2.1860298751939053,
+      "learning_rate": 9.405352296772829e-07,
+      "loss": 0.0331,
+      "step": 7295
+    },
+    {
+      "epoch": 1.6581818181818182,
+      "grad_norm": 2.387716036077203,
+      "learning_rate": 9.40458185740275e-07,
+      "loss": 0.0891,
+      "step": 7296
+    },
+    {
+      "epoch": 1.658409090909091,
+      "grad_norm": 1.4011707048433553,
+      "learning_rate": 9.403811353705244e-07,
+      "loss": 0.0377,
+      "step": 7297
+    },
+    {
+      "epoch": 1.6586363636363637,
+      "grad_norm": 2.067205640797631,
+      "learning_rate": 9.40304078569602e-07,
+      "loss": 0.0707,
+      "step": 7298
+    },
+    {
+      "epoch": 1.6588636363636362,
+      "grad_norm": 1.0376164493046613,
+      "learning_rate": 9.402270153390796e-07,
+      "loss": 0.0112,
+      "step": 7299
+    },
+    {
+      "epoch": 1.6590909090909092,
+      "grad_norm": 1.7132362438767355,
+      "learning_rate": 9.401499456805285e-07,
+      "loss": 0.0592,
+      "step": 7300
+    },
+    {
+      "epoch": 1.6593181818181817,
+      "grad_norm": 1.5023847220606765,
+      "learning_rate": 9.4007286959552e-07,
+      "loss": 0.0649,
+      "step": 7301
+    },
+    {
+      "epoch": 1.6595454545454547,
+      "grad_norm": 1.4982497165313517,
+      "learning_rate": 9.399957870856261e-07,
+      "loss": 0.0411,
+      "step": 7302
+    },
+    {
+      "epoch": 1.6597727272727272,
+      "grad_norm": 1.8059044183957071,
+      "learning_rate": 9.399186981524186e-07,
+      "loss": 0.0491,
+      "step": 7303
+    },
+    {
+      "epoch": 1.6600000000000001,
+      "grad_norm": 1.4422322434076909,
+      "learning_rate": 9.398416027974694e-07,
+      "loss": 0.0258,
+      "step": 7304
+    },
+    {
+      "epoch": 1.6602272727272727,
+      "grad_norm": 1.2410147825616635,
+      "learning_rate": 9.397645010223506e-07,
+      "loss": 0.031,
+      "step": 7305
+    },
+    {
+      "epoch": 1.6604545454545454,
+      "grad_norm": 2.4385229096295755,
+      "learning_rate": 9.396873928286346e-07,
+      "loss": 0.0562,
+      "step": 7306
+    },
+    {
+      "epoch": 1.6606818181818181,
+      "grad_norm": 1.2044430332086316,
+      "learning_rate": 9.396102782178934e-07,
+      "loss": 0.0261,
+      "step": 7307
+    },
+    {
+      "epoch": 1.660909090909091,
+      "grad_norm": 1.2727848673274471,
+      "learning_rate": 9.395331571917001e-07,
+      "loss": 0.0391,
+      "step": 7308
+    },
+    {
+      "epoch": 1.6611363636363636,
+      "grad_norm": 1.36340554994059,
+      "learning_rate": 9.394560297516269e-07,
+      "loss": 0.0347,
+      "step": 7309
+    },
+    {
+      "epoch": 1.6613636363636364,
+      "grad_norm": 1.8669972407761821,
+      "learning_rate": 9.393788958992465e-07,
+      "loss": 0.0404,
+      "step": 7310
+    },
+    {
+      "epoch": 1.6615909090909091,
+      "grad_norm": 2.0901590662766796,
+      "learning_rate": 9.393017556361321e-07,
+      "loss": 0.0462,
+      "step": 7311
+    },
+    {
+      "epoch": 1.6618181818181819,
+      "grad_norm": 13.444947839845884,
+      "learning_rate": 9.392246089638565e-07,
+      "loss": 0.0541,
+      "step": 7312
+    },
+    {
+      "epoch": 1.6620454545454546,
+      "grad_norm": 4.374770242551673,
+      "learning_rate": 9.391474558839929e-07,
+      "loss": 0.0793,
+      "step": 7313
+    },
+    {
+      "epoch": 1.6622727272727271,
+      "grad_norm": 1.9671026535650966,
+      "learning_rate": 9.390702963981147e-07,
+      "loss": 0.0432,
+      "step": 7314
+    },
+    {
+      "epoch": 1.6625,
+      "grad_norm": 1.6225979859961648,
+      "learning_rate": 9.389931305077951e-07,
+      "loss": 0.032,
+      "step": 7315
+    },
+    {
+      "epoch": 1.6627272727272726,
+      "grad_norm": 1.6388938587141717,
+      "learning_rate": 9.389159582146079e-07,
+      "loss": 0.0314,
+      "step": 7316
+    },
+    {
+      "epoch": 1.6629545454545456,
+      "grad_norm": 1.365641774936706,
+      "learning_rate": 9.388387795201268e-07,
+      "loss": 0.0311,
+      "step": 7317
+    },
+    {
+      "epoch": 1.663181818181818,
+      "grad_norm": 1.5104729276532334,
+      "learning_rate": 9.387615944259252e-07,
+      "loss": 0.0377,
+      "step": 7318
+    },
+    {
+      "epoch": 1.663409090909091,
+      "grad_norm": 1.6427330178107948,
+      "learning_rate": 9.386844029335773e-07,
+      "loss": 0.043,
+      "step": 7319
+    },
+    {
+      "epoch": 1.6636363636363636,
+      "grad_norm": 2.2093462677437774,
+      "learning_rate": 9.386072050446574e-07,
+      "loss": 0.0771,
+      "step": 7320
+    },
+    {
+      "epoch": 1.6638636363636363,
+      "grad_norm": 2.1241467324417336,
+      "learning_rate": 9.385300007607392e-07,
+      "loss": 0.0804,
+      "step": 7321
+    },
+    {
+      "epoch": 1.664090909090909,
+      "grad_norm": 1.8996674467272072,
+      "learning_rate": 9.384527900833973e-07,
+      "loss": 0.0614,
+      "step": 7322
+    },
+    {
+      "epoch": 1.6643181818181818,
+      "grad_norm": 2.1967400287680388,
+      "learning_rate": 9.383755730142064e-07,
+      "loss": 0.0345,
+      "step": 7323
+    },
+    {
+      "epoch": 1.6645454545454546,
+      "grad_norm": 1.9410916171903962,
+      "learning_rate": 9.382983495547407e-07,
+      "loss": 0.0737,
+      "step": 7324
+    },
+    {
+      "epoch": 1.6647727272727273,
+      "grad_norm": 2.5222933819749467,
+      "learning_rate": 9.382211197065752e-07,
+      "loss": 0.0615,
+      "step": 7325
+    },
+    {
+      "epoch": 1.665,
+      "grad_norm": 1.4383145234794692,
+      "learning_rate": 9.381438834712846e-07,
+      "loss": 0.047,
+      "step": 7326
+    },
+    {
+      "epoch": 1.6652272727272728,
+      "grad_norm": 2.0326895727586547,
+      "learning_rate": 9.380666408504438e-07,
+      "loss": 0.0469,
+      "step": 7327
+    },
+    {
+      "epoch": 1.6654545454545455,
+      "grad_norm": 2.0374393260417367,
+      "learning_rate": 9.37989391845628e-07,
+      "loss": 0.054,
+      "step": 7328
+    },
+    {
+      "epoch": 1.665681818181818,
+      "grad_norm": 1.3188432932987642,
+      "learning_rate": 9.379121364584126e-07,
+      "loss": 0.0286,
+      "step": 7329
+    },
+    {
+      "epoch": 1.665909090909091,
+      "grad_norm": 2.5101287785977746,
+      "learning_rate": 9.378348746903727e-07,
+      "loss": 0.0387,
+      "step": 7330
+    },
+    {
+      "epoch": 1.6661363636363635,
+      "grad_norm": 1.0062131189962462,
+      "learning_rate": 9.377576065430841e-07,
+      "loss": 0.0243,
+      "step": 7331
+    },
+    {
+      "epoch": 1.6663636363636365,
+      "grad_norm": 1.4915919145810745,
+      "learning_rate": 9.376803320181221e-07,
+      "loss": 0.047,
+      "step": 7332
+    },
+    {
+      "epoch": 1.666590909090909,
+      "grad_norm": 2.5112476225383404,
+      "learning_rate": 9.376030511170628e-07,
+      "loss": 0.0405,
+      "step": 7333
+    },
+    {
+      "epoch": 1.666818181818182,
+      "grad_norm": 2.163841951953429,
+      "learning_rate": 9.375257638414818e-07,
+      "loss": 0.0565,
+      "step": 7334
+    },
+    {
+      "epoch": 1.6670454545454545,
+      "grad_norm": 2.506927639512883,
+      "learning_rate": 9.374484701929552e-07,
+      "loss": 0.0587,
+      "step": 7335
+    },
+    {
+      "epoch": 1.6672727272727272,
+      "grad_norm": 2.0519524518801124,
+      "learning_rate": 9.373711701730593e-07,
+      "loss": 0.0362,
+      "step": 7336
+    },
+    {
+      "epoch": 1.6675,
+      "grad_norm": 2.214629461170852,
+      "learning_rate": 9.3729386378337e-07,
+      "loss": 0.0569,
+      "step": 7337
+    },
+    {
+      "epoch": 1.6677272727272727,
+      "grad_norm": 0.7533021557139618,
+      "learning_rate": 9.372165510254644e-07,
+      "loss": 0.0238,
+      "step": 7338
+    },
+    {
+      "epoch": 1.6679545454545455,
+      "grad_norm": 1.9822137157438324,
+      "learning_rate": 9.371392319009184e-07,
+      "loss": 0.036,
+      "step": 7339
+    },
+    {
+      "epoch": 1.6681818181818182,
+      "grad_norm": 2.3986638196889754,
+      "learning_rate": 9.37061906411309e-07,
+      "loss": 0.0543,
+      "step": 7340
+    },
+    {
+      "epoch": 1.668409090909091,
+      "grad_norm": 1.9603646687615675,
+      "learning_rate": 9.369845745582129e-07,
+      "loss": 0.0311,
+      "step": 7341
+    },
+    {
+      "epoch": 1.6686363636363635,
+      "grad_norm": 1.8367435340509823,
+      "learning_rate": 9.36907236343207e-07,
+      "loss": 0.0472,
+      "step": 7342
+    },
+    {
+      "epoch": 1.6688636363636364,
+      "grad_norm": 2.4346115866960854,
+      "learning_rate": 9.368298917678684e-07,
+      "loss": 0.0483,
+      "step": 7343
+    },
+    {
+      "epoch": 1.669090909090909,
+      "grad_norm": 1.942935165316651,
+      "learning_rate": 9.367525408337742e-07,
+      "loss": 0.0555,
+      "step": 7344
+    },
+    {
+      "epoch": 1.669318181818182,
+      "grad_norm": 2.3481007622318635,
+      "learning_rate": 9.36675183542502e-07,
+      "loss": 0.0538,
+      "step": 7345
+    },
+    {
+      "epoch": 1.6695454545454544,
+      "grad_norm": 2.259565000189565,
+      "learning_rate": 9.365978198956289e-07,
+      "loss": 0.051,
+      "step": 7346
+    },
+    {
+      "epoch": 1.6697727272727274,
+      "grad_norm": 2.0801337154347594,
+      "learning_rate": 9.365204498947328e-07,
+      "loss": 0.0419,
+      "step": 7347
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.8968451177122612,
+      "learning_rate": 9.364430735413911e-07,
+      "loss": 0.045,
+      "step": 7348
+    },
+    {
+      "epoch": 1.670227272727273,
+      "grad_norm": 1.071507784061628,
+      "learning_rate": 9.363656908371818e-07,
+      "loss": 0.0227,
+      "step": 7349
+    },
+    {
+      "epoch": 1.6704545454545454,
+      "grad_norm": 1.4245326305398327,
+      "learning_rate": 9.36288301783683e-07,
+      "loss": 0.0286,
+      "step": 7350
+    },
+    {
+      "epoch": 1.6706818181818182,
+      "grad_norm": 2.0696158684077206,
+      "learning_rate": 9.362109063824725e-07,
+      "loss": 0.0523,
+      "step": 7351
+    },
+    {
+      "epoch": 1.670909090909091,
+      "grad_norm": 1.6331535931938375,
+      "learning_rate": 9.361335046351287e-07,
+      "loss": 0.0424,
+      "step": 7352
+    },
+    {
+      "epoch": 1.6711363636363636,
+      "grad_norm": 2.451060547772885,
+      "learning_rate": 9.3605609654323e-07,
+      "loss": 0.0809,
+      "step": 7353
+    },
+    {
+      "epoch": 1.6713636363636364,
+      "grad_norm": 1.8291630400100427,
+      "learning_rate": 9.35978682108355e-07,
+      "loss": 0.0564,
+      "step": 7354
+    },
+    {
+      "epoch": 1.6715909090909091,
+      "grad_norm": 1.207192322121505,
+      "learning_rate": 9.359012613320817e-07,
+      "loss": 0.0206,
+      "step": 7355
+    },
+    {
+      "epoch": 1.6718181818181819,
+      "grad_norm": 2.3829261197007683,
+      "learning_rate": 9.358238342159898e-07,
+      "loss": 0.0611,
+      "step": 7356
+    },
+    {
+      "epoch": 1.6720454545454544,
+      "grad_norm": 1.1193375140277018,
+      "learning_rate": 9.357464007616572e-07,
+      "loss": 0.0213,
+      "step": 7357
+    },
+    {
+      "epoch": 1.6722727272727274,
+      "grad_norm": 2.164817299171392,
+      "learning_rate": 9.356689609706635e-07,
+      "loss": 0.0417,
+      "step": 7358
+    },
+    {
+      "epoch": 1.6724999999999999,
+      "grad_norm": 1.9196722500248125,
+      "learning_rate": 9.355915148445877e-07,
+      "loss": 0.0327,
+      "step": 7359
+    },
+    {
+      "epoch": 1.6727272727272728,
+      "grad_norm": 1.7050998754184783,
+      "learning_rate": 9.355140623850091e-07,
+      "loss": 0.0271,
+      "step": 7360
+    },
+    {
+      "epoch": 1.6729545454545454,
+      "grad_norm": 1.4621955554514614,
+      "learning_rate": 9.354366035935071e-07,
+      "loss": 0.0184,
+      "step": 7361
+    },
+    {
+      "epoch": 1.6731818181818183,
+      "grad_norm": 1.6377828940638481,
+      "learning_rate": 9.353591384716611e-07,
+      "loss": 0.0307,
+      "step": 7362
+    },
+    {
+      "epoch": 1.6734090909090908,
+      "grad_norm": 1.3421844268648477,
+      "learning_rate": 9.352816670210508e-07,
+      "loss": 0.0304,
+      "step": 7363
+    },
+    {
+      "epoch": 1.6736363636363636,
+      "grad_norm": 1.824879708354052,
+      "learning_rate": 9.35204189243256e-07,
+      "loss": 0.0399,
+      "step": 7364
+    },
+    {
+      "epoch": 1.6738636363636363,
+      "grad_norm": 1.8022807716909093,
+      "learning_rate": 9.351267051398567e-07,
+      "loss": 0.0518,
+      "step": 7365
+    },
+    {
+      "epoch": 1.674090909090909,
+      "grad_norm": 2.6326359535647303,
+      "learning_rate": 9.350492147124328e-07,
+      "loss": 0.0836,
+      "step": 7366
+    },
+    {
+      "epoch": 1.6743181818181818,
+      "grad_norm": 2.140454921433608,
+      "learning_rate": 9.349717179625644e-07,
+      "loss": 0.0321,
+      "step": 7367
+    },
+    {
+      "epoch": 1.6745454545454546,
+      "grad_norm": 1.947278313314387,
+      "learning_rate": 9.348942148918319e-07,
+      "loss": 0.0649,
+      "step": 7368
+    },
+    {
+      "epoch": 1.6747727272727273,
+      "grad_norm": 1.9154538410660864,
+      "learning_rate": 9.348167055018156e-07,
+      "loss": 0.0855,
+      "step": 7369
+    },
+    {
+      "epoch": 1.675,
+      "grad_norm": 3.7483689355141454,
+      "learning_rate": 9.347391897940967e-07,
+      "loss": 0.0792,
+      "step": 7370
+    },
+    {
+      "epoch": 1.6752272727272728,
+      "grad_norm": 1.8185180191737902,
+      "learning_rate": 9.346616677702548e-07,
+      "loss": 0.0473,
+      "step": 7371
+    },
+    {
+      "epoch": 1.6754545454545453,
+      "grad_norm": 1.8972958234734039,
+      "learning_rate": 9.345841394318715e-07,
+      "loss": 0.029,
+      "step": 7372
+    },
+    {
+      "epoch": 1.6756818181818183,
+      "grad_norm": 2.997010625513006,
+      "learning_rate": 9.345066047805273e-07,
+      "loss": 0.0593,
+      "step": 7373
+    },
+    {
+      "epoch": 1.6759090909090908,
+      "grad_norm": 1.6812799386144974,
+      "learning_rate": 9.344290638178037e-07,
+      "loss": 0.0521,
+      "step": 7374
+    },
+    {
+      "epoch": 1.6761363636363638,
+      "grad_norm": 2.0470496381087977,
+      "learning_rate": 9.343515165452816e-07,
+      "loss": 0.0496,
+      "step": 7375
+    },
+    {
+      "epoch": 1.6763636363636363,
+      "grad_norm": 1.932579301077024,
+      "learning_rate": 9.342739629645423e-07,
+      "loss": 0.0885,
+      "step": 7376
+    },
+    {
+      "epoch": 1.6765909090909092,
+      "grad_norm": 1.3850111048481026,
+      "learning_rate": 9.341964030771674e-07,
+      "loss": 0.0302,
+      "step": 7377
+    },
+    {
+      "epoch": 1.6768181818181818,
+      "grad_norm": 1.4720757604069918,
+      "learning_rate": 9.341188368847384e-07,
+      "loss": 0.0294,
+      "step": 7378
+    },
+    {
+      "epoch": 1.6770454545454545,
+      "grad_norm": 2.348639321989614,
+      "learning_rate": 9.340412643888372e-07,
+      "loss": 0.0309,
+      "step": 7379
+    },
+    {
+      "epoch": 1.6772727272727272,
+      "grad_norm": 3.727984145759993,
+      "learning_rate": 9.339636855910451e-07,
+      "loss": 0.054,
+      "step": 7380
+    },
+    {
+      "epoch": 1.6775,
+      "grad_norm": 1.6037168675891338,
+      "learning_rate": 9.338861004929447e-07,
+      "loss": 0.0474,
+      "step": 7381
+    },
+    {
+      "epoch": 1.6777272727272727,
+      "grad_norm": 1.4516168681476525,
+      "learning_rate": 9.338085090961177e-07,
+      "loss": 0.0347,
+      "step": 7382
+    },
+    {
+      "epoch": 1.6779545454545455,
+      "grad_norm": 2.163044649187232,
+      "learning_rate": 9.337309114021466e-07,
+      "loss": 0.0337,
+      "step": 7383
+    },
+    {
+      "epoch": 1.6781818181818182,
+      "grad_norm": 2.5189512784582684,
+      "learning_rate": 9.336533074126134e-07,
+      "loss": 0.0475,
+      "step": 7384
+    },
+    {
+      "epoch": 1.678409090909091,
+      "grad_norm": 1.9108358529066496,
+      "learning_rate": 9.33575697129101e-07,
+      "loss": 0.0827,
+      "step": 7385
+    },
+    {
+      "epoch": 1.6786363636363637,
+      "grad_norm": 0.6013830930401313,
+      "learning_rate": 9.334980805531918e-07,
+      "loss": 0.0174,
+      "step": 7386
+    },
+    {
+      "epoch": 1.6788636363636362,
+      "grad_norm": 3.4263146696957008,
+      "learning_rate": 9.334204576864684e-07,
+      "loss": 0.1516,
+      "step": 7387
+    },
+    {
+      "epoch": 1.6790909090909092,
+      "grad_norm": 1.9167356073014061,
+      "learning_rate": 9.333428285305139e-07,
+      "loss": 0.0477,
+      "step": 7388
+    },
+    {
+      "epoch": 1.6793181818181817,
+      "grad_norm": 1.844946141630317,
+      "learning_rate": 9.332651930869111e-07,
+      "loss": 0.0766,
+      "step": 7389
+    },
+    {
+      "epoch": 1.6795454545454547,
+      "grad_norm": 2.181686305340196,
+      "learning_rate": 9.331875513572432e-07,
+      "loss": 0.0521,
+      "step": 7390
+    },
+    {
+      "epoch": 1.6797727272727272,
+      "grad_norm": 3.8448014084773146,
+      "learning_rate": 9.331099033430935e-07,
+      "loss": 0.0997,
+      "step": 7391
+    },
+    {
+      "epoch": 1.6800000000000002,
+      "grad_norm": 3.025554031489333,
+      "learning_rate": 9.330322490460454e-07,
+      "loss": 0.0486,
+      "step": 7392
+    },
+    {
+      "epoch": 1.6802272727272727,
+      "grad_norm": 1.2372462325137352,
+      "learning_rate": 9.329545884676823e-07,
+      "loss": 0.035,
+      "step": 7393
+    },
+    {
+      "epoch": 1.6804545454545454,
+      "grad_norm": 3.44807265988987,
+      "learning_rate": 9.32876921609588e-07,
+      "loss": 0.0845,
+      "step": 7394
+    },
+    {
+      "epoch": 1.6806818181818182,
+      "grad_norm": 2.8004148023371593,
+      "learning_rate": 9.32799248473346e-07,
+      "loss": 0.0739,
+      "step": 7395
+    },
+    {
+      "epoch": 1.680909090909091,
+      "grad_norm": 1.7115712726696053,
+      "learning_rate": 9.327215690605403e-07,
+      "loss": 0.0322,
+      "step": 7396
+    },
+    {
+      "epoch": 1.6811363636363637,
+      "grad_norm": 1.5265013582176767,
+      "learning_rate": 9.32643883372755e-07,
+      "loss": 0.0663,
+      "step": 7397
+    },
+    {
+      "epoch": 1.6813636363636364,
+      "grad_norm": 2.0130623154888267,
+      "learning_rate": 9.325661914115744e-07,
+      "loss": 0.0393,
+      "step": 7398
+    },
+    {
+      "epoch": 1.6815909090909091,
+      "grad_norm": 1.4189736893834453,
+      "learning_rate": 9.324884931785823e-07,
+      "loss": 0.0325,
+      "step": 7399
+    },
+    {
+      "epoch": 1.6818181818181817,
+      "grad_norm": 2.0903866943318214,
+      "learning_rate": 9.324107886753636e-07,
+      "loss": 0.0572,
+      "step": 7400
+    },
+    {
+      "epoch": 1.6820454545454546,
+      "grad_norm": 2.2307361943176516,
+      "learning_rate": 9.323330779035025e-07,
+      "loss": 0.0665,
+      "step": 7401
+    },
+    {
+      "epoch": 1.6822727272727271,
+      "grad_norm": 2.9835688844446486,
+      "learning_rate": 9.322553608645837e-07,
+      "loss": 0.0765,
+      "step": 7402
+    },
+    {
+      "epoch": 1.6825,
+      "grad_norm": 1.9243050264578443,
+      "learning_rate": 9.321776375601923e-07,
+      "loss": 0.0374,
+      "step": 7403
+    },
+    {
+      "epoch": 1.6827272727272726,
+      "grad_norm": 2.5266907558208245,
+      "learning_rate": 9.320999079919127e-07,
+      "loss": 0.0939,
+      "step": 7404
+    },
+    {
+      "epoch": 1.6829545454545456,
+      "grad_norm": 1.8334845701153812,
+      "learning_rate": 9.320221721613302e-07,
+      "loss": 0.068,
+      "step": 7405
+    },
+    {
+      "epoch": 1.6831818181818181,
+      "grad_norm": 2.466190313310103,
+      "learning_rate": 9.319444300700301e-07,
+      "loss": 0.0461,
+      "step": 7406
+    },
+    {
+      "epoch": 1.6834090909090909,
+      "grad_norm": 2.4793362949264095,
+      "learning_rate": 9.318666817195978e-07,
+      "loss": 0.0532,
+      "step": 7407
+    },
+    {
+      "epoch": 1.6836363636363636,
+      "grad_norm": 1.419889722015816,
+      "learning_rate": 9.317889271116182e-07,
+      "loss": 0.0333,
+      "step": 7408
+    },
+    {
+      "epoch": 1.6838636363636363,
+      "grad_norm": 2.208578469759838,
+      "learning_rate": 9.317111662476772e-07,
+      "loss": 0.0643,
+      "step": 7409
+    },
+    {
+      "epoch": 1.684090909090909,
+      "grad_norm": 2.5522075390665835,
+      "learning_rate": 9.316333991293605e-07,
+      "loss": 0.067,
+      "step": 7410
+    },
+    {
+      "epoch": 1.6843181818181818,
+      "grad_norm": 5.609558426863095,
+      "learning_rate": 9.315556257582536e-07,
+      "loss": 0.1262,
+      "step": 7411
+    },
+    {
+      "epoch": 1.6845454545454546,
+      "grad_norm": 1.5016344478223091,
+      "learning_rate": 9.31477846135943e-07,
+      "loss": 0.053,
+      "step": 7412
+    },
+    {
+      "epoch": 1.6847727272727273,
+      "grad_norm": 1.0022086222387918,
+      "learning_rate": 9.314000602640142e-07,
+      "loss": 0.0131,
+      "step": 7413
+    },
+    {
+      "epoch": 1.685,
+      "grad_norm": 3.190312664796673,
+      "learning_rate": 9.313222681440538e-07,
+      "loss": 0.0457,
+      "step": 7414
+    },
+    {
+      "epoch": 1.6852272727272726,
+      "grad_norm": 1.600464610206276,
+      "learning_rate": 9.31244469777648e-07,
+      "loss": 0.0389,
+      "step": 7415
+    },
+    {
+      "epoch": 1.6854545454545455,
+      "grad_norm": 1.3574955229280026,
+      "learning_rate": 9.311666651663831e-07,
+      "loss": 0.0334,
+      "step": 7416
+    },
+    {
+      "epoch": 1.685681818181818,
+      "grad_norm": 2.723645677020909,
+      "learning_rate": 9.310888543118457e-07,
+      "loss": 0.078,
+      "step": 7417
+    },
+    {
+      "epoch": 1.685909090909091,
+      "grad_norm": 2.085488480492263,
+      "learning_rate": 9.310110372156227e-07,
+      "loss": 0.0726,
+      "step": 7418
+    },
+    {
+      "epoch": 1.6861363636363635,
+      "grad_norm": 1.6250816704406563,
+      "learning_rate": 9.309332138793006e-07,
+      "loss": 0.0483,
+      "step": 7419
+    },
+    {
+      "epoch": 1.6863636363636365,
+      "grad_norm": 0.781229730517054,
+      "learning_rate": 9.308553843044665e-07,
+      "loss": 0.0158,
+      "step": 7420
+    },
+    {
+      "epoch": 1.686590909090909,
+      "grad_norm": 2.665170390068854,
+      "learning_rate": 9.307775484927078e-07,
+      "loss": 0.0747,
+      "step": 7421
+    },
+    {
+      "epoch": 1.6868181818181818,
+      "grad_norm": 1.9215163073274666,
+      "learning_rate": 9.306997064456113e-07,
+      "loss": 0.0323,
+      "step": 7422
+    },
+    {
+      "epoch": 1.6870454545454545,
+      "grad_norm": 1.2016738726270975,
+      "learning_rate": 9.306218581647642e-07,
+      "loss": 0.0219,
+      "step": 7423
+    },
+    {
+      "epoch": 1.6872727272727273,
+      "grad_norm": 2.591165012316031,
+      "learning_rate": 9.305440036517545e-07,
+      "loss": 0.0604,
+      "step": 7424
+    },
+    {
+      "epoch": 1.6875,
+      "grad_norm": 1.9520409567481884,
+      "learning_rate": 9.304661429081696e-07,
+      "loss": 0.04,
+      "step": 7425
+    },
+    {
+      "epoch": 1.6877272727272727,
+      "grad_norm": 2.532342391446377,
+      "learning_rate": 9.30388275935597e-07,
+      "loss": 0.0654,
+      "step": 7426
+    },
+    {
+      "epoch": 1.6879545454545455,
+      "grad_norm": 2.5037793615570316,
+      "learning_rate": 9.303104027356246e-07,
+      "loss": 0.0898,
+      "step": 7427
+    },
+    {
+      "epoch": 1.6881818181818182,
+      "grad_norm": 2.8890983200022453,
+      "learning_rate": 9.302325233098406e-07,
+      "loss": 0.0592,
+      "step": 7428
+    },
+    {
+      "epoch": 1.688409090909091,
+      "grad_norm": 1.4008167036125259,
+      "learning_rate": 9.301546376598327e-07,
+      "loss": 0.0291,
+      "step": 7429
+    },
+    {
+      "epoch": 1.6886363636363635,
+      "grad_norm": 1.2375945568070692,
+      "learning_rate": 9.300767457871896e-07,
+      "loss": 0.0159,
+      "step": 7430
+    },
+    {
+      "epoch": 1.6888636363636365,
+      "grad_norm": 2.2280479661015717,
+      "learning_rate": 9.299988476934995e-07,
+      "loss": 0.0392,
+      "step": 7431
+    },
+    {
+      "epoch": 1.689090909090909,
+      "grad_norm": 3.0778355675046303,
+      "learning_rate": 9.299209433803504e-07,
+      "loss": 0.047,
+      "step": 7432
+    },
+    {
+      "epoch": 1.689318181818182,
+      "grad_norm": 1.028681338864647,
+      "learning_rate": 9.298430328493317e-07,
+      "loss": 0.0147,
+      "step": 7433
+    },
+    {
+      "epoch": 1.6895454545454545,
+      "grad_norm": 1.4445066491969543,
+      "learning_rate": 9.297651161020315e-07,
+      "loss": 0.0341,
+      "step": 7434
+    },
+    {
+      "epoch": 1.6897727272727274,
+      "grad_norm": 1.9745998779642608,
+      "learning_rate": 9.296871931400391e-07,
+      "loss": 0.0823,
+      "step": 7435
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 2.0457246081527556,
+      "learning_rate": 9.29609263964943e-07,
+      "loss": 0.0564,
+      "step": 7436
+    },
+    {
+      "epoch": 1.6902272727272727,
+      "grad_norm": 0.9792239470830714,
+      "learning_rate": 9.295313285783328e-07,
+      "loss": 0.0075,
+      "step": 7437
+    },
+    {
+      "epoch": 1.6904545454545454,
+      "grad_norm": 1.8099435404112998,
+      "learning_rate": 9.294533869817974e-07,
+      "loss": 0.0306,
+      "step": 7438
+    },
+    {
+      "epoch": 1.6906818181818182,
+      "grad_norm": 1.8438565641547264,
+      "learning_rate": 9.293754391769263e-07,
+      "loss": 0.0369,
+      "step": 7439
+    },
+    {
+      "epoch": 1.690909090909091,
+      "grad_norm": 3.1682113843378805,
+      "learning_rate": 9.292974851653092e-07,
+      "loss": 0.0228,
+      "step": 7440
+    },
+    {
+      "epoch": 1.6911363636363637,
+      "grad_norm": 1.5579457291639176,
+      "learning_rate": 9.292195249485353e-07,
+      "loss": 0.0542,
+      "step": 7441
+    },
+    {
+      "epoch": 1.6913636363636364,
+      "grad_norm": 1.9279137426536248,
+      "learning_rate": 9.291415585281946e-07,
+      "loss": 0.0391,
+      "step": 7442
+    },
+    {
+      "epoch": 1.6915909090909091,
+      "grad_norm": 1.7194584952822154,
+      "learning_rate": 9.290635859058768e-07,
+      "loss": 0.0669,
+      "step": 7443
+    },
+    {
+      "epoch": 1.6918181818181819,
+      "grad_norm": 2.7926330311217042,
+      "learning_rate": 9.289856070831722e-07,
+      "loss": 0.0533,
+      "step": 7444
+    },
+    {
+      "epoch": 1.6920454545454544,
+      "grad_norm": 2.051194009070036,
+      "learning_rate": 9.289076220616705e-07,
+      "loss": 0.0322,
+      "step": 7445
+    },
+    {
+      "epoch": 1.6922727272727274,
+      "grad_norm": 1.309138884537518,
+      "learning_rate": 9.288296308429626e-07,
+      "loss": 0.0166,
+      "step": 7446
+    },
+    {
+      "epoch": 1.6925,
+      "grad_norm": 3.45262486801975,
+      "learning_rate": 9.287516334286382e-07,
+      "loss": 0.0522,
+      "step": 7447
+    },
+    {
+      "epoch": 1.6927272727272729,
+      "grad_norm": 1.6778585265651311,
+      "learning_rate": 9.286736298202879e-07,
+      "loss": 0.052,
+      "step": 7448
+    },
+    {
+      "epoch": 1.6929545454545454,
+      "grad_norm": 1.4453600302889311,
+      "learning_rate": 9.285956200195029e-07,
+      "loss": 0.0395,
+      "step": 7449
+    },
+    {
+      "epoch": 1.6931818181818183,
+      "grad_norm": 1.8494188989793485,
+      "learning_rate": 9.285176040278732e-07,
+      "loss": 0.0282,
+      "step": 7450
+    },
+    {
+      "epoch": 1.6934090909090909,
+      "grad_norm": 2.419119779392227,
+      "learning_rate": 9.284395818469902e-07,
+      "loss": 0.03,
+      "step": 7451
+    },
+    {
+      "epoch": 1.6936363636363636,
+      "grad_norm": 2.8601487000967056,
+      "learning_rate": 9.283615534784446e-07,
+      "loss": 0.0758,
+      "step": 7452
+    },
+    {
+      "epoch": 1.6938636363636363,
+      "grad_norm": 2.1055074536834475,
+      "learning_rate": 9.282835189238279e-07,
+      "loss": 0.046,
+      "step": 7453
+    },
+    {
+      "epoch": 1.694090909090909,
+      "grad_norm": 1.9279268805303964,
+      "learning_rate": 9.28205478184731e-07,
+      "loss": 0.0379,
+      "step": 7454
+    },
+    {
+      "epoch": 1.6943181818181818,
+      "grad_norm": 2.3262119582375838,
+      "learning_rate": 9.281274312627455e-07,
+      "loss": 0.0895,
+      "step": 7455
+    },
+    {
+      "epoch": 1.6945454545454546,
+      "grad_norm": 1.7252259431533383,
+      "learning_rate": 9.280493781594628e-07,
+      "loss": 0.0407,
+      "step": 7456
+    },
+    {
+      "epoch": 1.6947727272727273,
+      "grad_norm": 1.356113346145792,
+      "learning_rate": 9.279713188764746e-07,
+      "loss": 0.049,
+      "step": 7457
+    },
+    {
+      "epoch": 1.6949999999999998,
+      "grad_norm": 1.755529908295906,
+      "learning_rate": 9.278932534153726e-07,
+      "loss": 0.0369,
+      "step": 7458
+    },
+    {
+      "epoch": 1.6952272727272728,
+      "grad_norm": 1.5067467824849627,
+      "learning_rate": 9.278151817777488e-07,
+      "loss": 0.0432,
+      "step": 7459
+    },
+    {
+      "epoch": 1.6954545454545453,
+      "grad_norm": 2.220398661179403,
+      "learning_rate": 9.277371039651951e-07,
+      "loss": 0.0608,
+      "step": 7460
+    },
+    {
+      "epoch": 1.6956818181818183,
+      "grad_norm": 2.5563469220934465,
+      "learning_rate": 9.276590199793037e-07,
+      "loss": 0.0888,
+      "step": 7461
+    },
+    {
+      "epoch": 1.6959090909090908,
+      "grad_norm": 1.8090414523733453,
+      "learning_rate": 9.27580929821667e-07,
+      "loss": 0.0452,
+      "step": 7462
+    },
+    {
+      "epoch": 1.6961363636363638,
+      "grad_norm": 2.2084299846801194,
+      "learning_rate": 9.275028334938771e-07,
+      "loss": 0.0483,
+      "step": 7463
+    },
+    {
+      "epoch": 1.6963636363636363,
+      "grad_norm": 2.3922988857437484,
+      "learning_rate": 9.274247309975268e-07,
+      "loss": 0.0722,
+      "step": 7464
+    },
+    {
+      "epoch": 1.696590909090909,
+      "grad_norm": 1.970072696415226,
+      "learning_rate": 9.273466223342084e-07,
+      "loss": 0.0343,
+      "step": 7465
+    },
+    {
+      "epoch": 1.6968181818181818,
+      "grad_norm": 1.8062555806186729,
+      "learning_rate": 9.272685075055152e-07,
+      "loss": 0.056,
+      "step": 7466
+    },
+    {
+      "epoch": 1.6970454545454545,
+      "grad_norm": 1.52183170977674,
+      "learning_rate": 9.271903865130397e-07,
+      "loss": 0.0384,
+      "step": 7467
+    },
+    {
+      "epoch": 1.6972727272727273,
+      "grad_norm": 1.6765539142361645,
+      "learning_rate": 9.27112259358375e-07,
+      "loss": 0.0313,
+      "step": 7468
+    },
+    {
+      "epoch": 1.6975,
+      "grad_norm": 1.935005897479298,
+      "learning_rate": 9.270341260431141e-07,
+      "loss": 0.0404,
+      "step": 7469
+    },
+    {
+      "epoch": 1.6977272727272728,
+      "grad_norm": 0.9936965483598326,
+      "learning_rate": 9.269559865688506e-07,
+      "loss": 0.0162,
+      "step": 7470
+    },
+    {
+      "epoch": 1.6979545454545455,
+      "grad_norm": 1.8367963000738075,
+      "learning_rate": 9.268778409371778e-07,
+      "loss": 0.0429,
+      "step": 7471
+    },
+    {
+      "epoch": 1.6981818181818182,
+      "grad_norm": 1.826603265017403,
+      "learning_rate": 9.267996891496891e-07,
+      "loss": 0.041,
+      "step": 7472
+    },
+    {
+      "epoch": 1.6984090909090908,
+      "grad_norm": 0.8403448975060999,
+      "learning_rate": 9.267215312079782e-07,
+      "loss": 0.0164,
+      "step": 7473
+    },
+    {
+      "epoch": 1.6986363636363637,
+      "grad_norm": 1.3993961076045194,
+      "learning_rate": 9.266433671136388e-07,
+      "loss": 0.0227,
+      "step": 7474
+    },
+    {
+      "epoch": 1.6988636363636362,
+      "grad_norm": 2.8168217839291096,
+      "learning_rate": 9.265651968682651e-07,
+      "loss": 0.0539,
+      "step": 7475
+    },
+    {
+      "epoch": 1.6990909090909092,
+      "grad_norm": 2.594820827357807,
+      "learning_rate": 9.264870204734508e-07,
+      "loss": 0.1262,
+      "step": 7476
+    },
+    {
+      "epoch": 1.6993181818181817,
+      "grad_norm": 2.0777413784852485,
+      "learning_rate": 9.264088379307903e-07,
+      "loss": 0.0607,
+      "step": 7477
+    },
+    {
+      "epoch": 1.6995454545454547,
+      "grad_norm": 1.7115487071310145,
+      "learning_rate": 9.263306492418776e-07,
+      "loss": 0.0759,
+      "step": 7478
+    },
+    {
+      "epoch": 1.6997727272727272,
+      "grad_norm": 1.3542745066235304,
+      "learning_rate": 9.262524544083075e-07,
+      "loss": 0.031,
+      "step": 7479
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 4.233223590635224,
+      "learning_rate": 9.261742534316739e-07,
+      "loss": 0.0897,
+      "step": 7480
+    },
+    {
+      "epoch": 1.7002272727272727,
+      "grad_norm": 1.9246040600958316,
+      "learning_rate": 9.260960463135721e-07,
+      "loss": 0.0555,
+      "step": 7481
+    },
+    {
+      "epoch": 1.7004545454545454,
+      "grad_norm": 2.396870450153261,
+      "learning_rate": 9.260178330555966e-07,
+      "loss": 0.0713,
+      "step": 7482
+    },
+    {
+      "epoch": 1.7006818181818182,
+      "grad_norm": 2.2044606228707333,
+      "learning_rate": 9.259396136593423e-07,
+      "loss": 0.0385,
+      "step": 7483
+    },
+    {
+      "epoch": 1.700909090909091,
+      "grad_norm": 1.8736048638993077,
+      "learning_rate": 9.258613881264043e-07,
+      "loss": 0.0669,
+      "step": 7484
+    },
+    {
+      "epoch": 1.7011363636363637,
+      "grad_norm": 1.4263320208668973,
+      "learning_rate": 9.257831564583778e-07,
+      "loss": 0.0319,
+      "step": 7485
+    },
+    {
+      "epoch": 1.7013636363636364,
+      "grad_norm": 1.9428829502695273,
+      "learning_rate": 9.257049186568579e-07,
+      "loss": 0.0374,
+      "step": 7486
+    },
+    {
+      "epoch": 1.7015909090909092,
+      "grad_norm": 1.5170315235251202,
+      "learning_rate": 9.256266747234401e-07,
+      "loss": 0.0285,
+      "step": 7487
+    },
+    {
+      "epoch": 1.7018181818181817,
+      "grad_norm": 1.4857133779637564,
+      "learning_rate": 9.255484246597198e-07,
+      "loss": 0.0538,
+      "step": 7488
+    },
+    {
+      "epoch": 1.7020454545454546,
+      "grad_norm": 2.387887907155917,
+      "learning_rate": 9.254701684672929e-07,
+      "loss": 0.0507,
+      "step": 7489
+    },
+    {
+      "epoch": 1.7022727272727272,
+      "grad_norm": 1.7295367592305018,
+      "learning_rate": 9.253919061477552e-07,
+      "loss": 0.0426,
+      "step": 7490
+    },
+    {
+      "epoch": 1.7025000000000001,
+      "grad_norm": 0.9241245189223286,
+      "learning_rate": 9.253136377027022e-07,
+      "loss": 0.0165,
+      "step": 7491
+    },
+    {
+      "epoch": 1.7027272727272726,
+      "grad_norm": 1.6355160433178186,
+      "learning_rate": 9.252353631337305e-07,
+      "loss": 0.0297,
+      "step": 7492
+    },
+    {
+      "epoch": 1.7029545454545456,
+      "grad_norm": 2.7005399729832877,
+      "learning_rate": 9.251570824424358e-07,
+      "loss": 0.0603,
+      "step": 7493
+    },
+    {
+      "epoch": 1.7031818181818181,
+      "grad_norm": 1.666295759954741,
+      "learning_rate": 9.250787956304144e-07,
+      "loss": 0.0567,
+      "step": 7494
+    },
+    {
+      "epoch": 1.7034090909090909,
+      "grad_norm": 1.5151815151406733,
+      "learning_rate": 9.250005026992631e-07,
+      "loss": 0.0534,
+      "step": 7495
+    },
+    {
+      "epoch": 1.7036363636363636,
+      "grad_norm": 2.1396953465989315,
+      "learning_rate": 9.249222036505779e-07,
+      "loss": 0.039,
+      "step": 7496
+    },
+    {
+      "epoch": 1.7038636363636364,
+      "grad_norm": 1.6108502304405126,
+      "learning_rate": 9.248438984859561e-07,
+      "loss": 0.0326,
+      "step": 7497
+    },
+    {
+      "epoch": 1.704090909090909,
+      "grad_norm": 2.202572606522634,
+      "learning_rate": 9.247655872069938e-07,
+      "loss": 0.0541,
+      "step": 7498
+    },
+    {
+      "epoch": 1.7043181818181818,
+      "grad_norm": 1.1743598217803972,
+      "learning_rate": 9.246872698152884e-07,
+      "loss": 0.0264,
+      "step": 7499
+    },
+    {
+      "epoch": 1.7045454545454546,
+      "grad_norm": 1.8169182241749096,
+      "learning_rate": 9.246089463124365e-07,
+      "loss": 0.0616,
+      "step": 7500
+    },
+    {
+      "epoch": 1.704772727272727,
+      "grad_norm": 2.10588648072524,
+      "learning_rate": 9.245306167000357e-07,
+      "loss": 0.0465,
+      "step": 7501
+    },
+    {
+      "epoch": 1.705,
+      "grad_norm": 2.435494364981861,
+      "learning_rate": 9.24452280979683e-07,
+      "loss": 0.0412,
+      "step": 7502
+    },
+    {
+      "epoch": 1.7052272727272726,
+      "grad_norm": 1.2656235106786113,
+      "learning_rate": 9.243739391529757e-07,
+      "loss": 0.0267,
+      "step": 7503
+    },
+    {
+      "epoch": 1.7054545454545456,
+      "grad_norm": 0.41945981540036487,
+      "learning_rate": 9.242955912215117e-07,
+      "loss": 0.01,
+      "step": 7504
+    },
+    {
+      "epoch": 1.705681818181818,
+      "grad_norm": 1.1598718661175542,
+      "learning_rate": 9.242172371868882e-07,
+      "loss": 0.0279,
+      "step": 7505
+    },
+    {
+      "epoch": 1.705909090909091,
+      "grad_norm": 1.2470923033963537,
+      "learning_rate": 9.241388770507036e-07,
+      "loss": 0.0246,
+      "step": 7506
+    },
+    {
+      "epoch": 1.7061363636363636,
+      "grad_norm": 2.3938906387441166,
+      "learning_rate": 9.240605108145552e-07,
+      "loss": 0.0506,
+      "step": 7507
+    },
+    {
+      "epoch": 1.7063636363636365,
+      "grad_norm": 1.8195227695574383,
+      "learning_rate": 9.239821384800413e-07,
+      "loss": 0.0438,
+      "step": 7508
+    },
+    {
+      "epoch": 1.706590909090909,
+      "grad_norm": 2.1668385738812135,
+      "learning_rate": 9.239037600487599e-07,
+      "loss": 0.0509,
+      "step": 7509
+    },
+    {
+      "epoch": 1.7068181818181818,
+      "grad_norm": 1.3559680453725473,
+      "learning_rate": 9.238253755223095e-07,
+      "loss": 0.0351,
+      "step": 7510
+    },
+    {
+      "epoch": 1.7070454545454545,
+      "grad_norm": 1.8137172818682408,
+      "learning_rate": 9.237469849022881e-07,
+      "loss": 0.0395,
+      "step": 7511
+    },
+    {
+      "epoch": 1.7072727272727273,
+      "grad_norm": 1.9692003384946686,
+      "learning_rate": 9.236685881902949e-07,
+      "loss": 0.0805,
+      "step": 7512
+    },
+    {
+      "epoch": 1.7075,
+      "grad_norm": 2.7476134638946683,
+      "learning_rate": 9.235901853879279e-07,
+      "loss": 0.0433,
+      "step": 7513
+    },
+    {
+      "epoch": 1.7077272727272728,
+      "grad_norm": 1.0852699444694398,
+      "learning_rate": 9.235117764967861e-07,
+      "loss": 0.023,
+      "step": 7514
+    },
+    {
+      "epoch": 1.7079545454545455,
+      "grad_norm": 2.4529117758898855,
+      "learning_rate": 9.234333615184685e-07,
+      "loss": 0.0232,
+      "step": 7515
+    },
+    {
+      "epoch": 1.708181818181818,
+      "grad_norm": 1.8402987839415061,
+      "learning_rate": 9.233549404545737e-07,
+      "loss": 0.0378,
+      "step": 7516
+    },
+    {
+      "epoch": 1.708409090909091,
+      "grad_norm": 1.8815654733495868,
+      "learning_rate": 9.232765133067015e-07,
+      "loss": 0.0534,
+      "step": 7517
+    },
+    {
+      "epoch": 1.7086363636363635,
+      "grad_norm": 1.138891591453972,
+      "learning_rate": 9.231980800764506e-07,
+      "loss": 0.028,
+      "step": 7518
+    },
+    {
+      "epoch": 1.7088636363636365,
+      "grad_norm": 0.6056214286779696,
+      "learning_rate": 9.231196407654208e-07,
+      "loss": 0.0268,
+      "step": 7519
+    },
+    {
+      "epoch": 1.709090909090909,
+      "grad_norm": 2.0272698735080352,
+      "learning_rate": 9.230411953752113e-07,
+      "loss": 0.0515,
+      "step": 7520
+    },
+    {
+      "epoch": 1.709318181818182,
+      "grad_norm": 1.5979692521432007,
+      "learning_rate": 9.229627439074219e-07,
+      "loss": 0.0384,
+      "step": 7521
+    },
+    {
+      "epoch": 1.7095454545454545,
+      "grad_norm": 1.7717099927125506,
+      "learning_rate": 9.228842863636523e-07,
+      "loss": 0.0229,
+      "step": 7522
+    },
+    {
+      "epoch": 1.7097727272727272,
+      "grad_norm": 1.935645706434609,
+      "learning_rate": 9.228058227455025e-07,
+      "loss": 0.0419,
+      "step": 7523
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 2.587221359647727,
+      "learning_rate": 9.227273530545723e-07,
+      "loss": 0.0767,
+      "step": 7524
+    },
+    {
+      "epoch": 1.7102272727272727,
+      "grad_norm": 2.1709828951683603,
+      "learning_rate": 9.226488772924621e-07,
+      "loss": 0.0445,
+      "step": 7525
+    },
+    {
+      "epoch": 1.7104545454545454,
+      "grad_norm": 1.7233007885237395,
+      "learning_rate": 9.225703954607718e-07,
+      "loss": 0.0482,
+      "step": 7526
+    },
+    {
+      "epoch": 1.7106818181818182,
+      "grad_norm": 1.478657026097904,
+      "learning_rate": 9.224919075611022e-07,
+      "loss": 0.0363,
+      "step": 7527
+    },
+    {
+      "epoch": 1.710909090909091,
+      "grad_norm": 0.6288439733016561,
+      "learning_rate": 9.224134135950536e-07,
+      "loss": 0.0258,
+      "step": 7528
+    },
+    {
+      "epoch": 1.7111363636363637,
+      "grad_norm": 1.4358280030614807,
+      "learning_rate": 9.223349135642265e-07,
+      "loss": 0.0274,
+      "step": 7529
+    },
+    {
+      "epoch": 1.7113636363636364,
+      "grad_norm": 1.2148523560791984,
+      "learning_rate": 9.222564074702219e-07,
+      "loss": 0.0304,
+      "step": 7530
+    },
+    {
+      "epoch": 1.711590909090909,
+      "grad_norm": 1.2091609265368073,
+      "learning_rate": 9.221778953146405e-07,
+      "loss": 0.029,
+      "step": 7531
+    },
+    {
+      "epoch": 1.711818181818182,
+      "grad_norm": 2.753762515876599,
+      "learning_rate": 9.220993770990834e-07,
+      "loss": 0.0764,
+      "step": 7532
+    },
+    {
+      "epoch": 1.7120454545454544,
+      "grad_norm": 1.4564552777140167,
+      "learning_rate": 9.220208528251516e-07,
+      "loss": 0.0218,
+      "step": 7533
+    },
+    {
+      "epoch": 1.7122727272727274,
+      "grad_norm": 1.9322531638252667,
+      "learning_rate": 9.219423224944465e-07,
+      "loss": 0.029,
+      "step": 7534
+    },
+    {
+      "epoch": 1.7125,
+      "grad_norm": 2.561976198584319,
+      "learning_rate": 9.218637861085692e-07,
+      "loss": 0.0456,
+      "step": 7535
+    },
+    {
+      "epoch": 1.7127272727272729,
+      "grad_norm": 2.408894466317291,
+      "learning_rate": 9.217852436691216e-07,
+      "loss": 0.0877,
+      "step": 7536
+    },
+    {
+      "epoch": 1.7129545454545454,
+      "grad_norm": 2.442171871515233,
+      "learning_rate": 9.217066951777052e-07,
+      "loss": 0.0666,
+      "step": 7537
+    },
+    {
+      "epoch": 1.7131818181818181,
+      "grad_norm": 2.1182903548636065,
+      "learning_rate": 9.216281406359215e-07,
+      "loss": 0.0557,
+      "step": 7538
+    },
+    {
+      "epoch": 1.7134090909090909,
+      "grad_norm": 1.4944576545485095,
+      "learning_rate": 9.215495800453726e-07,
+      "loss": 0.0562,
+      "step": 7539
+    },
+    {
+      "epoch": 1.7136363636363636,
+      "grad_norm": 1.7508566543201383,
+      "learning_rate": 9.214710134076604e-07,
+      "loss": 0.0735,
+      "step": 7540
+    },
+    {
+      "epoch": 1.7138636363636364,
+      "grad_norm": 1.6888292778964005,
+      "learning_rate": 9.213924407243868e-07,
+      "loss": 0.028,
+      "step": 7541
+    },
+    {
+      "epoch": 1.714090909090909,
+      "grad_norm": 2.693797864932574,
+      "learning_rate": 9.213138619971546e-07,
+      "loss": 0.0662,
+      "step": 7542
+    },
+    {
+      "epoch": 1.7143181818181819,
+      "grad_norm": 2.899386490422322,
+      "learning_rate": 9.212352772275655e-07,
+      "loss": 0.1013,
+      "step": 7543
+    },
+    {
+      "epoch": 1.7145454545454546,
+      "grad_norm": 1.7208696780217776,
+      "learning_rate": 9.211566864172225e-07,
+      "loss": 0.0224,
+      "step": 7544
+    },
+    {
+      "epoch": 1.7147727272727273,
+      "grad_norm": 1.3971158161208512,
+      "learning_rate": 9.21078089567728e-07,
+      "loss": 0.0696,
+      "step": 7545
+    },
+    {
+      "epoch": 1.7149999999999999,
+      "grad_norm": 1.5127483477417665,
+      "learning_rate": 9.209994866806846e-07,
+      "loss": 0.0303,
+      "step": 7546
+    },
+    {
+      "epoch": 1.7152272727272728,
+      "grad_norm": 1.4513125666592266,
+      "learning_rate": 9.209208777576952e-07,
+      "loss": 0.0324,
+      "step": 7547
+    },
+    {
+      "epoch": 1.7154545454545453,
+      "grad_norm": 2.5292185810914156,
+      "learning_rate": 9.208422628003629e-07,
+      "loss": 0.0569,
+      "step": 7548
+    },
+    {
+      "epoch": 1.7156818181818183,
+      "grad_norm": 1.6330181667562664,
+      "learning_rate": 9.207636418102906e-07,
+      "loss": 0.0352,
+      "step": 7549
+    },
+    {
+      "epoch": 1.7159090909090908,
+      "grad_norm": 0.9498224447625182,
+      "learning_rate": 9.206850147890819e-07,
+      "loss": 0.0246,
+      "step": 7550
+    },
+    {
+      "epoch": 1.7161363636363638,
+      "grad_norm": 1.3785043517705668,
+      "learning_rate": 9.206063817383399e-07,
+      "loss": 0.0362,
+      "step": 7551
+    },
+    {
+      "epoch": 1.7163636363636363,
+      "grad_norm": 1.6055983386215162,
+      "learning_rate": 9.205277426596681e-07,
+      "loss": 0.0301,
+      "step": 7552
+    },
+    {
+      "epoch": 1.716590909090909,
+      "grad_norm": 1.597855919062168,
+      "learning_rate": 9.204490975546698e-07,
+      "loss": 0.0233,
+      "step": 7553
+    },
+    {
+      "epoch": 1.7168181818181818,
+      "grad_norm": 1.9719357588876318,
+      "learning_rate": 9.203704464249491e-07,
+      "loss": 0.0589,
+      "step": 7554
+    },
+    {
+      "epoch": 1.7170454545454545,
+      "grad_norm": 1.889908294308924,
+      "learning_rate": 9.202917892721097e-07,
+      "loss": 0.052,
+      "step": 7555
+    },
+    {
+      "epoch": 1.7172727272727273,
+      "grad_norm": 2.235390057482794,
+      "learning_rate": 9.202131260977555e-07,
+      "loss": 0.0503,
+      "step": 7556
+    },
+    {
+      "epoch": 1.7175,
+      "grad_norm": 2.804488854630809,
+      "learning_rate": 9.201344569034906e-07,
+      "loss": 0.0869,
+      "step": 7557
+    },
+    {
+      "epoch": 1.7177272727272728,
+      "grad_norm": 1.6691733761506757,
+      "learning_rate": 9.200557816909194e-07,
+      "loss": 0.0249,
+      "step": 7558
+    },
+    {
+      "epoch": 1.7179545454545453,
+      "grad_norm": 1.0623647801161507,
+      "learning_rate": 9.199771004616459e-07,
+      "loss": 0.0266,
+      "step": 7559
+    },
+    {
+      "epoch": 1.7181818181818183,
+      "grad_norm": 1.1826952410585887,
+      "learning_rate": 9.198984132172748e-07,
+      "loss": 0.0261,
+      "step": 7560
+    },
+    {
+      "epoch": 1.7184090909090908,
+      "grad_norm": 1.8115391319241485,
+      "learning_rate": 9.198197199594105e-07,
+      "loss": 0.0518,
+      "step": 7561
+    },
+    {
+      "epoch": 1.7186363636363637,
+      "grad_norm": 1.619706364056777,
+      "learning_rate": 9.197410206896578e-07,
+      "loss": 0.0476,
+      "step": 7562
+    },
+    {
+      "epoch": 1.7188636363636363,
+      "grad_norm": 1.0141101694809727,
+      "learning_rate": 9.196623154096217e-07,
+      "loss": 0.0153,
+      "step": 7563
+    },
+    {
+      "epoch": 1.7190909090909092,
+      "grad_norm": 2.4369996406605337,
+      "learning_rate": 9.195836041209065e-07,
+      "loss": 0.0837,
+      "step": 7564
+    },
+    {
+      "epoch": 1.7193181818181817,
+      "grad_norm": 3.28070914377262,
+      "learning_rate": 9.19504886825118e-07,
+      "loss": 0.0675,
+      "step": 7565
+    },
+    {
+      "epoch": 1.7195454545454547,
+      "grad_norm": 1.9430777808422963,
+      "learning_rate": 9.194261635238609e-07,
+      "loss": 0.0366,
+      "step": 7566
+    },
+    {
+      "epoch": 1.7197727272727272,
+      "grad_norm": 2.4064843559807936,
+      "learning_rate": 9.193474342187408e-07,
+      "loss": 0.0613,
+      "step": 7567
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 3.104667104610887,
+      "learning_rate": 9.19268698911363e-07,
+      "loss": 0.0899,
+      "step": 7568
+    },
+    {
+      "epoch": 1.7202272727272727,
+      "grad_norm": 2.742852585472886,
+      "learning_rate": 9.191899576033331e-07,
+      "loss": 0.063,
+      "step": 7569
+    },
+    {
+      "epoch": 1.7204545454545455,
+      "grad_norm": 1.9880852704930696,
+      "learning_rate": 9.191112102962566e-07,
+      "loss": 0.0499,
+      "step": 7570
+    },
+    {
+      "epoch": 1.7206818181818182,
+      "grad_norm": 1.63601644287955,
+      "learning_rate": 9.190324569917393e-07,
+      "loss": 0.044,
+      "step": 7571
+    },
+    {
+      "epoch": 1.720909090909091,
+      "grad_norm": 1.6088632154560687,
+      "learning_rate": 9.189536976913873e-07,
+      "loss": 0.0235,
+      "step": 7572
+    },
+    {
+      "epoch": 1.7211363636363637,
+      "grad_norm": 2.7699582925407147,
+      "learning_rate": 9.188749323968066e-07,
+      "loss": 0.0823,
+      "step": 7573
+    },
+    {
+      "epoch": 1.7213636363636362,
+      "grad_norm": 1.437866690716271,
+      "learning_rate": 9.187961611096035e-07,
+      "loss": 0.0217,
+      "step": 7574
+    },
+    {
+      "epoch": 1.7215909090909092,
+      "grad_norm": 1.7023340883016929,
+      "learning_rate": 9.18717383831384e-07,
+      "loss": 0.0752,
+      "step": 7575
+    },
+    {
+      "epoch": 1.7218181818181817,
+      "grad_norm": 0.5454268067289567,
+      "learning_rate": 9.186386005637546e-07,
+      "loss": 0.0185,
+      "step": 7576
+    },
+    {
+      "epoch": 1.7220454545454547,
+      "grad_norm": 2.6473284141677906,
+      "learning_rate": 9.185598113083219e-07,
+      "loss": 0.0517,
+      "step": 7577
+    },
+    {
+      "epoch": 1.7222727272727272,
+      "grad_norm": 1.7778324757213702,
+      "learning_rate": 9.184810160666926e-07,
+      "loss": 0.0438,
+      "step": 7578
+    },
+    {
+      "epoch": 1.7225000000000001,
+      "grad_norm": 1.3463397163574438,
+      "learning_rate": 9.184022148404729e-07,
+      "loss": 0.0356,
+      "step": 7579
+    },
+    {
+      "epoch": 1.7227272727272727,
+      "grad_norm": 1.3889190538245169,
+      "learning_rate": 9.183234076312706e-07,
+      "loss": 0.0331,
+      "step": 7580
+    },
+    {
+      "epoch": 1.7229545454545454,
+      "grad_norm": 1.9155053321593158,
+      "learning_rate": 9.182445944406923e-07,
+      "loss": 0.041,
+      "step": 7581
+    },
+    {
+      "epoch": 1.7231818181818181,
+      "grad_norm": 1.658309041366462,
+      "learning_rate": 9.181657752703449e-07,
+      "loss": 0.0272,
+      "step": 7582
+    },
+    {
+      "epoch": 1.723409090909091,
+      "grad_norm": 2.2975102625743347,
+      "learning_rate": 9.180869501218362e-07,
+      "loss": 0.0359,
+      "step": 7583
+    },
+    {
+      "epoch": 1.7236363636363636,
+      "grad_norm": 1.9087152811358277,
+      "learning_rate": 9.18008118996773e-07,
+      "loss": 0.0632,
+      "step": 7584
+    },
+    {
+      "epoch": 1.7238636363636364,
+      "grad_norm": 2.5168452414108056,
+      "learning_rate": 9.179292818967632e-07,
+      "loss": 0.0387,
+      "step": 7585
+    },
+    {
+      "epoch": 1.7240909090909091,
+      "grad_norm": 3.20513286546622,
+      "learning_rate": 9.178504388234143e-07,
+      "loss": 0.0939,
+      "step": 7586
+    },
+    {
+      "epoch": 1.7243181818181819,
+      "grad_norm": 1.9300600110318118,
+      "learning_rate": 9.177715897783339e-07,
+      "loss": 0.0614,
+      "step": 7587
+    },
+    {
+      "epoch": 1.7245454545454546,
+      "grad_norm": 1.6268253268014061,
+      "learning_rate": 9.176927347631301e-07,
+      "loss": 0.041,
+      "step": 7588
+    },
+    {
+      "epoch": 1.7247727272727271,
+      "grad_norm": 1.6407080972238215,
+      "learning_rate": 9.176138737794109e-07,
+      "loss": 0.0303,
+      "step": 7589
+    },
+    {
+      "epoch": 1.725,
+      "grad_norm": 2.994670030114365,
+      "learning_rate": 9.175350068287842e-07,
+      "loss": 0.0339,
+      "step": 7590
+    },
+    {
+      "epoch": 1.7252272727272726,
+      "grad_norm": 1.4053372687087287,
+      "learning_rate": 9.174561339128584e-07,
+      "loss": 0.0206,
+      "step": 7591
+    },
+    {
+      "epoch": 1.7254545454545456,
+      "grad_norm": 1.0841127023119752,
+      "learning_rate": 9.173772550332419e-07,
+      "loss": 0.0337,
+      "step": 7592
+    },
+    {
+      "epoch": 1.725681818181818,
+      "grad_norm": 2.043247663097329,
+      "learning_rate": 9.172983701915431e-07,
+      "loss": 0.0276,
+      "step": 7593
+    },
+    {
+      "epoch": 1.725909090909091,
+      "grad_norm": 1.213285218567271,
+      "learning_rate": 9.172194793893703e-07,
+      "loss": 0.0331,
+      "step": 7594
+    },
+    {
+      "epoch": 1.7261363636363636,
+      "grad_norm": 2.2192657229672643,
+      "learning_rate": 9.171405826283328e-07,
+      "loss": 0.0378,
+      "step": 7595
+    },
+    {
+      "epoch": 1.7263636363636363,
+      "grad_norm": 1.928211599427828,
+      "learning_rate": 9.170616799100391e-07,
+      "loss": 0.039,
+      "step": 7596
+    },
+    {
+      "epoch": 1.726590909090909,
+      "grad_norm": 1.9628295221021508,
+      "learning_rate": 9.169827712360981e-07,
+      "loss": 0.0348,
+      "step": 7597
+    },
+    {
+      "epoch": 1.7268181818181818,
+      "grad_norm": 1.5857776010575788,
+      "learning_rate": 9.169038566081191e-07,
+      "loss": 0.0618,
+      "step": 7598
+    },
+    {
+      "epoch": 1.7270454545454546,
+      "grad_norm": 1.5661383822804507,
+      "learning_rate": 9.168249360277112e-07,
+      "loss": 0.041,
+      "step": 7599
+    },
+    {
+      "epoch": 1.7272727272727273,
+      "grad_norm": 1.4330939054047653,
+      "learning_rate": 9.167460094964839e-07,
+      "loss": 0.0365,
+      "step": 7600
+    },
+    {
+      "epoch": 1.7275,
+      "grad_norm": 1.5601565498134204,
+      "learning_rate": 9.166670770160461e-07,
+      "loss": 0.0302,
+      "step": 7601
+    },
+    {
+      "epoch": 1.7277272727272728,
+      "grad_norm": 2.3991871665999955,
+      "learning_rate": 9.16588138588008e-07,
+      "loss": 0.0417,
+      "step": 7602
+    },
+    {
+      "epoch": 1.7279545454545455,
+      "grad_norm": 1.3052056822102804,
+      "learning_rate": 9.165091942139791e-07,
+      "loss": 0.0315,
+      "step": 7603
+    },
+    {
+      "epoch": 1.728181818181818,
+      "grad_norm": 1.8665357519332784,
+      "learning_rate": 9.164302438955689e-07,
+      "loss": 0.0452,
+      "step": 7604
+    },
+    {
+      "epoch": 1.728409090909091,
+      "grad_norm": 2.6408021550073544,
+      "learning_rate": 9.16351287634388e-07,
+      "loss": 0.0691,
+      "step": 7605
+    },
+    {
+      "epoch": 1.7286363636363635,
+      "grad_norm": 1.1456390024022323,
+      "learning_rate": 9.162723254320458e-07,
+      "loss": 0.0199,
+      "step": 7606
+    },
+    {
+      "epoch": 1.7288636363636365,
+      "grad_norm": 2.529194575611235,
+      "learning_rate": 9.161933572901528e-07,
+      "loss": 0.0559,
+      "step": 7607
+    },
+    {
+      "epoch": 1.729090909090909,
+      "grad_norm": 1.4445349872958857,
+      "learning_rate": 9.161143832103192e-07,
+      "loss": 0.0335,
+      "step": 7608
+    },
+    {
+      "epoch": 1.729318181818182,
+      "grad_norm": 2.871420460901955,
+      "learning_rate": 9.160354031941554e-07,
+      "loss": 0.066,
+      "step": 7609
+    },
+    {
+      "epoch": 1.7295454545454545,
+      "grad_norm": 1.6301921393196028,
+      "learning_rate": 9.15956417243272e-07,
+      "loss": 0.0408,
+      "step": 7610
+    },
+    {
+      "epoch": 1.7297727272727272,
+      "grad_norm": 1.7652414502953313,
+      "learning_rate": 9.158774253592798e-07,
+      "loss": 0.0319,
+      "step": 7611
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 2.3181058597234,
+      "learning_rate": 9.157984275437892e-07,
+      "loss": 0.0447,
+      "step": 7612
+    },
+    {
+      "epoch": 1.7302272727272727,
+      "grad_norm": 1.92731237113378,
+      "learning_rate": 9.157194237984116e-07,
+      "loss": 0.0321,
+      "step": 7613
+    },
+    {
+      "epoch": 1.7304545454545455,
+      "grad_norm": 1.124005898871169,
+      "learning_rate": 9.156404141247577e-07,
+      "loss": 0.0243,
+      "step": 7614
+    },
+    {
+      "epoch": 1.7306818181818182,
+      "grad_norm": 2.6030665348732476,
+      "learning_rate": 9.155613985244386e-07,
+      "loss": 0.0738,
+      "step": 7615
+    },
+    {
+      "epoch": 1.730909090909091,
+      "grad_norm": 2.007888670816116,
+      "learning_rate": 9.154823769990656e-07,
+      "loss": 0.0599,
+      "step": 7616
+    },
+    {
+      "epoch": 1.7311363636363635,
+      "grad_norm": 1.6042500652936535,
+      "learning_rate": 9.154033495502504e-07,
+      "loss": 0.0845,
+      "step": 7617
+    },
+    {
+      "epoch": 1.7313636363636364,
+      "grad_norm": 2.8550112251012,
+      "learning_rate": 9.15324316179604e-07,
+      "loss": 0.0529,
+      "step": 7618
+    },
+    {
+      "epoch": 1.731590909090909,
+      "grad_norm": 1.8184578632844084,
+      "learning_rate": 9.152452768887384e-07,
+      "loss": 0.0736,
+      "step": 7619
+    },
+    {
+      "epoch": 1.731818181818182,
+      "grad_norm": 1.999000954300751,
+      "learning_rate": 9.151662316792653e-07,
+      "loss": 0.0462,
+      "step": 7620
+    },
+    {
+      "epoch": 1.7320454545454544,
+      "grad_norm": 1.985269834499706,
+      "learning_rate": 9.150871805527965e-07,
+      "loss": 0.067,
+      "step": 7621
+    },
+    {
+      "epoch": 1.7322727272727274,
+      "grad_norm": 3.186116295826077,
+      "learning_rate": 9.15008123510944e-07,
+      "loss": 0.0775,
+      "step": 7622
+    },
+    {
+      "epoch": 1.7325,
+      "grad_norm": 2.979263880123448,
+      "learning_rate": 9.149290605553199e-07,
+      "loss": 0.072,
+      "step": 7623
+    },
+    {
+      "epoch": 1.732727272727273,
+      "grad_norm": 2.852639120583224,
+      "learning_rate": 9.148499916875364e-07,
+      "loss": 0.0493,
+      "step": 7624
+    },
+    {
+      "epoch": 1.7329545454545454,
+      "grad_norm": 2.146934909913895,
+      "learning_rate": 9.147709169092059e-07,
+      "loss": 0.1034,
+      "step": 7625
+    },
+    {
+      "epoch": 1.7331818181818182,
+      "grad_norm": 2.797743793812036,
+      "learning_rate": 9.146918362219409e-07,
+      "loss": 0.0399,
+      "step": 7626
+    },
+    {
+      "epoch": 1.733409090909091,
+      "grad_norm": 1.1001513247613943,
+      "learning_rate": 9.146127496273541e-07,
+      "loss": 0.0269,
+      "step": 7627
+    },
+    {
+      "epoch": 1.7336363636363636,
+      "grad_norm": 1.104711186357593,
+      "learning_rate": 9.145336571270579e-07,
+      "loss": 0.0196,
+      "step": 7628
+    },
+    {
+      "epoch": 1.7338636363636364,
+      "grad_norm": 2.273150261514639,
+      "learning_rate": 9.144545587226655e-07,
+      "loss": 0.0408,
+      "step": 7629
+    },
+    {
+      "epoch": 1.7340909090909091,
+      "grad_norm": 1.5342538051623618,
+      "learning_rate": 9.143754544157894e-07,
+      "loss": 0.0234,
+      "step": 7630
+    },
+    {
+      "epoch": 1.7343181818181819,
+      "grad_norm": 2.6392423157418983,
+      "learning_rate": 9.142963442080432e-07,
+      "loss": 0.1052,
+      "step": 7631
+    },
+    {
+      "epoch": 1.7345454545454544,
+      "grad_norm": 1.4062399918062678,
+      "learning_rate": 9.142172281010397e-07,
+      "loss": 0.0283,
+      "step": 7632
+    },
+    {
+      "epoch": 1.7347727272727274,
+      "grad_norm": 1.061450263300308,
+      "learning_rate": 9.141381060963924e-07,
+      "loss": 0.0335,
+      "step": 7633
+    },
+    {
+      "epoch": 1.7349999999999999,
+      "grad_norm": 1.286182510254549,
+      "learning_rate": 9.140589781957146e-07,
+      "loss": 0.034,
+      "step": 7634
+    },
+    {
+      "epoch": 1.7352272727272728,
+      "grad_norm": 2.630317123793079,
+      "learning_rate": 9.1397984440062e-07,
+      "loss": 0.0546,
+      "step": 7635
+    },
+    {
+      "epoch": 1.7354545454545454,
+      "grad_norm": 2.6892249760199296,
+      "learning_rate": 9.139007047127222e-07,
+      "loss": 0.1051,
+      "step": 7636
+    },
+    {
+      "epoch": 1.7356818181818183,
+      "grad_norm": 1.9901405581296434,
+      "learning_rate": 9.138215591336349e-07,
+      "loss": 0.0541,
+      "step": 7637
+    },
+    {
+      "epoch": 1.7359090909090908,
+      "grad_norm": 1.3731414301584983,
+      "learning_rate": 9.137424076649725e-07,
+      "loss": 0.0332,
+      "step": 7638
+    },
+    {
+      "epoch": 1.7361363636363636,
+      "grad_norm": 1.720681594302477,
+      "learning_rate": 9.136632503083484e-07,
+      "loss": 0.0444,
+      "step": 7639
+    },
+    {
+      "epoch": 1.7363636363636363,
+      "grad_norm": 1.6531340280855662,
+      "learning_rate": 9.13584087065377e-07,
+      "loss": 0.0267,
+      "step": 7640
+    },
+    {
+      "epoch": 1.736590909090909,
+      "grad_norm": 2.2193165719231187,
+      "learning_rate": 9.135049179376726e-07,
+      "loss": 0.0403,
+      "step": 7641
+    },
+    {
+      "epoch": 1.7368181818181818,
+      "grad_norm": 1.886131753654549,
+      "learning_rate": 9.134257429268497e-07,
+      "loss": 0.0326,
+      "step": 7642
+    },
+    {
+      "epoch": 1.7370454545454546,
+      "grad_norm": 2.332879360149857,
+      "learning_rate": 9.133465620345227e-07,
+      "loss": 0.0639,
+      "step": 7643
+    },
+    {
+      "epoch": 1.7372727272727273,
+      "grad_norm": 1.942330304913573,
+      "learning_rate": 9.132673752623062e-07,
+      "loss": 0.0561,
+      "step": 7644
+    },
+    {
+      "epoch": 1.7375,
+      "grad_norm": 0.9990905350901553,
+      "learning_rate": 9.131881826118151e-07,
+      "loss": 0.0331,
+      "step": 7645
+    },
+    {
+      "epoch": 1.7377272727272728,
+      "grad_norm": 1.9976598970756694,
+      "learning_rate": 9.131089840846641e-07,
+      "loss": 0.0373,
+      "step": 7646
+    },
+    {
+      "epoch": 1.7379545454545453,
+      "grad_norm": 1.4416750541522627,
+      "learning_rate": 9.130297796824683e-07,
+      "loss": 0.0441,
+      "step": 7647
+    },
+    {
+      "epoch": 1.7381818181818183,
+      "grad_norm": 2.0546403074735187,
+      "learning_rate": 9.129505694068428e-07,
+      "loss": 0.0444,
+      "step": 7648
+    },
+    {
+      "epoch": 1.7384090909090908,
+      "grad_norm": 2.369092406957017,
+      "learning_rate": 9.128713532594027e-07,
+      "loss": 0.0687,
+      "step": 7649
+    },
+    {
+      "epoch": 1.7386363636363638,
+      "grad_norm": 1.7831177553087343,
+      "learning_rate": 9.127921312417636e-07,
+      "loss": 0.0316,
+      "step": 7650
+    },
+    {
+      "epoch": 1.7388636363636363,
+      "grad_norm": 2.124118981724053,
+      "learning_rate": 9.12712903355541e-07,
+      "loss": 0.0521,
+      "step": 7651
+    },
+    {
+      "epoch": 1.7390909090909092,
+      "grad_norm": 2.175202537251037,
+      "learning_rate": 9.126336696023502e-07,
+      "loss": 0.044,
+      "step": 7652
+    },
+    {
+      "epoch": 1.7393181818181818,
+      "grad_norm": 1.5789336266443892,
+      "learning_rate": 9.125544299838073e-07,
+      "loss": 0.0426,
+      "step": 7653
+    },
+    {
+      "epoch": 1.7395454545454545,
+      "grad_norm": 2.1772102861796383,
+      "learning_rate": 9.124751845015276e-07,
+      "loss": 0.0483,
+      "step": 7654
+    },
+    {
+      "epoch": 1.7397727272727272,
+      "grad_norm": 2.2946109736173863,
+      "learning_rate": 9.123959331571275e-07,
+      "loss": 0.0544,
+      "step": 7655
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.9454435995988268,
+      "learning_rate": 9.123166759522229e-07,
+      "loss": 0.0493,
+      "step": 7656
+    },
+    {
+      "epoch": 1.7402272727272727,
+      "grad_norm": 2.9552108141293734,
+      "learning_rate": 9.122374128884301e-07,
+      "loss": 0.0988,
+      "step": 7657
+    },
+    {
+      "epoch": 1.7404545454545455,
+      "grad_norm": 1.4918133206891906,
+      "learning_rate": 9.121581439673652e-07,
+      "loss": 0.0373,
+      "step": 7658
+    },
+    {
+      "epoch": 1.7406818181818182,
+      "grad_norm": 2.056551491419689,
+      "learning_rate": 9.120788691906449e-07,
+      "loss": 0.0535,
+      "step": 7659
+    },
+    {
+      "epoch": 1.740909090909091,
+      "grad_norm": 1.761458532583131,
+      "learning_rate": 9.119995885598855e-07,
+      "loss": 0.0471,
+      "step": 7660
+    },
+    {
+      "epoch": 1.7411363636363637,
+      "grad_norm": 1.1568684584161368,
+      "learning_rate": 9.119203020767038e-07,
+      "loss": 0.0242,
+      "step": 7661
+    },
+    {
+      "epoch": 1.7413636363636362,
+      "grad_norm": 2.3996232085482383,
+      "learning_rate": 9.118410097427167e-07,
+      "loss": 0.0564,
+      "step": 7662
+    },
+    {
+      "epoch": 1.7415909090909092,
+      "grad_norm": 2.2032772710725608,
+      "learning_rate": 9.117617115595409e-07,
+      "loss": 0.0569,
+      "step": 7663
+    },
+    {
+      "epoch": 1.7418181818181817,
+      "grad_norm": 1.1402633983753165,
+      "learning_rate": 9.116824075287936e-07,
+      "loss": 0.0502,
+      "step": 7664
+    },
+    {
+      "epoch": 1.7420454545454547,
+      "grad_norm": 1.334202644899938,
+      "learning_rate": 9.116030976520918e-07,
+      "loss": 0.0559,
+      "step": 7665
+    },
+    {
+      "epoch": 1.7422727272727272,
+      "grad_norm": 2.4455810860820058,
+      "learning_rate": 9.11523781931053e-07,
+      "loss": 0.0739,
+      "step": 7666
+    },
+    {
+      "epoch": 1.7425000000000002,
+      "grad_norm": 1.239253868237358,
+      "learning_rate": 9.114444603672941e-07,
+      "loss": 0.0411,
+      "step": 7667
+    },
+    {
+      "epoch": 1.7427272727272727,
+      "grad_norm": 1.9388270986122595,
+      "learning_rate": 9.113651329624332e-07,
+      "loss": 0.0327,
+      "step": 7668
+    },
+    {
+      "epoch": 1.7429545454545454,
+      "grad_norm": 1.263141445078995,
+      "learning_rate": 9.112857997180874e-07,
+      "loss": 0.0267,
+      "step": 7669
+    },
+    {
+      "epoch": 1.7431818181818182,
+      "grad_norm": 2.3278549628626894,
+      "learning_rate": 9.112064606358747e-07,
+      "loss": 0.1055,
+      "step": 7670
+    },
+    {
+      "epoch": 1.743409090909091,
+      "grad_norm": 1.4570181453448892,
+      "learning_rate": 9.111271157174131e-07,
+      "loss": 0.0385,
+      "step": 7671
+    },
+    {
+      "epoch": 1.7436363636363637,
+      "grad_norm": 1.504779677338791,
+      "learning_rate": 9.110477649643204e-07,
+      "loss": 0.048,
+      "step": 7672
+    },
+    {
+      "epoch": 1.7438636363636364,
+      "grad_norm": 1.6547428036078036,
+      "learning_rate": 9.109684083782148e-07,
+      "loss": 0.0212,
+      "step": 7673
+    },
+    {
+      "epoch": 1.7440909090909091,
+      "grad_norm": 3.0342887134836665,
+      "learning_rate": 9.108890459607145e-07,
+      "loss": 0.0888,
+      "step": 7674
+    },
+    {
+      "epoch": 1.7443181818181817,
+      "grad_norm": 1.6775768234892632,
+      "learning_rate": 9.108096777134377e-07,
+      "loss": 0.0335,
+      "step": 7675
+    },
+    {
+      "epoch": 1.7445454545454546,
+      "grad_norm": 1.4399765123259312,
+      "learning_rate": 9.107303036380028e-07,
+      "loss": 0.0307,
+      "step": 7676
+    },
+    {
+      "epoch": 1.7447727272727271,
+      "grad_norm": 1.7799906106455685,
+      "learning_rate": 9.106509237360288e-07,
+      "loss": 0.0312,
+      "step": 7677
+    },
+    {
+      "epoch": 1.745,
+      "grad_norm": 1.863081908665759,
+      "learning_rate": 9.105715380091339e-07,
+      "loss": 0.0381,
+      "step": 7678
+    },
+    {
+      "epoch": 1.7452272727272726,
+      "grad_norm": 2.810208156019642,
+      "learning_rate": 9.104921464589373e-07,
+      "loss": 0.1,
+      "step": 7679
+    },
+    {
+      "epoch": 1.7454545454545456,
+      "grad_norm": 1.146544011158368,
+      "learning_rate": 9.104127490870576e-07,
+      "loss": 0.0208,
+      "step": 7680
+    },
+    {
+      "epoch": 1.7456818181818181,
+      "grad_norm": 2.5767042860795293,
+      "learning_rate": 9.103333458951142e-07,
+      "loss": 0.0848,
+      "step": 7681
+    },
+    {
+      "epoch": 1.7459090909090909,
+      "grad_norm": 2.4178713211307374,
+      "learning_rate": 9.10253936884726e-07,
+      "loss": 0.0599,
+      "step": 7682
+    },
+    {
+      "epoch": 1.7461363636363636,
+      "grad_norm": 2.9686268461085907,
+      "learning_rate": 9.101745220575123e-07,
+      "loss": 0.0645,
+      "step": 7683
+    },
+    {
+      "epoch": 1.7463636363636363,
+      "grad_norm": 1.2168567272168722,
+      "learning_rate": 9.100951014150927e-07,
+      "loss": 0.0317,
+      "step": 7684
+    },
+    {
+      "epoch": 1.746590909090909,
+      "grad_norm": 1.593230718108367,
+      "learning_rate": 9.100156749590866e-07,
+      "loss": 0.0416,
+      "step": 7685
+    },
+    {
+      "epoch": 1.7468181818181818,
+      "grad_norm": 2.8557124294776823,
+      "learning_rate": 9.099362426911135e-07,
+      "loss": 0.0899,
+      "step": 7686
+    },
+    {
+      "epoch": 1.7470454545454546,
+      "grad_norm": 1.4830562811518515,
+      "learning_rate": 9.098568046127934e-07,
+      "loss": 0.0343,
+      "step": 7687
+    },
+    {
+      "epoch": 1.7472727272727273,
+      "grad_norm": 2.6942392081918087,
+      "learning_rate": 9.097773607257461e-07,
+      "loss": 0.0413,
+      "step": 7688
+    },
+    {
+      "epoch": 1.7475,
+      "grad_norm": 2.0519276149672727,
+      "learning_rate": 9.096979110315917e-07,
+      "loss": 0.079,
+      "step": 7689
+    },
+    {
+      "epoch": 1.7477272727272726,
+      "grad_norm": 2.10856660459233,
+      "learning_rate": 9.096184555319501e-07,
+      "loss": 0.0511,
+      "step": 7690
+    },
+    {
+      "epoch": 1.7479545454545455,
+      "grad_norm": 2.7888886544298956,
+      "learning_rate": 9.095389942284416e-07,
+      "loss": 0.1,
+      "step": 7691
+    },
+    {
+      "epoch": 1.748181818181818,
+      "grad_norm": 2.3574653948422535,
+      "learning_rate": 9.094595271226867e-07,
+      "loss": 0.0807,
+      "step": 7692
+    },
+    {
+      "epoch": 1.748409090909091,
+      "grad_norm": 1.6138569958759346,
+      "learning_rate": 9.093800542163056e-07,
+      "loss": 0.0193,
+      "step": 7693
+    },
+    {
+      "epoch": 1.7486363636363635,
+      "grad_norm": 2.141597652800344,
+      "learning_rate": 9.093005755109193e-07,
+      "loss": 0.0319,
+      "step": 7694
+    },
+    {
+      "epoch": 1.7488636363636365,
+      "grad_norm": 2.1651119105639873,
+      "learning_rate": 9.092210910081481e-07,
+      "loss": 0.0334,
+      "step": 7695
+    },
+    {
+      "epoch": 1.749090909090909,
+      "grad_norm": 1.1963101143430592,
+      "learning_rate": 9.091416007096131e-07,
+      "loss": 0.0213,
+      "step": 7696
+    },
+    {
+      "epoch": 1.7493181818181818,
+      "grad_norm": 1.8205046285629187,
+      "learning_rate": 9.090621046169353e-07,
+      "loss": 0.0313,
+      "step": 7697
+    },
+    {
+      "epoch": 1.7495454545454545,
+      "grad_norm": 1.6369727423288736,
+      "learning_rate": 9.089826027317352e-07,
+      "loss": 0.0311,
+      "step": 7698
+    },
+    {
+      "epoch": 1.7497727272727273,
+      "grad_norm": 1.7986399419784713,
+      "learning_rate": 9.089030950556349e-07,
+      "loss": 0.0298,
+      "step": 7699
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.9324325423730984,
+      "learning_rate": 9.088235815902549e-07,
+      "loss": 0.061,
+      "step": 7700
+    },
+    {
+      "epoch": 1.7502272727272727,
+      "grad_norm": 1.7936863221593742,
+      "learning_rate": 9.087440623372168e-07,
+      "loss": 0.0773,
+      "step": 7701
+    },
+    {
+      "epoch": 1.7504545454545455,
+      "grad_norm": 1.5290334928637097,
+      "learning_rate": 9.086645372981424e-07,
+      "loss": 0.0288,
+      "step": 7702
+    },
+    {
+      "epoch": 1.7506818181818182,
+      "grad_norm": 2.2507467217790875,
+      "learning_rate": 9.085850064746533e-07,
+      "loss": 0.0444,
+      "step": 7703
+    },
+    {
+      "epoch": 1.750909090909091,
+      "grad_norm": 1.1355267657457864,
+      "learning_rate": 9.08505469868371e-07,
+      "loss": 0.0334,
+      "step": 7704
+    },
+    {
+      "epoch": 1.7511363636363635,
+      "grad_norm": 1.8549741559478843,
+      "learning_rate": 9.084259274809177e-07,
+      "loss": 0.0607,
+      "step": 7705
+    },
+    {
+      "epoch": 1.7513636363636365,
+      "grad_norm": 2.1433584422822936,
+      "learning_rate": 9.083463793139152e-07,
+      "loss": 0.0475,
+      "step": 7706
+    },
+    {
+      "epoch": 1.751590909090909,
+      "grad_norm": 1.542695580782565,
+      "learning_rate": 9.082668253689858e-07,
+      "loss": 0.0382,
+      "step": 7707
+    },
+    {
+      "epoch": 1.751818181818182,
+      "grad_norm": 2.1924393950808847,
+      "learning_rate": 9.081872656477514e-07,
+      "loss": 0.0584,
+      "step": 7708
+    },
+    {
+      "epoch": 1.7520454545454545,
+      "grad_norm": 1.8375115102210988,
+      "learning_rate": 9.081077001518349e-07,
+      "loss": 0.0525,
+      "step": 7709
+    },
+    {
+      "epoch": 1.7522727272727274,
+      "grad_norm": 2.9476376453789848,
+      "learning_rate": 9.080281288828584e-07,
+      "loss": 0.0447,
+      "step": 7710
+    },
+    {
+      "epoch": 1.7525,
+      "grad_norm": 1.666405561664288,
+      "learning_rate": 9.079485518424444e-07,
+      "loss": 0.0612,
+      "step": 7711
+    },
+    {
+      "epoch": 1.7527272727272727,
+      "grad_norm": 1.3527567985475566,
+      "learning_rate": 9.078689690322161e-07,
+      "loss": 0.0296,
+      "step": 7712
+    },
+    {
+      "epoch": 1.7529545454545454,
+      "grad_norm": 2.32759985120955,
+      "learning_rate": 9.077893804537957e-07,
+      "loss": 0.0946,
+      "step": 7713
+    },
+    {
+      "epoch": 1.7531818181818182,
+      "grad_norm": 2.0994522250760514,
+      "learning_rate": 9.077097861088069e-07,
+      "loss": 0.0554,
+      "step": 7714
+    },
+    {
+      "epoch": 1.753409090909091,
+      "grad_norm": 3.2083671687461397,
+      "learning_rate": 9.07630185998872e-07,
+      "loss": 0.1051,
+      "step": 7715
+    },
+    {
+      "epoch": 1.7536363636363637,
+      "grad_norm": 1.4393560429173193,
+      "learning_rate": 9.075505801256145e-07,
+      "loss": 0.0486,
+      "step": 7716
+    },
+    {
+      "epoch": 1.7538636363636364,
+      "grad_norm": 3.2466838578310737,
+      "learning_rate": 9.074709684906579e-07,
+      "loss": 0.1185,
+      "step": 7717
+    },
+    {
+      "epoch": 1.7540909090909091,
+      "grad_norm": 2.526026425886204,
+      "learning_rate": 9.073913510956253e-07,
+      "loss": 0.0392,
+      "step": 7718
+    },
+    {
+      "epoch": 1.7543181818181819,
+      "grad_norm": 2.0981522149998137,
+      "learning_rate": 9.073117279421406e-07,
+      "loss": 0.0631,
+      "step": 7719
+    },
+    {
+      "epoch": 1.7545454545454544,
+      "grad_norm": 2.5891414633569747,
+      "learning_rate": 9.072320990318272e-07,
+      "loss": 0.0828,
+      "step": 7720
+    },
+    {
+      "epoch": 1.7547727272727274,
+      "grad_norm": 2.1800398030044374,
+      "learning_rate": 9.071524643663089e-07,
+      "loss": 0.0552,
+      "step": 7721
+    },
+    {
+      "epoch": 1.755,
+      "grad_norm": 2.5749386185592162,
+      "learning_rate": 9.070728239472096e-07,
+      "loss": 0.0734,
+      "step": 7722
+    },
+    {
+      "epoch": 1.7552272727272729,
+      "grad_norm": 1.384634745645385,
+      "learning_rate": 9.069931777761533e-07,
+      "loss": 0.0268,
+      "step": 7723
+    },
+    {
+      "epoch": 1.7554545454545454,
+      "grad_norm": 1.4436309033331705,
+      "learning_rate": 9.06913525854764e-07,
+      "loss": 0.0295,
+      "step": 7724
+    },
+    {
+      "epoch": 1.7556818181818183,
+      "grad_norm": 1.5808354572379617,
+      "learning_rate": 9.068338681846664e-07,
+      "loss": 0.0565,
+      "step": 7725
+    },
+    {
+      "epoch": 1.7559090909090909,
+      "grad_norm": 1.5886648147754534,
+      "learning_rate": 9.067542047674843e-07,
+      "loss": 0.0487,
+      "step": 7726
+    },
+    {
+      "epoch": 1.7561363636363636,
+      "grad_norm": 2.110810605776701,
+      "learning_rate": 9.066745356048424e-07,
+      "loss": 0.0674,
+      "step": 7727
+    },
+    {
+      "epoch": 1.7563636363636363,
+      "grad_norm": 1.809988298790082,
+      "learning_rate": 9.065948606983655e-07,
+      "loss": 0.0399,
+      "step": 7728
+    },
+    {
+      "epoch": 1.756590909090909,
+      "grad_norm": 1.6553740879213414,
+      "learning_rate": 9.06515180049678e-07,
+      "loss": 0.0332,
+      "step": 7729
+    },
+    {
+      "epoch": 1.7568181818181818,
+      "grad_norm": 2.4430598808201327,
+      "learning_rate": 9.06435493660405e-07,
+      "loss": 0.0668,
+      "step": 7730
+    },
+    {
+      "epoch": 1.7570454545454546,
+      "grad_norm": 3.5576972400188063,
+      "learning_rate": 9.063558015321711e-07,
+      "loss": 0.0942,
+      "step": 7731
+    },
+    {
+      "epoch": 1.7572727272727273,
+      "grad_norm": 1.9874332004506128,
+      "learning_rate": 9.062761036666015e-07,
+      "loss": 0.0609,
+      "step": 7732
+    },
+    {
+      "epoch": 1.7574999999999998,
+      "grad_norm": 4.058276736941139,
+      "learning_rate": 9.061964000653215e-07,
+      "loss": 0.0701,
+      "step": 7733
+    },
+    {
+      "epoch": 1.7577272727272728,
+      "grad_norm": 2.0479479679671724,
+      "learning_rate": 9.061166907299566e-07,
+      "loss": 0.0638,
+      "step": 7734
+    },
+    {
+      "epoch": 1.7579545454545453,
+      "grad_norm": 2.2961700989246907,
+      "learning_rate": 9.060369756621315e-07,
+      "loss": 0.0573,
+      "step": 7735
+    },
+    {
+      "epoch": 1.7581818181818183,
+      "grad_norm": 2.212348104294835,
+      "learning_rate": 9.059572548634725e-07,
+      "loss": 0.0691,
+      "step": 7736
+    },
+    {
+      "epoch": 1.7584090909090908,
+      "grad_norm": 2.0714878778742527,
+      "learning_rate": 9.058775283356047e-07,
+      "loss": 0.043,
+      "step": 7737
+    },
+    {
+      "epoch": 1.7586363636363638,
+      "grad_norm": 1.5322850109685582,
+      "learning_rate": 9.057977960801542e-07,
+      "loss": 0.033,
+      "step": 7738
+    },
+    {
+      "epoch": 1.7588636363636363,
+      "grad_norm": 1.8180394932770072,
+      "learning_rate": 9.057180580987468e-07,
+      "loss": 0.0506,
+      "step": 7739
+    },
+    {
+      "epoch": 1.759090909090909,
+      "grad_norm": 3.1589464076485396,
+      "learning_rate": 9.056383143930084e-07,
+      "loss": 0.0449,
+      "step": 7740
+    },
+    {
+      "epoch": 1.7593181818181818,
+      "grad_norm": 2.0736567719088574,
+      "learning_rate": 9.055585649645651e-07,
+      "loss": 0.0641,
+      "step": 7741
+    },
+    {
+      "epoch": 1.7595454545454545,
+      "grad_norm": 3.7841412841480317,
+      "learning_rate": 9.054788098150433e-07,
+      "loss": 0.0944,
+      "step": 7742
+    },
+    {
+      "epoch": 1.7597727272727273,
+      "grad_norm": 1.899054825211368,
+      "learning_rate": 9.053990489460693e-07,
+      "loss": 0.0453,
+      "step": 7743
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.8954855495567273,
+      "learning_rate": 9.053192823592693e-07,
+      "loss": 0.0651,
+      "step": 7744
+    },
+    {
+      "epoch": 1.7602272727272728,
+      "grad_norm": 2.2815288300592353,
+      "learning_rate": 9.052395100562704e-07,
+      "loss": 0.0416,
+      "step": 7745
+    },
+    {
+      "epoch": 1.7604545454545455,
+      "grad_norm": 1.3487489821007361,
+      "learning_rate": 9.051597320386986e-07,
+      "loss": 0.019,
+      "step": 7746
+    },
+    {
+      "epoch": 1.7606818181818182,
+      "grad_norm": 2.6963098305924156,
+      "learning_rate": 9.050799483081814e-07,
+      "loss": 0.08,
+      "step": 7747
+    },
+    {
+      "epoch": 1.7609090909090908,
+      "grad_norm": 2.1151235634326007,
+      "learning_rate": 9.050001588663453e-07,
+      "loss": 0.0563,
+      "step": 7748
+    },
+    {
+      "epoch": 1.7611363636363637,
+      "grad_norm": 2.979180030204256,
+      "learning_rate": 9.049203637148174e-07,
+      "loss": 0.0931,
+      "step": 7749
+    },
+    {
+      "epoch": 1.7613636363636362,
+      "grad_norm": 2.847048136998814,
+      "learning_rate": 9.048405628552251e-07,
+      "loss": 0.0443,
+      "step": 7750
+    },
+    {
+      "epoch": 1.7615909090909092,
+      "grad_norm": 2.1914513119200296,
+      "learning_rate": 9.047607562891954e-07,
+      "loss": 0.0614,
+      "step": 7751
+    },
+    {
+      "epoch": 1.7618181818181817,
+      "grad_norm": 3.4026796458247888,
+      "learning_rate": 9.04680944018356e-07,
+      "loss": 0.1124,
+      "step": 7752
+    },
+    {
+      "epoch": 1.7620454545454547,
+      "grad_norm": 1.1586656530802637,
+      "learning_rate": 9.04601126044334e-07,
+      "loss": 0.0264,
+      "step": 7753
+    },
+    {
+      "epoch": 1.7622727272727272,
+      "grad_norm": 2.159108666936158,
+      "learning_rate": 9.045213023687572e-07,
+      "loss": 0.0553,
+      "step": 7754
+    },
+    {
+      "epoch": 1.7625,
+      "grad_norm": 1.3112697877256994,
+      "learning_rate": 9.044414729932536e-07,
+      "loss": 0.0291,
+      "step": 7755
+    },
+    {
+      "epoch": 1.7627272727272727,
+      "grad_norm": 0.7809900624968678,
+      "learning_rate": 9.043616379194508e-07,
+      "loss": 0.0183,
+      "step": 7756
+    },
+    {
+      "epoch": 1.7629545454545454,
+      "grad_norm": 2.448140946129195,
+      "learning_rate": 9.042817971489767e-07,
+      "loss": 0.0399,
+      "step": 7757
+    },
+    {
+      "epoch": 1.7631818181818182,
+      "grad_norm": 1.5041861682219728,
+      "learning_rate": 9.042019506834598e-07,
+      "loss": 0.0497,
+      "step": 7758
+    },
+    {
+      "epoch": 1.763409090909091,
+      "grad_norm": 1.4758963950896482,
+      "learning_rate": 9.041220985245278e-07,
+      "loss": 0.0309,
+      "step": 7759
+    },
+    {
+      "epoch": 1.7636363636363637,
+      "grad_norm": 1.5493060050160898,
+      "learning_rate": 9.040422406738092e-07,
+      "loss": 0.0298,
+      "step": 7760
+    },
+    {
+      "epoch": 1.7638636363636364,
+      "grad_norm": 1.8916588245618189,
+      "learning_rate": 9.039623771329326e-07,
+      "loss": 0.052,
+      "step": 7761
+    },
+    {
+      "epoch": 1.7640909090909092,
+      "grad_norm": 1.7246732538522682,
+      "learning_rate": 9.038825079035265e-07,
+      "loss": 0.06,
+      "step": 7762
+    },
+    {
+      "epoch": 1.7643181818181817,
+      "grad_norm": 3.7937406194404177,
+      "learning_rate": 9.038026329872196e-07,
+      "loss": 0.092,
+      "step": 7763
+    },
+    {
+      "epoch": 1.7645454545454546,
+      "grad_norm": 1.1605911139980654,
+      "learning_rate": 9.037227523856404e-07,
+      "loss": 0.0375,
+      "step": 7764
+    },
+    {
+      "epoch": 1.7647727272727272,
+      "grad_norm": 1.254540644897159,
+      "learning_rate": 9.036428661004182e-07,
+      "loss": 0.0416,
+      "step": 7765
+    },
+    {
+      "epoch": 1.7650000000000001,
+      "grad_norm": 3.6176860650551403,
+      "learning_rate": 9.035629741331817e-07,
+      "loss": 0.0697,
+      "step": 7766
+    },
+    {
+      "epoch": 1.7652272727272726,
+      "grad_norm": 1.4801645819648062,
+      "learning_rate": 9.034830764855604e-07,
+      "loss": 0.0247,
+      "step": 7767
+    },
+    {
+      "epoch": 1.7654545454545456,
+      "grad_norm": 1.344978809300701,
+      "learning_rate": 9.03403173159183e-07,
+      "loss": 0.0479,
+      "step": 7768
+    },
+    {
+      "epoch": 1.7656818181818181,
+      "grad_norm": 3.6102261940511307,
+      "learning_rate": 9.033232641556794e-07,
+      "loss": 0.0877,
+      "step": 7769
+    },
+    {
+      "epoch": 1.7659090909090909,
+      "grad_norm": 3.0088235465763424,
+      "learning_rate": 9.032433494766789e-07,
+      "loss": 0.0481,
+      "step": 7770
+    },
+    {
+      "epoch": 1.7661363636363636,
+      "grad_norm": 2.5722789281017655,
+      "learning_rate": 9.031634291238109e-07,
+      "loss": 0.0686,
+      "step": 7771
+    },
+    {
+      "epoch": 1.7663636363636364,
+      "grad_norm": 2.8024818305275137,
+      "learning_rate": 9.030835030987057e-07,
+      "loss": 0.0603,
+      "step": 7772
+    },
+    {
+      "epoch": 1.766590909090909,
+      "grad_norm": 1.9972807670931432,
+      "learning_rate": 9.030035714029922e-07,
+      "loss": 0.0349,
+      "step": 7773
+    },
+    {
+      "epoch": 1.7668181818181818,
+      "grad_norm": 3.4898485427509156,
+      "learning_rate": 9.029236340383013e-07,
+      "loss": 0.0676,
+      "step": 7774
+    },
+    {
+      "epoch": 1.7670454545454546,
+      "grad_norm": 2.5494895596719447,
+      "learning_rate": 9.028436910062624e-07,
+      "loss": 0.0966,
+      "step": 7775
+    },
+    {
+      "epoch": 1.767272727272727,
+      "grad_norm": 3.4177384989018917,
+      "learning_rate": 9.027637423085061e-07,
+      "loss": 0.0715,
+      "step": 7776
+    },
+    {
+      "epoch": 1.7675,
+      "grad_norm": 2.28691829981354,
+      "learning_rate": 9.026837879466622e-07,
+      "loss": 0.0321,
+      "step": 7777
+    },
+    {
+      "epoch": 1.7677272727272726,
+      "grad_norm": 2.3129715482665563,
+      "learning_rate": 9.026038279223618e-07,
+      "loss": 0.0227,
+      "step": 7778
+    },
+    {
+      "epoch": 1.7679545454545456,
+      "grad_norm": 1.6159147450198397,
+      "learning_rate": 9.025238622372349e-07,
+      "loss": 0.0506,
+      "step": 7779
+    },
+    {
+      "epoch": 1.768181818181818,
+      "grad_norm": 2.391529663168054,
+      "learning_rate": 9.024438908929122e-07,
+      "loss": 0.0448,
+      "step": 7780
+    },
+    {
+      "epoch": 1.768409090909091,
+      "grad_norm": 2.8022845766102527,
+      "learning_rate": 9.023639138910246e-07,
+      "loss": 0.0994,
+      "step": 7781
+    },
+    {
+      "epoch": 1.7686363636363636,
+      "grad_norm": 1.6209511324856483,
+      "learning_rate": 9.022839312332026e-07,
+      "loss": 0.0407,
+      "step": 7782
+    },
+    {
+      "epoch": 1.7688636363636365,
+      "grad_norm": 1.7722012567394674,
+      "learning_rate": 9.022039429210779e-07,
+      "loss": 0.0524,
+      "step": 7783
+    },
+    {
+      "epoch": 1.769090909090909,
+      "grad_norm": 1.2390009249532223,
+      "learning_rate": 9.021239489562811e-07,
+      "loss": 0.0379,
+      "step": 7784
+    },
+    {
+      "epoch": 1.7693181818181818,
+      "grad_norm": 1.5436880405435158,
+      "learning_rate": 9.020439493404434e-07,
+      "loss": 0.017,
+      "step": 7785
+    },
+    {
+      "epoch": 1.7695454545454545,
+      "grad_norm": 2.485218137566382,
+      "learning_rate": 9.019639440751964e-07,
+      "loss": 0.0346,
+      "step": 7786
+    },
+    {
+      "epoch": 1.7697727272727273,
+      "grad_norm": 0.9363855547286567,
+      "learning_rate": 9.018839331621714e-07,
+      "loss": 0.0144,
+      "step": 7787
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 2.0627861320969876,
+      "learning_rate": 9.018039166029999e-07,
+      "loss": 0.0421,
+      "step": 7788
+    },
+    {
+      "epoch": 1.7702272727272728,
+      "grad_norm": 1.0523163502642623,
+      "learning_rate": 9.017238943993137e-07,
+      "loss": 0.0303,
+      "step": 7789
+    },
+    {
+      "epoch": 1.7704545454545455,
+      "grad_norm": 1.2848897016057923,
+      "learning_rate": 9.016438665527443e-07,
+      "loss": 0.0414,
+      "step": 7790
+    },
+    {
+      "epoch": 1.770681818181818,
+      "grad_norm": 1.9156922785758586,
+      "learning_rate": 9.015638330649239e-07,
+      "loss": 0.069,
+      "step": 7791
+    },
+    {
+      "epoch": 1.770909090909091,
+      "grad_norm": 1.2137996027369078,
+      "learning_rate": 9.014837939374845e-07,
+      "loss": 0.0249,
+      "step": 7792
+    },
+    {
+      "epoch": 1.7711363636363635,
+      "grad_norm": 1.3029841889656097,
+      "learning_rate": 9.014037491720581e-07,
+      "loss": 0.0371,
+      "step": 7793
+    },
+    {
+      "epoch": 1.7713636363636365,
+      "grad_norm": 1.7075432801415538,
+      "learning_rate": 9.013236987702773e-07,
+      "loss": 0.0334,
+      "step": 7794
+    },
+    {
+      "epoch": 1.771590909090909,
+      "grad_norm": 1.7968484706093562,
+      "learning_rate": 9.012436427337739e-07,
+      "loss": 0.0461,
+      "step": 7795
+    },
+    {
+      "epoch": 1.771818181818182,
+      "grad_norm": 3.8173458422087947,
+      "learning_rate": 9.011635810641809e-07,
+      "loss": 0.0773,
+      "step": 7796
+    },
+    {
+      "epoch": 1.7720454545454545,
+      "grad_norm": 1.0850453266856526,
+      "learning_rate": 9.010835137631306e-07,
+      "loss": 0.0358,
+      "step": 7797
+    },
+    {
+      "epoch": 1.7722727272727272,
+      "grad_norm": 2.394965007260899,
+      "learning_rate": 9.01003440832256e-07,
+      "loss": 0.0523,
+      "step": 7798
+    },
+    {
+      "epoch": 1.7725,
+      "grad_norm": 1.2676887359951772,
+      "learning_rate": 9.009233622731895e-07,
+      "loss": 0.0206,
+      "step": 7799
+    },
+    {
+      "epoch": 1.7727272727272727,
+      "grad_norm": 1.8178914868204468,
+      "learning_rate": 9.008432780875642e-07,
+      "loss": 0.0383,
+      "step": 7800
+    },
+    {
+      "epoch": 1.7729545454545454,
+      "grad_norm": 1.605073333151607,
+      "learning_rate": 9.007631882770133e-07,
+      "loss": 0.0377,
+      "step": 7801
+    },
+    {
+      "epoch": 1.7731818181818182,
+      "grad_norm": 3.063995590481074,
+      "learning_rate": 9.0068309284317e-07,
+      "loss": 0.0832,
+      "step": 7802
+    },
+    {
+      "epoch": 1.773409090909091,
+      "grad_norm": 1.476613202326439,
+      "learning_rate": 9.006029917876674e-07,
+      "loss": 0.0325,
+      "step": 7803
+    },
+    {
+      "epoch": 1.7736363636363637,
+      "grad_norm": 1.36985348408591,
+      "learning_rate": 9.00522885112139e-07,
+      "loss": 0.0456,
+      "step": 7804
+    },
+    {
+      "epoch": 1.7738636363636364,
+      "grad_norm": 2.2014964549924563,
+      "learning_rate": 9.004427728182184e-07,
+      "loss": 0.0392,
+      "step": 7805
+    },
+    {
+      "epoch": 1.774090909090909,
+      "grad_norm": 2.9069146103856585,
+      "learning_rate": 9.00362654907539e-07,
+      "loss": 0.0723,
+      "step": 7806
+    },
+    {
+      "epoch": 1.774318181818182,
+      "grad_norm": 3.547882513157458,
+      "learning_rate": 9.002825313817347e-07,
+      "loss": 0.1128,
+      "step": 7807
+    },
+    {
+      "epoch": 1.7745454545454544,
+      "grad_norm": 1.7360978068108377,
+      "learning_rate": 9.002024022424393e-07,
+      "loss": 0.053,
+      "step": 7808
+    },
+    {
+      "epoch": 1.7747727272727274,
+      "grad_norm": 2.4125668983041777,
+      "learning_rate": 9.001222674912868e-07,
+      "loss": 0.0469,
+      "step": 7809
+    },
+    {
+      "epoch": 1.775,
+      "grad_norm": 2.9968364840445374,
+      "learning_rate": 9.000421271299113e-07,
+      "loss": 0.0444,
+      "step": 7810
+    },
+    {
+      "epoch": 1.7752272727272729,
+      "grad_norm": 2.417747305510907,
+      "learning_rate": 8.99961981159947e-07,
+      "loss": 0.0475,
+      "step": 7811
+    },
+    {
+      "epoch": 1.7754545454545454,
+      "grad_norm": 2.3381532800549727,
+      "learning_rate": 8.998818295830283e-07,
+      "loss": 0.05,
+      "step": 7812
+    },
+    {
+      "epoch": 1.7756818181818181,
+      "grad_norm": 1.407882305224817,
+      "learning_rate": 8.998016724007895e-07,
+      "loss": 0.072,
+      "step": 7813
+    },
+    {
+      "epoch": 1.7759090909090909,
+      "grad_norm": 1.849418517940874,
+      "learning_rate": 8.997215096148651e-07,
+      "loss": 0.0314,
+      "step": 7814
+    },
+    {
+      "epoch": 1.7761363636363636,
+      "grad_norm": 3.3868989166866497,
+      "learning_rate": 8.996413412268898e-07,
+      "loss": 0.065,
+      "step": 7815
+    },
+    {
+      "epoch": 1.7763636363636364,
+      "grad_norm": 1.7956786107576086,
+      "learning_rate": 8.995611672384985e-07,
+      "loss": 0.0344,
+      "step": 7816
+    },
+    {
+      "epoch": 1.776590909090909,
+      "grad_norm": 1.4732976324870206,
+      "learning_rate": 8.99480987651326e-07,
+      "loss": 0.0531,
+      "step": 7817
+    },
+    {
+      "epoch": 1.7768181818181819,
+      "grad_norm": 2.7551732896599703,
+      "learning_rate": 8.994008024670072e-07,
+      "loss": 0.0991,
+      "step": 7818
+    },
+    {
+      "epoch": 1.7770454545454546,
+      "grad_norm": 2.1559585457779686,
+      "learning_rate": 8.993206116871775e-07,
+      "loss": 0.0279,
+      "step": 7819
+    },
+    {
+      "epoch": 1.7772727272727273,
+      "grad_norm": 3.049067654943818,
+      "learning_rate": 8.992404153134719e-07,
+      "loss": 0.1118,
+      "step": 7820
+    },
+    {
+      "epoch": 1.7774999999999999,
+      "grad_norm": 2.1765676475446876,
+      "learning_rate": 8.991602133475258e-07,
+      "loss": 0.0452,
+      "step": 7821
+    },
+    {
+      "epoch": 1.7777272727272728,
+      "grad_norm": 1.43776992477586,
+      "learning_rate": 8.990800057909745e-07,
+      "loss": 0.0503,
+      "step": 7822
+    },
+    {
+      "epoch": 1.7779545454545453,
+      "grad_norm": 1.857740214080316,
+      "learning_rate": 8.989997926454538e-07,
+      "loss": 0.0895,
+      "step": 7823
+    },
+    {
+      "epoch": 1.7781818181818183,
+      "grad_norm": 1.4827166813711763,
+      "learning_rate": 8.989195739125995e-07,
+      "loss": 0.0339,
+      "step": 7824
+    },
+    {
+      "epoch": 1.7784090909090908,
+      "grad_norm": 1.9750230112459777,
+      "learning_rate": 8.988393495940469e-07,
+      "loss": 0.0617,
+      "step": 7825
+    },
+    {
+      "epoch": 1.7786363636363638,
+      "grad_norm": 1.138022704038579,
+      "learning_rate": 8.987591196914325e-07,
+      "loss": 0.0285,
+      "step": 7826
+    },
+    {
+      "epoch": 1.7788636363636363,
+      "grad_norm": 2.1852582450323412,
+      "learning_rate": 8.986788842063918e-07,
+      "loss": 0.0777,
+      "step": 7827
+    },
+    {
+      "epoch": 1.779090909090909,
+      "grad_norm": 1.924959185550527,
+      "learning_rate": 8.985986431405613e-07,
+      "loss": 0.059,
+      "step": 7828
+    },
+    {
+      "epoch": 1.7793181818181818,
+      "grad_norm": 1.475135623805425,
+      "learning_rate": 8.985183964955773e-07,
+      "loss": 0.0544,
+      "step": 7829
+    },
+    {
+      "epoch": 1.7795454545454545,
+      "grad_norm": 1.2349559194357875,
+      "learning_rate": 8.984381442730756e-07,
+      "loss": 0.0293,
+      "step": 7830
+    },
+    {
+      "epoch": 1.7797727272727273,
+      "grad_norm": 2.728725560204543,
+      "learning_rate": 8.983578864746935e-07,
+      "loss": 0.0935,
+      "step": 7831
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.6355227671748191,
+      "learning_rate": 8.98277623102067e-07,
+      "loss": 0.0406,
+      "step": 7832
+    },
+    {
+      "epoch": 1.7802272727272728,
+      "grad_norm": 1.1004354226495177,
+      "learning_rate": 8.981973541568331e-07,
+      "loss": 0.0409,
+      "step": 7833
+    },
+    {
+      "epoch": 1.7804545454545453,
+      "grad_norm": 1.7223723317946846,
+      "learning_rate": 8.981170796406285e-07,
+      "loss": 0.0501,
+      "step": 7834
+    },
+    {
+      "epoch": 1.7806818181818183,
+      "grad_norm": 1.9056647442074197,
+      "learning_rate": 8.980367995550903e-07,
+      "loss": 0.0668,
+      "step": 7835
+    },
+    {
+      "epoch": 1.7809090909090908,
+      "grad_norm": 1.9568337991847666,
+      "learning_rate": 8.979565139018554e-07,
+      "loss": 0.0575,
+      "step": 7836
+    },
+    {
+      "epoch": 1.7811363636363637,
+      "grad_norm": 1.6365695264398303,
+      "learning_rate": 8.978762226825608e-07,
+      "loss": 0.0409,
+      "step": 7837
+    },
+    {
+      "epoch": 1.7813636363636363,
+      "grad_norm": 1.817751572443709,
+      "learning_rate": 8.977959258988439e-07,
+      "loss": 0.0385,
+      "step": 7838
+    },
+    {
+      "epoch": 1.7815909090909092,
+      "grad_norm": 1.512338690362937,
+      "learning_rate": 8.977156235523423e-07,
+      "loss": 0.0218,
+      "step": 7839
+    },
+    {
+      "epoch": 1.7818181818181817,
+      "grad_norm": 2.304113594638741,
+      "learning_rate": 8.976353156446934e-07,
+      "loss": 0.0586,
+      "step": 7840
+    },
+    {
+      "epoch": 1.7820454545454547,
+      "grad_norm": 1.1379168714354186,
+      "learning_rate": 8.975550021775346e-07,
+      "loss": 0.0162,
+      "step": 7841
+    },
+    {
+      "epoch": 1.7822727272727272,
+      "grad_norm": 2.4344622657178485,
+      "learning_rate": 8.974746831525041e-07,
+      "loss": 0.0791,
+      "step": 7842
+    },
+    {
+      "epoch": 1.7825,
+      "grad_norm": 1.4255848853695057,
+      "learning_rate": 8.973943585712392e-07,
+      "loss": 0.0449,
+      "step": 7843
+    },
+    {
+      "epoch": 1.7827272727272727,
+      "grad_norm": 1.7980507688101182,
+      "learning_rate": 8.973140284353784e-07,
+      "loss": 0.0654,
+      "step": 7844
+    },
+    {
+      "epoch": 1.7829545454545455,
+      "grad_norm": 1.5722063743630719,
+      "learning_rate": 8.97233692746559e-07,
+      "loss": 0.0378,
+      "step": 7845
+    },
+    {
+      "epoch": 1.7831818181818182,
+      "grad_norm": 1.633565234829344,
+      "learning_rate": 8.9715335150642e-07,
+      "loss": 0.0306,
+      "step": 7846
+    },
+    {
+      "epoch": 1.783409090909091,
+      "grad_norm": 1.4808715653746238,
+      "learning_rate": 8.970730047165993e-07,
+      "loss": 0.0372,
+      "step": 7847
+    },
+    {
+      "epoch": 1.7836363636363637,
+      "grad_norm": 2.1268516558119774,
+      "learning_rate": 8.969926523787355e-07,
+      "loss": 0.0467,
+      "step": 7848
+    },
+    {
+      "epoch": 1.7838636363636362,
+      "grad_norm": 1.23231537197342,
+      "learning_rate": 8.969122944944668e-07,
+      "loss": 0.032,
+      "step": 7849
+    },
+    {
+      "epoch": 1.7840909090909092,
+      "grad_norm": 1.7833375708344226,
+      "learning_rate": 8.968319310654321e-07,
+      "loss": 0.0499,
+      "step": 7850
+    },
+    {
+      "epoch": 1.7843181818181817,
+      "grad_norm": 2.7543991747104473,
+      "learning_rate": 8.967515620932702e-07,
+      "loss": 0.0543,
+      "step": 7851
+    },
+    {
+      "epoch": 1.7845454545454547,
+      "grad_norm": 2.3508674504457794,
+      "learning_rate": 8.966711875796197e-07,
+      "loss": 0.0947,
+      "step": 7852
+    },
+    {
+      "epoch": 1.7847727272727272,
+      "grad_norm": 1.336350474231818,
+      "learning_rate": 8.965908075261197e-07,
+      "loss": 0.0254,
+      "step": 7853
+    },
+    {
+      "epoch": 1.7850000000000001,
+      "grad_norm": 2.616413252299649,
+      "learning_rate": 8.965104219344094e-07,
+      "loss": 0.055,
+      "step": 7854
+    },
+    {
+      "epoch": 1.7852272727272727,
+      "grad_norm": 1.8629930521643099,
+      "learning_rate": 8.96430030806128e-07,
+      "loss": 0.0396,
+      "step": 7855
+    },
+    {
+      "epoch": 1.7854545454545454,
+      "grad_norm": 1.852368266982457,
+      "learning_rate": 8.963496341429145e-07,
+      "loss": 0.0477,
+      "step": 7856
+    },
+    {
+      "epoch": 1.7856818181818181,
+      "grad_norm": 1.9475293499515842,
+      "learning_rate": 8.96269231946409e-07,
+      "loss": 0.0298,
+      "step": 7857
+    },
+    {
+      "epoch": 1.785909090909091,
+      "grad_norm": 1.487379401707851,
+      "learning_rate": 8.961888242182502e-07,
+      "loss": 0.0466,
+      "step": 7858
+    },
+    {
+      "epoch": 1.7861363636363636,
+      "grad_norm": 1.277842767921913,
+      "learning_rate": 8.961084109600785e-07,
+      "loss": 0.0224,
+      "step": 7859
+    },
+    {
+      "epoch": 1.7863636363636364,
+      "grad_norm": 1.4697244095433515,
+      "learning_rate": 8.960279921735329e-07,
+      "loss": 0.0351,
+      "step": 7860
+    },
+    {
+      "epoch": 1.7865909090909091,
+      "grad_norm": 2.8535837705835037,
+      "learning_rate": 8.959475678602539e-07,
+      "loss": 0.0815,
+      "step": 7861
+    },
+    {
+      "epoch": 1.7868181818181819,
+      "grad_norm": 1.723818598533904,
+      "learning_rate": 8.958671380218814e-07,
+      "loss": 0.0359,
+      "step": 7862
+    },
+    {
+      "epoch": 1.7870454545454546,
+      "grad_norm": 2.439439334387233,
+      "learning_rate": 8.957867026600553e-07,
+      "loss": 0.0551,
+      "step": 7863
+    },
+    {
+      "epoch": 1.7872727272727271,
+      "grad_norm": 3.3163045214332527,
+      "learning_rate": 8.957062617764159e-07,
+      "loss": 0.0445,
+      "step": 7864
+    },
+    {
+      "epoch": 1.7875,
+      "grad_norm": 3.934310650645701,
+      "learning_rate": 8.956258153726035e-07,
+      "loss": 0.07,
+      "step": 7865
+    },
+    {
+      "epoch": 1.7877272727272726,
+      "grad_norm": 1.5597325564292228,
+      "learning_rate": 8.95545363450259e-07,
+      "loss": 0.0308,
+      "step": 7866
+    },
+    {
+      "epoch": 1.7879545454545456,
+      "grad_norm": 3.3745456471278246,
+      "learning_rate": 8.954649060110221e-07,
+      "loss": 0.0495,
+      "step": 7867
+    },
+    {
+      "epoch": 1.788181818181818,
+      "grad_norm": 2.0748947953450334,
+      "learning_rate": 8.95384443056534e-07,
+      "loss": 0.0443,
+      "step": 7868
+    },
+    {
+      "epoch": 1.788409090909091,
+      "grad_norm": 0.9787468268846439,
+      "learning_rate": 8.953039745884354e-07,
+      "loss": 0.0233,
+      "step": 7869
+    },
+    {
+      "epoch": 1.7886363636363636,
+      "grad_norm": 1.8628027278554704,
+      "learning_rate": 8.952235006083674e-07,
+      "loss": 0.0652,
+      "step": 7870
+    },
+    {
+      "epoch": 1.7888636363636363,
+      "grad_norm": 2.059791600175135,
+      "learning_rate": 8.951430211179704e-07,
+      "loss": 0.0495,
+      "step": 7871
+    },
+    {
+      "epoch": 1.789090909090909,
+      "grad_norm": 1.8895670065199026,
+      "learning_rate": 8.950625361188861e-07,
+      "loss": 0.0426,
+      "step": 7872
+    },
+    {
+      "epoch": 1.7893181818181818,
+      "grad_norm": 3.7571070090477097,
+      "learning_rate": 8.949820456127557e-07,
+      "loss": 0.0421,
+      "step": 7873
+    },
+    {
+      "epoch": 1.7895454545454546,
+      "grad_norm": 1.6840878363947187,
+      "learning_rate": 8.949015496012202e-07,
+      "loss": 0.0655,
+      "step": 7874
+    },
+    {
+      "epoch": 1.7897727272727273,
+      "grad_norm": 1.4499566405663284,
+      "learning_rate": 8.948210480859212e-07,
+      "loss": 0.0264,
+      "step": 7875
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 3.3202501130989104,
+      "learning_rate": 8.947405410685004e-07,
+      "loss": 0.1303,
+      "step": 7876
+    },
+    {
+      "epoch": 1.7902272727272728,
+      "grad_norm": 2.4506869176525585,
+      "learning_rate": 8.946600285505995e-07,
+      "loss": 0.0572,
+      "step": 7877
+    },
+    {
+      "epoch": 1.7904545454545455,
+      "grad_norm": 5.3950548171260095,
+      "learning_rate": 8.9457951053386e-07,
+      "loss": 0.1275,
+      "step": 7878
+    },
+    {
+      "epoch": 1.790681818181818,
+      "grad_norm": 1.4405369990814976,
+      "learning_rate": 8.94498987019924e-07,
+      "loss": 0.0283,
+      "step": 7879
+    },
+    {
+      "epoch": 1.790909090909091,
+      "grad_norm": 3.1281461241771016,
+      "learning_rate": 8.944184580104334e-07,
+      "loss": 0.0895,
+      "step": 7880
+    },
+    {
+      "epoch": 1.7911363636363635,
+      "grad_norm": 1.7490986906202552,
+      "learning_rate": 8.943379235070307e-07,
+      "loss": 0.0373,
+      "step": 7881
+    },
+    {
+      "epoch": 1.7913636363636365,
+      "grad_norm": 1.4901225592140592,
+      "learning_rate": 8.942573835113575e-07,
+      "loss": 0.0375,
+      "step": 7882
+    },
+    {
+      "epoch": 1.791590909090909,
+      "grad_norm": 1.8011083351094166,
+      "learning_rate": 8.941768380250568e-07,
+      "loss": 0.0334,
+      "step": 7883
+    },
+    {
+      "epoch": 1.791818181818182,
+      "grad_norm": 0.8940488836485314,
+      "learning_rate": 8.940962870497705e-07,
+      "loss": 0.0241,
+      "step": 7884
+    },
+    {
+      "epoch": 1.7920454545454545,
+      "grad_norm": 1.271994350603335,
+      "learning_rate": 8.940157305871417e-07,
+      "loss": 0.0346,
+      "step": 7885
+    },
+    {
+      "epoch": 1.7922727272727272,
+      "grad_norm": 2.2602745802838373,
+      "learning_rate": 8.939351686388127e-07,
+      "loss": 0.0564,
+      "step": 7886
+    },
+    {
+      "epoch": 1.7925,
+      "grad_norm": 1.3290383654029327,
+      "learning_rate": 8.938546012064265e-07,
+      "loss": 0.0423,
+      "step": 7887
+    },
+    {
+      "epoch": 1.7927272727272727,
+      "grad_norm": 1.6730189382202283,
+      "learning_rate": 8.937740282916259e-07,
+      "loss": 0.0279,
+      "step": 7888
+    },
+    {
+      "epoch": 1.7929545454545455,
+      "grad_norm": 1.8073847484112688,
+      "learning_rate": 8.936934498960538e-07,
+      "loss": 0.0591,
+      "step": 7889
+    },
+    {
+      "epoch": 1.7931818181818182,
+      "grad_norm": 2.0241755948695292,
+      "learning_rate": 8.936128660213536e-07,
+      "loss": 0.0322,
+      "step": 7890
+    },
+    {
+      "epoch": 1.793409090909091,
+      "grad_norm": 1.5783737558212185,
+      "learning_rate": 8.935322766691684e-07,
+      "loss": 0.0583,
+      "step": 7891
+    },
+    {
+      "epoch": 1.7936363636363635,
+      "grad_norm": 2.208665643431723,
+      "learning_rate": 8.934516818411418e-07,
+      "loss": 0.0596,
+      "step": 7892
+    },
+    {
+      "epoch": 1.7938636363636364,
+      "grad_norm": 1.3161330312748196,
+      "learning_rate": 8.933710815389168e-07,
+      "loss": 0.0313,
+      "step": 7893
+    },
+    {
+      "epoch": 1.794090909090909,
+      "grad_norm": 3.0189362149638477,
+      "learning_rate": 8.932904757641375e-07,
+      "loss": 0.1057,
+      "step": 7894
+    },
+    {
+      "epoch": 1.794318181818182,
+      "grad_norm": 1.643413811708223,
+      "learning_rate": 8.932098645184472e-07,
+      "loss": 0.025,
+      "step": 7895
+    },
+    {
+      "epoch": 1.7945454545454544,
+      "grad_norm": 2.317294209559219,
+      "learning_rate": 8.931292478034899e-07,
+      "loss": 0.059,
+      "step": 7896
+    },
+    {
+      "epoch": 1.7947727272727274,
+      "grad_norm": 2.612836594455801,
+      "learning_rate": 8.930486256209095e-07,
+      "loss": 0.07,
+      "step": 7897
+    },
+    {
+      "epoch": 1.795,
+      "grad_norm": 1.9830624339810572,
+      "learning_rate": 8.929679979723499e-07,
+      "loss": 0.0682,
+      "step": 7898
+    },
+    {
+      "epoch": 1.795227272727273,
+      "grad_norm": 2.463257856960012,
+      "learning_rate": 8.928873648594554e-07,
+      "loss": 0.0856,
+      "step": 7899
+    },
+    {
+      "epoch": 1.7954545454545454,
+      "grad_norm": 1.1714966076514697,
+      "learning_rate": 8.9280672628387e-07,
+      "loss": 0.0442,
+      "step": 7900
+    },
+    {
+      "epoch": 1.7956818181818182,
+      "grad_norm": 1.4792328321697932,
+      "learning_rate": 8.927260822472384e-07,
+      "loss": 0.026,
+      "step": 7901
+    },
+    {
+      "epoch": 1.795909090909091,
+      "grad_norm": 2.7385696552128693,
+      "learning_rate": 8.926454327512049e-07,
+      "loss": 0.0897,
+      "step": 7902
+    },
+    {
+      "epoch": 1.7961363636363636,
+      "grad_norm": 2.633374816885308,
+      "learning_rate": 8.92564777797414e-07,
+      "loss": 0.0553,
+      "step": 7903
+    },
+    {
+      "epoch": 1.7963636363636364,
+      "grad_norm": 2.135351948355052,
+      "learning_rate": 8.924841173875107e-07,
+      "loss": 0.0349,
+      "step": 7904
+    },
+    {
+      "epoch": 1.7965909090909091,
+      "grad_norm": 2.4797892857005572,
+      "learning_rate": 8.924034515231393e-07,
+      "loss": 0.0651,
+      "step": 7905
+    },
+    {
+      "epoch": 1.7968181818181819,
+      "grad_norm": 1.7118050807272398,
+      "learning_rate": 8.92322780205945e-07,
+      "loss": 0.0605,
+      "step": 7906
+    },
+    {
+      "epoch": 1.7970454545454544,
+      "grad_norm": 1.8182142869916402,
+      "learning_rate": 8.92242103437573e-07,
+      "loss": 0.0695,
+      "step": 7907
+    },
+    {
+      "epoch": 1.7972727272727274,
+      "grad_norm": 1.207286859732638,
+      "learning_rate": 8.921614212196685e-07,
+      "loss": 0.0315,
+      "step": 7908
+    },
+    {
+      "epoch": 1.7974999999999999,
+      "grad_norm": 1.66968734897308,
+      "learning_rate": 8.920807335538761e-07,
+      "loss": 0.0301,
+      "step": 7909
+    },
+    {
+      "epoch": 1.7977272727272728,
+      "grad_norm": 2.1424959292653276,
+      "learning_rate": 8.920000404418421e-07,
+      "loss": 0.0345,
+      "step": 7910
+    },
+    {
+      "epoch": 1.7979545454545454,
+      "grad_norm": 2.7246051048496254,
+      "learning_rate": 8.91919341885211e-07,
+      "loss": 0.0892,
+      "step": 7911
+    },
+    {
+      "epoch": 1.7981818181818183,
+      "grad_norm": 1.902397780041678,
+      "learning_rate": 8.918386378856292e-07,
+      "loss": 0.0387,
+      "step": 7912
+    },
+    {
+      "epoch": 1.7984090909090908,
+      "grad_norm": 2.0424761762865056,
+      "learning_rate": 8.917579284447418e-07,
+      "loss": 0.0592,
+      "step": 7913
+    },
+    {
+      "epoch": 1.7986363636363636,
+      "grad_norm": 2.2570519225485333,
+      "learning_rate": 8.916772135641949e-07,
+      "loss": 0.0574,
+      "step": 7914
+    },
+    {
+      "epoch": 1.7988636363636363,
+      "grad_norm": 2.2282289872762773,
+      "learning_rate": 8.915964932456343e-07,
+      "loss": 0.0388,
+      "step": 7915
+    },
+    {
+      "epoch": 1.799090909090909,
+      "grad_norm": 2.5973547345485812,
+      "learning_rate": 8.915157674907062e-07,
+      "loss": 0.0634,
+      "step": 7916
+    },
+    {
+      "epoch": 1.7993181818181818,
+      "grad_norm": 2.0374627833114265,
+      "learning_rate": 8.914350363010567e-07,
+      "loss": 0.0383,
+      "step": 7917
+    },
+    {
+      "epoch": 1.7995454545454546,
+      "grad_norm": 2.231289260503037,
+      "learning_rate": 8.91354299678332e-07,
+      "loss": 0.0916,
+      "step": 7918
+    },
+    {
+      "epoch": 1.7997727272727273,
+      "grad_norm": 2.003402260585324,
+      "learning_rate": 8.912735576241783e-07,
+      "loss": 0.0315,
+      "step": 7919
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 2.1664685117041156,
+      "learning_rate": 8.911928101402423e-07,
+      "loss": 0.0759,
+      "step": 7920
+    },
+    {
+      "epoch": 1.8002272727272728,
+      "grad_norm": 2.5982424199177867,
+      "learning_rate": 8.911120572281705e-07,
+      "loss": 0.1027,
+      "step": 7921
+    },
+    {
+      "epoch": 1.8004545454545453,
+      "grad_norm": 1.9009914016027256,
+      "learning_rate": 8.910312988896095e-07,
+      "loss": 0.0431,
+      "step": 7922
+    },
+    {
+      "epoch": 1.8006818181818183,
+      "grad_norm": 1.140079353844668,
+      "learning_rate": 8.909505351262064e-07,
+      "loss": 0.0254,
+      "step": 7923
+    },
+    {
+      "epoch": 1.8009090909090908,
+      "grad_norm": 1.4082652618440161,
+      "learning_rate": 8.908697659396077e-07,
+      "loss": 0.0289,
+      "step": 7924
+    },
+    {
+      "epoch": 1.8011363636363638,
+      "grad_norm": 2.7767711324036384,
+      "learning_rate": 8.907889913314609e-07,
+      "loss": 0.0421,
+      "step": 7925
+    },
+    {
+      "epoch": 1.8013636363636363,
+      "grad_norm": 1.5309111699011773,
+      "learning_rate": 8.907082113034126e-07,
+      "loss": 0.0657,
+      "step": 7926
+    },
+    {
+      "epoch": 1.8015909090909092,
+      "grad_norm": 1.7092781810995985,
+      "learning_rate": 8.906274258571105e-07,
+      "loss": 0.0336,
+      "step": 7927
+    },
+    {
+      "epoch": 1.8018181818181818,
+      "grad_norm": 1.6178697096153027,
+      "learning_rate": 8.905466349942017e-07,
+      "loss": 0.044,
+      "step": 7928
+    },
+    {
+      "epoch": 1.8020454545454545,
+      "grad_norm": 1.4531154453894939,
+      "learning_rate": 8.904658387163337e-07,
+      "loss": 0.0249,
+      "step": 7929
+    },
+    {
+      "epoch": 1.8022727272727272,
+      "grad_norm": 0.8530892509433383,
+      "learning_rate": 8.903850370251541e-07,
+      "loss": 0.0165,
+      "step": 7930
+    },
+    {
+      "epoch": 1.8025,
+      "grad_norm": 2.7923348710402123,
+      "learning_rate": 8.903042299223108e-07,
+      "loss": 0.0713,
+      "step": 7931
+    },
+    {
+      "epoch": 1.8027272727272727,
+      "grad_norm": 1.526463379833029,
+      "learning_rate": 8.902234174094513e-07,
+      "loss": 0.0439,
+      "step": 7932
+    },
+    {
+      "epoch": 1.8029545454545455,
+      "grad_norm": 1.5034015699034304,
+      "learning_rate": 8.901425994882234e-07,
+      "loss": 0.0386,
+      "step": 7933
+    },
+    {
+      "epoch": 1.8031818181818182,
+      "grad_norm": 1.971189058497508,
+      "learning_rate": 8.900617761602758e-07,
+      "loss": 0.0492,
+      "step": 7934
+    },
+    {
+      "epoch": 1.803409090909091,
+      "grad_norm": 2.2945513895133476,
+      "learning_rate": 8.899809474272558e-07,
+      "loss": 0.0373,
+      "step": 7935
+    },
+    {
+      "epoch": 1.8036363636363637,
+      "grad_norm": 2.245347507647015,
+      "learning_rate": 8.89900113290812e-07,
+      "loss": 0.0402,
+      "step": 7936
+    },
+    {
+      "epoch": 1.8038636363636362,
+      "grad_norm": 2.0329537063843226,
+      "learning_rate": 8.898192737525928e-07,
+      "loss": 0.0498,
+      "step": 7937
+    },
+    {
+      "epoch": 1.8040909090909092,
+      "grad_norm": 2.0102204713891334,
+      "learning_rate": 8.897384288142466e-07,
+      "loss": 0.0599,
+      "step": 7938
+    },
+    {
+      "epoch": 1.8043181818181817,
+      "grad_norm": 1.2078386183461767,
+      "learning_rate": 8.896575784774221e-07,
+      "loss": 0.0264,
+      "step": 7939
+    },
+    {
+      "epoch": 1.8045454545454547,
+      "grad_norm": 1.7137702153457892,
+      "learning_rate": 8.895767227437676e-07,
+      "loss": 0.0471,
+      "step": 7940
+    },
+    {
+      "epoch": 1.8047727272727272,
+      "grad_norm": 3.910860533054162,
+      "learning_rate": 8.894958616149324e-07,
+      "loss": 0.1348,
+      "step": 7941
+    },
+    {
+      "epoch": 1.8050000000000002,
+      "grad_norm": 1.7668939499998237,
+      "learning_rate": 8.894149950925651e-07,
+      "loss": 0.0575,
+      "step": 7942
+    },
+    {
+      "epoch": 1.8052272727272727,
+      "grad_norm": 2.121918072517211,
+      "learning_rate": 8.893341231783145e-07,
+      "loss": 0.0296,
+      "step": 7943
+    },
+    {
+      "epoch": 1.8054545454545454,
+      "grad_norm": 1.85136061466438,
+      "learning_rate": 8.892532458738302e-07,
+      "loss": 0.0609,
+      "step": 7944
+    },
+    {
+      "epoch": 1.8056818181818182,
+      "grad_norm": 1.9426072241883041,
+      "learning_rate": 8.891723631807613e-07,
+      "loss": 0.0678,
+      "step": 7945
+    },
+    {
+      "epoch": 1.805909090909091,
+      "grad_norm": 1.5653721839944663,
+      "learning_rate": 8.890914751007568e-07,
+      "loss": 0.0557,
+      "step": 7946
+    },
+    {
+      "epoch": 1.8061363636363637,
+      "grad_norm": 1.738571386137214,
+      "learning_rate": 8.890105816354666e-07,
+      "loss": 0.0322,
+      "step": 7947
+    },
+    {
+      "epoch": 1.8063636363636364,
+      "grad_norm": 1.4043316638544674,
+      "learning_rate": 8.889296827865398e-07,
+      "loss": 0.0419,
+      "step": 7948
+    },
+    {
+      "epoch": 1.8065909090909091,
+      "grad_norm": 1.652543282734235,
+      "learning_rate": 8.888487785556267e-07,
+      "loss": 0.0458,
+      "step": 7949
+    },
+    {
+      "epoch": 1.8068181818181817,
+      "grad_norm": 1.8314751038128971,
+      "learning_rate": 8.887678689443764e-07,
+      "loss": 0.0754,
+      "step": 7950
+    },
+    {
+      "epoch": 1.8070454545454546,
+      "grad_norm": 1.9553678374151817,
+      "learning_rate": 8.88686953954439e-07,
+      "loss": 0.0487,
+      "step": 7951
+    },
+    {
+      "epoch": 1.8072727272727271,
+      "grad_norm": 1.7679219736982703,
+      "learning_rate": 8.886060335874648e-07,
+      "loss": 0.0478,
+      "step": 7952
+    },
+    {
+      "epoch": 1.8075,
+      "grad_norm": 3.072290015020925,
+      "learning_rate": 8.885251078451035e-07,
+      "loss": 0.0992,
+      "step": 7953
+    },
+    {
+      "epoch": 1.8077272727272726,
+      "grad_norm": 2.2792705654993193,
+      "learning_rate": 8.884441767290059e-07,
+      "loss": 0.0729,
+      "step": 7954
+    },
+    {
+      "epoch": 1.8079545454545456,
+      "grad_norm": 2.6050375814225197,
+      "learning_rate": 8.883632402408215e-07,
+      "loss": 0.0867,
+      "step": 7955
+    },
+    {
+      "epoch": 1.8081818181818181,
+      "grad_norm": 1.950764760180847,
+      "learning_rate": 8.882822983822014e-07,
+      "loss": 0.0715,
+      "step": 7956
+    },
+    {
+      "epoch": 1.8084090909090909,
+      "grad_norm": 1.4780765028055698,
+      "learning_rate": 8.882013511547958e-07,
+      "loss": 0.0251,
+      "step": 7957
+    },
+    {
+      "epoch": 1.8086363636363636,
+      "grad_norm": 1.824492084640513,
+      "learning_rate": 8.881203985602554e-07,
+      "loss": 0.0508,
+      "step": 7958
+    },
+    {
+      "epoch": 1.8088636363636363,
+      "grad_norm": 1.3026583667729699,
+      "learning_rate": 8.880394406002312e-07,
+      "loss": 0.0607,
+      "step": 7959
+    },
+    {
+      "epoch": 1.809090909090909,
+      "grad_norm": 1.6193776920820435,
+      "learning_rate": 8.879584772763739e-07,
+      "loss": 0.0265,
+      "step": 7960
+    },
+    {
+      "epoch": 1.8093181818181818,
+      "grad_norm": 1.8648401725599226,
+      "learning_rate": 8.878775085903344e-07,
+      "loss": 0.0469,
+      "step": 7961
+    },
+    {
+      "epoch": 1.8095454545454546,
+      "grad_norm": 2.4668350419132237,
+      "learning_rate": 8.877965345437639e-07,
+      "loss": 0.1022,
+      "step": 7962
+    },
+    {
+      "epoch": 1.8097727272727273,
+      "grad_norm": 1.9084462826495985,
+      "learning_rate": 8.877155551383138e-07,
+      "loss": 0.0483,
+      "step": 7963
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 2.186178875041265,
+      "learning_rate": 8.876345703756348e-07,
+      "loss": 0.0573,
+      "step": 7964
+    },
+    {
+      "epoch": 1.8102272727272726,
+      "grad_norm": 2.3828586936737457,
+      "learning_rate": 8.875535802573792e-07,
+      "loss": 0.051,
+      "step": 7965
+    },
+    {
+      "epoch": 1.8104545454545455,
+      "grad_norm": 1.5036484487200994,
+      "learning_rate": 8.874725847851976e-07,
+      "loss": 0.03,
+      "step": 7966
+    },
+    {
+      "epoch": 1.810681818181818,
+      "grad_norm": 1.4829215157261708,
+      "learning_rate": 8.873915839607422e-07,
+      "loss": 0.0369,
+      "step": 7967
+    },
+    {
+      "epoch": 1.810909090909091,
+      "grad_norm": 2.2235782208050376,
+      "learning_rate": 8.873105777856647e-07,
+      "loss": 0.0559,
+      "step": 7968
+    },
+    {
+      "epoch": 1.8111363636363635,
+      "grad_norm": 1.5711338638597638,
+      "learning_rate": 8.872295662616171e-07,
+      "loss": 0.0298,
+      "step": 7969
+    },
+    {
+      "epoch": 1.8113636363636365,
+      "grad_norm": 1.6777988170115539,
+      "learning_rate": 8.871485493902508e-07,
+      "loss": 0.0442,
+      "step": 7970
+    },
+    {
+      "epoch": 1.811590909090909,
+      "grad_norm": 1.864328693019573,
+      "learning_rate": 8.870675271732184e-07,
+      "loss": 0.0734,
+      "step": 7971
+    },
+    {
+      "epoch": 1.8118181818181818,
+      "grad_norm": 3.2521752332172027,
+      "learning_rate": 8.869864996121719e-07,
+      "loss": 0.0644,
+      "step": 7972
+    },
+    {
+      "epoch": 1.8120454545454545,
+      "grad_norm": 2.17964277382715,
+      "learning_rate": 8.869054667087637e-07,
+      "loss": 0.0817,
+      "step": 7973
+    },
+    {
+      "epoch": 1.8122727272727273,
+      "grad_norm": 2.633904264733123,
+      "learning_rate": 8.86824428464646e-07,
+      "loss": 0.0717,
+      "step": 7974
+    },
+    {
+      "epoch": 1.8125,
+      "grad_norm": 2.1486139686034247,
+      "learning_rate": 8.867433848814714e-07,
+      "loss": 0.0575,
+      "step": 7975
+    },
+    {
+      "epoch": 1.8127272727272727,
+      "grad_norm": 1.7414786902463975,
+      "learning_rate": 8.866623359608927e-07,
+      "loss": 0.039,
+      "step": 7976
+    },
+    {
+      "epoch": 1.8129545454545455,
+      "grad_norm": 0.9954111147521801,
+      "learning_rate": 8.865812817045623e-07,
+      "loss": 0.0277,
+      "step": 7977
+    },
+    {
+      "epoch": 1.8131818181818182,
+      "grad_norm": 2.97547960284437,
+      "learning_rate": 8.865002221141333e-07,
+      "loss": 0.0782,
+      "step": 7978
+    },
+    {
+      "epoch": 1.813409090909091,
+      "grad_norm": 2.0396859376701375,
+      "learning_rate": 8.864191571912587e-07,
+      "loss": 0.0389,
+      "step": 7979
+    },
+    {
+      "epoch": 1.8136363636363635,
+      "grad_norm": 0.9329194580416932,
+      "learning_rate": 8.863380869375912e-07,
+      "loss": 0.0363,
+      "step": 7980
+    },
+    {
+      "epoch": 1.8138636363636365,
+      "grad_norm": 2.097731942300446,
+      "learning_rate": 8.862570113547843e-07,
+      "loss": 0.0704,
+      "step": 7981
+    },
+    {
+      "epoch": 1.814090909090909,
+      "grad_norm": 2.0170442961953046,
+      "learning_rate": 8.86175930444491e-07,
+      "loss": 0.0525,
+      "step": 7982
+    },
+    {
+      "epoch": 1.814318181818182,
+      "grad_norm": 1.9259723124673813,
+      "learning_rate": 8.860948442083648e-07,
+      "loss": 0.041,
+      "step": 7983
+    },
+    {
+      "epoch": 1.8145454545454545,
+      "grad_norm": 1.8890812028179549,
+      "learning_rate": 8.860137526480595e-07,
+      "loss": 0.0329,
+      "step": 7984
+    },
+    {
+      "epoch": 1.8147727272727274,
+      "grad_norm": 2.2074846597781517,
+      "learning_rate": 8.859326557652282e-07,
+      "loss": 0.0439,
+      "step": 7985
+    },
+    {
+      "epoch": 1.815,
+      "grad_norm": 1.5491165938097662,
+      "learning_rate": 8.858515535615249e-07,
+      "loss": 0.0544,
+      "step": 7986
+    },
+    {
+      "epoch": 1.8152272727272727,
+      "grad_norm": 1.0149681811471278,
+      "learning_rate": 8.857704460386035e-07,
+      "loss": 0.0273,
+      "step": 7987
+    },
+    {
+      "epoch": 1.8154545454545454,
+      "grad_norm": 2.446819234241308,
+      "learning_rate": 8.856893331981175e-07,
+      "loss": 0.0501,
+      "step": 7988
+    },
+    {
+      "epoch": 1.8156818181818182,
+      "grad_norm": 1.3568441637902509,
+      "learning_rate": 8.856082150417212e-07,
+      "loss": 0.0483,
+      "step": 7989
+    },
+    {
+      "epoch": 1.815909090909091,
+      "grad_norm": 2.18829931450672,
+      "learning_rate": 8.855270915710689e-07,
+      "loss": 0.0499,
+      "step": 7990
+    },
+    {
+      "epoch": 1.8161363636363637,
+      "grad_norm": 3.3254800701452516,
+      "learning_rate": 8.854459627878146e-07,
+      "loss": 0.1027,
+      "step": 7991
+    },
+    {
+      "epoch": 1.8163636363636364,
+      "grad_norm": 2.621379535258039,
+      "learning_rate": 8.853648286936129e-07,
+      "loss": 0.1119,
+      "step": 7992
+    },
+    {
+      "epoch": 1.8165909090909091,
+      "grad_norm": 1.3365940376417142,
+      "learning_rate": 8.85283689290118e-07,
+      "loss": 0.0236,
+      "step": 7993
+    },
+    {
+      "epoch": 1.8168181818181819,
+      "grad_norm": 1.8025090258884442,
+      "learning_rate": 8.852025445789845e-07,
+      "loss": 0.0254,
+      "step": 7994
+    },
+    {
+      "epoch": 1.8170454545454544,
+      "grad_norm": 2.0337313761531024,
+      "learning_rate": 8.851213945618672e-07,
+      "loss": 0.0406,
+      "step": 7995
+    },
+    {
+      "epoch": 1.8172727272727274,
+      "grad_norm": 1.6644411268043338,
+      "learning_rate": 8.85040239240421e-07,
+      "loss": 0.0465,
+      "step": 7996
+    },
+    {
+      "epoch": 1.8175,
+      "grad_norm": 1.2011490568611811,
+      "learning_rate": 8.849590786163006e-07,
+      "loss": 0.0312,
+      "step": 7997
+    },
+    {
+      "epoch": 1.8177272727272729,
+      "grad_norm": 0.6979052749974061,
+      "learning_rate": 8.848779126911609e-07,
+      "loss": 0.0265,
+      "step": 7998
+    },
+    {
+      "epoch": 1.8179545454545454,
+      "grad_norm": 1.7191979406116815,
+      "learning_rate": 8.847967414666574e-07,
+      "loss": 0.0308,
+      "step": 7999
+    },
+    {
+      "epoch": 1.8181818181818183,
+      "grad_norm": 3.269414263348343,
+      "learning_rate": 8.84715564944445e-07,
+      "loss": 0.0999,
+      "step": 8000
+    },
+    {
+      "epoch": 1.8184090909090909,
+      "grad_norm": 1.5129348498537667,
+      "learning_rate": 8.846343831261792e-07,
+      "loss": 0.0358,
+      "step": 8001
+    },
+    {
+      "epoch": 1.8186363636363636,
+      "grad_norm": 2.7781863980218695,
+      "learning_rate": 8.845531960135154e-07,
+      "loss": 0.047,
+      "step": 8002
+    },
+    {
+      "epoch": 1.8188636363636363,
+      "grad_norm": 1.7809153541145593,
+      "learning_rate": 8.844720036081089e-07,
+      "loss": 0.0338,
+      "step": 8003
+    },
+    {
+      "epoch": 1.819090909090909,
+      "grad_norm": 1.7140324614687368,
+      "learning_rate": 8.843908059116156e-07,
+      "loss": 0.04,
+      "step": 8004
+    },
+    {
+      "epoch": 1.8193181818181818,
+      "grad_norm": 1.9721273812997364,
+      "learning_rate": 8.843096029256913e-07,
+      "loss": 0.0581,
+      "step": 8005
+    },
+    {
+      "epoch": 1.8195454545454546,
+      "grad_norm": 2.8153910559183544,
+      "learning_rate": 8.842283946519918e-07,
+      "loss": 0.0729,
+      "step": 8006
+    },
+    {
+      "epoch": 1.8197727272727273,
+      "grad_norm": 2.2091237032757434,
+      "learning_rate": 8.841471810921733e-07,
+      "loss": 0.0722,
+      "step": 8007
+    },
+    {
+      "epoch": 1.8199999999999998,
+      "grad_norm": 1.5825113652915053,
+      "learning_rate": 8.840659622478914e-07,
+      "loss": 0.0203,
+      "step": 8008
+    },
+    {
+      "epoch": 1.8202272727272728,
+      "grad_norm": 1.9965791096577876,
+      "learning_rate": 8.839847381208027e-07,
+      "loss": 0.0343,
+      "step": 8009
+    },
+    {
+      "epoch": 1.8204545454545453,
+      "grad_norm": 1.72174517066227,
+      "learning_rate": 8.839035087125633e-07,
+      "loss": 0.0365,
+      "step": 8010
+    },
+    {
+      "epoch": 1.8206818181818183,
+      "grad_norm": 2.059053021480631,
+      "learning_rate": 8.838222740248297e-07,
+      "loss": 0.066,
+      "step": 8011
+    },
+    {
+      "epoch": 1.8209090909090908,
+      "grad_norm": 1.8604010588746789,
+      "learning_rate": 8.837410340592584e-07,
+      "loss": 0.0995,
+      "step": 8012
+    },
+    {
+      "epoch": 1.8211363636363638,
+      "grad_norm": 1.0085893859068866,
+      "learning_rate": 8.83659788817506e-07,
+      "loss": 0.0176,
+      "step": 8013
+    },
+    {
+      "epoch": 1.8213636363636363,
+      "grad_norm": 1.4156991921172546,
+      "learning_rate": 8.835785383012292e-07,
+      "loss": 0.0375,
+      "step": 8014
+    },
+    {
+      "epoch": 1.821590909090909,
+      "grad_norm": 2.530610287984006,
+      "learning_rate": 8.834972825120851e-07,
+      "loss": 0.0556,
+      "step": 8015
+    },
+    {
+      "epoch": 1.8218181818181818,
+      "grad_norm": 2.2911865526433393,
+      "learning_rate": 8.834160214517302e-07,
+      "loss": 0.0532,
+      "step": 8016
+    },
+    {
+      "epoch": 1.8220454545454545,
+      "grad_norm": 2.0659026498557975,
+      "learning_rate": 8.833347551218219e-07,
+      "loss": 0.0671,
+      "step": 8017
+    },
+    {
+      "epoch": 1.8222727272727273,
+      "grad_norm": 1.7885625634399538,
+      "learning_rate": 8.832534835240173e-07,
+      "loss": 0.0488,
+      "step": 8018
+    },
+    {
+      "epoch": 1.8225,
+      "grad_norm": 2.3646220849989796,
+      "learning_rate": 8.831722066599735e-07,
+      "loss": 0.0925,
+      "step": 8019
+    },
+    {
+      "epoch": 1.8227272727272728,
+      "grad_norm": 1.9717661947914191,
+      "learning_rate": 8.83090924531348e-07,
+      "loss": 0.039,
+      "step": 8020
+    },
+    {
+      "epoch": 1.8229545454545455,
+      "grad_norm": 2.108198110645839,
+      "learning_rate": 8.830096371397982e-07,
+      "loss": 0.0488,
+      "step": 8021
+    },
+    {
+      "epoch": 1.8231818181818182,
+      "grad_norm": 1.021543401791637,
+      "learning_rate": 8.829283444869821e-07,
+      "loss": 0.0237,
+      "step": 8022
+    },
+    {
+      "epoch": 1.8234090909090908,
+      "grad_norm": 2.460584506200457,
+      "learning_rate": 8.828470465745568e-07,
+      "loss": 0.0761,
+      "step": 8023
+    },
+    {
+      "epoch": 1.8236363636363637,
+      "grad_norm": 2.044743220892855,
+      "learning_rate": 8.827657434041806e-07,
+      "loss": 0.0536,
+      "step": 8024
+    },
+    {
+      "epoch": 1.8238636363636362,
+      "grad_norm": 2.2915351904751082,
+      "learning_rate": 8.82684434977511e-07,
+      "loss": 0.0953,
+      "step": 8025
+    },
+    {
+      "epoch": 1.8240909090909092,
+      "grad_norm": 2.3980567645591275,
+      "learning_rate": 8.826031212962064e-07,
+      "loss": 0.079,
+      "step": 8026
+    },
+    {
+      "epoch": 1.8243181818181817,
+      "grad_norm": 1.5799546293818567,
+      "learning_rate": 8.825218023619245e-07,
+      "loss": 0.0453,
+      "step": 8027
+    },
+    {
+      "epoch": 1.8245454545454547,
+      "grad_norm": 1.802658025320612,
+      "learning_rate": 8.824404781763238e-07,
+      "loss": 0.0498,
+      "step": 8028
+    },
+    {
+      "epoch": 1.8247727272727272,
+      "grad_norm": 2.775235726626137,
+      "learning_rate": 8.823591487410628e-07,
+      "loss": 0.1195,
+      "step": 8029
+    },
+    {
+      "epoch": 1.825,
+      "grad_norm": 1.2504167664368324,
+      "learning_rate": 8.822778140577996e-07,
+      "loss": 0.0292,
+      "step": 8030
+    },
+    {
+      "epoch": 1.8252272727272727,
+      "grad_norm": 2.6953548321653167,
+      "learning_rate": 8.821964741281932e-07,
+      "loss": 0.053,
+      "step": 8031
+    },
+    {
+      "epoch": 1.8254545454545454,
+      "grad_norm": 2.1406517261561526,
+      "learning_rate": 8.821151289539016e-07,
+      "loss": 0.037,
+      "step": 8032
+    },
+    {
+      "epoch": 1.8256818181818182,
+      "grad_norm": 1.8181837030081278,
+      "learning_rate": 8.820337785365843e-07,
+      "loss": 0.0364,
+      "step": 8033
+    },
+    {
+      "epoch": 1.825909090909091,
+      "grad_norm": 2.433363862903433,
+      "learning_rate": 8.819524228778998e-07,
+      "loss": 0.0635,
+      "step": 8034
+    },
+    {
+      "epoch": 1.8261363636363637,
+      "grad_norm": 2.089612479693195,
+      "learning_rate": 8.818710619795068e-07,
+      "loss": 0.0503,
+      "step": 8035
+    },
+    {
+      "epoch": 1.8263636363636364,
+      "grad_norm": 2.4489962966656162,
+      "learning_rate": 8.81789695843065e-07,
+      "loss": 0.0814,
+      "step": 8036
+    },
+    {
+      "epoch": 1.8265909090909092,
+      "grad_norm": 2.2192197414852783,
+      "learning_rate": 8.817083244702332e-07,
+      "loss": 0.0435,
+      "step": 8037
+    },
+    {
+      "epoch": 1.8268181818181817,
+      "grad_norm": 1.9647680145779065,
+      "learning_rate": 8.816269478626708e-07,
+      "loss": 0.0474,
+      "step": 8038
+    },
+    {
+      "epoch": 1.8270454545454546,
+      "grad_norm": 1.6576213659897545,
+      "learning_rate": 8.815455660220373e-07,
+      "loss": 0.0433,
+      "step": 8039
+    },
+    {
+      "epoch": 1.8272727272727272,
+      "grad_norm": 1.3933298626596564,
+      "learning_rate": 8.814641789499923e-07,
+      "loss": 0.0343,
+      "step": 8040
+    },
+    {
+      "epoch": 1.8275000000000001,
+      "grad_norm": 1.9993440400334543,
+      "learning_rate": 8.81382786648195e-07,
+      "loss": 0.0348,
+      "step": 8041
+    },
+    {
+      "epoch": 1.8277272727272726,
+      "grad_norm": 2.5239832797193786,
+      "learning_rate": 8.813013891183053e-07,
+      "loss": 0.0545,
+      "step": 8042
+    },
+    {
+      "epoch": 1.8279545454545456,
+      "grad_norm": 1.9888781177731916,
+      "learning_rate": 8.812199863619833e-07,
+      "loss": 0.0317,
+      "step": 8043
+    },
+    {
+      "epoch": 1.8281818181818181,
+      "grad_norm": 1.715822640103994,
+      "learning_rate": 8.811385783808889e-07,
+      "loss": 0.0365,
+      "step": 8044
+    },
+    {
+      "epoch": 1.8284090909090909,
+      "grad_norm": 2.6647205620481915,
+      "learning_rate": 8.810571651766819e-07,
+      "loss": 0.1026,
+      "step": 8045
+    },
+    {
+      "epoch": 1.8286363636363636,
+      "grad_norm": 1.7649590041583536,
+      "learning_rate": 8.809757467510227e-07,
+      "loss": 0.0566,
+      "step": 8046
+    },
+    {
+      "epoch": 1.8288636363636364,
+      "grad_norm": 2.7694993553894602,
+      "learning_rate": 8.808943231055713e-07,
+      "loss": 0.0787,
+      "step": 8047
+    },
+    {
+      "epoch": 1.829090909090909,
+      "grad_norm": 2.3610229539239085,
+      "learning_rate": 8.808128942419883e-07,
+      "loss": 0.0572,
+      "step": 8048
+    },
+    {
+      "epoch": 1.8293181818181818,
+      "grad_norm": 1.7714589858565182,
+      "learning_rate": 8.80731460161934e-07,
+      "loss": 0.0377,
+      "step": 8049
+    },
+    {
+      "epoch": 1.8295454545454546,
+      "grad_norm": 1.706694190342187,
+      "learning_rate": 8.806500208670691e-07,
+      "loss": 0.0333,
+      "step": 8050
+    },
+    {
+      "epoch": 1.829772727272727,
+      "grad_norm": 2.6403144593905314,
+      "learning_rate": 8.805685763590543e-07,
+      "loss": 0.0434,
+      "step": 8051
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.5184745292966897,
+      "learning_rate": 8.804871266395504e-07,
+      "loss": 0.0331,
+      "step": 8052
+    },
+    {
+      "epoch": 1.8302272727272726,
+      "grad_norm": 1.329888507644357,
+      "learning_rate": 8.804056717102183e-07,
+      "loss": 0.0311,
+      "step": 8053
+    },
+    {
+      "epoch": 1.8304545454545456,
+      "grad_norm": 2.5675813174946462,
+      "learning_rate": 8.803242115727188e-07,
+      "loss": 0.0348,
+      "step": 8054
+    },
+    {
+      "epoch": 1.830681818181818,
+      "grad_norm": 1.8199326624188874,
+      "learning_rate": 8.802427462287134e-07,
+      "loss": 0.0694,
+      "step": 8055
+    },
+    {
+      "epoch": 1.830909090909091,
+      "grad_norm": 1.9853218091383065,
+      "learning_rate": 8.80161275679863e-07,
+      "loss": 0.0514,
+      "step": 8056
+    },
+    {
+      "epoch": 1.8311363636363636,
+      "grad_norm": 1.2735187608620961,
+      "learning_rate": 8.80079799927829e-07,
+      "loss": 0.0404,
+      "step": 8057
+    },
+    {
+      "epoch": 1.8313636363636365,
+      "grad_norm": 1.9858219292362114,
+      "learning_rate": 8.799983189742729e-07,
+      "loss": 0.0694,
+      "step": 8058
+    },
+    {
+      "epoch": 1.831590909090909,
+      "grad_norm": 1.5573039014347494,
+      "learning_rate": 8.799168328208562e-07,
+      "loss": 0.0339,
+      "step": 8059
+    },
+    {
+      "epoch": 1.8318181818181818,
+      "grad_norm": 1.330648622612751,
+      "learning_rate": 8.798353414692407e-07,
+      "loss": 0.0184,
+      "step": 8060
+    },
+    {
+      "epoch": 1.8320454545454545,
+      "grad_norm": 1.8823544705830342,
+      "learning_rate": 8.797538449210878e-07,
+      "loss": 0.0459,
+      "step": 8061
+    },
+    {
+      "epoch": 1.8322727272727273,
+      "grad_norm": 0.9476809783657248,
+      "learning_rate": 8.796723431780596e-07,
+      "loss": 0.0133,
+      "step": 8062
+    },
+    {
+      "epoch": 1.8325,
+      "grad_norm": 1.8449534037996522,
+      "learning_rate": 8.795908362418182e-07,
+      "loss": 0.0462,
+      "step": 8063
+    },
+    {
+      "epoch": 1.8327272727272728,
+      "grad_norm": 2.24563107447383,
+      "learning_rate": 8.795093241140254e-07,
+      "loss": 0.0594,
+      "step": 8064
+    },
+    {
+      "epoch": 1.8329545454545455,
+      "grad_norm": 2.012434382287865,
+      "learning_rate": 8.794278067963433e-07,
+      "loss": 0.0408,
+      "step": 8065
+    },
+    {
+      "epoch": 1.833181818181818,
+      "grad_norm": 2.1875756123631414,
+      "learning_rate": 8.793462842904344e-07,
+      "loss": 0.0572,
+      "step": 8066
+    },
+    {
+      "epoch": 1.833409090909091,
+      "grad_norm": 1.4152841549265025,
+      "learning_rate": 8.792647565979612e-07,
+      "loss": 0.0213,
+      "step": 8067
+    },
+    {
+      "epoch": 1.8336363636363635,
+      "grad_norm": 1.1014375115307176,
+      "learning_rate": 8.79183223720586e-07,
+      "loss": 0.0379,
+      "step": 8068
+    },
+    {
+      "epoch": 1.8338636363636365,
+      "grad_norm": 2.704420960863485,
+      "learning_rate": 8.791016856599713e-07,
+      "loss": 0.0356,
+      "step": 8069
+    },
+    {
+      "epoch": 1.834090909090909,
+      "grad_norm": 1.7386995352495578,
+      "learning_rate": 8.790201424177799e-07,
+      "loss": 0.0221,
+      "step": 8070
+    },
+    {
+      "epoch": 1.834318181818182,
+      "grad_norm": 1.7136448224797023,
+      "learning_rate": 8.789385939956747e-07,
+      "loss": 0.0409,
+      "step": 8071
+    },
+    {
+      "epoch": 1.8345454545454545,
+      "grad_norm": 1.4601782059642152,
+      "learning_rate": 8.788570403953185e-07,
+      "loss": 0.0588,
+      "step": 8072
+    },
+    {
+      "epoch": 1.8347727272727272,
+      "grad_norm": 1.4217940860496496,
+      "learning_rate": 8.787754816183743e-07,
+      "loss": 0.0311,
+      "step": 8073
+    },
+    {
+      "epoch": 1.835,
+      "grad_norm": 1.0138787576563042,
+      "learning_rate": 8.786939176665055e-07,
+      "loss": 0.0263,
+      "step": 8074
+    },
+    {
+      "epoch": 1.8352272727272727,
+      "grad_norm": 1.8239781250991853,
+      "learning_rate": 8.78612348541375e-07,
+      "loss": 0.0466,
+      "step": 8075
+    },
+    {
+      "epoch": 1.8354545454545454,
+      "grad_norm": 2.0583614882878063,
+      "learning_rate": 8.785307742446464e-07,
+      "loss": 0.0413,
+      "step": 8076
+    },
+    {
+      "epoch": 1.8356818181818182,
+      "grad_norm": 3.123472098220273,
+      "learning_rate": 8.784491947779829e-07,
+      "loss": 0.0643,
+      "step": 8077
+    },
+    {
+      "epoch": 1.835909090909091,
+      "grad_norm": 2.402526746277211,
+      "learning_rate": 8.783676101430482e-07,
+      "loss": 0.0616,
+      "step": 8078
+    },
+    {
+      "epoch": 1.8361363636363637,
+      "grad_norm": 2.671472306227745,
+      "learning_rate": 8.782860203415058e-07,
+      "loss": 0.0814,
+      "step": 8079
+    },
+    {
+      "epoch": 1.8363636363636364,
+      "grad_norm": 1.722970992030125,
+      "learning_rate": 8.782044253750197e-07,
+      "loss": 0.0478,
+      "step": 8080
+    },
+    {
+      "epoch": 1.836590909090909,
+      "grad_norm": 2.989263202796664,
+      "learning_rate": 8.781228252452534e-07,
+      "loss": 0.0619,
+      "step": 8081
+    },
+    {
+      "epoch": 1.836818181818182,
+      "grad_norm": 1.770124018095723,
+      "learning_rate": 8.780412199538712e-07,
+      "loss": 0.0358,
+      "step": 8082
+    },
+    {
+      "epoch": 1.8370454545454544,
+      "grad_norm": 1.5083546759671853,
+      "learning_rate": 8.779596095025373e-07,
+      "loss": 0.0396,
+      "step": 8083
+    },
+    {
+      "epoch": 1.8372727272727274,
+      "grad_norm": 2.2409835481045866,
+      "learning_rate": 8.778779938929153e-07,
+      "loss": 0.0543,
+      "step": 8084
+    },
+    {
+      "epoch": 1.8375,
+      "grad_norm": 2.1907161675588487,
+      "learning_rate": 8.7779637312667e-07,
+      "loss": 0.0525,
+      "step": 8085
+    },
+    {
+      "epoch": 1.8377272727272729,
+      "grad_norm": 0.9731146925729535,
+      "learning_rate": 8.777147472054657e-07,
+      "loss": 0.0261,
+      "step": 8086
+    },
+    {
+      "epoch": 1.8379545454545454,
+      "grad_norm": 1.7511860477006975,
+      "learning_rate": 8.776331161309668e-07,
+      "loss": 0.0356,
+      "step": 8087
+    },
+    {
+      "epoch": 1.8381818181818181,
+      "grad_norm": 2.7966477613038534,
+      "learning_rate": 8.775514799048378e-07,
+      "loss": 0.0722,
+      "step": 8088
+    },
+    {
+      "epoch": 1.8384090909090909,
+      "grad_norm": 2.402948678602388,
+      "learning_rate": 8.774698385287436e-07,
+      "loss": 0.0704,
+      "step": 8089
+    },
+    {
+      "epoch": 1.8386363636363636,
+      "grad_norm": 2.1073768594196562,
+      "learning_rate": 8.77388192004349e-07,
+      "loss": 0.0317,
+      "step": 8090
+    },
+    {
+      "epoch": 1.8388636363636364,
+      "grad_norm": 2.7267735032720117,
+      "learning_rate": 8.773065403333186e-07,
+      "loss": 0.0557,
+      "step": 8091
+    },
+    {
+      "epoch": 1.839090909090909,
+      "grad_norm": 3.107159594826314,
+      "learning_rate": 8.772248835173181e-07,
+      "loss": 0.0583,
+      "step": 8092
+    },
+    {
+      "epoch": 1.8393181818181819,
+      "grad_norm": 2.8399389692503063,
+      "learning_rate": 8.771432215580118e-07,
+      "loss": 0.0801,
+      "step": 8093
+    },
+    {
+      "epoch": 1.8395454545454546,
+      "grad_norm": 1.7297973092928314,
+      "learning_rate": 8.770615544570656e-07,
+      "loss": 0.047,
+      "step": 8094
+    },
+    {
+      "epoch": 1.8397727272727273,
+      "grad_norm": 1.5288159511807191,
+      "learning_rate": 8.769798822161443e-07,
+      "loss": 0.0334,
+      "step": 8095
+    },
+    {
+      "epoch": 1.8399999999999999,
+      "grad_norm": 1.8171827475658662,
+      "learning_rate": 8.768982048369136e-07,
+      "loss": 0.0359,
+      "step": 8096
+    },
+    {
+      "epoch": 1.8402272727272728,
+      "grad_norm": 1.3604068770726048,
+      "learning_rate": 8.768165223210393e-07,
+      "loss": 0.0383,
+      "step": 8097
+    },
+    {
+      "epoch": 1.8404545454545453,
+      "grad_norm": 1.6501596981638254,
+      "learning_rate": 8.767348346701865e-07,
+      "loss": 0.0513,
+      "step": 8098
+    },
+    {
+      "epoch": 1.8406818181818183,
+      "grad_norm": 2.0834268870383092,
+      "learning_rate": 8.766531418860215e-07,
+      "loss": 0.0326,
+      "step": 8099
+    },
+    {
+      "epoch": 1.8409090909090908,
+      "grad_norm": 1.5988917623440368,
+      "learning_rate": 8.765714439702098e-07,
+      "loss": 0.0349,
+      "step": 8100
+    },
+    {
+      "epoch": 1.8411363636363638,
+      "grad_norm": 2.411114223533715,
+      "learning_rate": 8.764897409244176e-07,
+      "loss": 0.0477,
+      "step": 8101
+    },
+    {
+      "epoch": 1.8413636363636363,
+      "grad_norm": 3.186097105339078,
+      "learning_rate": 8.764080327503107e-07,
+      "loss": 0.0891,
+      "step": 8102
+    },
+    {
+      "epoch": 1.841590909090909,
+      "grad_norm": 1.875435775534947,
+      "learning_rate": 8.763263194495553e-07,
+      "loss": 0.0479,
+      "step": 8103
+    },
+    {
+      "epoch": 1.8418181818181818,
+      "grad_norm": 2.2869852317308195,
+      "learning_rate": 8.762446010238179e-07,
+      "loss": 0.0543,
+      "step": 8104
+    },
+    {
+      "epoch": 1.8420454545454545,
+      "grad_norm": 0.7873683760402905,
+      "learning_rate": 8.761628774747648e-07,
+      "loss": 0.035,
+      "step": 8105
+    },
+    {
+      "epoch": 1.8422727272727273,
+      "grad_norm": 1.807199242586263,
+      "learning_rate": 8.760811488040624e-07,
+      "loss": 0.0262,
+      "step": 8106
+    },
+    {
+      "epoch": 1.8425,
+      "grad_norm": 1.211349519738688,
+      "learning_rate": 8.759994150133772e-07,
+      "loss": 0.0597,
+      "step": 8107
+    },
+    {
+      "epoch": 1.8427272727272728,
+      "grad_norm": 1.9921782715487306,
+      "learning_rate": 8.759176761043764e-07,
+      "loss": 0.0343,
+      "step": 8108
+    },
+    {
+      "epoch": 1.8429545454545453,
+      "grad_norm": 1.8141327247779144,
+      "learning_rate": 8.758359320787262e-07,
+      "loss": 0.0693,
+      "step": 8109
+    },
+    {
+      "epoch": 1.8431818181818183,
+      "grad_norm": 1.8639382595064096,
+      "learning_rate": 8.757541829380937e-07,
+      "loss": 0.0681,
+      "step": 8110
+    },
+    {
+      "epoch": 1.8434090909090908,
+      "grad_norm": 1.7100458167715096,
+      "learning_rate": 8.756724286841458e-07,
+      "loss": 0.0522,
+      "step": 8111
+    },
+    {
+      "epoch": 1.8436363636363637,
+      "grad_norm": 0.8784285730292124,
+      "learning_rate": 8.755906693185501e-07,
+      "loss": 0.024,
+      "step": 8112
+    },
+    {
+      "epoch": 1.8438636363636363,
+      "grad_norm": 1.6769938817899217,
+      "learning_rate": 8.755089048429734e-07,
+      "loss": 0.0304,
+      "step": 8113
+    },
+    {
+      "epoch": 1.8440909090909092,
+      "grad_norm": 2.035708045413266,
+      "learning_rate": 8.75427135259083e-07,
+      "loss": 0.0284,
+      "step": 8114
+    },
+    {
+      "epoch": 1.8443181818181817,
+      "grad_norm": 1.7048293869799978,
+      "learning_rate": 8.753453605685465e-07,
+      "loss": 0.0318,
+      "step": 8115
+    },
+    {
+      "epoch": 1.8445454545454547,
+      "grad_norm": 1.810715329065708,
+      "learning_rate": 8.752635807730314e-07,
+      "loss": 0.0742,
+      "step": 8116
+    },
+    {
+      "epoch": 1.8447727272727272,
+      "grad_norm": 1.59746765542727,
+      "learning_rate": 8.751817958742051e-07,
+      "loss": 0.0415,
+      "step": 8117
+    },
+    {
+      "epoch": 1.845,
+      "grad_norm": 1.9484650641223171,
+      "learning_rate": 8.751000058737355e-07,
+      "loss": 0.0459,
+      "step": 8118
+    },
+    {
+      "epoch": 1.8452272727272727,
+      "grad_norm": 2.7949696433282174,
+      "learning_rate": 8.750182107732904e-07,
+      "loss": 0.0892,
+      "step": 8119
+    },
+    {
+      "epoch": 1.8454545454545455,
+      "grad_norm": 1.6092855043684418,
+      "learning_rate": 8.749364105745381e-07,
+      "loss": 0.0528,
+      "step": 8120
+    },
+    {
+      "epoch": 1.8456818181818182,
+      "grad_norm": 1.4398518713682242,
+      "learning_rate": 8.748546052791462e-07,
+      "loss": 0.0325,
+      "step": 8121
+    },
+    {
+      "epoch": 1.845909090909091,
+      "grad_norm": 1.1249317701647037,
+      "learning_rate": 8.74772794888783e-07,
+      "loss": 0.0328,
+      "step": 8122
+    },
+    {
+      "epoch": 1.8461363636363637,
+      "grad_norm": 2.3317972000080545,
+      "learning_rate": 8.746909794051169e-07,
+      "loss": 0.077,
+      "step": 8123
+    },
+    {
+      "epoch": 1.8463636363636362,
+      "grad_norm": 3.0001085123967077,
+      "learning_rate": 8.746091588298159e-07,
+      "loss": 0.0812,
+      "step": 8124
+    },
+    {
+      "epoch": 1.8465909090909092,
+      "grad_norm": 1.9038231169503914,
+      "learning_rate": 8.745273331645487e-07,
+      "loss": 0.0473,
+      "step": 8125
+    },
+    {
+      "epoch": 1.8468181818181817,
+      "grad_norm": 2.2118049236239696,
+      "learning_rate": 8.744455024109838e-07,
+      "loss": 0.0365,
+      "step": 8126
+    },
+    {
+      "epoch": 1.8470454545454547,
+      "grad_norm": 3.368938387305072,
+      "learning_rate": 8.743636665707903e-07,
+      "loss": 0.0467,
+      "step": 8127
+    },
+    {
+      "epoch": 1.8472727272727272,
+      "grad_norm": 1.5658303981374628,
+      "learning_rate": 8.742818256456362e-07,
+      "loss": 0.046,
+      "step": 8128
+    },
+    {
+      "epoch": 1.8475000000000001,
+      "grad_norm": 1.543874949679967,
+      "learning_rate": 8.741999796371909e-07,
+      "loss": 0.0236,
+      "step": 8129
+    },
+    {
+      "epoch": 1.8477272727272727,
+      "grad_norm": 2.403151667263669,
+      "learning_rate": 8.741181285471234e-07,
+      "loss": 0.05,
+      "step": 8130
+    },
+    {
+      "epoch": 1.8479545454545454,
+      "grad_norm": 1.618152744157936,
+      "learning_rate": 8.740362723771024e-07,
+      "loss": 0.0478,
+      "step": 8131
+    },
+    {
+      "epoch": 1.8481818181818181,
+      "grad_norm": 2.1946984150497046,
+      "learning_rate": 8.739544111287976e-07,
+      "loss": 0.0361,
+      "step": 8132
+    },
+    {
+      "epoch": 1.848409090909091,
+      "grad_norm": 2.2055109190067346,
+      "learning_rate": 8.738725448038779e-07,
+      "loss": 0.0366,
+      "step": 8133
+    },
+    {
+      "epoch": 1.8486363636363636,
+      "grad_norm": 1.678453674035477,
+      "learning_rate": 8.737906734040127e-07,
+      "loss": 0.0563,
+      "step": 8134
+    },
+    {
+      "epoch": 1.8488636363636364,
+      "grad_norm": 1.9571573291674438,
+      "learning_rate": 8.737087969308718e-07,
+      "loss": 0.0329,
+      "step": 8135
+    },
+    {
+      "epoch": 1.8490909090909091,
+      "grad_norm": 1.815424045059717,
+      "learning_rate": 8.736269153861246e-07,
+      "loss": 0.0279,
+      "step": 8136
+    },
+    {
+      "epoch": 1.8493181818181819,
+      "grad_norm": 2.05063647123803,
+      "learning_rate": 8.735450287714409e-07,
+      "loss": 0.038,
+      "step": 8137
+    },
+    {
+      "epoch": 1.8495454545454546,
+      "grad_norm": 2.2482943564875573,
+      "learning_rate": 8.734631370884904e-07,
+      "loss": 0.0702,
+      "step": 8138
+    },
+    {
+      "epoch": 1.8497727272727271,
+      "grad_norm": 2.46331540497991,
+      "learning_rate": 8.73381240338943e-07,
+      "loss": 0.0723,
+      "step": 8139
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 2.7141093449324902,
+      "learning_rate": 8.732993385244688e-07,
+      "loss": 0.0311,
+      "step": 8140
+    },
+    {
+      "epoch": 1.8502272727272726,
+      "grad_norm": 1.6200641469435357,
+      "learning_rate": 8.732174316467379e-07,
+      "loss": 0.0529,
+      "step": 8141
+    },
+    {
+      "epoch": 1.8504545454545456,
+      "grad_norm": 4.135861212210806,
+      "learning_rate": 8.731355197074206e-07,
+      "loss": 0.0968,
+      "step": 8142
+    },
+    {
+      "epoch": 1.850681818181818,
+      "grad_norm": 2.090742831929552,
+      "learning_rate": 8.730536027081872e-07,
+      "loss": 0.0589,
+      "step": 8143
+    },
+    {
+      "epoch": 1.850909090909091,
+      "grad_norm": 1.9744546554579592,
+      "learning_rate": 8.72971680650708e-07,
+      "loss": 0.0562,
+      "step": 8144
+    },
+    {
+      "epoch": 1.8511363636363636,
+      "grad_norm": 2.088352556188155,
+      "learning_rate": 8.728897535366537e-07,
+      "loss": 0.0557,
+      "step": 8145
+    },
+    {
+      "epoch": 1.8513636363636363,
+      "grad_norm": 2.035296286709105,
+      "learning_rate": 8.728078213676948e-07,
+      "loss": 0.0348,
+      "step": 8146
+    },
+    {
+      "epoch": 1.851590909090909,
+      "grad_norm": 1.5130687258456348,
+      "learning_rate": 8.727258841455022e-07,
+      "loss": 0.0433,
+      "step": 8147
+    },
+    {
+      "epoch": 1.8518181818181818,
+      "grad_norm": 1.3528946462800318,
+      "learning_rate": 8.726439418717464e-07,
+      "loss": 0.0558,
+      "step": 8148
+    },
+    {
+      "epoch": 1.8520454545454546,
+      "grad_norm": 2.142407954059985,
+      "learning_rate": 8.725619945480987e-07,
+      "loss": 0.0585,
+      "step": 8149
+    },
+    {
+      "epoch": 1.8522727272727273,
+      "grad_norm": 1.5488184529336833,
+      "learning_rate": 8.7248004217623e-07,
+      "loss": 0.0465,
+      "step": 8150
+    },
+    {
+      "epoch": 1.8525,
+      "grad_norm": 1.8615441878173435,
+      "learning_rate": 8.723980847578116e-07,
+      "loss": 0.0409,
+      "step": 8151
+    },
+    {
+      "epoch": 1.8527272727272728,
+      "grad_norm": 1.4533269274177942,
+      "learning_rate": 8.723161222945147e-07,
+      "loss": 0.0293,
+      "step": 8152
+    },
+    {
+      "epoch": 1.8529545454545455,
+      "grad_norm": 2.3427900646338444,
+      "learning_rate": 8.722341547880103e-07,
+      "loss": 0.0685,
+      "step": 8153
+    },
+    {
+      "epoch": 1.853181818181818,
+      "grad_norm": 2.03876830139978,
+      "learning_rate": 8.721521822399704e-07,
+      "loss": 0.065,
+      "step": 8154
+    },
+    {
+      "epoch": 1.853409090909091,
+      "grad_norm": 1.5851731845293189,
+      "learning_rate": 8.720702046520663e-07,
+      "loss": 0.023,
+      "step": 8155
+    },
+    {
+      "epoch": 1.8536363636363635,
+      "grad_norm": 1.717359944085746,
+      "learning_rate": 8.719882220259697e-07,
+      "loss": 0.0425,
+      "step": 8156
+    },
+    {
+      "epoch": 1.8538636363636365,
+      "grad_norm": 1.8007240421214594,
+      "learning_rate": 8.719062343633522e-07,
+      "loss": 0.0664,
+      "step": 8157
+    },
+    {
+      "epoch": 1.854090909090909,
+      "grad_norm": 3.3076020686660685,
+      "learning_rate": 8.718242416658861e-07,
+      "loss": 0.0654,
+      "step": 8158
+    },
+    {
+      "epoch": 1.854318181818182,
+      "grad_norm": 2.251428276923691,
+      "learning_rate": 8.717422439352429e-07,
+      "loss": 0.0492,
+      "step": 8159
+    },
+    {
+      "epoch": 1.8545454545454545,
+      "grad_norm": 1.662989035501785,
+      "learning_rate": 8.716602411730951e-07,
+      "loss": 0.0563,
+      "step": 8160
+    },
+    {
+      "epoch": 1.8547727272727272,
+      "grad_norm": 1.9599923124535856,
+      "learning_rate": 8.715782333811143e-07,
+      "loss": 0.0315,
+      "step": 8161
+    },
+    {
+      "epoch": 1.855,
+      "grad_norm": 1.380594298546293,
+      "learning_rate": 8.714962205609734e-07,
+      "loss": 0.0319,
+      "step": 8162
+    },
+    {
+      "epoch": 1.8552272727272727,
+      "grad_norm": 2.425468082659738,
+      "learning_rate": 8.714142027143445e-07,
+      "loss": 0.0492,
+      "step": 8163
+    },
+    {
+      "epoch": 1.8554545454545455,
+      "grad_norm": 2.7514269499802415,
+      "learning_rate": 8.713321798429e-07,
+      "loss": 0.107,
+      "step": 8164
+    },
+    {
+      "epoch": 1.8556818181818182,
+      "grad_norm": 1.623672539361061,
+      "learning_rate": 8.712501519483127e-07,
+      "loss": 0.0411,
+      "step": 8165
+    },
+    {
+      "epoch": 1.855909090909091,
+      "grad_norm": 1.6121153637190382,
+      "learning_rate": 8.711681190322552e-07,
+      "loss": 0.0611,
+      "step": 8166
+    },
+    {
+      "epoch": 1.8561363636363635,
+      "grad_norm": 1.736175124714783,
+      "learning_rate": 8.710860810964003e-07,
+      "loss": 0.0356,
+      "step": 8167
+    },
+    {
+      "epoch": 1.8563636363636364,
+      "grad_norm": 2.0944011928984567,
+      "learning_rate": 8.710040381424208e-07,
+      "loss": 0.0433,
+      "step": 8168
+    },
+    {
+      "epoch": 1.856590909090909,
+      "grad_norm": 2.0836649489440524,
+      "learning_rate": 8.7092199017199e-07,
+      "loss": 0.0363,
+      "step": 8169
+    },
+    {
+      "epoch": 1.856818181818182,
+      "grad_norm": 1.8846587770645318,
+      "learning_rate": 8.708399371867805e-07,
+      "loss": 0.0737,
+      "step": 8170
+    },
+    {
+      "epoch": 1.8570454545454544,
+      "grad_norm": 1.7617879020238814,
+      "learning_rate": 8.707578791884659e-07,
+      "loss": 0.059,
+      "step": 8171
+    },
+    {
+      "epoch": 1.8572727272727274,
+      "grad_norm": 1.622120255091733,
+      "learning_rate": 8.706758161787193e-07,
+      "loss": 0.0398,
+      "step": 8172
+    },
+    {
+      "epoch": 1.8575,
+      "grad_norm": 2.451463048861813,
+      "learning_rate": 8.705937481592143e-07,
+      "loss": 0.0684,
+      "step": 8173
+    },
+    {
+      "epoch": 1.857727272727273,
+      "grad_norm": 1.7387651879635684,
+      "learning_rate": 8.705116751316244e-07,
+      "loss": 0.0484,
+      "step": 8174
+    },
+    {
+      "epoch": 1.8579545454545454,
+      "grad_norm": 2.691665317718088,
+      "learning_rate": 8.70429597097623e-07,
+      "loss": 0.0793,
+      "step": 8175
+    },
+    {
+      "epoch": 1.8581818181818182,
+      "grad_norm": 1.1308691789881817,
+      "learning_rate": 8.703475140588841e-07,
+      "loss": 0.0322,
+      "step": 8176
+    },
+    {
+      "epoch": 1.858409090909091,
+      "grad_norm": 2.4233966414303967,
+      "learning_rate": 8.702654260170812e-07,
+      "loss": 0.0497,
+      "step": 8177
+    },
+    {
+      "epoch": 1.8586363636363636,
+      "grad_norm": 1.6072086082413624,
+      "learning_rate": 8.701833329738885e-07,
+      "loss": 0.0612,
+      "step": 8178
+    },
+    {
+      "epoch": 1.8588636363636364,
+      "grad_norm": 2.363205950375358,
+      "learning_rate": 8.701012349309798e-07,
+      "loss": 0.0743,
+      "step": 8179
+    },
+    {
+      "epoch": 1.8590909090909091,
+      "grad_norm": 2.4583751968473835,
+      "learning_rate": 8.700191318900293e-07,
+      "loss": 0.0426,
+      "step": 8180
+    },
+    {
+      "epoch": 1.8593181818181819,
+      "grad_norm": 1.4809260955760606,
+      "learning_rate": 8.699370238527112e-07,
+      "loss": 0.0474,
+      "step": 8181
+    },
+    {
+      "epoch": 1.8595454545454544,
+      "grad_norm": 1.855509821684292,
+      "learning_rate": 8.698549108207001e-07,
+      "loss": 0.0554,
+      "step": 8182
+    },
+    {
+      "epoch": 1.8597727272727274,
+      "grad_norm": 0.6847443711483117,
+      "learning_rate": 8.697727927956699e-07,
+      "loss": 0.0291,
+      "step": 8183
+    },
+    {
+      "epoch": 1.8599999999999999,
+      "grad_norm": 1.6170474169239775,
+      "learning_rate": 8.696906697792957e-07,
+      "loss": 0.0464,
+      "step": 8184
+    },
+    {
+      "epoch": 1.8602272727272728,
+      "grad_norm": 1.758225802034341,
+      "learning_rate": 8.696085417732517e-07,
+      "loss": 0.0397,
+      "step": 8185
+    },
+    {
+      "epoch": 1.8604545454545454,
+      "grad_norm": 2.994365200818807,
+      "learning_rate": 8.695264087792127e-07,
+      "loss": 0.0942,
+      "step": 8186
+    },
+    {
+      "epoch": 1.8606818181818183,
+      "grad_norm": 1.8594239876304124,
+      "learning_rate": 8.694442707988538e-07,
+      "loss": 0.061,
+      "step": 8187
+    },
+    {
+      "epoch": 1.8609090909090908,
+      "grad_norm": 2.100784484715475,
+      "learning_rate": 8.693621278338496e-07,
+      "loss": 0.0509,
+      "step": 8188
+    },
+    {
+      "epoch": 1.8611363636363636,
+      "grad_norm": 2.473477878156028,
+      "learning_rate": 8.692799798858754e-07,
+      "loss": 0.056,
+      "step": 8189
+    },
+    {
+      "epoch": 1.8613636363636363,
+      "grad_norm": 2.1998378530907097,
+      "learning_rate": 8.691978269566062e-07,
+      "loss": 0.0293,
+      "step": 8190
+    },
+    {
+      "epoch": 1.861590909090909,
+      "grad_norm": 2.0947622512433117,
+      "learning_rate": 8.691156690477174e-07,
+      "loss": 0.0245,
+      "step": 8191
+    },
+    {
+      "epoch": 1.8618181818181818,
+      "grad_norm": 1.745690213909743,
+      "learning_rate": 8.690335061608841e-07,
+      "loss": 0.0548,
+      "step": 8192
+    },
+    {
+      "epoch": 1.8620454545454546,
+      "grad_norm": 1.3314791138845607,
+      "learning_rate": 8.689513382977821e-07,
+      "loss": 0.0487,
+      "step": 8193
+    },
+    {
+      "epoch": 1.8622727272727273,
+      "grad_norm": 3.35653853433555,
+      "learning_rate": 8.688691654600864e-07,
+      "loss": 0.088,
+      "step": 8194
+    },
+    {
+      "epoch": 1.8625,
+      "grad_norm": 1.3073197383811421,
+      "learning_rate": 8.687869876494731e-07,
+      "loss": 0.0259,
+      "step": 8195
+    },
+    {
+      "epoch": 1.8627272727272728,
+      "grad_norm": 1.8379636588676154,
+      "learning_rate": 8.687048048676178e-07,
+      "loss": 0.0428,
+      "step": 8196
+    },
+    {
+      "epoch": 1.8629545454545453,
+      "grad_norm": 1.535691836860078,
+      "learning_rate": 8.686226171161963e-07,
+      "loss": 0.0434,
+      "step": 8197
+    },
+    {
+      "epoch": 1.8631818181818183,
+      "grad_norm": 2.5658189440595227,
+      "learning_rate": 8.685404243968848e-07,
+      "loss": 0.0746,
+      "step": 8198
+    },
+    {
+      "epoch": 1.8634090909090908,
+      "grad_norm": 2.160159534724591,
+      "learning_rate": 8.684582267113589e-07,
+      "loss": 0.0484,
+      "step": 8199
+    },
+    {
+      "epoch": 1.8636363636363638,
+      "grad_norm": 2.28116345211753,
+      "learning_rate": 8.683760240612953e-07,
+      "loss": 0.0565,
+      "step": 8200
+    },
+    {
+      "epoch": 1.8638636363636363,
+      "grad_norm": 2.1442396058548057,
+      "learning_rate": 8.682938164483698e-07,
+      "loss": 0.0548,
+      "step": 8201
+    },
+    {
+      "epoch": 1.8640909090909092,
+      "grad_norm": 1.2887925970385776,
+      "learning_rate": 8.68211603874259e-07,
+      "loss": 0.0393,
+      "step": 8202
+    },
+    {
+      "epoch": 1.8643181818181818,
+      "grad_norm": 1.7526881807819348,
+      "learning_rate": 8.681293863406391e-07,
+      "loss": 0.0447,
+      "step": 8203
+    },
+    {
+      "epoch": 1.8645454545454545,
+      "grad_norm": 1.2886507406520862,
+      "learning_rate": 8.680471638491872e-07,
+      "loss": 0.0253,
+      "step": 8204
+    },
+    {
+      "epoch": 1.8647727272727272,
+      "grad_norm": 1.7632679460151008,
+      "learning_rate": 8.679649364015793e-07,
+      "loss": 0.0247,
+      "step": 8205
+    },
+    {
+      "epoch": 1.865,
+      "grad_norm": 1.4963524490076627,
+      "learning_rate": 8.678827039994928e-07,
+      "loss": 0.0573,
+      "step": 8206
+    },
+    {
+      "epoch": 1.8652272727272727,
+      "grad_norm": 1.8508588936480814,
+      "learning_rate": 8.67800466644604e-07,
+      "loss": 0.0351,
+      "step": 8207
+    },
+    {
+      "epoch": 1.8654545454545455,
+      "grad_norm": 1.3695308514360276,
+      "learning_rate": 8.677182243385901e-07,
+      "loss": 0.0302,
+      "step": 8208
+    },
+    {
+      "epoch": 1.8656818181818182,
+      "grad_norm": 1.2896427852631427,
+      "learning_rate": 8.676359770831282e-07,
+      "loss": 0.0262,
+      "step": 8209
+    },
+    {
+      "epoch": 1.865909090909091,
+      "grad_norm": 1.9592101148371677,
+      "learning_rate": 8.675537248798953e-07,
+      "loss": 0.0785,
+      "step": 8210
+    },
+    {
+      "epoch": 1.8661363636363637,
+      "grad_norm": 1.2258630129571717,
+      "learning_rate": 8.67471467730569e-07,
+      "loss": 0.0243,
+      "step": 8211
+    },
+    {
+      "epoch": 1.8663636363636362,
+      "grad_norm": 2.663474924998656,
+      "learning_rate": 8.673892056368263e-07,
+      "loss": 0.0514,
+      "step": 8212
+    },
+    {
+      "epoch": 1.8665909090909092,
+      "grad_norm": 1.8587820169958962,
+      "learning_rate": 8.673069386003449e-07,
+      "loss": 0.0298,
+      "step": 8213
+    },
+    {
+      "epoch": 1.8668181818181817,
+      "grad_norm": 1.3501153174460792,
+      "learning_rate": 8.672246666228024e-07,
+      "loss": 0.0424,
+      "step": 8214
+    },
+    {
+      "epoch": 1.8670454545454547,
+      "grad_norm": 1.8130291859479615,
+      "learning_rate": 8.671423897058761e-07,
+      "loss": 0.0375,
+      "step": 8215
+    },
+    {
+      "epoch": 1.8672727272727272,
+      "grad_norm": 2.1531552288176483,
+      "learning_rate": 8.670601078512443e-07,
+      "loss": 0.0677,
+      "step": 8216
+    },
+    {
+      "epoch": 1.8675000000000002,
+      "grad_norm": 2.872098823514195,
+      "learning_rate": 8.669778210605845e-07,
+      "loss": 0.0788,
+      "step": 8217
+    },
+    {
+      "epoch": 1.8677272727272727,
+      "grad_norm": 1.0724832716856993,
+      "learning_rate": 8.668955293355746e-07,
+      "loss": 0.0179,
+      "step": 8218
+    },
+    {
+      "epoch": 1.8679545454545454,
+      "grad_norm": 1.1828128356694667,
+      "learning_rate": 8.66813232677893e-07,
+      "loss": 0.0251,
+      "step": 8219
+    },
+    {
+      "epoch": 1.8681818181818182,
+      "grad_norm": 2.0337747530652335,
+      "learning_rate": 8.667309310892177e-07,
+      "loss": 0.04,
+      "step": 8220
+    },
+    {
+      "epoch": 1.868409090909091,
+      "grad_norm": 3.5251295066275397,
+      "learning_rate": 8.666486245712271e-07,
+      "loss": 0.089,
+      "step": 8221
+    },
+    {
+      "epoch": 1.8686363636363637,
+      "grad_norm": 1.984721530790709,
+      "learning_rate": 8.665663131255994e-07,
+      "loss": 0.0216,
+      "step": 8222
+    },
+    {
+      "epoch": 1.8688636363636364,
+      "grad_norm": 2.1442256691823087,
+      "learning_rate": 8.664839967540131e-07,
+      "loss": 0.0268,
+      "step": 8223
+    },
+    {
+      "epoch": 1.8690909090909091,
+      "grad_norm": 1.13372567554243,
+      "learning_rate": 8.664016754581468e-07,
+      "loss": 0.0351,
+      "step": 8224
+    },
+    {
+      "epoch": 1.8693181818181817,
+      "grad_norm": 1.8223946702138556,
+      "learning_rate": 8.663193492396793e-07,
+      "loss": 0.0458,
+      "step": 8225
+    },
+    {
+      "epoch": 1.8695454545454546,
+      "grad_norm": 1.1831612048741404,
+      "learning_rate": 8.662370181002892e-07,
+      "loss": 0.0242,
+      "step": 8226
+    },
+    {
+      "epoch": 1.8697727272727271,
+      "grad_norm": 1.8360093239199908,
+      "learning_rate": 8.661546820416554e-07,
+      "loss": 0.0456,
+      "step": 8227
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 2.3354668623764914,
+      "learning_rate": 8.66072341065457e-07,
+      "loss": 0.0464,
+      "step": 8228
+    },
+    {
+      "epoch": 1.8702272727272726,
+      "grad_norm": 0.5332542048756175,
+      "learning_rate": 8.65989995173373e-07,
+      "loss": 0.0175,
+      "step": 8229
+    },
+    {
+      "epoch": 1.8704545454545456,
+      "grad_norm": 1.6900064744964585,
+      "learning_rate": 8.659076443670825e-07,
+      "loss": 0.0329,
+      "step": 8230
+    },
+    {
+      "epoch": 1.8706818181818181,
+      "grad_norm": 1.9805762943419887,
+      "learning_rate": 8.658252886482649e-07,
+      "loss": 0.0607,
+      "step": 8231
+    },
+    {
+      "epoch": 1.8709090909090909,
+      "grad_norm": 2.886267807853451,
+      "learning_rate": 8.657429280185995e-07,
+      "loss": 0.0848,
+      "step": 8232
+    },
+    {
+      "epoch": 1.8711363636363636,
+      "grad_norm": 1.53845561695173,
+      "learning_rate": 8.656605624797656e-07,
+      "loss": 0.0308,
+      "step": 8233
+    },
+    {
+      "epoch": 1.8713636363636363,
+      "grad_norm": 1.3815316631346295,
+      "learning_rate": 8.655781920334432e-07,
+      "loss": 0.0339,
+      "step": 8234
+    },
+    {
+      "epoch": 1.871590909090909,
+      "grad_norm": 1.7233975134798776,
+      "learning_rate": 8.654958166813117e-07,
+      "loss": 0.0714,
+      "step": 8235
+    },
+    {
+      "epoch": 1.8718181818181818,
+      "grad_norm": 2.4473991781756714,
+      "learning_rate": 8.654134364250509e-07,
+      "loss": 0.0605,
+      "step": 8236
+    },
+    {
+      "epoch": 1.8720454545454546,
+      "grad_norm": 2.421861588044951,
+      "learning_rate": 8.653310512663408e-07,
+      "loss": 0.0543,
+      "step": 8237
+    },
+    {
+      "epoch": 1.8722727272727273,
+      "grad_norm": 2.001172621572925,
+      "learning_rate": 8.652486612068612e-07,
+      "loss": 0.0568,
+      "step": 8238
+    },
+    {
+      "epoch": 1.8725,
+      "grad_norm": 1.9032420420585654,
+      "learning_rate": 8.651662662482924e-07,
+      "loss": 0.0493,
+      "step": 8239
+    },
+    {
+      "epoch": 1.8727272727272726,
+      "grad_norm": 2.130551922778231,
+      "learning_rate": 8.650838663923142e-07,
+      "loss": 0.0367,
+      "step": 8240
+    },
+    {
+      "epoch": 1.8729545454545455,
+      "grad_norm": 2.021811383732857,
+      "learning_rate": 8.650014616406072e-07,
+      "loss": 0.0298,
+      "step": 8241
+    },
+    {
+      "epoch": 1.873181818181818,
+      "grad_norm": 1.4535980954567524,
+      "learning_rate": 8.649190519948518e-07,
+      "loss": 0.0327,
+      "step": 8242
+    },
+    {
+      "epoch": 1.873409090909091,
+      "grad_norm": 1.6657942219446615,
+      "learning_rate": 8.648366374567282e-07,
+      "loss": 0.031,
+      "step": 8243
+    },
+    {
+      "epoch": 1.8736363636363635,
+      "grad_norm": 1.9185371294080553,
+      "learning_rate": 8.647542180279174e-07,
+      "loss": 0.0498,
+      "step": 8244
+    },
+    {
+      "epoch": 1.8738636363636365,
+      "grad_norm": 2.707117585869523,
+      "learning_rate": 8.646717937100996e-07,
+      "loss": 0.0674,
+      "step": 8245
+    },
+    {
+      "epoch": 1.874090909090909,
+      "grad_norm": 2.7134332767657456,
+      "learning_rate": 8.64589364504956e-07,
+      "loss": 0.0445,
+      "step": 8246
+    },
+    {
+      "epoch": 1.8743181818181818,
+      "grad_norm": 2.8325122055412177,
+      "learning_rate": 8.64506930414167e-07,
+      "loss": 0.1029,
+      "step": 8247
+    },
+    {
+      "epoch": 1.8745454545454545,
+      "grad_norm": 2.3990882377200493,
+      "learning_rate": 8.64424491439414e-07,
+      "loss": 0.1059,
+      "step": 8248
+    },
+    {
+      "epoch": 1.8747727272727273,
+      "grad_norm": 1.9006136295047613,
+      "learning_rate": 8.643420475823779e-07,
+      "loss": 0.0643,
+      "step": 8249
+    },
+    {
+      "epoch": 1.875,
+      "grad_norm": 2.2269622777499922,
+      "learning_rate": 8.642595988447398e-07,
+      "loss": 0.0767,
+      "step": 8250
+    },
+    {
+      "epoch": 1.8752272727272727,
+      "grad_norm": 1.8224253676102213,
+      "learning_rate": 8.641771452281812e-07,
+      "loss": 0.052,
+      "step": 8251
+    },
+    {
+      "epoch": 1.8754545454545455,
+      "grad_norm": 2.2956271736541547,
+      "learning_rate": 8.640946867343833e-07,
+      "loss": 0.0452,
+      "step": 8252
+    },
+    {
+      "epoch": 1.8756818181818182,
+      "grad_norm": 2.758857292023191,
+      "learning_rate": 8.640122233650277e-07,
+      "loss": 0.0924,
+      "step": 8253
+    },
+    {
+      "epoch": 1.875909090909091,
+      "grad_norm": 2.312716947098934,
+      "learning_rate": 8.639297551217957e-07,
+      "loss": 0.0893,
+      "step": 8254
+    },
+    {
+      "epoch": 1.8761363636363635,
+      "grad_norm": 1.9304198794454825,
+      "learning_rate": 8.638472820063692e-07,
+      "loss": 0.0424,
+      "step": 8255
+    },
+    {
+      "epoch": 1.8763636363636365,
+      "grad_norm": 2.0957288388525086,
+      "learning_rate": 8.637648040204299e-07,
+      "loss": 0.046,
+      "step": 8256
+    },
+    {
+      "epoch": 1.876590909090909,
+      "grad_norm": 1.9037145166740543,
+      "learning_rate": 8.636823211656601e-07,
+      "loss": 0.0668,
+      "step": 8257
+    },
+    {
+      "epoch": 1.876818181818182,
+      "grad_norm": 0.9010827968779203,
+      "learning_rate": 8.63599833443741e-07,
+      "loss": 0.0387,
+      "step": 8258
+    },
+    {
+      "epoch": 1.8770454545454545,
+      "grad_norm": 1.6523172759544884,
+      "learning_rate": 8.63517340856355e-07,
+      "loss": 0.0433,
+      "step": 8259
+    },
+    {
+      "epoch": 1.8772727272727274,
+      "grad_norm": 1.371629631654666,
+      "learning_rate": 8.634348434051844e-07,
+      "loss": 0.0374,
+      "step": 8260
+    },
+    {
+      "epoch": 1.8775,
+      "grad_norm": 1.454014182850163,
+      "learning_rate": 8.633523410919114e-07,
+      "loss": 0.018,
+      "step": 8261
+    },
+    {
+      "epoch": 1.8777272727272727,
+      "grad_norm": 2.009810849139226,
+      "learning_rate": 8.632698339182182e-07,
+      "loss": 0.074,
+      "step": 8262
+    },
+    {
+      "epoch": 1.8779545454545454,
+      "grad_norm": 2.1477358498891843,
+      "learning_rate": 8.631873218857875e-07,
+      "loss": 0.0562,
+      "step": 8263
+    },
+    {
+      "epoch": 1.8781818181818182,
+      "grad_norm": 2.0377997862766684,
+      "learning_rate": 8.631048049963018e-07,
+      "loss": 0.0446,
+      "step": 8264
+    },
+    {
+      "epoch": 1.878409090909091,
+      "grad_norm": 2.7345031606864176,
+      "learning_rate": 8.630222832514437e-07,
+      "loss": 0.1068,
+      "step": 8265
+    },
+    {
+      "epoch": 1.8786363636363637,
+      "grad_norm": 1.4276199816524182,
+      "learning_rate": 8.62939756652896e-07,
+      "loss": 0.0329,
+      "step": 8266
+    },
+    {
+      "epoch": 1.8788636363636364,
+      "grad_norm": 4.137081957841838,
+      "learning_rate": 8.628572252023415e-07,
+      "loss": 0.0863,
+      "step": 8267
+    },
+    {
+      "epoch": 1.8790909090909091,
+      "grad_norm": 1.6049366822462405,
+      "learning_rate": 8.627746889014634e-07,
+      "loss": 0.0235,
+      "step": 8268
+    },
+    {
+      "epoch": 1.8793181818181819,
+      "grad_norm": 2.053440541487594,
+      "learning_rate": 8.626921477519443e-07,
+      "loss": 0.0586,
+      "step": 8269
+    },
+    {
+      "epoch": 1.8795454545454544,
+      "grad_norm": 1.6348319955940054,
+      "learning_rate": 8.626096017554677e-07,
+      "loss": 0.0225,
+      "step": 8270
+    },
+    {
+      "epoch": 1.8797727272727274,
+      "grad_norm": 1.5532165380160796,
+      "learning_rate": 8.625270509137168e-07,
+      "loss": 0.0422,
+      "step": 8271
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.7909683610462022,
+      "learning_rate": 8.624444952283752e-07,
+      "loss": 0.0457,
+      "step": 8272
+    },
+    {
+      "epoch": 1.8802272727272729,
+      "grad_norm": 1.9301448681541453,
+      "learning_rate": 8.623619347011259e-07,
+      "loss": 0.0591,
+      "step": 8273
+    },
+    {
+      "epoch": 1.8804545454545454,
+      "grad_norm": 1.9699409300287705,
+      "learning_rate": 8.622793693336524e-07,
+      "loss": 0.0398,
+      "step": 8274
+    },
+    {
+      "epoch": 1.8806818181818183,
+      "grad_norm": 2.2528649013214954,
+      "learning_rate": 8.621967991276389e-07,
+      "loss": 0.0616,
+      "step": 8275
+    },
+    {
+      "epoch": 1.8809090909090909,
+      "grad_norm": 1.4970531860621858,
+      "learning_rate": 8.621142240847688e-07,
+      "loss": 0.0425,
+      "step": 8276
+    },
+    {
+      "epoch": 1.8811363636363636,
+      "grad_norm": 1.8409048364783929,
+      "learning_rate": 8.620316442067261e-07,
+      "loss": 0.0453,
+      "step": 8277
+    },
+    {
+      "epoch": 1.8813636363636363,
+      "grad_norm": 1.5388002389628679,
+      "learning_rate": 8.619490594951946e-07,
+      "loss": 0.0274,
+      "step": 8278
+    },
+    {
+      "epoch": 1.881590909090909,
+      "grad_norm": 1.0889573037951383,
+      "learning_rate": 8.618664699518582e-07,
+      "loss": 0.0363,
+      "step": 8279
+    },
+    {
+      "epoch": 1.8818181818181818,
+      "grad_norm": 1.8237163166714025,
+      "learning_rate": 8.617838755784013e-07,
+      "loss": 0.0451,
+      "step": 8280
+    },
+    {
+      "epoch": 1.8820454545454546,
+      "grad_norm": 1.6750826014007179,
+      "learning_rate": 8.617012763765084e-07,
+      "loss": 0.063,
+      "step": 8281
+    },
+    {
+      "epoch": 1.8822727272727273,
+      "grad_norm": 2.32450292460013,
+      "learning_rate": 8.616186723478631e-07,
+      "loss": 0.0391,
+      "step": 8282
+    },
+    {
+      "epoch": 1.8824999999999998,
+      "grad_norm": 2.249109482086292,
+      "learning_rate": 8.615360634941506e-07,
+      "loss": 0.03,
+      "step": 8283
+    },
+    {
+      "epoch": 1.8827272727272728,
+      "grad_norm": 2.5244270925453063,
+      "learning_rate": 8.61453449817055e-07,
+      "loss": 0.0627,
+      "step": 8284
+    },
+    {
+      "epoch": 1.8829545454545453,
+      "grad_norm": 1.757597069528191,
+      "learning_rate": 8.613708313182609e-07,
+      "loss": 0.061,
+      "step": 8285
+    },
+    {
+      "epoch": 1.8831818181818183,
+      "grad_norm": 2.908847790947674,
+      "learning_rate": 8.612882079994534e-07,
+      "loss": 0.0748,
+      "step": 8286
+    },
+    {
+      "epoch": 1.8834090909090908,
+      "grad_norm": 1.744666060733719,
+      "learning_rate": 8.612055798623168e-07,
+      "loss": 0.0347,
+      "step": 8287
+    },
+    {
+      "epoch": 1.8836363636363638,
+      "grad_norm": 2.1047715851132263,
+      "learning_rate": 8.611229469085367e-07,
+      "loss": 0.0859,
+      "step": 8288
+    },
+    {
+      "epoch": 1.8838636363636363,
+      "grad_norm": 1.619646513198746,
+      "learning_rate": 8.610403091397976e-07,
+      "loss": 0.0295,
+      "step": 8289
+    },
+    {
+      "epoch": 1.884090909090909,
+      "grad_norm": 2.033533444442194,
+      "learning_rate": 8.609576665577848e-07,
+      "loss": 0.0626,
+      "step": 8290
+    },
+    {
+      "epoch": 1.8843181818181818,
+      "grad_norm": 1.1600303363623403,
+      "learning_rate": 8.608750191641835e-07,
+      "loss": 0.0517,
+      "step": 8291
+    },
+    {
+      "epoch": 1.8845454545454545,
+      "grad_norm": 1.5059623060637428,
+      "learning_rate": 8.607923669606792e-07,
+      "loss": 0.0388,
+      "step": 8292
+    },
+    {
+      "epoch": 1.8847727272727273,
+      "grad_norm": 4.0367414832237785,
+      "learning_rate": 8.60709709948957e-07,
+      "loss": 0.119,
+      "step": 8293
+    },
+    {
+      "epoch": 1.885,
+      "grad_norm": 2.9485896152810063,
+      "learning_rate": 8.606270481307027e-07,
+      "loss": 0.0464,
+      "step": 8294
+    },
+    {
+      "epoch": 1.8852272727272728,
+      "grad_norm": 2.221972045447405,
+      "learning_rate": 8.605443815076018e-07,
+      "loss": 0.039,
+      "step": 8295
+    },
+    {
+      "epoch": 1.8854545454545455,
+      "grad_norm": 2.1778960006723427,
+      "learning_rate": 8.604617100813399e-07,
+      "loss": 0.0506,
+      "step": 8296
+    },
+    {
+      "epoch": 1.8856818181818182,
+      "grad_norm": 2.5393602027644455,
+      "learning_rate": 8.603790338536032e-07,
+      "loss": 0.0693,
+      "step": 8297
+    },
+    {
+      "epoch": 1.8859090909090908,
+      "grad_norm": 2.424056113076695,
+      "learning_rate": 8.602963528260772e-07,
+      "loss": 0.0602,
+      "step": 8298
+    },
+    {
+      "epoch": 1.8861363636363637,
+      "grad_norm": 2.041867360194765,
+      "learning_rate": 8.60213667000448e-07,
+      "loss": 0.0315,
+      "step": 8299
+    },
+    {
+      "epoch": 1.8863636363636362,
+      "grad_norm": 1.5940436472228157,
+      "learning_rate": 8.601309763784019e-07,
+      "loss": 0.0508,
+      "step": 8300
+    },
+    {
+      "epoch": 1.8865909090909092,
+      "grad_norm": 2.4185317754171005,
+      "learning_rate": 8.600482809616249e-07,
+      "loss": 0.0444,
+      "step": 8301
+    },
+    {
+      "epoch": 1.8868181818181817,
+      "grad_norm": 1.0536746558967267,
+      "learning_rate": 8.599655807518035e-07,
+      "loss": 0.0148,
+      "step": 8302
+    },
+    {
+      "epoch": 1.8870454545454547,
+      "grad_norm": 2.6066201440030006,
+      "learning_rate": 8.598828757506239e-07,
+      "loss": 0.0599,
+      "step": 8303
+    },
+    {
+      "epoch": 1.8872727272727272,
+      "grad_norm": 1.7181715382680995,
+      "learning_rate": 8.598001659597727e-07,
+      "loss": 0.0371,
+      "step": 8304
+    },
+    {
+      "epoch": 1.8875,
+      "grad_norm": 2.622211421134461,
+      "learning_rate": 8.597174513809365e-07,
+      "loss": 0.0562,
+      "step": 8305
+    },
+    {
+      "epoch": 1.8877272727272727,
+      "grad_norm": 1.3795662408381044,
+      "learning_rate": 8.59634732015802e-07,
+      "loss": 0.0422,
+      "step": 8306
+    },
+    {
+      "epoch": 1.8879545454545454,
+      "grad_norm": 1.6974769238437668,
+      "learning_rate": 8.59552007866056e-07,
+      "loss": 0.0405,
+      "step": 8307
+    },
+    {
+      "epoch": 1.8881818181818182,
+      "grad_norm": 2.258735780229916,
+      "learning_rate": 8.594692789333852e-07,
+      "loss": 0.0701,
+      "step": 8308
+    },
+    {
+      "epoch": 1.888409090909091,
+      "grad_norm": 1.4348264474957506,
+      "learning_rate": 8.593865452194769e-07,
+      "loss": 0.0345,
+      "step": 8309
+    },
+    {
+      "epoch": 1.8886363636363637,
+      "grad_norm": 1.5607901600709748,
+      "learning_rate": 8.59303806726018e-07,
+      "loss": 0.042,
+      "step": 8310
+    },
+    {
+      "epoch": 1.8888636363636364,
+      "grad_norm": 1.446851795760569,
+      "learning_rate": 8.592210634546958e-07,
+      "loss": 0.0235,
+      "step": 8311
+    },
+    {
+      "epoch": 1.8890909090909092,
+      "grad_norm": 1.5216838256038947,
+      "learning_rate": 8.591383154071974e-07,
+      "loss": 0.0524,
+      "step": 8312
+    },
+    {
+      "epoch": 1.8893181818181817,
+      "grad_norm": 2.7891578987687504,
+      "learning_rate": 8.590555625852101e-07,
+      "loss": 0.0592,
+      "step": 8313
+    },
+    {
+      "epoch": 1.8895454545454546,
+      "grad_norm": 2.3139590473439795,
+      "learning_rate": 8.589728049904219e-07,
+      "loss": 0.0537,
+      "step": 8314
+    },
+    {
+      "epoch": 1.8897727272727272,
+      "grad_norm": 2.213939779575695,
+      "learning_rate": 8.588900426245198e-07,
+      "loss": 0.0498,
+      "step": 8315
+    },
+    {
+      "epoch": 1.8900000000000001,
+      "grad_norm": 2.8230549371779197,
+      "learning_rate": 8.588072754891917e-07,
+      "loss": 0.0605,
+      "step": 8316
+    },
+    {
+      "epoch": 1.8902272727272726,
+      "grad_norm": 1.9416365060246457,
+      "learning_rate": 8.587245035861253e-07,
+      "loss": 0.0423,
+      "step": 8317
+    },
+    {
+      "epoch": 1.8904545454545456,
+      "grad_norm": 2.373859275708118,
+      "learning_rate": 8.586417269170088e-07,
+      "loss": 0.043,
+      "step": 8318
+    },
+    {
+      "epoch": 1.8906818181818181,
+      "grad_norm": 1.3274789494520765,
+      "learning_rate": 8.585589454835296e-07,
+      "loss": 0.0248,
+      "step": 8319
+    },
+    {
+      "epoch": 1.8909090909090909,
+      "grad_norm": 1.958417725774166,
+      "learning_rate": 8.58476159287376e-07,
+      "loss": 0.0261,
+      "step": 8320
+    },
+    {
+      "epoch": 1.8911363636363636,
+      "grad_norm": 1.7462594632175557,
+      "learning_rate": 8.583933683302364e-07,
+      "loss": 0.0378,
+      "step": 8321
+    },
+    {
+      "epoch": 1.8913636363636364,
+      "grad_norm": 3.9660304948799556,
+      "learning_rate": 8.583105726137987e-07,
+      "loss": 0.0677,
+      "step": 8322
+    },
+    {
+      "epoch": 1.891590909090909,
+      "grad_norm": 2.113921139888728,
+      "learning_rate": 8.582277721397514e-07,
+      "loss": 0.0474,
+      "step": 8323
+    },
+    {
+      "epoch": 1.8918181818181818,
+      "grad_norm": 1.6810102584716706,
+      "learning_rate": 8.58144966909783e-07,
+      "loss": 0.0399,
+      "step": 8324
+    },
+    {
+      "epoch": 1.8920454545454546,
+      "grad_norm": 2.9998084972268395,
+      "learning_rate": 8.58062156925582e-07,
+      "loss": 0.0753,
+      "step": 8325
+    },
+    {
+      "epoch": 1.892272727272727,
+      "grad_norm": 2.434717860644698,
+      "learning_rate": 8.579793421888369e-07,
+      "loss": 0.0407,
+      "step": 8326
+    },
+    {
+      "epoch": 1.8925,
+      "grad_norm": 2.542096085943469,
+      "learning_rate": 8.578965227012367e-07,
+      "loss": 0.0926,
+      "step": 8327
+    },
+    {
+      "epoch": 1.8927272727272726,
+      "grad_norm": 2.7541326383361153,
+      "learning_rate": 8.5781369846447e-07,
+      "loss": 0.1184,
+      "step": 8328
+    },
+    {
+      "epoch": 1.8929545454545456,
+      "grad_norm": 1.6050577915218764,
+      "learning_rate": 8.57730869480226e-07,
+      "loss": 0.0553,
+      "step": 8329
+    },
+    {
+      "epoch": 1.893181818181818,
+      "grad_norm": 1.2961707961186926,
+      "learning_rate": 8.576480357501933e-07,
+      "loss": 0.0471,
+      "step": 8330
+    },
+    {
+      "epoch": 1.893409090909091,
+      "grad_norm": 2.1882846012893764,
+      "learning_rate": 8.575651972760613e-07,
+      "loss": 0.0413,
+      "step": 8331
+    },
+    {
+      "epoch": 1.8936363636363636,
+      "grad_norm": 2.042478832900928,
+      "learning_rate": 8.574823540595192e-07,
+      "loss": 0.0579,
+      "step": 8332
+    },
+    {
+      "epoch": 1.8938636363636365,
+      "grad_norm": 2.6881483036278957,
+      "learning_rate": 8.573995061022564e-07,
+      "loss": 0.0805,
+      "step": 8333
+    },
+    {
+      "epoch": 1.894090909090909,
+      "grad_norm": 2.4094822135621414,
+      "learning_rate": 8.573166534059622e-07,
+      "loss": 0.0511,
+      "step": 8334
+    },
+    {
+      "epoch": 1.8943181818181818,
+      "grad_norm": 1.9340782524700257,
+      "learning_rate": 8.572337959723262e-07,
+      "loss": 0.042,
+      "step": 8335
+    },
+    {
+      "epoch": 1.8945454545454545,
+      "grad_norm": 1.5519577710132717,
+      "learning_rate": 8.57150933803038e-07,
+      "loss": 0.0306,
+      "step": 8336
+    },
+    {
+      "epoch": 1.8947727272727273,
+      "grad_norm": 2.272527402105157,
+      "learning_rate": 8.57068066899787e-07,
+      "loss": 0.0417,
+      "step": 8337
+    },
+    {
+      "epoch": 1.895,
+      "grad_norm": 1.8183829649867738,
+      "learning_rate": 8.569851952642634e-07,
+      "loss": 0.03,
+      "step": 8338
+    },
+    {
+      "epoch": 1.8952272727272728,
+      "grad_norm": 1.7775357715489473,
+      "learning_rate": 8.569023188981569e-07,
+      "loss": 0.038,
+      "step": 8339
+    },
+    {
+      "epoch": 1.8954545454545455,
+      "grad_norm": 1.7337724744652474,
+      "learning_rate": 8.568194378031577e-07,
+      "loss": 0.0486,
+      "step": 8340
+    },
+    {
+      "epoch": 1.895681818181818,
+      "grad_norm": 2.0350651743194077,
+      "learning_rate": 8.567365519809556e-07,
+      "loss": 0.0753,
+      "step": 8341
+    },
+    {
+      "epoch": 1.895909090909091,
+      "grad_norm": 2.4621988349998505,
+      "learning_rate": 8.566536614332409e-07,
+      "loss": 0.0713,
+      "step": 8342
+    },
+    {
+      "epoch": 1.8961363636363635,
+      "grad_norm": 2.0238133215354015,
+      "learning_rate": 8.56570766161704e-07,
+      "loss": 0.0743,
+      "step": 8343
+    },
+    {
+      "epoch": 1.8963636363636365,
+      "grad_norm": 2.365442889062235,
+      "learning_rate": 8.56487866168035e-07,
+      "loss": 0.0367,
+      "step": 8344
+    },
+    {
+      "epoch": 1.896590909090909,
+      "grad_norm": 2.5367624627526584,
+      "learning_rate": 8.564049614539248e-07,
+      "loss": 0.0512,
+      "step": 8345
+    },
+    {
+      "epoch": 1.896818181818182,
+      "grad_norm": 2.3276192216463603,
+      "learning_rate": 8.563220520210636e-07,
+      "loss": 0.041,
+      "step": 8346
+    },
+    {
+      "epoch": 1.8970454545454545,
+      "grad_norm": 2.3149358017622528,
+      "learning_rate": 8.562391378711421e-07,
+      "loss": 0.106,
+      "step": 8347
+    },
+    {
+      "epoch": 1.8972727272727272,
+      "grad_norm": 1.5126871654213891,
+      "learning_rate": 8.561562190058512e-07,
+      "loss": 0.0336,
+      "step": 8348
+    },
+    {
+      "epoch": 1.8975,
+      "grad_norm": 1.9568599932427562,
+      "learning_rate": 8.560732954268819e-07,
+      "loss": 0.0702,
+      "step": 8349
+    },
+    {
+      "epoch": 1.8977272727272727,
+      "grad_norm": 2.4889516605856037,
+      "learning_rate": 8.559903671359249e-07,
+      "loss": 0.0617,
+      "step": 8350
+    },
+    {
+      "epoch": 1.8979545454545454,
+      "grad_norm": 1.1994905884187006,
+      "learning_rate": 8.559074341346713e-07,
+      "loss": 0.0234,
+      "step": 8351
+    },
+    {
+      "epoch": 1.8981818181818182,
+      "grad_norm": 2.5841881075326594,
+      "learning_rate": 8.558244964248122e-07,
+      "loss": 0.0613,
+      "step": 8352
+    },
+    {
+      "epoch": 1.898409090909091,
+      "grad_norm": 2.2614454728300184,
+      "learning_rate": 8.55741554008039e-07,
+      "loss": 0.045,
+      "step": 8353
+    },
+    {
+      "epoch": 1.8986363636363637,
+      "grad_norm": 2.0164218118177284,
+      "learning_rate": 8.556586068860428e-07,
+      "loss": 0.0344,
+      "step": 8354
+    },
+    {
+      "epoch": 1.8988636363636364,
+      "grad_norm": 1.6148576010444158,
+      "learning_rate": 8.555756550605154e-07,
+      "loss": 0.0293,
+      "step": 8355
+    },
+    {
+      "epoch": 1.899090909090909,
+      "grad_norm": 1.5665974997621075,
+      "learning_rate": 8.554926985331481e-07,
+      "loss": 0.0343,
+      "step": 8356
+    },
+    {
+      "epoch": 1.899318181818182,
+      "grad_norm": 1.9077691041782063,
+      "learning_rate": 8.554097373056325e-07,
+      "loss": 0.0169,
+      "step": 8357
+    },
+    {
+      "epoch": 1.8995454545454544,
+      "grad_norm": 2.372895823687386,
+      "learning_rate": 8.553267713796605e-07,
+      "loss": 0.0469,
+      "step": 8358
+    },
+    {
+      "epoch": 1.8997727272727274,
+      "grad_norm": 0.9779305722949622,
+      "learning_rate": 8.552438007569237e-07,
+      "loss": 0.0398,
+      "step": 8359
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.7287149545146088,
+      "learning_rate": 8.551608254391141e-07,
+      "loss": 0.108,
+      "step": 8360
+    },
+    {
+      "epoch": 1.9002272727272729,
+      "grad_norm": 1.2800891077472027,
+      "learning_rate": 8.550778454279239e-07,
+      "loss": 0.026,
+      "step": 8361
+    },
+    {
+      "epoch": 1.9004545454545454,
+      "grad_norm": 1.8212153657806283,
+      "learning_rate": 8.549948607250448e-07,
+      "loss": 0.0556,
+      "step": 8362
+    },
+    {
+      "epoch": 1.9006818181818181,
+      "grad_norm": 1.8573503115287555,
+      "learning_rate": 8.549118713321694e-07,
+      "loss": 0.0556,
+      "step": 8363
+    },
+    {
+      "epoch": 1.9009090909090909,
+      "grad_norm": 2.581405860294612,
+      "learning_rate": 8.548288772509897e-07,
+      "loss": 0.0738,
+      "step": 8364
+    },
+    {
+      "epoch": 1.9011363636363636,
+      "grad_norm": 2.4567796954159067,
+      "learning_rate": 8.547458784831984e-07,
+      "loss": 0.0648,
+      "step": 8365
+    },
+    {
+      "epoch": 1.9013636363636364,
+      "grad_norm": 1.4369150815008078,
+      "learning_rate": 8.546628750304877e-07,
+      "loss": 0.053,
+      "step": 8366
+    },
+    {
+      "epoch": 1.901590909090909,
+      "grad_norm": 1.8405216024432143,
+      "learning_rate": 8.545798668945505e-07,
+      "loss": 0.0495,
+      "step": 8367
+    },
+    {
+      "epoch": 1.9018181818181819,
+      "grad_norm": 2.390629352835323,
+      "learning_rate": 8.544968540770791e-07,
+      "loss": 0.0623,
+      "step": 8368
+    },
+    {
+      "epoch": 1.9020454545454546,
+      "grad_norm": 7.725872714801204,
+      "learning_rate": 8.544138365797664e-07,
+      "loss": 0.0493,
+      "step": 8369
+    },
+    {
+      "epoch": 1.9022727272727273,
+      "grad_norm": 2.3742754864262463,
+      "learning_rate": 8.543308144043055e-07,
+      "loss": 0.0516,
+      "step": 8370
+    },
+    {
+      "epoch": 1.9024999999999999,
+      "grad_norm": 3.4607836417428963,
+      "learning_rate": 8.542477875523893e-07,
+      "loss": 0.0955,
+      "step": 8371
+    },
+    {
+      "epoch": 1.9027272727272728,
+      "grad_norm": 2.544372148900421,
+      "learning_rate": 8.541647560257106e-07,
+      "loss": 0.0428,
+      "step": 8372
+    },
+    {
+      "epoch": 1.9029545454545453,
+      "grad_norm": 2.660006992878164,
+      "learning_rate": 8.540817198259628e-07,
+      "loss": 0.057,
+      "step": 8373
+    },
+    {
+      "epoch": 1.9031818181818183,
+      "grad_norm": 1.3617415705486335,
+      "learning_rate": 8.53998678954839e-07,
+      "loss": 0.036,
+      "step": 8374
+    },
+    {
+      "epoch": 1.9034090909090908,
+      "grad_norm": 1.22710835803522,
+      "learning_rate": 8.539156334140326e-07,
+      "loss": 0.0352,
+      "step": 8375
+    },
+    {
+      "epoch": 1.9036363636363638,
+      "grad_norm": 1.4189594168182371,
+      "learning_rate": 8.538325832052373e-07,
+      "loss": 0.0379,
+      "step": 8376
+    },
+    {
+      "epoch": 1.9038636363636363,
+      "grad_norm": 2.2912978009225324,
+      "learning_rate": 8.53749528330146e-07,
+      "loss": 0.0708,
+      "step": 8377
+    },
+    {
+      "epoch": 1.904090909090909,
+      "grad_norm": 1.3216422667859904,
+      "learning_rate": 8.53666468790453e-07,
+      "loss": 0.0436,
+      "step": 8378
+    },
+    {
+      "epoch": 1.9043181818181818,
+      "grad_norm": 2.0680838114414897,
+      "learning_rate": 8.535834045878519e-07,
+      "loss": 0.0395,
+      "step": 8379
+    },
+    {
+      "epoch": 1.9045454545454545,
+      "grad_norm": 1.751538406789565,
+      "learning_rate": 8.535003357240363e-07,
+      "loss": 0.0476,
+      "step": 8380
+    },
+    {
+      "epoch": 1.9047727272727273,
+      "grad_norm": 2.2301197151507095,
+      "learning_rate": 8.534172622007e-07,
+      "loss": 0.0465,
+      "step": 8381
+    },
+    {
+      "epoch": 1.905,
+      "grad_norm": 2.1267654292767055,
+      "learning_rate": 8.533341840195375e-07,
+      "loss": 0.0725,
+      "step": 8382
+    },
+    {
+      "epoch": 1.9052272727272728,
+      "grad_norm": 3.7871835672874288,
+      "learning_rate": 8.532511011822425e-07,
+      "loss": 0.1499,
+      "step": 8383
+    },
+    {
+      "epoch": 1.9054545454545453,
+      "grad_norm": 2.116539925382068,
+      "learning_rate": 8.531680136905094e-07,
+      "loss": 0.0443,
+      "step": 8384
+    },
+    {
+      "epoch": 1.9056818181818183,
+      "grad_norm": 2.5149811528801824,
+      "learning_rate": 8.530849215460322e-07,
+      "loss": 0.091,
+      "step": 8385
+    },
+    {
+      "epoch": 1.9059090909090908,
+      "grad_norm": 2.2870765574356855,
+      "learning_rate": 8.530018247505058e-07,
+      "loss": 0.0411,
+      "step": 8386
+    },
+    {
+      "epoch": 1.9061363636363637,
+      "grad_norm": 1.6664622075878854,
+      "learning_rate": 8.529187233056244e-07,
+      "loss": 0.0459,
+      "step": 8387
+    },
+    {
+      "epoch": 1.9063636363636363,
+      "grad_norm": 1.4224000492951647,
+      "learning_rate": 8.528356172130825e-07,
+      "loss": 0.0504,
+      "step": 8388
+    },
+    {
+      "epoch": 1.9065909090909092,
+      "grad_norm": 2.107460664310557,
+      "learning_rate": 8.527525064745751e-07,
+      "loss": 0.0578,
+      "step": 8389
+    },
+    {
+      "epoch": 1.9068181818181817,
+      "grad_norm": 2.1295116288936735,
+      "learning_rate": 8.526693910917966e-07,
+      "loss": 0.0539,
+      "step": 8390
+    },
+    {
+      "epoch": 1.9070454545454547,
+      "grad_norm": 1.8055882445506104,
+      "learning_rate": 8.525862710664421e-07,
+      "loss": 0.0378,
+      "step": 8391
+    },
+    {
+      "epoch": 1.9072727272727272,
+      "grad_norm": 2.3728148350886755,
+      "learning_rate": 8.525031464002062e-07,
+      "loss": 0.0676,
+      "step": 8392
+    },
+    {
+      "epoch": 1.9075,
+      "grad_norm": 0.862892677056585,
+      "learning_rate": 8.524200170947846e-07,
+      "loss": 0.0188,
+      "step": 8393
+    },
+    {
+      "epoch": 1.9077272727272727,
+      "grad_norm": 3.1860120213323495,
+      "learning_rate": 8.523368831518719e-07,
+      "loss": 0.0901,
+      "step": 8394
+    },
+    {
+      "epoch": 1.9079545454545455,
+      "grad_norm": 1.343783463057835,
+      "learning_rate": 8.522537445731638e-07,
+      "loss": 0.0747,
+      "step": 8395
+    },
+    {
+      "epoch": 1.9081818181818182,
+      "grad_norm": 3.7278759711901226,
+      "learning_rate": 8.521706013603551e-07,
+      "loss": 0.095,
+      "step": 8396
+    },
+    {
+      "epoch": 1.908409090909091,
+      "grad_norm": 1.2483848229427708,
+      "learning_rate": 8.520874535151415e-07,
+      "loss": 0.0709,
+      "step": 8397
+    },
+    {
+      "epoch": 1.9086363636363637,
+      "grad_norm": 2.4291125782734624,
+      "learning_rate": 8.520043010392187e-07,
+      "loss": 0.074,
+      "step": 8398
+    },
+    {
+      "epoch": 1.9088636363636362,
+      "grad_norm": 1.9640760410921299,
+      "learning_rate": 8.519211439342821e-07,
+      "loss": 0.0478,
+      "step": 8399
+    },
+    {
+      "epoch": 1.9090909090909092,
+      "grad_norm": 0.9993475214688049,
+      "learning_rate": 8.518379822020276e-07,
+      "loss": 0.0293,
+      "step": 8400
+    },
+    {
+      "epoch": 1.9093181818181817,
+      "grad_norm": 1.7367937377180527,
+      "learning_rate": 8.517548158441506e-07,
+      "loss": 0.0466,
+      "step": 8401
+    },
+    {
+      "epoch": 1.9095454545454547,
+      "grad_norm": 1.903550603766607,
+      "learning_rate": 8.516716448623477e-07,
+      "loss": 0.0337,
+      "step": 8402
+    },
+    {
+      "epoch": 1.9097727272727272,
+      "grad_norm": 1.9959996592647018,
+      "learning_rate": 8.515884692583143e-07,
+      "loss": 0.0418,
+      "step": 8403
+    },
+    {
+      "epoch": 1.9100000000000001,
+      "grad_norm": 1.7350309835759516,
+      "learning_rate": 8.515052890337468e-07,
+      "loss": 0.0402,
+      "step": 8404
+    },
+    {
+      "epoch": 1.9102272727272727,
+      "grad_norm": 1.580152123267052,
+      "learning_rate": 8.514221041903412e-07,
+      "loss": 0.0269,
+      "step": 8405
+    },
+    {
+      "epoch": 1.9104545454545454,
+      "grad_norm": 0.9715344606663674,
+      "learning_rate": 8.51338914729794e-07,
+      "loss": 0.0144,
+      "step": 8406
+    },
+    {
+      "epoch": 1.9106818181818181,
+      "grad_norm": 1.716302436381525,
+      "learning_rate": 8.512557206538012e-07,
+      "loss": 0.0481,
+      "step": 8407
+    },
+    {
+      "epoch": 1.910909090909091,
+      "grad_norm": 1.319761563736713,
+      "learning_rate": 8.511725219640596e-07,
+      "loss": 0.0376,
+      "step": 8408
+    },
+    {
+      "epoch": 1.9111363636363636,
+      "grad_norm": 1.8872343853397695,
+      "learning_rate": 8.510893186622659e-07,
+      "loss": 0.0681,
+      "step": 8409
+    },
+    {
+      "epoch": 1.9113636363636364,
+      "grad_norm": 1.6879158011302695,
+      "learning_rate": 8.510061107501165e-07,
+      "loss": 0.029,
+      "step": 8410
+    },
+    {
+      "epoch": 1.9115909090909091,
+      "grad_norm": 2.0894371673535623,
+      "learning_rate": 8.509228982293083e-07,
+      "loss": 0.0378,
+      "step": 8411
+    },
+    {
+      "epoch": 1.9118181818181819,
+      "grad_norm": 2.068568527640439,
+      "learning_rate": 8.508396811015378e-07,
+      "loss": 0.055,
+      "step": 8412
+    },
+    {
+      "epoch": 1.9120454545454546,
+      "grad_norm": 2.0322518867632,
+      "learning_rate": 8.507564593685025e-07,
+      "loss": 0.0459,
+      "step": 8413
+    },
+    {
+      "epoch": 1.9122727272727271,
+      "grad_norm": 1.8179197769725477,
+      "learning_rate": 8.506732330318988e-07,
+      "loss": 0.0469,
+      "step": 8414
+    },
+    {
+      "epoch": 1.9125,
+      "grad_norm": 2.032386623537287,
+      "learning_rate": 8.505900020934243e-07,
+      "loss": 0.039,
+      "step": 8415
+    },
+    {
+      "epoch": 1.9127272727272726,
+      "grad_norm": 1.9450961138630973,
+      "learning_rate": 8.505067665547761e-07,
+      "loss": 0.0459,
+      "step": 8416
+    },
+    {
+      "epoch": 1.9129545454545456,
+      "grad_norm": 2.7470902647822246,
+      "learning_rate": 8.504235264176517e-07,
+      "loss": 0.0807,
+      "step": 8417
+    },
+    {
+      "epoch": 1.913181818181818,
+      "grad_norm": 1.2577777986237995,
+      "learning_rate": 8.503402816837482e-07,
+      "loss": 0.0258,
+      "step": 8418
+    },
+    {
+      "epoch": 1.913409090909091,
+      "grad_norm": 2.0199897244191227,
+      "learning_rate": 8.50257032354763e-07,
+      "loss": 0.0365,
+      "step": 8419
+    },
+    {
+      "epoch": 1.9136363636363636,
+      "grad_norm": 1.6197561266363751,
+      "learning_rate": 8.501737784323943e-07,
+      "loss": 0.0249,
+      "step": 8420
+    },
+    {
+      "epoch": 1.9138636363636363,
+      "grad_norm": 1.848373065363987,
+      "learning_rate": 8.500905199183393e-07,
+      "loss": 0.0362,
+      "step": 8421
+    },
+    {
+      "epoch": 1.914090909090909,
+      "grad_norm": 1.5617011541345902,
+      "learning_rate": 8.500072568142957e-07,
+      "loss": 0.0406,
+      "step": 8422
+    },
+    {
+      "epoch": 1.9143181818181818,
+      "grad_norm": 0.6383303584488915,
+      "learning_rate": 8.499239891219618e-07,
+      "loss": 0.0192,
+      "step": 8423
+    },
+    {
+      "epoch": 1.9145454545454546,
+      "grad_norm": 1.7638153991859706,
+      "learning_rate": 8.498407168430354e-07,
+      "loss": 0.0633,
+      "step": 8424
+    },
+    {
+      "epoch": 1.9147727272727273,
+      "grad_norm": 8.663529588288588,
+      "learning_rate": 8.497574399792144e-07,
+      "loss": 0.04,
+      "step": 8425
+    },
+    {
+      "epoch": 1.915,
+      "grad_norm": 1.1254838420787971,
+      "learning_rate": 8.496741585321971e-07,
+      "loss": 0.0251,
+      "step": 8426
+    },
+    {
+      "epoch": 1.9152272727272728,
+      "grad_norm": 1.6921639926788343,
+      "learning_rate": 8.495908725036818e-07,
+      "loss": 0.0615,
+      "step": 8427
+    },
+    {
+      "epoch": 1.9154545454545455,
+      "grad_norm": 2.2149674448914283,
+      "learning_rate": 8.495075818953668e-07,
+      "loss": 0.0722,
+      "step": 8428
+    },
+    {
+      "epoch": 1.915681818181818,
+      "grad_norm": 1.5868886321623155,
+      "learning_rate": 8.494242867089505e-07,
+      "loss": 0.0509,
+      "step": 8429
+    },
+    {
+      "epoch": 1.915909090909091,
+      "grad_norm": 1.8455811307596584,
+      "learning_rate": 8.493409869461313e-07,
+      "loss": 0.0452,
+      "step": 8430
+    },
+    {
+      "epoch": 1.9161363636363635,
+      "grad_norm": 1.4560785148142192,
+      "learning_rate": 8.492576826086083e-07,
+      "loss": 0.049,
+      "step": 8431
+    },
+    {
+      "epoch": 1.9163636363636365,
+      "grad_norm": 1.6215057874883383,
+      "learning_rate": 8.491743736980795e-07,
+      "loss": 0.042,
+      "step": 8432
+    },
+    {
+      "epoch": 1.916590909090909,
+      "grad_norm": 2.0968881965323347,
+      "learning_rate": 8.490910602162446e-07,
+      "loss": 0.0462,
+      "step": 8433
+    },
+    {
+      "epoch": 1.916818181818182,
+      "grad_norm": 1.3505977427493807,
+      "learning_rate": 8.490077421648018e-07,
+      "loss": 0.0368,
+      "step": 8434
+    },
+    {
+      "epoch": 1.9170454545454545,
+      "grad_norm": 2.5748278382178293,
+      "learning_rate": 8.489244195454504e-07,
+      "loss": 0.0613,
+      "step": 8435
+    },
+    {
+      "epoch": 1.9172727272727272,
+      "grad_norm": 1.6730513394942241,
+      "learning_rate": 8.488410923598893e-07,
+      "loss": 0.0274,
+      "step": 8436
+    },
+    {
+      "epoch": 1.9175,
+      "grad_norm": 2.148207797980926,
+      "learning_rate": 8.487577606098178e-07,
+      "loss": 0.049,
+      "step": 8437
+    },
+    {
+      "epoch": 1.9177272727272727,
+      "grad_norm": 2.315080040132232,
+      "learning_rate": 8.486744242969352e-07,
+      "loss": 0.0407,
+      "step": 8438
+    },
+    {
+      "epoch": 1.9179545454545455,
+      "grad_norm": 2.1114645481848426,
+      "learning_rate": 8.485910834229411e-07,
+      "loss": 0.0616,
+      "step": 8439
+    },
+    {
+      "epoch": 1.9181818181818182,
+      "grad_norm": 2.763033396649998,
+      "learning_rate": 8.485077379895346e-07,
+      "loss": 0.0943,
+      "step": 8440
+    },
+    {
+      "epoch": 1.918409090909091,
+      "grad_norm": 1.7402269434610242,
+      "learning_rate": 8.484243879984154e-07,
+      "loss": 0.036,
+      "step": 8441
+    },
+    {
+      "epoch": 1.9186363636363635,
+      "grad_norm": 2.104643895485957,
+      "learning_rate": 8.483410334512831e-07,
+      "loss": 0.0455,
+      "step": 8442
+    },
+    {
+      "epoch": 1.9188636363636364,
+      "grad_norm": 2.4043621266716024,
+      "learning_rate": 8.482576743498375e-07,
+      "loss": 0.0604,
+      "step": 8443
+    },
+    {
+      "epoch": 1.919090909090909,
+      "grad_norm": 1.7657723566395112,
+      "learning_rate": 8.481743106957785e-07,
+      "loss": 0.0444,
+      "step": 8444
+    },
+    {
+      "epoch": 1.919318181818182,
+      "grad_norm": 2.51365364450366,
+      "learning_rate": 8.480909424908059e-07,
+      "loss": 0.0901,
+      "step": 8445
+    },
+    {
+      "epoch": 1.9195454545454544,
+      "grad_norm": 1.0055451823390824,
+      "learning_rate": 8.480075697366197e-07,
+      "loss": 0.0272,
+      "step": 8446
+    },
+    {
+      "epoch": 1.9197727272727274,
+      "grad_norm": 2.229240349796799,
+      "learning_rate": 8.4792419243492e-07,
+      "loss": 0.0666,
+      "step": 8447
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.430663336698422,
+      "learning_rate": 8.478408105874075e-07,
+      "loss": 0.0289,
+      "step": 8448
+    },
+    {
+      "epoch": 1.920227272727273,
+      "grad_norm": 2.5454951494411393,
+      "learning_rate": 8.477574241957817e-07,
+      "loss": 0.0634,
+      "step": 8449
+    },
+    {
+      "epoch": 1.9204545454545454,
+      "grad_norm": 2.3045859243463296,
+      "learning_rate": 8.476740332617438e-07,
+      "loss": 0.0399,
+      "step": 8450
+    },
+    {
+      "epoch": 1.9206818181818182,
+      "grad_norm": 1.8419184618063815,
+      "learning_rate": 8.475906377869936e-07,
+      "loss": 0.0324,
+      "step": 8451
+    },
+    {
+      "epoch": 1.920909090909091,
+      "grad_norm": 1.9451195915084836,
+      "learning_rate": 8.475072377732318e-07,
+      "loss": 0.0547,
+      "step": 8452
+    },
+    {
+      "epoch": 1.9211363636363636,
+      "grad_norm": 2.2171805525304946,
+      "learning_rate": 8.474238332221596e-07,
+      "loss": 0.0852,
+      "step": 8453
+    },
+    {
+      "epoch": 1.9213636363636364,
+      "grad_norm": 2.479771658401893,
+      "learning_rate": 8.473404241354771e-07,
+      "loss": 0.0648,
+      "step": 8454
+    },
+    {
+      "epoch": 1.9215909090909091,
+      "grad_norm": 1.9601025455854142,
+      "learning_rate": 8.472570105148858e-07,
+      "loss": 0.0682,
+      "step": 8455
+    },
+    {
+      "epoch": 1.9218181818181819,
+      "grad_norm": 2.1718774205617293,
+      "learning_rate": 8.471735923620861e-07,
+      "loss": 0.0682,
+      "step": 8456
+    },
+    {
+      "epoch": 1.9220454545454544,
+      "grad_norm": 1.1928705434858187,
+      "learning_rate": 8.470901696787793e-07,
+      "loss": 0.0139,
+      "step": 8457
+    },
+    {
+      "epoch": 1.9222727272727274,
+      "grad_norm": 2.5597712562020547,
+      "learning_rate": 8.470067424666664e-07,
+      "loss": 0.0527,
+      "step": 8458
+    },
+    {
+      "epoch": 1.9224999999999999,
+      "grad_norm": 2.0609138970578695,
+      "learning_rate": 8.469233107274488e-07,
+      "loss": 0.0455,
+      "step": 8459
+    },
+    {
+      "epoch": 1.9227272727272728,
+      "grad_norm": 1.938559222184419,
+      "learning_rate": 8.468398744628277e-07,
+      "loss": 0.0398,
+      "step": 8460
+    },
+    {
+      "epoch": 1.9229545454545454,
+      "grad_norm": 0.7980466981189329,
+      "learning_rate": 8.467564336745046e-07,
+      "loss": 0.0166,
+      "step": 8461
+    },
+    {
+      "epoch": 1.9231818181818183,
+      "grad_norm": 1.7184808575964532,
+      "learning_rate": 8.466729883641808e-07,
+      "loss": 0.0363,
+      "step": 8462
+    },
+    {
+      "epoch": 1.9234090909090908,
+      "grad_norm": 1.9558304783468503,
+      "learning_rate": 8.465895385335582e-07,
+      "loss": 0.0359,
+      "step": 8463
+    },
+    {
+      "epoch": 1.9236363636363636,
+      "grad_norm": 2.4913703385548125,
+      "learning_rate": 8.465060841843381e-07,
+      "loss": 0.0639,
+      "step": 8464
+    },
+    {
+      "epoch": 1.9238636363636363,
+      "grad_norm": 1.8699692136989936,
+      "learning_rate": 8.464226253182227e-07,
+      "loss": 0.0591,
+      "step": 8465
+    },
+    {
+      "epoch": 1.924090909090909,
+      "grad_norm": 4.153768300050856,
+      "learning_rate": 8.463391619369136e-07,
+      "loss": 0.0658,
+      "step": 8466
+    },
+    {
+      "epoch": 1.9243181818181818,
+      "grad_norm": 2.3986738729438213,
+      "learning_rate": 8.46255694042113e-07,
+      "loss": 0.0418,
+      "step": 8467
+    },
+    {
+      "epoch": 1.9245454545454546,
+      "grad_norm": 1.1784203269331133,
+      "learning_rate": 8.461722216355225e-07,
+      "loss": 0.0178,
+      "step": 8468
+    },
+    {
+      "epoch": 1.9247727272727273,
+      "grad_norm": 1.8875549963180733,
+      "learning_rate": 8.460887447188445e-07,
+      "loss": 0.0441,
+      "step": 8469
+    },
+    {
+      "epoch": 1.925,
+      "grad_norm": 1.865225989351236,
+      "learning_rate": 8.460052632937818e-07,
+      "loss": 0.0368,
+      "step": 8470
+    },
+    {
+      "epoch": 1.9252272727272728,
+      "grad_norm": 1.4596200835550577,
+      "learning_rate": 8.459217773620358e-07,
+      "loss": 0.0285,
+      "step": 8471
+    },
+    {
+      "epoch": 1.9254545454545453,
+      "grad_norm": 1.7550000583349146,
+      "learning_rate": 8.458382869253096e-07,
+      "loss": 0.0459,
+      "step": 8472
+    },
+    {
+      "epoch": 1.9256818181818183,
+      "grad_norm": 1.5801261350170883,
+      "learning_rate": 8.457547919853052e-07,
+      "loss": 0.0349,
+      "step": 8473
+    },
+    {
+      "epoch": 1.9259090909090908,
+      "grad_norm": 1.3071425280958981,
+      "learning_rate": 8.456712925437256e-07,
+      "loss": 0.0261,
+      "step": 8474
+    },
+    {
+      "epoch": 1.9261363636363638,
+      "grad_norm": 1.6201870904564784,
+      "learning_rate": 8.455877886022733e-07,
+      "loss": 0.0633,
+      "step": 8475
+    },
+    {
+      "epoch": 1.9263636363636363,
+      "grad_norm": 2.091549483906913,
+      "learning_rate": 8.45504280162651e-07,
+      "loss": 0.0457,
+      "step": 8476
+    },
+    {
+      "epoch": 1.9265909090909092,
+      "grad_norm": 2.099780268229586,
+      "learning_rate": 8.45420767226562e-07,
+      "loss": 0.0511,
+      "step": 8477
+    },
+    {
+      "epoch": 1.9268181818181818,
+      "grad_norm": 2.8415831179323505,
+      "learning_rate": 8.453372497957088e-07,
+      "loss": 0.0737,
+      "step": 8478
+    },
+    {
+      "epoch": 1.9270454545454545,
+      "grad_norm": 1.6578895554359387,
+      "learning_rate": 8.452537278717949e-07,
+      "loss": 0.0611,
+      "step": 8479
+    },
+    {
+      "epoch": 1.9272727272727272,
+      "grad_norm": 3.141403801693656,
+      "learning_rate": 8.451702014565231e-07,
+      "loss": 0.0643,
+      "step": 8480
+    },
+    {
+      "epoch": 1.9275,
+      "grad_norm": 2.5755900849537667,
+      "learning_rate": 8.45086670551597e-07,
+      "loss": 0.0982,
+      "step": 8481
+    },
+    {
+      "epoch": 1.9277272727272727,
+      "grad_norm": 1.556694996758955,
+      "learning_rate": 8.450031351587196e-07,
+      "loss": 0.041,
+      "step": 8482
+    },
+    {
+      "epoch": 1.9279545454545455,
+      "grad_norm": 2.9978450470898683,
+      "learning_rate": 8.449195952795942e-07,
+      "loss": 0.0641,
+      "step": 8483
+    },
+    {
+      "epoch": 1.9281818181818182,
+      "grad_norm": 2.201849684890536,
+      "learning_rate": 8.448360509159248e-07,
+      "loss": 0.0643,
+      "step": 8484
+    },
+    {
+      "epoch": 1.928409090909091,
+      "grad_norm": 3.4499382172224826,
+      "learning_rate": 8.447525020694149e-07,
+      "loss": 0.075,
+      "step": 8485
+    },
+    {
+      "epoch": 1.9286363636363637,
+      "grad_norm": 1.6861381027391058,
+      "learning_rate": 8.44668948741768e-07,
+      "loss": 0.0434,
+      "step": 8486
+    },
+    {
+      "epoch": 1.9288636363636362,
+      "grad_norm": 1.7674163859758831,
+      "learning_rate": 8.445853909346879e-07,
+      "loss": 0.0508,
+      "step": 8487
+    },
+    {
+      "epoch": 1.9290909090909092,
+      "grad_norm": 5.702881610101776,
+      "learning_rate": 8.445018286498788e-07,
+      "loss": 0.1131,
+      "step": 8488
+    },
+    {
+      "epoch": 1.9293181818181817,
+      "grad_norm": 1.8538782684170443,
+      "learning_rate": 8.444182618890443e-07,
+      "loss": 0.0467,
+      "step": 8489
+    },
+    {
+      "epoch": 1.9295454545454547,
+      "grad_norm": 1.6715984463254758,
+      "learning_rate": 8.443346906538885e-07,
+      "loss": 0.0465,
+      "step": 8490
+    },
+    {
+      "epoch": 1.9297727272727272,
+      "grad_norm": 1.5676436065272432,
+      "learning_rate": 8.442511149461159e-07,
+      "loss": 0.0288,
+      "step": 8491
+    },
+    {
+      "epoch": 1.9300000000000002,
+      "grad_norm": 2.359509342484606,
+      "learning_rate": 8.441675347674307e-07,
+      "loss": 0.0647,
+      "step": 8492
+    },
+    {
+      "epoch": 1.9302272727272727,
+      "grad_norm": 1.473042174758475,
+      "learning_rate": 8.440839501195369e-07,
+      "loss": 0.0531,
+      "step": 8493
+    },
+    {
+      "epoch": 1.9304545454545454,
+      "grad_norm": 1.3513029493439879,
+      "learning_rate": 8.440003610041394e-07,
+      "loss": 0.0406,
+      "step": 8494
+    },
+    {
+      "epoch": 1.9306818181818182,
+      "grad_norm": 1.7641773351784698,
+      "learning_rate": 8.439167674229422e-07,
+      "loss": 0.0514,
+      "step": 8495
+    },
+    {
+      "epoch": 1.930909090909091,
+      "grad_norm": 2.464940353494242,
+      "learning_rate": 8.438331693776504e-07,
+      "loss": 0.0699,
+      "step": 8496
+    },
+    {
+      "epoch": 1.9311363636363637,
+      "grad_norm": 1.3161541258983513,
+      "learning_rate": 8.437495668699683e-07,
+      "loss": 0.0221,
+      "step": 8497
+    },
+    {
+      "epoch": 1.9313636363636364,
+      "grad_norm": 1.016985069649912,
+      "learning_rate": 8.436659599016009e-07,
+      "loss": 0.0388,
+      "step": 8498
+    },
+    {
+      "epoch": 1.9315909090909091,
+      "grad_norm": 2.6528017438714686,
+      "learning_rate": 8.435823484742532e-07,
+      "loss": 0.0681,
+      "step": 8499
+    },
+    {
+      "epoch": 1.9318181818181817,
+      "grad_norm": 2.1269456611268067,
+      "learning_rate": 8.4349873258963e-07,
+      "loss": 0.0825,
+      "step": 8500
+    },
+    {
+      "epoch": 1.9320454545454546,
+      "grad_norm": 2.031780173752752,
+      "learning_rate": 8.434151122494366e-07,
+      "loss": 0.0718,
+      "step": 8501
+    },
+    {
+      "epoch": 1.9322727272727271,
+      "grad_norm": 1.562884324455671,
+      "learning_rate": 8.433314874553778e-07,
+      "loss": 0.0304,
+      "step": 8502
+    },
+    {
+      "epoch": 1.9325,
+      "grad_norm": 1.883281960426498,
+      "learning_rate": 8.432478582091594e-07,
+      "loss": 0.0699,
+      "step": 8503
+    },
+    {
+      "epoch": 1.9327272727272726,
+      "grad_norm": 1.005475492605776,
+      "learning_rate": 8.43164224512486e-07,
+      "loss": 0.0325,
+      "step": 8504
+    },
+    {
+      "epoch": 1.9329545454545456,
+      "grad_norm": 2.000490796640838,
+      "learning_rate": 8.430805863670636e-07,
+      "loss": 0.0303,
+      "step": 8505
+    },
+    {
+      "epoch": 1.9331818181818181,
+      "grad_norm": 1.749242117552764,
+      "learning_rate": 8.429969437745977e-07,
+      "loss": 0.0422,
+      "step": 8506
+    },
+    {
+      "epoch": 1.9334090909090909,
+      "grad_norm": 2.6289415679542882,
+      "learning_rate": 8.429132967367938e-07,
+      "loss": 0.0744,
+      "step": 8507
+    },
+    {
+      "epoch": 1.9336363636363636,
+      "grad_norm": 1.7544860277190786,
+      "learning_rate": 8.428296452553575e-07,
+      "loss": 0.0617,
+      "step": 8508
+    },
+    {
+      "epoch": 1.9338636363636363,
+      "grad_norm": 1.891943167681188,
+      "learning_rate": 8.427459893319947e-07,
+      "loss": 0.0355,
+      "step": 8509
+    },
+    {
+      "epoch": 1.934090909090909,
+      "grad_norm": 3.4110276100395653,
+      "learning_rate": 8.426623289684114e-07,
+      "loss": 0.0913,
+      "step": 8510
+    },
+    {
+      "epoch": 1.9343181818181818,
+      "grad_norm": 1.7107786678252135,
+      "learning_rate": 8.425786641663135e-07,
+      "loss": 0.0154,
+      "step": 8511
+    },
+    {
+      "epoch": 1.9345454545454546,
+      "grad_norm": 2.416481757331445,
+      "learning_rate": 8.424949949274071e-07,
+      "loss": 0.058,
+      "step": 8512
+    },
+    {
+      "epoch": 1.9347727272727273,
+      "grad_norm": 2.7891984478006733,
+      "learning_rate": 8.424113212533981e-07,
+      "loss": 0.0703,
+      "step": 8513
+    },
+    {
+      "epoch": 1.935,
+      "grad_norm": 2.5702670380468757,
+      "learning_rate": 8.423276431459931e-07,
+      "loss": 0.0493,
+      "step": 8514
+    },
+    {
+      "epoch": 1.9352272727272726,
+      "grad_norm": 1.730528248506474,
+      "learning_rate": 8.422439606068981e-07,
+      "loss": 0.06,
+      "step": 8515
+    },
+    {
+      "epoch": 1.9354545454545455,
+      "grad_norm": 2.0269984115990773,
+      "learning_rate": 8.421602736378201e-07,
+      "loss": 0.0491,
+      "step": 8516
+    },
+    {
+      "epoch": 1.935681818181818,
+      "grad_norm": 2.045669595775401,
+      "learning_rate": 8.420765822404651e-07,
+      "loss": 0.0492,
+      "step": 8517
+    },
+    {
+      "epoch": 1.935909090909091,
+      "grad_norm": 2.6332555781168305,
+      "learning_rate": 8.419928864165399e-07,
+      "loss": 0.0611,
+      "step": 8518
+    },
+    {
+      "epoch": 1.9361363636363635,
+      "grad_norm": 1.7620089625165722,
+      "learning_rate": 8.419091861677512e-07,
+      "loss": 0.0635,
+      "step": 8519
+    },
+    {
+      "epoch": 1.9363636363636365,
+      "grad_norm": 2.3758710479385945,
+      "learning_rate": 8.418254814958055e-07,
+      "loss": 0.0482,
+      "step": 8520
+    },
+    {
+      "epoch": 1.936590909090909,
+      "grad_norm": 1.3747845845302173,
+      "learning_rate": 8.417417724024101e-07,
+      "loss": 0.0222,
+      "step": 8521
+    },
+    {
+      "epoch": 1.9368181818181818,
+      "grad_norm": 1.671290097126096,
+      "learning_rate": 8.41658058889272e-07,
+      "loss": 0.0556,
+      "step": 8522
+    },
+    {
+      "epoch": 1.9370454545454545,
+      "grad_norm": 1.945452625557999,
+      "learning_rate": 8.41574340958098e-07,
+      "loss": 0.0698,
+      "step": 8523
+    },
+    {
+      "epoch": 1.9372727272727273,
+      "grad_norm": 3.2259778477974517,
+      "learning_rate": 8.414906186105953e-07,
+      "loss": 0.0611,
+      "step": 8524
+    },
+    {
+      "epoch": 1.9375,
+      "grad_norm": 3.262490420829914,
+      "learning_rate": 8.414068918484713e-07,
+      "loss": 0.0601,
+      "step": 8525
+    },
+    {
+      "epoch": 1.9377272727272727,
+      "grad_norm": 1.4042689678657398,
+      "learning_rate": 8.413231606734332e-07,
+      "loss": 0.0253,
+      "step": 8526
+    },
+    {
+      "epoch": 1.9379545454545455,
+      "grad_norm": 1.3175419283387337,
+      "learning_rate": 8.412394250871886e-07,
+      "loss": 0.0308,
+      "step": 8527
+    },
+    {
+      "epoch": 1.9381818181818182,
+      "grad_norm": 1.663016475606808,
+      "learning_rate": 8.411556850914446e-07,
+      "loss": 0.0327,
+      "step": 8528
+    },
+    {
+      "epoch": 1.938409090909091,
+      "grad_norm": 2.3847269328971423,
+      "learning_rate": 8.410719406879092e-07,
+      "loss": 0.0992,
+      "step": 8529
+    },
+    {
+      "epoch": 1.9386363636363635,
+      "grad_norm": 3.091644008888873,
+      "learning_rate": 8.409881918782899e-07,
+      "loss": 0.0879,
+      "step": 8530
+    },
+    {
+      "epoch": 1.9388636363636365,
+      "grad_norm": 1.8797249895407997,
+      "learning_rate": 8.409044386642945e-07,
+      "loss": 0.056,
+      "step": 8531
+    },
+    {
+      "epoch": 1.939090909090909,
+      "grad_norm": 2.334481838004716,
+      "learning_rate": 8.408206810476312e-07,
+      "loss": 0.0541,
+      "step": 8532
+    },
+    {
+      "epoch": 1.939318181818182,
+      "grad_norm": 2.1890812177665184,
+      "learning_rate": 8.407369190300075e-07,
+      "loss": 0.0759,
+      "step": 8533
+    },
+    {
+      "epoch": 1.9395454545454545,
+      "grad_norm": 2.304715776068879,
+      "learning_rate": 8.406531526131317e-07,
+      "loss": 0.0506,
+      "step": 8534
+    },
+    {
+      "epoch": 1.9397727272727274,
+      "grad_norm": 2.0067588773696245,
+      "learning_rate": 8.405693817987119e-07,
+      "loss": 0.0763,
+      "step": 8535
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.3026379714266623,
+      "learning_rate": 8.404856065884564e-07,
+      "loss": 0.029,
+      "step": 8536
+    },
+    {
+      "epoch": 1.9402272727272727,
+      "grad_norm": 1.2128177224042247,
+      "learning_rate": 8.404018269840731e-07,
+      "loss": 0.0189,
+      "step": 8537
+    },
+    {
+      "epoch": 1.9404545454545454,
+      "grad_norm": 1.965527148533003,
+      "learning_rate": 8.403180429872713e-07,
+      "loss": 0.0668,
+      "step": 8538
+    },
+    {
+      "epoch": 1.9406818181818182,
+      "grad_norm": 1.764215591897118,
+      "learning_rate": 8.402342545997586e-07,
+      "loss": 0.0552,
+      "step": 8539
+    },
+    {
+      "epoch": 1.940909090909091,
+      "grad_norm": 2.3746862267197963,
+      "learning_rate": 8.401504618232439e-07,
+      "loss": 0.0517,
+      "step": 8540
+    },
+    {
+      "epoch": 1.9411363636363637,
+      "grad_norm": 1.8724458651016196,
+      "learning_rate": 8.400666646594362e-07,
+      "loss": 0.076,
+      "step": 8541
+    },
+    {
+      "epoch": 1.9413636363636364,
+      "grad_norm": 1.4200194635043795,
+      "learning_rate": 8.399828631100438e-07,
+      "loss": 0.0388,
+      "step": 8542
+    },
+    {
+      "epoch": 1.9415909090909091,
+      "grad_norm": 1.9439483546483187,
+      "learning_rate": 8.398990571767758e-07,
+      "loss": 0.0345,
+      "step": 8543
+    },
+    {
+      "epoch": 1.9418181818181819,
+      "grad_norm": 1.574692968098551,
+      "learning_rate": 8.398152468613412e-07,
+      "loss": 0.0355,
+      "step": 8544
+    },
+    {
+      "epoch": 1.9420454545454544,
+      "grad_norm": 2.0372770485741656,
+      "learning_rate": 8.397314321654487e-07,
+      "loss": 0.0385,
+      "step": 8545
+    },
+    {
+      "epoch": 1.9422727272727274,
+      "grad_norm": 1.9483378434170946,
+      "learning_rate": 8.39647613090808e-07,
+      "loss": 0.0452,
+      "step": 8546
+    },
+    {
+      "epoch": 1.9425,
+      "grad_norm": 2.133617000677784,
+      "learning_rate": 8.395637896391279e-07,
+      "loss": 0.0558,
+      "step": 8547
+    },
+    {
+      "epoch": 1.9427272727272729,
+      "grad_norm": 1.3562217568327133,
+      "learning_rate": 8.394799618121177e-07,
+      "loss": 0.0351,
+      "step": 8548
+    },
+    {
+      "epoch": 1.9429545454545454,
+      "grad_norm": 3.473905862323096,
+      "learning_rate": 8.393961296114871e-07,
+      "loss": 0.0659,
+      "step": 8549
+    },
+    {
+      "epoch": 1.9431818181818183,
+      "grad_norm": 2.8819164880299035,
+      "learning_rate": 8.393122930389451e-07,
+      "loss": 0.0977,
+      "step": 8550
+    },
+    {
+      "epoch": 1.9434090909090909,
+      "grad_norm": 2.133713019933783,
+      "learning_rate": 8.392284520962017e-07,
+      "loss": 0.0388,
+      "step": 8551
+    },
+    {
+      "epoch": 1.9436363636363636,
+      "grad_norm": 1.44123100704359,
+      "learning_rate": 8.391446067849664e-07,
+      "loss": 0.0284,
+      "step": 8552
+    },
+    {
+      "epoch": 1.9438636363636363,
+      "grad_norm": 2.184294242851373,
+      "learning_rate": 8.390607571069489e-07,
+      "loss": 0.1101,
+      "step": 8553
+    },
+    {
+      "epoch": 1.944090909090909,
+      "grad_norm": 1.9918563208830702,
+      "learning_rate": 8.389769030638592e-07,
+      "loss": 0.0657,
+      "step": 8554
+    },
+    {
+      "epoch": 1.9443181818181818,
+      "grad_norm": 1.976173765687963,
+      "learning_rate": 8.388930446574071e-07,
+      "loss": 0.0525,
+      "step": 8555
+    },
+    {
+      "epoch": 1.9445454545454546,
+      "grad_norm": 2.613407675877445,
+      "learning_rate": 8.388091818893029e-07,
+      "loss": 0.039,
+      "step": 8556
+    },
+    {
+      "epoch": 1.9447727272727273,
+      "grad_norm": 3.1741655019004176,
+      "learning_rate": 8.387253147612562e-07,
+      "loss": 0.0569,
+      "step": 8557
+    },
+    {
+      "epoch": 1.9449999999999998,
+      "grad_norm": 1.8384827452184722,
+      "learning_rate": 8.386414432749776e-07,
+      "loss": 0.0538,
+      "step": 8558
+    },
+    {
+      "epoch": 1.9452272727272728,
+      "grad_norm": 1.8877128578711035,
+      "learning_rate": 8.385575674321772e-07,
+      "loss": 0.0605,
+      "step": 8559
+    },
+    {
+      "epoch": 1.9454545454545453,
+      "grad_norm": 2.446880161885613,
+      "learning_rate": 8.384736872345654e-07,
+      "loss": 0.0483,
+      "step": 8560
+    },
+    {
+      "epoch": 1.9456818181818183,
+      "grad_norm": 1.5105150045412834,
+      "learning_rate": 8.383898026838527e-07,
+      "loss": 0.0432,
+      "step": 8561
+    },
+    {
+      "epoch": 1.9459090909090908,
+      "grad_norm": 2.4067291729697295,
+      "learning_rate": 8.383059137817498e-07,
+      "loss": 0.0642,
+      "step": 8562
+    },
+    {
+      "epoch": 1.9461363636363638,
+      "grad_norm": 1.649163933573683,
+      "learning_rate": 8.382220205299669e-07,
+      "loss": 0.0337,
+      "step": 8563
+    },
+    {
+      "epoch": 1.9463636363636363,
+      "grad_norm": 2.1444198132374557,
+      "learning_rate": 8.381381229302154e-07,
+      "loss": 0.0518,
+      "step": 8564
+    },
+    {
+      "epoch": 1.946590909090909,
+      "grad_norm": 1.106549659223853,
+      "learning_rate": 8.380542209842053e-07,
+      "loss": 0.0159,
+      "step": 8565
+    },
+    {
+      "epoch": 1.9468181818181818,
+      "grad_norm": 1.5029439977594667,
+      "learning_rate": 8.379703146936482e-07,
+      "loss": 0.0451,
+      "step": 8566
+    },
+    {
+      "epoch": 1.9470454545454545,
+      "grad_norm": 2.4230171889743684,
+      "learning_rate": 8.378864040602547e-07,
+      "loss": 0.0752,
+      "step": 8567
+    },
+    {
+      "epoch": 1.9472727272727273,
+      "grad_norm": 2.739063624120671,
+      "learning_rate": 8.37802489085736e-07,
+      "loss": 0.0425,
+      "step": 8568
+    },
+    {
+      "epoch": 1.9475,
+      "grad_norm": 2.6061761478011585,
+      "learning_rate": 8.377185697718035e-07,
+      "loss": 0.0906,
+      "step": 8569
+    },
+    {
+      "epoch": 1.9477272727272728,
+      "grad_norm": 0.9005890342414239,
+      "learning_rate": 8.376346461201682e-07,
+      "loss": 0.0248,
+      "step": 8570
+    },
+    {
+      "epoch": 1.9479545454545455,
+      "grad_norm": 1.7838948687365808,
+      "learning_rate": 8.375507181325415e-07,
+      "loss": 0.0544,
+      "step": 8571
+    },
+    {
+      "epoch": 1.9481818181818182,
+      "grad_norm": 1.7506785880660465,
+      "learning_rate": 8.374667858106347e-07,
+      "loss": 0.0408,
+      "step": 8572
+    },
+    {
+      "epoch": 1.9484090909090908,
+      "grad_norm": 2.1779334819774148,
+      "learning_rate": 8.373828491561596e-07,
+      "loss": 0.0289,
+      "step": 8573
+    },
+    {
+      "epoch": 1.9486363636363637,
+      "grad_norm": 10.244687971624,
+      "learning_rate": 8.372989081708276e-07,
+      "loss": 0.225,
+      "step": 8574
+    },
+    {
+      "epoch": 1.9488636363636362,
+      "grad_norm": 1.2802348534380472,
+      "learning_rate": 8.372149628563506e-07,
+      "loss": 0.0294,
+      "step": 8575
+    },
+    {
+      "epoch": 1.9490909090909092,
+      "grad_norm": 1.8712210181971716,
+      "learning_rate": 8.371310132144401e-07,
+      "loss": 0.0425,
+      "step": 8576
+    },
+    {
+      "epoch": 1.9493181818181817,
+      "grad_norm": 1.8983698479595252,
+      "learning_rate": 8.370470592468083e-07,
+      "loss": 0.0458,
+      "step": 8577
+    },
+    {
+      "epoch": 1.9495454545454547,
+      "grad_norm": 2.8491142435008414,
+      "learning_rate": 8.36963100955167e-07,
+      "loss": 0.089,
+      "step": 8578
+    },
+    {
+      "epoch": 1.9497727272727272,
+      "grad_norm": 2.843444240343028,
+      "learning_rate": 8.368791383412282e-07,
+      "loss": 0.0685,
+      "step": 8579
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 2.0610814301863667,
+      "learning_rate": 8.367951714067044e-07,
+      "loss": 0.0523,
+      "step": 8580
+    },
+    {
+      "epoch": 1.9502272727272727,
+      "grad_norm": 3.075944796189844,
+      "learning_rate": 8.367112001533073e-07,
+      "loss": 0.0853,
+      "step": 8581
+    },
+    {
+      "epoch": 1.9504545454545454,
+      "grad_norm": 2.223305995172519,
+      "learning_rate": 8.366272245827495e-07,
+      "loss": 0.0522,
+      "step": 8582
+    },
+    {
+      "epoch": 1.9506818181818182,
+      "grad_norm": 2.194528315155391,
+      "learning_rate": 8.365432446967434e-07,
+      "loss": 0.0626,
+      "step": 8583
+    },
+    {
+      "epoch": 1.950909090909091,
+      "grad_norm": 1.4640271372720306,
+      "learning_rate": 8.364592604970015e-07,
+      "loss": 0.0385,
+      "step": 8584
+    },
+    {
+      "epoch": 1.9511363636363637,
+      "grad_norm": 1.9679015634604617,
+      "learning_rate": 8.363752719852364e-07,
+      "loss": 0.03,
+      "step": 8585
+    },
+    {
+      "epoch": 1.9513636363636364,
+      "grad_norm": 0.8513523504580124,
+      "learning_rate": 8.362912791631607e-07,
+      "loss": 0.0141,
+      "step": 8586
+    },
+    {
+      "epoch": 1.9515909090909092,
+      "grad_norm": 1.4605368275740336,
+      "learning_rate": 8.362072820324872e-07,
+      "loss": 0.0291,
+      "step": 8587
+    },
+    {
+      "epoch": 1.9518181818181817,
+      "grad_norm": 1.952993251705836,
+      "learning_rate": 8.361232805949287e-07,
+      "loss": 0.0526,
+      "step": 8588
+    },
+    {
+      "epoch": 1.9520454545454546,
+      "grad_norm": 0.5921362821182155,
+      "learning_rate": 8.360392748521982e-07,
+      "loss": 0.0232,
+      "step": 8589
+    },
+    {
+      "epoch": 1.9522727272727272,
+      "grad_norm": 4.169368269345234,
+      "learning_rate": 8.359552648060088e-07,
+      "loss": 0.1113,
+      "step": 8590
+    },
+    {
+      "epoch": 1.9525000000000001,
+      "grad_norm": 1.3362019281446398,
+      "learning_rate": 8.358712504580735e-07,
+      "loss": 0.0488,
+      "step": 8591
+    },
+    {
+      "epoch": 1.9527272727272726,
+      "grad_norm": 1.9915263788164532,
+      "learning_rate": 8.357872318101054e-07,
+      "loss": 0.0703,
+      "step": 8592
+    },
+    {
+      "epoch": 1.9529545454545456,
+      "grad_norm": 1.9871324989812926,
+      "learning_rate": 8.357032088638182e-07,
+      "loss": 0.0887,
+      "step": 8593
+    },
+    {
+      "epoch": 1.9531818181818181,
+      "grad_norm": 2.214493808307907,
+      "learning_rate": 8.356191816209247e-07,
+      "loss": 0.0434,
+      "step": 8594
+    },
+    {
+      "epoch": 1.9534090909090909,
+      "grad_norm": 2.6098815542429006,
+      "learning_rate": 8.355351500831388e-07,
+      "loss": 0.0686,
+      "step": 8595
+    },
+    {
+      "epoch": 1.9536363636363636,
+      "grad_norm": 2.1159674820617105,
+      "learning_rate": 8.354511142521738e-07,
+      "loss": 0.052,
+      "step": 8596
+    },
+    {
+      "epoch": 1.9538636363636364,
+      "grad_norm": 2.40786662322715,
+      "learning_rate": 8.353670741297433e-07,
+      "loss": 0.0525,
+      "step": 8597
+    },
+    {
+      "epoch": 1.954090909090909,
+      "grad_norm": 1.8352304487246403,
+      "learning_rate": 8.352830297175612e-07,
+      "loss": 0.0691,
+      "step": 8598
+    },
+    {
+      "epoch": 1.9543181818181818,
+      "grad_norm": 0.8297767672984012,
+      "learning_rate": 8.351989810173414e-07,
+      "loss": 0.0189,
+      "step": 8599
+    },
+    {
+      "epoch": 1.9545454545454546,
+      "grad_norm": 1.307653576733176,
+      "learning_rate": 8.351149280307977e-07,
+      "loss": 0.0415,
+      "step": 8600
+    },
+    {
+      "epoch": 1.954772727272727,
+      "grad_norm": 1.7698734541612955,
+      "learning_rate": 8.350308707596438e-07,
+      "loss": 0.0532,
+      "step": 8601
+    },
+    {
+      "epoch": 1.955,
+      "grad_norm": 2.0667658744692323,
+      "learning_rate": 8.349468092055943e-07,
+      "loss": 0.0489,
+      "step": 8602
+    },
+    {
+      "epoch": 1.9552272727272726,
+      "grad_norm": 2.0023324428410647,
+      "learning_rate": 8.348627433703629e-07,
+      "loss": 0.0657,
+      "step": 8603
+    },
+    {
+      "epoch": 1.9554545454545456,
+      "grad_norm": 2.0316756759222576,
+      "learning_rate": 8.347786732556643e-07,
+      "loss": 0.0348,
+      "step": 8604
+    },
+    {
+      "epoch": 1.955681818181818,
+      "grad_norm": 2.69228628860869,
+      "learning_rate": 8.346945988632123e-07,
+      "loss": 0.0445,
+      "step": 8605
+    },
+    {
+      "epoch": 1.955909090909091,
+      "grad_norm": 2.5795732780005936,
+      "learning_rate": 8.346105201947217e-07,
+      "loss": 0.0559,
+      "step": 8606
+    },
+    {
+      "epoch": 1.9561363636363636,
+      "grad_norm": 1.5197767304729204,
+      "learning_rate": 8.345264372519068e-07,
+      "loss": 0.0376,
+      "step": 8607
+    },
+    {
+      "epoch": 1.9563636363636365,
+      "grad_norm": 3.8154376667578456,
+      "learning_rate": 8.344423500364826e-07,
+      "loss": 0.0571,
+      "step": 8608
+    },
+    {
+      "epoch": 1.956590909090909,
+      "grad_norm": 2.7333613448477574,
+      "learning_rate": 8.343582585501632e-07,
+      "loss": 0.0633,
+      "step": 8609
+    },
+    {
+      "epoch": 1.9568181818181818,
+      "grad_norm": 3.0236659379382727,
+      "learning_rate": 8.342741627946637e-07,
+      "loss": 0.1148,
+      "step": 8610
+    },
+    {
+      "epoch": 1.9570454545454545,
+      "grad_norm": 2.0647448937081974,
+      "learning_rate": 8.34190062771699e-07,
+      "loss": 0.0576,
+      "step": 8611
+    },
+    {
+      "epoch": 1.9572727272727273,
+      "grad_norm": 1.6800868041242216,
+      "learning_rate": 8.34105958482984e-07,
+      "loss": 0.0228,
+      "step": 8612
+    },
+    {
+      "epoch": 1.9575,
+      "grad_norm": 2.8492058883566966,
+      "learning_rate": 8.340218499302335e-07,
+      "loss": 0.0763,
+      "step": 8613
+    },
+    {
+      "epoch": 1.9577272727272728,
+      "grad_norm": 0.7828676633521581,
+      "learning_rate": 8.33937737115163e-07,
+      "loss": 0.0273,
+      "step": 8614
+    },
+    {
+      "epoch": 1.9579545454545455,
+      "grad_norm": 0.6918433329406963,
+      "learning_rate": 8.338536200394876e-07,
+      "loss": 0.019,
+      "step": 8615
+    },
+    {
+      "epoch": 1.958181818181818,
+      "grad_norm": 1.9437033225750657,
+      "learning_rate": 8.337694987049225e-07,
+      "loss": 0.0488,
+      "step": 8616
+    },
+    {
+      "epoch": 1.958409090909091,
+      "grad_norm": 2.739060152880892,
+      "learning_rate": 8.33685373113183e-07,
+      "loss": 0.0839,
+      "step": 8617
+    },
+    {
+      "epoch": 1.9586363636363635,
+      "grad_norm": 2.0797868324353725,
+      "learning_rate": 8.336012432659846e-07,
+      "loss": 0.0314,
+      "step": 8618
+    },
+    {
+      "epoch": 1.9588636363636365,
+      "grad_norm": 1.1601597159070347,
+      "learning_rate": 8.335171091650431e-07,
+      "loss": 0.0271,
+      "step": 8619
+    },
+    {
+      "epoch": 1.959090909090909,
+      "grad_norm": 1.0130745722693182,
+      "learning_rate": 8.334329708120738e-07,
+      "loss": 0.0354,
+      "step": 8620
+    },
+    {
+      "epoch": 1.959318181818182,
+      "grad_norm": 2.4853845915413637,
+      "learning_rate": 8.333488282087926e-07,
+      "loss": 0.0684,
+      "step": 8621
+    },
+    {
+      "epoch": 1.9595454545454545,
+      "grad_norm": 2.0814233738634194,
+      "learning_rate": 8.332646813569155e-07,
+      "loss": 0.0423,
+      "step": 8622
+    },
+    {
+      "epoch": 1.9597727272727272,
+      "grad_norm": 1.9972174064800392,
+      "learning_rate": 8.331805302581581e-07,
+      "loss": 0.0589,
+      "step": 8623
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.770574840216226,
+      "learning_rate": 8.330963749142365e-07,
+      "loss": 0.0299,
+      "step": 8624
+    },
+    {
+      "epoch": 1.9602272727272727,
+      "grad_norm": 1.5224257342115308,
+      "learning_rate": 8.330122153268668e-07,
+      "loss": 0.041,
+      "step": 8625
+    },
+    {
+      "epoch": 1.9604545454545454,
+      "grad_norm": 1.8772754575548842,
+      "learning_rate": 8.329280514977652e-07,
+      "loss": 0.0469,
+      "step": 8626
+    },
+    {
+      "epoch": 1.9606818181818182,
+      "grad_norm": 1.9314438389298587,
+      "learning_rate": 8.328438834286476e-07,
+      "loss": 0.0306,
+      "step": 8627
+    },
+    {
+      "epoch": 1.960909090909091,
+      "grad_norm": 1.7474741597798473,
+      "learning_rate": 8.327597111212308e-07,
+      "loss": 0.0275,
+      "step": 8628
+    },
+    {
+      "epoch": 1.9611363636363637,
+      "grad_norm": 1.2824214337217361,
+      "learning_rate": 8.32675534577231e-07,
+      "loss": 0.0277,
+      "step": 8629
+    },
+    {
+      "epoch": 1.9613636363636364,
+      "grad_norm": 2.68938825954371,
+      "learning_rate": 8.325913537983648e-07,
+      "loss": 0.0585,
+      "step": 8630
+    },
+    {
+      "epoch": 1.961590909090909,
+      "grad_norm": 0.9111202283940623,
+      "learning_rate": 8.325071687863487e-07,
+      "loss": 0.0366,
+      "step": 8631
+    },
+    {
+      "epoch": 1.961818181818182,
+      "grad_norm": 1.7131712822204825,
+      "learning_rate": 8.324229795428994e-07,
+      "loss": 0.0359,
+      "step": 8632
+    },
+    {
+      "epoch": 1.9620454545454544,
+      "grad_norm": 0.45297661441800297,
+      "learning_rate": 8.323387860697338e-07,
+      "loss": 0.0154,
+      "step": 8633
+    },
+    {
+      "epoch": 1.9622727272727274,
+      "grad_norm": 1.6033505429590804,
+      "learning_rate": 8.322545883685685e-07,
+      "loss": 0.0379,
+      "step": 8634
+    },
+    {
+      "epoch": 1.9625,
+      "grad_norm": 1.1817577148966671,
+      "learning_rate": 8.321703864411205e-07,
+      "loss": 0.0292,
+      "step": 8635
+    },
+    {
+      "epoch": 1.9627272727272729,
+      "grad_norm": 1.8414383090058324,
+      "learning_rate": 8.320861802891068e-07,
+      "loss": 0.0589,
+      "step": 8636
+    },
+    {
+      "epoch": 1.9629545454545454,
+      "grad_norm": 2.8183994340030503,
+      "learning_rate": 8.32001969914245e-07,
+      "loss": 0.0683,
+      "step": 8637
+    },
+    {
+      "epoch": 1.9631818181818181,
+      "grad_norm": 2.614753185267649,
+      "learning_rate": 8.319177553182515e-07,
+      "loss": 0.0669,
+      "step": 8638
+    },
+    {
+      "epoch": 1.9634090909090909,
+      "grad_norm": 2.230619966909436,
+      "learning_rate": 8.318335365028443e-07,
+      "loss": 0.1141,
+      "step": 8639
+    },
+    {
+      "epoch": 1.9636363636363636,
+      "grad_norm": 1.31346370417812,
+      "learning_rate": 8.317493134697403e-07,
+      "loss": 0.0428,
+      "step": 8640
+    },
+    {
+      "epoch": 1.9638636363636364,
+      "grad_norm": 1.3417497021504632,
+      "learning_rate": 8.316650862206572e-07,
+      "loss": 0.0382,
+      "step": 8641
+    },
+    {
+      "epoch": 1.964090909090909,
+      "grad_norm": 1.0966070881861423,
+      "learning_rate": 8.315808547573123e-07,
+      "loss": 0.0339,
+      "step": 8642
+    },
+    {
+      "epoch": 1.9643181818181819,
+      "grad_norm": 1.4647371734375148,
+      "learning_rate": 8.314966190814233e-07,
+      "loss": 0.0259,
+      "step": 8643
+    },
+    {
+      "epoch": 1.9645454545454546,
+      "grad_norm": 2.4605773218779077,
+      "learning_rate": 8.314123791947082e-07,
+      "loss": 0.0964,
+      "step": 8644
+    },
+    {
+      "epoch": 1.9647727272727273,
+      "grad_norm": 2.038278500067985,
+      "learning_rate": 8.313281350988844e-07,
+      "loss": 0.0728,
+      "step": 8645
+    },
+    {
+      "epoch": 1.9649999999999999,
+      "grad_norm": 1.4116163285185672,
+      "learning_rate": 8.312438867956702e-07,
+      "loss": 0.0178,
+      "step": 8646
+    },
+    {
+      "epoch": 1.9652272727272728,
+      "grad_norm": 1.7081281528690533,
+      "learning_rate": 8.311596342867831e-07,
+      "loss": 0.0313,
+      "step": 8647
+    },
+    {
+      "epoch": 1.9654545454545453,
+      "grad_norm": 1.3109461109023066,
+      "learning_rate": 8.310753775739416e-07,
+      "loss": 0.0281,
+      "step": 8648
+    },
+    {
+      "epoch": 1.9656818181818183,
+      "grad_norm": 1.5589721133667471,
+      "learning_rate": 8.309911166588635e-07,
+      "loss": 0.0331,
+      "step": 8649
+    },
+    {
+      "epoch": 1.9659090909090908,
+      "grad_norm": 2.635171646668848,
+      "learning_rate": 8.309068515432671e-07,
+      "loss": 0.0613,
+      "step": 8650
+    },
+    {
+      "epoch": 1.9661363636363638,
+      "grad_norm": 1.7465915169925073,
+      "learning_rate": 8.308225822288709e-07,
+      "loss": 0.0281,
+      "step": 8651
+    },
+    {
+      "epoch": 1.9663636363636363,
+      "grad_norm": 2.129911441954809,
+      "learning_rate": 8.307383087173932e-07,
+      "loss": 0.0558,
+      "step": 8652
+    },
+    {
+      "epoch": 1.966590909090909,
+      "grad_norm": 2.373592027289962,
+      "learning_rate": 8.306540310105524e-07,
+      "loss": 0.0538,
+      "step": 8653
+    },
+    {
+      "epoch": 1.9668181818181818,
+      "grad_norm": 1.5873988886654453,
+      "learning_rate": 8.30569749110067e-07,
+      "loss": 0.0271,
+      "step": 8654
+    },
+    {
+      "epoch": 1.9670454545454545,
+      "grad_norm": 2.467746999425034,
+      "learning_rate": 8.304854630176559e-07,
+      "loss": 0.0541,
+      "step": 8655
+    },
+    {
+      "epoch": 1.9672727272727273,
+      "grad_norm": 2.5748403515444123,
+      "learning_rate": 8.304011727350377e-07,
+      "loss": 0.114,
+      "step": 8656
+    },
+    {
+      "epoch": 1.9675,
+      "grad_norm": 1.7846645139963844,
+      "learning_rate": 8.303168782639314e-07,
+      "loss": 0.0664,
+      "step": 8657
+    },
+    {
+      "epoch": 1.9677272727272728,
+      "grad_norm": 2.4064039968711706,
+      "learning_rate": 8.302325796060554e-07,
+      "loss": 0.0453,
+      "step": 8658
+    },
+    {
+      "epoch": 1.9679545454545453,
+      "grad_norm": 1.7848309735714918,
+      "learning_rate": 8.301482767631293e-07,
+      "loss": 0.0414,
+      "step": 8659
+    },
+    {
+      "epoch": 1.9681818181818183,
+      "grad_norm": 2.52038968259213,
+      "learning_rate": 8.300639697368719e-07,
+      "loss": 0.124,
+      "step": 8660
+    },
+    {
+      "epoch": 1.9684090909090908,
+      "grad_norm": 1.5850407837865357,
+      "learning_rate": 8.299796585290024e-07,
+      "loss": 0.0285,
+      "step": 8661
+    },
+    {
+      "epoch": 1.9686363636363637,
+      "grad_norm": 1.908559230299604,
+      "learning_rate": 8.298953431412401e-07,
+      "loss": 0.0401,
+      "step": 8662
+    },
+    {
+      "epoch": 1.9688636363636363,
+      "grad_norm": 0.5401232945992206,
+      "learning_rate": 8.298110235753043e-07,
+      "loss": 0.0213,
+      "step": 8663
+    },
+    {
+      "epoch": 1.9690909090909092,
+      "grad_norm": 1.129439896690663,
+      "learning_rate": 8.297266998329142e-07,
+      "loss": 0.0188,
+      "step": 8664
+    },
+    {
+      "epoch": 1.9693181818181817,
+      "grad_norm": 1.5627845322222578,
+      "learning_rate": 8.296423719157898e-07,
+      "loss": 0.0596,
+      "step": 8665
+    },
+    {
+      "epoch": 1.9695454545454547,
+      "grad_norm": 1.8335074611440407,
+      "learning_rate": 8.295580398256503e-07,
+      "loss": 0.0396,
+      "step": 8666
+    },
+    {
+      "epoch": 1.9697727272727272,
+      "grad_norm": 1.7384925971792413,
+      "learning_rate": 8.294737035642155e-07,
+      "loss": 0.0494,
+      "step": 8667
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 2.6608431083627475,
+      "learning_rate": 8.293893631332052e-07,
+      "loss": 0.0601,
+      "step": 8668
+    },
+    {
+      "epoch": 1.9702272727272727,
+      "grad_norm": 2.635778940448981,
+      "learning_rate": 8.29305018534339e-07,
+      "loss": 0.0679,
+      "step": 8669
+    },
+    {
+      "epoch": 1.9704545454545455,
+      "grad_norm": 1.8827776134347307,
+      "learning_rate": 8.292206697693374e-07,
+      "loss": 0.0378,
+      "step": 8670
+    },
+    {
+      "epoch": 1.9706818181818182,
+      "grad_norm": 1.8607058273933548,
+      "learning_rate": 8.291363168399198e-07,
+      "loss": 0.0506,
+      "step": 8671
+    },
+    {
+      "epoch": 1.970909090909091,
+      "grad_norm": 1.7125721581914697,
+      "learning_rate": 8.290519597478068e-07,
+      "loss": 0.0276,
+      "step": 8672
+    },
+    {
+      "epoch": 1.9711363636363637,
+      "grad_norm": 3.065670833043553,
+      "learning_rate": 8.289675984947181e-07,
+      "loss": 0.0556,
+      "step": 8673
+    },
+    {
+      "epoch": 1.9713636363636362,
+      "grad_norm": 2.153483211620786,
+      "learning_rate": 8.288832330823742e-07,
+      "loss": 0.0721,
+      "step": 8674
+    },
+    {
+      "epoch": 1.9715909090909092,
+      "grad_norm": 1.7073218370686833,
+      "learning_rate": 8.287988635124956e-07,
+      "loss": 0.0469,
+      "step": 8675
+    },
+    {
+      "epoch": 1.9718181818181817,
+      "grad_norm": 1.883855455086181,
+      "learning_rate": 8.287144897868027e-07,
+      "loss": 0.0385,
+      "step": 8676
+    },
+    {
+      "epoch": 1.9720454545454547,
+      "grad_norm": 2.3168482498589165,
+      "learning_rate": 8.28630111907016e-07,
+      "loss": 0.0529,
+      "step": 8677
+    },
+    {
+      "epoch": 1.9722727272727272,
+      "grad_norm": 1.7901050942442338,
+      "learning_rate": 8.285457298748559e-07,
+      "loss": 0.031,
+      "step": 8678
+    },
+    {
+      "epoch": 1.9725000000000001,
+      "grad_norm": 1.5454398826836493,
+      "learning_rate": 8.284613436920433e-07,
+      "loss": 0.0326,
+      "step": 8679
+    },
+    {
+      "epoch": 1.9727272727272727,
+      "grad_norm": 2.0040272560391887,
+      "learning_rate": 8.283769533602989e-07,
+      "loss": 0.0659,
+      "step": 8680
+    },
+    {
+      "epoch": 1.9729545454545454,
+      "grad_norm": 2.2789662104438904,
+      "learning_rate": 8.282925588813437e-07,
+      "loss": 0.03,
+      "step": 8681
+    },
+    {
+      "epoch": 1.9731818181818181,
+      "grad_norm": 2.024025760462103,
+      "learning_rate": 8.282081602568986e-07,
+      "loss": 0.0576,
+      "step": 8682
+    },
+    {
+      "epoch": 1.973409090909091,
+      "grad_norm": 0.7139080822566534,
+      "learning_rate": 8.281237574886845e-07,
+      "loss": 0.0267,
+      "step": 8683
+    },
+    {
+      "epoch": 1.9736363636363636,
+      "grad_norm": 2.4352323014402044,
+      "learning_rate": 8.280393505784226e-07,
+      "loss": 0.0455,
+      "step": 8684
+    },
+    {
+      "epoch": 1.9738636363636364,
+      "grad_norm": 1.291923330572298,
+      "learning_rate": 8.279549395278343e-07,
+      "loss": 0.0264,
+      "step": 8685
+    },
+    {
+      "epoch": 1.9740909090909091,
+      "grad_norm": 2.5434353507984486,
+      "learning_rate": 8.278705243386404e-07,
+      "loss": 0.0429,
+      "step": 8686
+    },
+    {
+      "epoch": 1.9743181818181819,
+      "grad_norm": 3.263825571616958,
+      "learning_rate": 8.277861050125628e-07,
+      "loss": 0.0719,
+      "step": 8687
+    },
+    {
+      "epoch": 1.9745454545454546,
+      "grad_norm": 1.7212804338757035,
+      "learning_rate": 8.277016815513226e-07,
+      "loss": 0.0383,
+      "step": 8688
+    },
+    {
+      "epoch": 1.9747727272727271,
+      "grad_norm": 1.4445036502841477,
+      "learning_rate": 8.276172539566416e-07,
+      "loss": 0.0348,
+      "step": 8689
+    },
+    {
+      "epoch": 1.975,
+      "grad_norm": 2.362991237873058,
+      "learning_rate": 8.275328222302414e-07,
+      "loss": 0.0622,
+      "step": 8690
+    },
+    {
+      "epoch": 1.9752272727272726,
+      "grad_norm": 0.790337505604815,
+      "learning_rate": 8.274483863738435e-07,
+      "loss": 0.0285,
+      "step": 8691
+    },
+    {
+      "epoch": 1.9754545454545456,
+      "grad_norm": 2.2131500967498674,
+      "learning_rate": 8.273639463891699e-07,
+      "loss": 0.0752,
+      "step": 8692
+    },
+    {
+      "epoch": 1.975681818181818,
+      "grad_norm": 2.4862019536172832,
+      "learning_rate": 8.272795022779422e-07,
+      "loss": 0.0705,
+      "step": 8693
+    },
+    {
+      "epoch": 1.975909090909091,
+      "grad_norm": 2.2240316457616585,
+      "learning_rate": 8.271950540418828e-07,
+      "loss": 0.0647,
+      "step": 8694
+    },
+    {
+      "epoch": 1.9761363636363636,
+      "grad_norm": 0.5900486982674069,
+      "learning_rate": 8.271106016827132e-07,
+      "loss": 0.0198,
+      "step": 8695
+    },
+    {
+      "epoch": 1.9763636363636363,
+      "grad_norm": 2.895627882182519,
+      "learning_rate": 8.270261452021559e-07,
+      "loss": 0.0747,
+      "step": 8696
+    },
+    {
+      "epoch": 1.976590909090909,
+      "grad_norm": 2.219156987105715,
+      "learning_rate": 8.269416846019332e-07,
+      "loss": 0.0505,
+      "step": 8697
+    },
+    {
+      "epoch": 1.9768181818181818,
+      "grad_norm": 2.9591757374838603,
+      "learning_rate": 8.268572198837673e-07,
+      "loss": 0.0594,
+      "step": 8698
+    },
+    {
+      "epoch": 1.9770454545454546,
+      "grad_norm": 1.8187916751564694,
+      "learning_rate": 8.267727510493803e-07,
+      "loss": 0.0682,
+      "step": 8699
+    },
+    {
+      "epoch": 1.9772727272727273,
+      "grad_norm": 2.5053116912859514,
+      "learning_rate": 8.26688278100495e-07,
+      "loss": 0.0527,
+      "step": 8700
+    },
+    {
+      "epoch": 1.9775,
+      "grad_norm": 2.5570979058889565,
+      "learning_rate": 8.266038010388339e-07,
+      "loss": 0.088,
+      "step": 8701
+    },
+    {
+      "epoch": 1.9777272727272728,
+      "grad_norm": 1.7560170276627918,
+      "learning_rate": 8.265193198661195e-07,
+      "loss": 0.0435,
+      "step": 8702
+    },
+    {
+      "epoch": 1.9779545454545455,
+      "grad_norm": 2.237100735479056,
+      "learning_rate": 8.264348345840745e-07,
+      "loss": 0.0668,
+      "step": 8703
+    },
+    {
+      "epoch": 1.978181818181818,
+      "grad_norm": 1.5915644715761759,
+      "learning_rate": 8.263503451944218e-07,
+      "loss": 0.0396,
+      "step": 8704
+    },
+    {
+      "epoch": 1.978409090909091,
+      "grad_norm": 2.8740484047155475,
+      "learning_rate": 8.262658516988846e-07,
+      "loss": 0.0647,
+      "step": 8705
+    },
+    {
+      "epoch": 1.9786363636363635,
+      "grad_norm": 1.245191504486044,
+      "learning_rate": 8.261813540991853e-07,
+      "loss": 0.0144,
+      "step": 8706
+    },
+    {
+      "epoch": 1.9788636363636365,
+      "grad_norm": 2.131970722801339,
+      "learning_rate": 8.260968523970474e-07,
+      "loss": 0.0414,
+      "step": 8707
+    },
+    {
+      "epoch": 1.979090909090909,
+      "grad_norm": 2.3987471424652265,
+      "learning_rate": 8.260123465941935e-07,
+      "loss": 0.0606,
+      "step": 8708
+    },
+    {
+      "epoch": 1.979318181818182,
+      "grad_norm": 1.8997526988883378,
+      "learning_rate": 8.259278366923477e-07,
+      "loss": 0.062,
+      "step": 8709
+    },
+    {
+      "epoch": 1.9795454545454545,
+      "grad_norm": 1.016876362301328,
+      "learning_rate": 8.258433226932323e-07,
+      "loss": 0.0183,
+      "step": 8710
+    },
+    {
+      "epoch": 1.9797727272727272,
+      "grad_norm": 2.4917677799109113,
+      "learning_rate": 8.257588045985712e-07,
+      "loss": 0.0347,
+      "step": 8711
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 2.311773951681663,
+      "learning_rate": 8.25674282410088e-07,
+      "loss": 0.1049,
+      "step": 8712
+    },
+    {
+      "epoch": 1.9802272727272727,
+      "grad_norm": 2.039211485747705,
+      "learning_rate": 8.255897561295059e-07,
+      "loss": 0.0606,
+      "step": 8713
+    },
+    {
+      "epoch": 1.9804545454545455,
+      "grad_norm": 2.1623149227184433,
+      "learning_rate": 8.25505225758549e-07,
+      "loss": 0.0408,
+      "step": 8714
+    },
+    {
+      "epoch": 1.9806818181818182,
+      "grad_norm": 2.8220607681433165,
+      "learning_rate": 8.254206912989405e-07,
+      "loss": 0.0582,
+      "step": 8715
+    },
+    {
+      "epoch": 1.980909090909091,
+      "grad_norm": 0.9445100911294585,
+      "learning_rate": 8.253361527524047e-07,
+      "loss": 0.0245,
+      "step": 8716
+    },
+    {
+      "epoch": 1.9811363636363635,
+      "grad_norm": 1.4267048872598347,
+      "learning_rate": 8.252516101206651e-07,
+      "loss": 0.046,
+      "step": 8717
+    },
+    {
+      "epoch": 1.9813636363636364,
+      "grad_norm": 1.6794596811254268,
+      "learning_rate": 8.251670634054457e-07,
+      "loss": 0.0286,
+      "step": 8718
+    },
+    {
+      "epoch": 1.981590909090909,
+      "grad_norm": 1.727901946118178,
+      "learning_rate": 8.250825126084708e-07,
+      "loss": 0.0357,
+      "step": 8719
+    },
+    {
+      "epoch": 1.981818181818182,
+      "grad_norm": 2.5996935091473254,
+      "learning_rate": 8.249979577314644e-07,
+      "loss": 0.0361,
+      "step": 8720
+    },
+    {
+      "epoch": 1.9820454545454544,
+      "grad_norm": 1.9649498555813165,
+      "learning_rate": 8.249133987761507e-07,
+      "loss": 0.0357,
+      "step": 8721
+    },
+    {
+      "epoch": 1.9822727272727274,
+      "grad_norm": 0.8461492386466994,
+      "learning_rate": 8.24828835744254e-07,
+      "loss": 0.0222,
+      "step": 8722
+    },
+    {
+      "epoch": 1.9825,
+      "grad_norm": 1.7362916596511984,
+      "learning_rate": 8.247442686374987e-07,
+      "loss": 0.0484,
+      "step": 8723
+    },
+    {
+      "epoch": 1.982727272727273,
+      "grad_norm": 1.775455451821214,
+      "learning_rate": 8.246596974576095e-07,
+      "loss": 0.0331,
+      "step": 8724
+    },
+    {
+      "epoch": 1.9829545454545454,
+      "grad_norm": 2.4862309736463026,
+      "learning_rate": 8.245751222063106e-07,
+      "loss": 0.0732,
+      "step": 8725
+    },
+    {
+      "epoch": 1.9831818181818182,
+      "grad_norm": 2.0699632870302915,
+      "learning_rate": 8.244905428853267e-07,
+      "loss": 0.0661,
+      "step": 8726
+    },
+    {
+      "epoch": 1.983409090909091,
+      "grad_norm": 2.8745107980127567,
+      "learning_rate": 8.244059594963827e-07,
+      "loss": 0.0511,
+      "step": 8727
+    },
+    {
+      "epoch": 1.9836363636363636,
+      "grad_norm": 2.1776426266657576,
+      "learning_rate": 8.243213720412031e-07,
+      "loss": 0.0535,
+      "step": 8728
+    },
+    {
+      "epoch": 1.9838636363636364,
+      "grad_norm": 0.9212171399017491,
+      "learning_rate": 8.242367805215133e-07,
+      "loss": 0.0127,
+      "step": 8729
+    },
+    {
+      "epoch": 1.9840909090909091,
+      "grad_norm": 1.1960600942751947,
+      "learning_rate": 8.241521849390377e-07,
+      "loss": 0.0213,
+      "step": 8730
+    },
+    {
+      "epoch": 1.9843181818181819,
+      "grad_norm": 2.0777429392136515,
+      "learning_rate": 8.240675852955019e-07,
+      "loss": 0.0469,
+      "step": 8731
+    },
+    {
+      "epoch": 1.9845454545454544,
+      "grad_norm": 2.2778422726531904,
+      "learning_rate": 8.239829815926305e-07,
+      "loss": 0.0517,
+      "step": 8732
+    },
+    {
+      "epoch": 1.9847727272727274,
+      "grad_norm": 3.06794804333878,
+      "learning_rate": 8.238983738321491e-07,
+      "loss": 0.0571,
+      "step": 8733
+    },
+    {
+      "epoch": 1.9849999999999999,
+      "grad_norm": 1.7934294522621639,
+      "learning_rate": 8.238137620157827e-07,
+      "loss": 0.0574,
+      "step": 8734
+    },
+    {
+      "epoch": 1.9852272727272728,
+      "grad_norm": 2.3264339861634253,
+      "learning_rate": 8.237291461452568e-07,
+      "loss": 0.0541,
+      "step": 8735
+    },
+    {
+      "epoch": 1.9854545454545454,
+      "grad_norm": 2.064919584056717,
+      "learning_rate": 8.236445262222972e-07,
+      "loss": 0.0485,
+      "step": 8736
+    },
+    {
+      "epoch": 1.9856818181818183,
+      "grad_norm": 1.5258855648011806,
+      "learning_rate": 8.235599022486289e-07,
+      "loss": 0.042,
+      "step": 8737
+    },
+    {
+      "epoch": 1.9859090909090908,
+      "grad_norm": 1.8839930676397276,
+      "learning_rate": 8.23475274225978e-07,
+      "loss": 0.0302,
+      "step": 8738
+    },
+    {
+      "epoch": 1.9861363636363636,
+      "grad_norm": 5.541671697317397,
+      "learning_rate": 8.233906421560699e-07,
+      "loss": 0.0547,
+      "step": 8739
+    },
+    {
+      "epoch": 1.9863636363636363,
+      "grad_norm": 1.1706971541153537,
+      "learning_rate": 8.233060060406305e-07,
+      "loss": 0.0319,
+      "step": 8740
+    },
+    {
+      "epoch": 1.986590909090909,
+      "grad_norm": 2.2071707789692083,
+      "learning_rate": 8.232213658813856e-07,
+      "loss": 0.0534,
+      "step": 8741
+    },
+    {
+      "epoch": 1.9868181818181818,
+      "grad_norm": 1.362872721011993,
+      "learning_rate": 8.231367216800612e-07,
+      "loss": 0.028,
+      "step": 8742
+    },
+    {
+      "epoch": 1.9870454545454546,
+      "grad_norm": 1.8106905855184288,
+      "learning_rate": 8.230520734383835e-07,
+      "loss": 0.0751,
+      "step": 8743
+    },
+    {
+      "epoch": 1.9872727272727273,
+      "grad_norm": 1.2294096789627724,
+      "learning_rate": 8.229674211580785e-07,
+      "loss": 0.0463,
+      "step": 8744
+    },
+    {
+      "epoch": 1.9875,
+      "grad_norm": 1.30581459520116,
+      "learning_rate": 8.228827648408724e-07,
+      "loss": 0.0348,
+      "step": 8745
+    },
+    {
+      "epoch": 1.9877272727272728,
+      "grad_norm": 1.8350142316518787,
+      "learning_rate": 8.227981044884914e-07,
+      "loss": 0.037,
+      "step": 8746
+    },
+    {
+      "epoch": 1.9879545454545453,
+      "grad_norm": 1.6285016165849195,
+      "learning_rate": 8.227134401026621e-07,
+      "loss": 0.0513,
+      "step": 8747
+    },
+    {
+      "epoch": 1.9881818181818183,
+      "grad_norm": 1.8910596027343785,
+      "learning_rate": 8.226287716851107e-07,
+      "loss": 0.0587,
+      "step": 8748
+    },
+    {
+      "epoch": 1.9884090909090908,
+      "grad_norm": 0.8297264028021923,
+      "learning_rate": 8.22544099237564e-07,
+      "loss": 0.0186,
+      "step": 8749
+    },
+    {
+      "epoch": 1.9886363636363638,
+      "grad_norm": 2.4526126367994276,
+      "learning_rate": 8.224594227617483e-07,
+      "loss": 0.0423,
+      "step": 8750
+    },
+    {
+      "epoch": 1.9888636363636363,
+      "grad_norm": 1.7113994683256857,
+      "learning_rate": 8.223747422593907e-07,
+      "loss": 0.0299,
+      "step": 8751
+    },
+    {
+      "epoch": 1.9890909090909092,
+      "grad_norm": 1.9326944729007134,
+      "learning_rate": 8.222900577322176e-07,
+      "loss": 0.0463,
+      "step": 8752
+    },
+    {
+      "epoch": 1.9893181818181818,
+      "grad_norm": 1.454410026310381,
+      "learning_rate": 8.222053691819562e-07,
+      "loss": 0.0226,
+      "step": 8753
+    },
+    {
+      "epoch": 1.9895454545454545,
+      "grad_norm": 1.3083358086713468,
+      "learning_rate": 8.221206766103332e-07,
+      "loss": 0.0305,
+      "step": 8754
+    },
+    {
+      "epoch": 1.9897727272727272,
+      "grad_norm": 1.6736458341437197,
+      "learning_rate": 8.220359800190755e-07,
+      "loss": 0.0682,
+      "step": 8755
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 2.148762744734874,
+      "learning_rate": 8.219512794099108e-07,
+      "loss": 0.059,
+      "step": 8756
+    },
+    {
+      "epoch": 1.9902272727272727,
+      "grad_norm": 1.9933058279217286,
+      "learning_rate": 8.218665747845655e-07,
+      "loss": 0.0665,
+      "step": 8757
+    },
+    {
+      "epoch": 1.9904545454545455,
+      "grad_norm": 1.2425840243619992,
+      "learning_rate": 8.217818661447675e-07,
+      "loss": 0.043,
+      "step": 8758
+    },
+    {
+      "epoch": 1.9906818181818182,
+      "grad_norm": 1.7651459450576843,
+      "learning_rate": 8.216971534922439e-07,
+      "loss": 0.0336,
+      "step": 8759
+    },
+    {
+      "epoch": 1.990909090909091,
+      "grad_norm": 1.3144204643216117,
+      "learning_rate": 8.216124368287223e-07,
+      "loss": 0.0694,
+      "step": 8760
+    },
+    {
+      "epoch": 1.9911363636363637,
+      "grad_norm": 2.2177501779872504,
+      "learning_rate": 8.215277161559301e-07,
+      "loss": 0.0483,
+      "step": 8761
+    },
+    {
+      "epoch": 1.9913636363636362,
+      "grad_norm": 2.2309212921892327,
+      "learning_rate": 8.214429914755949e-07,
+      "loss": 0.0734,
+      "step": 8762
+    },
+    {
+      "epoch": 1.9915909090909092,
+      "grad_norm": 1.797202703440963,
+      "learning_rate": 8.213582627894442e-07,
+      "loss": 0.0531,
+      "step": 8763
+    },
+    {
+      "epoch": 1.9918181818181817,
+      "grad_norm": 1.7890316580331702,
+      "learning_rate": 8.212735300992062e-07,
+      "loss": 0.0377,
+      "step": 8764
+    },
+    {
+      "epoch": 1.9920454545454547,
+      "grad_norm": 1.570326329753888,
+      "learning_rate": 8.211887934066083e-07,
+      "loss": 0.0414,
+      "step": 8765
+    },
+    {
+      "epoch": 1.9922727272727272,
+      "grad_norm": 1.7370158714308974,
+      "learning_rate": 8.211040527133787e-07,
+      "loss": 0.0456,
+      "step": 8766
+    },
+    {
+      "epoch": 1.9925000000000002,
+      "grad_norm": 2.194479992470471,
+      "learning_rate": 8.210193080212454e-07,
+      "loss": 0.0923,
+      "step": 8767
+    },
+    {
+      "epoch": 1.9927272727272727,
+      "grad_norm": 1.7363274020398745,
+      "learning_rate": 8.209345593319364e-07,
+      "loss": 0.0406,
+      "step": 8768
+    },
+    {
+      "epoch": 1.9929545454545454,
+      "grad_norm": 2.080654727046625,
+      "learning_rate": 8.2084980664718e-07,
+      "loss": 0.0694,
+      "step": 8769
+    },
+    {
+      "epoch": 1.9931818181818182,
+      "grad_norm": 2.115229263611499,
+      "learning_rate": 8.207650499687043e-07,
+      "loss": 0.0444,
+      "step": 8770
+    },
+    {
+      "epoch": 1.993409090909091,
+      "grad_norm": 2.1626130355713666,
+      "learning_rate": 8.206802892982376e-07,
+      "loss": 0.0497,
+      "step": 8771
+    },
+    {
+      "epoch": 1.9936363636363637,
+      "grad_norm": 1.2344958372757207,
+      "learning_rate": 8.205955246375083e-07,
+      "loss": 0.038,
+      "step": 8772
+    },
+    {
+      "epoch": 1.9938636363636364,
+      "grad_norm": 1.2021525527144783,
+      "learning_rate": 8.205107559882454e-07,
+      "loss": 0.0306,
+      "step": 8773
+    },
+    {
+      "epoch": 1.9940909090909091,
+      "grad_norm": 3.479760210891923,
+      "learning_rate": 8.204259833521769e-07,
+      "loss": 0.1088,
+      "step": 8774
+    },
+    {
+      "epoch": 1.9943181818181817,
+      "grad_norm": 2.1343099450078897,
+      "learning_rate": 8.203412067310317e-07,
+      "loss": 0.0622,
+      "step": 8775
+    },
+    {
+      "epoch": 1.9945454545454546,
+      "grad_norm": 1.7771584473725968,
+      "learning_rate": 8.202564261265385e-07,
+      "loss": 0.0326,
+      "step": 8776
+    },
+    {
+      "epoch": 1.9947727272727271,
+      "grad_norm": 2.1632256791207864,
+      "learning_rate": 8.201716415404259e-07,
+      "loss": 0.0634,
+      "step": 8777
+    },
+    {
+      "epoch": 1.995,
+      "grad_norm": 2.169906253096477,
+      "learning_rate": 8.200868529744234e-07,
+      "loss": 0.0519,
+      "step": 8778
+    },
+    {
+      "epoch": 1.9952272727272726,
+      "grad_norm": 1.1771961871541654,
+      "learning_rate": 8.200020604302594e-07,
+      "loss": 0.0302,
+      "step": 8779
+    },
+    {
+      "epoch": 1.9954545454545456,
+      "grad_norm": 1.5781771104987972,
+      "learning_rate": 8.199172639096632e-07,
+      "loss": 0.0545,
+      "step": 8780
+    },
+    {
+      "epoch": 1.9956818181818181,
+      "grad_norm": 3.3077430371689185,
+      "learning_rate": 8.19832463414364e-07,
+      "loss": 0.0646,
+      "step": 8781
+    },
+    {
+      "epoch": 1.9959090909090909,
+      "grad_norm": 1.9021928266666333,
+      "learning_rate": 8.19747658946091e-07,
+      "loss": 0.0361,
+      "step": 8782
+    },
+    {
+      "epoch": 1.9961363636363636,
+      "grad_norm": 2.2009002323185203,
+      "learning_rate": 8.196628505065733e-07,
+      "loss": 0.0479,
+      "step": 8783
+    },
+    {
+      "epoch": 1.9963636363636363,
+      "grad_norm": 1.3396806313702245,
+      "learning_rate": 8.195780380975407e-07,
+      "loss": 0.0359,
+      "step": 8784
+    },
+    {
+      "epoch": 1.996590909090909,
+      "grad_norm": 2.853219751348542,
+      "learning_rate": 8.194932217207223e-07,
+      "loss": 0.0817,
+      "step": 8785
+    },
+    {
+      "epoch": 1.9968181818181818,
+      "grad_norm": 1.907704529649111,
+      "learning_rate": 8.194084013778478e-07,
+      "loss": 0.0536,
+      "step": 8786
+    },
+    {
+      "epoch": 1.9970454545454546,
+      "grad_norm": 1.6568499059218196,
+      "learning_rate": 8.193235770706468e-07,
+      "loss": 0.0717,
+      "step": 8787
+    },
+    {
+      "epoch": 1.9972727272727273,
+      "grad_norm": 1.8448281328867744,
+      "learning_rate": 8.192387488008491e-07,
+      "loss": 0.0897,
+      "step": 8788
+    },
+    {
+      "epoch": 1.9975,
+      "grad_norm": 1.7779236316796818,
+      "learning_rate": 8.191539165701845e-07,
+      "loss": 0.0425,
+      "step": 8789
+    },
+    {
+      "epoch": 1.9977272727272726,
+      "grad_norm": 1.52296431013828,
+      "learning_rate": 8.190690803803828e-07,
+      "loss": 0.0349,
+      "step": 8790
+    },
+    {
+      "epoch": 1.9979545454545455,
+      "grad_norm": 2.007936796441165,
+      "learning_rate": 8.189842402331741e-07,
+      "loss": 0.0931,
+      "step": 8791
+    },
+    {
+      "epoch": 1.998181818181818,
+      "grad_norm": 1.7750712847838028,
+      "learning_rate": 8.18899396130288e-07,
+      "loss": 0.0467,
+      "step": 8792
+    },
+    {
+      "epoch": 1.998409090909091,
+      "grad_norm": 1.7785560062476349,
+      "learning_rate": 8.188145480734553e-07,
+      "loss": 0.0305,
+      "step": 8793
+    },
+    {
+      "epoch": 1.9986363636363635,
+      "grad_norm": 1.7817827645782018,
+      "learning_rate": 8.187296960644055e-07,
+      "loss": 0.0387,
+      "step": 8794
+    },
+    {
+      "epoch": 1.9988636363636365,
+      "grad_norm": 1.1995325801788492,
+      "learning_rate": 8.186448401048694e-07,
+      "loss": 0.0201,
+      "step": 8795
+    },
+    {
+      "epoch": 1.999090909090909,
+      "grad_norm": 2.3198064485501226,
+      "learning_rate": 8.185599801965772e-07,
+      "loss": 0.0456,
+      "step": 8796
+    },
+    {
+      "epoch": 1.9993181818181818,
+      "grad_norm": 1.2146136148248292,
+      "learning_rate": 8.184751163412592e-07,
+      "loss": 0.0344,
+      "step": 8797
+    },
+    {
+      "epoch": 1.9995454545454545,
+      "grad_norm": 1.6126600860479723,
+      "learning_rate": 8.18390248540646e-07,
+      "loss": 0.0479,
+      "step": 8798
+    },
+    {
+      "epoch": 1.9997727272727273,
+      "grad_norm": 1.748476692669491,
+      "learning_rate": 8.183053767964682e-07,
+      "loss": 0.0458,
+      "step": 8799
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.6529701310291995,
+      "learning_rate": 8.182205011104569e-07,
+      "loss": 0.0318,
+      "step": 8800
+    },
+    {
+      "epoch": 2.0002272727272725,
+      "grad_norm": 1.0642262143891825,
+      "learning_rate": 8.181356214843422e-07,
+      "loss": 0.024,
+      "step": 8801
+    },
+    {
+      "epoch": 2.0004545454545455,
+      "grad_norm": 1.179476610366187,
+      "learning_rate": 8.180507379198553e-07,
+      "loss": 0.0218,
+      "step": 8802
+    },
+    {
+      "epoch": 2.000681818181818,
+      "grad_norm": 1.5552713356769985,
+      "learning_rate": 8.17965850418727e-07,
+      "loss": 0.0407,
+      "step": 8803
+    },
+    {
+      "epoch": 2.000909090909091,
+      "grad_norm": 1.4354922396465075,
+      "learning_rate": 8.178809589826886e-07,
+      "loss": 0.026,
+      "step": 8804
+    },
+    {
+      "epoch": 2.0011363636363635,
+      "grad_norm": 1.6688477451615222,
+      "learning_rate": 8.177960636134706e-07,
+      "loss": 0.0313,
+      "step": 8805
+    },
+    {
+      "epoch": 2.0013636363636365,
+      "grad_norm": 1.346711017328622,
+      "learning_rate": 8.177111643128048e-07,
+      "loss": 0.0348,
+      "step": 8806
+    },
+    {
+      "epoch": 2.001590909090909,
+      "grad_norm": 1.4730411534745296,
+      "learning_rate": 8.17626261082422e-07,
+      "loss": 0.0368,
+      "step": 8807
+    },
+    {
+      "epoch": 2.001818181818182,
+      "grad_norm": 0.8570989132023453,
+      "learning_rate": 8.175413539240538e-07,
+      "loss": 0.018,
+      "step": 8808
+    },
+    {
+      "epoch": 2.0020454545454545,
+      "grad_norm": 0.8069690381100304,
+      "learning_rate": 8.174564428394313e-07,
+      "loss": 0.0111,
+      "step": 8809
+    },
+    {
+      "epoch": 2.0022727272727274,
+      "grad_norm": 1.5691099199285297,
+      "learning_rate": 8.173715278302863e-07,
+      "loss": 0.024,
+      "step": 8810
+    },
+    {
+      "epoch": 2.0025,
+      "grad_norm": 1.001249652705722,
+      "learning_rate": 8.172866088983501e-07,
+      "loss": 0.0227,
+      "step": 8811
+    },
+    {
+      "epoch": 2.002727272727273,
+      "grad_norm": 1.1377227051372358,
+      "learning_rate": 8.172016860453547e-07,
+      "loss": 0.0186,
+      "step": 8812
+    },
+    {
+      "epoch": 2.0029545454545454,
+      "grad_norm": 1.866873787506207,
+      "learning_rate": 8.171167592730315e-07,
+      "loss": 0.0209,
+      "step": 8813
+    },
+    {
+      "epoch": 2.0031818181818184,
+      "grad_norm": 1.0349670754684888,
+      "learning_rate": 8.170318285831123e-07,
+      "loss": 0.014,
+      "step": 8814
+    },
+    {
+      "epoch": 2.003409090909091,
+      "grad_norm": 1.5548658547832415,
+      "learning_rate": 8.169468939773294e-07,
+      "loss": 0.0212,
+      "step": 8815
+    },
+    {
+      "epoch": 2.0036363636363634,
+      "grad_norm": 2.057256169504386,
+      "learning_rate": 8.168619554574142e-07,
+      "loss": 0.0202,
+      "step": 8816
+    },
+    {
+      "epoch": 2.0038636363636364,
+      "grad_norm": 0.9096165988009467,
+      "learning_rate": 8.167770130250991e-07,
+      "loss": 0.0132,
+      "step": 8817
+    },
+    {
+      "epoch": 2.004090909090909,
+      "grad_norm": 1.6039109442341877,
+      "learning_rate": 8.16692066682116e-07,
+      "loss": 0.0212,
+      "step": 8818
+    },
+    {
+      "epoch": 2.004318181818182,
+      "grad_norm": 1.0308973261607612,
+      "learning_rate": 8.166071164301975e-07,
+      "loss": 0.0158,
+      "step": 8819
+    },
+    {
+      "epoch": 2.0045454545454544,
+      "grad_norm": 2.2214955982833287,
+      "learning_rate": 8.165221622710754e-07,
+      "loss": 0.0684,
+      "step": 8820
+    },
+    {
+      "epoch": 2.0047727272727274,
+      "grad_norm": 1.3651513589660331,
+      "learning_rate": 8.164372042064825e-07,
+      "loss": 0.0324,
+      "step": 8821
+    },
+    {
+      "epoch": 2.005,
+      "grad_norm": 0.9540266823301249,
+      "learning_rate": 8.163522422381507e-07,
+      "loss": 0.0175,
+      "step": 8822
+    },
+    {
+      "epoch": 2.005227272727273,
+      "grad_norm": 1.1887260145090845,
+      "learning_rate": 8.162672763678132e-07,
+      "loss": 0.0168,
+      "step": 8823
+    },
+    {
+      "epoch": 2.0054545454545454,
+      "grad_norm": 1.3026144623384117,
+      "learning_rate": 8.161823065972021e-07,
+      "loss": 0.0114,
+      "step": 8824
+    },
+    {
+      "epoch": 2.0056818181818183,
+      "grad_norm": 0.9652205937050906,
+      "learning_rate": 8.1609733292805e-07,
+      "loss": 0.008,
+      "step": 8825
+    },
+    {
+      "epoch": 2.005909090909091,
+      "grad_norm": 1.6267698718772405,
+      "learning_rate": 8.160123553620901e-07,
+      "loss": 0.0215,
+      "step": 8826
+    },
+    {
+      "epoch": 2.006136363636364,
+      "grad_norm": 2.7077820439431295,
+      "learning_rate": 8.159273739010551e-07,
+      "loss": 0.017,
+      "step": 8827
+    },
+    {
+      "epoch": 2.0063636363636363,
+      "grad_norm": 2.916046077456705,
+      "learning_rate": 8.158423885466777e-07,
+      "loss": 0.0388,
+      "step": 8828
+    },
+    {
+      "epoch": 2.006590909090909,
+      "grad_norm": 1.1596411407342335,
+      "learning_rate": 8.157573993006912e-07,
+      "loss": 0.0253,
+      "step": 8829
+    },
+    {
+      "epoch": 2.006818181818182,
+      "grad_norm": 1.717346487258726,
+      "learning_rate": 8.156724061648285e-07,
+      "loss": 0.0392,
+      "step": 8830
+    },
+    {
+      "epoch": 2.0070454545454544,
+      "grad_norm": 1.1263032786135667,
+      "learning_rate": 8.155874091408228e-07,
+      "loss": 0.0259,
+      "step": 8831
+    },
+    {
+      "epoch": 2.0072727272727273,
+      "grad_norm": 2.0205779890167825,
+      "learning_rate": 8.155024082304072e-07,
+      "loss": 0.0241,
+      "step": 8832
+    },
+    {
+      "epoch": 2.0075,
+      "grad_norm": 2.0768048159988615,
+      "learning_rate": 8.154174034353152e-07,
+      "loss": 0.021,
+      "step": 8833
+    },
+    {
+      "epoch": 2.007727272727273,
+      "grad_norm": 1.3783850564851525,
+      "learning_rate": 8.153323947572803e-07,
+      "loss": 0.0246,
+      "step": 8834
+    },
+    {
+      "epoch": 2.0079545454545453,
+      "grad_norm": 1.0961525764195479,
+      "learning_rate": 8.152473821980358e-07,
+      "loss": 0.007,
+      "step": 8835
+    },
+    {
+      "epoch": 2.0081818181818183,
+      "grad_norm": 3.743315251456867,
+      "learning_rate": 8.15162365759315e-07,
+      "loss": 0.0288,
+      "step": 8836
+    },
+    {
+      "epoch": 2.008409090909091,
+      "grad_norm": 2.0521411484145347,
+      "learning_rate": 8.15077345442852e-07,
+      "loss": 0.0174,
+      "step": 8837
+    },
+    {
+      "epoch": 2.0086363636363638,
+      "grad_norm": 3.2010915979398264,
+      "learning_rate": 8.149923212503803e-07,
+      "loss": 0.0173,
+      "step": 8838
+    },
+    {
+      "epoch": 2.0088636363636363,
+      "grad_norm": 4.085121779985305,
+      "learning_rate": 8.149072931836338e-07,
+      "loss": 0.069,
+      "step": 8839
+    },
+    {
+      "epoch": 2.0090909090909093,
+      "grad_norm": 2.084159750365483,
+      "learning_rate": 8.14822261244346e-07,
+      "loss": 0.0483,
+      "step": 8840
+    },
+    {
+      "epoch": 2.009318181818182,
+      "grad_norm": 1.9302699272402608,
+      "learning_rate": 8.147372254342514e-07,
+      "loss": 0.0481,
+      "step": 8841
+    },
+    {
+      "epoch": 2.0095454545454547,
+      "grad_norm": 1.6281574632786706,
+      "learning_rate": 8.146521857550838e-07,
+      "loss": 0.032,
+      "step": 8842
+    },
+    {
+      "epoch": 2.0097727272727273,
+      "grad_norm": 1.7983107432807255,
+      "learning_rate": 8.145671422085771e-07,
+      "loss": 0.0186,
+      "step": 8843
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.3611575631550414,
+      "learning_rate": 8.144820947964658e-07,
+      "loss": 0.0185,
+      "step": 8844
+    },
+    {
+      "epoch": 2.0102272727272728,
+      "grad_norm": 1.7388619997374173,
+      "learning_rate": 8.143970435204838e-07,
+      "loss": 0.0527,
+      "step": 8845
+    },
+    {
+      "epoch": 2.0104545454545453,
+      "grad_norm": 2.832968452517632,
+      "learning_rate": 8.143119883823658e-07,
+      "loss": 0.0415,
+      "step": 8846
+    },
+    {
+      "epoch": 2.0106818181818182,
+      "grad_norm": 1.608696032899749,
+      "learning_rate": 8.14226929383846e-07,
+      "loss": 0.03,
+      "step": 8847
+    },
+    {
+      "epoch": 2.0109090909090908,
+      "grad_norm": 1.9832218871606027,
+      "learning_rate": 8.141418665266591e-07,
+      "loss": 0.0157,
+      "step": 8848
+    },
+    {
+      "epoch": 2.0111363636363637,
+      "grad_norm": 0.9730618025070376,
+      "learning_rate": 8.140567998125395e-07,
+      "loss": 0.0223,
+      "step": 8849
+    },
+    {
+      "epoch": 2.0113636363636362,
+      "grad_norm": 2.523161159823314,
+      "learning_rate": 8.13971729243222e-07,
+      "loss": 0.0361,
+      "step": 8850
+    },
+    {
+      "epoch": 2.011590909090909,
+      "grad_norm": 3.146706904904031,
+      "learning_rate": 8.138866548204412e-07,
+      "loss": 0.0404,
+      "step": 8851
+    },
+    {
+      "epoch": 2.0118181818181817,
+      "grad_norm": 1.9138980745623788,
+      "learning_rate": 8.13801576545932e-07,
+      "loss": 0.0128,
+      "step": 8852
+    },
+    {
+      "epoch": 2.0120454545454547,
+      "grad_norm": 2.370585985675945,
+      "learning_rate": 8.137164944214293e-07,
+      "loss": 0.0585,
+      "step": 8853
+    },
+    {
+      "epoch": 2.012272727272727,
+      "grad_norm": 1.6464602173230276,
+      "learning_rate": 8.13631408448668e-07,
+      "loss": 0.0282,
+      "step": 8854
+    },
+    {
+      "epoch": 2.0125,
+      "grad_norm": 1.5847117813912173,
+      "learning_rate": 8.13546318629383e-07,
+      "loss": 0.013,
+      "step": 8855
+    },
+    {
+      "epoch": 2.0127272727272727,
+      "grad_norm": 1.8060621745613803,
+      "learning_rate": 8.134612249653098e-07,
+      "loss": 0.0169,
+      "step": 8856
+    },
+    {
+      "epoch": 2.0129545454545457,
+      "grad_norm": 1.7708152690421881,
+      "learning_rate": 8.133761274581837e-07,
+      "loss": 0.0149,
+      "step": 8857
+    },
+    {
+      "epoch": 2.013181818181818,
+      "grad_norm": 1.560534277184312,
+      "learning_rate": 8.132910261097393e-07,
+      "loss": 0.0507,
+      "step": 8858
+    },
+    {
+      "epoch": 2.0134090909090907,
+      "grad_norm": 2.4373853361013444,
+      "learning_rate": 8.132059209217126e-07,
+      "loss": 0.0279,
+      "step": 8859
+    },
+    {
+      "epoch": 2.0136363636363637,
+      "grad_norm": 1.3246543309800571,
+      "learning_rate": 8.131208118958388e-07,
+      "loss": 0.0107,
+      "step": 8860
+    },
+    {
+      "epoch": 2.013863636363636,
+      "grad_norm": 2.136113383289482,
+      "learning_rate": 8.130356990338535e-07,
+      "loss": 0.0301,
+      "step": 8861
+    },
+    {
+      "epoch": 2.014090909090909,
+      "grad_norm": 1.8964736398114062,
+      "learning_rate": 8.129505823374921e-07,
+      "loss": 0.0293,
+      "step": 8862
+    },
+    {
+      "epoch": 2.0143181818181817,
+      "grad_norm": 4.0623699825891855,
+      "learning_rate": 8.128654618084906e-07,
+      "loss": 0.0241,
+      "step": 8863
+    },
+    {
+      "epoch": 2.0145454545454546,
+      "grad_norm": 2.8043094618519815,
+      "learning_rate": 8.127803374485843e-07,
+      "loss": 0.0494,
+      "step": 8864
+    },
+    {
+      "epoch": 2.014772727272727,
+      "grad_norm": 2.9899829653234833,
+      "learning_rate": 8.126952092595098e-07,
+      "loss": 0.0286,
+      "step": 8865
+    },
+    {
+      "epoch": 2.015,
+      "grad_norm": 4.220157847745995,
+      "learning_rate": 8.126100772430021e-07,
+      "loss": 0.0887,
+      "step": 8866
+    },
+    {
+      "epoch": 2.0152272727272726,
+      "grad_norm": 2.0158695700325833,
+      "learning_rate": 8.125249414007977e-07,
+      "loss": 0.0341,
+      "step": 8867
+    },
+    {
+      "epoch": 2.0154545454545456,
+      "grad_norm": 1.8188202752323543,
+      "learning_rate": 8.124398017346327e-07,
+      "loss": 0.033,
+      "step": 8868
+    },
+    {
+      "epoch": 2.015681818181818,
+      "grad_norm": 1.1190676166338007,
+      "learning_rate": 8.123546582462432e-07,
+      "loss": 0.0174,
+      "step": 8869
+    },
+    {
+      "epoch": 2.015909090909091,
+      "grad_norm": 1.0787887025092133,
+      "learning_rate": 8.122695109373652e-07,
+      "loss": 0.0046,
+      "step": 8870
+    },
+    {
+      "epoch": 2.0161363636363636,
+      "grad_norm": 3.4165968465518834,
+      "learning_rate": 8.121843598097351e-07,
+      "loss": 0.0361,
+      "step": 8871
+    },
+    {
+      "epoch": 2.0163636363636366,
+      "grad_norm": 3.825233589888443,
+      "learning_rate": 8.120992048650896e-07,
+      "loss": 0.0313,
+      "step": 8872
+    },
+    {
+      "epoch": 2.016590909090909,
+      "grad_norm": 2.7960709064138505,
+      "learning_rate": 8.120140461051647e-07,
+      "loss": 0.0229,
+      "step": 8873
+    },
+    {
+      "epoch": 2.0168181818181816,
+      "grad_norm": 1.6424738292521888,
+      "learning_rate": 8.119288835316973e-07,
+      "loss": 0.0246,
+      "step": 8874
+    },
+    {
+      "epoch": 2.0170454545454546,
+      "grad_norm": 1.2112652060527727,
+      "learning_rate": 8.118437171464236e-07,
+      "loss": 0.0162,
+      "step": 8875
+    },
+    {
+      "epoch": 2.017272727272727,
+      "grad_norm": 0.9411419912863566,
+      "learning_rate": 8.117585469510808e-07,
+      "loss": 0.0155,
+      "step": 8876
+    },
+    {
+      "epoch": 2.0175,
+      "grad_norm": 3.0099862245961786,
+      "learning_rate": 8.116733729474053e-07,
+      "loss": 0.0361,
+      "step": 8877
+    },
+    {
+      "epoch": 2.0177272727272726,
+      "grad_norm": 1.6470717974477138,
+      "learning_rate": 8.11588195137134e-07,
+      "loss": 0.016,
+      "step": 8878
+    },
+    {
+      "epoch": 2.0179545454545456,
+      "grad_norm": 1.7286693113613112,
+      "learning_rate": 8.11503013522004e-07,
+      "loss": 0.0409,
+      "step": 8879
+    },
+    {
+      "epoch": 2.018181818181818,
+      "grad_norm": 1.8261326718090458,
+      "learning_rate": 8.11417828103752e-07,
+      "loss": 0.0277,
+      "step": 8880
+    },
+    {
+      "epoch": 2.018409090909091,
+      "grad_norm": 2.3487195386929107,
+      "learning_rate": 8.113326388841154e-07,
+      "loss": 0.0537,
+      "step": 8881
+    },
+    {
+      "epoch": 2.0186363636363636,
+      "grad_norm": 3.498011210133813,
+      "learning_rate": 8.112474458648311e-07,
+      "loss": 0.0277,
+      "step": 8882
+    },
+    {
+      "epoch": 2.0188636363636365,
+      "grad_norm": 3.0182352313125147,
+      "learning_rate": 8.111622490476367e-07,
+      "loss": 0.0528,
+      "step": 8883
+    },
+    {
+      "epoch": 2.019090909090909,
+      "grad_norm": 2.912218664579697,
+      "learning_rate": 8.11077048434269e-07,
+      "loss": 0.0315,
+      "step": 8884
+    },
+    {
+      "epoch": 2.019318181818182,
+      "grad_norm": 3.1415898899580594,
+      "learning_rate": 8.109918440264658e-07,
+      "loss": 0.0469,
+      "step": 8885
+    },
+    {
+      "epoch": 2.0195454545454545,
+      "grad_norm": 1.8617428915703156,
+      "learning_rate": 8.109066358259642e-07,
+      "loss": 0.0246,
+      "step": 8886
+    },
+    {
+      "epoch": 2.019772727272727,
+      "grad_norm": 1.5237862277819,
+      "learning_rate": 8.108214238345025e-07,
+      "loss": 0.0195,
+      "step": 8887
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 2.963188289444044,
+      "learning_rate": 8.107362080538173e-07,
+      "loss": 0.0431,
+      "step": 8888
+    },
+    {
+      "epoch": 2.0202272727272725,
+      "grad_norm": 1.9393724225373339,
+      "learning_rate": 8.106509884856469e-07,
+      "loss": 0.0191,
+      "step": 8889
+    },
+    {
+      "epoch": 2.0204545454545455,
+      "grad_norm": 2.8188266404099807,
+      "learning_rate": 8.10565765131729e-07,
+      "loss": 0.0259,
+      "step": 8890
+    },
+    {
+      "epoch": 2.020681818181818,
+      "grad_norm": 2.07038889770858,
+      "learning_rate": 8.104805379938013e-07,
+      "loss": 0.0382,
+      "step": 8891
+    },
+    {
+      "epoch": 2.020909090909091,
+      "grad_norm": 1.5158586129960885,
+      "learning_rate": 8.10395307073602e-07,
+      "loss": 0.0215,
+      "step": 8892
+    },
+    {
+      "epoch": 2.0211363636363635,
+      "grad_norm": 1.8542000659837785,
+      "learning_rate": 8.103100723728689e-07,
+      "loss": 0.0324,
+      "step": 8893
+    },
+    {
+      "epoch": 2.0213636363636365,
+      "grad_norm": 1.2277862477760493,
+      "learning_rate": 8.102248338933401e-07,
+      "loss": 0.0265,
+      "step": 8894
+    },
+    {
+      "epoch": 2.021590909090909,
+      "grad_norm": 2.073781157186795,
+      "learning_rate": 8.101395916367539e-07,
+      "loss": 0.0241,
+      "step": 8895
+    },
+    {
+      "epoch": 2.021818181818182,
+      "grad_norm": 2.632244111864261,
+      "learning_rate": 8.100543456048482e-07,
+      "loss": 0.0212,
+      "step": 8896
+    },
+    {
+      "epoch": 2.0220454545454545,
+      "grad_norm": 2.220109830462071,
+      "learning_rate": 8.099690957993617e-07,
+      "loss": 0.0301,
+      "step": 8897
+    },
+    {
+      "epoch": 2.0222727272727274,
+      "grad_norm": 2.8703063130454316,
+      "learning_rate": 8.098838422220328e-07,
+      "loss": 0.0203,
+      "step": 8898
+    },
+    {
+      "epoch": 2.0225,
+      "grad_norm": 0.9658761860259077,
+      "learning_rate": 8.097985848745995e-07,
+      "loss": 0.0156,
+      "step": 8899
+    },
+    {
+      "epoch": 2.022727272727273,
+      "grad_norm": 2.0695302569593093,
+      "learning_rate": 8.097133237588008e-07,
+      "loss": 0.0246,
+      "step": 8900
+    },
+    {
+      "epoch": 2.0229545454545454,
+      "grad_norm": 1.5045713991590146,
+      "learning_rate": 8.096280588763751e-07,
+      "loss": 0.0297,
+      "step": 8901
+    },
+    {
+      "epoch": 2.023181818181818,
+      "grad_norm": 2.493296534351663,
+      "learning_rate": 8.095427902290612e-07,
+      "loss": 0.0814,
+      "step": 8902
+    },
+    {
+      "epoch": 2.023409090909091,
+      "grad_norm": 3.032117996877638,
+      "learning_rate": 8.09457517818598e-07,
+      "loss": 0.08,
+      "step": 8903
+    },
+    {
+      "epoch": 2.0236363636363635,
+      "grad_norm": 1.793357585265989,
+      "learning_rate": 8.093722416467241e-07,
+      "loss": 0.0172,
+      "step": 8904
+    },
+    {
+      "epoch": 2.0238636363636364,
+      "grad_norm": 0.9913595704103181,
+      "learning_rate": 8.092869617151786e-07,
+      "loss": 0.0208,
+      "step": 8905
+    },
+    {
+      "epoch": 2.024090909090909,
+      "grad_norm": 1.7479841732939403,
+      "learning_rate": 8.092016780257003e-07,
+      "loss": 0.0184,
+      "step": 8906
+    },
+    {
+      "epoch": 2.024318181818182,
+      "grad_norm": 2.383138133373193,
+      "learning_rate": 8.091163905800286e-07,
+      "loss": 0.0294,
+      "step": 8907
+    },
+    {
+      "epoch": 2.0245454545454544,
+      "grad_norm": 2.2078466792034157,
+      "learning_rate": 8.090310993799025e-07,
+      "loss": 0.0471,
+      "step": 8908
+    },
+    {
+      "epoch": 2.0247727272727274,
+      "grad_norm": 1.7347176923364722,
+      "learning_rate": 8.089458044270609e-07,
+      "loss": 0.0173,
+      "step": 8909
+    },
+    {
+      "epoch": 2.025,
+      "grad_norm": 0.925764398982417,
+      "learning_rate": 8.088605057232437e-07,
+      "loss": 0.0087,
+      "step": 8910
+    },
+    {
+      "epoch": 2.025227272727273,
+      "grad_norm": 1.5957462510408085,
+      "learning_rate": 8.0877520327019e-07,
+      "loss": 0.0119,
+      "step": 8911
+    },
+    {
+      "epoch": 2.0254545454545454,
+      "grad_norm": 1.2302283490144261,
+      "learning_rate": 8.086898970696394e-07,
+      "loss": 0.0188,
+      "step": 8912
+    },
+    {
+      "epoch": 2.0256818181818184,
+      "grad_norm": 1.0073602473437357,
+      "learning_rate": 8.086045871233312e-07,
+      "loss": 0.0209,
+      "step": 8913
+    },
+    {
+      "epoch": 2.025909090909091,
+      "grad_norm": 2.1748520929258315,
+      "learning_rate": 8.085192734330053e-07,
+      "loss": 0.0205,
+      "step": 8914
+    },
+    {
+      "epoch": 2.026136363636364,
+      "grad_norm": 2.614795040546885,
+      "learning_rate": 8.08433956000401e-07,
+      "loss": 0.037,
+      "step": 8915
+    },
+    {
+      "epoch": 2.0263636363636364,
+      "grad_norm": 1.0508874902635847,
+      "learning_rate": 8.083486348272584e-07,
+      "loss": 0.0219,
+      "step": 8916
+    },
+    {
+      "epoch": 2.026590909090909,
+      "grad_norm": 0.9282681419379135,
+      "learning_rate": 8.082633099153173e-07,
+      "loss": 0.0174,
+      "step": 8917
+    },
+    {
+      "epoch": 2.026818181818182,
+      "grad_norm": 1.721714303546085,
+      "learning_rate": 8.081779812663177e-07,
+      "loss": 0.025,
+      "step": 8918
+    },
+    {
+      "epoch": 2.0270454545454544,
+      "grad_norm": 2.2491256653334957,
+      "learning_rate": 8.080926488819993e-07,
+      "loss": 0.0263,
+      "step": 8919
+    },
+    {
+      "epoch": 2.0272727272727273,
+      "grad_norm": 1.1162979322722322,
+      "learning_rate": 8.080073127641025e-07,
+      "loss": 0.0136,
+      "step": 8920
+    },
+    {
+      "epoch": 2.0275,
+      "grad_norm": 1.3729426533283442,
+      "learning_rate": 8.079219729143672e-07,
+      "loss": 0.0157,
+      "step": 8921
+    },
+    {
+      "epoch": 2.027727272727273,
+      "grad_norm": 1.6775716390179225,
+      "learning_rate": 8.078366293345339e-07,
+      "loss": 0.0215,
+      "step": 8922
+    },
+    {
+      "epoch": 2.0279545454545453,
+      "grad_norm": 1.4708076626652982,
+      "learning_rate": 8.077512820263426e-07,
+      "loss": 0.0239,
+      "step": 8923
+    },
+    {
+      "epoch": 2.0281818181818183,
+      "grad_norm": 2.5509070361589803,
+      "learning_rate": 8.076659309915339e-07,
+      "loss": 0.0384,
+      "step": 8924
+    },
+    {
+      "epoch": 2.028409090909091,
+      "grad_norm": 1.598339685720091,
+      "learning_rate": 8.075805762318482e-07,
+      "loss": 0.0166,
+      "step": 8925
+    },
+    {
+      "epoch": 2.028636363636364,
+      "grad_norm": 1.2565034156585466,
+      "learning_rate": 8.07495217749026e-07,
+      "loss": 0.0156,
+      "step": 8926
+    },
+    {
+      "epoch": 2.0288636363636363,
+      "grad_norm": 2.448913007335206,
+      "learning_rate": 8.07409855544808e-07,
+      "loss": 0.0399,
+      "step": 8927
+    },
+    {
+      "epoch": 2.0290909090909093,
+      "grad_norm": 2.631951058505373,
+      "learning_rate": 8.073244896209347e-07,
+      "loss": 0.042,
+      "step": 8928
+    },
+    {
+      "epoch": 2.029318181818182,
+      "grad_norm": 1.635192128181818,
+      "learning_rate": 8.072391199791471e-07,
+      "loss": 0.0236,
+      "step": 8929
+    },
+    {
+      "epoch": 2.0295454545454548,
+      "grad_norm": 1.1261884114773426,
+      "learning_rate": 8.071537466211858e-07,
+      "loss": 0.013,
+      "step": 8930
+    },
+    {
+      "epoch": 2.0297727272727273,
+      "grad_norm": 1.7585484650142076,
+      "learning_rate": 8.070683695487919e-07,
+      "loss": 0.0334,
+      "step": 8931
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.1371285719332789,
+      "learning_rate": 8.069829887637063e-07,
+      "loss": 0.0186,
+      "step": 8932
+    },
+    {
+      "epoch": 2.0302272727272728,
+      "grad_norm": 3.6716287996954238,
+      "learning_rate": 8.068976042676699e-07,
+      "loss": 0.0369,
+      "step": 8933
+    },
+    {
+      "epoch": 2.0304545454545453,
+      "grad_norm": 1.8779143389196666,
+      "learning_rate": 8.068122160624242e-07,
+      "loss": 0.0278,
+      "step": 8934
+    },
+    {
+      "epoch": 2.0306818181818183,
+      "grad_norm": 1.1485147302043919,
+      "learning_rate": 8.067268241497102e-07,
+      "loss": 0.0176,
+      "step": 8935
+    },
+    {
+      "epoch": 2.0309090909090908,
+      "grad_norm": 2.466820857603386,
+      "learning_rate": 8.066414285312693e-07,
+      "loss": 0.0236,
+      "step": 8936
+    },
+    {
+      "epoch": 2.0311363636363637,
+      "grad_norm": 2.8097117202964195,
+      "learning_rate": 8.065560292088425e-07,
+      "loss": 0.0355,
+      "step": 8937
+    },
+    {
+      "epoch": 2.0313636363636363,
+      "grad_norm": 2.245882497838707,
+      "learning_rate": 8.06470626184172e-07,
+      "loss": 0.0246,
+      "step": 8938
+    },
+    {
+      "epoch": 2.0315909090909092,
+      "grad_norm": 1.8898648519224202,
+      "learning_rate": 8.063852194589982e-07,
+      "loss": 0.0137,
+      "step": 8939
+    },
+    {
+      "epoch": 2.0318181818181817,
+      "grad_norm": 1.7148629851358677,
+      "learning_rate": 8.062998090350637e-07,
+      "loss": 0.0235,
+      "step": 8940
+    },
+    {
+      "epoch": 2.0320454545454547,
+      "grad_norm": 2.1249195622289774,
+      "learning_rate": 8.062143949141099e-07,
+      "loss": 0.0313,
+      "step": 8941
+    },
+    {
+      "epoch": 2.0322727272727272,
+      "grad_norm": 1.8405345015626307,
+      "learning_rate": 8.061289770978783e-07,
+      "loss": 0.0171,
+      "step": 8942
+    },
+    {
+      "epoch": 2.0325,
+      "grad_norm": 1.0274171847019964,
+      "learning_rate": 8.060435555881109e-07,
+      "loss": 0.0096,
+      "step": 8943
+    },
+    {
+      "epoch": 2.0327272727272727,
+      "grad_norm": 1.7582541118997015,
+      "learning_rate": 8.059581303865497e-07,
+      "loss": 0.0169,
+      "step": 8944
+    },
+    {
+      "epoch": 2.0329545454545452,
+      "grad_norm": 3.3912244249693995,
+      "learning_rate": 8.058727014949362e-07,
+      "loss": 0.0399,
+      "step": 8945
+    },
+    {
+      "epoch": 2.033181818181818,
+      "grad_norm": 1.5602187336546136,
+      "learning_rate": 8.057872689150129e-07,
+      "loss": 0.0167,
+      "step": 8946
+    },
+    {
+      "epoch": 2.0334090909090907,
+      "grad_norm": 2.0501131136974156,
+      "learning_rate": 8.057018326485218e-07,
+      "loss": 0.0161,
+      "step": 8947
+    },
+    {
+      "epoch": 2.0336363636363637,
+      "grad_norm": 1.7688399065169667,
+      "learning_rate": 8.056163926972049e-07,
+      "loss": 0.0146,
+      "step": 8948
+    },
+    {
+      "epoch": 2.033863636363636,
+      "grad_norm": 1.4287278604108569,
+      "learning_rate": 8.05530949062805e-07,
+      "loss": 0.036,
+      "step": 8949
+    },
+    {
+      "epoch": 2.034090909090909,
+      "grad_norm": 2.300456842138099,
+      "learning_rate": 8.054455017470638e-07,
+      "loss": 0.0267,
+      "step": 8950
+    },
+    {
+      "epoch": 2.0343181818181817,
+      "grad_norm": 2.781408279434715,
+      "learning_rate": 8.053600507517244e-07,
+      "loss": 0.0252,
+      "step": 8951
+    },
+    {
+      "epoch": 2.0345454545454547,
+      "grad_norm": 2.000037280880295,
+      "learning_rate": 8.052745960785285e-07,
+      "loss": 0.0226,
+      "step": 8952
+    },
+    {
+      "epoch": 2.034772727272727,
+      "grad_norm": 2.550519188207765,
+      "learning_rate": 8.051891377292194e-07,
+      "loss": 0.0186,
+      "step": 8953
+    },
+    {
+      "epoch": 2.035,
+      "grad_norm": 4.3952888779956725,
+      "learning_rate": 8.05103675705539e-07,
+      "loss": 0.0371,
+      "step": 8954
+    },
+    {
+      "epoch": 2.0352272727272727,
+      "grad_norm": 1.856648391523915,
+      "learning_rate": 8.050182100092309e-07,
+      "loss": 0.0319,
+      "step": 8955
+    },
+    {
+      "epoch": 2.0354545454545456,
+      "grad_norm": 2.4371788449863376,
+      "learning_rate": 8.049327406420371e-07,
+      "loss": 0.0189,
+      "step": 8956
+    },
+    {
+      "epoch": 2.035681818181818,
+      "grad_norm": 3.133495240210206,
+      "learning_rate": 8.048472676057009e-07,
+      "loss": 0.021,
+      "step": 8957
+    },
+    {
+      "epoch": 2.035909090909091,
+      "grad_norm": 2.07792132627012,
+      "learning_rate": 8.047617909019654e-07,
+      "loss": 0.0201,
+      "step": 8958
+    },
+    {
+      "epoch": 2.0361363636363636,
+      "grad_norm": 1.8721506946360966,
+      "learning_rate": 8.046763105325731e-07,
+      "loss": 0.0186,
+      "step": 8959
+    },
+    {
+      "epoch": 2.036363636363636,
+      "grad_norm": 1.5184186976470155,
+      "learning_rate": 8.045908264992674e-07,
+      "loss": 0.0077,
+      "step": 8960
+    },
+    {
+      "epoch": 2.036590909090909,
+      "grad_norm": 2.1893713653289266,
+      "learning_rate": 8.045053388037916e-07,
+      "loss": 0.0422,
+      "step": 8961
+    },
+    {
+      "epoch": 2.0368181818181816,
+      "grad_norm": 4.3927537403795505,
+      "learning_rate": 8.044198474478885e-07,
+      "loss": 0.069,
+      "step": 8962
+    },
+    {
+      "epoch": 2.0370454545454546,
+      "grad_norm": 2.1468473132676094,
+      "learning_rate": 8.043343524333019e-07,
+      "loss": 0.0273,
+      "step": 8963
+    },
+    {
+      "epoch": 2.037272727272727,
+      "grad_norm": 1.9596371164658941,
+      "learning_rate": 8.04248853761775e-07,
+      "loss": 0.025,
+      "step": 8964
+    },
+    {
+      "epoch": 2.0375,
+      "grad_norm": 1.4983438250890506,
+      "learning_rate": 8.04163351435051e-07,
+      "loss": 0.0242,
+      "step": 8965
+    },
+    {
+      "epoch": 2.0377272727272726,
+      "grad_norm": 1.9412423083121408,
+      "learning_rate": 8.040778454548741e-07,
+      "loss": 0.0258,
+      "step": 8966
+    },
+    {
+      "epoch": 2.0379545454545456,
+      "grad_norm": 1.1598184546636703,
+      "learning_rate": 8.039923358229873e-07,
+      "loss": 0.0113,
+      "step": 8967
+    },
+    {
+      "epoch": 2.038181818181818,
+      "grad_norm": 1.7262469609438738,
+      "learning_rate": 8.039068225411345e-07,
+      "loss": 0.0333,
+      "step": 8968
+    },
+    {
+      "epoch": 2.038409090909091,
+      "grad_norm": 2.7663630370845698,
+      "learning_rate": 8.038213056110594e-07,
+      "loss": 0.0256,
+      "step": 8969
+    },
+    {
+      "epoch": 2.0386363636363636,
+      "grad_norm": 1.3046633113778867,
+      "learning_rate": 8.037357850345059e-07,
+      "loss": 0.0152,
+      "step": 8970
+    },
+    {
+      "epoch": 2.0388636363636365,
+      "grad_norm": 1.7343213211977309,
+      "learning_rate": 8.036502608132181e-07,
+      "loss": 0.0147,
+      "step": 8971
+    },
+    {
+      "epoch": 2.039090909090909,
+      "grad_norm": 1.8146037543279956,
+      "learning_rate": 8.035647329489396e-07,
+      "loss": 0.0179,
+      "step": 8972
+    },
+    {
+      "epoch": 2.039318181818182,
+      "grad_norm": 1.6204455250239433,
+      "learning_rate": 8.034792014434149e-07,
+      "loss": 0.0172,
+      "step": 8973
+    },
+    {
+      "epoch": 2.0395454545454546,
+      "grad_norm": 2.1878990012431765,
+      "learning_rate": 8.033936662983878e-07,
+      "loss": 0.0233,
+      "step": 8974
+    },
+    {
+      "epoch": 2.039772727272727,
+      "grad_norm": 0.6037630249341628,
+      "learning_rate": 8.033081275156027e-07,
+      "loss": 0.0079,
+      "step": 8975
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 3.9954128178011015,
+      "learning_rate": 8.032225850968037e-07,
+      "loss": 0.0449,
+      "step": 8976
+    },
+    {
+      "epoch": 2.0402272727272726,
+      "grad_norm": 2.0527595431154886,
+      "learning_rate": 8.031370390437352e-07,
+      "loss": 0.0138,
+      "step": 8977
+    },
+    {
+      "epoch": 2.0404545454545455,
+      "grad_norm": 2.415489834406922,
+      "learning_rate": 8.030514893581417e-07,
+      "loss": 0.0488,
+      "step": 8978
+    },
+    {
+      "epoch": 2.040681818181818,
+      "grad_norm": 1.5688250355913895,
+      "learning_rate": 8.029659360417677e-07,
+      "loss": 0.0117,
+      "step": 8979
+    },
+    {
+      "epoch": 2.040909090909091,
+      "grad_norm": 1.5808354198340206,
+      "learning_rate": 8.02880379096358e-07,
+      "loss": 0.0193,
+      "step": 8980
+    },
+    {
+      "epoch": 2.0411363636363635,
+      "grad_norm": 2.0932707009923055,
+      "learning_rate": 8.027948185236568e-07,
+      "loss": 0.0282,
+      "step": 8981
+    },
+    {
+      "epoch": 2.0413636363636365,
+      "grad_norm": 2.041255333195398,
+      "learning_rate": 8.027092543254094e-07,
+      "loss": 0.0261,
+      "step": 8982
+    },
+    {
+      "epoch": 2.041590909090909,
+      "grad_norm": 1.278835088835247,
+      "learning_rate": 8.026236865033599e-07,
+      "loss": 0.0099,
+      "step": 8983
+    },
+    {
+      "epoch": 2.041818181818182,
+      "grad_norm": 1.9092561156887853,
+      "learning_rate": 8.025381150592537e-07,
+      "loss": 0.0155,
+      "step": 8984
+    },
+    {
+      "epoch": 2.0420454545454545,
+      "grad_norm": 1.8954286069894728,
+      "learning_rate": 8.024525399948356e-07,
+      "loss": 0.0466,
+      "step": 8985
+    },
+    {
+      "epoch": 2.0422727272727275,
+      "grad_norm": 1.0919184675633073,
+      "learning_rate": 8.023669613118507e-07,
+      "loss": 0.0137,
+      "step": 8986
+    },
+    {
+      "epoch": 2.0425,
+      "grad_norm": 1.534268454395415,
+      "learning_rate": 8.022813790120439e-07,
+      "loss": 0.0201,
+      "step": 8987
+    },
+    {
+      "epoch": 2.042727272727273,
+      "grad_norm": 1.2605430839608562,
+      "learning_rate": 8.021957930971606e-07,
+      "loss": 0.0142,
+      "step": 8988
+    },
+    {
+      "epoch": 2.0429545454545455,
+      "grad_norm": 1.6838707816946814,
+      "learning_rate": 8.021102035689459e-07,
+      "loss": 0.0191,
+      "step": 8989
+    },
+    {
+      "epoch": 2.043181818181818,
+      "grad_norm": 3.0015218044875507,
+      "learning_rate": 8.020246104291451e-07,
+      "loss": 0.034,
+      "step": 8990
+    },
+    {
+      "epoch": 2.043409090909091,
+      "grad_norm": 2.1724163966818986,
+      "learning_rate": 8.019390136795039e-07,
+      "loss": 0.0293,
+      "step": 8991
+    },
+    {
+      "epoch": 2.0436363636363635,
+      "grad_norm": 2.0442013412206927,
+      "learning_rate": 8.018534133217672e-07,
+      "loss": 0.0298,
+      "step": 8992
+    },
+    {
+      "epoch": 2.0438636363636364,
+      "grad_norm": 0.850978791462434,
+      "learning_rate": 8.01767809357681e-07,
+      "loss": 0.0059,
+      "step": 8993
+    },
+    {
+      "epoch": 2.044090909090909,
+      "grad_norm": 1.6769577504326383,
+      "learning_rate": 8.016822017889907e-07,
+      "loss": 0.0202,
+      "step": 8994
+    },
+    {
+      "epoch": 2.044318181818182,
+      "grad_norm": 1.6458526967779405,
+      "learning_rate": 8.015965906174422e-07,
+      "loss": 0.0136,
+      "step": 8995
+    },
+    {
+      "epoch": 2.0445454545454544,
+      "grad_norm": 1.4604017930492739,
+      "learning_rate": 8.01510975844781e-07,
+      "loss": 0.0181,
+      "step": 8996
+    },
+    {
+      "epoch": 2.0447727272727274,
+      "grad_norm": 2.226845415228313,
+      "learning_rate": 8.014253574727534e-07,
+      "loss": 0.0245,
+      "step": 8997
+    },
+    {
+      "epoch": 2.045,
+      "grad_norm": 2.3200047643241284,
+      "learning_rate": 8.013397355031045e-07,
+      "loss": 0.0393,
+      "step": 8998
+    },
+    {
+      "epoch": 2.045227272727273,
+      "grad_norm": 2.2409811293936235,
+      "learning_rate": 8.012541099375809e-07,
+      "loss": 0.0157,
+      "step": 8999
+    },
+    {
+      "epoch": 2.0454545454545454,
+      "grad_norm": 1.8357585759879784,
+      "learning_rate": 8.011684807779286e-07,
+      "loss": 0.0174,
+      "step": 9000
+    },
+    {
+      "epoch": 2.0456818181818184,
+      "grad_norm": 0.9431738501272071,
+      "learning_rate": 8.010828480258937e-07,
+      "loss": 0.0075,
+      "step": 9001
+    },
+    {
+      "epoch": 2.045909090909091,
+      "grad_norm": 1.3698850878838027,
+      "learning_rate": 8.009972116832223e-07,
+      "loss": 0.0172,
+      "step": 9002
+    },
+    {
+      "epoch": 2.0461363636363634,
+      "grad_norm": 2.4025122643350687,
+      "learning_rate": 8.009115717516605e-07,
+      "loss": 0.0313,
+      "step": 9003
+    },
+    {
+      "epoch": 2.0463636363636364,
+      "grad_norm": 2.3598273653984254,
+      "learning_rate": 8.00825928232955e-07,
+      "loss": 0.0302,
+      "step": 9004
+    },
+    {
+      "epoch": 2.046590909090909,
+      "grad_norm": 1.3794674272494922,
+      "learning_rate": 8.007402811288522e-07,
+      "loss": 0.0108,
+      "step": 9005
+    },
+    {
+      "epoch": 2.046818181818182,
+      "grad_norm": 2.823614823022368,
+      "learning_rate": 8.006546304410984e-07,
+      "loss": 0.0288,
+      "step": 9006
+    },
+    {
+      "epoch": 2.0470454545454544,
+      "grad_norm": 1.9492275352795152,
+      "learning_rate": 8.005689761714401e-07,
+      "loss": 0.0194,
+      "step": 9007
+    },
+    {
+      "epoch": 2.0472727272727274,
+      "grad_norm": 1.7957121553808888,
+      "learning_rate": 8.004833183216242e-07,
+      "loss": 0.024,
+      "step": 9008
+    },
+    {
+      "epoch": 2.0475,
+      "grad_norm": 1.7874726272649921,
+      "learning_rate": 8.003976568933972e-07,
+      "loss": 0.0287,
+      "step": 9009
+    },
+    {
+      "epoch": 2.047727272727273,
+      "grad_norm": 1.5954976141447965,
+      "learning_rate": 8.003119918885061e-07,
+      "loss": 0.0137,
+      "step": 9010
+    },
+    {
+      "epoch": 2.0479545454545454,
+      "grad_norm": 1.9684532200694211,
+      "learning_rate": 8.002263233086975e-07,
+      "loss": 0.0322,
+      "step": 9011
+    },
+    {
+      "epoch": 2.0481818181818183,
+      "grad_norm": 3.9237557734899027,
+      "learning_rate": 8.001406511557186e-07,
+      "loss": 0.0481,
+      "step": 9012
+    },
+    {
+      "epoch": 2.048409090909091,
+      "grad_norm": 1.7027586603857174,
+      "learning_rate": 8.000549754313163e-07,
+      "loss": 0.0218,
+      "step": 9013
+    },
+    {
+      "epoch": 2.048636363636364,
+      "grad_norm": 1.5216721199719656,
+      "learning_rate": 7.999692961372374e-07,
+      "loss": 0.0309,
+      "step": 9014
+    },
+    {
+      "epoch": 2.0488636363636363,
+      "grad_norm": 1.678927861783464,
+      "learning_rate": 7.998836132752295e-07,
+      "loss": 0.0168,
+      "step": 9015
+    },
+    {
+      "epoch": 2.0490909090909093,
+      "grad_norm": 2.1590549473266543,
+      "learning_rate": 7.997979268470396e-07,
+      "loss": 0.0286,
+      "step": 9016
+    },
+    {
+      "epoch": 2.049318181818182,
+      "grad_norm": 1.928653770261641,
+      "learning_rate": 7.997122368544151e-07,
+      "loss": 0.0213,
+      "step": 9017
+    },
+    {
+      "epoch": 2.0495454545454543,
+      "grad_norm": 2.181601018196589,
+      "learning_rate": 7.996265432991031e-07,
+      "loss": 0.0318,
+      "step": 9018
+    },
+    {
+      "epoch": 2.0497727272727273,
+      "grad_norm": 3.0480760827135156,
+      "learning_rate": 7.995408461828515e-07,
+      "loss": 0.0386,
+      "step": 9019
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.6916742146935588,
+      "learning_rate": 7.994551455074073e-07,
+      "loss": 0.0105,
+      "step": 9020
+    },
+    {
+      "epoch": 2.050227272727273,
+      "grad_norm": 1.706808480227056,
+      "learning_rate": 7.993694412745184e-07,
+      "loss": 0.0147,
+      "step": 9021
+    },
+    {
+      "epoch": 2.0504545454545453,
+      "grad_norm": 1.3359008962616985,
+      "learning_rate": 7.992837334859324e-07,
+      "loss": 0.0137,
+      "step": 9022
+    },
+    {
+      "epoch": 2.0506818181818183,
+      "grad_norm": 1.6523868108690722,
+      "learning_rate": 7.991980221433969e-07,
+      "loss": 0.0098,
+      "step": 9023
+    },
+    {
+      "epoch": 2.050909090909091,
+      "grad_norm": 1.3869569895693057,
+      "learning_rate": 7.9911230724866e-07,
+      "loss": 0.021,
+      "step": 9024
+    },
+    {
+      "epoch": 2.0511363636363638,
+      "grad_norm": 2.459119094695033,
+      "learning_rate": 7.990265888034693e-07,
+      "loss": 0.0172,
+      "step": 9025
+    },
+    {
+      "epoch": 2.0513636363636363,
+      "grad_norm": 1.1580072257108236,
+      "learning_rate": 7.98940866809573e-07,
+      "loss": 0.0163,
+      "step": 9026
+    },
+    {
+      "epoch": 2.0515909090909092,
+      "grad_norm": 1.2633320750414327,
+      "learning_rate": 7.988551412687188e-07,
+      "loss": 0.0191,
+      "step": 9027
+    },
+    {
+      "epoch": 2.0518181818181818,
+      "grad_norm": 2.894456799382323,
+      "learning_rate": 7.98769412182655e-07,
+      "loss": 0.0402,
+      "step": 9028
+    },
+    {
+      "epoch": 2.0520454545454547,
+      "grad_norm": 3.0679098623322005,
+      "learning_rate": 7.986836795531298e-07,
+      "loss": 0.051,
+      "step": 9029
+    },
+    {
+      "epoch": 2.0522727272727272,
+      "grad_norm": 1.8615226559568283,
+      "learning_rate": 7.985979433818914e-07,
+      "loss": 0.0201,
+      "step": 9030
+    },
+    {
+      "epoch": 2.0525,
+      "grad_norm": 2.4839917054965994,
+      "learning_rate": 7.985122036706879e-07,
+      "loss": 0.0631,
+      "step": 9031
+    },
+    {
+      "epoch": 2.0527272727272727,
+      "grad_norm": 2.977283870213277,
+      "learning_rate": 7.98426460421268e-07,
+      "loss": 0.0184,
+      "step": 9032
+    },
+    {
+      "epoch": 2.0529545454545453,
+      "grad_norm": 0.8368669777098757,
+      "learning_rate": 7.983407136353799e-07,
+      "loss": 0.011,
+      "step": 9033
+    },
+    {
+      "epoch": 2.053181818181818,
+      "grad_norm": 0.8353549162765513,
+      "learning_rate": 7.982549633147722e-07,
+      "loss": 0.0091,
+      "step": 9034
+    },
+    {
+      "epoch": 2.0534090909090907,
+      "grad_norm": 1.3882770080102982,
+      "learning_rate": 7.981692094611937e-07,
+      "loss": 0.0159,
+      "step": 9035
+    },
+    {
+      "epoch": 2.0536363636363637,
+      "grad_norm": 2.7999522516537407,
+      "learning_rate": 7.980834520763929e-07,
+      "loss": 0.013,
+      "step": 9036
+    },
+    {
+      "epoch": 2.0538636363636362,
+      "grad_norm": 1.9110840470627486,
+      "learning_rate": 7.979976911621186e-07,
+      "loss": 0.0339,
+      "step": 9037
+    },
+    {
+      "epoch": 2.054090909090909,
+      "grad_norm": 2.652325624485484,
+      "learning_rate": 7.979119267201192e-07,
+      "loss": 0.0315,
+      "step": 9038
+    },
+    {
+      "epoch": 2.0543181818181817,
+      "grad_norm": 1.3436929946400216,
+      "learning_rate": 7.978261587521443e-07,
+      "loss": 0.0123,
+      "step": 9039
+    },
+    {
+      "epoch": 2.0545454545454547,
+      "grad_norm": 1.576334157595129,
+      "learning_rate": 7.977403872599423e-07,
+      "loss": 0.0213,
+      "step": 9040
+    },
+    {
+      "epoch": 2.054772727272727,
+      "grad_norm": 1.9416411459476273,
+      "learning_rate": 7.976546122452627e-07,
+      "loss": 0.018,
+      "step": 9041
+    },
+    {
+      "epoch": 2.055,
+      "grad_norm": 1.9624172011193253,
+      "learning_rate": 7.975688337098542e-07,
+      "loss": 0.0321,
+      "step": 9042
+    },
+    {
+      "epoch": 2.0552272727272727,
+      "grad_norm": 1.7806043851265236,
+      "learning_rate": 7.974830516554663e-07,
+      "loss": 0.019,
+      "step": 9043
+    },
+    {
+      "epoch": 2.0554545454545456,
+      "grad_norm": 1.9612023951736808,
+      "learning_rate": 7.973972660838478e-07,
+      "loss": 0.0196,
+      "step": 9044
+    },
+    {
+      "epoch": 2.055681818181818,
+      "grad_norm": 3.6606299924876082,
+      "learning_rate": 7.973114769967482e-07,
+      "loss": 0.0371,
+      "step": 9045
+    },
+    {
+      "epoch": 2.0559090909090907,
+      "grad_norm": 18.80453856523345,
+      "learning_rate": 7.972256843959172e-07,
+      "loss": 0.0279,
+      "step": 9046
+    },
+    {
+      "epoch": 2.0561363636363637,
+      "grad_norm": 1.947574102158957,
+      "learning_rate": 7.97139888283104e-07,
+      "loss": 0.0262,
+      "step": 9047
+    },
+    {
+      "epoch": 2.056363636363636,
+      "grad_norm": 2.3152922064335213,
+      "learning_rate": 7.970540886600582e-07,
+      "loss": 0.0191,
+      "step": 9048
+    },
+    {
+      "epoch": 2.056590909090909,
+      "grad_norm": 0.588221119827585,
+      "learning_rate": 7.969682855285294e-07,
+      "loss": 0.006,
+      "step": 9049
+    },
+    {
+      "epoch": 2.0568181818181817,
+      "grad_norm": 1.8627774824949737,
+      "learning_rate": 7.968824788902671e-07,
+      "loss": 0.0291,
+      "step": 9050
+    },
+    {
+      "epoch": 2.0570454545454546,
+      "grad_norm": 1.5176945784419336,
+      "learning_rate": 7.967966687470211e-07,
+      "loss": 0.0111,
+      "step": 9051
+    },
+    {
+      "epoch": 2.057272727272727,
+      "grad_norm": 2.2757601809049333,
+      "learning_rate": 7.967108551005415e-07,
+      "loss": 0.044,
+      "step": 9052
+    },
+    {
+      "epoch": 2.0575,
+      "grad_norm": 3.327977141735081,
+      "learning_rate": 7.966250379525779e-07,
+      "loss": 0.0639,
+      "step": 9053
+    },
+    {
+      "epoch": 2.0577272727272726,
+      "grad_norm": 2.2664423357151207,
+      "learning_rate": 7.965392173048806e-07,
+      "loss": 0.0195,
+      "step": 9054
+    },
+    {
+      "epoch": 2.0579545454545456,
+      "grad_norm": 2.325111257528079,
+      "learning_rate": 7.964533931591991e-07,
+      "loss": 0.0457,
+      "step": 9055
+    },
+    {
+      "epoch": 2.058181818181818,
+      "grad_norm": 1.631819088261587,
+      "learning_rate": 7.963675655172839e-07,
+      "loss": 0.0173,
+      "step": 9056
+    },
+    {
+      "epoch": 2.058409090909091,
+      "grad_norm": 0.8025893171911359,
+      "learning_rate": 7.962817343808851e-07,
+      "loss": 0.0058,
+      "step": 9057
+    },
+    {
+      "epoch": 2.0586363636363636,
+      "grad_norm": 3.0458217360101556,
+      "learning_rate": 7.961958997517528e-07,
+      "loss": 0.052,
+      "step": 9058
+    },
+    {
+      "epoch": 2.0588636363636366,
+      "grad_norm": 1.508546989047244,
+      "learning_rate": 7.961100616316376e-07,
+      "loss": 0.0169,
+      "step": 9059
+    },
+    {
+      "epoch": 2.059090909090909,
+      "grad_norm": 1.1284467423621212,
+      "learning_rate": 7.960242200222896e-07,
+      "loss": 0.0156,
+      "step": 9060
+    },
+    {
+      "epoch": 2.0593181818181816,
+      "grad_norm": 4.716097985921757,
+      "learning_rate": 7.959383749254595e-07,
+      "loss": 0.0522,
+      "step": 9061
+    },
+    {
+      "epoch": 2.0595454545454546,
+      "grad_norm": 2.5544715584600812,
+      "learning_rate": 7.958525263428976e-07,
+      "loss": 0.024,
+      "step": 9062
+    },
+    {
+      "epoch": 2.059772727272727,
+      "grad_norm": 0.8255906306359435,
+      "learning_rate": 7.957666742763548e-07,
+      "loss": 0.009,
+      "step": 9063
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.3775829850303762,
+      "learning_rate": 7.956808187275813e-07,
+      "loss": 0.0126,
+      "step": 9064
+    },
+    {
+      "epoch": 2.0602272727272726,
+      "grad_norm": 1.5007972728837908,
+      "learning_rate": 7.955949596983284e-07,
+      "loss": 0.0195,
+      "step": 9065
+    },
+    {
+      "epoch": 2.0604545454545455,
+      "grad_norm": 4.217656657296637,
+      "learning_rate": 7.955090971903464e-07,
+      "loss": 0.046,
+      "step": 9066
+    },
+    {
+      "epoch": 2.060681818181818,
+      "grad_norm": 1.5080108400323347,
+      "learning_rate": 7.954232312053867e-07,
+      "loss": 0.0203,
+      "step": 9067
+    },
+    {
+      "epoch": 2.060909090909091,
+      "grad_norm": 1.4184943376211048,
+      "learning_rate": 7.953373617451998e-07,
+      "loss": 0.0168,
+      "step": 9068
+    },
+    {
+      "epoch": 2.0611363636363635,
+      "grad_norm": 2.110899563642247,
+      "learning_rate": 7.95251488811537e-07,
+      "loss": 0.0184,
+      "step": 9069
+    },
+    {
+      "epoch": 2.0613636363636365,
+      "grad_norm": 1.0658352388570218,
+      "learning_rate": 7.951656124061493e-07,
+      "loss": 0.0105,
+      "step": 9070
+    },
+    {
+      "epoch": 2.061590909090909,
+      "grad_norm": 1.3486199460168429,
+      "learning_rate": 7.95079732530788e-07,
+      "loss": 0.0206,
+      "step": 9071
+    },
+    {
+      "epoch": 2.061818181818182,
+      "grad_norm": 0.6018898979867817,
+      "learning_rate": 7.949938491872043e-07,
+      "loss": 0.0097,
+      "step": 9072
+    },
+    {
+      "epoch": 2.0620454545454545,
+      "grad_norm": 0.9214892244277431,
+      "learning_rate": 7.949079623771492e-07,
+      "loss": 0.007,
+      "step": 9073
+    },
+    {
+      "epoch": 2.0622727272727275,
+      "grad_norm": 2.3183763417987358,
+      "learning_rate": 7.948220721023745e-07,
+      "loss": 0.0235,
+      "step": 9074
+    },
+    {
+      "epoch": 2.0625,
+      "grad_norm": 1.2111743055732138,
+      "learning_rate": 7.947361783646315e-07,
+      "loss": 0.0118,
+      "step": 9075
+    },
+    {
+      "epoch": 2.0627272727272725,
+      "grad_norm": 0.8578277706526317,
+      "learning_rate": 7.946502811656714e-07,
+      "loss": 0.0053,
+      "step": 9076
+    },
+    {
+      "epoch": 2.0629545454545455,
+      "grad_norm": 0.824695684193645,
+      "learning_rate": 7.945643805072465e-07,
+      "loss": 0.0184,
+      "step": 9077
+    },
+    {
+      "epoch": 2.063181818181818,
+      "grad_norm": 2.8038462054537794,
+      "learning_rate": 7.944784763911079e-07,
+      "loss": 0.0702,
+      "step": 9078
+    },
+    {
+      "epoch": 2.063409090909091,
+      "grad_norm": 2.0162093023981162,
+      "learning_rate": 7.943925688190074e-07,
+      "loss": 0.0192,
+      "step": 9079
+    },
+    {
+      "epoch": 2.0636363636363635,
+      "grad_norm": 1.8683459894624694,
+      "learning_rate": 7.943066577926969e-07,
+      "loss": 0.0239,
+      "step": 9080
+    },
+    {
+      "epoch": 2.0638636363636365,
+      "grad_norm": 1.7085449139535915,
+      "learning_rate": 7.942207433139284e-07,
+      "loss": 0.0208,
+      "step": 9081
+    },
+    {
+      "epoch": 2.064090909090909,
+      "grad_norm": 1.5261477327808528,
+      "learning_rate": 7.941348253844536e-07,
+      "loss": 0.0209,
+      "step": 9082
+    },
+    {
+      "epoch": 2.064318181818182,
+      "grad_norm": 2.8471415760151966,
+      "learning_rate": 7.940489040060246e-07,
+      "loss": 0.059,
+      "step": 9083
+    },
+    {
+      "epoch": 2.0645454545454545,
+      "grad_norm": 1.8385967347627095,
+      "learning_rate": 7.939629791803937e-07,
+      "loss": 0.0214,
+      "step": 9084
+    },
+    {
+      "epoch": 2.0647727272727274,
+      "grad_norm": 5.205857516928785,
+      "learning_rate": 7.938770509093128e-07,
+      "loss": 0.0245,
+      "step": 9085
+    },
+    {
+      "epoch": 2.065,
+      "grad_norm": 2.711076326719968,
+      "learning_rate": 7.937911191945342e-07,
+      "loss": 0.0365,
+      "step": 9086
+    },
+    {
+      "epoch": 2.065227272727273,
+      "grad_norm": 1.320880403041551,
+      "learning_rate": 7.937051840378105e-07,
+      "loss": 0.0128,
+      "step": 9087
+    },
+    {
+      "epoch": 2.0654545454545454,
+      "grad_norm": 1.5862869367510672,
+      "learning_rate": 7.936192454408934e-07,
+      "loss": 0.0229,
+      "step": 9088
+    },
+    {
+      "epoch": 2.0656818181818184,
+      "grad_norm": 2.380679187703778,
+      "learning_rate": 7.935333034055358e-07,
+      "loss": 0.0573,
+      "step": 9089
+    },
+    {
+      "epoch": 2.065909090909091,
+      "grad_norm": 1.5317010448340034,
+      "learning_rate": 7.934473579334901e-07,
+      "loss": 0.0209,
+      "step": 9090
+    },
+    {
+      "epoch": 2.0661363636363634,
+      "grad_norm": 4.304495947402892,
+      "learning_rate": 7.933614090265089e-07,
+      "loss": 0.0767,
+      "step": 9091
+    },
+    {
+      "epoch": 2.0663636363636364,
+      "grad_norm": 2.5236772974316444,
+      "learning_rate": 7.932754566863449e-07,
+      "loss": 0.0475,
+      "step": 9092
+    },
+    {
+      "epoch": 2.066590909090909,
+      "grad_norm": 2.3900083720616006,
+      "learning_rate": 7.931895009147507e-07,
+      "loss": 0.0245,
+      "step": 9093
+    },
+    {
+      "epoch": 2.066818181818182,
+      "grad_norm": 1.7004866984494125,
+      "learning_rate": 7.931035417134793e-07,
+      "loss": 0.0242,
+      "step": 9094
+    },
+    {
+      "epoch": 2.0670454545454544,
+      "grad_norm": 1.7355323424979718,
+      "learning_rate": 7.930175790842833e-07,
+      "loss": 0.0379,
+      "step": 9095
+    },
+    {
+      "epoch": 2.0672727272727274,
+      "grad_norm": 1.213822668491732,
+      "learning_rate": 7.929316130289158e-07,
+      "loss": 0.0128,
+      "step": 9096
+    },
+    {
+      "epoch": 2.0675,
+      "grad_norm": 2.009455270248327,
+      "learning_rate": 7.928456435491297e-07,
+      "loss": 0.033,
+      "step": 9097
+    },
+    {
+      "epoch": 2.067727272727273,
+      "grad_norm": 1.6193105838499888,
+      "learning_rate": 7.92759670646678e-07,
+      "loss": 0.0259,
+      "step": 9098
+    },
+    {
+      "epoch": 2.0679545454545454,
+      "grad_norm": 1.6899710729144033,
+      "learning_rate": 7.926736943233141e-07,
+      "loss": 0.0172,
+      "step": 9099
+    },
+    {
+      "epoch": 2.0681818181818183,
+      "grad_norm": 1.3498399938352923,
+      "learning_rate": 7.925877145807908e-07,
+      "loss": 0.0221,
+      "step": 9100
+    },
+    {
+      "epoch": 2.068409090909091,
+      "grad_norm": 1.3227340281476458,
+      "learning_rate": 7.925017314208619e-07,
+      "loss": 0.0231,
+      "step": 9101
+    },
+    {
+      "epoch": 2.068636363636364,
+      "grad_norm": 1.7848153375240485,
+      "learning_rate": 7.924157448452803e-07,
+      "loss": 0.0299,
+      "step": 9102
+    },
+    {
+      "epoch": 2.0688636363636363,
+      "grad_norm": 2.566718323430619,
+      "learning_rate": 7.923297548557998e-07,
+      "loss": 0.0348,
+      "step": 9103
+    },
+    {
+      "epoch": 2.0690909090909093,
+      "grad_norm": 0.7624428539970265,
+      "learning_rate": 7.922437614541735e-07,
+      "loss": 0.0057,
+      "step": 9104
+    },
+    {
+      "epoch": 2.069318181818182,
+      "grad_norm": 3.2059878551425216,
+      "learning_rate": 7.921577646421554e-07,
+      "loss": 0.0185,
+      "step": 9105
+    },
+    {
+      "epoch": 2.0695454545454544,
+      "grad_norm": 1.4816220776407463,
+      "learning_rate": 7.920717644214986e-07,
+      "loss": 0.0289,
+      "step": 9106
+    },
+    {
+      "epoch": 2.0697727272727273,
+      "grad_norm": 1.8277229612985495,
+      "learning_rate": 7.91985760793957e-07,
+      "loss": 0.0321,
+      "step": 9107
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.28441363529791,
+      "learning_rate": 7.918997537612844e-07,
+      "loss": 0.0106,
+      "step": 9108
+    },
+    {
+      "epoch": 2.070227272727273,
+      "grad_norm": 1.805494764983086,
+      "learning_rate": 7.918137433252349e-07,
+      "loss": 0.0232,
+      "step": 9109
+    },
+    {
+      "epoch": 2.0704545454545453,
+      "grad_norm": 2.652765955464138,
+      "learning_rate": 7.91727729487562e-07,
+      "loss": 0.0159,
+      "step": 9110
+    },
+    {
+      "epoch": 2.0706818181818183,
+      "grad_norm": 1.247086617983288,
+      "learning_rate": 7.916417122500199e-07,
+      "loss": 0.0177,
+      "step": 9111
+    },
+    {
+      "epoch": 2.070909090909091,
+      "grad_norm": 2.5164898261385584,
+      "learning_rate": 7.915556916143624e-07,
+      "loss": 0.0436,
+      "step": 9112
+    },
+    {
+      "epoch": 2.0711363636363638,
+      "grad_norm": 1.8277626115133676,
+      "learning_rate": 7.914696675823439e-07,
+      "loss": 0.0446,
+      "step": 9113
+    },
+    {
+      "epoch": 2.0713636363636363,
+      "grad_norm": 2.166523999873084,
+      "learning_rate": 7.913836401557183e-07,
+      "loss": 0.0128,
+      "step": 9114
+    },
+    {
+      "epoch": 2.0715909090909093,
+      "grad_norm": 1.3151218651351122,
+      "learning_rate": 7.912976093362402e-07,
+      "loss": 0.0125,
+      "step": 9115
+    },
+    {
+      "epoch": 2.071818181818182,
+      "grad_norm": 1.6770327941706016,
+      "learning_rate": 7.912115751256637e-07,
+      "loss": 0.0173,
+      "step": 9116
+    },
+    {
+      "epoch": 2.0720454545454547,
+      "grad_norm": 2.3438056326162515,
+      "learning_rate": 7.91125537525743e-07,
+      "loss": 0.0489,
+      "step": 9117
+    },
+    {
+      "epoch": 2.0722727272727273,
+      "grad_norm": 2.462313707459912,
+      "learning_rate": 7.910394965382332e-07,
+      "loss": 0.0498,
+      "step": 9118
+    },
+    {
+      "epoch": 2.0725,
+      "grad_norm": 2.2796585054692224,
+      "learning_rate": 7.909534521648879e-07,
+      "loss": 0.0544,
+      "step": 9119
+    },
+    {
+      "epoch": 2.0727272727272728,
+      "grad_norm": 1.9426193793349356,
+      "learning_rate": 7.908674044074625e-07,
+      "loss": 0.0217,
+      "step": 9120
+    },
+    {
+      "epoch": 2.0729545454545453,
+      "grad_norm": 1.9678111563321383,
+      "learning_rate": 7.907813532677111e-07,
+      "loss": 0.0261,
+      "step": 9121
+    },
+    {
+      "epoch": 2.0731818181818182,
+      "grad_norm": 3.206593663145863,
+      "learning_rate": 7.906952987473889e-07,
+      "loss": 0.0629,
+      "step": 9122
+    },
+    {
+      "epoch": 2.0734090909090908,
+      "grad_norm": 1.462903238938699,
+      "learning_rate": 7.906092408482505e-07,
+      "loss": 0.0236,
+      "step": 9123
+    },
+    {
+      "epoch": 2.0736363636363637,
+      "grad_norm": 0.9814436400652117,
+      "learning_rate": 7.905231795720507e-07,
+      "loss": 0.0084,
+      "step": 9124
+    },
+    {
+      "epoch": 2.0738636363636362,
+      "grad_norm": 1.4836718619532092,
+      "learning_rate": 7.904371149205444e-07,
+      "loss": 0.0136,
+      "step": 9125
+    },
+    {
+      "epoch": 2.074090909090909,
+      "grad_norm": 1.3704852230605964,
+      "learning_rate": 7.90351046895487e-07,
+      "loss": 0.0125,
+      "step": 9126
+    },
+    {
+      "epoch": 2.0743181818181817,
+      "grad_norm": 1.8264220937720146,
+      "learning_rate": 7.90264975498633e-07,
+      "loss": 0.0244,
+      "step": 9127
+    },
+    {
+      "epoch": 2.0745454545454547,
+      "grad_norm": 1.4053573786038915,
+      "learning_rate": 7.90178900731738e-07,
+      "loss": 0.031,
+      "step": 9128
+    },
+    {
+      "epoch": 2.074772727272727,
+      "grad_norm": 1.5599878582929658,
+      "learning_rate": 7.900928225965569e-07,
+      "loss": 0.0353,
+      "step": 9129
+    },
+    {
+      "epoch": 2.075,
+      "grad_norm": 0.7146245830138042,
+      "learning_rate": 7.900067410948451e-07,
+      "loss": 0.0108,
+      "step": 9130
+    },
+    {
+      "epoch": 2.0752272727272727,
+      "grad_norm": 2.1838032823131077,
+      "learning_rate": 7.899206562283582e-07,
+      "loss": 0.0196,
+      "step": 9131
+    },
+    {
+      "epoch": 2.0754545454545457,
+      "grad_norm": 1.5619141086842634,
+      "learning_rate": 7.898345679988512e-07,
+      "loss": 0.0202,
+      "step": 9132
+    },
+    {
+      "epoch": 2.075681818181818,
+      "grad_norm": 0.7798730080638594,
+      "learning_rate": 7.8974847640808e-07,
+      "loss": 0.0169,
+      "step": 9133
+    },
+    {
+      "epoch": 2.0759090909090907,
+      "grad_norm": 1.852454414307674,
+      "learning_rate": 7.896623814577998e-07,
+      "loss": 0.0212,
+      "step": 9134
+    },
+    {
+      "epoch": 2.0761363636363637,
+      "grad_norm": 1.1132843975991762,
+      "learning_rate": 7.895762831497662e-07,
+      "loss": 0.0223,
+      "step": 9135
+    },
+    {
+      "epoch": 2.076363636363636,
+      "grad_norm": 1.6237212167750252,
+      "learning_rate": 7.894901814857353e-07,
+      "loss": 0.0196,
+      "step": 9136
+    },
+    {
+      "epoch": 2.076590909090909,
+      "grad_norm": 1.385576295554586,
+      "learning_rate": 7.894040764674625e-07,
+      "loss": 0.0334,
+      "step": 9137
+    },
+    {
+      "epoch": 2.0768181818181817,
+      "grad_norm": 1.8188979314618174,
+      "learning_rate": 7.893179680967039e-07,
+      "loss": 0.0281,
+      "step": 9138
+    },
+    {
+      "epoch": 2.0770454545454546,
+      "grad_norm": 1.177519737566869,
+      "learning_rate": 7.892318563752152e-07,
+      "loss": 0.0145,
+      "step": 9139
+    },
+    {
+      "epoch": 2.077272727272727,
+      "grad_norm": 2.304158201073416,
+      "learning_rate": 7.891457413047525e-07,
+      "loss": 0.0412,
+      "step": 9140
+    },
+    {
+      "epoch": 2.0775,
+      "grad_norm": 1.4489535588798184,
+      "learning_rate": 7.890596228870716e-07,
+      "loss": 0.0228,
+      "step": 9141
+    },
+    {
+      "epoch": 2.0777272727272726,
+      "grad_norm": 1.7292623272886805,
+      "learning_rate": 7.88973501123929e-07,
+      "loss": 0.0202,
+      "step": 9142
+    },
+    {
+      "epoch": 2.0779545454545456,
+      "grad_norm": 2.9678326478101384,
+      "learning_rate": 7.888873760170803e-07,
+      "loss": 0.0347,
+      "step": 9143
+    },
+    {
+      "epoch": 2.078181818181818,
+      "grad_norm": 0.9158976597491179,
+      "learning_rate": 7.888012475682824e-07,
+      "loss": 0.0144,
+      "step": 9144
+    },
+    {
+      "epoch": 2.078409090909091,
+      "grad_norm": 1.2575426406086054,
+      "learning_rate": 7.88715115779291e-07,
+      "loss": 0.0219,
+      "step": 9145
+    },
+    {
+      "epoch": 2.0786363636363636,
+      "grad_norm": 1.6625239102460565,
+      "learning_rate": 7.88628980651863e-07,
+      "loss": 0.0191,
+      "step": 9146
+    },
+    {
+      "epoch": 2.0788636363636366,
+      "grad_norm": 1.1521050973522162,
+      "learning_rate": 7.885428421877547e-07,
+      "loss": 0.0094,
+      "step": 9147
+    },
+    {
+      "epoch": 2.079090909090909,
+      "grad_norm": 1.250145232844473,
+      "learning_rate": 7.884567003887224e-07,
+      "loss": 0.0147,
+      "step": 9148
+    },
+    {
+      "epoch": 2.0793181818181816,
+      "grad_norm": 1.965140930193578,
+      "learning_rate": 7.883705552565229e-07,
+      "loss": 0.0321,
+      "step": 9149
+    },
+    {
+      "epoch": 2.0795454545454546,
+      "grad_norm": 1.773400425634013,
+      "learning_rate": 7.882844067929126e-07,
+      "loss": 0.0292,
+      "step": 9150
+    },
+    {
+      "epoch": 2.079772727272727,
+      "grad_norm": 1.57997547828074,
+      "learning_rate": 7.881982549996485e-07,
+      "loss": 0.0199,
+      "step": 9151
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.9116928678689986,
+      "learning_rate": 7.881120998784872e-07,
+      "loss": 0.0227,
+      "step": 9152
+    },
+    {
+      "epoch": 2.0802272727272726,
+      "grad_norm": 0.8432515821845026,
+      "learning_rate": 7.880259414311857e-07,
+      "loss": 0.0065,
+      "step": 9153
+    },
+    {
+      "epoch": 2.0804545454545456,
+      "grad_norm": 2.3423658056048424,
+      "learning_rate": 7.879397796595008e-07,
+      "loss": 0.0387,
+      "step": 9154
+    },
+    {
+      "epoch": 2.080681818181818,
+      "grad_norm": 1.985394592850441,
+      "learning_rate": 7.878536145651895e-07,
+      "loss": 0.0192,
+      "step": 9155
+    },
+    {
+      "epoch": 2.080909090909091,
+      "grad_norm": 2.1983040938995604,
+      "learning_rate": 7.87767446150009e-07,
+      "loss": 0.0377,
+      "step": 9156
+    },
+    {
+      "epoch": 2.0811363636363636,
+      "grad_norm": 1.4916459141694922,
+      "learning_rate": 7.876812744157161e-07,
+      "loss": 0.0166,
+      "step": 9157
+    },
+    {
+      "epoch": 2.0813636363636365,
+      "grad_norm": 2.069277401432296,
+      "learning_rate": 7.875950993640684e-07,
+      "loss": 0.0343,
+      "step": 9158
+    },
+    {
+      "epoch": 2.081590909090909,
+      "grad_norm": 2.7209068504382485,
+      "learning_rate": 7.875089209968228e-07,
+      "loss": 0.0155,
+      "step": 9159
+    },
+    {
+      "epoch": 2.081818181818182,
+      "grad_norm": 1.3965112056419386,
+      "learning_rate": 7.874227393157369e-07,
+      "loss": 0.0257,
+      "step": 9160
+    },
+    {
+      "epoch": 2.0820454545454545,
+      "grad_norm": 1.1792329274653242,
+      "learning_rate": 7.873365543225678e-07,
+      "loss": 0.0123,
+      "step": 9161
+    },
+    {
+      "epoch": 2.082272727272727,
+      "grad_norm": 2.4877146567633193,
+      "learning_rate": 7.872503660190735e-07,
+      "loss": 0.0269,
+      "step": 9162
+    },
+    {
+      "epoch": 2.0825,
+      "grad_norm": 4.960990275203934,
+      "learning_rate": 7.871641744070109e-07,
+      "loss": 0.077,
+      "step": 9163
+    },
+    {
+      "epoch": 2.0827272727272725,
+      "grad_norm": 3.1378470808807295,
+      "learning_rate": 7.870779794881381e-07,
+      "loss": 0.04,
+      "step": 9164
+    },
+    {
+      "epoch": 2.0829545454545455,
+      "grad_norm": 1.5271331828545918,
+      "learning_rate": 7.869917812642122e-07,
+      "loss": 0.0144,
+      "step": 9165
+    },
+    {
+      "epoch": 2.083181818181818,
+      "grad_norm": 1.3086193281360319,
+      "learning_rate": 7.869055797369915e-07,
+      "loss": 0.0145,
+      "step": 9166
+    },
+    {
+      "epoch": 2.083409090909091,
+      "grad_norm": 1.1126037438119445,
+      "learning_rate": 7.868193749082334e-07,
+      "loss": 0.0094,
+      "step": 9167
+    },
+    {
+      "epoch": 2.0836363636363635,
+      "grad_norm": 1.569799085352959,
+      "learning_rate": 7.86733166779696e-07,
+      "loss": 0.0246,
+      "step": 9168
+    },
+    {
+      "epoch": 2.0838636363636365,
+      "grad_norm": 2.093094014541389,
+      "learning_rate": 7.866469553531373e-07,
+      "loss": 0.016,
+      "step": 9169
+    },
+    {
+      "epoch": 2.084090909090909,
+      "grad_norm": 1.3945980385519934,
+      "learning_rate": 7.865607406303151e-07,
+      "loss": 0.0197,
+      "step": 9170
+    },
+    {
+      "epoch": 2.084318181818182,
+      "grad_norm": 1.8362500879852859,
+      "learning_rate": 7.864745226129877e-07,
+      "loss": 0.0213,
+      "step": 9171
+    },
+    {
+      "epoch": 2.0845454545454545,
+      "grad_norm": 1.6168658364225363,
+      "learning_rate": 7.863883013029128e-07,
+      "loss": 0.0157,
+      "step": 9172
+    },
+    {
+      "epoch": 2.0847727272727274,
+      "grad_norm": 2.7619780367951434,
+      "learning_rate": 7.863020767018491e-07,
+      "loss": 0.0353,
+      "step": 9173
+    },
+    {
+      "epoch": 2.085,
+      "grad_norm": 2.1459100005540233,
+      "learning_rate": 7.862158488115545e-07,
+      "loss": 0.0255,
+      "step": 9174
+    },
+    {
+      "epoch": 2.085227272727273,
+      "grad_norm": 1.8448513953664736,
+      "learning_rate": 7.861296176337875e-07,
+      "loss": 0.0166,
+      "step": 9175
+    },
+    {
+      "epoch": 2.0854545454545454,
+      "grad_norm": 2.232041057402966,
+      "learning_rate": 7.860433831703066e-07,
+      "loss": 0.0429,
+      "step": 9176
+    },
+    {
+      "epoch": 2.085681818181818,
+      "grad_norm": 1.5024941141977946,
+      "learning_rate": 7.859571454228701e-07,
+      "loss": 0.0253,
+      "step": 9177
+    },
+    {
+      "epoch": 2.085909090909091,
+      "grad_norm": 1.5398010738730155,
+      "learning_rate": 7.858709043932365e-07,
+      "loss": 0.0241,
+      "step": 9178
+    },
+    {
+      "epoch": 2.0861363636363635,
+      "grad_norm": 2.018326411406427,
+      "learning_rate": 7.857846600831646e-07,
+      "loss": 0.0214,
+      "step": 9179
+    },
+    {
+      "epoch": 2.0863636363636364,
+      "grad_norm": 2.07060996270715,
+      "learning_rate": 7.856984124944129e-07,
+      "loss": 0.0219,
+      "step": 9180
+    },
+    {
+      "epoch": 2.086590909090909,
+      "grad_norm": 3.309347737476696,
+      "learning_rate": 7.856121616287403e-07,
+      "loss": 0.0784,
+      "step": 9181
+    },
+    {
+      "epoch": 2.086818181818182,
+      "grad_norm": 1.408450102232457,
+      "learning_rate": 7.855259074879054e-07,
+      "loss": 0.0132,
+      "step": 9182
+    },
+    {
+      "epoch": 2.0870454545454544,
+      "grad_norm": 1.4499146002662882,
+      "learning_rate": 7.854396500736673e-07,
+      "loss": 0.0118,
+      "step": 9183
+    },
+    {
+      "epoch": 2.0872727272727274,
+      "grad_norm": 2.7741807268081944,
+      "learning_rate": 7.853533893877849e-07,
+      "loss": 0.0342,
+      "step": 9184
+    },
+    {
+      "epoch": 2.0875,
+      "grad_norm": 2.326600290625171,
+      "learning_rate": 7.852671254320171e-07,
+      "loss": 0.0166,
+      "step": 9185
+    },
+    {
+      "epoch": 2.087727272727273,
+      "grad_norm": 0.7726838849900746,
+      "learning_rate": 7.85180858208123e-07,
+      "loss": 0.0088,
+      "step": 9186
+    },
+    {
+      "epoch": 2.0879545454545454,
+      "grad_norm": 0.8026774702211673,
+      "learning_rate": 7.850945877178618e-07,
+      "loss": 0.0063,
+      "step": 9187
+    },
+    {
+      "epoch": 2.0881818181818184,
+      "grad_norm": 2.0574351637299433,
+      "learning_rate": 7.850083139629926e-07,
+      "loss": 0.0351,
+      "step": 9188
+    },
+    {
+      "epoch": 2.088409090909091,
+      "grad_norm": 1.7898650187724963,
+      "learning_rate": 7.849220369452747e-07,
+      "loss": 0.0173,
+      "step": 9189
+    },
+    {
+      "epoch": 2.088636363636364,
+      "grad_norm": 4.619301658685381,
+      "learning_rate": 7.848357566664674e-07,
+      "loss": 0.0853,
+      "step": 9190
+    },
+    {
+      "epoch": 2.0888636363636364,
+      "grad_norm": 1.7455517694308795,
+      "learning_rate": 7.847494731283302e-07,
+      "loss": 0.0284,
+      "step": 9191
+    },
+    {
+      "epoch": 2.089090909090909,
+      "grad_norm": 1.5852232380666396,
+      "learning_rate": 7.846631863326226e-07,
+      "loss": 0.0275,
+      "step": 9192
+    },
+    {
+      "epoch": 2.089318181818182,
+      "grad_norm": 1.6551393159385186,
+      "learning_rate": 7.845768962811041e-07,
+      "loss": 0.0135,
+      "step": 9193
+    },
+    {
+      "epoch": 2.0895454545454544,
+      "grad_norm": 2.5659299094532177,
+      "learning_rate": 7.844906029755342e-07,
+      "loss": 0.0155,
+      "step": 9194
+    },
+    {
+      "epoch": 2.0897727272727273,
+      "grad_norm": 1.3563287711268588,
+      "learning_rate": 7.844043064176729e-07,
+      "loss": 0.0141,
+      "step": 9195
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.763396207190361,
+      "learning_rate": 7.843180066092795e-07,
+      "loss": 0.0077,
+      "step": 9196
+    },
+    {
+      "epoch": 2.090227272727273,
+      "grad_norm": 2.6694908751859767,
+      "learning_rate": 7.842317035521139e-07,
+      "loss": 0.0405,
+      "step": 9197
+    },
+    {
+      "epoch": 2.0904545454545453,
+      "grad_norm": 1.6896486992929185,
+      "learning_rate": 7.841453972479363e-07,
+      "loss": 0.0304,
+      "step": 9198
+    },
+    {
+      "epoch": 2.0906818181818183,
+      "grad_norm": 1.9952912271818823,
+      "learning_rate": 7.840590876985063e-07,
+      "loss": 0.0217,
+      "step": 9199
+    },
+    {
+      "epoch": 2.090909090909091,
+      "grad_norm": 1.8283572535581507,
+      "learning_rate": 7.839727749055839e-07,
+      "loss": 0.0302,
+      "step": 9200
+    },
+    {
+      "epoch": 2.091136363636364,
+      "grad_norm": 2.626198082804542,
+      "learning_rate": 7.838864588709296e-07,
+      "loss": 0.0341,
+      "step": 9201
+    },
+    {
+      "epoch": 2.0913636363636363,
+      "grad_norm": 1.1872115310531208,
+      "learning_rate": 7.838001395963028e-07,
+      "loss": 0.0148,
+      "step": 9202
+    },
+    {
+      "epoch": 2.0915909090909093,
+      "grad_norm": 0.8484911733752426,
+      "learning_rate": 7.837138170834643e-07,
+      "loss": 0.0062,
+      "step": 9203
+    },
+    {
+      "epoch": 2.091818181818182,
+      "grad_norm": 1.6269996369721136,
+      "learning_rate": 7.836274913341743e-07,
+      "loss": 0.0216,
+      "step": 9204
+    },
+    {
+      "epoch": 2.0920454545454543,
+      "grad_norm": 1.331177700128449,
+      "learning_rate": 7.835411623501927e-07,
+      "loss": 0.014,
+      "step": 9205
+    },
+    {
+      "epoch": 2.0922727272727273,
+      "grad_norm": 1.1020516790168764,
+      "learning_rate": 7.834548301332805e-07,
+      "loss": 0.0071,
+      "step": 9206
+    },
+    {
+      "epoch": 2.0925,
+      "grad_norm": 3.1227033985798287,
+      "learning_rate": 7.833684946851977e-07,
+      "loss": 0.0596,
+      "step": 9207
+    },
+    {
+      "epoch": 2.0927272727272728,
+      "grad_norm": 1.809322948026916,
+      "learning_rate": 7.832821560077052e-07,
+      "loss": 0.0192,
+      "step": 9208
+    },
+    {
+      "epoch": 2.0929545454545453,
+      "grad_norm": 1.502525498307183,
+      "learning_rate": 7.831958141025632e-07,
+      "loss": 0.0155,
+      "step": 9209
+    },
+    {
+      "epoch": 2.0931818181818183,
+      "grad_norm": 2.422379855458984,
+      "learning_rate": 7.831094689715328e-07,
+      "loss": 0.0163,
+      "step": 9210
+    },
+    {
+      "epoch": 2.0934090909090908,
+      "grad_norm": 1.922210618644154,
+      "learning_rate": 7.830231206163742e-07,
+      "loss": 0.0128,
+      "step": 9211
+    },
+    {
+      "epoch": 2.0936363636363637,
+      "grad_norm": 2.2762185705095233,
+      "learning_rate": 7.829367690388487e-07,
+      "loss": 0.0539,
+      "step": 9212
+    },
+    {
+      "epoch": 2.0938636363636363,
+      "grad_norm": 3.4913087043313453,
+      "learning_rate": 7.828504142407168e-07,
+      "loss": 0.0368,
+      "step": 9213
+    },
+    {
+      "epoch": 2.0940909090909092,
+      "grad_norm": 1.5810169707402406,
+      "learning_rate": 7.827640562237396e-07,
+      "loss": 0.0221,
+      "step": 9214
+    },
+    {
+      "epoch": 2.0943181818181817,
+      "grad_norm": 1.79344279543514,
+      "learning_rate": 7.826776949896781e-07,
+      "loss": 0.0209,
+      "step": 9215
+    },
+    {
+      "epoch": 2.0945454545454547,
+      "grad_norm": 2.4579423722778175,
+      "learning_rate": 7.825913305402932e-07,
+      "loss": 0.0142,
+      "step": 9216
+    },
+    {
+      "epoch": 2.0947727272727272,
+      "grad_norm": 2.5522878485492733,
+      "learning_rate": 7.825049628773464e-07,
+      "loss": 0.029,
+      "step": 9217
+    },
+    {
+      "epoch": 2.095,
+      "grad_norm": 0.6768194168688795,
+      "learning_rate": 7.824185920025984e-07,
+      "loss": 0.005,
+      "step": 9218
+    },
+    {
+      "epoch": 2.0952272727272727,
+      "grad_norm": 2.655423384768992,
+      "learning_rate": 7.823322179178108e-07,
+      "loss": 0.0215,
+      "step": 9219
+    },
+    {
+      "epoch": 2.0954545454545457,
+      "grad_norm": 6.213970316576767,
+      "learning_rate": 7.822458406247447e-07,
+      "loss": 0.0327,
+      "step": 9220
+    },
+    {
+      "epoch": 2.095681818181818,
+      "grad_norm": 1.4268958456326482,
+      "learning_rate": 7.821594601251617e-07,
+      "loss": 0.0144,
+      "step": 9221
+    },
+    {
+      "epoch": 2.0959090909090907,
+      "grad_norm": 1.77110653049449,
+      "learning_rate": 7.820730764208229e-07,
+      "loss": 0.0133,
+      "step": 9222
+    },
+    {
+      "epoch": 2.0961363636363637,
+      "grad_norm": 2.426785046787907,
+      "learning_rate": 7.819866895134903e-07,
+      "loss": 0.0315,
+      "step": 9223
+    },
+    {
+      "epoch": 2.096363636363636,
+      "grad_norm": 2.004076312284428,
+      "learning_rate": 7.819002994049252e-07,
+      "loss": 0.0208,
+      "step": 9224
+    },
+    {
+      "epoch": 2.096590909090909,
+      "grad_norm": 2.6462414464402713,
+      "learning_rate": 7.81813906096889e-07,
+      "loss": 0.0308,
+      "step": 9225
+    },
+    {
+      "epoch": 2.0968181818181817,
+      "grad_norm": 1.4639556552220208,
+      "learning_rate": 7.817275095911439e-07,
+      "loss": 0.0172,
+      "step": 9226
+    },
+    {
+      "epoch": 2.0970454545454547,
+      "grad_norm": 1.1853555708489225,
+      "learning_rate": 7.816411098894515e-07,
+      "loss": 0.0178,
+      "step": 9227
+    },
+    {
+      "epoch": 2.097272727272727,
+      "grad_norm": 0.7028744577232585,
+      "learning_rate": 7.815547069935733e-07,
+      "loss": 0.0132,
+      "step": 9228
+    },
+    {
+      "epoch": 2.0975,
+      "grad_norm": 1.4817194426188935,
+      "learning_rate": 7.814683009052718e-07,
+      "loss": 0.0194,
+      "step": 9229
+    },
+    {
+      "epoch": 2.0977272727272727,
+      "grad_norm": 1.828614956182983,
+      "learning_rate": 7.813818916263086e-07,
+      "loss": 0.0278,
+      "step": 9230
+    },
+    {
+      "epoch": 2.0979545454545456,
+      "grad_norm": 1.3817999013232374,
+      "learning_rate": 7.812954791584458e-07,
+      "loss": 0.0153,
+      "step": 9231
+    },
+    {
+      "epoch": 2.098181818181818,
+      "grad_norm": 1.4560799763795234,
+      "learning_rate": 7.812090635034456e-07,
+      "loss": 0.0275,
+      "step": 9232
+    },
+    {
+      "epoch": 2.098409090909091,
+      "grad_norm": 2.71472744372923,
+      "learning_rate": 7.811226446630699e-07,
+      "loss": 0.0566,
+      "step": 9233
+    },
+    {
+      "epoch": 2.0986363636363636,
+      "grad_norm": 1.3054583811025349,
+      "learning_rate": 7.810362226390812e-07,
+      "loss": 0.0166,
+      "step": 9234
+    },
+    {
+      "epoch": 2.098863636363636,
+      "grad_norm": 1.9118696627754945,
+      "learning_rate": 7.809497974332415e-07,
+      "loss": 0.017,
+      "step": 9235
+    },
+    {
+      "epoch": 2.099090909090909,
+      "grad_norm": 1.9733831894600748,
+      "learning_rate": 7.808633690473136e-07,
+      "loss": 0.0281,
+      "step": 9236
+    },
+    {
+      "epoch": 2.0993181818181816,
+      "grad_norm": 1.5496999011054644,
+      "learning_rate": 7.807769374830596e-07,
+      "loss": 0.0097,
+      "step": 9237
+    },
+    {
+      "epoch": 2.0995454545454546,
+      "grad_norm": 3.208187816136752,
+      "learning_rate": 7.806905027422421e-07,
+      "loss": 0.0958,
+      "step": 9238
+    },
+    {
+      "epoch": 2.099772727272727,
+      "grad_norm": 2.911629372927346,
+      "learning_rate": 7.806040648266237e-07,
+      "loss": 0.0456,
+      "step": 9239
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.502819244643515,
+      "learning_rate": 7.805176237379668e-07,
+      "loss": 0.0198,
+      "step": 9240
+    },
+    {
+      "epoch": 2.1002272727272726,
+      "grad_norm": 2.905485244689414,
+      "learning_rate": 7.804311794780345e-07,
+      "loss": 0.0425,
+      "step": 9241
+    },
+    {
+      "epoch": 2.1004545454545456,
+      "grad_norm": 1.8576170680538338,
+      "learning_rate": 7.803447320485889e-07,
+      "loss": 0.0366,
+      "step": 9242
+    },
+    {
+      "epoch": 2.100681818181818,
+      "grad_norm": 1.9941688477815238,
+      "learning_rate": 7.802582814513934e-07,
+      "loss": 0.0212,
+      "step": 9243
+    },
+    {
+      "epoch": 2.100909090909091,
+      "grad_norm": 2.398310177576153,
+      "learning_rate": 7.801718276882106e-07,
+      "loss": 0.0467,
+      "step": 9244
+    },
+    {
+      "epoch": 2.1011363636363636,
+      "grad_norm": 1.3320725381767868,
+      "learning_rate": 7.800853707608038e-07,
+      "loss": 0.0207,
+      "step": 9245
+    },
+    {
+      "epoch": 2.1013636363636365,
+      "grad_norm": 2.893830984981929,
+      "learning_rate": 7.799989106709354e-07,
+      "loss": 0.032,
+      "step": 9246
+    },
+    {
+      "epoch": 2.101590909090909,
+      "grad_norm": 1.544731793207706,
+      "learning_rate": 7.79912447420369e-07,
+      "loss": 0.0206,
+      "step": 9247
+    },
+    {
+      "epoch": 2.101818181818182,
+      "grad_norm": 1.4906129811021203,
+      "learning_rate": 7.798259810108675e-07,
+      "loss": 0.0265,
+      "step": 9248
+    },
+    {
+      "epoch": 2.1020454545454546,
+      "grad_norm": 2.6589707248993935,
+      "learning_rate": 7.79739511444194e-07,
+      "loss": 0.0284,
+      "step": 9249
+    },
+    {
+      "epoch": 2.102272727272727,
+      "grad_norm": 5.5287510584327615,
+      "learning_rate": 7.79653038722112e-07,
+      "loss": 0.053,
+      "step": 9250
+    },
+    {
+      "epoch": 2.1025,
+      "grad_norm": 2.5045576762252932,
+      "learning_rate": 7.795665628463847e-07,
+      "loss": 0.0269,
+      "step": 9251
+    },
+    {
+      "epoch": 2.1027272727272726,
+      "grad_norm": 2.76544840764494,
+      "learning_rate": 7.794800838187756e-07,
+      "loss": 0.0544,
+      "step": 9252
+    },
+    {
+      "epoch": 2.1029545454545455,
+      "grad_norm": 2.058472752627887,
+      "learning_rate": 7.793936016410481e-07,
+      "loss": 0.0403,
+      "step": 9253
+    },
+    {
+      "epoch": 2.103181818181818,
+      "grad_norm": 1.6311927795604595,
+      "learning_rate": 7.793071163149657e-07,
+      "loss": 0.0191,
+      "step": 9254
+    },
+    {
+      "epoch": 2.103409090909091,
+      "grad_norm": 1.1450462471501555,
+      "learning_rate": 7.792206278422918e-07,
+      "loss": 0.0148,
+      "step": 9255
+    },
+    {
+      "epoch": 2.1036363636363635,
+      "grad_norm": 1.3730599888934079,
+      "learning_rate": 7.791341362247906e-07,
+      "loss": 0.0163,
+      "step": 9256
+    },
+    {
+      "epoch": 2.1038636363636365,
+      "grad_norm": 1.7435389908507433,
+      "learning_rate": 7.790476414642252e-07,
+      "loss": 0.0251,
+      "step": 9257
+    },
+    {
+      "epoch": 2.104090909090909,
+      "grad_norm": 1.4923529147793206,
+      "learning_rate": 7.789611435623596e-07,
+      "loss": 0.0328,
+      "step": 9258
+    },
+    {
+      "epoch": 2.104318181818182,
+      "grad_norm": 0.8324045245610106,
+      "learning_rate": 7.788746425209579e-07,
+      "loss": 0.0193,
+      "step": 9259
+    },
+    {
+      "epoch": 2.1045454545454545,
+      "grad_norm": 1.660082286705851,
+      "learning_rate": 7.787881383417836e-07,
+      "loss": 0.0204,
+      "step": 9260
+    },
+    {
+      "epoch": 2.1047727272727275,
+      "grad_norm": 1.8395088267505182,
+      "learning_rate": 7.78701631026601e-07,
+      "loss": 0.0241,
+      "step": 9261
+    },
+    {
+      "epoch": 2.105,
+      "grad_norm": 1.850577740692887,
+      "learning_rate": 7.786151205771739e-07,
+      "loss": 0.0211,
+      "step": 9262
+    },
+    {
+      "epoch": 2.105227272727273,
+      "grad_norm": 2.817511311909845,
+      "learning_rate": 7.785286069952666e-07,
+      "loss": 0.0257,
+      "step": 9263
+    },
+    {
+      "epoch": 2.1054545454545455,
+      "grad_norm": 1.4547877506993565,
+      "learning_rate": 7.784420902826431e-07,
+      "loss": 0.0155,
+      "step": 9264
+    },
+    {
+      "epoch": 2.105681818181818,
+      "grad_norm": 3.2748660164657672,
+      "learning_rate": 7.783555704410675e-07,
+      "loss": 0.0259,
+      "step": 9265
+    },
+    {
+      "epoch": 2.105909090909091,
+      "grad_norm": 0.92770649017058,
+      "learning_rate": 7.782690474723045e-07,
+      "loss": 0.0077,
+      "step": 9266
+    },
+    {
+      "epoch": 2.1061363636363635,
+      "grad_norm": 1.418781031871454,
+      "learning_rate": 7.781825213781183e-07,
+      "loss": 0.0177,
+      "step": 9267
+    },
+    {
+      "epoch": 2.1063636363636364,
+      "grad_norm": 2.45361196089032,
+      "learning_rate": 7.780959921602731e-07,
+      "loss": 0.0373,
+      "step": 9268
+    },
+    {
+      "epoch": 2.106590909090909,
+      "grad_norm": 3.5742266831272507,
+      "learning_rate": 7.780094598205334e-07,
+      "loss": 0.0179,
+      "step": 9269
+    },
+    {
+      "epoch": 2.106818181818182,
+      "grad_norm": 2.0567336228720396,
+      "learning_rate": 7.779229243606642e-07,
+      "loss": 0.0232,
+      "step": 9270
+    },
+    {
+      "epoch": 2.1070454545454544,
+      "grad_norm": 4.865924133883742,
+      "learning_rate": 7.778363857824295e-07,
+      "loss": 0.0366,
+      "step": 9271
+    },
+    {
+      "epoch": 2.1072727272727274,
+      "grad_norm": 2.120005624224915,
+      "learning_rate": 7.777498440875946e-07,
+      "loss": 0.0267,
+      "step": 9272
+    },
+    {
+      "epoch": 2.1075,
+      "grad_norm": 1.72347317132233,
+      "learning_rate": 7.776632992779234e-07,
+      "loss": 0.0094,
+      "step": 9273
+    },
+    {
+      "epoch": 2.107727272727273,
+      "grad_norm": 1.2420760336957566,
+      "learning_rate": 7.775767513551816e-07,
+      "loss": 0.0204,
+      "step": 9274
+    },
+    {
+      "epoch": 2.1079545454545454,
+      "grad_norm": 1.5194935624545276,
+      "learning_rate": 7.774902003211334e-07,
+      "loss": 0.0103,
+      "step": 9275
+    },
+    {
+      "epoch": 2.1081818181818184,
+      "grad_norm": 1.5321205238096365,
+      "learning_rate": 7.774036461775442e-07,
+      "loss": 0.0196,
+      "step": 9276
+    },
+    {
+      "epoch": 2.108409090909091,
+      "grad_norm": 1.51893760092697,
+      "learning_rate": 7.773170889261785e-07,
+      "loss": 0.0176,
+      "step": 9277
+    },
+    {
+      "epoch": 2.1086363636363634,
+      "grad_norm": 1.6501027059482547,
+      "learning_rate": 7.77230528568802e-07,
+      "loss": 0.0356,
+      "step": 9278
+    },
+    {
+      "epoch": 2.1088636363636364,
+      "grad_norm": 1.4107736513587112,
+      "learning_rate": 7.771439651071791e-07,
+      "loss": 0.0214,
+      "step": 9279
+    },
+    {
+      "epoch": 2.109090909090909,
+      "grad_norm": 3.148798257894106,
+      "learning_rate": 7.770573985430755e-07,
+      "loss": 0.0501,
+      "step": 9280
+    },
+    {
+      "epoch": 2.109318181818182,
+      "grad_norm": 2.1677880635946174,
+      "learning_rate": 7.769708288782562e-07,
+      "loss": 0.0421,
+      "step": 9281
+    },
+    {
+      "epoch": 2.1095454545454544,
+      "grad_norm": 3.7110665449968474,
+      "learning_rate": 7.768842561144864e-07,
+      "loss": 0.0306,
+      "step": 9282
+    },
+    {
+      "epoch": 2.1097727272727274,
+      "grad_norm": 2.462908469321759,
+      "learning_rate": 7.76797680253532e-07,
+      "loss": 0.066,
+      "step": 9283
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 3.3593642613243517,
+      "learning_rate": 7.767111012971578e-07,
+      "loss": 0.0403,
+      "step": 9284
+    },
+    {
+      "epoch": 2.110227272727273,
+      "grad_norm": 1.9340834705923202,
+      "learning_rate": 7.766245192471297e-07,
+      "loss": 0.0153,
+      "step": 9285
+    },
+    {
+      "epoch": 2.1104545454545454,
+      "grad_norm": 2.499818521012918,
+      "learning_rate": 7.765379341052132e-07,
+      "loss": 0.0159,
+      "step": 9286
+    },
+    {
+      "epoch": 2.1106818181818183,
+      "grad_norm": 1.1335500486162078,
+      "learning_rate": 7.764513458731738e-07,
+      "loss": 0.0101,
+      "step": 9287
+    },
+    {
+      "epoch": 2.110909090909091,
+      "grad_norm": 2.199355286222464,
+      "learning_rate": 7.76364754552777e-07,
+      "loss": 0.0256,
+      "step": 9288
+    },
+    {
+      "epoch": 2.111136363636364,
+      "grad_norm": 1.2674417164526328,
+      "learning_rate": 7.76278160145789e-07,
+      "loss": 0.0147,
+      "step": 9289
+    },
+    {
+      "epoch": 2.1113636363636363,
+      "grad_norm": 2.260459153487845,
+      "learning_rate": 7.761915626539754e-07,
+      "loss": 0.016,
+      "step": 9290
+    },
+    {
+      "epoch": 2.1115909090909093,
+      "grad_norm": 1.2513912961001314,
+      "learning_rate": 7.761049620791019e-07,
+      "loss": 0.0128,
+      "step": 9291
+    },
+    {
+      "epoch": 2.111818181818182,
+      "grad_norm": 2.329927488258217,
+      "learning_rate": 7.760183584229348e-07,
+      "loss": 0.0218,
+      "step": 9292
+    },
+    {
+      "epoch": 2.1120454545454543,
+      "grad_norm": 0.8781849542444989,
+      "learning_rate": 7.759317516872399e-07,
+      "loss": 0.0231,
+      "step": 9293
+    },
+    {
+      "epoch": 2.1122727272727273,
+      "grad_norm": 2.283910784159098,
+      "learning_rate": 7.758451418737833e-07,
+      "loss": 0.0155,
+      "step": 9294
+    },
+    {
+      "epoch": 2.1125,
+      "grad_norm": 1.3444337457847946,
+      "learning_rate": 7.75758528984331e-07,
+      "loss": 0.03,
+      "step": 9295
+    },
+    {
+      "epoch": 2.112727272727273,
+      "grad_norm": 1.6857885026870454,
+      "learning_rate": 7.756719130206493e-07,
+      "loss": 0.0371,
+      "step": 9296
+    },
+    {
+      "epoch": 2.1129545454545453,
+      "grad_norm": 1.8649735097727278,
+      "learning_rate": 7.755852939845043e-07,
+      "loss": 0.0289,
+      "step": 9297
+    },
+    {
+      "epoch": 2.1131818181818183,
+      "grad_norm": 1.7456500324115949,
+      "learning_rate": 7.754986718776626e-07,
+      "loss": 0.0282,
+      "step": 9298
+    },
+    {
+      "epoch": 2.113409090909091,
+      "grad_norm": 1.2374331977849857,
+      "learning_rate": 7.754120467018903e-07,
+      "loss": 0.0068,
+      "step": 9299
+    },
+    {
+      "epoch": 2.1136363636363638,
+      "grad_norm": 1.9978627920244965,
+      "learning_rate": 7.753254184589542e-07,
+      "loss": 0.0403,
+      "step": 9300
+    },
+    {
+      "epoch": 2.1138636363636363,
+      "grad_norm": 1.6105200779007887,
+      "learning_rate": 7.752387871506203e-07,
+      "loss": 0.015,
+      "step": 9301
+    },
+    {
+      "epoch": 2.1140909090909092,
+      "grad_norm": 2.302286572767389,
+      "learning_rate": 7.751521527786556e-07,
+      "loss": 0.0153,
+      "step": 9302
+    },
+    {
+      "epoch": 2.1143181818181818,
+      "grad_norm": 1.8046526444318962,
+      "learning_rate": 7.750655153448265e-07,
+      "loss": 0.0163,
+      "step": 9303
+    },
+    {
+      "epoch": 2.1145454545454547,
+      "grad_norm": 1.129073852879504,
+      "learning_rate": 7.749788748508997e-07,
+      "loss": 0.0146,
+      "step": 9304
+    },
+    {
+      "epoch": 2.1147727272727272,
+      "grad_norm": 1.5892465399463276,
+      "learning_rate": 7.748922312986421e-07,
+      "loss": 0.0165,
+      "step": 9305
+    },
+    {
+      "epoch": 2.115,
+      "grad_norm": 1.5359923895648941,
+      "learning_rate": 7.748055846898201e-07,
+      "loss": 0.0262,
+      "step": 9306
+    },
+    {
+      "epoch": 2.1152272727272727,
+      "grad_norm": 1.9437719201699803,
+      "learning_rate": 7.747189350262013e-07,
+      "loss": 0.0125,
+      "step": 9307
+    },
+    {
+      "epoch": 2.1154545454545453,
+      "grad_norm": 0.7980996455148012,
+      "learning_rate": 7.74632282309552e-07,
+      "loss": 0.0135,
+      "step": 9308
+    },
+    {
+      "epoch": 2.115681818181818,
+      "grad_norm": 2.8530294278770083,
+      "learning_rate": 7.745456265416394e-07,
+      "loss": 0.0401,
+      "step": 9309
+    },
+    {
+      "epoch": 2.1159090909090907,
+      "grad_norm": 2.6836590647382605,
+      "learning_rate": 7.744589677242306e-07,
+      "loss": 0.023,
+      "step": 9310
+    },
+    {
+      "epoch": 2.1161363636363637,
+      "grad_norm": 0.9883494202927874,
+      "learning_rate": 7.743723058590926e-07,
+      "loss": 0.0152,
+      "step": 9311
+    },
+    {
+      "epoch": 2.1163636363636362,
+      "grad_norm": 3.7979340922500557,
+      "learning_rate": 7.742856409479928e-07,
+      "loss": 0.0213,
+      "step": 9312
+    },
+    {
+      "epoch": 2.116590909090909,
+      "grad_norm": 2.138517830795839,
+      "learning_rate": 7.741989729926983e-07,
+      "loss": 0.0268,
+      "step": 9313
+    },
+    {
+      "epoch": 2.1168181818181817,
+      "grad_norm": 1.5857190097288645,
+      "learning_rate": 7.741123019949765e-07,
+      "loss": 0.0124,
+      "step": 9314
+    },
+    {
+      "epoch": 2.1170454545454547,
+      "grad_norm": 3.6636577304673024,
+      "learning_rate": 7.740256279565947e-07,
+      "loss": 0.0273,
+      "step": 9315
+    },
+    {
+      "epoch": 2.117272727272727,
+      "grad_norm": 1.6992912985974338,
+      "learning_rate": 7.739389508793204e-07,
+      "loss": 0.0204,
+      "step": 9316
+    },
+    {
+      "epoch": 2.1175,
+      "grad_norm": 3.1606599306203886,
+      "learning_rate": 7.738522707649211e-07,
+      "loss": 0.0374,
+      "step": 9317
+    },
+    {
+      "epoch": 2.1177272727272727,
+      "grad_norm": 2.1396626296764194,
+      "learning_rate": 7.737655876151641e-07,
+      "loss": 0.0125,
+      "step": 9318
+    },
+    {
+      "epoch": 2.1179545454545456,
+      "grad_norm": 1.5666184318036531,
+      "learning_rate": 7.736789014318173e-07,
+      "loss": 0.0211,
+      "step": 9319
+    },
+    {
+      "epoch": 2.118181818181818,
+      "grad_norm": 2.198859830516765,
+      "learning_rate": 7.735922122166485e-07,
+      "loss": 0.0216,
+      "step": 9320
+    },
+    {
+      "epoch": 2.1184090909090907,
+      "grad_norm": 4.028353602733989,
+      "learning_rate": 7.73505519971425e-07,
+      "loss": 0.0467,
+      "step": 9321
+    },
+    {
+      "epoch": 2.1186363636363637,
+      "grad_norm": 2.998476171913057,
+      "learning_rate": 7.734188246979152e-07,
+      "loss": 0.0312,
+      "step": 9322
+    },
+    {
+      "epoch": 2.118863636363636,
+      "grad_norm": 2.304011504397092,
+      "learning_rate": 7.733321263978862e-07,
+      "loss": 0.0354,
+      "step": 9323
+    },
+    {
+      "epoch": 2.119090909090909,
+      "grad_norm": 0.7418002915244701,
+      "learning_rate": 7.732454250731068e-07,
+      "loss": 0.0072,
+      "step": 9324
+    },
+    {
+      "epoch": 2.1193181818181817,
+      "grad_norm": 3.347105861941187,
+      "learning_rate": 7.731587207253444e-07,
+      "loss": 0.0403,
+      "step": 9325
+    },
+    {
+      "epoch": 2.1195454545454546,
+      "grad_norm": 2.146144014427624,
+      "learning_rate": 7.730720133563671e-07,
+      "loss": 0.0345,
+      "step": 9326
+    },
+    {
+      "epoch": 2.119772727272727,
+      "grad_norm": 4.109113453615138,
+      "learning_rate": 7.729853029679431e-07,
+      "loss": 0.0612,
+      "step": 9327
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.4392441261184397,
+      "learning_rate": 7.728985895618406e-07,
+      "loss": 0.0343,
+      "step": 9328
+    },
+    {
+      "epoch": 2.1202272727272726,
+      "grad_norm": 1.9302975718192947,
+      "learning_rate": 7.728118731398281e-07,
+      "loss": 0.0095,
+      "step": 9329
+    },
+    {
+      "epoch": 2.1204545454545456,
+      "grad_norm": 1.217351724169575,
+      "learning_rate": 7.727251537036734e-07,
+      "loss": 0.0181,
+      "step": 9330
+    },
+    {
+      "epoch": 2.120681818181818,
+      "grad_norm": 1.3620500314741062,
+      "learning_rate": 7.726384312551452e-07,
+      "loss": 0.0152,
+      "step": 9331
+    },
+    {
+      "epoch": 2.120909090909091,
+      "grad_norm": 3.1177909526949965,
+      "learning_rate": 7.725517057960117e-07,
+      "loss": 0.0588,
+      "step": 9332
+    },
+    {
+      "epoch": 2.1211363636363636,
+      "grad_norm": 1.7968797618123244,
+      "learning_rate": 7.724649773280416e-07,
+      "loss": 0.0305,
+      "step": 9333
+    },
+    {
+      "epoch": 2.1213636363636366,
+      "grad_norm": 1.2428222908024664,
+      "learning_rate": 7.723782458530031e-07,
+      "loss": 0.0156,
+      "step": 9334
+    },
+    {
+      "epoch": 2.121590909090909,
+      "grad_norm": 2.1942679242205467,
+      "learning_rate": 7.722915113726653e-07,
+      "loss": 0.035,
+      "step": 9335
+    },
+    {
+      "epoch": 2.1218181818181816,
+      "grad_norm": 1.5673578194327031,
+      "learning_rate": 7.722047738887967e-07,
+      "loss": 0.0183,
+      "step": 9336
+    },
+    {
+      "epoch": 2.1220454545454546,
+      "grad_norm": 2.0873626079630525,
+      "learning_rate": 7.721180334031657e-07,
+      "loss": 0.0292,
+      "step": 9337
+    },
+    {
+      "epoch": 2.122272727272727,
+      "grad_norm": 1.6025148496180852,
+      "learning_rate": 7.720312899175415e-07,
+      "loss": 0.019,
+      "step": 9338
+    },
+    {
+      "epoch": 2.1225,
+      "grad_norm": 1.5299780239950609,
+      "learning_rate": 7.719445434336928e-07,
+      "loss": 0.0284,
+      "step": 9339
+    },
+    {
+      "epoch": 2.1227272727272726,
+      "grad_norm": 1.734170338514141,
+      "learning_rate": 7.718577939533885e-07,
+      "loss": 0.0238,
+      "step": 9340
+    },
+    {
+      "epoch": 2.1229545454545455,
+      "grad_norm": 1.7360946978066571,
+      "learning_rate": 7.717710414783974e-07,
+      "loss": 0.0228,
+      "step": 9341
+    },
+    {
+      "epoch": 2.123181818181818,
+      "grad_norm": 1.523762431299079,
+      "learning_rate": 7.716842860104889e-07,
+      "loss": 0.0134,
+      "step": 9342
+    },
+    {
+      "epoch": 2.123409090909091,
+      "grad_norm": 3.0110014633236024,
+      "learning_rate": 7.715975275514318e-07,
+      "loss": 0.0341,
+      "step": 9343
+    },
+    {
+      "epoch": 2.1236363636363635,
+      "grad_norm": 2.196348514534903,
+      "learning_rate": 7.715107661029955e-07,
+      "loss": 0.0177,
+      "step": 9344
+    },
+    {
+      "epoch": 2.1238636363636365,
+      "grad_norm": 1.930049242847441,
+      "learning_rate": 7.714240016669488e-07,
+      "loss": 0.0481,
+      "step": 9345
+    },
+    {
+      "epoch": 2.124090909090909,
+      "grad_norm": 1.5853709389378214,
+      "learning_rate": 7.713372342450616e-07,
+      "loss": 0.0163,
+      "step": 9346
+    },
+    {
+      "epoch": 2.124318181818182,
+      "grad_norm": 1.8663348750385031,
+      "learning_rate": 7.712504638391027e-07,
+      "loss": 0.0171,
+      "step": 9347
+    },
+    {
+      "epoch": 2.1245454545454545,
+      "grad_norm": 1.6419744942911925,
+      "learning_rate": 7.711636904508417e-07,
+      "loss": 0.0353,
+      "step": 9348
+    },
+    {
+      "epoch": 2.1247727272727275,
+      "grad_norm": 4.970826273924447,
+      "learning_rate": 7.71076914082048e-07,
+      "loss": 0.0669,
+      "step": 9349
+    },
+    {
+      "epoch": 2.125,
+      "grad_norm": 2.756037811575983,
+      "learning_rate": 7.709901347344912e-07,
+      "loss": 0.038,
+      "step": 9350
+    },
+    {
+      "epoch": 2.1252272727272725,
+      "grad_norm": 1.9995234905239556,
+      "learning_rate": 7.709033524099411e-07,
+      "loss": 0.0306,
+      "step": 9351
+    },
+    {
+      "epoch": 2.1254545454545455,
+      "grad_norm": 1.6478544394335213,
+      "learning_rate": 7.708165671101666e-07,
+      "loss": 0.0177,
+      "step": 9352
+    },
+    {
+      "epoch": 2.125681818181818,
+      "grad_norm": 2.742762670133434,
+      "learning_rate": 7.707297788369384e-07,
+      "loss": 0.0236,
+      "step": 9353
+    },
+    {
+      "epoch": 2.125909090909091,
+      "grad_norm": 2.794888374849228,
+      "learning_rate": 7.706429875920256e-07,
+      "loss": 0.0463,
+      "step": 9354
+    },
+    {
+      "epoch": 2.1261363636363635,
+      "grad_norm": 1.7359128187022566,
+      "learning_rate": 7.705561933771983e-07,
+      "loss": 0.0251,
+      "step": 9355
+    },
+    {
+      "epoch": 2.1263636363636365,
+      "grad_norm": 0.7916134381611711,
+      "learning_rate": 7.704693961942262e-07,
+      "loss": 0.0047,
+      "step": 9356
+    },
+    {
+      "epoch": 2.126590909090909,
+      "grad_norm": 2.329289403314928,
+      "learning_rate": 7.703825960448791e-07,
+      "loss": 0.0301,
+      "step": 9357
+    },
+    {
+      "epoch": 2.126818181818182,
+      "grad_norm": 0.6126086314174083,
+      "learning_rate": 7.702957929309274e-07,
+      "loss": 0.0045,
+      "step": 9358
+    },
+    {
+      "epoch": 2.1270454545454545,
+      "grad_norm": 2.031700629578357,
+      "learning_rate": 7.70208986854141e-07,
+      "loss": 0.0354,
+      "step": 9359
+    },
+    {
+      "epoch": 2.1272727272727274,
+      "grad_norm": 2.5624593247719525,
+      "learning_rate": 7.701221778162901e-07,
+      "loss": 0.0266,
+      "step": 9360
+    },
+    {
+      "epoch": 2.1275,
+      "grad_norm": 2.4090277822893125,
+      "learning_rate": 7.700353658191448e-07,
+      "loss": 0.029,
+      "step": 9361
+    },
+    {
+      "epoch": 2.127727272727273,
+      "grad_norm": 1.1209365527065012,
+      "learning_rate": 7.699485508644753e-07,
+      "loss": 0.0288,
+      "step": 9362
+    },
+    {
+      "epoch": 2.1279545454545454,
+      "grad_norm": 1.372089045572056,
+      "learning_rate": 7.69861732954052e-07,
+      "loss": 0.0106,
+      "step": 9363
+    },
+    {
+      "epoch": 2.128181818181818,
+      "grad_norm": 1.4947954629920837,
+      "learning_rate": 7.697749120896452e-07,
+      "loss": 0.0236,
+      "step": 9364
+    },
+    {
+      "epoch": 2.128409090909091,
+      "grad_norm": 1.0003033043106213,
+      "learning_rate": 7.696880882730254e-07,
+      "loss": 0.0099,
+      "step": 9365
+    },
+    {
+      "epoch": 2.1286363636363634,
+      "grad_norm": 1.2874096104901354,
+      "learning_rate": 7.696012615059631e-07,
+      "loss": 0.0282,
+      "step": 9366
+    },
+    {
+      "epoch": 2.1288636363636364,
+      "grad_norm": 1.900548626812327,
+      "learning_rate": 7.695144317902286e-07,
+      "loss": 0.0188,
+      "step": 9367
+    },
+    {
+      "epoch": 2.129090909090909,
+      "grad_norm": 1.8518489760952312,
+      "learning_rate": 7.694275991275931e-07,
+      "loss": 0.021,
+      "step": 9368
+    },
+    {
+      "epoch": 2.129318181818182,
+      "grad_norm": 0.7102722839070048,
+      "learning_rate": 7.693407635198266e-07,
+      "loss": 0.012,
+      "step": 9369
+    },
+    {
+      "epoch": 2.1295454545454544,
+      "grad_norm": 3.800822431810771,
+      "learning_rate": 7.692539249687002e-07,
+      "loss": 0.0196,
+      "step": 9370
+    },
+    {
+      "epoch": 2.1297727272727274,
+      "grad_norm": 1.2523559682662981,
+      "learning_rate": 7.691670834759846e-07,
+      "loss": 0.0168,
+      "step": 9371
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.596797042612363,
+      "learning_rate": 7.690802390434506e-07,
+      "loss": 0.0359,
+      "step": 9372
+    },
+    {
+      "epoch": 2.130227272727273,
+      "grad_norm": 2.119416793477916,
+      "learning_rate": 7.689933916728692e-07,
+      "loss": 0.0393,
+      "step": 9373
+    },
+    {
+      "epoch": 2.1304545454545454,
+      "grad_norm": 2.606561405855427,
+      "learning_rate": 7.689065413660113e-07,
+      "loss": 0.048,
+      "step": 9374
+    },
+    {
+      "epoch": 2.1306818181818183,
+      "grad_norm": 2.8649013697968257,
+      "learning_rate": 7.688196881246481e-07,
+      "loss": 0.0192,
+      "step": 9375
+    },
+    {
+      "epoch": 2.130909090909091,
+      "grad_norm": 4.190526306189065,
+      "learning_rate": 7.687328319505505e-07,
+      "loss": 0.0708,
+      "step": 9376
+    },
+    {
+      "epoch": 2.131136363636364,
+      "grad_norm": 1.6847524286566475,
+      "learning_rate": 7.686459728454897e-07,
+      "loss": 0.0217,
+      "step": 9377
+    },
+    {
+      "epoch": 2.1313636363636363,
+      "grad_norm": 2.1777502323813107,
+      "learning_rate": 7.685591108112369e-07,
+      "loss": 0.0357,
+      "step": 9378
+    },
+    {
+      "epoch": 2.1315909090909093,
+      "grad_norm": 2.1641252704408687,
+      "learning_rate": 7.684722458495632e-07,
+      "loss": 0.0353,
+      "step": 9379
+    },
+    {
+      "epoch": 2.131818181818182,
+      "grad_norm": 1.4921807891423844,
+      "learning_rate": 7.683853779622403e-07,
+      "loss": 0.0173,
+      "step": 9380
+    },
+    {
+      "epoch": 2.1320454545454544,
+      "grad_norm": 1.5035580633223455,
+      "learning_rate": 7.682985071510392e-07,
+      "loss": 0.0263,
+      "step": 9381
+    },
+    {
+      "epoch": 2.1322727272727273,
+      "grad_norm": 2.2641354577542105,
+      "learning_rate": 7.682116334177318e-07,
+      "loss": 0.0313,
+      "step": 9382
+    },
+    {
+      "epoch": 2.1325,
+      "grad_norm": 0.8191165238039134,
+      "learning_rate": 7.681247567640892e-07,
+      "loss": 0.0114,
+      "step": 9383
+    },
+    {
+      "epoch": 2.132727272727273,
+      "grad_norm": 1.2776491088440913,
+      "learning_rate": 7.680378771918832e-07,
+      "loss": 0.0121,
+      "step": 9384
+    },
+    {
+      "epoch": 2.1329545454545453,
+      "grad_norm": 2.7633991803472266,
+      "learning_rate": 7.679509947028853e-07,
+      "loss": 0.0432,
+      "step": 9385
+    },
+    {
+      "epoch": 2.1331818181818183,
+      "grad_norm": 1.5131423616011586,
+      "learning_rate": 7.678641092988673e-07,
+      "loss": 0.0265,
+      "step": 9386
+    },
+    {
+      "epoch": 2.133409090909091,
+      "grad_norm": 3.4140869058889005,
+      "learning_rate": 7.677772209816008e-07,
+      "loss": 0.0407,
+      "step": 9387
+    },
+    {
+      "epoch": 2.1336363636363638,
+      "grad_norm": 2.011372920564729,
+      "learning_rate": 7.676903297528574e-07,
+      "loss": 0.0152,
+      "step": 9388
+    },
+    {
+      "epoch": 2.1338636363636363,
+      "grad_norm": 1.6268776663541913,
+      "learning_rate": 7.676034356144097e-07,
+      "loss": 0.0315,
+      "step": 9389
+    },
+    {
+      "epoch": 2.1340909090909093,
+      "grad_norm": 1.6678833980441592,
+      "learning_rate": 7.675165385680289e-07,
+      "loss": 0.0245,
+      "step": 9390
+    },
+    {
+      "epoch": 2.134318181818182,
+      "grad_norm": 1.7381060605723597,
+      "learning_rate": 7.674296386154873e-07,
+      "loss": 0.0131,
+      "step": 9391
+    },
+    {
+      "epoch": 2.1345454545454547,
+      "grad_norm": 1.4999670124415059,
+      "learning_rate": 7.673427357585569e-07,
+      "loss": 0.0126,
+      "step": 9392
+    },
+    {
+      "epoch": 2.1347727272727273,
+      "grad_norm": 2.905174622091691,
+      "learning_rate": 7.672558299990098e-07,
+      "loss": 0.0491,
+      "step": 9393
+    },
+    {
+      "epoch": 2.135,
+      "grad_norm": 1.171877174955428,
+      "learning_rate": 7.671689213386181e-07,
+      "loss": 0.0106,
+      "step": 9394
+    },
+    {
+      "epoch": 2.1352272727272728,
+      "grad_norm": 1.8080665423541766,
+      "learning_rate": 7.670820097791542e-07,
+      "loss": 0.044,
+      "step": 9395
+    },
+    {
+      "epoch": 2.1354545454545453,
+      "grad_norm": 2.00331619101598,
+      "learning_rate": 7.6699509532239e-07,
+      "loss": 0.0388,
+      "step": 9396
+    },
+    {
+      "epoch": 2.1356818181818182,
+      "grad_norm": 1.4722417968481833,
+      "learning_rate": 7.669081779700983e-07,
+      "loss": 0.0149,
+      "step": 9397
+    },
+    {
+      "epoch": 2.1359090909090908,
+      "grad_norm": 0.6268691779402586,
+      "learning_rate": 7.668212577240512e-07,
+      "loss": 0.0124,
+      "step": 9398
+    },
+    {
+      "epoch": 2.1361363636363637,
+      "grad_norm": 1.6300532331198267,
+      "learning_rate": 7.667343345860214e-07,
+      "loss": 0.0186,
+      "step": 9399
+    },
+    {
+      "epoch": 2.1363636363636362,
+      "grad_norm": 2.589363066917316,
+      "learning_rate": 7.666474085577809e-07,
+      "loss": 0.0208,
+      "step": 9400
+    },
+    {
+      "epoch": 2.136590909090909,
+      "grad_norm": 0.26114886027174977,
+      "learning_rate": 7.665604796411029e-07,
+      "loss": 0.0045,
+      "step": 9401
+    },
+    {
+      "epoch": 2.1368181818181817,
+      "grad_norm": 1.1057913700420616,
+      "learning_rate": 7.664735478377596e-07,
+      "loss": 0.0119,
+      "step": 9402
+    },
+    {
+      "epoch": 2.1370454545454547,
+      "grad_norm": 1.5260707047882074,
+      "learning_rate": 7.663866131495238e-07,
+      "loss": 0.0183,
+      "step": 9403
+    },
+    {
+      "epoch": 2.137272727272727,
+      "grad_norm": 2.6759648785562393,
+      "learning_rate": 7.662996755781684e-07,
+      "loss": 0.0227,
+      "step": 9404
+    },
+    {
+      "epoch": 2.1375,
+      "grad_norm": 1.254969370788069,
+      "learning_rate": 7.662127351254661e-07,
+      "loss": 0.0132,
+      "step": 9405
+    },
+    {
+      "epoch": 2.1377272727272727,
+      "grad_norm": 0.9457919658051469,
+      "learning_rate": 7.661257917931899e-07,
+      "loss": 0.0162,
+      "step": 9406
+    },
+    {
+      "epoch": 2.137954545454545,
+      "grad_norm": 3.1475980365474774,
+      "learning_rate": 7.660388455831125e-07,
+      "loss": 0.0716,
+      "step": 9407
+    },
+    {
+      "epoch": 2.138181818181818,
+      "grad_norm": 1.3073245763652395,
+      "learning_rate": 7.65951896497007e-07,
+      "loss": 0.0141,
+      "step": 9408
+    },
+    {
+      "epoch": 2.1384090909090907,
+      "grad_norm": 1.0785358174206527,
+      "learning_rate": 7.658649445366463e-07,
+      "loss": 0.0103,
+      "step": 9409
+    },
+    {
+      "epoch": 2.1386363636363637,
+      "grad_norm": 1.7472221410995739,
+      "learning_rate": 7.657779897038037e-07,
+      "loss": 0.0167,
+      "step": 9410
+    },
+    {
+      "epoch": 2.138863636363636,
+      "grad_norm": 1.8462862946475163,
+      "learning_rate": 7.656910320002522e-07,
+      "loss": 0.0121,
+      "step": 9411
+    },
+    {
+      "epoch": 2.139090909090909,
+      "grad_norm": 2.1921116494513617,
+      "learning_rate": 7.656040714277655e-07,
+      "loss": 0.024,
+      "step": 9412
+    },
+    {
+      "epoch": 2.1393181818181817,
+      "grad_norm": 1.3080519879871308,
+      "learning_rate": 7.655171079881161e-07,
+      "loss": 0.0097,
+      "step": 9413
+    },
+    {
+      "epoch": 2.1395454545454546,
+      "grad_norm": 1.4177708117490497,
+      "learning_rate": 7.654301416830777e-07,
+      "loss": 0.0151,
+      "step": 9414
+    },
+    {
+      "epoch": 2.139772727272727,
+      "grad_norm": 1.6146114312984943,
+      "learning_rate": 7.653431725144239e-07,
+      "loss": 0.0175,
+      "step": 9415
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.4311589150190085,
+      "learning_rate": 7.65256200483928e-07,
+      "loss": 0.0174,
+      "step": 9416
+    },
+    {
+      "epoch": 2.1402272727272726,
+      "grad_norm": 1.9866932680209917,
+      "learning_rate": 7.651692255933634e-07,
+      "loss": 0.0323,
+      "step": 9417
+    },
+    {
+      "epoch": 2.1404545454545456,
+      "grad_norm": 3.1084767967402103,
+      "learning_rate": 7.650822478445036e-07,
+      "loss": 0.0314,
+      "step": 9418
+    },
+    {
+      "epoch": 2.140681818181818,
+      "grad_norm": 1.4850680936778395,
+      "learning_rate": 7.649952672391226e-07,
+      "loss": 0.0091,
+      "step": 9419
+    },
+    {
+      "epoch": 2.140909090909091,
+      "grad_norm": 0.5748049685548319,
+      "learning_rate": 7.649082837789937e-07,
+      "loss": 0.006,
+      "step": 9420
+    },
+    {
+      "epoch": 2.1411363636363636,
+      "grad_norm": 2.5342152041405046,
+      "learning_rate": 7.64821297465891e-07,
+      "loss": 0.0209,
+      "step": 9421
+    },
+    {
+      "epoch": 2.1413636363636366,
+      "grad_norm": 1.4962420962150613,
+      "learning_rate": 7.64734308301588e-07,
+      "loss": 0.0164,
+      "step": 9422
+    },
+    {
+      "epoch": 2.141590909090909,
+      "grad_norm": 0.9667300092113447,
+      "learning_rate": 7.646473162878587e-07,
+      "loss": 0.0067,
+      "step": 9423
+    },
+    {
+      "epoch": 2.1418181818181816,
+      "grad_norm": 2.1652434702918875,
+      "learning_rate": 7.64560321426477e-07,
+      "loss": 0.0247,
+      "step": 9424
+    },
+    {
+      "epoch": 2.1420454545454546,
+      "grad_norm": 4.246630378937667,
+      "learning_rate": 7.644733237192169e-07,
+      "loss": 0.0517,
+      "step": 9425
+    },
+    {
+      "epoch": 2.142272727272727,
+      "grad_norm": 1.531518018824639,
+      "learning_rate": 7.643863231678523e-07,
+      "loss": 0.0181,
+      "step": 9426
+    },
+    {
+      "epoch": 2.1425,
+      "grad_norm": 1.4311224493048884,
+      "learning_rate": 7.642993197741573e-07,
+      "loss": 0.0169,
+      "step": 9427
+    },
+    {
+      "epoch": 2.1427272727272726,
+      "grad_norm": 1.2012024278817937,
+      "learning_rate": 7.642123135399064e-07,
+      "loss": 0.0113,
+      "step": 9428
+    },
+    {
+      "epoch": 2.1429545454545456,
+      "grad_norm": 3.143117339011605,
+      "learning_rate": 7.641253044668733e-07,
+      "loss": 0.0529,
+      "step": 9429
+    },
+    {
+      "epoch": 2.143181818181818,
+      "grad_norm": 1.0911669105277675,
+      "learning_rate": 7.640382925568329e-07,
+      "loss": 0.0092,
+      "step": 9430
+    },
+    {
+      "epoch": 2.143409090909091,
+      "grad_norm": 0.8418479608664992,
+      "learning_rate": 7.639512778115589e-07,
+      "loss": 0.01,
+      "step": 9431
+    },
+    {
+      "epoch": 2.1436363636363636,
+      "grad_norm": 2.04797547094016,
+      "learning_rate": 7.63864260232826e-07,
+      "loss": 0.0153,
+      "step": 9432
+    },
+    {
+      "epoch": 2.1438636363636365,
+      "grad_norm": 1.61667290067131,
+      "learning_rate": 7.637772398224084e-07,
+      "loss": 0.0209,
+      "step": 9433
+    },
+    {
+      "epoch": 2.144090909090909,
+      "grad_norm": 1.2762125389454364,
+      "learning_rate": 7.63690216582081e-07,
+      "loss": 0.0319,
+      "step": 9434
+    },
+    {
+      "epoch": 2.144318181818182,
+      "grad_norm": 1.6632252586097844,
+      "learning_rate": 7.63603190513618e-07,
+      "loss": 0.0089,
+      "step": 9435
+    },
+    {
+      "epoch": 2.1445454545454545,
+      "grad_norm": 2.8379034536922947,
+      "learning_rate": 7.635161616187941e-07,
+      "loss": 0.0566,
+      "step": 9436
+    },
+    {
+      "epoch": 2.144772727272727,
+      "grad_norm": 2.0762519889754643,
+      "learning_rate": 7.634291298993843e-07,
+      "loss": 0.0158,
+      "step": 9437
+    },
+    {
+      "epoch": 2.145,
+      "grad_norm": 2.451584673527537,
+      "learning_rate": 7.633420953571626e-07,
+      "loss": 0.0307,
+      "step": 9438
+    },
+    {
+      "epoch": 2.1452272727272725,
+      "grad_norm": 2.6427800840092113,
+      "learning_rate": 7.632550579939046e-07,
+      "loss": 0.0331,
+      "step": 9439
+    },
+    {
+      "epoch": 2.1454545454545455,
+      "grad_norm": 2.6802314578555237,
+      "learning_rate": 7.631680178113846e-07,
+      "loss": 0.0674,
+      "step": 9440
+    },
+    {
+      "epoch": 2.145681818181818,
+      "grad_norm": 2.3184580373466637,
+      "learning_rate": 7.630809748113777e-07,
+      "loss": 0.0506,
+      "step": 9441
+    },
+    {
+      "epoch": 2.145909090909091,
+      "grad_norm": 1.7782197403253461,
+      "learning_rate": 7.629939289956587e-07,
+      "loss": 0.0157,
+      "step": 9442
+    },
+    {
+      "epoch": 2.1461363636363635,
+      "grad_norm": 3.2684732578590956,
+      "learning_rate": 7.629068803660029e-07,
+      "loss": 0.12,
+      "step": 9443
+    },
+    {
+      "epoch": 2.1463636363636365,
+      "grad_norm": 3.1633219246671715,
+      "learning_rate": 7.628198289241852e-07,
+      "loss": 0.0305,
+      "step": 9444
+    },
+    {
+      "epoch": 2.146590909090909,
+      "grad_norm": 1.8327881456781283,
+      "learning_rate": 7.627327746719808e-07,
+      "loss": 0.0335,
+      "step": 9445
+    },
+    {
+      "epoch": 2.146818181818182,
+      "grad_norm": 2.9742154920769424,
+      "learning_rate": 7.626457176111647e-07,
+      "loss": 0.0541,
+      "step": 9446
+    },
+    {
+      "epoch": 2.1470454545454545,
+      "grad_norm": 1.0638882593451753,
+      "learning_rate": 7.625586577435124e-07,
+      "loss": 0.0135,
+      "step": 9447
+    },
+    {
+      "epoch": 2.1472727272727274,
+      "grad_norm": 1.2147601797668008,
+      "learning_rate": 7.624715950707989e-07,
+      "loss": 0.0158,
+      "step": 9448
+    },
+    {
+      "epoch": 2.1475,
+      "grad_norm": 3.0707850081733308,
+      "learning_rate": 7.623845295947998e-07,
+      "loss": 0.0405,
+      "step": 9449
+    },
+    {
+      "epoch": 2.147727272727273,
+      "grad_norm": 1.2856101001688736,
+      "learning_rate": 7.622974613172907e-07,
+      "loss": 0.0112,
+      "step": 9450
+    },
+    {
+      "epoch": 2.1479545454545454,
+      "grad_norm": 2.354485779308135,
+      "learning_rate": 7.622103902400466e-07,
+      "loss": 0.0255,
+      "step": 9451
+    },
+    {
+      "epoch": 2.1481818181818184,
+      "grad_norm": 1.8771356476474699,
+      "learning_rate": 7.621233163648434e-07,
+      "loss": 0.0488,
+      "step": 9452
+    },
+    {
+      "epoch": 2.148409090909091,
+      "grad_norm": 3.2226734690266214,
+      "learning_rate": 7.620362396934564e-07,
+      "loss": 0.0243,
+      "step": 9453
+    },
+    {
+      "epoch": 2.1486363636363635,
+      "grad_norm": 1.7284036688799642,
+      "learning_rate": 7.619491602276616e-07,
+      "loss": 0.0121,
+      "step": 9454
+    },
+    {
+      "epoch": 2.1488636363636364,
+      "grad_norm": 1.0811359180267317,
+      "learning_rate": 7.618620779692343e-07,
+      "loss": 0.0261,
+      "step": 9455
+    },
+    {
+      "epoch": 2.149090909090909,
+      "grad_norm": 2.016334445808552,
+      "learning_rate": 7.617749929199506e-07,
+      "loss": 0.0245,
+      "step": 9456
+    },
+    {
+      "epoch": 2.149318181818182,
+      "grad_norm": 1.4445946574073918,
+      "learning_rate": 7.616879050815859e-07,
+      "loss": 0.0216,
+      "step": 9457
+    },
+    {
+      "epoch": 2.1495454545454544,
+      "grad_norm": 1.0886699631886387,
+      "learning_rate": 7.616008144559167e-07,
+      "loss": 0.0106,
+      "step": 9458
+    },
+    {
+      "epoch": 2.1497727272727274,
+      "grad_norm": 3.168222421284589,
+      "learning_rate": 7.615137210447185e-07,
+      "loss": 0.0509,
+      "step": 9459
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 2.4628614171131926,
+      "learning_rate": 7.614266248497672e-07,
+      "loss": 0.0294,
+      "step": 9460
+    },
+    {
+      "epoch": 2.150227272727273,
+      "grad_norm": 1.7034376121963375,
+      "learning_rate": 7.613395258728392e-07,
+      "loss": 0.0213,
+      "step": 9461
+    },
+    {
+      "epoch": 2.1504545454545454,
+      "grad_norm": 1.681268788272346,
+      "learning_rate": 7.612524241157103e-07,
+      "loss": 0.0156,
+      "step": 9462
+    },
+    {
+      "epoch": 2.1506818181818184,
+      "grad_norm": 1.1537910156771578,
+      "learning_rate": 7.611653195801569e-07,
+      "loss": 0.0145,
+      "step": 9463
+    },
+    {
+      "epoch": 2.150909090909091,
+      "grad_norm": 2.1592831127015333,
+      "learning_rate": 7.610782122679548e-07,
+      "loss": 0.0148,
+      "step": 9464
+    },
+    {
+      "epoch": 2.151136363636364,
+      "grad_norm": 1.5887449255755106,
+      "learning_rate": 7.609911021808809e-07,
+      "loss": 0.0197,
+      "step": 9465
+    },
+    {
+      "epoch": 2.1513636363636364,
+      "grad_norm": 4.168831524838901,
+      "learning_rate": 7.60903989320711e-07,
+      "loss": 0.0787,
+      "step": 9466
+    },
+    {
+      "epoch": 2.151590909090909,
+      "grad_norm": 4.622614203701564,
+      "learning_rate": 7.608168736892217e-07,
+      "loss": 0.028,
+      "step": 9467
+    },
+    {
+      "epoch": 2.151818181818182,
+      "grad_norm": 2.0922143770141024,
+      "learning_rate": 7.607297552881894e-07,
+      "loss": 0.0165,
+      "step": 9468
+    },
+    {
+      "epoch": 2.1520454545454544,
+      "grad_norm": 1.3167332194649353,
+      "learning_rate": 7.606426341193906e-07,
+      "loss": 0.0119,
+      "step": 9469
+    },
+    {
+      "epoch": 2.1522727272727273,
+      "grad_norm": 2.336689787845125,
+      "learning_rate": 7.605555101846019e-07,
+      "loss": 0.0365,
+      "step": 9470
+    },
+    {
+      "epoch": 2.1525,
+      "grad_norm": 2.159798710318973,
+      "learning_rate": 7.604683834855995e-07,
+      "loss": 0.0409,
+      "step": 9471
+    },
+    {
+      "epoch": 2.152727272727273,
+      "grad_norm": 1.6112732262366727,
+      "learning_rate": 7.603812540241608e-07,
+      "loss": 0.0347,
+      "step": 9472
+    },
+    {
+      "epoch": 2.1529545454545453,
+      "grad_norm": 1.2417300945292054,
+      "learning_rate": 7.60294121802062e-07,
+      "loss": 0.0162,
+      "step": 9473
+    },
+    {
+      "epoch": 2.1531818181818183,
+      "grad_norm": 2.2691399472291653,
+      "learning_rate": 7.602069868210802e-07,
+      "loss": 0.0312,
+      "step": 9474
+    },
+    {
+      "epoch": 2.153409090909091,
+      "grad_norm": 1.225408795840806,
+      "learning_rate": 7.601198490829918e-07,
+      "loss": 0.0101,
+      "step": 9475
+    },
+    {
+      "epoch": 2.153636363636364,
+      "grad_norm": 1.291667306621604,
+      "learning_rate": 7.600327085895742e-07,
+      "loss": 0.0123,
+      "step": 9476
+    },
+    {
+      "epoch": 2.1538636363636363,
+      "grad_norm": 1.0700621589984567,
+      "learning_rate": 7.59945565342604e-07,
+      "loss": 0.025,
+      "step": 9477
+    },
+    {
+      "epoch": 2.1540909090909093,
+      "grad_norm": 2.316699248853042,
+      "learning_rate": 7.598584193438581e-07,
+      "loss": 0.017,
+      "step": 9478
+    },
+    {
+      "epoch": 2.154318181818182,
+      "grad_norm": 1.2237160299334577,
+      "learning_rate": 7.597712705951139e-07,
+      "loss": 0.017,
+      "step": 9479
+    },
+    {
+      "epoch": 2.1545454545454543,
+      "grad_norm": 7.4179175525654175,
+      "learning_rate": 7.596841190981486e-07,
+      "loss": 0.0617,
+      "step": 9480
+    },
+    {
+      "epoch": 2.1547727272727273,
+      "grad_norm": 0.9936545251002082,
+      "learning_rate": 7.595969648547388e-07,
+      "loss": 0.0107,
+      "step": 9481
+    },
+    {
+      "epoch": 2.155,
+      "grad_norm": 2.611735622562155,
+      "learning_rate": 7.595098078666621e-07,
+      "loss": 0.0138,
+      "step": 9482
+    },
+    {
+      "epoch": 2.1552272727272728,
+      "grad_norm": 2.4013420312781104,
+      "learning_rate": 7.59422648135696e-07,
+      "loss": 0.0152,
+      "step": 9483
+    },
+    {
+      "epoch": 2.1554545454545453,
+      "grad_norm": 1.147943105527006,
+      "learning_rate": 7.593354856636174e-07,
+      "loss": 0.0209,
+      "step": 9484
+    },
+    {
+      "epoch": 2.1556818181818183,
+      "grad_norm": 2.2098690974874784,
+      "learning_rate": 7.59248320452204e-07,
+      "loss": 0.0233,
+      "step": 9485
+    },
+    {
+      "epoch": 2.1559090909090908,
+      "grad_norm": 1.4322431320668665,
+      "learning_rate": 7.591611525032328e-07,
+      "loss": 0.0168,
+      "step": 9486
+    },
+    {
+      "epoch": 2.1561363636363637,
+      "grad_norm": 2.0220711662663993,
+      "learning_rate": 7.590739818184821e-07,
+      "loss": 0.0266,
+      "step": 9487
+    },
+    {
+      "epoch": 2.1563636363636363,
+      "grad_norm": 2.4098101612319036,
+      "learning_rate": 7.589868083997289e-07,
+      "loss": 0.0321,
+      "step": 9488
+    },
+    {
+      "epoch": 2.1565909090909092,
+      "grad_norm": 1.685182577579371,
+      "learning_rate": 7.588996322487509e-07,
+      "loss": 0.0185,
+      "step": 9489
+    },
+    {
+      "epoch": 2.1568181818181817,
+      "grad_norm": 1.7289944008677334,
+      "learning_rate": 7.588124533673257e-07,
+      "loss": 0.0185,
+      "step": 9490
+    },
+    {
+      "epoch": 2.1570454545454547,
+      "grad_norm": 1.4452607932842316,
+      "learning_rate": 7.587252717572312e-07,
+      "loss": 0.0133,
+      "step": 9491
+    },
+    {
+      "epoch": 2.1572727272727272,
+      "grad_norm": 2.445026829727573,
+      "learning_rate": 7.58638087420245e-07,
+      "loss": 0.0192,
+      "step": 9492
+    },
+    {
+      "epoch": 2.1575,
+      "grad_norm": 1.3873824552900882,
+      "learning_rate": 7.585509003581451e-07,
+      "loss": 0.0067,
+      "step": 9493
+    },
+    {
+      "epoch": 2.1577272727272727,
+      "grad_norm": 0.8546391309646049,
+      "learning_rate": 7.584637105727094e-07,
+      "loss": 0.0075,
+      "step": 9494
+    },
+    {
+      "epoch": 2.1579545454545457,
+      "grad_norm": 2.4900962869114,
+      "learning_rate": 7.583765180657157e-07,
+      "loss": 0.0193,
+      "step": 9495
+    },
+    {
+      "epoch": 2.158181818181818,
+      "grad_norm": 1.8896444012996825,
+      "learning_rate": 7.582893228389423e-07,
+      "loss": 0.0193,
+      "step": 9496
+    },
+    {
+      "epoch": 2.1584090909090907,
+      "grad_norm": 2.0002576720686123,
+      "learning_rate": 7.582021248941669e-07,
+      "loss": 0.0358,
+      "step": 9497
+    },
+    {
+      "epoch": 2.1586363636363637,
+      "grad_norm": 2.4655790907924917,
+      "learning_rate": 7.58114924233168e-07,
+      "loss": 0.037,
+      "step": 9498
+    },
+    {
+      "epoch": 2.158863636363636,
+      "grad_norm": 1.4499721166216093,
+      "learning_rate": 7.580277208577233e-07,
+      "loss": 0.0145,
+      "step": 9499
+    },
+    {
+      "epoch": 2.159090909090909,
+      "grad_norm": 0.8224065340500074,
+      "learning_rate": 7.579405147696116e-07,
+      "loss": 0.0053,
+      "step": 9500
+    },
+    {
+      "epoch": 2.1593181818181817,
+      "grad_norm": 1.2043884202972308,
+      "learning_rate": 7.578533059706104e-07,
+      "loss": 0.0105,
+      "step": 9501
+    },
+    {
+      "epoch": 2.1595454545454547,
+      "grad_norm": 1.23978331576653,
+      "learning_rate": 7.57766094462499e-07,
+      "loss": 0.0092,
+      "step": 9502
+    },
+    {
+      "epoch": 2.159772727272727,
+      "grad_norm": 2.461277724408853,
+      "learning_rate": 7.57678880247055e-07,
+      "loss": 0.0608,
+      "step": 9503
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 5.6840469184538005,
+      "learning_rate": 7.575916633260572e-07,
+      "loss": 0.0197,
+      "step": 9504
+    },
+    {
+      "epoch": 2.1602272727272727,
+      "grad_norm": 1.1583676450588933,
+      "learning_rate": 7.575044437012844e-07,
+      "loss": 0.0149,
+      "step": 9505
+    },
+    {
+      "epoch": 2.1604545454545456,
+      "grad_norm": 6.206545741110995,
+      "learning_rate": 7.574172213745144e-07,
+      "loss": 0.0182,
+      "step": 9506
+    },
+    {
+      "epoch": 2.160681818181818,
+      "grad_norm": 1.8522456300922556,
+      "learning_rate": 7.573299963475265e-07,
+      "loss": 0.0297,
+      "step": 9507
+    },
+    {
+      "epoch": 2.160909090909091,
+      "grad_norm": 1.365702295616251,
+      "learning_rate": 7.57242768622099e-07,
+      "loss": 0.0209,
+      "step": 9508
+    },
+    {
+      "epoch": 2.1611363636363636,
+      "grad_norm": 1.540365418188812,
+      "learning_rate": 7.571555382000106e-07,
+      "loss": 0.0221,
+      "step": 9509
+    },
+    {
+      "epoch": 2.161363636363636,
+      "grad_norm": 1.4382702785425119,
+      "learning_rate": 7.570683050830405e-07,
+      "loss": 0.0172,
+      "step": 9510
+    },
+    {
+      "epoch": 2.161590909090909,
+      "grad_norm": 2.0077291872808396,
+      "learning_rate": 7.569810692729672e-07,
+      "loss": 0.0398,
+      "step": 9511
+    },
+    {
+      "epoch": 2.1618181818181816,
+      "grad_norm": 2.524055077936866,
+      "learning_rate": 7.568938307715695e-07,
+      "loss": 0.0306,
+      "step": 9512
+    },
+    {
+      "epoch": 2.1620454545454546,
+      "grad_norm": 1.6708930788059346,
+      "learning_rate": 7.568065895806267e-07,
+      "loss": 0.017,
+      "step": 9513
+    },
+    {
+      "epoch": 2.162272727272727,
+      "grad_norm": 2.1540455033976085,
+      "learning_rate": 7.567193457019173e-07,
+      "loss": 0.0267,
+      "step": 9514
+    },
+    {
+      "epoch": 2.1625,
+      "grad_norm": 2.5815218960919295,
+      "learning_rate": 7.566320991372207e-07,
+      "loss": 0.045,
+      "step": 9515
+    },
+    {
+      "epoch": 2.1627272727272726,
+      "grad_norm": 2.6186058821410882,
+      "learning_rate": 7.56544849888316e-07,
+      "loss": 0.0257,
+      "step": 9516
+    },
+    {
+      "epoch": 2.1629545454545456,
+      "grad_norm": 1.8492011019648205,
+      "learning_rate": 7.564575979569824e-07,
+      "loss": 0.0174,
+      "step": 9517
+    },
+    {
+      "epoch": 2.163181818181818,
+      "grad_norm": 1.5712355704893655,
+      "learning_rate": 7.563703433449991e-07,
+      "loss": 0.015,
+      "step": 9518
+    },
+    {
+      "epoch": 2.163409090909091,
+      "grad_norm": 1.2674322035621381,
+      "learning_rate": 7.562830860541452e-07,
+      "loss": 0.0185,
+      "step": 9519
+    },
+    {
+      "epoch": 2.1636363636363636,
+      "grad_norm": 1.8798112538807918,
+      "learning_rate": 7.561958260862003e-07,
+      "loss": 0.0222,
+      "step": 9520
+    },
+    {
+      "epoch": 2.1638636363636365,
+      "grad_norm": 1.8935708136329656,
+      "learning_rate": 7.561085634429434e-07,
+      "loss": 0.0282,
+      "step": 9521
+    },
+    {
+      "epoch": 2.164090909090909,
+      "grad_norm": 2.1303643344032106,
+      "learning_rate": 7.560212981261543e-07,
+      "loss": 0.0209,
+      "step": 9522
+    },
+    {
+      "epoch": 2.1643181818181816,
+      "grad_norm": 2.3541581040701143,
+      "learning_rate": 7.559340301376123e-07,
+      "loss": 0.0407,
+      "step": 9523
+    },
+    {
+      "epoch": 2.1645454545454546,
+      "grad_norm": 1.1523656414485075,
+      "learning_rate": 7.558467594790969e-07,
+      "loss": 0.0161,
+      "step": 9524
+    },
+    {
+      "epoch": 2.164772727272727,
+      "grad_norm": 2.3707675859344786,
+      "learning_rate": 7.557594861523878e-07,
+      "loss": 0.0236,
+      "step": 9525
+    },
+    {
+      "epoch": 2.165,
+      "grad_norm": 1.509616097779685,
+      "learning_rate": 7.556722101592648e-07,
+      "loss": 0.0118,
+      "step": 9526
+    },
+    {
+      "epoch": 2.1652272727272726,
+      "grad_norm": 1.4627649880258733,
+      "learning_rate": 7.555849315015073e-07,
+      "loss": 0.0177,
+      "step": 9527
+    },
+    {
+      "epoch": 2.1654545454545455,
+      "grad_norm": 1.59491529723371,
+      "learning_rate": 7.554976501808954e-07,
+      "loss": 0.0283,
+      "step": 9528
+    },
+    {
+      "epoch": 2.165681818181818,
+      "grad_norm": 1.3006423652669914,
+      "learning_rate": 7.554103661992087e-07,
+      "loss": 0.0115,
+      "step": 9529
+    },
+    {
+      "epoch": 2.165909090909091,
+      "grad_norm": 0.9129970373908144,
+      "learning_rate": 7.553230795582271e-07,
+      "loss": 0.0098,
+      "step": 9530
+    },
+    {
+      "epoch": 2.1661363636363635,
+      "grad_norm": 2.630833924096012,
+      "learning_rate": 7.552357902597306e-07,
+      "loss": 0.0223,
+      "step": 9531
+    },
+    {
+      "epoch": 2.1663636363636365,
+      "grad_norm": 0.9753052334395441,
+      "learning_rate": 7.55148498305499e-07,
+      "loss": 0.0183,
+      "step": 9532
+    },
+    {
+      "epoch": 2.166590909090909,
+      "grad_norm": 1.0207261743260008,
+      "learning_rate": 7.550612036973126e-07,
+      "loss": 0.01,
+      "step": 9533
+    },
+    {
+      "epoch": 2.166818181818182,
+      "grad_norm": 1.1459709853725635,
+      "learning_rate": 7.549739064369512e-07,
+      "loss": 0.015,
+      "step": 9534
+    },
+    {
+      "epoch": 2.1670454545454545,
+      "grad_norm": 2.5094984617583957,
+      "learning_rate": 7.548866065261954e-07,
+      "loss": 0.0217,
+      "step": 9535
+    },
+    {
+      "epoch": 2.1672727272727275,
+      "grad_norm": 1.837963364826961,
+      "learning_rate": 7.547993039668248e-07,
+      "loss": 0.0164,
+      "step": 9536
+    },
+    {
+      "epoch": 2.1675,
+      "grad_norm": 1.16539153606161,
+      "learning_rate": 7.5471199876062e-07,
+      "loss": 0.0158,
+      "step": 9537
+    },
+    {
+      "epoch": 2.167727272727273,
+      "grad_norm": 4.247288839543823,
+      "learning_rate": 7.546246909093613e-07,
+      "loss": 0.0444,
+      "step": 9538
+    },
+    {
+      "epoch": 2.1679545454545455,
+      "grad_norm": 1.0427532564094824,
+      "learning_rate": 7.54537380414829e-07,
+      "loss": 0.0054,
+      "step": 9539
+    },
+    {
+      "epoch": 2.168181818181818,
+      "grad_norm": 0.6176106658494027,
+      "learning_rate": 7.544500672788034e-07,
+      "loss": 0.0113,
+      "step": 9540
+    },
+    {
+      "epoch": 2.168409090909091,
+      "grad_norm": 2.436833091692273,
+      "learning_rate": 7.543627515030652e-07,
+      "loss": 0.0149,
+      "step": 9541
+    },
+    {
+      "epoch": 2.1686363636363635,
+      "grad_norm": 1.9026156421894203,
+      "learning_rate": 7.54275433089395e-07,
+      "loss": 0.0184,
+      "step": 9542
+    },
+    {
+      "epoch": 2.1688636363636364,
+      "grad_norm": 1.4411344658917913,
+      "learning_rate": 7.541881120395729e-07,
+      "loss": 0.0181,
+      "step": 9543
+    },
+    {
+      "epoch": 2.169090909090909,
+      "grad_norm": 2.198289092558246,
+      "learning_rate": 7.5410078835538e-07,
+      "loss": 0.026,
+      "step": 9544
+    },
+    {
+      "epoch": 2.169318181818182,
+      "grad_norm": 1.9539151903035237,
+      "learning_rate": 7.540134620385967e-07,
+      "loss": 0.0406,
+      "step": 9545
+    },
+    {
+      "epoch": 2.1695454545454544,
+      "grad_norm": 2.7157705021675,
+      "learning_rate": 7.539261330910038e-07,
+      "loss": 0.0418,
+      "step": 9546
+    },
+    {
+      "epoch": 2.1697727272727274,
+      "grad_norm": 2.035082997400166,
+      "learning_rate": 7.538388015143821e-07,
+      "loss": 0.0159,
+      "step": 9547
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.635028046302899,
+      "learning_rate": 7.537514673105124e-07,
+      "loss": 0.0169,
+      "step": 9548
+    },
+    {
+      "epoch": 2.170227272727273,
+      "grad_norm": 1.5946152802668783,
+      "learning_rate": 7.536641304811759e-07,
+      "loss": 0.0116,
+      "step": 9549
+    },
+    {
+      "epoch": 2.1704545454545454,
+      "grad_norm": 4.5751958834023165,
+      "learning_rate": 7.535767910281531e-07,
+      "loss": 0.0287,
+      "step": 9550
+    },
+    {
+      "epoch": 2.1706818181818184,
+      "grad_norm": 1.7178513771472552,
+      "learning_rate": 7.534894489532255e-07,
+      "loss": 0.0447,
+      "step": 9551
+    },
+    {
+      "epoch": 2.170909090909091,
+      "grad_norm": 1.9636015865812928,
+      "learning_rate": 7.534021042581736e-07,
+      "loss": 0.0251,
+      "step": 9552
+    },
+    {
+      "epoch": 2.1711363636363634,
+      "grad_norm": 2.9250102828029836,
+      "learning_rate": 7.533147569447791e-07,
+      "loss": 0.0204,
+      "step": 9553
+    },
+    {
+      "epoch": 2.1713636363636364,
+      "grad_norm": 1.70008678078672,
+      "learning_rate": 7.532274070148225e-07,
+      "loss": 0.0263,
+      "step": 9554
+    },
+    {
+      "epoch": 2.171590909090909,
+      "grad_norm": 4.761437083749199,
+      "learning_rate": 7.531400544700854e-07,
+      "loss": 0.0296,
+      "step": 9555
+    },
+    {
+      "epoch": 2.171818181818182,
+      "grad_norm": 2.2634571942064565,
+      "learning_rate": 7.530526993123491e-07,
+      "loss": 0.0363,
+      "step": 9556
+    },
+    {
+      "epoch": 2.1720454545454544,
+      "grad_norm": 2.048836833286836,
+      "learning_rate": 7.52965341543395e-07,
+      "loss": 0.0227,
+      "step": 9557
+    },
+    {
+      "epoch": 2.1722727272727274,
+      "grad_norm": 2.255786403600417,
+      "learning_rate": 7.528779811650041e-07,
+      "loss": 0.0248,
+      "step": 9558
+    },
+    {
+      "epoch": 2.1725,
+      "grad_norm": 2.9335720932815468,
+      "learning_rate": 7.527906181789582e-07,
+      "loss": 0.0452,
+      "step": 9559
+    },
+    {
+      "epoch": 2.172727272727273,
+      "grad_norm": 0.7388483909562038,
+      "learning_rate": 7.527032525870388e-07,
+      "loss": 0.0112,
+      "step": 9560
+    },
+    {
+      "epoch": 2.1729545454545454,
+      "grad_norm": 2.2522675141661215,
+      "learning_rate": 7.526158843910271e-07,
+      "loss": 0.0263,
+      "step": 9561
+    },
+    {
+      "epoch": 2.1731818181818183,
+      "grad_norm": 3.8978326007146973,
+      "learning_rate": 7.525285135927049e-07,
+      "loss": 0.0738,
+      "step": 9562
+    },
+    {
+      "epoch": 2.173409090909091,
+      "grad_norm": 2.599936766593295,
+      "learning_rate": 7.524411401938538e-07,
+      "loss": 0.0422,
+      "step": 9563
+    },
+    {
+      "epoch": 2.173636363636364,
+      "grad_norm": 2.3934497109627055,
+      "learning_rate": 7.523537641962555e-07,
+      "loss": 0.0369,
+      "step": 9564
+    },
+    {
+      "epoch": 2.1738636363636363,
+      "grad_norm": 1.7851062713433605,
+      "learning_rate": 7.52266385601692e-07,
+      "loss": 0.0161,
+      "step": 9565
+    },
+    {
+      "epoch": 2.1740909090909093,
+      "grad_norm": 4.141191175650336,
+      "learning_rate": 7.521790044119447e-07,
+      "loss": 0.0312,
+      "step": 9566
+    },
+    {
+      "epoch": 2.174318181818182,
+      "grad_norm": 1.2851816559567923,
+      "learning_rate": 7.520916206287955e-07,
+      "loss": 0.0175,
+      "step": 9567
+    },
+    {
+      "epoch": 2.174545454545455,
+      "grad_norm": 1.2026185112146341,
+      "learning_rate": 7.520042342540266e-07,
+      "loss": 0.0193,
+      "step": 9568
+    },
+    {
+      "epoch": 2.1747727272727273,
+      "grad_norm": 1.1788939726308687,
+      "learning_rate": 7.519168452894198e-07,
+      "loss": 0.0067,
+      "step": 9569
+    },
+    {
+      "epoch": 2.175,
+      "grad_norm": 1.909852333065936,
+      "learning_rate": 7.518294537367569e-07,
+      "loss": 0.0175,
+      "step": 9570
+    },
+    {
+      "epoch": 2.175227272727273,
+      "grad_norm": 0.8125018199180292,
+      "learning_rate": 7.517420595978203e-07,
+      "loss": 0.0092,
+      "step": 9571
+    },
+    {
+      "epoch": 2.1754545454545453,
+      "grad_norm": 1.1464135672831117,
+      "learning_rate": 7.51654662874392e-07,
+      "loss": 0.0162,
+      "step": 9572
+    },
+    {
+      "epoch": 2.1756818181818183,
+      "grad_norm": 2.486233650506863,
+      "learning_rate": 7.515672635682543e-07,
+      "loss": 0.0248,
+      "step": 9573
+    },
+    {
+      "epoch": 2.175909090909091,
+      "grad_norm": 1.4362690850674997,
+      "learning_rate": 7.514798616811891e-07,
+      "loss": 0.0168,
+      "step": 9574
+    },
+    {
+      "epoch": 2.1761363636363638,
+      "grad_norm": 1.322510207354574,
+      "learning_rate": 7.51392457214979e-07,
+      "loss": 0.0219,
+      "step": 9575
+    },
+    {
+      "epoch": 2.1763636363636363,
+      "grad_norm": 1.7271184227813845,
+      "learning_rate": 7.513050501714061e-07,
+      "loss": 0.0202,
+      "step": 9576
+    },
+    {
+      "epoch": 2.1765909090909092,
+      "grad_norm": 2.149878262621506,
+      "learning_rate": 7.512176405522528e-07,
+      "loss": 0.0448,
+      "step": 9577
+    },
+    {
+      "epoch": 2.1768181818181818,
+      "grad_norm": 1.9274833302538146,
+      "learning_rate": 7.511302283593017e-07,
+      "loss": 0.0221,
+      "step": 9578
+    },
+    {
+      "epoch": 2.1770454545454547,
+      "grad_norm": 1.0158174371605826,
+      "learning_rate": 7.510428135943354e-07,
+      "loss": 0.0097,
+      "step": 9579
+    },
+    {
+      "epoch": 2.1772727272727272,
+      "grad_norm": 2.379226475488944,
+      "learning_rate": 7.509553962591359e-07,
+      "loss": 0.0275,
+      "step": 9580
+    },
+    {
+      "epoch": 2.1775,
+      "grad_norm": 1.72365575593215,
+      "learning_rate": 7.508679763554865e-07,
+      "loss": 0.0199,
+      "step": 9581
+    },
+    {
+      "epoch": 2.1777272727272727,
+      "grad_norm": 2.336680544497769,
+      "learning_rate": 7.507805538851691e-07,
+      "loss": 0.0461,
+      "step": 9582
+    },
+    {
+      "epoch": 2.1779545454545453,
+      "grad_norm": 1.833266270999137,
+      "learning_rate": 7.506931288499669e-07,
+      "loss": 0.0255,
+      "step": 9583
+    },
+    {
+      "epoch": 2.178181818181818,
+      "grad_norm": 3.8562873109877276,
+      "learning_rate": 7.506057012516627e-07,
+      "loss": 0.0783,
+      "step": 9584
+    },
+    {
+      "epoch": 2.1784090909090907,
+      "grad_norm": 2.3965565063087713,
+      "learning_rate": 7.505182710920387e-07,
+      "loss": 0.0309,
+      "step": 9585
+    },
+    {
+      "epoch": 2.1786363636363637,
+      "grad_norm": 1.3100982936269245,
+      "learning_rate": 7.504308383728785e-07,
+      "loss": 0.0106,
+      "step": 9586
+    },
+    {
+      "epoch": 2.1788636363636362,
+      "grad_norm": 1.4453082437664924,
+      "learning_rate": 7.503434030959645e-07,
+      "loss": 0.0159,
+      "step": 9587
+    },
+    {
+      "epoch": 2.179090909090909,
+      "grad_norm": 1.4511817719824032,
+      "learning_rate": 7.5025596526308e-07,
+      "loss": 0.0189,
+      "step": 9588
+    },
+    {
+      "epoch": 2.1793181818181817,
+      "grad_norm": 2.6707437778150704,
+      "learning_rate": 7.501685248760078e-07,
+      "loss": 0.0527,
+      "step": 9589
+    },
+    {
+      "epoch": 2.1795454545454547,
+      "grad_norm": 1.8523811390468379,
+      "learning_rate": 7.500810819365312e-07,
+      "loss": 0.0216,
+      "step": 9590
+    },
+    {
+      "epoch": 2.179772727272727,
+      "grad_norm": 1.2741441860332843,
+      "learning_rate": 7.499936364464328e-07,
+      "loss": 0.0128,
+      "step": 9591
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.7259029666267236,
+      "learning_rate": 7.499061884074961e-07,
+      "loss": 0.0297,
+      "step": 9592
+    },
+    {
+      "epoch": 2.1802272727272727,
+      "grad_norm": 2.312575002399278,
+      "learning_rate": 7.498187378215044e-07,
+      "loss": 0.0328,
+      "step": 9593
+    },
+    {
+      "epoch": 2.1804545454545456,
+      "grad_norm": 2.066690292301009,
+      "learning_rate": 7.497312846902409e-07,
+      "loss": 0.0193,
+      "step": 9594
+    },
+    {
+      "epoch": 2.180681818181818,
+      "grad_norm": 3.200244016676321,
+      "learning_rate": 7.496438290154889e-07,
+      "loss": 0.0829,
+      "step": 9595
+    },
+    {
+      "epoch": 2.1809090909090907,
+      "grad_norm": 2.4635730621602967,
+      "learning_rate": 7.495563707990318e-07,
+      "loss": 0.0205,
+      "step": 9596
+    },
+    {
+      "epoch": 2.1811363636363637,
+      "grad_norm": 3.0503195804141923,
+      "learning_rate": 7.49468910042653e-07,
+      "loss": 0.0258,
+      "step": 9597
+    },
+    {
+      "epoch": 2.181363636363636,
+      "grad_norm": 2.713380996005117,
+      "learning_rate": 7.493814467481359e-07,
+      "loss": 0.0789,
+      "step": 9598
+    },
+    {
+      "epoch": 2.181590909090909,
+      "grad_norm": 3.1565733244665797,
+      "learning_rate": 7.492939809172642e-07,
+      "loss": 0.0368,
+      "step": 9599
+    },
+    {
+      "epoch": 2.1818181818181817,
+      "grad_norm": 1.9501567721200437,
+      "learning_rate": 7.492065125518212e-07,
+      "loss": 0.0145,
+      "step": 9600
+    },
+    {
+      "epoch": 2.1820454545454546,
+      "grad_norm": 1.7742928554917743,
+      "learning_rate": 7.491190416535909e-07,
+      "loss": 0.0174,
+      "step": 9601
+    },
+    {
+      "epoch": 2.182272727272727,
+      "grad_norm": 2.0374219563750957,
+      "learning_rate": 7.490315682243568e-07,
+      "loss": 0.0202,
+      "step": 9602
+    },
+    {
+      "epoch": 2.1825,
+      "grad_norm": 2.0604334370695394,
+      "learning_rate": 7.489440922659028e-07,
+      "loss": 0.0242,
+      "step": 9603
+    },
+    {
+      "epoch": 2.1827272727272726,
+      "grad_norm": 2.8430128813742352,
+      "learning_rate": 7.488566137800123e-07,
+      "loss": 0.0443,
+      "step": 9604
+    },
+    {
+      "epoch": 2.1829545454545456,
+      "grad_norm": 2.7069872208906394,
+      "learning_rate": 7.487691327684693e-07,
+      "loss": 0.0377,
+      "step": 9605
+    },
+    {
+      "epoch": 2.183181818181818,
+      "grad_norm": 2.327459060878834,
+      "learning_rate": 7.486816492330579e-07,
+      "loss": 0.063,
+      "step": 9606
+    },
+    {
+      "epoch": 2.183409090909091,
+      "grad_norm": 2.62698105305305,
+      "learning_rate": 7.485941631755618e-07,
+      "loss": 0.0371,
+      "step": 9607
+    },
+    {
+      "epoch": 2.1836363636363636,
+      "grad_norm": 1.4587141067511218,
+      "learning_rate": 7.485066745977651e-07,
+      "loss": 0.0187,
+      "step": 9608
+    },
+    {
+      "epoch": 2.1838636363636366,
+      "grad_norm": 1.8212153290458084,
+      "learning_rate": 7.484191835014518e-07,
+      "loss": 0.0239,
+      "step": 9609
+    },
+    {
+      "epoch": 2.184090909090909,
+      "grad_norm": 1.6437024372464655,
+      "learning_rate": 7.483316898884061e-07,
+      "loss": 0.0243,
+      "step": 9610
+    },
+    {
+      "epoch": 2.184318181818182,
+      "grad_norm": 3.34447697049652,
+      "learning_rate": 7.482441937604121e-07,
+      "loss": 0.0635,
+      "step": 9611
+    },
+    {
+      "epoch": 2.1845454545454546,
+      "grad_norm": 1.8098941926784216,
+      "learning_rate": 7.481566951192541e-07,
+      "loss": 0.0257,
+      "step": 9612
+    },
+    {
+      "epoch": 2.184772727272727,
+      "grad_norm": 2.4811825994551833,
+      "learning_rate": 7.48069193966716e-07,
+      "loss": 0.0725,
+      "step": 9613
+    },
+    {
+      "epoch": 2.185,
+      "grad_norm": 2.7818280141314946,
+      "learning_rate": 7.479816903045824e-07,
+      "loss": 0.0234,
+      "step": 9614
+    },
+    {
+      "epoch": 2.1852272727272726,
+      "grad_norm": 3.31197934949866,
+      "learning_rate": 7.478941841346375e-07,
+      "loss": 0.0591,
+      "step": 9615
+    },
+    {
+      "epoch": 2.1854545454545455,
+      "grad_norm": 1.6104931560110785,
+      "learning_rate": 7.478066754586659e-07,
+      "loss": 0.0233,
+      "step": 9616
+    },
+    {
+      "epoch": 2.185681818181818,
+      "grad_norm": 2.148125533120134,
+      "learning_rate": 7.47719164278452e-07,
+      "loss": 0.0343,
+      "step": 9617
+    },
+    {
+      "epoch": 2.185909090909091,
+      "grad_norm": 2.0077566365383395,
+      "learning_rate": 7.476316505957801e-07,
+      "loss": 0.0287,
+      "step": 9618
+    },
+    {
+      "epoch": 2.1861363636363635,
+      "grad_norm": 2.5270281770617653,
+      "learning_rate": 7.475441344124351e-07,
+      "loss": 0.02,
+      "step": 9619
+    },
+    {
+      "epoch": 2.1863636363636365,
+      "grad_norm": 1.4373609944216141,
+      "learning_rate": 7.474566157302013e-07,
+      "loss": 0.0245,
+      "step": 9620
+    },
+    {
+      "epoch": 2.186590909090909,
+      "grad_norm": 3.4823866882824017,
+      "learning_rate": 7.473690945508636e-07,
+      "loss": 0.028,
+      "step": 9621
+    },
+    {
+      "epoch": 2.186818181818182,
+      "grad_norm": 1.037578544809167,
+      "learning_rate": 7.472815708762063e-07,
+      "loss": 0.0151,
+      "step": 9622
+    },
+    {
+      "epoch": 2.1870454545454545,
+      "grad_norm": 0.8153908806453481,
+      "learning_rate": 7.471940447080145e-07,
+      "loss": 0.0111,
+      "step": 9623
+    },
+    {
+      "epoch": 2.1872727272727275,
+      "grad_norm": 0.37901271235172407,
+      "learning_rate": 7.47106516048073e-07,
+      "loss": 0.0029,
+      "step": 9624
+    },
+    {
+      "epoch": 2.1875,
+      "grad_norm": 2.135449213709016,
+      "learning_rate": 7.470189848981667e-07,
+      "loss": 0.0313,
+      "step": 9625
+    },
+    {
+      "epoch": 2.1877272727272725,
+      "grad_norm": 1.2107220528815645,
+      "learning_rate": 7.469314512600804e-07,
+      "loss": 0.0133,
+      "step": 9626
+    },
+    {
+      "epoch": 2.1879545454545455,
+      "grad_norm": 1.188492505363367,
+      "learning_rate": 7.468439151355988e-07,
+      "loss": 0.0152,
+      "step": 9627
+    },
+    {
+      "epoch": 2.188181818181818,
+      "grad_norm": 2.0684195093658375,
+      "learning_rate": 7.467563765265077e-07,
+      "loss": 0.0368,
+      "step": 9628
+    },
+    {
+      "epoch": 2.188409090909091,
+      "grad_norm": 1.5893339897544332,
+      "learning_rate": 7.466688354345912e-07,
+      "loss": 0.0346,
+      "step": 9629
+    },
+    {
+      "epoch": 2.1886363636363635,
+      "grad_norm": 2.154770230225279,
+      "learning_rate": 7.465812918616351e-07,
+      "loss": 0.0296,
+      "step": 9630
+    },
+    {
+      "epoch": 2.1888636363636365,
+      "grad_norm": 2.0096619246856604,
+      "learning_rate": 7.464937458094243e-07,
+      "loss": 0.0179,
+      "step": 9631
+    },
+    {
+      "epoch": 2.189090909090909,
+      "grad_norm": 1.0649397912512777,
+      "learning_rate": 7.464061972797442e-07,
+      "loss": 0.0087,
+      "step": 9632
+    },
+    {
+      "epoch": 2.189318181818182,
+      "grad_norm": 2.48223111684632,
+      "learning_rate": 7.463186462743797e-07,
+      "loss": 0.0302,
+      "step": 9633
+    },
+    {
+      "epoch": 2.1895454545454545,
+      "grad_norm": 1.6186898773940859,
+      "learning_rate": 7.462310927951167e-07,
+      "loss": 0.0279,
+      "step": 9634
+    },
+    {
+      "epoch": 2.1897727272727274,
+      "grad_norm": 2.3407854763406055,
+      "learning_rate": 7.461435368437399e-07,
+      "loss": 0.0542,
+      "step": 9635
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.73599367269467,
+      "learning_rate": 7.460559784220351e-07,
+      "loss": 0.023,
+      "step": 9636
+    },
+    {
+      "epoch": 2.190227272727273,
+      "grad_norm": 2.219371769880806,
+      "learning_rate": 7.459684175317879e-07,
+      "loss": 0.0231,
+      "step": 9637
+    },
+    {
+      "epoch": 2.1904545454545454,
+      "grad_norm": 1.3729862181901435,
+      "learning_rate": 7.458808541747834e-07,
+      "loss": 0.0214,
+      "step": 9638
+    },
+    {
+      "epoch": 2.190681818181818,
+      "grad_norm": 1.003754837797275,
+      "learning_rate": 7.457932883528074e-07,
+      "loss": 0.0109,
+      "step": 9639
+    },
+    {
+      "epoch": 2.190909090909091,
+      "grad_norm": 1.9500629599219457,
+      "learning_rate": 7.457057200676455e-07,
+      "loss": 0.0213,
+      "step": 9640
+    },
+    {
+      "epoch": 2.1911363636363634,
+      "grad_norm": 3.7001930905295226,
+      "learning_rate": 7.456181493210836e-07,
+      "loss": 0.0274,
+      "step": 9641
+    },
+    {
+      "epoch": 2.1913636363636364,
+      "grad_norm": 0.7373453336125293,
+      "learning_rate": 7.455305761149071e-07,
+      "loss": 0.0082,
+      "step": 9642
+    },
+    {
+      "epoch": 2.191590909090909,
+      "grad_norm": 1.859895089805158,
+      "learning_rate": 7.454430004509018e-07,
+      "loss": 0.0226,
+      "step": 9643
+    },
+    {
+      "epoch": 2.191818181818182,
+      "grad_norm": 1.4763536795163146,
+      "learning_rate": 7.453554223308536e-07,
+      "loss": 0.0178,
+      "step": 9644
+    },
+    {
+      "epoch": 2.1920454545454544,
+      "grad_norm": 1.224606229169292,
+      "learning_rate": 7.452678417565484e-07,
+      "loss": 0.0125,
+      "step": 9645
+    },
+    {
+      "epoch": 2.1922727272727274,
+      "grad_norm": 0.8758393352893433,
+      "learning_rate": 7.451802587297721e-07,
+      "loss": 0.0093,
+      "step": 9646
+    },
+    {
+      "epoch": 2.1925,
+      "grad_norm": 2.300299533738891,
+      "learning_rate": 7.450926732523108e-07,
+      "loss": 0.0295,
+      "step": 9647
+    },
+    {
+      "epoch": 2.192727272727273,
+      "grad_norm": 0.9219741134381937,
+      "learning_rate": 7.450050853259501e-07,
+      "loss": 0.0094,
+      "step": 9648
+    },
+    {
+      "epoch": 2.1929545454545454,
+      "grad_norm": 1.7870427763357584,
+      "learning_rate": 7.449174949524765e-07,
+      "loss": 0.0232,
+      "step": 9649
+    },
+    {
+      "epoch": 2.1931818181818183,
+      "grad_norm": 1.403202895796912,
+      "learning_rate": 7.448299021336759e-07,
+      "loss": 0.0144,
+      "step": 9650
+    },
+    {
+      "epoch": 2.193409090909091,
+      "grad_norm": 4.120343763006819,
+      "learning_rate": 7.447423068713347e-07,
+      "loss": 0.0664,
+      "step": 9651
+    },
+    {
+      "epoch": 2.193636363636364,
+      "grad_norm": 4.190475623844785,
+      "learning_rate": 7.446547091672389e-07,
+      "loss": 0.0398,
+      "step": 9652
+    },
+    {
+      "epoch": 2.1938636363636363,
+      "grad_norm": 1.6947894207108154,
+      "learning_rate": 7.445671090231748e-07,
+      "loss": 0.018,
+      "step": 9653
+    },
+    {
+      "epoch": 2.1940909090909093,
+      "grad_norm": 2.9362856584020354,
+      "learning_rate": 7.444795064409286e-07,
+      "loss": 0.0539,
+      "step": 9654
+    },
+    {
+      "epoch": 2.194318181818182,
+      "grad_norm": 0.9008395095981849,
+      "learning_rate": 7.44391901422287e-07,
+      "loss": 0.0095,
+      "step": 9655
+    },
+    {
+      "epoch": 2.1945454545454544,
+      "grad_norm": 2.069537707106069,
+      "learning_rate": 7.443042939690362e-07,
+      "loss": 0.0121,
+      "step": 9656
+    },
+    {
+      "epoch": 2.1947727272727273,
+      "grad_norm": 1.535562560776091,
+      "learning_rate": 7.442166840829627e-07,
+      "loss": 0.0323,
+      "step": 9657
+    },
+    {
+      "epoch": 2.195,
+      "grad_norm": 2.341981392957456,
+      "learning_rate": 7.441290717658531e-07,
+      "loss": 0.028,
+      "step": 9658
+    },
+    {
+      "epoch": 2.195227272727273,
+      "grad_norm": 3.907617113306359,
+      "learning_rate": 7.440414570194939e-07,
+      "loss": 0.0233,
+      "step": 9659
+    },
+    {
+      "epoch": 2.1954545454545453,
+      "grad_norm": 3.772356371343251,
+      "learning_rate": 7.439538398456715e-07,
+      "loss": 0.0212,
+      "step": 9660
+    },
+    {
+      "epoch": 2.1956818181818183,
+      "grad_norm": 2.9682264426088025,
+      "learning_rate": 7.438662202461728e-07,
+      "loss": 0.024,
+      "step": 9661
+    },
+    {
+      "epoch": 2.195909090909091,
+      "grad_norm": 1.0609287281234525,
+      "learning_rate": 7.437785982227847e-07,
+      "loss": 0.0158,
+      "step": 9662
+    },
+    {
+      "epoch": 2.1961363636363638,
+      "grad_norm": 0.815793593477776,
+      "learning_rate": 7.436909737772938e-07,
+      "loss": 0.0116,
+      "step": 9663
+    },
+    {
+      "epoch": 2.1963636363636363,
+      "grad_norm": 1.609038722170744,
+      "learning_rate": 7.436033469114867e-07,
+      "loss": 0.0244,
+      "step": 9664
+    },
+    {
+      "epoch": 2.1965909090909093,
+      "grad_norm": 2.075384107898709,
+      "learning_rate": 7.435157176271505e-07,
+      "loss": 0.0316,
+      "step": 9665
+    },
+    {
+      "epoch": 2.196818181818182,
+      "grad_norm": 2.778444383568077,
+      "learning_rate": 7.43428085926072e-07,
+      "loss": 0.0395,
+      "step": 9666
+    },
+    {
+      "epoch": 2.1970454545454547,
+      "grad_norm": 0.7212135140798391,
+      "learning_rate": 7.433404518100384e-07,
+      "loss": 0.0136,
+      "step": 9667
+    },
+    {
+      "epoch": 2.1972727272727273,
+      "grad_norm": 2.0832116486629597,
+      "learning_rate": 7.432528152808362e-07,
+      "loss": 0.0204,
+      "step": 9668
+    },
+    {
+      "epoch": 2.1975,
+      "grad_norm": 2.6310492988434233,
+      "learning_rate": 7.431651763402531e-07,
+      "loss": 0.0392,
+      "step": 9669
+    },
+    {
+      "epoch": 2.1977272727272728,
+      "grad_norm": 2.3241024642956534,
+      "learning_rate": 7.430775349900758e-07,
+      "loss": 0.0372,
+      "step": 9670
+    },
+    {
+      "epoch": 2.1979545454545453,
+      "grad_norm": 1.8209008561946671,
+      "learning_rate": 7.429898912320916e-07,
+      "loss": 0.029,
+      "step": 9671
+    },
+    {
+      "epoch": 2.1981818181818182,
+      "grad_norm": 1.7167846542243708,
+      "learning_rate": 7.429022450680876e-07,
+      "loss": 0.0333,
+      "step": 9672
+    },
+    {
+      "epoch": 2.1984090909090908,
+      "grad_norm": 1.5865664886519046,
+      "learning_rate": 7.428145964998512e-07,
+      "loss": 0.0174,
+      "step": 9673
+    },
+    {
+      "epoch": 2.1986363636363637,
+      "grad_norm": 0.8997857476235719,
+      "learning_rate": 7.427269455291697e-07,
+      "loss": 0.0097,
+      "step": 9674
+    },
+    {
+      "epoch": 2.1988636363636362,
+      "grad_norm": 0.6804431397286285,
+      "learning_rate": 7.426392921578302e-07,
+      "loss": 0.0062,
+      "step": 9675
+    },
+    {
+      "epoch": 2.199090909090909,
+      "grad_norm": 1.6931392806120016,
+      "learning_rate": 7.425516363876204e-07,
+      "loss": 0.0185,
+      "step": 9676
+    },
+    {
+      "epoch": 2.1993181818181817,
+      "grad_norm": 1.2444420926090425,
+      "learning_rate": 7.424639782203276e-07,
+      "loss": 0.0165,
+      "step": 9677
+    },
+    {
+      "epoch": 2.1995454545454547,
+      "grad_norm": 1.403136704141494,
+      "learning_rate": 7.423763176577394e-07,
+      "loss": 0.016,
+      "step": 9678
+    },
+    {
+      "epoch": 2.199772727272727,
+      "grad_norm": 2.0211240813524065,
+      "learning_rate": 7.422886547016433e-07,
+      "loss": 0.0207,
+      "step": 9679
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.050494833723123,
+      "learning_rate": 7.422009893538271e-07,
+      "loss": 0.0329,
+      "step": 9680
+    },
+    {
+      "epoch": 2.2002272727272727,
+      "grad_norm": 4.368196424371244,
+      "learning_rate": 7.421133216160781e-07,
+      "loss": 0.055,
+      "step": 9681
+    },
+    {
+      "epoch": 2.200454545454545,
+      "grad_norm": 0.936263375611105,
+      "learning_rate": 7.42025651490184e-07,
+      "loss": 0.0145,
+      "step": 9682
+    },
+    {
+      "epoch": 2.200681818181818,
+      "grad_norm": 2.158431605363282,
+      "learning_rate": 7.419379789779328e-07,
+      "loss": 0.0149,
+      "step": 9683
+    },
+    {
+      "epoch": 2.2009090909090907,
+      "grad_norm": 0.26646553069820506,
+      "learning_rate": 7.418503040811122e-07,
+      "loss": 0.0044,
+      "step": 9684
+    },
+    {
+      "epoch": 2.2011363636363637,
+      "grad_norm": 2.5203469887097927,
+      "learning_rate": 7.417626268015102e-07,
+      "loss": 0.0315,
+      "step": 9685
+    },
+    {
+      "epoch": 2.201363636363636,
+      "grad_norm": 0.847785702752143,
+      "learning_rate": 7.416749471409144e-07,
+      "loss": 0.0047,
+      "step": 9686
+    },
+    {
+      "epoch": 2.201590909090909,
+      "grad_norm": 1.8178397718758441,
+      "learning_rate": 7.41587265101113e-07,
+      "loss": 0.0213,
+      "step": 9687
+    },
+    {
+      "epoch": 2.2018181818181817,
+      "grad_norm": 2.375495823834582,
+      "learning_rate": 7.414995806838937e-07,
+      "loss": 0.0495,
+      "step": 9688
+    },
+    {
+      "epoch": 2.2020454545454546,
+      "grad_norm": 4.642741372006696,
+      "learning_rate": 7.414118938910448e-07,
+      "loss": 0.0368,
+      "step": 9689
+    },
+    {
+      "epoch": 2.202272727272727,
+      "grad_norm": 2.45069743928174,
+      "learning_rate": 7.413242047243543e-07,
+      "loss": 0.0367,
+      "step": 9690
+    },
+    {
+      "epoch": 2.2025,
+      "grad_norm": 3.14101802906977,
+      "learning_rate": 7.412365131856103e-07,
+      "loss": 0.0359,
+      "step": 9691
+    },
+    {
+      "epoch": 2.2027272727272726,
+      "grad_norm": 1.3408509447519532,
+      "learning_rate": 7.411488192766009e-07,
+      "loss": 0.0202,
+      "step": 9692
+    },
+    {
+      "epoch": 2.2029545454545456,
+      "grad_norm": 4.022075762911413,
+      "learning_rate": 7.410611229991146e-07,
+      "loss": 0.0369,
+      "step": 9693
+    },
+    {
+      "epoch": 2.203181818181818,
+      "grad_norm": 6.281377246510745,
+      "learning_rate": 7.409734243549394e-07,
+      "loss": 0.0717,
+      "step": 9694
+    },
+    {
+      "epoch": 2.203409090909091,
+      "grad_norm": 1.6545639426501328,
+      "learning_rate": 7.408857233458638e-07,
+      "loss": 0.0123,
+      "step": 9695
+    },
+    {
+      "epoch": 2.2036363636363636,
+      "grad_norm": 3.3266475060987832,
+      "learning_rate": 7.407980199736762e-07,
+      "loss": 0.0376,
+      "step": 9696
+    },
+    {
+      "epoch": 2.2038636363636366,
+      "grad_norm": 1.7402938730255864,
+      "learning_rate": 7.407103142401651e-07,
+      "loss": 0.0124,
+      "step": 9697
+    },
+    {
+      "epoch": 2.204090909090909,
+      "grad_norm": 1.9215138297473455,
+      "learning_rate": 7.406226061471185e-07,
+      "loss": 0.0213,
+      "step": 9698
+    },
+    {
+      "epoch": 2.2043181818181816,
+      "grad_norm": 1.6739819801504026,
+      "learning_rate": 7.405348956963252e-07,
+      "loss": 0.0098,
+      "step": 9699
+    },
+    {
+      "epoch": 2.2045454545454546,
+      "grad_norm": 1.2039929051379044,
+      "learning_rate": 7.404471828895741e-07,
+      "loss": 0.0088,
+      "step": 9700
+    },
+    {
+      "epoch": 2.204772727272727,
+      "grad_norm": 1.6077305620551483,
+      "learning_rate": 7.403594677286534e-07,
+      "loss": 0.0216,
+      "step": 9701
+    },
+    {
+      "epoch": 2.205,
+      "grad_norm": 1.8235018593995038,
+      "learning_rate": 7.402717502153519e-07,
+      "loss": 0.0238,
+      "step": 9702
+    },
+    {
+      "epoch": 2.2052272727272726,
+      "grad_norm": 1.906628737355214,
+      "learning_rate": 7.401840303514584e-07,
+      "loss": 0.0218,
+      "step": 9703
+    },
+    {
+      "epoch": 2.2054545454545456,
+      "grad_norm": 0.9775457873625856,
+      "learning_rate": 7.400963081387616e-07,
+      "loss": 0.0141,
+      "step": 9704
+    },
+    {
+      "epoch": 2.205681818181818,
+      "grad_norm": 1.950617446519274,
+      "learning_rate": 7.400085835790502e-07,
+      "loss": 0.0295,
+      "step": 9705
+    },
+    {
+      "epoch": 2.205909090909091,
+      "grad_norm": 2.8993491951475394,
+      "learning_rate": 7.39920856674113e-07,
+      "loss": 0.0514,
+      "step": 9706
+    },
+    {
+      "epoch": 2.2061363636363636,
+      "grad_norm": 2.318895796200592,
+      "learning_rate": 7.398331274257392e-07,
+      "loss": 0.0187,
+      "step": 9707
+    },
+    {
+      "epoch": 2.2063636363636365,
+      "grad_norm": 1.9476239731737148,
+      "learning_rate": 7.397453958357174e-07,
+      "loss": 0.0239,
+      "step": 9708
+    },
+    {
+      "epoch": 2.206590909090909,
+      "grad_norm": 1.9910592827871314,
+      "learning_rate": 7.396576619058371e-07,
+      "loss": 0.0318,
+      "step": 9709
+    },
+    {
+      "epoch": 2.206818181818182,
+      "grad_norm": 1.9242996877965826,
+      "learning_rate": 7.395699256378869e-07,
+      "loss": 0.0316,
+      "step": 9710
+    },
+    {
+      "epoch": 2.2070454545454545,
+      "grad_norm": 2.429027978799468,
+      "learning_rate": 7.394821870336561e-07,
+      "loss": 0.0175,
+      "step": 9711
+    },
+    {
+      "epoch": 2.207272727272727,
+      "grad_norm": 2.913250559185877,
+      "learning_rate": 7.393944460949336e-07,
+      "loss": 0.0226,
+      "step": 9712
+    },
+    {
+      "epoch": 2.2075,
+      "grad_norm": 2.180957210432053,
+      "learning_rate": 7.393067028235089e-07,
+      "loss": 0.0221,
+      "step": 9713
+    },
+    {
+      "epoch": 2.2077272727272725,
+      "grad_norm": 1.6715221273328946,
+      "learning_rate": 7.39218957221171e-07,
+      "loss": 0.0248,
+      "step": 9714
+    },
+    {
+      "epoch": 2.2079545454545455,
+      "grad_norm": 2.008465879095117,
+      "learning_rate": 7.391312092897095e-07,
+      "loss": 0.0276,
+      "step": 9715
+    },
+    {
+      "epoch": 2.208181818181818,
+      "grad_norm": 1.1530506461709946,
+      "learning_rate": 7.390434590309135e-07,
+      "loss": 0.0113,
+      "step": 9716
+    },
+    {
+      "epoch": 2.208409090909091,
+      "grad_norm": 1.2655179103772258,
+      "learning_rate": 7.389557064465722e-07,
+      "loss": 0.0099,
+      "step": 9717
+    },
+    {
+      "epoch": 2.2086363636363635,
+      "grad_norm": 3.125280114812933,
+      "learning_rate": 7.388679515384754e-07,
+      "loss": 0.0708,
+      "step": 9718
+    },
+    {
+      "epoch": 2.2088636363636365,
+      "grad_norm": 2.036047612863548,
+      "learning_rate": 7.387801943084123e-07,
+      "loss": 0.032,
+      "step": 9719
+    },
+    {
+      "epoch": 2.209090909090909,
+      "grad_norm": 2.3257207214868987,
+      "learning_rate": 7.386924347581728e-07,
+      "loss": 0.0219,
+      "step": 9720
+    },
+    {
+      "epoch": 2.209318181818182,
+      "grad_norm": 2.1109937951902085,
+      "learning_rate": 7.386046728895459e-07,
+      "loss": 0.0256,
+      "step": 9721
+    },
+    {
+      "epoch": 2.2095454545454545,
+      "grad_norm": 2.5589013362785447,
+      "learning_rate": 7.385169087043216e-07,
+      "loss": 0.0571,
+      "step": 9722
+    },
+    {
+      "epoch": 2.2097727272727274,
+      "grad_norm": 1.5700217470404685,
+      "learning_rate": 7.384291422042896e-07,
+      "loss": 0.0255,
+      "step": 9723
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.1402325542958776,
+      "learning_rate": 7.383413733912395e-07,
+      "loss": 0.0148,
+      "step": 9724
+    },
+    {
+      "epoch": 2.210227272727273,
+      "grad_norm": 1.781397540180257,
+      "learning_rate": 7.382536022669609e-07,
+      "loss": 0.0265,
+      "step": 9725
+    },
+    {
+      "epoch": 2.2104545454545454,
+      "grad_norm": 1.6796395452017117,
+      "learning_rate": 7.38165828833244e-07,
+      "loss": 0.0273,
+      "step": 9726
+    },
+    {
+      "epoch": 2.2106818181818184,
+      "grad_norm": 2.073148026673671,
+      "learning_rate": 7.380780530918783e-07,
+      "loss": 0.0149,
+      "step": 9727
+    },
+    {
+      "epoch": 2.210909090909091,
+      "grad_norm": 1.9581501722456238,
+      "learning_rate": 7.379902750446537e-07,
+      "loss": 0.0193,
+      "step": 9728
+    },
+    {
+      "epoch": 2.2111363636363635,
+      "grad_norm": 3.5075272536404776,
+      "learning_rate": 7.379024946933603e-07,
+      "loss": 0.0407,
+      "step": 9729
+    },
+    {
+      "epoch": 2.2113636363636364,
+      "grad_norm": 0.8615865858737627,
+      "learning_rate": 7.378147120397882e-07,
+      "loss": 0.0196,
+      "step": 9730
+    },
+    {
+      "epoch": 2.211590909090909,
+      "grad_norm": 3.268746264436619,
+      "learning_rate": 7.377269270857274e-07,
+      "loss": 0.0537,
+      "step": 9731
+    },
+    {
+      "epoch": 2.211818181818182,
+      "grad_norm": 2.1799304963163824,
+      "learning_rate": 7.376391398329677e-07,
+      "loss": 0.028,
+      "step": 9732
+    },
+    {
+      "epoch": 2.2120454545454544,
+      "grad_norm": 2.165256276863638,
+      "learning_rate": 7.375513502832995e-07,
+      "loss": 0.0514,
+      "step": 9733
+    },
+    {
+      "epoch": 2.2122727272727274,
+      "grad_norm": 2.9061997480979493,
+      "learning_rate": 7.374635584385129e-07,
+      "loss": 0.0292,
+      "step": 9734
+    },
+    {
+      "epoch": 2.2125,
+      "grad_norm": 1.2233683117038714,
+      "learning_rate": 7.373757643003981e-07,
+      "loss": 0.0113,
+      "step": 9735
+    },
+    {
+      "epoch": 2.212727272727273,
+      "grad_norm": 1.3298477108928586,
+      "learning_rate": 7.372879678707455e-07,
+      "loss": 0.0218,
+      "step": 9736
+    },
+    {
+      "epoch": 2.2129545454545454,
+      "grad_norm": 1.9931470502533886,
+      "learning_rate": 7.372001691513453e-07,
+      "loss": 0.0222,
+      "step": 9737
+    },
+    {
+      "epoch": 2.2131818181818184,
+      "grad_norm": 1.4073515621255932,
+      "learning_rate": 7.371123681439877e-07,
+      "loss": 0.021,
+      "step": 9738
+    },
+    {
+      "epoch": 2.213409090909091,
+      "grad_norm": 2.1840422234302994,
+      "learning_rate": 7.370245648504636e-07,
+      "loss": 0.0189,
+      "step": 9739
+    },
+    {
+      "epoch": 2.213636363636364,
+      "grad_norm": 1.7486339950296887,
+      "learning_rate": 7.369367592725632e-07,
+      "loss": 0.015,
+      "step": 9740
+    },
+    {
+      "epoch": 2.2138636363636364,
+      "grad_norm": 1.4415002830181673,
+      "learning_rate": 7.368489514120768e-07,
+      "loss": 0.0135,
+      "step": 9741
+    },
+    {
+      "epoch": 2.214090909090909,
+      "grad_norm": 1.4282899900700614,
+      "learning_rate": 7.367611412707953e-07,
+      "loss": 0.0236,
+      "step": 9742
+    },
+    {
+      "epoch": 2.214318181818182,
+      "grad_norm": 1.6940902110164877,
+      "learning_rate": 7.366733288505091e-07,
+      "loss": 0.0184,
+      "step": 9743
+    },
+    {
+      "epoch": 2.2145454545454544,
+      "grad_norm": 2.6349864352180847,
+      "learning_rate": 7.365855141530088e-07,
+      "loss": 0.0323,
+      "step": 9744
+    },
+    {
+      "epoch": 2.2147727272727273,
+      "grad_norm": 1.342196827796498,
+      "learning_rate": 7.364976971800852e-07,
+      "loss": 0.0206,
+      "step": 9745
+    },
+    {
+      "epoch": 2.215,
+      "grad_norm": 2.5819483408154063,
+      "learning_rate": 7.364098779335294e-07,
+      "loss": 0.0326,
+      "step": 9746
+    },
+    {
+      "epoch": 2.215227272727273,
+      "grad_norm": 2.5558035332552858,
+      "learning_rate": 7.363220564151313e-07,
+      "loss": 0.0312,
+      "step": 9747
+    },
+    {
+      "epoch": 2.2154545454545453,
+      "grad_norm": 1.9767427808720013,
+      "learning_rate": 7.362342326266827e-07,
+      "loss": 0.0233,
+      "step": 9748
+    },
+    {
+      "epoch": 2.2156818181818183,
+      "grad_norm": 1.8353683370021416,
+      "learning_rate": 7.361464065699739e-07,
+      "loss": 0.0128,
+      "step": 9749
+    },
+    {
+      "epoch": 2.215909090909091,
+      "grad_norm": 1.0544486060838694,
+      "learning_rate": 7.360585782467958e-07,
+      "loss": 0.0095,
+      "step": 9750
+    },
+    {
+      "epoch": 2.216136363636364,
+      "grad_norm": 2.519672546819916,
+      "learning_rate": 7.359707476589398e-07,
+      "loss": 0.0407,
+      "step": 9751
+    },
+    {
+      "epoch": 2.2163636363636363,
+      "grad_norm": 1.157714155267152,
+      "learning_rate": 7.358829148081963e-07,
+      "loss": 0.0132,
+      "step": 9752
+    },
+    {
+      "epoch": 2.2165909090909093,
+      "grad_norm": 1.1175272506728786,
+      "learning_rate": 7.35795079696357e-07,
+      "loss": 0.0126,
+      "step": 9753
+    },
+    {
+      "epoch": 2.216818181818182,
+      "grad_norm": 1.318150517732887,
+      "learning_rate": 7.357072423252126e-07,
+      "loss": 0.014,
+      "step": 9754
+    },
+    {
+      "epoch": 2.2170454545454543,
+      "grad_norm": 1.7649368018480278,
+      "learning_rate": 7.356194026965544e-07,
+      "loss": 0.0309,
+      "step": 9755
+    },
+    {
+      "epoch": 2.2172727272727273,
+      "grad_norm": 0.8376943772570926,
+      "learning_rate": 7.355315608121736e-07,
+      "loss": 0.0048,
+      "step": 9756
+    },
+    {
+      "epoch": 2.2175,
+      "grad_norm": 0.7108441983454811,
+      "learning_rate": 7.354437166738617e-07,
+      "loss": 0.0073,
+      "step": 9757
+    },
+    {
+      "epoch": 2.2177272727272728,
+      "grad_norm": 2.122769756712854,
+      "learning_rate": 7.353558702834095e-07,
+      "loss": 0.0214,
+      "step": 9758
+    },
+    {
+      "epoch": 2.2179545454545453,
+      "grad_norm": 1.664770919914316,
+      "learning_rate": 7.352680216426084e-07,
+      "loss": 0.032,
+      "step": 9759
+    },
+    {
+      "epoch": 2.2181818181818183,
+      "grad_norm": 1.7690171693188899,
+      "learning_rate": 7.351801707532502e-07,
+      "loss": 0.0266,
+      "step": 9760
+    },
+    {
+      "epoch": 2.2184090909090908,
+      "grad_norm": 3.0095584835315985,
+      "learning_rate": 7.35092317617126e-07,
+      "loss": 0.028,
+      "step": 9761
+    },
+    {
+      "epoch": 2.2186363636363637,
+      "grad_norm": 1.126361028358107,
+      "learning_rate": 7.350044622360275e-07,
+      "loss": 0.0115,
+      "step": 9762
+    },
+    {
+      "epoch": 2.2188636363636363,
+      "grad_norm": 1.5801998300793894,
+      "learning_rate": 7.349166046117461e-07,
+      "loss": 0.0124,
+      "step": 9763
+    },
+    {
+      "epoch": 2.2190909090909092,
+      "grad_norm": 2.175722340875782,
+      "learning_rate": 7.348287447460733e-07,
+      "loss": 0.0324,
+      "step": 9764
+    },
+    {
+      "epoch": 2.2193181818181817,
+      "grad_norm": 2.816303716976785,
+      "learning_rate": 7.347408826408009e-07,
+      "loss": 0.0257,
+      "step": 9765
+    },
+    {
+      "epoch": 2.2195454545454547,
+      "grad_norm": 1.6680790932976393,
+      "learning_rate": 7.346530182977203e-07,
+      "loss": 0.0247,
+      "step": 9766
+    },
+    {
+      "epoch": 2.2197727272727272,
+      "grad_norm": 1.9882001840498265,
+      "learning_rate": 7.345651517186233e-07,
+      "loss": 0.0263,
+      "step": 9767
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 2.028790315113887,
+      "learning_rate": 7.344772829053019e-07,
+      "loss": 0.0283,
+      "step": 9768
+    },
+    {
+      "epoch": 2.2202272727272727,
+      "grad_norm": 1.588614558446675,
+      "learning_rate": 7.343894118595477e-07,
+      "loss": 0.0208,
+      "step": 9769
+    },
+    {
+      "epoch": 2.2204545454545457,
+      "grad_norm": 1.617805995170962,
+      "learning_rate": 7.343015385831524e-07,
+      "loss": 0.0187,
+      "step": 9770
+    },
+    {
+      "epoch": 2.220681818181818,
+      "grad_norm": 2.6268506335067228,
+      "learning_rate": 7.342136630779082e-07,
+      "loss": 0.0226,
+      "step": 9771
+    },
+    {
+      "epoch": 2.2209090909090907,
+      "grad_norm": 2.637339643179205,
+      "learning_rate": 7.341257853456067e-07,
+      "loss": 0.0225,
+      "step": 9772
+    },
+    {
+      "epoch": 2.2211363636363637,
+      "grad_norm": 2.077927153431144,
+      "learning_rate": 7.340379053880401e-07,
+      "loss": 0.0314,
+      "step": 9773
+    },
+    {
+      "epoch": 2.221363636363636,
+      "grad_norm": 1.0062953178009233,
+      "learning_rate": 7.339500232070006e-07,
+      "loss": 0.0203,
+      "step": 9774
+    },
+    {
+      "epoch": 2.221590909090909,
+      "grad_norm": 1.1807495736865379,
+      "learning_rate": 7.338621388042798e-07,
+      "loss": 0.0086,
+      "step": 9775
+    },
+    {
+      "epoch": 2.2218181818181817,
+      "grad_norm": 1.8988799724813858,
+      "learning_rate": 7.3377425218167e-07,
+      "loss": 0.0387,
+      "step": 9776
+    },
+    {
+      "epoch": 2.2220454545454547,
+      "grad_norm": 2.0641791440051227,
+      "learning_rate": 7.336863633409637e-07,
+      "loss": 0.0227,
+      "step": 9777
+    },
+    {
+      "epoch": 2.222272727272727,
+      "grad_norm": 1.9012130286841387,
+      "learning_rate": 7.335984722839528e-07,
+      "loss": 0.0217,
+      "step": 9778
+    },
+    {
+      "epoch": 2.2225,
+      "grad_norm": 2.4784701744170996,
+      "learning_rate": 7.335105790124295e-07,
+      "loss": 0.0349,
+      "step": 9779
+    },
+    {
+      "epoch": 2.2227272727272727,
+      "grad_norm": 2.284977219327871,
+      "learning_rate": 7.334226835281862e-07,
+      "loss": 0.0277,
+      "step": 9780
+    },
+    {
+      "epoch": 2.2229545454545456,
+      "grad_norm": 2.0026787892316604,
+      "learning_rate": 7.333347858330153e-07,
+      "loss": 0.0203,
+      "step": 9781
+    },
+    {
+      "epoch": 2.223181818181818,
+      "grad_norm": 1.7630424288009097,
+      "learning_rate": 7.332468859287088e-07,
+      "loss": 0.018,
+      "step": 9782
+    },
+    {
+      "epoch": 2.223409090909091,
+      "grad_norm": 1.9974391837424654,
+      "learning_rate": 7.331589838170597e-07,
+      "loss": 0.0267,
+      "step": 9783
+    },
+    {
+      "epoch": 2.2236363636363636,
+      "grad_norm": 1.6098816189458185,
+      "learning_rate": 7.330710794998603e-07,
+      "loss": 0.0212,
+      "step": 9784
+    },
+    {
+      "epoch": 2.223863636363636,
+      "grad_norm": 2.492513620583543,
+      "learning_rate": 7.329831729789029e-07,
+      "loss": 0.0399,
+      "step": 9785
+    },
+    {
+      "epoch": 2.224090909090909,
+      "grad_norm": 2.403410599333524,
+      "learning_rate": 7.328952642559804e-07,
+      "loss": 0.0363,
+      "step": 9786
+    },
+    {
+      "epoch": 2.2243181818181816,
+      "grad_norm": 1.3827579556863292,
+      "learning_rate": 7.328073533328851e-07,
+      "loss": 0.0084,
+      "step": 9787
+    },
+    {
+      "epoch": 2.2245454545454546,
+      "grad_norm": 2.7679936637599636,
+      "learning_rate": 7.327194402114099e-07,
+      "loss": 0.067,
+      "step": 9788
+    },
+    {
+      "epoch": 2.224772727272727,
+      "grad_norm": 1.6090921597086534,
+      "learning_rate": 7.326315248933471e-07,
+      "loss": 0.025,
+      "step": 9789
+    },
+    {
+      "epoch": 2.225,
+      "grad_norm": 1.902218284755912,
+      "learning_rate": 7.3254360738049e-07,
+      "loss": 0.0215,
+      "step": 9790
+    },
+    {
+      "epoch": 2.2252272727272726,
+      "grad_norm": 2.209820366449165,
+      "learning_rate": 7.32455687674631e-07,
+      "loss": 0.0164,
+      "step": 9791
+    },
+    {
+      "epoch": 2.2254545454545456,
+      "grad_norm": 1.7556781846132337,
+      "learning_rate": 7.323677657775633e-07,
+      "loss": 0.042,
+      "step": 9792
+    },
+    {
+      "epoch": 2.225681818181818,
+      "grad_norm": 1.6288646291456752,
+      "learning_rate": 7.322798416910792e-07,
+      "loss": 0.0256,
+      "step": 9793
+    },
+    {
+      "epoch": 2.225909090909091,
+      "grad_norm": 2.957279725353443,
+      "learning_rate": 7.321919154169721e-07,
+      "loss": 0.0329,
+      "step": 9794
+    },
+    {
+      "epoch": 2.2261363636363636,
+      "grad_norm": 1.6980794765165899,
+      "learning_rate": 7.32103986957035e-07,
+      "loss": 0.0266,
+      "step": 9795
+    },
+    {
+      "epoch": 2.2263636363636365,
+      "grad_norm": 2.2542503007243697,
+      "learning_rate": 7.320160563130605e-07,
+      "loss": 0.026,
+      "step": 9796
+    },
+    {
+      "epoch": 2.226590909090909,
+      "grad_norm": 2.512868206150587,
+      "learning_rate": 7.319281234868419e-07,
+      "loss": 0.029,
+      "step": 9797
+    },
+    {
+      "epoch": 2.2268181818181816,
+      "grad_norm": 1.5688122827749618,
+      "learning_rate": 7.318401884801725e-07,
+      "loss": 0.0215,
+      "step": 9798
+    },
+    {
+      "epoch": 2.2270454545454546,
+      "grad_norm": 0.8586651657190552,
+      "learning_rate": 7.317522512948452e-07,
+      "loss": 0.007,
+      "step": 9799
+    },
+    {
+      "epoch": 2.227272727272727,
+      "grad_norm": 1.7240440976792066,
+      "learning_rate": 7.316643119326533e-07,
+      "loss": 0.0152,
+      "step": 9800
+    },
+    {
+      "epoch": 2.2275,
+      "grad_norm": 1.711938982471081,
+      "learning_rate": 7.3157637039539e-07,
+      "loss": 0.0116,
+      "step": 9801
+    },
+    {
+      "epoch": 2.2277272727272726,
+      "grad_norm": 2.0273483561422356,
+      "learning_rate": 7.314884266848485e-07,
+      "loss": 0.0308,
+      "step": 9802
+    },
+    {
+      "epoch": 2.2279545454545455,
+      "grad_norm": 2.32330293779499,
+      "learning_rate": 7.314004808028223e-07,
+      "loss": 0.031,
+      "step": 9803
+    },
+    {
+      "epoch": 2.228181818181818,
+      "grad_norm": 1.6096790922216617,
+      "learning_rate": 7.313125327511046e-07,
+      "loss": 0.0131,
+      "step": 9804
+    },
+    {
+      "epoch": 2.228409090909091,
+      "grad_norm": 2.486861768708967,
+      "learning_rate": 7.312245825314888e-07,
+      "loss": 0.0531,
+      "step": 9805
+    },
+    {
+      "epoch": 2.2286363636363635,
+      "grad_norm": 1.5044705584650266,
+      "learning_rate": 7.311366301457684e-07,
+      "loss": 0.0136,
+      "step": 9806
+    },
+    {
+      "epoch": 2.2288636363636365,
+      "grad_norm": 1.8468884032977335,
+      "learning_rate": 7.310486755957372e-07,
+      "loss": 0.0406,
+      "step": 9807
+    },
+    {
+      "epoch": 2.229090909090909,
+      "grad_norm": 1.7724212760028772,
+      "learning_rate": 7.309607188831884e-07,
+      "loss": 0.029,
+      "step": 9808
+    },
+    {
+      "epoch": 2.229318181818182,
+      "grad_norm": 2.6656445432509175,
+      "learning_rate": 7.308727600099155e-07,
+      "loss": 0.0389,
+      "step": 9809
+    },
+    {
+      "epoch": 2.2295454545454545,
+      "grad_norm": 1.2925629863426957,
+      "learning_rate": 7.307847989777125e-07,
+      "loss": 0.0221,
+      "step": 9810
+    },
+    {
+      "epoch": 2.2297727272727275,
+      "grad_norm": 1.820889765859273,
+      "learning_rate": 7.306968357883729e-07,
+      "loss": 0.0283,
+      "step": 9811
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.8886935396265105,
+      "learning_rate": 7.306088704436903e-07,
+      "loss": 0.0217,
+      "step": 9812
+    },
+    {
+      "epoch": 2.230227272727273,
+      "grad_norm": 3.2418206287687243,
+      "learning_rate": 7.305209029454586e-07,
+      "loss": 0.0665,
+      "step": 9813
+    },
+    {
+      "epoch": 2.2304545454545455,
+      "grad_norm": 0.9793650384957027,
+      "learning_rate": 7.304329332954718e-07,
+      "loss": 0.0073,
+      "step": 9814
+    },
+    {
+      "epoch": 2.230681818181818,
+      "grad_norm": 2.06097076862386,
+      "learning_rate": 7.303449614955234e-07,
+      "loss": 0.0334,
+      "step": 9815
+    },
+    {
+      "epoch": 2.230909090909091,
+      "grad_norm": 2.187587948845686,
+      "learning_rate": 7.302569875474076e-07,
+      "loss": 0.0253,
+      "step": 9816
+    },
+    {
+      "epoch": 2.2311363636363635,
+      "grad_norm": 3.9442293151886436,
+      "learning_rate": 7.301690114529181e-07,
+      "loss": 0.0465,
+      "step": 9817
+    },
+    {
+      "epoch": 2.2313636363636364,
+      "grad_norm": 1.9098681970546232,
+      "learning_rate": 7.300810332138488e-07,
+      "loss": 0.0298,
+      "step": 9818
+    },
+    {
+      "epoch": 2.231590909090909,
+      "grad_norm": 1.9102548946827405,
+      "learning_rate": 7.299930528319942e-07,
+      "loss": 0.0198,
+      "step": 9819
+    },
+    {
+      "epoch": 2.231818181818182,
+      "grad_norm": 1.6839987992284666,
+      "learning_rate": 7.299050703091478e-07,
+      "loss": 0.0337,
+      "step": 9820
+    },
+    {
+      "epoch": 2.2320454545454544,
+      "grad_norm": 1.6244350564129508,
+      "learning_rate": 7.298170856471042e-07,
+      "loss": 0.0169,
+      "step": 9821
+    },
+    {
+      "epoch": 2.2322727272727274,
+      "grad_norm": 1.8261078217537419,
+      "learning_rate": 7.297290988476572e-07,
+      "loss": 0.015,
+      "step": 9822
+    },
+    {
+      "epoch": 2.2325,
+      "grad_norm": 2.7231861606051777,
+      "learning_rate": 7.296411099126014e-07,
+      "loss": 0.0422,
+      "step": 9823
+    },
+    {
+      "epoch": 2.232727272727273,
+      "grad_norm": 2.0102051451083867,
+      "learning_rate": 7.295531188437305e-07,
+      "loss": 0.0105,
+      "step": 9824
+    },
+    {
+      "epoch": 2.2329545454545454,
+      "grad_norm": 1.6516364897803086,
+      "learning_rate": 7.294651256428393e-07,
+      "loss": 0.03,
+      "step": 9825
+    },
+    {
+      "epoch": 2.2331818181818184,
+      "grad_norm": 2.767440251004509,
+      "learning_rate": 7.29377130311722e-07,
+      "loss": 0.0203,
+      "step": 9826
+    },
+    {
+      "epoch": 2.233409090909091,
+      "grad_norm": 3.7583330756205,
+      "learning_rate": 7.292891328521727e-07,
+      "loss": 0.0495,
+      "step": 9827
+    },
+    {
+      "epoch": 2.2336363636363634,
+      "grad_norm": 1.1107568269257517,
+      "learning_rate": 7.292011332659863e-07,
+      "loss": 0.016,
+      "step": 9828
+    },
+    {
+      "epoch": 2.2338636363636364,
+      "grad_norm": 1.4753062355192894,
+      "learning_rate": 7.291131315549567e-07,
+      "loss": 0.0156,
+      "step": 9829
+    },
+    {
+      "epoch": 2.234090909090909,
+      "grad_norm": 2.8621372720789195,
+      "learning_rate": 7.29025127720879e-07,
+      "loss": 0.036,
+      "step": 9830
+    },
+    {
+      "epoch": 2.234318181818182,
+      "grad_norm": 1.0895240276891107,
+      "learning_rate": 7.289371217655473e-07,
+      "loss": 0.0105,
+      "step": 9831
+    },
+    {
+      "epoch": 2.2345454545454544,
+      "grad_norm": 1.110669432829355,
+      "learning_rate": 7.288491136907565e-07,
+      "loss": 0.0181,
+      "step": 9832
+    },
+    {
+      "epoch": 2.2347727272727274,
+      "grad_norm": 1.3615578161808317,
+      "learning_rate": 7.28761103498301e-07,
+      "loss": 0.0218,
+      "step": 9833
+    },
+    {
+      "epoch": 2.235,
+      "grad_norm": 1.635258134508564,
+      "learning_rate": 7.286730911899756e-07,
+      "loss": 0.0222,
+      "step": 9834
+    },
+    {
+      "epoch": 2.235227272727273,
+      "grad_norm": 2.107720273224647,
+      "learning_rate": 7.285850767675749e-07,
+      "loss": 0.0346,
+      "step": 9835
+    },
+    {
+      "epoch": 2.2354545454545454,
+      "grad_norm": 1.3053476446200525,
+      "learning_rate": 7.28497060232894e-07,
+      "loss": 0.0145,
+      "step": 9836
+    },
+    {
+      "epoch": 2.2356818181818183,
+      "grad_norm": 2.4410838890151574,
+      "learning_rate": 7.284090415877272e-07,
+      "loss": 0.0378,
+      "step": 9837
+    },
+    {
+      "epoch": 2.235909090909091,
+      "grad_norm": 3.399332064238748,
+      "learning_rate": 7.283210208338698e-07,
+      "loss": 0.0353,
+      "step": 9838
+    },
+    {
+      "epoch": 2.236136363636364,
+      "grad_norm": 1.2261538021982874,
+      "learning_rate": 7.282329979731165e-07,
+      "loss": 0.0176,
+      "step": 9839
+    },
+    {
+      "epoch": 2.2363636363636363,
+      "grad_norm": 1.6186032595856743,
+      "learning_rate": 7.281449730072622e-07,
+      "loss": 0.0262,
+      "step": 9840
+    },
+    {
+      "epoch": 2.2365909090909093,
+      "grad_norm": 3.0117739524796465,
+      "learning_rate": 7.280569459381022e-07,
+      "loss": 0.021,
+      "step": 9841
+    },
+    {
+      "epoch": 2.236818181818182,
+      "grad_norm": 0.9545987297819722,
+      "learning_rate": 7.27968916767431e-07,
+      "loss": 0.0136,
+      "step": 9842
+    },
+    {
+      "epoch": 2.237045454545455,
+      "grad_norm": 2.061553015333028,
+      "learning_rate": 7.27880885497044e-07,
+      "loss": 0.0171,
+      "step": 9843
+    },
+    {
+      "epoch": 2.2372727272727273,
+      "grad_norm": 3.2170440233071105,
+      "learning_rate": 7.277928521287363e-07,
+      "loss": 0.0225,
+      "step": 9844
+    },
+    {
+      "epoch": 2.2375,
+      "grad_norm": 1.820683571441575,
+      "learning_rate": 7.27704816664303e-07,
+      "loss": 0.0299,
+      "step": 9845
+    },
+    {
+      "epoch": 2.237727272727273,
+      "grad_norm": 2.2095199438809745,
+      "learning_rate": 7.276167791055394e-07,
+      "loss": 0.0201,
+      "step": 9846
+    },
+    {
+      "epoch": 2.2379545454545453,
+      "grad_norm": 1.782658701472118,
+      "learning_rate": 7.275287394542408e-07,
+      "loss": 0.0238,
+      "step": 9847
+    },
+    {
+      "epoch": 2.2381818181818183,
+      "grad_norm": 2.224517008469978,
+      "learning_rate": 7.27440697712202e-07,
+      "loss": 0.0266,
+      "step": 9848
+    },
+    {
+      "epoch": 2.238409090909091,
+      "grad_norm": 2.5629079499266956,
+      "learning_rate": 7.273526538812188e-07,
+      "loss": 0.0397,
+      "step": 9849
+    },
+    {
+      "epoch": 2.2386363636363638,
+      "grad_norm": 2.0583496397674943,
+      "learning_rate": 7.272646079630865e-07,
+      "loss": 0.0258,
+      "step": 9850
+    },
+    {
+      "epoch": 2.2388636363636363,
+      "grad_norm": 1.6989178700555634,
+      "learning_rate": 7.271765599596001e-07,
+      "loss": 0.0293,
+      "step": 9851
+    },
+    {
+      "epoch": 2.2390909090909092,
+      "grad_norm": 1.6890850026689614,
+      "learning_rate": 7.270885098725557e-07,
+      "loss": 0.0298,
+      "step": 9852
+    },
+    {
+      "epoch": 2.2393181818181818,
+      "grad_norm": 0.9989719638270679,
+      "learning_rate": 7.270004577037483e-07,
+      "loss": 0.0116,
+      "step": 9853
+    },
+    {
+      "epoch": 2.2395454545454547,
+      "grad_norm": 1.2621674952832622,
+      "learning_rate": 7.269124034549738e-07,
+      "loss": 0.0169,
+      "step": 9854
+    },
+    {
+      "epoch": 2.2397727272727272,
+      "grad_norm": 1.7372004616542096,
+      "learning_rate": 7.268243471280274e-07,
+      "loss": 0.0279,
+      "step": 9855
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.6969999967853433,
+      "learning_rate": 7.267362887247052e-07,
+      "loss": 0.0241,
+      "step": 9856
+    },
+    {
+      "epoch": 2.2402272727272727,
+      "grad_norm": 1.0637851627151313,
+      "learning_rate": 7.266482282468024e-07,
+      "loss": 0.0081,
+      "step": 9857
+    },
+    {
+      "epoch": 2.2404545454545453,
+      "grad_norm": 1.1840491469736976,
+      "learning_rate": 7.265601656961147e-07,
+      "loss": 0.0061,
+      "step": 9858
+    },
+    {
+      "epoch": 2.240681818181818,
+      "grad_norm": 2.226041223160367,
+      "learning_rate": 7.264721010744383e-07,
+      "loss": 0.034,
+      "step": 9859
+    },
+    {
+      "epoch": 2.2409090909090907,
+      "grad_norm": 1.9093816943437691,
+      "learning_rate": 7.263840343835688e-07,
+      "loss": 0.0217,
+      "step": 9860
+    },
+    {
+      "epoch": 2.2411363636363637,
+      "grad_norm": 2.1443978951271703,
+      "learning_rate": 7.262959656253019e-07,
+      "loss": 0.0302,
+      "step": 9861
+    },
+    {
+      "epoch": 2.2413636363636362,
+      "grad_norm": 0.7538398263246207,
+      "learning_rate": 7.262078948014334e-07,
+      "loss": 0.008,
+      "step": 9862
+    },
+    {
+      "epoch": 2.241590909090909,
+      "grad_norm": 1.9900981160773998,
+      "learning_rate": 7.261198219137596e-07,
+      "loss": 0.0284,
+      "step": 9863
+    },
+    {
+      "epoch": 2.2418181818181817,
+      "grad_norm": 2.9541129967832167,
+      "learning_rate": 7.260317469640759e-07,
+      "loss": 0.0392,
+      "step": 9864
+    },
+    {
+      "epoch": 2.2420454545454547,
+      "grad_norm": 1.9388793209601318,
+      "learning_rate": 7.259436699541787e-07,
+      "loss": 0.0174,
+      "step": 9865
+    },
+    {
+      "epoch": 2.242272727272727,
+      "grad_norm": 2.8193029374135956,
+      "learning_rate": 7.258555908858641e-07,
+      "loss": 0.0225,
+      "step": 9866
+    },
+    {
+      "epoch": 2.2425,
+      "grad_norm": 1.119523445194285,
+      "learning_rate": 7.25767509760928e-07,
+      "loss": 0.012,
+      "step": 9867
+    },
+    {
+      "epoch": 2.2427272727272727,
+      "grad_norm": 1.6583532773412286,
+      "learning_rate": 7.256794265811666e-07,
+      "loss": 0.0248,
+      "step": 9868
+    },
+    {
+      "epoch": 2.2429545454545456,
+      "grad_norm": 0.6539100610009385,
+      "learning_rate": 7.25591341348376e-07,
+      "loss": 0.0042,
+      "step": 9869
+    },
+    {
+      "epoch": 2.243181818181818,
+      "grad_norm": 2.193263257815197,
+      "learning_rate": 7.255032540643525e-07,
+      "loss": 0.0486,
+      "step": 9870
+    },
+    {
+      "epoch": 2.2434090909090907,
+      "grad_norm": 1.6413685553372912,
+      "learning_rate": 7.254151647308924e-07,
+      "loss": 0.0219,
+      "step": 9871
+    },
+    {
+      "epoch": 2.2436363636363637,
+      "grad_norm": 2.117804472521402,
+      "learning_rate": 7.253270733497918e-07,
+      "loss": 0.0233,
+      "step": 9872
+    },
+    {
+      "epoch": 2.243863636363636,
+      "grad_norm": 2.4472962988166316,
+      "learning_rate": 7.252389799228471e-07,
+      "loss": 0.0185,
+      "step": 9873
+    },
+    {
+      "epoch": 2.244090909090909,
+      "grad_norm": 2.5750670989977835,
+      "learning_rate": 7.251508844518547e-07,
+      "loss": 0.0335,
+      "step": 9874
+    },
+    {
+      "epoch": 2.2443181818181817,
+      "grad_norm": 1.3834709154404974,
+      "learning_rate": 7.25062786938611e-07,
+      "loss": 0.0109,
+      "step": 9875
+    },
+    {
+      "epoch": 2.2445454545454546,
+      "grad_norm": 2.6683669024349737,
+      "learning_rate": 7.249746873849127e-07,
+      "loss": 0.0246,
+      "step": 9876
+    },
+    {
+      "epoch": 2.244772727272727,
+      "grad_norm": 1.8981366255346284,
+      "learning_rate": 7.248865857925561e-07,
+      "loss": 0.0205,
+      "step": 9877
+    },
+    {
+      "epoch": 2.245,
+      "grad_norm": 1.6166685932815343,
+      "learning_rate": 7.247984821633376e-07,
+      "loss": 0.0422,
+      "step": 9878
+    },
+    {
+      "epoch": 2.2452272727272726,
+      "grad_norm": 1.4459876368751687,
+      "learning_rate": 7.24710376499054e-07,
+      "loss": 0.0143,
+      "step": 9879
+    },
+    {
+      "epoch": 2.2454545454545456,
+      "grad_norm": 1.9875892267287996,
+      "learning_rate": 7.246222688015019e-07,
+      "loss": 0.0278,
+      "step": 9880
+    },
+    {
+      "epoch": 2.245681818181818,
+      "grad_norm": 0.9977799091659192,
+      "learning_rate": 7.245341590724779e-07,
+      "loss": 0.0154,
+      "step": 9881
+    },
+    {
+      "epoch": 2.245909090909091,
+      "grad_norm": 1.1891807472278497,
+      "learning_rate": 7.244460473137788e-07,
+      "loss": 0.017,
+      "step": 9882
+    },
+    {
+      "epoch": 2.2461363636363636,
+      "grad_norm": 1.3631432121643192,
+      "learning_rate": 7.243579335272012e-07,
+      "loss": 0.0111,
+      "step": 9883
+    },
+    {
+      "epoch": 2.2463636363636366,
+      "grad_norm": 1.929428162823112,
+      "learning_rate": 7.242698177145422e-07,
+      "loss": 0.0435,
+      "step": 9884
+    },
+    {
+      "epoch": 2.246590909090909,
+      "grad_norm": 1.073060471151976,
+      "learning_rate": 7.241816998775983e-07,
+      "loss": 0.0123,
+      "step": 9885
+    },
+    {
+      "epoch": 2.246818181818182,
+      "grad_norm": 1.916087651620491,
+      "learning_rate": 7.240935800181664e-07,
+      "loss": 0.0261,
+      "step": 9886
+    },
+    {
+      "epoch": 2.2470454545454546,
+      "grad_norm": 2.175877892995899,
+      "learning_rate": 7.240054581380438e-07,
+      "loss": 0.0261,
+      "step": 9887
+    },
+    {
+      "epoch": 2.247272727272727,
+      "grad_norm": 1.3328947914668647,
+      "learning_rate": 7.239173342390269e-07,
+      "loss": 0.014,
+      "step": 9888
+    },
+    {
+      "epoch": 2.2475,
+      "grad_norm": 1.7258524057163966,
+      "learning_rate": 7.238292083229131e-07,
+      "loss": 0.0111,
+      "step": 9889
+    },
+    {
+      "epoch": 2.2477272727272726,
+      "grad_norm": 1.979991226419733,
+      "learning_rate": 7.237410803914994e-07,
+      "loss": 0.0193,
+      "step": 9890
+    },
+    {
+      "epoch": 2.2479545454545455,
+      "grad_norm": 1.7047459806159702,
+      "learning_rate": 7.236529504465828e-07,
+      "loss": 0.0339,
+      "step": 9891
+    },
+    {
+      "epoch": 2.248181818181818,
+      "grad_norm": 1.2888395772339,
+      "learning_rate": 7.235648184899602e-07,
+      "loss": 0.0174,
+      "step": 9892
+    },
+    {
+      "epoch": 2.248409090909091,
+      "grad_norm": 3.0568330551909475,
+      "learning_rate": 7.234766845234294e-07,
+      "loss": 0.0372,
+      "step": 9893
+    },
+    {
+      "epoch": 2.2486363636363635,
+      "grad_norm": 0.9323646288410712,
+      "learning_rate": 7.233885485487868e-07,
+      "loss": 0.012,
+      "step": 9894
+    },
+    {
+      "epoch": 2.2488636363636365,
+      "grad_norm": 1.3427700133052003,
+      "learning_rate": 7.233004105678302e-07,
+      "loss": 0.0147,
+      "step": 9895
+    },
+    {
+      "epoch": 2.249090909090909,
+      "grad_norm": 10.591429332056437,
+      "learning_rate": 7.232122705823567e-07,
+      "loss": 0.0137,
+      "step": 9896
+    },
+    {
+      "epoch": 2.249318181818182,
+      "grad_norm": 1.3404965337579684,
+      "learning_rate": 7.231241285941637e-07,
+      "loss": 0.0194,
+      "step": 9897
+    },
+    {
+      "epoch": 2.2495454545454545,
+      "grad_norm": 1.09160240268655,
+      "learning_rate": 7.230359846050486e-07,
+      "loss": 0.0128,
+      "step": 9898
+    },
+    {
+      "epoch": 2.2497727272727275,
+      "grad_norm": 2.142676882171102,
+      "learning_rate": 7.229478386168086e-07,
+      "loss": 0.0276,
+      "step": 9899
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.9782185760558654,
+      "learning_rate": 7.228596906312414e-07,
+      "loss": 0.0109,
+      "step": 9900
+    },
+    {
+      "epoch": 2.2502272727272725,
+      "grad_norm": 2.5342296105690045,
+      "learning_rate": 7.227715406501444e-07,
+      "loss": 0.0348,
+      "step": 9901
+    },
+    {
+      "epoch": 2.2504545454545455,
+      "grad_norm": 1.9823389341345214,
+      "learning_rate": 7.226833886753151e-07,
+      "loss": 0.0381,
+      "step": 9902
+    },
+    {
+      "epoch": 2.250681818181818,
+      "grad_norm": 3.457350759813856,
+      "learning_rate": 7.22595234708551e-07,
+      "loss": 0.0222,
+      "step": 9903
+    },
+    {
+      "epoch": 2.250909090909091,
+      "grad_norm": 1.9340461737113686,
+      "learning_rate": 7.225070787516499e-07,
+      "loss": 0.0251,
+      "step": 9904
+    },
+    {
+      "epoch": 2.2511363636363635,
+      "grad_norm": 2.0264597853561916,
+      "learning_rate": 7.224189208064092e-07,
+      "loss": 0.0119,
+      "step": 9905
+    },
+    {
+      "epoch": 2.2513636363636365,
+      "grad_norm": 1.5513742967572934,
+      "learning_rate": 7.223307608746269e-07,
+      "loss": 0.013,
+      "step": 9906
+    },
+    {
+      "epoch": 2.251590909090909,
+      "grad_norm": 3.373041389718195,
+      "learning_rate": 7.222425989581006e-07,
+      "loss": 0.0306,
+      "step": 9907
+    },
+    {
+      "epoch": 2.251818181818182,
+      "grad_norm": 2.4034194583978814,
+      "learning_rate": 7.22154435058628e-07,
+      "loss": 0.0384,
+      "step": 9908
+    },
+    {
+      "epoch": 2.2520454545454545,
+      "grad_norm": 1.1141498282429818,
+      "learning_rate": 7.22066269178007e-07,
+      "loss": 0.0109,
+      "step": 9909
+    },
+    {
+      "epoch": 2.2522727272727274,
+      "grad_norm": 3.5958671935051467,
+      "learning_rate": 7.219781013180354e-07,
+      "loss": 0.0355,
+      "step": 9910
+    },
+    {
+      "epoch": 2.2525,
+      "grad_norm": 2.5996902158367714,
+      "learning_rate": 7.21889931480511e-07,
+      "loss": 0.0315,
+      "step": 9911
+    },
+    {
+      "epoch": 2.252727272727273,
+      "grad_norm": 1.9812145040232219,
+      "learning_rate": 7.218017596672319e-07,
+      "loss": 0.0303,
+      "step": 9912
+    },
+    {
+      "epoch": 2.2529545454545454,
+      "grad_norm": 1.5218167410588646,
+      "learning_rate": 7.217135858799963e-07,
+      "loss": 0.0292,
+      "step": 9913
+    },
+    {
+      "epoch": 2.253181818181818,
+      "grad_norm": 1.90579509826111,
+      "learning_rate": 7.216254101206018e-07,
+      "loss": 0.0179,
+      "step": 9914
+    },
+    {
+      "epoch": 2.253409090909091,
+      "grad_norm": 16.088016158229888,
+      "learning_rate": 7.215372323908467e-07,
+      "loss": 0.0309,
+      "step": 9915
+    },
+    {
+      "epoch": 2.253636363636364,
+      "grad_norm": 3.4785678844916856,
+      "learning_rate": 7.214490526925289e-07,
+      "loss": 0.0527,
+      "step": 9916
+    },
+    {
+      "epoch": 2.2538636363636364,
+      "grad_norm": 1.5734694637973263,
+      "learning_rate": 7.213608710274465e-07,
+      "loss": 0.0169,
+      "step": 9917
+    },
+    {
+      "epoch": 2.254090909090909,
+      "grad_norm": 4.225498775470921,
+      "learning_rate": 7.212726873973981e-07,
+      "loss": 0.0306,
+      "step": 9918
+    },
+    {
+      "epoch": 2.254318181818182,
+      "grad_norm": 0.539831349522088,
+      "learning_rate": 7.211845018041814e-07,
+      "loss": 0.0036,
+      "step": 9919
+    },
+    {
+      "epoch": 2.2545454545454544,
+      "grad_norm": 2.078505358292161,
+      "learning_rate": 7.210963142495951e-07,
+      "loss": 0.0228,
+      "step": 9920
+    },
+    {
+      "epoch": 2.2547727272727274,
+      "grad_norm": 1.1728265568864487,
+      "learning_rate": 7.210081247354372e-07,
+      "loss": 0.0099,
+      "step": 9921
+    },
+    {
+      "epoch": 2.255,
+      "grad_norm": 1.5379610143521998,
+      "learning_rate": 7.209199332635063e-07,
+      "loss": 0.027,
+      "step": 9922
+    },
+    {
+      "epoch": 2.255227272727273,
+      "grad_norm": 2.4868322713082134,
+      "learning_rate": 7.208317398356004e-07,
+      "loss": 0.0188,
+      "step": 9923
+    },
+    {
+      "epoch": 2.2554545454545454,
+      "grad_norm": 1.349124138010001,
+      "learning_rate": 7.207435444535184e-07,
+      "loss": 0.0179,
+      "step": 9924
+    },
+    {
+      "epoch": 2.2556818181818183,
+      "grad_norm": 1.8362314750503035,
+      "learning_rate": 7.206553471190584e-07,
+      "loss": 0.0136,
+      "step": 9925
+    },
+    {
+      "epoch": 2.255909090909091,
+      "grad_norm": 1.2397128719441204,
+      "learning_rate": 7.205671478340189e-07,
+      "loss": 0.0209,
+      "step": 9926
+    },
+    {
+      "epoch": 2.256136363636364,
+      "grad_norm": 1.8087140036834444,
+      "learning_rate": 7.204789466001983e-07,
+      "loss": 0.0156,
+      "step": 9927
+    },
+    {
+      "epoch": 2.2563636363636363,
+      "grad_norm": 1.7049767256804929,
+      "learning_rate": 7.203907434193958e-07,
+      "loss": 0.0364,
+      "step": 9928
+    },
+    {
+      "epoch": 2.2565909090909093,
+      "grad_norm": 1.589100216556357,
+      "learning_rate": 7.203025382934095e-07,
+      "loss": 0.0242,
+      "step": 9929
+    },
+    {
+      "epoch": 2.256818181818182,
+      "grad_norm": 2.1338807195862324,
+      "learning_rate": 7.202143312240381e-07,
+      "loss": 0.0354,
+      "step": 9930
+    },
+    {
+      "epoch": 2.2570454545454544,
+      "grad_norm": 1.9618639144184744,
+      "learning_rate": 7.201261222130807e-07,
+      "loss": 0.0275,
+      "step": 9931
+    },
+    {
+      "epoch": 2.2572727272727273,
+      "grad_norm": 0.6562884525290988,
+      "learning_rate": 7.200379112623354e-07,
+      "loss": 0.0051,
+      "step": 9932
+    },
+    {
+      "epoch": 2.2575,
+      "grad_norm": 0.9975111937690145,
+      "learning_rate": 7.199496983736014e-07,
+      "loss": 0.0052,
+      "step": 9933
+    },
+    {
+      "epoch": 2.257727272727273,
+      "grad_norm": 4.3841924391484035,
+      "learning_rate": 7.198614835486773e-07,
+      "loss": 0.0543,
+      "step": 9934
+    },
+    {
+      "epoch": 2.2579545454545453,
+      "grad_norm": 1.2683753951504162,
+      "learning_rate": 7.19773266789362e-07,
+      "loss": 0.0171,
+      "step": 9935
+    },
+    {
+      "epoch": 2.2581818181818183,
+      "grad_norm": 0.779698331779589,
+      "learning_rate": 7.196850480974546e-07,
+      "loss": 0.0058,
+      "step": 9936
+    },
+    {
+      "epoch": 2.258409090909091,
+      "grad_norm": 0.7650262498403476,
+      "learning_rate": 7.195968274747539e-07,
+      "loss": 0.0076,
+      "step": 9937
+    },
+    {
+      "epoch": 2.2586363636363638,
+      "grad_norm": 1.9237896909506718,
+      "learning_rate": 7.195086049230587e-07,
+      "loss": 0.0298,
+      "step": 9938
+    },
+    {
+      "epoch": 2.2588636363636363,
+      "grad_norm": 1.7536417432467424,
+      "learning_rate": 7.194203804441684e-07,
+      "loss": 0.0146,
+      "step": 9939
+    },
+    {
+      "epoch": 2.2590909090909093,
+      "grad_norm": 1.5774473801974949,
+      "learning_rate": 7.193321540398816e-07,
+      "loss": 0.029,
+      "step": 9940
+    },
+    {
+      "epoch": 2.259318181818182,
+      "grad_norm": 1.8886490015747113,
+      "learning_rate": 7.192439257119976e-07,
+      "loss": 0.0226,
+      "step": 9941
+    },
+    {
+      "epoch": 2.2595454545454547,
+      "grad_norm": 2.768270664876381,
+      "learning_rate": 7.191556954623156e-07,
+      "loss": 0.0407,
+      "step": 9942
+    },
+    {
+      "epoch": 2.2597727272727273,
+      "grad_norm": 1.5673572386228656,
+      "learning_rate": 7.190674632926348e-07,
+      "loss": 0.0109,
+      "step": 9943
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.4693866770913082,
+      "learning_rate": 7.189792292047543e-07,
+      "loss": 0.0203,
+      "step": 9944
+    },
+    {
+      "epoch": 2.2602272727272728,
+      "grad_norm": 2.3863158874318344,
+      "learning_rate": 7.188909932004733e-07,
+      "loss": 0.0321,
+      "step": 9945
+    },
+    {
+      "epoch": 2.2604545454545453,
+      "grad_norm": 1.9944356740862335,
+      "learning_rate": 7.188027552815912e-07,
+      "loss": 0.0303,
+      "step": 9946
+    },
+    {
+      "epoch": 2.2606818181818182,
+      "grad_norm": 4.299351502145729,
+      "learning_rate": 7.187145154499072e-07,
+      "loss": 0.0338,
+      "step": 9947
+    },
+    {
+      "epoch": 2.2609090909090908,
+      "grad_norm": 2.478496052265079,
+      "learning_rate": 7.186262737072208e-07,
+      "loss": 0.0571,
+      "step": 9948
+    },
+    {
+      "epoch": 2.2611363636363637,
+      "grad_norm": 2.079434356720784,
+      "learning_rate": 7.185380300553311e-07,
+      "loss": 0.03,
+      "step": 9949
+    },
+    {
+      "epoch": 2.2613636363636362,
+      "grad_norm": 2.1866254730678234,
+      "learning_rate": 7.184497844960381e-07,
+      "loss": 0.0216,
+      "step": 9950
+    },
+    {
+      "epoch": 2.261590909090909,
+      "grad_norm": 1.5253273536793455,
+      "learning_rate": 7.183615370311407e-07,
+      "loss": 0.0099,
+      "step": 9951
+    },
+    {
+      "epoch": 2.2618181818181817,
+      "grad_norm": 2.0197155171731356,
+      "learning_rate": 7.182732876624389e-07,
+      "loss": 0.0381,
+      "step": 9952
+    },
+    {
+      "epoch": 2.2620454545454547,
+      "grad_norm": 1.258565951574646,
+      "learning_rate": 7.18185036391732e-07,
+      "loss": 0.0224,
+      "step": 9953
+    },
+    {
+      "epoch": 2.262272727272727,
+      "grad_norm": 2.5444189034902207,
+      "learning_rate": 7.180967832208195e-07,
+      "loss": 0.0339,
+      "step": 9954
+    },
+    {
+      "epoch": 2.2625,
+      "grad_norm": 0.7788399062885538,
+      "learning_rate": 7.180085281515014e-07,
+      "loss": 0.0089,
+      "step": 9955
+    },
+    {
+      "epoch": 2.2627272727272727,
+      "grad_norm": 2.44510545017995,
+      "learning_rate": 7.179202711855771e-07,
+      "loss": 0.0525,
+      "step": 9956
+    },
+    {
+      "epoch": 2.262954545454545,
+      "grad_norm": 2.0637066828783874,
+      "learning_rate": 7.178320123248461e-07,
+      "loss": 0.0169,
+      "step": 9957
+    },
+    {
+      "epoch": 2.263181818181818,
+      "grad_norm": 1.3482706622750404,
+      "learning_rate": 7.177437515711086e-07,
+      "loss": 0.0106,
+      "step": 9958
+    },
+    {
+      "epoch": 2.263409090909091,
+      "grad_norm": 2.8516039586794433,
+      "learning_rate": 7.176554889261643e-07,
+      "loss": 0.0336,
+      "step": 9959
+    },
+    {
+      "epoch": 2.2636363636363637,
+      "grad_norm": 1.9542213707190281,
+      "learning_rate": 7.175672243918128e-07,
+      "loss": 0.0076,
+      "step": 9960
+    },
+    {
+      "epoch": 2.263863636363636,
+      "grad_norm": 2.274131735675503,
+      "learning_rate": 7.174789579698542e-07,
+      "loss": 0.0228,
+      "step": 9961
+    },
+    {
+      "epoch": 2.264090909090909,
+      "grad_norm": 3.401355824984848,
+      "learning_rate": 7.173906896620882e-07,
+      "loss": 0.0307,
+      "step": 9962
+    },
+    {
+      "epoch": 2.2643181818181817,
+      "grad_norm": 1.3038221024034835,
+      "learning_rate": 7.173024194703149e-07,
+      "loss": 0.0146,
+      "step": 9963
+    },
+    {
+      "epoch": 2.2645454545454546,
+      "grad_norm": 1.5930210156164648,
+      "learning_rate": 7.172141473963341e-07,
+      "loss": 0.0377,
+      "step": 9964
+    },
+    {
+      "epoch": 2.264772727272727,
+      "grad_norm": 0.8826316139602581,
+      "learning_rate": 7.171258734419461e-07,
+      "loss": 0.0096,
+      "step": 9965
+    },
+    {
+      "epoch": 2.265,
+      "grad_norm": 2.4114676556672814,
+      "learning_rate": 7.170375976089509e-07,
+      "loss": 0.0314,
+      "step": 9966
+    },
+    {
+      "epoch": 2.2652272727272726,
+      "grad_norm": 2.044000908548546,
+      "learning_rate": 7.169493198991484e-07,
+      "loss": 0.0143,
+      "step": 9967
+    },
+    {
+      "epoch": 2.2654545454545456,
+      "grad_norm": 1.291452395649828,
+      "learning_rate": 7.16861040314339e-07,
+      "loss": 0.0095,
+      "step": 9968
+    },
+    {
+      "epoch": 2.265681818181818,
+      "grad_norm": 3.3980128509688226,
+      "learning_rate": 7.167727588563225e-07,
+      "loss": 0.0502,
+      "step": 9969
+    },
+    {
+      "epoch": 2.265909090909091,
+      "grad_norm": 2.7308103509756387,
+      "learning_rate": 7.166844755268997e-07,
+      "loss": 0.0188,
+      "step": 9970
+    },
+    {
+      "epoch": 2.2661363636363636,
+      "grad_norm": 2.694253531370601,
+      "learning_rate": 7.1659619032787e-07,
+      "loss": 0.0411,
+      "step": 9971
+    },
+    {
+      "epoch": 2.2663636363636366,
+      "grad_norm": 1.0641414823314366,
+      "learning_rate": 7.165079032610345e-07,
+      "loss": 0.0103,
+      "step": 9972
+    },
+    {
+      "epoch": 2.266590909090909,
+      "grad_norm": 2.069582763695594,
+      "learning_rate": 7.16419614328193e-07,
+      "loss": 0.0243,
+      "step": 9973
+    },
+    {
+      "epoch": 2.2668181818181816,
+      "grad_norm": 5.02675745226715,
+      "learning_rate": 7.163313235311461e-07,
+      "loss": 0.0603,
+      "step": 9974
+    },
+    {
+      "epoch": 2.2670454545454546,
+      "grad_norm": 1.286036383682505,
+      "learning_rate": 7.162430308716945e-07,
+      "loss": 0.0199,
+      "step": 9975
+    },
+    {
+      "epoch": 2.267272727272727,
+      "grad_norm": 2.5843055755204047,
+      "learning_rate": 7.16154736351638e-07,
+      "loss": 0.074,
+      "step": 9976
+    },
+    {
+      "epoch": 2.2675,
+      "grad_norm": 2.514073160837038,
+      "learning_rate": 7.160664399727777e-07,
+      "loss": 0.0327,
+      "step": 9977
+    },
+    {
+      "epoch": 2.2677272727272726,
+      "grad_norm": 1.5233460589210663,
+      "learning_rate": 7.159781417369136e-07,
+      "loss": 0.0194,
+      "step": 9978
+    },
+    {
+      "epoch": 2.2679545454545456,
+      "grad_norm": 1.6161468781617936,
+      "learning_rate": 7.158898416458464e-07,
+      "loss": 0.026,
+      "step": 9979
+    },
+    {
+      "epoch": 2.268181818181818,
+      "grad_norm": 2.0570024778819733,
+      "learning_rate": 7.158015397013768e-07,
+      "loss": 0.0334,
+      "step": 9980
+    },
+    {
+      "epoch": 2.268409090909091,
+      "grad_norm": 1.7038131829132743,
+      "learning_rate": 7.157132359053056e-07,
+      "loss": 0.0373,
+      "step": 9981
+    },
+    {
+      "epoch": 2.2686363636363636,
+      "grad_norm": 3.2469313593780287,
+      "learning_rate": 7.15624930259433e-07,
+      "loss": 0.0714,
+      "step": 9982
+    },
+    {
+      "epoch": 2.2688636363636365,
+      "grad_norm": 2.7762331602083044,
+      "learning_rate": 7.155366227655601e-07,
+      "loss": 0.0248,
+      "step": 9983
+    },
+    {
+      "epoch": 2.269090909090909,
+      "grad_norm": 2.9154833837749807,
+      "learning_rate": 7.154483134254875e-07,
+      "loss": 0.0289,
+      "step": 9984
+    },
+    {
+      "epoch": 2.269318181818182,
+      "grad_norm": 2.5443514450852733,
+      "learning_rate": 7.15360002241016e-07,
+      "loss": 0.033,
+      "step": 9985
+    },
+    {
+      "epoch": 2.2695454545454545,
+      "grad_norm": 2.889105651454126,
+      "learning_rate": 7.152716892139464e-07,
+      "loss": 0.0247,
+      "step": 9986
+    },
+    {
+      "epoch": 2.269772727272727,
+      "grad_norm": 1.4917476023478127,
+      "learning_rate": 7.151833743460795e-07,
+      "loss": 0.0144,
+      "step": 9987
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 2.0287685105792557,
+      "learning_rate": 7.150950576392163e-07,
+      "loss": 0.0313,
+      "step": 9988
+    },
+    {
+      "epoch": 2.2702272727272725,
+      "grad_norm": 0.9196939373749398,
+      "learning_rate": 7.150067390951576e-07,
+      "loss": 0.0114,
+      "step": 9989
+    },
+    {
+      "epoch": 2.2704545454545455,
+      "grad_norm": 2.9505925884418387,
+      "learning_rate": 7.149184187157048e-07,
+      "loss": 0.0365,
+      "step": 9990
+    },
+    {
+      "epoch": 2.270681818181818,
+      "grad_norm": 2.091141109481235,
+      "learning_rate": 7.148300965026582e-07,
+      "loss": 0.0169,
+      "step": 9991
+    },
+    {
+      "epoch": 2.270909090909091,
+      "grad_norm": 1.9847583220758487,
+      "learning_rate": 7.147417724578194e-07,
+      "loss": 0.0172,
+      "step": 9992
+    },
+    {
+      "epoch": 2.2711363636363635,
+      "grad_norm": 3.1515231294440667,
+      "learning_rate": 7.146534465829894e-07,
+      "loss": 0.0314,
+      "step": 9993
+    },
+    {
+      "epoch": 2.2713636363636365,
+      "grad_norm": 1.609859524709798,
+      "learning_rate": 7.145651188799689e-07,
+      "loss": 0.0256,
+      "step": 9994
+    },
+    {
+      "epoch": 2.271590909090909,
+      "grad_norm": 1.8933440672450472,
+      "learning_rate": 7.144767893505596e-07,
+      "loss": 0.0223,
+      "step": 9995
+    },
+    {
+      "epoch": 2.271818181818182,
+      "grad_norm": 4.2937476242545385,
+      "learning_rate": 7.143884579965624e-07,
+      "loss": 0.0497,
+      "step": 9996
+    },
+    {
+      "epoch": 2.2720454545454545,
+      "grad_norm": 2.5934726540256396,
+      "learning_rate": 7.143001248197787e-07,
+      "loss": 0.0283,
+      "step": 9997
+    },
+    {
+      "epoch": 2.2722727272727274,
+      "grad_norm": 2.0861708432187442,
+      "learning_rate": 7.142117898220094e-07,
+      "loss": 0.0223,
+      "step": 9998
+    },
+    {
+      "epoch": 2.2725,
+      "grad_norm": 1.9856143371838364,
+      "learning_rate": 7.141234530050564e-07,
+      "loss": 0.0284,
+      "step": 9999
+    },
+    {
+      "epoch": 2.2727272727272725,
+      "grad_norm": 1.0690837197511363,
+      "learning_rate": 7.140351143707206e-07,
+      "loss": 0.0185,
+      "step": 10000
+    },
+    {
+      "epoch": 2.2729545454545454,
+      "grad_norm": 0.7596962477187721,
+      "learning_rate": 7.139467739208034e-07,
+      "loss": 0.009,
+      "step": 10001
+    },
+    {
+      "epoch": 2.2731818181818184,
+      "grad_norm": 2.155992185596243,
+      "learning_rate": 7.138584316571062e-07,
+      "loss": 0.0412,
+      "step": 10002
+    },
+    {
+      "epoch": 2.273409090909091,
+      "grad_norm": 0.8168793441625807,
+      "learning_rate": 7.137700875814307e-07,
+      "loss": 0.0093,
+      "step": 10003
+    },
+    {
+      "epoch": 2.2736363636363635,
+      "grad_norm": 1.0610088627506213,
+      "learning_rate": 7.136817416955783e-07,
+      "loss": 0.0171,
+      "step": 10004
+    },
+    {
+      "epoch": 2.2738636363636364,
+      "grad_norm": 1.4841730198626406,
+      "learning_rate": 7.135933940013505e-07,
+      "loss": 0.0119,
+      "step": 10005
+    },
+    {
+      "epoch": 2.274090909090909,
+      "grad_norm": 1.3920216795820273,
+      "learning_rate": 7.135050445005487e-07,
+      "loss": 0.0202,
+      "step": 10006
+    },
+    {
+      "epoch": 2.274318181818182,
+      "grad_norm": 1.473317083979157,
+      "learning_rate": 7.134166931949746e-07,
+      "loss": 0.011,
+      "step": 10007
+    },
+    {
+      "epoch": 2.2745454545454544,
+      "grad_norm": 1.0775453663342525,
+      "learning_rate": 7.133283400864301e-07,
+      "loss": 0.01,
+      "step": 10008
+    },
+    {
+      "epoch": 2.2747727272727274,
+      "grad_norm": 2.645744716882427,
+      "learning_rate": 7.132399851767165e-07,
+      "loss": 0.0295,
+      "step": 10009
+    },
+    {
+      "epoch": 2.275,
+      "grad_norm": 2.807733915279644,
+      "learning_rate": 7.131516284676355e-07,
+      "loss": 0.0573,
+      "step": 10010
+    },
+    {
+      "epoch": 2.275227272727273,
+      "grad_norm": 2.4066240845341467,
+      "learning_rate": 7.130632699609892e-07,
+      "loss": 0.0202,
+      "step": 10011
+    },
+    {
+      "epoch": 2.2754545454545454,
+      "grad_norm": 3.7752472759109286,
+      "learning_rate": 7.129749096585792e-07,
+      "loss": 0.0435,
+      "step": 10012
+    },
+    {
+      "epoch": 2.2756818181818184,
+      "grad_norm": 1.6626831167000062,
+      "learning_rate": 7.128865475622072e-07,
+      "loss": 0.0146,
+      "step": 10013
+    },
+    {
+      "epoch": 2.275909090909091,
+      "grad_norm": 1.4659637683822897,
+      "learning_rate": 7.127981836736754e-07,
+      "loss": 0.0257,
+      "step": 10014
+    },
+    {
+      "epoch": 2.276136363636364,
+      "grad_norm": 2.8110676611650005,
+      "learning_rate": 7.127098179947853e-07,
+      "loss": 0.0282,
+      "step": 10015
+    },
+    {
+      "epoch": 2.2763636363636364,
+      "grad_norm": 2.520589530850877,
+      "learning_rate": 7.12621450527339e-07,
+      "loss": 0.0231,
+      "step": 10016
+    },
+    {
+      "epoch": 2.276590909090909,
+      "grad_norm": 1.1265981085246901,
+      "learning_rate": 7.125330812731385e-07,
+      "loss": 0.0163,
+      "step": 10017
+    },
+    {
+      "epoch": 2.276818181818182,
+      "grad_norm": 2.432252070809586,
+      "learning_rate": 7.124447102339856e-07,
+      "loss": 0.0595,
+      "step": 10018
+    },
+    {
+      "epoch": 2.2770454545454544,
+      "grad_norm": 1.2945164346315556,
+      "learning_rate": 7.123563374116826e-07,
+      "loss": 0.0166,
+      "step": 10019
+    },
+    {
+      "epoch": 2.2772727272727273,
+      "grad_norm": 2.0093743318440676,
+      "learning_rate": 7.122679628080315e-07,
+      "loss": 0.0211,
+      "step": 10020
+    },
+    {
+      "epoch": 2.2775,
+      "grad_norm": 1.160461001553267,
+      "learning_rate": 7.121795864248345e-07,
+      "loss": 0.0099,
+      "step": 10021
+    },
+    {
+      "epoch": 2.277727272727273,
+      "grad_norm": 1.6926670292027455,
+      "learning_rate": 7.120912082638935e-07,
+      "loss": 0.0143,
+      "step": 10022
+    },
+    {
+      "epoch": 2.2779545454545453,
+      "grad_norm": 1.5016110038871062,
+      "learning_rate": 7.12002828327011e-07,
+      "loss": 0.0201,
+      "step": 10023
+    },
+    {
+      "epoch": 2.2781818181818183,
+      "grad_norm": 1.5911760578655199,
+      "learning_rate": 7.119144466159888e-07,
+      "loss": 0.0153,
+      "step": 10024
+    },
+    {
+      "epoch": 2.278409090909091,
+      "grad_norm": 0.7105578232596462,
+      "learning_rate": 7.118260631326295e-07,
+      "loss": 0.0066,
+      "step": 10025
+    },
+    {
+      "epoch": 2.278636363636364,
+      "grad_norm": 1.4136913096281556,
+      "learning_rate": 7.117376778787353e-07,
+      "loss": 0.0165,
+      "step": 10026
+    },
+    {
+      "epoch": 2.2788636363636363,
+      "grad_norm": 2.294374575167304,
+      "learning_rate": 7.116492908561085e-07,
+      "loss": 0.0481,
+      "step": 10027
+    },
+    {
+      "epoch": 2.2790909090909093,
+      "grad_norm": 1.9339001438033772,
+      "learning_rate": 7.115609020665515e-07,
+      "loss": 0.0252,
+      "step": 10028
+    },
+    {
+      "epoch": 2.279318181818182,
+      "grad_norm": 1.894342796649742,
+      "learning_rate": 7.114725115118668e-07,
+      "loss": 0.0125,
+      "step": 10029
+    },
+    {
+      "epoch": 2.2795454545454543,
+      "grad_norm": 2.2353334292278686,
+      "learning_rate": 7.113841191938567e-07,
+      "loss": 0.0513,
+      "step": 10030
+    },
+    {
+      "epoch": 2.2797727272727273,
+      "grad_norm": 2.131653594103603,
+      "learning_rate": 7.112957251143236e-07,
+      "loss": 0.0191,
+      "step": 10031
+    },
+    {
+      "epoch": 2.2800000000000002,
+      "grad_norm": 1.6348707764301413,
+      "learning_rate": 7.112073292750703e-07,
+      "loss": 0.0168,
+      "step": 10032
+    },
+    {
+      "epoch": 2.2802272727272728,
+      "grad_norm": 1.3227059292170222,
+      "learning_rate": 7.111189316778988e-07,
+      "loss": 0.0206,
+      "step": 10033
+    },
+    {
+      "epoch": 2.2804545454545453,
+      "grad_norm": 1.4901803197750771,
+      "learning_rate": 7.110305323246124e-07,
+      "loss": 0.022,
+      "step": 10034
+    },
+    {
+      "epoch": 2.2806818181818183,
+      "grad_norm": 0.9622324505517489,
+      "learning_rate": 7.109421312170131e-07,
+      "loss": 0.0122,
+      "step": 10035
+    },
+    {
+      "epoch": 2.2809090909090908,
+      "grad_norm": 1.3305212833491407,
+      "learning_rate": 7.10853728356904e-07,
+      "loss": 0.0097,
+      "step": 10036
+    },
+    {
+      "epoch": 2.2811363636363637,
+      "grad_norm": 1.8229766681675383,
+      "learning_rate": 7.107653237460875e-07,
+      "loss": 0.0167,
+      "step": 10037
+    },
+    {
+      "epoch": 2.2813636363636363,
+      "grad_norm": 3.090203731748964,
+      "learning_rate": 7.106769173863666e-07,
+      "loss": 0.0313,
+      "step": 10038
+    },
+    {
+      "epoch": 2.2815909090909092,
+      "grad_norm": 2.4532234027431117,
+      "learning_rate": 7.105885092795437e-07,
+      "loss": 0.0265,
+      "step": 10039
+    },
+    {
+      "epoch": 2.2818181818181817,
+      "grad_norm": 1.2143025189128247,
+      "learning_rate": 7.105000994274218e-07,
+      "loss": 0.0103,
+      "step": 10040
+    },
+    {
+      "epoch": 2.2820454545454547,
+      "grad_norm": 2.3545345705176133,
+      "learning_rate": 7.104116878318037e-07,
+      "loss": 0.0176,
+      "step": 10041
+    },
+    {
+      "epoch": 2.2822727272727272,
+      "grad_norm": 1.5029634402351264,
+      "learning_rate": 7.103232744944922e-07,
+      "loss": 0.0227,
+      "step": 10042
+    },
+    {
+      "epoch": 2.2824999999999998,
+      "grad_norm": 2.0690481115948063,
+      "learning_rate": 7.102348594172905e-07,
+      "loss": 0.0272,
+      "step": 10043
+    },
+    {
+      "epoch": 2.2827272727272727,
+      "grad_norm": 1.5656567712482117,
+      "learning_rate": 7.10146442602001e-07,
+      "loss": 0.0078,
+      "step": 10044
+    },
+    {
+      "epoch": 2.2829545454545457,
+      "grad_norm": 1.6968034930879088,
+      "learning_rate": 7.100580240504273e-07,
+      "loss": 0.0063,
+      "step": 10045
+    },
+    {
+      "epoch": 2.283181818181818,
+      "grad_norm": 3.164695372339886,
+      "learning_rate": 7.09969603764372e-07,
+      "loss": 0.0714,
+      "step": 10046
+    },
+    {
+      "epoch": 2.2834090909090907,
+      "grad_norm": 1.4888262928664617,
+      "learning_rate": 7.098811817456383e-07,
+      "loss": 0.0212,
+      "step": 10047
+    },
+    {
+      "epoch": 2.2836363636363637,
+      "grad_norm": 2.2490362766994596,
+      "learning_rate": 7.09792757996029e-07,
+      "loss": 0.0636,
+      "step": 10048
+    },
+    {
+      "epoch": 2.283863636363636,
+      "grad_norm": 1.3052985795229617,
+      "learning_rate": 7.097043325173478e-07,
+      "loss": 0.0228,
+      "step": 10049
+    },
+    {
+      "epoch": 2.284090909090909,
+      "grad_norm": 3.1465323114512636,
+      "learning_rate": 7.096159053113971e-07,
+      "loss": 0.035,
+      "step": 10050
+    },
+    {
+      "epoch": 2.2843181818181817,
+      "grad_norm": 1.1840458012849404,
+      "learning_rate": 7.095274763799805e-07,
+      "loss": 0.0113,
+      "step": 10051
+    },
+    {
+      "epoch": 2.2845454545454547,
+      "grad_norm": 2.300001262734724,
+      "learning_rate": 7.094390457249015e-07,
+      "loss": 0.0291,
+      "step": 10052
+    },
+    {
+      "epoch": 2.284772727272727,
+      "grad_norm": 1.7948182111763815,
+      "learning_rate": 7.093506133479629e-07,
+      "loss": 0.0313,
+      "step": 10053
+    },
+    {
+      "epoch": 2.285,
+      "grad_norm": 2.3286706353461386,
+      "learning_rate": 7.092621792509682e-07,
+      "loss": 0.0378,
+      "step": 10054
+    },
+    {
+      "epoch": 2.2852272727272727,
+      "grad_norm": 1.3623571501870424,
+      "learning_rate": 7.091737434357204e-07,
+      "loss": 0.018,
+      "step": 10055
+    },
+    {
+      "epoch": 2.2854545454545456,
+      "grad_norm": 1.034305860706768,
+      "learning_rate": 7.090853059040234e-07,
+      "loss": 0.0077,
+      "step": 10056
+    },
+    {
+      "epoch": 2.285681818181818,
+      "grad_norm": 3.0806130105322613,
+      "learning_rate": 7.089968666576801e-07,
+      "loss": 0.0497,
+      "step": 10057
+    },
+    {
+      "epoch": 2.285909090909091,
+      "grad_norm": 3.1145298843026983,
+      "learning_rate": 7.089084256984944e-07,
+      "loss": 0.0302,
+      "step": 10058
+    },
+    {
+      "epoch": 2.2861363636363636,
+      "grad_norm": 4.270853185414852,
+      "learning_rate": 7.088199830282693e-07,
+      "loss": 0.0669,
+      "step": 10059
+    },
+    {
+      "epoch": 2.286363636363636,
+      "grad_norm": 1.3850067368189434,
+      "learning_rate": 7.087315386488086e-07,
+      "loss": 0.0223,
+      "step": 10060
+    },
+    {
+      "epoch": 2.286590909090909,
+      "grad_norm": 2.2506722812245528,
+      "learning_rate": 7.086430925619157e-07,
+      "loss": 0.0232,
+      "step": 10061
+    },
+    {
+      "epoch": 2.2868181818181816,
+      "grad_norm": 2.1904018972283668,
+      "learning_rate": 7.085546447693943e-07,
+      "loss": 0.0344,
+      "step": 10062
+    },
+    {
+      "epoch": 2.2870454545454546,
+      "grad_norm": 2.255564257997422,
+      "learning_rate": 7.084661952730478e-07,
+      "loss": 0.0291,
+      "step": 10063
+    },
+    {
+      "epoch": 2.287272727272727,
+      "grad_norm": 0.8800869963568904,
+      "learning_rate": 7.0837774407468e-07,
+      "loss": 0.0127,
+      "step": 10064
+    },
+    {
+      "epoch": 2.2875,
+      "grad_norm": 1.0793822592690296,
+      "learning_rate": 7.082892911760947e-07,
+      "loss": 0.0109,
+      "step": 10065
+    },
+    {
+      "epoch": 2.2877272727272726,
+      "grad_norm": 2.236046506302884,
+      "learning_rate": 7.082008365790953e-07,
+      "loss": 0.0268,
+      "step": 10066
+    },
+    {
+      "epoch": 2.2879545454545456,
+      "grad_norm": 2.3475230140071717,
+      "learning_rate": 7.081123802854858e-07,
+      "loss": 0.0233,
+      "step": 10067
+    },
+    {
+      "epoch": 2.288181818181818,
+      "grad_norm": 2.0117581390774837,
+      "learning_rate": 7.080239222970697e-07,
+      "loss": 0.0184,
+      "step": 10068
+    },
+    {
+      "epoch": 2.288409090909091,
+      "grad_norm": 1.271110995658505,
+      "learning_rate": 7.079354626156512e-07,
+      "loss": 0.0134,
+      "step": 10069
+    },
+    {
+      "epoch": 2.2886363636363636,
+      "grad_norm": 1.371108489930817,
+      "learning_rate": 7.078470012430337e-07,
+      "loss": 0.0212,
+      "step": 10070
+    },
+    {
+      "epoch": 2.2888636363636365,
+      "grad_norm": 3.9864824015089333,
+      "learning_rate": 7.077585381810213e-07,
+      "loss": 0.0242,
+      "step": 10071
+    },
+    {
+      "epoch": 2.289090909090909,
+      "grad_norm": 4.487508910751766,
+      "learning_rate": 7.07670073431418e-07,
+      "loss": 0.0605,
+      "step": 10072
+    },
+    {
+      "epoch": 2.2893181818181816,
+      "grad_norm": 1.1652916875951287,
+      "learning_rate": 7.075816069960278e-07,
+      "loss": 0.013,
+      "step": 10073
+    },
+    {
+      "epoch": 2.2895454545454546,
+      "grad_norm": 2.624773522038925,
+      "learning_rate": 7.074931388766545e-07,
+      "loss": 0.0166,
+      "step": 10074
+    },
+    {
+      "epoch": 2.2897727272727275,
+      "grad_norm": 1.2534809150591524,
+      "learning_rate": 7.074046690751022e-07,
+      "loss": 0.0078,
+      "step": 10075
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.4959237699923806,
+      "learning_rate": 7.073161975931748e-07,
+      "loss": 0.029,
+      "step": 10076
+    },
+    {
+      "epoch": 2.2902272727272726,
+      "grad_norm": 1.1446692466930999,
+      "learning_rate": 7.072277244326767e-07,
+      "loss": 0.0148,
+      "step": 10077
+    },
+    {
+      "epoch": 2.2904545454545455,
+      "grad_norm": 1.971218624184188,
+      "learning_rate": 7.071392495954117e-07,
+      "loss": 0.0207,
+      "step": 10078
+    },
+    {
+      "epoch": 2.290681818181818,
+      "grad_norm": 1.728965900802292,
+      "learning_rate": 7.070507730831841e-07,
+      "loss": 0.0384,
+      "step": 10079
+    },
+    {
+      "epoch": 2.290909090909091,
+      "grad_norm": 1.9933483335929756,
+      "learning_rate": 7.069622948977982e-07,
+      "loss": 0.0227,
+      "step": 10080
+    },
+    {
+      "epoch": 2.2911363636363635,
+      "grad_norm": 2.234356535273388,
+      "learning_rate": 7.068738150410582e-07,
+      "loss": 0.0123,
+      "step": 10081
+    },
+    {
+      "epoch": 2.2913636363636365,
+      "grad_norm": 2.004626881673958,
+      "learning_rate": 7.067853335147682e-07,
+      "loss": 0.0175,
+      "step": 10082
+    },
+    {
+      "epoch": 2.291590909090909,
+      "grad_norm": 1.375684266307635,
+      "learning_rate": 7.066968503207324e-07,
+      "loss": 0.0072,
+      "step": 10083
+    },
+    {
+      "epoch": 2.291818181818182,
+      "grad_norm": 4.360312110203321,
+      "learning_rate": 7.066083654607554e-07,
+      "loss": 0.0517,
+      "step": 10084
+    },
+    {
+      "epoch": 2.2920454545454545,
+      "grad_norm": 4.063230643185981,
+      "learning_rate": 7.065198789366414e-07,
+      "loss": 0.0285,
+      "step": 10085
+    },
+    {
+      "epoch": 2.2922727272727275,
+      "grad_norm": 2.110615042157599,
+      "learning_rate": 7.064313907501949e-07,
+      "loss": 0.0244,
+      "step": 10086
+    },
+    {
+      "epoch": 2.2925,
+      "grad_norm": 2.393061310417336,
+      "learning_rate": 7.063429009032202e-07,
+      "loss": 0.0326,
+      "step": 10087
+    },
+    {
+      "epoch": 2.292727272727273,
+      "grad_norm": 0.8366651133041886,
+      "learning_rate": 7.06254409397522e-07,
+      "loss": 0.0169,
+      "step": 10088
+    },
+    {
+      "epoch": 2.2929545454545455,
+      "grad_norm": 0.5066293472532815,
+      "learning_rate": 7.061659162349045e-07,
+      "loss": 0.0068,
+      "step": 10089
+    },
+    {
+      "epoch": 2.293181818181818,
+      "grad_norm": 3.141777074633708,
+      "learning_rate": 7.060774214171725e-07,
+      "loss": 0.0342,
+      "step": 10090
+    },
+    {
+      "epoch": 2.293409090909091,
+      "grad_norm": 2.1118661492486903,
+      "learning_rate": 7.059889249461303e-07,
+      "loss": 0.0131,
+      "step": 10091
+    },
+    {
+      "epoch": 2.2936363636363635,
+      "grad_norm": 2.594436267524619,
+      "learning_rate": 7.059004268235828e-07,
+      "loss": 0.0306,
+      "step": 10092
+    },
+    {
+      "epoch": 2.2938636363636364,
+      "grad_norm": 1.583072835636141,
+      "learning_rate": 7.058119270513342e-07,
+      "loss": 0.0144,
+      "step": 10093
+    },
+    {
+      "epoch": 2.294090909090909,
+      "grad_norm": 1.7967851934493106,
+      "learning_rate": 7.057234256311896e-07,
+      "loss": 0.0229,
+      "step": 10094
+    },
+    {
+      "epoch": 2.294318181818182,
+      "grad_norm": 1.0502391245694538,
+      "learning_rate": 7.056349225649535e-07,
+      "loss": 0.0094,
+      "step": 10095
+    },
+    {
+      "epoch": 2.2945454545454544,
+      "grad_norm": 0.6093625377022994,
+      "learning_rate": 7.055464178544306e-07,
+      "loss": 0.0114,
+      "step": 10096
+    },
+    {
+      "epoch": 2.2947727272727274,
+      "grad_norm": 4.6391713349407215,
+      "learning_rate": 7.054579115014257e-07,
+      "loss": 0.0378,
+      "step": 10097
+    },
+    {
+      "epoch": 2.295,
+      "grad_norm": 1.140771562183799,
+      "learning_rate": 7.053694035077438e-07,
+      "loss": 0.0104,
+      "step": 10098
+    },
+    {
+      "epoch": 2.295227272727273,
+      "grad_norm": 1.795137886254852,
+      "learning_rate": 7.052808938751894e-07,
+      "loss": 0.0359,
+      "step": 10099
+    },
+    {
+      "epoch": 2.2954545454545454,
+      "grad_norm": 1.9833701864163384,
+      "learning_rate": 7.051923826055676e-07,
+      "loss": 0.0304,
+      "step": 10100
+    },
+    {
+      "epoch": 2.2956818181818184,
+      "grad_norm": 2.1990773326263664,
+      "learning_rate": 7.051038697006832e-07,
+      "loss": 0.0273,
+      "step": 10101
+    },
+    {
+      "epoch": 2.295909090909091,
+      "grad_norm": 0.8218767817504598,
+      "learning_rate": 7.050153551623409e-07,
+      "loss": 0.0057,
+      "step": 10102
+    },
+    {
+      "epoch": 2.2961363636363634,
+      "grad_norm": 2.9770856843058757,
+      "learning_rate": 7.049268389923461e-07,
+      "loss": 0.0598,
+      "step": 10103
+    },
+    {
+      "epoch": 2.2963636363636364,
+      "grad_norm": 1.6892075185909634,
+      "learning_rate": 7.048383211925038e-07,
+      "loss": 0.0207,
+      "step": 10104
+    },
+    {
+      "epoch": 2.296590909090909,
+      "grad_norm": 1.8547066551249782,
+      "learning_rate": 7.047498017646186e-07,
+      "loss": 0.0196,
+      "step": 10105
+    },
+    {
+      "epoch": 2.296818181818182,
+      "grad_norm": 1.4693114485625893,
+      "learning_rate": 7.04661280710496e-07,
+      "loss": 0.0246,
+      "step": 10106
+    },
+    {
+      "epoch": 2.2970454545454544,
+      "grad_norm": 1.0867688014867685,
+      "learning_rate": 7.045727580319407e-07,
+      "loss": 0.01,
+      "step": 10107
+    },
+    {
+      "epoch": 2.2972727272727274,
+      "grad_norm": 3.214836469683178,
+      "learning_rate": 7.044842337307581e-07,
+      "loss": 0.0736,
+      "step": 10108
+    },
+    {
+      "epoch": 2.2975,
+      "grad_norm": 2.8696391038849813,
+      "learning_rate": 7.043957078087534e-07,
+      "loss": 0.0448,
+      "step": 10109
+    },
+    {
+      "epoch": 2.297727272727273,
+      "grad_norm": 3.2152078768148695,
+      "learning_rate": 7.043071802677315e-07,
+      "loss": 0.0479,
+      "step": 10110
+    },
+    {
+      "epoch": 2.2979545454545454,
+      "grad_norm": 1.2259339503860422,
+      "learning_rate": 7.042186511094981e-07,
+      "loss": 0.019,
+      "step": 10111
+    },
+    {
+      "epoch": 2.2981818181818183,
+      "grad_norm": 2.4996326845315413,
+      "learning_rate": 7.04130120335858e-07,
+      "loss": 0.0376,
+      "step": 10112
+    },
+    {
+      "epoch": 2.298409090909091,
+      "grad_norm": 1.8686836353201217,
+      "learning_rate": 7.040415879486169e-07,
+      "loss": 0.0236,
+      "step": 10113
+    },
+    {
+      "epoch": 2.298636363636364,
+      "grad_norm": 2.206160966344599,
+      "learning_rate": 7.039530539495798e-07,
+      "loss": 0.03,
+      "step": 10114
+    },
+    {
+      "epoch": 2.2988636363636363,
+      "grad_norm": 2.1846061713898273,
+      "learning_rate": 7.038645183405522e-07,
+      "loss": 0.0259,
+      "step": 10115
+    },
+    {
+      "epoch": 2.299090909090909,
+      "grad_norm": 2.930718773882941,
+      "learning_rate": 7.037759811233395e-07,
+      "loss": 0.023,
+      "step": 10116
+    },
+    {
+      "epoch": 2.299318181818182,
+      "grad_norm": 2.056240683783111,
+      "learning_rate": 7.036874422997471e-07,
+      "loss": 0.0439,
+      "step": 10117
+    },
+    {
+      "epoch": 2.299545454545455,
+      "grad_norm": 2.3024245977923803,
+      "learning_rate": 7.035989018715803e-07,
+      "loss": 0.0152,
+      "step": 10118
+    },
+    {
+      "epoch": 2.2997727272727273,
+      "grad_norm": 3.947153122541735,
+      "learning_rate": 7.03510359840645e-07,
+      "loss": 0.0394,
+      "step": 10119
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.108408397191899,
+      "learning_rate": 7.034218162087466e-07,
+      "loss": 0.025,
+      "step": 10120
+    },
+    {
+      "epoch": 2.300227272727273,
+      "grad_norm": 1.81364436179116,
+      "learning_rate": 7.033332709776902e-07,
+      "loss": 0.0154,
+      "step": 10121
+    },
+    {
+      "epoch": 2.3004545454545453,
+      "grad_norm": 1.65548753134716,
+      "learning_rate": 7.032447241492821e-07,
+      "loss": 0.0256,
+      "step": 10122
+    },
+    {
+      "epoch": 2.3006818181818183,
+      "grad_norm": 2.304516487712148,
+      "learning_rate": 7.031561757253273e-07,
+      "loss": 0.0182,
+      "step": 10123
+    },
+    {
+      "epoch": 2.300909090909091,
+      "grad_norm": 1.340844848919544,
+      "learning_rate": 7.030676257076317e-07,
+      "loss": 0.0308,
+      "step": 10124
+    },
+    {
+      "epoch": 2.3011363636363638,
+      "grad_norm": 2.003312516932486,
+      "learning_rate": 7.029790740980011e-07,
+      "loss": 0.0346,
+      "step": 10125
+    },
+    {
+      "epoch": 2.3013636363636363,
+      "grad_norm": 2.2566985461670424,
+      "learning_rate": 7.028905208982413e-07,
+      "loss": 0.0249,
+      "step": 10126
+    },
+    {
+      "epoch": 2.3015909090909092,
+      "grad_norm": 2.7243014012143925,
+      "learning_rate": 7.028019661101577e-07,
+      "loss": 0.034,
+      "step": 10127
+    },
+    {
+      "epoch": 2.3018181818181818,
+      "grad_norm": 1.7469817380363866,
+      "learning_rate": 7.027134097355564e-07,
+      "loss": 0.0189,
+      "step": 10128
+    },
+    {
+      "epoch": 2.3020454545454547,
+      "grad_norm": 1.682385678141751,
+      "learning_rate": 7.026248517762431e-07,
+      "loss": 0.0197,
+      "step": 10129
+    },
+    {
+      "epoch": 2.3022727272727272,
+      "grad_norm": 1.8243944211940342,
+      "learning_rate": 7.025362922340235e-07,
+      "loss": 0.0146,
+      "step": 10130
+    },
+    {
+      "epoch": 2.3025,
+      "grad_norm": 2.6094034320217268,
+      "learning_rate": 7.024477311107038e-07,
+      "loss": 0.0434,
+      "step": 10131
+    },
+    {
+      "epoch": 2.3027272727272727,
+      "grad_norm": 2.046880940637163,
+      "learning_rate": 7.023591684080895e-07,
+      "loss": 0.0427,
+      "step": 10132
+    },
+    {
+      "epoch": 2.3029545454545453,
+      "grad_norm": 2.027977157291915,
+      "learning_rate": 7.022706041279871e-07,
+      "loss": 0.0197,
+      "step": 10133
+    },
+    {
+      "epoch": 2.303181818181818,
+      "grad_norm": 2.058433207478304,
+      "learning_rate": 7.021820382722021e-07,
+      "loss": 0.0428,
+      "step": 10134
+    },
+    {
+      "epoch": 2.3034090909090907,
+      "grad_norm": 1.6007629769755414,
+      "learning_rate": 7.020934708425408e-07,
+      "loss": 0.0139,
+      "step": 10135
+    },
+    {
+      "epoch": 2.3036363636363637,
+      "grad_norm": 2.714344497109982,
+      "learning_rate": 7.02004901840809e-07,
+      "loss": 0.0281,
+      "step": 10136
+    },
+    {
+      "epoch": 2.3038636363636362,
+      "grad_norm": 1.7822967567113384,
+      "learning_rate": 7.019163312688131e-07,
+      "loss": 0.0194,
+      "step": 10137
+    },
+    {
+      "epoch": 2.304090909090909,
+      "grad_norm": 0.6499597733852378,
+      "learning_rate": 7.018277591283588e-07,
+      "loss": 0.0118,
+      "step": 10138
+    },
+    {
+      "epoch": 2.3043181818181817,
+      "grad_norm": 1.594272793097855,
+      "learning_rate": 7.017391854212525e-07,
+      "loss": 0.0195,
+      "step": 10139
+    },
+    {
+      "epoch": 2.3045454545454547,
+      "grad_norm": 2.266042106837464,
+      "learning_rate": 7.016506101493003e-07,
+      "loss": 0.0332,
+      "step": 10140
+    },
+    {
+      "epoch": 2.304772727272727,
+      "grad_norm": 1.6255522323055775,
+      "learning_rate": 7.015620333143085e-07,
+      "loss": 0.0295,
+      "step": 10141
+    },
+    {
+      "epoch": 2.305,
+      "grad_norm": 2.457853231444558,
+      "learning_rate": 7.014734549180834e-07,
+      "loss": 0.039,
+      "step": 10142
+    },
+    {
+      "epoch": 2.3052272727272727,
+      "grad_norm": 1.5214799633641407,
+      "learning_rate": 7.01384874962431e-07,
+      "loss": 0.0162,
+      "step": 10143
+    },
+    {
+      "epoch": 2.3054545454545456,
+      "grad_norm": 1.154904459193959,
+      "learning_rate": 7.012962934491578e-07,
+      "loss": 0.0108,
+      "step": 10144
+    },
+    {
+      "epoch": 2.305681818181818,
+      "grad_norm": 0.9432610545049077,
+      "learning_rate": 7.0120771038007e-07,
+      "loss": 0.0072,
+      "step": 10145
+    },
+    {
+      "epoch": 2.3059090909090907,
+      "grad_norm": 1.5910295284422453,
+      "learning_rate": 7.011191257569739e-07,
+      "loss": 0.0102,
+      "step": 10146
+    },
+    {
+      "epoch": 2.3061363636363637,
+      "grad_norm": 1.1659398017604108,
+      "learning_rate": 7.010305395816762e-07,
+      "loss": 0.017,
+      "step": 10147
+    },
+    {
+      "epoch": 2.306363636363636,
+      "grad_norm": 0.9254173846846587,
+      "learning_rate": 7.009419518559833e-07,
+      "loss": 0.0162,
+      "step": 10148
+    },
+    {
+      "epoch": 2.306590909090909,
+      "grad_norm": 1.2516488386457245,
+      "learning_rate": 7.008533625817013e-07,
+      "loss": 0.0105,
+      "step": 10149
+    },
+    {
+      "epoch": 2.3068181818181817,
+      "grad_norm": 1.001253670110457,
+      "learning_rate": 7.007647717606371e-07,
+      "loss": 0.0198,
+      "step": 10150
+    },
+    {
+      "epoch": 2.3070454545454546,
+      "grad_norm": 1.7580860116292272,
+      "learning_rate": 7.006761793945968e-07,
+      "loss": 0.0228,
+      "step": 10151
+    },
+    {
+      "epoch": 2.307272727272727,
+      "grad_norm": 2.03254140924999,
+      "learning_rate": 7.005875854853872e-07,
+      "loss": 0.0268,
+      "step": 10152
+    },
+    {
+      "epoch": 2.3075,
+      "grad_norm": 1.5011777971181899,
+      "learning_rate": 7.00498990034815e-07,
+      "loss": 0.0222,
+      "step": 10153
+    },
+    {
+      "epoch": 2.3077272727272726,
+      "grad_norm": 2.5003321553580466,
+      "learning_rate": 7.004103930446866e-07,
+      "loss": 0.0306,
+      "step": 10154
+    },
+    {
+      "epoch": 2.3079545454545456,
+      "grad_norm": 1.48289670165159,
+      "learning_rate": 7.003217945168086e-07,
+      "loss": 0.0143,
+      "step": 10155
+    },
+    {
+      "epoch": 2.308181818181818,
+      "grad_norm": 2.3369445543771676,
+      "learning_rate": 7.00233194452988e-07,
+      "loss": 0.017,
+      "step": 10156
+    },
+    {
+      "epoch": 2.308409090909091,
+      "grad_norm": 0.9452651613969041,
+      "learning_rate": 7.001445928550313e-07,
+      "loss": 0.01,
+      "step": 10157
+    },
+    {
+      "epoch": 2.3086363636363636,
+      "grad_norm": 2.006392856508558,
+      "learning_rate": 7.000559897247452e-07,
+      "loss": 0.0354,
+      "step": 10158
+    },
+    {
+      "epoch": 2.308863636363636,
+      "grad_norm": 1.208003950349878,
+      "learning_rate": 6.999673850639366e-07,
+      "loss": 0.0137,
+      "step": 10159
+    },
+    {
+      "epoch": 2.309090909090909,
+      "grad_norm": 2.0059866998294584,
+      "learning_rate": 6.998787788744122e-07,
+      "loss": 0.0385,
+      "step": 10160
+    },
+    {
+      "epoch": 2.309318181818182,
+      "grad_norm": 1.36579581617133,
+      "learning_rate": 6.997901711579787e-07,
+      "loss": 0.0278,
+      "step": 10161
+    },
+    {
+      "epoch": 2.3095454545454546,
+      "grad_norm": 1.5225439619646464,
+      "learning_rate": 6.997015619164433e-07,
+      "loss": 0.0111,
+      "step": 10162
+    },
+    {
+      "epoch": 2.309772727272727,
+      "grad_norm": 4.425975302571897,
+      "learning_rate": 6.996129511516126e-07,
+      "loss": 0.0536,
+      "step": 10163
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.1771239224787728,
+      "learning_rate": 6.995243388652937e-07,
+      "loss": 0.0253,
+      "step": 10164
+    },
+    {
+      "epoch": 2.3102272727272726,
+      "grad_norm": 2.094681497002624,
+      "learning_rate": 6.994357250592937e-07,
+      "loss": 0.0237,
+      "step": 10165
+    },
+    {
+      "epoch": 2.3104545454545455,
+      "grad_norm": 1.9326467263667348,
+      "learning_rate": 6.993471097354193e-07,
+      "loss": 0.0112,
+      "step": 10166
+    },
+    {
+      "epoch": 2.310681818181818,
+      "grad_norm": 1.9274169934705503,
+      "learning_rate": 6.992584928954775e-07,
+      "loss": 0.0314,
+      "step": 10167
+    },
+    {
+      "epoch": 2.310909090909091,
+      "grad_norm": 1.8269537004531968,
+      "learning_rate": 6.991698745412758e-07,
+      "loss": 0.0347,
+      "step": 10168
+    },
+    {
+      "epoch": 2.3111363636363635,
+      "grad_norm": 2.41844849635047,
+      "learning_rate": 6.990812546746206e-07,
+      "loss": 0.022,
+      "step": 10169
+    },
+    {
+      "epoch": 2.3113636363636365,
+      "grad_norm": 2.778803852436074,
+      "learning_rate": 6.989926332973194e-07,
+      "loss": 0.0202,
+      "step": 10170
+    },
+    {
+      "epoch": 2.311590909090909,
+      "grad_norm": 1.1330021474902467,
+      "learning_rate": 6.989040104111794e-07,
+      "loss": 0.0101,
+      "step": 10171
+    },
+    {
+      "epoch": 2.311818181818182,
+      "grad_norm": 2.1600268434506584,
+      "learning_rate": 6.988153860180079e-07,
+      "loss": 0.0187,
+      "step": 10172
+    },
+    {
+      "epoch": 2.3120454545454545,
+      "grad_norm": 2.4905734118713907,
+      "learning_rate": 6.987267601196116e-07,
+      "loss": 0.0322,
+      "step": 10173
+    },
+    {
+      "epoch": 2.3122727272727275,
+      "grad_norm": 3.0332258000935877,
+      "learning_rate": 6.986381327177981e-07,
+      "loss": 0.0438,
+      "step": 10174
+    },
+    {
+      "epoch": 2.3125,
+      "grad_norm": 2.7156085011331568,
+      "learning_rate": 6.985495038143748e-07,
+      "loss": 0.0554,
+      "step": 10175
+    },
+    {
+      "epoch": 2.3127272727272725,
+      "grad_norm": 1.945060303213118,
+      "learning_rate": 6.984608734111486e-07,
+      "loss": 0.0216,
+      "step": 10176
+    },
+    {
+      "epoch": 2.3129545454545455,
+      "grad_norm": 1.1250284975117217,
+      "learning_rate": 6.983722415099271e-07,
+      "loss": 0.0095,
+      "step": 10177
+    },
+    {
+      "epoch": 2.313181818181818,
+      "grad_norm": 1.5916410508673053,
+      "learning_rate": 6.982836081125176e-07,
+      "loss": 0.042,
+      "step": 10178
+    },
+    {
+      "epoch": 2.313409090909091,
+      "grad_norm": 1.0192320788482578,
+      "learning_rate": 6.981949732207274e-07,
+      "loss": 0.0058,
+      "step": 10179
+    },
+    {
+      "epoch": 2.3136363636363635,
+      "grad_norm": 3.8888082522721783,
+      "learning_rate": 6.981063368363642e-07,
+      "loss": 0.0256,
+      "step": 10180
+    },
+    {
+      "epoch": 2.3138636363636365,
+      "grad_norm": 1.3046433658976748,
+      "learning_rate": 6.980176989612352e-07,
+      "loss": 0.0252,
+      "step": 10181
+    },
+    {
+      "epoch": 2.314090909090909,
+      "grad_norm": 2.3077224900504496,
+      "learning_rate": 6.979290595971477e-07,
+      "loss": 0.0225,
+      "step": 10182
+    },
+    {
+      "epoch": 2.314318181818182,
+      "grad_norm": 1.8520798177417412,
+      "learning_rate": 6.978404187459098e-07,
+      "loss": 0.0231,
+      "step": 10183
+    },
+    {
+      "epoch": 2.3145454545454545,
+      "grad_norm": 1.5642119797711749,
+      "learning_rate": 6.977517764093283e-07,
+      "loss": 0.0129,
+      "step": 10184
+    },
+    {
+      "epoch": 2.3147727272727274,
+      "grad_norm": 1.768935955039873,
+      "learning_rate": 6.976631325892114e-07,
+      "loss": 0.0139,
+      "step": 10185
+    },
+    {
+      "epoch": 2.315,
+      "grad_norm": 3.182150780035839,
+      "learning_rate": 6.975744872873663e-07,
+      "loss": 0.038,
+      "step": 10186
+    },
+    {
+      "epoch": 2.315227272727273,
+      "grad_norm": 1.3397723765397993,
+      "learning_rate": 6.974858405056009e-07,
+      "loss": 0.0172,
+      "step": 10187
+    },
+    {
+      "epoch": 2.3154545454545454,
+      "grad_norm": 3.282267156713617,
+      "learning_rate": 6.973971922457227e-07,
+      "loss": 0.0361,
+      "step": 10188
+    },
+    {
+      "epoch": 2.315681818181818,
+      "grad_norm": 2.1657702424839296,
+      "learning_rate": 6.973085425095395e-07,
+      "loss": 0.0242,
+      "step": 10189
+    },
+    {
+      "epoch": 2.315909090909091,
+      "grad_norm": 1.1723530614064883,
+      "learning_rate": 6.972198912988591e-07,
+      "loss": 0.0079,
+      "step": 10190
+    },
+    {
+      "epoch": 2.316136363636364,
+      "grad_norm": 1.8846717864539122,
+      "learning_rate": 6.97131238615489e-07,
+      "loss": 0.0329,
+      "step": 10191
+    },
+    {
+      "epoch": 2.3163636363636364,
+      "grad_norm": 1.264357249886009,
+      "learning_rate": 6.97042584461237e-07,
+      "loss": 0.0229,
+      "step": 10192
+    },
+    {
+      "epoch": 2.316590909090909,
+      "grad_norm": 1.353709627204435,
+      "learning_rate": 6.969539288379112e-07,
+      "loss": 0.0207,
+      "step": 10193
+    },
+    {
+      "epoch": 2.316818181818182,
+      "grad_norm": 2.578432306355291,
+      "learning_rate": 6.968652717473191e-07,
+      "loss": 0.0373,
+      "step": 10194
+    },
+    {
+      "epoch": 2.3170454545454544,
+      "grad_norm": 2.0631920339973084,
+      "learning_rate": 6.967766131912687e-07,
+      "loss": 0.0213,
+      "step": 10195
+    },
+    {
+      "epoch": 2.3172727272727274,
+      "grad_norm": 1.133393028245033,
+      "learning_rate": 6.966879531715681e-07,
+      "loss": 0.0204,
+      "step": 10196
+    },
+    {
+      "epoch": 2.3175,
+      "grad_norm": 0.9728743944256625,
+      "learning_rate": 6.965992916900251e-07,
+      "loss": 0.0069,
+      "step": 10197
+    },
+    {
+      "epoch": 2.317727272727273,
+      "grad_norm": 6.6424616537980725,
+      "learning_rate": 6.965106287484475e-07,
+      "loss": 0.027,
+      "step": 10198
+    },
+    {
+      "epoch": 2.3179545454545454,
+      "grad_norm": 0.5360623101481427,
+      "learning_rate": 6.964219643486435e-07,
+      "loss": 0.0121,
+      "step": 10199
+    },
+    {
+      "epoch": 2.3181818181818183,
+      "grad_norm": 2.0701004981949436,
+      "learning_rate": 6.963332984924209e-07,
+      "loss": 0.0188,
+      "step": 10200
+    },
+    {
+      "epoch": 2.318409090909091,
+      "grad_norm": 1.6090719961159785,
+      "learning_rate": 6.962446311815879e-07,
+      "loss": 0.0295,
+      "step": 10201
+    },
+    {
+      "epoch": 2.318636363636364,
+      "grad_norm": 2.1611528835496285,
+      "learning_rate": 6.961559624179528e-07,
+      "loss": 0.0292,
+      "step": 10202
+    },
+    {
+      "epoch": 2.3188636363636363,
+      "grad_norm": 2.9046211172541585,
+      "learning_rate": 6.960672922033233e-07,
+      "loss": 0.0268,
+      "step": 10203
+    },
+    {
+      "epoch": 2.3190909090909093,
+      "grad_norm": 2.6850925059180835,
+      "learning_rate": 6.959786205395077e-07,
+      "loss": 0.0345,
+      "step": 10204
+    },
+    {
+      "epoch": 2.319318181818182,
+      "grad_norm": 1.6972262102673203,
+      "learning_rate": 6.958899474283144e-07,
+      "loss": 0.0298,
+      "step": 10205
+    },
+    {
+      "epoch": 2.3195454545454544,
+      "grad_norm": 4.075823431640161,
+      "learning_rate": 6.958012728715513e-07,
+      "loss": 0.0353,
+      "step": 10206
+    },
+    {
+      "epoch": 2.3197727272727273,
+      "grad_norm": 1.3762654110940182,
+      "learning_rate": 6.957125968710267e-07,
+      "loss": 0.0161,
+      "step": 10207
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.6727196271989737,
+      "learning_rate": 6.956239194285488e-07,
+      "loss": 0.0187,
+      "step": 10208
+    },
+    {
+      "epoch": 2.320227272727273,
+      "grad_norm": 2.00524860935543,
+      "learning_rate": 6.955352405459261e-07,
+      "loss": 0.0285,
+      "step": 10209
+    },
+    {
+      "epoch": 2.3204545454545453,
+      "grad_norm": 2.9896244385776827,
+      "learning_rate": 6.954465602249668e-07,
+      "loss": 0.0309,
+      "step": 10210
+    },
+    {
+      "epoch": 2.3206818181818183,
+      "grad_norm": 2.8319109114466268,
+      "learning_rate": 6.953578784674791e-07,
+      "loss": 0.0508,
+      "step": 10211
+    },
+    {
+      "epoch": 2.320909090909091,
+      "grad_norm": 2.084428565192749,
+      "learning_rate": 6.952691952752717e-07,
+      "loss": 0.0243,
+      "step": 10212
+    },
+    {
+      "epoch": 2.3211363636363638,
+      "grad_norm": 2.8157083134051373,
+      "learning_rate": 6.951805106501527e-07,
+      "loss": 0.0458,
+      "step": 10213
+    },
+    {
+      "epoch": 2.3213636363636363,
+      "grad_norm": 1.5697850777737952,
+      "learning_rate": 6.950918245939307e-07,
+      "loss": 0.016,
+      "step": 10214
+    },
+    {
+      "epoch": 2.3215909090909093,
+      "grad_norm": 2.1451226792437423,
+      "learning_rate": 6.95003137108414e-07,
+      "loss": 0.0237,
+      "step": 10215
+    },
+    {
+      "epoch": 2.321818181818182,
+      "grad_norm": 0.9233031033872057,
+      "learning_rate": 6.949144481954114e-07,
+      "loss": 0.0108,
+      "step": 10216
+    },
+    {
+      "epoch": 2.3220454545454547,
+      "grad_norm": 1.840186404916936,
+      "learning_rate": 6.948257578567311e-07,
+      "loss": 0.0209,
+      "step": 10217
+    },
+    {
+      "epoch": 2.3222727272727273,
+      "grad_norm": 2.6349812631780596,
+      "learning_rate": 6.947370660941819e-07,
+      "loss": 0.0162,
+      "step": 10218
+    },
+    {
+      "epoch": 2.3225,
+      "grad_norm": 1.2892118994312418,
+      "learning_rate": 6.94648372909572e-07,
+      "loss": 0.0166,
+      "step": 10219
+    },
+    {
+      "epoch": 2.3227272727272728,
+      "grad_norm": 1.32285836152764,
+      "learning_rate": 6.945596783047105e-07,
+      "loss": 0.0177,
+      "step": 10220
+    },
+    {
+      "epoch": 2.3229545454545453,
+      "grad_norm": 1.666161469215167,
+      "learning_rate": 6.944709822814058e-07,
+      "loss": 0.0241,
+      "step": 10221
+    },
+    {
+      "epoch": 2.3231818181818182,
+      "grad_norm": 1.353820424235833,
+      "learning_rate": 6.943822848414665e-07,
+      "loss": 0.0139,
+      "step": 10222
+    },
+    {
+      "epoch": 2.3234090909090908,
+      "grad_norm": 1.2212779429844844,
+      "learning_rate": 6.942935859867013e-07,
+      "loss": 0.0122,
+      "step": 10223
+    },
+    {
+      "epoch": 2.3236363636363637,
+      "grad_norm": 1.6210105185297863,
+      "learning_rate": 6.94204885718919e-07,
+      "loss": 0.0169,
+      "step": 10224
+    },
+    {
+      "epoch": 2.3238636363636362,
+      "grad_norm": 0.3908883457206313,
+      "learning_rate": 6.941161840399285e-07,
+      "loss": 0.0061,
+      "step": 10225
+    },
+    {
+      "epoch": 2.324090909090909,
+      "grad_norm": 0.8780623159165349,
+      "learning_rate": 6.940274809515384e-07,
+      "loss": 0.0102,
+      "step": 10226
+    },
+    {
+      "epoch": 2.3243181818181817,
+      "grad_norm": 0.7735177678653464,
+      "learning_rate": 6.939387764555575e-07,
+      "loss": 0.0055,
+      "step": 10227
+    },
+    {
+      "epoch": 2.3245454545454547,
+      "grad_norm": 0.7451914129833414,
+      "learning_rate": 6.938500705537947e-07,
+      "loss": 0.028,
+      "step": 10228
+    },
+    {
+      "epoch": 2.324772727272727,
+      "grad_norm": 0.9273938605236286,
+      "learning_rate": 6.937613632480588e-07,
+      "loss": 0.0077,
+      "step": 10229
+    },
+    {
+      "epoch": 2.325,
+      "grad_norm": 2.617522821791786,
+      "learning_rate": 6.936726545401586e-07,
+      "loss": 0.0306,
+      "step": 10230
+    },
+    {
+      "epoch": 2.3252272727272727,
+      "grad_norm": 0.9184821200242077,
+      "learning_rate": 6.935839444319033e-07,
+      "loss": 0.0077,
+      "step": 10231
+    },
+    {
+      "epoch": 2.325454545454545,
+      "grad_norm": 2.3827895878022227,
+      "learning_rate": 6.934952329251018e-07,
+      "loss": 0.0336,
+      "step": 10232
+    },
+    {
+      "epoch": 2.325681818181818,
+      "grad_norm": 3.4400921841423,
+      "learning_rate": 6.934065200215629e-07,
+      "loss": 0.0244,
+      "step": 10233
+    },
+    {
+      "epoch": 2.325909090909091,
+      "grad_norm": 2.2548831667953166,
+      "learning_rate": 6.933178057230959e-07,
+      "loss": 0.0435,
+      "step": 10234
+    },
+    {
+      "epoch": 2.3261363636363637,
+      "grad_norm": 1.281830164078099,
+      "learning_rate": 6.932290900315095e-07,
+      "loss": 0.0101,
+      "step": 10235
+    },
+    {
+      "epoch": 2.326363636363636,
+      "grad_norm": 1.570314133777208,
+      "learning_rate": 6.931403729486128e-07,
+      "loss": 0.018,
+      "step": 10236
+    },
+    {
+      "epoch": 2.326590909090909,
+      "grad_norm": 2.529857626995442,
+      "learning_rate": 6.930516544762153e-07,
+      "loss": 0.0235,
+      "step": 10237
+    },
+    {
+      "epoch": 2.3268181818181817,
+      "grad_norm": 1.3952195838756773,
+      "learning_rate": 6.929629346161255e-07,
+      "loss": 0.0131,
+      "step": 10238
+    },
+    {
+      "epoch": 2.3270454545454546,
+      "grad_norm": 2.42264380656511,
+      "learning_rate": 6.928742133701531e-07,
+      "loss": 0.0207,
+      "step": 10239
+    },
+    {
+      "epoch": 2.327272727272727,
+      "grad_norm": 2.050988293862686,
+      "learning_rate": 6.927854907401072e-07,
+      "loss": 0.0158,
+      "step": 10240
+    },
+    {
+      "epoch": 2.3275,
+      "grad_norm": 1.679512152005922,
+      "learning_rate": 6.926967667277967e-07,
+      "loss": 0.0225,
+      "step": 10241
+    },
+    {
+      "epoch": 2.3277272727272726,
+      "grad_norm": 2.5988328874031343,
+      "learning_rate": 6.926080413350312e-07,
+      "loss": 0.0188,
+      "step": 10242
+    },
+    {
+      "epoch": 2.3279545454545456,
+      "grad_norm": 1.256223907185132,
+      "learning_rate": 6.925193145636195e-07,
+      "loss": 0.0116,
+      "step": 10243
+    },
+    {
+      "epoch": 2.328181818181818,
+      "grad_norm": 1.8900604707815638,
+      "learning_rate": 6.924305864153714e-07,
+      "loss": 0.025,
+      "step": 10244
+    },
+    {
+      "epoch": 2.328409090909091,
+      "grad_norm": 1.5914953075855605,
+      "learning_rate": 6.923418568920959e-07,
+      "loss": 0.0139,
+      "step": 10245
+    },
+    {
+      "epoch": 2.3286363636363636,
+      "grad_norm": 1.561707722893316,
+      "learning_rate": 6.922531259956026e-07,
+      "loss": 0.0119,
+      "step": 10246
+    },
+    {
+      "epoch": 2.3288636363636366,
+      "grad_norm": 3.036428916156409,
+      "learning_rate": 6.921643937277006e-07,
+      "loss": 0.0363,
+      "step": 10247
+    },
+    {
+      "epoch": 2.329090909090909,
+      "grad_norm": 2.377208276088008,
+      "learning_rate": 6.920756600901994e-07,
+      "loss": 0.0341,
+      "step": 10248
+    },
+    {
+      "epoch": 2.3293181818181816,
+      "grad_norm": 1.3174488397117168,
+      "learning_rate": 6.919869250849087e-07,
+      "loss": 0.0088,
+      "step": 10249
+    },
+    {
+      "epoch": 2.3295454545454546,
+      "grad_norm": 1.8044669966635432,
+      "learning_rate": 6.918981887136375e-07,
+      "loss": 0.0257,
+      "step": 10250
+    },
+    {
+      "epoch": 2.329772727272727,
+      "grad_norm": 2.39465444206805,
+      "learning_rate": 6.918094509781957e-07,
+      "loss": 0.0308,
+      "step": 10251
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.025330428568252,
+      "learning_rate": 6.917207118803924e-07,
+      "loss": 0.0176,
+      "step": 10252
+    },
+    {
+      "epoch": 2.3302272727272726,
+      "grad_norm": 2.620502003156049,
+      "learning_rate": 6.916319714220375e-07,
+      "loss": 0.0286,
+      "step": 10253
+    },
+    {
+      "epoch": 2.3304545454545456,
+      "grad_norm": 3.4064512482820635,
+      "learning_rate": 6.915432296049404e-07,
+      "loss": 0.0279,
+      "step": 10254
+    },
+    {
+      "epoch": 2.330681818181818,
+      "grad_norm": 2.9310109646439604,
+      "learning_rate": 6.914544864309109e-07,
+      "loss": 0.0337,
+      "step": 10255
+    },
+    {
+      "epoch": 2.330909090909091,
+      "grad_norm": 3.6846982271503093,
+      "learning_rate": 6.913657419017585e-07,
+      "loss": 0.0348,
+      "step": 10256
+    },
+    {
+      "epoch": 2.3311363636363636,
+      "grad_norm": 5.565079453973392,
+      "learning_rate": 6.912769960192928e-07,
+      "loss": 0.025,
+      "step": 10257
+    },
+    {
+      "epoch": 2.3313636363636365,
+      "grad_norm": 1.626641781974056,
+      "learning_rate": 6.911882487853234e-07,
+      "loss": 0.0424,
+      "step": 10258
+    },
+    {
+      "epoch": 2.331590909090909,
+      "grad_norm": 1.6632653661166348,
+      "learning_rate": 6.910995002016603e-07,
+      "loss": 0.0175,
+      "step": 10259
+    },
+    {
+      "epoch": 2.331818181818182,
+      "grad_norm": 1.2436441556183824,
+      "learning_rate": 6.910107502701128e-07,
+      "loss": 0.01,
+      "step": 10260
+    },
+    {
+      "epoch": 2.3320454545454545,
+      "grad_norm": 2.774785727859219,
+      "learning_rate": 6.909219989924912e-07,
+      "loss": 0.0183,
+      "step": 10261
+    },
+    {
+      "epoch": 2.332272727272727,
+      "grad_norm": 1.162855248312347,
+      "learning_rate": 6.908332463706051e-07,
+      "loss": 0.0098,
+      "step": 10262
+    },
+    {
+      "epoch": 2.3325,
+      "grad_norm": 3.2689700487190683,
+      "learning_rate": 6.90744492406264e-07,
+      "loss": 0.0381,
+      "step": 10263
+    },
+    {
+      "epoch": 2.3327272727272725,
+      "grad_norm": 2.0456183590001453,
+      "learning_rate": 6.906557371012781e-07,
+      "loss": 0.022,
+      "step": 10264
+    },
+    {
+      "epoch": 2.3329545454545455,
+      "grad_norm": 3.9990178445126623,
+      "learning_rate": 6.905669804574573e-07,
+      "loss": 0.0414,
+      "step": 10265
+    },
+    {
+      "epoch": 2.333181818181818,
+      "grad_norm": 2.2146727756765396,
+      "learning_rate": 6.904782224766112e-07,
+      "loss": 0.029,
+      "step": 10266
+    },
+    {
+      "epoch": 2.333409090909091,
+      "grad_norm": 2.5345881526235945,
+      "learning_rate": 6.903894631605501e-07,
+      "loss": 0.0365,
+      "step": 10267
+    },
+    {
+      "epoch": 2.3336363636363635,
+      "grad_norm": 2.276112348811552,
+      "learning_rate": 6.903007025110836e-07,
+      "loss": 0.0377,
+      "step": 10268
+    },
+    {
+      "epoch": 2.3338636363636365,
+      "grad_norm": 1.312836010994017,
+      "learning_rate": 6.90211940530022e-07,
+      "loss": 0.0142,
+      "step": 10269
+    },
+    {
+      "epoch": 2.334090909090909,
+      "grad_norm": 1.735598583833651,
+      "learning_rate": 6.901231772191749e-07,
+      "loss": 0.0131,
+      "step": 10270
+    },
+    {
+      "epoch": 2.334318181818182,
+      "grad_norm": 1.3186572686277938,
+      "learning_rate": 6.900344125803529e-07,
+      "loss": 0.0218,
+      "step": 10271
+    },
+    {
+      "epoch": 2.3345454545454545,
+      "grad_norm": 1.7891135667685978,
+      "learning_rate": 6.899456466153657e-07,
+      "loss": 0.0258,
+      "step": 10272
+    },
+    {
+      "epoch": 2.3347727272727274,
+      "grad_norm": 1.3217594254422187,
+      "learning_rate": 6.898568793260234e-07,
+      "loss": 0.0272,
+      "step": 10273
+    },
+    {
+      "epoch": 2.335,
+      "grad_norm": 1.9869022976117998,
+      "learning_rate": 6.897681107141361e-07,
+      "loss": 0.0161,
+      "step": 10274
+    },
+    {
+      "epoch": 2.3352272727272725,
+      "grad_norm": 0.9010803038142289,
+      "learning_rate": 6.896793407815142e-07,
+      "loss": 0.0069,
+      "step": 10275
+    },
+    {
+      "epoch": 2.3354545454545454,
+      "grad_norm": 2.104070343030142,
+      "learning_rate": 6.895905695299676e-07,
+      "loss": 0.0353,
+      "step": 10276
+    },
+    {
+      "epoch": 2.3356818181818184,
+      "grad_norm": 1.8350803268335882,
+      "learning_rate": 6.895017969613066e-07,
+      "loss": 0.0488,
+      "step": 10277
+    },
+    {
+      "epoch": 2.335909090909091,
+      "grad_norm": 2.0375362138117046,
+      "learning_rate": 6.894130230773415e-07,
+      "loss": 0.0295,
+      "step": 10278
+    },
+    {
+      "epoch": 2.3361363636363635,
+      "grad_norm": 0.9606934905154186,
+      "learning_rate": 6.893242478798825e-07,
+      "loss": 0.0151,
+      "step": 10279
+    },
+    {
+      "epoch": 2.3363636363636364,
+      "grad_norm": 1.5812702716492268,
+      "learning_rate": 6.892354713707397e-07,
+      "loss": 0.012,
+      "step": 10280
+    },
+    {
+      "epoch": 2.336590909090909,
+      "grad_norm": 1.7726397908232823,
+      "learning_rate": 6.891466935517237e-07,
+      "loss": 0.0191,
+      "step": 10281
+    },
+    {
+      "epoch": 2.336818181818182,
+      "grad_norm": 1.782694621322098,
+      "learning_rate": 6.890579144246448e-07,
+      "loss": 0.0196,
+      "step": 10282
+    },
+    {
+      "epoch": 2.3370454545454544,
+      "grad_norm": 1.1788905636555271,
+      "learning_rate": 6.889691339913131e-07,
+      "loss": 0.0132,
+      "step": 10283
+    },
+    {
+      "epoch": 2.3372727272727274,
+      "grad_norm": 2.2705356331652404,
+      "learning_rate": 6.888803522535393e-07,
+      "loss": 0.0295,
+      "step": 10284
+    },
+    {
+      "epoch": 2.3375,
+      "grad_norm": 1.9528410915324046,
+      "learning_rate": 6.887915692131335e-07,
+      "loss": 0.0146,
+      "step": 10285
+    },
+    {
+      "epoch": 2.337727272727273,
+      "grad_norm": 0.8674189424290032,
+      "learning_rate": 6.887027848719063e-07,
+      "loss": 0.0089,
+      "step": 10286
+    },
+    {
+      "epoch": 2.3379545454545454,
+      "grad_norm": 1.7971750533790989,
+      "learning_rate": 6.886139992316683e-07,
+      "loss": 0.0227,
+      "step": 10287
+    },
+    {
+      "epoch": 2.3381818181818184,
+      "grad_norm": 1.5868130534123794,
+      "learning_rate": 6.885252122942298e-07,
+      "loss": 0.0215,
+      "step": 10288
+    },
+    {
+      "epoch": 2.338409090909091,
+      "grad_norm": 2.4694364867187852,
+      "learning_rate": 6.884364240614017e-07,
+      "loss": 0.0451,
+      "step": 10289
+    },
+    {
+      "epoch": 2.338636363636364,
+      "grad_norm": 1.7037880288334863,
+      "learning_rate": 6.883476345349939e-07,
+      "loss": 0.0186,
+      "step": 10290
+    },
+    {
+      "epoch": 2.3388636363636364,
+      "grad_norm": 0.9994382060071195,
+      "learning_rate": 6.882588437168173e-07,
+      "loss": 0.0111,
+      "step": 10291
+    },
+    {
+      "epoch": 2.339090909090909,
+      "grad_norm": 1.9910166011096697,
+      "learning_rate": 6.881700516086826e-07,
+      "loss": 0.0289,
+      "step": 10292
+    },
+    {
+      "epoch": 2.339318181818182,
+      "grad_norm": 0.9969949095830234,
+      "learning_rate": 6.880812582124005e-07,
+      "loss": 0.0105,
+      "step": 10293
+    },
+    {
+      "epoch": 2.3395454545454544,
+      "grad_norm": 1.584613903098352,
+      "learning_rate": 6.879924635297813e-07,
+      "loss": 0.0315,
+      "step": 10294
+    },
+    {
+      "epoch": 2.3397727272727273,
+      "grad_norm": 1.9414372035640555,
+      "learning_rate": 6.87903667562636e-07,
+      "loss": 0.0258,
+      "step": 10295
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.5852355774334637,
+      "learning_rate": 6.878148703127751e-07,
+      "loss": 0.0111,
+      "step": 10296
+    },
+    {
+      "epoch": 2.340227272727273,
+      "grad_norm": 1.9749513928474196,
+      "learning_rate": 6.877260717820093e-07,
+      "loss": 0.0219,
+      "step": 10297
+    },
+    {
+      "epoch": 2.3404545454545453,
+      "grad_norm": 0.9550508939290977,
+      "learning_rate": 6.876372719721496e-07,
+      "loss": 0.0084,
+      "step": 10298
+    },
+    {
+      "epoch": 2.3406818181818183,
+      "grad_norm": 1.319208411354759,
+      "learning_rate": 6.875484708850066e-07,
+      "loss": 0.0083,
+      "step": 10299
+    },
+    {
+      "epoch": 2.340909090909091,
+      "grad_norm": 2.4985861880461684,
+      "learning_rate": 6.874596685223913e-07,
+      "loss": 0.0486,
+      "step": 10300
+    },
+    {
+      "epoch": 2.341136363636364,
+      "grad_norm": 2.13617722334338,
+      "learning_rate": 6.873708648861143e-07,
+      "loss": 0.0322,
+      "step": 10301
+    },
+    {
+      "epoch": 2.3413636363636363,
+      "grad_norm": 2.752925357879335,
+      "learning_rate": 6.872820599779866e-07,
+      "loss": 0.0326,
+      "step": 10302
+    },
+    {
+      "epoch": 2.3415909090909093,
+      "grad_norm": 2.530693118845808,
+      "learning_rate": 6.871932537998189e-07,
+      "loss": 0.0417,
+      "step": 10303
+    },
+    {
+      "epoch": 2.341818181818182,
+      "grad_norm": 3.1455263403415095,
+      "learning_rate": 6.871044463534224e-07,
+      "loss": 0.035,
+      "step": 10304
+    },
+    {
+      "epoch": 2.3420454545454543,
+      "grad_norm": 1.704741649212477,
+      "learning_rate": 6.870156376406078e-07,
+      "loss": 0.0283,
+      "step": 10305
+    },
+    {
+      "epoch": 2.3422727272727273,
+      "grad_norm": 1.0813727980902443,
+      "learning_rate": 6.869268276631862e-07,
+      "loss": 0.0124,
+      "step": 10306
+    },
+    {
+      "epoch": 2.3425000000000002,
+      "grad_norm": 1.5053119033737452,
+      "learning_rate": 6.868380164229683e-07,
+      "loss": 0.0175,
+      "step": 10307
+    },
+    {
+      "epoch": 2.3427272727272728,
+      "grad_norm": 2.591122469928812,
+      "learning_rate": 6.867492039217656e-07,
+      "loss": 0.055,
+      "step": 10308
+    },
+    {
+      "epoch": 2.3429545454545453,
+      "grad_norm": 1.372310606303812,
+      "learning_rate": 6.86660390161389e-07,
+      "loss": 0.0175,
+      "step": 10309
+    },
+    {
+      "epoch": 2.3431818181818183,
+      "grad_norm": 1.0426476088529781,
+      "learning_rate": 6.865715751436493e-07,
+      "loss": 0.013,
+      "step": 10310
+    },
+    {
+      "epoch": 2.3434090909090908,
+      "grad_norm": 4.7567100582104045,
+      "learning_rate": 6.86482758870358e-07,
+      "loss": 0.0222,
+      "step": 10311
+    },
+    {
+      "epoch": 2.3436363636363637,
+      "grad_norm": 0.7231645046738713,
+      "learning_rate": 6.863939413433257e-07,
+      "loss": 0.0173,
+      "step": 10312
+    },
+    {
+      "epoch": 2.3438636363636363,
+      "grad_norm": 2.4840242375469885,
+      "learning_rate": 6.863051225643641e-07,
+      "loss": 0.0583,
+      "step": 10313
+    },
+    {
+      "epoch": 2.3440909090909092,
+      "grad_norm": 1.605000842507234,
+      "learning_rate": 6.862163025352838e-07,
+      "loss": 0.0277,
+      "step": 10314
+    },
+    {
+      "epoch": 2.3443181818181817,
+      "grad_norm": 1.633071370863063,
+      "learning_rate": 6.861274812578964e-07,
+      "loss": 0.0177,
+      "step": 10315
+    },
+    {
+      "epoch": 2.3445454545454547,
+      "grad_norm": 1.4971237199674812,
+      "learning_rate": 6.86038658734013e-07,
+      "loss": 0.0127,
+      "step": 10316
+    },
+    {
+      "epoch": 2.3447727272727272,
+      "grad_norm": 1.6103816367442327,
+      "learning_rate": 6.859498349654451e-07,
+      "loss": 0.0166,
+      "step": 10317
+    },
+    {
+      "epoch": 2.3449999999999998,
+      "grad_norm": 2.4232455400538044,
+      "learning_rate": 6.858610099540033e-07,
+      "loss": 0.0201,
+      "step": 10318
+    },
+    {
+      "epoch": 2.3452272727272727,
+      "grad_norm": 0.7966438743651423,
+      "learning_rate": 6.857721837014998e-07,
+      "loss": 0.0144,
+      "step": 10319
+    },
+    {
+      "epoch": 2.3454545454545457,
+      "grad_norm": 1.4094179637387767,
+      "learning_rate": 6.856833562097451e-07,
+      "loss": 0.0127,
+      "step": 10320
+    },
+    {
+      "epoch": 2.345681818181818,
+      "grad_norm": 1.9931557016277788,
+      "learning_rate": 6.85594527480551e-07,
+      "loss": 0.0512,
+      "step": 10321
+    },
+    {
+      "epoch": 2.3459090909090907,
+      "grad_norm": 1.7709326448078513,
+      "learning_rate": 6.855056975157287e-07,
+      "loss": 0.014,
+      "step": 10322
+    },
+    {
+      "epoch": 2.3461363636363637,
+      "grad_norm": 1.741354282069605,
+      "learning_rate": 6.854168663170898e-07,
+      "loss": 0.0161,
+      "step": 10323
+    },
+    {
+      "epoch": 2.346363636363636,
+      "grad_norm": 1.7413177902687087,
+      "learning_rate": 6.853280338864457e-07,
+      "loss": 0.019,
+      "step": 10324
+    },
+    {
+      "epoch": 2.346590909090909,
+      "grad_norm": 2.238118493029072,
+      "learning_rate": 6.852392002256076e-07,
+      "loss": 0.0218,
+      "step": 10325
+    },
+    {
+      "epoch": 2.3468181818181817,
+      "grad_norm": 2.177651030013857,
+      "learning_rate": 6.851503653363871e-07,
+      "loss": 0.0289,
+      "step": 10326
+    },
+    {
+      "epoch": 2.3470454545454547,
+      "grad_norm": 1.62983010521057,
+      "learning_rate": 6.850615292205956e-07,
+      "loss": 0.017,
+      "step": 10327
+    },
+    {
+      "epoch": 2.347272727272727,
+      "grad_norm": 2.1853446680599933,
+      "learning_rate": 6.84972691880045e-07,
+      "loss": 0.0249,
+      "step": 10328
+    },
+    {
+      "epoch": 2.3475,
+      "grad_norm": 2.076168017008325,
+      "learning_rate": 6.848838533165462e-07,
+      "loss": 0.0175,
+      "step": 10329
+    },
+    {
+      "epoch": 2.3477272727272727,
+      "grad_norm": 1.21682527958761,
+      "learning_rate": 6.847950135319116e-07,
+      "loss": 0.013,
+      "step": 10330
+    },
+    {
+      "epoch": 2.3479545454545456,
+      "grad_norm": 3.4376412884793064,
+      "learning_rate": 6.847061725279521e-07,
+      "loss": 0.0793,
+      "step": 10331
+    },
+    {
+      "epoch": 2.348181818181818,
+      "grad_norm": 1.605578999087237,
+      "learning_rate": 6.846173303064796e-07,
+      "loss": 0.0145,
+      "step": 10332
+    },
+    {
+      "epoch": 2.348409090909091,
+      "grad_norm": 3.371515277219833,
+      "learning_rate": 6.845284868693059e-07,
+      "loss": 0.0438,
+      "step": 10333
+    },
+    {
+      "epoch": 2.3486363636363636,
+      "grad_norm": 6.987805319340352,
+      "learning_rate": 6.844396422182423e-07,
+      "loss": 0.0628,
+      "step": 10334
+    },
+    {
+      "epoch": 2.348863636363636,
+      "grad_norm": 2.900324792072915,
+      "learning_rate": 6.843507963551008e-07,
+      "loss": 0.0318,
+      "step": 10335
+    },
+    {
+      "epoch": 2.349090909090909,
+      "grad_norm": 1.57539267712402,
+      "learning_rate": 6.84261949281693e-07,
+      "loss": 0.0142,
+      "step": 10336
+    },
+    {
+      "epoch": 2.3493181818181816,
+      "grad_norm": 1.1086917937786545,
+      "learning_rate": 6.841731009998305e-07,
+      "loss": 0.0145,
+      "step": 10337
+    },
+    {
+      "epoch": 2.3495454545454546,
+      "grad_norm": 0.822161390363829,
+      "learning_rate": 6.840842515113254e-07,
+      "loss": 0.013,
+      "step": 10338
+    },
+    {
+      "epoch": 2.349772727272727,
+      "grad_norm": 1.6256925424780015,
+      "learning_rate": 6.839954008179893e-07,
+      "loss": 0.0111,
+      "step": 10339
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 2.513713905546605,
+      "learning_rate": 6.83906548921634e-07,
+      "loss": 0.0198,
+      "step": 10340
+    },
+    {
+      "epoch": 2.3502272727272726,
+      "grad_norm": 2.2900449195503016,
+      "learning_rate": 6.838176958240716e-07,
+      "loss": 0.0269,
+      "step": 10341
+    },
+    {
+      "epoch": 2.3504545454545456,
+      "grad_norm": 1.7661958354145355,
+      "learning_rate": 6.837288415271136e-07,
+      "loss": 0.0269,
+      "step": 10342
+    },
+    {
+      "epoch": 2.350681818181818,
+      "grad_norm": 2.458836878802293,
+      "learning_rate": 6.836399860325721e-07,
+      "loss": 0.0184,
+      "step": 10343
+    },
+    {
+      "epoch": 2.350909090909091,
+      "grad_norm": 1.1384898567952122,
+      "learning_rate": 6.835511293422589e-07,
+      "loss": 0.0119,
+      "step": 10344
+    },
+    {
+      "epoch": 2.3511363636363636,
+      "grad_norm": 1.7638821054929348,
+      "learning_rate": 6.834622714579861e-07,
+      "loss": 0.0197,
+      "step": 10345
+    },
+    {
+      "epoch": 2.3513636363636365,
+      "grad_norm": 2.129832697067119,
+      "learning_rate": 6.833734123815657e-07,
+      "loss": 0.0322,
+      "step": 10346
+    },
+    {
+      "epoch": 2.351590909090909,
+      "grad_norm": 1.8452526533654567,
+      "learning_rate": 6.832845521148094e-07,
+      "loss": 0.0215,
+      "step": 10347
+    },
+    {
+      "epoch": 2.3518181818181816,
+      "grad_norm": 1.4163397643276292,
+      "learning_rate": 6.831956906595295e-07,
+      "loss": 0.0246,
+      "step": 10348
+    },
+    {
+      "epoch": 2.3520454545454546,
+      "grad_norm": 1.2939756029151679,
+      "learning_rate": 6.831068280175379e-07,
+      "loss": 0.0105,
+      "step": 10349
+    },
+    {
+      "epoch": 2.3522727272727275,
+      "grad_norm": 2.0694059167729457,
+      "learning_rate": 6.830179641906469e-07,
+      "loss": 0.0243,
+      "step": 10350
+    },
+    {
+      "epoch": 2.3525,
+      "grad_norm": 1.1151585233984496,
+      "learning_rate": 6.829290991806681e-07,
+      "loss": 0.014,
+      "step": 10351
+    },
+    {
+      "epoch": 2.3527272727272726,
+      "grad_norm": 4.213266288740668,
+      "learning_rate": 6.828402329894141e-07,
+      "loss": 0.0728,
+      "step": 10352
+    },
+    {
+      "epoch": 2.3529545454545455,
+      "grad_norm": 1.403223778142821,
+      "learning_rate": 6.827513656186967e-07,
+      "loss": 0.0264,
+      "step": 10353
+    },
+    {
+      "epoch": 2.353181818181818,
+      "grad_norm": 1.682384505603896,
+      "learning_rate": 6.826624970703281e-07,
+      "loss": 0.0255,
+      "step": 10354
+    },
+    {
+      "epoch": 2.353409090909091,
+      "grad_norm": 4.603372325663765,
+      "learning_rate": 6.825736273461208e-07,
+      "loss": 0.0491,
+      "step": 10355
+    },
+    {
+      "epoch": 2.3536363636363635,
+      "grad_norm": 1.5250783983012062,
+      "learning_rate": 6.824847564478868e-07,
+      "loss": 0.0117,
+      "step": 10356
+    },
+    {
+      "epoch": 2.3538636363636365,
+      "grad_norm": 1.6350531513538233,
+      "learning_rate": 6.823958843774383e-07,
+      "loss": 0.0164,
+      "step": 10357
+    },
+    {
+      "epoch": 2.354090909090909,
+      "grad_norm": 2.475932542493826,
+      "learning_rate": 6.823070111365876e-07,
+      "loss": 0.0256,
+      "step": 10358
+    },
+    {
+      "epoch": 2.354318181818182,
+      "grad_norm": 2.8067455264892187,
+      "learning_rate": 6.822181367271469e-07,
+      "loss": 0.0448,
+      "step": 10359
+    },
+    {
+      "epoch": 2.3545454545454545,
+      "grad_norm": 2.6217131185260394,
+      "learning_rate": 6.821292611509285e-07,
+      "loss": 0.025,
+      "step": 10360
+    },
+    {
+      "epoch": 2.3547727272727275,
+      "grad_norm": 0.5759718071138455,
+      "learning_rate": 6.820403844097449e-07,
+      "loss": 0.0043,
+      "step": 10361
+    },
+    {
+      "epoch": 2.355,
+      "grad_norm": 1.5642935967710285,
+      "learning_rate": 6.819515065054084e-07,
+      "loss": 0.0322,
+      "step": 10362
+    },
+    {
+      "epoch": 2.355227272727273,
+      "grad_norm": 1.6141763105382023,
+      "learning_rate": 6.818626274397314e-07,
+      "loss": 0.0316,
+      "step": 10363
+    },
+    {
+      "epoch": 2.3554545454545455,
+      "grad_norm": 0.8828494347452065,
+      "learning_rate": 6.81773747214526e-07,
+      "loss": 0.0061,
+      "step": 10364
+    },
+    {
+      "epoch": 2.355681818181818,
+      "grad_norm": 1.484321525007599,
+      "learning_rate": 6.816848658316048e-07,
+      "loss": 0.0218,
+      "step": 10365
+    },
+    {
+      "epoch": 2.355909090909091,
+      "grad_norm": 1.4767301606723333,
+      "learning_rate": 6.815959832927805e-07,
+      "loss": 0.0193,
+      "step": 10366
+    },
+    {
+      "epoch": 2.3561363636363635,
+      "grad_norm": 2.9468104166252393,
+      "learning_rate": 6.815070995998654e-07,
+      "loss": 0.0478,
+      "step": 10367
+    },
+    {
+      "epoch": 2.3563636363636364,
+      "grad_norm": 2.0737257569158807,
+      "learning_rate": 6.814182147546718e-07,
+      "loss": 0.0393,
+      "step": 10368
+    },
+    {
+      "epoch": 2.356590909090909,
+      "grad_norm": 2.5452036396778834,
+      "learning_rate": 6.813293287590124e-07,
+      "loss": 0.0356,
+      "step": 10369
+    },
+    {
+      "epoch": 2.356818181818182,
+      "grad_norm": 1.7570346552671898,
+      "learning_rate": 6.812404416146997e-07,
+      "loss": 0.0193,
+      "step": 10370
+    },
+    {
+      "epoch": 2.3570454545454544,
+      "grad_norm": 1.8880895321199558,
+      "learning_rate": 6.811515533235463e-07,
+      "loss": 0.0314,
+      "step": 10371
+    },
+    {
+      "epoch": 2.3572727272727274,
+      "grad_norm": 1.7183487139511144,
+      "learning_rate": 6.81062663887365e-07,
+      "loss": 0.024,
+      "step": 10372
+    },
+    {
+      "epoch": 2.3575,
+      "grad_norm": 2.4087584180112196,
+      "learning_rate": 6.809737733079678e-07,
+      "loss": 0.0129,
+      "step": 10373
+    },
+    {
+      "epoch": 2.357727272727273,
+      "grad_norm": 2.649628866931803,
+      "learning_rate": 6.808848815871678e-07,
+      "loss": 0.027,
+      "step": 10374
+    },
+    {
+      "epoch": 2.3579545454545454,
+      "grad_norm": 2.5179119336542777,
+      "learning_rate": 6.807959887267777e-07,
+      "loss": 0.0471,
+      "step": 10375
+    },
+    {
+      "epoch": 2.3581818181818184,
+      "grad_norm": 1.149243983353638,
+      "learning_rate": 6.8070709472861e-07,
+      "loss": 0.0154,
+      "step": 10376
+    },
+    {
+      "epoch": 2.358409090909091,
+      "grad_norm": 1.5050023417259855,
+      "learning_rate": 6.806181995944774e-07,
+      "loss": 0.0156,
+      "step": 10377
+    },
+    {
+      "epoch": 2.3586363636363634,
+      "grad_norm": 2.406576978906348,
+      "learning_rate": 6.805293033261928e-07,
+      "loss": 0.0325,
+      "step": 10378
+    },
+    {
+      "epoch": 2.3588636363636364,
+      "grad_norm": 1.6246665706819936,
+      "learning_rate": 6.804404059255688e-07,
+      "loss": 0.0221,
+      "step": 10379
+    },
+    {
+      "epoch": 2.359090909090909,
+      "grad_norm": 0.4463606672471651,
+      "learning_rate": 6.803515073944181e-07,
+      "loss": 0.0098,
+      "step": 10380
+    },
+    {
+      "epoch": 2.359318181818182,
+      "grad_norm": 1.546661010811789,
+      "learning_rate": 6.802626077345538e-07,
+      "loss": 0.0104,
+      "step": 10381
+    },
+    {
+      "epoch": 2.3595454545454544,
+      "grad_norm": 2.827425525805514,
+      "learning_rate": 6.801737069477883e-07,
+      "loss": 0.0294,
+      "step": 10382
+    },
+    {
+      "epoch": 2.3597727272727274,
+      "grad_norm": 2.8189849998103713,
+      "learning_rate": 6.800848050359346e-07,
+      "loss": 0.0311,
+      "step": 10383
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 2.1462998662322748,
+      "learning_rate": 6.799959020008059e-07,
+      "loss": 0.0137,
+      "step": 10384
+    },
+    {
+      "epoch": 2.360227272727273,
+      "grad_norm": 2.332384671642455,
+      "learning_rate": 6.799069978442147e-07,
+      "loss": 0.0333,
+      "step": 10385
+    },
+    {
+      "epoch": 2.3604545454545454,
+      "grad_norm": 1.6212712212897649,
+      "learning_rate": 6.798180925679738e-07,
+      "loss": 0.0135,
+      "step": 10386
+    },
+    {
+      "epoch": 2.3606818181818183,
+      "grad_norm": 1.4023326033681116,
+      "learning_rate": 6.797291861738965e-07,
+      "loss": 0.0309,
+      "step": 10387
+    },
+    {
+      "epoch": 2.360909090909091,
+      "grad_norm": 3.078948645456555,
+      "learning_rate": 6.796402786637958e-07,
+      "loss": 0.0349,
+      "step": 10388
+    },
+    {
+      "epoch": 2.361136363636364,
+      "grad_norm": 1.060957217189683,
+      "learning_rate": 6.795513700394842e-07,
+      "loss": 0.0158,
+      "step": 10389
+    },
+    {
+      "epoch": 2.3613636363636363,
+      "grad_norm": 0.945695422397778,
+      "learning_rate": 6.794624603027752e-07,
+      "loss": 0.0144,
+      "step": 10390
+    },
+    {
+      "epoch": 2.361590909090909,
+      "grad_norm": 1.6284613909270518,
+      "learning_rate": 6.793735494554815e-07,
+      "loss": 0.0212,
+      "step": 10391
+    },
+    {
+      "epoch": 2.361818181818182,
+      "grad_norm": 1.6684713892834986,
+      "learning_rate": 6.792846374994165e-07,
+      "loss": 0.0173,
+      "step": 10392
+    },
+    {
+      "epoch": 2.362045454545455,
+      "grad_norm": 2.542619327846744,
+      "learning_rate": 6.791957244363928e-07,
+      "loss": 0.0456,
+      "step": 10393
+    },
+    {
+      "epoch": 2.3622727272727273,
+      "grad_norm": 2.7061477387115813,
+      "learning_rate": 6.791068102682238e-07,
+      "loss": 0.0184,
+      "step": 10394
+    },
+    {
+      "epoch": 2.3625,
+      "grad_norm": 1.6550072868354018,
+      "learning_rate": 6.790178949967226e-07,
+      "loss": 0.0135,
+      "step": 10395
+    },
+    {
+      "epoch": 2.362727272727273,
+      "grad_norm": 2.085055453280482,
+      "learning_rate": 6.789289786237023e-07,
+      "loss": 0.0315,
+      "step": 10396
+    },
+    {
+      "epoch": 2.3629545454545453,
+      "grad_norm": 2.239364837964624,
+      "learning_rate": 6.78840061150976e-07,
+      "loss": 0.0281,
+      "step": 10397
+    },
+    {
+      "epoch": 2.3631818181818183,
+      "grad_norm": 2.383354840902857,
+      "learning_rate": 6.787511425803567e-07,
+      "loss": 0.0429,
+      "step": 10398
+    },
+    {
+      "epoch": 2.363409090909091,
+      "grad_norm": 1.7646489587710144,
+      "learning_rate": 6.78662222913658e-07,
+      "loss": 0.0271,
+      "step": 10399
+    },
+    {
+      "epoch": 2.3636363636363638,
+      "grad_norm": 1.2773564502739885,
+      "learning_rate": 6.78573302152693e-07,
+      "loss": 0.014,
+      "step": 10400
+    },
+    {
+      "epoch": 2.3638636363636363,
+      "grad_norm": 0.8919280332744008,
+      "learning_rate": 6.78484380299275e-07,
+      "loss": 0.0106,
+      "step": 10401
+    },
+    {
+      "epoch": 2.3640909090909092,
+      "grad_norm": 1.738296712691308,
+      "learning_rate": 6.783954573552169e-07,
+      "loss": 0.0225,
+      "step": 10402
+    },
+    {
+      "epoch": 2.3643181818181818,
+      "grad_norm": 1.9878310975529434,
+      "learning_rate": 6.783065333223327e-07,
+      "loss": 0.0285,
+      "step": 10403
+    },
+    {
+      "epoch": 2.3645454545454547,
+      "grad_norm": 1.966448338389457,
+      "learning_rate": 6.782176082024349e-07,
+      "loss": 0.0136,
+      "step": 10404
+    },
+    {
+      "epoch": 2.3647727272727272,
+      "grad_norm": 2.7378467348879574,
+      "learning_rate": 6.781286819973372e-07,
+      "loss": 0.028,
+      "step": 10405
+    },
+    {
+      "epoch": 2.365,
+      "grad_norm": 1.574756901730208,
+      "learning_rate": 6.780397547088531e-07,
+      "loss": 0.0326,
+      "step": 10406
+    },
+    {
+      "epoch": 2.3652272727272727,
+      "grad_norm": 2.89837340531719,
+      "learning_rate": 6.77950826338796e-07,
+      "loss": 0.0391,
+      "step": 10407
+    },
+    {
+      "epoch": 2.3654545454545453,
+      "grad_norm": 2.523673466026929,
+      "learning_rate": 6.77861896888979e-07,
+      "loss": 0.0428,
+      "step": 10408
+    },
+    {
+      "epoch": 2.365681818181818,
+      "grad_norm": 1.207023643529972,
+      "learning_rate": 6.777729663612158e-07,
+      "loss": 0.0164,
+      "step": 10409
+    },
+    {
+      "epoch": 2.3659090909090907,
+      "grad_norm": 1.4593343500611675,
+      "learning_rate": 6.776840347573198e-07,
+      "loss": 0.0179,
+      "step": 10410
+    },
+    {
+      "epoch": 2.3661363636363637,
+      "grad_norm": 2.239298767767106,
+      "learning_rate": 6.775951020791043e-07,
+      "loss": 0.0201,
+      "step": 10411
+    },
+    {
+      "epoch": 2.3663636363636362,
+      "grad_norm": 1.1677565776854115,
+      "learning_rate": 6.77506168328383e-07,
+      "loss": 0.0108,
+      "step": 10412
+    },
+    {
+      "epoch": 2.366590909090909,
+      "grad_norm": 2.7628796117511114,
+      "learning_rate": 6.774172335069691e-07,
+      "loss": 0.0178,
+      "step": 10413
+    },
+    {
+      "epoch": 2.3668181818181817,
+      "grad_norm": 1.8645901890809482,
+      "learning_rate": 6.773282976166766e-07,
+      "loss": 0.0106,
+      "step": 10414
+    },
+    {
+      "epoch": 2.3670454545454547,
+      "grad_norm": 1.161040614761707,
+      "learning_rate": 6.772393606593188e-07,
+      "loss": 0.0158,
+      "step": 10415
+    },
+    {
+      "epoch": 2.367272727272727,
+      "grad_norm": 3.3485352331455744,
+      "learning_rate": 6.771504226367093e-07,
+      "loss": 0.0738,
+      "step": 10416
+    },
+    {
+      "epoch": 2.3675,
+      "grad_norm": 1.883793927655888,
+      "learning_rate": 6.770614835506616e-07,
+      "loss": 0.0187,
+      "step": 10417
+    },
+    {
+      "epoch": 2.3677272727272727,
+      "grad_norm": 1.435783079670041,
+      "learning_rate": 6.769725434029895e-07,
+      "loss": 0.014,
+      "step": 10418
+    },
+    {
+      "epoch": 2.3679545454545456,
+      "grad_norm": 1.4629459697044707,
+      "learning_rate": 6.768836021955066e-07,
+      "loss": 0.0165,
+      "step": 10419
+    },
+    {
+      "epoch": 2.368181818181818,
+      "grad_norm": 2.347242306424479,
+      "learning_rate": 6.767946599300264e-07,
+      "loss": 0.0147,
+      "step": 10420
+    },
+    {
+      "epoch": 2.3684090909090907,
+      "grad_norm": 2.4837393850899314,
+      "learning_rate": 6.767057166083628e-07,
+      "loss": 0.0323,
+      "step": 10421
+    },
+    {
+      "epoch": 2.3686363636363637,
+      "grad_norm": 1.665823881284852,
+      "learning_rate": 6.766167722323295e-07,
+      "loss": 0.0202,
+      "step": 10422
+    },
+    {
+      "epoch": 2.368863636363636,
+      "grad_norm": 2.830741062319838,
+      "learning_rate": 6.765278268037401e-07,
+      "loss": 0.0412,
+      "step": 10423
+    },
+    {
+      "epoch": 2.369090909090909,
+      "grad_norm": 2.6149552347883263,
+      "learning_rate": 6.764388803244083e-07,
+      "loss": 0.027,
+      "step": 10424
+    },
+    {
+      "epoch": 2.3693181818181817,
+      "grad_norm": 1.2450590531731431,
+      "learning_rate": 6.763499327961482e-07,
+      "loss": 0.012,
+      "step": 10425
+    },
+    {
+      "epoch": 2.3695454545454546,
+      "grad_norm": 1.661523733374938,
+      "learning_rate": 6.762609842207733e-07,
+      "loss": 0.0247,
+      "step": 10426
+    },
+    {
+      "epoch": 2.369772727272727,
+      "grad_norm": 4.928694899578975,
+      "learning_rate": 6.761720346000974e-07,
+      "loss": 0.0608,
+      "step": 10427
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.071550333581827,
+      "learning_rate": 6.760830839359346e-07,
+      "loss": 0.0565,
+      "step": 10428
+    },
+    {
+      "epoch": 2.3702272727272726,
+      "grad_norm": 2.281324514455298,
+      "learning_rate": 6.759941322300987e-07,
+      "loss": 0.0256,
+      "step": 10429
+    },
+    {
+      "epoch": 2.3704545454545456,
+      "grad_norm": 1.8249831189413053,
+      "learning_rate": 6.759051794844032e-07,
+      "loss": 0.0229,
+      "step": 10430
+    },
+    {
+      "epoch": 2.370681818181818,
+      "grad_norm": 5.398053747247302,
+      "learning_rate": 6.758162257006623e-07,
+      "loss": 0.0278,
+      "step": 10431
+    },
+    {
+      "epoch": 2.370909090909091,
+      "grad_norm": 1.9684757701978408,
+      "learning_rate": 6.757272708806902e-07,
+      "loss": 0.0228,
+      "step": 10432
+    },
+    {
+      "epoch": 2.3711363636363636,
+      "grad_norm": 1.1731670868769557,
+      "learning_rate": 6.756383150263002e-07,
+      "loss": 0.0071,
+      "step": 10433
+    },
+    {
+      "epoch": 2.371363636363636,
+      "grad_norm": 1.6139440916235148,
+      "learning_rate": 6.755493581393069e-07,
+      "loss": 0.028,
+      "step": 10434
+    },
+    {
+      "epoch": 2.371590909090909,
+      "grad_norm": 3.8874552611238813,
+      "learning_rate": 6.754604002215238e-07,
+      "loss": 0.0641,
+      "step": 10435
+    },
+    {
+      "epoch": 2.371818181818182,
+      "grad_norm": 0.8931314556848413,
+      "learning_rate": 6.753714412747652e-07,
+      "loss": 0.0073,
+      "step": 10436
+    },
+    {
+      "epoch": 2.3720454545454546,
+      "grad_norm": 1.8246649900595833,
+      "learning_rate": 6.752824813008449e-07,
+      "loss": 0.0281,
+      "step": 10437
+    },
+    {
+      "epoch": 2.372272727272727,
+      "grad_norm": 1.8452553500160407,
+      "learning_rate": 6.751935203015772e-07,
+      "loss": 0.0209,
+      "step": 10438
+    },
+    {
+      "epoch": 2.3725,
+      "grad_norm": 1.861096223411626,
+      "learning_rate": 6.75104558278776e-07,
+      "loss": 0.025,
+      "step": 10439
+    },
+    {
+      "epoch": 2.3727272727272726,
+      "grad_norm": 3.03811775085987,
+      "learning_rate": 6.750155952342556e-07,
+      "loss": 0.0304,
+      "step": 10440
+    },
+    {
+      "epoch": 2.3729545454545455,
+      "grad_norm": 2.2370466822110173,
+      "learning_rate": 6.749266311698298e-07,
+      "loss": 0.0205,
+      "step": 10441
+    },
+    {
+      "epoch": 2.373181818181818,
+      "grad_norm": 2.2113798573412207,
+      "learning_rate": 6.748376660873129e-07,
+      "loss": 0.0245,
+      "step": 10442
+    },
+    {
+      "epoch": 2.373409090909091,
+      "grad_norm": 2.372578910038477,
+      "learning_rate": 6.74748699988519e-07,
+      "loss": 0.0299,
+      "step": 10443
+    },
+    {
+      "epoch": 2.3736363636363635,
+      "grad_norm": 2.411180728202771,
+      "learning_rate": 6.746597328752625e-07,
+      "loss": 0.0274,
+      "step": 10444
+    },
+    {
+      "epoch": 2.3738636363636365,
+      "grad_norm": 3.8427099738793626,
+      "learning_rate": 6.745707647493572e-07,
+      "loss": 0.0319,
+      "step": 10445
+    },
+    {
+      "epoch": 2.374090909090909,
+      "grad_norm": 1.818135783078723,
+      "learning_rate": 6.744817956126176e-07,
+      "loss": 0.0184,
+      "step": 10446
+    },
+    {
+      "epoch": 2.374318181818182,
+      "grad_norm": 0.8555351634396402,
+      "learning_rate": 6.743928254668581e-07,
+      "loss": 0.0117,
+      "step": 10447
+    },
+    {
+      "epoch": 2.3745454545454545,
+      "grad_norm": 1.4517662570681906,
+      "learning_rate": 6.743038543138925e-07,
+      "loss": 0.0178,
+      "step": 10448
+    },
+    {
+      "epoch": 2.3747727272727275,
+      "grad_norm": 2.5431767845305404,
+      "learning_rate": 6.742148821555353e-07,
+      "loss": 0.0428,
+      "step": 10449
+    },
+    {
+      "epoch": 2.375,
+      "grad_norm": 3.3767763755501554,
+      "learning_rate": 6.741259089936008e-07,
+      "loss": 0.0369,
+      "step": 10450
+    },
+    {
+      "epoch": 2.3752272727272725,
+      "grad_norm": 1.971146421831263,
+      "learning_rate": 6.740369348299033e-07,
+      "loss": 0.0279,
+      "step": 10451
+    },
+    {
+      "epoch": 2.3754545454545455,
+      "grad_norm": 1.4647473133588995,
+      "learning_rate": 6.73947959666257e-07,
+      "loss": 0.0157,
+      "step": 10452
+    },
+    {
+      "epoch": 2.375681818181818,
+      "grad_norm": 4.967025556032205,
+      "learning_rate": 6.738589835044766e-07,
+      "loss": 0.0411,
+      "step": 10453
+    },
+    {
+      "epoch": 2.375909090909091,
+      "grad_norm": 1.6162076330918882,
+      "learning_rate": 6.737700063463762e-07,
+      "loss": 0.0152,
+      "step": 10454
+    },
+    {
+      "epoch": 2.3761363636363635,
+      "grad_norm": 1.5352049799029075,
+      "learning_rate": 6.736810281937703e-07,
+      "loss": 0.0204,
+      "step": 10455
+    },
+    {
+      "epoch": 2.3763636363636365,
+      "grad_norm": 1.1583312076819972,
+      "learning_rate": 6.735920490484734e-07,
+      "loss": 0.0125,
+      "step": 10456
+    },
+    {
+      "epoch": 2.376590909090909,
+      "grad_norm": 0.977774688646982,
+      "learning_rate": 6.735030689122998e-07,
+      "loss": 0.0069,
+      "step": 10457
+    },
+    {
+      "epoch": 2.376818181818182,
+      "grad_norm": 1.5434651062955187,
+      "learning_rate": 6.73414087787064e-07,
+      "loss": 0.0169,
+      "step": 10458
+    },
+    {
+      "epoch": 2.3770454545454545,
+      "grad_norm": 4.768394015153867,
+      "learning_rate": 6.733251056745804e-07,
+      "loss": 0.0512,
+      "step": 10459
+    },
+    {
+      "epoch": 2.3772727272727274,
+      "grad_norm": 2.292376045597535,
+      "learning_rate": 6.732361225766637e-07,
+      "loss": 0.0476,
+      "step": 10460
+    },
+    {
+      "epoch": 2.3775,
+      "grad_norm": 2.048590376318825,
+      "learning_rate": 6.731471384951283e-07,
+      "loss": 0.0491,
+      "step": 10461
+    },
+    {
+      "epoch": 2.377727272727273,
+      "grad_norm": 2.2408672471206685,
+      "learning_rate": 6.730581534317888e-07,
+      "loss": 0.0266,
+      "step": 10462
+    },
+    {
+      "epoch": 2.3779545454545454,
+      "grad_norm": 2.3906711753257923,
+      "learning_rate": 6.729691673884596e-07,
+      "loss": 0.0319,
+      "step": 10463
+    },
+    {
+      "epoch": 2.378181818181818,
+      "grad_norm": 1.52189754251198,
+      "learning_rate": 6.728801803669555e-07,
+      "loss": 0.0105,
+      "step": 10464
+    },
+    {
+      "epoch": 2.378409090909091,
+      "grad_norm": 3.575464385476703,
+      "learning_rate": 6.72791192369091e-07,
+      "loss": 0.0439,
+      "step": 10465
+    },
+    {
+      "epoch": 2.378636363636364,
+      "grad_norm": 2.2952439134597125,
+      "learning_rate": 6.727022033966804e-07,
+      "loss": 0.0562,
+      "step": 10466
+    },
+    {
+      "epoch": 2.3788636363636364,
+      "grad_norm": 2.0064089593664565,
+      "learning_rate": 6.72613213451539e-07,
+      "loss": 0.026,
+      "step": 10467
+    },
+    {
+      "epoch": 2.379090909090909,
+      "grad_norm": 2.023071310487869,
+      "learning_rate": 6.725242225354808e-07,
+      "loss": 0.0386,
+      "step": 10468
+    },
+    {
+      "epoch": 2.379318181818182,
+      "grad_norm": 3.8216165218219853,
+      "learning_rate": 6.724352306503211e-07,
+      "loss": 0.0364,
+      "step": 10469
+    },
+    {
+      "epoch": 2.3795454545454544,
+      "grad_norm": 2.0699681346259413,
+      "learning_rate": 6.723462377978741e-07,
+      "loss": 0.024,
+      "step": 10470
+    },
+    {
+      "epoch": 2.3797727272727274,
+      "grad_norm": 1.6626696397315017,
+      "learning_rate": 6.722572439799549e-07,
+      "loss": 0.0136,
+      "step": 10471
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.6201598775353776,
+      "learning_rate": 6.721682491983779e-07,
+      "loss": 0.0384,
+      "step": 10472
+    },
+    {
+      "epoch": 2.380227272727273,
+      "grad_norm": 1.9524738536751318,
+      "learning_rate": 6.72079253454958e-07,
+      "loss": 0.0174,
+      "step": 10473
+    },
+    {
+      "epoch": 2.3804545454545454,
+      "grad_norm": 1.7471383502913158,
+      "learning_rate": 6.719902567515101e-07,
+      "loss": 0.0172,
+      "step": 10474
+    },
+    {
+      "epoch": 2.3806818181818183,
+      "grad_norm": 1.791265695878381,
+      "learning_rate": 6.719012590898488e-07,
+      "loss": 0.0181,
+      "step": 10475
+    },
+    {
+      "epoch": 2.380909090909091,
+      "grad_norm": 2.227688202587054,
+      "learning_rate": 6.718122604717891e-07,
+      "loss": 0.0473,
+      "step": 10476
+    },
+    {
+      "epoch": 2.381136363636364,
+      "grad_norm": 3.528083197285125,
+      "learning_rate": 6.717232608991455e-07,
+      "loss": 0.025,
+      "step": 10477
+    },
+    {
+      "epoch": 2.3813636363636363,
+      "grad_norm": 1.3348981273020644,
+      "learning_rate": 6.716342603737333e-07,
+      "loss": 0.0189,
+      "step": 10478
+    },
+    {
+      "epoch": 2.3815909090909093,
+      "grad_norm": 1.9822433519612614,
+      "learning_rate": 6.715452588973671e-07,
+      "loss": 0.0198,
+      "step": 10479
+    },
+    {
+      "epoch": 2.381818181818182,
+      "grad_norm": 1.065765005286305,
+      "learning_rate": 6.71456256471862e-07,
+      "loss": 0.013,
+      "step": 10480
+    },
+    {
+      "epoch": 2.3820454545454544,
+      "grad_norm": 1.7450775638979223,
+      "learning_rate": 6.713672530990327e-07,
+      "loss": 0.0087,
+      "step": 10481
+    },
+    {
+      "epoch": 2.3822727272727273,
+      "grad_norm": 3.7209840404121195,
+      "learning_rate": 6.712782487806942e-07,
+      "loss": 0.0595,
+      "step": 10482
+    },
+    {
+      "epoch": 2.3825,
+      "grad_norm": 1.9638753468578096,
+      "learning_rate": 6.711892435186613e-07,
+      "loss": 0.0198,
+      "step": 10483
+    },
+    {
+      "epoch": 2.382727272727273,
+      "grad_norm": 1.4860586890221448,
+      "learning_rate": 6.711002373147494e-07,
+      "loss": 0.0198,
+      "step": 10484
+    },
+    {
+      "epoch": 2.3829545454545453,
+      "grad_norm": 2.606759681735011,
+      "learning_rate": 6.710112301707732e-07,
+      "loss": 0.0257,
+      "step": 10485
+    },
+    {
+      "epoch": 2.3831818181818183,
+      "grad_norm": 1.5939392808026018,
+      "learning_rate": 6.709222220885478e-07,
+      "loss": 0.031,
+      "step": 10486
+    },
+    {
+      "epoch": 2.383409090909091,
+      "grad_norm": 1.7914977607767677,
+      "learning_rate": 6.70833213069888e-07,
+      "loss": 0.02,
+      "step": 10487
+    },
+    {
+      "epoch": 2.3836363636363638,
+      "grad_norm": 0.7821970140877345,
+      "learning_rate": 6.70744203116609e-07,
+      "loss": 0.009,
+      "step": 10488
+    },
+    {
+      "epoch": 2.3838636363636363,
+      "grad_norm": 1.660696347535654,
+      "learning_rate": 6.70655192230526e-07,
+      "loss": 0.0162,
+      "step": 10489
+    },
+    {
+      "epoch": 2.3840909090909093,
+      "grad_norm": 1.834870825878584,
+      "learning_rate": 6.705661804134538e-07,
+      "loss": 0.0181,
+      "step": 10490
+    },
+    {
+      "epoch": 2.384318181818182,
+      "grad_norm": 1.7265492867357637,
+      "learning_rate": 6.704771676672081e-07,
+      "loss": 0.0153,
+      "step": 10491
+    },
+    {
+      "epoch": 2.3845454545454547,
+      "grad_norm": 2.247126989419924,
+      "learning_rate": 6.703881539936033e-07,
+      "loss": 0.0187,
+      "step": 10492
+    },
+    {
+      "epoch": 2.3847727272727273,
+      "grad_norm": 1.480039231165826,
+      "learning_rate": 6.702991393944551e-07,
+      "loss": 0.0132,
+      "step": 10493
+    },
+    {
+      "epoch": 2.385,
+      "grad_norm": 1.0757921478482695,
+      "learning_rate": 6.702101238715782e-07,
+      "loss": 0.0095,
+      "step": 10494
+    },
+    {
+      "epoch": 2.3852272727272728,
+      "grad_norm": 2.5187558342334526,
+      "learning_rate": 6.701211074267881e-07,
+      "loss": 0.034,
+      "step": 10495
+    },
+    {
+      "epoch": 2.3854545454545453,
+      "grad_norm": 1.1181493236286746,
+      "learning_rate": 6.700320900618997e-07,
+      "loss": 0.0178,
+      "step": 10496
+    },
+    {
+      "epoch": 2.3856818181818182,
+      "grad_norm": 2.5011454727605065,
+      "learning_rate": 6.699430717787286e-07,
+      "loss": 0.0255,
+      "step": 10497
+    },
+    {
+      "epoch": 2.3859090909090908,
+      "grad_norm": 1.7519059010890268,
+      "learning_rate": 6.698540525790898e-07,
+      "loss": 0.0132,
+      "step": 10498
+    },
+    {
+      "epoch": 2.3861363636363637,
+      "grad_norm": 1.6783982001349285,
+      "learning_rate": 6.697650324647987e-07,
+      "loss": 0.0252,
+      "step": 10499
+    },
+    {
+      "epoch": 2.3863636363636362,
+      "grad_norm": 4.392302801099856,
+      "learning_rate": 6.696760114376704e-07,
+      "loss": 0.0291,
+      "step": 10500
+    },
+    {
+      "epoch": 2.386590909090909,
+      "grad_norm": 1.9229083141372885,
+      "learning_rate": 6.695869894995202e-07,
+      "loss": 0.0183,
+      "step": 10501
+    },
+    {
+      "epoch": 2.3868181818181817,
+      "grad_norm": 3.6481663606071697,
+      "learning_rate": 6.694979666521638e-07,
+      "loss": 0.0649,
+      "step": 10502
+    },
+    {
+      "epoch": 2.3870454545454547,
+      "grad_norm": 2.2484680158690185,
+      "learning_rate": 6.69408942897416e-07,
+      "loss": 0.0215,
+      "step": 10503
+    },
+    {
+      "epoch": 2.387272727272727,
+      "grad_norm": 1.8832511655909951,
+      "learning_rate": 6.693199182370924e-07,
+      "loss": 0.0369,
+      "step": 10504
+    },
+    {
+      "epoch": 2.3875,
+      "grad_norm": 3.310734962421799,
+      "learning_rate": 6.692308926730084e-07,
+      "loss": 0.0491,
+      "step": 10505
+    },
+    {
+      "epoch": 2.3877272727272727,
+      "grad_norm": 1.4628762942416478,
+      "learning_rate": 6.691418662069793e-07,
+      "loss": 0.0182,
+      "step": 10506
+    },
+    {
+      "epoch": 2.387954545454545,
+      "grad_norm": 1.669250434902276,
+      "learning_rate": 6.690528388408205e-07,
+      "loss": 0.0127,
+      "step": 10507
+    },
+    {
+      "epoch": 2.388181818181818,
+      "grad_norm": 1.8575900257524687,
+      "learning_rate": 6.689638105763476e-07,
+      "loss": 0.0187,
+      "step": 10508
+    },
+    {
+      "epoch": 2.388409090909091,
+      "grad_norm": 2.9246620970162507,
+      "learning_rate": 6.688747814153758e-07,
+      "loss": 0.0551,
+      "step": 10509
+    },
+    {
+      "epoch": 2.3886363636363637,
+      "grad_norm": 1.09889534588469,
+      "learning_rate": 6.687857513597205e-07,
+      "loss": 0.0192,
+      "step": 10510
+    },
+    {
+      "epoch": 2.388863636363636,
+      "grad_norm": 2.0837162244555616,
+      "learning_rate": 6.686967204111976e-07,
+      "loss": 0.0284,
+      "step": 10511
+    },
+    {
+      "epoch": 2.389090909090909,
+      "grad_norm": 2.015694664861503,
+      "learning_rate": 6.686076885716222e-07,
+      "loss": 0.0245,
+      "step": 10512
+    },
+    {
+      "epoch": 2.3893181818181817,
+      "grad_norm": 2.389194725965936,
+      "learning_rate": 6.6851865584281e-07,
+      "loss": 0.0272,
+      "step": 10513
+    },
+    {
+      "epoch": 2.3895454545454546,
+      "grad_norm": 4.701024596883792,
+      "learning_rate": 6.684296222265766e-07,
+      "loss": 0.0548,
+      "step": 10514
+    },
+    {
+      "epoch": 2.389772727272727,
+      "grad_norm": 0.7947623065816933,
+      "learning_rate": 6.683405877247374e-07,
+      "loss": 0.0086,
+      "step": 10515
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 2.6876958275383895,
+      "learning_rate": 6.682515523391078e-07,
+      "loss": 0.0216,
+      "step": 10516
+    },
+    {
+      "epoch": 2.3902272727272726,
+      "grad_norm": 1.0143620320759705,
+      "learning_rate": 6.681625160715039e-07,
+      "loss": 0.0081,
+      "step": 10517
+    },
+    {
+      "epoch": 2.3904545454545456,
+      "grad_norm": 1.8631698062693256,
+      "learning_rate": 6.680734789237407e-07,
+      "loss": 0.0226,
+      "step": 10518
+    },
+    {
+      "epoch": 2.390681818181818,
+      "grad_norm": 1.368719669441996,
+      "learning_rate": 6.67984440897634e-07,
+      "loss": 0.0132,
+      "step": 10519
+    },
+    {
+      "epoch": 2.390909090909091,
+      "grad_norm": 1.05911587535068,
+      "learning_rate": 6.678954019949999e-07,
+      "loss": 0.0077,
+      "step": 10520
+    },
+    {
+      "epoch": 2.3911363636363636,
+      "grad_norm": 1.7261531098169294,
+      "learning_rate": 6.678063622176536e-07,
+      "loss": 0.0287,
+      "step": 10521
+    },
+    {
+      "epoch": 2.3913636363636366,
+      "grad_norm": 1.163208014664585,
+      "learning_rate": 6.67717321567411e-07,
+      "loss": 0.0104,
+      "step": 10522
+    },
+    {
+      "epoch": 2.391590909090909,
+      "grad_norm": 1.2898645410243,
+      "learning_rate": 6.676282800460876e-07,
+      "loss": 0.009,
+      "step": 10523
+    },
+    {
+      "epoch": 2.3918181818181816,
+      "grad_norm": 1.6443809775883778,
+      "learning_rate": 6.675392376554993e-07,
+      "loss": 0.0312,
+      "step": 10524
+    },
+    {
+      "epoch": 2.3920454545454546,
+      "grad_norm": 1.7490266082654748,
+      "learning_rate": 6.674501943974616e-07,
+      "loss": 0.014,
+      "step": 10525
+    },
+    {
+      "epoch": 2.392272727272727,
+      "grad_norm": 1.9940879122566249,
+      "learning_rate": 6.673611502737902e-07,
+      "loss": 0.0158,
+      "step": 10526
+    },
+    {
+      "epoch": 2.3925,
+      "grad_norm": 2.0924286935724568,
+      "learning_rate": 6.672721052863012e-07,
+      "loss": 0.0324,
+      "step": 10527
+    },
+    {
+      "epoch": 2.3927272727272726,
+      "grad_norm": 1.6582637386107688,
+      "learning_rate": 6.671830594368104e-07,
+      "loss": 0.0157,
+      "step": 10528
+    },
+    {
+      "epoch": 2.3929545454545456,
+      "grad_norm": 2.1383961485676455,
+      "learning_rate": 6.670940127271333e-07,
+      "loss": 0.0443,
+      "step": 10529
+    },
+    {
+      "epoch": 2.393181818181818,
+      "grad_norm": 0.9607762668700498,
+      "learning_rate": 6.67004965159086e-07,
+      "loss": 0.0059,
+      "step": 10530
+    },
+    {
+      "epoch": 2.393409090909091,
+      "grad_norm": 1.282011561456881,
+      "learning_rate": 6.669159167344839e-07,
+      "loss": 0.012,
+      "step": 10531
+    },
+    {
+      "epoch": 2.3936363636363636,
+      "grad_norm": 1.7260512143486784,
+      "learning_rate": 6.668268674551432e-07,
+      "loss": 0.025,
+      "step": 10532
+    },
+    {
+      "epoch": 2.3938636363636365,
+      "grad_norm": 1.5540729290776363,
+      "learning_rate": 6.667378173228798e-07,
+      "loss": 0.0115,
+      "step": 10533
+    },
+    {
+      "epoch": 2.394090909090909,
+      "grad_norm": 1.7129977188784975,
+      "learning_rate": 6.666487663395095e-07,
+      "loss": 0.0165,
+      "step": 10534
+    },
+    {
+      "epoch": 2.394318181818182,
+      "grad_norm": 3.132557438655731,
+      "learning_rate": 6.665597145068481e-07,
+      "loss": 0.0392,
+      "step": 10535
+    },
+    {
+      "epoch": 2.3945454545454545,
+      "grad_norm": 2.914917093019858,
+      "learning_rate": 6.664706618267115e-07,
+      "loss": 0.046,
+      "step": 10536
+    },
+    {
+      "epoch": 2.394772727272727,
+      "grad_norm": 1.3272042806079478,
+      "learning_rate": 6.66381608300916e-07,
+      "loss": 0.0264,
+      "step": 10537
+    },
+    {
+      "epoch": 2.395,
+      "grad_norm": 5.893468401086836,
+      "learning_rate": 6.662925539312773e-07,
+      "loss": 0.046,
+      "step": 10538
+    },
+    {
+      "epoch": 2.3952272727272725,
+      "grad_norm": 1.312541468707092,
+      "learning_rate": 6.662034987196113e-07,
+      "loss": 0.0179,
+      "step": 10539
+    },
+    {
+      "epoch": 2.3954545454545455,
+      "grad_norm": 3.3549569272147264,
+      "learning_rate": 6.66114442667734e-07,
+      "loss": 0.0499,
+      "step": 10540
+    },
+    {
+      "epoch": 2.395681818181818,
+      "grad_norm": 3.978031359077238,
+      "learning_rate": 6.660253857774616e-07,
+      "loss": 0.0628,
+      "step": 10541
+    },
+    {
+      "epoch": 2.395909090909091,
+      "grad_norm": 1.6368339251746695,
+      "learning_rate": 6.659363280506099e-07,
+      "loss": 0.026,
+      "step": 10542
+    },
+    {
+      "epoch": 2.3961363636363635,
+      "grad_norm": 1.635771789646828,
+      "learning_rate": 6.658472694889952e-07,
+      "loss": 0.014,
+      "step": 10543
+    },
+    {
+      "epoch": 2.3963636363636365,
+      "grad_norm": 2.062803165520184,
+      "learning_rate": 6.657582100944333e-07,
+      "loss": 0.0203,
+      "step": 10544
+    },
+    {
+      "epoch": 2.396590909090909,
+      "grad_norm": 1.8860054706921092,
+      "learning_rate": 6.656691498687405e-07,
+      "loss": 0.0258,
+      "step": 10545
+    },
+    {
+      "epoch": 2.396818181818182,
+      "grad_norm": 0.7291385920667206,
+      "learning_rate": 6.655800888137327e-07,
+      "loss": 0.0069,
+      "step": 10546
+    },
+    {
+      "epoch": 2.3970454545454545,
+      "grad_norm": 2.2666239093790157,
+      "learning_rate": 6.654910269312262e-07,
+      "loss": 0.0312,
+      "step": 10547
+    },
+    {
+      "epoch": 2.3972727272727274,
+      "grad_norm": 4.986569920950137,
+      "learning_rate": 6.65401964223037e-07,
+      "loss": 0.0468,
+      "step": 10548
+    },
+    {
+      "epoch": 2.3975,
+      "grad_norm": 1.4794154018129184,
+      "learning_rate": 6.653129006909811e-07,
+      "loss": 0.0108,
+      "step": 10549
+    },
+    {
+      "epoch": 2.3977272727272725,
+      "grad_norm": 1.8726856859862464,
+      "learning_rate": 6.652238363368749e-07,
+      "loss": 0.02,
+      "step": 10550
+    },
+    {
+      "epoch": 2.3979545454545454,
+      "grad_norm": 1.4194007128641457,
+      "learning_rate": 6.651347711625344e-07,
+      "loss": 0.0115,
+      "step": 10551
+    },
+    {
+      "epoch": 2.3981818181818184,
+      "grad_norm": 1.241883738667639,
+      "learning_rate": 6.650457051697762e-07,
+      "loss": 0.0129,
+      "step": 10552
+    },
+    {
+      "epoch": 2.398409090909091,
+      "grad_norm": 1.5632683961443659,
+      "learning_rate": 6.64956638360416e-07,
+      "loss": 0.0107,
+      "step": 10553
+    },
+    {
+      "epoch": 2.3986363636363635,
+      "grad_norm": 3.233002337423147,
+      "learning_rate": 6.648675707362702e-07,
+      "loss": 0.0336,
+      "step": 10554
+    },
+    {
+      "epoch": 2.3988636363636364,
+      "grad_norm": 1.932516253811545,
+      "learning_rate": 6.647785022991553e-07,
+      "loss": 0.02,
+      "step": 10555
+    },
+    {
+      "epoch": 2.399090909090909,
+      "grad_norm": 3.2795885626454377,
+      "learning_rate": 6.646894330508872e-07,
+      "loss": 0.0405,
+      "step": 10556
+    },
+    {
+      "epoch": 2.399318181818182,
+      "grad_norm": 2.082932147472518,
+      "learning_rate": 6.646003629932822e-07,
+      "loss": 0.0201,
+      "step": 10557
+    },
+    {
+      "epoch": 2.3995454545454544,
+      "grad_norm": 1.714988576266295,
+      "learning_rate": 6.645112921281568e-07,
+      "loss": 0.0344,
+      "step": 10558
+    },
+    {
+      "epoch": 2.3997727272727274,
+      "grad_norm": 2.0886804349040897,
+      "learning_rate": 6.644222204573272e-07,
+      "loss": 0.0252,
+      "step": 10559
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 2.115399571848598,
+      "learning_rate": 6.643331479826099e-07,
+      "loss": 0.0388,
+      "step": 10560
+    },
+    {
+      "epoch": 2.400227272727273,
+      "grad_norm": 1.2646418705416642,
+      "learning_rate": 6.642440747058209e-07,
+      "loss": 0.0203,
+      "step": 10561
+    },
+    {
+      "epoch": 2.4004545454545454,
+      "grad_norm": 1.9221190029711164,
+      "learning_rate": 6.64155000628777e-07,
+      "loss": 0.0309,
+      "step": 10562
+    },
+    {
+      "epoch": 2.4006818181818184,
+      "grad_norm": 2.549516414259786,
+      "learning_rate": 6.640659257532941e-07,
+      "loss": 0.0213,
+      "step": 10563
+    },
+    {
+      "epoch": 2.400909090909091,
+      "grad_norm": 2.1391864843292177,
+      "learning_rate": 6.639768500811889e-07,
+      "loss": 0.0395,
+      "step": 10564
+    },
+    {
+      "epoch": 2.401136363636364,
+      "grad_norm": 1.1208020973721204,
+      "learning_rate": 6.638877736142776e-07,
+      "loss": 0.0193,
+      "step": 10565
+    },
+    {
+      "epoch": 2.4013636363636364,
+      "grad_norm": 1.5432489922075165,
+      "learning_rate": 6.637986963543769e-07,
+      "loss": 0.0102,
+      "step": 10566
+    },
+    {
+      "epoch": 2.401590909090909,
+      "grad_norm": 2.0345405662961658,
+      "learning_rate": 6.637096183033031e-07,
+      "loss": 0.0298,
+      "step": 10567
+    },
+    {
+      "epoch": 2.401818181818182,
+      "grad_norm": 1.7299241643588712,
+      "learning_rate": 6.636205394628725e-07,
+      "loss": 0.0254,
+      "step": 10568
+    },
+    {
+      "epoch": 2.4020454545454544,
+      "grad_norm": 1.1077410140362258,
+      "learning_rate": 6.635314598349018e-07,
+      "loss": 0.0116,
+      "step": 10569
+    },
+    {
+      "epoch": 2.4022727272727273,
+      "grad_norm": 1.9066479006537835,
+      "learning_rate": 6.634423794212076e-07,
+      "loss": 0.0273,
+      "step": 10570
+    },
+    {
+      "epoch": 2.4025,
+      "grad_norm": 1.36283834732374,
+      "learning_rate": 6.63353298223606e-07,
+      "loss": 0.0181,
+      "step": 10571
+    },
+    {
+      "epoch": 2.402727272727273,
+      "grad_norm": 3.4070929962324787,
+      "learning_rate": 6.632642162439137e-07,
+      "loss": 0.0568,
+      "step": 10572
+    },
+    {
+      "epoch": 2.4029545454545453,
+      "grad_norm": 3.5621709013789813,
+      "learning_rate": 6.631751334839475e-07,
+      "loss": 0.0335,
+      "step": 10573
+    },
+    {
+      "epoch": 2.4031818181818183,
+      "grad_norm": 2.467809710406873,
+      "learning_rate": 6.630860499455236e-07,
+      "loss": 0.0248,
+      "step": 10574
+    },
+    {
+      "epoch": 2.403409090909091,
+      "grad_norm": 2.30632520248931,
+      "learning_rate": 6.629969656304587e-07,
+      "loss": 0.0238,
+      "step": 10575
+    },
+    {
+      "epoch": 2.403636363636364,
+      "grad_norm": 1.2530602686768602,
+      "learning_rate": 6.629078805405695e-07,
+      "loss": 0.0156,
+      "step": 10576
+    },
+    {
+      "epoch": 2.4038636363636363,
+      "grad_norm": 1.1252309525811925,
+      "learning_rate": 6.628187946776722e-07,
+      "loss": 0.0097,
+      "step": 10577
+    },
+    {
+      "epoch": 2.4040909090909093,
+      "grad_norm": 2.1939196380062564,
+      "learning_rate": 6.627297080435838e-07,
+      "loss": 0.0339,
+      "step": 10578
+    },
+    {
+      "epoch": 2.404318181818182,
+      "grad_norm": 1.7345100648507592,
+      "learning_rate": 6.626406206401211e-07,
+      "loss": 0.0258,
+      "step": 10579
+    },
+    {
+      "epoch": 2.4045454545454543,
+      "grad_norm": 1.2985797907584455,
+      "learning_rate": 6.625515324690999e-07,
+      "loss": 0.0128,
+      "step": 10580
+    },
+    {
+      "epoch": 2.4047727272727273,
+      "grad_norm": 1.3971330685941878,
+      "learning_rate": 6.62462443532338e-07,
+      "loss": 0.0126,
+      "step": 10581
+    },
+    {
+      "epoch": 2.4050000000000002,
+      "grad_norm": 1.072372384926917,
+      "learning_rate": 6.623733538316513e-07,
+      "loss": 0.0103,
+      "step": 10582
+    },
+    {
+      "epoch": 2.4052272727272728,
+      "grad_norm": 1.435520130353622,
+      "learning_rate": 6.622842633688569e-07,
+      "loss": 0.019,
+      "step": 10583
+    },
+    {
+      "epoch": 2.4054545454545453,
+      "grad_norm": 2.564588071473851,
+      "learning_rate": 6.62195172145771e-07,
+      "loss": 0.0383,
+      "step": 10584
+    },
+    {
+      "epoch": 2.4056818181818183,
+      "grad_norm": 2.9549206689587924,
+      "learning_rate": 6.62106080164211e-07,
+      "loss": 0.0291,
+      "step": 10585
+    },
+    {
+      "epoch": 2.4059090909090908,
+      "grad_norm": 1.0748962618478552,
+      "learning_rate": 6.62016987425993e-07,
+      "loss": 0.02,
+      "step": 10586
+    },
+    {
+      "epoch": 2.4061363636363637,
+      "grad_norm": 1.774196386781618,
+      "learning_rate": 6.619278939329341e-07,
+      "loss": 0.0178,
+      "step": 10587
+    },
+    {
+      "epoch": 2.4063636363636363,
+      "grad_norm": 1.1257621037549275,
+      "learning_rate": 6.61838799686851e-07,
+      "loss": 0.0123,
+      "step": 10588
+    },
+    {
+      "epoch": 2.4065909090909092,
+      "grad_norm": 1.37054471188642,
+      "learning_rate": 6.617497046895606e-07,
+      "loss": 0.0242,
+      "step": 10589
+    },
+    {
+      "epoch": 2.4068181818181817,
+      "grad_norm": 1.9515206500730562,
+      "learning_rate": 6.616606089428797e-07,
+      "loss": 0.0285,
+      "step": 10590
+    },
+    {
+      "epoch": 2.4070454545454547,
+      "grad_norm": 0.9412917758939807,
+      "learning_rate": 6.615715124486248e-07,
+      "loss": 0.0194,
+      "step": 10591
+    },
+    {
+      "epoch": 2.4072727272727272,
+      "grad_norm": 2.15451319258753,
+      "learning_rate": 6.614824152086132e-07,
+      "loss": 0.0242,
+      "step": 10592
+    },
+    {
+      "epoch": 2.4074999999999998,
+      "grad_norm": 2.62888204296807,
+      "learning_rate": 6.613933172246613e-07,
+      "loss": 0.0277,
+      "step": 10593
+    },
+    {
+      "epoch": 2.4077272727272727,
+      "grad_norm": 1.834715711170738,
+      "learning_rate": 6.613042184985863e-07,
+      "loss": 0.0322,
+      "step": 10594
+    },
+    {
+      "epoch": 2.4079545454545457,
+      "grad_norm": 1.1773568999065054,
+      "learning_rate": 6.612151190322047e-07,
+      "loss": 0.0144,
+      "step": 10595
+    },
+    {
+      "epoch": 2.408181818181818,
+      "grad_norm": 2.155228325489354,
+      "learning_rate": 6.61126018827334e-07,
+      "loss": 0.0196,
+      "step": 10596
+    },
+    {
+      "epoch": 2.4084090909090907,
+      "grad_norm": 2.160837402327591,
+      "learning_rate": 6.610369178857905e-07,
+      "loss": 0.0258,
+      "step": 10597
+    },
+    {
+      "epoch": 2.4086363636363637,
+      "grad_norm": 0.9990467779336093,
+      "learning_rate": 6.609478162093915e-07,
+      "loss": 0.0104,
+      "step": 10598
+    },
+    {
+      "epoch": 2.408863636363636,
+      "grad_norm": 3.0051682591459317,
+      "learning_rate": 6.608587137999537e-07,
+      "loss": 0.0342,
+      "step": 10599
+    },
+    {
+      "epoch": 2.409090909090909,
+      "grad_norm": 0.7310027965745622,
+      "learning_rate": 6.607696106592942e-07,
+      "loss": 0.0067,
+      "step": 10600
+    },
+    {
+      "epoch": 2.4093181818181817,
+      "grad_norm": 0.5322721843468345,
+      "learning_rate": 6.606805067892301e-07,
+      "loss": 0.0037,
+      "step": 10601
+    },
+    {
+      "epoch": 2.4095454545454547,
+      "grad_norm": 2.057961398204372,
+      "learning_rate": 6.605914021915781e-07,
+      "loss": 0.0233,
+      "step": 10602
+    },
+    {
+      "epoch": 2.409772727272727,
+      "grad_norm": 2.442114828787302,
+      "learning_rate": 6.605022968681553e-07,
+      "loss": 0.0175,
+      "step": 10603
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.9275633001425588,
+      "learning_rate": 6.604131908207787e-07,
+      "loss": 0.0081,
+      "step": 10604
+    },
+    {
+      "epoch": 2.4102272727272727,
+      "grad_norm": 2.314298314970485,
+      "learning_rate": 6.603240840512655e-07,
+      "loss": 0.0369,
+      "step": 10605
+    },
+    {
+      "epoch": 2.4104545454545456,
+      "grad_norm": 1.5728594419748048,
+      "learning_rate": 6.602349765614325e-07,
+      "loss": 0.0154,
+      "step": 10606
+    },
+    {
+      "epoch": 2.410681818181818,
+      "grad_norm": 1.6650942220868052,
+      "learning_rate": 6.601458683530969e-07,
+      "loss": 0.0253,
+      "step": 10607
+    },
+    {
+      "epoch": 2.410909090909091,
+      "grad_norm": 14.107868622940467,
+      "learning_rate": 6.600567594280756e-07,
+      "loss": 0.0353,
+      "step": 10608
+    },
+    {
+      "epoch": 2.4111363636363636,
+      "grad_norm": 1.7283191346453322,
+      "learning_rate": 6.59967649788186e-07,
+      "loss": 0.0255,
+      "step": 10609
+    },
+    {
+      "epoch": 2.411363636363636,
+      "grad_norm": 1.3025596534660129,
+      "learning_rate": 6.598785394352447e-07,
+      "loss": 0.0124,
+      "step": 10610
+    },
+    {
+      "epoch": 2.411590909090909,
+      "grad_norm": 0.8354589891930302,
+      "learning_rate": 6.597894283710692e-07,
+      "loss": 0.0071,
+      "step": 10611
+    },
+    {
+      "epoch": 2.4118181818181816,
+      "grad_norm": 2.067462283409752,
+      "learning_rate": 6.597003165974769e-07,
+      "loss": 0.0285,
+      "step": 10612
+    },
+    {
+      "epoch": 2.4120454545454546,
+      "grad_norm": 1.308947390692318,
+      "learning_rate": 6.596112041162845e-07,
+      "loss": 0.0154,
+      "step": 10613
+    },
+    {
+      "epoch": 2.412272727272727,
+      "grad_norm": 1.9088714635737454,
+      "learning_rate": 6.595220909293092e-07,
+      "loss": 0.0213,
+      "step": 10614
+    },
+    {
+      "epoch": 2.4125,
+      "grad_norm": 1.7114061819345765,
+      "learning_rate": 6.594329770383681e-07,
+      "loss": 0.0212,
+      "step": 10615
+    },
+    {
+      "epoch": 2.4127272727272726,
+      "grad_norm": 1.463679010536529,
+      "learning_rate": 6.593438624452787e-07,
+      "loss": 0.0111,
+      "step": 10616
+    },
+    {
+      "epoch": 2.4129545454545456,
+      "grad_norm": 2.6892593084464815,
+      "learning_rate": 6.59254747151858e-07,
+      "loss": 0.0113,
+      "step": 10617
+    },
+    {
+      "epoch": 2.413181818181818,
+      "grad_norm": 1.552263193300739,
+      "learning_rate": 6.591656311599231e-07,
+      "loss": 0.0146,
+      "step": 10618
+    },
+    {
+      "epoch": 2.413409090909091,
+      "grad_norm": 2.2339505637410575,
+      "learning_rate": 6.590765144712916e-07,
+      "loss": 0.0455,
+      "step": 10619
+    },
+    {
+      "epoch": 2.4136363636363636,
+      "grad_norm": 1.5439677945288508,
+      "learning_rate": 6.589873970877804e-07,
+      "loss": 0.0098,
+      "step": 10620
+    },
+    {
+      "epoch": 2.4138636363636365,
+      "grad_norm": 1.8072493186524776,
+      "learning_rate": 6.588982790112068e-07,
+      "loss": 0.0268,
+      "step": 10621
+    },
+    {
+      "epoch": 2.414090909090909,
+      "grad_norm": 2.456356488073788,
+      "learning_rate": 6.588091602433883e-07,
+      "loss": 0.0234,
+      "step": 10622
+    },
+    {
+      "epoch": 2.4143181818181816,
+      "grad_norm": 1.7344118780314493,
+      "learning_rate": 6.58720040786142e-07,
+      "loss": 0.0222,
+      "step": 10623
+    },
+    {
+      "epoch": 2.4145454545454546,
+      "grad_norm": 1.9435928620510456,
+      "learning_rate": 6.586309206412852e-07,
+      "loss": 0.0222,
+      "step": 10624
+    },
+    {
+      "epoch": 2.4147727272727275,
+      "grad_norm": 1.6639143352796038,
+      "learning_rate": 6.585417998106352e-07,
+      "loss": 0.0379,
+      "step": 10625
+    },
+    {
+      "epoch": 2.415,
+      "grad_norm": 1.9246982279063356,
+      "learning_rate": 6.584526782960095e-07,
+      "loss": 0.0237,
+      "step": 10626
+    },
+    {
+      "epoch": 2.4152272727272726,
+      "grad_norm": 2.0929673837621725,
+      "learning_rate": 6.583635560992253e-07,
+      "loss": 0.0189,
+      "step": 10627
+    },
+    {
+      "epoch": 2.4154545454545455,
+      "grad_norm": 1.6906036926718013,
+      "learning_rate": 6.582744332221e-07,
+      "loss": 0.0147,
+      "step": 10628
+    },
+    {
+      "epoch": 2.415681818181818,
+      "grad_norm": 3.0945470510585134,
+      "learning_rate": 6.58185309666451e-07,
+      "loss": 0.0471,
+      "step": 10629
+    },
+    {
+      "epoch": 2.415909090909091,
+      "grad_norm": 2.9807769366684718,
+      "learning_rate": 6.580961854340956e-07,
+      "loss": 0.0257,
+      "step": 10630
+    },
+    {
+      "epoch": 2.4161363636363635,
+      "grad_norm": 1.5884474436843703,
+      "learning_rate": 6.580070605268513e-07,
+      "loss": 0.0213,
+      "step": 10631
+    },
+    {
+      "epoch": 2.4163636363636365,
+      "grad_norm": 0.6062117954377475,
+      "learning_rate": 6.579179349465353e-07,
+      "loss": 0.0099,
+      "step": 10632
+    },
+    {
+      "epoch": 2.416590909090909,
+      "grad_norm": 2.5493514443042664,
+      "learning_rate": 6.578288086949652e-07,
+      "loss": 0.0564,
+      "step": 10633
+    },
+    {
+      "epoch": 2.416818181818182,
+      "grad_norm": 2.036296869645202,
+      "learning_rate": 6.577396817739584e-07,
+      "loss": 0.0261,
+      "step": 10634
+    },
+    {
+      "epoch": 2.4170454545454545,
+      "grad_norm": 1.9424782041822208,
+      "learning_rate": 6.576505541853325e-07,
+      "loss": 0.0304,
+      "step": 10635
+    },
+    {
+      "epoch": 2.4172727272727275,
+      "grad_norm": 2.521750558288975,
+      "learning_rate": 6.575614259309048e-07,
+      "loss": 0.0193,
+      "step": 10636
+    },
+    {
+      "epoch": 2.4175,
+      "grad_norm": 2.2381011201595102,
+      "learning_rate": 6.574722970124927e-07,
+      "loss": 0.0326,
+      "step": 10637
+    },
+    {
+      "epoch": 2.417727272727273,
+      "grad_norm": 2.0551986038581953,
+      "learning_rate": 6.57383167431914e-07,
+      "loss": 0.0198,
+      "step": 10638
+    },
+    {
+      "epoch": 2.4179545454545455,
+      "grad_norm": 1.8167969223816964,
+      "learning_rate": 6.572940371909859e-07,
+      "loss": 0.022,
+      "step": 10639
+    },
+    {
+      "epoch": 2.418181818181818,
+      "grad_norm": 1.6766627715454168,
+      "learning_rate": 6.57204906291526e-07,
+      "loss": 0.0216,
+      "step": 10640
+    },
+    {
+      "epoch": 2.418409090909091,
+      "grad_norm": 1.7942548041077824,
+      "learning_rate": 6.571157747353519e-07,
+      "loss": 0.0253,
+      "step": 10641
+    },
+    {
+      "epoch": 2.4186363636363635,
+      "grad_norm": 1.9614040732618432,
+      "learning_rate": 6.570266425242813e-07,
+      "loss": 0.0341,
+      "step": 10642
+    },
+    {
+      "epoch": 2.4188636363636364,
+      "grad_norm": 1.8041861188019934,
+      "learning_rate": 6.569375096601314e-07,
+      "loss": 0.0175,
+      "step": 10643
+    },
+    {
+      "epoch": 2.419090909090909,
+      "grad_norm": 1.9977853950388342,
+      "learning_rate": 6.568483761447201e-07,
+      "loss": 0.0375,
+      "step": 10644
+    },
+    {
+      "epoch": 2.419318181818182,
+      "grad_norm": 3.1592408805822942,
+      "learning_rate": 6.567592419798647e-07,
+      "loss": 0.1464,
+      "step": 10645
+    },
+    {
+      "epoch": 2.4195454545454544,
+      "grad_norm": 1.617970616985487,
+      "learning_rate": 6.56670107167383e-07,
+      "loss": 0.0209,
+      "step": 10646
+    },
+    {
+      "epoch": 2.4197727272727274,
+      "grad_norm": 2.14193815109533,
+      "learning_rate": 6.565809717090927e-07,
+      "loss": 0.0667,
+      "step": 10647
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.6188897358230292,
+      "learning_rate": 6.56491835606811e-07,
+      "loss": 0.0151,
+      "step": 10648
+    },
+    {
+      "epoch": 2.420227272727273,
+      "grad_norm": 1.150457872454334,
+      "learning_rate": 6.56402698862356e-07,
+      "loss": 0.0098,
+      "step": 10649
+    },
+    {
+      "epoch": 2.4204545454545454,
+      "grad_norm": 2.481885377928937,
+      "learning_rate": 6.563135614775452e-07,
+      "loss": 0.0246,
+      "step": 10650
+    },
+    {
+      "epoch": 2.4206818181818184,
+      "grad_norm": 2.060021367830767,
+      "learning_rate": 6.562244234541964e-07,
+      "loss": 0.0276,
+      "step": 10651
+    },
+    {
+      "epoch": 2.420909090909091,
+      "grad_norm": 1.3464149836680712,
+      "learning_rate": 6.561352847941269e-07,
+      "loss": 0.0109,
+      "step": 10652
+    },
+    {
+      "epoch": 2.4211363636363634,
+      "grad_norm": 1.582860118965766,
+      "learning_rate": 6.560461454991548e-07,
+      "loss": 0.0276,
+      "step": 10653
+    },
+    {
+      "epoch": 2.4213636363636364,
+      "grad_norm": 2.080684191371454,
+      "learning_rate": 6.559570055710975e-07,
+      "loss": 0.0218,
+      "step": 10654
+    },
+    {
+      "epoch": 2.421590909090909,
+      "grad_norm": 2.257025519028812,
+      "learning_rate": 6.558678650117726e-07,
+      "loss": 0.0363,
+      "step": 10655
+    },
+    {
+      "epoch": 2.421818181818182,
+      "grad_norm": 2.523426455104573,
+      "learning_rate": 6.557787238229984e-07,
+      "loss": 0.0571,
+      "step": 10656
+    },
+    {
+      "epoch": 2.4220454545454544,
+      "grad_norm": 3.187077277583304,
+      "learning_rate": 6.556895820065923e-07,
+      "loss": 0.0507,
+      "step": 10657
+    },
+    {
+      "epoch": 2.4222727272727274,
+      "grad_norm": 1.8826139333605771,
+      "learning_rate": 6.556004395643721e-07,
+      "loss": 0.0187,
+      "step": 10658
+    },
+    {
+      "epoch": 2.4225,
+      "grad_norm": 1.6111210399125493,
+      "learning_rate": 6.555112964981554e-07,
+      "loss": 0.0139,
+      "step": 10659
+    },
+    {
+      "epoch": 2.422727272727273,
+      "grad_norm": 1.9994373701504364,
+      "learning_rate": 6.554221528097604e-07,
+      "loss": 0.0226,
+      "step": 10660
+    },
+    {
+      "epoch": 2.4229545454545454,
+      "grad_norm": 4.1816572835278505,
+      "learning_rate": 6.553330085010044e-07,
+      "loss": 0.0739,
+      "step": 10661
+    },
+    {
+      "epoch": 2.4231818181818183,
+      "grad_norm": 1.2549443249385248,
+      "learning_rate": 6.552438635737055e-07,
+      "loss": 0.0242,
+      "step": 10662
+    },
+    {
+      "epoch": 2.423409090909091,
+      "grad_norm": 1.4025156759860382,
+      "learning_rate": 6.551547180296816e-07,
+      "loss": 0.0144,
+      "step": 10663
+    },
+    {
+      "epoch": 2.423636363636364,
+      "grad_norm": 1.6164676770826543,
+      "learning_rate": 6.550655718707502e-07,
+      "loss": 0.0204,
+      "step": 10664
+    },
+    {
+      "epoch": 2.4238636363636363,
+      "grad_norm": 1.0196918835894104,
+      "learning_rate": 6.549764250987293e-07,
+      "loss": 0.0093,
+      "step": 10665
+    },
+    {
+      "epoch": 2.424090909090909,
+      "grad_norm": 2.4327131204246046,
+      "learning_rate": 6.548872777154369e-07,
+      "loss": 0.0255,
+      "step": 10666
+    },
+    {
+      "epoch": 2.424318181818182,
+      "grad_norm": 1.8531575343129996,
+      "learning_rate": 6.547981297226909e-07,
+      "loss": 0.0496,
+      "step": 10667
+    },
+    {
+      "epoch": 2.424545454545455,
+      "grad_norm": 2.793986737344444,
+      "learning_rate": 6.547089811223088e-07,
+      "loss": 0.0372,
+      "step": 10668
+    },
+    {
+      "epoch": 2.4247727272727273,
+      "grad_norm": 1.2637736054772986,
+      "learning_rate": 6.546198319161089e-07,
+      "loss": 0.0103,
+      "step": 10669
+    },
+    {
+      "epoch": 2.425,
+      "grad_norm": 1.8029564087685677,
+      "learning_rate": 6.545306821059088e-07,
+      "loss": 0.0277,
+      "step": 10670
+    },
+    {
+      "epoch": 2.425227272727273,
+      "grad_norm": 3.2841325247376614,
+      "learning_rate": 6.544415316935267e-07,
+      "loss": 0.0473,
+      "step": 10671
+    },
+    {
+      "epoch": 2.4254545454545453,
+      "grad_norm": 1.4427211846510524,
+      "learning_rate": 6.543523806807804e-07,
+      "loss": 0.0128,
+      "step": 10672
+    },
+    {
+      "epoch": 2.4256818181818183,
+      "grad_norm": 2.9533502870728214,
+      "learning_rate": 6.54263229069488e-07,
+      "loss": 0.0196,
+      "step": 10673
+    },
+    {
+      "epoch": 2.425909090909091,
+      "grad_norm": 3.163268717629234,
+      "learning_rate": 6.541740768614669e-07,
+      "loss": 0.0326,
+      "step": 10674
+    },
+    {
+      "epoch": 2.4261363636363638,
+      "grad_norm": 2.119724509997459,
+      "learning_rate": 6.540849240585359e-07,
+      "loss": 0.036,
+      "step": 10675
+    },
+    {
+      "epoch": 2.4263636363636363,
+      "grad_norm": 2.120810260511264,
+      "learning_rate": 6.539957706625123e-07,
+      "loss": 0.0452,
+      "step": 10676
+    },
+    {
+      "epoch": 2.4265909090909092,
+      "grad_norm": 1.2400729010367024,
+      "learning_rate": 6.539066166752144e-07,
+      "loss": 0.0201,
+      "step": 10677
+    },
+    {
+      "epoch": 2.4268181818181818,
+      "grad_norm": 1.2274944905713743,
+      "learning_rate": 6.538174620984603e-07,
+      "loss": 0.0127,
+      "step": 10678
+    },
+    {
+      "epoch": 2.4270454545454547,
+      "grad_norm": 1.641331491285657,
+      "learning_rate": 6.537283069340676e-07,
+      "loss": 0.0266,
+      "step": 10679
+    },
+    {
+      "epoch": 2.4272727272727272,
+      "grad_norm": 3.097010853737966,
+      "learning_rate": 6.536391511838549e-07,
+      "loss": 0.0313,
+      "step": 10680
+    },
+    {
+      "epoch": 2.4275,
+      "grad_norm": 3.905184859276694,
+      "learning_rate": 6.535499948496396e-07,
+      "loss": 0.0523,
+      "step": 10681
+    },
+    {
+      "epoch": 2.4277272727272727,
+      "grad_norm": 2.0791078583972196,
+      "learning_rate": 6.534608379332405e-07,
+      "loss": 0.0232,
+      "step": 10682
+    },
+    {
+      "epoch": 2.4279545454545453,
+      "grad_norm": 1.043784583538037,
+      "learning_rate": 6.53371680436475e-07,
+      "loss": 0.026,
+      "step": 10683
+    },
+    {
+      "epoch": 2.428181818181818,
+      "grad_norm": 2.8821721549378574,
+      "learning_rate": 6.532825223611615e-07,
+      "loss": 0.0318,
+      "step": 10684
+    },
+    {
+      "epoch": 2.4284090909090907,
+      "grad_norm": 1.6682614411628751,
+      "learning_rate": 6.53193363709118e-07,
+      "loss": 0.0383,
+      "step": 10685
+    },
+    {
+      "epoch": 2.4286363636363637,
+      "grad_norm": 1.7611473055847118,
+      "learning_rate": 6.531042044821625e-07,
+      "loss": 0.0222,
+      "step": 10686
+    },
+    {
+      "epoch": 2.4288636363636362,
+      "grad_norm": 2.587361442493617,
+      "learning_rate": 6.530150446821135e-07,
+      "loss": 0.0717,
+      "step": 10687
+    },
+    {
+      "epoch": 2.429090909090909,
+      "grad_norm": 3.2645044176356204,
+      "learning_rate": 6.529258843107887e-07,
+      "loss": 0.0189,
+      "step": 10688
+    },
+    {
+      "epoch": 2.4293181818181817,
+      "grad_norm": 3.712282187712964,
+      "learning_rate": 6.528367233700066e-07,
+      "loss": 0.0202,
+      "step": 10689
+    },
+    {
+      "epoch": 2.4295454545454547,
+      "grad_norm": 1.7218999517838218,
+      "learning_rate": 6.527475618615848e-07,
+      "loss": 0.0133,
+      "step": 10690
+    },
+    {
+      "epoch": 2.429772727272727,
+      "grad_norm": 1.8658872303348375,
+      "learning_rate": 6.526583997873422e-07,
+      "loss": 0.031,
+      "step": 10691
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.3704084465687822,
+      "learning_rate": 6.525692371490963e-07,
+      "loss": 0.0234,
+      "step": 10692
+    },
+    {
+      "epoch": 2.4302272727272727,
+      "grad_norm": 1.7041559968336677,
+      "learning_rate": 6.524800739486656e-07,
+      "loss": 0.0133,
+      "step": 10693
+    },
+    {
+      "epoch": 2.4304545454545456,
+      "grad_norm": 2.1026420212039874,
+      "learning_rate": 6.523909101878683e-07,
+      "loss": 0.0273,
+      "step": 10694
+    },
+    {
+      "epoch": 2.430681818181818,
+      "grad_norm": 1.1887994215927422,
+      "learning_rate": 6.523017458685226e-07,
+      "loss": 0.0103,
+      "step": 10695
+    },
+    {
+      "epoch": 2.4309090909090907,
+      "grad_norm": 1.7044236900625465,
+      "learning_rate": 6.522125809924465e-07,
+      "loss": 0.0222,
+      "step": 10696
+    },
+    {
+      "epoch": 2.4311363636363637,
+      "grad_norm": 1.476554137905608,
+      "learning_rate": 6.521234155614586e-07,
+      "loss": 0.0178,
+      "step": 10697
+    },
+    {
+      "epoch": 2.431363636363636,
+      "grad_norm": 2.0787220773764705,
+      "learning_rate": 6.520342495773768e-07,
+      "loss": 0.0333,
+      "step": 10698
+    },
+    {
+      "epoch": 2.431590909090909,
+      "grad_norm": 1.8551273729109365,
+      "learning_rate": 6.519450830420197e-07,
+      "loss": 0.0208,
+      "step": 10699
+    },
+    {
+      "epoch": 2.4318181818181817,
+      "grad_norm": 1.946179012398003,
+      "learning_rate": 6.518559159572052e-07,
+      "loss": 0.0492,
+      "step": 10700
+    },
+    {
+      "epoch": 2.4320454545454546,
+      "grad_norm": 2.8999213196664555,
+      "learning_rate": 6.517667483247517e-07,
+      "loss": 0.037,
+      "step": 10701
+    },
+    {
+      "epoch": 2.432272727272727,
+      "grad_norm": 1.4329661336033352,
+      "learning_rate": 6.516775801464774e-07,
+      "loss": 0.0112,
+      "step": 10702
+    },
+    {
+      "epoch": 2.4325,
+      "grad_norm": 3.0332146872100028,
+      "learning_rate": 6.515884114242008e-07,
+      "loss": 0.0358,
+      "step": 10703
+    },
+    {
+      "epoch": 2.4327272727272726,
+      "grad_norm": 0.8421969463440356,
+      "learning_rate": 6.514992421597402e-07,
+      "loss": 0.0068,
+      "step": 10704
+    },
+    {
+      "epoch": 2.4329545454545456,
+      "grad_norm": 2.408887571090738,
+      "learning_rate": 6.514100723549137e-07,
+      "loss": 0.0273,
+      "step": 10705
+    },
+    {
+      "epoch": 2.433181818181818,
+      "grad_norm": 2.000249870312162,
+      "learning_rate": 6.513209020115399e-07,
+      "loss": 0.041,
+      "step": 10706
+    },
+    {
+      "epoch": 2.433409090909091,
+      "grad_norm": 1.8303961589457416,
+      "learning_rate": 6.512317311314368e-07,
+      "loss": 0.0222,
+      "step": 10707
+    },
+    {
+      "epoch": 2.4336363636363636,
+      "grad_norm": 2.3334045784008475,
+      "learning_rate": 6.511425597164229e-07,
+      "loss": 0.0445,
+      "step": 10708
+    },
+    {
+      "epoch": 2.433863636363636,
+      "grad_norm": 2.3237932995223387,
+      "learning_rate": 6.510533877683167e-07,
+      "loss": 0.0323,
+      "step": 10709
+    },
+    {
+      "epoch": 2.434090909090909,
+      "grad_norm": 4.003027001263201,
+      "learning_rate": 6.509642152889366e-07,
+      "loss": 0.0564,
+      "step": 10710
+    },
+    {
+      "epoch": 2.434318181818182,
+      "grad_norm": 1.3010353490032103,
+      "learning_rate": 6.508750422801006e-07,
+      "loss": 0.0291,
+      "step": 10711
+    },
+    {
+      "epoch": 2.4345454545454546,
+      "grad_norm": 1.2276952593344537,
+      "learning_rate": 6.507858687436274e-07,
+      "loss": 0.0125,
+      "step": 10712
+    },
+    {
+      "epoch": 2.434772727272727,
+      "grad_norm": 2.6031668566636834,
+      "learning_rate": 6.506966946813357e-07,
+      "loss": 0.0427,
+      "step": 10713
+    },
+    {
+      "epoch": 2.435,
+      "grad_norm": 1.7101356154259095,
+      "learning_rate": 6.506075200950432e-07,
+      "loss": 0.0272,
+      "step": 10714
+    },
+    {
+      "epoch": 2.4352272727272726,
+      "grad_norm": 0.9191620509119441,
+      "learning_rate": 6.505183449865689e-07,
+      "loss": 0.0147,
+      "step": 10715
+    },
+    {
+      "epoch": 2.4354545454545455,
+      "grad_norm": 2.4953769492244846,
+      "learning_rate": 6.504291693577309e-07,
+      "loss": 0.0375,
+      "step": 10716
+    },
+    {
+      "epoch": 2.435681818181818,
+      "grad_norm": 2.762021079808424,
+      "learning_rate": 6.503399932103477e-07,
+      "loss": 0.021,
+      "step": 10717
+    },
+    {
+      "epoch": 2.435909090909091,
+      "grad_norm": 2.756080890767236,
+      "learning_rate": 6.502508165462379e-07,
+      "loss": 0.0618,
+      "step": 10718
+    },
+    {
+      "epoch": 2.4361363636363635,
+      "grad_norm": 1.1791926909078547,
+      "learning_rate": 6.5016163936722e-07,
+      "loss": 0.0124,
+      "step": 10719
+    },
+    {
+      "epoch": 2.4363636363636365,
+      "grad_norm": 2.4279846581834135,
+      "learning_rate": 6.500724616751122e-07,
+      "loss": 0.0424,
+      "step": 10720
+    },
+    {
+      "epoch": 2.436590909090909,
+      "grad_norm": 2.457655289583442,
+      "learning_rate": 6.499832834717335e-07,
+      "loss": 0.0249,
+      "step": 10721
+    },
+    {
+      "epoch": 2.436818181818182,
+      "grad_norm": 1.6639357071222696,
+      "learning_rate": 6.498941047589017e-07,
+      "loss": 0.017,
+      "step": 10722
+    },
+    {
+      "epoch": 2.4370454545454545,
+      "grad_norm": 2.9674888532752393,
+      "learning_rate": 6.498049255384357e-07,
+      "loss": 0.0546,
+      "step": 10723
+    },
+    {
+      "epoch": 2.4372727272727275,
+      "grad_norm": 2.921079311001886,
+      "learning_rate": 6.497157458121542e-07,
+      "loss": 0.0851,
+      "step": 10724
+    },
+    {
+      "epoch": 2.4375,
+      "grad_norm": 3.290527728227217,
+      "learning_rate": 6.496265655818752e-07,
+      "loss": 0.0208,
+      "step": 10725
+    },
+    {
+      "epoch": 2.4377272727272725,
+      "grad_norm": 1.583970323771248,
+      "learning_rate": 6.495373848494179e-07,
+      "loss": 0.0173,
+      "step": 10726
+    },
+    {
+      "epoch": 2.4379545454545455,
+      "grad_norm": 3.609531101910938,
+      "learning_rate": 6.494482036166004e-07,
+      "loss": 0.0752,
+      "step": 10727
+    },
+    {
+      "epoch": 2.438181818181818,
+      "grad_norm": 2.8532504382731863,
+      "learning_rate": 6.493590218852415e-07,
+      "loss": 0.0315,
+      "step": 10728
+    },
+    {
+      "epoch": 2.438409090909091,
+      "grad_norm": 2.690407789683469,
+      "learning_rate": 6.492698396571595e-07,
+      "loss": 0.0288,
+      "step": 10729
+    },
+    {
+      "epoch": 2.4386363636363635,
+      "grad_norm": 1.634791412724901,
+      "learning_rate": 6.491806569341733e-07,
+      "loss": 0.0168,
+      "step": 10730
+    },
+    {
+      "epoch": 2.4388636363636365,
+      "grad_norm": 2.8478371055548304,
+      "learning_rate": 6.49091473718101e-07,
+      "loss": 0.012,
+      "step": 10731
+    },
+    {
+      "epoch": 2.439090909090909,
+      "grad_norm": 3.5171565143662065,
+      "learning_rate": 6.490022900107617e-07,
+      "loss": 0.0574,
+      "step": 10732
+    },
+    {
+      "epoch": 2.439318181818182,
+      "grad_norm": 2.636225078420494,
+      "learning_rate": 6.489131058139737e-07,
+      "loss": 0.0153,
+      "step": 10733
+    },
+    {
+      "epoch": 2.4395454545454545,
+      "grad_norm": 2.051892365352072,
+      "learning_rate": 6.488239211295559e-07,
+      "loss": 0.0315,
+      "step": 10734
+    },
+    {
+      "epoch": 2.4397727272727274,
+      "grad_norm": 2.2579104861758474,
+      "learning_rate": 6.487347359593269e-07,
+      "loss": 0.027,
+      "step": 10735
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.0448243664316412,
+      "learning_rate": 6.48645550305105e-07,
+      "loss": 0.0236,
+      "step": 10736
+    },
+    {
+      "epoch": 2.440227272727273,
+      "grad_norm": 1.4158688638580392,
+      "learning_rate": 6.485563641687093e-07,
+      "loss": 0.0105,
+      "step": 10737
+    },
+    {
+      "epoch": 2.4404545454545454,
+      "grad_norm": 2.402914129231504,
+      "learning_rate": 6.48467177551958e-07,
+      "loss": 0.0273,
+      "step": 10738
+    },
+    {
+      "epoch": 2.440681818181818,
+      "grad_norm": 1.9009570266427531,
+      "learning_rate": 6.483779904566701e-07,
+      "loss": 0.0175,
+      "step": 10739
+    },
+    {
+      "epoch": 2.440909090909091,
+      "grad_norm": 2.419930148748,
+      "learning_rate": 6.482888028846641e-07,
+      "loss": 0.0402,
+      "step": 10740
+    },
+    {
+      "epoch": 2.441136363636364,
+      "grad_norm": 1.6617009624647823,
+      "learning_rate": 6.48199614837759e-07,
+      "loss": 0.0155,
+      "step": 10741
+    },
+    {
+      "epoch": 2.4413636363636364,
+      "grad_norm": 2.3385113776760647,
+      "learning_rate": 6.481104263177732e-07,
+      "loss": 0.0319,
+      "step": 10742
+    },
+    {
+      "epoch": 2.441590909090909,
+      "grad_norm": 2.9994989197220336,
+      "learning_rate": 6.480212373265254e-07,
+      "loss": 0.0261,
+      "step": 10743
+    },
+    {
+      "epoch": 2.441818181818182,
+      "grad_norm": 2.248451921746013,
+      "learning_rate": 6.479320478658345e-07,
+      "loss": 0.0226,
+      "step": 10744
+    },
+    {
+      "epoch": 2.4420454545454544,
+      "grad_norm": 3.0712995665326366,
+      "learning_rate": 6.47842857937519e-07,
+      "loss": 0.0513,
+      "step": 10745
+    },
+    {
+      "epoch": 2.4422727272727274,
+      "grad_norm": 1.793073626578348,
+      "learning_rate": 6.477536675433979e-07,
+      "loss": 0.0431,
+      "step": 10746
+    },
+    {
+      "epoch": 2.4425,
+      "grad_norm": 2.3620399148441953,
+      "learning_rate": 6.476644766852897e-07,
+      "loss": 0.0498,
+      "step": 10747
+    },
+    {
+      "epoch": 2.442727272727273,
+      "grad_norm": 1.2436484385469355,
+      "learning_rate": 6.475752853650135e-07,
+      "loss": 0.0126,
+      "step": 10748
+    },
+    {
+      "epoch": 2.4429545454545454,
+      "grad_norm": 2.4256692326882665,
+      "learning_rate": 6.474860935843877e-07,
+      "loss": 0.0392,
+      "step": 10749
+    },
+    {
+      "epoch": 2.4431818181818183,
+      "grad_norm": 1.8141145538940902,
+      "learning_rate": 6.473969013452313e-07,
+      "loss": 0.0259,
+      "step": 10750
+    },
+    {
+      "epoch": 2.443409090909091,
+      "grad_norm": 2.826513026154679,
+      "learning_rate": 6.473077086493629e-07,
+      "loss": 0.0326,
+      "step": 10751
+    },
+    {
+      "epoch": 2.443636363636364,
+      "grad_norm": 1.9889517303908322,
+      "learning_rate": 6.472185154986015e-07,
+      "loss": 0.0262,
+      "step": 10752
+    },
+    {
+      "epoch": 2.4438636363636363,
+      "grad_norm": 1.8936573442044073,
+      "learning_rate": 6.471293218947657e-07,
+      "loss": 0.0246,
+      "step": 10753
+    },
+    {
+      "epoch": 2.4440909090909093,
+      "grad_norm": 1.4276958543822618,
+      "learning_rate": 6.470401278396744e-07,
+      "loss": 0.0163,
+      "step": 10754
+    },
+    {
+      "epoch": 2.444318181818182,
+      "grad_norm": 1.6187048680346157,
+      "learning_rate": 6.469509333351465e-07,
+      "loss": 0.023,
+      "step": 10755
+    },
+    {
+      "epoch": 2.4445454545454544,
+      "grad_norm": 1.1256482896447357,
+      "learning_rate": 6.468617383830007e-07,
+      "loss": 0.0093,
+      "step": 10756
+    },
+    {
+      "epoch": 2.4447727272727273,
+      "grad_norm": 2.0806563754663308,
+      "learning_rate": 6.467725429850561e-07,
+      "loss": 0.0293,
+      "step": 10757
+    },
+    {
+      "epoch": 2.445,
+      "grad_norm": 1.6173440702006592,
+      "learning_rate": 6.466833471431313e-07,
+      "loss": 0.0149,
+      "step": 10758
+    },
+    {
+      "epoch": 2.445227272727273,
+      "grad_norm": 1.6216384460785982,
+      "learning_rate": 6.465941508590454e-07,
+      "loss": 0.0235,
+      "step": 10759
+    },
+    {
+      "epoch": 2.4454545454545453,
+      "grad_norm": 1.7965821258098478,
+      "learning_rate": 6.465049541346168e-07,
+      "loss": 0.0269,
+      "step": 10760
+    },
+    {
+      "epoch": 2.4456818181818183,
+      "grad_norm": 1.0858417683269765,
+      "learning_rate": 6.464157569716648e-07,
+      "loss": 0.0113,
+      "step": 10761
+    },
+    {
+      "epoch": 2.445909090909091,
+      "grad_norm": 5.22217497820639,
+      "learning_rate": 6.46326559372008e-07,
+      "loss": 0.0422,
+      "step": 10762
+    },
+    {
+      "epoch": 2.4461363636363638,
+      "grad_norm": 1.4139609003714353,
+      "learning_rate": 6.462373613374656e-07,
+      "loss": 0.0141,
+      "step": 10763
+    },
+    {
+      "epoch": 2.4463636363636363,
+      "grad_norm": 1.7029096600741096,
+      "learning_rate": 6.461481628698563e-07,
+      "loss": 0.0156,
+      "step": 10764
+    },
+    {
+      "epoch": 2.4465909090909093,
+      "grad_norm": 1.8101573144742007,
+      "learning_rate": 6.460589639709991e-07,
+      "loss": 0.022,
+      "step": 10765
+    },
+    {
+      "epoch": 2.446818181818182,
+      "grad_norm": 1.7751317141610348,
+      "learning_rate": 6.459697646427128e-07,
+      "loss": 0.0118,
+      "step": 10766
+    },
+    {
+      "epoch": 2.4470454545454547,
+      "grad_norm": 3.4989415083172997,
+      "learning_rate": 6.458805648868164e-07,
+      "loss": 0.0458,
+      "step": 10767
+    },
+    {
+      "epoch": 2.4472727272727273,
+      "grad_norm": 1.3121955533723961,
+      "learning_rate": 6.45791364705129e-07,
+      "loss": 0.017,
+      "step": 10768
+    },
+    {
+      "epoch": 2.4475,
+      "grad_norm": 1.9922502830593254,
+      "learning_rate": 6.457021640994693e-07,
+      "loss": 0.0378,
+      "step": 10769
+    },
+    {
+      "epoch": 2.4477272727272728,
+      "grad_norm": 1.283456293782596,
+      "learning_rate": 6.456129630716564e-07,
+      "loss": 0.0123,
+      "step": 10770
+    },
+    {
+      "epoch": 2.4479545454545453,
+      "grad_norm": 2.5673015719514063,
+      "learning_rate": 6.455237616235091e-07,
+      "loss": 0.0234,
+      "step": 10771
+    },
+    {
+      "epoch": 2.4481818181818182,
+      "grad_norm": 2.0377059456226787,
+      "learning_rate": 6.454345597568468e-07,
+      "loss": 0.0428,
+      "step": 10772
+    },
+    {
+      "epoch": 2.4484090909090908,
+      "grad_norm": 1.997379309119622,
+      "learning_rate": 6.453453574734879e-07,
+      "loss": 0.0317,
+      "step": 10773
+    },
+    {
+      "epoch": 2.4486363636363637,
+      "grad_norm": 4.618365777583543,
+      "learning_rate": 6.452561547752518e-07,
+      "loss": 0.0629,
+      "step": 10774
+    },
+    {
+      "epoch": 2.4488636363636362,
+      "grad_norm": 1.9413640765903704,
+      "learning_rate": 6.451669516639573e-07,
+      "loss": 0.0167,
+      "step": 10775
+    },
+    {
+      "epoch": 2.449090909090909,
+      "grad_norm": 2.072969026114925,
+      "learning_rate": 6.450777481414236e-07,
+      "loss": 0.0726,
+      "step": 10776
+    },
+    {
+      "epoch": 2.4493181818181817,
+      "grad_norm": 1.4785839499150457,
+      "learning_rate": 6.449885442094693e-07,
+      "loss": 0.0196,
+      "step": 10777
+    },
+    {
+      "epoch": 2.4495454545454547,
+      "grad_norm": 2.5270492745289466,
+      "learning_rate": 6.448993398699138e-07,
+      "loss": 0.0296,
+      "step": 10778
+    },
+    {
+      "epoch": 2.449772727272727,
+      "grad_norm": 2.2461940393483344,
+      "learning_rate": 6.448101351245762e-07,
+      "loss": 0.0205,
+      "step": 10779
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.878762628885306,
+      "learning_rate": 6.447209299752754e-07,
+      "loss": 0.0274,
+      "step": 10780
+    },
+    {
+      "epoch": 2.4502272727272727,
+      "grad_norm": 0.9945938818601292,
+      "learning_rate": 6.446317244238302e-07,
+      "loss": 0.008,
+      "step": 10781
+    },
+    {
+      "epoch": 2.450454545454545,
+      "grad_norm": 3.9951342257172144,
+      "learning_rate": 6.4454251847206e-07,
+      "loss": 0.0351,
+      "step": 10782
+    },
+    {
+      "epoch": 2.450681818181818,
+      "grad_norm": 3.186322044248222,
+      "learning_rate": 6.44453312121784e-07,
+      "loss": 0.0296,
+      "step": 10783
+    },
+    {
+      "epoch": 2.450909090909091,
+      "grad_norm": 3.178400692367429,
+      "learning_rate": 6.443641053748208e-07,
+      "loss": 0.0475,
+      "step": 10784
+    },
+    {
+      "epoch": 2.4511363636363637,
+      "grad_norm": 2.171653403711302,
+      "learning_rate": 6.442748982329897e-07,
+      "loss": 0.0211,
+      "step": 10785
+    },
+    {
+      "epoch": 2.451363636363636,
+      "grad_norm": 2.1219362280655605,
+      "learning_rate": 6.441856906981096e-07,
+      "loss": 0.0194,
+      "step": 10786
+    },
+    {
+      "epoch": 2.451590909090909,
+      "grad_norm": 2.468509713309802,
+      "learning_rate": 6.440964827720002e-07,
+      "loss": 0.0302,
+      "step": 10787
+    },
+    {
+      "epoch": 2.4518181818181817,
+      "grad_norm": 2.5176585768929534,
+      "learning_rate": 6.4400727445648e-07,
+      "loss": 0.0218,
+      "step": 10788
+    },
+    {
+      "epoch": 2.4520454545454546,
+      "grad_norm": 1.320433287152039,
+      "learning_rate": 6.439180657533682e-07,
+      "loss": 0.0245,
+      "step": 10789
+    },
+    {
+      "epoch": 2.452272727272727,
+      "grad_norm": 2.28643012429246,
+      "learning_rate": 6.438288566644843e-07,
+      "loss": 0.0192,
+      "step": 10790
+    },
+    {
+      "epoch": 2.4525,
+      "grad_norm": 2.039970585681391,
+      "learning_rate": 6.437396471916469e-07,
+      "loss": 0.0229,
+      "step": 10791
+    },
+    {
+      "epoch": 2.4527272727272726,
+      "grad_norm": 1.6729059990248967,
+      "learning_rate": 6.436504373366756e-07,
+      "loss": 0.0099,
+      "step": 10792
+    },
+    {
+      "epoch": 2.4529545454545456,
+      "grad_norm": 1.3072814050112362,
+      "learning_rate": 6.435612271013892e-07,
+      "loss": 0.0162,
+      "step": 10793
+    },
+    {
+      "epoch": 2.453181818181818,
+      "grad_norm": 1.4723809429693377,
+      "learning_rate": 6.434720164876072e-07,
+      "loss": 0.0359,
+      "step": 10794
+    },
+    {
+      "epoch": 2.453409090909091,
+      "grad_norm": 1.2786047237139888,
+      "learning_rate": 6.433828054971484e-07,
+      "loss": 0.021,
+      "step": 10795
+    },
+    {
+      "epoch": 2.4536363636363636,
+      "grad_norm": 2.352803057387099,
+      "learning_rate": 6.432935941318322e-07,
+      "loss": 0.0427,
+      "step": 10796
+    },
+    {
+      "epoch": 2.4538636363636366,
+      "grad_norm": 2.3434082314797076,
+      "learning_rate": 6.432043823934777e-07,
+      "loss": 0.0134,
+      "step": 10797
+    },
+    {
+      "epoch": 2.454090909090909,
+      "grad_norm": 1.1396684921191633,
+      "learning_rate": 6.431151702839041e-07,
+      "loss": 0.0124,
+      "step": 10798
+    },
+    {
+      "epoch": 2.4543181818181816,
+      "grad_norm": 2.3853717955035876,
+      "learning_rate": 6.430259578049304e-07,
+      "loss": 0.0242,
+      "step": 10799
+    },
+    {
+      "epoch": 2.4545454545454546,
+      "grad_norm": 1.9005404260543104,
+      "learning_rate": 6.429367449583761e-07,
+      "loss": 0.0253,
+      "step": 10800
+    },
+    {
+      "epoch": 2.454772727272727,
+      "grad_norm": 1.3817212007594866,
+      "learning_rate": 6.428475317460601e-07,
+      "loss": 0.0239,
+      "step": 10801
+    },
+    {
+      "epoch": 2.455,
+      "grad_norm": 1.637098496148658,
+      "learning_rate": 6.427583181698021e-07,
+      "loss": 0.0315,
+      "step": 10802
+    },
+    {
+      "epoch": 2.4552272727272726,
+      "grad_norm": 1.920981598808276,
+      "learning_rate": 6.426691042314209e-07,
+      "loss": 0.023,
+      "step": 10803
+    },
+    {
+      "epoch": 2.4554545454545456,
+      "grad_norm": 1.9313496084130797,
+      "learning_rate": 6.425798899327356e-07,
+      "loss": 0.009,
+      "step": 10804
+    },
+    {
+      "epoch": 2.455681818181818,
+      "grad_norm": 1.7004793948567671,
+      "learning_rate": 6.424906752755661e-07,
+      "loss": 0.0235,
+      "step": 10805
+    },
+    {
+      "epoch": 2.455909090909091,
+      "grad_norm": 1.7036792493336275,
+      "learning_rate": 6.424014602617309e-07,
+      "loss": 0.0295,
+      "step": 10806
+    },
+    {
+      "epoch": 2.4561363636363636,
+      "grad_norm": 1.678239519764456,
+      "learning_rate": 6.423122448930496e-07,
+      "loss": 0.0173,
+      "step": 10807
+    },
+    {
+      "epoch": 2.4563636363636365,
+      "grad_norm": 1.6655793609264735,
+      "learning_rate": 6.422230291713416e-07,
+      "loss": 0.0119,
+      "step": 10808
+    },
+    {
+      "epoch": 2.456590909090909,
+      "grad_norm": 2.06833774462078,
+      "learning_rate": 6.421338130984259e-07,
+      "loss": 0.0083,
+      "step": 10809
+    },
+    {
+      "epoch": 2.456818181818182,
+      "grad_norm": 1.8466209610554303,
+      "learning_rate": 6.420445966761219e-07,
+      "loss": 0.0088,
+      "step": 10810
+    },
+    {
+      "epoch": 2.4570454545454545,
+      "grad_norm": 1.6299364043107418,
+      "learning_rate": 6.419553799062486e-07,
+      "loss": 0.0255,
+      "step": 10811
+    },
+    {
+      "epoch": 2.457272727272727,
+      "grad_norm": 2.3379158028295053,
+      "learning_rate": 6.418661627906259e-07,
+      "loss": 0.0365,
+      "step": 10812
+    },
+    {
+      "epoch": 2.4575,
+      "grad_norm": 1.029323231120894,
+      "learning_rate": 6.417769453310724e-07,
+      "loss": 0.0086,
+      "step": 10813
+    },
+    {
+      "epoch": 2.4577272727272725,
+      "grad_norm": 1.6032102592957005,
+      "learning_rate": 6.41687727529408e-07,
+      "loss": 0.0388,
+      "step": 10814
+    },
+    {
+      "epoch": 2.4579545454545455,
+      "grad_norm": 1.586215643567616,
+      "learning_rate": 6.415985093874514e-07,
+      "loss": 0.0213,
+      "step": 10815
+    },
+    {
+      "epoch": 2.458181818181818,
+      "grad_norm": 1.4456236436789403,
+      "learning_rate": 6.415092909070224e-07,
+      "loss": 0.0102,
+      "step": 10816
+    },
+    {
+      "epoch": 2.458409090909091,
+      "grad_norm": 1.3444045633126083,
+      "learning_rate": 6.414200720899401e-07,
+      "loss": 0.0215,
+      "step": 10817
+    },
+    {
+      "epoch": 2.4586363636363635,
+      "grad_norm": 1.494097877010675,
+      "learning_rate": 6.41330852938024e-07,
+      "loss": 0.0125,
+      "step": 10818
+    },
+    {
+      "epoch": 2.4588636363636365,
+      "grad_norm": 1.6741639915808122,
+      "learning_rate": 6.412416334530931e-07,
+      "loss": 0.0171,
+      "step": 10819
+    },
+    {
+      "epoch": 2.459090909090909,
+      "grad_norm": 1.4519173392137474,
+      "learning_rate": 6.411524136369674e-07,
+      "loss": 0.0262,
+      "step": 10820
+    },
+    {
+      "epoch": 2.459318181818182,
+      "grad_norm": 1.96431123455104,
+      "learning_rate": 6.410631934914653e-07,
+      "loss": 0.0212,
+      "step": 10821
+    },
+    {
+      "epoch": 2.4595454545454545,
+      "grad_norm": 1.4175747834317671,
+      "learning_rate": 6.409739730184068e-07,
+      "loss": 0.0135,
+      "step": 10822
+    },
+    {
+      "epoch": 2.4597727272727274,
+      "grad_norm": 1.8123080041858566,
+      "learning_rate": 6.408847522196112e-07,
+      "loss": 0.0306,
+      "step": 10823
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.5626394838000255,
+      "learning_rate": 6.407955310968978e-07,
+      "loss": 0.0247,
+      "step": 10824
+    },
+    {
+      "epoch": 2.4602272727272725,
+      "grad_norm": 2.476703905911519,
+      "learning_rate": 6.407063096520861e-07,
+      "loss": 0.0363,
+      "step": 10825
+    },
+    {
+      "epoch": 2.4604545454545454,
+      "grad_norm": 2.0355003624230625,
+      "learning_rate": 6.406170878869951e-07,
+      "loss": 0.0225,
+      "step": 10826
+    },
+    {
+      "epoch": 2.4606818181818184,
+      "grad_norm": 2.0433805350645735,
+      "learning_rate": 6.405278658034445e-07,
+      "loss": 0.0391,
+      "step": 10827
+    },
+    {
+      "epoch": 2.460909090909091,
+      "grad_norm": 1.6527777585196108,
+      "learning_rate": 6.404386434032536e-07,
+      "loss": 0.0148,
+      "step": 10828
+    },
+    {
+      "epoch": 2.4611363636363635,
+      "grad_norm": 4.207461643586708,
+      "learning_rate": 6.40349420688242e-07,
+      "loss": 0.0154,
+      "step": 10829
+    },
+    {
+      "epoch": 2.4613636363636364,
+      "grad_norm": 1.6170823199133135,
+      "learning_rate": 6.402601976602288e-07,
+      "loss": 0.0206,
+      "step": 10830
+    },
+    {
+      "epoch": 2.461590909090909,
+      "grad_norm": 1.6641028463570948,
+      "learning_rate": 6.401709743210335e-07,
+      "loss": 0.0312,
+      "step": 10831
+    },
+    {
+      "epoch": 2.461818181818182,
+      "grad_norm": 1.6849334940354075,
+      "learning_rate": 6.400817506724756e-07,
+      "loss": 0.0129,
+      "step": 10832
+    },
+    {
+      "epoch": 2.4620454545454544,
+      "grad_norm": 1.8396244882466593,
+      "learning_rate": 6.399925267163746e-07,
+      "loss": 0.0258,
+      "step": 10833
+    },
+    {
+      "epoch": 2.4622727272727274,
+      "grad_norm": 0.9294104320855019,
+      "learning_rate": 6.399033024545498e-07,
+      "loss": 0.006,
+      "step": 10834
+    },
+    {
+      "epoch": 2.4625,
+      "grad_norm": 3.6837345659321463,
+      "learning_rate": 6.398140778888205e-07,
+      "loss": 0.0612,
+      "step": 10835
+    },
+    {
+      "epoch": 2.462727272727273,
+      "grad_norm": 2.0713834478658066,
+      "learning_rate": 6.397248530210065e-07,
+      "loss": 0.031,
+      "step": 10836
+    },
+    {
+      "epoch": 2.4629545454545454,
+      "grad_norm": 1.961494229037976,
+      "learning_rate": 6.396356278529269e-07,
+      "loss": 0.0266,
+      "step": 10837
+    },
+    {
+      "epoch": 2.4631818181818184,
+      "grad_norm": 1.3236818880538272,
+      "learning_rate": 6.395464023864013e-07,
+      "loss": 0.0123,
+      "step": 10838
+    },
+    {
+      "epoch": 2.463409090909091,
+      "grad_norm": 2.201602786419833,
+      "learning_rate": 6.394571766232492e-07,
+      "loss": 0.0337,
+      "step": 10839
+    },
+    {
+      "epoch": 2.463636363636364,
+      "grad_norm": 1.6191088217649339,
+      "learning_rate": 6.393679505652903e-07,
+      "loss": 0.0199,
+      "step": 10840
+    },
+    {
+      "epoch": 2.4638636363636364,
+      "grad_norm": 2.3227233683148496,
+      "learning_rate": 6.392787242143437e-07,
+      "loss": 0.0145,
+      "step": 10841
+    },
+    {
+      "epoch": 2.464090909090909,
+      "grad_norm": 1.3506072235154343,
+      "learning_rate": 6.391894975722289e-07,
+      "loss": 0.0116,
+      "step": 10842
+    },
+    {
+      "epoch": 2.464318181818182,
+      "grad_norm": 1.4397876345267224,
+      "learning_rate": 6.391002706407655e-07,
+      "loss": 0.0112,
+      "step": 10843
+    },
+    {
+      "epoch": 2.4645454545454544,
+      "grad_norm": 0.4679026366626368,
+      "learning_rate": 6.39011043421773e-07,
+      "loss": 0.0034,
+      "step": 10844
+    },
+    {
+      "epoch": 2.4647727272727273,
+      "grad_norm": 1.7803033173539207,
+      "learning_rate": 6.38921815917071e-07,
+      "loss": 0.0375,
+      "step": 10845
+    },
+    {
+      "epoch": 2.465,
+      "grad_norm": 2.277535787744011,
+      "learning_rate": 6.388325881284785e-07,
+      "loss": 0.0256,
+      "step": 10846
+    },
+    {
+      "epoch": 2.465227272727273,
+      "grad_norm": 1.5783516318532844,
+      "learning_rate": 6.387433600578158e-07,
+      "loss": 0.0319,
+      "step": 10847
+    },
+    {
+      "epoch": 2.4654545454545453,
+      "grad_norm": 5.3594746889583575,
+      "learning_rate": 6.386541317069019e-07,
+      "loss": 0.0198,
+      "step": 10848
+    },
+    {
+      "epoch": 2.4656818181818183,
+      "grad_norm": 1.3310102747791799,
+      "learning_rate": 6.385649030775563e-07,
+      "loss": 0.0118,
+      "step": 10849
+    },
+    {
+      "epoch": 2.465909090909091,
+      "grad_norm": 2.4998190306803005,
+      "learning_rate": 6.384756741715989e-07,
+      "loss": 0.0355,
+      "step": 10850
+    },
+    {
+      "epoch": 2.466136363636364,
+      "grad_norm": 3.7760218385578006,
+      "learning_rate": 6.383864449908489e-07,
+      "loss": 0.0404,
+      "step": 10851
+    },
+    {
+      "epoch": 2.4663636363636363,
+      "grad_norm": 2.214597469909013,
+      "learning_rate": 6.382972155371259e-07,
+      "loss": 0.0225,
+      "step": 10852
+    },
+    {
+      "epoch": 2.4665909090909093,
+      "grad_norm": 1.1609386482858246,
+      "learning_rate": 6.382079858122494e-07,
+      "loss": 0.0077,
+      "step": 10853
+    },
+    {
+      "epoch": 2.466818181818182,
+      "grad_norm": 5.50827360615021,
+      "learning_rate": 6.38118755818039e-07,
+      "loss": 0.028,
+      "step": 10854
+    },
+    {
+      "epoch": 2.4670454545454543,
+      "grad_norm": 1.544047467243418,
+      "learning_rate": 6.380295255563145e-07,
+      "loss": 0.0109,
+      "step": 10855
+    },
+    {
+      "epoch": 2.4672727272727273,
+      "grad_norm": 1.735979049572167,
+      "learning_rate": 6.379402950288952e-07,
+      "loss": 0.0235,
+      "step": 10856
+    },
+    {
+      "epoch": 2.4675000000000002,
+      "grad_norm": 1.0805566354563003,
+      "learning_rate": 6.378510642376006e-07,
+      "loss": 0.0093,
+      "step": 10857
+    },
+    {
+      "epoch": 2.4677272727272728,
+      "grad_norm": 2.4912428472465926,
+      "learning_rate": 6.377618331842504e-07,
+      "loss": 0.0265,
+      "step": 10858
+    },
+    {
+      "epoch": 2.4679545454545453,
+      "grad_norm": 2.450078365921936,
+      "learning_rate": 6.376726018706642e-07,
+      "loss": 0.0219,
+      "step": 10859
+    },
+    {
+      "epoch": 2.4681818181818183,
+      "grad_norm": 1.7859878318276436,
+      "learning_rate": 6.375833702986616e-07,
+      "loss": 0.0138,
+      "step": 10860
+    },
+    {
+      "epoch": 2.4684090909090908,
+      "grad_norm": 1.6339660689965443,
+      "learning_rate": 6.374941384700618e-07,
+      "loss": 0.0119,
+      "step": 10861
+    },
+    {
+      "epoch": 2.4686363636363637,
+      "grad_norm": 4.330336503920811,
+      "learning_rate": 6.374049063866851e-07,
+      "loss": 0.0749,
+      "step": 10862
+    },
+    {
+      "epoch": 2.4688636363636363,
+      "grad_norm": 3.5895724117731294,
+      "learning_rate": 6.373156740503505e-07,
+      "loss": 0.0137,
+      "step": 10863
+    },
+    {
+      "epoch": 2.4690909090909092,
+      "grad_norm": 2.681230055499921,
+      "learning_rate": 6.37226441462878e-07,
+      "loss": 0.0538,
+      "step": 10864
+    },
+    {
+      "epoch": 2.4693181818181817,
+      "grad_norm": 1.9875011962883677,
+      "learning_rate": 6.371372086260868e-07,
+      "loss": 0.0278,
+      "step": 10865
+    },
+    {
+      "epoch": 2.4695454545454547,
+      "grad_norm": 1.108076683509661,
+      "learning_rate": 6.37047975541797e-07,
+      "loss": 0.0128,
+      "step": 10866
+    },
+    {
+      "epoch": 2.4697727272727272,
+      "grad_norm": 2.0426415340521653,
+      "learning_rate": 6.369587422118278e-07,
+      "loss": 0.0266,
+      "step": 10867
+    },
+    {
+      "epoch": 2.4699999999999998,
+      "grad_norm": 1.7468257813764008,
+      "learning_rate": 6.36869508637999e-07,
+      "loss": 0.0308,
+      "step": 10868
+    },
+    {
+      "epoch": 2.4702272727272727,
+      "grad_norm": 1.3116599892444665,
+      "learning_rate": 6.367802748221302e-07,
+      "loss": 0.0137,
+      "step": 10869
+    },
+    {
+      "epoch": 2.4704545454545457,
+      "grad_norm": 2.3462401028482587,
+      "learning_rate": 6.366910407660409e-07,
+      "loss": 0.0382,
+      "step": 10870
+    },
+    {
+      "epoch": 2.470681818181818,
+      "grad_norm": 1.1074512340516565,
+      "learning_rate": 6.366018064715512e-07,
+      "loss": 0.0071,
+      "step": 10871
+    },
+    {
+      "epoch": 2.4709090909090907,
+      "grad_norm": 1.7552858853262974,
+      "learning_rate": 6.365125719404802e-07,
+      "loss": 0.0226,
+      "step": 10872
+    },
+    {
+      "epoch": 2.4711363636363637,
+      "grad_norm": 1.4494868467399737,
+      "learning_rate": 6.364233371746479e-07,
+      "loss": 0.0145,
+      "step": 10873
+    },
+    {
+      "epoch": 2.471363636363636,
+      "grad_norm": 1.4755755943877176,
+      "learning_rate": 6.363341021758739e-07,
+      "loss": 0.0421,
+      "step": 10874
+    },
+    {
+      "epoch": 2.471590909090909,
+      "grad_norm": 1.8067823422282276,
+      "learning_rate": 6.362448669459777e-07,
+      "loss": 0.0362,
+      "step": 10875
+    },
+    {
+      "epoch": 2.4718181818181817,
+      "grad_norm": 2.016124526266312,
+      "learning_rate": 6.361556314867789e-07,
+      "loss": 0.0187,
+      "step": 10876
+    },
+    {
+      "epoch": 2.4720454545454547,
+      "grad_norm": 2.609063258490142,
+      "learning_rate": 6.360663958000975e-07,
+      "loss": 0.0549,
+      "step": 10877
+    },
+    {
+      "epoch": 2.472272727272727,
+      "grad_norm": 4.670065392237402,
+      "learning_rate": 6.359771598877528e-07,
+      "loss": 0.09,
+      "step": 10878
+    },
+    {
+      "epoch": 2.4725,
+      "grad_norm": 4.797067592526729,
+      "learning_rate": 6.358879237515648e-07,
+      "loss": 0.05,
+      "step": 10879
+    },
+    {
+      "epoch": 2.4727272727272727,
+      "grad_norm": 2.0072990812176315,
+      "learning_rate": 6.357986873933529e-07,
+      "loss": 0.0316,
+      "step": 10880
+    },
+    {
+      "epoch": 2.4729545454545456,
+      "grad_norm": 1.093552308737013,
+      "learning_rate": 6.35709450814937e-07,
+      "loss": 0.018,
+      "step": 10881
+    },
+    {
+      "epoch": 2.473181818181818,
+      "grad_norm": 1.619949510974502,
+      "learning_rate": 6.356202140181368e-07,
+      "loss": 0.0195,
+      "step": 10882
+    },
+    {
+      "epoch": 2.473409090909091,
+      "grad_norm": 3.1406271991364854,
+      "learning_rate": 6.355309770047719e-07,
+      "loss": 0.0396,
+      "step": 10883
+    },
+    {
+      "epoch": 2.4736363636363636,
+      "grad_norm": 1.2784234871475921,
+      "learning_rate": 6.354417397766619e-07,
+      "loss": 0.0208,
+      "step": 10884
+    },
+    {
+      "epoch": 2.473863636363636,
+      "grad_norm": 1.45789205686941,
+      "learning_rate": 6.353525023356266e-07,
+      "loss": 0.0153,
+      "step": 10885
+    },
+    {
+      "epoch": 2.474090909090909,
+      "grad_norm": 0.7961441844956687,
+      "learning_rate": 6.352632646834858e-07,
+      "loss": 0.0121,
+      "step": 10886
+    },
+    {
+      "epoch": 2.4743181818181816,
+      "grad_norm": 2.294897366607033,
+      "learning_rate": 6.351740268220591e-07,
+      "loss": 0.0283,
+      "step": 10887
+    },
+    {
+      "epoch": 2.4745454545454546,
+      "grad_norm": 2.2025377894003655,
+      "learning_rate": 6.350847887531664e-07,
+      "loss": 0.0129,
+      "step": 10888
+    },
+    {
+      "epoch": 2.474772727272727,
+      "grad_norm": 2.754588596630246,
+      "learning_rate": 6.34995550478627e-07,
+      "loss": 0.0364,
+      "step": 10889
+    },
+    {
+      "epoch": 2.475,
+      "grad_norm": 1.3896366048327755,
+      "learning_rate": 6.349063120002611e-07,
+      "loss": 0.0192,
+      "step": 10890
+    },
+    {
+      "epoch": 2.4752272727272726,
+      "grad_norm": 1.3360399727069436,
+      "learning_rate": 6.34817073319888e-07,
+      "loss": 0.0165,
+      "step": 10891
+    },
+    {
+      "epoch": 2.4754545454545456,
+      "grad_norm": 2.401385945374518,
+      "learning_rate": 6.347278344393276e-07,
+      "loss": 0.0266,
+      "step": 10892
+    },
+    {
+      "epoch": 2.475681818181818,
+      "grad_norm": 1.773388145063138,
+      "learning_rate": 6.346385953604001e-07,
+      "loss": 0.0159,
+      "step": 10893
+    },
+    {
+      "epoch": 2.475909090909091,
+      "grad_norm": 6.957632788294299,
+      "learning_rate": 6.345493560849245e-07,
+      "loss": 0.0692,
+      "step": 10894
+    },
+    {
+      "epoch": 2.4761363636363636,
+      "grad_norm": 1.6018743352443778,
+      "learning_rate": 6.34460116614721e-07,
+      "loss": 0.0208,
+      "step": 10895
+    },
+    {
+      "epoch": 2.4763636363636365,
+      "grad_norm": 2.218024290490397,
+      "learning_rate": 6.34370876951609e-07,
+      "loss": 0.0175,
+      "step": 10896
+    },
+    {
+      "epoch": 2.476590909090909,
+      "grad_norm": 1.6803343997714613,
+      "learning_rate": 6.342816370974088e-07,
+      "loss": 0.0141,
+      "step": 10897
+    },
+    {
+      "epoch": 2.4768181818181816,
+      "grad_norm": 2.395290862121279,
+      "learning_rate": 6.341923970539395e-07,
+      "loss": 0.0391,
+      "step": 10898
+    },
+    {
+      "epoch": 2.4770454545454546,
+      "grad_norm": 2.0464605397528213,
+      "learning_rate": 6.341031568230215e-07,
+      "loss": 0.0386,
+      "step": 10899
+    },
+    {
+      "epoch": 2.4772727272727275,
+      "grad_norm": 4.590430442685313,
+      "learning_rate": 6.34013916406474e-07,
+      "loss": 0.0299,
+      "step": 10900
+    },
+    {
+      "epoch": 2.4775,
+      "grad_norm": 1.8658031411170217,
+      "learning_rate": 6.339246758061171e-07,
+      "loss": 0.0329,
+      "step": 10901
+    },
+    {
+      "epoch": 2.4777272727272726,
+      "grad_norm": 2.005484863270229,
+      "learning_rate": 6.338354350237706e-07,
+      "loss": 0.0321,
+      "step": 10902
+    },
+    {
+      "epoch": 2.4779545454545455,
+      "grad_norm": 0.6939994433521477,
+      "learning_rate": 6.33746194061254e-07,
+      "loss": 0.0047,
+      "step": 10903
+    },
+    {
+      "epoch": 2.478181818181818,
+      "grad_norm": 2.0244604001420274,
+      "learning_rate": 6.336569529203875e-07,
+      "loss": 0.022,
+      "step": 10904
+    },
+    {
+      "epoch": 2.478409090909091,
+      "grad_norm": 1.5150497340710496,
+      "learning_rate": 6.335677116029904e-07,
+      "loss": 0.0208,
+      "step": 10905
+    },
+    {
+      "epoch": 2.4786363636363635,
+      "grad_norm": 1.640897456419249,
+      "learning_rate": 6.334784701108826e-07,
+      "loss": 0.0148,
+      "step": 10906
+    },
+    {
+      "epoch": 2.4788636363636365,
+      "grad_norm": 1.497122850226125,
+      "learning_rate": 6.333892284458841e-07,
+      "loss": 0.013,
+      "step": 10907
+    },
+    {
+      "epoch": 2.479090909090909,
+      "grad_norm": 1.7253450415887293,
+      "learning_rate": 6.332999866098148e-07,
+      "loss": 0.0118,
+      "step": 10908
+    },
+    {
+      "epoch": 2.479318181818182,
+      "grad_norm": 1.082734931054992,
+      "learning_rate": 6.332107446044941e-07,
+      "loss": 0.0038,
+      "step": 10909
+    },
+    {
+      "epoch": 2.4795454545454545,
+      "grad_norm": 1.3121907967750044,
+      "learning_rate": 6.331215024317421e-07,
+      "loss": 0.0124,
+      "step": 10910
+    },
+    {
+      "epoch": 2.4797727272727275,
+      "grad_norm": 1.7461268977240676,
+      "learning_rate": 6.330322600933784e-07,
+      "loss": 0.0196,
+      "step": 10911
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 2.2555412981451064,
+      "learning_rate": 6.32943017591223e-07,
+      "loss": 0.0269,
+      "step": 10912
+    },
+    {
+      "epoch": 2.480227272727273,
+      "grad_norm": 2.8269019772820765,
+      "learning_rate": 6.328537749270955e-07,
+      "loss": 0.0432,
+      "step": 10913
+    },
+    {
+      "epoch": 2.4804545454545455,
+      "grad_norm": 1.2585649889201407,
+      "learning_rate": 6.327645321028158e-07,
+      "loss": 0.0149,
+      "step": 10914
+    },
+    {
+      "epoch": 2.480681818181818,
+      "grad_norm": 1.5829306621577142,
+      "learning_rate": 6.326752891202039e-07,
+      "loss": 0.0158,
+      "step": 10915
+    },
+    {
+      "epoch": 2.480909090909091,
+      "grad_norm": 1.672087419707205,
+      "learning_rate": 6.325860459810793e-07,
+      "loss": 0.0221,
+      "step": 10916
+    },
+    {
+      "epoch": 2.4811363636363635,
+      "grad_norm": 2.5908527284867704,
+      "learning_rate": 6.324968026872621e-07,
+      "loss": 0.0235,
+      "step": 10917
+    },
+    {
+      "epoch": 2.4813636363636364,
+      "grad_norm": 1.850646994491693,
+      "learning_rate": 6.32407559240572e-07,
+      "loss": 0.0432,
+      "step": 10918
+    },
+    {
+      "epoch": 2.481590909090909,
+      "grad_norm": 3.1826978397026733,
+      "learning_rate": 6.323183156428289e-07,
+      "loss": 0.0408,
+      "step": 10919
+    },
+    {
+      "epoch": 2.481818181818182,
+      "grad_norm": 1.8490040549223437,
+      "learning_rate": 6.322290718958525e-07,
+      "loss": 0.0207,
+      "step": 10920
+    },
+    {
+      "epoch": 2.4820454545454544,
+      "grad_norm": 2.4182741138458987,
+      "learning_rate": 6.321398280014627e-07,
+      "loss": 0.0289,
+      "step": 10921
+    },
+    {
+      "epoch": 2.4822727272727274,
+      "grad_norm": 2.877771108951923,
+      "learning_rate": 6.320505839614793e-07,
+      "loss": 0.045,
+      "step": 10922
+    },
+    {
+      "epoch": 2.4825,
+      "grad_norm": 2.6543127045040995,
+      "learning_rate": 6.319613397777223e-07,
+      "loss": 0.0277,
+      "step": 10923
+    },
+    {
+      "epoch": 2.482727272727273,
+      "grad_norm": 1.7990100798959336,
+      "learning_rate": 6.318720954520114e-07,
+      "loss": 0.0135,
+      "step": 10924
+    },
+    {
+      "epoch": 2.4829545454545454,
+      "grad_norm": 1.9365270856430536,
+      "learning_rate": 6.317828509861664e-07,
+      "loss": 0.0302,
+      "step": 10925
+    },
+    {
+      "epoch": 2.4831818181818184,
+      "grad_norm": 1.8227791682869945,
+      "learning_rate": 6.316936063820075e-07,
+      "loss": 0.0353,
+      "step": 10926
+    },
+    {
+      "epoch": 2.483409090909091,
+      "grad_norm": 0.6690299521208757,
+      "learning_rate": 6.31604361641354e-07,
+      "loss": 0.005,
+      "step": 10927
+    },
+    {
+      "epoch": 2.4836363636363634,
+      "grad_norm": 1.8032028158501205,
+      "learning_rate": 6.315151167660263e-07,
+      "loss": 0.0124,
+      "step": 10928
+    },
+    {
+      "epoch": 2.4838636363636364,
+      "grad_norm": 2.0279606357391997,
+      "learning_rate": 6.314258717578439e-07,
+      "loss": 0.0482,
+      "step": 10929
+    },
+    {
+      "epoch": 2.484090909090909,
+      "grad_norm": 1.6809290020426497,
+      "learning_rate": 6.313366266186266e-07,
+      "loss": 0.0151,
+      "step": 10930
+    },
+    {
+      "epoch": 2.484318181818182,
+      "grad_norm": 1.5673742274634324,
+      "learning_rate": 6.312473813501946e-07,
+      "loss": 0.0274,
+      "step": 10931
+    },
+    {
+      "epoch": 2.4845454545454544,
+      "grad_norm": 2.195789405126306,
+      "learning_rate": 6.311581359543676e-07,
+      "loss": 0.0501,
+      "step": 10932
+    },
+    {
+      "epoch": 2.4847727272727274,
+      "grad_norm": 3.5571416923014367,
+      "learning_rate": 6.310688904329653e-07,
+      "loss": 0.03,
+      "step": 10933
+    },
+    {
+      "epoch": 2.485,
+      "grad_norm": 1.6208490764385686,
+      "learning_rate": 6.309796447878079e-07,
+      "loss": 0.0137,
+      "step": 10934
+    },
+    {
+      "epoch": 2.485227272727273,
+      "grad_norm": 1.9751364572784658,
+      "learning_rate": 6.308903990207152e-07,
+      "loss": 0.0286,
+      "step": 10935
+    },
+    {
+      "epoch": 2.4854545454545454,
+      "grad_norm": 2.212639525382802,
+      "learning_rate": 6.308011531335068e-07,
+      "loss": 0.0249,
+      "step": 10936
+    },
+    {
+      "epoch": 2.4856818181818183,
+      "grad_norm": 1.8036364722180689,
+      "learning_rate": 6.307119071280027e-07,
+      "loss": 0.0264,
+      "step": 10937
+    },
+    {
+      "epoch": 2.485909090909091,
+      "grad_norm": 2.1931317719680985,
+      "learning_rate": 6.306226610060229e-07,
+      "loss": 0.0469,
+      "step": 10938
+    },
+    {
+      "epoch": 2.486136363636364,
+      "grad_norm": 1.4347162643259832,
+      "learning_rate": 6.305334147693875e-07,
+      "loss": 0.0154,
+      "step": 10939
+    },
+    {
+      "epoch": 2.4863636363636363,
+      "grad_norm": 2.41045530606326,
+      "learning_rate": 6.304441684199159e-07,
+      "loss": 0.0324,
+      "step": 10940
+    },
+    {
+      "epoch": 2.486590909090909,
+      "grad_norm": 3.6770810843823316,
+      "learning_rate": 6.303549219594282e-07,
+      "loss": 0.0257,
+      "step": 10941
+    },
+    {
+      "epoch": 2.486818181818182,
+      "grad_norm": 2.8191495903415986,
+      "learning_rate": 6.302656753897443e-07,
+      "loss": 0.0347,
+      "step": 10942
+    },
+    {
+      "epoch": 2.487045454545455,
+      "grad_norm": 1.8322487938831942,
+      "learning_rate": 6.301764287126843e-07,
+      "loss": 0.0193,
+      "step": 10943
+    },
+    {
+      "epoch": 2.4872727272727273,
+      "grad_norm": 2.307556091477001,
+      "learning_rate": 6.300871819300676e-07,
+      "loss": 0.0231,
+      "step": 10944
+    },
+    {
+      "epoch": 2.4875,
+      "grad_norm": 2.022289523396304,
+      "learning_rate": 6.299979350437143e-07,
+      "loss": 0.0401,
+      "step": 10945
+    },
+    {
+      "epoch": 2.487727272727273,
+      "grad_norm": 2.1351445928815562,
+      "learning_rate": 6.299086880554446e-07,
+      "loss": 0.0222,
+      "step": 10946
+    },
+    {
+      "epoch": 2.4879545454545453,
+      "grad_norm": 3.191389442171109,
+      "learning_rate": 6.298194409670782e-07,
+      "loss": 0.0276,
+      "step": 10947
+    },
+    {
+      "epoch": 2.4881818181818183,
+      "grad_norm": 1.6383480409038969,
+      "learning_rate": 6.29730193780435e-07,
+      "loss": 0.0166,
+      "step": 10948
+    },
+    {
+      "epoch": 2.488409090909091,
+      "grad_norm": 2.220960652936976,
+      "learning_rate": 6.296409464973348e-07,
+      "loss": 0.024,
+      "step": 10949
+    },
+    {
+      "epoch": 2.4886363636363638,
+      "grad_norm": 2.7523387052720576,
+      "learning_rate": 6.295516991195979e-07,
+      "loss": 0.0307,
+      "step": 10950
+    },
+    {
+      "epoch": 2.4888636363636363,
+      "grad_norm": 4.569190930639715,
+      "learning_rate": 6.294624516490435e-07,
+      "loss": 0.0607,
+      "step": 10951
+    },
+    {
+      "epoch": 2.4890909090909092,
+      "grad_norm": 2.440879490236729,
+      "learning_rate": 6.29373204087492e-07,
+      "loss": 0.0555,
+      "step": 10952
+    },
+    {
+      "epoch": 2.4893181818181818,
+      "grad_norm": 2.8593507053110208,
+      "learning_rate": 6.292839564367634e-07,
+      "loss": 0.0421,
+      "step": 10953
+    },
+    {
+      "epoch": 2.4895454545454547,
+      "grad_norm": 1.2814436345198816,
+      "learning_rate": 6.291947086986775e-07,
+      "loss": 0.0151,
+      "step": 10954
+    },
+    {
+      "epoch": 2.4897727272727272,
+      "grad_norm": 3.4040467597864814,
+      "learning_rate": 6.291054608750539e-07,
+      "loss": 0.0353,
+      "step": 10955
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.9874050719755476,
+      "learning_rate": 6.290162129677131e-07,
+      "loss": 0.0234,
+      "step": 10956
+    },
+    {
+      "epoch": 2.4902272727272727,
+      "grad_norm": 2.6188597797436026,
+      "learning_rate": 6.289269649784745e-07,
+      "loss": 0.052,
+      "step": 10957
+    },
+    {
+      "epoch": 2.4904545454545453,
+      "grad_norm": 2.5843217564915077,
+      "learning_rate": 6.288377169091582e-07,
+      "loss": 0.0147,
+      "step": 10958
+    },
+    {
+      "epoch": 2.490681818181818,
+      "grad_norm": 3.6317270773218433,
+      "learning_rate": 6.287484687615844e-07,
+      "loss": 0.0615,
+      "step": 10959
+    },
+    {
+      "epoch": 2.4909090909090907,
+      "grad_norm": 2.140983813457729,
+      "learning_rate": 6.286592205375724e-07,
+      "loss": 0.0257,
+      "step": 10960
+    },
+    {
+      "epoch": 2.4911363636363637,
+      "grad_norm": 2.711719336621933,
+      "learning_rate": 6.285699722389428e-07,
+      "loss": 0.0314,
+      "step": 10961
+    },
+    {
+      "epoch": 2.4913636363636362,
+      "grad_norm": 2.1496145646402196,
+      "learning_rate": 6.284807238675151e-07,
+      "loss": 0.0309,
+      "step": 10962
+    },
+    {
+      "epoch": 2.491590909090909,
+      "grad_norm": 2.45782840090567,
+      "learning_rate": 6.283914754251095e-07,
+      "loss": 0.0318,
+      "step": 10963
+    },
+    {
+      "epoch": 2.4918181818181817,
+      "grad_norm": 1.3354811606335324,
+      "learning_rate": 6.283022269135458e-07,
+      "loss": 0.0262,
+      "step": 10964
+    },
+    {
+      "epoch": 2.4920454545454547,
+      "grad_norm": 1.7582746110842964,
+      "learning_rate": 6.28212978334644e-07,
+      "loss": 0.025,
+      "step": 10965
+    },
+    {
+      "epoch": 2.492272727272727,
+      "grad_norm": 2.126324712590461,
+      "learning_rate": 6.281237296902237e-07,
+      "loss": 0.0222,
+      "step": 10966
+    },
+    {
+      "epoch": 2.4925,
+      "grad_norm": 1.450109743494687,
+      "learning_rate": 6.280344809821053e-07,
+      "loss": 0.0226,
+      "step": 10967
+    },
+    {
+      "epoch": 2.4927272727272727,
+      "grad_norm": 1.90172938492032,
+      "learning_rate": 6.279452322121084e-07,
+      "loss": 0.0251,
+      "step": 10968
+    },
+    {
+      "epoch": 2.4929545454545456,
+      "grad_norm": 2.480014033723489,
+      "learning_rate": 6.278559833820531e-07,
+      "loss": 0.0316,
+      "step": 10969
+    },
+    {
+      "epoch": 2.493181818181818,
+      "grad_norm": 1.904853802405333,
+      "learning_rate": 6.277667344937595e-07,
+      "loss": 0.0324,
+      "step": 10970
+    },
+    {
+      "epoch": 2.4934090909090907,
+      "grad_norm": 1.5485876216289889,
+      "learning_rate": 6.276774855490472e-07,
+      "loss": 0.0149,
+      "step": 10971
+    },
+    {
+      "epoch": 2.4936363636363637,
+      "grad_norm": 3.992112380058109,
+      "learning_rate": 6.275882365497364e-07,
+      "loss": 0.0345,
+      "step": 10972
+    },
+    {
+      "epoch": 2.493863636363636,
+      "grad_norm": 1.7205798087219142,
+      "learning_rate": 6.274989874976468e-07,
+      "loss": 0.0198,
+      "step": 10973
+    },
+    {
+      "epoch": 2.494090909090909,
+      "grad_norm": 3.2602214224721187,
+      "learning_rate": 6.274097383945986e-07,
+      "loss": 0.0376,
+      "step": 10974
+    },
+    {
+      "epoch": 2.4943181818181817,
+      "grad_norm": 2.330711142637699,
+      "learning_rate": 6.273204892424116e-07,
+      "loss": 0.0342,
+      "step": 10975
+    },
+    {
+      "epoch": 2.4945454545454546,
+      "grad_norm": 2.0147361693535846,
+      "learning_rate": 6.272312400429057e-07,
+      "loss": 0.0205,
+      "step": 10976
+    },
+    {
+      "epoch": 2.494772727272727,
+      "grad_norm": 1.6837468099828634,
+      "learning_rate": 6.271419907979011e-07,
+      "loss": 0.0291,
+      "step": 10977
+    },
+    {
+      "epoch": 2.495,
+      "grad_norm": 1.3566414292881623,
+      "learning_rate": 6.270527415092175e-07,
+      "loss": 0.0144,
+      "step": 10978
+    },
+    {
+      "epoch": 2.4952272727272726,
+      "grad_norm": 1.425637123502602,
+      "learning_rate": 6.269634921786747e-07,
+      "loss": 0.0174,
+      "step": 10979
+    },
+    {
+      "epoch": 2.4954545454545456,
+      "grad_norm": 3.0490357295947934,
+      "learning_rate": 6.268742428080932e-07,
+      "loss": 0.0486,
+      "step": 10980
+    },
+    {
+      "epoch": 2.495681818181818,
+      "grad_norm": 1.7432915260341049,
+      "learning_rate": 6.267849933992926e-07,
+      "loss": 0.0165,
+      "step": 10981
+    },
+    {
+      "epoch": 2.495909090909091,
+      "grad_norm": 1.7012116071028471,
+      "learning_rate": 6.266957439540926e-07,
+      "loss": 0.0499,
+      "step": 10982
+    },
+    {
+      "epoch": 2.4961363636363636,
+      "grad_norm": 1.8185291879852554,
+      "learning_rate": 6.266064944743135e-07,
+      "loss": 0.0213,
+      "step": 10983
+    },
+    {
+      "epoch": 2.496363636363636,
+      "grad_norm": 1.1980837653232,
+      "learning_rate": 6.265172449617752e-07,
+      "loss": 0.0106,
+      "step": 10984
+    },
+    {
+      "epoch": 2.496590909090909,
+      "grad_norm": 1.4399089192341408,
+      "learning_rate": 6.264279954182978e-07,
+      "loss": 0.0186,
+      "step": 10985
+    },
+    {
+      "epoch": 2.496818181818182,
+      "grad_norm": 2.1569017321590263,
+      "learning_rate": 6.263387458457009e-07,
+      "loss": 0.0265,
+      "step": 10986
+    },
+    {
+      "epoch": 2.4970454545454546,
+      "grad_norm": 2.2495543054266385,
+      "learning_rate": 6.262494962458047e-07,
+      "loss": 0.0425,
+      "step": 10987
+    },
+    {
+      "epoch": 2.497272727272727,
+      "grad_norm": 1.8967537525735507,
+      "learning_rate": 6.26160246620429e-07,
+      "loss": 0.0224,
+      "step": 10988
+    },
+    {
+      "epoch": 2.4975,
+      "grad_norm": 1.0994432115047899,
+      "learning_rate": 6.26070996971394e-07,
+      "loss": 0.0107,
+      "step": 10989
+    },
+    {
+      "epoch": 2.4977272727272726,
+      "grad_norm": 1.3452127956292532,
+      "learning_rate": 6.259817473005193e-07,
+      "loss": 0.0218,
+      "step": 10990
+    },
+    {
+      "epoch": 2.4979545454545455,
+      "grad_norm": 1.291062716748942,
+      "learning_rate": 6.258924976096252e-07,
+      "loss": 0.0145,
+      "step": 10991
+    },
+    {
+      "epoch": 2.498181818181818,
+      "grad_norm": 1.27028193466812,
+      "learning_rate": 6.258032479005316e-07,
+      "loss": 0.0132,
+      "step": 10992
+    },
+    {
+      "epoch": 2.498409090909091,
+      "grad_norm": 1.901656648776654,
+      "learning_rate": 6.257139981750582e-07,
+      "loss": 0.0364,
+      "step": 10993
+    },
+    {
+      "epoch": 2.4986363636363635,
+      "grad_norm": 2.4352458099245684,
+      "learning_rate": 6.256247484350253e-07,
+      "loss": 0.0218,
+      "step": 10994
+    },
+    {
+      "epoch": 2.4988636363636365,
+      "grad_norm": 2.6880138984984,
+      "learning_rate": 6.255354986822526e-07,
+      "loss": 0.0214,
+      "step": 10995
+    },
+    {
+      "epoch": 2.499090909090909,
+      "grad_norm": 2.4440483783944336,
+      "learning_rate": 6.254462489185601e-07,
+      "loss": 0.0273,
+      "step": 10996
+    },
+    {
+      "epoch": 2.499318181818182,
+      "grad_norm": 3.752930885437377,
+      "learning_rate": 6.25356999145768e-07,
+      "loss": 0.0559,
+      "step": 10997
+    },
+    {
+      "epoch": 2.4995454545454545,
+      "grad_norm": 3.426698384487745,
+      "learning_rate": 6.252677493656958e-07,
+      "loss": 0.0386,
+      "step": 10998
+    },
+    {
+      "epoch": 2.4997727272727275,
+      "grad_norm": 1.5899082625713845,
+      "learning_rate": 6.251784995801638e-07,
+      "loss": 0.0205,
+      "step": 10999
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.9089637585113608,
+      "learning_rate": 6.25089249790992e-07,
+      "loss": 0.015,
+      "step": 11000
+    },
+    {
+      "epoch": 2.5002272727272725,
+      "grad_norm": 2.3729385504267384,
+      "learning_rate": 6.25e-07,
+      "loss": 0.0186,
+      "step": 11001
+    },
+    {
+      "epoch": 2.5004545454545455,
+      "grad_norm": 2.0624570295273394,
+      "learning_rate": 6.249107502090082e-07,
+      "loss": 0.0224,
+      "step": 11002
+    },
+    {
+      "epoch": 2.5006818181818184,
+      "grad_norm": 0.5251757620676692,
+      "learning_rate": 6.248215004198363e-07,
+      "loss": 0.0122,
+      "step": 11003
+    },
+    {
+      "epoch": 2.500909090909091,
+      "grad_norm": 2.1922707524746614,
+      "learning_rate": 6.247322506343043e-07,
+      "loss": 0.0429,
+      "step": 11004
+    },
+    {
+      "epoch": 2.5011363636363635,
+      "grad_norm": 4.246414260473672,
+      "learning_rate": 6.246430008542323e-07,
+      "loss": 0.0431,
+      "step": 11005
+    },
+    {
+      "epoch": 2.5013636363636365,
+      "grad_norm": 2.0120089789642863,
+      "learning_rate": 6.2455375108144e-07,
+      "loss": 0.042,
+      "step": 11006
+    },
+    {
+      "epoch": 2.501590909090909,
+      "grad_norm": 1.6325234617712343,
+      "learning_rate": 6.244645013177475e-07,
+      "loss": 0.0113,
+      "step": 11007
+    },
+    {
+      "epoch": 2.501818181818182,
+      "grad_norm": 1.3548109895315057,
+      "learning_rate": 6.24375251564975e-07,
+      "loss": 0.0168,
+      "step": 11008
+    },
+    {
+      "epoch": 2.5020454545454545,
+      "grad_norm": 2.522689539341824,
+      "learning_rate": 6.242860018249419e-07,
+      "loss": 0.0343,
+      "step": 11009
+    },
+    {
+      "epoch": 2.5022727272727274,
+      "grad_norm": 2.079546756844252,
+      "learning_rate": 6.241967520994686e-07,
+      "loss": 0.0208,
+      "step": 11010
+    },
+    {
+      "epoch": 2.5025,
+      "grad_norm": 1.853086642669776,
+      "learning_rate": 6.241075023903749e-07,
+      "loss": 0.016,
+      "step": 11011
+    },
+    {
+      "epoch": 2.502727272727273,
+      "grad_norm": 0.8950208190078854,
+      "learning_rate": 6.240182526994809e-07,
+      "loss": 0.012,
+      "step": 11012
+    },
+    {
+      "epoch": 2.5029545454545454,
+      "grad_norm": 1.530832971736629,
+      "learning_rate": 6.239290030286062e-07,
+      "loss": 0.0186,
+      "step": 11013
+    },
+    {
+      "epoch": 2.503181818181818,
+      "grad_norm": 3.921663035700055,
+      "learning_rate": 6.238397533795711e-07,
+      "loss": 0.0546,
+      "step": 11014
+    },
+    {
+      "epoch": 2.503409090909091,
+      "grad_norm": 1.3861793353234257,
+      "learning_rate": 6.237505037541955e-07,
+      "loss": 0.009,
+      "step": 11015
+    },
+    {
+      "epoch": 2.503636363636364,
+      "grad_norm": 2.2774035297550115,
+      "learning_rate": 6.236612541542992e-07,
+      "loss": 0.0517,
+      "step": 11016
+    },
+    {
+      "epoch": 2.5038636363636364,
+      "grad_norm": 2.114616429763541,
+      "learning_rate": 6.235720045817026e-07,
+      "loss": 0.0227,
+      "step": 11017
+    },
+    {
+      "epoch": 2.504090909090909,
+      "grad_norm": 1.7796572222249776,
+      "learning_rate": 6.23482755038225e-07,
+      "loss": 0.0237,
+      "step": 11018
+    },
+    {
+      "epoch": 2.504318181818182,
+      "grad_norm": 2.482714036328514,
+      "learning_rate": 6.233935055256866e-07,
+      "loss": 0.0345,
+      "step": 11019
+    },
+    {
+      "epoch": 2.5045454545454544,
+      "grad_norm": 1.6040305424186077,
+      "learning_rate": 6.233042560459076e-07,
+      "loss": 0.018,
+      "step": 11020
+    },
+    {
+      "epoch": 2.5047727272727274,
+      "grad_norm": 1.7754714122118131,
+      "learning_rate": 6.232150066007079e-07,
+      "loss": 0.0344,
+      "step": 11021
+    },
+    {
+      "epoch": 2.505,
+      "grad_norm": 2.001218777302332,
+      "learning_rate": 6.23125757191907e-07,
+      "loss": 0.0202,
+      "step": 11022
+    },
+    {
+      "epoch": 2.505227272727273,
+      "grad_norm": 2.000668137673269,
+      "learning_rate": 6.230365078213254e-07,
+      "loss": 0.0217,
+      "step": 11023
+    },
+    {
+      "epoch": 2.5054545454545454,
+      "grad_norm": 1.3818821641739307,
+      "learning_rate": 6.229472584907827e-07,
+      "loss": 0.0102,
+      "step": 11024
+    },
+    {
+      "epoch": 2.5056818181818183,
+      "grad_norm": 4.1094606515023,
+      "learning_rate": 6.22858009202099e-07,
+      "loss": 0.0629,
+      "step": 11025
+    },
+    {
+      "epoch": 2.505909090909091,
+      "grad_norm": 1.0847675025688277,
+      "learning_rate": 6.227687599570946e-07,
+      "loss": 0.0088,
+      "step": 11026
+    },
+    {
+      "epoch": 2.5061363636363634,
+      "grad_norm": 2.083951272365241,
+      "learning_rate": 6.226795107575887e-07,
+      "loss": 0.0372,
+      "step": 11027
+    },
+    {
+      "epoch": 2.5063636363636363,
+      "grad_norm": 2.607733868020019,
+      "learning_rate": 6.225902616054016e-07,
+      "loss": 0.0352,
+      "step": 11028
+    },
+    {
+      "epoch": 2.5065909090909093,
+      "grad_norm": 1.3063395263764805,
+      "learning_rate": 6.225010125023532e-07,
+      "loss": 0.0144,
+      "step": 11029
+    },
+    {
+      "epoch": 2.506818181818182,
+      "grad_norm": 1.2563417530577403,
+      "learning_rate": 6.224117634502639e-07,
+      "loss": 0.0171,
+      "step": 11030
+    },
+    {
+      "epoch": 2.5070454545454544,
+      "grad_norm": 2.4055069419160864,
+      "learning_rate": 6.223225144509529e-07,
+      "loss": 0.0193,
+      "step": 11031
+    },
+    {
+      "epoch": 2.5072727272727273,
+      "grad_norm": 1.7807326900246292,
+      "learning_rate": 6.222332655062407e-07,
+      "loss": 0.0144,
+      "step": 11032
+    },
+    {
+      "epoch": 2.5075,
+      "grad_norm": 1.592565593698499,
+      "learning_rate": 6.221440166179469e-07,
+      "loss": 0.0215,
+      "step": 11033
+    },
+    {
+      "epoch": 2.507727272727273,
+      "grad_norm": 2.049580153727903,
+      "learning_rate": 6.220547677878917e-07,
+      "loss": 0.0269,
+      "step": 11034
+    },
+    {
+      "epoch": 2.5079545454545453,
+      "grad_norm": 0.9429997811443767,
+      "learning_rate": 6.21965519017895e-07,
+      "loss": 0.0072,
+      "step": 11035
+    },
+    {
+      "epoch": 2.5081818181818183,
+      "grad_norm": 1.3296283355755634,
+      "learning_rate": 6.218762703097764e-07,
+      "loss": 0.0234,
+      "step": 11036
+    },
+    {
+      "epoch": 2.508409090909091,
+      "grad_norm": 1.8476259691606716,
+      "learning_rate": 6.217870216653562e-07,
+      "loss": 0.0262,
+      "step": 11037
+    },
+    {
+      "epoch": 2.5086363636363638,
+      "grad_norm": 2.220436666685711,
+      "learning_rate": 6.216977730864542e-07,
+      "loss": 0.0318,
+      "step": 11038
+    },
+    {
+      "epoch": 2.5088636363636363,
+      "grad_norm": 1.8255671191125311,
+      "learning_rate": 6.216085245748907e-07,
+      "loss": 0.0204,
+      "step": 11039
+    },
+    {
+      "epoch": 2.509090909090909,
+      "grad_norm": 1.052919560547335,
+      "learning_rate": 6.21519276132485e-07,
+      "loss": 0.0131,
+      "step": 11040
+    },
+    {
+      "epoch": 2.509318181818182,
+      "grad_norm": 1.7033439379112496,
+      "learning_rate": 6.214300277610574e-07,
+      "loss": 0.0181,
+      "step": 11041
+    },
+    {
+      "epoch": 2.5095454545454547,
+      "grad_norm": 1.854567460157462,
+      "learning_rate": 6.213407794624277e-07,
+      "loss": 0.016,
+      "step": 11042
+    },
+    {
+      "epoch": 2.5097727272727273,
+      "grad_norm": 1.1349965165589944,
+      "learning_rate": 6.21251531238416e-07,
+      "loss": 0.0099,
+      "step": 11043
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 2.1943446548357075,
+      "learning_rate": 6.21162283090842e-07,
+      "loss": 0.0253,
+      "step": 11044
+    },
+    {
+      "epoch": 2.5102272727272728,
+      "grad_norm": 2.571402974223069,
+      "learning_rate": 6.210730350215257e-07,
+      "loss": 0.0305,
+      "step": 11045
+    },
+    {
+      "epoch": 2.5104545454545457,
+      "grad_norm": 1.3973317379683032,
+      "learning_rate": 6.209837870322871e-07,
+      "loss": 0.0142,
+      "step": 11046
+    },
+    {
+      "epoch": 2.5106818181818182,
+      "grad_norm": 1.7250864886856754,
+      "learning_rate": 6.208945391249462e-07,
+      "loss": 0.0139,
+      "step": 11047
+    },
+    {
+      "epoch": 2.5109090909090908,
+      "grad_norm": 1.4403455150124305,
+      "learning_rate": 6.20805291301323e-07,
+      "loss": 0.024,
+      "step": 11048
+    },
+    {
+      "epoch": 2.5111363636363637,
+      "grad_norm": 2.0014367889767213,
+      "learning_rate": 6.207160435632369e-07,
+      "loss": 0.035,
+      "step": 11049
+    },
+    {
+      "epoch": 2.5113636363636362,
+      "grad_norm": 1.5475488420641572,
+      "learning_rate": 6.206267959125081e-07,
+      "loss": 0.014,
+      "step": 11050
+    },
+    {
+      "epoch": 2.511590909090909,
+      "grad_norm": 1.4611670266283783,
+      "learning_rate": 6.205375483509566e-07,
+      "loss": 0.0188,
+      "step": 11051
+    },
+    {
+      "epoch": 2.5118181818181817,
+      "grad_norm": 2.0420394141661533,
+      "learning_rate": 6.204483008804026e-07,
+      "loss": 0.0418,
+      "step": 11052
+    },
+    {
+      "epoch": 2.5120454545454547,
+      "grad_norm": 2.4195485638246876,
+      "learning_rate": 6.203590535026654e-07,
+      "loss": 0.0326,
+      "step": 11053
+    },
+    {
+      "epoch": 2.512272727272727,
+      "grad_norm": 1.626809350594232,
+      "learning_rate": 6.202698062195651e-07,
+      "loss": 0.0229,
+      "step": 11054
+    },
+    {
+      "epoch": 2.5125,
+      "grad_norm": 1.007876840930417,
+      "learning_rate": 6.201805590329218e-07,
+      "loss": 0.0068,
+      "step": 11055
+    },
+    {
+      "epoch": 2.5127272727272727,
+      "grad_norm": 2.455337066558525,
+      "learning_rate": 6.200913119445557e-07,
+      "loss": 0.017,
+      "step": 11056
+    },
+    {
+      "epoch": 2.512954545454545,
+      "grad_norm": 3.6329141573483614,
+      "learning_rate": 6.200020649562859e-07,
+      "loss": 0.0552,
+      "step": 11057
+    },
+    {
+      "epoch": 2.513181818181818,
+      "grad_norm": 1.4603280436551227,
+      "learning_rate": 6.199128180699326e-07,
+      "loss": 0.0204,
+      "step": 11058
+    },
+    {
+      "epoch": 2.513409090909091,
+      "grad_norm": 1.696946187995458,
+      "learning_rate": 6.198235712873159e-07,
+      "loss": 0.041,
+      "step": 11059
+    },
+    {
+      "epoch": 2.5136363636363637,
+      "grad_norm": 1.3613518386397179,
+      "learning_rate": 6.197343246102558e-07,
+      "loss": 0.02,
+      "step": 11060
+    },
+    {
+      "epoch": 2.513863636363636,
+      "grad_norm": 2.74773652936439,
+      "learning_rate": 6.196450780405721e-07,
+      "loss": 0.0422,
+      "step": 11061
+    },
+    {
+      "epoch": 2.514090909090909,
+      "grad_norm": 1.2807425835319755,
+      "learning_rate": 6.195558315800844e-07,
+      "loss": 0.0112,
+      "step": 11062
+    },
+    {
+      "epoch": 2.5143181818181817,
+      "grad_norm": 2.9619453195298613,
+      "learning_rate": 6.194665852306127e-07,
+      "loss": 0.0486,
+      "step": 11063
+    },
+    {
+      "epoch": 2.5145454545454546,
+      "grad_norm": 1.6563082869048027,
+      "learning_rate": 6.193773389939771e-07,
+      "loss": 0.0198,
+      "step": 11064
+    },
+    {
+      "epoch": 2.514772727272727,
+      "grad_norm": 1.7195792176789457,
+      "learning_rate": 6.192880928719974e-07,
+      "loss": 0.0264,
+      "step": 11065
+    },
+    {
+      "epoch": 2.515,
+      "grad_norm": 1.8276175738072824,
+      "learning_rate": 6.191988468664934e-07,
+      "loss": 0.0193,
+      "step": 11066
+    },
+    {
+      "epoch": 2.5152272727272726,
+      "grad_norm": 1.5055201087417431,
+      "learning_rate": 6.191096009792851e-07,
+      "loss": 0.0175,
+      "step": 11067
+    },
+    {
+      "epoch": 2.5154545454545456,
+      "grad_norm": 1.5392629586103264,
+      "learning_rate": 6.190203552121921e-07,
+      "loss": 0.0105,
+      "step": 11068
+    },
+    {
+      "epoch": 2.515681818181818,
+      "grad_norm": 2.022391717874961,
+      "learning_rate": 6.189311095670347e-07,
+      "loss": 0.0186,
+      "step": 11069
+    },
+    {
+      "epoch": 2.5159090909090907,
+      "grad_norm": 1.9641523254303574,
+      "learning_rate": 6.188418640456327e-07,
+      "loss": 0.0229,
+      "step": 11070
+    },
+    {
+      "epoch": 2.5161363636363636,
+      "grad_norm": 2.9780925429725813,
+      "learning_rate": 6.187526186498055e-07,
+      "loss": 0.0271,
+      "step": 11071
+    },
+    {
+      "epoch": 2.5163636363636366,
+      "grad_norm": 1.9243081426769544,
+      "learning_rate": 6.186633733813735e-07,
+      "loss": 0.0174,
+      "step": 11072
+    },
+    {
+      "epoch": 2.516590909090909,
+      "grad_norm": 2.8882852810054107,
+      "learning_rate": 6.185741282421563e-07,
+      "loss": 0.0487,
+      "step": 11073
+    },
+    {
+      "epoch": 2.5168181818181816,
+      "grad_norm": 1.8574397702425491,
+      "learning_rate": 6.18484883233974e-07,
+      "loss": 0.0317,
+      "step": 11074
+    },
+    {
+      "epoch": 2.5170454545454546,
+      "grad_norm": 2.1602414903417975,
+      "learning_rate": 6.183956383586461e-07,
+      "loss": 0.0374,
+      "step": 11075
+    },
+    {
+      "epoch": 2.517272727272727,
+      "grad_norm": 2.50575826621375,
+      "learning_rate": 6.183063936179927e-07,
+      "loss": 0.0558,
+      "step": 11076
+    },
+    {
+      "epoch": 2.5175,
+      "grad_norm": 1.6016585838491226,
+      "learning_rate": 6.182171490138336e-07,
+      "loss": 0.0052,
+      "step": 11077
+    },
+    {
+      "epoch": 2.5177272727272726,
+      "grad_norm": 2.2261170236033574,
+      "learning_rate": 6.181279045479889e-07,
+      "loss": 0.0411,
+      "step": 11078
+    },
+    {
+      "epoch": 2.5179545454545456,
+      "grad_norm": 3.2863489938726524,
+      "learning_rate": 6.18038660222278e-07,
+      "loss": 0.027,
+      "step": 11079
+    },
+    {
+      "epoch": 2.518181818181818,
+      "grad_norm": 1.5247168075170523,
+      "learning_rate": 6.179494160385209e-07,
+      "loss": 0.0102,
+      "step": 11080
+    },
+    {
+      "epoch": 2.518409090909091,
+      "grad_norm": 0.48460016395849415,
+      "learning_rate": 6.178601719985375e-07,
+      "loss": 0.0071,
+      "step": 11081
+    },
+    {
+      "epoch": 2.5186363636363636,
+      "grad_norm": 4.386422603405153,
+      "learning_rate": 6.177709281041477e-07,
+      "loss": 0.0186,
+      "step": 11082
+    },
+    {
+      "epoch": 2.5188636363636365,
+      "grad_norm": 0.9835745626107808,
+      "learning_rate": 6.176816843571714e-07,
+      "loss": 0.0119,
+      "step": 11083
+    },
+    {
+      "epoch": 2.519090909090909,
+      "grad_norm": 2.0809264695218923,
+      "learning_rate": 6.175924407594282e-07,
+      "loss": 0.0274,
+      "step": 11084
+    },
+    {
+      "epoch": 2.519318181818182,
+      "grad_norm": 1.289872514143293,
+      "learning_rate": 6.17503197312738e-07,
+      "loss": 0.014,
+      "step": 11085
+    },
+    {
+      "epoch": 2.5195454545454545,
+      "grad_norm": 2.2690059545668384,
+      "learning_rate": 6.174139540189207e-07,
+      "loss": 0.0191,
+      "step": 11086
+    },
+    {
+      "epoch": 2.519772727272727,
+      "grad_norm": 1.5112941412553949,
+      "learning_rate": 6.173247108797964e-07,
+      "loss": 0.0199,
+      "step": 11087
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.2653816424949453,
+      "learning_rate": 6.172354678971843e-07,
+      "loss": 0.0561,
+      "step": 11088
+    },
+    {
+      "epoch": 2.520227272727273,
+      "grad_norm": 2.5355760652494106,
+      "learning_rate": 6.171462250729046e-07,
+      "loss": 0.024,
+      "step": 11089
+    },
+    {
+      "epoch": 2.5204545454545455,
+      "grad_norm": 1.968137299645445,
+      "learning_rate": 6.170569824087772e-07,
+      "loss": 0.022,
+      "step": 11090
+    },
+    {
+      "epoch": 2.520681818181818,
+      "grad_norm": 1.386775706757669,
+      "learning_rate": 6.169677399066217e-07,
+      "loss": 0.021,
+      "step": 11091
+    },
+    {
+      "epoch": 2.520909090909091,
+      "grad_norm": 1.3516755651822703,
+      "learning_rate": 6.168784975682582e-07,
+      "loss": 0.0127,
+      "step": 11092
+    },
+    {
+      "epoch": 2.5211363636363635,
+      "grad_norm": 0.9901752311743511,
+      "learning_rate": 6.167892553955061e-07,
+      "loss": 0.0107,
+      "step": 11093
+    },
+    {
+      "epoch": 2.5213636363636365,
+      "grad_norm": 0.41608728580900456,
+      "learning_rate": 6.167000133901854e-07,
+      "loss": 0.0048,
+      "step": 11094
+    },
+    {
+      "epoch": 2.521590909090909,
+      "grad_norm": 2.688036443735637,
+      "learning_rate": 6.166107715541159e-07,
+      "loss": 0.0325,
+      "step": 11095
+    },
+    {
+      "epoch": 2.521818181818182,
+      "grad_norm": 1.7676512644000244,
+      "learning_rate": 6.165215298891176e-07,
+      "loss": 0.0188,
+      "step": 11096
+    },
+    {
+      "epoch": 2.5220454545454545,
+      "grad_norm": 2.029372802275238,
+      "learning_rate": 6.164322883970099e-07,
+      "loss": 0.0143,
+      "step": 11097
+    },
+    {
+      "epoch": 2.5222727272727274,
+      "grad_norm": 2.4760701729099557,
+      "learning_rate": 6.163430470796127e-07,
+      "loss": 0.0416,
+      "step": 11098
+    },
+    {
+      "epoch": 2.5225,
+      "grad_norm": 2.7487175050636194,
+      "learning_rate": 6.16253805938746e-07,
+      "loss": 0.0224,
+      "step": 11099
+    },
+    {
+      "epoch": 2.5227272727272725,
+      "grad_norm": 1.779354977302831,
+      "learning_rate": 6.161645649762297e-07,
+      "loss": 0.0303,
+      "step": 11100
+    },
+    {
+      "epoch": 2.5229545454545454,
+      "grad_norm": 1.6899616026291102,
+      "learning_rate": 6.160753241938832e-07,
+      "loss": 0.0196,
+      "step": 11101
+    },
+    {
+      "epoch": 2.5231818181818184,
+      "grad_norm": 2.0016617770669773,
+      "learning_rate": 6.159860835935262e-07,
+      "loss": 0.0323,
+      "step": 11102
+    },
+    {
+      "epoch": 2.523409090909091,
+      "grad_norm": 2.7896666865681907,
+      "learning_rate": 6.158968431769787e-07,
+      "loss": 0.0408,
+      "step": 11103
+    },
+    {
+      "epoch": 2.5236363636363635,
+      "grad_norm": 2.201040800661145,
+      "learning_rate": 6.158076029460606e-07,
+      "loss": 0.0439,
+      "step": 11104
+    },
+    {
+      "epoch": 2.5238636363636364,
+      "grad_norm": 0.7282319141973532,
+      "learning_rate": 6.157183629025915e-07,
+      "loss": 0.0078,
+      "step": 11105
+    },
+    {
+      "epoch": 2.524090909090909,
+      "grad_norm": 1.5234963131357695,
+      "learning_rate": 6.156291230483911e-07,
+      "loss": 0.0131,
+      "step": 11106
+    },
+    {
+      "epoch": 2.524318181818182,
+      "grad_norm": 1.5233115467117948,
+      "learning_rate": 6.155398833852792e-07,
+      "loss": 0.0252,
+      "step": 11107
+    },
+    {
+      "epoch": 2.5245454545454544,
+      "grad_norm": 1.4255850523543154,
+      "learning_rate": 6.154506439150756e-07,
+      "loss": 0.0173,
+      "step": 11108
+    },
+    {
+      "epoch": 2.5247727272727274,
+      "grad_norm": 0.8514421677055347,
+      "learning_rate": 6.153614046396003e-07,
+      "loss": 0.0108,
+      "step": 11109
+    },
+    {
+      "epoch": 2.525,
+      "grad_norm": 1.5050575659141783,
+      "learning_rate": 6.152721655606725e-07,
+      "loss": 0.0153,
+      "step": 11110
+    },
+    {
+      "epoch": 2.525227272727273,
+      "grad_norm": 2.0809825249566245,
+      "learning_rate": 6.151829266801122e-07,
+      "loss": 0.0144,
+      "step": 11111
+    },
+    {
+      "epoch": 2.5254545454545454,
+      "grad_norm": 1.8793904274205775,
+      "learning_rate": 6.150936879997392e-07,
+      "loss": 0.0376,
+      "step": 11112
+    },
+    {
+      "epoch": 2.525681818181818,
+      "grad_norm": 1.3964484606863727,
+      "learning_rate": 6.15004449521373e-07,
+      "loss": 0.0134,
+      "step": 11113
+    },
+    {
+      "epoch": 2.525909090909091,
+      "grad_norm": 1.5418617563259307,
+      "learning_rate": 6.149152112468339e-07,
+      "loss": 0.0237,
+      "step": 11114
+    },
+    {
+      "epoch": 2.526136363636364,
+      "grad_norm": 2.721571140147393,
+      "learning_rate": 6.148259731779411e-07,
+      "loss": 0.0194,
+      "step": 11115
+    },
+    {
+      "epoch": 2.5263636363636364,
+      "grad_norm": 0.6920830810365226,
+      "learning_rate": 6.147367353165143e-07,
+      "loss": 0.0144,
+      "step": 11116
+    },
+    {
+      "epoch": 2.526590909090909,
+      "grad_norm": 2.3848391972297014,
+      "learning_rate": 6.146474976643734e-07,
+      "loss": 0.0209,
+      "step": 11117
+    },
+    {
+      "epoch": 2.526818181818182,
+      "grad_norm": 2.046214108590808,
+      "learning_rate": 6.145582602233383e-07,
+      "loss": 0.0171,
+      "step": 11118
+    },
+    {
+      "epoch": 2.527045454545455,
+      "grad_norm": 1.55781159490251,
+      "learning_rate": 6.144690229952283e-07,
+      "loss": 0.0158,
+      "step": 11119
+    },
+    {
+      "epoch": 2.5272727272727273,
+      "grad_norm": 1.7245896967188814,
+      "learning_rate": 6.143797859818633e-07,
+      "loss": 0.024,
+      "step": 11120
+    },
+    {
+      "epoch": 2.5275,
+      "grad_norm": 3.710477357923557,
+      "learning_rate": 6.14290549185063e-07,
+      "loss": 0.0395,
+      "step": 11121
+    },
+    {
+      "epoch": 2.527727272727273,
+      "grad_norm": 1.8555352419925077,
+      "learning_rate": 6.142013126066473e-07,
+      "loss": 0.0334,
+      "step": 11122
+    },
+    {
+      "epoch": 2.5279545454545453,
+      "grad_norm": 1.5529442717960988,
+      "learning_rate": 6.141120762484356e-07,
+      "loss": 0.0378,
+      "step": 11123
+    },
+    {
+      "epoch": 2.5281818181818183,
+      "grad_norm": 1.885023847755054,
+      "learning_rate": 6.140228401122475e-07,
+      "loss": 0.021,
+      "step": 11124
+    },
+    {
+      "epoch": 2.528409090909091,
+      "grad_norm": 2.225507694574082,
+      "learning_rate": 6.139336041999027e-07,
+      "loss": 0.0312,
+      "step": 11125
+    },
+    {
+      "epoch": 2.528636363636364,
+      "grad_norm": 1.434695610400901,
+      "learning_rate": 6.138443685132212e-07,
+      "loss": 0.011,
+      "step": 11126
+    },
+    {
+      "epoch": 2.5288636363636363,
+      "grad_norm": 2.507308286647576,
+      "learning_rate": 6.137551330540226e-07,
+      "loss": 0.0266,
+      "step": 11127
+    },
+    {
+      "epoch": 2.5290909090909093,
+      "grad_norm": 1.750141906093758,
+      "learning_rate": 6.136658978241263e-07,
+      "loss": 0.0207,
+      "step": 11128
+    },
+    {
+      "epoch": 2.529318181818182,
+      "grad_norm": 1.2822814980522717,
+      "learning_rate": 6.135766628253522e-07,
+      "loss": 0.0132,
+      "step": 11129
+    },
+    {
+      "epoch": 2.5295454545454543,
+      "grad_norm": 1.2022685991710436,
+      "learning_rate": 6.134874280595198e-07,
+      "loss": 0.015,
+      "step": 11130
+    },
+    {
+      "epoch": 2.5297727272727273,
+      "grad_norm": 1.48898377244451,
+      "learning_rate": 6.133981935284491e-07,
+      "loss": 0.0205,
+      "step": 11131
+    },
+    {
+      "epoch": 2.5300000000000002,
+      "grad_norm": 2.1413751108548853,
+      "learning_rate": 6.133089592339592e-07,
+      "loss": 0.0319,
+      "step": 11132
+    },
+    {
+      "epoch": 2.5302272727272728,
+      "grad_norm": 1.6283071871686032,
+      "learning_rate": 6.1321972517787e-07,
+      "loss": 0.0156,
+      "step": 11133
+    },
+    {
+      "epoch": 2.5304545454545453,
+      "grad_norm": 2.239318375075804,
+      "learning_rate": 6.131304913620012e-07,
+      "loss": 0.044,
+      "step": 11134
+    },
+    {
+      "epoch": 2.5306818181818183,
+      "grad_norm": 1.3637867082687287,
+      "learning_rate": 6.130412577881726e-07,
+      "loss": 0.018,
+      "step": 11135
+    },
+    {
+      "epoch": 2.5309090909090908,
+      "grad_norm": 1.6394347500518052,
+      "learning_rate": 6.129520244582033e-07,
+      "loss": 0.0181,
+      "step": 11136
+    },
+    {
+      "epoch": 2.5311363636363637,
+      "grad_norm": 1.0383115600338864,
+      "learning_rate": 6.128627913739133e-07,
+      "loss": 0.0085,
+      "step": 11137
+    },
+    {
+      "epoch": 2.5313636363636363,
+      "grad_norm": 1.9333168781265573,
+      "learning_rate": 6.127735585371221e-07,
+      "loss": 0.0235,
+      "step": 11138
+    },
+    {
+      "epoch": 2.5315909090909092,
+      "grad_norm": 2.3309296086231104,
+      "learning_rate": 6.126843259496496e-07,
+      "loss": 0.0241,
+      "step": 11139
+    },
+    {
+      "epoch": 2.5318181818181817,
+      "grad_norm": 2.2756388530694993,
+      "learning_rate": 6.125950936133152e-07,
+      "loss": 0.0333,
+      "step": 11140
+    },
+    {
+      "epoch": 2.5320454545454547,
+      "grad_norm": 1.6295005491603376,
+      "learning_rate": 6.125058615299384e-07,
+      "loss": 0.032,
+      "step": 11141
+    },
+    {
+      "epoch": 2.5322727272727272,
+      "grad_norm": 1.3496177802285398,
+      "learning_rate": 6.124166297013386e-07,
+      "loss": 0.0193,
+      "step": 11142
+    },
+    {
+      "epoch": 2.5324999999999998,
+      "grad_norm": 2.2762213063487597,
+      "learning_rate": 6.123273981293359e-07,
+      "loss": 0.0489,
+      "step": 11143
+    },
+    {
+      "epoch": 2.5327272727272727,
+      "grad_norm": 1.475607093102924,
+      "learning_rate": 6.122381668157499e-07,
+      "loss": 0.0284,
+      "step": 11144
+    },
+    {
+      "epoch": 2.5329545454545457,
+      "grad_norm": 1.4230106467090922,
+      "learning_rate": 6.121489357623997e-07,
+      "loss": 0.0252,
+      "step": 11145
+    },
+    {
+      "epoch": 2.533181818181818,
+      "grad_norm": 1.4386552068923992,
+      "learning_rate": 6.12059704971105e-07,
+      "loss": 0.0135,
+      "step": 11146
+    },
+    {
+      "epoch": 2.5334090909090907,
+      "grad_norm": 1.97221917463957,
+      "learning_rate": 6.119704744436857e-07,
+      "loss": 0.0336,
+      "step": 11147
+    },
+    {
+      "epoch": 2.5336363636363637,
+      "grad_norm": 1.6661424134447476,
+      "learning_rate": 6.11881244181961e-07,
+      "loss": 0.0266,
+      "step": 11148
+    },
+    {
+      "epoch": 2.533863636363636,
+      "grad_norm": 1.5586991981937146,
+      "learning_rate": 6.117920141877507e-07,
+      "loss": 0.0188,
+      "step": 11149
+    },
+    {
+      "epoch": 2.534090909090909,
+      "grad_norm": 1.5085097742854152,
+      "learning_rate": 6.117027844628743e-07,
+      "loss": 0.0175,
+      "step": 11150
+    },
+    {
+      "epoch": 2.5343181818181817,
+      "grad_norm": 1.5480449947400279,
+      "learning_rate": 6.116135550091513e-07,
+      "loss": 0.0218,
+      "step": 11151
+    },
+    {
+      "epoch": 2.5345454545454547,
+      "grad_norm": 2.515082658732752,
+      "learning_rate": 6.115243258284012e-07,
+      "loss": 0.0288,
+      "step": 11152
+    },
+    {
+      "epoch": 2.534772727272727,
+      "grad_norm": 1.6455891487442067,
+      "learning_rate": 6.114350969224439e-07,
+      "loss": 0.0107,
+      "step": 11153
+    },
+    {
+      "epoch": 2.535,
+      "grad_norm": 2.9944217169130707,
+      "learning_rate": 6.113458682930984e-07,
+      "loss": 0.0524,
+      "step": 11154
+    },
+    {
+      "epoch": 2.5352272727272727,
+      "grad_norm": 1.7260583739062134,
+      "learning_rate": 6.112566399421844e-07,
+      "loss": 0.0135,
+      "step": 11155
+    },
+    {
+      "epoch": 2.535454545454545,
+      "grad_norm": 2.1936371416357674,
+      "learning_rate": 6.111674118715215e-07,
+      "loss": 0.0095,
+      "step": 11156
+    },
+    {
+      "epoch": 2.535681818181818,
+      "grad_norm": 1.7704817620190587,
+      "learning_rate": 6.110781840829294e-07,
+      "loss": 0.0419,
+      "step": 11157
+    },
+    {
+      "epoch": 2.535909090909091,
+      "grad_norm": 2.001990680169987,
+      "learning_rate": 6.109889565782273e-07,
+      "loss": 0.0194,
+      "step": 11158
+    },
+    {
+      "epoch": 2.5361363636363636,
+      "grad_norm": 2.083296469146469,
+      "learning_rate": 6.108997293592346e-07,
+      "loss": 0.0189,
+      "step": 11159
+    },
+    {
+      "epoch": 2.536363636363636,
+      "grad_norm": 2.4036235802337997,
+      "learning_rate": 6.108105024277712e-07,
+      "loss": 0.0496,
+      "step": 11160
+    },
+    {
+      "epoch": 2.536590909090909,
+      "grad_norm": 1.5532424532197098,
+      "learning_rate": 6.107212757856564e-07,
+      "loss": 0.0105,
+      "step": 11161
+    },
+    {
+      "epoch": 2.536818181818182,
+      "grad_norm": 3.4577960011243403,
+      "learning_rate": 6.1063204943471e-07,
+      "loss": 0.0627,
+      "step": 11162
+    },
+    {
+      "epoch": 2.5370454545454546,
+      "grad_norm": 1.344840254613391,
+      "learning_rate": 6.105428233767509e-07,
+      "loss": 0.0173,
+      "step": 11163
+    },
+    {
+      "epoch": 2.537272727272727,
+      "grad_norm": 2.5246660379141157,
+      "learning_rate": 6.104535976135989e-07,
+      "loss": 0.0357,
+      "step": 11164
+    },
+    {
+      "epoch": 2.5375,
+      "grad_norm": 1.1877422928167891,
+      "learning_rate": 6.103643721470732e-07,
+      "loss": 0.0206,
+      "step": 11165
+    },
+    {
+      "epoch": 2.5377272727272726,
+      "grad_norm": 1.612643454448893,
+      "learning_rate": 6.102751469789937e-07,
+      "loss": 0.0161,
+      "step": 11166
+    },
+    {
+      "epoch": 2.5379545454545456,
+      "grad_norm": 3.36154911789723,
+      "learning_rate": 6.101859221111797e-07,
+      "loss": 0.0414,
+      "step": 11167
+    },
+    {
+      "epoch": 2.538181818181818,
+      "grad_norm": 1.412941556549793,
+      "learning_rate": 6.100966975454504e-07,
+      "loss": 0.0341,
+      "step": 11168
+    },
+    {
+      "epoch": 2.538409090909091,
+      "grad_norm": 2.1093380092096607,
+      "learning_rate": 6.100074732836256e-07,
+      "loss": 0.0327,
+      "step": 11169
+    },
+    {
+      "epoch": 2.5386363636363636,
+      "grad_norm": 3.4596160198448698,
+      "learning_rate": 6.099182493275245e-07,
+      "loss": 0.086,
+      "step": 11170
+    },
+    {
+      "epoch": 2.5388636363636365,
+      "grad_norm": 1.5859625607332204,
+      "learning_rate": 6.098290256789666e-07,
+      "loss": 0.0168,
+      "step": 11171
+    },
+    {
+      "epoch": 2.539090909090909,
+      "grad_norm": 2.601389449206152,
+      "learning_rate": 6.097398023397714e-07,
+      "loss": 0.025,
+      "step": 11172
+    },
+    {
+      "epoch": 2.5393181818181816,
+      "grad_norm": 3.5002431252108934,
+      "learning_rate": 6.096505793117581e-07,
+      "loss": 0.027,
+      "step": 11173
+    },
+    {
+      "epoch": 2.5395454545454546,
+      "grad_norm": 2.5387509336811673,
+      "learning_rate": 6.095613565967463e-07,
+      "loss": 0.0499,
+      "step": 11174
+    },
+    {
+      "epoch": 2.5397727272727275,
+      "grad_norm": 1.9792677814732953,
+      "learning_rate": 6.094721341965558e-07,
+      "loss": 0.0336,
+      "step": 11175
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.753462284038357,
+      "learning_rate": 6.093829121130052e-07,
+      "loss": 0.0296,
+      "step": 11176
+    },
+    {
+      "epoch": 2.5402272727272726,
+      "grad_norm": 2.3726993304312427,
+      "learning_rate": 6.092936903479141e-07,
+      "loss": 0.0572,
+      "step": 11177
+    },
+    {
+      "epoch": 2.5404545454545455,
+      "grad_norm": 1.6494973304499527,
+      "learning_rate": 6.092044689031023e-07,
+      "loss": 0.025,
+      "step": 11178
+    },
+    {
+      "epoch": 2.540681818181818,
+      "grad_norm": 0.9827718842289213,
+      "learning_rate": 6.09115247780389e-07,
+      "loss": 0.017,
+      "step": 11179
+    },
+    {
+      "epoch": 2.540909090909091,
+      "grad_norm": 2.656171907244583,
+      "learning_rate": 6.090260269815933e-07,
+      "loss": 0.0314,
+      "step": 11180
+    },
+    {
+      "epoch": 2.5411363636363635,
+      "grad_norm": 1.9126907970950426,
+      "learning_rate": 6.089368065085349e-07,
+      "loss": 0.0402,
+      "step": 11181
+    },
+    {
+      "epoch": 2.5413636363636365,
+      "grad_norm": 2.485112833526169,
+      "learning_rate": 6.088475863630328e-07,
+      "loss": 0.0415,
+      "step": 11182
+    },
+    {
+      "epoch": 2.541590909090909,
+      "grad_norm": 1.8281603745781514,
+      "learning_rate": 6.087583665469069e-07,
+      "loss": 0.0304,
+      "step": 11183
+    },
+    {
+      "epoch": 2.541818181818182,
+      "grad_norm": 1.6002580859217175,
+      "learning_rate": 6.086691470619763e-07,
+      "loss": 0.0186,
+      "step": 11184
+    },
+    {
+      "epoch": 2.5420454545454545,
+      "grad_norm": 1.911204731764562,
+      "learning_rate": 6.0857992791006e-07,
+      "loss": 0.0239,
+      "step": 11185
+    },
+    {
+      "epoch": 2.542272727272727,
+      "grad_norm": 1.7299453694344968,
+      "learning_rate": 6.084907090929778e-07,
+      "loss": 0.0274,
+      "step": 11186
+    },
+    {
+      "epoch": 2.5425,
+      "grad_norm": 1.3555323553605563,
+      "learning_rate": 6.084014906125488e-07,
+      "loss": 0.0261,
+      "step": 11187
+    },
+    {
+      "epoch": 2.542727272727273,
+      "grad_norm": 3.741228632157337,
+      "learning_rate": 6.083122724705925e-07,
+      "loss": 0.0332,
+      "step": 11188
+    },
+    {
+      "epoch": 2.5429545454545455,
+      "grad_norm": 2.162061258968342,
+      "learning_rate": 6.082230546689278e-07,
+      "loss": 0.0194,
+      "step": 11189
+    },
+    {
+      "epoch": 2.543181818181818,
+      "grad_norm": 1.7189856981938547,
+      "learning_rate": 6.081338372093743e-07,
+      "loss": 0.0293,
+      "step": 11190
+    },
+    {
+      "epoch": 2.543409090909091,
+      "grad_norm": 2.60923243964131,
+      "learning_rate": 6.080446200937514e-07,
+      "loss": 0.0575,
+      "step": 11191
+    },
+    {
+      "epoch": 2.5436363636363635,
+      "grad_norm": 1.2722591901698297,
+      "learning_rate": 6.079554033238782e-07,
+      "loss": 0.0152,
+      "step": 11192
+    },
+    {
+      "epoch": 2.5438636363636364,
+      "grad_norm": 1.12070378726027,
+      "learning_rate": 6.078661869015744e-07,
+      "loss": 0.0126,
+      "step": 11193
+    },
+    {
+      "epoch": 2.544090909090909,
+      "grad_norm": 1.4165173117684482,
+      "learning_rate": 6.077769708286586e-07,
+      "loss": 0.019,
+      "step": 11194
+    },
+    {
+      "epoch": 2.544318181818182,
+      "grad_norm": 2.230578952909692,
+      "learning_rate": 6.076877551069505e-07,
+      "loss": 0.0254,
+      "step": 11195
+    },
+    {
+      "epoch": 2.5445454545454544,
+      "grad_norm": 1.3974367850722789,
+      "learning_rate": 6.075985397382691e-07,
+      "loss": 0.0171,
+      "step": 11196
+    },
+    {
+      "epoch": 2.5447727272727274,
+      "grad_norm": 3.0034950092811803,
+      "learning_rate": 6.075093247244342e-07,
+      "loss": 0.0529,
+      "step": 11197
+    },
+    {
+      "epoch": 2.545,
+      "grad_norm": 3.0227324439384797,
+      "learning_rate": 6.074201100672645e-07,
+      "loss": 0.0278,
+      "step": 11198
+    },
+    {
+      "epoch": 2.545227272727273,
+      "grad_norm": 1.2267719266275539,
+      "learning_rate": 6.073308957685793e-07,
+      "loss": 0.0243,
+      "step": 11199
+    },
+    {
+      "epoch": 2.5454545454545454,
+      "grad_norm": 3.4133460004532195,
+      "learning_rate": 6.07241681830198e-07,
+      "loss": 0.0397,
+      "step": 11200
+    },
+    {
+      "epoch": 2.5456818181818184,
+      "grad_norm": 1.2144687760278325,
+      "learning_rate": 6.0715246825394e-07,
+      "loss": 0.0169,
+      "step": 11201
+    },
+    {
+      "epoch": 2.545909090909091,
+      "grad_norm": 1.8772239310372631,
+      "learning_rate": 6.070632550416241e-07,
+      "loss": 0.0178,
+      "step": 11202
+    },
+    {
+      "epoch": 2.5461363636363634,
+      "grad_norm": 2.0831738650931775,
+      "learning_rate": 6.069740421950697e-07,
+      "loss": 0.0358,
+      "step": 11203
+    },
+    {
+      "epoch": 2.5463636363636364,
+      "grad_norm": 1.4861358459437737,
+      "learning_rate": 6.068848297160961e-07,
+      "loss": 0.0286,
+      "step": 11204
+    },
+    {
+      "epoch": 2.5465909090909093,
+      "grad_norm": 1.6737919532726584,
+      "learning_rate": 6.067956176065224e-07,
+      "loss": 0.0273,
+      "step": 11205
+    },
+    {
+      "epoch": 2.546818181818182,
+      "grad_norm": 1.4663442047190396,
+      "learning_rate": 6.067064058681681e-07,
+      "loss": 0.0154,
+      "step": 11206
+    },
+    {
+      "epoch": 2.5470454545454544,
+      "grad_norm": 0.8851213416316988,
+      "learning_rate": 6.066171945028518e-07,
+      "loss": 0.0114,
+      "step": 11207
+    },
+    {
+      "epoch": 2.5472727272727274,
+      "grad_norm": 2.571275264502912,
+      "learning_rate": 6.06527983512393e-07,
+      "loss": 0.0314,
+      "step": 11208
+    },
+    {
+      "epoch": 2.5475,
+      "grad_norm": 0.9987553212838233,
+      "learning_rate": 6.064387728986109e-07,
+      "loss": 0.011,
+      "step": 11209
+    },
+    {
+      "epoch": 2.547727272727273,
+      "grad_norm": 2.3238375965450615,
+      "learning_rate": 6.063495626633245e-07,
+      "loss": 0.0396,
+      "step": 11210
+    },
+    {
+      "epoch": 2.5479545454545454,
+      "grad_norm": 2.2335421091554886,
+      "learning_rate": 6.062603528083532e-07,
+      "loss": 0.0197,
+      "step": 11211
+    },
+    {
+      "epoch": 2.5481818181818183,
+      "grad_norm": 1.4639588168466546,
+      "learning_rate": 6.061711433355158e-07,
+      "loss": 0.0126,
+      "step": 11212
+    },
+    {
+      "epoch": 2.548409090909091,
+      "grad_norm": 2.7825604293579365,
+      "learning_rate": 6.060819342466318e-07,
+      "loss": 0.0573,
+      "step": 11213
+    },
+    {
+      "epoch": 2.548636363636364,
+      "grad_norm": 3.3671669625116016,
+      "learning_rate": 6.0599272554352e-07,
+      "loss": 0.0454,
+      "step": 11214
+    },
+    {
+      "epoch": 2.5488636363636363,
+      "grad_norm": 2.8202003012813037,
+      "learning_rate": 6.059035172280001e-07,
+      "loss": 0.0307,
+      "step": 11215
+    },
+    {
+      "epoch": 2.549090909090909,
+      "grad_norm": 1.5021763937550394,
+      "learning_rate": 6.058143093018905e-07,
+      "loss": 0.0238,
+      "step": 11216
+    },
+    {
+      "epoch": 2.549318181818182,
+      "grad_norm": 0.5584982735333354,
+      "learning_rate": 6.057251017670105e-07,
+      "loss": 0.0034,
+      "step": 11217
+    },
+    {
+      "epoch": 2.549545454545455,
+      "grad_norm": 2.819173242066571,
+      "learning_rate": 6.056358946251795e-07,
+      "loss": 0.0634,
+      "step": 11218
+    },
+    {
+      "epoch": 2.5497727272727273,
+      "grad_norm": 1.874478467578541,
+      "learning_rate": 6.055466878782164e-07,
+      "loss": 0.0292,
+      "step": 11219
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 2.669065098047097,
+      "learning_rate": 6.054574815279401e-07,
+      "loss": 0.03,
+      "step": 11220
+    },
+    {
+      "epoch": 2.550227272727273,
+      "grad_norm": 1.7559654296326324,
+      "learning_rate": 6.053682755761699e-07,
+      "loss": 0.0255,
+      "step": 11221
+    },
+    {
+      "epoch": 2.5504545454545453,
+      "grad_norm": 2.2717833095197646,
+      "learning_rate": 6.052790700247247e-07,
+      "loss": 0.0315,
+      "step": 11222
+    },
+    {
+      "epoch": 2.5506818181818183,
+      "grad_norm": 1.2510228531704985,
+      "learning_rate": 6.051898648754241e-07,
+      "loss": 0.0116,
+      "step": 11223
+    },
+    {
+      "epoch": 2.550909090909091,
+      "grad_norm": 2.7518368469204777,
+      "learning_rate": 6.051006601300865e-07,
+      "loss": 0.0297,
+      "step": 11224
+    },
+    {
+      "epoch": 2.5511363636363638,
+      "grad_norm": 1.0783524268105462,
+      "learning_rate": 6.05011455790531e-07,
+      "loss": 0.0121,
+      "step": 11225
+    },
+    {
+      "epoch": 2.5513636363636363,
+      "grad_norm": 3.184030352273646,
+      "learning_rate": 6.049222518585767e-07,
+      "loss": 0.0378,
+      "step": 11226
+    },
+    {
+      "epoch": 2.5515909090909092,
+      "grad_norm": 1.1189186989989046,
+      "learning_rate": 6.048330483360428e-07,
+      "loss": 0.0114,
+      "step": 11227
+    },
+    {
+      "epoch": 2.5518181818181818,
+      "grad_norm": 0.9235944980225251,
+      "learning_rate": 6.047438452247486e-07,
+      "loss": 0.0116,
+      "step": 11228
+    },
+    {
+      "epoch": 2.5520454545454543,
+      "grad_norm": 1.0116424045507681,
+      "learning_rate": 6.046546425265123e-07,
+      "loss": 0.0202,
+      "step": 11229
+    },
+    {
+      "epoch": 2.5522727272727272,
+      "grad_norm": 2.6007707868249588,
+      "learning_rate": 6.045654402431534e-07,
+      "loss": 0.015,
+      "step": 11230
+    },
+    {
+      "epoch": 2.5525,
+      "grad_norm": 1.7505957469589952,
+      "learning_rate": 6.044762383764909e-07,
+      "loss": 0.0207,
+      "step": 11231
+    },
+    {
+      "epoch": 2.5527272727272727,
+      "grad_norm": 2.277662987566612,
+      "learning_rate": 6.043870369283438e-07,
+      "loss": 0.0511,
+      "step": 11232
+    },
+    {
+      "epoch": 2.5529545454545453,
+      "grad_norm": 1.5130191600823428,
+      "learning_rate": 6.042978359005308e-07,
+      "loss": 0.0166,
+      "step": 11233
+    },
+    {
+      "epoch": 2.553181818181818,
+      "grad_norm": 1.287522421532347,
+      "learning_rate": 6.042086352948711e-07,
+      "loss": 0.0138,
+      "step": 11234
+    },
+    {
+      "epoch": 2.553409090909091,
+      "grad_norm": 1.227286241625068,
+      "learning_rate": 6.041194351131837e-07,
+      "loss": 0.0166,
+      "step": 11235
+    },
+    {
+      "epoch": 2.5536363636363637,
+      "grad_norm": 1.669042634276714,
+      "learning_rate": 6.040302353572872e-07,
+      "loss": 0.0166,
+      "step": 11236
+    },
+    {
+      "epoch": 2.5538636363636362,
+      "grad_norm": 3.004522746200978,
+      "learning_rate": 6.039410360290012e-07,
+      "loss": 0.0116,
+      "step": 11237
+    },
+    {
+      "epoch": 2.554090909090909,
+      "grad_norm": 1.5593299423173217,
+      "learning_rate": 6.03851837130144e-07,
+      "loss": 0.0323,
+      "step": 11238
+    },
+    {
+      "epoch": 2.5543181818181817,
+      "grad_norm": 1.9847943672737687,
+      "learning_rate": 6.037626386625346e-07,
+      "loss": 0.0223,
+      "step": 11239
+    },
+    {
+      "epoch": 2.5545454545454547,
+      "grad_norm": 1.4908494491906792,
+      "learning_rate": 6.036734406279921e-07,
+      "loss": 0.0201,
+      "step": 11240
+    },
+    {
+      "epoch": 2.554772727272727,
+      "grad_norm": 1.4315456420868373,
+      "learning_rate": 6.035842430283355e-07,
+      "loss": 0.0114,
+      "step": 11241
+    },
+    {
+      "epoch": 2.555,
+      "grad_norm": 2.5960828460205194,
+      "learning_rate": 6.034950458653834e-07,
+      "loss": 0.0307,
+      "step": 11242
+    },
+    {
+      "epoch": 2.5552272727272727,
+      "grad_norm": 1.2179607701508652,
+      "learning_rate": 6.034058491409549e-07,
+      "loss": 0.0183,
+      "step": 11243
+    },
+    {
+      "epoch": 2.5554545454545456,
+      "grad_norm": 3.0953459966537262,
+      "learning_rate": 6.033166528568688e-07,
+      "loss": 0.0325,
+      "step": 11244
+    },
+    {
+      "epoch": 2.555681818181818,
+      "grad_norm": 2.1995492691393115,
+      "learning_rate": 6.032274570149442e-07,
+      "loss": 0.0305,
+      "step": 11245
+    },
+    {
+      "epoch": 2.5559090909090907,
+      "grad_norm": 2.3490363638017193,
+      "learning_rate": 6.031382616169994e-07,
+      "loss": 0.0249,
+      "step": 11246
+    },
+    {
+      "epoch": 2.5561363636363637,
+      "grad_norm": 5.803983726705048,
+      "learning_rate": 6.030490666648536e-07,
+      "loss": 0.0567,
+      "step": 11247
+    },
+    {
+      "epoch": 2.5563636363636366,
+      "grad_norm": 2.629118202687729,
+      "learning_rate": 6.029598721603258e-07,
+      "loss": 0.0206,
+      "step": 11248
+    },
+    {
+      "epoch": 2.556590909090909,
+      "grad_norm": 2.143654044058778,
+      "learning_rate": 6.028706781052344e-07,
+      "loss": 0.0196,
+      "step": 11249
+    },
+    {
+      "epoch": 2.5568181818181817,
+      "grad_norm": 2.195055178775719,
+      "learning_rate": 6.027814845013988e-07,
+      "loss": 0.0277,
+      "step": 11250
+    },
+    {
+      "epoch": 2.5570454545454546,
+      "grad_norm": 2.0362859751649904,
+      "learning_rate": 6.026922913506373e-07,
+      "loss": 0.027,
+      "step": 11251
+    },
+    {
+      "epoch": 2.557272727272727,
+      "grad_norm": 2.008444587531715,
+      "learning_rate": 6.026030986547689e-07,
+      "loss": 0.0214,
+      "step": 11252
+    },
+    {
+      "epoch": 2.5575,
+      "grad_norm": 2.1757764084013718,
+      "learning_rate": 6.025139064156124e-07,
+      "loss": 0.0286,
+      "step": 11253
+    },
+    {
+      "epoch": 2.5577272727272726,
+      "grad_norm": 2.1914674977957147,
+      "learning_rate": 6.024247146349868e-07,
+      "loss": 0.0416,
+      "step": 11254
+    },
+    {
+      "epoch": 2.5579545454545456,
+      "grad_norm": 1.3019000189240015,
+      "learning_rate": 6.023355233147104e-07,
+      "loss": 0.0187,
+      "step": 11255
+    },
+    {
+      "epoch": 2.558181818181818,
+      "grad_norm": 1.7083898385641048,
+      "learning_rate": 6.022463324566022e-07,
+      "loss": 0.0192,
+      "step": 11256
+    },
+    {
+      "epoch": 2.558409090909091,
+      "grad_norm": 1.622690601379426,
+      "learning_rate": 6.021571420624811e-07,
+      "loss": 0.0111,
+      "step": 11257
+    },
+    {
+      "epoch": 2.5586363636363636,
+      "grad_norm": 2.6731089848162584,
+      "learning_rate": 6.020679521341659e-07,
+      "loss": 0.0281,
+      "step": 11258
+    },
+    {
+      "epoch": 2.558863636363636,
+      "grad_norm": 0.8342395927470239,
+      "learning_rate": 6.019787626734749e-07,
+      "loss": 0.0119,
+      "step": 11259
+    },
+    {
+      "epoch": 2.559090909090909,
+      "grad_norm": 1.9950793997091683,
+      "learning_rate": 6.018895736822271e-07,
+      "loss": 0.017,
+      "step": 11260
+    },
+    {
+      "epoch": 2.559318181818182,
+      "grad_norm": 2.28177957518864,
+      "learning_rate": 6.018003851622411e-07,
+      "loss": 0.0198,
+      "step": 11261
+    },
+    {
+      "epoch": 2.5595454545454546,
+      "grad_norm": 0.7822669982332394,
+      "learning_rate": 6.017111971153359e-07,
+      "loss": 0.0066,
+      "step": 11262
+    },
+    {
+      "epoch": 2.559772727272727,
+      "grad_norm": 1.7319882445523076,
+      "learning_rate": 6.016220095433301e-07,
+      "loss": 0.0292,
+      "step": 11263
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 2.8276955483822026,
+      "learning_rate": 6.015328224480422e-07,
+      "loss": 0.0364,
+      "step": 11264
+    },
+    {
+      "epoch": 2.5602272727272726,
+      "grad_norm": 1.4622984912939168,
+      "learning_rate": 6.014436358312909e-07,
+      "loss": 0.0168,
+      "step": 11265
+    },
+    {
+      "epoch": 2.5604545454545455,
+      "grad_norm": 2.0078002484079986,
+      "learning_rate": 6.01354449694895e-07,
+      "loss": 0.0239,
+      "step": 11266
+    },
+    {
+      "epoch": 2.560681818181818,
+      "grad_norm": 1.1795285141605214,
+      "learning_rate": 6.012652640406734e-07,
+      "loss": 0.0086,
+      "step": 11267
+    },
+    {
+      "epoch": 2.560909090909091,
+      "grad_norm": 2.0441673115707357,
+      "learning_rate": 6.011760788704443e-07,
+      "loss": 0.0195,
+      "step": 11268
+    },
+    {
+      "epoch": 2.5611363636363635,
+      "grad_norm": 1.5537315401107876,
+      "learning_rate": 6.010868941860264e-07,
+      "loss": 0.0153,
+      "step": 11269
+    },
+    {
+      "epoch": 2.5613636363636365,
+      "grad_norm": 2.2562061331852066,
+      "learning_rate": 6.009977099892384e-07,
+      "loss": 0.0235,
+      "step": 11270
+    },
+    {
+      "epoch": 2.561590909090909,
+      "grad_norm": 0.8685618693610683,
+      "learning_rate": 6.009085262818992e-07,
+      "loss": 0.0106,
+      "step": 11271
+    },
+    {
+      "epoch": 2.5618181818181816,
+      "grad_norm": 1.9717830987461014,
+      "learning_rate": 6.008193430658272e-07,
+      "loss": 0.0256,
+      "step": 11272
+    },
+    {
+      "epoch": 2.5620454545454545,
+      "grad_norm": 1.8714202987331745,
+      "learning_rate": 6.007301603428408e-07,
+      "loss": 0.0152,
+      "step": 11273
+    },
+    {
+      "epoch": 2.5622727272727275,
+      "grad_norm": 1.1065425179714714,
+      "learning_rate": 6.006409781147587e-07,
+      "loss": 0.0146,
+      "step": 11274
+    },
+    {
+      "epoch": 2.5625,
+      "grad_norm": 1.4057903958185163,
+      "learning_rate": 6.005517963833996e-07,
+      "loss": 0.0128,
+      "step": 11275
+    },
+    {
+      "epoch": 2.5627272727272725,
+      "grad_norm": 0.9334240851638629,
+      "learning_rate": 6.004626151505824e-07,
+      "loss": 0.0067,
+      "step": 11276
+    },
+    {
+      "epoch": 2.5629545454545455,
+      "grad_norm": 1.4755155117922842,
+      "learning_rate": 6.003734344181249e-07,
+      "loss": 0.019,
+      "step": 11277
+    },
+    {
+      "epoch": 2.5631818181818184,
+      "grad_norm": 1.0249028148240822,
+      "learning_rate": 6.00284254187846e-07,
+      "loss": 0.0113,
+      "step": 11278
+    },
+    {
+      "epoch": 2.563409090909091,
+      "grad_norm": 1.9277616868107912,
+      "learning_rate": 6.001950744615645e-07,
+      "loss": 0.0236,
+      "step": 11279
+    },
+    {
+      "epoch": 2.5636363636363635,
+      "grad_norm": 2.0672988434985777,
+      "learning_rate": 6.001058952410986e-07,
+      "loss": 0.022,
+      "step": 11280
+    },
+    {
+      "epoch": 2.5638636363636365,
+      "grad_norm": 2.282742916468752,
+      "learning_rate": 6.000167165282669e-07,
+      "loss": 0.0221,
+      "step": 11281
+    },
+    {
+      "epoch": 2.564090909090909,
+      "grad_norm": 2.663859050115134,
+      "learning_rate": 5.999275383248879e-07,
+      "loss": 0.023,
+      "step": 11282
+    },
+    {
+      "epoch": 2.564318181818182,
+      "grad_norm": 2.3290369303131744,
+      "learning_rate": 5.998383606327801e-07,
+      "loss": 0.0287,
+      "step": 11283
+    },
+    {
+      "epoch": 2.5645454545454545,
+      "grad_norm": 1.6373399678272096,
+      "learning_rate": 5.997491834537622e-07,
+      "loss": 0.0208,
+      "step": 11284
+    },
+    {
+      "epoch": 2.5647727272727274,
+      "grad_norm": 2.5584562342753885,
+      "learning_rate": 5.996600067896524e-07,
+      "loss": 0.02,
+      "step": 11285
+    },
+    {
+      "epoch": 2.565,
+      "grad_norm": 3.9137565292216268,
+      "learning_rate": 5.995708306422693e-07,
+      "loss": 0.0605,
+      "step": 11286
+    },
+    {
+      "epoch": 2.565227272727273,
+      "grad_norm": 3.7984714255901104,
+      "learning_rate": 5.994816550134313e-07,
+      "loss": 0.0313,
+      "step": 11287
+    },
+    {
+      "epoch": 2.5654545454545454,
+      "grad_norm": 2.901882762332469,
+      "learning_rate": 5.993924799049567e-07,
+      "loss": 0.0329,
+      "step": 11288
+    },
+    {
+      "epoch": 2.565681818181818,
+      "grad_norm": 1.3555920976165259,
+      "learning_rate": 5.993033053186646e-07,
+      "loss": 0.0137,
+      "step": 11289
+    },
+    {
+      "epoch": 2.565909090909091,
+      "grad_norm": 1.5127829870939589,
+      "learning_rate": 5.992141312563727e-07,
+      "loss": 0.0363,
+      "step": 11290
+    },
+    {
+      "epoch": 2.566136363636364,
+      "grad_norm": 2.002721261100375,
+      "learning_rate": 5.991249577198995e-07,
+      "loss": 0.0421,
+      "step": 11291
+    },
+    {
+      "epoch": 2.5663636363636364,
+      "grad_norm": 2.061913141839439,
+      "learning_rate": 5.990357847110636e-07,
+      "loss": 0.0247,
+      "step": 11292
+    },
+    {
+      "epoch": 2.566590909090909,
+      "grad_norm": 2.0491743476594175,
+      "learning_rate": 5.989466122316833e-07,
+      "loss": 0.0322,
+      "step": 11293
+    },
+    {
+      "epoch": 2.566818181818182,
+      "grad_norm": 0.8411784751707322,
+      "learning_rate": 5.988574402835772e-07,
+      "loss": 0.0079,
+      "step": 11294
+    },
+    {
+      "epoch": 2.5670454545454544,
+      "grad_norm": 1.559817498695652,
+      "learning_rate": 5.987682688685633e-07,
+      "loss": 0.0307,
+      "step": 11295
+    },
+    {
+      "epoch": 2.5672727272727274,
+      "grad_norm": 0.8843296409249856,
+      "learning_rate": 5.986790979884603e-07,
+      "loss": 0.0102,
+      "step": 11296
+    },
+    {
+      "epoch": 2.5675,
+      "grad_norm": 6.284665721698099,
+      "learning_rate": 5.985899276450864e-07,
+      "loss": 0.0917,
+      "step": 11297
+    },
+    {
+      "epoch": 2.567727272727273,
+      "grad_norm": 2.137427011668303,
+      "learning_rate": 5.985007578402601e-07,
+      "loss": 0.0194,
+      "step": 11298
+    },
+    {
+      "epoch": 2.5679545454545454,
+      "grad_norm": 2.123023451476279,
+      "learning_rate": 5.984115885757994e-07,
+      "loss": 0.0268,
+      "step": 11299
+    },
+    {
+      "epoch": 2.5681818181818183,
+      "grad_norm": 2.0661069779787344,
+      "learning_rate": 5.983224198535228e-07,
+      "loss": 0.0156,
+      "step": 11300
+    },
+    {
+      "epoch": 2.568409090909091,
+      "grad_norm": 2.607112840003923,
+      "learning_rate": 5.982332516752486e-07,
+      "loss": 0.0362,
+      "step": 11301
+    },
+    {
+      "epoch": 2.5686363636363634,
+      "grad_norm": 1.721378901955717,
+      "learning_rate": 5.981440840427951e-07,
+      "loss": 0.026,
+      "step": 11302
+    },
+    {
+      "epoch": 2.5688636363636363,
+      "grad_norm": 1.4590553299852929,
+      "learning_rate": 5.980549169579806e-07,
+      "loss": 0.02,
+      "step": 11303
+    },
+    {
+      "epoch": 2.5690909090909093,
+      "grad_norm": 3.745372730873593,
+      "learning_rate": 5.979657504226233e-07,
+      "loss": 0.0474,
+      "step": 11304
+    },
+    {
+      "epoch": 2.569318181818182,
+      "grad_norm": 1.329085377117397,
+      "learning_rate": 5.978765844385416e-07,
+      "loss": 0.0154,
+      "step": 11305
+    },
+    {
+      "epoch": 2.5695454545454544,
+      "grad_norm": 1.6689642305623333,
+      "learning_rate": 5.977874190075534e-07,
+      "loss": 0.019,
+      "step": 11306
+    },
+    {
+      "epoch": 2.5697727272727273,
+      "grad_norm": 1.9764181018673512,
+      "learning_rate": 5.976982541314776e-07,
+      "loss": 0.0192,
+      "step": 11307
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.825046945294248,
+      "learning_rate": 5.97609089812132e-07,
+      "loss": 0.0193,
+      "step": 11308
+    },
+    {
+      "epoch": 2.570227272727273,
+      "grad_norm": 2.3494927360068787,
+      "learning_rate": 5.975199260513346e-07,
+      "loss": 0.0358,
+      "step": 11309
+    },
+    {
+      "epoch": 2.5704545454545453,
+      "grad_norm": 1.4114847862530082,
+      "learning_rate": 5.974307628509038e-07,
+      "loss": 0.0236,
+      "step": 11310
+    },
+    {
+      "epoch": 2.5706818181818183,
+      "grad_norm": 1.3470406488017057,
+      "learning_rate": 5.973416002126581e-07,
+      "loss": 0.0169,
+      "step": 11311
+    },
+    {
+      "epoch": 2.570909090909091,
+      "grad_norm": 1.286122390130578,
+      "learning_rate": 5.972524381384153e-07,
+      "loss": 0.0077,
+      "step": 11312
+    },
+    {
+      "epoch": 2.5711363636363638,
+      "grad_norm": 1.2528809056777668,
+      "learning_rate": 5.971632766299937e-07,
+      "loss": 0.0118,
+      "step": 11313
+    },
+    {
+      "epoch": 2.5713636363636363,
+      "grad_norm": 1.697399552286226,
+      "learning_rate": 5.970741156892114e-07,
+      "loss": 0.0337,
+      "step": 11314
+    },
+    {
+      "epoch": 2.571590909090909,
+      "grad_norm": 2.793468925896672,
+      "learning_rate": 5.969849553178866e-07,
+      "loss": 0.0628,
+      "step": 11315
+    },
+    {
+      "epoch": 2.571818181818182,
+      "grad_norm": 1.9622818483909124,
+      "learning_rate": 5.968957955178376e-07,
+      "loss": 0.0311,
+      "step": 11316
+    },
+    {
+      "epoch": 2.5720454545454547,
+      "grad_norm": 1.6481852265985697,
+      "learning_rate": 5.968066362908822e-07,
+      "loss": 0.0204,
+      "step": 11317
+    },
+    {
+      "epoch": 2.5722727272727273,
+      "grad_norm": 3.365312384773325,
+      "learning_rate": 5.967174776388386e-07,
+      "loss": 0.0466,
+      "step": 11318
+    },
+    {
+      "epoch": 2.5725,
+      "grad_norm": 1.6118503632016825,
+      "learning_rate": 5.96628319563525e-07,
+      "loss": 0.01,
+      "step": 11319
+    },
+    {
+      "epoch": 2.5727272727272728,
+      "grad_norm": 2.104824684943451,
+      "learning_rate": 5.965391620667599e-07,
+      "loss": 0.0256,
+      "step": 11320
+    },
+    {
+      "epoch": 2.5729545454545457,
+      "grad_norm": 2.9723552842022687,
+      "learning_rate": 5.964500051503605e-07,
+      "loss": 0.0411,
+      "step": 11321
+    },
+    {
+      "epoch": 2.5731818181818182,
+      "grad_norm": 1.4800339801633522,
+      "learning_rate": 5.963608488161454e-07,
+      "loss": 0.0236,
+      "step": 11322
+    },
+    {
+      "epoch": 2.5734090909090908,
+      "grad_norm": 1.2185104147819392,
+      "learning_rate": 5.962716930659325e-07,
+      "loss": 0.0199,
+      "step": 11323
+    },
+    {
+      "epoch": 2.5736363636363637,
+      "grad_norm": 1.1887338572761774,
+      "learning_rate": 5.9618253790154e-07,
+      "loss": 0.0128,
+      "step": 11324
+    },
+    {
+      "epoch": 2.5738636363636362,
+      "grad_norm": 1.3593301162372902,
+      "learning_rate": 5.960933833247857e-07,
+      "loss": 0.0282,
+      "step": 11325
+    },
+    {
+      "epoch": 2.574090909090909,
+      "grad_norm": 1.7649320365445815,
+      "learning_rate": 5.960042293374879e-07,
+      "loss": 0.0152,
+      "step": 11326
+    },
+    {
+      "epoch": 2.5743181818181817,
+      "grad_norm": 1.8588672409128748,
+      "learning_rate": 5.959150759414642e-07,
+      "loss": 0.0283,
+      "step": 11327
+    },
+    {
+      "epoch": 2.5745454545454547,
+      "grad_norm": 1.0890240439929333,
+      "learning_rate": 5.958259231385331e-07,
+      "loss": 0.0096,
+      "step": 11328
+    },
+    {
+      "epoch": 2.574772727272727,
+      "grad_norm": 1.878244353347734,
+      "learning_rate": 5.957367709305124e-07,
+      "loss": 0.0215,
+      "step": 11329
+    },
+    {
+      "epoch": 2.575,
+      "grad_norm": 1.4271993973567834,
+      "learning_rate": 5.956476193192198e-07,
+      "loss": 0.0106,
+      "step": 11330
+    },
+    {
+      "epoch": 2.5752272727272727,
+      "grad_norm": 2.8914186081788813,
+      "learning_rate": 5.955584683064734e-07,
+      "loss": 0.0273,
+      "step": 11331
+    },
+    {
+      "epoch": 2.575454545454545,
+      "grad_norm": 2.305159114901488,
+      "learning_rate": 5.954693178940913e-07,
+      "loss": 0.0377,
+      "step": 11332
+    },
+    {
+      "epoch": 2.575681818181818,
+      "grad_norm": 1.4186707287219569,
+      "learning_rate": 5.953801680838914e-07,
+      "loss": 0.0199,
+      "step": 11333
+    },
+    {
+      "epoch": 2.575909090909091,
+      "grad_norm": 2.3124610425085255,
+      "learning_rate": 5.952910188776914e-07,
+      "loss": 0.0251,
+      "step": 11334
+    },
+    {
+      "epoch": 2.5761363636363637,
+      "grad_norm": 1.6645995892677996,
+      "learning_rate": 5.952018702773093e-07,
+      "loss": 0.03,
+      "step": 11335
+    },
+    {
+      "epoch": 2.576363636363636,
+      "grad_norm": 3.565156215686591,
+      "learning_rate": 5.951127222845632e-07,
+      "loss": 0.0739,
+      "step": 11336
+    },
+    {
+      "epoch": 2.576590909090909,
+      "grad_norm": 1.7839214557680432,
+      "learning_rate": 5.950235749012708e-07,
+      "loss": 0.0183,
+      "step": 11337
+    },
+    {
+      "epoch": 2.5768181818181817,
+      "grad_norm": 2.091678174247532,
+      "learning_rate": 5.949344281292502e-07,
+      "loss": 0.0192,
+      "step": 11338
+    },
+    {
+      "epoch": 2.5770454545454546,
+      "grad_norm": 1.621725628588719,
+      "learning_rate": 5.948452819703186e-07,
+      "loss": 0.0268,
+      "step": 11339
+    },
+    {
+      "epoch": 2.577272727272727,
+      "grad_norm": 3.309796965035904,
+      "learning_rate": 5.947561364262946e-07,
+      "loss": 0.0396,
+      "step": 11340
+    },
+    {
+      "epoch": 2.5775,
+      "grad_norm": 3.5046927251362656,
+      "learning_rate": 5.946669914989956e-07,
+      "loss": 0.06,
+      "step": 11341
+    },
+    {
+      "epoch": 2.5777272727272726,
+      "grad_norm": 1.9748185310463424,
+      "learning_rate": 5.945778471902399e-07,
+      "loss": 0.0238,
+      "step": 11342
+    },
+    {
+      "epoch": 2.5779545454545456,
+      "grad_norm": 3.233450543140924,
+      "learning_rate": 5.944887035018447e-07,
+      "loss": 0.0344,
+      "step": 11343
+    },
+    {
+      "epoch": 2.578181818181818,
+      "grad_norm": 1.8579161623056408,
+      "learning_rate": 5.943995604356281e-07,
+      "loss": 0.042,
+      "step": 11344
+    },
+    {
+      "epoch": 2.5784090909090907,
+      "grad_norm": 3.134003973526431,
+      "learning_rate": 5.943104179934077e-07,
+      "loss": 0.0272,
+      "step": 11345
+    },
+    {
+      "epoch": 2.5786363636363636,
+      "grad_norm": 1.4278008077519797,
+      "learning_rate": 5.942212761770018e-07,
+      "loss": 0.0193,
+      "step": 11346
+    },
+    {
+      "epoch": 2.5788636363636366,
+      "grad_norm": 1.28137142291755,
+      "learning_rate": 5.941321349882275e-07,
+      "loss": 0.0132,
+      "step": 11347
+    },
+    {
+      "epoch": 2.579090909090909,
+      "grad_norm": 2.128000797893107,
+      "learning_rate": 5.940429944289028e-07,
+      "loss": 0.0193,
+      "step": 11348
+    },
+    {
+      "epoch": 2.5793181818181816,
+      "grad_norm": 1.809460326026823,
+      "learning_rate": 5.939538545008454e-07,
+      "loss": 0.0195,
+      "step": 11349
+    },
+    {
+      "epoch": 2.5795454545454546,
+      "grad_norm": 2.563377443258387,
+      "learning_rate": 5.938647152058732e-07,
+      "loss": 0.0259,
+      "step": 11350
+    },
+    {
+      "epoch": 2.579772727272727,
+      "grad_norm": 1.199382376252082,
+      "learning_rate": 5.93775576545804e-07,
+      "loss": 0.0075,
+      "step": 11351
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.280645126642147,
+      "learning_rate": 5.93686438522455e-07,
+      "loss": 0.0366,
+      "step": 11352
+    },
+    {
+      "epoch": 2.5802272727272726,
+      "grad_norm": 2.1954181117514846,
+      "learning_rate": 5.935973011376441e-07,
+      "loss": 0.0438,
+      "step": 11353
+    },
+    {
+      "epoch": 2.5804545454545456,
+      "grad_norm": 1.3980394360015518,
+      "learning_rate": 5.935081643931891e-07,
+      "loss": 0.0145,
+      "step": 11354
+    },
+    {
+      "epoch": 2.580681818181818,
+      "grad_norm": 0.7243829052826335,
+      "learning_rate": 5.934190282909076e-07,
+      "loss": 0.0165,
+      "step": 11355
+    },
+    {
+      "epoch": 2.580909090909091,
+      "grad_norm": 1.3791904994341213,
+      "learning_rate": 5.933298928326171e-07,
+      "loss": 0.0119,
+      "step": 11356
+    },
+    {
+      "epoch": 2.5811363636363636,
+      "grad_norm": 1.850408110272767,
+      "learning_rate": 5.932407580201354e-07,
+      "loss": 0.0273,
+      "step": 11357
+    },
+    {
+      "epoch": 2.5813636363636365,
+      "grad_norm": 1.91172824581042,
+      "learning_rate": 5.9315162385528e-07,
+      "loss": 0.0209,
+      "step": 11358
+    },
+    {
+      "epoch": 2.581590909090909,
+      "grad_norm": 1.6735949989713712,
+      "learning_rate": 5.930624903398686e-07,
+      "loss": 0.0313,
+      "step": 11359
+    },
+    {
+      "epoch": 2.581818181818182,
+      "grad_norm": 2.250965968754418,
+      "learning_rate": 5.92973357475719e-07,
+      "loss": 0.0428,
+      "step": 11360
+    },
+    {
+      "epoch": 2.5820454545454545,
+      "grad_norm": 1.0920733486682481,
+      "learning_rate": 5.928842252646482e-07,
+      "loss": 0.0088,
+      "step": 11361
+    },
+    {
+      "epoch": 2.582272727272727,
+      "grad_norm": 2.0222062886902963,
+      "learning_rate": 5.927950937084742e-07,
+      "loss": 0.0289,
+      "step": 11362
+    },
+    {
+      "epoch": 2.5825,
+      "grad_norm": 1.6239498918088153,
+      "learning_rate": 5.927059628090143e-07,
+      "loss": 0.0399,
+      "step": 11363
+    },
+    {
+      "epoch": 2.582727272727273,
+      "grad_norm": 2.657473242632913,
+      "learning_rate": 5.926168325680863e-07,
+      "loss": 0.0549,
+      "step": 11364
+    },
+    {
+      "epoch": 2.5829545454545455,
+      "grad_norm": 1.652005406162505,
+      "learning_rate": 5.925277029875074e-07,
+      "loss": 0.031,
+      "step": 11365
+    },
+    {
+      "epoch": 2.583181818181818,
+      "grad_norm": 2.7677901686223465,
+      "learning_rate": 5.924385740690953e-07,
+      "loss": 0.0474,
+      "step": 11366
+    },
+    {
+      "epoch": 2.583409090909091,
+      "grad_norm": 7.471567166970074,
+      "learning_rate": 5.923494458146676e-07,
+      "loss": 0.015,
+      "step": 11367
+    },
+    {
+      "epoch": 2.5836363636363635,
+      "grad_norm": 2.171860518938641,
+      "learning_rate": 5.922603182260417e-07,
+      "loss": 0.0245,
+      "step": 11368
+    },
+    {
+      "epoch": 2.5838636363636365,
+      "grad_norm": 2.3947320969423833,
+      "learning_rate": 5.921711913050349e-07,
+      "loss": 0.0288,
+      "step": 11369
+    },
+    {
+      "epoch": 2.584090909090909,
+      "grad_norm": 1.3558870172962343,
+      "learning_rate": 5.920820650534648e-07,
+      "loss": 0.0177,
+      "step": 11370
+    },
+    {
+      "epoch": 2.584318181818182,
+      "grad_norm": 1.641268046188856,
+      "learning_rate": 5.919929394731488e-07,
+      "loss": 0.0259,
+      "step": 11371
+    },
+    {
+      "epoch": 2.5845454545454545,
+      "grad_norm": 1.8676933841869552,
+      "learning_rate": 5.919038145659045e-07,
+      "loss": 0.0222,
+      "step": 11372
+    },
+    {
+      "epoch": 2.5847727272727274,
+      "grad_norm": 2.336593627846066,
+      "learning_rate": 5.918146903335493e-07,
+      "loss": 0.0316,
+      "step": 11373
+    },
+    {
+      "epoch": 2.585,
+      "grad_norm": 0.7632985105992801,
+      "learning_rate": 5.917255667779002e-07,
+      "loss": 0.0051,
+      "step": 11374
+    },
+    {
+      "epoch": 2.5852272727272725,
+      "grad_norm": 2.244334155112635,
+      "learning_rate": 5.916364439007748e-07,
+      "loss": 0.0224,
+      "step": 11375
+    },
+    {
+      "epoch": 2.5854545454545454,
+      "grad_norm": 1.2593382346723498,
+      "learning_rate": 5.915473217039906e-07,
+      "loss": 0.0099,
+      "step": 11376
+    },
+    {
+      "epoch": 2.5856818181818184,
+      "grad_norm": 1.6404898593146806,
+      "learning_rate": 5.914582001893649e-07,
+      "loss": 0.0219,
+      "step": 11377
+    },
+    {
+      "epoch": 2.585909090909091,
+      "grad_norm": 2.1492527324685686,
+      "learning_rate": 5.913690793587149e-07,
+      "loss": 0.0093,
+      "step": 11378
+    },
+    {
+      "epoch": 2.5861363636363635,
+      "grad_norm": 1.697360044667629,
+      "learning_rate": 5.912799592138581e-07,
+      "loss": 0.0425,
+      "step": 11379
+    },
+    {
+      "epoch": 2.5863636363636364,
+      "grad_norm": 1.4405225507114474,
+      "learning_rate": 5.911908397566118e-07,
+      "loss": 0.0059,
+      "step": 11380
+    },
+    {
+      "epoch": 2.586590909090909,
+      "grad_norm": 1.5932218759290397,
+      "learning_rate": 5.911017209887935e-07,
+      "loss": 0.0122,
+      "step": 11381
+    },
+    {
+      "epoch": 2.586818181818182,
+      "grad_norm": 2.5380795840932544,
+      "learning_rate": 5.910126029122199e-07,
+      "loss": 0.0288,
+      "step": 11382
+    },
+    {
+      "epoch": 2.5870454545454544,
+      "grad_norm": 2.241557194574368,
+      "learning_rate": 5.909234855287086e-07,
+      "loss": 0.0418,
+      "step": 11383
+    },
+    {
+      "epoch": 2.5872727272727274,
+      "grad_norm": 1.6932113925948573,
+      "learning_rate": 5.90834368840077e-07,
+      "loss": 0.0155,
+      "step": 11384
+    },
+    {
+      "epoch": 2.5875,
+      "grad_norm": 1.400893646233047,
+      "learning_rate": 5.907452528481422e-07,
+      "loss": 0.0126,
+      "step": 11385
+    },
+    {
+      "epoch": 2.587727272727273,
+      "grad_norm": 1.516669717267036,
+      "learning_rate": 5.906561375547216e-07,
+      "loss": 0.0141,
+      "step": 11386
+    },
+    {
+      "epoch": 2.5879545454545454,
+      "grad_norm": 2.0369164339981114,
+      "learning_rate": 5.905670229616321e-07,
+      "loss": 0.0423,
+      "step": 11387
+    },
+    {
+      "epoch": 2.588181818181818,
+      "grad_norm": 1.145841818038638,
+      "learning_rate": 5.904779090706911e-07,
+      "loss": 0.0136,
+      "step": 11388
+    },
+    {
+      "epoch": 2.588409090909091,
+      "grad_norm": 1.2213199570670088,
+      "learning_rate": 5.903887958837156e-07,
+      "loss": 0.0173,
+      "step": 11389
+    },
+    {
+      "epoch": 2.588636363636364,
+      "grad_norm": 2.616559873338896,
+      "learning_rate": 5.902996834025234e-07,
+      "loss": 0.0485,
+      "step": 11390
+    },
+    {
+      "epoch": 2.5888636363636364,
+      "grad_norm": 1.8052867036714488,
+      "learning_rate": 5.902105716289309e-07,
+      "loss": 0.0223,
+      "step": 11391
+    },
+    {
+      "epoch": 2.589090909090909,
+      "grad_norm": 2.0412690994821534,
+      "learning_rate": 5.901214605647555e-07,
+      "loss": 0.0196,
+      "step": 11392
+    },
+    {
+      "epoch": 2.589318181818182,
+      "grad_norm": 2.712035919162745,
+      "learning_rate": 5.900323502118142e-07,
+      "loss": 0.0264,
+      "step": 11393
+    },
+    {
+      "epoch": 2.589545454545455,
+      "grad_norm": 2.0594434437929534,
+      "learning_rate": 5.899432405719245e-07,
+      "loss": 0.0289,
+      "step": 11394
+    },
+    {
+      "epoch": 2.5897727272727273,
+      "grad_norm": 2.3758733396739666,
+      "learning_rate": 5.898541316469035e-07,
+      "loss": 0.0331,
+      "step": 11395
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 2.588459653536939,
+      "learning_rate": 5.897650234385678e-07,
+      "loss": 0.0222,
+      "step": 11396
+    },
+    {
+      "epoch": 2.590227272727273,
+      "grad_norm": 1.004772941260272,
+      "learning_rate": 5.896759159487347e-07,
+      "loss": 0.0065,
+      "step": 11397
+    },
+    {
+      "epoch": 2.5904545454545453,
+      "grad_norm": 2.2195329727348976,
+      "learning_rate": 5.895868091792214e-07,
+      "loss": 0.04,
+      "step": 11398
+    },
+    {
+      "epoch": 2.5906818181818183,
+      "grad_norm": 2.2220736659637326,
+      "learning_rate": 5.89497703131845e-07,
+      "loss": 0.0421,
+      "step": 11399
+    },
+    {
+      "epoch": 2.590909090909091,
+      "grad_norm": 1.5606164139661989,
+      "learning_rate": 5.894085978084221e-07,
+      "loss": 0.0131,
+      "step": 11400
+    },
+    {
+      "epoch": 2.591136363636364,
+      "grad_norm": 3.910957272749161,
+      "learning_rate": 5.8931949321077e-07,
+      "loss": 0.0211,
+      "step": 11401
+    },
+    {
+      "epoch": 2.5913636363636363,
+      "grad_norm": 1.1989330464307337,
+      "learning_rate": 5.892303893407058e-07,
+      "loss": 0.0159,
+      "step": 11402
+    },
+    {
+      "epoch": 2.5915909090909093,
+      "grad_norm": 1.379391791233135,
+      "learning_rate": 5.891412862000466e-07,
+      "loss": 0.0306,
+      "step": 11403
+    },
+    {
+      "epoch": 2.591818181818182,
+      "grad_norm": 1.4316754044640123,
+      "learning_rate": 5.890521837906088e-07,
+      "loss": 0.0193,
+      "step": 11404
+    },
+    {
+      "epoch": 2.5920454545454543,
+      "grad_norm": 1.2708803354276563,
+      "learning_rate": 5.889630821142097e-07,
+      "loss": 0.0087,
+      "step": 11405
+    },
+    {
+      "epoch": 2.5922727272727273,
+      "grad_norm": 2.219275061087988,
+      "learning_rate": 5.888739811726662e-07,
+      "loss": 0.045,
+      "step": 11406
+    },
+    {
+      "epoch": 2.5925000000000002,
+      "grad_norm": 4.147211608130537,
+      "learning_rate": 5.887848809677954e-07,
+      "loss": 0.0662,
+      "step": 11407
+    },
+    {
+      "epoch": 2.5927272727272728,
+      "grad_norm": 1.4767327654544498,
+      "learning_rate": 5.88695781501414e-07,
+      "loss": 0.0126,
+      "step": 11408
+    },
+    {
+      "epoch": 2.5929545454545453,
+      "grad_norm": 1.6677972319059695,
+      "learning_rate": 5.886066827753389e-07,
+      "loss": 0.0143,
+      "step": 11409
+    },
+    {
+      "epoch": 2.5931818181818183,
+      "grad_norm": 2.1180472590334043,
+      "learning_rate": 5.88517584791387e-07,
+      "loss": 0.0373,
+      "step": 11410
+    },
+    {
+      "epoch": 2.5934090909090908,
+      "grad_norm": 2.567328647789524,
+      "learning_rate": 5.884284875513752e-07,
+      "loss": 0.0378,
+      "step": 11411
+    },
+    {
+      "epoch": 2.5936363636363637,
+      "grad_norm": 1.6860935046139747,
+      "learning_rate": 5.883393910571207e-07,
+      "loss": 0.0137,
+      "step": 11412
+    },
+    {
+      "epoch": 2.5938636363636363,
+      "grad_norm": 0.9862991387269457,
+      "learning_rate": 5.882502953104396e-07,
+      "loss": 0.0109,
+      "step": 11413
+    },
+    {
+      "epoch": 2.5940909090909092,
+      "grad_norm": 1.3047265836978346,
+      "learning_rate": 5.881612003131491e-07,
+      "loss": 0.0153,
+      "step": 11414
+    },
+    {
+      "epoch": 2.5943181818181817,
+      "grad_norm": 2.1496083978708405,
+      "learning_rate": 5.88072106067066e-07,
+      "loss": 0.0578,
+      "step": 11415
+    },
+    {
+      "epoch": 2.5945454545454547,
+      "grad_norm": 1.5900601385579716,
+      "learning_rate": 5.879830125740073e-07,
+      "loss": 0.0294,
+      "step": 11416
+    },
+    {
+      "epoch": 2.5947727272727272,
+      "grad_norm": 1.4485615389016482,
+      "learning_rate": 5.878939198357893e-07,
+      "loss": 0.0177,
+      "step": 11417
+    },
+    {
+      "epoch": 2.5949999999999998,
+      "grad_norm": 1.6682522961608102,
+      "learning_rate": 5.878048278542292e-07,
+      "loss": 0.0341,
+      "step": 11418
+    },
+    {
+      "epoch": 2.5952272727272727,
+      "grad_norm": 1.317166977776951,
+      "learning_rate": 5.877157366311434e-07,
+      "loss": 0.0112,
+      "step": 11419
+    },
+    {
+      "epoch": 2.5954545454545457,
+      "grad_norm": 0.9950360334416849,
+      "learning_rate": 5.876266461683487e-07,
+      "loss": 0.0097,
+      "step": 11420
+    },
+    {
+      "epoch": 2.595681818181818,
+      "grad_norm": 2.676826851634453,
+      "learning_rate": 5.875375564676623e-07,
+      "loss": 0.0368,
+      "step": 11421
+    },
+    {
+      "epoch": 2.5959090909090907,
+      "grad_norm": 2.8358983692474866,
+      "learning_rate": 5.874484675309002e-07,
+      "loss": 0.0247,
+      "step": 11422
+    },
+    {
+      "epoch": 2.5961363636363637,
+      "grad_norm": 1.0510327019652927,
+      "learning_rate": 5.873593793598792e-07,
+      "loss": 0.011,
+      "step": 11423
+    },
+    {
+      "epoch": 2.596363636363636,
+      "grad_norm": 1.6658048972987,
+      "learning_rate": 5.872702919564162e-07,
+      "loss": 0.0226,
+      "step": 11424
+    },
+    {
+      "epoch": 2.596590909090909,
+      "grad_norm": 3.9481646657178078,
+      "learning_rate": 5.871812053223281e-07,
+      "loss": 0.0389,
+      "step": 11425
+    },
+    {
+      "epoch": 2.5968181818181817,
+      "grad_norm": 3.5336522549140534,
+      "learning_rate": 5.870921194594309e-07,
+      "loss": 0.0473,
+      "step": 11426
+    },
+    {
+      "epoch": 2.5970454545454547,
+      "grad_norm": 2.0519458454877966,
+      "learning_rate": 5.870030343695416e-07,
+      "loss": 0.0254,
+      "step": 11427
+    },
+    {
+      "epoch": 2.597272727272727,
+      "grad_norm": 1.0740754568687232,
+      "learning_rate": 5.869139500544766e-07,
+      "loss": 0.0103,
+      "step": 11428
+    },
+    {
+      "epoch": 2.5975,
+      "grad_norm": 1.3342530882268115,
+      "learning_rate": 5.868248665160526e-07,
+      "loss": 0.0126,
+      "step": 11429
+    },
+    {
+      "epoch": 2.5977272727272727,
+      "grad_norm": 1.1306368657736794,
+      "learning_rate": 5.867357837560864e-07,
+      "loss": 0.0126,
+      "step": 11430
+    },
+    {
+      "epoch": 2.597954545454545,
+      "grad_norm": 1.861639577041052,
+      "learning_rate": 5.866467017763941e-07,
+      "loss": 0.0298,
+      "step": 11431
+    },
+    {
+      "epoch": 2.598181818181818,
+      "grad_norm": 1.848236828049054,
+      "learning_rate": 5.865576205787926e-07,
+      "loss": 0.0277,
+      "step": 11432
+    },
+    {
+      "epoch": 2.598409090909091,
+      "grad_norm": 1.7167647986288765,
+      "learning_rate": 5.864685401650982e-07,
+      "loss": 0.0187,
+      "step": 11433
+    },
+    {
+      "epoch": 2.5986363636363636,
+      "grad_norm": 2.115072219879336,
+      "learning_rate": 5.863794605371277e-07,
+      "loss": 0.0313,
+      "step": 11434
+    },
+    {
+      "epoch": 2.598863636363636,
+      "grad_norm": 1.504759998828156,
+      "learning_rate": 5.862903816966972e-07,
+      "loss": 0.0158,
+      "step": 11435
+    },
+    {
+      "epoch": 2.599090909090909,
+      "grad_norm": 3.4303096048771065,
+      "learning_rate": 5.862013036456233e-07,
+      "loss": 0.0282,
+      "step": 11436
+    },
+    {
+      "epoch": 2.599318181818182,
+      "grad_norm": 0.6270887266550819,
+      "learning_rate": 5.861122263857226e-07,
+      "loss": 0.0081,
+      "step": 11437
+    },
+    {
+      "epoch": 2.5995454545454546,
+      "grad_norm": 2.1097023681405447,
+      "learning_rate": 5.860231499188113e-07,
+      "loss": 0.0176,
+      "step": 11438
+    },
+    {
+      "epoch": 2.599772727272727,
+      "grad_norm": 2.212154465567929,
+      "learning_rate": 5.859340742467062e-07,
+      "loss": 0.0176,
+      "step": 11439
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.0330220529152414,
+      "learning_rate": 5.858449993712232e-07,
+      "loss": 0.0115,
+      "step": 11440
+    },
+    {
+      "epoch": 2.6002272727272726,
+      "grad_norm": 1.1245746075552143,
+      "learning_rate": 5.857559252941792e-07,
+      "loss": 0.016,
+      "step": 11441
+    },
+    {
+      "epoch": 2.6004545454545456,
+      "grad_norm": 2.3274650620123842,
+      "learning_rate": 5.856668520173902e-07,
+      "loss": 0.0239,
+      "step": 11442
+    },
+    {
+      "epoch": 2.600681818181818,
+      "grad_norm": 2.037794392443987,
+      "learning_rate": 5.85577779542673e-07,
+      "loss": 0.0412,
+      "step": 11443
+    },
+    {
+      "epoch": 2.600909090909091,
+      "grad_norm": 1.8319866741207902,
+      "learning_rate": 5.854887078718433e-07,
+      "loss": 0.0144,
+      "step": 11444
+    },
+    {
+      "epoch": 2.6011363636363636,
+      "grad_norm": 1.5487852422229915,
+      "learning_rate": 5.85399637006718e-07,
+      "loss": 0.0242,
+      "step": 11445
+    },
+    {
+      "epoch": 2.6013636363636365,
+      "grad_norm": 1.9107924125268325,
+      "learning_rate": 5.85310566949113e-07,
+      "loss": 0.0152,
+      "step": 11446
+    },
+    {
+      "epoch": 2.601590909090909,
+      "grad_norm": 1.500646914856896,
+      "learning_rate": 5.85221497700845e-07,
+      "loss": 0.0205,
+      "step": 11447
+    },
+    {
+      "epoch": 2.6018181818181816,
+      "grad_norm": 3.542051035620886,
+      "learning_rate": 5.8513242926373e-07,
+      "loss": 0.0405,
+      "step": 11448
+    },
+    {
+      "epoch": 2.6020454545454546,
+      "grad_norm": 1.1799435551592488,
+      "learning_rate": 5.850433616395842e-07,
+      "loss": 0.018,
+      "step": 11449
+    },
+    {
+      "epoch": 2.6022727272727275,
+      "grad_norm": 2.0602321288205347,
+      "learning_rate": 5.84954294830224e-07,
+      "loss": 0.0186,
+      "step": 11450
+    },
+    {
+      "epoch": 2.6025,
+      "grad_norm": 3.887200940721231,
+      "learning_rate": 5.848652288374655e-07,
+      "loss": 0.0329,
+      "step": 11451
+    },
+    {
+      "epoch": 2.6027272727272726,
+      "grad_norm": 2.399849152968134,
+      "learning_rate": 5.847761636631253e-07,
+      "loss": 0.0226,
+      "step": 11452
+    },
+    {
+      "epoch": 2.6029545454545455,
+      "grad_norm": 2.2171692405521397,
+      "learning_rate": 5.84687099309019e-07,
+      "loss": 0.0358,
+      "step": 11453
+    },
+    {
+      "epoch": 2.603181818181818,
+      "grad_norm": 2.130267152846488,
+      "learning_rate": 5.845980357769632e-07,
+      "loss": 0.0613,
+      "step": 11454
+    },
+    {
+      "epoch": 2.603409090909091,
+      "grad_norm": 2.1874449713007644,
+      "learning_rate": 5.845089730687739e-07,
+      "loss": 0.0305,
+      "step": 11455
+    },
+    {
+      "epoch": 2.6036363636363635,
+      "grad_norm": 1.5516998414484664,
+      "learning_rate": 5.844199111862675e-07,
+      "loss": 0.0123,
+      "step": 11456
+    },
+    {
+      "epoch": 2.6038636363636365,
+      "grad_norm": 1.7673715611303047,
+      "learning_rate": 5.843308501312598e-07,
+      "loss": 0.0114,
+      "step": 11457
+    },
+    {
+      "epoch": 2.604090909090909,
+      "grad_norm": 1.7609594130675472,
+      "learning_rate": 5.842417899055669e-07,
+      "loss": 0.0226,
+      "step": 11458
+    },
+    {
+      "epoch": 2.604318181818182,
+      "grad_norm": 2.1750854593348485,
+      "learning_rate": 5.841527305110049e-07,
+      "loss": 0.0335,
+      "step": 11459
+    },
+    {
+      "epoch": 2.6045454545454545,
+      "grad_norm": 1.9867684010201705,
+      "learning_rate": 5.840636719493902e-07,
+      "loss": 0.0164,
+      "step": 11460
+    },
+    {
+      "epoch": 2.604772727272727,
+      "grad_norm": 1.1241889179003897,
+      "learning_rate": 5.839746142225386e-07,
+      "loss": 0.0206,
+      "step": 11461
+    },
+    {
+      "epoch": 2.605,
+      "grad_norm": 1.6782386097136095,
+      "learning_rate": 5.838855573322661e-07,
+      "loss": 0.0186,
+      "step": 11462
+    },
+    {
+      "epoch": 2.605227272727273,
+      "grad_norm": 2.2679347480969345,
+      "learning_rate": 5.837965012803889e-07,
+      "loss": 0.0203,
+      "step": 11463
+    },
+    {
+      "epoch": 2.6054545454545455,
+      "grad_norm": 1.7915776993916754,
+      "learning_rate": 5.837074460687228e-07,
+      "loss": 0.0143,
+      "step": 11464
+    },
+    {
+      "epoch": 2.605681818181818,
+      "grad_norm": 1.2059195066131345,
+      "learning_rate": 5.836183916990843e-07,
+      "loss": 0.0163,
+      "step": 11465
+    },
+    {
+      "epoch": 2.605909090909091,
+      "grad_norm": 4.2785496080320415,
+      "learning_rate": 5.835293381732887e-07,
+      "loss": 0.0745,
+      "step": 11466
+    },
+    {
+      "epoch": 2.6061363636363635,
+      "grad_norm": 2.0983678238774135,
+      "learning_rate": 5.834402854931521e-07,
+      "loss": 0.0282,
+      "step": 11467
+    },
+    {
+      "epoch": 2.6063636363636364,
+      "grad_norm": 1.7901719657621742,
+      "learning_rate": 5.833512336604907e-07,
+      "loss": 0.0138,
+      "step": 11468
+    },
+    {
+      "epoch": 2.606590909090909,
+      "grad_norm": 2.8343481587866246,
+      "learning_rate": 5.832621826771205e-07,
+      "loss": 0.0451,
+      "step": 11469
+    },
+    {
+      "epoch": 2.606818181818182,
+      "grad_norm": 1.7547874964088976,
+      "learning_rate": 5.831731325448569e-07,
+      "loss": 0.0256,
+      "step": 11470
+    },
+    {
+      "epoch": 2.6070454545454544,
+      "grad_norm": 1.0330509617107295,
+      "learning_rate": 5.830840832655163e-07,
+      "loss": 0.0075,
+      "step": 11471
+    },
+    {
+      "epoch": 2.6072727272727274,
+      "grad_norm": 1.7048644911010096,
+      "learning_rate": 5.829950348409143e-07,
+      "loss": 0.0143,
+      "step": 11472
+    },
+    {
+      "epoch": 2.6075,
+      "grad_norm": 2.0126762231902653,
+      "learning_rate": 5.829059872728668e-07,
+      "loss": 0.0193,
+      "step": 11473
+    },
+    {
+      "epoch": 2.607727272727273,
+      "grad_norm": 1.0296671423055848,
+      "learning_rate": 5.828169405631899e-07,
+      "loss": 0.0146,
+      "step": 11474
+    },
+    {
+      "epoch": 2.6079545454545454,
+      "grad_norm": 2.9008922616508213,
+      "learning_rate": 5.827278947136989e-07,
+      "loss": 0.0335,
+      "step": 11475
+    },
+    {
+      "epoch": 2.6081818181818184,
+      "grad_norm": 1.2024923635882854,
+      "learning_rate": 5.826388497262099e-07,
+      "loss": 0.0093,
+      "step": 11476
+    },
+    {
+      "epoch": 2.608409090909091,
+      "grad_norm": 4.194933596159661,
+      "learning_rate": 5.825498056025385e-07,
+      "loss": 0.0319,
+      "step": 11477
+    },
+    {
+      "epoch": 2.6086363636363634,
+      "grad_norm": 1.7872595660231037,
+      "learning_rate": 5.824607623445011e-07,
+      "loss": 0.016,
+      "step": 11478
+    },
+    {
+      "epoch": 2.6088636363636364,
+      "grad_norm": 1.6996827304419584,
+      "learning_rate": 5.823717199539127e-07,
+      "loss": 0.0208,
+      "step": 11479
+    },
+    {
+      "epoch": 2.6090909090909093,
+      "grad_norm": 1.7794786220394385,
+      "learning_rate": 5.822826784325892e-07,
+      "loss": 0.0234,
+      "step": 11480
+    },
+    {
+      "epoch": 2.609318181818182,
+      "grad_norm": 1.6169624453829055,
+      "learning_rate": 5.821936377823464e-07,
+      "loss": 0.0238,
+      "step": 11481
+    },
+    {
+      "epoch": 2.6095454545454544,
+      "grad_norm": 1.2504824575205153,
+      "learning_rate": 5.821045980050001e-07,
+      "loss": 0.0077,
+      "step": 11482
+    },
+    {
+      "epoch": 2.6097727272727274,
+      "grad_norm": 2.8440663637049566,
+      "learning_rate": 5.82015559102366e-07,
+      "loss": 0.0472,
+      "step": 11483
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.393363185475453,
+      "learning_rate": 5.819265210762595e-07,
+      "loss": 0.0463,
+      "step": 11484
+    },
+    {
+      "epoch": 2.610227272727273,
+      "grad_norm": 0.8866797301747242,
+      "learning_rate": 5.818374839284964e-07,
+      "loss": 0.0075,
+      "step": 11485
+    },
+    {
+      "epoch": 2.6104545454545454,
+      "grad_norm": 0.9170092830957094,
+      "learning_rate": 5.817484476608923e-07,
+      "loss": 0.0039,
+      "step": 11486
+    },
+    {
+      "epoch": 2.6106818181818183,
+      "grad_norm": 1.0598528647957022,
+      "learning_rate": 5.81659412275263e-07,
+      "loss": 0.0191,
+      "step": 11487
+    },
+    {
+      "epoch": 2.610909090909091,
+      "grad_norm": 3.123980441764899,
+      "learning_rate": 5.815703777734237e-07,
+      "loss": 0.0533,
+      "step": 11488
+    },
+    {
+      "epoch": 2.611136363636364,
+      "grad_norm": 2.0816452939157557,
+      "learning_rate": 5.814813441571901e-07,
+      "loss": 0.0141,
+      "step": 11489
+    },
+    {
+      "epoch": 2.6113636363636363,
+      "grad_norm": 1.949369753692559,
+      "learning_rate": 5.813923114283779e-07,
+      "loss": 0.0437,
+      "step": 11490
+    },
+    {
+      "epoch": 2.611590909090909,
+      "grad_norm": 1.4791865174596872,
+      "learning_rate": 5.813032795888027e-07,
+      "loss": 0.0122,
+      "step": 11491
+    },
+    {
+      "epoch": 2.611818181818182,
+      "grad_norm": 2.0152630738594364,
+      "learning_rate": 5.812142486402796e-07,
+      "loss": 0.0201,
+      "step": 11492
+    },
+    {
+      "epoch": 2.612045454545455,
+      "grad_norm": 2.631789751371069,
+      "learning_rate": 5.811252185846245e-07,
+      "loss": 0.0205,
+      "step": 11493
+    },
+    {
+      "epoch": 2.6122727272727273,
+      "grad_norm": 1.6006361912341929,
+      "learning_rate": 5.810361894236525e-07,
+      "loss": 0.0183,
+      "step": 11494
+    },
+    {
+      "epoch": 2.6125,
+      "grad_norm": 2.6923422059634086,
+      "learning_rate": 5.809471611591795e-07,
+      "loss": 0.0298,
+      "step": 11495
+    },
+    {
+      "epoch": 2.612727272727273,
+      "grad_norm": 1.3258869648447076,
+      "learning_rate": 5.80858133793021e-07,
+      "loss": 0.0335,
+      "step": 11496
+    },
+    {
+      "epoch": 2.6129545454545453,
+      "grad_norm": 1.230407881544965,
+      "learning_rate": 5.807691073269918e-07,
+      "loss": 0.012,
+      "step": 11497
+    },
+    {
+      "epoch": 2.6131818181818183,
+      "grad_norm": 1.4112985400311324,
+      "learning_rate": 5.806800817629078e-07,
+      "loss": 0.0137,
+      "step": 11498
+    },
+    {
+      "epoch": 2.613409090909091,
+      "grad_norm": 1.408133099636492,
+      "learning_rate": 5.805910571025842e-07,
+      "loss": 0.0236,
+      "step": 11499
+    },
+    {
+      "epoch": 2.6136363636363638,
+      "grad_norm": 1.5351882122080875,
+      "learning_rate": 5.805020333478365e-07,
+      "loss": 0.0106,
+      "step": 11500
+    },
+    {
+      "epoch": 2.6138636363636363,
+      "grad_norm": 1.864237443928545,
+      "learning_rate": 5.804130105004799e-07,
+      "loss": 0.0464,
+      "step": 11501
+    },
+    {
+      "epoch": 2.6140909090909092,
+      "grad_norm": 2.570678925732285,
+      "learning_rate": 5.803239885623298e-07,
+      "loss": 0.0358,
+      "step": 11502
+    },
+    {
+      "epoch": 2.6143181818181818,
+      "grad_norm": 1.20304139618018,
+      "learning_rate": 5.802349675352014e-07,
+      "loss": 0.0157,
+      "step": 11503
+    },
+    {
+      "epoch": 2.6145454545454543,
+      "grad_norm": 2.256594724460587,
+      "learning_rate": 5.801459474209106e-07,
+      "loss": 0.0172,
+      "step": 11504
+    },
+    {
+      "epoch": 2.6147727272727272,
+      "grad_norm": 1.1166666772603981,
+      "learning_rate": 5.800569282212718e-07,
+      "loss": 0.0114,
+      "step": 11505
+    },
+    {
+      "epoch": 2.615,
+      "grad_norm": 1.309454158716034,
+      "learning_rate": 5.799679099381005e-07,
+      "loss": 0.0229,
+      "step": 11506
+    },
+    {
+      "epoch": 2.6152272727272727,
+      "grad_norm": 1.9092282593546717,
+      "learning_rate": 5.798788925732121e-07,
+      "loss": 0.0455,
+      "step": 11507
+    },
+    {
+      "epoch": 2.6154545454545453,
+      "grad_norm": 1.034271403508559,
+      "learning_rate": 5.797898761284219e-07,
+      "loss": 0.0108,
+      "step": 11508
+    },
+    {
+      "epoch": 2.615681818181818,
+      "grad_norm": 1.8024219413864784,
+      "learning_rate": 5.797008606055453e-07,
+      "loss": 0.0391,
+      "step": 11509
+    },
+    {
+      "epoch": 2.615909090909091,
+      "grad_norm": 1.8608710061286333,
+      "learning_rate": 5.796118460063969e-07,
+      "loss": 0.0169,
+      "step": 11510
+    },
+    {
+      "epoch": 2.6161363636363637,
+      "grad_norm": 1.5523502014681394,
+      "learning_rate": 5.795228323327921e-07,
+      "loss": 0.0189,
+      "step": 11511
+    },
+    {
+      "epoch": 2.6163636363636362,
+      "grad_norm": 0.7319977646147056,
+      "learning_rate": 5.79433819586546e-07,
+      "loss": 0.0088,
+      "step": 11512
+    },
+    {
+      "epoch": 2.616590909090909,
+      "grad_norm": 1.7443304613092754,
+      "learning_rate": 5.793448077694742e-07,
+      "loss": 0.014,
+      "step": 11513
+    },
+    {
+      "epoch": 2.6168181818181817,
+      "grad_norm": 2.122159474345853,
+      "learning_rate": 5.792557968833912e-07,
+      "loss": 0.0203,
+      "step": 11514
+    },
+    {
+      "epoch": 2.6170454545454547,
+      "grad_norm": 1.179500745631835,
+      "learning_rate": 5.791667869301122e-07,
+      "loss": 0.0092,
+      "step": 11515
+    },
+    {
+      "epoch": 2.617272727272727,
+      "grad_norm": 2.6854913527036284,
+      "learning_rate": 5.790777779114524e-07,
+      "loss": 0.0568,
+      "step": 11516
+    },
+    {
+      "epoch": 2.6175,
+      "grad_norm": 2.200767512535004,
+      "learning_rate": 5.789887698292268e-07,
+      "loss": 0.0281,
+      "step": 11517
+    },
+    {
+      "epoch": 2.6177272727272727,
+      "grad_norm": 2.11272002801924,
+      "learning_rate": 5.788997626852509e-07,
+      "loss": 0.026,
+      "step": 11518
+    },
+    {
+      "epoch": 2.6179545454545456,
+      "grad_norm": 1.3765391090653092,
+      "learning_rate": 5.788107564813388e-07,
+      "loss": 0.0117,
+      "step": 11519
+    },
+    {
+      "epoch": 2.618181818181818,
+      "grad_norm": 1.7252656151979304,
+      "learning_rate": 5.78721751219306e-07,
+      "loss": 0.0129,
+      "step": 11520
+    },
+    {
+      "epoch": 2.6184090909090907,
+      "grad_norm": 2.2070943380574586,
+      "learning_rate": 5.786327469009675e-07,
+      "loss": 0.0286,
+      "step": 11521
+    },
+    {
+      "epoch": 2.6186363636363637,
+      "grad_norm": 2.293128888855362,
+      "learning_rate": 5.785437435281383e-07,
+      "loss": 0.0359,
+      "step": 11522
+    },
+    {
+      "epoch": 2.6188636363636366,
+      "grad_norm": 2.0119082215098225,
+      "learning_rate": 5.784547411026331e-07,
+      "loss": 0.0218,
+      "step": 11523
+    },
+    {
+      "epoch": 2.619090909090909,
+      "grad_norm": 1.2807135942835586,
+      "learning_rate": 5.783657396262668e-07,
+      "loss": 0.0079,
+      "step": 11524
+    },
+    {
+      "epoch": 2.6193181818181817,
+      "grad_norm": 1.1469570885855223,
+      "learning_rate": 5.782767391008546e-07,
+      "loss": 0.0181,
+      "step": 11525
+    },
+    {
+      "epoch": 2.6195454545454546,
+      "grad_norm": 1.449181484739175,
+      "learning_rate": 5.781877395282112e-07,
+      "loss": 0.0166,
+      "step": 11526
+    },
+    {
+      "epoch": 2.619772727272727,
+      "grad_norm": 1.562379954193578,
+      "learning_rate": 5.780987409101515e-07,
+      "loss": 0.0207,
+      "step": 11527
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.1798834250894514,
+      "learning_rate": 5.780097432484902e-07,
+      "loss": 0.0115,
+      "step": 11528
+    },
+    {
+      "epoch": 2.6202272727272726,
+      "grad_norm": 2.5759545465541778,
+      "learning_rate": 5.779207465450422e-07,
+      "loss": 0.0329,
+      "step": 11529
+    },
+    {
+      "epoch": 2.6204545454545456,
+      "grad_norm": 3.2688803612865227,
+      "learning_rate": 5.778317508016223e-07,
+      "loss": 0.0406,
+      "step": 11530
+    },
+    {
+      "epoch": 2.620681818181818,
+      "grad_norm": 0.9383499364946275,
+      "learning_rate": 5.777427560200454e-07,
+      "loss": 0.0088,
+      "step": 11531
+    },
+    {
+      "epoch": 2.620909090909091,
+      "grad_norm": 1.4640595448298679,
+      "learning_rate": 5.776537622021261e-07,
+      "loss": 0.0124,
+      "step": 11532
+    },
+    {
+      "epoch": 2.6211363636363636,
+      "grad_norm": 1.4838665577702828,
+      "learning_rate": 5.775647693496791e-07,
+      "loss": 0.0118,
+      "step": 11533
+    },
+    {
+      "epoch": 2.621363636363636,
+      "grad_norm": 1.8304967518831508,
+      "learning_rate": 5.774757774645192e-07,
+      "loss": 0.0116,
+      "step": 11534
+    },
+    {
+      "epoch": 2.621590909090909,
+      "grad_norm": 2.890582005988172,
+      "learning_rate": 5.773867865484615e-07,
+      "loss": 0.0277,
+      "step": 11535
+    },
+    {
+      "epoch": 2.621818181818182,
+      "grad_norm": 3.3931749325518163,
+      "learning_rate": 5.772977966033197e-07,
+      "loss": 0.026,
+      "step": 11536
+    },
+    {
+      "epoch": 2.6220454545454546,
+      "grad_norm": 0.8341254519451861,
+      "learning_rate": 5.772088076309093e-07,
+      "loss": 0.0064,
+      "step": 11537
+    },
+    {
+      "epoch": 2.622272727272727,
+      "grad_norm": 1.6252593546097411,
+      "learning_rate": 5.771198196330446e-07,
+      "loss": 0.0146,
+      "step": 11538
+    },
+    {
+      "epoch": 2.6225,
+      "grad_norm": 1.3147624942605147,
+      "learning_rate": 5.770308326115404e-07,
+      "loss": 0.0157,
+      "step": 11539
+    },
+    {
+      "epoch": 2.6227272727272726,
+      "grad_norm": 1.7810171447399008,
+      "learning_rate": 5.769418465682115e-07,
+      "loss": 0.0251,
+      "step": 11540
+    },
+    {
+      "epoch": 2.6229545454545455,
+      "grad_norm": 2.0655848323874593,
+      "learning_rate": 5.768528615048718e-07,
+      "loss": 0.0276,
+      "step": 11541
+    },
+    {
+      "epoch": 2.623181818181818,
+      "grad_norm": 1.992035160259169,
+      "learning_rate": 5.767638774233364e-07,
+      "loss": 0.0178,
+      "step": 11542
+    },
+    {
+      "epoch": 2.623409090909091,
+      "grad_norm": 0.8993136503193377,
+      "learning_rate": 5.766748943254197e-07,
+      "loss": 0.0082,
+      "step": 11543
+    },
+    {
+      "epoch": 2.6236363636363635,
+      "grad_norm": 1.3221838265906283,
+      "learning_rate": 5.765859122129362e-07,
+      "loss": 0.0111,
+      "step": 11544
+    },
+    {
+      "epoch": 2.6238636363636365,
+      "grad_norm": 2.6785302526118544,
+      "learning_rate": 5.764969310877003e-07,
+      "loss": 0.0462,
+      "step": 11545
+    },
+    {
+      "epoch": 2.624090909090909,
+      "grad_norm": 1.6110988543332545,
+      "learning_rate": 5.764079509515267e-07,
+      "loss": 0.021,
+      "step": 11546
+    },
+    {
+      "epoch": 2.6243181818181816,
+      "grad_norm": 1.7388094246873504,
+      "learning_rate": 5.763189718062297e-07,
+      "loss": 0.023,
+      "step": 11547
+    },
+    {
+      "epoch": 2.6245454545454545,
+      "grad_norm": 3.8838701049273627,
+      "learning_rate": 5.76229993653624e-07,
+      "loss": 0.0622,
+      "step": 11548
+    },
+    {
+      "epoch": 2.6247727272727275,
+      "grad_norm": 3.056969924117314,
+      "learning_rate": 5.761410164955237e-07,
+      "loss": 0.0352,
+      "step": 11549
+    },
+    {
+      "epoch": 2.625,
+      "grad_norm": 2.632039739191141,
+      "learning_rate": 5.760520403337431e-07,
+      "loss": 0.0384,
+      "step": 11550
+    },
+    {
+      "epoch": 2.6252272727272725,
+      "grad_norm": 1.675298976344395,
+      "learning_rate": 5.759630651700971e-07,
+      "loss": 0.0279,
+      "step": 11551
+    },
+    {
+      "epoch": 2.6254545454545455,
+      "grad_norm": 1.6604339686782252,
+      "learning_rate": 5.758740910063994e-07,
+      "loss": 0.0165,
+      "step": 11552
+    },
+    {
+      "epoch": 2.6256818181818184,
+      "grad_norm": 1.2329196888951583,
+      "learning_rate": 5.75785117844465e-07,
+      "loss": 0.0148,
+      "step": 11553
+    },
+    {
+      "epoch": 2.625909090909091,
+      "grad_norm": 2.3883368305940196,
+      "learning_rate": 5.756961456861077e-07,
+      "loss": 0.0328,
+      "step": 11554
+    },
+    {
+      "epoch": 2.6261363636363635,
+      "grad_norm": 3.5637961082865366,
+      "learning_rate": 5.756071745331421e-07,
+      "loss": 0.0342,
+      "step": 11555
+    },
+    {
+      "epoch": 2.6263636363636365,
+      "grad_norm": 1.3831946219176912,
+      "learning_rate": 5.755182043873824e-07,
+      "loss": 0.0115,
+      "step": 11556
+    },
+    {
+      "epoch": 2.626590909090909,
+      "grad_norm": 2.4604120627011477,
+      "learning_rate": 5.754292352506431e-07,
+      "loss": 0.0331,
+      "step": 11557
+    },
+    {
+      "epoch": 2.626818181818182,
+      "grad_norm": 1.6676973537384183,
+      "learning_rate": 5.753402671247379e-07,
+      "loss": 0.0227,
+      "step": 11558
+    },
+    {
+      "epoch": 2.6270454545454545,
+      "grad_norm": 1.800927024619778,
+      "learning_rate": 5.752513000114812e-07,
+      "loss": 0.0246,
+      "step": 11559
+    },
+    {
+      "epoch": 2.6272727272727274,
+      "grad_norm": 1.2467210883621342,
+      "learning_rate": 5.751623339126873e-07,
+      "loss": 0.0251,
+      "step": 11560
+    },
+    {
+      "epoch": 2.6275,
+      "grad_norm": 1.846224544356583,
+      "learning_rate": 5.750733688301703e-07,
+      "loss": 0.0431,
+      "step": 11561
+    },
+    {
+      "epoch": 2.627727272727273,
+      "grad_norm": 2.096389679106739,
+      "learning_rate": 5.749844047657447e-07,
+      "loss": 0.0144,
+      "step": 11562
+    },
+    {
+      "epoch": 2.6279545454545454,
+      "grad_norm": 1.6400397720103466,
+      "learning_rate": 5.748954417212242e-07,
+      "loss": 0.0244,
+      "step": 11563
+    },
+    {
+      "epoch": 2.628181818181818,
+      "grad_norm": 1.443366209177607,
+      "learning_rate": 5.748064796984229e-07,
+      "loss": 0.0196,
+      "step": 11564
+    },
+    {
+      "epoch": 2.628409090909091,
+      "grad_norm": 1.5274004035708098,
+      "learning_rate": 5.747175186991552e-07,
+      "loss": 0.0224,
+      "step": 11565
+    },
+    {
+      "epoch": 2.628636363636364,
+      "grad_norm": 1.5245670235088709,
+      "learning_rate": 5.746285587252351e-07,
+      "loss": 0.0216,
+      "step": 11566
+    },
+    {
+      "epoch": 2.6288636363636364,
+      "grad_norm": 1.7356273006703022,
+      "learning_rate": 5.745395997784764e-07,
+      "loss": 0.0136,
+      "step": 11567
+    },
+    {
+      "epoch": 2.629090909090909,
+      "grad_norm": 2.9021902605222163,
+      "learning_rate": 5.744506418606933e-07,
+      "loss": 0.017,
+      "step": 11568
+    },
+    {
+      "epoch": 2.629318181818182,
+      "grad_norm": 2.4182290985650123,
+      "learning_rate": 5.743616849736998e-07,
+      "loss": 0.0268,
+      "step": 11569
+    },
+    {
+      "epoch": 2.6295454545454544,
+      "grad_norm": 3.084680043283037,
+      "learning_rate": 5.742727291193102e-07,
+      "loss": 0.0412,
+      "step": 11570
+    },
+    {
+      "epoch": 2.6297727272727274,
+      "grad_norm": 3.2606138817000336,
+      "learning_rate": 5.741837742993379e-07,
+      "loss": 0.0482,
+      "step": 11571
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.6278340235008635,
+      "learning_rate": 5.74094820515597e-07,
+      "loss": 0.0144,
+      "step": 11572
+    },
+    {
+      "epoch": 2.630227272727273,
+      "grad_norm": 3.552610774276409,
+      "learning_rate": 5.740058677699016e-07,
+      "loss": 0.0282,
+      "step": 11573
+    },
+    {
+      "epoch": 2.6304545454545454,
+      "grad_norm": 1.9798923912506576,
+      "learning_rate": 5.739169160640655e-07,
+      "loss": 0.022,
+      "step": 11574
+    },
+    {
+      "epoch": 2.6306818181818183,
+      "grad_norm": 1.4692025142230565,
+      "learning_rate": 5.738279653999027e-07,
+      "loss": 0.0146,
+      "step": 11575
+    },
+    {
+      "epoch": 2.630909090909091,
+      "grad_norm": 1.8175582734930806,
+      "learning_rate": 5.737390157792269e-07,
+      "loss": 0.0189,
+      "step": 11576
+    },
+    {
+      "epoch": 2.6311363636363634,
+      "grad_norm": 2.7068264282433248,
+      "learning_rate": 5.736500672038519e-07,
+      "loss": 0.0275,
+      "step": 11577
+    },
+    {
+      "epoch": 2.6313636363636363,
+      "grad_norm": 3.4022145719260446,
+      "learning_rate": 5.735611196755917e-07,
+      "loss": 0.0395,
+      "step": 11578
+    },
+    {
+      "epoch": 2.6315909090909093,
+      "grad_norm": 1.8984097789325836,
+      "learning_rate": 5.734721731962602e-07,
+      "loss": 0.0171,
+      "step": 11579
+    },
+    {
+      "epoch": 2.631818181818182,
+      "grad_norm": 1.8463611043745107,
+      "learning_rate": 5.733832277676709e-07,
+      "loss": 0.0236,
+      "step": 11580
+    },
+    {
+      "epoch": 2.6320454545454544,
+      "grad_norm": 1.2404625470662873,
+      "learning_rate": 5.732942833916374e-07,
+      "loss": 0.0137,
+      "step": 11581
+    },
+    {
+      "epoch": 2.6322727272727273,
+      "grad_norm": 3.2148175813687794,
+      "learning_rate": 5.732053400699738e-07,
+      "loss": 0.0275,
+      "step": 11582
+    },
+    {
+      "epoch": 2.6325,
+      "grad_norm": 1.97110456947996,
+      "learning_rate": 5.731163978044936e-07,
+      "loss": 0.0234,
+      "step": 11583
+    },
+    {
+      "epoch": 2.632727272727273,
+      "grad_norm": 4.27592280507297,
+      "learning_rate": 5.730274565970108e-07,
+      "loss": 0.0525,
+      "step": 11584
+    },
+    {
+      "epoch": 2.6329545454545453,
+      "grad_norm": 1.5804268572845313,
+      "learning_rate": 5.729385164493385e-07,
+      "loss": 0.0176,
+      "step": 11585
+    },
+    {
+      "epoch": 2.6331818181818183,
+      "grad_norm": 1.640360265763609,
+      "learning_rate": 5.72849577363291e-07,
+      "loss": 0.0178,
+      "step": 11586
+    },
+    {
+      "epoch": 2.633409090909091,
+      "grad_norm": 2.1772872839828414,
+      "learning_rate": 5.727606393406813e-07,
+      "loss": 0.0181,
+      "step": 11587
+    },
+    {
+      "epoch": 2.6336363636363638,
+      "grad_norm": 2.2399876975503035,
+      "learning_rate": 5.726717023833237e-07,
+      "loss": 0.0335,
+      "step": 11588
+    },
+    {
+      "epoch": 2.6338636363636363,
+      "grad_norm": 1.7016122403946066,
+      "learning_rate": 5.725827664930311e-07,
+      "loss": 0.0204,
+      "step": 11589
+    },
+    {
+      "epoch": 2.634090909090909,
+      "grad_norm": 1.641905156190159,
+      "learning_rate": 5.724938316716172e-07,
+      "loss": 0.0146,
+      "step": 11590
+    },
+    {
+      "epoch": 2.634318181818182,
+      "grad_norm": 3.467972008425736,
+      "learning_rate": 5.724048979208958e-07,
+      "loss": 0.0448,
+      "step": 11591
+    },
+    {
+      "epoch": 2.6345454545454547,
+      "grad_norm": 2.0815249300787144,
+      "learning_rate": 5.723159652426806e-07,
+      "loss": 0.032,
+      "step": 11592
+    },
+    {
+      "epoch": 2.6347727272727273,
+      "grad_norm": 0.5141594796137389,
+      "learning_rate": 5.722270336387844e-07,
+      "loss": 0.0048,
+      "step": 11593
+    },
+    {
+      "epoch": 2.635,
+      "grad_norm": 1.4811097203011065,
+      "learning_rate": 5.721381031110211e-07,
+      "loss": 0.0201,
+      "step": 11594
+    },
+    {
+      "epoch": 2.6352272727272728,
+      "grad_norm": 2.084333877810082,
+      "learning_rate": 5.720491736612041e-07,
+      "loss": 0.0311,
+      "step": 11595
+    },
+    {
+      "epoch": 2.6354545454545457,
+      "grad_norm": 1.7816454987279313,
+      "learning_rate": 5.719602452911469e-07,
+      "loss": 0.0187,
+      "step": 11596
+    },
+    {
+      "epoch": 2.6356818181818182,
+      "grad_norm": 2.526580827714817,
+      "learning_rate": 5.718713180026629e-07,
+      "loss": 0.0472,
+      "step": 11597
+    },
+    {
+      "epoch": 2.6359090909090908,
+      "grad_norm": 1.2881328497936817,
+      "learning_rate": 5.717823917975653e-07,
+      "loss": 0.0084,
+      "step": 11598
+    },
+    {
+      "epoch": 2.6361363636363637,
+      "grad_norm": 1.668687792872824,
+      "learning_rate": 5.716934666776676e-07,
+      "loss": 0.016,
+      "step": 11599
+    },
+    {
+      "epoch": 2.6363636363636362,
+      "grad_norm": 1.7643460997738845,
+      "learning_rate": 5.71604542644783e-07,
+      "loss": 0.0318,
+      "step": 11600
+    },
+    {
+      "epoch": 2.636590909090909,
+      "grad_norm": 0.9783322621410996,
+      "learning_rate": 5.715156197007253e-07,
+      "loss": 0.0146,
+      "step": 11601
+    },
+    {
+      "epoch": 2.6368181818181817,
+      "grad_norm": 1.5926684253757712,
+      "learning_rate": 5.714266978473072e-07,
+      "loss": 0.0148,
+      "step": 11602
+    },
+    {
+      "epoch": 2.6370454545454547,
+      "grad_norm": 1.6268827583825014,
+      "learning_rate": 5.713377770863422e-07,
+      "loss": 0.0195,
+      "step": 11603
+    },
+    {
+      "epoch": 2.637272727272727,
+      "grad_norm": 3.110926134495698,
+      "learning_rate": 5.712488574196434e-07,
+      "loss": 0.0215,
+      "step": 11604
+    },
+    {
+      "epoch": 2.6375,
+      "grad_norm": 1.5600426543676298,
+      "learning_rate": 5.711599388490242e-07,
+      "loss": 0.0177,
+      "step": 11605
+    },
+    {
+      "epoch": 2.6377272727272727,
+      "grad_norm": 1.479428876996329,
+      "learning_rate": 5.71071021376298e-07,
+      "loss": 0.0134,
+      "step": 11606
+    },
+    {
+      "epoch": 2.637954545454545,
+      "grad_norm": 1.637959510600137,
+      "learning_rate": 5.709821050032776e-07,
+      "loss": 0.0142,
+      "step": 11607
+    },
+    {
+      "epoch": 2.638181818181818,
+      "grad_norm": 2.564305296911466,
+      "learning_rate": 5.708931897317763e-07,
+      "loss": 0.0335,
+      "step": 11608
+    },
+    {
+      "epoch": 2.638409090909091,
+      "grad_norm": 1.6381300605688132,
+      "learning_rate": 5.708042755636072e-07,
+      "loss": 0.0121,
+      "step": 11609
+    },
+    {
+      "epoch": 2.6386363636363637,
+      "grad_norm": 6.27417872234224,
+      "learning_rate": 5.707153625005839e-07,
+      "loss": 0.042,
+      "step": 11610
+    },
+    {
+      "epoch": 2.638863636363636,
+      "grad_norm": 1.5895634251540949,
+      "learning_rate": 5.706264505445186e-07,
+      "loss": 0.0242,
+      "step": 11611
+    },
+    {
+      "epoch": 2.639090909090909,
+      "grad_norm": 1.506190630424679,
+      "learning_rate": 5.705375396972249e-07,
+      "loss": 0.0199,
+      "step": 11612
+    },
+    {
+      "epoch": 2.6393181818181817,
+      "grad_norm": 2.492493643861563,
+      "learning_rate": 5.704486299605159e-07,
+      "loss": 0.0285,
+      "step": 11613
+    },
+    {
+      "epoch": 2.6395454545454546,
+      "grad_norm": 1.6363103397115253,
+      "learning_rate": 5.703597213362045e-07,
+      "loss": 0.0235,
+      "step": 11614
+    },
+    {
+      "epoch": 2.639772727272727,
+      "grad_norm": 2.303403893775863,
+      "learning_rate": 5.702708138261036e-07,
+      "loss": 0.0373,
+      "step": 11615
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 2.428457114606061,
+      "learning_rate": 5.701819074320264e-07,
+      "loss": 0.03,
+      "step": 11616
+    },
+    {
+      "epoch": 2.6402272727272726,
+      "grad_norm": 1.691143899024865,
+      "learning_rate": 5.700930021557855e-07,
+      "loss": 0.0195,
+      "step": 11617
+    },
+    {
+      "epoch": 2.6404545454545456,
+      "grad_norm": 3.112934590084424,
+      "learning_rate": 5.700040979991942e-07,
+      "loss": 0.0312,
+      "step": 11618
+    },
+    {
+      "epoch": 2.640681818181818,
+      "grad_norm": 3.6030926609950735,
+      "learning_rate": 5.699151949640657e-07,
+      "loss": 0.0368,
+      "step": 11619
+    },
+    {
+      "epoch": 2.6409090909090907,
+      "grad_norm": 2.4777981646313045,
+      "learning_rate": 5.69826293052212e-07,
+      "loss": 0.0323,
+      "step": 11620
+    },
+    {
+      "epoch": 2.6411363636363636,
+      "grad_norm": 1.2281317405812793,
+      "learning_rate": 5.697373922654464e-07,
+      "loss": 0.0088,
+      "step": 11621
+    },
+    {
+      "epoch": 2.6413636363636366,
+      "grad_norm": 0.7822472357912046,
+      "learning_rate": 5.696484926055819e-07,
+      "loss": 0.0065,
+      "step": 11622
+    },
+    {
+      "epoch": 2.641590909090909,
+      "grad_norm": 1.6471897403156397,
+      "learning_rate": 5.695595940744315e-07,
+      "loss": 0.0186,
+      "step": 11623
+    },
+    {
+      "epoch": 2.6418181818181816,
+      "grad_norm": 1.922785473711195,
+      "learning_rate": 5.694706966738074e-07,
+      "loss": 0.0205,
+      "step": 11624
+    },
+    {
+      "epoch": 2.6420454545454546,
+      "grad_norm": 0.9403897835204048,
+      "learning_rate": 5.693818004055227e-07,
+      "loss": 0.0066,
+      "step": 11625
+    },
+    {
+      "epoch": 2.642272727272727,
+      "grad_norm": 1.9313041744755135,
+      "learning_rate": 5.692929052713901e-07,
+      "loss": 0.0224,
+      "step": 11626
+    },
+    {
+      "epoch": 2.6425,
+      "grad_norm": 1.5049944056969493,
+      "learning_rate": 5.692040112732225e-07,
+      "loss": 0.0161,
+      "step": 11627
+    },
+    {
+      "epoch": 2.6427272727272726,
+      "grad_norm": 1.0016097679889324,
+      "learning_rate": 5.691151184128323e-07,
+      "loss": 0.0075,
+      "step": 11628
+    },
+    {
+      "epoch": 2.6429545454545456,
+      "grad_norm": 1.8131747181708904,
+      "learning_rate": 5.690262266920323e-07,
+      "loss": 0.0243,
+      "step": 11629
+    },
+    {
+      "epoch": 2.643181818181818,
+      "grad_norm": 1.2345973495170306,
+      "learning_rate": 5.689373361126351e-07,
+      "loss": 0.0168,
+      "step": 11630
+    },
+    {
+      "epoch": 2.643409090909091,
+      "grad_norm": 1.8638407706086748,
+      "learning_rate": 5.688484466764537e-07,
+      "loss": 0.0329,
+      "step": 11631
+    },
+    {
+      "epoch": 2.6436363636363636,
+      "grad_norm": 3.190821756797882,
+      "learning_rate": 5.687595583853006e-07,
+      "loss": 0.0289,
+      "step": 11632
+    },
+    {
+      "epoch": 2.6438636363636365,
+      "grad_norm": 0.5203648481406806,
+      "learning_rate": 5.686706712409878e-07,
+      "loss": 0.0072,
+      "step": 11633
+    },
+    {
+      "epoch": 2.644090909090909,
+      "grad_norm": 2.6870542525355283,
+      "learning_rate": 5.685817852453284e-07,
+      "loss": 0.0204,
+      "step": 11634
+    },
+    {
+      "epoch": 2.644318181818182,
+      "grad_norm": 2.9972765068693628,
+      "learning_rate": 5.684929004001348e-07,
+      "loss": 0.0448,
+      "step": 11635
+    },
+    {
+      "epoch": 2.6445454545454545,
+      "grad_norm": 2.3401679175107724,
+      "learning_rate": 5.684040167072197e-07,
+      "loss": 0.0192,
+      "step": 11636
+    },
+    {
+      "epoch": 2.644772727272727,
+      "grad_norm": 3.971733033757204,
+      "learning_rate": 5.683151341683953e-07,
+      "loss": 0.0176,
+      "step": 11637
+    },
+    {
+      "epoch": 2.645,
+      "grad_norm": 2.2224448828032743,
+      "learning_rate": 5.682262527854743e-07,
+      "loss": 0.0244,
+      "step": 11638
+    },
+    {
+      "epoch": 2.645227272727273,
+      "grad_norm": 1.5363864293884315,
+      "learning_rate": 5.681373725602688e-07,
+      "loss": 0.0275,
+      "step": 11639
+    },
+    {
+      "epoch": 2.6454545454545455,
+      "grad_norm": 2.8171087131408554,
+      "learning_rate": 5.680484934945917e-07,
+      "loss": 0.0353,
+      "step": 11640
+    },
+    {
+      "epoch": 2.645681818181818,
+      "grad_norm": 2.45960568832658,
+      "learning_rate": 5.679596155902553e-07,
+      "loss": 0.0172,
+      "step": 11641
+    },
+    {
+      "epoch": 2.645909090909091,
+      "grad_norm": 1.7821067363331184,
+      "learning_rate": 5.678707388490717e-07,
+      "loss": 0.0296,
+      "step": 11642
+    },
+    {
+      "epoch": 2.6461363636363635,
+      "grad_norm": 0.8094265977143834,
+      "learning_rate": 5.677818632728533e-07,
+      "loss": 0.0101,
+      "step": 11643
+    },
+    {
+      "epoch": 2.6463636363636365,
+      "grad_norm": 1.3375417318965521,
+      "learning_rate": 5.676929888634126e-07,
+      "loss": 0.0165,
+      "step": 11644
+    },
+    {
+      "epoch": 2.646590909090909,
+      "grad_norm": 2.747542237012872,
+      "learning_rate": 5.67604115622562e-07,
+      "loss": 0.0471,
+      "step": 11645
+    },
+    {
+      "epoch": 2.646818181818182,
+      "grad_norm": 1.533605199739648,
+      "learning_rate": 5.675152435521133e-07,
+      "loss": 0.0158,
+      "step": 11646
+    },
+    {
+      "epoch": 2.6470454545454545,
+      "grad_norm": 1.5638513129237475,
+      "learning_rate": 5.674263726538793e-07,
+      "loss": 0.0166,
+      "step": 11647
+    },
+    {
+      "epoch": 2.6472727272727274,
+      "grad_norm": 1.9747005056568678,
+      "learning_rate": 5.673375029296719e-07,
+      "loss": 0.0214,
+      "step": 11648
+    },
+    {
+      "epoch": 2.6475,
+      "grad_norm": 2.678026610726486,
+      "learning_rate": 5.672486343813036e-07,
+      "loss": 0.0562,
+      "step": 11649
+    },
+    {
+      "epoch": 2.6477272727272725,
+      "grad_norm": 1.8626868073787266,
+      "learning_rate": 5.671597670105862e-07,
+      "loss": 0.022,
+      "step": 11650
+    },
+    {
+      "epoch": 2.6479545454545454,
+      "grad_norm": 1.0401889050693067,
+      "learning_rate": 5.67070900819332e-07,
+      "loss": 0.0164,
+      "step": 11651
+    },
+    {
+      "epoch": 2.6481818181818184,
+      "grad_norm": 2.0734932788997478,
+      "learning_rate": 5.669820358093533e-07,
+      "loss": 0.0153,
+      "step": 11652
+    },
+    {
+      "epoch": 2.648409090909091,
+      "grad_norm": 2.542237157084967,
+      "learning_rate": 5.668931719824621e-07,
+      "loss": 0.0468,
+      "step": 11653
+    },
+    {
+      "epoch": 2.6486363636363635,
+      "grad_norm": 1.547166025948475,
+      "learning_rate": 5.668043093404708e-07,
+      "loss": 0.0158,
+      "step": 11654
+    },
+    {
+      "epoch": 2.6488636363636364,
+      "grad_norm": 2.381518947311937,
+      "learning_rate": 5.667154478851908e-07,
+      "loss": 0.0365,
+      "step": 11655
+    },
+    {
+      "epoch": 2.649090909090909,
+      "grad_norm": 1.8496177773519333,
+      "learning_rate": 5.666265876184345e-07,
+      "loss": 0.0268,
+      "step": 11656
+    },
+    {
+      "epoch": 2.649318181818182,
+      "grad_norm": 1.1816399332759047,
+      "learning_rate": 5.66537728542014e-07,
+      "loss": 0.0134,
+      "step": 11657
+    },
+    {
+      "epoch": 2.6495454545454544,
+      "grad_norm": 1.3935507964757454,
+      "learning_rate": 5.664488706577412e-07,
+      "loss": 0.0112,
+      "step": 11658
+    },
+    {
+      "epoch": 2.6497727272727274,
+      "grad_norm": 1.401779742027964,
+      "learning_rate": 5.663600139674281e-07,
+      "loss": 0.0111,
+      "step": 11659
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.2408637066113388,
+      "learning_rate": 5.662711584728866e-07,
+      "loss": 0.0102,
+      "step": 11660
+    },
+    {
+      "epoch": 2.650227272727273,
+      "grad_norm": 2.5593388712627463,
+      "learning_rate": 5.661823041759286e-07,
+      "loss": 0.026,
+      "step": 11661
+    },
+    {
+      "epoch": 2.6504545454545454,
+      "grad_norm": 1.874001054771184,
+      "learning_rate": 5.660934510783659e-07,
+      "loss": 0.0131,
+      "step": 11662
+    },
+    {
+      "epoch": 2.650681818181818,
+      "grad_norm": 1.9940802384472804,
+      "learning_rate": 5.66004599182011e-07,
+      "loss": 0.0275,
+      "step": 11663
+    },
+    {
+      "epoch": 2.650909090909091,
+      "grad_norm": 1.249893591659597,
+      "learning_rate": 5.659157484886749e-07,
+      "loss": 0.0155,
+      "step": 11664
+    },
+    {
+      "epoch": 2.651136363636364,
+      "grad_norm": 1.7393820571867116,
+      "learning_rate": 5.658268990001696e-07,
+      "loss": 0.0302,
+      "step": 11665
+    },
+    {
+      "epoch": 2.6513636363636364,
+      "grad_norm": 1.7437971666009464,
+      "learning_rate": 5.657380507183073e-07,
+      "loss": 0.0143,
+      "step": 11666
+    },
+    {
+      "epoch": 2.651590909090909,
+      "grad_norm": 2.1477113679531543,
+      "learning_rate": 5.656492036448996e-07,
+      "loss": 0.023,
+      "step": 11667
+    },
+    {
+      "epoch": 2.651818181818182,
+      "grad_norm": 3.0187399203732634,
+      "learning_rate": 5.65560357781758e-07,
+      "loss": 0.0359,
+      "step": 11668
+    },
+    {
+      "epoch": 2.652045454545455,
+      "grad_norm": 2.2827218414205754,
+      "learning_rate": 5.654715131306944e-07,
+      "loss": 0.0709,
+      "step": 11669
+    },
+    {
+      "epoch": 2.6522727272727273,
+      "grad_norm": 5.303754444887579,
+      "learning_rate": 5.653826696935206e-07,
+      "loss": 0.052,
+      "step": 11670
+    },
+    {
+      "epoch": 2.6525,
+      "grad_norm": 2.0212402250952866,
+      "learning_rate": 5.652938274720482e-07,
+      "loss": 0.0194,
+      "step": 11671
+    },
+    {
+      "epoch": 2.652727272727273,
+      "grad_norm": 1.7238121649391074,
+      "learning_rate": 5.652049864680888e-07,
+      "loss": 0.0268,
+      "step": 11672
+    },
+    {
+      "epoch": 2.6529545454545453,
+      "grad_norm": 1.521831210583644,
+      "learning_rate": 5.65116146683454e-07,
+      "loss": 0.0144,
+      "step": 11673
+    },
+    {
+      "epoch": 2.6531818181818183,
+      "grad_norm": 1.3417469964586324,
+      "learning_rate": 5.650273081199552e-07,
+      "loss": 0.0081,
+      "step": 11674
+    },
+    {
+      "epoch": 2.653409090909091,
+      "grad_norm": 1.4380875886651443,
+      "learning_rate": 5.649384707794043e-07,
+      "loss": 0.02,
+      "step": 11675
+    },
+    {
+      "epoch": 2.653636363636364,
+      "grad_norm": 3.6399004729330833,
+      "learning_rate": 5.648496346636132e-07,
+      "loss": 0.0327,
+      "step": 11676
+    },
+    {
+      "epoch": 2.6538636363636363,
+      "grad_norm": 2.869969609646063,
+      "learning_rate": 5.647607997743927e-07,
+      "loss": 0.0521,
+      "step": 11677
+    },
+    {
+      "epoch": 2.6540909090909093,
+      "grad_norm": 2.0630440531009246,
+      "learning_rate": 5.646719661135545e-07,
+      "loss": 0.0313,
+      "step": 11678
+    },
+    {
+      "epoch": 2.654318181818182,
+      "grad_norm": 5.125080080864651,
+      "learning_rate": 5.645831336829102e-07,
+      "loss": 0.0303,
+      "step": 11679
+    },
+    {
+      "epoch": 2.6545454545454543,
+      "grad_norm": 2.0021513006111458,
+      "learning_rate": 5.644943024842714e-07,
+      "loss": 0.0182,
+      "step": 11680
+    },
+    {
+      "epoch": 2.6547727272727273,
+      "grad_norm": 2.3941887371632107,
+      "learning_rate": 5.644054725194492e-07,
+      "loss": 0.0331,
+      "step": 11681
+    },
+    {
+      "epoch": 2.6550000000000002,
+      "grad_norm": 2.910441419262804,
+      "learning_rate": 5.64316643790255e-07,
+      "loss": 0.0404,
+      "step": 11682
+    },
+    {
+      "epoch": 2.6552272727272728,
+      "grad_norm": 2.317869967105031,
+      "learning_rate": 5.642278162985005e-07,
+      "loss": 0.0416,
+      "step": 11683
+    },
+    {
+      "epoch": 2.6554545454545453,
+      "grad_norm": 1.5944804863852433,
+      "learning_rate": 5.641389900459967e-07,
+      "loss": 0.0206,
+      "step": 11684
+    },
+    {
+      "epoch": 2.6556818181818183,
+      "grad_norm": 1.9850710373163156,
+      "learning_rate": 5.640501650345552e-07,
+      "loss": 0.0298,
+      "step": 11685
+    },
+    {
+      "epoch": 2.6559090909090908,
+      "grad_norm": 1.7157326313544718,
+      "learning_rate": 5.639613412659872e-07,
+      "loss": 0.0149,
+      "step": 11686
+    },
+    {
+      "epoch": 2.6561363636363637,
+      "grad_norm": 2.4250198158403524,
+      "learning_rate": 5.638725187421037e-07,
+      "loss": 0.034,
+      "step": 11687
+    },
+    {
+      "epoch": 2.6563636363636363,
+      "grad_norm": 1.1147744479390593,
+      "learning_rate": 5.637836974647163e-07,
+      "loss": 0.009,
+      "step": 11688
+    },
+    {
+      "epoch": 2.6565909090909092,
+      "grad_norm": 3.51374834286157,
+      "learning_rate": 5.636948774356363e-07,
+      "loss": 0.072,
+      "step": 11689
+    },
+    {
+      "epoch": 2.6568181818181817,
+      "grad_norm": 0.675484621717108,
+      "learning_rate": 5.636060586566744e-07,
+      "loss": 0.0036,
+      "step": 11690
+    },
+    {
+      "epoch": 2.6570454545454547,
+      "grad_norm": 1.9230672186243043,
+      "learning_rate": 5.635172411296423e-07,
+      "loss": 0.0302,
+      "step": 11691
+    },
+    {
+      "epoch": 2.6572727272727272,
+      "grad_norm": 2.2639913729697736,
+      "learning_rate": 5.634284248563507e-07,
+      "loss": 0.0234,
+      "step": 11692
+    },
+    {
+      "epoch": 2.6574999999999998,
+      "grad_norm": 1.4399427183142417,
+      "learning_rate": 5.633396098386113e-07,
+      "loss": 0.0147,
+      "step": 11693
+    },
+    {
+      "epoch": 2.6577272727272727,
+      "grad_norm": 2.234757510705809,
+      "learning_rate": 5.632507960782346e-07,
+      "loss": 0.0488,
+      "step": 11694
+    },
+    {
+      "epoch": 2.6579545454545457,
+      "grad_norm": 2.185788109214478,
+      "learning_rate": 5.631619835770318e-07,
+      "loss": 0.05,
+      "step": 11695
+    },
+    {
+      "epoch": 2.658181818181818,
+      "grad_norm": 5.633799815721707,
+      "learning_rate": 5.630731723368141e-07,
+      "loss": 0.0493,
+      "step": 11696
+    },
+    {
+      "epoch": 2.6584090909090907,
+      "grad_norm": 1.8236865956307349,
+      "learning_rate": 5.629843623593923e-07,
+      "loss": 0.0177,
+      "step": 11697
+    },
+    {
+      "epoch": 2.6586363636363637,
+      "grad_norm": 1.7696618743371961,
+      "learning_rate": 5.62895553646578e-07,
+      "loss": 0.0191,
+      "step": 11698
+    },
+    {
+      "epoch": 2.658863636363636,
+      "grad_norm": 1.4851714295733327,
+      "learning_rate": 5.628067462001813e-07,
+      "loss": 0.0221,
+      "step": 11699
+    },
+    {
+      "epoch": 2.659090909090909,
+      "grad_norm": 1.9192272161523851,
+      "learning_rate": 5.627179400220136e-07,
+      "loss": 0.0532,
+      "step": 11700
+    },
+    {
+      "epoch": 2.6593181818181817,
+      "grad_norm": 1.843204561311486,
+      "learning_rate": 5.626291351138858e-07,
+      "loss": 0.0339,
+      "step": 11701
+    },
+    {
+      "epoch": 2.6595454545454547,
+      "grad_norm": 1.2696632871317362,
+      "learning_rate": 5.62540331477609e-07,
+      "loss": 0.0203,
+      "step": 11702
+    },
+    {
+      "epoch": 2.659772727272727,
+      "grad_norm": 1.4617805035731177,
+      "learning_rate": 5.624515291149936e-07,
+      "loss": 0.012,
+      "step": 11703
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.1649317158644226,
+      "learning_rate": 5.623627280278505e-07,
+      "loss": 0.0336,
+      "step": 11704
+    },
+    {
+      "epoch": 2.6602272727272727,
+      "grad_norm": 2.101738249471541,
+      "learning_rate": 5.622739282179907e-07,
+      "loss": 0.0285,
+      "step": 11705
+    },
+    {
+      "epoch": 2.660454545454545,
+      "grad_norm": 1.9310553260553964,
+      "learning_rate": 5.62185129687225e-07,
+      "loss": 0.0227,
+      "step": 11706
+    },
+    {
+      "epoch": 2.660681818181818,
+      "grad_norm": 1.7321479747964172,
+      "learning_rate": 5.620963324373643e-07,
+      "loss": 0.0138,
+      "step": 11707
+    },
+    {
+      "epoch": 2.660909090909091,
+      "grad_norm": 1.142396096902504,
+      "learning_rate": 5.620075364702189e-07,
+      "loss": 0.0079,
+      "step": 11708
+    },
+    {
+      "epoch": 2.6611363636363636,
+      "grad_norm": 3.102471670960751,
+      "learning_rate": 5.619187417875997e-07,
+      "loss": 0.0333,
+      "step": 11709
+    },
+    {
+      "epoch": 2.661363636363636,
+      "grad_norm": 2.3750658592747085,
+      "learning_rate": 5.618299483913174e-07,
+      "loss": 0.0168,
+      "step": 11710
+    },
+    {
+      "epoch": 2.661590909090909,
+      "grad_norm": 2.0713974408866545,
+      "learning_rate": 5.617411562831829e-07,
+      "loss": 0.0206,
+      "step": 11711
+    },
+    {
+      "epoch": 2.661818181818182,
+      "grad_norm": 2.037987983971663,
+      "learning_rate": 5.616523654650063e-07,
+      "loss": 0.039,
+      "step": 11712
+    },
+    {
+      "epoch": 2.6620454545454546,
+      "grad_norm": 1.7804784783121421,
+      "learning_rate": 5.615635759385985e-07,
+      "loss": 0.014,
+      "step": 11713
+    },
+    {
+      "epoch": 2.662272727272727,
+      "grad_norm": 2.7741747069409617,
+      "learning_rate": 5.614747877057702e-07,
+      "loss": 0.0419,
+      "step": 11714
+    },
+    {
+      "epoch": 2.6625,
+      "grad_norm": 1.1808491020759337,
+      "learning_rate": 5.613860007683319e-07,
+      "loss": 0.0176,
+      "step": 11715
+    },
+    {
+      "epoch": 2.6627272727272726,
+      "grad_norm": 1.8628431636041491,
+      "learning_rate": 5.612972151280939e-07,
+      "loss": 0.0211,
+      "step": 11716
+    },
+    {
+      "epoch": 2.6629545454545456,
+      "grad_norm": 1.80893634796343,
+      "learning_rate": 5.612084307868667e-07,
+      "loss": 0.0331,
+      "step": 11717
+    },
+    {
+      "epoch": 2.663181818181818,
+      "grad_norm": 2.817902256285721,
+      "learning_rate": 5.61119647746461e-07,
+      "loss": 0.0418,
+      "step": 11718
+    },
+    {
+      "epoch": 2.663409090909091,
+      "grad_norm": 1.158316414550328,
+      "learning_rate": 5.61030866008687e-07,
+      "loss": 0.0186,
+      "step": 11719
+    },
+    {
+      "epoch": 2.6636363636363636,
+      "grad_norm": 2.6115188095721904,
+      "learning_rate": 5.609420855753555e-07,
+      "loss": 0.0371,
+      "step": 11720
+    },
+    {
+      "epoch": 2.6638636363636365,
+      "grad_norm": 1.185256526796921,
+      "learning_rate": 5.608533064482764e-07,
+      "loss": 0.0135,
+      "step": 11721
+    },
+    {
+      "epoch": 2.664090909090909,
+      "grad_norm": 1.5835394248493464,
+      "learning_rate": 5.607645286292604e-07,
+      "loss": 0.0154,
+      "step": 11722
+    },
+    {
+      "epoch": 2.6643181818181816,
+      "grad_norm": 4.5669218586589215,
+      "learning_rate": 5.606757521201176e-07,
+      "loss": 0.0352,
+      "step": 11723
+    },
+    {
+      "epoch": 2.6645454545454546,
+      "grad_norm": 1.68076961030417,
+      "learning_rate": 5.605869769226589e-07,
+      "loss": 0.0335,
+      "step": 11724
+    },
+    {
+      "epoch": 2.6647727272727275,
+      "grad_norm": 2.458847327564256,
+      "learning_rate": 5.604982030386936e-07,
+      "loss": 0.0251,
+      "step": 11725
+    },
+    {
+      "epoch": 2.665,
+      "grad_norm": 1.292374141186261,
+      "learning_rate": 5.604094304700326e-07,
+      "loss": 0.0149,
+      "step": 11726
+    },
+    {
+      "epoch": 2.6652272727272726,
+      "grad_norm": 1.3718629615325926,
+      "learning_rate": 5.60320659218486e-07,
+      "loss": 0.0133,
+      "step": 11727
+    },
+    {
+      "epoch": 2.6654545454545455,
+      "grad_norm": 1.004003610767137,
+      "learning_rate": 5.60231889285864e-07,
+      "loss": 0.0084,
+      "step": 11728
+    },
+    {
+      "epoch": 2.665681818181818,
+      "grad_norm": 1.9140007428558863,
+      "learning_rate": 5.601431206739769e-07,
+      "loss": 0.052,
+      "step": 11729
+    },
+    {
+      "epoch": 2.665909090909091,
+      "grad_norm": 1.7369144746657705,
+      "learning_rate": 5.600543533846345e-07,
+      "loss": 0.0135,
+      "step": 11730
+    },
+    {
+      "epoch": 2.6661363636363635,
+      "grad_norm": 2.574057922829251,
+      "learning_rate": 5.599655874196473e-07,
+      "loss": 0.0219,
+      "step": 11731
+    },
+    {
+      "epoch": 2.6663636363636365,
+      "grad_norm": 2.9302920013248377,
+      "learning_rate": 5.598768227808251e-07,
+      "loss": 0.0227,
+      "step": 11732
+    },
+    {
+      "epoch": 2.666590909090909,
+      "grad_norm": 1.8672461584225577,
+      "learning_rate": 5.597880594699783e-07,
+      "loss": 0.0149,
+      "step": 11733
+    },
+    {
+      "epoch": 2.666818181818182,
+      "grad_norm": 1.6424303546436871,
+      "learning_rate": 5.596992974889165e-07,
+      "loss": 0.0186,
+      "step": 11734
+    },
+    {
+      "epoch": 2.6670454545454545,
+      "grad_norm": 1.4719765271237462,
+      "learning_rate": 5.596105368394501e-07,
+      "loss": 0.0241,
+      "step": 11735
+    },
+    {
+      "epoch": 2.667272727272727,
+      "grad_norm": 1.2003475837753352,
+      "learning_rate": 5.595217775233889e-07,
+      "loss": 0.011,
+      "step": 11736
+    },
+    {
+      "epoch": 2.6675,
+      "grad_norm": 1.9456495414136576,
+      "learning_rate": 5.594330195425429e-07,
+      "loss": 0.018,
+      "step": 11737
+    },
+    {
+      "epoch": 2.667727272727273,
+      "grad_norm": 1.2461231476041363,
+      "learning_rate": 5.593442628987221e-07,
+      "loss": 0.0198,
+      "step": 11738
+    },
+    {
+      "epoch": 2.6679545454545455,
+      "grad_norm": 1.9145554562655327,
+      "learning_rate": 5.592555075937362e-07,
+      "loss": 0.0277,
+      "step": 11739
+    },
+    {
+      "epoch": 2.668181818181818,
+      "grad_norm": 1.7700291540876243,
+      "learning_rate": 5.591667536293951e-07,
+      "loss": 0.0194,
+      "step": 11740
+    },
+    {
+      "epoch": 2.668409090909091,
+      "grad_norm": 1.6869580663325903,
+      "learning_rate": 5.590780010075088e-07,
+      "loss": 0.0253,
+      "step": 11741
+    },
+    {
+      "epoch": 2.6686363636363635,
+      "grad_norm": 2.4204297391496814,
+      "learning_rate": 5.589892497298873e-07,
+      "loss": 0.0532,
+      "step": 11742
+    },
+    {
+      "epoch": 2.6688636363636364,
+      "grad_norm": 1.6265701994143735,
+      "learning_rate": 5.5890049979834e-07,
+      "loss": 0.0211,
+      "step": 11743
+    },
+    {
+      "epoch": 2.669090909090909,
+      "grad_norm": 1.4368382113852427,
+      "learning_rate": 5.588117512146767e-07,
+      "loss": 0.0181,
+      "step": 11744
+    },
+    {
+      "epoch": 2.669318181818182,
+      "grad_norm": 1.790686890886762,
+      "learning_rate": 5.587230039807073e-07,
+      "loss": 0.0181,
+      "step": 11745
+    },
+    {
+      "epoch": 2.6695454545454544,
+      "grad_norm": 2.0759181127882633,
+      "learning_rate": 5.586342580982419e-07,
+      "loss": 0.0186,
+      "step": 11746
+    },
+    {
+      "epoch": 2.6697727272727274,
+      "grad_norm": 1.3803804799008,
+      "learning_rate": 5.585455135690893e-07,
+      "loss": 0.0197,
+      "step": 11747
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.060701273446355,
+      "learning_rate": 5.584567703950597e-07,
+      "loss": 0.0048,
+      "step": 11748
+    },
+    {
+      "epoch": 2.670227272727273,
+      "grad_norm": 2.065153321108138,
+      "learning_rate": 5.583680285779626e-07,
+      "loss": 0.062,
+      "step": 11749
+    },
+    {
+      "epoch": 2.6704545454545454,
+      "grad_norm": 2.2144713499968227,
+      "learning_rate": 5.582792881196078e-07,
+      "loss": 0.0203,
+      "step": 11750
+    },
+    {
+      "epoch": 2.6706818181818184,
+      "grad_norm": 3.1849836551611626,
+      "learning_rate": 5.581905490218047e-07,
+      "loss": 0.0367,
+      "step": 11751
+    },
+    {
+      "epoch": 2.670909090909091,
+      "grad_norm": 2.383897492502974,
+      "learning_rate": 5.581018112863627e-07,
+      "loss": 0.023,
+      "step": 11752
+    },
+    {
+      "epoch": 2.6711363636363634,
+      "grad_norm": 0.8597196090334893,
+      "learning_rate": 5.580130749150915e-07,
+      "loss": 0.0166,
+      "step": 11753
+    },
+    {
+      "epoch": 2.6713636363636364,
+      "grad_norm": 2.7156516615504143,
+      "learning_rate": 5.579243399098006e-07,
+      "loss": 0.0255,
+      "step": 11754
+    },
+    {
+      "epoch": 2.6715909090909093,
+      "grad_norm": 3.4919408266315077,
+      "learning_rate": 5.578356062722997e-07,
+      "loss": 0.0278,
+      "step": 11755
+    },
+    {
+      "epoch": 2.671818181818182,
+      "grad_norm": 2.6407365150526485,
+      "learning_rate": 5.577468740043978e-07,
+      "loss": 0.0377,
+      "step": 11756
+    },
+    {
+      "epoch": 2.6720454545454544,
+      "grad_norm": 0.751993660020604,
+      "learning_rate": 5.576581431079043e-07,
+      "loss": 0.0092,
+      "step": 11757
+    },
+    {
+      "epoch": 2.6722727272727274,
+      "grad_norm": 0.8757454236023996,
+      "learning_rate": 5.575694135846287e-07,
+      "loss": 0.0184,
+      "step": 11758
+    },
+    {
+      "epoch": 2.6725,
+      "grad_norm": 3.521923532402986,
+      "learning_rate": 5.574806854363807e-07,
+      "loss": 0.0385,
+      "step": 11759
+    },
+    {
+      "epoch": 2.672727272727273,
+      "grad_norm": 0.9626983382755927,
+      "learning_rate": 5.573919586649691e-07,
+      "loss": 0.0086,
+      "step": 11760
+    },
+    {
+      "epoch": 2.6729545454545454,
+      "grad_norm": 2.1167857937030687,
+      "learning_rate": 5.573032332722035e-07,
+      "loss": 0.0224,
+      "step": 11761
+    },
+    {
+      "epoch": 2.6731818181818183,
+      "grad_norm": 0.5589008875327882,
+      "learning_rate": 5.572145092598929e-07,
+      "loss": 0.0025,
+      "step": 11762
+    },
+    {
+      "epoch": 2.673409090909091,
+      "grad_norm": 1.2296604462766012,
+      "learning_rate": 5.571257866298469e-07,
+      "loss": 0.0137,
+      "step": 11763
+    },
+    {
+      "epoch": 2.673636363636364,
+      "grad_norm": 3.1865008838064086,
+      "learning_rate": 5.570370653838746e-07,
+      "loss": 0.0331,
+      "step": 11764
+    },
+    {
+      "epoch": 2.6738636363636363,
+      "grad_norm": 1.0654309818025864,
+      "learning_rate": 5.569483455237849e-07,
+      "loss": 0.0195,
+      "step": 11765
+    },
+    {
+      "epoch": 2.674090909090909,
+      "grad_norm": 1.7903892217947472,
+      "learning_rate": 5.568596270513873e-07,
+      "loss": 0.0241,
+      "step": 11766
+    },
+    {
+      "epoch": 2.674318181818182,
+      "grad_norm": 1.314382943227654,
+      "learning_rate": 5.567709099684906e-07,
+      "loss": 0.0262,
+      "step": 11767
+    },
+    {
+      "epoch": 2.674545454545455,
+      "grad_norm": 2.0279168577938034,
+      "learning_rate": 5.566821942769045e-07,
+      "loss": 0.0312,
+      "step": 11768
+    },
+    {
+      "epoch": 2.6747727272727273,
+      "grad_norm": 1.539058308705292,
+      "learning_rate": 5.565934799784372e-07,
+      "loss": 0.0148,
+      "step": 11769
+    },
+    {
+      "epoch": 2.675,
+      "grad_norm": 2.6074328937348925,
+      "learning_rate": 5.565047670748984e-07,
+      "loss": 0.0382,
+      "step": 11770
+    },
+    {
+      "epoch": 2.675227272727273,
+      "grad_norm": 1.7300643659116544,
+      "learning_rate": 5.564160555680967e-07,
+      "loss": 0.0096,
+      "step": 11771
+    },
+    {
+      "epoch": 2.6754545454545453,
+      "grad_norm": 0.8771764752426321,
+      "learning_rate": 5.563273454598415e-07,
+      "loss": 0.0064,
+      "step": 11772
+    },
+    {
+      "epoch": 2.6756818181818183,
+      "grad_norm": 1.4218015295449462,
+      "learning_rate": 5.562386367519415e-07,
+      "loss": 0.0136,
+      "step": 11773
+    },
+    {
+      "epoch": 2.675909090909091,
+      "grad_norm": 1.3578695828885632,
+      "learning_rate": 5.561499294462055e-07,
+      "loss": 0.0159,
+      "step": 11774
+    },
+    {
+      "epoch": 2.6761363636363638,
+      "grad_norm": 1.2044872201860461,
+      "learning_rate": 5.560612235444426e-07,
+      "loss": 0.0165,
+      "step": 11775
+    },
+    {
+      "epoch": 2.6763636363636363,
+      "grad_norm": 1.6556172742732596,
+      "learning_rate": 5.559725190484617e-07,
+      "loss": 0.0192,
+      "step": 11776
+    },
+    {
+      "epoch": 2.6765909090909092,
+      "grad_norm": 2.318123706038659,
+      "learning_rate": 5.558838159600718e-07,
+      "loss": 0.0186,
+      "step": 11777
+    },
+    {
+      "epoch": 2.6768181818181818,
+      "grad_norm": 1.2501311067844543,
+      "learning_rate": 5.557951142810811e-07,
+      "loss": 0.0169,
+      "step": 11778
+    },
+    {
+      "epoch": 2.6770454545454543,
+      "grad_norm": 2.1608249751901663,
+      "learning_rate": 5.557064140132988e-07,
+      "loss": 0.0205,
+      "step": 11779
+    },
+    {
+      "epoch": 2.6772727272727272,
+      "grad_norm": 1.8539768962100007,
+      "learning_rate": 5.556177151585337e-07,
+      "loss": 0.0166,
+      "step": 11780
+    },
+    {
+      "epoch": 2.6775,
+      "grad_norm": 2.5416802098485154,
+      "learning_rate": 5.555290177185945e-07,
+      "loss": 0.0283,
+      "step": 11781
+    },
+    {
+      "epoch": 2.6777272727272727,
+      "grad_norm": 1.3676811693763737,
+      "learning_rate": 5.554403216952897e-07,
+      "loss": 0.0135,
+      "step": 11782
+    },
+    {
+      "epoch": 2.6779545454545453,
+      "grad_norm": 1.1452923767180896,
+      "learning_rate": 5.55351627090428e-07,
+      "loss": 0.0081,
+      "step": 11783
+    },
+    {
+      "epoch": 2.678181818181818,
+      "grad_norm": 0.9488523775152115,
+      "learning_rate": 5.552629339058183e-07,
+      "loss": 0.0117,
+      "step": 11784
+    },
+    {
+      "epoch": 2.678409090909091,
+      "grad_norm": 1.6051953677117465,
+      "learning_rate": 5.55174242143269e-07,
+      "loss": 0.018,
+      "step": 11785
+    },
+    {
+      "epoch": 2.6786363636363637,
+      "grad_norm": 1.3824103645026506,
+      "learning_rate": 5.550855518045889e-07,
+      "loss": 0.009,
+      "step": 11786
+    },
+    {
+      "epoch": 2.6788636363636362,
+      "grad_norm": 1.5405533464522534,
+      "learning_rate": 5.549968628915862e-07,
+      "loss": 0.0223,
+      "step": 11787
+    },
+    {
+      "epoch": 2.679090909090909,
+      "grad_norm": 1.924130698580498,
+      "learning_rate": 5.549081754060694e-07,
+      "loss": 0.0242,
+      "step": 11788
+    },
+    {
+      "epoch": 2.6793181818181817,
+      "grad_norm": 3.4565809127903235,
+      "learning_rate": 5.548194893498473e-07,
+      "loss": 0.0361,
+      "step": 11789
+    },
+    {
+      "epoch": 2.6795454545454547,
+      "grad_norm": 1.2641205604357373,
+      "learning_rate": 5.547308047247286e-07,
+      "loss": 0.0076,
+      "step": 11790
+    },
+    {
+      "epoch": 2.679772727272727,
+      "grad_norm": 3.879607946191189,
+      "learning_rate": 5.54642121532521e-07,
+      "loss": 0.0367,
+      "step": 11791
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.4436080121048696,
+      "learning_rate": 5.545534397750334e-07,
+      "loss": 0.0182,
+      "step": 11792
+    },
+    {
+      "epoch": 2.6802272727272727,
+      "grad_norm": 1.4637749824723374,
+      "learning_rate": 5.544647594540739e-07,
+      "loss": 0.0134,
+      "step": 11793
+    },
+    {
+      "epoch": 2.6804545454545456,
+      "grad_norm": 0.865098460967603,
+      "learning_rate": 5.543760805714513e-07,
+      "loss": 0.008,
+      "step": 11794
+    },
+    {
+      "epoch": 2.680681818181818,
+      "grad_norm": 2.099268696334918,
+      "learning_rate": 5.542874031289734e-07,
+      "loss": 0.0598,
+      "step": 11795
+    },
+    {
+      "epoch": 2.6809090909090907,
+      "grad_norm": 1.0620080936776275,
+      "learning_rate": 5.541987271284489e-07,
+      "loss": 0.0112,
+      "step": 11796
+    },
+    {
+      "epoch": 2.6811363636363637,
+      "grad_norm": 1.2134726684652228,
+      "learning_rate": 5.541100525716857e-07,
+      "loss": 0.0201,
+      "step": 11797
+    },
+    {
+      "epoch": 2.6813636363636366,
+      "grad_norm": 3.568616312070019,
+      "learning_rate": 5.540213794604923e-07,
+      "loss": 0.0411,
+      "step": 11798
+    },
+    {
+      "epoch": 2.681590909090909,
+      "grad_norm": 1.0730991393539329,
+      "learning_rate": 5.53932707796677e-07,
+      "loss": 0.0133,
+      "step": 11799
+    },
+    {
+      "epoch": 2.6818181818181817,
+      "grad_norm": 2.051456862166433,
+      "learning_rate": 5.538440375820474e-07,
+      "loss": 0.0253,
+      "step": 11800
+    },
+    {
+      "epoch": 2.6820454545454546,
+      "grad_norm": 1.1897919407336337,
+      "learning_rate": 5.537553688184122e-07,
+      "loss": 0.0096,
+      "step": 11801
+    },
+    {
+      "epoch": 2.682272727272727,
+      "grad_norm": 1.7942929709334432,
+      "learning_rate": 5.536667015075792e-07,
+      "loss": 0.0209,
+      "step": 11802
+    },
+    {
+      "epoch": 2.6825,
+      "grad_norm": 2.00214584908174,
+      "learning_rate": 5.535780356513568e-07,
+      "loss": 0.0143,
+      "step": 11803
+    },
+    {
+      "epoch": 2.6827272727272726,
+      "grad_norm": 0.9934076589666009,
+      "learning_rate": 5.534893712515528e-07,
+      "loss": 0.012,
+      "step": 11804
+    },
+    {
+      "epoch": 2.6829545454545456,
+      "grad_norm": 1.7266581708563211,
+      "learning_rate": 5.534007083099752e-07,
+      "loss": 0.018,
+      "step": 11805
+    },
+    {
+      "epoch": 2.683181818181818,
+      "grad_norm": 1.4972431226756482,
+      "learning_rate": 5.53312046828432e-07,
+      "loss": 0.0144,
+      "step": 11806
+    },
+    {
+      "epoch": 2.683409090909091,
+      "grad_norm": 5.134834487636778,
+      "learning_rate": 5.532233868087314e-07,
+      "loss": 0.0411,
+      "step": 11807
+    },
+    {
+      "epoch": 2.6836363636363636,
+      "grad_norm": 1.286519734194517,
+      "learning_rate": 5.531347282526813e-07,
+      "loss": 0.0145,
+      "step": 11808
+    },
+    {
+      "epoch": 2.683863636363636,
+      "grad_norm": 0.6568639175828387,
+      "learning_rate": 5.530460711620892e-07,
+      "loss": 0.0088,
+      "step": 11809
+    },
+    {
+      "epoch": 2.684090909090909,
+      "grad_norm": 3.29361466013791,
+      "learning_rate": 5.529574155387632e-07,
+      "loss": 0.0672,
+      "step": 11810
+    },
+    {
+      "epoch": 2.684318181818182,
+      "grad_norm": 1.853412377707427,
+      "learning_rate": 5.528687613845112e-07,
+      "loss": 0.0309,
+      "step": 11811
+    },
+    {
+      "epoch": 2.6845454545454546,
+      "grad_norm": 1.8110742440033245,
+      "learning_rate": 5.527801087011412e-07,
+      "loss": 0.0241,
+      "step": 11812
+    },
+    {
+      "epoch": 2.684772727272727,
+      "grad_norm": 2.087332566966979,
+      "learning_rate": 5.526914574904606e-07,
+      "loss": 0.0176,
+      "step": 11813
+    },
+    {
+      "epoch": 2.685,
+      "grad_norm": 2.356740500241892,
+      "learning_rate": 5.526028077542774e-07,
+      "loss": 0.0423,
+      "step": 11814
+    },
+    {
+      "epoch": 2.6852272727272726,
+      "grad_norm": 2.3206070261579326,
+      "learning_rate": 5.525141594943992e-07,
+      "loss": 0.0332,
+      "step": 11815
+    },
+    {
+      "epoch": 2.6854545454545455,
+      "grad_norm": 1.2727575147492545,
+      "learning_rate": 5.52425512712634e-07,
+      "loss": 0.0145,
+      "step": 11816
+    },
+    {
+      "epoch": 2.685681818181818,
+      "grad_norm": 2.9742494948699747,
+      "learning_rate": 5.523368674107888e-07,
+      "loss": 0.0477,
+      "step": 11817
+    },
+    {
+      "epoch": 2.685909090909091,
+      "grad_norm": 1.8112457729716072,
+      "learning_rate": 5.522482235906719e-07,
+      "loss": 0.0345,
+      "step": 11818
+    },
+    {
+      "epoch": 2.6861363636363635,
+      "grad_norm": 1.3728673809335794,
+      "learning_rate": 5.521595812540904e-07,
+      "loss": 0.0154,
+      "step": 11819
+    },
+    {
+      "epoch": 2.6863636363636365,
+      "grad_norm": 1.5990048861174673,
+      "learning_rate": 5.520709404028523e-07,
+      "loss": 0.0104,
+      "step": 11820
+    },
+    {
+      "epoch": 2.686590909090909,
+      "grad_norm": 2.4597303088546414,
+      "learning_rate": 5.519823010387651e-07,
+      "loss": 0.0323,
+      "step": 11821
+    },
+    {
+      "epoch": 2.6868181818181816,
+      "grad_norm": 2.6593452729050044,
+      "learning_rate": 5.51893663163636e-07,
+      "loss": 0.0269,
+      "step": 11822
+    },
+    {
+      "epoch": 2.6870454545454545,
+      "grad_norm": 2.4466077494380447,
+      "learning_rate": 5.518050267792727e-07,
+      "loss": 0.0334,
+      "step": 11823
+    },
+    {
+      "epoch": 2.6872727272727275,
+      "grad_norm": 1.96592270099336,
+      "learning_rate": 5.517163918874825e-07,
+      "loss": 0.0344,
+      "step": 11824
+    },
+    {
+      "epoch": 2.6875,
+      "grad_norm": 2.464891663951074,
+      "learning_rate": 5.516277584900731e-07,
+      "loss": 0.0364,
+      "step": 11825
+    },
+    {
+      "epoch": 2.6877272727272725,
+      "grad_norm": 2.4527671971438694,
+      "learning_rate": 5.515391265888516e-07,
+      "loss": 0.0486,
+      "step": 11826
+    },
+    {
+      "epoch": 2.6879545454545455,
+      "grad_norm": 1.3478447258883162,
+      "learning_rate": 5.514504961856254e-07,
+      "loss": 0.0073,
+      "step": 11827
+    },
+    {
+      "epoch": 2.6881818181818184,
+      "grad_norm": 2.2552271108855315,
+      "learning_rate": 5.513618672822018e-07,
+      "loss": 0.0106,
+      "step": 11828
+    },
+    {
+      "epoch": 2.688409090909091,
+      "grad_norm": 1.9465557891376832,
+      "learning_rate": 5.512732398803884e-07,
+      "loss": 0.017,
+      "step": 11829
+    },
+    {
+      "epoch": 2.6886363636363635,
+      "grad_norm": 1.7892539739541222,
+      "learning_rate": 5.511846139819925e-07,
+      "loss": 0.0218,
+      "step": 11830
+    },
+    {
+      "epoch": 2.6888636363636365,
+      "grad_norm": 1.6098864043599865,
+      "learning_rate": 5.510959895888207e-07,
+      "loss": 0.0279,
+      "step": 11831
+    },
+    {
+      "epoch": 2.689090909090909,
+      "grad_norm": 2.551348968368078,
+      "learning_rate": 5.510073667026807e-07,
+      "loss": 0.049,
+      "step": 11832
+    },
+    {
+      "epoch": 2.689318181818182,
+      "grad_norm": 2.691833457252325,
+      "learning_rate": 5.509187453253796e-07,
+      "loss": 0.0401,
+      "step": 11833
+    },
+    {
+      "epoch": 2.6895454545454545,
+      "grad_norm": 1.12395012937306,
+      "learning_rate": 5.508301254587247e-07,
+      "loss": 0.0361,
+      "step": 11834
+    },
+    {
+      "epoch": 2.6897727272727274,
+      "grad_norm": 2.7069228644092367,
+      "learning_rate": 5.507415071045227e-07,
+      "loss": 0.0282,
+      "step": 11835
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.465865712158709,
+      "learning_rate": 5.50652890264581e-07,
+      "loss": 0.0149,
+      "step": 11836
+    },
+    {
+      "epoch": 2.690227272727273,
+      "grad_norm": 0.9407382493888214,
+      "learning_rate": 5.505642749407064e-07,
+      "loss": 0.0123,
+      "step": 11837
+    },
+    {
+      "epoch": 2.6904545454545454,
+      "grad_norm": 2.049574558677045,
+      "learning_rate": 5.504756611347065e-07,
+      "loss": 0.0256,
+      "step": 11838
+    },
+    {
+      "epoch": 2.690681818181818,
+      "grad_norm": 2.079344905862657,
+      "learning_rate": 5.503870488483876e-07,
+      "loss": 0.0573,
+      "step": 11839
+    },
+    {
+      "epoch": 2.690909090909091,
+      "grad_norm": 2.575027219789114,
+      "learning_rate": 5.502984380835569e-07,
+      "loss": 0.0375,
+      "step": 11840
+    },
+    {
+      "epoch": 2.691136363636364,
+      "grad_norm": 1.2680304252645322,
+      "learning_rate": 5.502098288420215e-07,
+      "loss": 0.0262,
+      "step": 11841
+    },
+    {
+      "epoch": 2.6913636363636364,
+      "grad_norm": 2.769954757589217,
+      "learning_rate": 5.501212211255879e-07,
+      "loss": 0.0334,
+      "step": 11842
+    },
+    {
+      "epoch": 2.691590909090909,
+      "grad_norm": 2.3826973782105902,
+      "learning_rate": 5.500326149360637e-07,
+      "loss": 0.0258,
+      "step": 11843
+    },
+    {
+      "epoch": 2.691818181818182,
+      "grad_norm": 2.823478352184705,
+      "learning_rate": 5.49944010275255e-07,
+      "loss": 0.0273,
+      "step": 11844
+    },
+    {
+      "epoch": 2.6920454545454544,
+      "grad_norm": 0.7349457891550677,
+      "learning_rate": 5.498554071449688e-07,
+      "loss": 0.0089,
+      "step": 11845
+    },
+    {
+      "epoch": 2.6922727272727274,
+      "grad_norm": 1.0498070291382995,
+      "learning_rate": 5.497668055470121e-07,
+      "loss": 0.0069,
+      "step": 11846
+    },
+    {
+      "epoch": 2.6925,
+      "grad_norm": 3.686853500899227,
+      "learning_rate": 5.496782054831915e-07,
+      "loss": 0.0357,
+      "step": 11847
+    },
+    {
+      "epoch": 2.692727272727273,
+      "grad_norm": 2.7092418220675545,
+      "learning_rate": 5.495896069553136e-07,
+      "loss": 0.0279,
+      "step": 11848
+    },
+    {
+      "epoch": 2.6929545454545454,
+      "grad_norm": 1.8378569383629655,
+      "learning_rate": 5.495010099651851e-07,
+      "loss": 0.0149,
+      "step": 11849
+    },
+    {
+      "epoch": 2.6931818181818183,
+      "grad_norm": 1.7850943848147114,
+      "learning_rate": 5.494124145146129e-07,
+      "loss": 0.0086,
+      "step": 11850
+    },
+    {
+      "epoch": 2.693409090909091,
+      "grad_norm": 3.0449039036133825,
+      "learning_rate": 5.493238206054032e-07,
+      "loss": 0.0372,
+      "step": 11851
+    },
+    {
+      "epoch": 2.6936363636363634,
+      "grad_norm": 3.1779907883406,
+      "learning_rate": 5.492352282393633e-07,
+      "loss": 0.047,
+      "step": 11852
+    },
+    {
+      "epoch": 2.6938636363636363,
+      "grad_norm": 1.7407674569925233,
+      "learning_rate": 5.491466374182989e-07,
+      "loss": 0.0195,
+      "step": 11853
+    },
+    {
+      "epoch": 2.6940909090909093,
+      "grad_norm": 2.7721690399016596,
+      "learning_rate": 5.490580481440169e-07,
+      "loss": 0.0668,
+      "step": 11854
+    },
+    {
+      "epoch": 2.694318181818182,
+      "grad_norm": 2.1216943913302857,
+      "learning_rate": 5.489694604183238e-07,
+      "loss": 0.0298,
+      "step": 11855
+    },
+    {
+      "epoch": 2.6945454545454544,
+      "grad_norm": 1.3226270992222864,
+      "learning_rate": 5.488808742430262e-07,
+      "loss": 0.0108,
+      "step": 11856
+    },
+    {
+      "epoch": 2.6947727272727273,
+      "grad_norm": 0.6825055957455329,
+      "learning_rate": 5.487922896199302e-07,
+      "loss": 0.0083,
+      "step": 11857
+    },
+    {
+      "epoch": 2.695,
+      "grad_norm": 1.9477143632218312,
+      "learning_rate": 5.487037065508424e-07,
+      "loss": 0.0161,
+      "step": 11858
+    },
+    {
+      "epoch": 2.695227272727273,
+      "grad_norm": 1.7802766124615317,
+      "learning_rate": 5.486151250375691e-07,
+      "loss": 0.0191,
+      "step": 11859
+    },
+    {
+      "epoch": 2.6954545454545453,
+      "grad_norm": 2.599712073530218,
+      "learning_rate": 5.485265450819169e-07,
+      "loss": 0.0312,
+      "step": 11860
+    },
+    {
+      "epoch": 2.6956818181818183,
+      "grad_norm": 1.9565570978003288,
+      "learning_rate": 5.484379666856917e-07,
+      "loss": 0.0245,
+      "step": 11861
+    },
+    {
+      "epoch": 2.695909090909091,
+      "grad_norm": 1.7998605226383655,
+      "learning_rate": 5.483493898506999e-07,
+      "loss": 0.0154,
+      "step": 11862
+    },
+    {
+      "epoch": 2.6961363636363638,
+      "grad_norm": 2.9265183934580117,
+      "learning_rate": 5.482608145787477e-07,
+      "loss": 0.0511,
+      "step": 11863
+    },
+    {
+      "epoch": 2.6963636363636363,
+      "grad_norm": 2.2681672512608015,
+      "learning_rate": 5.481722408716414e-07,
+      "loss": 0.0339,
+      "step": 11864
+    },
+    {
+      "epoch": 2.696590909090909,
+      "grad_norm": 1.0554961037008843,
+      "learning_rate": 5.480836687311873e-07,
+      "loss": 0.0078,
+      "step": 11865
+    },
+    {
+      "epoch": 2.696818181818182,
+      "grad_norm": 3.5155102830643825,
+      "learning_rate": 5.479950981591911e-07,
+      "loss": 0.0583,
+      "step": 11866
+    },
+    {
+      "epoch": 2.6970454545454547,
+      "grad_norm": 1.8118167982844071,
+      "learning_rate": 5.479065291574593e-07,
+      "loss": 0.0189,
+      "step": 11867
+    },
+    {
+      "epoch": 2.6972727272727273,
+      "grad_norm": 1.7800115263146008,
+      "learning_rate": 5.478179617277979e-07,
+      "loss": 0.0094,
+      "step": 11868
+    },
+    {
+      "epoch": 2.6975,
+      "grad_norm": 2.369798284771426,
+      "learning_rate": 5.477293958720133e-07,
+      "loss": 0.0296,
+      "step": 11869
+    },
+    {
+      "epoch": 2.6977272727272728,
+      "grad_norm": 3.854248397722478,
+      "learning_rate": 5.476408315919107e-07,
+      "loss": 0.0442,
+      "step": 11870
+    },
+    {
+      "epoch": 2.6979545454545457,
+      "grad_norm": 2.6389893222423098,
+      "learning_rate": 5.475522688892963e-07,
+      "loss": 0.0311,
+      "step": 11871
+    },
+    {
+      "epoch": 2.6981818181818182,
+      "grad_norm": 0.7350713546421357,
+      "learning_rate": 5.474637077659766e-07,
+      "loss": 0.0161,
+      "step": 11872
+    },
+    {
+      "epoch": 2.6984090909090908,
+      "grad_norm": 1.5737493887673313,
+      "learning_rate": 5.47375148223757e-07,
+      "loss": 0.025,
+      "step": 11873
+    },
+    {
+      "epoch": 2.6986363636363637,
+      "grad_norm": 1.3425881988510835,
+      "learning_rate": 5.472865902644439e-07,
+      "loss": 0.0166,
+      "step": 11874
+    },
+    {
+      "epoch": 2.6988636363636362,
+      "grad_norm": 1.170400466093782,
+      "learning_rate": 5.471980338898424e-07,
+      "loss": 0.0162,
+      "step": 11875
+    },
+    {
+      "epoch": 2.699090909090909,
+      "grad_norm": 1.7792942727402106,
+      "learning_rate": 5.471094791017588e-07,
+      "loss": 0.0245,
+      "step": 11876
+    },
+    {
+      "epoch": 2.6993181818181817,
+      "grad_norm": 1.3297205412636777,
+      "learning_rate": 5.470209259019988e-07,
+      "loss": 0.0089,
+      "step": 11877
+    },
+    {
+      "epoch": 2.6995454545454547,
+      "grad_norm": 1.9773673879183635,
+      "learning_rate": 5.469323742923685e-07,
+      "loss": 0.0282,
+      "step": 11878
+    },
+    {
+      "epoch": 2.699772727272727,
+      "grad_norm": 2.057377709511012,
+      "learning_rate": 5.468438242746729e-07,
+      "loss": 0.0263,
+      "step": 11879
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.9782740885201728,
+      "learning_rate": 5.467552758507181e-07,
+      "loss": 0.0319,
+      "step": 11880
+    },
+    {
+      "epoch": 2.7002272727272727,
+      "grad_norm": 1.8655583987913558,
+      "learning_rate": 5.466667290223098e-07,
+      "loss": 0.0231,
+      "step": 11881
+    },
+    {
+      "epoch": 2.700454545454545,
+      "grad_norm": 1.9610033758768237,
+      "learning_rate": 5.465781837912539e-07,
+      "loss": 0.0219,
+      "step": 11882
+    },
+    {
+      "epoch": 2.700681818181818,
+      "grad_norm": 2.0900314412492382,
+      "learning_rate": 5.464896401593553e-07,
+      "loss": 0.0358,
+      "step": 11883
+    },
+    {
+      "epoch": 2.700909090909091,
+      "grad_norm": 2.8864604399899974,
+      "learning_rate": 5.464010981284198e-07,
+      "loss": 0.0484,
+      "step": 11884
+    },
+    {
+      "epoch": 2.7011363636363637,
+      "grad_norm": 1.2108107528343512,
+      "learning_rate": 5.463125577002531e-07,
+      "loss": 0.0258,
+      "step": 11885
+    },
+    {
+      "epoch": 2.701363636363636,
+      "grad_norm": 1.2875966389084297,
+      "learning_rate": 5.462240188766607e-07,
+      "loss": 0.0246,
+      "step": 11886
+    },
+    {
+      "epoch": 2.701590909090909,
+      "grad_norm": 1.5713054395859964,
+      "learning_rate": 5.46135481659448e-07,
+      "loss": 0.0176,
+      "step": 11887
+    },
+    {
+      "epoch": 2.7018181818181817,
+      "grad_norm": 2.784112826270527,
+      "learning_rate": 5.460469460504205e-07,
+      "loss": 0.0365,
+      "step": 11888
+    },
+    {
+      "epoch": 2.7020454545454546,
+      "grad_norm": 1.2981919896267413,
+      "learning_rate": 5.459584120513833e-07,
+      "loss": 0.0176,
+      "step": 11889
+    },
+    {
+      "epoch": 2.702272727272727,
+      "grad_norm": 2.77568274868232,
+      "learning_rate": 5.45869879664142e-07,
+      "loss": 0.0344,
+      "step": 11890
+    },
+    {
+      "epoch": 2.7025,
+      "grad_norm": 2.0985816326121123,
+      "learning_rate": 5.457813488905022e-07,
+      "loss": 0.0401,
+      "step": 11891
+    },
+    {
+      "epoch": 2.7027272727272726,
+      "grad_norm": 2.259692140567951,
+      "learning_rate": 5.456928197322686e-07,
+      "loss": 0.0209,
+      "step": 11892
+    },
+    {
+      "epoch": 2.7029545454545456,
+      "grad_norm": 2.5610546288879505,
+      "learning_rate": 5.456042921912469e-07,
+      "loss": 0.0217,
+      "step": 11893
+    },
+    {
+      "epoch": 2.703181818181818,
+      "grad_norm": 2.261683479576444,
+      "learning_rate": 5.45515766269242e-07,
+      "loss": 0.0417,
+      "step": 11894
+    },
+    {
+      "epoch": 2.7034090909090907,
+      "grad_norm": 1.6390505442502892,
+      "learning_rate": 5.454272419680596e-07,
+      "loss": 0.0136,
+      "step": 11895
+    },
+    {
+      "epoch": 2.7036363636363636,
+      "grad_norm": 2.691203977901802,
+      "learning_rate": 5.453387192895044e-07,
+      "loss": 0.0435,
+      "step": 11896
+    },
+    {
+      "epoch": 2.7038636363636366,
+      "grad_norm": 0.941634226340322,
+      "learning_rate": 5.452501982353816e-07,
+      "loss": 0.0103,
+      "step": 11897
+    },
+    {
+      "epoch": 2.704090909090909,
+      "grad_norm": 1.1705717310113553,
+      "learning_rate": 5.451616788074964e-07,
+      "loss": 0.0169,
+      "step": 11898
+    },
+    {
+      "epoch": 2.7043181818181816,
+      "grad_norm": 1.479990196024721,
+      "learning_rate": 5.450731610076538e-07,
+      "loss": 0.0136,
+      "step": 11899
+    },
+    {
+      "epoch": 2.7045454545454546,
+      "grad_norm": 2.019018173414066,
+      "learning_rate": 5.449846448376593e-07,
+      "loss": 0.0256,
+      "step": 11900
+    },
+    {
+      "epoch": 2.704772727272727,
+      "grad_norm": 1.7925196096467828,
+      "learning_rate": 5.448961302993171e-07,
+      "loss": 0.0148,
+      "step": 11901
+    },
+    {
+      "epoch": 2.705,
+      "grad_norm": 1.5296682116526819,
+      "learning_rate": 5.448076173944326e-07,
+      "loss": 0.015,
+      "step": 11902
+    },
+    {
+      "epoch": 2.7052272727272726,
+      "grad_norm": 3.744869311345074,
+      "learning_rate": 5.447191061248107e-07,
+      "loss": 0.0586,
+      "step": 11903
+    },
+    {
+      "epoch": 2.7054545454545456,
+      "grad_norm": 1.6647808191141402,
+      "learning_rate": 5.446305964922566e-07,
+      "loss": 0.0321,
+      "step": 11904
+    },
+    {
+      "epoch": 2.705681818181818,
+      "grad_norm": 1.9156775065609495,
+      "learning_rate": 5.445420884985744e-07,
+      "loss": 0.0311,
+      "step": 11905
+    },
+    {
+      "epoch": 2.705909090909091,
+      "grad_norm": 2.1251937017035223,
+      "learning_rate": 5.444535821455696e-07,
+      "loss": 0.0566,
+      "step": 11906
+    },
+    {
+      "epoch": 2.7061363636363636,
+      "grad_norm": 2.4573604433931875,
+      "learning_rate": 5.443650774350467e-07,
+      "loss": 0.036,
+      "step": 11907
+    },
+    {
+      "epoch": 2.7063636363636365,
+      "grad_norm": 2.407537619233222,
+      "learning_rate": 5.442765743688105e-07,
+      "loss": 0.018,
+      "step": 11908
+    },
+    {
+      "epoch": 2.706590909090909,
+      "grad_norm": 1.385468306927487,
+      "learning_rate": 5.441880729486659e-07,
+      "loss": 0.0176,
+      "step": 11909
+    },
+    {
+      "epoch": 2.706818181818182,
+      "grad_norm": 1.576146976937451,
+      "learning_rate": 5.440995731764174e-07,
+      "loss": 0.0231,
+      "step": 11910
+    },
+    {
+      "epoch": 2.7070454545454545,
+      "grad_norm": 1.5055887770423255,
+      "learning_rate": 5.440110750538698e-07,
+      "loss": 0.0129,
+      "step": 11911
+    },
+    {
+      "epoch": 2.707272727272727,
+      "grad_norm": 2.4491767767696366,
+      "learning_rate": 5.439225785828276e-07,
+      "loss": 0.0275,
+      "step": 11912
+    },
+    {
+      "epoch": 2.7075,
+      "grad_norm": 1.8306882265582076,
+      "learning_rate": 5.438340837650957e-07,
+      "loss": 0.0259,
+      "step": 11913
+    },
+    {
+      "epoch": 2.707727272727273,
+      "grad_norm": 1.8176503350997193,
+      "learning_rate": 5.437455906024783e-07,
+      "loss": 0.026,
+      "step": 11914
+    },
+    {
+      "epoch": 2.7079545454545455,
+      "grad_norm": 2.1318733524014717,
+      "learning_rate": 5.4365709909678e-07,
+      "loss": 0.0264,
+      "step": 11915
+    },
+    {
+      "epoch": 2.708181818181818,
+      "grad_norm": 2.2699654350328693,
+      "learning_rate": 5.435686092498053e-07,
+      "loss": 0.026,
+      "step": 11916
+    },
+    {
+      "epoch": 2.708409090909091,
+      "grad_norm": 7.320508901875231,
+      "learning_rate": 5.434801210633589e-07,
+      "loss": 0.0393,
+      "step": 11917
+    },
+    {
+      "epoch": 2.7086363636363635,
+      "grad_norm": 1.3869083528045174,
+      "learning_rate": 5.433916345392449e-07,
+      "loss": 0.0264,
+      "step": 11918
+    },
+    {
+      "epoch": 2.7088636363636365,
+      "grad_norm": 0.9836832130464696,
+      "learning_rate": 5.433031496792678e-07,
+      "loss": 0.0091,
+      "step": 11919
+    },
+    {
+      "epoch": 2.709090909090909,
+      "grad_norm": 2.0420103065835575,
+      "learning_rate": 5.432146664852321e-07,
+      "loss": 0.0311,
+      "step": 11920
+    },
+    {
+      "epoch": 2.709318181818182,
+      "grad_norm": 1.066542150666885,
+      "learning_rate": 5.431261849589419e-07,
+      "loss": 0.0182,
+      "step": 11921
+    },
+    {
+      "epoch": 2.7095454545454545,
+      "grad_norm": 2.1548878023207068,
+      "learning_rate": 5.43037705102202e-07,
+      "loss": 0.0212,
+      "step": 11922
+    },
+    {
+      "epoch": 2.7097727272727274,
+      "grad_norm": 1.500922467210162,
+      "learning_rate": 5.42949226916816e-07,
+      "loss": 0.008,
+      "step": 11923
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.4568471469046864,
+      "learning_rate": 5.428607504045885e-07,
+      "loss": 0.0115,
+      "step": 11924
+    },
+    {
+      "epoch": 2.7102272727272725,
+      "grad_norm": 2.7984872654916693,
+      "learning_rate": 5.427722755673235e-07,
+      "loss": 0.0164,
+      "step": 11925
+    },
+    {
+      "epoch": 2.7104545454545454,
+      "grad_norm": 1.6776234350109438,
+      "learning_rate": 5.426838024068254e-07,
+      "loss": 0.0161,
+      "step": 11926
+    },
+    {
+      "epoch": 2.7106818181818184,
+      "grad_norm": 1.7483559878477815,
+      "learning_rate": 5.42595330924898e-07,
+      "loss": 0.0245,
+      "step": 11927
+    },
+    {
+      "epoch": 2.710909090909091,
+      "grad_norm": 1.538442641577486,
+      "learning_rate": 5.425068611233457e-07,
+      "loss": 0.0173,
+      "step": 11928
+    },
+    {
+      "epoch": 2.7111363636363635,
+      "grad_norm": 1.6357550819391282,
+      "learning_rate": 5.424183930039723e-07,
+      "loss": 0.0283,
+      "step": 11929
+    },
+    {
+      "epoch": 2.7113636363636364,
+      "grad_norm": 1.1121130841328983,
+      "learning_rate": 5.42329926568582e-07,
+      "loss": 0.0118,
+      "step": 11930
+    },
+    {
+      "epoch": 2.711590909090909,
+      "grad_norm": 1.7655317997290105,
+      "learning_rate": 5.422414618189788e-07,
+      "loss": 0.0214,
+      "step": 11931
+    },
+    {
+      "epoch": 2.711818181818182,
+      "grad_norm": 2.1855293392338884,
+      "learning_rate": 5.421529987569665e-07,
+      "loss": 0.0369,
+      "step": 11932
+    },
+    {
+      "epoch": 2.7120454545454544,
+      "grad_norm": 2.3071105827289426,
+      "learning_rate": 5.42064537384349e-07,
+      "loss": 0.028,
+      "step": 11933
+    },
+    {
+      "epoch": 2.7122727272727274,
+      "grad_norm": 5.204242742753489,
+      "learning_rate": 5.419760777029303e-07,
+      "loss": 0.0608,
+      "step": 11934
+    },
+    {
+      "epoch": 2.7125,
+      "grad_norm": 0.4389426902605691,
+      "learning_rate": 5.418876197145146e-07,
+      "loss": 0.0032,
+      "step": 11935
+    },
+    {
+      "epoch": 2.712727272727273,
+      "grad_norm": 2.34761877790151,
+      "learning_rate": 5.417991634209049e-07,
+      "loss": 0.0432,
+      "step": 11936
+    },
+    {
+      "epoch": 2.7129545454545454,
+      "grad_norm": 2.216333833092801,
+      "learning_rate": 5.417107088239055e-07,
+      "loss": 0.0204,
+      "step": 11937
+    },
+    {
+      "epoch": 2.713181818181818,
+      "grad_norm": 2.207844692190625,
+      "learning_rate": 5.4162225592532e-07,
+      "loss": 0.0237,
+      "step": 11938
+    },
+    {
+      "epoch": 2.713409090909091,
+      "grad_norm": 1.3723476286914094,
+      "learning_rate": 5.415338047269523e-07,
+      "loss": 0.0117,
+      "step": 11939
+    },
+    {
+      "epoch": 2.713636363636364,
+      "grad_norm": 1.101504398712854,
+      "learning_rate": 5.414453552306059e-07,
+      "loss": 0.017,
+      "step": 11940
+    },
+    {
+      "epoch": 2.7138636363636364,
+      "grad_norm": 2.051766550986719,
+      "learning_rate": 5.413569074380845e-07,
+      "loss": 0.0222,
+      "step": 11941
+    },
+    {
+      "epoch": 2.714090909090909,
+      "grad_norm": 1.2255816672253006,
+      "learning_rate": 5.412684613511915e-07,
+      "loss": 0.0189,
+      "step": 11942
+    },
+    {
+      "epoch": 2.714318181818182,
+      "grad_norm": 1.4775973499215411,
+      "learning_rate": 5.411800169717307e-07,
+      "loss": 0.0153,
+      "step": 11943
+    },
+    {
+      "epoch": 2.714545454545455,
+      "grad_norm": 1.1844197858104801,
+      "learning_rate": 5.41091574301506e-07,
+      "loss": 0.0178,
+      "step": 11944
+    },
+    {
+      "epoch": 2.7147727272727273,
+      "grad_norm": 1.7461754747819074,
+      "learning_rate": 5.4100313334232e-07,
+      "loss": 0.0125,
+      "step": 11945
+    },
+    {
+      "epoch": 2.715,
+      "grad_norm": 2.0510212595553616,
+      "learning_rate": 5.409146940959768e-07,
+      "loss": 0.0297,
+      "step": 11946
+    },
+    {
+      "epoch": 2.715227272727273,
+      "grad_norm": 1.9479626033050519,
+      "learning_rate": 5.408262565642797e-07,
+      "loss": 0.0347,
+      "step": 11947
+    },
+    {
+      "epoch": 2.7154545454545453,
+      "grad_norm": 1.8413793823695976,
+      "learning_rate": 5.407378207490321e-07,
+      "loss": 0.0245,
+      "step": 11948
+    },
+    {
+      "epoch": 2.7156818181818183,
+      "grad_norm": 1.4257464709434533,
+      "learning_rate": 5.406493866520373e-07,
+      "loss": 0.0204,
+      "step": 11949
+    },
+    {
+      "epoch": 2.715909090909091,
+      "grad_norm": 1.3534394865814556,
+      "learning_rate": 5.405609542750986e-07,
+      "loss": 0.0108,
+      "step": 11950
+    },
+    {
+      "epoch": 2.716136363636364,
+      "grad_norm": 1.862598242054347,
+      "learning_rate": 5.404725236200195e-07,
+      "loss": 0.0185,
+      "step": 11951
+    },
+    {
+      "epoch": 2.7163636363636363,
+      "grad_norm": 1.5884011093551775,
+      "learning_rate": 5.403840946886029e-07,
+      "loss": 0.0335,
+      "step": 11952
+    },
+    {
+      "epoch": 2.7165909090909093,
+      "grad_norm": 2.644162877536427,
+      "learning_rate": 5.402956674826526e-07,
+      "loss": 0.0204,
+      "step": 11953
+    },
+    {
+      "epoch": 2.716818181818182,
+      "grad_norm": 1.4164645598417525,
+      "learning_rate": 5.402072420039711e-07,
+      "loss": 0.0201,
+      "step": 11954
+    },
+    {
+      "epoch": 2.7170454545454543,
+      "grad_norm": 1.420113286220467,
+      "learning_rate": 5.401188182543618e-07,
+      "loss": 0.0182,
+      "step": 11955
+    },
+    {
+      "epoch": 2.7172727272727273,
+      "grad_norm": 2.2934652780186147,
+      "learning_rate": 5.40030396235628e-07,
+      "loss": 0.0228,
+      "step": 11956
+    },
+    {
+      "epoch": 2.7175000000000002,
+      "grad_norm": 2.5966076555123787,
+      "learning_rate": 5.399419759495729e-07,
+      "loss": 0.0199,
+      "step": 11957
+    },
+    {
+      "epoch": 2.7177272727272728,
+      "grad_norm": 1.8169324364637032,
+      "learning_rate": 5.398535573979991e-07,
+      "loss": 0.023,
+      "step": 11958
+    },
+    {
+      "epoch": 2.7179545454545453,
+      "grad_norm": 2.4143318005354684,
+      "learning_rate": 5.397651405827097e-07,
+      "loss": 0.0339,
+      "step": 11959
+    },
+    {
+      "epoch": 2.7181818181818183,
+      "grad_norm": 2.15867241861139,
+      "learning_rate": 5.396767255055078e-07,
+      "loss": 0.0208,
+      "step": 11960
+    },
+    {
+      "epoch": 2.7184090909090908,
+      "grad_norm": 1.438106994613542,
+      "learning_rate": 5.395883121681965e-07,
+      "loss": 0.0172,
+      "step": 11961
+    },
+    {
+      "epoch": 2.7186363636363637,
+      "grad_norm": 0.7231406385241563,
+      "learning_rate": 5.394999005725784e-07,
+      "loss": 0.0097,
+      "step": 11962
+    },
+    {
+      "epoch": 2.7188636363636363,
+      "grad_norm": 1.3250319979206584,
+      "learning_rate": 5.394114907204565e-07,
+      "loss": 0.0221,
+      "step": 11963
+    },
+    {
+      "epoch": 2.7190909090909092,
+      "grad_norm": 1.6940022913995807,
+      "learning_rate": 5.393230826136336e-07,
+      "loss": 0.0147,
+      "step": 11964
+    },
+    {
+      "epoch": 2.7193181818181817,
+      "grad_norm": 1.9694159193549208,
+      "learning_rate": 5.392346762539125e-07,
+      "loss": 0.0148,
+      "step": 11965
+    },
+    {
+      "epoch": 2.7195454545454547,
+      "grad_norm": 3.193402622875964,
+      "learning_rate": 5.391462716430963e-07,
+      "loss": 0.0436,
+      "step": 11966
+    },
+    {
+      "epoch": 2.7197727272727272,
+      "grad_norm": 2.0965667739660234,
+      "learning_rate": 5.390578687829871e-07,
+      "loss": 0.0282,
+      "step": 11967
+    },
+    {
+      "epoch": 2.7199999999999998,
+      "grad_norm": 3.61910213796242,
+      "learning_rate": 5.389694676753878e-07,
+      "loss": 0.0271,
+      "step": 11968
+    },
+    {
+      "epoch": 2.7202272727272727,
+      "grad_norm": 1.5800885304247374,
+      "learning_rate": 5.388810683221013e-07,
+      "loss": 0.0129,
+      "step": 11969
+    },
+    {
+      "epoch": 2.7204545454545457,
+      "grad_norm": 2.624948565344363,
+      "learning_rate": 5.387926707249301e-07,
+      "loss": 0.0211,
+      "step": 11970
+    },
+    {
+      "epoch": 2.720681818181818,
+      "grad_norm": 2.8142755791202223,
+      "learning_rate": 5.387042748856766e-07,
+      "loss": 0.0511,
+      "step": 11971
+    },
+    {
+      "epoch": 2.7209090909090907,
+      "grad_norm": 1.7072965293344635,
+      "learning_rate": 5.386158808061435e-07,
+      "loss": 0.0326,
+      "step": 11972
+    },
+    {
+      "epoch": 2.7211363636363637,
+      "grad_norm": 2.4835403274444414,
+      "learning_rate": 5.385274884881333e-07,
+      "loss": 0.0383,
+      "step": 11973
+    },
+    {
+      "epoch": 2.721363636363636,
+      "grad_norm": 1.7372370455636874,
+      "learning_rate": 5.384390979334485e-07,
+      "loss": 0.0191,
+      "step": 11974
+    },
+    {
+      "epoch": 2.721590909090909,
+      "grad_norm": 2.3410676692599997,
+      "learning_rate": 5.383507091438918e-07,
+      "loss": 0.0177,
+      "step": 11975
+    },
+    {
+      "epoch": 2.7218181818181817,
+      "grad_norm": 1.6782620793140886,
+      "learning_rate": 5.382623221212648e-07,
+      "loss": 0.0196,
+      "step": 11976
+    },
+    {
+      "epoch": 2.7220454545454547,
+      "grad_norm": 1.662391895554596,
+      "learning_rate": 5.381739368673706e-07,
+      "loss": 0.0275,
+      "step": 11977
+    },
+    {
+      "epoch": 2.722272727272727,
+      "grad_norm": 2.606297351208909,
+      "learning_rate": 5.380855533840114e-07,
+      "loss": 0.0472,
+      "step": 11978
+    },
+    {
+      "epoch": 2.7225,
+      "grad_norm": 1.8882289568414539,
+      "learning_rate": 5.379971716729894e-07,
+      "loss": 0.0232,
+      "step": 11979
+    },
+    {
+      "epoch": 2.7227272727272727,
+      "grad_norm": 1.2821473448337455,
+      "learning_rate": 5.379087917361068e-07,
+      "loss": 0.0138,
+      "step": 11980
+    },
+    {
+      "epoch": 2.722954545454545,
+      "grad_norm": 1.7730955360920233,
+      "learning_rate": 5.378204135751656e-07,
+      "loss": 0.0236,
+      "step": 11981
+    },
+    {
+      "epoch": 2.723181818181818,
+      "grad_norm": 1.8924816974402754,
+      "learning_rate": 5.377320371919685e-07,
+      "loss": 0.0224,
+      "step": 11982
+    },
+    {
+      "epoch": 2.723409090909091,
+      "grad_norm": 1.657384658067239,
+      "learning_rate": 5.376436625883177e-07,
+      "loss": 0.0153,
+      "step": 11983
+    },
+    {
+      "epoch": 2.7236363636363636,
+      "grad_norm": 1.5274154064357424,
+      "learning_rate": 5.375552897660146e-07,
+      "loss": 0.0236,
+      "step": 11984
+    },
+    {
+      "epoch": 2.723863636363636,
+      "grad_norm": 1.550656948946884,
+      "learning_rate": 5.374669187268617e-07,
+      "loss": 0.0266,
+      "step": 11985
+    },
+    {
+      "epoch": 2.724090909090909,
+      "grad_norm": 1.7876047656941907,
+      "learning_rate": 5.373785494726612e-07,
+      "loss": 0.02,
+      "step": 11986
+    },
+    {
+      "epoch": 2.724318181818182,
+      "grad_norm": 0.9053164668831135,
+      "learning_rate": 5.372901820052147e-07,
+      "loss": 0.0093,
+      "step": 11987
+    },
+    {
+      "epoch": 2.7245454545454546,
+      "grad_norm": 1.4187086524116534,
+      "learning_rate": 5.372018163263249e-07,
+      "loss": 0.0111,
+      "step": 11988
+    },
+    {
+      "epoch": 2.724772727272727,
+      "grad_norm": 1.7287790347553937,
+      "learning_rate": 5.371134524377929e-07,
+      "loss": 0.0167,
+      "step": 11989
+    },
+    {
+      "epoch": 2.725,
+      "grad_norm": 1.6533728907948317,
+      "learning_rate": 5.370250903414209e-07,
+      "loss": 0.011,
+      "step": 11990
+    },
+    {
+      "epoch": 2.7252272727272726,
+      "grad_norm": 1.4128698791476713,
+      "learning_rate": 5.369367300390109e-07,
+      "loss": 0.0119,
+      "step": 11991
+    },
+    {
+      "epoch": 2.7254545454545456,
+      "grad_norm": 1.4206507899835046,
+      "learning_rate": 5.368483715323647e-07,
+      "loss": 0.0155,
+      "step": 11992
+    },
+    {
+      "epoch": 2.725681818181818,
+      "grad_norm": 0.9940980378870988,
+      "learning_rate": 5.367600148232837e-07,
+      "loss": 0.0101,
+      "step": 11993
+    },
+    {
+      "epoch": 2.725909090909091,
+      "grad_norm": 1.4987805087122523,
+      "learning_rate": 5.366716599135701e-07,
+      "loss": 0.0242,
+      "step": 11994
+    },
+    {
+      "epoch": 2.7261363636363636,
+      "grad_norm": 2.605229324280742,
+      "learning_rate": 5.365833068050255e-07,
+      "loss": 0.0166,
+      "step": 11995
+    },
+    {
+      "epoch": 2.7263636363636365,
+      "grad_norm": 4.734567741132346,
+      "learning_rate": 5.364949554994514e-07,
+      "loss": 0.0693,
+      "step": 11996
+    },
+    {
+      "epoch": 2.726590909090909,
+      "grad_norm": 1.7186057800791688,
+      "learning_rate": 5.364066059986499e-07,
+      "loss": 0.0144,
+      "step": 11997
+    },
+    {
+      "epoch": 2.7268181818181816,
+      "grad_norm": 2.2407470288003535,
+      "learning_rate": 5.36318258304422e-07,
+      "loss": 0.0268,
+      "step": 11998
+    },
+    {
+      "epoch": 2.7270454545454546,
+      "grad_norm": 1.8551963154281461,
+      "learning_rate": 5.362299124185694e-07,
+      "loss": 0.0234,
+      "step": 11999
+    },
+    {
+      "epoch": 2.7272727272727275,
+      "grad_norm": 3.2326795976643496,
+      "learning_rate": 5.361415683428939e-07,
+      "loss": 0.0918,
+      "step": 12000
+    },
+    {
+      "epoch": 2.7275,
+      "grad_norm": 1.2920099093891402,
+      "learning_rate": 5.360532260791969e-07,
+      "loss": 0.024,
+      "step": 12001
+    },
+    {
+      "epoch": 2.7277272727272726,
+      "grad_norm": 1.077304041707251,
+      "learning_rate": 5.359648856292797e-07,
+      "loss": 0.0105,
+      "step": 12002
+    },
+    {
+      "epoch": 2.7279545454545455,
+      "grad_norm": 4.3350076611579045,
+      "learning_rate": 5.358765469949438e-07,
+      "loss": 0.0482,
+      "step": 12003
+    },
+    {
+      "epoch": 2.728181818181818,
+      "grad_norm": 1.656186585796026,
+      "learning_rate": 5.357882101779906e-07,
+      "loss": 0.0165,
+      "step": 12004
+    },
+    {
+      "epoch": 2.728409090909091,
+      "grad_norm": 0.7124654352471834,
+      "learning_rate": 5.356998751802217e-07,
+      "loss": 0.0167,
+      "step": 12005
+    },
+    {
+      "epoch": 2.7286363636363635,
+      "grad_norm": 1.365097153576992,
+      "learning_rate": 5.356115420034378e-07,
+      "loss": 0.0108,
+      "step": 12006
+    },
+    {
+      "epoch": 2.7288636363636365,
+      "grad_norm": 2.761628959086201,
+      "learning_rate": 5.355232106494407e-07,
+      "loss": 0.0389,
+      "step": 12007
+    },
+    {
+      "epoch": 2.729090909090909,
+      "grad_norm": 1.8945641908830009,
+      "learning_rate": 5.354348811200312e-07,
+      "loss": 0.0223,
+      "step": 12008
+    },
+    {
+      "epoch": 2.729318181818182,
+      "grad_norm": 2.2953665417980806,
+      "learning_rate": 5.353465534170109e-07,
+      "loss": 0.0341,
+      "step": 12009
+    },
+    {
+      "epoch": 2.7295454545454545,
+      "grad_norm": 1.7315528749786624,
+      "learning_rate": 5.352582275421808e-07,
+      "loss": 0.0188,
+      "step": 12010
+    },
+    {
+      "epoch": 2.729772727272727,
+      "grad_norm": 3.096879126065366,
+      "learning_rate": 5.35169903497342e-07,
+      "loss": 0.034,
+      "step": 12011
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 2.3180323739411857,
+      "learning_rate": 5.350815812842955e-07,
+      "loss": 0.0201,
+      "step": 12012
+    },
+    {
+      "epoch": 2.730227272727273,
+      "grad_norm": 2.382815883278521,
+      "learning_rate": 5.349932609048424e-07,
+      "loss": 0.0165,
+      "step": 12013
+    },
+    {
+      "epoch": 2.7304545454545455,
+      "grad_norm": 1.2275822992572454,
+      "learning_rate": 5.349049423607839e-07,
+      "loss": 0.0098,
+      "step": 12014
+    },
+    {
+      "epoch": 2.730681818181818,
+      "grad_norm": 1.6029549948638506,
+      "learning_rate": 5.348166256539207e-07,
+      "loss": 0.0201,
+      "step": 12015
+    },
+    {
+      "epoch": 2.730909090909091,
+      "grad_norm": 1.0160834092001356,
+      "learning_rate": 5.347283107860538e-07,
+      "loss": 0.0096,
+      "step": 12016
+    },
+    {
+      "epoch": 2.7311363636363635,
+      "grad_norm": 3.5211578127798933,
+      "learning_rate": 5.346399977589842e-07,
+      "loss": 0.036,
+      "step": 12017
+    },
+    {
+      "epoch": 2.7313636363636364,
+      "grad_norm": 2.6717108898755417,
+      "learning_rate": 5.345516865745129e-07,
+      "loss": 0.0446,
+      "step": 12018
+    },
+    {
+      "epoch": 2.731590909090909,
+      "grad_norm": 2.2938949958784094,
+      "learning_rate": 5.344633772344402e-07,
+      "loss": 0.0275,
+      "step": 12019
+    },
+    {
+      "epoch": 2.731818181818182,
+      "grad_norm": 2.844275266016365,
+      "learning_rate": 5.343750697405673e-07,
+      "loss": 0.0355,
+      "step": 12020
+    },
+    {
+      "epoch": 2.7320454545454544,
+      "grad_norm": 2.5305830854149245,
+      "learning_rate": 5.342867640946948e-07,
+      "loss": 0.045,
+      "step": 12021
+    },
+    {
+      "epoch": 2.7322727272727274,
+      "grad_norm": 1.5836434574424136,
+      "learning_rate": 5.341984602986233e-07,
+      "loss": 0.0134,
+      "step": 12022
+    },
+    {
+      "epoch": 2.7325,
+      "grad_norm": 2.4174068773109587,
+      "learning_rate": 5.341101583541539e-07,
+      "loss": 0.0328,
+      "step": 12023
+    },
+    {
+      "epoch": 2.732727272727273,
+      "grad_norm": 2.166757956434984,
+      "learning_rate": 5.340218582630867e-07,
+      "loss": 0.0226,
+      "step": 12024
+    },
+    {
+      "epoch": 2.7329545454545454,
+      "grad_norm": 1.9818367483764066,
+      "learning_rate": 5.339335600272225e-07,
+      "loss": 0.0326,
+      "step": 12025
+    },
+    {
+      "epoch": 2.7331818181818184,
+      "grad_norm": 0.6482799098102466,
+      "learning_rate": 5.33845263648362e-07,
+      "loss": 0.0057,
+      "step": 12026
+    },
+    {
+      "epoch": 2.733409090909091,
+      "grad_norm": 3.5860083270471685,
+      "learning_rate": 5.337569691283057e-07,
+      "loss": 0.0336,
+      "step": 12027
+    },
+    {
+      "epoch": 2.7336363636363634,
+      "grad_norm": 1.2770227353825998,
+      "learning_rate": 5.33668676468854e-07,
+      "loss": 0.0147,
+      "step": 12028
+    },
+    {
+      "epoch": 2.7338636363636364,
+      "grad_norm": 3.729521542629489,
+      "learning_rate": 5.335803856718071e-07,
+      "loss": 0.0414,
+      "step": 12029
+    },
+    {
+      "epoch": 2.7340909090909093,
+      "grad_norm": 1.898657670472316,
+      "learning_rate": 5.334920967389657e-07,
+      "loss": 0.0207,
+      "step": 12030
+    },
+    {
+      "epoch": 2.734318181818182,
+      "grad_norm": 1.7059390997606043,
+      "learning_rate": 5.334038096721301e-07,
+      "loss": 0.0147,
+      "step": 12031
+    },
+    {
+      "epoch": 2.7345454545454544,
+      "grad_norm": 1.9931257523214247,
+      "learning_rate": 5.333155244731008e-07,
+      "loss": 0.0271,
+      "step": 12032
+    },
+    {
+      "epoch": 2.7347727272727274,
+      "grad_norm": 1.0335031209198746,
+      "learning_rate": 5.332272411436776e-07,
+      "loss": 0.0071,
+      "step": 12033
+    },
+    {
+      "epoch": 2.735,
+      "grad_norm": 2.631626497172306,
+      "learning_rate": 5.331389596856612e-07,
+      "loss": 0.0348,
+      "step": 12034
+    },
+    {
+      "epoch": 2.735227272727273,
+      "grad_norm": 1.7080126497659323,
+      "learning_rate": 5.330506801008516e-07,
+      "loss": 0.031,
+      "step": 12035
+    },
+    {
+      "epoch": 2.7354545454545454,
+      "grad_norm": 1.0497252737611134,
+      "learning_rate": 5.329624023910494e-07,
+      "loss": 0.0111,
+      "step": 12036
+    },
+    {
+      "epoch": 2.7356818181818183,
+      "grad_norm": 2.757270403112357,
+      "learning_rate": 5.328741265580541e-07,
+      "loss": 0.0274,
+      "step": 12037
+    },
+    {
+      "epoch": 2.735909090909091,
+      "grad_norm": 1.252710198228303,
+      "learning_rate": 5.32785852603666e-07,
+      "loss": 0.0179,
+      "step": 12038
+    },
+    {
+      "epoch": 2.736136363636364,
+      "grad_norm": 1.5914417727118972,
+      "learning_rate": 5.326975805296852e-07,
+      "loss": 0.0222,
+      "step": 12039
+    },
+    {
+      "epoch": 2.7363636363636363,
+      "grad_norm": 2.180981411269662,
+      "learning_rate": 5.32609310337912e-07,
+      "loss": 0.0292,
+      "step": 12040
+    },
+    {
+      "epoch": 2.736590909090909,
+      "grad_norm": 2.450945906520751,
+      "learning_rate": 5.325210420301461e-07,
+      "loss": 0.0366,
+      "step": 12041
+    },
+    {
+      "epoch": 2.736818181818182,
+      "grad_norm": 0.8939961907636059,
+      "learning_rate": 5.324327756081874e-07,
+      "loss": 0.0065,
+      "step": 12042
+    },
+    {
+      "epoch": 2.737045454545455,
+      "grad_norm": 2.219171189940429,
+      "learning_rate": 5.323445110738358e-07,
+      "loss": 0.0231,
+      "step": 12043
+    },
+    {
+      "epoch": 2.7372727272727273,
+      "grad_norm": 0.9418829796527453,
+      "learning_rate": 5.322562484288914e-07,
+      "loss": 0.0097,
+      "step": 12044
+    },
+    {
+      "epoch": 2.7375,
+      "grad_norm": 2.177334679144043,
+      "learning_rate": 5.321679876751541e-07,
+      "loss": 0.0476,
+      "step": 12045
+    },
+    {
+      "epoch": 2.737727272727273,
+      "grad_norm": 2.3324194179851734,
+      "learning_rate": 5.320797288144231e-07,
+      "loss": 0.0697,
+      "step": 12046
+    },
+    {
+      "epoch": 2.7379545454545453,
+      "grad_norm": 2.049650713720358,
+      "learning_rate": 5.319914718484988e-07,
+      "loss": 0.0179,
+      "step": 12047
+    },
+    {
+      "epoch": 2.7381818181818183,
+      "grad_norm": 2.296159010621013,
+      "learning_rate": 5.319032167791805e-07,
+      "loss": 0.0327,
+      "step": 12048
+    },
+    {
+      "epoch": 2.738409090909091,
+      "grad_norm": 1.6788756132463862,
+      "learning_rate": 5.318149636082683e-07,
+      "loss": 0.0122,
+      "step": 12049
+    },
+    {
+      "epoch": 2.7386363636363638,
+      "grad_norm": 1.3147149774006066,
+      "learning_rate": 5.317267123375613e-07,
+      "loss": 0.0125,
+      "step": 12050
+    },
+    {
+      "epoch": 2.7388636363636363,
+      "grad_norm": 4.119250068548314,
+      "learning_rate": 5.316384629688594e-07,
+      "loss": 0.0492,
+      "step": 12051
+    },
+    {
+      "epoch": 2.7390909090909092,
+      "grad_norm": 1.434322460734525,
+      "learning_rate": 5.315502155039621e-07,
+      "loss": 0.0154,
+      "step": 12052
+    },
+    {
+      "epoch": 2.7393181818181818,
+      "grad_norm": 1.3246698653104603,
+      "learning_rate": 5.31461969944669e-07,
+      "loss": 0.0143,
+      "step": 12053
+    },
+    {
+      "epoch": 2.7395454545454543,
+      "grad_norm": 1.828472483277219,
+      "learning_rate": 5.313737262927796e-07,
+      "loss": 0.015,
+      "step": 12054
+    },
+    {
+      "epoch": 2.7397727272727272,
+      "grad_norm": 2.1066043487974135,
+      "learning_rate": 5.31285484550093e-07,
+      "loss": 0.0318,
+      "step": 12055
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 2.7224403609357113,
+      "learning_rate": 5.31197244718409e-07,
+      "loss": 0.056,
+      "step": 12056
+    },
+    {
+      "epoch": 2.7402272727272727,
+      "grad_norm": 2.444381855961367,
+      "learning_rate": 5.311090067995268e-07,
+      "loss": 0.0391,
+      "step": 12057
+    },
+    {
+      "epoch": 2.7404545454545453,
+      "grad_norm": 2.5306896237156247,
+      "learning_rate": 5.31020770795246e-07,
+      "loss": 0.0391,
+      "step": 12058
+    },
+    {
+      "epoch": 2.740681818181818,
+      "grad_norm": 2.670539522417963,
+      "learning_rate": 5.309325367073655e-07,
+      "loss": 0.0575,
+      "step": 12059
+    },
+    {
+      "epoch": 2.740909090909091,
+      "grad_norm": 1.6049258269378741,
+      "learning_rate": 5.308443045376845e-07,
+      "loss": 0.0101,
+      "step": 12060
+    },
+    {
+      "epoch": 2.7411363636363637,
+      "grad_norm": 1.4433532205702997,
+      "learning_rate": 5.307560742880025e-07,
+      "loss": 0.0297,
+      "step": 12061
+    },
+    {
+      "epoch": 2.7413636363636362,
+      "grad_norm": 1.5389920706348896,
+      "learning_rate": 5.306678459601187e-07,
+      "loss": 0.0108,
+      "step": 12062
+    },
+    {
+      "epoch": 2.741590909090909,
+      "grad_norm": 1.5567872997685719,
+      "learning_rate": 5.305796195558319e-07,
+      "loss": 0.0147,
+      "step": 12063
+    },
+    {
+      "epoch": 2.7418181818181817,
+      "grad_norm": 2.160184507726518,
+      "learning_rate": 5.304913950769415e-07,
+      "loss": 0.0242,
+      "step": 12064
+    },
+    {
+      "epoch": 2.7420454545454547,
+      "grad_norm": 0.46590013082725984,
+      "learning_rate": 5.304031725252463e-07,
+      "loss": 0.0074,
+      "step": 12065
+    },
+    {
+      "epoch": 2.742272727272727,
+      "grad_norm": 1.4640724815050767,
+      "learning_rate": 5.303149519025455e-07,
+      "loss": 0.0106,
+      "step": 12066
+    },
+    {
+      "epoch": 2.7425,
+      "grad_norm": 1.1593221664956164,
+      "learning_rate": 5.302267332106383e-07,
+      "loss": 0.019,
+      "step": 12067
+    },
+    {
+      "epoch": 2.7427272727272727,
+      "grad_norm": 1.3531254449359635,
+      "learning_rate": 5.301385164513228e-07,
+      "loss": 0.0218,
+      "step": 12068
+    },
+    {
+      "epoch": 2.7429545454545456,
+      "grad_norm": 2.419351044896246,
+      "learning_rate": 5.300503016263988e-07,
+      "loss": 0.0386,
+      "step": 12069
+    },
+    {
+      "epoch": 2.743181818181818,
+      "grad_norm": 1.5416889477241646,
+      "learning_rate": 5.299620887376647e-07,
+      "loss": 0.0152,
+      "step": 12070
+    },
+    {
+      "epoch": 2.7434090909090907,
+      "grad_norm": 1.3000412500275569,
+      "learning_rate": 5.298738777869198e-07,
+      "loss": 0.0151,
+      "step": 12071
+    },
+    {
+      "epoch": 2.7436363636363637,
+      "grad_norm": 7.599825310712453,
+      "learning_rate": 5.29785668775962e-07,
+      "loss": 0.0437,
+      "step": 12072
+    },
+    {
+      "epoch": 2.7438636363636366,
+      "grad_norm": 2.076190490363368,
+      "learning_rate": 5.296974617065906e-07,
+      "loss": 0.024,
+      "step": 12073
+    },
+    {
+      "epoch": 2.744090909090909,
+      "grad_norm": 1.8585613804840313,
+      "learning_rate": 5.296092565806043e-07,
+      "loss": 0.0302,
+      "step": 12074
+    },
+    {
+      "epoch": 2.7443181818181817,
+      "grad_norm": 1.6212418563807975,
+      "learning_rate": 5.295210533998016e-07,
+      "loss": 0.03,
+      "step": 12075
+    },
+    {
+      "epoch": 2.7445454545454546,
+      "grad_norm": 1.3218595161968463,
+      "learning_rate": 5.294328521659814e-07,
+      "loss": 0.0146,
+      "step": 12076
+    },
+    {
+      "epoch": 2.744772727272727,
+      "grad_norm": 1.6269446068919513,
+      "learning_rate": 5.293446528809418e-07,
+      "loss": 0.0301,
+      "step": 12077
+    },
+    {
+      "epoch": 2.745,
+      "grad_norm": 1.698151662715053,
+      "learning_rate": 5.292564555464817e-07,
+      "loss": 0.0219,
+      "step": 12078
+    },
+    {
+      "epoch": 2.7452272727272726,
+      "grad_norm": 1.3128184369987097,
+      "learning_rate": 5.291682601643996e-07,
+      "loss": 0.0292,
+      "step": 12079
+    },
+    {
+      "epoch": 2.7454545454545456,
+      "grad_norm": 2.4354008750214846,
+      "learning_rate": 5.290800667364939e-07,
+      "loss": 0.023,
+      "step": 12080
+    },
+    {
+      "epoch": 2.745681818181818,
+      "grad_norm": 2.3460385259596794,
+      "learning_rate": 5.28991875264563e-07,
+      "loss": 0.0192,
+      "step": 12081
+    },
+    {
+      "epoch": 2.745909090909091,
+      "grad_norm": 2.198662757561689,
+      "learning_rate": 5.289036857504051e-07,
+      "loss": 0.0242,
+      "step": 12082
+    },
+    {
+      "epoch": 2.7461363636363636,
+      "grad_norm": 1.6510811146267295,
+      "learning_rate": 5.288154981958187e-07,
+      "loss": 0.0286,
+      "step": 12083
+    },
+    {
+      "epoch": 2.746363636363636,
+      "grad_norm": 0.7073239178950238,
+      "learning_rate": 5.287273126026022e-07,
+      "loss": 0.0117,
+      "step": 12084
+    },
+    {
+      "epoch": 2.746590909090909,
+      "grad_norm": 2.27535912172285,
+      "learning_rate": 5.286391289725537e-07,
+      "loss": 0.0319,
+      "step": 12085
+    },
+    {
+      "epoch": 2.746818181818182,
+      "grad_norm": 2.3248395098524153,
+      "learning_rate": 5.285509473074714e-07,
+      "loss": 0.0464,
+      "step": 12086
+    },
+    {
+      "epoch": 2.7470454545454546,
+      "grad_norm": 2.39789831378696,
+      "learning_rate": 5.284627676091535e-07,
+      "loss": 0.0349,
+      "step": 12087
+    },
+    {
+      "epoch": 2.747272727272727,
+      "grad_norm": 2.089126303736408,
+      "learning_rate": 5.283745898793983e-07,
+      "loss": 0.0192,
+      "step": 12088
+    },
+    {
+      "epoch": 2.7475,
+      "grad_norm": 2.088324381807859,
+      "learning_rate": 5.282864141200041e-07,
+      "loss": 0.0333,
+      "step": 12089
+    },
+    {
+      "epoch": 2.7477272727272726,
+      "grad_norm": 1.8212620279080594,
+      "learning_rate": 5.281982403327682e-07,
+      "loss": 0.0221,
+      "step": 12090
+    },
+    {
+      "epoch": 2.7479545454545455,
+      "grad_norm": 1.0118909406148946,
+      "learning_rate": 5.281100685194891e-07,
+      "loss": 0.0111,
+      "step": 12091
+    },
+    {
+      "epoch": 2.748181818181818,
+      "grad_norm": 1.8990897312242414,
+      "learning_rate": 5.280218986819648e-07,
+      "loss": 0.0249,
+      "step": 12092
+    },
+    {
+      "epoch": 2.748409090909091,
+      "grad_norm": 3.1129551622497984,
+      "learning_rate": 5.279337308219933e-07,
+      "loss": 0.0422,
+      "step": 12093
+    },
+    {
+      "epoch": 2.7486363636363635,
+      "grad_norm": 1.662390815986015,
+      "learning_rate": 5.278455649413723e-07,
+      "loss": 0.0219,
+      "step": 12094
+    },
+    {
+      "epoch": 2.7488636363636365,
+      "grad_norm": 1.3785262750874028,
+      "learning_rate": 5.277574010418997e-07,
+      "loss": 0.0258,
+      "step": 12095
+    },
+    {
+      "epoch": 2.749090909090909,
+      "grad_norm": 1.8616536071227565,
+      "learning_rate": 5.276692391253732e-07,
+      "loss": 0.0197,
+      "step": 12096
+    },
+    {
+      "epoch": 2.7493181818181816,
+      "grad_norm": 1.546496879214579,
+      "learning_rate": 5.275810791935908e-07,
+      "loss": 0.0259,
+      "step": 12097
+    },
+    {
+      "epoch": 2.7495454545454545,
+      "grad_norm": 2.0839563288061695,
+      "learning_rate": 5.274929212483504e-07,
+      "loss": 0.0226,
+      "step": 12098
+    },
+    {
+      "epoch": 2.7497727272727275,
+      "grad_norm": 2.220563246003215,
+      "learning_rate": 5.274047652914492e-07,
+      "loss": 0.0242,
+      "step": 12099
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.8160764004354184,
+      "learning_rate": 5.27316611324685e-07,
+      "loss": 0.0338,
+      "step": 12100
+    },
+    {
+      "epoch": 2.7502272727272725,
+      "grad_norm": 1.148829235279144,
+      "learning_rate": 5.272284593498556e-07,
+      "loss": 0.0109,
+      "step": 12101
+    },
+    {
+      "epoch": 2.7504545454545455,
+      "grad_norm": 2.000407812144859,
+      "learning_rate": 5.271403093687589e-07,
+      "loss": 0.0311,
+      "step": 12102
+    },
+    {
+      "epoch": 2.7506818181818184,
+      "grad_norm": 1.4665730985198615,
+      "learning_rate": 5.270521613831916e-07,
+      "loss": 0.0206,
+      "step": 12103
+    },
+    {
+      "epoch": 2.750909090909091,
+      "grad_norm": 2.270664712711041,
+      "learning_rate": 5.269640153949516e-07,
+      "loss": 0.0207,
+      "step": 12104
+    },
+    {
+      "epoch": 2.7511363636363635,
+      "grad_norm": 1.3892375574539348,
+      "learning_rate": 5.268758714058364e-07,
+      "loss": 0.0103,
+      "step": 12105
+    },
+    {
+      "epoch": 2.7513636363636365,
+      "grad_norm": 1.4554701097229086,
+      "learning_rate": 5.267877294176435e-07,
+      "loss": 0.015,
+      "step": 12106
+    },
+    {
+      "epoch": 2.751590909090909,
+      "grad_norm": 2.7673717406396507,
+      "learning_rate": 5.266995894321699e-07,
+      "loss": 0.0335,
+      "step": 12107
+    },
+    {
+      "epoch": 2.751818181818182,
+      "grad_norm": 2.163231582471017,
+      "learning_rate": 5.266114514512134e-07,
+      "loss": 0.034,
+      "step": 12108
+    },
+    {
+      "epoch": 2.7520454545454545,
+      "grad_norm": 2.680626628223964,
+      "learning_rate": 5.265233154765708e-07,
+      "loss": 0.0605,
+      "step": 12109
+    },
+    {
+      "epoch": 2.7522727272727274,
+      "grad_norm": 1.7647599827294196,
+      "learning_rate": 5.264351815100398e-07,
+      "loss": 0.0243,
+      "step": 12110
+    },
+    {
+      "epoch": 2.7525,
+      "grad_norm": 1.836386980150448,
+      "learning_rate": 5.263470495534176e-07,
+      "loss": 0.024,
+      "step": 12111
+    },
+    {
+      "epoch": 2.752727272727273,
+      "grad_norm": 2.2110355049284114,
+      "learning_rate": 5.262589196085008e-07,
+      "loss": 0.0287,
+      "step": 12112
+    },
+    {
+      "epoch": 2.7529545454545454,
+      "grad_norm": 2.226763900908649,
+      "learning_rate": 5.261707916770871e-07,
+      "loss": 0.024,
+      "step": 12113
+    },
+    {
+      "epoch": 2.753181818181818,
+      "grad_norm": 2.899089455138853,
+      "learning_rate": 5.260826657609732e-07,
+      "loss": 0.0114,
+      "step": 12114
+    },
+    {
+      "epoch": 2.753409090909091,
+      "grad_norm": 1.0439765554927463,
+      "learning_rate": 5.259945418619565e-07,
+      "loss": 0.0097,
+      "step": 12115
+    },
+    {
+      "epoch": 2.753636363636364,
+      "grad_norm": 1.9070596363403451,
+      "learning_rate": 5.259064199818337e-07,
+      "loss": 0.0353,
+      "step": 12116
+    },
+    {
+      "epoch": 2.7538636363636364,
+      "grad_norm": 0.9579886290929541,
+      "learning_rate": 5.258183001224019e-07,
+      "loss": 0.009,
+      "step": 12117
+    },
+    {
+      "epoch": 2.754090909090909,
+      "grad_norm": 1.0135347300377917,
+      "learning_rate": 5.257301822854579e-07,
+      "loss": 0.0158,
+      "step": 12118
+    },
+    {
+      "epoch": 2.754318181818182,
+      "grad_norm": 1.7761203826544638,
+      "learning_rate": 5.256420664727988e-07,
+      "loss": 0.0181,
+      "step": 12119
+    },
+    {
+      "epoch": 2.7545454545454544,
+      "grad_norm": 3.6539366768465826,
+      "learning_rate": 5.255539526862214e-07,
+      "loss": 0.0411,
+      "step": 12120
+    },
+    {
+      "epoch": 2.7547727272727274,
+      "grad_norm": 37.86539783184717,
+      "learning_rate": 5.254658409275223e-07,
+      "loss": 0.0064,
+      "step": 12121
+    },
+    {
+      "epoch": 2.755,
+      "grad_norm": 3.379342778880706,
+      "learning_rate": 5.253777311984982e-07,
+      "loss": 0.0389,
+      "step": 12122
+    },
+    {
+      "epoch": 2.755227272727273,
+      "grad_norm": 2.6162991987772757,
+      "learning_rate": 5.25289623500946e-07,
+      "loss": 0.0483,
+      "step": 12123
+    },
+    {
+      "epoch": 2.7554545454545454,
+      "grad_norm": 1.9121794224187236,
+      "learning_rate": 5.252015178366626e-07,
+      "loss": 0.0296,
+      "step": 12124
+    },
+    {
+      "epoch": 2.7556818181818183,
+      "grad_norm": 1.5932835035587254,
+      "learning_rate": 5.251134142074442e-07,
+      "loss": 0.0293,
+      "step": 12125
+    },
+    {
+      "epoch": 2.755909090909091,
+      "grad_norm": 2.1254498561534074,
+      "learning_rate": 5.250253126150875e-07,
+      "loss": 0.0163,
+      "step": 12126
+    },
+    {
+      "epoch": 2.7561363636363634,
+      "grad_norm": 0.9646145202772739,
+      "learning_rate": 5.249372130613891e-07,
+      "loss": 0.0107,
+      "step": 12127
+    },
+    {
+      "epoch": 2.7563636363636363,
+      "grad_norm": 2.1701225637205335,
+      "learning_rate": 5.248491155481455e-07,
+      "loss": 0.0218,
+      "step": 12128
+    },
+    {
+      "epoch": 2.7565909090909093,
+      "grad_norm": 3.0574160845516696,
+      "learning_rate": 5.247610200771532e-07,
+      "loss": 0.0443,
+      "step": 12129
+    },
+    {
+      "epoch": 2.756818181818182,
+      "grad_norm": 1.2572350456224815,
+      "learning_rate": 5.246729266502085e-07,
+      "loss": 0.0146,
+      "step": 12130
+    },
+    {
+      "epoch": 2.7570454545454544,
+      "grad_norm": 1.0074634581169504,
+      "learning_rate": 5.245848352691077e-07,
+      "loss": 0.0143,
+      "step": 12131
+    },
+    {
+      "epoch": 2.7572727272727273,
+      "grad_norm": 1.395882130509205,
+      "learning_rate": 5.244967459356475e-07,
+      "loss": 0.0193,
+      "step": 12132
+    },
+    {
+      "epoch": 2.7575,
+      "grad_norm": 1.9941069445143609,
+      "learning_rate": 5.244086586516242e-07,
+      "loss": 0.0192,
+      "step": 12133
+    },
+    {
+      "epoch": 2.757727272727273,
+      "grad_norm": 2.7817764427058176,
+      "learning_rate": 5.243205734188336e-07,
+      "loss": 0.0499,
+      "step": 12134
+    },
+    {
+      "epoch": 2.7579545454545453,
+      "grad_norm": 0.7665074105898195,
+      "learning_rate": 5.242324902390721e-07,
+      "loss": 0.0105,
+      "step": 12135
+    },
+    {
+      "epoch": 2.7581818181818183,
+      "grad_norm": 3.4454070009361937,
+      "learning_rate": 5.24144409114136e-07,
+      "loss": 0.0576,
+      "step": 12136
+    },
+    {
+      "epoch": 2.758409090909091,
+      "grad_norm": 1.1068326723339792,
+      "learning_rate": 5.240563300458215e-07,
+      "loss": 0.0139,
+      "step": 12137
+    },
+    {
+      "epoch": 2.7586363636363638,
+      "grad_norm": 2.4669543336542565,
+      "learning_rate": 5.239682530359243e-07,
+      "loss": 0.0198,
+      "step": 12138
+    },
+    {
+      "epoch": 2.7588636363636363,
+      "grad_norm": 3.5130461650934532,
+      "learning_rate": 5.238801780862407e-07,
+      "loss": 0.0252,
+      "step": 12139
+    },
+    {
+      "epoch": 2.759090909090909,
+      "grad_norm": 1.5565472769226703,
+      "learning_rate": 5.237921051985667e-07,
+      "loss": 0.0106,
+      "step": 12140
+    },
+    {
+      "epoch": 2.759318181818182,
+      "grad_norm": 1.6059226308186154,
+      "learning_rate": 5.237040343746986e-07,
+      "loss": 0.0182,
+      "step": 12141
+    },
+    {
+      "epoch": 2.7595454545454547,
+      "grad_norm": 3.1330084621066656,
+      "learning_rate": 5.236159656164315e-07,
+      "loss": 0.0415,
+      "step": 12142
+    },
+    {
+      "epoch": 2.7597727272727273,
+      "grad_norm": 4.944602516394686,
+      "learning_rate": 5.235278989255618e-07,
+      "loss": 0.0443,
+      "step": 12143
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.7975817801185303,
+      "learning_rate": 5.234398343038854e-07,
+      "loss": 0.0196,
+      "step": 12144
+    },
+    {
+      "epoch": 2.7602272727272728,
+      "grad_norm": 1.6923697492818788,
+      "learning_rate": 5.233517717531978e-07,
+      "loss": 0.0136,
+      "step": 12145
+    },
+    {
+      "epoch": 2.7604545454545457,
+      "grad_norm": 2.3375389824452277,
+      "learning_rate": 5.232637112752951e-07,
+      "loss": 0.0448,
+      "step": 12146
+    },
+    {
+      "epoch": 2.7606818181818182,
+      "grad_norm": 2.5121221832774743,
+      "learning_rate": 5.231756528719727e-07,
+      "loss": 0.0282,
+      "step": 12147
+    },
+    {
+      "epoch": 2.7609090909090908,
+      "grad_norm": 2.2931140170975426,
+      "learning_rate": 5.230875965450263e-07,
+      "loss": 0.0327,
+      "step": 12148
+    },
+    {
+      "epoch": 2.7611363636363637,
+      "grad_norm": 1.8756823615843,
+      "learning_rate": 5.229995422962517e-07,
+      "loss": 0.0192,
+      "step": 12149
+    },
+    {
+      "epoch": 2.7613636363636362,
+      "grad_norm": 0.8836024327389775,
+      "learning_rate": 5.229114901274447e-07,
+      "loss": 0.0076,
+      "step": 12150
+    },
+    {
+      "epoch": 2.761590909090909,
+      "grad_norm": 1.3733290590091258,
+      "learning_rate": 5.228234400404002e-07,
+      "loss": 0.0128,
+      "step": 12151
+    },
+    {
+      "epoch": 2.7618181818181817,
+      "grad_norm": 1.5156975371756862,
+      "learning_rate": 5.227353920369138e-07,
+      "loss": 0.0123,
+      "step": 12152
+    },
+    {
+      "epoch": 2.7620454545454547,
+      "grad_norm": 0.9840264223953956,
+      "learning_rate": 5.226473461187813e-07,
+      "loss": 0.0163,
+      "step": 12153
+    },
+    {
+      "epoch": 2.762272727272727,
+      "grad_norm": 1.9551046822478273,
+      "learning_rate": 5.225593022877981e-07,
+      "loss": 0.0179,
+      "step": 12154
+    },
+    {
+      "epoch": 2.7625,
+      "grad_norm": 2.4976398849130734,
+      "learning_rate": 5.224712605457596e-07,
+      "loss": 0.0245,
+      "step": 12155
+    },
+    {
+      "epoch": 2.7627272727272727,
+      "grad_norm": 1.447379702863345,
+      "learning_rate": 5.223832208944607e-07,
+      "loss": 0.0159,
+      "step": 12156
+    },
+    {
+      "epoch": 2.762954545454545,
+      "grad_norm": 1.746855357888169,
+      "learning_rate": 5.222951833356971e-07,
+      "loss": 0.0147,
+      "step": 12157
+    },
+    {
+      "epoch": 2.763181818181818,
+      "grad_norm": 2.7111915973619505,
+      "learning_rate": 5.222071478712637e-07,
+      "loss": 0.0189,
+      "step": 12158
+    },
+    {
+      "epoch": 2.763409090909091,
+      "grad_norm": 4.0385242366886445,
+      "learning_rate": 5.221191145029561e-07,
+      "loss": 0.0325,
+      "step": 12159
+    },
+    {
+      "epoch": 2.7636363636363637,
+      "grad_norm": 1.967740117079476,
+      "learning_rate": 5.220310832325691e-07,
+      "loss": 0.0218,
+      "step": 12160
+    },
+    {
+      "epoch": 2.763863636363636,
+      "grad_norm": 2.104500287578427,
+      "learning_rate": 5.21943054061898e-07,
+      "loss": 0.0246,
+      "step": 12161
+    },
+    {
+      "epoch": 2.764090909090909,
+      "grad_norm": 2.0831379662248013,
+      "learning_rate": 5.218550269927378e-07,
+      "loss": 0.034,
+      "step": 12162
+    },
+    {
+      "epoch": 2.7643181818181817,
+      "grad_norm": 2.1917073743929247,
+      "learning_rate": 5.217670020268838e-07,
+      "loss": 0.0161,
+      "step": 12163
+    },
+    {
+      "epoch": 2.7645454545454546,
+      "grad_norm": 2.175419289364609,
+      "learning_rate": 5.216789791661305e-07,
+      "loss": 0.0232,
+      "step": 12164
+    },
+    {
+      "epoch": 2.764772727272727,
+      "grad_norm": 1.3099305912442354,
+      "learning_rate": 5.21590958412273e-07,
+      "loss": 0.0229,
+      "step": 12165
+    },
+    {
+      "epoch": 2.765,
+      "grad_norm": 2.24169234335686,
+      "learning_rate": 5.215029397671064e-07,
+      "loss": 0.0222,
+      "step": 12166
+    },
+    {
+      "epoch": 2.7652272727272726,
+      "grad_norm": 1.8783718431046446,
+      "learning_rate": 5.214149232324252e-07,
+      "loss": 0.0151,
+      "step": 12167
+    },
+    {
+      "epoch": 2.7654545454545456,
+      "grad_norm": 1.9338559460623508,
+      "learning_rate": 5.213269088100247e-07,
+      "loss": 0.0231,
+      "step": 12168
+    },
+    {
+      "epoch": 2.765681818181818,
+      "grad_norm": 2.3969338683902808,
+      "learning_rate": 5.212388965016992e-07,
+      "loss": 0.0303,
+      "step": 12169
+    },
+    {
+      "epoch": 2.7659090909090907,
+      "grad_norm": 1.4705319602570286,
+      "learning_rate": 5.211508863092437e-07,
+      "loss": 0.0169,
+      "step": 12170
+    },
+    {
+      "epoch": 2.7661363636363636,
+      "grad_norm": 1.799654159294105,
+      "learning_rate": 5.210628782344528e-07,
+      "loss": 0.0285,
+      "step": 12171
+    },
+    {
+      "epoch": 2.7663636363636366,
+      "grad_norm": 0.9959950785404856,
+      "learning_rate": 5.209748722791214e-07,
+      "loss": 0.0086,
+      "step": 12172
+    },
+    {
+      "epoch": 2.766590909090909,
+      "grad_norm": 1.4455343550793456,
+      "learning_rate": 5.208868684450435e-07,
+      "loss": 0.0257,
+      "step": 12173
+    },
+    {
+      "epoch": 2.7668181818181816,
+      "grad_norm": 0.7406752363705416,
+      "learning_rate": 5.20798866734014e-07,
+      "loss": 0.0039,
+      "step": 12174
+    },
+    {
+      "epoch": 2.7670454545454546,
+      "grad_norm": 1.1959067451989303,
+      "learning_rate": 5.207108671478274e-07,
+      "loss": 0.0188,
+      "step": 12175
+    },
+    {
+      "epoch": 2.767272727272727,
+      "grad_norm": 1.0087513631354148,
+      "learning_rate": 5.206228696882782e-07,
+      "loss": 0.0066,
+      "step": 12176
+    },
+    {
+      "epoch": 2.7675,
+      "grad_norm": 1.4610705826689623,
+      "learning_rate": 5.205348743571609e-07,
+      "loss": 0.0169,
+      "step": 12177
+    },
+    {
+      "epoch": 2.7677272727272726,
+      "grad_norm": 2.4165502293998666,
+      "learning_rate": 5.204468811562696e-07,
+      "loss": 0.0369,
+      "step": 12178
+    },
+    {
+      "epoch": 2.7679545454545456,
+      "grad_norm": 2.483157094947669,
+      "learning_rate": 5.203588900873989e-07,
+      "loss": 0.0409,
+      "step": 12179
+    },
+    {
+      "epoch": 2.768181818181818,
+      "grad_norm": 2.3981601774506913,
+      "learning_rate": 5.202709011523429e-07,
+      "loss": 0.0171,
+      "step": 12180
+    },
+    {
+      "epoch": 2.768409090909091,
+      "grad_norm": 1.9386599356370104,
+      "learning_rate": 5.201829143528962e-07,
+      "loss": 0.038,
+      "step": 12181
+    },
+    {
+      "epoch": 2.7686363636363636,
+      "grad_norm": 2.9097641598228665,
+      "learning_rate": 5.200949296908524e-07,
+      "loss": 0.0463,
+      "step": 12182
+    },
+    {
+      "epoch": 2.7688636363636365,
+      "grad_norm": 4.356071518675691,
+      "learning_rate": 5.20006947168006e-07,
+      "loss": 0.0307,
+      "step": 12183
+    },
+    {
+      "epoch": 2.769090909090909,
+      "grad_norm": 1.445760254796352,
+      "learning_rate": 5.199189667861513e-07,
+      "loss": 0.0252,
+      "step": 12184
+    },
+    {
+      "epoch": 2.769318181818182,
+      "grad_norm": 3.315516809172481,
+      "learning_rate": 5.198309885470823e-07,
+      "loss": 0.0301,
+      "step": 12185
+    },
+    {
+      "epoch": 2.7695454545454545,
+      "grad_norm": 1.3331441313324337,
+      "learning_rate": 5.197430124525927e-07,
+      "loss": 0.0352,
+      "step": 12186
+    },
+    {
+      "epoch": 2.769772727272727,
+      "grad_norm": 1.8011812266750895,
+      "learning_rate": 5.196550385044767e-07,
+      "loss": 0.0245,
+      "step": 12187
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 2.230075472049039,
+      "learning_rate": 5.195670667045284e-07,
+      "loss": 0.0253,
+      "step": 12188
+    },
+    {
+      "epoch": 2.770227272727273,
+      "grad_norm": 2.1129888729078448,
+      "learning_rate": 5.194790970545413e-07,
+      "loss": 0.0335,
+      "step": 12189
+    },
+    {
+      "epoch": 2.7704545454545455,
+      "grad_norm": 1.0461386015675511,
+      "learning_rate": 5.193911295563098e-07,
+      "loss": 0.0075,
+      "step": 12190
+    },
+    {
+      "epoch": 2.770681818181818,
+      "grad_norm": 1.8089998221683528,
+      "learning_rate": 5.193031642116273e-07,
+      "loss": 0.05,
+      "step": 12191
+    },
+    {
+      "epoch": 2.770909090909091,
+      "grad_norm": 0.7608372353063501,
+      "learning_rate": 5.192152010222876e-07,
+      "loss": 0.0177,
+      "step": 12192
+    },
+    {
+      "epoch": 2.7711363636363635,
+      "grad_norm": 2.0449286548884027,
+      "learning_rate": 5.191272399900845e-07,
+      "loss": 0.0228,
+      "step": 12193
+    },
+    {
+      "epoch": 2.7713636363636365,
+      "grad_norm": 1.283055591576752,
+      "learning_rate": 5.19039281116812e-07,
+      "loss": 0.0127,
+      "step": 12194
+    },
+    {
+      "epoch": 2.771590909090909,
+      "grad_norm": 1.1353269753379547,
+      "learning_rate": 5.189513244042631e-07,
+      "loss": 0.0149,
+      "step": 12195
+    },
+    {
+      "epoch": 2.771818181818182,
+      "grad_norm": 1.01169978554706,
+      "learning_rate": 5.188633698542317e-07,
+      "loss": 0.0119,
+      "step": 12196
+    },
+    {
+      "epoch": 2.7720454545454545,
+      "grad_norm": 1.4043574486314179,
+      "learning_rate": 5.187754174685114e-07,
+      "loss": 0.0157,
+      "step": 12197
+    },
+    {
+      "epoch": 2.7722727272727274,
+      "grad_norm": 1.4763290127197353,
+      "learning_rate": 5.186874672488956e-07,
+      "loss": 0.0127,
+      "step": 12198
+    },
+    {
+      "epoch": 2.7725,
+      "grad_norm": 2.2274977485401855,
+      "learning_rate": 5.18599519197178e-07,
+      "loss": 0.0304,
+      "step": 12199
+    },
+    {
+      "epoch": 2.7727272727272725,
+      "grad_norm": 1.0670426560310269,
+      "learning_rate": 5.185115733151517e-07,
+      "loss": 0.0134,
+      "step": 12200
+    },
+    {
+      "epoch": 2.7729545454545454,
+      "grad_norm": 2.1606423790761244,
+      "learning_rate": 5.184236296046101e-07,
+      "loss": 0.0229,
+      "step": 12201
+    },
+    {
+      "epoch": 2.7731818181818184,
+      "grad_norm": 0.8656855097123708,
+      "learning_rate": 5.183356880673468e-07,
+      "loss": 0.0092,
+      "step": 12202
+    },
+    {
+      "epoch": 2.773409090909091,
+      "grad_norm": 1.8521851309128476,
+      "learning_rate": 5.182477487051551e-07,
+      "loss": 0.0185,
+      "step": 12203
+    },
+    {
+      "epoch": 2.7736363636363635,
+      "grad_norm": 1.738169814527573,
+      "learning_rate": 5.181598115198277e-07,
+      "loss": 0.0147,
+      "step": 12204
+    },
+    {
+      "epoch": 2.7738636363636364,
+      "grad_norm": 2.1762620780437825,
+      "learning_rate": 5.180718765131582e-07,
+      "loss": 0.0212,
+      "step": 12205
+    },
+    {
+      "epoch": 2.774090909090909,
+      "grad_norm": 1.8077287036205343,
+      "learning_rate": 5.179839436869396e-07,
+      "loss": 0.022,
+      "step": 12206
+    },
+    {
+      "epoch": 2.774318181818182,
+      "grad_norm": 1.4402731349552433,
+      "learning_rate": 5.178960130429654e-07,
+      "loss": 0.0134,
+      "step": 12207
+    },
+    {
+      "epoch": 2.7745454545454544,
+      "grad_norm": 1.4661684624677842,
+      "learning_rate": 5.178080845830281e-07,
+      "loss": 0.0231,
+      "step": 12208
+    },
+    {
+      "epoch": 2.7747727272727274,
+      "grad_norm": 1.491626304976318,
+      "learning_rate": 5.17720158308921e-07,
+      "loss": 0.01,
+      "step": 12209
+    },
+    {
+      "epoch": 2.775,
+      "grad_norm": 7.837585445795834,
+      "learning_rate": 5.17632234222437e-07,
+      "loss": 0.0346,
+      "step": 12210
+    },
+    {
+      "epoch": 2.775227272727273,
+      "grad_norm": 2.1537324186806157,
+      "learning_rate": 5.17544312325369e-07,
+      "loss": 0.0209,
+      "step": 12211
+    },
+    {
+      "epoch": 2.7754545454545454,
+      "grad_norm": 2.0632663078332487,
+      "learning_rate": 5.174563926195102e-07,
+      "loss": 0.0212,
+      "step": 12212
+    },
+    {
+      "epoch": 2.775681818181818,
+      "grad_norm": 3.151676825309715,
+      "learning_rate": 5.17368475106653e-07,
+      "loss": 0.0334,
+      "step": 12213
+    },
+    {
+      "epoch": 2.775909090909091,
+      "grad_norm": 3.262147630751991,
+      "learning_rate": 5.172805597885903e-07,
+      "loss": 0.0432,
+      "step": 12214
+    },
+    {
+      "epoch": 2.776136363636364,
+      "grad_norm": 2.921348843497236,
+      "learning_rate": 5.17192646667115e-07,
+      "loss": 0.0205,
+      "step": 12215
+    },
+    {
+      "epoch": 2.7763636363636364,
+      "grad_norm": 1.889605639157288,
+      "learning_rate": 5.171047357440199e-07,
+      "loss": 0.0217,
+      "step": 12216
+    },
+    {
+      "epoch": 2.776590909090909,
+      "grad_norm": 1.952370561068969,
+      "learning_rate": 5.170168270210972e-07,
+      "loss": 0.0284,
+      "step": 12217
+    },
+    {
+      "epoch": 2.776818181818182,
+      "grad_norm": 2.9857471760459484,
+      "learning_rate": 5.169289205001399e-07,
+      "loss": 0.0438,
+      "step": 12218
+    },
+    {
+      "epoch": 2.777045454545455,
+      "grad_norm": 2.359602264680585,
+      "learning_rate": 5.168410161829404e-07,
+      "loss": 0.0259,
+      "step": 12219
+    },
+    {
+      "epoch": 2.7772727272727273,
+      "grad_norm": 0.6238977296611492,
+      "learning_rate": 5.167531140712911e-07,
+      "loss": 0.0101,
+      "step": 12220
+    },
+    {
+      "epoch": 2.7775,
+      "grad_norm": 1.7131568074209718,
+      "learning_rate": 5.16665214166985e-07,
+      "loss": 0.022,
+      "step": 12221
+    },
+    {
+      "epoch": 2.777727272727273,
+      "grad_norm": 8.888326272887216,
+      "learning_rate": 5.16577316471814e-07,
+      "loss": 0.0451,
+      "step": 12222
+    },
+    {
+      "epoch": 2.7779545454545453,
+      "grad_norm": 2.580028273309004,
+      "learning_rate": 5.164894209875706e-07,
+      "loss": 0.0211,
+      "step": 12223
+    },
+    {
+      "epoch": 2.7781818181818183,
+      "grad_norm": 1.4390028142940876,
+      "learning_rate": 5.164015277160473e-07,
+      "loss": 0.0106,
+      "step": 12224
+    },
+    {
+      "epoch": 2.778409090909091,
+      "grad_norm": 1.6112186086119213,
+      "learning_rate": 5.163136366590366e-07,
+      "loss": 0.0167,
+      "step": 12225
+    },
+    {
+      "epoch": 2.778636363636364,
+      "grad_norm": 2.450409472423976,
+      "learning_rate": 5.162257478183301e-07,
+      "loss": 0.0195,
+      "step": 12226
+    },
+    {
+      "epoch": 2.7788636363636363,
+      "grad_norm": 1.6899028004631038,
+      "learning_rate": 5.161378611957204e-07,
+      "loss": 0.0158,
+      "step": 12227
+    },
+    {
+      "epoch": 2.7790909090909093,
+      "grad_norm": 1.3102832707076892,
+      "learning_rate": 5.160499767929997e-07,
+      "loss": 0.0177,
+      "step": 12228
+    },
+    {
+      "epoch": 2.779318181818182,
+      "grad_norm": 1.9229856241668988,
+      "learning_rate": 5.159620946119601e-07,
+      "loss": 0.0212,
+      "step": 12229
+    },
+    {
+      "epoch": 2.7795454545454543,
+      "grad_norm": 1.6191040146223195,
+      "learning_rate": 5.158742146543935e-07,
+      "loss": 0.0172,
+      "step": 12230
+    },
+    {
+      "epoch": 2.7797727272727273,
+      "grad_norm": 2.1519193812440465,
+      "learning_rate": 5.15786336922092e-07,
+      "loss": 0.0283,
+      "step": 12231
+    },
+    {
+      "epoch": 2.7800000000000002,
+      "grad_norm": 2.3395193906832215,
+      "learning_rate": 5.156984614168477e-07,
+      "loss": 0.017,
+      "step": 12232
+    },
+    {
+      "epoch": 2.7802272727272728,
+      "grad_norm": 2.3423285845470447,
+      "learning_rate": 5.156105881404524e-07,
+      "loss": 0.0161,
+      "step": 12233
+    },
+    {
+      "epoch": 2.7804545454545453,
+      "grad_norm": 6.291375436344016,
+      "learning_rate": 5.155227170946984e-07,
+      "loss": 0.0633,
+      "step": 12234
+    },
+    {
+      "epoch": 2.7806818181818183,
+      "grad_norm": 0.6185195662408011,
+      "learning_rate": 5.154348482813769e-07,
+      "loss": 0.0043,
+      "step": 12235
+    },
+    {
+      "epoch": 2.7809090909090908,
+      "grad_norm": 1.1185799877166311,
+      "learning_rate": 5.153469817022798e-07,
+      "loss": 0.0225,
+      "step": 12236
+    },
+    {
+      "epoch": 2.7811363636363637,
+      "grad_norm": 1.9978703522776093,
+      "learning_rate": 5.152591173591992e-07,
+      "loss": 0.0132,
+      "step": 12237
+    },
+    {
+      "epoch": 2.7813636363636363,
+      "grad_norm": 1.8097578807773491,
+      "learning_rate": 5.15171255253927e-07,
+      "loss": 0.0159,
+      "step": 12238
+    },
+    {
+      "epoch": 2.7815909090909092,
+      "grad_norm": 1.0356815612622665,
+      "learning_rate": 5.150833953882541e-07,
+      "loss": 0.0218,
+      "step": 12239
+    },
+    {
+      "epoch": 2.7818181818181817,
+      "grad_norm": 1.0092126326259478,
+      "learning_rate": 5.149955377639726e-07,
+      "loss": 0.0119,
+      "step": 12240
+    },
+    {
+      "epoch": 2.7820454545454547,
+      "grad_norm": 0.8913315772742072,
+      "learning_rate": 5.14907682382874e-07,
+      "loss": 0.0108,
+      "step": 12241
+    },
+    {
+      "epoch": 2.7822727272727272,
+      "grad_norm": 2.567709954895636,
+      "learning_rate": 5.148198292467499e-07,
+      "loss": 0.0263,
+      "step": 12242
+    },
+    {
+      "epoch": 2.7824999999999998,
+      "grad_norm": 2.157681478633534,
+      "learning_rate": 5.147319783573917e-07,
+      "loss": 0.0209,
+      "step": 12243
+    },
+    {
+      "epoch": 2.7827272727272727,
+      "grad_norm": 1.320891909825229,
+      "learning_rate": 5.146441297165907e-07,
+      "loss": 0.0172,
+      "step": 12244
+    },
+    {
+      "epoch": 2.7829545454545457,
+      "grad_norm": 1.1192738520294092,
+      "learning_rate": 5.145562833261384e-07,
+      "loss": 0.0053,
+      "step": 12245
+    },
+    {
+      "epoch": 2.783181818181818,
+      "grad_norm": 2.94448741364395,
+      "learning_rate": 5.144684391878264e-07,
+      "loss": 0.0327,
+      "step": 12246
+    },
+    {
+      "epoch": 2.7834090909090907,
+      "grad_norm": 2.159509546944381,
+      "learning_rate": 5.143805973034458e-07,
+      "loss": 0.0147,
+      "step": 12247
+    },
+    {
+      "epoch": 2.7836363636363637,
+      "grad_norm": 0.888331848422377,
+      "learning_rate": 5.142927576747876e-07,
+      "loss": 0.0086,
+      "step": 12248
+    },
+    {
+      "epoch": 2.783863636363636,
+      "grad_norm": 1.821519961923764,
+      "learning_rate": 5.142049203036433e-07,
+      "loss": 0.0222,
+      "step": 12249
+    },
+    {
+      "epoch": 2.784090909090909,
+      "grad_norm": 2.8507171988906745,
+      "learning_rate": 5.141170851918038e-07,
+      "loss": 0.0435,
+      "step": 12250
+    },
+    {
+      "epoch": 2.7843181818181817,
+      "grad_norm": 1.7163590220270917,
+      "learning_rate": 5.140292523410606e-07,
+      "loss": 0.024,
+      "step": 12251
+    },
+    {
+      "epoch": 2.7845454545454547,
+      "grad_norm": 1.7123381064517609,
+      "learning_rate": 5.139414217532044e-07,
+      "loss": 0.0365,
+      "step": 12252
+    },
+    {
+      "epoch": 2.784772727272727,
+      "grad_norm": 3.528324668106626,
+      "learning_rate": 5.138535934300263e-07,
+      "loss": 0.0517,
+      "step": 12253
+    },
+    {
+      "epoch": 2.785,
+      "grad_norm": 3.9097934232287277,
+      "learning_rate": 5.137657673733175e-07,
+      "loss": 0.0154,
+      "step": 12254
+    },
+    {
+      "epoch": 2.7852272727272727,
+      "grad_norm": 1.4605156218506459,
+      "learning_rate": 5.136779435848686e-07,
+      "loss": 0.0159,
+      "step": 12255
+    },
+    {
+      "epoch": 2.785454545454545,
+      "grad_norm": 2.8109463710455573,
+      "learning_rate": 5.13590122066471e-07,
+      "loss": 0.0551,
+      "step": 12256
+    },
+    {
+      "epoch": 2.785681818181818,
+      "grad_norm": 1.1856562373861046,
+      "learning_rate": 5.135023028199149e-07,
+      "loss": 0.0132,
+      "step": 12257
+    },
+    {
+      "epoch": 2.785909090909091,
+      "grad_norm": 1.720117467817091,
+      "learning_rate": 5.134144858469914e-07,
+      "loss": 0.0199,
+      "step": 12258
+    },
+    {
+      "epoch": 2.7861363636363636,
+      "grad_norm": 0.7609689368407956,
+      "learning_rate": 5.13326671149491e-07,
+      "loss": 0.0068,
+      "step": 12259
+    },
+    {
+      "epoch": 2.786363636363636,
+      "grad_norm": 2.130871247065307,
+      "learning_rate": 5.13238858729205e-07,
+      "loss": 0.0386,
+      "step": 12260
+    },
+    {
+      "epoch": 2.786590909090909,
+      "grad_norm": 3.4205094814784625,
+      "learning_rate": 5.131510485879235e-07,
+      "loss": 0.0458,
+      "step": 12261
+    },
+    {
+      "epoch": 2.786818181818182,
+      "grad_norm": 1.7147939107826125,
+      "learning_rate": 5.13063240727437e-07,
+      "loss": 0.0166,
+      "step": 12262
+    },
+    {
+      "epoch": 2.7870454545454546,
+      "grad_norm": 2.910585335717366,
+      "learning_rate": 5.129754351495365e-07,
+      "loss": 0.0339,
+      "step": 12263
+    },
+    {
+      "epoch": 2.787272727272727,
+      "grad_norm": 1.766473711593712,
+      "learning_rate": 5.128876318560126e-07,
+      "loss": 0.0151,
+      "step": 12264
+    },
+    {
+      "epoch": 2.7875,
+      "grad_norm": 1.8328041924222107,
+      "learning_rate": 5.127998308486549e-07,
+      "loss": 0.0268,
+      "step": 12265
+    },
+    {
+      "epoch": 2.7877272727272726,
+      "grad_norm": 2.368865897105551,
+      "learning_rate": 5.127120321292547e-07,
+      "loss": 0.0185,
+      "step": 12266
+    },
+    {
+      "epoch": 2.7879545454545456,
+      "grad_norm": 1.7568654934606627,
+      "learning_rate": 5.12624235699602e-07,
+      "loss": 0.02,
+      "step": 12267
+    },
+    {
+      "epoch": 2.788181818181818,
+      "grad_norm": 3.500368592011555,
+      "learning_rate": 5.125364415614871e-07,
+      "loss": 0.0447,
+      "step": 12268
+    },
+    {
+      "epoch": 2.788409090909091,
+      "grad_norm": 2.839862226476313,
+      "learning_rate": 5.124486497167009e-07,
+      "loss": 0.0355,
+      "step": 12269
+    },
+    {
+      "epoch": 2.7886363636363636,
+      "grad_norm": 1.6807303670900233,
+      "learning_rate": 5.123608601670325e-07,
+      "loss": 0.0245,
+      "step": 12270
+    },
+    {
+      "epoch": 2.7888636363636365,
+      "grad_norm": 2.763917427090092,
+      "learning_rate": 5.122730729142729e-07,
+      "loss": 0.0311,
+      "step": 12271
+    },
+    {
+      "epoch": 2.789090909090909,
+      "grad_norm": 2.7170270067324265,
+      "learning_rate": 5.121852879602118e-07,
+      "loss": 0.0133,
+      "step": 12272
+    },
+    {
+      "epoch": 2.7893181818181816,
+      "grad_norm": 2.570471626064527,
+      "learning_rate": 5.120975053066398e-07,
+      "loss": 0.023,
+      "step": 12273
+    },
+    {
+      "epoch": 2.7895454545454546,
+      "grad_norm": 2.6216371402172687,
+      "learning_rate": 5.120097249553464e-07,
+      "loss": 0.0203,
+      "step": 12274
+    },
+    {
+      "epoch": 2.7897727272727275,
+      "grad_norm": 1.1977451372769619,
+      "learning_rate": 5.119219469081219e-07,
+      "loss": 0.013,
+      "step": 12275
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 2.1831970189391474,
+      "learning_rate": 5.118341711667562e-07,
+      "loss": 0.023,
+      "step": 12276
+    },
+    {
+      "epoch": 2.7902272727272726,
+      "grad_norm": 2.539468692773852,
+      "learning_rate": 5.117463977330391e-07,
+      "loss": 0.0325,
+      "step": 12277
+    },
+    {
+      "epoch": 2.7904545454545455,
+      "grad_norm": 4.292113385644927,
+      "learning_rate": 5.116586266087608e-07,
+      "loss": 0.0316,
+      "step": 12278
+    },
+    {
+      "epoch": 2.790681818181818,
+      "grad_norm": 2.374109698620887,
+      "learning_rate": 5.115708577957106e-07,
+      "loss": 0.0143,
+      "step": 12279
+    },
+    {
+      "epoch": 2.790909090909091,
+      "grad_norm": 2.2974410146894964,
+      "learning_rate": 5.114830912956785e-07,
+      "loss": 0.0229,
+      "step": 12280
+    },
+    {
+      "epoch": 2.7911363636363635,
+      "grad_norm": 1.8682130311823428,
+      "learning_rate": 5.113953271104542e-07,
+      "loss": 0.024,
+      "step": 12281
+    },
+    {
+      "epoch": 2.7913636363636365,
+      "grad_norm": 3.403164342038426,
+      "learning_rate": 5.113075652418276e-07,
+      "loss": 0.0563,
+      "step": 12282
+    },
+    {
+      "epoch": 2.791590909090909,
+      "grad_norm": 3.3720699730796038,
+      "learning_rate": 5.112198056915878e-07,
+      "loss": 0.0548,
+      "step": 12283
+    },
+    {
+      "epoch": 2.791818181818182,
+      "grad_norm": 0.9059064234519231,
+      "learning_rate": 5.111320484615248e-07,
+      "loss": 0.0162,
+      "step": 12284
+    },
+    {
+      "epoch": 2.7920454545454545,
+      "grad_norm": 3.1032617189927323,
+      "learning_rate": 5.110442935534279e-07,
+      "loss": 0.0396,
+      "step": 12285
+    },
+    {
+      "epoch": 2.792272727272727,
+      "grad_norm": 3.2578925367748175,
+      "learning_rate": 5.10956540969087e-07,
+      "loss": 0.0367,
+      "step": 12286
+    },
+    {
+      "epoch": 2.7925,
+      "grad_norm": 2.5932279544870958,
+      "learning_rate": 5.108687907102908e-07,
+      "loss": 0.014,
+      "step": 12287
+    },
+    {
+      "epoch": 2.792727272727273,
+      "grad_norm": 1.652184175529946,
+      "learning_rate": 5.107810427788291e-07,
+      "loss": 0.0104,
+      "step": 12288
+    },
+    {
+      "epoch": 2.7929545454545455,
+      "grad_norm": 2.0129948260556767,
+      "learning_rate": 5.106932971764914e-07,
+      "loss": 0.0365,
+      "step": 12289
+    },
+    {
+      "epoch": 2.793181818181818,
+      "grad_norm": 0.5761927758679061,
+      "learning_rate": 5.106055539050665e-07,
+      "loss": 0.0046,
+      "step": 12290
+    },
+    {
+      "epoch": 2.793409090909091,
+      "grad_norm": 1.0650905077019484,
+      "learning_rate": 5.105178129663443e-07,
+      "loss": 0.0122,
+      "step": 12291
+    },
+    {
+      "epoch": 2.7936363636363635,
+      "grad_norm": 2.903807048853938,
+      "learning_rate": 5.104300743621134e-07,
+      "loss": 0.0198,
+      "step": 12292
+    },
+    {
+      "epoch": 2.7938636363636364,
+      "grad_norm": 1.7893681869481977,
+      "learning_rate": 5.10342338094163e-07,
+      "loss": 0.022,
+      "step": 12293
+    },
+    {
+      "epoch": 2.794090909090909,
+      "grad_norm": 1.9360729948219038,
+      "learning_rate": 5.102546041642825e-07,
+      "loss": 0.0216,
+      "step": 12294
+    },
+    {
+      "epoch": 2.794318181818182,
+      "grad_norm": 1.3548187238156728,
+      "learning_rate": 5.10166872574261e-07,
+      "loss": 0.0115,
+      "step": 12295
+    },
+    {
+      "epoch": 2.7945454545454544,
+      "grad_norm": 1.624625825869628,
+      "learning_rate": 5.100791433258871e-07,
+      "loss": 0.0123,
+      "step": 12296
+    },
+    {
+      "epoch": 2.7947727272727274,
+      "grad_norm": 2.781516367976031,
+      "learning_rate": 5.0999141642095e-07,
+      "loss": 0.0409,
+      "step": 12297
+    },
+    {
+      "epoch": 2.795,
+      "grad_norm": 2.103016092918984,
+      "learning_rate": 5.099036918612386e-07,
+      "loss": 0.0219,
+      "step": 12298
+    },
+    {
+      "epoch": 2.795227272727273,
+      "grad_norm": 1.605502029014961,
+      "learning_rate": 5.098159696485417e-07,
+      "loss": 0.0169,
+      "step": 12299
+    },
+    {
+      "epoch": 2.7954545454545454,
+      "grad_norm": 4.716142721942775,
+      "learning_rate": 5.097282497846483e-07,
+      "loss": 0.058,
+      "step": 12300
+    },
+    {
+      "epoch": 2.7956818181818184,
+      "grad_norm": 1.5446658328410345,
+      "learning_rate": 5.096405322713467e-07,
+      "loss": 0.0121,
+      "step": 12301
+    },
+    {
+      "epoch": 2.795909090909091,
+      "grad_norm": 1.85600979649271,
+      "learning_rate": 5.09552817110426e-07,
+      "loss": 0.019,
+      "step": 12302
+    },
+    {
+      "epoch": 2.7961363636363634,
+      "grad_norm": 2.0211700708247475,
+      "learning_rate": 5.094651043036748e-07,
+      "loss": 0.0373,
+      "step": 12303
+    },
+    {
+      "epoch": 2.7963636363636364,
+      "grad_norm": 3.0566200047867045,
+      "learning_rate": 5.093773938528818e-07,
+      "loss": 0.0262,
+      "step": 12304
+    },
+    {
+      "epoch": 2.7965909090909093,
+      "grad_norm": 1.3798682749358577,
+      "learning_rate": 5.092896857598352e-07,
+      "loss": 0.0096,
+      "step": 12305
+    },
+    {
+      "epoch": 2.796818181818182,
+      "grad_norm": 2.3622260874110883,
+      "learning_rate": 5.092019800263239e-07,
+      "loss": 0.0249,
+      "step": 12306
+    },
+    {
+      "epoch": 2.7970454545454544,
+      "grad_norm": 1.2474935381620338,
+      "learning_rate": 5.091142766541362e-07,
+      "loss": 0.0132,
+      "step": 12307
+    },
+    {
+      "epoch": 2.7972727272727274,
+      "grad_norm": 1.1972322897099055,
+      "learning_rate": 5.090265756450609e-07,
+      "loss": 0.0165,
+      "step": 12308
+    },
+    {
+      "epoch": 2.7975,
+      "grad_norm": 2.1283105194565697,
+      "learning_rate": 5.089388770008857e-07,
+      "loss": 0.0362,
+      "step": 12309
+    },
+    {
+      "epoch": 2.797727272727273,
+      "grad_norm": 2.9569403276196695,
+      "learning_rate": 5.088511807233992e-07,
+      "loss": 0.0449,
+      "step": 12310
+    },
+    {
+      "epoch": 2.7979545454545454,
+      "grad_norm": 2.1895729193001547,
+      "learning_rate": 5.0876348681439e-07,
+      "loss": 0.0251,
+      "step": 12311
+    },
+    {
+      "epoch": 2.7981818181818183,
+      "grad_norm": 1.2012141250255266,
+      "learning_rate": 5.086757952756459e-07,
+      "loss": 0.0111,
+      "step": 12312
+    },
+    {
+      "epoch": 2.798409090909091,
+      "grad_norm": 1.610284987892696,
+      "learning_rate": 5.085881061089555e-07,
+      "loss": 0.015,
+      "step": 12313
+    },
+    {
+      "epoch": 2.798636363636364,
+      "grad_norm": 2.9390574347769793,
+      "learning_rate": 5.085004193161065e-07,
+      "loss": 0.0327,
+      "step": 12314
+    },
+    {
+      "epoch": 2.7988636363636363,
+      "grad_norm": 1.7426551135446025,
+      "learning_rate": 5.084127348988871e-07,
+      "loss": 0.0255,
+      "step": 12315
+    },
+    {
+      "epoch": 2.799090909090909,
+      "grad_norm": 1.631923301614982,
+      "learning_rate": 5.083250528590857e-07,
+      "loss": 0.0209,
+      "step": 12316
+    },
+    {
+      "epoch": 2.799318181818182,
+      "grad_norm": 0.5121639877942092,
+      "learning_rate": 5.082373731984901e-07,
+      "loss": 0.0079,
+      "step": 12317
+    },
+    {
+      "epoch": 2.799545454545455,
+      "grad_norm": 1.9746389960363855,
+      "learning_rate": 5.081496959188879e-07,
+      "loss": 0.0188,
+      "step": 12318
+    },
+    {
+      "epoch": 2.7997727272727273,
+      "grad_norm": 1.9245720354565012,
+      "learning_rate": 5.080620210220674e-07,
+      "loss": 0.0235,
+      "step": 12319
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 2.1047670503997673,
+      "learning_rate": 5.079743485098161e-07,
+      "loss": 0.0189,
+      "step": 12320
+    },
+    {
+      "epoch": 2.800227272727273,
+      "grad_norm": 1.7945499916408885,
+      "learning_rate": 5.07886678383922e-07,
+      "loss": 0.0155,
+      "step": 12321
+    },
+    {
+      "epoch": 2.8004545454545453,
+      "grad_norm": 2.0819819907105757,
+      "learning_rate": 5.077990106461732e-07,
+      "loss": 0.0291,
+      "step": 12322
+    },
+    {
+      "epoch": 2.8006818181818183,
+      "grad_norm": 2.1179233521173857,
+      "learning_rate": 5.077113452983569e-07,
+      "loss": 0.0325,
+      "step": 12323
+    },
+    {
+      "epoch": 2.800909090909091,
+      "grad_norm": 1.0170249427499636,
+      "learning_rate": 5.076236823422607e-07,
+      "loss": 0.0119,
+      "step": 12324
+    },
+    {
+      "epoch": 2.8011363636363638,
+      "grad_norm": 1.1927566709895707,
+      "learning_rate": 5.075360217796725e-07,
+      "loss": 0.0155,
+      "step": 12325
+    },
+    {
+      "epoch": 2.8013636363636363,
+      "grad_norm": 1.9186580828982809,
+      "learning_rate": 5.074483636123798e-07,
+      "loss": 0.0149,
+      "step": 12326
+    },
+    {
+      "epoch": 2.8015909090909092,
+      "grad_norm": 1.253435824816612,
+      "learning_rate": 5.073607078421699e-07,
+      "loss": 0.0277,
+      "step": 12327
+    },
+    {
+      "epoch": 2.8018181818181818,
+      "grad_norm": 2.4549028562196495,
+      "learning_rate": 5.072730544708305e-07,
+      "loss": 0.0158,
+      "step": 12328
+    },
+    {
+      "epoch": 2.8020454545454543,
+      "grad_norm": 1.7968130255465955,
+      "learning_rate": 5.071854035001489e-07,
+      "loss": 0.0256,
+      "step": 12329
+    },
+    {
+      "epoch": 2.8022727272727272,
+      "grad_norm": 3.0063092657498363,
+      "learning_rate": 5.070977549319127e-07,
+      "loss": 0.0371,
+      "step": 12330
+    },
+    {
+      "epoch": 2.8025,
+      "grad_norm": 1.1541556660473145,
+      "learning_rate": 5.070101087679087e-07,
+      "loss": 0.0074,
+      "step": 12331
+    },
+    {
+      "epoch": 2.8027272727272727,
+      "grad_norm": 1.8897051632903166,
+      "learning_rate": 5.069224650099244e-07,
+      "loss": 0.0179,
+      "step": 12332
+    },
+    {
+      "epoch": 2.8029545454545453,
+      "grad_norm": 1.341399475716582,
+      "learning_rate": 5.06834823659747e-07,
+      "loss": 0.0198,
+      "step": 12333
+    },
+    {
+      "epoch": 2.803181818181818,
+      "grad_norm": 3.5418213912155267,
+      "learning_rate": 5.067471847191638e-07,
+      "loss": 0.0218,
+      "step": 12334
+    },
+    {
+      "epoch": 2.803409090909091,
+      "grad_norm": 0.8460164748502756,
+      "learning_rate": 5.066595481899619e-07,
+      "loss": 0.0095,
+      "step": 12335
+    },
+    {
+      "epoch": 2.8036363636363637,
+      "grad_norm": 1.1830695899939756,
+      "learning_rate": 5.065719140739281e-07,
+      "loss": 0.011,
+      "step": 12336
+    },
+    {
+      "epoch": 2.8038636363636362,
+      "grad_norm": 0.7343894439529204,
+      "learning_rate": 5.064842823728496e-07,
+      "loss": 0.0066,
+      "step": 12337
+    },
+    {
+      "epoch": 2.804090909090909,
+      "grad_norm": 1.69952913457759,
+      "learning_rate": 5.063966530885134e-07,
+      "loss": 0.0169,
+      "step": 12338
+    },
+    {
+      "epoch": 2.8043181818181817,
+      "grad_norm": 2.7650975326364717,
+      "learning_rate": 5.063090262227066e-07,
+      "loss": 0.026,
+      "step": 12339
+    },
+    {
+      "epoch": 2.8045454545454547,
+      "grad_norm": 1.119461739207456,
+      "learning_rate": 5.062214017772156e-07,
+      "loss": 0.0273,
+      "step": 12340
+    },
+    {
+      "epoch": 2.804772727272727,
+      "grad_norm": 1.9692025316776554,
+      "learning_rate": 5.061337797538273e-07,
+      "loss": 0.0203,
+      "step": 12341
+    },
+    {
+      "epoch": 2.805,
+      "grad_norm": 1.3332938065544717,
+      "learning_rate": 5.060461601543286e-07,
+      "loss": 0.015,
+      "step": 12342
+    },
+    {
+      "epoch": 2.8052272727272727,
+      "grad_norm": 0.6738745146963842,
+      "learning_rate": 5.059585429805064e-07,
+      "loss": 0.0026,
+      "step": 12343
+    },
+    {
+      "epoch": 2.8054545454545456,
+      "grad_norm": 1.3130502372617134,
+      "learning_rate": 5.058709282341471e-07,
+      "loss": 0.0229,
+      "step": 12344
+    },
+    {
+      "epoch": 2.805681818181818,
+      "grad_norm": 2.1870664548677747,
+      "learning_rate": 5.057833159170375e-07,
+      "loss": 0.0309,
+      "step": 12345
+    },
+    {
+      "epoch": 2.8059090909090907,
+      "grad_norm": 2.1514199925071824,
+      "learning_rate": 5.056957060309639e-07,
+      "loss": 0.0292,
+      "step": 12346
+    },
+    {
+      "epoch": 2.8061363636363637,
+      "grad_norm": 1.3482865409758356,
+      "learning_rate": 5.056080985777131e-07,
+      "loss": 0.0101,
+      "step": 12347
+    },
+    {
+      "epoch": 2.8063636363636366,
+      "grad_norm": 3.2170112205283705,
+      "learning_rate": 5.055204935590717e-07,
+      "loss": 0.0378,
+      "step": 12348
+    },
+    {
+      "epoch": 2.806590909090909,
+      "grad_norm": 1.6328843909523254,
+      "learning_rate": 5.054328909768254e-07,
+      "loss": 0.007,
+      "step": 12349
+    },
+    {
+      "epoch": 2.8068181818181817,
+      "grad_norm": 2.5651124755063894,
+      "learning_rate": 5.053452908327613e-07,
+      "loss": 0.0269,
+      "step": 12350
+    },
+    {
+      "epoch": 2.8070454545454546,
+      "grad_norm": 1.367677116567435,
+      "learning_rate": 5.052576931286654e-07,
+      "loss": 0.0199,
+      "step": 12351
+    },
+    {
+      "epoch": 2.807272727272727,
+      "grad_norm": 1.7406236619813005,
+      "learning_rate": 5.051700978663243e-07,
+      "loss": 0.0331,
+      "step": 12352
+    },
+    {
+      "epoch": 2.8075,
+      "grad_norm": 1.6529651847828084,
+      "learning_rate": 5.050825050475237e-07,
+      "loss": 0.0139,
+      "step": 12353
+    },
+    {
+      "epoch": 2.8077272727272726,
+      "grad_norm": 1.8885180586240509,
+      "learning_rate": 5.049949146740501e-07,
+      "loss": 0.0109,
+      "step": 12354
+    },
+    {
+      "epoch": 2.8079545454545456,
+      "grad_norm": 1.863644272941076,
+      "learning_rate": 5.049073267476895e-07,
+      "loss": 0.0256,
+      "step": 12355
+    },
+    {
+      "epoch": 2.808181818181818,
+      "grad_norm": 1.4997229853281162,
+      "learning_rate": 5.04819741270228e-07,
+      "loss": 0.0356,
+      "step": 12356
+    },
+    {
+      "epoch": 2.808409090909091,
+      "grad_norm": 4.665860129695058,
+      "learning_rate": 5.047321582434518e-07,
+      "loss": 0.0262,
+      "step": 12357
+    },
+    {
+      "epoch": 2.8086363636363636,
+      "grad_norm": 2.547046097506418,
+      "learning_rate": 5.046445776691465e-07,
+      "loss": 0.0264,
+      "step": 12358
+    },
+    {
+      "epoch": 2.808863636363636,
+      "grad_norm": 2.056946578225029,
+      "learning_rate": 5.045569995490984e-07,
+      "loss": 0.0487,
+      "step": 12359
+    },
+    {
+      "epoch": 2.809090909090909,
+      "grad_norm": 3.6915232714055715,
+      "learning_rate": 5.04469423885093e-07,
+      "loss": 0.0326,
+      "step": 12360
+    },
+    {
+      "epoch": 2.809318181818182,
+      "grad_norm": 0.9120158713363457,
+      "learning_rate": 5.043818506789168e-07,
+      "loss": 0.0095,
+      "step": 12361
+    },
+    {
+      "epoch": 2.8095454545454546,
+      "grad_norm": 2.5518587910150603,
+      "learning_rate": 5.042942799323547e-07,
+      "loss": 0.0415,
+      "step": 12362
+    },
+    {
+      "epoch": 2.809772727272727,
+      "grad_norm": 2.7015431975262305,
+      "learning_rate": 5.042067116471929e-07,
+      "loss": 0.0383,
+      "step": 12363
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.9593773330336228,
+      "learning_rate": 5.041191458252168e-07,
+      "loss": 0.0235,
+      "step": 12364
+    },
+    {
+      "epoch": 2.8102272727272726,
+      "grad_norm": 1.4958631141489345,
+      "learning_rate": 5.040315824682124e-07,
+      "loss": 0.0302,
+      "step": 12365
+    },
+    {
+      "epoch": 2.8104545454545455,
+      "grad_norm": 2.1783605714505154,
+      "learning_rate": 5.039440215779651e-07,
+      "loss": 0.0278,
+      "step": 12366
+    },
+    {
+      "epoch": 2.810681818181818,
+      "grad_norm": 2.3487195398751206,
+      "learning_rate": 5.038564631562603e-07,
+      "loss": 0.0291,
+      "step": 12367
+    },
+    {
+      "epoch": 2.810909090909091,
+      "grad_norm": 2.074590303503796,
+      "learning_rate": 5.037689072048835e-07,
+      "loss": 0.0264,
+      "step": 12368
+    },
+    {
+      "epoch": 2.8111363636363635,
+      "grad_norm": 1.1389046437957224,
+      "learning_rate": 5.036813537256203e-07,
+      "loss": 0.0157,
+      "step": 12369
+    },
+    {
+      "epoch": 2.8113636363636365,
+      "grad_norm": 1.6922571710744967,
+      "learning_rate": 5.035938027202561e-07,
+      "loss": 0.0162,
+      "step": 12370
+    },
+    {
+      "epoch": 2.811590909090909,
+      "grad_norm": 1.0899095445671283,
+      "learning_rate": 5.035062541905759e-07,
+      "loss": 0.0153,
+      "step": 12371
+    },
+    {
+      "epoch": 2.8118181818181816,
+      "grad_norm": 3.055959987077733,
+      "learning_rate": 5.034187081383652e-07,
+      "loss": 0.0159,
+      "step": 12372
+    },
+    {
+      "epoch": 2.8120454545454545,
+      "grad_norm": 2.5174606629525913,
+      "learning_rate": 5.033311645654089e-07,
+      "loss": 0.0229,
+      "step": 12373
+    },
+    {
+      "epoch": 2.8122727272727275,
+      "grad_norm": 3.045506986499561,
+      "learning_rate": 5.032436234734928e-07,
+      "loss": 0.0386,
+      "step": 12374
+    },
+    {
+      "epoch": 2.8125,
+      "grad_norm": 3.645022277286079,
+      "learning_rate": 5.031560848644013e-07,
+      "loss": 0.0341,
+      "step": 12375
+    },
+    {
+      "epoch": 2.8127272727272725,
+      "grad_norm": 2.265842061017137,
+      "learning_rate": 5.0306854873992e-07,
+      "loss": 0.0207,
+      "step": 12376
+    },
+    {
+      "epoch": 2.8129545454545455,
+      "grad_norm": 3.028987876792556,
+      "learning_rate": 5.029810151018335e-07,
+      "loss": 0.0447,
+      "step": 12377
+    },
+    {
+      "epoch": 2.8131818181818184,
+      "grad_norm": 0.8550436367103501,
+      "learning_rate": 5.028934839519271e-07,
+      "loss": 0.018,
+      "step": 12378
+    },
+    {
+      "epoch": 2.813409090909091,
+      "grad_norm": 2.7064618797635758,
+      "learning_rate": 5.028059552919857e-07,
+      "loss": 0.0308,
+      "step": 12379
+    },
+    {
+      "epoch": 2.8136363636363635,
+      "grad_norm": 0.737362066491427,
+      "learning_rate": 5.02718429123794e-07,
+      "loss": 0.0136,
+      "step": 12380
+    },
+    {
+      "epoch": 2.8138636363636365,
+      "grad_norm": 2.3034775438589232,
+      "learning_rate": 5.026309054491367e-07,
+      "loss": 0.0231,
+      "step": 12381
+    },
+    {
+      "epoch": 2.814090909090909,
+      "grad_norm": 1.4119555973402367,
+      "learning_rate": 5.025433842697988e-07,
+      "loss": 0.0123,
+      "step": 12382
+    },
+    {
+      "epoch": 2.814318181818182,
+      "grad_norm": 2.571910819981167,
+      "learning_rate": 5.024558655875652e-07,
+      "loss": 0.0578,
+      "step": 12383
+    },
+    {
+      "epoch": 2.8145454545454545,
+      "grad_norm": 1.6048738516898846,
+      "learning_rate": 5.0236834940422e-07,
+      "loss": 0.0203,
+      "step": 12384
+    },
+    {
+      "epoch": 2.8147727272727274,
+      "grad_norm": 3.1011581558233656,
+      "learning_rate": 5.022808357215482e-07,
+      "loss": 0.0313,
+      "step": 12385
+    },
+    {
+      "epoch": 2.815,
+      "grad_norm": 1.1597159004809134,
+      "learning_rate": 5.021933245413342e-07,
+      "loss": 0.0105,
+      "step": 12386
+    },
+    {
+      "epoch": 2.815227272727273,
+      "grad_norm": 1.5241024512126058,
+      "learning_rate": 5.021058158653626e-07,
+      "loss": 0.0275,
+      "step": 12387
+    },
+    {
+      "epoch": 2.8154545454545454,
+      "grad_norm": 2.987084848055443,
+      "learning_rate": 5.020183096954177e-07,
+      "loss": 0.0371,
+      "step": 12388
+    },
+    {
+      "epoch": 2.815681818181818,
+      "grad_norm": 0.8055313210167637,
+      "learning_rate": 5.019308060332841e-07,
+      "loss": 0.0077,
+      "step": 12389
+    },
+    {
+      "epoch": 2.815909090909091,
+      "grad_norm": 1.4296893505549695,
+      "learning_rate": 5.018433048807461e-07,
+      "loss": 0.0152,
+      "step": 12390
+    },
+    {
+      "epoch": 2.816136363636364,
+      "grad_norm": 2.2784817082765136,
+      "learning_rate": 5.017558062395879e-07,
+      "loss": 0.0365,
+      "step": 12391
+    },
+    {
+      "epoch": 2.8163636363636364,
+      "grad_norm": 2.1488103763739157,
+      "learning_rate": 5.016683101115941e-07,
+      "loss": 0.0155,
+      "step": 12392
+    },
+    {
+      "epoch": 2.816590909090909,
+      "grad_norm": 2.293682693395599,
+      "learning_rate": 5.015808164985483e-07,
+      "loss": 0.0308,
+      "step": 12393
+    },
+    {
+      "epoch": 2.816818181818182,
+      "grad_norm": 2.022177100395668,
+      "learning_rate": 5.014933254022351e-07,
+      "loss": 0.0425,
+      "step": 12394
+    },
+    {
+      "epoch": 2.8170454545454544,
+      "grad_norm": 2.201827612485937,
+      "learning_rate": 5.014058368244384e-07,
+      "loss": 0.0336,
+      "step": 12395
+    },
+    {
+      "epoch": 2.8172727272727274,
+      "grad_norm": 0.8134972495109075,
+      "learning_rate": 5.013183507669424e-07,
+      "loss": 0.0105,
+      "step": 12396
+    },
+    {
+      "epoch": 2.8175,
+      "grad_norm": 1.4422043625533303,
+      "learning_rate": 5.01230867231531e-07,
+      "loss": 0.0105,
+      "step": 12397
+    },
+    {
+      "epoch": 2.817727272727273,
+      "grad_norm": 1.4311908406867144,
+      "learning_rate": 5.01143386219988e-07,
+      "loss": 0.0124,
+      "step": 12398
+    },
+    {
+      "epoch": 2.8179545454545454,
+      "grad_norm": 2.0738662821923852,
+      "learning_rate": 5.010559077340975e-07,
+      "loss": 0.0348,
+      "step": 12399
+    },
+    {
+      "epoch": 2.8181818181818183,
+      "grad_norm": 1.856508584995078,
+      "learning_rate": 5.009684317756432e-07,
+      "loss": 0.0177,
+      "step": 12400
+    },
+    {
+      "epoch": 2.818409090909091,
+      "grad_norm": 0.9316456494845348,
+      "learning_rate": 5.008809583464093e-07,
+      "loss": 0.0219,
+      "step": 12401
+    },
+    {
+      "epoch": 2.8186363636363634,
+      "grad_norm": 0.5577726084175885,
+      "learning_rate": 5.007934874481789e-07,
+      "loss": 0.0081,
+      "step": 12402
+    },
+    {
+      "epoch": 2.8188636363636363,
+      "grad_norm": 1.6603923295662535,
+      "learning_rate": 5.00706019082736e-07,
+      "loss": 0.017,
+      "step": 12403
+    },
+    {
+      "epoch": 2.8190909090909093,
+      "grad_norm": 1.1423680602573378,
+      "learning_rate": 5.006185532518641e-07,
+      "loss": 0.0086,
+      "step": 12404
+    },
+    {
+      "epoch": 2.819318181818182,
+      "grad_norm": 1.7604188748965792,
+      "learning_rate": 5.005310899573473e-07,
+      "loss": 0.0137,
+      "step": 12405
+    },
+    {
+      "epoch": 2.8195454545454544,
+      "grad_norm": 12.884985101111651,
+      "learning_rate": 5.004436292009684e-07,
+      "loss": 0.0389,
+      "step": 12406
+    },
+    {
+      "epoch": 2.8197727272727273,
+      "grad_norm": 1.6443081070015502,
+      "learning_rate": 5.003561709845112e-07,
+      "loss": 0.0176,
+      "step": 12407
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 2.312086544753676,
+      "learning_rate": 5.002687153097591e-07,
+      "loss": 0.0244,
+      "step": 12408
+    },
+    {
+      "epoch": 2.820227272727273,
+      "grad_norm": 1.7580351357997932,
+      "learning_rate": 5.001812621784957e-07,
+      "loss": 0.0167,
+      "step": 12409
+    },
+    {
+      "epoch": 2.8204545454545453,
+      "grad_norm": 3.0777613160331208,
+      "learning_rate": 5.00093811592504e-07,
+      "loss": 0.0522,
+      "step": 12410
+    },
+    {
+      "epoch": 2.8206818181818183,
+      "grad_norm": 1.7561644182265912,
+      "learning_rate": 5.000063635535673e-07,
+      "loss": 0.0199,
+      "step": 12411
+    },
+    {
+      "epoch": 2.820909090909091,
+      "grad_norm": 1.0210405909584108,
+      "learning_rate": 4.999189180634691e-07,
+      "loss": 0.0134,
+      "step": 12412
+    },
+    {
+      "epoch": 2.8211363636363638,
+      "grad_norm": 1.6758412845913644,
+      "learning_rate": 4.998314751239922e-07,
+      "loss": 0.033,
+      "step": 12413
+    },
+    {
+      "epoch": 2.8213636363636363,
+      "grad_norm": 2.5140664587116612,
+      "learning_rate": 4.997440347369202e-07,
+      "loss": 0.0373,
+      "step": 12414
+    },
+    {
+      "epoch": 2.821590909090909,
+      "grad_norm": 1.5257632731059998,
+      "learning_rate": 4.996565969040357e-07,
+      "loss": 0.0168,
+      "step": 12415
+    },
+    {
+      "epoch": 2.821818181818182,
+      "grad_norm": 1.604101580738139,
+      "learning_rate": 4.995691616271217e-07,
+      "loss": 0.0148,
+      "step": 12416
+    },
+    {
+      "epoch": 2.8220454545454547,
+      "grad_norm": 3.3948898644131216,
+      "learning_rate": 4.994817289079613e-07,
+      "loss": 0.1044,
+      "step": 12417
+    },
+    {
+      "epoch": 2.8222727272727273,
+      "grad_norm": 0.6526562702959491,
+      "learning_rate": 4.993942987483377e-07,
+      "loss": 0.0057,
+      "step": 12418
+    },
+    {
+      "epoch": 2.8225,
+      "grad_norm": 1.7424825324599158,
+      "learning_rate": 4.993068711500333e-07,
+      "loss": 0.0176,
+      "step": 12419
+    },
+    {
+      "epoch": 2.8227272727272728,
+      "grad_norm": 1.1799957313688363,
+      "learning_rate": 4.99219446114831e-07,
+      "loss": 0.0176,
+      "step": 12420
+    },
+    {
+      "epoch": 2.8229545454545457,
+      "grad_norm": 2.6714533174168955,
+      "learning_rate": 4.991320236445138e-07,
+      "loss": 0.0432,
+      "step": 12421
+    },
+    {
+      "epoch": 2.8231818181818182,
+      "grad_norm": 1.9023781975813556,
+      "learning_rate": 4.990446037408641e-07,
+      "loss": 0.0139,
+      "step": 12422
+    },
+    {
+      "epoch": 2.8234090909090908,
+      "grad_norm": 3.155483962582216,
+      "learning_rate": 4.98957186405665e-07,
+      "loss": 0.0384,
+      "step": 12423
+    },
+    {
+      "epoch": 2.8236363636363637,
+      "grad_norm": 2.8786375605516046,
+      "learning_rate": 4.988697716406984e-07,
+      "loss": 0.0297,
+      "step": 12424
+    },
+    {
+      "epoch": 2.8238636363636362,
+      "grad_norm": 1.5032755339879877,
+      "learning_rate": 4.987823594477473e-07,
+      "loss": 0.0154,
+      "step": 12425
+    },
+    {
+      "epoch": 2.824090909090909,
+      "grad_norm": 1.6048069199533879,
+      "learning_rate": 4.986949498285941e-07,
+      "loss": 0.0138,
+      "step": 12426
+    },
+    {
+      "epoch": 2.8243181818181817,
+      "grad_norm": 2.2092573955906407,
+      "learning_rate": 4.986075427850213e-07,
+      "loss": 0.0209,
+      "step": 12427
+    },
+    {
+      "epoch": 2.8245454545454547,
+      "grad_norm": 2.378062233630698,
+      "learning_rate": 4.98520138318811e-07,
+      "loss": 0.0359,
+      "step": 12428
+    },
+    {
+      "epoch": 2.824772727272727,
+      "grad_norm": 2.9828785102671245,
+      "learning_rate": 4.984327364317458e-07,
+      "loss": 0.1089,
+      "step": 12429
+    },
+    {
+      "epoch": 2.825,
+      "grad_norm": 1.4540048722141283,
+      "learning_rate": 4.98345337125608e-07,
+      "loss": 0.0188,
+      "step": 12430
+    },
+    {
+      "epoch": 2.8252272727272727,
+      "grad_norm": 0.9254622914100107,
+      "learning_rate": 4.982579404021799e-07,
+      "loss": 0.0075,
+      "step": 12431
+    },
+    {
+      "epoch": 2.825454545454545,
+      "grad_norm": 0.5278817199390405,
+      "learning_rate": 4.981705462632433e-07,
+      "loss": 0.0108,
+      "step": 12432
+    },
+    {
+      "epoch": 2.825681818181818,
+      "grad_norm": 1.9072709452959762,
+      "learning_rate": 4.980831547105803e-07,
+      "loss": 0.0222,
+      "step": 12433
+    },
+    {
+      "epoch": 2.825909090909091,
+      "grad_norm": 2.244497521126506,
+      "learning_rate": 4.979957657459735e-07,
+      "loss": 0.0261,
+      "step": 12434
+    },
+    {
+      "epoch": 2.8261363636363637,
+      "grad_norm": 2.140112556279605,
+      "learning_rate": 4.979083793712045e-07,
+      "loss": 0.0354,
+      "step": 12435
+    },
+    {
+      "epoch": 2.826363636363636,
+      "grad_norm": 2.0492820841451698,
+      "learning_rate": 4.978209955880556e-07,
+      "loss": 0.0298,
+      "step": 12436
+    },
+    {
+      "epoch": 2.826590909090909,
+      "grad_norm": 0.6094400074072962,
+      "learning_rate": 4.977336143983083e-07,
+      "loss": 0.0064,
+      "step": 12437
+    },
+    {
+      "epoch": 2.8268181818181817,
+      "grad_norm": 1.3337093075053033,
+      "learning_rate": 4.976462358037446e-07,
+      "loss": 0.0084,
+      "step": 12438
+    },
+    {
+      "epoch": 2.8270454545454546,
+      "grad_norm": 2.6362302914728004,
+      "learning_rate": 4.975588598061463e-07,
+      "loss": 0.036,
+      "step": 12439
+    },
+    {
+      "epoch": 2.827272727272727,
+      "grad_norm": 1.5195512428202278,
+      "learning_rate": 4.974714864072954e-07,
+      "loss": 0.0251,
+      "step": 12440
+    },
+    {
+      "epoch": 2.8275,
+      "grad_norm": 3.0146319873942033,
+      "learning_rate": 4.97384115608973e-07,
+      "loss": 0.0425,
+      "step": 12441
+    },
+    {
+      "epoch": 2.8277272727272726,
+      "grad_norm": 1.1670417840748488,
+      "learning_rate": 4.972967474129614e-07,
+      "loss": 0.0149,
+      "step": 12442
+    },
+    {
+      "epoch": 2.8279545454545456,
+      "grad_norm": 1.7177574743340627,
+      "learning_rate": 4.972093818210419e-07,
+      "loss": 0.0246,
+      "step": 12443
+    },
+    {
+      "epoch": 2.828181818181818,
+      "grad_norm": 1.6223530912513606,
+      "learning_rate": 4.971220188349959e-07,
+      "loss": 0.0239,
+      "step": 12444
+    },
+    {
+      "epoch": 2.8284090909090907,
+      "grad_norm": 3.3763014249962766,
+      "learning_rate": 4.970346584566053e-07,
+      "loss": 0.0411,
+      "step": 12445
+    },
+    {
+      "epoch": 2.8286363636363636,
+      "grad_norm": 2.3713755212913403,
+      "learning_rate": 4.96947300687651e-07,
+      "loss": 0.0277,
+      "step": 12446
+    },
+    {
+      "epoch": 2.8288636363636366,
+      "grad_norm": 1.7271060490854289,
+      "learning_rate": 4.968599455299147e-07,
+      "loss": 0.0248,
+      "step": 12447
+    },
+    {
+      "epoch": 2.829090909090909,
+      "grad_norm": 1.9507539524814745,
+      "learning_rate": 4.967725929851777e-07,
+      "loss": 0.0432,
+      "step": 12448
+    },
+    {
+      "epoch": 2.8293181818181816,
+      "grad_norm": 2.7022887899918184,
+      "learning_rate": 4.966852430552213e-07,
+      "loss": 0.0445,
+      "step": 12449
+    },
+    {
+      "epoch": 2.8295454545454546,
+      "grad_norm": 0.6890684802365633,
+      "learning_rate": 4.965978957418265e-07,
+      "loss": 0.0062,
+      "step": 12450
+    },
+    {
+      "epoch": 2.829772727272727,
+      "grad_norm": 1.3273356952338362,
+      "learning_rate": 4.965105510467747e-07,
+      "loss": 0.0209,
+      "step": 12451
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.9706965829570215,
+      "learning_rate": 4.964232089718469e-07,
+      "loss": 0.0094,
+      "step": 12452
+    },
+    {
+      "epoch": 2.8302272727272726,
+      "grad_norm": 2.8854650220353952,
+      "learning_rate": 4.963358695188244e-07,
+      "loss": 0.0263,
+      "step": 12453
+    },
+    {
+      "epoch": 2.8304545454545456,
+      "grad_norm": 2.1336089361881423,
+      "learning_rate": 4.962485326894878e-07,
+      "loss": 0.0268,
+      "step": 12454
+    },
+    {
+      "epoch": 2.830681818181818,
+      "grad_norm": 1.9569625483535447,
+      "learning_rate": 4.961611984856181e-07,
+      "loss": 0.0164,
+      "step": 12455
+    },
+    {
+      "epoch": 2.830909090909091,
+      "grad_norm": 1.2677988266907203,
+      "learning_rate": 4.960738669089964e-07,
+      "loss": 0.0146,
+      "step": 12456
+    },
+    {
+      "epoch": 2.8311363636363636,
+      "grad_norm": 1.2668730593878104,
+      "learning_rate": 4.959865379614035e-07,
+      "loss": 0.0215,
+      "step": 12457
+    },
+    {
+      "epoch": 2.8313636363636365,
+      "grad_norm": 1.511897225213577,
+      "learning_rate": 4.958992116446203e-07,
+      "loss": 0.0264,
+      "step": 12458
+    },
+    {
+      "epoch": 2.831590909090909,
+      "grad_norm": 1.1943060350486507,
+      "learning_rate": 4.958118879604274e-07,
+      "loss": 0.0174,
+      "step": 12459
+    },
+    {
+      "epoch": 2.831818181818182,
+      "grad_norm": 2.5359254221652985,
+      "learning_rate": 4.957245669106052e-07,
+      "loss": 0.0781,
+      "step": 12460
+    },
+    {
+      "epoch": 2.8320454545454545,
+      "grad_norm": 1.4943600995162085,
+      "learning_rate": 4.956372484969348e-07,
+      "loss": 0.0141,
+      "step": 12461
+    },
+    {
+      "epoch": 2.832272727272727,
+      "grad_norm": 1.8291769853740725,
+      "learning_rate": 4.955499327211967e-07,
+      "loss": 0.0198,
+      "step": 12462
+    },
+    {
+      "epoch": 2.8325,
+      "grad_norm": 2.8701494531204275,
+      "learning_rate": 4.954626195851712e-07,
+      "loss": 0.0365,
+      "step": 12463
+    },
+    {
+      "epoch": 2.832727272727273,
+      "grad_norm": 1.4054556354503325,
+      "learning_rate": 4.953753090906388e-07,
+      "loss": 0.011,
+      "step": 12464
+    },
+    {
+      "epoch": 2.8329545454545455,
+      "grad_norm": 1.189604759738078,
+      "learning_rate": 4.952880012393801e-07,
+      "loss": 0.0364,
+      "step": 12465
+    },
+    {
+      "epoch": 2.833181818181818,
+      "grad_norm": 1.226600310047444,
+      "learning_rate": 4.952006960331752e-07,
+      "loss": 0.0152,
+      "step": 12466
+    },
+    {
+      "epoch": 2.833409090909091,
+      "grad_norm": 1.6349481525542133,
+      "learning_rate": 4.951133934738049e-07,
+      "loss": 0.0181,
+      "step": 12467
+    },
+    {
+      "epoch": 2.8336363636363635,
+      "grad_norm": 1.391651839310393,
+      "learning_rate": 4.95026093563049e-07,
+      "loss": 0.0174,
+      "step": 12468
+    },
+    {
+      "epoch": 2.8338636363636365,
+      "grad_norm": 1.7217099221860221,
+      "learning_rate": 4.949387963026876e-07,
+      "loss": 0.0193,
+      "step": 12469
+    },
+    {
+      "epoch": 2.834090909090909,
+      "grad_norm": 1.9024241116538916,
+      "learning_rate": 4.94851501694501e-07,
+      "loss": 0.0367,
+      "step": 12470
+    },
+    {
+      "epoch": 2.834318181818182,
+      "grad_norm": 3.9538592391566905,
+      "learning_rate": 4.947642097402696e-07,
+      "loss": 0.0344,
+      "step": 12471
+    },
+    {
+      "epoch": 2.8345454545454545,
+      "grad_norm": 2.146795518681181,
+      "learning_rate": 4.94676920441773e-07,
+      "loss": 0.0336,
+      "step": 12472
+    },
+    {
+      "epoch": 2.8347727272727274,
+      "grad_norm": 2.5771708094592087,
+      "learning_rate": 4.945896338007914e-07,
+      "loss": 0.0271,
+      "step": 12473
+    },
+    {
+      "epoch": 2.835,
+      "grad_norm": 1.40178595009767,
+      "learning_rate": 4.945023498191046e-07,
+      "loss": 0.0253,
+      "step": 12474
+    },
+    {
+      "epoch": 2.8352272727272725,
+      "grad_norm": 1.988159359768931,
+      "learning_rate": 4.944150684984929e-07,
+      "loss": 0.0187,
+      "step": 12475
+    },
+    {
+      "epoch": 2.8354545454545454,
+      "grad_norm": 0.8171345538397876,
+      "learning_rate": 4.943277898407354e-07,
+      "loss": 0.0084,
+      "step": 12476
+    },
+    {
+      "epoch": 2.8356818181818184,
+      "grad_norm": 0.9808398661260559,
+      "learning_rate": 4.942405138476124e-07,
+      "loss": 0.007,
+      "step": 12477
+    },
+    {
+      "epoch": 2.835909090909091,
+      "grad_norm": 1.9039667357307766,
+      "learning_rate": 4.941532405209033e-07,
+      "loss": 0.016,
+      "step": 12478
+    },
+    {
+      "epoch": 2.8361363636363635,
+      "grad_norm": 1.6898652746950817,
+      "learning_rate": 4.940659698623879e-07,
+      "loss": 0.0211,
+      "step": 12479
+    },
+    {
+      "epoch": 2.8363636363636364,
+      "grad_norm": 1.5653607859534275,
+      "learning_rate": 4.93978701873846e-07,
+      "loss": 0.019,
+      "step": 12480
+    },
+    {
+      "epoch": 2.836590909090909,
+      "grad_norm": 1.647706814889891,
+      "learning_rate": 4.938914365570568e-07,
+      "loss": 0.0274,
+      "step": 12481
+    },
+    {
+      "epoch": 2.836818181818182,
+      "grad_norm": 3.127982138173281,
+      "learning_rate": 4.938041739137999e-07,
+      "loss": 0.0711,
+      "step": 12482
+    },
+    {
+      "epoch": 2.8370454545454544,
+      "grad_norm": 1.3579723000063249,
+      "learning_rate": 4.937169139458549e-07,
+      "loss": 0.0181,
+      "step": 12483
+    },
+    {
+      "epoch": 2.8372727272727274,
+      "grad_norm": 2.322113022544986,
+      "learning_rate": 4.936296566550012e-07,
+      "loss": 0.0271,
+      "step": 12484
+    },
+    {
+      "epoch": 2.8375,
+      "grad_norm": 2.522943486300571,
+      "learning_rate": 4.935424020430178e-07,
+      "loss": 0.029,
+      "step": 12485
+    },
+    {
+      "epoch": 2.837727272727273,
+      "grad_norm": 2.1255126042367665,
+      "learning_rate": 4.934551501116841e-07,
+      "loss": 0.0312,
+      "step": 12486
+    },
+    {
+      "epoch": 2.8379545454545454,
+      "grad_norm": 2.68471478910752,
+      "learning_rate": 4.933679008627794e-07,
+      "loss": 0.0362,
+      "step": 12487
+    },
+    {
+      "epoch": 2.838181818181818,
+      "grad_norm": 1.6748710922540553,
+      "learning_rate": 4.932806542980828e-07,
+      "loss": 0.0188,
+      "step": 12488
+    },
+    {
+      "epoch": 2.838409090909091,
+      "grad_norm": 2.676002297187591,
+      "learning_rate": 4.931934104193737e-07,
+      "loss": 0.0288,
+      "step": 12489
+    },
+    {
+      "epoch": 2.838636363636364,
+      "grad_norm": 1.7197966079126092,
+      "learning_rate": 4.931061692284307e-07,
+      "loss": 0.0116,
+      "step": 12490
+    },
+    {
+      "epoch": 2.8388636363636364,
+      "grad_norm": 2.096266331304726,
+      "learning_rate": 4.930189307270331e-07,
+      "loss": 0.0196,
+      "step": 12491
+    },
+    {
+      "epoch": 2.839090909090909,
+      "grad_norm": 1.4282687999195447,
+      "learning_rate": 4.929316949169596e-07,
+      "loss": 0.0135,
+      "step": 12492
+    },
+    {
+      "epoch": 2.839318181818182,
+      "grad_norm": 2.2944902284355457,
+      "learning_rate": 4.928444617999895e-07,
+      "loss": 0.0173,
+      "step": 12493
+    },
+    {
+      "epoch": 2.839545454545455,
+      "grad_norm": 2.7227972531347726,
+      "learning_rate": 4.927572313779012e-07,
+      "loss": 0.0219,
+      "step": 12494
+    },
+    {
+      "epoch": 2.8397727272727273,
+      "grad_norm": 1.7616170990481306,
+      "learning_rate": 4.926700036524737e-07,
+      "loss": 0.0224,
+      "step": 12495
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 2.160719170654452,
+      "learning_rate": 4.925827786254857e-07,
+      "loss": 0.0339,
+      "step": 12496
+    },
+    {
+      "epoch": 2.840227272727273,
+      "grad_norm": 2.5714277041520734,
+      "learning_rate": 4.92495556298716e-07,
+      "loss": 0.0293,
+      "step": 12497
+    },
+    {
+      "epoch": 2.8404545454545453,
+      "grad_norm": 2.5077463598032215,
+      "learning_rate": 4.924083366739429e-07,
+      "loss": 0.0241,
+      "step": 12498
+    },
+    {
+      "epoch": 2.8406818181818183,
+      "grad_norm": 3.104606430904353,
+      "learning_rate": 4.923211197529451e-07,
+      "loss": 0.0569,
+      "step": 12499
+    },
+    {
+      "epoch": 2.840909090909091,
+      "grad_norm": 2.3177086445863564,
+      "learning_rate": 4.922339055375013e-07,
+      "loss": 0.0347,
+      "step": 12500
+    },
+    {
+      "epoch": 2.841136363636364,
+      "grad_norm": 1.1581743672734357,
+      "learning_rate": 4.921466940293896e-07,
+      "loss": 0.0102,
+      "step": 12501
+    },
+    {
+      "epoch": 2.8413636363636363,
+      "grad_norm": 1.8447703239243645,
+      "learning_rate": 4.920594852303888e-07,
+      "loss": 0.0215,
+      "step": 12502
+    },
+    {
+      "epoch": 2.8415909090909093,
+      "grad_norm": 1.3471891476034283,
+      "learning_rate": 4.919722791422769e-07,
+      "loss": 0.0171,
+      "step": 12503
+    },
+    {
+      "epoch": 2.841818181818182,
+      "grad_norm": 2.563674095368351,
+      "learning_rate": 4.918850757668322e-07,
+      "loss": 0.026,
+      "step": 12504
+    },
+    {
+      "epoch": 2.8420454545454543,
+      "grad_norm": 1.624315807678263,
+      "learning_rate": 4.917978751058331e-07,
+      "loss": 0.0257,
+      "step": 12505
+    },
+    {
+      "epoch": 2.8422727272727273,
+      "grad_norm": 2.4176513784931735,
+      "learning_rate": 4.91710677161058e-07,
+      "loss": 0.0428,
+      "step": 12506
+    },
+    {
+      "epoch": 2.8425000000000002,
+      "grad_norm": 1.5312743658876211,
+      "learning_rate": 4.916234819342845e-07,
+      "loss": 0.0185,
+      "step": 12507
+    },
+    {
+      "epoch": 2.8427272727272728,
+      "grad_norm": 1.5820327950596458,
+      "learning_rate": 4.915362894272909e-07,
+      "loss": 0.0203,
+      "step": 12508
+    },
+    {
+      "epoch": 2.8429545454545453,
+      "grad_norm": 3.17091599873079,
+      "learning_rate": 4.914490996418551e-07,
+      "loss": 0.0567,
+      "step": 12509
+    },
+    {
+      "epoch": 2.8431818181818183,
+      "grad_norm": 1.6481806576801643,
+      "learning_rate": 4.913619125797553e-07,
+      "loss": 0.0266,
+      "step": 12510
+    },
+    {
+      "epoch": 2.8434090909090908,
+      "grad_norm": 1.4749274272895798,
+      "learning_rate": 4.912747282427691e-07,
+      "loss": 0.0222,
+      "step": 12511
+    },
+    {
+      "epoch": 2.8436363636363637,
+      "grad_norm": 2.4015324522557795,
+      "learning_rate": 4.911875466326745e-07,
+      "loss": 0.0173,
+      "step": 12512
+    },
+    {
+      "epoch": 2.8438636363636363,
+      "grad_norm": 1.7653047243061228,
+      "learning_rate": 4.911003677512494e-07,
+      "loss": 0.0278,
+      "step": 12513
+    },
+    {
+      "epoch": 2.8440909090909092,
+      "grad_norm": 1.7631330575944275,
+      "learning_rate": 4.910131916002712e-07,
+      "loss": 0.0184,
+      "step": 12514
+    },
+    {
+      "epoch": 2.8443181818181817,
+      "grad_norm": 1.5302790438859168,
+      "learning_rate": 4.909260181815182e-07,
+      "loss": 0.0117,
+      "step": 12515
+    },
+    {
+      "epoch": 2.8445454545454547,
+      "grad_norm": 4.047616446208139,
+      "learning_rate": 4.908388474967673e-07,
+      "loss": 0.0354,
+      "step": 12516
+    },
+    {
+      "epoch": 2.8447727272727272,
+      "grad_norm": 2.0079106086055023,
+      "learning_rate": 4.907516795477962e-07,
+      "loss": 0.018,
+      "step": 12517
+    },
+    {
+      "epoch": 2.8449999999999998,
+      "grad_norm": 1.9173383959418613,
+      "learning_rate": 4.906645143363827e-07,
+      "loss": 0.021,
+      "step": 12518
+    },
+    {
+      "epoch": 2.8452272727272727,
+      "grad_norm": 2.3969333270410216,
+      "learning_rate": 4.905773518643044e-07,
+      "loss": 0.023,
+      "step": 12519
+    },
+    {
+      "epoch": 2.8454545454545457,
+      "grad_norm": 1.7946702551046891,
+      "learning_rate": 4.90490192133338e-07,
+      "loss": 0.0235,
+      "step": 12520
+    },
+    {
+      "epoch": 2.845681818181818,
+      "grad_norm": 0.9718764388661658,
+      "learning_rate": 4.904030351452614e-07,
+      "loss": 0.012,
+      "step": 12521
+    },
+    {
+      "epoch": 2.8459090909090907,
+      "grad_norm": 2.361270476437923,
+      "learning_rate": 4.903158809018518e-07,
+      "loss": 0.0317,
+      "step": 12522
+    },
+    {
+      "epoch": 2.8461363636363637,
+      "grad_norm": 3.7848204175755176,
+      "learning_rate": 4.902287294048862e-07,
+      "loss": 0.0176,
+      "step": 12523
+    },
+    {
+      "epoch": 2.846363636363636,
+      "grad_norm": 1.7857003021118227,
+      "learning_rate": 4.90141580656142e-07,
+      "loss": 0.0304,
+      "step": 12524
+    },
+    {
+      "epoch": 2.846590909090909,
+      "grad_norm": 1.3308103759054686,
+      "learning_rate": 4.900544346573963e-07,
+      "loss": 0.0097,
+      "step": 12525
+    },
+    {
+      "epoch": 2.8468181818181817,
+      "grad_norm": 2.0527591757978363,
+      "learning_rate": 4.89967291410426e-07,
+      "loss": 0.0215,
+      "step": 12526
+    },
+    {
+      "epoch": 2.8470454545454547,
+      "grad_norm": 1.1294764704084916,
+      "learning_rate": 4.898801509170082e-07,
+      "loss": 0.0113,
+      "step": 12527
+    },
+    {
+      "epoch": 2.847272727272727,
+      "grad_norm": 1.7164193970884516,
+      "learning_rate": 4.897930131789202e-07,
+      "loss": 0.0259,
+      "step": 12528
+    },
+    {
+      "epoch": 2.8475,
+      "grad_norm": 1.7168390074102953,
+      "learning_rate": 4.897058781979381e-07,
+      "loss": 0.0161,
+      "step": 12529
+    },
+    {
+      "epoch": 2.8477272727272727,
+      "grad_norm": 1.4410805592038254,
+      "learning_rate": 4.896187459758394e-07,
+      "loss": 0.0104,
+      "step": 12530
+    },
+    {
+      "epoch": 2.847954545454545,
+      "grad_norm": 2.4205114529455205,
+      "learning_rate": 4.895316165144005e-07,
+      "loss": 0.0386,
+      "step": 12531
+    },
+    {
+      "epoch": 2.848181818181818,
+      "grad_norm": 1.1626962919451884,
+      "learning_rate": 4.894444898153985e-07,
+      "loss": 0.0173,
+      "step": 12532
+    },
+    {
+      "epoch": 2.848409090909091,
+      "grad_norm": 2.0248293395517325,
+      "learning_rate": 4.893573658806097e-07,
+      "loss": 0.0381,
+      "step": 12533
+    },
+    {
+      "epoch": 2.8486363636363636,
+      "grad_norm": 1.1693684033393965,
+      "learning_rate": 4.892702447118109e-07,
+      "loss": 0.016,
+      "step": 12534
+    },
+    {
+      "epoch": 2.848863636363636,
+      "grad_norm": 3.7141481178083997,
+      "learning_rate": 4.891831263107784e-07,
+      "loss": 0.0316,
+      "step": 12535
+    },
+    {
+      "epoch": 2.849090909090909,
+      "grad_norm": 2.2816164188927117,
+      "learning_rate": 4.890960106792891e-07,
+      "loss": 0.0154,
+      "step": 12536
+    },
+    {
+      "epoch": 2.849318181818182,
+      "grad_norm": 2.2398564017620406,
+      "learning_rate": 4.890088978191195e-07,
+      "loss": 0.0234,
+      "step": 12537
+    },
+    {
+      "epoch": 2.8495454545454546,
+      "grad_norm": 4.1423871179469955,
+      "learning_rate": 4.889217877320453e-07,
+      "loss": 0.043,
+      "step": 12538
+    },
+    {
+      "epoch": 2.849772727272727,
+      "grad_norm": 1.6446261778997477,
+      "learning_rate": 4.888346804198433e-07,
+      "loss": 0.0226,
+      "step": 12539
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.2737796383428128,
+      "learning_rate": 4.887475758842898e-07,
+      "loss": 0.009,
+      "step": 12540
+    },
+    {
+      "epoch": 2.8502272727272726,
+      "grad_norm": 2.875973296991617,
+      "learning_rate": 4.886604741271611e-07,
+      "loss": 0.0262,
+      "step": 12541
+    },
+    {
+      "epoch": 2.8504545454545456,
+      "grad_norm": 1.928800201950148,
+      "learning_rate": 4.885733751502329e-07,
+      "loss": 0.0351,
+      "step": 12542
+    },
+    {
+      "epoch": 2.850681818181818,
+      "grad_norm": 1.9017093438982124,
+      "learning_rate": 4.884862789552818e-07,
+      "loss": 0.0214,
+      "step": 12543
+    },
+    {
+      "epoch": 2.850909090909091,
+      "grad_norm": 2.5169497315156795,
+      "learning_rate": 4.883991855440834e-07,
+      "loss": 0.0254,
+      "step": 12544
+    },
+    {
+      "epoch": 2.8511363636363636,
+      "grad_norm": 2.5089303293238183,
+      "learning_rate": 4.883120949184141e-07,
+      "loss": 0.0393,
+      "step": 12545
+    },
+    {
+      "epoch": 2.8513636363636365,
+      "grad_norm": 1.9429759324496065,
+      "learning_rate": 4.882250070800497e-07,
+      "loss": 0.0237,
+      "step": 12546
+    },
+    {
+      "epoch": 2.851590909090909,
+      "grad_norm": 1.4337986311200677,
+      "learning_rate": 4.881379220307659e-07,
+      "loss": 0.0129,
+      "step": 12547
+    },
+    {
+      "epoch": 2.8518181818181816,
+      "grad_norm": 0.5787193587977623,
+      "learning_rate": 4.880508397723386e-07,
+      "loss": 0.0078,
+      "step": 12548
+    },
+    {
+      "epoch": 2.8520454545454546,
+      "grad_norm": 2.0387685308646124,
+      "learning_rate": 4.879637603065436e-07,
+      "loss": 0.0285,
+      "step": 12549
+    },
+    {
+      "epoch": 2.8522727272727275,
+      "grad_norm": 1.641361793330782,
+      "learning_rate": 4.87876683635157e-07,
+      "loss": 0.0248,
+      "step": 12550
+    },
+    {
+      "epoch": 2.8525,
+      "grad_norm": 1.768288418497092,
+      "learning_rate": 4.877896097599536e-07,
+      "loss": 0.0206,
+      "step": 12551
+    },
+    {
+      "epoch": 2.8527272727272726,
+      "grad_norm": 2.896386163987913,
+      "learning_rate": 4.877025386827096e-07,
+      "loss": 0.0303,
+      "step": 12552
+    },
+    {
+      "epoch": 2.8529545454545455,
+      "grad_norm": 1.570872617167085,
+      "learning_rate": 4.876154704052002e-07,
+      "loss": 0.0075,
+      "step": 12553
+    },
+    {
+      "epoch": 2.853181818181818,
+      "grad_norm": 1.8876263246671627,
+      "learning_rate": 4.875284049292013e-07,
+      "loss": 0.029,
+      "step": 12554
+    },
+    {
+      "epoch": 2.853409090909091,
+      "grad_norm": 1.2211790103891154,
+      "learning_rate": 4.874413422564878e-07,
+      "loss": 0.0186,
+      "step": 12555
+    },
+    {
+      "epoch": 2.8536363636363635,
+      "grad_norm": 2.362917300894924,
+      "learning_rate": 4.873542823888355e-07,
+      "loss": 0.0334,
+      "step": 12556
+    },
+    {
+      "epoch": 2.8538636363636365,
+      "grad_norm": 2.1611039025067025,
+      "learning_rate": 4.872672253280194e-07,
+      "loss": 0.0323,
+      "step": 12557
+    },
+    {
+      "epoch": 2.854090909090909,
+      "grad_norm": 1.173960810811248,
+      "learning_rate": 4.871801710758149e-07,
+      "loss": 0.0195,
+      "step": 12558
+    },
+    {
+      "epoch": 2.854318181818182,
+      "grad_norm": 1.8318871136786303,
+      "learning_rate": 4.870931196339974e-07,
+      "loss": 0.0104,
+      "step": 12559
+    },
+    {
+      "epoch": 2.8545454545454545,
+      "grad_norm": 1.6254200860505523,
+      "learning_rate": 4.870060710043415e-07,
+      "loss": 0.0264,
+      "step": 12560
+    },
+    {
+      "epoch": 2.854772727272727,
+      "grad_norm": 2.1640850616294474,
+      "learning_rate": 4.869190251886226e-07,
+      "loss": 0.0354,
+      "step": 12561
+    },
+    {
+      "epoch": 2.855,
+      "grad_norm": 0.9861575281922661,
+      "learning_rate": 4.868319821886155e-07,
+      "loss": 0.0115,
+      "step": 12562
+    },
+    {
+      "epoch": 2.855227272727273,
+      "grad_norm": 1.987202054803685,
+      "learning_rate": 4.867449420060957e-07,
+      "loss": 0.0288,
+      "step": 12563
+    },
+    {
+      "epoch": 2.8554545454545455,
+      "grad_norm": 1.7641753458964393,
+      "learning_rate": 4.866579046428375e-07,
+      "loss": 0.0233,
+      "step": 12564
+    },
+    {
+      "epoch": 2.855681818181818,
+      "grad_norm": 2.7508772233713925,
+      "learning_rate": 4.86570870100616e-07,
+      "loss": 0.0379,
+      "step": 12565
+    },
+    {
+      "epoch": 2.855909090909091,
+      "grad_norm": 1.9000322009009782,
+      "learning_rate": 4.86483838381206e-07,
+      "loss": 0.0282,
+      "step": 12566
+    },
+    {
+      "epoch": 2.8561363636363635,
+      "grad_norm": 1.693707016703422,
+      "learning_rate": 4.863968094863821e-07,
+      "loss": 0.0152,
+      "step": 12567
+    },
+    {
+      "epoch": 2.8563636363636364,
+      "grad_norm": 1.8391823321662424,
+      "learning_rate": 4.863097834179193e-07,
+      "loss": 0.0222,
+      "step": 12568
+    },
+    {
+      "epoch": 2.856590909090909,
+      "grad_norm": 1.6996943452263389,
+      "learning_rate": 4.862227601775918e-07,
+      "loss": 0.0156,
+      "step": 12569
+    },
+    {
+      "epoch": 2.856818181818182,
+      "grad_norm": 1.3340316316014085,
+      "learning_rate": 4.861357397671743e-07,
+      "loss": 0.0156,
+      "step": 12570
+    },
+    {
+      "epoch": 2.8570454545454544,
+      "grad_norm": 1.5844415952284268,
+      "learning_rate": 4.860487221884411e-07,
+      "loss": 0.012,
+      "step": 12571
+    },
+    {
+      "epoch": 2.8572727272727274,
+      "grad_norm": 1.4856192708678146,
+      "learning_rate": 4.859617074431674e-07,
+      "loss": 0.0117,
+      "step": 12572
+    },
+    {
+      "epoch": 2.8575,
+      "grad_norm": 2.1405506253715876,
+      "learning_rate": 4.858746955331268e-07,
+      "loss": 0.0203,
+      "step": 12573
+    },
+    {
+      "epoch": 2.857727272727273,
+      "grad_norm": 1.8418291421232673,
+      "learning_rate": 4.857876864600938e-07,
+      "loss": 0.0164,
+      "step": 12574
+    },
+    {
+      "epoch": 2.8579545454545454,
+      "grad_norm": 2.6685899862474085,
+      "learning_rate": 4.857006802258427e-07,
+      "loss": 0.0308,
+      "step": 12575
+    },
+    {
+      "epoch": 2.8581818181818184,
+      "grad_norm": 1.5153809428459226,
+      "learning_rate": 4.856136768321479e-07,
+      "loss": 0.017,
+      "step": 12576
+    },
+    {
+      "epoch": 2.858409090909091,
+      "grad_norm": 3.2353222398197437,
+      "learning_rate": 4.855266762807834e-07,
+      "loss": 0.0439,
+      "step": 12577
+    },
+    {
+      "epoch": 2.8586363636363634,
+      "grad_norm": 1.8064516894744196,
+      "learning_rate": 4.854396785735231e-07,
+      "loss": 0.0153,
+      "step": 12578
+    },
+    {
+      "epoch": 2.8588636363636364,
+      "grad_norm": 1.9073469329177009,
+      "learning_rate": 4.853526837121414e-07,
+      "loss": 0.0153,
+      "step": 12579
+    },
+    {
+      "epoch": 2.8590909090909093,
+      "grad_norm": 1.5108624450487917,
+      "learning_rate": 4.852656916984121e-07,
+      "loss": 0.028,
+      "step": 12580
+    },
+    {
+      "epoch": 2.859318181818182,
+      "grad_norm": 2.205453031509749,
+      "learning_rate": 4.851787025341093e-07,
+      "loss": 0.0235,
+      "step": 12581
+    },
+    {
+      "epoch": 2.8595454545454544,
+      "grad_norm": 1.6951180701912325,
+      "learning_rate": 4.850917162210064e-07,
+      "loss": 0.0211,
+      "step": 12582
+    },
+    {
+      "epoch": 2.8597727272727274,
+      "grad_norm": 1.330233568954747,
+      "learning_rate": 4.850047327608775e-07,
+      "loss": 0.0206,
+      "step": 12583
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.5486146985156481,
+      "learning_rate": 4.849177521554964e-07,
+      "loss": 0.0173,
+      "step": 12584
+    },
+    {
+      "epoch": 2.860227272727273,
+      "grad_norm": 1.8067894051055697,
+      "learning_rate": 4.848307744066369e-07,
+      "loss": 0.0242,
+      "step": 12585
+    },
+    {
+      "epoch": 2.8604545454545454,
+      "grad_norm": 3.131984911308389,
+      "learning_rate": 4.847437995160722e-07,
+      "loss": 0.0335,
+      "step": 12586
+    },
+    {
+      "epoch": 2.8606818181818183,
+      "grad_norm": 1.0095713327696236,
+      "learning_rate": 4.846568274855762e-07,
+      "loss": 0.0082,
+      "step": 12587
+    },
+    {
+      "epoch": 2.860909090909091,
+      "grad_norm": 1.9262018046580485,
+      "learning_rate": 4.845698583169223e-07,
+      "loss": 0.0158,
+      "step": 12588
+    },
+    {
+      "epoch": 2.861136363636364,
+      "grad_norm": 2.4890227864117733,
+      "learning_rate": 4.844828920118839e-07,
+      "loss": 0.0255,
+      "step": 12589
+    },
+    {
+      "epoch": 2.8613636363636363,
+      "grad_norm": 2.2182873371654934,
+      "learning_rate": 4.843959285722349e-07,
+      "loss": 0.0291,
+      "step": 12590
+    },
+    {
+      "epoch": 2.861590909090909,
+      "grad_norm": 2.11347067606543,
+      "learning_rate": 4.843089679997479e-07,
+      "loss": 0.0186,
+      "step": 12591
+    },
+    {
+      "epoch": 2.861818181818182,
+      "grad_norm": 0.9373382453134327,
+      "learning_rate": 4.842220102961964e-07,
+      "loss": 0.0047,
+      "step": 12592
+    },
+    {
+      "epoch": 2.862045454545455,
+      "grad_norm": 2.094202333260013,
+      "learning_rate": 4.841350554633538e-07,
+      "loss": 0.0198,
+      "step": 12593
+    },
+    {
+      "epoch": 2.8622727272727273,
+      "grad_norm": 2.5338685047384115,
+      "learning_rate": 4.840481035029933e-07,
+      "loss": 0.041,
+      "step": 12594
+    },
+    {
+      "epoch": 2.8625,
+      "grad_norm": 1.505582691950317,
+      "learning_rate": 4.839611544168877e-07,
+      "loss": 0.0194,
+      "step": 12595
+    },
+    {
+      "epoch": 2.862727272727273,
+      "grad_norm": 1.5027947803374448,
+      "learning_rate": 4.838742082068103e-07,
+      "loss": 0.0243,
+      "step": 12596
+    },
+    {
+      "epoch": 2.8629545454545453,
+      "grad_norm": 2.0172305927487857,
+      "learning_rate": 4.837872648745339e-07,
+      "loss": 0.0192,
+      "step": 12597
+    },
+    {
+      "epoch": 2.8631818181818183,
+      "grad_norm": 1.677246811908001,
+      "learning_rate": 4.837003244218319e-07,
+      "loss": 0.0193,
+      "step": 12598
+    },
+    {
+      "epoch": 2.863409090909091,
+      "grad_norm": 1.6839482812190043,
+      "learning_rate": 4.836133868504763e-07,
+      "loss": 0.0201,
+      "step": 12599
+    },
+    {
+      "epoch": 2.8636363636363638,
+      "grad_norm": 2.211596389059065,
+      "learning_rate": 4.835264521622406e-07,
+      "loss": 0.0261,
+      "step": 12600
+    },
+    {
+      "epoch": 2.8638636363636363,
+      "grad_norm": 2.417518026278159,
+      "learning_rate": 4.834395203588972e-07,
+      "loss": 0.0272,
+      "step": 12601
+    },
+    {
+      "epoch": 2.8640909090909092,
+      "grad_norm": 1.0971845330746637,
+      "learning_rate": 4.83352591442219e-07,
+      "loss": 0.0111,
+      "step": 12602
+    },
+    {
+      "epoch": 2.8643181818181818,
+      "grad_norm": 2.0481427988982364,
+      "learning_rate": 4.83265665413979e-07,
+      "loss": 0.018,
+      "step": 12603
+    },
+    {
+      "epoch": 2.8645454545454543,
+      "grad_norm": 1.5394547010455144,
+      "learning_rate": 4.83178742275949e-07,
+      "loss": 0.0206,
+      "step": 12604
+    },
+    {
+      "epoch": 2.8647727272727272,
+      "grad_norm": 3.780496625949409,
+      "learning_rate": 4.830918220299019e-07,
+      "loss": 0.0333,
+      "step": 12605
+    },
+    {
+      "epoch": 2.865,
+      "grad_norm": 2.138966337250704,
+      "learning_rate": 4.8300490467761e-07,
+      "loss": 0.0161,
+      "step": 12606
+    },
+    {
+      "epoch": 2.8652272727272727,
+      "grad_norm": 2.2462935133859725,
+      "learning_rate": 4.829179902208462e-07,
+      "loss": 0.0276,
+      "step": 12607
+    },
+    {
+      "epoch": 2.8654545454545453,
+      "grad_norm": 1.8584574272966634,
+      "learning_rate": 4.828310786613821e-07,
+      "loss": 0.0126,
+      "step": 12608
+    },
+    {
+      "epoch": 2.865681818181818,
+      "grad_norm": 1.320981317724673,
+      "learning_rate": 4.827441700009904e-07,
+      "loss": 0.013,
+      "step": 12609
+    },
+    {
+      "epoch": 2.865909090909091,
+      "grad_norm": 2.1390231751194047,
+      "learning_rate": 4.826572642414432e-07,
+      "loss": 0.0232,
+      "step": 12610
+    },
+    {
+      "epoch": 2.8661363636363637,
+      "grad_norm": 1.3483193932405575,
+      "learning_rate": 4.825703613845127e-07,
+      "loss": 0.0109,
+      "step": 12611
+    },
+    {
+      "epoch": 2.8663636363636362,
+      "grad_norm": 2.5607097616292442,
+      "learning_rate": 4.824834614319713e-07,
+      "loss": 0.0165,
+      "step": 12612
+    },
+    {
+      "epoch": 2.866590909090909,
+      "grad_norm": 3.5154014583678443,
+      "learning_rate": 4.823965643855906e-07,
+      "loss": 0.0306,
+      "step": 12613
+    },
+    {
+      "epoch": 2.8668181818181817,
+      "grad_norm": 2.7224373662928425,
+      "learning_rate": 4.823096702471427e-07,
+      "loss": 0.0241,
+      "step": 12614
+    },
+    {
+      "epoch": 2.8670454545454547,
+      "grad_norm": 2.0912477422261766,
+      "learning_rate": 4.822227790183995e-07,
+      "loss": 0.0246,
+      "step": 12615
+    },
+    {
+      "epoch": 2.867272727272727,
+      "grad_norm": 1.2549734842260498,
+      "learning_rate": 4.821358907011331e-07,
+      "loss": 0.0147,
+      "step": 12616
+    },
+    {
+      "epoch": 2.8675,
+      "grad_norm": 6.584073339891306,
+      "learning_rate": 4.820490052971149e-07,
+      "loss": 0.038,
+      "step": 12617
+    },
+    {
+      "epoch": 2.8677272727272727,
+      "grad_norm": 2.706472759521001,
+      "learning_rate": 4.81962122808117e-07,
+      "loss": 0.0398,
+      "step": 12618
+    },
+    {
+      "epoch": 2.8679545454545456,
+      "grad_norm": 2.3009144286697754,
+      "learning_rate": 4.818752432359108e-07,
+      "loss": 0.0221,
+      "step": 12619
+    },
+    {
+      "epoch": 2.868181818181818,
+      "grad_norm": 1.8161726728204524,
+      "learning_rate": 4.817883665822686e-07,
+      "loss": 0.0305,
+      "step": 12620
+    },
+    {
+      "epoch": 2.8684090909090907,
+      "grad_norm": 1.9795760760892007,
+      "learning_rate": 4.81701492848961e-07,
+      "loss": 0.0136,
+      "step": 12621
+    },
+    {
+      "epoch": 2.8686363636363637,
+      "grad_norm": 1.5174453701203594,
+      "learning_rate": 4.816146220377601e-07,
+      "loss": 0.0208,
+      "step": 12622
+    },
+    {
+      "epoch": 2.8688636363636366,
+      "grad_norm": 2.1972638266298357,
+      "learning_rate": 4.81527754150437e-07,
+      "loss": 0.0233,
+      "step": 12623
+    },
+    {
+      "epoch": 2.869090909090909,
+      "grad_norm": 2.7466639155123556,
+      "learning_rate": 4.814408891887635e-07,
+      "loss": 0.0375,
+      "step": 12624
+    },
+    {
+      "epoch": 2.8693181818181817,
+      "grad_norm": 2.341938720883978,
+      "learning_rate": 4.813540271545106e-07,
+      "loss": 0.0264,
+      "step": 12625
+    },
+    {
+      "epoch": 2.8695454545454546,
+      "grad_norm": 2.5973678560001563,
+      "learning_rate": 4.812671680494497e-07,
+      "loss": 0.0207,
+      "step": 12626
+    },
+    {
+      "epoch": 2.869772727272727,
+      "grad_norm": 1.5849452838280809,
+      "learning_rate": 4.81180311875352e-07,
+      "loss": 0.0141,
+      "step": 12627
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.424308641600252,
+      "learning_rate": 4.810934586339887e-07,
+      "loss": 0.0133,
+      "step": 12628
+    },
+    {
+      "epoch": 2.8702272727272726,
+      "grad_norm": 2.661495026511521,
+      "learning_rate": 4.810066083271312e-07,
+      "loss": 0.0232,
+      "step": 12629
+    },
+    {
+      "epoch": 2.8704545454545456,
+      "grad_norm": 1.6028839464811595,
+      "learning_rate": 4.809197609565496e-07,
+      "loss": 0.016,
+      "step": 12630
+    },
+    {
+      "epoch": 2.870681818181818,
+      "grad_norm": 3.3373629762355628,
+      "learning_rate": 4.808329165240155e-07,
+      "loss": 0.0298,
+      "step": 12631
+    },
+    {
+      "epoch": 2.870909090909091,
+      "grad_norm": 1.3334112598319874,
+      "learning_rate": 4.807460750313e-07,
+      "loss": 0.0146,
+      "step": 12632
+    },
+    {
+      "epoch": 2.8711363636363636,
+      "grad_norm": 1.3751076557530668,
+      "learning_rate": 4.806592364801737e-07,
+      "loss": 0.0229,
+      "step": 12633
+    },
+    {
+      "epoch": 2.871363636363636,
+      "grad_norm": 1.8814066266941893,
+      "learning_rate": 4.805724008724072e-07,
+      "loss": 0.0326,
+      "step": 12634
+    },
+    {
+      "epoch": 2.871590909090909,
+      "grad_norm": 1.5678607812705336,
+      "learning_rate": 4.804855682097715e-07,
+      "loss": 0.031,
+      "step": 12635
+    },
+    {
+      "epoch": 2.871818181818182,
+      "grad_norm": 0.8266448196353117,
+      "learning_rate": 4.803987384940371e-07,
+      "loss": 0.0057,
+      "step": 12636
+    },
+    {
+      "epoch": 2.8720454545454546,
+      "grad_norm": 2.8522664875045467,
+      "learning_rate": 4.803119117269747e-07,
+      "loss": 0.0253,
+      "step": 12637
+    },
+    {
+      "epoch": 2.872272727272727,
+      "grad_norm": 0.9350356796836354,
+      "learning_rate": 4.80225087910355e-07,
+      "loss": 0.0185,
+      "step": 12638
+    },
+    {
+      "epoch": 2.8725,
+      "grad_norm": 2.398640305921785,
+      "learning_rate": 4.801382670459482e-07,
+      "loss": 0.028,
+      "step": 12639
+    },
+    {
+      "epoch": 2.8727272727272726,
+      "grad_norm": 4.192562082024605,
+      "learning_rate": 4.800514491355249e-07,
+      "loss": 0.0112,
+      "step": 12640
+    },
+    {
+      "epoch": 2.8729545454545455,
+      "grad_norm": 1.8549993299644991,
+      "learning_rate": 4.799646341808553e-07,
+      "loss": 0.0201,
+      "step": 12641
+    },
+    {
+      "epoch": 2.873181818181818,
+      "grad_norm": 2.047954940246873,
+      "learning_rate": 4.798778221837102e-07,
+      "loss": 0.0205,
+      "step": 12642
+    },
+    {
+      "epoch": 2.873409090909091,
+      "grad_norm": 1.7298717222234592,
+      "learning_rate": 4.797910131458592e-07,
+      "loss": 0.0175,
+      "step": 12643
+    },
+    {
+      "epoch": 2.8736363636363635,
+      "grad_norm": 1.7260810183608202,
+      "learning_rate": 4.797042070690728e-07,
+      "loss": 0.015,
+      "step": 12644
+    },
+    {
+      "epoch": 2.8738636363636365,
+      "grad_norm": 1.082455687867897,
+      "learning_rate": 4.796174039551211e-07,
+      "loss": 0.0061,
+      "step": 12645
+    },
+    {
+      "epoch": 2.874090909090909,
+      "grad_norm": 0.9321883707838073,
+      "learning_rate": 4.795306038057742e-07,
+      "loss": 0.018,
+      "step": 12646
+    },
+    {
+      "epoch": 2.8743181818181816,
+      "grad_norm": 3.052168247371994,
+      "learning_rate": 4.79443806622802e-07,
+      "loss": 0.0402,
+      "step": 12647
+    },
+    {
+      "epoch": 2.8745454545454545,
+      "grad_norm": 3.3718972273171706,
+      "learning_rate": 4.793570124079746e-07,
+      "loss": 0.0305,
+      "step": 12648
+    },
+    {
+      "epoch": 2.8747727272727275,
+      "grad_norm": 2.2081287051195284,
+      "learning_rate": 4.792702211630618e-07,
+      "loss": 0.0181,
+      "step": 12649
+    },
+    {
+      "epoch": 2.875,
+      "grad_norm": 1.813304209933723,
+      "learning_rate": 4.791834328898332e-07,
+      "loss": 0.017,
+      "step": 12650
+    },
+    {
+      "epoch": 2.8752272727272725,
+      "grad_norm": 2.039943430492563,
+      "learning_rate": 4.790966475900594e-07,
+      "loss": 0.0171,
+      "step": 12651
+    },
+    {
+      "epoch": 2.8754545454545455,
+      "grad_norm": 1.4464129292464873,
+      "learning_rate": 4.790098652655089e-07,
+      "loss": 0.0233,
+      "step": 12652
+    },
+    {
+      "epoch": 2.8756818181818184,
+      "grad_norm": 1.283922812076184,
+      "learning_rate": 4.789230859179522e-07,
+      "loss": 0.0122,
+      "step": 12653
+    },
+    {
+      "epoch": 2.875909090909091,
+      "grad_norm": 1.7807545432445913,
+      "learning_rate": 4.788363095491585e-07,
+      "loss": 0.0161,
+      "step": 12654
+    },
+    {
+      "epoch": 2.8761363636363635,
+      "grad_norm": 1.5304132569437825,
+      "learning_rate": 4.787495361608976e-07,
+      "loss": 0.011,
+      "step": 12655
+    },
+    {
+      "epoch": 2.8763636363636365,
+      "grad_norm": 2.151816640187043,
+      "learning_rate": 4.786627657549388e-07,
+      "loss": 0.0337,
+      "step": 12656
+    },
+    {
+      "epoch": 2.876590909090909,
+      "grad_norm": 1.6874943391423196,
+      "learning_rate": 4.785759983330513e-07,
+      "loss": 0.0226,
+      "step": 12657
+    },
+    {
+      "epoch": 2.876818181818182,
+      "grad_norm": 2.1922934569739363,
+      "learning_rate": 4.784892338970047e-07,
+      "loss": 0.033,
+      "step": 12658
+    },
+    {
+      "epoch": 2.8770454545454545,
+      "grad_norm": 6.634619194063783,
+      "learning_rate": 4.784024724485683e-07,
+      "loss": 0.0269,
+      "step": 12659
+    },
+    {
+      "epoch": 2.8772727272727274,
+      "grad_norm": 2.093069385186525,
+      "learning_rate": 4.783157139895113e-07,
+      "loss": 0.0211,
+      "step": 12660
+    },
+    {
+      "epoch": 2.8775,
+      "grad_norm": 1.8971965285272865,
+      "learning_rate": 4.782289585216027e-07,
+      "loss": 0.0237,
+      "step": 12661
+    },
+    {
+      "epoch": 2.877727272727273,
+      "grad_norm": 0.94157241607602,
+      "learning_rate": 4.781422060466116e-07,
+      "loss": 0.0086,
+      "step": 12662
+    },
+    {
+      "epoch": 2.8779545454545454,
+      "grad_norm": 1.5472788048205115,
+      "learning_rate": 4.780554565663073e-07,
+      "loss": 0.0185,
+      "step": 12663
+    },
+    {
+      "epoch": 2.878181818181818,
+      "grad_norm": 1.9006414246880639,
+      "learning_rate": 4.779687100824587e-07,
+      "loss": 0.0226,
+      "step": 12664
+    },
+    {
+      "epoch": 2.878409090909091,
+      "grad_norm": 1.1827962263476994,
+      "learning_rate": 4.778819665968344e-07,
+      "loss": 0.0082,
+      "step": 12665
+    },
+    {
+      "epoch": 2.878636363636364,
+      "grad_norm": 2.126853766353373,
+      "learning_rate": 4.777952261112036e-07,
+      "loss": 0.0166,
+      "step": 12666
+    },
+    {
+      "epoch": 2.8788636363636364,
+      "grad_norm": 1.717000723321775,
+      "learning_rate": 4.777084886273348e-07,
+      "loss": 0.0222,
+      "step": 12667
+    },
+    {
+      "epoch": 2.879090909090909,
+      "grad_norm": 2.016957257045631,
+      "learning_rate": 4.776217541469969e-07,
+      "loss": 0.0311,
+      "step": 12668
+    },
+    {
+      "epoch": 2.879318181818182,
+      "grad_norm": 0.9453364811541899,
+      "learning_rate": 4.775350226719587e-07,
+      "loss": 0.0099,
+      "step": 12669
+    },
+    {
+      "epoch": 2.8795454545454544,
+      "grad_norm": 2.949972756365216,
+      "learning_rate": 4.774482942039884e-07,
+      "loss": 0.0232,
+      "step": 12670
+    },
+    {
+      "epoch": 2.8797727272727274,
+      "grad_norm": 2.0428600086600195,
+      "learning_rate": 4.77361568744855e-07,
+      "loss": 0.0539,
+      "step": 12671
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.1964009595226386,
+      "learning_rate": 4.772748462963267e-07,
+      "loss": 0.0068,
+      "step": 12672
+    },
+    {
+      "epoch": 2.880227272727273,
+      "grad_norm": 1.1996716418527744,
+      "learning_rate": 4.771881268601722e-07,
+      "loss": 0.011,
+      "step": 12673
+    },
+    {
+      "epoch": 2.8804545454545454,
+      "grad_norm": 0.820287231613384,
+      "learning_rate": 4.771014104381595e-07,
+      "loss": 0.006,
+      "step": 12674
+    },
+    {
+      "epoch": 2.8806818181818183,
+      "grad_norm": 1.7579980442444885,
+      "learning_rate": 4.770146970320571e-07,
+      "loss": 0.024,
+      "step": 12675
+    },
+    {
+      "epoch": 2.880909090909091,
+      "grad_norm": 5.400619247270469,
+      "learning_rate": 4.769279866436331e-07,
+      "loss": 0.0268,
+      "step": 12676
+    },
+    {
+      "epoch": 2.8811363636363634,
+      "grad_norm": 2.2094122914632197,
+      "learning_rate": 4.768412792746559e-07,
+      "loss": 0.0211,
+      "step": 12677
+    },
+    {
+      "epoch": 2.8813636363636363,
+      "grad_norm": 2.497071257677597,
+      "learning_rate": 4.7675457492689344e-07,
+      "loss": 0.0531,
+      "step": 12678
+    },
+    {
+      "epoch": 2.8815909090909093,
+      "grad_norm": 1.5884159724793538,
+      "learning_rate": 4.766678736021138e-07,
+      "loss": 0.0139,
+      "step": 12679
+    },
+    {
+      "epoch": 2.881818181818182,
+      "grad_norm": 2.0864258212113094,
+      "learning_rate": 4.76581175302085e-07,
+      "loss": 0.0227,
+      "step": 12680
+    },
+    {
+      "epoch": 2.8820454545454544,
+      "grad_norm": 2.2890882149466196,
+      "learning_rate": 4.764944800285749e-07,
+      "loss": 0.0309,
+      "step": 12681
+    },
+    {
+      "epoch": 2.8822727272727273,
+      "grad_norm": 1.9703141870272742,
+      "learning_rate": 4.764077877833518e-07,
+      "loss": 0.0219,
+      "step": 12682
+    },
+    {
+      "epoch": 2.8825,
+      "grad_norm": 2.797302256535186,
+      "learning_rate": 4.763210985681828e-07,
+      "loss": 0.0313,
+      "step": 12683
+    },
+    {
+      "epoch": 2.882727272727273,
+      "grad_norm": 1.427916706182776,
+      "learning_rate": 4.76234412384836e-07,
+      "loss": 0.018,
+      "step": 12684
+    },
+    {
+      "epoch": 2.8829545454545453,
+      "grad_norm": 2.702916465193195,
+      "learning_rate": 4.7614772923507897e-07,
+      "loss": 0.0432,
+      "step": 12685
+    },
+    {
+      "epoch": 2.8831818181818183,
+      "grad_norm": 1.3336977857445051,
+      "learning_rate": 4.7606104912067984e-07,
+      "loss": 0.0107,
+      "step": 12686
+    },
+    {
+      "epoch": 2.883409090909091,
+      "grad_norm": 1.8286605834707035,
+      "learning_rate": 4.7597437204340543e-07,
+      "loss": 0.0254,
+      "step": 12687
+    },
+    {
+      "epoch": 2.8836363636363638,
+      "grad_norm": 2.9606499667441395,
+      "learning_rate": 4.7588769800502356e-07,
+      "loss": 0.0579,
+      "step": 12688
+    },
+    {
+      "epoch": 2.8838636363636363,
+      "grad_norm": 2.4882378792893496,
+      "learning_rate": 4.758010270073017e-07,
+      "loss": 0.0258,
+      "step": 12689
+    },
+    {
+      "epoch": 2.884090909090909,
+      "grad_norm": 1.802387548403678,
+      "learning_rate": 4.757143590520072e-07,
+      "loss": 0.0411,
+      "step": 12690
+    },
+    {
+      "epoch": 2.884318181818182,
+      "grad_norm": 2.950902912481742,
+      "learning_rate": 4.756276941409075e-07,
+      "loss": 0.0618,
+      "step": 12691
+    },
+    {
+      "epoch": 2.8845454545454547,
+      "grad_norm": 1.5662783926230202,
+      "learning_rate": 4.7554103227576954e-07,
+      "loss": 0.0305,
+      "step": 12692
+    },
+    {
+      "epoch": 2.8847727272727273,
+      "grad_norm": 1.9435320751290397,
+      "learning_rate": 4.754543734583607e-07,
+      "loss": 0.0194,
+      "step": 12693
+    },
+    {
+      "epoch": 2.885,
+      "grad_norm": 0.9080091756814561,
+      "learning_rate": 4.753677176904481e-07,
+      "loss": 0.0077,
+      "step": 12694
+    },
+    {
+      "epoch": 2.8852272727272728,
+      "grad_norm": 2.6859885627696394,
+      "learning_rate": 4.7528106497379905e-07,
+      "loss": 0.0403,
+      "step": 12695
+    },
+    {
+      "epoch": 2.8854545454545457,
+      "grad_norm": 1.9275444927794632,
+      "learning_rate": 4.7519441531018e-07,
+      "loss": 0.0376,
+      "step": 12696
+    },
+    {
+      "epoch": 2.8856818181818182,
+      "grad_norm": 1.821056416937177,
+      "learning_rate": 4.751077687013582e-07,
+      "loss": 0.0191,
+      "step": 12697
+    },
+    {
+      "epoch": 2.8859090909090908,
+      "grad_norm": 3.631360071648419,
+      "learning_rate": 4.7502112514910044e-07,
+      "loss": 0.0519,
+      "step": 12698
+    },
+    {
+      "epoch": 2.8861363636363637,
+      "grad_norm": 6.488828377365266,
+      "learning_rate": 4.7493448465517373e-07,
+      "loss": 0.0333,
+      "step": 12699
+    },
+    {
+      "epoch": 2.8863636363636362,
+      "grad_norm": 2.0858626829188096,
+      "learning_rate": 4.7484784722134455e-07,
+      "loss": 0.0205,
+      "step": 12700
+    },
+    {
+      "epoch": 2.886590909090909,
+      "grad_norm": 1.8537193824569582,
+      "learning_rate": 4.7476121284937976e-07,
+      "loss": 0.0103,
+      "step": 12701
+    },
+    {
+      "epoch": 2.8868181818181817,
+      "grad_norm": 1.748111266822919,
+      "learning_rate": 4.746745815410459e-07,
+      "loss": 0.0249,
+      "step": 12702
+    },
+    {
+      "epoch": 2.8870454545454547,
+      "grad_norm": 1.832786871071193,
+      "learning_rate": 4.7458795329810965e-07,
+      "loss": 0.0447,
+      "step": 12703
+    },
+    {
+      "epoch": 2.887272727272727,
+      "grad_norm": 3.6909861742932053,
+      "learning_rate": 4.745013281223376e-07,
+      "loss": 0.0133,
+      "step": 12704
+    },
+    {
+      "epoch": 2.8875,
+      "grad_norm": 0.613053766709362,
+      "learning_rate": 4.7441470601549593e-07,
+      "loss": 0.0132,
+      "step": 12705
+    },
+    {
+      "epoch": 2.8877272727272727,
+      "grad_norm": 1.98590396935263,
+      "learning_rate": 4.7432808697935096e-07,
+      "loss": 0.0189,
+      "step": 12706
+    },
+    {
+      "epoch": 2.887954545454545,
+      "grad_norm": 1.913694320675549,
+      "learning_rate": 4.742414710156692e-07,
+      "loss": 0.0273,
+      "step": 12707
+    },
+    {
+      "epoch": 2.888181818181818,
+      "grad_norm": 1.1282990817994178,
+      "learning_rate": 4.7415485812621706e-07,
+      "loss": 0.0119,
+      "step": 12708
+    },
+    {
+      "epoch": 2.888409090909091,
+      "grad_norm": 2.024298222641397,
+      "learning_rate": 4.7406824831276025e-07,
+      "loss": 0.0287,
+      "step": 12709
+    },
+    {
+      "epoch": 2.8886363636363637,
+      "grad_norm": 1.54366230456827,
+      "learning_rate": 4.7398164157706527e-07,
+      "loss": 0.0153,
+      "step": 12710
+    },
+    {
+      "epoch": 2.888863636363636,
+      "grad_norm": 1.5795892586192364,
+      "learning_rate": 4.7389503792089806e-07,
+      "loss": 0.0134,
+      "step": 12711
+    },
+    {
+      "epoch": 2.889090909090909,
+      "grad_norm": 1.5627407700202474,
+      "learning_rate": 4.7380843734602463e-07,
+      "loss": 0.0212,
+      "step": 12712
+    },
+    {
+      "epoch": 2.8893181818181817,
+      "grad_norm": 1.6760684629041913,
+      "learning_rate": 4.737218398542112e-07,
+      "loss": 0.0166,
+      "step": 12713
+    },
+    {
+      "epoch": 2.8895454545454546,
+      "grad_norm": 4.448763999671923,
+      "learning_rate": 4.7363524544722303e-07,
+      "loss": 0.0397,
+      "step": 12714
+    },
+    {
+      "epoch": 2.889772727272727,
+      "grad_norm": 1.2178563928746065,
+      "learning_rate": 4.735486541268264e-07,
+      "loss": 0.0082,
+      "step": 12715
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.423312825433941,
+      "learning_rate": 4.7346206589478695e-07,
+      "loss": 0.0128,
+      "step": 12716
+    },
+    {
+      "epoch": 2.8902272727272726,
+      "grad_norm": 1.3463377566422405,
+      "learning_rate": 4.733754807528705e-07,
+      "loss": 0.0159,
+      "step": 12717
+    },
+    {
+      "epoch": 2.8904545454545456,
+      "grad_norm": 2.262854432451777,
+      "learning_rate": 4.7328889870284237e-07,
+      "loss": 0.0266,
+      "step": 12718
+    },
+    {
+      "epoch": 2.890681818181818,
+      "grad_norm": 0.8592950551541637,
+      "learning_rate": 4.7320231974646824e-07,
+      "loss": 0.0075,
+      "step": 12719
+    },
+    {
+      "epoch": 2.8909090909090907,
+      "grad_norm": 1.5856619902566487,
+      "learning_rate": 4.7311574388551356e-07,
+      "loss": 0.022,
+      "step": 12720
+    },
+    {
+      "epoch": 2.8911363636363636,
+      "grad_norm": 1.3281540436404364,
+      "learning_rate": 4.7302917112174405e-07,
+      "loss": 0.0196,
+      "step": 12721
+    },
+    {
+      "epoch": 2.8913636363636366,
+      "grad_norm": 1.0022041318740866,
+      "learning_rate": 4.7294260145692473e-07,
+      "loss": 0.0076,
+      "step": 12722
+    },
+    {
+      "epoch": 2.891590909090909,
+      "grad_norm": 2.016521601145222,
+      "learning_rate": 4.72856034892821e-07,
+      "loss": 0.0326,
+      "step": 12723
+    },
+    {
+      "epoch": 2.8918181818181816,
+      "grad_norm": 1.384059188878859,
+      "learning_rate": 4.7276947143119827e-07,
+      "loss": 0.0125,
+      "step": 12724
+    },
+    {
+      "epoch": 2.8920454545454546,
+      "grad_norm": 3.9246186525698916,
+      "learning_rate": 4.7268291107382146e-07,
+      "loss": 0.0885,
+      "step": 12725
+    },
+    {
+      "epoch": 2.892272727272727,
+      "grad_norm": 1.5852627345206889,
+      "learning_rate": 4.7259635382245617e-07,
+      "loss": 0.015,
+      "step": 12726
+    },
+    {
+      "epoch": 2.8925,
+      "grad_norm": 4.094680971039065,
+      "learning_rate": 4.725097996788668e-07,
+      "loss": 0.0323,
+      "step": 12727
+    },
+    {
+      "epoch": 2.8927272727272726,
+      "grad_norm": 0.907870892138521,
+      "learning_rate": 4.724232486448187e-07,
+      "loss": 0.0171,
+      "step": 12728
+    },
+    {
+      "epoch": 2.8929545454545456,
+      "grad_norm": 1.53720696854636,
+      "learning_rate": 4.7233670072207664e-07,
+      "loss": 0.0103,
+      "step": 12729
+    },
+    {
+      "epoch": 2.893181818181818,
+      "grad_norm": 1.8289499112372856,
+      "learning_rate": 4.7225015591240583e-07,
+      "loss": 0.0122,
+      "step": 12730
+    },
+    {
+      "epoch": 2.893409090909091,
+      "grad_norm": 2.3586598214313987,
+      "learning_rate": 4.721636142175706e-07,
+      "loss": 0.0667,
+      "step": 12731
+    },
+    {
+      "epoch": 2.8936363636363636,
+      "grad_norm": 2.0968645883472727,
+      "learning_rate": 4.72077075639336e-07,
+      "loss": 0.0207,
+      "step": 12732
+    },
+    {
+      "epoch": 2.8938636363636365,
+      "grad_norm": 1.6784989889059774,
+      "learning_rate": 4.7199054017946657e-07,
+      "loss": 0.0189,
+      "step": 12733
+    },
+    {
+      "epoch": 2.894090909090909,
+      "grad_norm": 1.2403758084777012,
+      "learning_rate": 4.7190400783972704e-07,
+      "loss": 0.0144,
+      "step": 12734
+    },
+    {
+      "epoch": 2.894318181818182,
+      "grad_norm": 5.2766564653650665,
+      "learning_rate": 4.71817478621882e-07,
+      "loss": 0.049,
+      "step": 12735
+    },
+    {
+      "epoch": 2.8945454545454545,
+      "grad_norm": 2.1696196877530216,
+      "learning_rate": 4.717309525276957e-07,
+      "loss": 0.0163,
+      "step": 12736
+    },
+    {
+      "epoch": 2.894772727272727,
+      "grad_norm": 2.278602090611038,
+      "learning_rate": 4.7164442955893256e-07,
+      "loss": 0.0208,
+      "step": 12737
+    },
+    {
+      "epoch": 2.895,
+      "grad_norm": 2.637614392956556,
+      "learning_rate": 4.7155790971735713e-07,
+      "loss": 0.0317,
+      "step": 12738
+    },
+    {
+      "epoch": 2.895227272727273,
+      "grad_norm": 1.8269984786289248,
+      "learning_rate": 4.714713930047337e-07,
+      "loss": 0.0161,
+      "step": 12739
+    },
+    {
+      "epoch": 2.8954545454545455,
+      "grad_norm": 1.4166760511336112,
+      "learning_rate": 4.7138487942282624e-07,
+      "loss": 0.0221,
+      "step": 12740
+    },
+    {
+      "epoch": 2.895681818181818,
+      "grad_norm": 1.7347846132235225,
+      "learning_rate": 4.7129836897339915e-07,
+      "loss": 0.0393,
+      "step": 12741
+    },
+    {
+      "epoch": 2.895909090909091,
+      "grad_norm": 2.819964884485168,
+      "learning_rate": 4.7121186165821646e-07,
+      "loss": 0.0218,
+      "step": 12742
+    },
+    {
+      "epoch": 2.8961363636363635,
+      "grad_norm": 2.0222177654159386,
+      "learning_rate": 4.7112535747904235e-07,
+      "loss": 0.0154,
+      "step": 12743
+    },
+    {
+      "epoch": 2.8963636363636365,
+      "grad_norm": 1.6390736462083846,
+      "learning_rate": 4.710388564376404e-07,
+      "loss": 0.0132,
+      "step": 12744
+    },
+    {
+      "epoch": 2.896590909090909,
+      "grad_norm": 3.5997627028789942,
+      "learning_rate": 4.709523585357749e-07,
+      "loss": 0.0581,
+      "step": 12745
+    },
+    {
+      "epoch": 2.896818181818182,
+      "grad_norm": 2.5507477494879653,
+      "learning_rate": 4.7086586377520954e-07,
+      "loss": 0.0419,
+      "step": 12746
+    },
+    {
+      "epoch": 2.8970454545454545,
+      "grad_norm": 3.870290073212872,
+      "learning_rate": 4.707793721577081e-07,
+      "loss": 0.0517,
+      "step": 12747
+    },
+    {
+      "epoch": 2.8972727272727274,
+      "grad_norm": 1.3094675662113664,
+      "learning_rate": 4.706928836850345e-07,
+      "loss": 0.0157,
+      "step": 12748
+    },
+    {
+      "epoch": 2.8975,
+      "grad_norm": 2.079906109111437,
+      "learning_rate": 4.7060639835895214e-07,
+      "loss": 0.0433,
+      "step": 12749
+    },
+    {
+      "epoch": 2.8977272727272725,
+      "grad_norm": 1.786988215880228,
+      "learning_rate": 4.705199161812245e-07,
+      "loss": 0.0185,
+      "step": 12750
+    },
+    {
+      "epoch": 2.8979545454545454,
+      "grad_norm": 1.8795184753531344,
+      "learning_rate": 4.7043343715361534e-07,
+      "loss": 0.0132,
+      "step": 12751
+    },
+    {
+      "epoch": 2.8981818181818184,
+      "grad_norm": 1.1025018481770807,
+      "learning_rate": 4.703469612778882e-07,
+      "loss": 0.0102,
+      "step": 12752
+    },
+    {
+      "epoch": 2.898409090909091,
+      "grad_norm": 1.6285461839712236,
+      "learning_rate": 4.7026048855580613e-07,
+      "loss": 0.0188,
+      "step": 12753
+    },
+    {
+      "epoch": 2.8986363636363635,
+      "grad_norm": 0.922963162349737,
+      "learning_rate": 4.701740189891327e-07,
+      "loss": 0.0093,
+      "step": 12754
+    },
+    {
+      "epoch": 2.8988636363636364,
+      "grad_norm": 2.029829040134524,
+      "learning_rate": 4.700875525796311e-07,
+      "loss": 0.0185,
+      "step": 12755
+    },
+    {
+      "epoch": 2.899090909090909,
+      "grad_norm": 2.0117777549657068,
+      "learning_rate": 4.700010893290649e-07,
+      "loss": 0.0213,
+      "step": 12756
+    },
+    {
+      "epoch": 2.899318181818182,
+      "grad_norm": 1.4371339365780162,
+      "learning_rate": 4.6991462923919656e-07,
+      "loss": 0.0172,
+      "step": 12757
+    },
+    {
+      "epoch": 2.8995454545454544,
+      "grad_norm": 2.552830193012968,
+      "learning_rate": 4.698281723117895e-07,
+      "loss": 0.0308,
+      "step": 12758
+    },
+    {
+      "epoch": 2.8997727272727274,
+      "grad_norm": 2.562341169774093,
+      "learning_rate": 4.6974171854860673e-07,
+      "loss": 0.0192,
+      "step": 12759
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.603115140335274,
+      "learning_rate": 4.6965526795141114e-07,
+      "loss": 0.0142,
+      "step": 12760
+    },
+    {
+      "epoch": 2.900227272727273,
+      "grad_norm": 2.2666253195926025,
+      "learning_rate": 4.695688205219659e-07,
+      "loss": 0.0236,
+      "step": 12761
+    },
+    {
+      "epoch": 2.9004545454545454,
+      "grad_norm": 2.2039963894961696,
+      "learning_rate": 4.6948237626203333e-07,
+      "loss": 0.0114,
+      "step": 12762
+    },
+    {
+      "epoch": 2.900681818181818,
+      "grad_norm": 1.0374208006411,
+      "learning_rate": 4.693959351733765e-07,
+      "loss": 0.0119,
+      "step": 12763
+    },
+    {
+      "epoch": 2.900909090909091,
+      "grad_norm": 5.116340223947326,
+      "learning_rate": 4.69309497257758e-07,
+      "loss": 0.022,
+      "step": 12764
+    },
+    {
+      "epoch": 2.901136363636364,
+      "grad_norm": 1.0242019105032165,
+      "learning_rate": 4.6922306251694067e-07,
+      "loss": 0.0067,
+      "step": 12765
+    },
+    {
+      "epoch": 2.9013636363636364,
+      "grad_norm": 2.1382626857778604,
+      "learning_rate": 4.691366309526867e-07,
+      "loss": 0.041,
+      "step": 12766
+    },
+    {
+      "epoch": 2.901590909090909,
+      "grad_norm": 2.219660589403809,
+      "learning_rate": 4.690502025667586e-07,
+      "loss": 0.0232,
+      "step": 12767
+    },
+    {
+      "epoch": 2.901818181818182,
+      "grad_norm": 2.183726076551549,
+      "learning_rate": 4.6896377736091897e-07,
+      "loss": 0.0564,
+      "step": 12768
+    },
+    {
+      "epoch": 2.902045454545455,
+      "grad_norm": 4.8893904794506255,
+      "learning_rate": 4.6887735533693016e-07,
+      "loss": 0.0284,
+      "step": 12769
+    },
+    {
+      "epoch": 2.9022727272727273,
+      "grad_norm": 1.6936211838666124,
+      "learning_rate": 4.6879093649655476e-07,
+      "loss": 0.0237,
+      "step": 12770
+    },
+    {
+      "epoch": 2.9025,
+      "grad_norm": 1.971614448403573,
+      "learning_rate": 4.6870452084155444e-07,
+      "loss": 0.0359,
+      "step": 12771
+    },
+    {
+      "epoch": 2.902727272727273,
+      "grad_norm": 0.7597331053713364,
+      "learning_rate": 4.6861810837369155e-07,
+      "loss": 0.0061,
+      "step": 12772
+    },
+    {
+      "epoch": 2.9029545454545453,
+      "grad_norm": 2.8189594712868624,
+      "learning_rate": 4.685316990947283e-07,
+      "loss": 0.0295,
+      "step": 12773
+    },
+    {
+      "epoch": 2.9031818181818183,
+      "grad_norm": 1.4155964084642894,
+      "learning_rate": 4.6844529300642677e-07,
+      "loss": 0.0084,
+      "step": 12774
+    },
+    {
+      "epoch": 2.903409090909091,
+      "grad_norm": 0.5245704457734446,
+      "learning_rate": 4.6835889011054875e-07,
+      "loss": 0.0099,
+      "step": 12775
+    },
+    {
+      "epoch": 2.903636363636364,
+      "grad_norm": 2.097816773840052,
+      "learning_rate": 4.682724904088562e-07,
+      "loss": 0.0187,
+      "step": 12776
+    },
+    {
+      "epoch": 2.9038636363636363,
+      "grad_norm": 1.5892562028094879,
+      "learning_rate": 4.6818609390311106e-07,
+      "loss": 0.0186,
+      "step": 12777
+    },
+    {
+      "epoch": 2.9040909090909093,
+      "grad_norm": 1.6979546674123438,
+      "learning_rate": 4.680997005950752e-07,
+      "loss": 0.0172,
+      "step": 12778
+    },
+    {
+      "epoch": 2.904318181818182,
+      "grad_norm": 2.45019299409277,
+      "learning_rate": 4.6801331048651e-07,
+      "loss": 0.0456,
+      "step": 12779
+    },
+    {
+      "epoch": 2.9045454545454543,
+      "grad_norm": 2.084470566565779,
+      "learning_rate": 4.6792692357917724e-07,
+      "loss": 0.0247,
+      "step": 12780
+    },
+    {
+      "epoch": 2.9047727272727273,
+      "grad_norm": 2.171491743497737,
+      "learning_rate": 4.6784053987483853e-07,
+      "loss": 0.0184,
+      "step": 12781
+    },
+    {
+      "epoch": 2.9050000000000002,
+      "grad_norm": 1.733452925154436,
+      "learning_rate": 4.677541593752554e-07,
+      "loss": 0.0273,
+      "step": 12782
+    },
+    {
+      "epoch": 2.9052272727272728,
+      "grad_norm": 2.105252036172478,
+      "learning_rate": 4.676677820821894e-07,
+      "loss": 0.0253,
+      "step": 12783
+    },
+    {
+      "epoch": 2.9054545454545453,
+      "grad_norm": 2.9814318571964993,
+      "learning_rate": 4.675814079974017e-07,
+      "loss": 0.0266,
+      "step": 12784
+    },
+    {
+      "epoch": 2.9056818181818183,
+      "grad_norm": 1.1294154574716144,
+      "learning_rate": 4.674950371226537e-07,
+      "loss": 0.0134,
+      "step": 12785
+    },
+    {
+      "epoch": 2.9059090909090908,
+      "grad_norm": 2.1031600229306364,
+      "learning_rate": 4.674086694597067e-07,
+      "loss": 0.0354,
+      "step": 12786
+    },
+    {
+      "epoch": 2.9061363636363637,
+      "grad_norm": 2.7856373061864246,
+      "learning_rate": 4.673223050103222e-07,
+      "loss": 0.0549,
+      "step": 12787
+    },
+    {
+      "epoch": 2.9063636363636363,
+      "grad_norm": 1.0416586410005952,
+      "learning_rate": 4.6723594377626064e-07,
+      "loss": 0.0078,
+      "step": 12788
+    },
+    {
+      "epoch": 2.9065909090909092,
+      "grad_norm": 1.0979450033049019,
+      "learning_rate": 4.671495857592834e-07,
+      "loss": 0.0089,
+      "step": 12789
+    },
+    {
+      "epoch": 2.9068181818181817,
+      "grad_norm": 1.4012563496533375,
+      "learning_rate": 4.6706323096115154e-07,
+      "loss": 0.017,
+      "step": 12790
+    },
+    {
+      "epoch": 2.9070454545454547,
+      "grad_norm": 3.1106989705748966,
+      "learning_rate": 4.6697687938362585e-07,
+      "loss": 0.0314,
+      "step": 12791
+    },
+    {
+      "epoch": 2.9072727272727272,
+      "grad_norm": 1.1587737821503639,
+      "learning_rate": 4.6689053102846755e-07,
+      "loss": 0.0135,
+      "step": 12792
+    },
+    {
+      "epoch": 2.9074999999999998,
+      "grad_norm": 0.6541535858423282,
+      "learning_rate": 4.6680418589743687e-07,
+      "loss": 0.0058,
+      "step": 12793
+    },
+    {
+      "epoch": 2.9077272727272727,
+      "grad_norm": 1.577939790942823,
+      "learning_rate": 4.6671784399229497e-07,
+      "loss": 0.0253,
+      "step": 12794
+    },
+    {
+      "epoch": 2.9079545454545457,
+      "grad_norm": 1.5765174173517542,
+      "learning_rate": 4.6663150531480233e-07,
+      "loss": 0.0411,
+      "step": 12795
+    },
+    {
+      "epoch": 2.908181818181818,
+      "grad_norm": 1.2664859124035734,
+      "learning_rate": 4.665451698667198e-07,
+      "loss": 0.0259,
+      "step": 12796
+    },
+    {
+      "epoch": 2.9084090909090907,
+      "grad_norm": 0.7297651769940211,
+      "learning_rate": 4.664588376498075e-07,
+      "loss": 0.0063,
+      "step": 12797
+    },
+    {
+      "epoch": 2.9086363636363637,
+      "grad_norm": 1.9368731998829012,
+      "learning_rate": 4.663725086658259e-07,
+      "loss": 0.0138,
+      "step": 12798
+    },
+    {
+      "epoch": 2.908863636363636,
+      "grad_norm": 2.1677992561707122,
+      "learning_rate": 4.6628618291653575e-07,
+      "loss": 0.0265,
+      "step": 12799
+    },
+    {
+      "epoch": 2.909090909090909,
+      "grad_norm": 1.568140184509729,
+      "learning_rate": 4.6619986040369744e-07,
+      "loss": 0.022,
+      "step": 12800
+    },
+    {
+      "epoch": 2.9093181818181817,
+      "grad_norm": 1.406576713182176,
+      "learning_rate": 4.6611354112907084e-07,
+      "loss": 0.01,
+      "step": 12801
+    },
+    {
+      "epoch": 2.9095454545454547,
+      "grad_norm": 1.1000907221185376,
+      "learning_rate": 4.6602722509441626e-07,
+      "loss": 0.0133,
+      "step": 12802
+    },
+    {
+      "epoch": 2.909772727272727,
+      "grad_norm": 0.8859887685597434,
+      "learning_rate": 4.6594091230149394e-07,
+      "loss": 0.0037,
+      "step": 12803
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.7152698825316168,
+      "learning_rate": 4.6585460275206385e-07,
+      "loss": 0.0373,
+      "step": 12804
+    },
+    {
+      "epoch": 2.9102272727272727,
+      "grad_norm": 1.4251525969837835,
+      "learning_rate": 4.657682964478863e-07,
+      "loss": 0.0158,
+      "step": 12805
+    },
+    {
+      "epoch": 2.910454545454545,
+      "grad_norm": 1.9993836500361735,
+      "learning_rate": 4.656819933907207e-07,
+      "loss": 0.019,
+      "step": 12806
+    },
+    {
+      "epoch": 2.910681818181818,
+      "grad_norm": 3.270670221278392,
+      "learning_rate": 4.6559569358232735e-07,
+      "loss": 0.0358,
+      "step": 12807
+    },
+    {
+      "epoch": 2.910909090909091,
+      "grad_norm": 2.7303826265747273,
+      "learning_rate": 4.655093970244658e-07,
+      "loss": 0.0418,
+      "step": 12808
+    },
+    {
+      "epoch": 2.9111363636363636,
+      "grad_norm": 1.8457628164705437,
+      "learning_rate": 4.6542310371889616e-07,
+      "loss": 0.0138,
+      "step": 12809
+    },
+    {
+      "epoch": 2.911363636363636,
+      "grad_norm": 1.7741408500933284,
+      "learning_rate": 4.653368136673776e-07,
+      "loss": 0.0267,
+      "step": 12810
+    },
+    {
+      "epoch": 2.911590909090909,
+      "grad_norm": 3.578110780607535,
+      "learning_rate": 4.6525052687167e-07,
+      "loss": 0.0436,
+      "step": 12811
+    },
+    {
+      "epoch": 2.911818181818182,
+      "grad_norm": 1.4885312126317694,
+      "learning_rate": 4.651642433335328e-07,
+      "loss": 0.0139,
+      "step": 12812
+    },
+    {
+      "epoch": 2.9120454545454546,
+      "grad_norm": 0.9196448728772708,
+      "learning_rate": 4.6507796305472553e-07,
+      "loss": 0.008,
+      "step": 12813
+    },
+    {
+      "epoch": 2.912272727272727,
+      "grad_norm": 2.76812601058781,
+      "learning_rate": 4.649916860370077e-07,
+      "loss": 0.0448,
+      "step": 12814
+    },
+    {
+      "epoch": 2.9125,
+      "grad_norm": 1.1279331181281627,
+      "learning_rate": 4.6490541228213845e-07,
+      "loss": 0.0139,
+      "step": 12815
+    },
+    {
+      "epoch": 2.9127272727272726,
+      "grad_norm": 1.094198039109162,
+      "learning_rate": 4.648191417918771e-07,
+      "loss": 0.0128,
+      "step": 12816
+    },
+    {
+      "epoch": 2.9129545454545456,
+      "grad_norm": 1.4416805284099805,
+      "learning_rate": 4.64732874567983e-07,
+      "loss": 0.0223,
+      "step": 12817
+    },
+    {
+      "epoch": 2.913181818181818,
+      "grad_norm": 1.3997224284348369,
+      "learning_rate": 4.6464661061221536e-07,
+      "loss": 0.0202,
+      "step": 12818
+    },
+    {
+      "epoch": 2.913409090909091,
+      "grad_norm": 1.2781072700734126,
+      "learning_rate": 4.6456034992633283e-07,
+      "loss": 0.0154,
+      "step": 12819
+    },
+    {
+      "epoch": 2.9136363636363636,
+      "grad_norm": 0.8343061238406108,
+      "learning_rate": 4.644740925120947e-07,
+      "loss": 0.0136,
+      "step": 12820
+    },
+    {
+      "epoch": 2.9138636363636365,
+      "grad_norm": 2.7027690946822016,
+      "learning_rate": 4.643878383712598e-07,
+      "loss": 0.0887,
+      "step": 12821
+    },
+    {
+      "epoch": 2.914090909090909,
+      "grad_norm": 1.3651361289905817,
+      "learning_rate": 4.643015875055873e-07,
+      "loss": 0.0104,
+      "step": 12822
+    },
+    {
+      "epoch": 2.9143181818181816,
+      "grad_norm": 1.2707162452521676,
+      "learning_rate": 4.642153399168356e-07,
+      "loss": 0.0182,
+      "step": 12823
+    },
+    {
+      "epoch": 2.9145454545454546,
+      "grad_norm": 1.0883885242312785,
+      "learning_rate": 4.641290956067637e-07,
+      "loss": 0.0124,
+      "step": 12824
+    },
+    {
+      "epoch": 2.9147727272727275,
+      "grad_norm": 2.7829203566950143,
+      "learning_rate": 4.6404285457713013e-07,
+      "loss": 0.0327,
+      "step": 12825
+    },
+    {
+      "epoch": 2.915,
+      "grad_norm": 2.4711641191996017,
+      "learning_rate": 4.639566168296936e-07,
+      "loss": 0.0192,
+      "step": 12826
+    },
+    {
+      "epoch": 2.9152272727272726,
+      "grad_norm": 1.3927848594464924,
+      "learning_rate": 4.638703823662127e-07,
+      "loss": 0.024,
+      "step": 12827
+    },
+    {
+      "epoch": 2.9154545454545455,
+      "grad_norm": 1.2352458205461332,
+      "learning_rate": 4.637841511884457e-07,
+      "loss": 0.0178,
+      "step": 12828
+    },
+    {
+      "epoch": 2.915681818181818,
+      "grad_norm": 3.13498389351569,
+      "learning_rate": 4.6369792329815103e-07,
+      "loss": 0.0542,
+      "step": 12829
+    },
+    {
+      "epoch": 2.915909090909091,
+      "grad_norm": 2.4579175918928735,
+      "learning_rate": 4.636116986970872e-07,
+      "loss": 0.0512,
+      "step": 12830
+    },
+    {
+      "epoch": 2.9161363636363635,
+      "grad_norm": 2.8454906578027144,
+      "learning_rate": 4.6352547738701266e-07,
+      "loss": 0.0307,
+      "step": 12831
+    },
+    {
+      "epoch": 2.9163636363636365,
+      "grad_norm": 1.987044051209683,
+      "learning_rate": 4.6343925936968503e-07,
+      "loss": 0.0336,
+      "step": 12832
+    },
+    {
+      "epoch": 2.916590909090909,
+      "grad_norm": 1.6964953836320904,
+      "learning_rate": 4.633530446468628e-07,
+      "loss": 0.0392,
+      "step": 12833
+    },
+    {
+      "epoch": 2.916818181818182,
+      "grad_norm": 2.217609883978251,
+      "learning_rate": 4.63266833220304e-07,
+      "loss": 0.0274,
+      "step": 12834
+    },
+    {
+      "epoch": 2.9170454545454545,
+      "grad_norm": 1.2051956181650447,
+      "learning_rate": 4.631806250917666e-07,
+      "loss": 0.0225,
+      "step": 12835
+    },
+    {
+      "epoch": 2.917272727272727,
+      "grad_norm": 3.0392291529574393,
+      "learning_rate": 4.6309442026300875e-07,
+      "loss": 0.0279,
+      "step": 12836
+    },
+    {
+      "epoch": 2.9175,
+      "grad_norm": 1.4506662724556942,
+      "learning_rate": 4.630082187357879e-07,
+      "loss": 0.0123,
+      "step": 12837
+    },
+    {
+      "epoch": 2.917727272727273,
+      "grad_norm": 2.1387404296890655,
+      "learning_rate": 4.6292202051186216e-07,
+      "loss": 0.0337,
+      "step": 12838
+    },
+    {
+      "epoch": 2.9179545454545455,
+      "grad_norm": 1.9430262375237846,
+      "learning_rate": 4.628358255929892e-07,
+      "loss": 0.0406,
+      "step": 12839
+    },
+    {
+      "epoch": 2.918181818181818,
+      "grad_norm": 1.9435359523402158,
+      "learning_rate": 4.627496339809269e-07,
+      "loss": 0.0347,
+      "step": 12840
+    },
+    {
+      "epoch": 2.918409090909091,
+      "grad_norm": 0.9235395774473552,
+      "learning_rate": 4.626634456774323e-07,
+      "loss": 0.0069,
+      "step": 12841
+    },
+    {
+      "epoch": 2.9186363636363635,
+      "grad_norm": 1.5604401072274419,
+      "learning_rate": 4.6257726068426335e-07,
+      "loss": 0.012,
+      "step": 12842
+    },
+    {
+      "epoch": 2.9188636363636364,
+      "grad_norm": 1.6115190970063829,
+      "learning_rate": 4.624910790031774e-07,
+      "loss": 0.0212,
+      "step": 12843
+    },
+    {
+      "epoch": 2.919090909090909,
+      "grad_norm": 1.6469483845510242,
+      "learning_rate": 4.624049006359319e-07,
+      "loss": 0.0275,
+      "step": 12844
+    },
+    {
+      "epoch": 2.919318181818182,
+      "grad_norm": 1.7599178172040992,
+      "learning_rate": 4.6231872558428413e-07,
+      "loss": 0.0167,
+      "step": 12845
+    },
+    {
+      "epoch": 2.9195454545454544,
+      "grad_norm": 2.7312812040488343,
+      "learning_rate": 4.6223255384999124e-07,
+      "loss": 0.0298,
+      "step": 12846
+    },
+    {
+      "epoch": 2.9197727272727274,
+      "grad_norm": 2.0814141477354178,
+      "learning_rate": 4.6214638543481055e-07,
+      "loss": 0.0247,
+      "step": 12847
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.2547290080782887,
+      "learning_rate": 4.620602203404993e-07,
+      "loss": 0.0153,
+      "step": 12848
+    },
+    {
+      "epoch": 2.920227272727273,
+      "grad_norm": 2.121890383485981,
+      "learning_rate": 4.619740585688146e-07,
+      "loss": 0.0124,
+      "step": 12849
+    },
+    {
+      "epoch": 2.9204545454545454,
+      "grad_norm": 2.2860695503416233,
+      "learning_rate": 4.6188790012151304e-07,
+      "loss": 0.0133,
+      "step": 12850
+    },
+    {
+      "epoch": 2.9206818181818184,
+      "grad_norm": 0.9403353421662599,
+      "learning_rate": 4.6180174500035173e-07,
+      "loss": 0.0154,
+      "step": 12851
+    },
+    {
+      "epoch": 2.920909090909091,
+      "grad_norm": 3.4124693363581433,
+      "learning_rate": 4.617155932070874e-07,
+      "loss": 0.0315,
+      "step": 12852
+    },
+    {
+      "epoch": 2.9211363636363634,
+      "grad_norm": 2.071544725993916,
+      "learning_rate": 4.6162944474347743e-07,
+      "loss": 0.0166,
+      "step": 12853
+    },
+    {
+      "epoch": 2.9213636363636364,
+      "grad_norm": 1.9751137881736387,
+      "learning_rate": 4.615432996112778e-07,
+      "loss": 0.0259,
+      "step": 12854
+    },
+    {
+      "epoch": 2.9215909090909093,
+      "grad_norm": 2.116593105098848,
+      "learning_rate": 4.6145715781224553e-07,
+      "loss": 0.0264,
+      "step": 12855
+    },
+    {
+      "epoch": 2.921818181818182,
+      "grad_norm": 1.668871927080665,
+      "learning_rate": 4.61371019348137e-07,
+      "loss": 0.0121,
+      "step": 12856
+    },
+    {
+      "epoch": 2.9220454545454544,
+      "grad_norm": 3.044154650802851,
+      "learning_rate": 4.6128488422070893e-07,
+      "loss": 0.0397,
+      "step": 12857
+    },
+    {
+      "epoch": 2.9222727272727274,
+      "grad_norm": 1.6484572319937987,
+      "learning_rate": 4.611987524317178e-07,
+      "loss": 0.0193,
+      "step": 12858
+    },
+    {
+      "epoch": 2.9225,
+      "grad_norm": 1.3472855498288892,
+      "learning_rate": 4.611126239829197e-07,
+      "loss": 0.0085,
+      "step": 12859
+    },
+    {
+      "epoch": 2.922727272727273,
+      "grad_norm": 1.8999665192806077,
+      "learning_rate": 4.6102649887607124e-07,
+      "loss": 0.0254,
+      "step": 12860
+    },
+    {
+      "epoch": 2.9229545454545454,
+      "grad_norm": 1.4364498740942635,
+      "learning_rate": 4.6094037711292843e-07,
+      "loss": 0.015,
+      "step": 12861
+    },
+    {
+      "epoch": 2.9231818181818183,
+      "grad_norm": 1.962676186653043,
+      "learning_rate": 4.6085425869524786e-07,
+      "loss": 0.0201,
+      "step": 12862
+    },
+    {
+      "epoch": 2.923409090909091,
+      "grad_norm": 2.505283998693439,
+      "learning_rate": 4.60768143624785e-07,
+      "loss": 0.029,
+      "step": 12863
+    },
+    {
+      "epoch": 2.923636363636364,
+      "grad_norm": 1.7598870520928318,
+      "learning_rate": 4.6068203190329635e-07,
+      "loss": 0.0318,
+      "step": 12864
+    },
+    {
+      "epoch": 2.9238636363636363,
+      "grad_norm": 2.2273725030757285,
+      "learning_rate": 4.605959235325376e-07,
+      "loss": 0.0159,
+      "step": 12865
+    },
+    {
+      "epoch": 2.924090909090909,
+      "grad_norm": 2.76367466279571,
+      "learning_rate": 4.6050981851426495e-07,
+      "loss": 0.0349,
+      "step": 12866
+    },
+    {
+      "epoch": 2.924318181818182,
+      "grad_norm": 2.0232070636777153,
+      "learning_rate": 4.6042371685023396e-07,
+      "loss": 0.013,
+      "step": 12867
+    },
+    {
+      "epoch": 2.924545454545455,
+      "grad_norm": 1.2857856072963612,
+      "learning_rate": 4.603376185422005e-07,
+      "loss": 0.0111,
+      "step": 12868
+    },
+    {
+      "epoch": 2.9247727272727273,
+      "grad_norm": 0.8657049657822482,
+      "learning_rate": 4.6025152359192026e-07,
+      "loss": 0.0175,
+      "step": 12869
+    },
+    {
+      "epoch": 2.925,
+      "grad_norm": 2.1025338596598453,
+      "learning_rate": 4.601654320011488e-07,
+      "loss": 0.0295,
+      "step": 12870
+    },
+    {
+      "epoch": 2.925227272727273,
+      "grad_norm": 2.4916154256629817,
+      "learning_rate": 4.600793437716421e-07,
+      "loss": 0.0302,
+      "step": 12871
+    },
+    {
+      "epoch": 2.9254545454545453,
+      "grad_norm": 3.0302348902690293,
+      "learning_rate": 4.5999325890515507e-07,
+      "loss": 0.0219,
+      "step": 12872
+    },
+    {
+      "epoch": 2.9256818181818183,
+      "grad_norm": 2.540174669787578,
+      "learning_rate": 4.599071774034433e-07,
+      "loss": 0.0324,
+      "step": 12873
+    },
+    {
+      "epoch": 2.925909090909091,
+      "grad_norm": 3.2003835571244124,
+      "learning_rate": 4.5982109926826223e-07,
+      "loss": 0.0287,
+      "step": 12874
+    },
+    {
+      "epoch": 2.9261363636363638,
+      "grad_norm": 1.3353710596792197,
+      "learning_rate": 4.5973502450136723e-07,
+      "loss": 0.0184,
+      "step": 12875
+    },
+    {
+      "epoch": 2.9263636363636363,
+      "grad_norm": 1.3020856717143703,
+      "learning_rate": 4.5964895310451325e-07,
+      "loss": 0.0167,
+      "step": 12876
+    },
+    {
+      "epoch": 2.9265909090909092,
+      "grad_norm": 1.6694349686908783,
+      "learning_rate": 4.595628850794556e-07,
+      "loss": 0.0209,
+      "step": 12877
+    },
+    {
+      "epoch": 2.9268181818181818,
+      "grad_norm": 2.0450133333592326,
+      "learning_rate": 4.594768204279494e-07,
+      "loss": 0.0056,
+      "step": 12878
+    },
+    {
+      "epoch": 2.9270454545454543,
+      "grad_norm": 2.607467334104472,
+      "learning_rate": 4.5939075915174975e-07,
+      "loss": 0.0265,
+      "step": 12879
+    },
+    {
+      "epoch": 2.9272727272727272,
+      "grad_norm": 2.4775188426709325,
+      "learning_rate": 4.5930470125261127e-07,
+      "loss": 0.0277,
+      "step": 12880
+    },
+    {
+      "epoch": 2.9275,
+      "grad_norm": 2.128687315845399,
+      "learning_rate": 4.59218646732289e-07,
+      "loss": 0.0238,
+      "step": 12881
+    },
+    {
+      "epoch": 2.9277272727272727,
+      "grad_norm": 1.8088331372199047,
+      "learning_rate": 4.5913259559253756e-07,
+      "loss": 0.0225,
+      "step": 12882
+    },
+    {
+      "epoch": 2.9279545454545453,
+      "grad_norm": 2.924679514336959,
+      "learning_rate": 4.5904654783511206e-07,
+      "loss": 0.0368,
+      "step": 12883
+    },
+    {
+      "epoch": 2.928181818181818,
+      "grad_norm": 1.8582533097470877,
+      "learning_rate": 4.589605034617672e-07,
+      "loss": 0.0169,
+      "step": 12884
+    },
+    {
+      "epoch": 2.928409090909091,
+      "grad_norm": 2.3007926958187936,
+      "learning_rate": 4.5887446247425705e-07,
+      "loss": 0.0293,
+      "step": 12885
+    },
+    {
+      "epoch": 2.9286363636363637,
+      "grad_norm": 2.1323135988469963,
+      "learning_rate": 4.5878842487433657e-07,
+      "loss": 0.0277,
+      "step": 12886
+    },
+    {
+      "epoch": 2.9288636363636362,
+      "grad_norm": 1.2570562850505678,
+      "learning_rate": 4.5870239066376e-07,
+      "loss": 0.0139,
+      "step": 12887
+    },
+    {
+      "epoch": 2.929090909090909,
+      "grad_norm": 2.153941996781355,
+      "learning_rate": 4.586163598442819e-07,
+      "loss": 0.0341,
+      "step": 12888
+    },
+    {
+      "epoch": 2.9293181818181817,
+      "grad_norm": 2.336172611223306,
+      "learning_rate": 4.5853033241765633e-07,
+      "loss": 0.0264,
+      "step": 12889
+    },
+    {
+      "epoch": 2.9295454545454547,
+      "grad_norm": 2.990663795956292,
+      "learning_rate": 4.5844430838563777e-07,
+      "loss": 0.0456,
+      "step": 12890
+    },
+    {
+      "epoch": 2.929772727272727,
+      "grad_norm": 1.3559800631112602,
+      "learning_rate": 4.5835828774998035e-07,
+      "loss": 0.0117,
+      "step": 12891
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.9731879512688983,
+      "learning_rate": 4.582722705124381e-07,
+      "loss": 0.0207,
+      "step": 12892
+    },
+    {
+      "epoch": 2.9302272727272727,
+      "grad_norm": 1.9505485425958204,
+      "learning_rate": 4.581862566747654e-07,
+      "loss": 0.0239,
+      "step": 12893
+    },
+    {
+      "epoch": 2.9304545454545456,
+      "grad_norm": 2.2619324077429734,
+      "learning_rate": 4.5810024623871575e-07,
+      "loss": 0.0424,
+      "step": 12894
+    },
+    {
+      "epoch": 2.930681818181818,
+      "grad_norm": 2.2629238293106537,
+      "learning_rate": 4.580142392060433e-07,
+      "loss": 0.0564,
+      "step": 12895
+    },
+    {
+      "epoch": 2.9309090909090907,
+      "grad_norm": 2.0081091167111174,
+      "learning_rate": 4.5792823557850167e-07,
+      "loss": 0.02,
+      "step": 12896
+    },
+    {
+      "epoch": 2.9311363636363637,
+      "grad_norm": 3.857434628708828,
+      "learning_rate": 4.5784223535784504e-07,
+      "loss": 0.0186,
+      "step": 12897
+    },
+    {
+      "epoch": 2.9313636363636366,
+      "grad_norm": 1.8772812167537936,
+      "learning_rate": 4.577562385458266e-07,
+      "loss": 0.0264,
+      "step": 12898
+    },
+    {
+      "epoch": 2.931590909090909,
+      "grad_norm": 0.4675051471115613,
+      "learning_rate": 4.5767024514420035e-07,
+      "loss": 0.006,
+      "step": 12899
+    },
+    {
+      "epoch": 2.9318181818181817,
+      "grad_norm": 2.28470224248399,
+      "learning_rate": 4.5758425515471965e-07,
+      "loss": 0.0376,
+      "step": 12900
+    },
+    {
+      "epoch": 2.9320454545454546,
+      "grad_norm": 3.7064476589150033,
+      "learning_rate": 4.5749826857913843e-07,
+      "loss": 0.0356,
+      "step": 12901
+    },
+    {
+      "epoch": 2.932272727272727,
+      "grad_norm": 1.6972566832330365,
+      "learning_rate": 4.5741228541920936e-07,
+      "loss": 0.0099,
+      "step": 12902
+    },
+    {
+      "epoch": 2.9325,
+      "grad_norm": 7.033760532469388,
+      "learning_rate": 4.5732630567668623e-07,
+      "loss": 0.0277,
+      "step": 12903
+    },
+    {
+      "epoch": 2.9327272727272726,
+      "grad_norm": 3.0572959959534427,
+      "learning_rate": 4.5724032935332223e-07,
+      "loss": 0.0724,
+      "step": 12904
+    },
+    {
+      "epoch": 2.9329545454545456,
+      "grad_norm": 2.403350964719268,
+      "learning_rate": 4.5715435645087057e-07,
+      "loss": 0.0236,
+      "step": 12905
+    },
+    {
+      "epoch": 2.933181818181818,
+      "grad_norm": 2.300228454336316,
+      "learning_rate": 4.570683869710846e-07,
+      "loss": 0.0203,
+      "step": 12906
+    },
+    {
+      "epoch": 2.933409090909091,
+      "grad_norm": 2.672099454516298,
+      "learning_rate": 4.56982420915717e-07,
+      "loss": 0.0583,
+      "step": 12907
+    },
+    {
+      "epoch": 2.9336363636363636,
+      "grad_norm": 2.5124141177878947,
+      "learning_rate": 4.5689645828652095e-07,
+      "loss": 0.0305,
+      "step": 12908
+    },
+    {
+      "epoch": 2.933863636363636,
+      "grad_norm": 1.9286197410674675,
+      "learning_rate": 4.568104990852494e-07,
+      "loss": 0.0201,
+      "step": 12909
+    },
+    {
+      "epoch": 2.934090909090909,
+      "grad_norm": 3.370516870183737,
+      "learning_rate": 4.567245433136555e-07,
+      "loss": 0.039,
+      "step": 12910
+    },
+    {
+      "epoch": 2.934318181818182,
+      "grad_norm": 2.4139120752338736,
+      "learning_rate": 4.5663859097349127e-07,
+      "loss": 0.0276,
+      "step": 12911
+    },
+    {
+      "epoch": 2.9345454545454546,
+      "grad_norm": 0.992004740706082,
+      "learning_rate": 4.565526420665101e-07,
+      "loss": 0.0121,
+      "step": 12912
+    },
+    {
+      "epoch": 2.934772727272727,
+      "grad_norm": 2.336699166981702,
+      "learning_rate": 4.5646669659446433e-07,
+      "loss": 0.0676,
+      "step": 12913
+    },
+    {
+      "epoch": 2.935,
+      "grad_norm": 1.4958753646143619,
+      "learning_rate": 4.5638075455910675e-07,
+      "loss": 0.0204,
+      "step": 12914
+    },
+    {
+      "epoch": 2.9352272727272726,
+      "grad_norm": 3.5576183913752817,
+      "learning_rate": 4.5629481596218995e-07,
+      "loss": 0.0372,
+      "step": 12915
+    },
+    {
+      "epoch": 2.9354545454545455,
+      "grad_norm": 3.124643841300237,
+      "learning_rate": 4.562088808054659e-07,
+      "loss": 0.033,
+      "step": 12916
+    },
+    {
+      "epoch": 2.935681818181818,
+      "grad_norm": 1.3854567998419345,
+      "learning_rate": 4.561229490906873e-07,
+      "loss": 0.0158,
+      "step": 12917
+    },
+    {
+      "epoch": 2.935909090909091,
+      "grad_norm": 1.1811537288975893,
+      "learning_rate": 4.5603702081960633e-07,
+      "loss": 0.0166,
+      "step": 12918
+    },
+    {
+      "epoch": 2.9361363636363635,
+      "grad_norm": 1.9120871691617016,
+      "learning_rate": 4.5595109599397543e-07,
+      "loss": 0.0317,
+      "step": 12919
+    },
+    {
+      "epoch": 2.9363636363636365,
+      "grad_norm": 1.751916099667407,
+      "learning_rate": 4.558651746155465e-07,
+      "loss": 0.0132,
+      "step": 12920
+    },
+    {
+      "epoch": 2.936590909090909,
+      "grad_norm": 1.8152551960893057,
+      "learning_rate": 4.557792566860718e-07,
+      "loss": 0.0168,
+      "step": 12921
+    },
+    {
+      "epoch": 2.9368181818181816,
+      "grad_norm": 2.8476738781355726,
+      "learning_rate": 4.5569334220730313e-07,
+      "loss": 0.035,
+      "step": 12922
+    },
+    {
+      "epoch": 2.9370454545454545,
+      "grad_norm": 1.8324963973072892,
+      "learning_rate": 4.55607431180993e-07,
+      "loss": 0.0467,
+      "step": 12923
+    },
+    {
+      "epoch": 2.9372727272727275,
+      "grad_norm": 1.3330452222824278,
+      "learning_rate": 4.555215236088925e-07,
+      "loss": 0.0128,
+      "step": 12924
+    },
+    {
+      "epoch": 2.9375,
+      "grad_norm": 1.1870851714473016,
+      "learning_rate": 4.554356194927538e-07,
+      "loss": 0.0143,
+      "step": 12925
+    },
+    {
+      "epoch": 2.9377272727272725,
+      "grad_norm": 2.3394276384402417,
+      "learning_rate": 4.5534971883432866e-07,
+      "loss": 0.0243,
+      "step": 12926
+    },
+    {
+      "epoch": 2.9379545454545455,
+      "grad_norm": 3.3543803679660584,
+      "learning_rate": 4.5526382163536874e-07,
+      "loss": 0.0292,
+      "step": 12927
+    },
+    {
+      "epoch": 2.9381818181818184,
+      "grad_norm": 2.474476953577468,
+      "learning_rate": 4.551779278976257e-07,
+      "loss": 0.065,
+      "step": 12928
+    },
+    {
+      "epoch": 2.938409090909091,
+      "grad_norm": 1.3841806733573023,
+      "learning_rate": 4.5509203762285097e-07,
+      "loss": 0.0165,
+      "step": 12929
+    },
+    {
+      "epoch": 2.9386363636363635,
+      "grad_norm": 1.477201371720455,
+      "learning_rate": 4.5500615081279593e-07,
+      "loss": 0.0186,
+      "step": 12930
+    },
+    {
+      "epoch": 2.9388636363636365,
+      "grad_norm": 1.8189480120679453,
+      "learning_rate": 4.54920267469212e-07,
+      "loss": 0.0219,
+      "step": 12931
+    },
+    {
+      "epoch": 2.939090909090909,
+      "grad_norm": 2.55985019900086,
+      "learning_rate": 4.5483438759385086e-07,
+      "loss": 0.0186,
+      "step": 12932
+    },
+    {
+      "epoch": 2.939318181818182,
+      "grad_norm": 1.2656943863825239,
+      "learning_rate": 4.5474851118846315e-07,
+      "loss": 0.0208,
+      "step": 12933
+    },
+    {
+      "epoch": 2.9395454545454545,
+      "grad_norm": 2.084049669893803,
+      "learning_rate": 4.546626382548003e-07,
+      "loss": 0.018,
+      "step": 12934
+    },
+    {
+      "epoch": 2.9397727272727274,
+      "grad_norm": 2.0738659923452136,
+      "learning_rate": 4.5457676879461343e-07,
+      "loss": 0.022,
+      "step": 12935
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.6516303801861543,
+      "learning_rate": 4.544909028096535e-07,
+      "loss": 0.0429,
+      "step": 12936
+    },
+    {
+      "epoch": 2.940227272727273,
+      "grad_norm": 1.310723879954941,
+      "learning_rate": 4.5440504030167185e-07,
+      "loss": 0.0246,
+      "step": 12937
+    },
+    {
+      "epoch": 2.9404545454545454,
+      "grad_norm": 1.2925214378202246,
+      "learning_rate": 4.543191812724188e-07,
+      "loss": 0.015,
+      "step": 12938
+    },
+    {
+      "epoch": 2.940681818181818,
+      "grad_norm": 1.2242733758751483,
+      "learning_rate": 4.5423332572364545e-07,
+      "loss": 0.0162,
+      "step": 12939
+    },
+    {
+      "epoch": 2.940909090909091,
+      "grad_norm": 3.2834069435169275,
+      "learning_rate": 4.541474736571025e-07,
+      "loss": 0.0369,
+      "step": 12940
+    },
+    {
+      "epoch": 2.941136363636364,
+      "grad_norm": 2.0963113823660158,
+      "learning_rate": 4.540616250745407e-07,
+      "loss": 0.0416,
+      "step": 12941
+    },
+    {
+      "epoch": 2.9413636363636364,
+      "grad_norm": 2.4574282703557735,
+      "learning_rate": 4.539757799777105e-07,
+      "loss": 0.0504,
+      "step": 12942
+    },
+    {
+      "epoch": 2.941590909090909,
+      "grad_norm": 1.3335547448260054,
+      "learning_rate": 4.5388993836836254e-07,
+      "loss": 0.0158,
+      "step": 12943
+    },
+    {
+      "epoch": 2.941818181818182,
+      "grad_norm": 1.8642876812492757,
+      "learning_rate": 4.538041002482472e-07,
+      "loss": 0.0241,
+      "step": 12944
+    },
+    {
+      "epoch": 2.9420454545454544,
+      "grad_norm": 1.6269661486562679,
+      "learning_rate": 4.5371826561911525e-07,
+      "loss": 0.0244,
+      "step": 12945
+    },
+    {
+      "epoch": 2.9422727272727274,
+      "grad_norm": 1.6931720963921066,
+      "learning_rate": 4.536324344827163e-07,
+      "loss": 0.0204,
+      "step": 12946
+    },
+    {
+      "epoch": 2.9425,
+      "grad_norm": 1.544929998590618,
+      "learning_rate": 4.5354660684080107e-07,
+      "loss": 0.0231,
+      "step": 12947
+    },
+    {
+      "epoch": 2.942727272727273,
+      "grad_norm": 0.6153149629105922,
+      "learning_rate": 4.534607826951196e-07,
+      "loss": 0.013,
+      "step": 12948
+    },
+    {
+      "epoch": 2.9429545454545454,
+      "grad_norm": 1.8872112418682931,
+      "learning_rate": 4.5337496204742213e-07,
+      "loss": 0.0133,
+      "step": 12949
+    },
+    {
+      "epoch": 2.9431818181818183,
+      "grad_norm": 2.9656986491243704,
+      "learning_rate": 4.5328914489945864e-07,
+      "loss": 0.0339,
+      "step": 12950
+    },
+    {
+      "epoch": 2.943409090909091,
+      "grad_norm": 1.9308941109003972,
+      "learning_rate": 4.532033312529789e-07,
+      "loss": 0.0179,
+      "step": 12951
+    },
+    {
+      "epoch": 2.9436363636363634,
+      "grad_norm": 2.036237856633111,
+      "learning_rate": 4.53117521109733e-07,
+      "loss": 0.0183,
+      "step": 12952
+    },
+    {
+      "epoch": 2.9438636363636363,
+      "grad_norm": 1.071455257577245,
+      "learning_rate": 4.5303171447147077e-07,
+      "loss": 0.0118,
+      "step": 12953
+    },
+    {
+      "epoch": 2.9440909090909093,
+      "grad_norm": 1.4866629984421145,
+      "learning_rate": 4.529459113399421e-07,
+      "loss": 0.012,
+      "step": 12954
+    },
+    {
+      "epoch": 2.944318181818182,
+      "grad_norm": 3.863831515195682,
+      "learning_rate": 4.528601117168962e-07,
+      "loss": 0.0635,
+      "step": 12955
+    },
+    {
+      "epoch": 2.9445454545454544,
+      "grad_norm": 3.5671254441704283,
+      "learning_rate": 4.527743156040829e-07,
+      "loss": 0.028,
+      "step": 12956
+    },
+    {
+      "epoch": 2.9447727272727273,
+      "grad_norm": 1.1045156304534025,
+      "learning_rate": 4.526885230032518e-07,
+      "loss": 0.0072,
+      "step": 12957
+    },
+    {
+      "epoch": 2.945,
+      "grad_norm": 1.7818114315842148,
+      "learning_rate": 4.5260273391615235e-07,
+      "loss": 0.0256,
+      "step": 12958
+    },
+    {
+      "epoch": 2.945227272727273,
+      "grad_norm": 2.0664987095472602,
+      "learning_rate": 4.5251694834453406e-07,
+      "loss": 0.0194,
+      "step": 12959
+    },
+    {
+      "epoch": 2.9454545454545453,
+      "grad_norm": 1.3104824669935247,
+      "learning_rate": 4.524311662901459e-07,
+      "loss": 0.0132,
+      "step": 12960
+    },
+    {
+      "epoch": 2.9456818181818183,
+      "grad_norm": 0.5391868766634489,
+      "learning_rate": 4.5234538775473737e-07,
+      "loss": 0.0066,
+      "step": 12961
+    },
+    {
+      "epoch": 2.945909090909091,
+      "grad_norm": 1.089852281978197,
+      "learning_rate": 4.5225961274005763e-07,
+      "loss": 0.0115,
+      "step": 12962
+    },
+    {
+      "epoch": 2.9461363636363638,
+      "grad_norm": 1.517019026402329,
+      "learning_rate": 4.52173841247856e-07,
+      "loss": 0.0082,
+      "step": 12963
+    },
+    {
+      "epoch": 2.9463636363636363,
+      "grad_norm": 1.804011497580289,
+      "learning_rate": 4.5208807327988093e-07,
+      "loss": 0.035,
+      "step": 12964
+    },
+    {
+      "epoch": 2.946590909090909,
+      "grad_norm": 1.1489140112836052,
+      "learning_rate": 4.5200230883788164e-07,
+      "loss": 0.014,
+      "step": 12965
+    },
+    {
+      "epoch": 2.946818181818182,
+      "grad_norm": 1.8220192934510266,
+      "learning_rate": 4.519165479236072e-07,
+      "loss": 0.0475,
+      "step": 12966
+    },
+    {
+      "epoch": 2.9470454545454547,
+      "grad_norm": 2.0902481600546596,
+      "learning_rate": 4.5183079053880655e-07,
+      "loss": 0.0171,
+      "step": 12967
+    },
+    {
+      "epoch": 2.9472727272727273,
+      "grad_norm": 1.021193505244001,
+      "learning_rate": 4.5174503668522795e-07,
+      "loss": 0.0113,
+      "step": 12968
+    },
+    {
+      "epoch": 2.9475,
+      "grad_norm": 1.5681107266831917,
+      "learning_rate": 4.5165928636462033e-07,
+      "loss": 0.0271,
+      "step": 12969
+    },
+    {
+      "epoch": 2.9477272727272728,
+      "grad_norm": 9.157627857754074,
+      "learning_rate": 4.515735395787323e-07,
+      "loss": 0.0181,
+      "step": 12970
+    },
+    {
+      "epoch": 2.9479545454545457,
+      "grad_norm": 4.165257514678662,
+      "learning_rate": 4.5148779632931225e-07,
+      "loss": 0.0384,
+      "step": 12971
+    },
+    {
+      "epoch": 2.9481818181818182,
+      "grad_norm": 1.9912460835740557,
+      "learning_rate": 4.514020566181089e-07,
+      "loss": 0.0286,
+      "step": 12972
+    },
+    {
+      "epoch": 2.9484090909090908,
+      "grad_norm": 2.3961946761947615,
+      "learning_rate": 4.5131632044687043e-07,
+      "loss": 0.0228,
+      "step": 12973
+    },
+    {
+      "epoch": 2.9486363636363637,
+      "grad_norm": 2.2092714162486535,
+      "learning_rate": 4.512305878173451e-07,
+      "loss": 0.0256,
+      "step": 12974
+    },
+    {
+      "epoch": 2.9488636363636362,
+      "grad_norm": 2.809641083333498,
+      "learning_rate": 4.5114485873128123e-07,
+      "loss": 0.0304,
+      "step": 12975
+    },
+    {
+      "epoch": 2.949090909090909,
+      "grad_norm": 1.6447562745988975,
+      "learning_rate": 4.510591331904274e-07,
+      "loss": 0.0248,
+      "step": 12976
+    },
+    {
+      "epoch": 2.9493181818181817,
+      "grad_norm": 3.6044700646187797,
+      "learning_rate": 4.5097341119653087e-07,
+      "loss": 0.0847,
+      "step": 12977
+    },
+    {
+      "epoch": 2.9495454545454547,
+      "grad_norm": 1.5540558554873487,
+      "learning_rate": 4.5088769275134017e-07,
+      "loss": 0.0164,
+      "step": 12978
+    },
+    {
+      "epoch": 2.949772727272727,
+      "grad_norm": 1.8406337991078807,
+      "learning_rate": 4.508019778566032e-07,
+      "loss": 0.0312,
+      "step": 12979
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.3956475922915823,
+      "learning_rate": 4.507162665140677e-07,
+      "loss": 0.0227,
+      "step": 12980
+    },
+    {
+      "epoch": 2.9502272727272727,
+      "grad_norm": 1.3842446139163476,
+      "learning_rate": 4.5063055872548183e-07,
+      "loss": 0.0493,
+      "step": 12981
+    },
+    {
+      "epoch": 2.950454545454545,
+      "grad_norm": 5.928980880496549,
+      "learning_rate": 4.505448544925929e-07,
+      "loss": 0.0251,
+      "step": 12982
+    },
+    {
+      "epoch": 2.950681818181818,
+      "grad_norm": 2.5612229876366435,
+      "learning_rate": 4.504591538171487e-07,
+      "loss": 0.0222,
+      "step": 12983
+    },
+    {
+      "epoch": 2.950909090909091,
+      "grad_norm": 3.2233494379268044,
+      "learning_rate": 4.5037345670089686e-07,
+      "loss": 0.0467,
+      "step": 12984
+    },
+    {
+      "epoch": 2.9511363636363637,
+      "grad_norm": 1.639602369157019,
+      "learning_rate": 4.502877631455853e-07,
+      "loss": 0.0273,
+      "step": 12985
+    },
+    {
+      "epoch": 2.951363636363636,
+      "grad_norm": 2.0444347774773632,
+      "learning_rate": 4.5020207315296065e-07,
+      "loss": 0.0309,
+      "step": 12986
+    },
+    {
+      "epoch": 2.951590909090909,
+      "grad_norm": 1.0514930446056137,
+      "learning_rate": 4.501163867247707e-07,
+      "loss": 0.0078,
+      "step": 12987
+    },
+    {
+      "epoch": 2.9518181818181817,
+      "grad_norm": 3.240722353115551,
+      "learning_rate": 4.500307038627627e-07,
+      "loss": 0.0418,
+      "step": 12988
+    },
+    {
+      "epoch": 2.9520454545454546,
+      "grad_norm": 2.233613783954144,
+      "learning_rate": 4.499450245686841e-07,
+      "loss": 0.032,
+      "step": 12989
+    },
+    {
+      "epoch": 2.952272727272727,
+      "grad_norm": 1.5743687712025913,
+      "learning_rate": 4.498593488442816e-07,
+      "loss": 0.0109,
+      "step": 12990
+    },
+    {
+      "epoch": 2.9525,
+      "grad_norm": 2.2168778179379816,
+      "learning_rate": 4.497736766913026e-07,
+      "loss": 0.0387,
+      "step": 12991
+    },
+    {
+      "epoch": 2.9527272727272726,
+      "grad_norm": 2.8654646533292696,
+      "learning_rate": 4.49688008111494e-07,
+      "loss": 0.0561,
+      "step": 12992
+    },
+    {
+      "epoch": 2.9529545454545456,
+      "grad_norm": 2.1191686084012926,
+      "learning_rate": 4.496023431066029e-07,
+      "loss": 0.0256,
+      "step": 12993
+    },
+    {
+      "epoch": 2.953181818181818,
+      "grad_norm": 2.6157943256583605,
+      "learning_rate": 4.4951668167837614e-07,
+      "loss": 0.0227,
+      "step": 12994
+    },
+    {
+      "epoch": 2.9534090909090907,
+      "grad_norm": 2.2118417820045955,
+      "learning_rate": 4.494310238285601e-07,
+      "loss": 0.0204,
+      "step": 12995
+    },
+    {
+      "epoch": 2.9536363636363636,
+      "grad_norm": 1.7860377420846998,
+      "learning_rate": 4.493453695589018e-07,
+      "loss": 0.0161,
+      "step": 12996
+    },
+    {
+      "epoch": 2.9538636363636366,
+      "grad_norm": 1.617037896335783,
+      "learning_rate": 4.4925971887114795e-07,
+      "loss": 0.0182,
+      "step": 12997
+    },
+    {
+      "epoch": 2.954090909090909,
+      "grad_norm": 2.2318010930604957,
+      "learning_rate": 4.4917407176704523e-07,
+      "loss": 0.0302,
+      "step": 12998
+    },
+    {
+      "epoch": 2.9543181818181816,
+      "grad_norm": 2.5526158509961676,
+      "learning_rate": 4.490884282483397e-07,
+      "loss": 0.0346,
+      "step": 12999
+    },
+    {
+      "epoch": 2.9545454545454546,
+      "grad_norm": 3.2729938848305973,
+      "learning_rate": 4.49002788316778e-07,
+      "loss": 0.0414,
+      "step": 13000
+    },
+    {
+      "epoch": 2.954772727272727,
+      "grad_norm": 0.6221517281603217,
+      "learning_rate": 4.4891715197410643e-07,
+      "loss": 0.0145,
+      "step": 13001
+    },
+    {
+      "epoch": 2.955,
+      "grad_norm": 1.3966572937896766,
+      "learning_rate": 4.488315192220716e-07,
+      "loss": 0.0116,
+      "step": 13002
+    },
+    {
+      "epoch": 2.9552272727272726,
+      "grad_norm": 2.651273944262639,
+      "learning_rate": 4.487458900624192e-07,
+      "loss": 0.0406,
+      "step": 13003
+    },
+    {
+      "epoch": 2.9554545454545456,
+      "grad_norm": 1.8694960449710663,
+      "learning_rate": 4.4866026449689557e-07,
+      "loss": 0.0126,
+      "step": 13004
+    },
+    {
+      "epoch": 2.955681818181818,
+      "grad_norm": 1.2797124262878428,
+      "learning_rate": 4.4857464252724694e-07,
+      "loss": 0.0182,
+      "step": 13005
+    },
+    {
+      "epoch": 2.955909090909091,
+      "grad_norm": 1.9580622913158323,
+      "learning_rate": 4.4848902415521897e-07,
+      "loss": 0.0374,
+      "step": 13006
+    },
+    {
+      "epoch": 2.9561363636363636,
+      "grad_norm": 2.2693919634739377,
+      "learning_rate": 4.484034093825581e-07,
+      "loss": 0.0428,
+      "step": 13007
+    },
+    {
+      "epoch": 2.9563636363636365,
+      "grad_norm": 1.869816406001042,
+      "learning_rate": 4.483177982110095e-07,
+      "loss": 0.0165,
+      "step": 13008
+    },
+    {
+      "epoch": 2.956590909090909,
+      "grad_norm": 1.735066981736339,
+      "learning_rate": 4.4823219064231926e-07,
+      "loss": 0.0196,
+      "step": 13009
+    },
+    {
+      "epoch": 2.956818181818182,
+      "grad_norm": 1.229329906874586,
+      "learning_rate": 4.4814658667823294e-07,
+      "loss": 0.0122,
+      "step": 13010
+    },
+    {
+      "epoch": 2.9570454545454545,
+      "grad_norm": 0.7485484977751607,
+      "learning_rate": 4.4806098632049655e-07,
+      "loss": 0.0112,
+      "step": 13011
+    },
+    {
+      "epoch": 2.957272727272727,
+      "grad_norm": 1.4429989916346215,
+      "learning_rate": 4.4797538957085507e-07,
+      "loss": 0.0105,
+      "step": 13012
+    },
+    {
+      "epoch": 2.9575,
+      "grad_norm": 2.012565600905032,
+      "learning_rate": 4.4788979643105424e-07,
+      "loss": 0.0175,
+      "step": 13013
+    },
+    {
+      "epoch": 2.957727272727273,
+      "grad_norm": 2.40705775306266,
+      "learning_rate": 4.478042069028396e-07,
+      "loss": 0.0559,
+      "step": 13014
+    },
+    {
+      "epoch": 2.9579545454545455,
+      "grad_norm": 3.8879796421119774,
+      "learning_rate": 4.4771862098795616e-07,
+      "loss": 0.1034,
+      "step": 13015
+    },
+    {
+      "epoch": 2.958181818181818,
+      "grad_norm": 1.1304023951675797,
+      "learning_rate": 4.476330386881496e-07,
+      "loss": 0.0079,
+      "step": 13016
+    },
+    {
+      "epoch": 2.958409090909091,
+      "grad_norm": 1.710088744970627,
+      "learning_rate": 4.4754746000516463e-07,
+      "loss": 0.0251,
+      "step": 13017
+    },
+    {
+      "epoch": 2.9586363636363635,
+      "grad_norm": 2.3494105721221525,
+      "learning_rate": 4.474618849407465e-07,
+      "loss": 0.0451,
+      "step": 13018
+    },
+    {
+      "epoch": 2.9588636363636365,
+      "grad_norm": 2.2253116176814083,
+      "learning_rate": 4.4737631349664024e-07,
+      "loss": 0.025,
+      "step": 13019
+    },
+    {
+      "epoch": 2.959090909090909,
+      "grad_norm": 1.5489100148738941,
+      "learning_rate": 4.47290745674591e-07,
+      "loss": 0.0113,
+      "step": 13020
+    },
+    {
+      "epoch": 2.959318181818182,
+      "grad_norm": 7.103398944540907,
+      "learning_rate": 4.4720518147634325e-07,
+      "loss": 0.0471,
+      "step": 13021
+    },
+    {
+      "epoch": 2.9595454545454545,
+      "grad_norm": 0.9368502819103321,
+      "learning_rate": 4.471196209036421e-07,
+      "loss": 0.0186,
+      "step": 13022
+    },
+    {
+      "epoch": 2.9597727272727274,
+      "grad_norm": 1.5940691617688603,
+      "learning_rate": 4.4703406395823225e-07,
+      "loss": 0.0421,
+      "step": 13023
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.454401749672922,
+      "learning_rate": 4.469485106418585e-07,
+      "loss": 0.0425,
+      "step": 13024
+    },
+    {
+      "epoch": 2.9602272727272725,
+      "grad_norm": 1.2543936679107273,
+      "learning_rate": 4.4686296095626495e-07,
+      "loss": 0.0341,
+      "step": 13025
+    },
+    {
+      "epoch": 2.9604545454545454,
+      "grad_norm": 1.6137238955077824,
+      "learning_rate": 4.4677741490319655e-07,
+      "loss": 0.0208,
+      "step": 13026
+    },
+    {
+      "epoch": 2.9606818181818184,
+      "grad_norm": 1.801545566690045,
+      "learning_rate": 4.466918724843975e-07,
+      "loss": 0.0347,
+      "step": 13027
+    },
+    {
+      "epoch": 2.960909090909091,
+      "grad_norm": 0.987361267981815,
+      "learning_rate": 4.466063337016123e-07,
+      "loss": 0.0163,
+      "step": 13028
+    },
+    {
+      "epoch": 2.9611363636363635,
+      "grad_norm": 1.2759549038323545,
+      "learning_rate": 4.465207985565854e-07,
+      "loss": 0.0106,
+      "step": 13029
+    },
+    {
+      "epoch": 2.9613636363636364,
+      "grad_norm": 1.8185040432942046,
+      "learning_rate": 4.4643526705106047e-07,
+      "loss": 0.0192,
+      "step": 13030
+    },
+    {
+      "epoch": 2.961590909090909,
+      "grad_norm": 2.431102522544093,
+      "learning_rate": 4.4634973918678207e-07,
+      "loss": 0.035,
+      "step": 13031
+    },
+    {
+      "epoch": 2.961818181818182,
+      "grad_norm": 0.903055456091314,
+      "learning_rate": 4.4626421496549417e-07,
+      "loss": 0.0063,
+      "step": 13032
+    },
+    {
+      "epoch": 2.9620454545454544,
+      "grad_norm": 2.3813041166630113,
+      "learning_rate": 4.461786943889408e-07,
+      "loss": 0.0343,
+      "step": 13033
+    },
+    {
+      "epoch": 2.9622727272727274,
+      "grad_norm": 1.898579764446547,
+      "learning_rate": 4.4609317745886573e-07,
+      "loss": 0.0359,
+      "step": 13034
+    },
+    {
+      "epoch": 2.9625,
+      "grad_norm": 2.2578935696899274,
+      "learning_rate": 4.4600766417701295e-07,
+      "loss": 0.0342,
+      "step": 13035
+    },
+    {
+      "epoch": 2.962727272727273,
+      "grad_norm": 1.5230578637505852,
+      "learning_rate": 4.4592215454512604e-07,
+      "loss": 0.033,
+      "step": 13036
+    },
+    {
+      "epoch": 2.9629545454545454,
+      "grad_norm": 1.4949335819821397,
+      "learning_rate": 4.458366485649489e-07,
+      "loss": 0.016,
+      "step": 13037
+    },
+    {
+      "epoch": 2.963181818181818,
+      "grad_norm": 1.403339309362203,
+      "learning_rate": 4.457511462382253e-07,
+      "loss": 0.0103,
+      "step": 13038
+    },
+    {
+      "epoch": 2.963409090909091,
+      "grad_norm": 1.8535987675357666,
+      "learning_rate": 4.456656475666983e-07,
+      "loss": 0.0309,
+      "step": 13039
+    },
+    {
+      "epoch": 2.963636363636364,
+      "grad_norm": 1.8716087319072137,
+      "learning_rate": 4.455801525521116e-07,
+      "loss": 0.0195,
+      "step": 13040
+    },
+    {
+      "epoch": 2.9638636363636364,
+      "grad_norm": 1.8875958478205124,
+      "learning_rate": 4.4549466119620864e-07,
+      "loss": 0.0249,
+      "step": 13041
+    },
+    {
+      "epoch": 2.964090909090909,
+      "grad_norm": 1.962399198009931,
+      "learning_rate": 4.4540917350073277e-07,
+      "loss": 0.0369,
+      "step": 13042
+    },
+    {
+      "epoch": 2.964318181818182,
+      "grad_norm": 1.4536452674682285,
+      "learning_rate": 4.4532368946742704e-07,
+      "loss": 0.0289,
+      "step": 13043
+    },
+    {
+      "epoch": 2.964545454545455,
+      "grad_norm": 1.9519456223917449,
+      "learning_rate": 4.452382090980348e-07,
+      "loss": 0.025,
+      "step": 13044
+    },
+    {
+      "epoch": 2.9647727272727273,
+      "grad_norm": 1.7460581095719512,
+      "learning_rate": 4.45152732394299e-07,
+      "loss": 0.0194,
+      "step": 13045
+    },
+    {
+      "epoch": 2.965,
+      "grad_norm": 1.6827586699956938,
+      "learning_rate": 4.450672593579632e-07,
+      "loss": 0.0149,
+      "step": 13046
+    },
+    {
+      "epoch": 2.965227272727273,
+      "grad_norm": 1.2532917911230372,
+      "learning_rate": 4.449817899907695e-07,
+      "loss": 0.0116,
+      "step": 13047
+    },
+    {
+      "epoch": 2.9654545454545453,
+      "grad_norm": 1.582905695681132,
+      "learning_rate": 4.448963242944611e-07,
+      "loss": 0.0211,
+      "step": 13048
+    },
+    {
+      "epoch": 2.9656818181818183,
+      "grad_norm": 2.1073566928285024,
+      "learning_rate": 4.448108622707808e-07,
+      "loss": 0.0288,
+      "step": 13049
+    },
+    {
+      "epoch": 2.965909090909091,
+      "grad_norm": 1.4453667787031648,
+      "learning_rate": 4.447254039214715e-07,
+      "loss": 0.0138,
+      "step": 13050
+    },
+    {
+      "epoch": 2.966136363636364,
+      "grad_norm": 2.687895631189147,
+      "learning_rate": 4.44639949248276e-07,
+      "loss": 0.0445,
+      "step": 13051
+    },
+    {
+      "epoch": 2.9663636363636363,
+      "grad_norm": 2.614766366835113,
+      "learning_rate": 4.445544982529362e-07,
+      "loss": 0.0266,
+      "step": 13052
+    },
+    {
+      "epoch": 2.9665909090909093,
+      "grad_norm": 1.5307992703345916,
+      "learning_rate": 4.4446905093719516e-07,
+      "loss": 0.0308,
+      "step": 13053
+    },
+    {
+      "epoch": 2.966818181818182,
+      "grad_norm": 1.1786335254455664,
+      "learning_rate": 4.4438360730279503e-07,
+      "loss": 0.0199,
+      "step": 13054
+    },
+    {
+      "epoch": 2.9670454545454543,
+      "grad_norm": 2.0447474875900418,
+      "learning_rate": 4.4429816735147844e-07,
+      "loss": 0.0263,
+      "step": 13055
+    },
+    {
+      "epoch": 2.9672727272727273,
+      "grad_norm": 1.9000490079032413,
+      "learning_rate": 4.4421273108498725e-07,
+      "loss": 0.0363,
+      "step": 13056
+    },
+    {
+      "epoch": 2.9675000000000002,
+      "grad_norm": 3.455332962546853,
+      "learning_rate": 4.44127298505064e-07,
+      "loss": 0.0514,
+      "step": 13057
+    },
+    {
+      "epoch": 2.9677272727272728,
+      "grad_norm": 2.2782718804577153,
+      "learning_rate": 4.4404186961345054e-07,
+      "loss": 0.0328,
+      "step": 13058
+    },
+    {
+      "epoch": 2.9679545454545453,
+      "grad_norm": 1.7857719830347316,
+      "learning_rate": 4.439564444118891e-07,
+      "loss": 0.0217,
+      "step": 13059
+    },
+    {
+      "epoch": 2.9681818181818183,
+      "grad_norm": 2.9669139035800653,
+      "learning_rate": 4.4387102290212187e-07,
+      "loss": 0.0536,
+      "step": 13060
+    },
+    {
+      "epoch": 2.9684090909090908,
+      "grad_norm": 1.1585592039090007,
+      "learning_rate": 4.4378560508589033e-07,
+      "loss": 0.0078,
+      "step": 13061
+    },
+    {
+      "epoch": 2.9686363636363637,
+      "grad_norm": 1.8403304078151626,
+      "learning_rate": 4.437001909649363e-07,
+      "loss": 0.0184,
+      "step": 13062
+    },
+    {
+      "epoch": 2.9688636363636363,
+      "grad_norm": 2.644591366723409,
+      "learning_rate": 4.4361478054100176e-07,
+      "loss": 0.0361,
+      "step": 13063
+    },
+    {
+      "epoch": 2.9690909090909092,
+      "grad_norm": 2.810439539981067,
+      "learning_rate": 4.435293738158284e-07,
+      "loss": 0.0302,
+      "step": 13064
+    },
+    {
+      "epoch": 2.9693181818181817,
+      "grad_norm": 0.6761943334882372,
+      "learning_rate": 4.434439707911575e-07,
+      "loss": 0.0044,
+      "step": 13065
+    },
+    {
+      "epoch": 2.9695454545454547,
+      "grad_norm": 1.341686206165455,
+      "learning_rate": 4.433585714687309e-07,
+      "loss": 0.0168,
+      "step": 13066
+    },
+    {
+      "epoch": 2.9697727272727272,
+      "grad_norm": 0.5798832322196704,
+      "learning_rate": 4.4327317585028984e-07,
+      "loss": 0.0093,
+      "step": 13067
+    },
+    {
+      "epoch": 2.9699999999999998,
+      "grad_norm": 2.012278569864771,
+      "learning_rate": 4.4318778393757605e-07,
+      "loss": 0.025,
+      "step": 13068
+    },
+    {
+      "epoch": 2.9702272727272727,
+      "grad_norm": 0.9966223369079135,
+      "learning_rate": 4.431023957323303e-07,
+      "loss": 0.0088,
+      "step": 13069
+    },
+    {
+      "epoch": 2.9704545454545457,
+      "grad_norm": 1.6820093860631464,
+      "learning_rate": 4.43017011236294e-07,
+      "loss": 0.0273,
+      "step": 13070
+    },
+    {
+      "epoch": 2.970681818181818,
+      "grad_norm": 2.289794223945765,
+      "learning_rate": 4.429316304512083e-07,
+      "loss": 0.034,
+      "step": 13071
+    },
+    {
+      "epoch": 2.9709090909090907,
+      "grad_norm": 1.7033127047275083,
+      "learning_rate": 4.428462533788143e-07,
+      "loss": 0.0192,
+      "step": 13072
+    },
+    {
+      "epoch": 2.9711363636363637,
+      "grad_norm": 2.6683988506869905,
+      "learning_rate": 4.4276088002085315e-07,
+      "loss": 0.0368,
+      "step": 13073
+    },
+    {
+      "epoch": 2.971363636363636,
+      "grad_norm": 2.12086588852245,
+      "learning_rate": 4.4267551037906546e-07,
+      "loss": 0.0172,
+      "step": 13074
+    },
+    {
+      "epoch": 2.971590909090909,
+      "grad_norm": 2.2328904876777913,
+      "learning_rate": 4.425901444551922e-07,
+      "loss": 0.0325,
+      "step": 13075
+    },
+    {
+      "epoch": 2.9718181818181817,
+      "grad_norm": 1.7245046062311977,
+      "learning_rate": 4.4250478225097404e-07,
+      "loss": 0.0196,
+      "step": 13076
+    },
+    {
+      "epoch": 2.9720454545454547,
+      "grad_norm": 3.519076910258693,
+      "learning_rate": 4.4241942376815214e-07,
+      "loss": 0.0471,
+      "step": 13077
+    },
+    {
+      "epoch": 2.972272727272727,
+      "grad_norm": 1.8034662638757928,
+      "learning_rate": 4.423340690084664e-07,
+      "loss": 0.0242,
+      "step": 13078
+    },
+    {
+      "epoch": 2.9725,
+      "grad_norm": 2.283095210526459,
+      "learning_rate": 4.422487179736575e-07,
+      "loss": 0.02,
+      "step": 13079
+    },
+    {
+      "epoch": 2.9727272727272727,
+      "grad_norm": 2.648775679547129,
+      "learning_rate": 4.421633706654662e-07,
+      "loss": 0.0471,
+      "step": 13080
+    },
+    {
+      "epoch": 2.972954545454545,
+      "grad_norm": 2.2029062470326264,
+      "learning_rate": 4.420780270856328e-07,
+      "loss": 0.0162,
+      "step": 13081
+    },
+    {
+      "epoch": 2.973181818181818,
+      "grad_norm": 1.0654978835370006,
+      "learning_rate": 4.4199268723589773e-07,
+      "loss": 0.0203,
+      "step": 13082
+    },
+    {
+      "epoch": 2.973409090909091,
+      "grad_norm": 3.0015966843618083,
+      "learning_rate": 4.419073511180009e-07,
+      "loss": 0.0296,
+      "step": 13083
+    },
+    {
+      "epoch": 2.9736363636363636,
+      "grad_norm": 1.4505566185969814,
+      "learning_rate": 4.4182201873368256e-07,
+      "loss": 0.0073,
+      "step": 13084
+    },
+    {
+      "epoch": 2.973863636363636,
+      "grad_norm": 0.6046091289915898,
+      "learning_rate": 4.4173669008468285e-07,
+      "loss": 0.0074,
+      "step": 13085
+    },
+    {
+      "epoch": 2.974090909090909,
+      "grad_norm": 1.1512691029408495,
+      "learning_rate": 4.416513651727418e-07,
+      "loss": 0.0088,
+      "step": 13086
+    },
+    {
+      "epoch": 2.974318181818182,
+      "grad_norm": 1.2873569203490587,
+      "learning_rate": 4.415660439995992e-07,
+      "loss": 0.0115,
+      "step": 13087
+    },
+    {
+      "epoch": 2.9745454545454546,
+      "grad_norm": 1.7895789803281932,
+      "learning_rate": 4.41480726566995e-07,
+      "loss": 0.026,
+      "step": 13088
+    },
+    {
+      "epoch": 2.974772727272727,
+      "grad_norm": 1.7603197874641205,
+      "learning_rate": 4.4139541287666894e-07,
+      "loss": 0.0235,
+      "step": 13089
+    },
+    {
+      "epoch": 2.975,
+      "grad_norm": 0.6789034406244145,
+      "learning_rate": 4.4131010293036095e-07,
+      "loss": 0.0189,
+      "step": 13090
+    },
+    {
+      "epoch": 2.9752272727272726,
+      "grad_norm": 0.9915936796294972,
+      "learning_rate": 4.4122479672981016e-07,
+      "loss": 0.0113,
+      "step": 13091
+    },
+    {
+      "epoch": 2.9754545454545456,
+      "grad_norm": 1.6392003090526064,
+      "learning_rate": 4.411394942767565e-07,
+      "loss": 0.0258,
+      "step": 13092
+    },
+    {
+      "epoch": 2.975681818181818,
+      "grad_norm": 1.3829462318004833,
+      "learning_rate": 4.4105419557293914e-07,
+      "loss": 0.0071,
+      "step": 13093
+    },
+    {
+      "epoch": 2.975909090909091,
+      "grad_norm": 1.2271126939487946,
+      "learning_rate": 4.409689006200978e-07,
+      "loss": 0.0109,
+      "step": 13094
+    },
+    {
+      "epoch": 2.9761363636363636,
+      "grad_norm": 1.8743272407106983,
+      "learning_rate": 4.4088360941997167e-07,
+      "loss": 0.0194,
+      "step": 13095
+    },
+    {
+      "epoch": 2.9763636363636365,
+      "grad_norm": 1.3491936471583517,
+      "learning_rate": 4.407983219742998e-07,
+      "loss": 0.0112,
+      "step": 13096
+    },
+    {
+      "epoch": 2.976590909090909,
+      "grad_norm": 1.22480823356288,
+      "learning_rate": 4.4071303828482154e-07,
+      "loss": 0.0081,
+      "step": 13097
+    },
+    {
+      "epoch": 2.9768181818181816,
+      "grad_norm": 1.9825240829510191,
+      "learning_rate": 4.40627758353276e-07,
+      "loss": 0.0166,
+      "step": 13098
+    },
+    {
+      "epoch": 2.9770454545454546,
+      "grad_norm": 2.5462538558310377,
+      "learning_rate": 4.405424821814024e-07,
+      "loss": 0.0405,
+      "step": 13099
+    },
+    {
+      "epoch": 2.9772727272727275,
+      "grad_norm": 9.548057575763531,
+      "learning_rate": 4.40457209770939e-07,
+      "loss": 0.0439,
+      "step": 13100
+    },
+    {
+      "epoch": 2.9775,
+      "grad_norm": 2.2325151131351375,
+      "learning_rate": 4.4037194112362506e-07,
+      "loss": 0.0237,
+      "step": 13101
+    },
+    {
+      "epoch": 2.9777272727272726,
+      "grad_norm": 3.9742472085694507,
+      "learning_rate": 4.402866762411994e-07,
+      "loss": 0.0367,
+      "step": 13102
+    },
+    {
+      "epoch": 2.9779545454545455,
+      "grad_norm": 1.5283061584746191,
+      "learning_rate": 4.4020141512540067e-07,
+      "loss": 0.0183,
+      "step": 13103
+    },
+    {
+      "epoch": 2.978181818181818,
+      "grad_norm": 1.0115635946752755,
+      "learning_rate": 4.4011615777796756e-07,
+      "loss": 0.013,
+      "step": 13104
+    },
+    {
+      "epoch": 2.978409090909091,
+      "grad_norm": 1.7526837828696782,
+      "learning_rate": 4.4003090420063844e-07,
+      "loss": 0.0177,
+      "step": 13105
+    },
+    {
+      "epoch": 2.9786363636363635,
+      "grad_norm": 2.394409663253924,
+      "learning_rate": 4.399456543951519e-07,
+      "loss": 0.0474,
+      "step": 13106
+    },
+    {
+      "epoch": 2.9788636363636365,
+      "grad_norm": 1.6321432532441298,
+      "learning_rate": 4.398604083632463e-07,
+      "loss": 0.0156,
+      "step": 13107
+    },
+    {
+      "epoch": 2.979090909090909,
+      "grad_norm": 3.5981553012640024,
+      "learning_rate": 4.397751661066601e-07,
+      "loss": 0.0453,
+      "step": 13108
+    },
+    {
+      "epoch": 2.979318181818182,
+      "grad_norm": 2.7206177435630345,
+      "learning_rate": 4.396899276271313e-07,
+      "loss": 0.023,
+      "step": 13109
+    },
+    {
+      "epoch": 2.9795454545454545,
+      "grad_norm": 1.1523390198815169,
+      "learning_rate": 4.3960469292639805e-07,
+      "loss": 0.0132,
+      "step": 13110
+    },
+    {
+      "epoch": 2.979772727272727,
+      "grad_norm": 1.374966996088673,
+      "learning_rate": 4.3951946200619865e-07,
+      "loss": 0.014,
+      "step": 13111
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.4175551308413539,
+      "learning_rate": 4.394342348682713e-07,
+      "loss": 0.0141,
+      "step": 13112
+    },
+    {
+      "epoch": 2.980227272727273,
+      "grad_norm": 3.3156092099574708,
+      "learning_rate": 4.3934901151435333e-07,
+      "loss": 0.0423,
+      "step": 13113
+    },
+    {
+      "epoch": 2.9804545454545455,
+      "grad_norm": 1.4629537732713813,
+      "learning_rate": 4.3926379194618283e-07,
+      "loss": 0.0171,
+      "step": 13114
+    },
+    {
+      "epoch": 2.980681818181818,
+      "grad_norm": 0.7212070656314739,
+      "learning_rate": 4.3917857616549776e-07,
+      "loss": 0.0096,
+      "step": 13115
+    },
+    {
+      "epoch": 2.980909090909091,
+      "grad_norm": 1.83637006880407,
+      "learning_rate": 4.3909336417403565e-07,
+      "loss": 0.0215,
+      "step": 13116
+    },
+    {
+      "epoch": 2.9811363636363635,
+      "grad_norm": 1.975469010522355,
+      "learning_rate": 4.390081559735344e-07,
+      "loss": 0.0245,
+      "step": 13117
+    },
+    {
+      "epoch": 2.9813636363636364,
+      "grad_norm": 1.8365574797454802,
+      "learning_rate": 4.389229515657311e-07,
+      "loss": 0.0224,
+      "step": 13118
+    },
+    {
+      "epoch": 2.981590909090909,
+      "grad_norm": 1.8849221632192863,
+      "learning_rate": 4.388377509523635e-07,
+      "loss": 0.0208,
+      "step": 13119
+    },
+    {
+      "epoch": 2.981818181818182,
+      "grad_norm": 2.6490900720572643,
+      "learning_rate": 4.3875255413516886e-07,
+      "loss": 0.0168,
+      "step": 13120
+    },
+    {
+      "epoch": 2.9820454545454544,
+      "grad_norm": 1.8026240968833545,
+      "learning_rate": 4.3866736111588486e-07,
+      "loss": 0.0332,
+      "step": 13121
+    },
+    {
+      "epoch": 2.9822727272727274,
+      "grad_norm": 2.4750680200607675,
+      "learning_rate": 4.3858217189624825e-07,
+      "loss": 0.0365,
+      "step": 13122
+    },
+    {
+      "epoch": 2.9825,
+      "grad_norm": 2.1349817509794593,
+      "learning_rate": 4.384969864779963e-07,
+      "loss": 0.0174,
+      "step": 13123
+    },
+    {
+      "epoch": 2.982727272727273,
+      "grad_norm": 2.8187011748264417,
+      "learning_rate": 4.384118048628662e-07,
+      "loss": 0.0178,
+      "step": 13124
+    },
+    {
+      "epoch": 2.9829545454545454,
+      "grad_norm": 1.42579361895096,
+      "learning_rate": 4.383266270525949e-07,
+      "loss": 0.0131,
+      "step": 13125
+    },
+    {
+      "epoch": 2.9831818181818184,
+      "grad_norm": 1.6181398680881534,
+      "learning_rate": 4.3824145304891945e-07,
+      "loss": 0.0159,
+      "step": 13126
+    },
+    {
+      "epoch": 2.983409090909091,
+      "grad_norm": 1.640871232603282,
+      "learning_rate": 4.3815628285357655e-07,
+      "loss": 0.0128,
+      "step": 13127
+    },
+    {
+      "epoch": 2.9836363636363634,
+      "grad_norm": 1.4843251974405638,
+      "learning_rate": 4.3807111646830295e-07,
+      "loss": 0.0175,
+      "step": 13128
+    },
+    {
+      "epoch": 2.9838636363636364,
+      "grad_norm": 2.6011424002456573,
+      "learning_rate": 4.379859538948354e-07,
+      "loss": 0.0394,
+      "step": 13129
+    },
+    {
+      "epoch": 2.9840909090909093,
+      "grad_norm": 2.5415967342253345,
+      "learning_rate": 4.3790079513491074e-07,
+      "loss": 0.0177,
+      "step": 13130
+    },
+    {
+      "epoch": 2.984318181818182,
+      "grad_norm": 0.8890124221214281,
+      "learning_rate": 4.378156401902651e-07,
+      "loss": 0.0123,
+      "step": 13131
+    },
+    {
+      "epoch": 2.9845454545454544,
+      "grad_norm": 2.6803575780147866,
+      "learning_rate": 4.377304890626351e-07,
+      "loss": 0.0281,
+      "step": 13132
+    },
+    {
+      "epoch": 2.9847727272727274,
+      "grad_norm": 2.0173936743860774,
+      "learning_rate": 4.37645341753757e-07,
+      "loss": 0.0235,
+      "step": 13133
+    },
+    {
+      "epoch": 2.985,
+      "grad_norm": 1.3955463502499739,
+      "learning_rate": 4.375601982653675e-07,
+      "loss": 0.0086,
+      "step": 13134
+    },
+    {
+      "epoch": 2.985227272727273,
+      "grad_norm": 2.948794512944229,
+      "learning_rate": 4.374750585992025e-07,
+      "loss": 0.0315,
+      "step": 13135
+    },
+    {
+      "epoch": 2.9854545454545454,
+      "grad_norm": 1.9272604610850006,
+      "learning_rate": 4.373899227569981e-07,
+      "loss": 0.0208,
+      "step": 13136
+    },
+    {
+      "epoch": 2.9856818181818183,
+      "grad_norm": 2.0392941943556773,
+      "learning_rate": 4.373047907404905e-07,
+      "loss": 0.0281,
+      "step": 13137
+    },
+    {
+      "epoch": 2.985909090909091,
+      "grad_norm": 3.1062275259768577,
+      "learning_rate": 4.3721966255141563e-07,
+      "loss": 0.058,
+      "step": 13138
+    },
+    {
+      "epoch": 2.986136363636364,
+      "grad_norm": 2.151898640142011,
+      "learning_rate": 4.371345381915097e-07,
+      "loss": 0.0291,
+      "step": 13139
+    },
+    {
+      "epoch": 2.9863636363636363,
+      "grad_norm": 1.7046241834336895,
+      "learning_rate": 4.3704941766250805e-07,
+      "loss": 0.0227,
+      "step": 13140
+    },
+    {
+      "epoch": 2.986590909090909,
+      "grad_norm": 1.337793562242567,
+      "learning_rate": 4.3696430096614674e-07,
+      "loss": 0.0156,
+      "step": 13141
+    },
+    {
+      "epoch": 2.986818181818182,
+      "grad_norm": 0.9378850322046771,
+      "learning_rate": 4.3687918810416133e-07,
+      "loss": 0.0114,
+      "step": 13142
+    },
+    {
+      "epoch": 2.987045454545455,
+      "grad_norm": 1.4259563126557426,
+      "learning_rate": 4.367940790782877e-07,
+      "loss": 0.0101,
+      "step": 13143
+    },
+    {
+      "epoch": 2.9872727272727273,
+      "grad_norm": 2.047941657292459,
+      "learning_rate": 4.3670897389026093e-07,
+      "loss": 0.0151,
+      "step": 13144
+    },
+    {
+      "epoch": 2.9875,
+      "grad_norm": 5.65421585242195,
+      "learning_rate": 4.366238725418166e-07,
+      "loss": 0.0564,
+      "step": 13145
+    },
+    {
+      "epoch": 2.987727272727273,
+      "grad_norm": 5.148596571795826,
+      "learning_rate": 4.3653877503469025e-07,
+      "loss": 0.0478,
+      "step": 13146
+    },
+    {
+      "epoch": 2.9879545454545453,
+      "grad_norm": 1.1039988218528225,
+      "learning_rate": 4.364536813706172e-07,
+      "loss": 0.0167,
+      "step": 13147
+    },
+    {
+      "epoch": 2.9881818181818183,
+      "grad_norm": 0.8884885098422453,
+      "learning_rate": 4.3636859155133227e-07,
+      "loss": 0.0081,
+      "step": 13148
+    },
+    {
+      "epoch": 2.988409090909091,
+      "grad_norm": 4.130740993724768,
+      "learning_rate": 4.3628350557857093e-07,
+      "loss": 0.047,
+      "step": 13149
+    },
+    {
+      "epoch": 2.9886363636363638,
+      "grad_norm": 1.7357339559505265,
+      "learning_rate": 4.3619842345406816e-07,
+      "loss": 0.0178,
+      "step": 13150
+    },
+    {
+      "epoch": 2.9888636363636363,
+      "grad_norm": 3.0510343372343334,
+      "learning_rate": 4.3611334517955893e-07,
+      "loss": 0.0268,
+      "step": 13151
+    },
+    {
+      "epoch": 2.9890909090909092,
+      "grad_norm": 1.0151115499460066,
+      "learning_rate": 4.3602827075677835e-07,
+      "loss": 0.012,
+      "step": 13152
+    },
+    {
+      "epoch": 2.9893181818181818,
+      "grad_norm": 1.1134450933926088,
+      "learning_rate": 4.359432001874607e-07,
+      "loss": 0.011,
+      "step": 13153
+    },
+    {
+      "epoch": 2.9895454545454543,
+      "grad_norm": 1.164745737121377,
+      "learning_rate": 4.358581334733411e-07,
+      "loss": 0.0094,
+      "step": 13154
+    },
+    {
+      "epoch": 2.9897727272727272,
+      "grad_norm": 2.083641922134021,
+      "learning_rate": 4.3577307061615406e-07,
+      "loss": 0.0272,
+      "step": 13155
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.978296330606882,
+      "learning_rate": 4.3568801161763445e-07,
+      "loss": 0.0352,
+      "step": 13156
+    },
+    {
+      "epoch": 2.9902272727272727,
+      "grad_norm": 2.7955845279213487,
+      "learning_rate": 4.356029564795164e-07,
+      "loss": 0.0455,
+      "step": 13157
+    },
+    {
+      "epoch": 2.9904545454545453,
+      "grad_norm": 2.604425728568055,
+      "learning_rate": 4.3551790520353444e-07,
+      "loss": 0.0475,
+      "step": 13158
+    },
+    {
+      "epoch": 2.990681818181818,
+      "grad_norm": 3.464288808537926,
+      "learning_rate": 4.35432857791423e-07,
+      "loss": 0.0268,
+      "step": 13159
+    },
+    {
+      "epoch": 2.990909090909091,
+      "grad_norm": 2.126153803557652,
+      "learning_rate": 4.3534781424491635e-07,
+      "loss": 0.0303,
+      "step": 13160
+    },
+    {
+      "epoch": 2.9911363636363637,
+      "grad_norm": 1.943462851802946,
+      "learning_rate": 4.352627745657488e-07,
+      "loss": 0.0263,
+      "step": 13161
+    },
+    {
+      "epoch": 2.9913636363636362,
+      "grad_norm": 1.6008848225780048,
+      "learning_rate": 4.351777387556541e-07,
+      "loss": 0.0194,
+      "step": 13162
+    },
+    {
+      "epoch": 2.991590909090909,
+      "grad_norm": 1.9047788551123612,
+      "learning_rate": 4.3509270681636626e-07,
+      "loss": 0.0266,
+      "step": 13163
+    },
+    {
+      "epoch": 2.9918181818181817,
+      "grad_norm": 1.4136898514428564,
+      "learning_rate": 4.350076787496197e-07,
+      "loss": 0.0158,
+      "step": 13164
+    },
+    {
+      "epoch": 2.9920454545454547,
+      "grad_norm": 0.5699990560925323,
+      "learning_rate": 4.349226545571482e-07,
+      "loss": 0.0033,
+      "step": 13165
+    },
+    {
+      "epoch": 2.992272727272727,
+      "grad_norm": 1.4059166843917859,
+      "learning_rate": 4.3483763424068514e-07,
+      "loss": 0.0322,
+      "step": 13166
+    },
+    {
+      "epoch": 2.9925,
+      "grad_norm": 1.8580635373935492,
+      "learning_rate": 4.3475261780196456e-07,
+      "loss": 0.0243,
+      "step": 13167
+    },
+    {
+      "epoch": 2.9927272727272727,
+      "grad_norm": 3.143345588249888,
+      "learning_rate": 4.3466760524271984e-07,
+      "loss": 0.0444,
+      "step": 13168
+    },
+    {
+      "epoch": 2.9929545454545456,
+      "grad_norm": 1.9874070025016535,
+      "learning_rate": 4.3458259656468495e-07,
+      "loss": 0.0267,
+      "step": 13169
+    },
+    {
+      "epoch": 2.993181818181818,
+      "grad_norm": 2.143442752331575,
+      "learning_rate": 4.3449759176959294e-07,
+      "loss": 0.031,
+      "step": 13170
+    },
+    {
+      "epoch": 2.9934090909090907,
+      "grad_norm": 1.9792630974487968,
+      "learning_rate": 4.3441259085917745e-07,
+      "loss": 0.0176,
+      "step": 13171
+    },
+    {
+      "epoch": 2.9936363636363637,
+      "grad_norm": 2.0388152470046976,
+      "learning_rate": 4.343275938351717e-07,
+      "loss": 0.0191,
+      "step": 13172
+    },
+    {
+      "epoch": 2.9938636363636366,
+      "grad_norm": 1.4680843215782073,
+      "learning_rate": 4.342426006993089e-07,
+      "loss": 0.0303,
+      "step": 13173
+    },
+    {
+      "epoch": 2.994090909090909,
+      "grad_norm": 1.472368450079791,
+      "learning_rate": 4.3415761145332254e-07,
+      "loss": 0.0146,
+      "step": 13174
+    },
+    {
+      "epoch": 2.9943181818181817,
+      "grad_norm": 1.5777286159380086,
+      "learning_rate": 4.3407262609894517e-07,
+      "loss": 0.0282,
+      "step": 13175
+    },
+    {
+      "epoch": 2.9945454545454546,
+      "grad_norm": 1.8369944065201929,
+      "learning_rate": 4.3398764463791e-07,
+      "loss": 0.0217,
+      "step": 13176
+    },
+    {
+      "epoch": 2.994772727272727,
+      "grad_norm": 15.105574288185291,
+      "learning_rate": 4.3390266707195013e-07,
+      "loss": 0.0306,
+      "step": 13177
+    },
+    {
+      "epoch": 2.995,
+      "grad_norm": 3.463696864935074,
+      "learning_rate": 4.3381769340279836e-07,
+      "loss": 0.0302,
+      "step": 13178
+    },
+    {
+      "epoch": 2.9952272727272726,
+      "grad_norm": 1.4157885578984382,
+      "learning_rate": 4.3373272363218715e-07,
+      "loss": 0.0133,
+      "step": 13179
+    },
+    {
+      "epoch": 2.9954545454545456,
+      "grad_norm": 1.0450764420704322,
+      "learning_rate": 4.336477577618495e-07,
+      "loss": 0.0137,
+      "step": 13180
+    },
+    {
+      "epoch": 2.995681818181818,
+      "grad_norm": 1.222246997406133,
+      "learning_rate": 4.335627957935178e-07,
+      "loss": 0.0106,
+      "step": 13181
+    },
+    {
+      "epoch": 2.995909090909091,
+      "grad_norm": 2.615558439838587,
+      "learning_rate": 4.334778377289246e-07,
+      "loss": 0.0368,
+      "step": 13182
+    },
+    {
+      "epoch": 2.9961363636363636,
+      "grad_norm": 1.1337921126775685,
+      "learning_rate": 4.3339288356980284e-07,
+      "loss": 0.011,
+      "step": 13183
+    },
+    {
+      "epoch": 2.996363636363636,
+      "grad_norm": 1.649961447524436,
+      "learning_rate": 4.3330793331788413e-07,
+      "loss": 0.0173,
+      "step": 13184
+    },
+    {
+      "epoch": 2.996590909090909,
+      "grad_norm": 1.886029297844586,
+      "learning_rate": 4.3322298697490114e-07,
+      "loss": 0.0187,
+      "step": 13185
+    },
+    {
+      "epoch": 2.996818181818182,
+      "grad_norm": 1.6415226489079353,
+      "learning_rate": 4.33138044542586e-07,
+      "loss": 0.0252,
+      "step": 13186
+    },
+    {
+      "epoch": 2.9970454545454546,
+      "grad_norm": 1.8416865849475803,
+      "learning_rate": 4.330531060226709e-07,
+      "loss": 0.0144,
+      "step": 13187
+    },
+    {
+      "epoch": 2.997272727272727,
+      "grad_norm": 2.106906602559252,
+      "learning_rate": 4.329681714168878e-07,
+      "loss": 0.0161,
+      "step": 13188
+    },
+    {
+      "epoch": 2.9975,
+      "grad_norm": 2.493524690507573,
+      "learning_rate": 4.3288324072696867e-07,
+      "loss": 0.0455,
+      "step": 13189
+    },
+    {
+      "epoch": 2.9977272727272726,
+      "grad_norm": 1.0682878939358231,
+      "learning_rate": 4.3279831395464545e-07,
+      "loss": 0.0112,
+      "step": 13190
+    },
+    {
+      "epoch": 2.9979545454545455,
+      "grad_norm": 5.869784146225063,
+      "learning_rate": 4.3271339110165013e-07,
+      "loss": 0.0572,
+      "step": 13191
+    },
+    {
+      "epoch": 2.998181818181818,
+      "grad_norm": 3.306529062493052,
+      "learning_rate": 4.326284721697139e-07,
+      "loss": 0.0523,
+      "step": 13192
+    },
+    {
+      "epoch": 2.998409090909091,
+      "grad_norm": 3.9531230607010004,
+      "learning_rate": 4.3254355716056877e-07,
+      "loss": 0.0285,
+      "step": 13193
+    },
+    {
+      "epoch": 2.9986363636363635,
+      "grad_norm": 2.0522168485796697,
+      "learning_rate": 4.324586460759464e-07,
+      "loss": 0.0216,
+      "step": 13194
+    },
+    {
+      "epoch": 2.9988636363636365,
+      "grad_norm": 1.7901337877294903,
+      "learning_rate": 4.3237373891757804e-07,
+      "loss": 0.0234,
+      "step": 13195
+    },
+    {
+      "epoch": 2.999090909090909,
+      "grad_norm": 1.320444829676736,
+      "learning_rate": 4.322888356871955e-07,
+      "loss": 0.0164,
+      "step": 13196
+    },
+    {
+      "epoch": 2.9993181818181816,
+      "grad_norm": 0.7205008094799169,
+      "learning_rate": 4.322039363865295e-07,
+      "loss": 0.0058,
+      "step": 13197
+    },
+    {
+      "epoch": 2.9995454545454545,
+      "grad_norm": 1.7468409936739184,
+      "learning_rate": 4.3211904101731166e-07,
+      "loss": 0.0192,
+      "step": 13198
+    },
+    {
+      "epoch": 2.9997727272727275,
+      "grad_norm": 2.6690943921571657,
+      "learning_rate": 4.3203414958127297e-07,
+      "loss": 0.0295,
+      "step": 13199
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.174245555780357,
+      "learning_rate": 4.319492620801449e-07,
+      "loss": 0.0101,
+      "step": 13200
+    },
+    {
+      "epoch": 3.0002272727272725,
+      "grad_norm": 0.6830200128017829,
+      "learning_rate": 4.3186437851565794e-07,
+      "loss": 0.0048,
+      "step": 13201
+    },
+    {
+      "epoch": 3.0004545454545455,
+      "grad_norm": 1.0569337328468333,
+      "learning_rate": 4.3177949888954325e-07,
+      "loss": 0.0041,
+      "step": 13202
+    },
+    {
+      "epoch": 3.000681818181818,
+      "grad_norm": 3.7034853316878045,
+      "learning_rate": 4.316946232035317e-07,
+      "loss": 0.0217,
+      "step": 13203
+    },
+    {
+      "epoch": 3.000909090909091,
+      "grad_norm": 1.535550153542299,
+      "learning_rate": 4.3160975145935394e-07,
+      "loss": 0.0152,
+      "step": 13204
+    },
+    {
+      "epoch": 3.0011363636363635,
+      "grad_norm": 1.3912811812365213,
+      "learning_rate": 4.3152488365874114e-07,
+      "loss": 0.0142,
+      "step": 13205
+    },
+    {
+      "epoch": 3.0013636363636365,
+      "grad_norm": 0.6659780994906302,
+      "learning_rate": 4.314400198034231e-07,
+      "loss": 0.0159,
+      "step": 13206
+    },
+    {
+      "epoch": 3.001590909090909,
+      "grad_norm": 1.2571238529105233,
+      "learning_rate": 4.3135515989513085e-07,
+      "loss": 0.0115,
+      "step": 13207
+    },
+    {
+      "epoch": 3.001818181818182,
+      "grad_norm": 0.6935865866945807,
+      "learning_rate": 4.3127030393559465e-07,
+      "loss": 0.0044,
+      "step": 13208
+    },
+    {
+      "epoch": 3.0020454545454545,
+      "grad_norm": 1.0475850560029667,
+      "learning_rate": 4.311854519265451e-07,
+      "loss": 0.0109,
+      "step": 13209
+    },
+    {
+      "epoch": 3.0022727272727274,
+      "grad_norm": 0.8548356964814834,
+      "learning_rate": 4.3110060386971215e-07,
+      "loss": 0.0076,
+      "step": 13210
+    },
+    {
+      "epoch": 3.0025,
+      "grad_norm": 1.2640939719628266,
+      "learning_rate": 4.3101575976682615e-07,
+      "loss": 0.0142,
+      "step": 13211
+    },
+    {
+      "epoch": 3.002727272727273,
+      "grad_norm": 0.7137743076333426,
+      "learning_rate": 4.309309196196173e-07,
+      "loss": 0.007,
+      "step": 13212
+    },
+    {
+      "epoch": 3.0029545454545454,
+      "grad_norm": 1.122395631507907,
+      "learning_rate": 4.3084608342981575e-07,
+      "loss": 0.0076,
+      "step": 13213
+    },
+    {
+      "epoch": 3.0031818181818184,
+      "grad_norm": 0.5175288441097742,
+      "learning_rate": 4.3076125119915106e-07,
+      "loss": 0.0045,
+      "step": 13214
+    },
+    {
+      "epoch": 3.003409090909091,
+      "grad_norm": 1.2628609712235859,
+      "learning_rate": 4.3067642292935336e-07,
+      "loss": 0.0216,
+      "step": 13215
+    },
+    {
+      "epoch": 3.0036363636363634,
+      "grad_norm": 0.797169724585137,
+      "learning_rate": 4.305915986221524e-07,
+      "loss": 0.0166,
+      "step": 13216
+    },
+    {
+      "epoch": 3.0038636363636364,
+      "grad_norm": 0.9219089902308175,
+      "learning_rate": 4.3050677827927774e-07,
+      "loss": 0.0074,
+      "step": 13217
+    },
+    {
+      "epoch": 3.004090909090909,
+      "grad_norm": 0.49487709540282576,
+      "learning_rate": 4.304219619024596e-07,
+      "loss": 0.0048,
+      "step": 13218
+    },
+    {
+      "epoch": 3.004318181818182,
+      "grad_norm": 1.54472411076556,
+      "learning_rate": 4.303371494934269e-07,
+      "loss": 0.0129,
+      "step": 13219
+    },
+    {
+      "epoch": 3.0045454545454544,
+      "grad_norm": 1.721288918603377,
+      "learning_rate": 4.3025234105390915e-07,
+      "loss": 0.0167,
+      "step": 13220
+    },
+    {
+      "epoch": 3.0047727272727274,
+      "grad_norm": 1.6836797078141938,
+      "learning_rate": 4.3016753658563615e-07,
+      "loss": 0.0079,
+      "step": 13221
+    },
+    {
+      "epoch": 3.005,
+      "grad_norm": 0.8753305047638699,
+      "learning_rate": 4.3008273609033704e-07,
+      "loss": 0.0028,
+      "step": 13222
+    },
+    {
+      "epoch": 3.005227272727273,
+      "grad_norm": 1.1558594011618248,
+      "learning_rate": 4.2999793956974083e-07,
+      "loss": 0.023,
+      "step": 13223
+    },
+    {
+      "epoch": 3.0054545454545454,
+      "grad_norm": 1.6574464178942094,
+      "learning_rate": 4.299131470255768e-07,
+      "loss": 0.0158,
+      "step": 13224
+    },
+    {
+      "epoch": 3.0056818181818183,
+      "grad_norm": 0.6239130714005162,
+      "learning_rate": 4.298283584595741e-07,
+      "loss": 0.0049,
+      "step": 13225
+    },
+    {
+      "epoch": 3.005909090909091,
+      "grad_norm": 0.9421889410750732,
+      "learning_rate": 4.297435738734616e-07,
+      "loss": 0.0062,
+      "step": 13226
+    },
+    {
+      "epoch": 3.006136363636364,
+      "grad_norm": 1.4940082549232832,
+      "learning_rate": 4.296587932689686e-07,
+      "loss": 0.0224,
+      "step": 13227
+    },
+    {
+      "epoch": 3.0063636363636363,
+      "grad_norm": 1.139290285280416,
+      "learning_rate": 4.2957401664782336e-07,
+      "loss": 0.0176,
+      "step": 13228
+    },
+    {
+      "epoch": 3.006590909090909,
+      "grad_norm": 1.3675246538059092,
+      "learning_rate": 4.2948924401175483e-07,
+      "loss": 0.0157,
+      "step": 13229
+    },
+    {
+      "epoch": 3.006818181818182,
+      "grad_norm": 0.6090883614342928,
+      "learning_rate": 4.2940447536249167e-07,
+      "loss": 0.0056,
+      "step": 13230
+    },
+    {
+      "epoch": 3.0070454545454544,
+      "grad_norm": 1.3105242359152625,
+      "learning_rate": 4.2931971070176266e-07,
+      "loss": 0.009,
+      "step": 13231
+    },
+    {
+      "epoch": 3.0072727272727273,
+      "grad_norm": 1.0442409747632209,
+      "learning_rate": 4.2923495003129596e-07,
+      "loss": 0.0088,
+      "step": 13232
+    },
+    {
+      "epoch": 3.0075,
+      "grad_norm": 0.4008030675400975,
+      "learning_rate": 4.291501933528202e-07,
+      "loss": 0.0023,
+      "step": 13233
+    },
+    {
+      "epoch": 3.007727272727273,
+      "grad_norm": 0.9730713547401315,
+      "learning_rate": 4.2906544066806367e-07,
+      "loss": 0.0151,
+      "step": 13234
+    },
+    {
+      "epoch": 3.0079545454545453,
+      "grad_norm": 1.6163652499332226,
+      "learning_rate": 4.2898069197875487e-07,
+      "loss": 0.0096,
+      "step": 13235
+    },
+    {
+      "epoch": 3.0081818181818183,
+      "grad_norm": 1.9158362400360283,
+      "learning_rate": 4.288959472866215e-07,
+      "loss": 0.014,
+      "step": 13236
+    },
+    {
+      "epoch": 3.008409090909091,
+      "grad_norm": 2.3412356391958227,
+      "learning_rate": 4.288112065933919e-07,
+      "loss": 0.0204,
+      "step": 13237
+    },
+    {
+      "epoch": 3.0086363636363638,
+      "grad_norm": 0.9338251969720147,
+      "learning_rate": 4.28726469900794e-07,
+      "loss": 0.0053,
+      "step": 13238
+    },
+    {
+      "epoch": 3.0088636363636363,
+      "grad_norm": 1.026716710811824,
+      "learning_rate": 4.2864173721055583e-07,
+      "loss": 0.0147,
+      "step": 13239
+    },
+    {
+      "epoch": 3.0090909090909093,
+      "grad_norm": 1.2548336886042128,
+      "learning_rate": 4.2855700852440537e-07,
+      "loss": 0.0147,
+      "step": 13240
+    },
+    {
+      "epoch": 3.009318181818182,
+      "grad_norm": 0.6397261116629579,
+      "learning_rate": 4.284722838440701e-07,
+      "loss": 0.0039,
+      "step": 13241
+    },
+    {
+      "epoch": 3.0095454545454547,
+      "grad_norm": 1.3217880020151234,
+      "learning_rate": 4.283875631712778e-07,
+      "loss": 0.0094,
+      "step": 13242
+    },
+    {
+      "epoch": 3.0097727272727273,
+      "grad_norm": 1.5613873189217364,
+      "learning_rate": 4.2830284650775606e-07,
+      "loss": 0.0121,
+      "step": 13243
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 2.072817761770811,
+      "learning_rate": 4.282181338552327e-07,
+      "loss": 0.0218,
+      "step": 13244
+    },
+    {
+      "epoch": 3.0102272727272728,
+      "grad_norm": 0.6735759263405294,
+      "learning_rate": 4.2813342521543467e-07,
+      "loss": 0.0067,
+      "step": 13245
+    },
+    {
+      "epoch": 3.0104545454545453,
+      "grad_norm": 0.7996902882175246,
+      "learning_rate": 4.280487205900894e-07,
+      "loss": 0.0075,
+      "step": 13246
+    },
+    {
+      "epoch": 3.0106818181818182,
+      "grad_norm": 1.0779774906733188,
+      "learning_rate": 4.2796401998092445e-07,
+      "loss": 0.0105,
+      "step": 13247
+    },
+    {
+      "epoch": 3.0109090909090908,
+      "grad_norm": 2.1558881058885695,
+      "learning_rate": 4.2787932338966694e-07,
+      "loss": 0.0094,
+      "step": 13248
+    },
+    {
+      "epoch": 3.0111363636363637,
+      "grad_norm": 1.2559439620443358,
+      "learning_rate": 4.277946308180441e-07,
+      "loss": 0.0079,
+      "step": 13249
+    },
+    {
+      "epoch": 3.0113636363636362,
+      "grad_norm": 0.7185789102613415,
+      "learning_rate": 4.277099422677826e-07,
+      "loss": 0.0051,
+      "step": 13250
+    },
+    {
+      "epoch": 3.011590909090909,
+      "grad_norm": 1.0173190188650072,
+      "learning_rate": 4.2762525774060953e-07,
+      "loss": 0.0067,
+      "step": 13251
+    },
+    {
+      "epoch": 3.0118181818181817,
+      "grad_norm": 2.375263271734533,
+      "learning_rate": 4.2754057723825174e-07,
+      "loss": 0.0194,
+      "step": 13252
+    },
+    {
+      "epoch": 3.0120454545454547,
+      "grad_norm": 1.9481119121462558,
+      "learning_rate": 4.2745590076243627e-07,
+      "loss": 0.0161,
+      "step": 13253
+    },
+    {
+      "epoch": 3.012272727272727,
+      "grad_norm": 2.1678004878653887,
+      "learning_rate": 4.2737122831488937e-07,
+      "loss": 0.0254,
+      "step": 13254
+    },
+    {
+      "epoch": 3.0125,
+      "grad_norm": 1.3400074780602849,
+      "learning_rate": 4.2728655989733804e-07,
+      "loss": 0.0105,
+      "step": 13255
+    },
+    {
+      "epoch": 3.0127272727272727,
+      "grad_norm": 2.075960811658736,
+      "learning_rate": 4.272018955115087e-07,
+      "loss": 0.0169,
+      "step": 13256
+    },
+    {
+      "epoch": 3.0129545454545457,
+      "grad_norm": 2.076533400195187,
+      "learning_rate": 4.2711723515912795e-07,
+      "loss": 0.0105,
+      "step": 13257
+    },
+    {
+      "epoch": 3.013181818181818,
+      "grad_norm": 0.8304598865521866,
+      "learning_rate": 4.2703257884192175e-07,
+      "loss": 0.0053,
+      "step": 13258
+    },
+    {
+      "epoch": 3.0134090909090907,
+      "grad_norm": 1.3583363171676814,
+      "learning_rate": 4.2694792656161666e-07,
+      "loss": 0.0107,
+      "step": 13259
+    },
+    {
+      "epoch": 3.0136363636363637,
+      "grad_norm": 1.2395507340712022,
+      "learning_rate": 4.268632783199389e-07,
+      "loss": 0.0066,
+      "step": 13260
+    },
+    {
+      "epoch": 3.013863636363636,
+      "grad_norm": 1.8898942620938775,
+      "learning_rate": 4.2677863411861456e-07,
+      "loss": 0.0089,
+      "step": 13261
+    },
+    {
+      "epoch": 3.014090909090909,
+      "grad_norm": 1.8731066750620262,
+      "learning_rate": 4.266939939593698e-07,
+      "loss": 0.0052,
+      "step": 13262
+    },
+    {
+      "epoch": 3.0143181818181817,
+      "grad_norm": 1.368119857844622,
+      "learning_rate": 4.2660935784393035e-07,
+      "loss": 0.0103,
+      "step": 13263
+    },
+    {
+      "epoch": 3.0145454545454546,
+      "grad_norm": 0.574943335271676,
+      "learning_rate": 4.2652472577402217e-07,
+      "loss": 0.0026,
+      "step": 13264
+    },
+    {
+      "epoch": 3.014772727272727,
+      "grad_norm": 0.7061528108512672,
+      "learning_rate": 4.264400977513711e-07,
+      "loss": 0.0051,
+      "step": 13265
+    },
+    {
+      "epoch": 3.015,
+      "grad_norm": 0.9994312378091221,
+      "learning_rate": 4.2635547377770315e-07,
+      "loss": 0.0058,
+      "step": 13266
+    },
+    {
+      "epoch": 3.0152272727272726,
+      "grad_norm": 0.8851704339548965,
+      "learning_rate": 4.262708538547433e-07,
+      "loss": 0.006,
+      "step": 13267
+    },
+    {
+      "epoch": 3.0154545454545456,
+      "grad_norm": 1.2029694473193824,
+      "learning_rate": 4.261862379842175e-07,
+      "loss": 0.0084,
+      "step": 13268
+    },
+    {
+      "epoch": 3.015681818181818,
+      "grad_norm": 1.1721633459116865,
+      "learning_rate": 4.261016261678512e-07,
+      "loss": 0.0109,
+      "step": 13269
+    },
+    {
+      "epoch": 3.015909090909091,
+      "grad_norm": 1.1375137095277763,
+      "learning_rate": 4.2601701840736973e-07,
+      "loss": 0.0061,
+      "step": 13270
+    },
+    {
+      "epoch": 3.0161363636363636,
+      "grad_norm": 5.048222036751239,
+      "learning_rate": 4.259324147044984e-07,
+      "loss": 0.0734,
+      "step": 13271
+    },
+    {
+      "epoch": 3.0163636363636366,
+      "grad_norm": 1.872835362821001,
+      "learning_rate": 4.258478150609624e-07,
+      "loss": 0.0073,
+      "step": 13272
+    },
+    {
+      "epoch": 3.016590909090909,
+      "grad_norm": 2.2654367620727056,
+      "learning_rate": 4.2576321947848685e-07,
+      "loss": 0.0144,
+      "step": 13273
+    },
+    {
+      "epoch": 3.0168181818181816,
+      "grad_norm": 2.272573289233585,
+      "learning_rate": 4.256786279587968e-07,
+      "loss": 0.0315,
+      "step": 13274
+    },
+    {
+      "epoch": 3.0170454545454546,
+      "grad_norm": 1.2948691186014079,
+      "learning_rate": 4.255940405036177e-07,
+      "loss": 0.0072,
+      "step": 13275
+    },
+    {
+      "epoch": 3.017272727272727,
+      "grad_norm": 3.3727110152088864,
+      "learning_rate": 4.255094571146735e-07,
+      "loss": 0.0272,
+      "step": 13276
+    },
+    {
+      "epoch": 3.0175,
+      "grad_norm": 1.8521249148783345,
+      "learning_rate": 4.254248777936896e-07,
+      "loss": 0.0211,
+      "step": 13277
+    },
+    {
+      "epoch": 3.0177272727272726,
+      "grad_norm": 1.6938641398348697,
+      "learning_rate": 4.253403025423906e-07,
+      "loss": 0.014,
+      "step": 13278
+    },
+    {
+      "epoch": 3.0179545454545456,
+      "grad_norm": 0.5735105685484073,
+      "learning_rate": 4.252557313625015e-07,
+      "loss": 0.0081,
+      "step": 13279
+    },
+    {
+      "epoch": 3.018181818181818,
+      "grad_norm": 0.31563583715223853,
+      "learning_rate": 4.251711642557462e-07,
+      "loss": 0.0012,
+      "step": 13280
+    },
+    {
+      "epoch": 3.018409090909091,
+      "grad_norm": 1.2116947542825567,
+      "learning_rate": 4.250866012238495e-07,
+      "loss": 0.0062,
+      "step": 13281
+    },
+    {
+      "epoch": 3.0186363636363636,
+      "grad_norm": 3.176400710528372,
+      "learning_rate": 4.2500204226853576e-07,
+      "loss": 0.0243,
+      "step": 13282
+    },
+    {
+      "epoch": 3.0188636363636365,
+      "grad_norm": 1.9127455587569675,
+      "learning_rate": 4.2491748739152935e-07,
+      "loss": 0.0152,
+      "step": 13283
+    },
+    {
+      "epoch": 3.019090909090909,
+      "grad_norm": 0.9671111623437612,
+      "learning_rate": 4.2483293659455443e-07,
+      "loss": 0.0055,
+      "step": 13284
+    },
+    {
+      "epoch": 3.019318181818182,
+      "grad_norm": 1.0215933744612335,
+      "learning_rate": 4.247483898793351e-07,
+      "loss": 0.0074,
+      "step": 13285
+    },
+    {
+      "epoch": 3.0195454545454545,
+      "grad_norm": 1.5500879494977913,
+      "learning_rate": 4.2466384724759543e-07,
+      "loss": 0.0126,
+      "step": 13286
+    },
+    {
+      "epoch": 3.019772727272727,
+      "grad_norm": 1.6033785036912331,
+      "learning_rate": 4.2457930870105946e-07,
+      "loss": 0.0082,
+      "step": 13287
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 1.1597405189150998,
+      "learning_rate": 4.244947742414513e-07,
+      "loss": 0.0077,
+      "step": 13288
+    },
+    {
+      "epoch": 3.0202272727272725,
+      "grad_norm": 1.4830187790980107,
+      "learning_rate": 4.244102438704942e-07,
+      "loss": 0.0116,
+      "step": 13289
+    },
+    {
+      "epoch": 3.0204545454545455,
+      "grad_norm": 0.7265699674556351,
+      "learning_rate": 4.2432571758991225e-07,
+      "loss": 0.0042,
+      "step": 13290
+    },
+    {
+      "epoch": 3.020681818181818,
+      "grad_norm": 0.9954829027725001,
+      "learning_rate": 4.242411954014289e-07,
+      "loss": 0.0076,
+      "step": 13291
+    },
+    {
+      "epoch": 3.020909090909091,
+      "grad_norm": 0.6364675611389962,
+      "learning_rate": 4.2415667730676804e-07,
+      "loss": 0.0041,
+      "step": 13292
+    },
+    {
+      "epoch": 3.0211363636363635,
+      "grad_norm": 3.654589530613306,
+      "learning_rate": 4.2407216330765277e-07,
+      "loss": 0.0274,
+      "step": 13293
+    },
+    {
+      "epoch": 3.0213636363636365,
+      "grad_norm": 1.1701262523675993,
+      "learning_rate": 4.2398765340580653e-07,
+      "loss": 0.0049,
+      "step": 13294
+    },
+    {
+      "epoch": 3.021590909090909,
+      "grad_norm": 1.3492396673592788,
+      "learning_rate": 4.2390314760295284e-07,
+      "loss": 0.0138,
+      "step": 13295
+    },
+    {
+      "epoch": 3.021818181818182,
+      "grad_norm": 1.4477182340528947,
+      "learning_rate": 4.238186459008148e-07,
+      "loss": 0.0072,
+      "step": 13296
+    },
+    {
+      "epoch": 3.0220454545454545,
+      "grad_norm": 1.7386944491303047,
+      "learning_rate": 4.237341483011158e-07,
+      "loss": 0.011,
+      "step": 13297
+    },
+    {
+      "epoch": 3.0222727272727274,
+      "grad_norm": 2.92409662017628,
+      "learning_rate": 4.236496548055783e-07,
+      "loss": 0.0149,
+      "step": 13298
+    },
+    {
+      "epoch": 3.0225,
+      "grad_norm": 0.6589325271633332,
+      "learning_rate": 4.235651654159257e-07,
+      "loss": 0.0029,
+      "step": 13299
+    },
+    {
+      "epoch": 3.022727272727273,
+      "grad_norm": 2.0281798391507064,
+      "learning_rate": 4.2348068013388077e-07,
+      "loss": 0.0117,
+      "step": 13300
+    },
+    {
+      "epoch": 3.0229545454545454,
+      "grad_norm": 0.45172728161201114,
+      "learning_rate": 4.233961989611664e-07,
+      "loss": 0.003,
+      "step": 13301
+    },
+    {
+      "epoch": 3.023181818181818,
+      "grad_norm": 1.1740373961340747,
+      "learning_rate": 4.2331172189950525e-07,
+      "loss": 0.0062,
+      "step": 13302
+    },
+    {
+      "epoch": 3.023409090909091,
+      "grad_norm": 1.3550848091549008,
+      "learning_rate": 4.232272489506199e-07,
+      "loss": 0.0128,
+      "step": 13303
+    },
+    {
+      "epoch": 3.0236363636363635,
+      "grad_norm": 1.306492441039216,
+      "learning_rate": 4.2314278011623297e-07,
+      "loss": 0.0093,
+      "step": 13304
+    },
+    {
+      "epoch": 3.0238636363636364,
+      "grad_norm": 0.9763451321518949,
+      "learning_rate": 4.230583153980669e-07,
+      "loss": 0.0083,
+      "step": 13305
+    },
+    {
+      "epoch": 3.024090909090909,
+      "grad_norm": 1.0767249675638226,
+      "learning_rate": 4.229738547978442e-07,
+      "loss": 0.01,
+      "step": 13306
+    },
+    {
+      "epoch": 3.024318181818182,
+      "grad_norm": 0.6495132504481576,
+      "learning_rate": 4.2288939831728693e-07,
+      "loss": 0.0035,
+      "step": 13307
+    },
+    {
+      "epoch": 3.0245454545454544,
+      "grad_norm": 1.2219724677237642,
+      "learning_rate": 4.2280494595811745e-07,
+      "loss": 0.0034,
+      "step": 13308
+    },
+    {
+      "epoch": 3.0247727272727274,
+      "grad_norm": 1.067697174343913,
+      "learning_rate": 4.227204977220578e-07,
+      "loss": 0.0066,
+      "step": 13309
+    },
+    {
+      "epoch": 3.025,
+      "grad_norm": 1.3756814782990872,
+      "learning_rate": 4.226360536108305e-07,
+      "loss": 0.01,
+      "step": 13310
+    },
+    {
+      "epoch": 3.025227272727273,
+      "grad_norm": 0.8151597682707018,
+      "learning_rate": 4.225516136261568e-07,
+      "loss": 0.0046,
+      "step": 13311
+    },
+    {
+      "epoch": 3.0254545454545454,
+      "grad_norm": 1.5537520224014705,
+      "learning_rate": 4.224671777697588e-07,
+      "loss": 0.0076,
+      "step": 13312
+    },
+    {
+      "epoch": 3.0256818181818184,
+      "grad_norm": 0.6785610907833647,
+      "learning_rate": 4.223827460433584e-07,
+      "loss": 0.0058,
+      "step": 13313
+    },
+    {
+      "epoch": 3.025909090909091,
+      "grad_norm": 1.770094821454295,
+      "learning_rate": 4.222983184486775e-07,
+      "loss": 0.0096,
+      "step": 13314
+    },
+    {
+      "epoch": 3.026136363636364,
+      "grad_norm": 1.0870458677762977,
+      "learning_rate": 4.2221389498743735e-07,
+      "loss": 0.0059,
+      "step": 13315
+    },
+    {
+      "epoch": 3.0263636363636364,
+      "grad_norm": 1.4846286184876767,
+      "learning_rate": 4.221294756613596e-07,
+      "loss": 0.0044,
+      "step": 13316
+    },
+    {
+      "epoch": 3.026590909090909,
+      "grad_norm": 1.213904428873342,
+      "learning_rate": 4.220450604721659e-07,
+      "loss": 0.0054,
+      "step": 13317
+    },
+    {
+      "epoch": 3.026818181818182,
+      "grad_norm": 0.4760945106370274,
+      "learning_rate": 4.2196064942157743e-07,
+      "loss": 0.0031,
+      "step": 13318
+    },
+    {
+      "epoch": 3.0270454545454544,
+      "grad_norm": 1.2839118991637732,
+      "learning_rate": 4.2187624251131577e-07,
+      "loss": 0.0151,
+      "step": 13319
+    },
+    {
+      "epoch": 3.0272727272727273,
+      "grad_norm": 0.8393959612414411,
+      "learning_rate": 4.2179183974310165e-07,
+      "loss": 0.0059,
+      "step": 13320
+    },
+    {
+      "epoch": 3.0275,
+      "grad_norm": 1.0460669661127229,
+      "learning_rate": 4.217074411186564e-07,
+      "loss": 0.0096,
+      "step": 13321
+    },
+    {
+      "epoch": 3.027727272727273,
+      "grad_norm": 1.577933327542485,
+      "learning_rate": 4.2162304663970117e-07,
+      "loss": 0.0142,
+      "step": 13322
+    },
+    {
+      "epoch": 3.0279545454545453,
+      "grad_norm": 0.7079586394037386,
+      "learning_rate": 4.2153865630795693e-07,
+      "loss": 0.0065,
+      "step": 13323
+    },
+    {
+      "epoch": 3.0281818181818183,
+      "grad_norm": 1.1666351271167192,
+      "learning_rate": 4.214542701251443e-07,
+      "loss": 0.0085,
+      "step": 13324
+    },
+    {
+      "epoch": 3.028409090909091,
+      "grad_norm": 0.39707655767234246,
+      "learning_rate": 4.2136988809298424e-07,
+      "loss": 0.0023,
+      "step": 13325
+    },
+    {
+      "epoch": 3.028636363636364,
+      "grad_norm": 0.9547792295481499,
+      "learning_rate": 4.212855102131974e-07,
+      "loss": 0.0051,
+      "step": 13326
+    },
+    {
+      "epoch": 3.0288636363636363,
+      "grad_norm": 2.543634239987202,
+      "learning_rate": 4.2120113648750434e-07,
+      "loss": 0.017,
+      "step": 13327
+    },
+    {
+      "epoch": 3.0290909090909093,
+      "grad_norm": 4.758014420162185,
+      "learning_rate": 4.2111676691762597e-07,
+      "loss": 0.0265,
+      "step": 13328
+    },
+    {
+      "epoch": 3.029318181818182,
+      "grad_norm": 0.7675587528825861,
+      "learning_rate": 4.210324015052822e-07,
+      "loss": 0.0053,
+      "step": 13329
+    },
+    {
+      "epoch": 3.0295454545454548,
+      "grad_norm": 0.4702151969740731,
+      "learning_rate": 4.209480402521934e-07,
+      "loss": 0.0019,
+      "step": 13330
+    },
+    {
+      "epoch": 3.0297727272727273,
+      "grad_norm": 1.5124727713371557,
+      "learning_rate": 4.2086368316008017e-07,
+      "loss": 0.02,
+      "step": 13331
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 1.4451364355127394,
+      "learning_rate": 4.2077933023066284e-07,
+      "loss": 0.0072,
+      "step": 13332
+    },
+    {
+      "epoch": 3.0302272727272728,
+      "grad_norm": 0.9487771168878743,
+      "learning_rate": 4.206949814656611e-07,
+      "loss": 0.0039,
+      "step": 13333
+    },
+    {
+      "epoch": 3.0304545454545453,
+      "grad_norm": 1.94252703419527,
+      "learning_rate": 4.2061063686679506e-07,
+      "loss": 0.0145,
+      "step": 13334
+    },
+    {
+      "epoch": 3.0306818181818183,
+      "grad_norm": 0.7724029834558229,
+      "learning_rate": 4.2052629643578467e-07,
+      "loss": 0.0078,
+      "step": 13335
+    },
+    {
+      "epoch": 3.0309090909090908,
+      "grad_norm": 0.7921982117682748,
+      "learning_rate": 4.204419601743499e-07,
+      "loss": 0.0066,
+      "step": 13336
+    },
+    {
+      "epoch": 3.0311363636363637,
+      "grad_norm": 2.034620679666645,
+      "learning_rate": 4.203576280842104e-07,
+      "loss": 0.008,
+      "step": 13337
+    },
+    {
+      "epoch": 3.0313636363636363,
+      "grad_norm": 1.594275581398991,
+      "learning_rate": 4.202733001670858e-07,
+      "loss": 0.0104,
+      "step": 13338
+    },
+    {
+      "epoch": 3.0315909090909092,
+      "grad_norm": 1.6159192988876678,
+      "learning_rate": 4.2018897642469586e-07,
+      "loss": 0.0136,
+      "step": 13339
+    },
+    {
+      "epoch": 3.0318181818181817,
+      "grad_norm": 1.1720818728356541,
+      "learning_rate": 4.2010465685875997e-07,
+      "loss": 0.0096,
+      "step": 13340
+    },
+    {
+      "epoch": 3.0320454545454547,
+      "grad_norm": 0.588035015053218,
+      "learning_rate": 4.200203414709979e-07,
+      "loss": 0.0026,
+      "step": 13341
+    },
+    {
+      "epoch": 3.0322727272727272,
+      "grad_norm": 0.7817002424231815,
+      "learning_rate": 4.1993603026312834e-07,
+      "loss": 0.0052,
+      "step": 13342
+    },
+    {
+      "epoch": 3.0325,
+      "grad_norm": 0.3851895707173541,
+      "learning_rate": 4.198517232368709e-07,
+      "loss": 0.0026,
+      "step": 13343
+    },
+    {
+      "epoch": 3.0327272727272727,
+      "grad_norm": 0.8630097191054225,
+      "learning_rate": 4.1976742039394467e-07,
+      "loss": 0.0032,
+      "step": 13344
+    },
+    {
+      "epoch": 3.0329545454545452,
+      "grad_norm": 1.0039347246107995,
+      "learning_rate": 4.19683121736069e-07,
+      "loss": 0.0097,
+      "step": 13345
+    },
+    {
+      "epoch": 3.033181818181818,
+      "grad_norm": 0.8257735992414195,
+      "learning_rate": 4.1959882726496245e-07,
+      "loss": 0.0116,
+      "step": 13346
+    },
+    {
+      "epoch": 3.0334090909090907,
+      "grad_norm": 1.2404877503432932,
+      "learning_rate": 4.1951453698234426e-07,
+      "loss": 0.0053,
+      "step": 13347
+    },
+    {
+      "epoch": 3.0336363636363637,
+      "grad_norm": 2.2315103236492293,
+      "learning_rate": 4.194302508899331e-07,
+      "loss": 0.0212,
+      "step": 13348
+    },
+    {
+      "epoch": 3.033863636363636,
+      "grad_norm": 1.5055247960959148,
+      "learning_rate": 4.1934596898944776e-07,
+      "loss": 0.006,
+      "step": 13349
+    },
+    {
+      "epoch": 3.034090909090909,
+      "grad_norm": 0.7215807072687465,
+      "learning_rate": 4.1926169128260715e-07,
+      "loss": 0.0031,
+      "step": 13350
+    },
+    {
+      "epoch": 3.0343181818181817,
+      "grad_norm": 0.7968232610873153,
+      "learning_rate": 4.1917741777112934e-07,
+      "loss": 0.0116,
+      "step": 13351
+    },
+    {
+      "epoch": 3.0345454545454547,
+      "grad_norm": 0.992642186368362,
+      "learning_rate": 4.1909314845673304e-07,
+      "loss": 0.0045,
+      "step": 13352
+    },
+    {
+      "epoch": 3.034772727272727,
+      "grad_norm": 1.5179916820040056,
+      "learning_rate": 4.1900888334113666e-07,
+      "loss": 0.01,
+      "step": 13353
+    },
+    {
+      "epoch": 3.035,
+      "grad_norm": 1.9479261622790254,
+      "learning_rate": 4.1892462242605877e-07,
+      "loss": 0.0094,
+      "step": 13354
+    },
+    {
+      "epoch": 3.0352272727272727,
+      "grad_norm": 3.1366330856636737,
+      "learning_rate": 4.18840365713217e-07,
+      "loss": 0.0086,
+      "step": 13355
+    },
+    {
+      "epoch": 3.0354545454545456,
+      "grad_norm": 0.4934643435432161,
+      "learning_rate": 4.1875611320433e-07,
+      "loss": 0.0059,
+      "step": 13356
+    },
+    {
+      "epoch": 3.035681818181818,
+      "grad_norm": 1.0740114396412515,
+      "learning_rate": 4.1867186490111565e-07,
+      "loss": 0.0076,
+      "step": 13357
+    },
+    {
+      "epoch": 3.035909090909091,
+      "grad_norm": 1.405491968139043,
+      "learning_rate": 4.185876208052921e-07,
+      "loss": 0.0153,
+      "step": 13358
+    },
+    {
+      "epoch": 3.0361363636363636,
+      "grad_norm": 2.303193555088964,
+      "learning_rate": 4.185033809185769e-07,
+      "loss": 0.0067,
+      "step": 13359
+    },
+    {
+      "epoch": 3.036363636363636,
+      "grad_norm": 1.6137980202027966,
+      "learning_rate": 4.1841914524268784e-07,
+      "loss": 0.0227,
+      "step": 13360
+    },
+    {
+      "epoch": 3.036590909090909,
+      "grad_norm": 1.3216803131433938,
+      "learning_rate": 4.18334913779343e-07,
+      "loss": 0.013,
+      "step": 13361
+    },
+    {
+      "epoch": 3.0368181818181816,
+      "grad_norm": 1.0954052771470604,
+      "learning_rate": 4.1825068653025975e-07,
+      "loss": 0.0058,
+      "step": 13362
+    },
+    {
+      "epoch": 3.0370454545454546,
+      "grad_norm": 0.8071619768011777,
+      "learning_rate": 4.181664634971559e-07,
+      "loss": 0.0067,
+      "step": 13363
+    },
+    {
+      "epoch": 3.037272727272727,
+      "grad_norm": 2.1511897428267432,
+      "learning_rate": 4.1808224468174853e-07,
+      "loss": 0.0123,
+      "step": 13364
+    },
+    {
+      "epoch": 3.0375,
+      "grad_norm": 1.9003325444905983,
+      "learning_rate": 4.179980300857552e-07,
+      "loss": 0.0104,
+      "step": 13365
+    },
+    {
+      "epoch": 3.0377272727272726,
+      "grad_norm": 1.1345705214989195,
+      "learning_rate": 4.179138197108931e-07,
+      "loss": 0.0115,
+      "step": 13366
+    },
+    {
+      "epoch": 3.0379545454545456,
+      "grad_norm": 2.6495543029567923,
+      "learning_rate": 4.178296135588797e-07,
+      "loss": 0.0253,
+      "step": 13367
+    },
+    {
+      "epoch": 3.038181818181818,
+      "grad_norm": 1.6430884272124964,
+      "learning_rate": 4.177454116314318e-07,
+      "loss": 0.0118,
+      "step": 13368
+    },
+    {
+      "epoch": 3.038409090909091,
+      "grad_norm": 2.8983535286107682,
+      "learning_rate": 4.176612139302665e-07,
+      "loss": 0.0192,
+      "step": 13369
+    },
+    {
+      "epoch": 3.0386363636363636,
+      "grad_norm": 2.272252489148228,
+      "learning_rate": 4.175770204571007e-07,
+      "loss": 0.0206,
+      "step": 13370
+    },
+    {
+      "epoch": 3.0388636363636365,
+      "grad_norm": 1.3688160974692032,
+      "learning_rate": 4.1749283121365137e-07,
+      "loss": 0.0228,
+      "step": 13371
+    },
+    {
+      "epoch": 3.039090909090909,
+      "grad_norm": 1.6403961097203201,
+      "learning_rate": 4.174086462016355e-07,
+      "loss": 0.0153,
+      "step": 13372
+    },
+    {
+      "epoch": 3.039318181818182,
+      "grad_norm": 1.5462982957471685,
+      "learning_rate": 4.1732446542276916e-07,
+      "loss": 0.014,
+      "step": 13373
+    },
+    {
+      "epoch": 3.0395454545454546,
+      "grad_norm": 0.664808029696496,
+      "learning_rate": 4.172402888787694e-07,
+      "loss": 0.0068,
+      "step": 13374
+    },
+    {
+      "epoch": 3.039772727272727,
+      "grad_norm": 1.5486913065416266,
+      "learning_rate": 4.171561165713525e-07,
+      "loss": 0.0061,
+      "step": 13375
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 2.473512866934722,
+      "learning_rate": 4.1707194850223515e-07,
+      "loss": 0.0287,
+      "step": 13376
+    },
+    {
+      "epoch": 3.0402272727272726,
+      "grad_norm": 0.731989199853223,
+      "learning_rate": 4.169877846731334e-07,
+      "loss": 0.0038,
+      "step": 13377
+    },
+    {
+      "epoch": 3.0404545454545455,
+      "grad_norm": 0.2996048267485832,
+      "learning_rate": 4.169036250857637e-07,
+      "loss": 0.0045,
+      "step": 13378
+    },
+    {
+      "epoch": 3.040681818181818,
+      "grad_norm": 1.7467593601369376,
+      "learning_rate": 4.16819469741842e-07,
+      "loss": 0.0117,
+      "step": 13379
+    },
+    {
+      "epoch": 3.040909090909091,
+      "grad_norm": 1.0912106734666709,
+      "learning_rate": 4.167353186430848e-07,
+      "loss": 0.0081,
+      "step": 13380
+    },
+    {
+      "epoch": 3.0411363636363635,
+      "grad_norm": 1.5700869521027017,
+      "learning_rate": 4.166511717912076e-07,
+      "loss": 0.0075,
+      "step": 13381
+    },
+    {
+      "epoch": 3.0413636363636365,
+      "grad_norm": 3.4857007896585737,
+      "learning_rate": 4.165670291879264e-07,
+      "loss": 0.0444,
+      "step": 13382
+    },
+    {
+      "epoch": 3.041590909090909,
+      "grad_norm": 1.9367608136276242,
+      "learning_rate": 4.164828908349572e-07,
+      "loss": 0.0145,
+      "step": 13383
+    },
+    {
+      "epoch": 3.041818181818182,
+      "grad_norm": 0.9042308380082255,
+      "learning_rate": 4.1639875673401555e-07,
+      "loss": 0.0073,
+      "step": 13384
+    },
+    {
+      "epoch": 3.0420454545454545,
+      "grad_norm": 1.4011173536252852,
+      "learning_rate": 4.1631462688681734e-07,
+      "loss": 0.0077,
+      "step": 13385
+    },
+    {
+      "epoch": 3.0422727272727275,
+      "grad_norm": 0.39696936433321134,
+      "learning_rate": 4.162305012950778e-07,
+      "loss": 0.0014,
+      "step": 13386
+    },
+    {
+      "epoch": 3.0425,
+      "grad_norm": 1.107483181872778,
+      "learning_rate": 4.1614637996051257e-07,
+      "loss": 0.0136,
+      "step": 13387
+    },
+    {
+      "epoch": 3.042727272727273,
+      "grad_norm": 2.3290748431049813,
+      "learning_rate": 4.16062262884837e-07,
+      "loss": 0.0101,
+      "step": 13388
+    },
+    {
+      "epoch": 3.0429545454545455,
+      "grad_norm": 1.8310799216458902,
+      "learning_rate": 4.1597815006976657e-07,
+      "loss": 0.0119,
+      "step": 13389
+    },
+    {
+      "epoch": 3.043181818181818,
+      "grad_norm": 2.4155679629596354,
+      "learning_rate": 4.1589404151701615e-07,
+      "loss": 0.0152,
+      "step": 13390
+    },
+    {
+      "epoch": 3.043409090909091,
+      "grad_norm": 2.8339772375377947,
+      "learning_rate": 4.158099372283011e-07,
+      "loss": 0.0466,
+      "step": 13391
+    },
+    {
+      "epoch": 3.0436363636363635,
+      "grad_norm": 1.7431553753119011,
+      "learning_rate": 4.157258372053363e-07,
+      "loss": 0.0133,
+      "step": 13392
+    },
+    {
+      "epoch": 3.0438636363636364,
+      "grad_norm": 1.037320181974709,
+      "learning_rate": 4.1564174144983707e-07,
+      "loss": 0.0079,
+      "step": 13393
+    },
+    {
+      "epoch": 3.044090909090909,
+      "grad_norm": 1.1265174986326996,
+      "learning_rate": 4.1555764996351776e-07,
+      "loss": 0.0115,
+      "step": 13394
+    },
+    {
+      "epoch": 3.044318181818182,
+      "grad_norm": 0.5385967023680895,
+      "learning_rate": 4.1547356274809326e-07,
+      "loss": 0.0027,
+      "step": 13395
+    },
+    {
+      "epoch": 3.0445454545454544,
+      "grad_norm": 2.527656565255388,
+      "learning_rate": 4.153894798052785e-07,
+      "loss": 0.0097,
+      "step": 13396
+    },
+    {
+      "epoch": 3.0447727272727274,
+      "grad_norm": 1.4210661832924523,
+      "learning_rate": 4.1530540113678785e-07,
+      "loss": 0.0097,
+      "step": 13397
+    },
+    {
+      "epoch": 3.045,
+      "grad_norm": 1.3971078687532537,
+      "learning_rate": 4.15221326744336e-07,
+      "loss": 0.0129,
+      "step": 13398
+    },
+    {
+      "epoch": 3.045227272727273,
+      "grad_norm": 1.7432440443957231,
+      "learning_rate": 4.1513725662963717e-07,
+      "loss": 0.0129,
+      "step": 13399
+    },
+    {
+      "epoch": 3.0454545454545454,
+      "grad_norm": 1.344500193404308,
+      "learning_rate": 4.150531907944058e-07,
+      "loss": 0.0081,
+      "step": 13400
+    },
+    {
+      "epoch": 3.0456818181818184,
+      "grad_norm": 1.7984499348614584,
+      "learning_rate": 4.149691292403561e-07,
+      "loss": 0.0136,
+      "step": 13401
+    },
+    {
+      "epoch": 3.045909090909091,
+      "grad_norm": 0.6581078894009353,
+      "learning_rate": 4.1488507196920267e-07,
+      "loss": 0.0054,
+      "step": 13402
+    },
+    {
+      "epoch": 3.0461363636363634,
+      "grad_norm": 1.8842207684303944,
+      "learning_rate": 4.148010189826588e-07,
+      "loss": 0.0046,
+      "step": 13403
+    },
+    {
+      "epoch": 3.0463636363636364,
+      "grad_norm": 0.728404832825665,
+      "learning_rate": 4.147169702824389e-07,
+      "loss": 0.0031,
+      "step": 13404
+    },
+    {
+      "epoch": 3.046590909090909,
+      "grad_norm": 1.0928171668949465,
+      "learning_rate": 4.1463292587025687e-07,
+      "loss": 0.0081,
+      "step": 13405
+    },
+    {
+      "epoch": 3.046818181818182,
+      "grad_norm": 1.4644754650533487,
+      "learning_rate": 4.145488857478265e-07,
+      "loss": 0.0141,
+      "step": 13406
+    },
+    {
+      "epoch": 3.0470454545454544,
+      "grad_norm": 0.9483361052722754,
+      "learning_rate": 4.144648499168616e-07,
+      "loss": 0.0098,
+      "step": 13407
+    },
+    {
+      "epoch": 3.0472727272727274,
+      "grad_norm": 1.7618816568220654,
+      "learning_rate": 4.1438081837907555e-07,
+      "loss": 0.0149,
+      "step": 13408
+    },
+    {
+      "epoch": 3.0475,
+      "grad_norm": 1.4341113679681032,
+      "learning_rate": 4.1429679113618204e-07,
+      "loss": 0.0107,
+      "step": 13409
+    },
+    {
+      "epoch": 3.047727272727273,
+      "grad_norm": 0.9411151525908193,
+      "learning_rate": 4.1421276818989455e-07,
+      "loss": 0.0052,
+      "step": 13410
+    },
+    {
+      "epoch": 3.0479545454545454,
+      "grad_norm": 0.7700716489663607,
+      "learning_rate": 4.1412874954192684e-07,
+      "loss": 0.0042,
+      "step": 13411
+    },
+    {
+      "epoch": 3.0481818181818183,
+      "grad_norm": 2.3678115701362774,
+      "learning_rate": 4.140447351939915e-07,
+      "loss": 0.0153,
+      "step": 13412
+    },
+    {
+      "epoch": 3.048409090909091,
+      "grad_norm": 2.92417306194958,
+      "learning_rate": 4.1396072514780203e-07,
+      "loss": 0.0189,
+      "step": 13413
+    },
+    {
+      "epoch": 3.048636363636364,
+      "grad_norm": 1.8974491007409922,
+      "learning_rate": 4.1387671940507133e-07,
+      "loss": 0.0182,
+      "step": 13414
+    },
+    {
+      "epoch": 3.0488636363636363,
+      "grad_norm": 2.774569563913146,
+      "learning_rate": 4.1379271796751307e-07,
+      "loss": 0.0216,
+      "step": 13415
+    },
+    {
+      "epoch": 3.0490909090909093,
+      "grad_norm": 2.7470175581721037,
+      "learning_rate": 4.1370872083683957e-07,
+      "loss": 0.0227,
+      "step": 13416
+    },
+    {
+      "epoch": 3.049318181818182,
+      "grad_norm": 2.220431794399996,
+      "learning_rate": 4.1362472801476386e-07,
+      "loss": 0.0356,
+      "step": 13417
+    },
+    {
+      "epoch": 3.0495454545454543,
+      "grad_norm": 3.3378934068541066,
+      "learning_rate": 4.135407395029987e-07,
+      "loss": 0.0467,
+      "step": 13418
+    },
+    {
+      "epoch": 3.0497727272727273,
+      "grad_norm": 1.6346864552230733,
+      "learning_rate": 4.1345675530325673e-07,
+      "loss": 0.0056,
+      "step": 13419
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 1.4357312527660298,
+      "learning_rate": 4.1337277541725077e-07,
+      "loss": 0.0187,
+      "step": 13420
+    },
+    {
+      "epoch": 3.050227272727273,
+      "grad_norm": 2.199189416893518,
+      "learning_rate": 4.132887998466929e-07,
+      "loss": 0.0167,
+      "step": 13421
+    },
+    {
+      "epoch": 3.0504545454545453,
+      "grad_norm": 1.9807209428855248,
+      "learning_rate": 4.132048285932959e-07,
+      "loss": 0.0226,
+      "step": 13422
+    },
+    {
+      "epoch": 3.0506818181818183,
+      "grad_norm": 0.7029731338928166,
+      "learning_rate": 4.131208616587718e-07,
+      "loss": 0.0037,
+      "step": 13423
+    },
+    {
+      "epoch": 3.050909090909091,
+      "grad_norm": 1.0481266269313565,
+      "learning_rate": 4.130368990448333e-07,
+      "loss": 0.0161,
+      "step": 13424
+    },
+    {
+      "epoch": 3.0511363636363638,
+      "grad_norm": 0.9742577170713057,
+      "learning_rate": 4.1295294075319194e-07,
+      "loss": 0.0089,
+      "step": 13425
+    },
+    {
+      "epoch": 3.0513636363636363,
+      "grad_norm": 0.3624107623194293,
+      "learning_rate": 4.128689867855601e-07,
+      "loss": 0.0022,
+      "step": 13426
+    },
+    {
+      "epoch": 3.0515909090909092,
+      "grad_norm": 2.038121764070667,
+      "learning_rate": 4.1278503714364967e-07,
+      "loss": 0.0142,
+      "step": 13427
+    },
+    {
+      "epoch": 3.0518181818181818,
+      "grad_norm": 0.43241840704389956,
+      "learning_rate": 4.1270109182917254e-07,
+      "loss": 0.0026,
+      "step": 13428
+    },
+    {
+      "epoch": 3.0520454545454547,
+      "grad_norm": 1.1035131935249778,
+      "learning_rate": 4.126171508438406e-07,
+      "loss": 0.0123,
+      "step": 13429
+    },
+    {
+      "epoch": 3.0522727272727272,
+      "grad_norm": 3.3744951720339187,
+      "learning_rate": 4.125332141893655e-07,
+      "loss": 0.0165,
+      "step": 13430
+    },
+    {
+      "epoch": 3.0525,
+      "grad_norm": 2.0655030878818312,
+      "learning_rate": 4.124492818674587e-07,
+      "loss": 0.0133,
+      "step": 13431
+    },
+    {
+      "epoch": 3.0527272727272727,
+      "grad_norm": 1.683850769839554,
+      "learning_rate": 4.12365353879832e-07,
+      "loss": 0.015,
+      "step": 13432
+    },
+    {
+      "epoch": 3.0529545454545453,
+      "grad_norm": 1.2279307054135007,
+      "learning_rate": 4.1228143022819675e-07,
+      "loss": 0.0094,
+      "step": 13433
+    },
+    {
+      "epoch": 3.053181818181818,
+      "grad_norm": 1.4255254836967244,
+      "learning_rate": 4.121975109142641e-07,
+      "loss": 0.0095,
+      "step": 13434
+    },
+    {
+      "epoch": 3.0534090909090907,
+      "grad_norm": 2.414066687558891,
+      "learning_rate": 4.121135959397455e-07,
+      "loss": 0.0131,
+      "step": 13435
+    },
+    {
+      "epoch": 3.0536363636363637,
+      "grad_norm": 4.391160477966368,
+      "learning_rate": 4.12029685306352e-07,
+      "loss": 0.0204,
+      "step": 13436
+    },
+    {
+      "epoch": 3.0538636363636362,
+      "grad_norm": 0.7845884073194624,
+      "learning_rate": 4.119457790157949e-07,
+      "loss": 0.0047,
+      "step": 13437
+    },
+    {
+      "epoch": 3.054090909090909,
+      "grad_norm": 1.9488661935006069,
+      "learning_rate": 4.1186187706978505e-07,
+      "loss": 0.0278,
+      "step": 13438
+    },
+    {
+      "epoch": 3.0543181818181817,
+      "grad_norm": 1.3869230730289004,
+      "learning_rate": 4.1177797947003327e-07,
+      "loss": 0.0123,
+      "step": 13439
+    },
+    {
+      "epoch": 3.0545454545454547,
+      "grad_norm": 2.2114877761205944,
+      "learning_rate": 4.1169408621825043e-07,
+      "loss": 0.0229,
+      "step": 13440
+    },
+    {
+      "epoch": 3.054772727272727,
+      "grad_norm": 1.032806579292452,
+      "learning_rate": 4.1161019731614736e-07,
+      "loss": 0.0069,
+      "step": 13441
+    },
+    {
+      "epoch": 3.055,
+      "grad_norm": 1.285619177758557,
+      "learning_rate": 4.115263127654349e-07,
+      "loss": 0.01,
+      "step": 13442
+    },
+    {
+      "epoch": 3.0552272727272727,
+      "grad_norm": 1.0796766245204357,
+      "learning_rate": 4.1144243256782295e-07,
+      "loss": 0.0071,
+      "step": 13443
+    },
+    {
+      "epoch": 3.0554545454545456,
+      "grad_norm": 1.1954239486368665,
+      "learning_rate": 4.113585567250225e-07,
+      "loss": 0.0074,
+      "step": 13444
+    },
+    {
+      "epoch": 3.055681818181818,
+      "grad_norm": 1.4673800570476208,
+      "learning_rate": 4.1127468523874376e-07,
+      "loss": 0.0144,
+      "step": 13445
+    },
+    {
+      "epoch": 3.0559090909090907,
+      "grad_norm": 0.8244004821761725,
+      "learning_rate": 4.1119081811069743e-07,
+      "loss": 0.009,
+      "step": 13446
+    },
+    {
+      "epoch": 3.0561363636363637,
+      "grad_norm": 4.0554496452338675,
+      "learning_rate": 4.1110695534259296e-07,
+      "loss": 0.022,
+      "step": 13447
+    },
+    {
+      "epoch": 3.056363636363636,
+      "grad_norm": 2.5115116971069638,
+      "learning_rate": 4.110230969361409e-07,
+      "loss": 0.0181,
+      "step": 13448
+    },
+    {
+      "epoch": 3.056590909090909,
+      "grad_norm": 3.1286271980899363,
+      "learning_rate": 4.1093924289305114e-07,
+      "loss": 0.0299,
+      "step": 13449
+    },
+    {
+      "epoch": 3.0568181818181817,
+      "grad_norm": 1.0124731637508206,
+      "learning_rate": 4.1085539321503374e-07,
+      "loss": 0.0065,
+      "step": 13450
+    },
+    {
+      "epoch": 3.0570454545454546,
+      "grad_norm": 1.6605202336095315,
+      "learning_rate": 4.107715479037985e-07,
+      "loss": 0.0137,
+      "step": 13451
+    },
+    {
+      "epoch": 3.057272727272727,
+      "grad_norm": 1.8028437154480752,
+      "learning_rate": 4.10687706961055e-07,
+      "loss": 0.0216,
+      "step": 13452
+    },
+    {
+      "epoch": 3.0575,
+      "grad_norm": 2.035732893986176,
+      "learning_rate": 4.106038703885132e-07,
+      "loss": 0.0113,
+      "step": 13453
+    },
+    {
+      "epoch": 3.0577272727272726,
+      "grad_norm": 0.6038531441345725,
+      "learning_rate": 4.1052003818788237e-07,
+      "loss": 0.0034,
+      "step": 13454
+    },
+    {
+      "epoch": 3.0579545454545456,
+      "grad_norm": 1.3073056193575945,
+      "learning_rate": 4.104362103608725e-07,
+      "loss": 0.017,
+      "step": 13455
+    },
+    {
+      "epoch": 3.058181818181818,
+      "grad_norm": 1.3133167714131146,
+      "learning_rate": 4.103523869091923e-07,
+      "loss": 0.0077,
+      "step": 13456
+    },
+    {
+      "epoch": 3.058409090909091,
+      "grad_norm": 1.2848591071324713,
+      "learning_rate": 4.1026856783455136e-07,
+      "loss": 0.0074,
+      "step": 13457
+    },
+    {
+      "epoch": 3.0586363636363636,
+      "grad_norm": 0.800326811448739,
+      "learning_rate": 4.1018475313865905e-07,
+      "loss": 0.0078,
+      "step": 13458
+    },
+    {
+      "epoch": 3.0588636363636366,
+      "grad_norm": 1.1100708001004098,
+      "learning_rate": 4.101009428232244e-07,
+      "loss": 0.0042,
+      "step": 13459
+    },
+    {
+      "epoch": 3.059090909090909,
+      "grad_norm": 1.0587273512253255,
+      "learning_rate": 4.1001713688995645e-07,
+      "loss": 0.0081,
+      "step": 13460
+    },
+    {
+      "epoch": 3.0593181818181816,
+      "grad_norm": 1.364041339853722,
+      "learning_rate": 4.0993333534056404e-07,
+      "loss": 0.01,
+      "step": 13461
+    },
+    {
+      "epoch": 3.0595454545454546,
+      "grad_norm": 0.8050800694040438,
+      "learning_rate": 4.0984953817675613e-07,
+      "loss": 0.006,
+      "step": 13462
+    },
+    {
+      "epoch": 3.059772727272727,
+      "grad_norm": 0.5925593769372894,
+      "learning_rate": 4.0976574540024156e-07,
+      "loss": 0.0032,
+      "step": 13463
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.7130666135164836,
+      "learning_rate": 4.0968195701272916e-07,
+      "loss": 0.0068,
+      "step": 13464
+    },
+    {
+      "epoch": 3.0602272727272726,
+      "grad_norm": 4.33826672062044,
+      "learning_rate": 4.095981730159269e-07,
+      "loss": 0.0491,
+      "step": 13465
+    },
+    {
+      "epoch": 3.0604545454545455,
+      "grad_norm": 0.8069885202918122,
+      "learning_rate": 4.095143934115439e-07,
+      "loss": 0.0036,
+      "step": 13466
+    },
+    {
+      "epoch": 3.060681818181818,
+      "grad_norm": 1.2278122061035641,
+      "learning_rate": 4.094306182012883e-07,
+      "loss": 0.0096,
+      "step": 13467
+    },
+    {
+      "epoch": 3.060909090909091,
+      "grad_norm": 1.135175644621067,
+      "learning_rate": 4.093468473868685e-07,
+      "loss": 0.011,
+      "step": 13468
+    },
+    {
+      "epoch": 3.0611363636363635,
+      "grad_norm": 2.881248725610049,
+      "learning_rate": 4.092630809699927e-07,
+      "loss": 0.0179,
+      "step": 13469
+    },
+    {
+      "epoch": 3.0613636363636365,
+      "grad_norm": 1.4571579814205062,
+      "learning_rate": 4.0917931895236894e-07,
+      "loss": 0.0082,
+      "step": 13470
+    },
+    {
+      "epoch": 3.061590909090909,
+      "grad_norm": 0.7201577342380838,
+      "learning_rate": 4.090955613357055e-07,
+      "loss": 0.0056,
+      "step": 13471
+    },
+    {
+      "epoch": 3.061818181818182,
+      "grad_norm": 2.4775546101038146,
+      "learning_rate": 4.090118081217102e-07,
+      "loss": 0.0213,
+      "step": 13472
+    },
+    {
+      "epoch": 3.0620454545454545,
+      "grad_norm": 1.1193193288554526,
+      "learning_rate": 4.0892805931209103e-07,
+      "loss": 0.0164,
+      "step": 13473
+    },
+    {
+      "epoch": 3.0622727272727275,
+      "grad_norm": 0.5678068606749695,
+      "learning_rate": 4.088443149085555e-07,
+      "loss": 0.0027,
+      "step": 13474
+    },
+    {
+      "epoch": 3.0625,
+      "grad_norm": 0.8081847221850784,
+      "learning_rate": 4.0876057491281165e-07,
+      "loss": 0.0043,
+      "step": 13475
+    },
+    {
+      "epoch": 3.0627272727272725,
+      "grad_norm": 0.8105694520101983,
+      "learning_rate": 4.086768393265668e-07,
+      "loss": 0.0046,
+      "step": 13476
+    },
+    {
+      "epoch": 3.0629545454545455,
+      "grad_norm": 0.9728868508553483,
+      "learning_rate": 4.0859310815152893e-07,
+      "loss": 0.0106,
+      "step": 13477
+    },
+    {
+      "epoch": 3.063181818181818,
+      "grad_norm": 1.7354901828227456,
+      "learning_rate": 4.0850938138940484e-07,
+      "loss": 0.0158,
+      "step": 13478
+    },
+    {
+      "epoch": 3.063409090909091,
+      "grad_norm": 1.5937240058423854,
+      "learning_rate": 4.084256590419022e-07,
+      "loss": 0.0098,
+      "step": 13479
+    },
+    {
+      "epoch": 3.0636363636363635,
+      "grad_norm": 1.9731275461736018,
+      "learning_rate": 4.0834194111072813e-07,
+      "loss": 0.0161,
+      "step": 13480
+    },
+    {
+      "epoch": 3.0638636363636365,
+      "grad_norm": 0.8753306632079465,
+      "learning_rate": 4.0825822759759006e-07,
+      "loss": 0.0066,
+      "step": 13481
+    },
+    {
+      "epoch": 3.064090909090909,
+      "grad_norm": 0.9158275475847595,
+      "learning_rate": 4.0817451850419463e-07,
+      "loss": 0.0062,
+      "step": 13482
+    },
+    {
+      "epoch": 3.064318181818182,
+      "grad_norm": 4.3012450371494415,
+      "learning_rate": 4.080908138322491e-07,
+      "loss": 0.0129,
+      "step": 13483
+    },
+    {
+      "epoch": 3.0645454545454545,
+      "grad_norm": 0.2902641711994485,
+      "learning_rate": 4.0800711358346024e-07,
+      "loss": 0.0016,
+      "step": 13484
+    },
+    {
+      "epoch": 3.0647727272727274,
+      "grad_norm": 0.6797508832789304,
+      "learning_rate": 4.07923417759535e-07,
+      "loss": 0.003,
+      "step": 13485
+    },
+    {
+      "epoch": 3.065,
+      "grad_norm": 0.3917987943067327,
+      "learning_rate": 4.078397263621802e-07,
+      "loss": 0.0023,
+      "step": 13486
+    },
+    {
+      "epoch": 3.065227272727273,
+      "grad_norm": 0.5196370533180509,
+      "learning_rate": 4.07756039393102e-07,
+      "loss": 0.0042,
+      "step": 13487
+    },
+    {
+      "epoch": 3.0654545454545454,
+      "grad_norm": 1.6413182360645153,
+      "learning_rate": 4.076723568540071e-07,
+      "loss": 0.0087,
+      "step": 13488
+    },
+    {
+      "epoch": 3.0656818181818184,
+      "grad_norm": 3.5209323897450635,
+      "learning_rate": 4.07588678746602e-07,
+      "loss": 0.0281,
+      "step": 13489
+    },
+    {
+      "epoch": 3.065909090909091,
+      "grad_norm": 1.0931130474813588,
+      "learning_rate": 4.075050050725932e-07,
+      "loss": 0.008,
+      "step": 13490
+    },
+    {
+      "epoch": 3.0661363636363634,
+      "grad_norm": 3.3350500744802076,
+      "learning_rate": 4.074213358336867e-07,
+      "loss": 0.0272,
+      "step": 13491
+    },
+    {
+      "epoch": 3.0663636363636364,
+      "grad_norm": 2.428977958701881,
+      "learning_rate": 4.0733767103158863e-07,
+      "loss": 0.0291,
+      "step": 13492
+    },
+    {
+      "epoch": 3.066590909090909,
+      "grad_norm": 3.3481263308171285,
+      "learning_rate": 4.0725401066800527e-07,
+      "loss": 0.0104,
+      "step": 13493
+    },
+    {
+      "epoch": 3.066818181818182,
+      "grad_norm": 1.7921053195159482,
+      "learning_rate": 4.0717035474464247e-07,
+      "loss": 0.0078,
+      "step": 13494
+    },
+    {
+      "epoch": 3.0670454545454544,
+      "grad_norm": 1.3289841492113197,
+      "learning_rate": 4.070867032632064e-07,
+      "loss": 0.0176,
+      "step": 13495
+    },
+    {
+      "epoch": 3.0672727272727274,
+      "grad_norm": 0.7049919995606067,
+      "learning_rate": 4.070030562254025e-07,
+      "loss": 0.0049,
+      "step": 13496
+    },
+    {
+      "epoch": 3.0675,
+      "grad_norm": 2.0809442901729773,
+      "learning_rate": 4.0691941363293646e-07,
+      "loss": 0.013,
+      "step": 13497
+    },
+    {
+      "epoch": 3.067727272727273,
+      "grad_norm": 1.273087970109055,
+      "learning_rate": 4.068357754875139e-07,
+      "loss": 0.0102,
+      "step": 13498
+    },
+    {
+      "epoch": 3.0679545454545454,
+      "grad_norm": 0.9625162180203619,
+      "learning_rate": 4.06752141790841e-07,
+      "loss": 0.006,
+      "step": 13499
+    },
+    {
+      "epoch": 3.0681818181818183,
+      "grad_norm": 1.1890147089068068,
+      "learning_rate": 4.0666851254462236e-07,
+      "loss": 0.0209,
+      "step": 13500
+    },
+    {
+      "epoch": 3.068409090909091,
+      "grad_norm": 2.8037391278654176,
+      "learning_rate": 4.0658488775056365e-07,
+      "loss": 0.0166,
+      "step": 13501
+    },
+    {
+      "epoch": 3.068636363636364,
+      "grad_norm": 3.3739420655393575,
+      "learning_rate": 4.065012674103701e-07,
+      "loss": 0.0373,
+      "step": 13502
+    },
+    {
+      "epoch": 3.0688636363636363,
+      "grad_norm": 3.0095361483308043,
+      "learning_rate": 4.06417651525747e-07,
+      "loss": 0.0336,
+      "step": 13503
+    },
+    {
+      "epoch": 3.0690909090909093,
+      "grad_norm": 2.190843652759724,
+      "learning_rate": 4.0633404009839924e-07,
+      "loss": 0.011,
+      "step": 13504
+    },
+    {
+      "epoch": 3.069318181818182,
+      "grad_norm": 1.2552116839127236,
+      "learning_rate": 4.062504331300319e-07,
+      "loss": 0.0064,
+      "step": 13505
+    },
+    {
+      "epoch": 3.0695454545454544,
+      "grad_norm": 1.2044854996429144,
+      "learning_rate": 4.0616683062234987e-07,
+      "loss": 0.0084,
+      "step": 13506
+    },
+    {
+      "epoch": 3.0697727272727273,
+      "grad_norm": 1.1215272972452894,
+      "learning_rate": 4.0608323257705784e-07,
+      "loss": 0.0095,
+      "step": 13507
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.9331932037753126,
+      "learning_rate": 4.0599963899586094e-07,
+      "loss": 0.0052,
+      "step": 13508
+    },
+    {
+      "epoch": 3.070227272727273,
+      "grad_norm": 1.7472318922461358,
+      "learning_rate": 4.0591604988046326e-07,
+      "loss": 0.0144,
+      "step": 13509
+    },
+    {
+      "epoch": 3.0704545454545453,
+      "grad_norm": 3.154988989414998,
+      "learning_rate": 4.0583246523256944e-07,
+      "loss": 0.0332,
+      "step": 13510
+    },
+    {
+      "epoch": 3.0706818181818183,
+      "grad_norm": 0.6736760437638778,
+      "learning_rate": 4.05748885053884e-07,
+      "loss": 0.0059,
+      "step": 13511
+    },
+    {
+      "epoch": 3.070909090909091,
+      "grad_norm": 1.093225390584526,
+      "learning_rate": 4.0566530934611157e-07,
+      "loss": 0.0114,
+      "step": 13512
+    },
+    {
+      "epoch": 3.0711363636363638,
+      "grad_norm": 1.1276762711795545,
+      "learning_rate": 4.0558173811095593e-07,
+      "loss": 0.01,
+      "step": 13513
+    },
+    {
+      "epoch": 3.0713636363636363,
+      "grad_norm": 0.5420819901705335,
+      "learning_rate": 4.0549817135012145e-07,
+      "loss": 0.0037,
+      "step": 13514
+    },
+    {
+      "epoch": 3.0715909090909093,
+      "grad_norm": 1.0432832005215098,
+      "learning_rate": 4.0541460906531214e-07,
+      "loss": 0.0055,
+      "step": 13515
+    },
+    {
+      "epoch": 3.071818181818182,
+      "grad_norm": 1.0126992937800554,
+      "learning_rate": 4.0533105125823244e-07,
+      "loss": 0.0119,
+      "step": 13516
+    },
+    {
+      "epoch": 3.0720454545454547,
+      "grad_norm": 0.9880773583161974,
+      "learning_rate": 4.052474979305854e-07,
+      "loss": 0.0047,
+      "step": 13517
+    },
+    {
+      "epoch": 3.0722727272727273,
+      "grad_norm": 1.2638849747197924,
+      "learning_rate": 4.0516394908407537e-07,
+      "loss": 0.0105,
+      "step": 13518
+    },
+    {
+      "epoch": 3.0725,
+      "grad_norm": 1.2970094486106996,
+      "learning_rate": 4.050804047204059e-07,
+      "loss": 0.0085,
+      "step": 13519
+    },
+    {
+      "epoch": 3.0727272727272728,
+      "grad_norm": 1.7091113597987821,
+      "learning_rate": 4.049968648412807e-07,
+      "loss": 0.0141,
+      "step": 13520
+    },
+    {
+      "epoch": 3.0729545454545453,
+      "grad_norm": 1.5338309958245626,
+      "learning_rate": 4.049133294484033e-07,
+      "loss": 0.0102,
+      "step": 13521
+    },
+    {
+      "epoch": 3.0731818181818182,
+      "grad_norm": 2.059423480632646,
+      "learning_rate": 4.0482979854347695e-07,
+      "loss": 0.0146,
+      "step": 13522
+    },
+    {
+      "epoch": 3.0734090909090908,
+      "grad_norm": 1.2916528139964036,
+      "learning_rate": 4.047462721282052e-07,
+      "loss": 0.0068,
+      "step": 13523
+    },
+    {
+      "epoch": 3.0736363636363637,
+      "grad_norm": 1.764446640610024,
+      "learning_rate": 4.0466275020429116e-07,
+      "loss": 0.0205,
+      "step": 13524
+    },
+    {
+      "epoch": 3.0738636363636362,
+      "grad_norm": 0.5433508230093456,
+      "learning_rate": 4.0457923277343834e-07,
+      "loss": 0.003,
+      "step": 13525
+    },
+    {
+      "epoch": 3.074090909090909,
+      "grad_norm": 1.6287027546484216,
+      "learning_rate": 4.044957198373492e-07,
+      "loss": 0.0299,
+      "step": 13526
+    },
+    {
+      "epoch": 3.0743181818181817,
+      "grad_norm": 0.45323665637276583,
+      "learning_rate": 4.044122113977269e-07,
+      "loss": 0.0025,
+      "step": 13527
+    },
+    {
+      "epoch": 3.0745454545454547,
+      "grad_norm": 1.3902781452319188,
+      "learning_rate": 4.043287074562746e-07,
+      "loss": 0.0157,
+      "step": 13528
+    },
+    {
+      "epoch": 3.074772727272727,
+      "grad_norm": 0.8266540998529932,
+      "learning_rate": 4.042452080146949e-07,
+      "loss": 0.0061,
+      "step": 13529
+    },
+    {
+      "epoch": 3.075,
+      "grad_norm": 2.391746931005692,
+      "learning_rate": 4.0416171307469084e-07,
+      "loss": 0.0227,
+      "step": 13530
+    },
+    {
+      "epoch": 3.0752272727272727,
+      "grad_norm": 1.2722375810034126,
+      "learning_rate": 4.040782226379644e-07,
+      "loss": 0.0069,
+      "step": 13531
+    },
+    {
+      "epoch": 3.0754545454545457,
+      "grad_norm": 1.6517668675169814,
+      "learning_rate": 4.0399473670621847e-07,
+      "loss": 0.0097,
+      "step": 13532
+    },
+    {
+      "epoch": 3.075681818181818,
+      "grad_norm": 1.7792181234021718,
+      "learning_rate": 4.039112552811554e-07,
+      "loss": 0.0098,
+      "step": 13533
+    },
+    {
+      "epoch": 3.0759090909090907,
+      "grad_norm": 1.6604056264801432,
+      "learning_rate": 4.038277783644777e-07,
+      "loss": 0.0176,
+      "step": 13534
+    },
+    {
+      "epoch": 3.0761363636363637,
+      "grad_norm": 0.9468878150692722,
+      "learning_rate": 4.037443059578873e-07,
+      "loss": 0.0077,
+      "step": 13535
+    },
+    {
+      "epoch": 3.076363636363636,
+      "grad_norm": 0.9431893522836888,
+      "learning_rate": 4.036608380630865e-07,
+      "loss": 0.0033,
+      "step": 13536
+    },
+    {
+      "epoch": 3.076590909090909,
+      "grad_norm": 0.6188086139973992,
+      "learning_rate": 4.035773746817774e-07,
+      "loss": 0.0022,
+      "step": 13537
+    },
+    {
+      "epoch": 3.0768181818181817,
+      "grad_norm": 0.647957544773077,
+      "learning_rate": 4.034939158156621e-07,
+      "loss": 0.0031,
+      "step": 13538
+    },
+    {
+      "epoch": 3.0770454545454546,
+      "grad_norm": 1.638202753448472,
+      "learning_rate": 4.034104614664422e-07,
+      "loss": 0.0105,
+      "step": 13539
+    },
+    {
+      "epoch": 3.077272727272727,
+      "grad_norm": 2.2215470891194085,
+      "learning_rate": 4.033270116358194e-07,
+      "loss": 0.0135,
+      "step": 13540
+    },
+    {
+      "epoch": 3.0775,
+      "grad_norm": 1.676077883423364,
+      "learning_rate": 4.032435663254956e-07,
+      "loss": 0.0098,
+      "step": 13541
+    },
+    {
+      "epoch": 3.0777272727272726,
+      "grad_norm": 4.61238464352617,
+      "learning_rate": 4.031601255371724e-07,
+      "loss": 0.0099,
+      "step": 13542
+    },
+    {
+      "epoch": 3.0779545454545456,
+      "grad_norm": 1.6450590144913566,
+      "learning_rate": 4.030766892725514e-07,
+      "loss": 0.0059,
+      "step": 13543
+    },
+    {
+      "epoch": 3.078181818181818,
+      "grad_norm": 1.7654635078228176,
+      "learning_rate": 4.0299325753333376e-07,
+      "loss": 0.0168,
+      "step": 13544
+    },
+    {
+      "epoch": 3.078409090909091,
+      "grad_norm": 1.3147858945948685,
+      "learning_rate": 4.029098303212209e-07,
+      "loss": 0.0128,
+      "step": 13545
+    },
+    {
+      "epoch": 3.0786363636363636,
+      "grad_norm": 0.9935929680756946,
+      "learning_rate": 4.02826407637914e-07,
+      "loss": 0.0089,
+      "step": 13546
+    },
+    {
+      "epoch": 3.0788636363636366,
+      "grad_norm": 2.668350198806893,
+      "learning_rate": 4.027429894851145e-07,
+      "loss": 0.0248,
+      "step": 13547
+    },
+    {
+      "epoch": 3.079090909090909,
+      "grad_norm": 3.8613060211462518,
+      "learning_rate": 4.02659575864523e-07,
+      "loss": 0.0339,
+      "step": 13548
+    },
+    {
+      "epoch": 3.0793181818181816,
+      "grad_norm": 1.0560541333461928,
+      "learning_rate": 4.0257616677784064e-07,
+      "loss": 0.0042,
+      "step": 13549
+    },
+    {
+      "epoch": 3.0795454545454546,
+      "grad_norm": 1.0570063269223904,
+      "learning_rate": 4.0249276222676826e-07,
+      "loss": 0.0042,
+      "step": 13550
+    },
+    {
+      "epoch": 3.079772727272727,
+      "grad_norm": 1.2367353207435887,
+      "learning_rate": 4.024093622130067e-07,
+      "loss": 0.0174,
+      "step": 13551
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 1.7430688907276077,
+      "learning_rate": 4.023259667382566e-07,
+      "loss": 0.0067,
+      "step": 13552
+    },
+    {
+      "epoch": 3.0802272727272726,
+      "grad_norm": 1.5244570270222975,
+      "learning_rate": 4.022425758042184e-07,
+      "loss": 0.0307,
+      "step": 13553
+    },
+    {
+      "epoch": 3.0804545454545456,
+      "grad_norm": 0.8968582321446642,
+      "learning_rate": 4.0215918941259273e-07,
+      "loss": 0.005,
+      "step": 13554
+    },
+    {
+      "epoch": 3.080681818181818,
+      "grad_norm": 2.860668196369034,
+      "learning_rate": 4.020758075650799e-07,
+      "loss": 0.0134,
+      "step": 13555
+    },
+    {
+      "epoch": 3.080909090909091,
+      "grad_norm": 0.5195147583588019,
+      "learning_rate": 4.0199243026338065e-07,
+      "loss": 0.0026,
+      "step": 13556
+    },
+    {
+      "epoch": 3.0811363636363636,
+      "grad_norm": 1.2017001376929732,
+      "learning_rate": 4.0190905750919433e-07,
+      "loss": 0.0131,
+      "step": 13557
+    },
+    {
+      "epoch": 3.0813636363636365,
+      "grad_norm": 0.8932930212949112,
+      "learning_rate": 4.018256893042217e-07,
+      "loss": 0.0093,
+      "step": 13558
+    },
+    {
+      "epoch": 3.081590909090909,
+      "grad_norm": 1.6496996493294422,
+      "learning_rate": 4.0174232565016254e-07,
+      "loss": 0.007,
+      "step": 13559
+    },
+    {
+      "epoch": 3.081818181818182,
+      "grad_norm": 1.5994044083476369,
+      "learning_rate": 4.0165896654871723e-07,
+      "loss": 0.0122,
+      "step": 13560
+    },
+    {
+      "epoch": 3.0820454545454545,
+      "grad_norm": 1.330786731553534,
+      "learning_rate": 4.0157561200158484e-07,
+      "loss": 0.0118,
+      "step": 13561
+    },
+    {
+      "epoch": 3.082272727272727,
+      "grad_norm": 1.6702007275431736,
+      "learning_rate": 4.014922620104656e-07,
+      "loss": 0.0111,
+      "step": 13562
+    },
+    {
+      "epoch": 3.0825,
+      "grad_norm": 0.8866064804312951,
+      "learning_rate": 4.0140891657705907e-07,
+      "loss": 0.0033,
+      "step": 13563
+    },
+    {
+      "epoch": 3.0827272727272725,
+      "grad_norm": 0.35588099601065776,
+      "learning_rate": 4.013255757030647e-07,
+      "loss": 0.0022,
+      "step": 13564
+    },
+    {
+      "epoch": 3.0829545454545455,
+      "grad_norm": 0.9495447614887693,
+      "learning_rate": 4.0124223939018237e-07,
+      "loss": 0.005,
+      "step": 13565
+    },
+    {
+      "epoch": 3.083181818181818,
+      "grad_norm": 1.5244606619510659,
+      "learning_rate": 4.011589076401109e-07,
+      "loss": 0.0137,
+      "step": 13566
+    },
+    {
+      "epoch": 3.083409090909091,
+      "grad_norm": 0.3512212196355577,
+      "learning_rate": 4.010755804545498e-07,
+      "loss": 0.0013,
+      "step": 13567
+    },
+    {
+      "epoch": 3.0836363636363635,
+      "grad_norm": 0.8864815736250822,
+      "learning_rate": 4.009922578351983e-07,
+      "loss": 0.0088,
+      "step": 13568
+    },
+    {
+      "epoch": 3.0838636363636365,
+      "grad_norm": 1.783985844949969,
+      "learning_rate": 4.009089397837558e-07,
+      "loss": 0.0084,
+      "step": 13569
+    },
+    {
+      "epoch": 3.084090909090909,
+      "grad_norm": 1.0506333933328895,
+      "learning_rate": 4.008256263019206e-07,
+      "loss": 0.0047,
+      "step": 13570
+    },
+    {
+      "epoch": 3.084318181818182,
+      "grad_norm": 3.030396986447236,
+      "learning_rate": 4.0074231739139197e-07,
+      "loss": 0.0113,
+      "step": 13571
+    },
+    {
+      "epoch": 3.0845454545454545,
+      "grad_norm": 1.5170449064627167,
+      "learning_rate": 4.006590130538688e-07,
+      "loss": 0.0175,
+      "step": 13572
+    },
+    {
+      "epoch": 3.0847727272727274,
+      "grad_norm": 1.6672823492801068,
+      "learning_rate": 4.005757132910497e-07,
+      "loss": 0.0094,
+      "step": 13573
+    },
+    {
+      "epoch": 3.085,
+      "grad_norm": 1.0469874675227302,
+      "learning_rate": 4.004924181046335e-07,
+      "loss": 0.0069,
+      "step": 13574
+    },
+    {
+      "epoch": 3.085227272727273,
+      "grad_norm": 2.42859290930973,
+      "learning_rate": 4.0040912749631846e-07,
+      "loss": 0.0092,
+      "step": 13575
+    },
+    {
+      "epoch": 3.0854545454545454,
+      "grad_norm": 1.9847911140814836,
+      "learning_rate": 4.0032584146780305e-07,
+      "loss": 0.02,
+      "step": 13576
+    },
+    {
+      "epoch": 3.085681818181818,
+      "grad_norm": 3.2892179458909494,
+      "learning_rate": 4.002425600207857e-07,
+      "loss": 0.0129,
+      "step": 13577
+    },
+    {
+      "epoch": 3.085909090909091,
+      "grad_norm": 1.8775219084101613,
+      "learning_rate": 4.0015928315696494e-07,
+      "loss": 0.0237,
+      "step": 13578
+    },
+    {
+      "epoch": 3.0861363636363635,
+      "grad_norm": 4.3115913860819575,
+      "learning_rate": 4.0007601087803843e-07,
+      "loss": 0.0157,
+      "step": 13579
+    },
+    {
+      "epoch": 3.0863636363636364,
+      "grad_norm": 2.529175782852872,
+      "learning_rate": 3.9999274318570437e-07,
+      "loss": 0.0128,
+      "step": 13580
+    },
+    {
+      "epoch": 3.086590909090909,
+      "grad_norm": 3.3048336884827214,
+      "learning_rate": 3.9990948008166097e-07,
+      "loss": 0.0205,
+      "step": 13581
+    },
+    {
+      "epoch": 3.086818181818182,
+      "grad_norm": 1.0104204694668444,
+      "learning_rate": 3.9982622156760605e-07,
+      "loss": 0.0061,
+      "step": 13582
+    },
+    {
+      "epoch": 3.0870454545454544,
+      "grad_norm": 1.2517886377058671,
+      "learning_rate": 3.9974296764523706e-07,
+      "loss": 0.0071,
+      "step": 13583
+    },
+    {
+      "epoch": 3.0872727272727274,
+      "grad_norm": 1.636040989870913,
+      "learning_rate": 3.9965971831625204e-07,
+      "loss": 0.0081,
+      "step": 13584
+    },
+    {
+      "epoch": 3.0875,
+      "grad_norm": 1.219096850574636,
+      "learning_rate": 3.995764735823485e-07,
+      "loss": 0.0045,
+      "step": 13585
+    },
+    {
+      "epoch": 3.087727272727273,
+      "grad_norm": 1.5280904741807086,
+      "learning_rate": 3.994932334452239e-07,
+      "loss": 0.01,
+      "step": 13586
+    },
+    {
+      "epoch": 3.0879545454545454,
+      "grad_norm": 2.187833135691207,
+      "learning_rate": 3.994099979065758e-07,
+      "loss": 0.0123,
+      "step": 13587
+    },
+    {
+      "epoch": 3.0881818181818184,
+      "grad_norm": 2.4378861217250942,
+      "learning_rate": 3.9932676696810126e-07,
+      "loss": 0.0208,
+      "step": 13588
+    },
+    {
+      "epoch": 3.088409090909091,
+      "grad_norm": 1.091835931239215,
+      "learning_rate": 3.9924354063149773e-07,
+      "loss": 0.0089,
+      "step": 13589
+    },
+    {
+      "epoch": 3.088636363636364,
+      "grad_norm": 1.823014633047369,
+      "learning_rate": 3.9916031889846227e-07,
+      "loss": 0.01,
+      "step": 13590
+    },
+    {
+      "epoch": 3.0888636363636364,
+      "grad_norm": 2.5315150282959684,
+      "learning_rate": 3.990771017706921e-07,
+      "loss": 0.0188,
+      "step": 13591
+    },
+    {
+      "epoch": 3.089090909090909,
+      "grad_norm": 0.7289462648793908,
+      "learning_rate": 3.9899388924988375e-07,
+      "loss": 0.0029,
+      "step": 13592
+    },
+    {
+      "epoch": 3.089318181818182,
+      "grad_norm": 1.1394054503829953,
+      "learning_rate": 3.9891068133773427e-07,
+      "loss": 0.0069,
+      "step": 13593
+    },
+    {
+      "epoch": 3.0895454545454544,
+      "grad_norm": 0.6216305084631434,
+      "learning_rate": 3.988274780359404e-07,
+      "loss": 0.0062,
+      "step": 13594
+    },
+    {
+      "epoch": 3.0897727272727273,
+      "grad_norm": 1.9215201014910965,
+      "learning_rate": 3.987442793461989e-07,
+      "loss": 0.0182,
+      "step": 13595
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 4.903261470517571,
+      "learning_rate": 3.986610852702064e-07,
+      "loss": 0.0178,
+      "step": 13596
+    },
+    {
+      "epoch": 3.090227272727273,
+      "grad_norm": 1.508620414033715,
+      "learning_rate": 3.98577895809659e-07,
+      "loss": 0.0125,
+      "step": 13597
+    },
+    {
+      "epoch": 3.0904545454545453,
+      "grad_norm": 0.5330619475934633,
+      "learning_rate": 3.984947109662534e-07,
+      "loss": 0.0027,
+      "step": 13598
+    },
+    {
+      "epoch": 3.0906818181818183,
+      "grad_norm": 0.8256575163907132,
+      "learning_rate": 3.984115307416858e-07,
+      "loss": 0.0056,
+      "step": 13599
+    },
+    {
+      "epoch": 3.090909090909091,
+      "grad_norm": 0.9484601994167245,
+      "learning_rate": 3.9832835513765266e-07,
+      "loss": 0.0043,
+      "step": 13600
+    },
+    {
+      "epoch": 3.091136363636364,
+      "grad_norm": 1.444215525210851,
+      "learning_rate": 3.9824518415584947e-07,
+      "loss": 0.0116,
+      "step": 13601
+    },
+    {
+      "epoch": 3.0913636363636363,
+      "grad_norm": 1.1807711311343945,
+      "learning_rate": 3.9816201779797266e-07,
+      "loss": 0.0058,
+      "step": 13602
+    },
+    {
+      "epoch": 3.0915909090909093,
+      "grad_norm": 0.7294966921869266,
+      "learning_rate": 3.9807885606571806e-07,
+      "loss": 0.0036,
+      "step": 13603
+    },
+    {
+      "epoch": 3.091818181818182,
+      "grad_norm": 1.1268792095964462,
+      "learning_rate": 3.9799569896078153e-07,
+      "loss": 0.0045,
+      "step": 13604
+    },
+    {
+      "epoch": 3.0920454545454543,
+      "grad_norm": 1.519629002360704,
+      "learning_rate": 3.979125464848586e-07,
+      "loss": 0.0041,
+      "step": 13605
+    },
+    {
+      "epoch": 3.0922727272727273,
+      "grad_norm": 0.9348120388591237,
+      "learning_rate": 3.978293986396452e-07,
+      "loss": 0.0034,
+      "step": 13606
+    },
+    {
+      "epoch": 3.0925,
+      "grad_norm": 2.761739605500276,
+      "learning_rate": 3.9774625542683646e-07,
+      "loss": 0.0103,
+      "step": 13607
+    },
+    {
+      "epoch": 3.0927272727272728,
+      "grad_norm": 1.4948615429764034,
+      "learning_rate": 3.976631168481281e-07,
+      "loss": 0.0139,
+      "step": 13608
+    },
+    {
+      "epoch": 3.0929545454545453,
+      "grad_norm": 2.900428247295143,
+      "learning_rate": 3.975799829052157e-07,
+      "loss": 0.0218,
+      "step": 13609
+    },
+    {
+      "epoch": 3.0931818181818183,
+      "grad_norm": 2.079679385423624,
+      "learning_rate": 3.974968535997939e-07,
+      "loss": 0.0303,
+      "step": 13610
+    },
+    {
+      "epoch": 3.0934090909090908,
+      "grad_norm": 2.0235660787571663,
+      "learning_rate": 3.974137289335581e-07,
+      "loss": 0.0158,
+      "step": 13611
+    },
+    {
+      "epoch": 3.0936363636363637,
+      "grad_norm": 1.3484832397387896,
+      "learning_rate": 3.9733060890820345e-07,
+      "loss": 0.0077,
+      "step": 13612
+    },
+    {
+      "epoch": 3.0938636363636363,
+      "grad_norm": 5.119495592527373,
+      "learning_rate": 3.972474935254252e-07,
+      "loss": 0.0148,
+      "step": 13613
+    },
+    {
+      "epoch": 3.0940909090909092,
+      "grad_norm": 2.684782197632722,
+      "learning_rate": 3.971643827869176e-07,
+      "loss": 0.0172,
+      "step": 13614
+    },
+    {
+      "epoch": 3.0943181818181817,
+      "grad_norm": 2.2998931654026964,
+      "learning_rate": 3.970812766943757e-07,
+      "loss": 0.015,
+      "step": 13615
+    },
+    {
+      "epoch": 3.0945454545454547,
+      "grad_norm": 0.4954438136716183,
+      "learning_rate": 3.9699817524949426e-07,
+      "loss": 0.0099,
+      "step": 13616
+    },
+    {
+      "epoch": 3.0947727272727272,
+      "grad_norm": 4.402783661527084,
+      "learning_rate": 3.969150784539677e-07,
+      "loss": 0.0261,
+      "step": 13617
+    },
+    {
+      "epoch": 3.095,
+      "grad_norm": 3.8330001254695585,
+      "learning_rate": 3.9683198630949085e-07,
+      "loss": 0.0246,
+      "step": 13618
+    },
+    {
+      "epoch": 3.0952272727272727,
+      "grad_norm": 1.0290211067287038,
+      "learning_rate": 3.967488988177577e-07,
+      "loss": 0.007,
+      "step": 13619
+    },
+    {
+      "epoch": 3.0954545454545457,
+      "grad_norm": 1.094689321323475,
+      "learning_rate": 3.966658159804626e-07,
+      "loss": 0.0169,
+      "step": 13620
+    },
+    {
+      "epoch": 3.095681818181818,
+      "grad_norm": 2.424933391703661,
+      "learning_rate": 3.965827377993e-07,
+      "loss": 0.0131,
+      "step": 13621
+    },
+    {
+      "epoch": 3.0959090909090907,
+      "grad_norm": 1.022244081111369,
+      "learning_rate": 3.9649966427596406e-07,
+      "loss": 0.0076,
+      "step": 13622
+    },
+    {
+      "epoch": 3.0961363636363637,
+      "grad_norm": 2.3622024700997764,
+      "learning_rate": 3.964165954121484e-07,
+      "loss": 0.0169,
+      "step": 13623
+    },
+    {
+      "epoch": 3.096363636363636,
+      "grad_norm": 1.3986891171853655,
+      "learning_rate": 3.9633353120954706e-07,
+      "loss": 0.0093,
+      "step": 13624
+    },
+    {
+      "epoch": 3.096590909090909,
+      "grad_norm": 1.9147785108766715,
+      "learning_rate": 3.9625047166985396e-07,
+      "loss": 0.0135,
+      "step": 13625
+    },
+    {
+      "epoch": 3.0968181818181817,
+      "grad_norm": 1.0411816032353358,
+      "learning_rate": 3.961674167947631e-07,
+      "loss": 0.0072,
+      "step": 13626
+    },
+    {
+      "epoch": 3.0970454545454547,
+      "grad_norm": 1.248049438667221,
+      "learning_rate": 3.9608436658596754e-07,
+      "loss": 0.0105,
+      "step": 13627
+    },
+    {
+      "epoch": 3.097272727272727,
+      "grad_norm": 1.4299873639542806,
+      "learning_rate": 3.960013210451612e-07,
+      "loss": 0.0187,
+      "step": 13628
+    },
+    {
+      "epoch": 3.0975,
+      "grad_norm": 1.2780633247724555,
+      "learning_rate": 3.959182801740374e-07,
+      "loss": 0.0133,
+      "step": 13629
+    },
+    {
+      "epoch": 3.0977272727272727,
+      "grad_norm": 4.861620428839823,
+      "learning_rate": 3.958352439742895e-07,
+      "loss": 0.0356,
+      "step": 13630
+    },
+    {
+      "epoch": 3.0979545454545456,
+      "grad_norm": 3.032442734135491,
+      "learning_rate": 3.9575221244761097e-07,
+      "loss": 0.0246,
+      "step": 13631
+    },
+    {
+      "epoch": 3.098181818181818,
+      "grad_norm": 0.46061450343414806,
+      "learning_rate": 3.956691855956946e-07,
+      "loss": 0.0024,
+      "step": 13632
+    },
+    {
+      "epoch": 3.098409090909091,
+      "grad_norm": 2.1081933391095995,
+      "learning_rate": 3.955861634202336e-07,
+      "loss": 0.0122,
+      "step": 13633
+    },
+    {
+      "epoch": 3.0986363636363636,
+      "grad_norm": 1.544395464450012,
+      "learning_rate": 3.9550314592292105e-07,
+      "loss": 0.0119,
+      "step": 13634
+    },
+    {
+      "epoch": 3.098863636363636,
+      "grad_norm": 1.575085408763518,
+      "learning_rate": 3.954201331054498e-07,
+      "loss": 0.0085,
+      "step": 13635
+    },
+    {
+      "epoch": 3.099090909090909,
+      "grad_norm": 0.4224634570431914,
+      "learning_rate": 3.9533712496951247e-07,
+      "loss": 0.0017,
+      "step": 13636
+    },
+    {
+      "epoch": 3.0993181818181816,
+      "grad_norm": 1.7786407976571457,
+      "learning_rate": 3.9525412151680176e-07,
+      "loss": 0.0108,
+      "step": 13637
+    },
+    {
+      "epoch": 3.0995454545454546,
+      "grad_norm": 1.5254591136843991,
+      "learning_rate": 3.951711227490103e-07,
+      "loss": 0.0145,
+      "step": 13638
+    },
+    {
+      "epoch": 3.099772727272727,
+      "grad_norm": 0.5889676278827376,
+      "learning_rate": 3.9508812866783096e-07,
+      "loss": 0.0019,
+      "step": 13639
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 1.3982672463449781,
+      "learning_rate": 3.950051392749553e-07,
+      "loss": 0.0145,
+      "step": 13640
+    },
+    {
+      "epoch": 3.1002272727272726,
+      "grad_norm": 0.6552879096637672,
+      "learning_rate": 3.9492215457207634e-07,
+      "loss": 0.0033,
+      "step": 13641
+    },
+    {
+      "epoch": 3.1004545454545456,
+      "grad_norm": 1.1731412631520162,
+      "learning_rate": 3.948391745608859e-07,
+      "loss": 0.0205,
+      "step": 13642
+    },
+    {
+      "epoch": 3.100681818181818,
+      "grad_norm": 1.416879282372834,
+      "learning_rate": 3.947561992430764e-07,
+      "loss": 0.0066,
+      "step": 13643
+    },
+    {
+      "epoch": 3.100909090909091,
+      "grad_norm": 2.0812472605254273,
+      "learning_rate": 3.9467322862033976e-07,
+      "loss": 0.0187,
+      "step": 13644
+    },
+    {
+      "epoch": 3.1011363636363636,
+      "grad_norm": 1.1956180456830772,
+      "learning_rate": 3.9459026269436763e-07,
+      "loss": 0.0091,
+      "step": 13645
+    },
+    {
+      "epoch": 3.1013636363636365,
+      "grad_norm": 0.9906311048595773,
+      "learning_rate": 3.9450730146685206e-07,
+      "loss": 0.0058,
+      "step": 13646
+    },
+    {
+      "epoch": 3.101590909090909,
+      "grad_norm": 1.189659339042713,
+      "learning_rate": 3.944243449394847e-07,
+      "loss": 0.0155,
+      "step": 13647
+    },
+    {
+      "epoch": 3.101818181818182,
+      "grad_norm": 1.57064070522977,
+      "learning_rate": 3.943413931139573e-07,
+      "loss": 0.0069,
+      "step": 13648
+    },
+    {
+      "epoch": 3.1020454545454546,
+      "grad_norm": 1.6564068023449585,
+      "learning_rate": 3.942584459919613e-07,
+      "loss": 0.0256,
+      "step": 13649
+    },
+    {
+      "epoch": 3.102272727272727,
+      "grad_norm": 2.0944298687839726,
+      "learning_rate": 3.94175503575188e-07,
+      "loss": 0.013,
+      "step": 13650
+    },
+    {
+      "epoch": 3.1025,
+      "grad_norm": 1.9337077348497016,
+      "learning_rate": 3.9409256586532885e-07,
+      "loss": 0.0109,
+      "step": 13651
+    },
+    {
+      "epoch": 3.1027272727272726,
+      "grad_norm": 2.1495349410863493,
+      "learning_rate": 3.940096328640752e-07,
+      "loss": 0.0085,
+      "step": 13652
+    },
+    {
+      "epoch": 3.1029545454545455,
+      "grad_norm": 2.118082305445234,
+      "learning_rate": 3.9392670457311836e-07,
+      "loss": 0.0298,
+      "step": 13653
+    },
+    {
+      "epoch": 3.103181818181818,
+      "grad_norm": 0.9522093308267223,
+      "learning_rate": 3.9384378099414895e-07,
+      "loss": 0.0044,
+      "step": 13654
+    },
+    {
+      "epoch": 3.103409090909091,
+      "grad_norm": 1.699573607944719,
+      "learning_rate": 3.93760862128858e-07,
+      "loss": 0.0118,
+      "step": 13655
+    },
+    {
+      "epoch": 3.1036363636363635,
+      "grad_norm": 0.4617850310545761,
+      "learning_rate": 3.936779479789366e-07,
+      "loss": 0.0027,
+      "step": 13656
+    },
+    {
+      "epoch": 3.1038636363636365,
+      "grad_norm": 2.8387780237640756,
+      "learning_rate": 3.9359503854607545e-07,
+      "loss": 0.0282,
+      "step": 13657
+    },
+    {
+      "epoch": 3.104090909090909,
+      "grad_norm": 1.806699153689498,
+      "learning_rate": 3.935121338319652e-07,
+      "loss": 0.0126,
+      "step": 13658
+    },
+    {
+      "epoch": 3.104318181818182,
+      "grad_norm": 1.5864541356327522,
+      "learning_rate": 3.934292338382962e-07,
+      "loss": 0.0138,
+      "step": 13659
+    },
+    {
+      "epoch": 3.1045454545454545,
+      "grad_norm": 2.2646234909399627,
+      "learning_rate": 3.9334633856675924e-07,
+      "loss": 0.014,
+      "step": 13660
+    },
+    {
+      "epoch": 3.1047727272727275,
+      "grad_norm": 2.1193362640271403,
+      "learning_rate": 3.9326344801904477e-07,
+      "loss": 0.0185,
+      "step": 13661
+    },
+    {
+      "epoch": 3.105,
+      "grad_norm": 2.04533478584222,
+      "learning_rate": 3.9318056219684263e-07,
+      "loss": 0.0071,
+      "step": 13662
+    },
+    {
+      "epoch": 3.105227272727273,
+      "grad_norm": 1.3183030874694017,
+      "learning_rate": 3.9309768110184327e-07,
+      "loss": 0.0082,
+      "step": 13663
+    },
+    {
+      "epoch": 3.1054545454545455,
+      "grad_norm": 1.0424120997847497,
+      "learning_rate": 3.9301480473573674e-07,
+      "loss": 0.0068,
+      "step": 13664
+    },
+    {
+      "epoch": 3.105681818181818,
+      "grad_norm": 1.7675403469726043,
+      "learning_rate": 3.929319331002131e-07,
+      "loss": 0.0091,
+      "step": 13665
+    },
+    {
+      "epoch": 3.105909090909091,
+      "grad_norm": 1.2563429591284774,
+      "learning_rate": 3.9284906619696235e-07,
+      "loss": 0.0061,
+      "step": 13666
+    },
+    {
+      "epoch": 3.1061363636363635,
+      "grad_norm": 2.4585502682524827,
+      "learning_rate": 3.92766204027674e-07,
+      "loss": 0.0113,
+      "step": 13667
+    },
+    {
+      "epoch": 3.1063636363636364,
+      "grad_norm": 1.5946201603763395,
+      "learning_rate": 3.9268334659403787e-07,
+      "loss": 0.0121,
+      "step": 13668
+    },
+    {
+      "epoch": 3.106590909090909,
+      "grad_norm": 0.9200639258956345,
+      "learning_rate": 3.926004938977437e-07,
+      "loss": 0.0051,
+      "step": 13669
+    },
+    {
+      "epoch": 3.106818181818182,
+      "grad_norm": 1.0600805342731385,
+      "learning_rate": 3.925176459404809e-07,
+      "loss": 0.0072,
+      "step": 13670
+    },
+    {
+      "epoch": 3.1070454545454544,
+      "grad_norm": 2.240120411301634,
+      "learning_rate": 3.9243480272393887e-07,
+      "loss": 0.0185,
+      "step": 13671
+    },
+    {
+      "epoch": 3.1072727272727274,
+      "grad_norm": 1.2780145362290842,
+      "learning_rate": 3.923519642498069e-07,
+      "loss": 0.0079,
+      "step": 13672
+    },
+    {
+      "epoch": 3.1075,
+      "grad_norm": 3.8994249108852252,
+      "learning_rate": 3.922691305197743e-07,
+      "loss": 0.0149,
+      "step": 13673
+    },
+    {
+      "epoch": 3.107727272727273,
+      "grad_norm": 0.8364976659445011,
+      "learning_rate": 3.9218630153553006e-07,
+      "loss": 0.0049,
+      "step": 13674
+    },
+    {
+      "epoch": 3.1079545454545454,
+      "grad_norm": 0.9303467600633872,
+      "learning_rate": 3.9210347729876354e-07,
+      "loss": 0.0057,
+      "step": 13675
+    },
+    {
+      "epoch": 3.1081818181818184,
+      "grad_norm": 0.6338886949071872,
+      "learning_rate": 3.920206578111633e-07,
+      "loss": 0.0044,
+      "step": 13676
+    },
+    {
+      "epoch": 3.108409090909091,
+      "grad_norm": 0.5208088511660851,
+      "learning_rate": 3.919378430744182e-07,
+      "loss": 0.0017,
+      "step": 13677
+    },
+    {
+      "epoch": 3.1086363636363634,
+      "grad_norm": 1.3865847494530064,
+      "learning_rate": 3.918550330902171e-07,
+      "loss": 0.0194,
+      "step": 13678
+    },
+    {
+      "epoch": 3.1088636363636364,
+      "grad_norm": 2.5445288954963,
+      "learning_rate": 3.917722278602488e-07,
+      "loss": 0.0207,
+      "step": 13679
+    },
+    {
+      "epoch": 3.109090909090909,
+      "grad_norm": 0.6247605472236869,
+      "learning_rate": 3.916894273862015e-07,
+      "loss": 0.0022,
+      "step": 13680
+    },
+    {
+      "epoch": 3.109318181818182,
+      "grad_norm": 1.51583809488982,
+      "learning_rate": 3.916066316697638e-07,
+      "loss": 0.0095,
+      "step": 13681
+    },
+    {
+      "epoch": 3.1095454545454544,
+      "grad_norm": 1.313821274898392,
+      "learning_rate": 3.91523840712624e-07,
+      "loss": 0.0099,
+      "step": 13682
+    },
+    {
+      "epoch": 3.1097727272727274,
+      "grad_norm": 1.11023078560332,
+      "learning_rate": 3.9144105451647074e-07,
+      "loss": 0.0085,
+      "step": 13683
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 2.1208639185960516,
+      "learning_rate": 3.9135827308299156e-07,
+      "loss": 0.013,
+      "step": 13684
+    },
+    {
+      "epoch": 3.110227272727273,
+      "grad_norm": 1.8986890253778606,
+      "learning_rate": 3.9127549641387487e-07,
+      "loss": 0.0087,
+      "step": 13685
+    },
+    {
+      "epoch": 3.1104545454545454,
+      "grad_norm": 1.1536241390411186,
+      "learning_rate": 3.9119272451080854e-07,
+      "loss": 0.0072,
+      "step": 13686
+    },
+    {
+      "epoch": 3.1106818181818183,
+      "grad_norm": 1.3862338471505056,
+      "learning_rate": 3.9110995737548035e-07,
+      "loss": 0.0039,
+      "step": 13687
+    },
+    {
+      "epoch": 3.110909090909091,
+      "grad_norm": 1.8359194248245807,
+      "learning_rate": 3.910271950095784e-07,
+      "loss": 0.0127,
+      "step": 13688
+    },
+    {
+      "epoch": 3.111136363636364,
+      "grad_norm": 0.433835442690269,
+      "learning_rate": 3.9094443741479e-07,
+      "loss": 0.0021,
+      "step": 13689
+    },
+    {
+      "epoch": 3.1113636363636363,
+      "grad_norm": 1.4083213222517632,
+      "learning_rate": 3.908616845928028e-07,
+      "loss": 0.0073,
+      "step": 13690
+    },
+    {
+      "epoch": 3.1115909090909093,
+      "grad_norm": 2.325091417005913,
+      "learning_rate": 3.9077893654530437e-07,
+      "loss": 0.0133,
+      "step": 13691
+    },
+    {
+      "epoch": 3.111818181818182,
+      "grad_norm": 1.3256444857044343,
+      "learning_rate": 3.906961932739823e-07,
+      "loss": 0.014,
+      "step": 13692
+    },
+    {
+      "epoch": 3.1120454545454543,
+      "grad_norm": 0.8140065084122858,
+      "learning_rate": 3.9061345478052335e-07,
+      "loss": 0.0052,
+      "step": 13693
+    },
+    {
+      "epoch": 3.1122727272727273,
+      "grad_norm": 2.1999134764875423,
+      "learning_rate": 3.9053072106661497e-07,
+      "loss": 0.0161,
+      "step": 13694
+    },
+    {
+      "epoch": 3.1125,
+      "grad_norm": 0.9714321363880315,
+      "learning_rate": 3.904479921339441e-07,
+      "loss": 0.0063,
+      "step": 13695
+    },
+    {
+      "epoch": 3.112727272727273,
+      "grad_norm": 0.624832137279872,
+      "learning_rate": 3.90365267984198e-07,
+      "loss": 0.004,
+      "step": 13696
+    },
+    {
+      "epoch": 3.1129545454545453,
+      "grad_norm": 0.7346189663666761,
+      "learning_rate": 3.9028254861906367e-07,
+      "loss": 0.0028,
+      "step": 13697
+    },
+    {
+      "epoch": 3.1131818181818183,
+      "grad_norm": 1.8295988905889946,
+      "learning_rate": 3.901998340402274e-07,
+      "loss": 0.014,
+      "step": 13698
+    },
+    {
+      "epoch": 3.113409090909091,
+      "grad_norm": 1.4822701176052266,
+      "learning_rate": 3.901171242493762e-07,
+      "loss": 0.0083,
+      "step": 13699
+    },
+    {
+      "epoch": 3.1136363636363638,
+      "grad_norm": 2.3524563372208482,
+      "learning_rate": 3.9003441924819664e-07,
+      "loss": 0.0401,
+      "step": 13700
+    },
+    {
+      "epoch": 3.1138636363636363,
+      "grad_norm": 2.01874629298479,
+      "learning_rate": 3.8995171903837526e-07,
+      "loss": 0.0126,
+      "step": 13701
+    },
+    {
+      "epoch": 3.1140909090909092,
+      "grad_norm": 3.1559531713792266,
+      "learning_rate": 3.8986902362159824e-07,
+      "loss": 0.019,
+      "step": 13702
+    },
+    {
+      "epoch": 3.1143181818181818,
+      "grad_norm": 1.4150910728120187,
+      "learning_rate": 3.89786332999552e-07,
+      "loss": 0.0112,
+      "step": 13703
+    },
+    {
+      "epoch": 3.1145454545454547,
+      "grad_norm": 1.3724494324979295,
+      "learning_rate": 3.8970364717392297e-07,
+      "loss": 0.0107,
+      "step": 13704
+    },
+    {
+      "epoch": 3.1147727272727272,
+      "grad_norm": 1.7436260190137316,
+      "learning_rate": 3.896209661463972e-07,
+      "loss": 0.0074,
+      "step": 13705
+    },
+    {
+      "epoch": 3.115,
+      "grad_norm": 1.2975519341144863,
+      "learning_rate": 3.8953828991866024e-07,
+      "loss": 0.0145,
+      "step": 13706
+    },
+    {
+      "epoch": 3.1152272727272727,
+      "grad_norm": 3.2309950947169166,
+      "learning_rate": 3.8945561849239844e-07,
+      "loss": 0.045,
+      "step": 13707
+    },
+    {
+      "epoch": 3.1154545454545453,
+      "grad_norm": 3.8106289043003647,
+      "learning_rate": 3.893729518692975e-07,
+      "loss": 0.0195,
+      "step": 13708
+    },
+    {
+      "epoch": 3.115681818181818,
+      "grad_norm": 1.0123128245214463,
+      "learning_rate": 3.892902900510431e-07,
+      "loss": 0.0049,
+      "step": 13709
+    },
+    {
+      "epoch": 3.1159090909090907,
+      "grad_norm": 1.2792331683667488,
+      "learning_rate": 3.892076330393211e-07,
+      "loss": 0.0056,
+      "step": 13710
+    },
+    {
+      "epoch": 3.1161363636363637,
+      "grad_norm": 2.3881480607449816,
+      "learning_rate": 3.8912498083581664e-07,
+      "loss": 0.013,
+      "step": 13711
+    },
+    {
+      "epoch": 3.1163636363636362,
+      "grad_norm": 1.429154634873166,
+      "learning_rate": 3.8904233344221535e-07,
+      "loss": 0.0057,
+      "step": 13712
+    },
+    {
+      "epoch": 3.116590909090909,
+      "grad_norm": 1.2397388332648207,
+      "learning_rate": 3.889596908602025e-07,
+      "loss": 0.0093,
+      "step": 13713
+    },
+    {
+      "epoch": 3.1168181818181817,
+      "grad_norm": 0.9666445644181703,
+      "learning_rate": 3.8887705309146365e-07,
+      "loss": 0.0048,
+      "step": 13714
+    },
+    {
+      "epoch": 3.1170454545454547,
+      "grad_norm": 0.9243082708391659,
+      "learning_rate": 3.887944201376834e-07,
+      "loss": 0.0044,
+      "step": 13715
+    },
+    {
+      "epoch": 3.117272727272727,
+      "grad_norm": 1.6522584139071295,
+      "learning_rate": 3.887117920005469e-07,
+      "loss": 0.0101,
+      "step": 13716
+    },
+    {
+      "epoch": 3.1175,
+      "grad_norm": 3.479490921438993,
+      "learning_rate": 3.8862916868173924e-07,
+      "loss": 0.0262,
+      "step": 13717
+    },
+    {
+      "epoch": 3.1177272727272727,
+      "grad_norm": 1.1987355980858698,
+      "learning_rate": 3.8854655018294525e-07,
+      "loss": 0.0095,
+      "step": 13718
+    },
+    {
+      "epoch": 3.1179545454545456,
+      "grad_norm": 3.052532989362841,
+      "learning_rate": 3.8846393650584973e-07,
+      "loss": 0.0075,
+      "step": 13719
+    },
+    {
+      "epoch": 3.118181818181818,
+      "grad_norm": 1.5915442068941164,
+      "learning_rate": 3.88381327652137e-07,
+      "loss": 0.0109,
+      "step": 13720
+    },
+    {
+      "epoch": 3.1184090909090907,
+      "grad_norm": 0.9432313957562047,
+      "learning_rate": 3.882987236234918e-07,
+      "loss": 0.0128,
+      "step": 13721
+    },
+    {
+      "epoch": 3.1186363636363637,
+      "grad_norm": 0.6426819592068483,
+      "learning_rate": 3.8821612442159867e-07,
+      "loss": 0.0037,
+      "step": 13722
+    },
+    {
+      "epoch": 3.118863636363636,
+      "grad_norm": 0.6523662990719653,
+      "learning_rate": 3.8813353004814203e-07,
+      "loss": 0.0036,
+      "step": 13723
+    },
+    {
+      "epoch": 3.119090909090909,
+      "grad_norm": 3.132249213476015,
+      "learning_rate": 3.8805094050480565e-07,
+      "loss": 0.0194,
+      "step": 13724
+    },
+    {
+      "epoch": 3.1193181818181817,
+      "grad_norm": 1.3870222352651465,
+      "learning_rate": 3.8796835579327406e-07,
+      "loss": 0.008,
+      "step": 13725
+    },
+    {
+      "epoch": 3.1195454545454546,
+      "grad_norm": 2.6682559662804235,
+      "learning_rate": 3.8788577591523116e-07,
+      "loss": 0.0163,
+      "step": 13726
+    },
+    {
+      "epoch": 3.119772727272727,
+      "grad_norm": 1.0330657723943142,
+      "learning_rate": 3.8780320087236133e-07,
+      "loss": 0.0116,
+      "step": 13727
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 1.6327051400753807,
+      "learning_rate": 3.877206306663477e-07,
+      "loss": 0.0151,
+      "step": 13728
+    },
+    {
+      "epoch": 3.1202272727272726,
+      "grad_norm": 1.0512268391118145,
+      "learning_rate": 3.8763806529887443e-07,
+      "loss": 0.0053,
+      "step": 13729
+    },
+    {
+      "epoch": 3.1204545454545456,
+      "grad_norm": 0.7940150694956496,
+      "learning_rate": 3.8755550477162505e-07,
+      "loss": 0.0043,
+      "step": 13730
+    },
+    {
+      "epoch": 3.120681818181818,
+      "grad_norm": 0.7793531199002539,
+      "learning_rate": 3.874729490862832e-07,
+      "loss": 0.0037,
+      "step": 13731
+    },
+    {
+      "epoch": 3.120909090909091,
+      "grad_norm": 1.4958117304827279,
+      "learning_rate": 3.873903982445324e-07,
+      "loss": 0.0115,
+      "step": 13732
+    },
+    {
+      "epoch": 3.1211363636363636,
+      "grad_norm": 1.0687722240175452,
+      "learning_rate": 3.8730785224805586e-07,
+      "loss": 0.0053,
+      "step": 13733
+    },
+    {
+      "epoch": 3.1213636363636366,
+      "grad_norm": 3.259899886302685,
+      "learning_rate": 3.872253110985368e-07,
+      "loss": 0.02,
+      "step": 13734
+    },
+    {
+      "epoch": 3.121590909090909,
+      "grad_norm": 3.3820963358878613,
+      "learning_rate": 3.871427747976585e-07,
+      "loss": 0.0282,
+      "step": 13735
+    },
+    {
+      "epoch": 3.1218181818181816,
+      "grad_norm": 1.9248435309662757,
+      "learning_rate": 3.8706024334710426e-07,
+      "loss": 0.0125,
+      "step": 13736
+    },
+    {
+      "epoch": 3.1220454545454546,
+      "grad_norm": 2.075426995045341,
+      "learning_rate": 3.8697771674855654e-07,
+      "loss": 0.011,
+      "step": 13737
+    },
+    {
+      "epoch": 3.122272727272727,
+      "grad_norm": 1.2217393687785636,
+      "learning_rate": 3.868951950036984e-07,
+      "loss": 0.0112,
+      "step": 13738
+    },
+    {
+      "epoch": 3.1225,
+      "grad_norm": 3.7522836543209457,
+      "learning_rate": 3.868126781142126e-07,
+      "loss": 0.0322,
+      "step": 13739
+    },
+    {
+      "epoch": 3.1227272727272726,
+      "grad_norm": 0.88917637611142,
+      "learning_rate": 3.8673016608178187e-07,
+      "loss": 0.008,
+      "step": 13740
+    },
+    {
+      "epoch": 3.1229545454545455,
+      "grad_norm": 0.7188242044413413,
+      "learning_rate": 3.8664765890808884e-07,
+      "loss": 0.0048,
+      "step": 13741
+    },
+    {
+      "epoch": 3.123181818181818,
+      "grad_norm": 5.664052566450942,
+      "learning_rate": 3.865651565948158e-07,
+      "loss": 0.0107,
+      "step": 13742
+    },
+    {
+      "epoch": 3.123409090909091,
+      "grad_norm": 1.2679939064657724,
+      "learning_rate": 3.864826591436451e-07,
+      "loss": 0.0171,
+      "step": 13743
+    },
+    {
+      "epoch": 3.1236363636363635,
+      "grad_norm": 1.3056760342789888,
+      "learning_rate": 3.864001665562592e-07,
+      "loss": 0.0058,
+      "step": 13744
+    },
+    {
+      "epoch": 3.1238636363636365,
+      "grad_norm": 1.9532413840455274,
+      "learning_rate": 3.8631767883434035e-07,
+      "loss": 0.0156,
+      "step": 13745
+    },
+    {
+      "epoch": 3.124090909090909,
+      "grad_norm": 1.8869593952485686,
+      "learning_rate": 3.8623519597957013e-07,
+      "loss": 0.0202,
+      "step": 13746
+    },
+    {
+      "epoch": 3.124318181818182,
+      "grad_norm": 1.6647938759938712,
+      "learning_rate": 3.861527179936309e-07,
+      "loss": 0.0188,
+      "step": 13747
+    },
+    {
+      "epoch": 3.1245454545454545,
+      "grad_norm": 1.306313001141913,
+      "learning_rate": 3.8607024487820445e-07,
+      "loss": 0.0059,
+      "step": 13748
+    },
+    {
+      "epoch": 3.1247727272727275,
+      "grad_norm": 1.5614920209809566,
+      "learning_rate": 3.859877766349726e-07,
+      "loss": 0.0141,
+      "step": 13749
+    },
+    {
+      "epoch": 3.125,
+      "grad_norm": 2.3004783066543064,
+      "learning_rate": 3.8590531326561697e-07,
+      "loss": 0.0195,
+      "step": 13750
+    },
+    {
+      "epoch": 3.1252272727272725,
+      "grad_norm": 1.0133224770228089,
+      "learning_rate": 3.8582285477181894e-07,
+      "loss": 0.0057,
+      "step": 13751
+    },
+    {
+      "epoch": 3.1254545454545455,
+      "grad_norm": 0.9898629940185318,
+      "learning_rate": 3.857404011552603e-07,
+      "loss": 0.0079,
+      "step": 13752
+    },
+    {
+      "epoch": 3.125681818181818,
+      "grad_norm": 0.949080026800604,
+      "learning_rate": 3.8565795241762225e-07,
+      "loss": 0.0068,
+      "step": 13753
+    },
+    {
+      "epoch": 3.125909090909091,
+      "grad_norm": 1.1515992683963059,
+      "learning_rate": 3.8557550856058626e-07,
+      "loss": 0.0055,
+      "step": 13754
+    },
+    {
+      "epoch": 3.1261363636363635,
+      "grad_norm": 1.0689504171348805,
+      "learning_rate": 3.854930695858332e-07,
+      "loss": 0.0033,
+      "step": 13755
+    },
+    {
+      "epoch": 3.1263636363636365,
+      "grad_norm": 1.5158955574242867,
+      "learning_rate": 3.8541063549504426e-07,
+      "loss": 0.0097,
+      "step": 13756
+    },
+    {
+      "epoch": 3.126590909090909,
+      "grad_norm": 1.19875037314914,
+      "learning_rate": 3.8532820628990044e-07,
+      "loss": 0.0085,
+      "step": 13757
+    },
+    {
+      "epoch": 3.126818181818182,
+      "grad_norm": 0.4854056096990115,
+      "learning_rate": 3.85245781972083e-07,
+      "loss": 0.0026,
+      "step": 13758
+    },
+    {
+      "epoch": 3.1270454545454545,
+      "grad_norm": 1.5321366724477323,
+      "learning_rate": 3.851633625432719e-07,
+      "loss": 0.0061,
+      "step": 13759
+    },
+    {
+      "epoch": 3.1272727272727274,
+      "grad_norm": 2.4799673865612046,
+      "learning_rate": 3.8508094800514843e-07,
+      "loss": 0.0224,
+      "step": 13760
+    },
+    {
+      "epoch": 3.1275,
+      "grad_norm": 1.8647249527821996,
+      "learning_rate": 3.849985383593929e-07,
+      "loss": 0.0096,
+      "step": 13761
+    },
+    {
+      "epoch": 3.127727272727273,
+      "grad_norm": 1.300609031046127,
+      "learning_rate": 3.84916133607686e-07,
+      "loss": 0.0102,
+      "step": 13762
+    },
+    {
+      "epoch": 3.1279545454545454,
+      "grad_norm": 1.2660061965223022,
+      "learning_rate": 3.848337337517079e-07,
+      "loss": 0.0095,
+      "step": 13763
+    },
+    {
+      "epoch": 3.128181818181818,
+      "grad_norm": 1.2472525159132637,
+      "learning_rate": 3.84751338793139e-07,
+      "loss": 0.0061,
+      "step": 13764
+    },
+    {
+      "epoch": 3.128409090909091,
+      "grad_norm": 1.3804711593144694,
+      "learning_rate": 3.8466894873365935e-07,
+      "loss": 0.0054,
+      "step": 13765
+    },
+    {
+      "epoch": 3.1286363636363634,
+      "grad_norm": 1.8945675768233954,
+      "learning_rate": 3.8458656357494913e-07,
+      "loss": 0.022,
+      "step": 13766
+    },
+    {
+      "epoch": 3.1288636363636364,
+      "grad_norm": 1.2143398415247935,
+      "learning_rate": 3.8450418331868854e-07,
+      "loss": 0.0084,
+      "step": 13767
+    },
+    {
+      "epoch": 3.129090909090909,
+      "grad_norm": 2.998343979223532,
+      "learning_rate": 3.84421807966557e-07,
+      "loss": 0.0162,
+      "step": 13768
+    },
+    {
+      "epoch": 3.129318181818182,
+      "grad_norm": 1.154883098372083,
+      "learning_rate": 3.843394375202346e-07,
+      "loss": 0.0053,
+      "step": 13769
+    },
+    {
+      "epoch": 3.1295454545454544,
+      "grad_norm": 1.209854450210286,
+      "learning_rate": 3.842570719814008e-07,
+      "loss": 0.012,
+      "step": 13770
+    },
+    {
+      "epoch": 3.1297727272727274,
+      "grad_norm": 2.442146540788601,
+      "learning_rate": 3.8417471135173543e-07,
+      "loss": 0.0291,
+      "step": 13771
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 1.20293180770176,
+      "learning_rate": 3.8409235563291775e-07,
+      "loss": 0.0069,
+      "step": 13772
+    },
+    {
+      "epoch": 3.130227272727273,
+      "grad_norm": 1.1251307950464755,
+      "learning_rate": 3.840100048266272e-07,
+      "loss": 0.0072,
+      "step": 13773
+    },
+    {
+      "epoch": 3.1304545454545454,
+      "grad_norm": 1.7237268023905434,
+      "learning_rate": 3.839276589345431e-07,
+      "loss": 0.0213,
+      "step": 13774
+    },
+    {
+      "epoch": 3.1306818181818183,
+      "grad_norm": 2.0671014835257666,
+      "learning_rate": 3.838453179583446e-07,
+      "loss": 0.0153,
+      "step": 13775
+    },
+    {
+      "epoch": 3.130909090909091,
+      "grad_norm": 1.0711309443998216,
+      "learning_rate": 3.837629818997111e-07,
+      "loss": 0.006,
+      "step": 13776
+    },
+    {
+      "epoch": 3.131136363636364,
+      "grad_norm": 1.3670942708234883,
+      "learning_rate": 3.8368065076032095e-07,
+      "loss": 0.0056,
+      "step": 13777
+    },
+    {
+      "epoch": 3.1313636363636363,
+      "grad_norm": 1.3277897340322529,
+      "learning_rate": 3.835983245418533e-07,
+      "loss": 0.0094,
+      "step": 13778
+    },
+    {
+      "epoch": 3.1315909090909093,
+      "grad_norm": 1.8647817251511178,
+      "learning_rate": 3.8351600324598683e-07,
+      "loss": 0.0103,
+      "step": 13779
+    },
+    {
+      "epoch": 3.131818181818182,
+      "grad_norm": 1.0165797479325227,
+      "learning_rate": 3.834336868744008e-07,
+      "loss": 0.0049,
+      "step": 13780
+    },
+    {
+      "epoch": 3.1320454545454544,
+      "grad_norm": 1.9330948427757935,
+      "learning_rate": 3.8335137542877307e-07,
+      "loss": 0.0118,
+      "step": 13781
+    },
+    {
+      "epoch": 3.1322727272727273,
+      "grad_norm": 0.9486728121640428,
+      "learning_rate": 3.8326906891078236e-07,
+      "loss": 0.007,
+      "step": 13782
+    },
+    {
+      "epoch": 3.1325,
+      "grad_norm": 0.8429797652907903,
+      "learning_rate": 3.831867673221071e-07,
+      "loss": 0.0031,
+      "step": 13783
+    },
+    {
+      "epoch": 3.132727272727273,
+      "grad_norm": 0.8169130502704005,
+      "learning_rate": 3.8310447066442556e-07,
+      "loss": 0.0035,
+      "step": 13784
+    },
+    {
+      "epoch": 3.1329545454545453,
+      "grad_norm": 0.7392412183909844,
+      "learning_rate": 3.8302217893941577e-07,
+      "loss": 0.0043,
+      "step": 13785
+    },
+    {
+      "epoch": 3.1331818181818183,
+      "grad_norm": 1.6689707753264846,
+      "learning_rate": 3.829398921487559e-07,
+      "loss": 0.0111,
+      "step": 13786
+    },
+    {
+      "epoch": 3.133409090909091,
+      "grad_norm": 1.4023727262835424,
+      "learning_rate": 3.828576102941239e-07,
+      "loss": 0.0082,
+      "step": 13787
+    },
+    {
+      "epoch": 3.1336363636363638,
+      "grad_norm": 4.891650041801063,
+      "learning_rate": 3.8277533337719775e-07,
+      "loss": 0.025,
+      "step": 13788
+    },
+    {
+      "epoch": 3.1338636363636363,
+      "grad_norm": 1.9568560293140391,
+      "learning_rate": 3.826930613996553e-07,
+      "loss": 0.0227,
+      "step": 13789
+    },
+    {
+      "epoch": 3.1340909090909093,
+      "grad_norm": 1.4938802323932174,
+      "learning_rate": 3.8261079436317393e-07,
+      "loss": 0.0081,
+      "step": 13790
+    },
+    {
+      "epoch": 3.134318181818182,
+      "grad_norm": 1.728805287814634,
+      "learning_rate": 3.825285322694312e-07,
+      "loss": 0.0103,
+      "step": 13791
+    },
+    {
+      "epoch": 3.1345454545454547,
+      "grad_norm": 1.931391210028627,
+      "learning_rate": 3.824462751201048e-07,
+      "loss": 0.0142,
+      "step": 13792
+    },
+    {
+      "epoch": 3.1347727272727273,
+      "grad_norm": 1.1189235847375154,
+      "learning_rate": 3.8236402291687205e-07,
+      "loss": 0.0067,
+      "step": 13793
+    },
+    {
+      "epoch": 3.135,
+      "grad_norm": 1.358219200590721,
+      "learning_rate": 3.8228177566141016e-07,
+      "loss": 0.0146,
+      "step": 13794
+    },
+    {
+      "epoch": 3.1352272727272728,
+      "grad_norm": 1.010030336967713,
+      "learning_rate": 3.821995333553963e-07,
+      "loss": 0.0037,
+      "step": 13795
+    },
+    {
+      "epoch": 3.1354545454545453,
+      "grad_norm": 1.4277233883112166,
+      "learning_rate": 3.821172960005074e-07,
+      "loss": 0.0119,
+      "step": 13796
+    },
+    {
+      "epoch": 3.1356818181818182,
+      "grad_norm": 1.4638274237669515,
+      "learning_rate": 3.8203506359842066e-07,
+      "loss": 0.0068,
+      "step": 13797
+    },
+    {
+      "epoch": 3.1359090909090908,
+      "grad_norm": 1.1534153896993637,
+      "learning_rate": 3.819528361508131e-07,
+      "loss": 0.005,
+      "step": 13798
+    },
+    {
+      "epoch": 3.1361363636363637,
+      "grad_norm": 0.7809898425282923,
+      "learning_rate": 3.81870613659361e-07,
+      "loss": 0.0055,
+      "step": 13799
+    },
+    {
+      "epoch": 3.1363636363636362,
+      "grad_norm": 1.3773394942290065,
+      "learning_rate": 3.8178839612574123e-07,
+      "loss": 0.0134,
+      "step": 13800
+    },
+    {
+      "epoch": 3.136590909090909,
+      "grad_norm": 3.423151622939819,
+      "learning_rate": 3.817061835516304e-07,
+      "loss": 0.0429,
+      "step": 13801
+    },
+    {
+      "epoch": 3.1368181818181817,
+      "grad_norm": 2.2761996743027577,
+      "learning_rate": 3.8162397593870496e-07,
+      "loss": 0.0176,
+      "step": 13802
+    },
+    {
+      "epoch": 3.1370454545454547,
+      "grad_norm": 1.8388076390520662,
+      "learning_rate": 3.8154177328864123e-07,
+      "loss": 0.0091,
+      "step": 13803
+    },
+    {
+      "epoch": 3.137272727272727,
+      "grad_norm": 2.163019836547882,
+      "learning_rate": 3.8145957560311543e-07,
+      "loss": 0.0208,
+      "step": 13804
+    },
+    {
+      "epoch": 3.1375,
+      "grad_norm": 1.7178423326230574,
+      "learning_rate": 3.813773828838038e-07,
+      "loss": 0.0152,
+      "step": 13805
+    },
+    {
+      "epoch": 3.1377272727272727,
+      "grad_norm": 0.970919758644059,
+      "learning_rate": 3.8129519513238256e-07,
+      "loss": 0.006,
+      "step": 13806
+    },
+    {
+      "epoch": 3.137954545454545,
+      "grad_norm": 3.992294257870713,
+      "learning_rate": 3.812130123505272e-07,
+      "loss": 0.0181,
+      "step": 13807
+    },
+    {
+      "epoch": 3.138181818181818,
+      "grad_norm": 2.0568070930090006,
+      "learning_rate": 3.8113083453991376e-07,
+      "loss": 0.024,
+      "step": 13808
+    },
+    {
+      "epoch": 3.1384090909090907,
+      "grad_norm": 1.9577625937639345,
+      "learning_rate": 3.810486617022181e-07,
+      "loss": 0.0068,
+      "step": 13809
+    },
+    {
+      "epoch": 3.1386363636363637,
+      "grad_norm": 1.0491267365359993,
+      "learning_rate": 3.809664938391159e-07,
+      "loss": 0.0061,
+      "step": 13810
+    },
+    {
+      "epoch": 3.138863636363636,
+      "grad_norm": 6.934666800868379,
+      "learning_rate": 3.8088433095228287e-07,
+      "loss": 0.0092,
+      "step": 13811
+    },
+    {
+      "epoch": 3.139090909090909,
+      "grad_norm": 2.3071926609546822,
+      "learning_rate": 3.808021730433939e-07,
+      "loss": 0.0183,
+      "step": 13812
+    },
+    {
+      "epoch": 3.1393181818181817,
+      "grad_norm": 0.698645518135427,
+      "learning_rate": 3.807200201141247e-07,
+      "loss": 0.0058,
+      "step": 13813
+    },
+    {
+      "epoch": 3.1395454545454546,
+      "grad_norm": 3.7786325927347977,
+      "learning_rate": 3.806378721661505e-07,
+      "loss": 0.0145,
+      "step": 13814
+    },
+    {
+      "epoch": 3.139772727272727,
+      "grad_norm": 2.143195008513661,
+      "learning_rate": 3.805557292011465e-07,
+      "loss": 0.0023,
+      "step": 13815
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 1.3376517245997548,
+      "learning_rate": 3.804735912207874e-07,
+      "loss": 0.0128,
+      "step": 13816
+    },
+    {
+      "epoch": 3.1402272727272726,
+      "grad_norm": 1.2887648235447176,
+      "learning_rate": 3.803914582267485e-07,
+      "loss": 0.0098,
+      "step": 13817
+    },
+    {
+      "epoch": 3.1404545454545456,
+      "grad_norm": 1.522527067900255,
+      "learning_rate": 3.803093302207045e-07,
+      "loss": 0.0083,
+      "step": 13818
+    },
+    {
+      "epoch": 3.140681818181818,
+      "grad_norm": 1.3612785293987704,
+      "learning_rate": 3.802272072043301e-07,
+      "loss": 0.0107,
+      "step": 13819
+    },
+    {
+      "epoch": 3.140909090909091,
+      "grad_norm": 1.618332945729497,
+      "learning_rate": 3.801450891793003e-07,
+      "loss": 0.0095,
+      "step": 13820
+    },
+    {
+      "epoch": 3.1411363636363636,
+      "grad_norm": 1.209931479623135,
+      "learning_rate": 3.80062976147289e-07,
+      "loss": 0.01,
+      "step": 13821
+    },
+    {
+      "epoch": 3.1413636363636366,
+      "grad_norm": 1.673675025732667,
+      "learning_rate": 3.7998086810997094e-07,
+      "loss": 0.0104,
+      "step": 13822
+    },
+    {
+      "epoch": 3.141590909090909,
+      "grad_norm": 3.1461607219636267,
+      "learning_rate": 3.7989876506902045e-07,
+      "loss": 0.0113,
+      "step": 13823
+    },
+    {
+      "epoch": 3.1418181818181816,
+      "grad_norm": 0.9389304056832336,
+      "learning_rate": 3.798166670261118e-07,
+      "loss": 0.0086,
+      "step": 13824
+    },
+    {
+      "epoch": 3.1420454545454546,
+      "grad_norm": 0.8877674031380438,
+      "learning_rate": 3.79734573982919e-07,
+      "loss": 0.007,
+      "step": 13825
+    },
+    {
+      "epoch": 3.142272727272727,
+      "grad_norm": 2.71283437328901,
+      "learning_rate": 3.79652485941116e-07,
+      "loss": 0.0176,
+      "step": 13826
+    },
+    {
+      "epoch": 3.1425,
+      "grad_norm": 0.6069565842606057,
+      "learning_rate": 3.79570402902377e-07,
+      "loss": 0.0032,
+      "step": 13827
+    },
+    {
+      "epoch": 3.1427272727272726,
+      "grad_norm": 2.219506321326004,
+      "learning_rate": 3.794883248683759e-07,
+      "loss": 0.031,
+      "step": 13828
+    },
+    {
+      "epoch": 3.1429545454545456,
+      "grad_norm": 0.25352698735473833,
+      "learning_rate": 3.794062518407858e-07,
+      "loss": 0.0011,
+      "step": 13829
+    },
+    {
+      "epoch": 3.143181818181818,
+      "grad_norm": 1.4388059377081652,
+      "learning_rate": 3.793241838212808e-07,
+      "loss": 0.0091,
+      "step": 13830
+    },
+    {
+      "epoch": 3.143409090909091,
+      "grad_norm": 0.8683506571719676,
+      "learning_rate": 3.7924212081153434e-07,
+      "loss": 0.0045,
+      "step": 13831
+    },
+    {
+      "epoch": 3.1436363636363636,
+      "grad_norm": 0.9501444045073345,
+      "learning_rate": 3.791600628132197e-07,
+      "loss": 0.0037,
+      "step": 13832
+    },
+    {
+      "epoch": 3.1438636363636365,
+      "grad_norm": 1.96631318394761,
+      "learning_rate": 3.7907800982801034e-07,
+      "loss": 0.0125,
+      "step": 13833
+    },
+    {
+      "epoch": 3.144090909090909,
+      "grad_norm": 2.9056935806171738,
+      "learning_rate": 3.7899596185757936e-07,
+      "loss": 0.0142,
+      "step": 13834
+    },
+    {
+      "epoch": 3.144318181818182,
+      "grad_norm": 0.5933463581823238,
+      "learning_rate": 3.789139189035999e-07,
+      "loss": 0.0037,
+      "step": 13835
+    },
+    {
+      "epoch": 3.1445454545454545,
+      "grad_norm": 1.3182301802651197,
+      "learning_rate": 3.7883188096774494e-07,
+      "loss": 0.0087,
+      "step": 13836
+    },
+    {
+      "epoch": 3.144772727272727,
+      "grad_norm": 1.1482519422538418,
+      "learning_rate": 3.787498480516875e-07,
+      "loss": 0.0073,
+      "step": 13837
+    },
+    {
+      "epoch": 3.145,
+      "grad_norm": 0.25362362744177064,
+      "learning_rate": 3.7866782015710007e-07,
+      "loss": 0.0011,
+      "step": 13838
+    },
+    {
+      "epoch": 3.1452272727272725,
+      "grad_norm": 0.875858382204396,
+      "learning_rate": 3.785857972856556e-07,
+      "loss": 0.0043,
+      "step": 13839
+    },
+    {
+      "epoch": 3.1454545454545455,
+      "grad_norm": 1.7448601550148617,
+      "learning_rate": 3.7850377943902666e-07,
+      "loss": 0.0098,
+      "step": 13840
+    },
+    {
+      "epoch": 3.145681818181818,
+      "grad_norm": 0.733637355701528,
+      "learning_rate": 3.7842176661888565e-07,
+      "loss": 0.0031,
+      "step": 13841
+    },
+    {
+      "epoch": 3.145909090909091,
+      "grad_norm": 1.988137368091579,
+      "learning_rate": 3.783397588269053e-07,
+      "loss": 0.0111,
+      "step": 13842
+    },
+    {
+      "epoch": 3.1461363636363635,
+      "grad_norm": 1.160936105167174,
+      "learning_rate": 3.7825775606475727e-07,
+      "loss": 0.0101,
+      "step": 13843
+    },
+    {
+      "epoch": 3.1463636363636365,
+      "grad_norm": 1.6441794910690068,
+      "learning_rate": 3.7817575833411403e-07,
+      "loss": 0.0155,
+      "step": 13844
+    },
+    {
+      "epoch": 3.146590909090909,
+      "grad_norm": 1.3290856560854223,
+      "learning_rate": 3.7809376563664777e-07,
+      "loss": 0.0276,
+      "step": 13845
+    },
+    {
+      "epoch": 3.146818181818182,
+      "grad_norm": 3.0740781461038638,
+      "learning_rate": 3.780117779740305e-07,
+      "loss": 0.0183,
+      "step": 13846
+    },
+    {
+      "epoch": 3.1470454545454545,
+      "grad_norm": 2.379375956120999,
+      "learning_rate": 3.779297953479338e-07,
+      "loss": 0.0134,
+      "step": 13847
+    },
+    {
+      "epoch": 3.1472727272727274,
+      "grad_norm": 0.8416323833294839,
+      "learning_rate": 3.778478177600296e-07,
+      "loss": 0.0047,
+      "step": 13848
+    },
+    {
+      "epoch": 3.1475,
+      "grad_norm": 0.8364709010394765,
+      "learning_rate": 3.777658452119896e-07,
+      "loss": 0.0044,
+      "step": 13849
+    },
+    {
+      "epoch": 3.147727272727273,
+      "grad_norm": 1.191985041603342,
+      "learning_rate": 3.7768387770548565e-07,
+      "loss": 0.0044,
+      "step": 13850
+    },
+    {
+      "epoch": 3.1479545454545454,
+      "grad_norm": 0.5439779341448465,
+      "learning_rate": 3.7760191524218864e-07,
+      "loss": 0.0069,
+      "step": 13851
+    },
+    {
+      "epoch": 3.1481818181818184,
+      "grad_norm": 2.0038029162053226,
+      "learning_rate": 3.775199578237701e-07,
+      "loss": 0.0152,
+      "step": 13852
+    },
+    {
+      "epoch": 3.148409090909091,
+      "grad_norm": 1.5053107314970862,
+      "learning_rate": 3.774380054519014e-07,
+      "loss": 0.0072,
+      "step": 13853
+    },
+    {
+      "epoch": 3.1486363636363635,
+      "grad_norm": 3.116440288383907,
+      "learning_rate": 3.773560581282537e-07,
+      "loss": 0.0201,
+      "step": 13854
+    },
+    {
+      "epoch": 3.1488636363636364,
+      "grad_norm": 0.7240287608621923,
+      "learning_rate": 3.772741158544982e-07,
+      "loss": 0.0049,
+      "step": 13855
+    },
+    {
+      "epoch": 3.149090909090909,
+      "grad_norm": 1.4569230085325353,
+      "learning_rate": 3.7719217863230545e-07,
+      "loss": 0.0093,
+      "step": 13856
+    },
+    {
+      "epoch": 3.149318181818182,
+      "grad_norm": 1.4896038300605825,
+      "learning_rate": 3.771102464633465e-07,
+      "loss": 0.013,
+      "step": 13857
+    },
+    {
+      "epoch": 3.1495454545454544,
+      "grad_norm": 3.162969118141545,
+      "learning_rate": 3.770283193492921e-07,
+      "loss": 0.0144,
+      "step": 13858
+    },
+    {
+      "epoch": 3.1497727272727274,
+      "grad_norm": 1.3809357936832471,
+      "learning_rate": 3.769463972918131e-07,
+      "loss": 0.0253,
+      "step": 13859
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 1.0564855285602448,
+      "learning_rate": 3.768644802925796e-07,
+      "loss": 0.0038,
+      "step": 13860
+    },
+    {
+      "epoch": 3.150227272727273,
+      "grad_norm": 1.5975304825055336,
+      "learning_rate": 3.767825683532622e-07,
+      "loss": 0.0091,
+      "step": 13861
+    },
+    {
+      "epoch": 3.1504545454545454,
+      "grad_norm": 1.6055955824290509,
+      "learning_rate": 3.7670066147553133e-07,
+      "loss": 0.0095,
+      "step": 13862
+    },
+    {
+      "epoch": 3.1506818181818184,
+      "grad_norm": 1.990233529700017,
+      "learning_rate": 3.76618759661057e-07,
+      "loss": 0.0136,
+      "step": 13863
+    },
+    {
+      "epoch": 3.150909090909091,
+      "grad_norm": 0.6598401315869212,
+      "learning_rate": 3.765368629115099e-07,
+      "loss": 0.0043,
+      "step": 13864
+    },
+    {
+      "epoch": 3.151136363636364,
+      "grad_norm": 0.7451994449029926,
+      "learning_rate": 3.764549712285594e-07,
+      "loss": 0.0082,
+      "step": 13865
+    },
+    {
+      "epoch": 3.1513636363636364,
+      "grad_norm": 1.5895165737164525,
+      "learning_rate": 3.7637308461387553e-07,
+      "loss": 0.007,
+      "step": 13866
+    },
+    {
+      "epoch": 3.151590909090909,
+      "grad_norm": 1.7762314686985912,
+      "learning_rate": 3.7629120306912833e-07,
+      "loss": 0.0129,
+      "step": 13867
+    },
+    {
+      "epoch": 3.151818181818182,
+      "grad_norm": 1.1370517896293666,
+      "learning_rate": 3.762093265959874e-07,
+      "loss": 0.0047,
+      "step": 13868
+    },
+    {
+      "epoch": 3.1520454545454544,
+      "grad_norm": 1.2806248002004958,
+      "learning_rate": 3.761274551961223e-07,
+      "loss": 0.0083,
+      "step": 13869
+    },
+    {
+      "epoch": 3.1522727272727273,
+      "grad_norm": 0.880861910612735,
+      "learning_rate": 3.760455888712026e-07,
+      "loss": 0.0216,
+      "step": 13870
+    },
+    {
+      "epoch": 3.1525,
+      "grad_norm": 3.409670272392007,
+      "learning_rate": 3.7596372762289766e-07,
+      "loss": 0.007,
+      "step": 13871
+    },
+    {
+      "epoch": 3.152727272727273,
+      "grad_norm": 1.792680339257585,
+      "learning_rate": 3.75881871452877e-07,
+      "loss": 0.0109,
+      "step": 13872
+    },
+    {
+      "epoch": 3.1529545454545453,
+      "grad_norm": 0.5815958339956084,
+      "learning_rate": 3.758000203628093e-07,
+      "loss": 0.0024,
+      "step": 13873
+    },
+    {
+      "epoch": 3.1531818181818183,
+      "grad_norm": 2.716374175652674,
+      "learning_rate": 3.75718174354364e-07,
+      "loss": 0.0209,
+      "step": 13874
+    },
+    {
+      "epoch": 3.153409090909091,
+      "grad_norm": 3.6164243197759833,
+      "learning_rate": 3.7563633342921e-07,
+      "loss": 0.0218,
+      "step": 13875
+    },
+    {
+      "epoch": 3.153636363636364,
+      "grad_norm": 1.3132103798901777,
+      "learning_rate": 3.755544975890162e-07,
+      "loss": 0.0051,
+      "step": 13876
+    },
+    {
+      "epoch": 3.1538636363636363,
+      "grad_norm": 2.883477010993954,
+      "learning_rate": 3.754726668354515e-07,
+      "loss": 0.0334,
+      "step": 13877
+    },
+    {
+      "epoch": 3.1540909090909093,
+      "grad_norm": 1.0471205744501484,
+      "learning_rate": 3.753908411701843e-07,
+      "loss": 0.0068,
+      "step": 13878
+    },
+    {
+      "epoch": 3.154318181818182,
+      "grad_norm": 2.0052898526706993,
+      "learning_rate": 3.753090205948834e-07,
+      "loss": 0.0051,
+      "step": 13879
+    },
+    {
+      "epoch": 3.1545454545454543,
+      "grad_norm": 1.8315760333233702,
+      "learning_rate": 3.7522720511121705e-07,
+      "loss": 0.0108,
+      "step": 13880
+    },
+    {
+      "epoch": 3.1547727272727273,
+      "grad_norm": 0.5640079723614654,
+      "learning_rate": 3.751453947208541e-07,
+      "loss": 0.0032,
+      "step": 13881
+    },
+    {
+      "epoch": 3.155,
+      "grad_norm": 1.5270363149689166,
+      "learning_rate": 3.7506358942546213e-07,
+      "loss": 0.0294,
+      "step": 13882
+    },
+    {
+      "epoch": 3.1552272727272728,
+      "grad_norm": 1.7384185377034977,
+      "learning_rate": 3.7498178922670966e-07,
+      "loss": 0.0081,
+      "step": 13883
+    },
+    {
+      "epoch": 3.1554545454545453,
+      "grad_norm": 2.566706249090651,
+      "learning_rate": 3.748999941262646e-07,
+      "loss": 0.0177,
+      "step": 13884
+    },
+    {
+      "epoch": 3.1556818181818183,
+      "grad_norm": 1.4090219972612215,
+      "learning_rate": 3.748182041257952e-07,
+      "loss": 0.0102,
+      "step": 13885
+    },
+    {
+      "epoch": 3.1559090909090908,
+      "grad_norm": 0.5049744070162598,
+      "learning_rate": 3.7473641922696895e-07,
+      "loss": 0.0027,
+      "step": 13886
+    },
+    {
+      "epoch": 3.1561363636363637,
+      "grad_norm": 1.251226223548957,
+      "learning_rate": 3.7465463943145373e-07,
+      "loss": 0.0118,
+      "step": 13887
+    },
+    {
+      "epoch": 3.1563636363636363,
+      "grad_norm": 1.1684103608101868,
+      "learning_rate": 3.7457286474091704e-07,
+      "loss": 0.0053,
+      "step": 13888
+    },
+    {
+      "epoch": 3.1565909090909092,
+      "grad_norm": 1.2439066294707688,
+      "learning_rate": 3.7449109515702674e-07,
+      "loss": 0.0035,
+      "step": 13889
+    },
+    {
+      "epoch": 3.1568181818181817,
+      "grad_norm": 0.9628657334930768,
+      "learning_rate": 3.7440933068145016e-07,
+      "loss": 0.0071,
+      "step": 13890
+    },
+    {
+      "epoch": 3.1570454545454547,
+      "grad_norm": 2.2278251505463054,
+      "learning_rate": 3.743275713158542e-07,
+      "loss": 0.0227,
+      "step": 13891
+    },
+    {
+      "epoch": 3.1572727272727272,
+      "grad_norm": 0.5304511795919893,
+      "learning_rate": 3.7424581706190633e-07,
+      "loss": 0.0062,
+      "step": 13892
+    },
+    {
+      "epoch": 3.1575,
+      "grad_norm": 2.5688557282875992,
+      "learning_rate": 3.7416406792127395e-07,
+      "loss": 0.0133,
+      "step": 13893
+    },
+    {
+      "epoch": 3.1577272727272727,
+      "grad_norm": 1.576671996527452,
+      "learning_rate": 3.74082323895624e-07,
+      "loss": 0.0107,
+      "step": 13894
+    },
+    {
+      "epoch": 3.1579545454545457,
+      "grad_norm": 2.49956639134313,
+      "learning_rate": 3.740005849866229e-07,
+      "loss": 0.0068,
+      "step": 13895
+    },
+    {
+      "epoch": 3.158181818181818,
+      "grad_norm": 1.6353691400284898,
+      "learning_rate": 3.739188511959378e-07,
+      "loss": 0.0096,
+      "step": 13896
+    },
+    {
+      "epoch": 3.1584090909090907,
+      "grad_norm": 1.159482158564537,
+      "learning_rate": 3.738371225252354e-07,
+      "loss": 0.0118,
+      "step": 13897
+    },
+    {
+      "epoch": 3.1586363636363637,
+      "grad_norm": 1.222518856093951,
+      "learning_rate": 3.737553989761822e-07,
+      "loss": 0.0125,
+      "step": 13898
+    },
+    {
+      "epoch": 3.158863636363636,
+      "grad_norm": 2.19217057996327,
+      "learning_rate": 3.7367368055044483e-07,
+      "loss": 0.0181,
+      "step": 13899
+    },
+    {
+      "epoch": 3.159090909090909,
+      "grad_norm": 1.4746138154678226,
+      "learning_rate": 3.735919672496896e-07,
+      "loss": 0.0162,
+      "step": 13900
+    },
+    {
+      "epoch": 3.1593181818181817,
+      "grad_norm": 3.080814182221813,
+      "learning_rate": 3.7351025907558266e-07,
+      "loss": 0.0114,
+      "step": 13901
+    },
+    {
+      "epoch": 3.1595454545454547,
+      "grad_norm": 0.7586924413534586,
+      "learning_rate": 3.7342855602979023e-07,
+      "loss": 0.0079,
+      "step": 13902
+    },
+    {
+      "epoch": 3.159772727272727,
+      "grad_norm": 1.7835852340919771,
+      "learning_rate": 3.7334685811397874e-07,
+      "loss": 0.0086,
+      "step": 13903
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 1.1326453057880965,
+      "learning_rate": 3.732651653298136e-07,
+      "loss": 0.0071,
+      "step": 13904
+    },
+    {
+      "epoch": 3.1602272727272727,
+      "grad_norm": 0.6495111487135538,
+      "learning_rate": 3.7318347767896093e-07,
+      "loss": 0.0039,
+      "step": 13905
+    },
+    {
+      "epoch": 3.1604545454545456,
+      "grad_norm": 1.0900081248116107,
+      "learning_rate": 3.7310179516308653e-07,
+      "loss": 0.0054,
+      "step": 13906
+    },
+    {
+      "epoch": 3.160681818181818,
+      "grad_norm": 1.7068150378027598,
+      "learning_rate": 3.730201177838559e-07,
+      "loss": 0.0101,
+      "step": 13907
+    },
+    {
+      "epoch": 3.160909090909091,
+      "grad_norm": 1.6611111418198257,
+      "learning_rate": 3.7293844554293475e-07,
+      "loss": 0.0072,
+      "step": 13908
+    },
+    {
+      "epoch": 3.1611363636363636,
+      "grad_norm": 1.7276178454511695,
+      "learning_rate": 3.7285677844198833e-07,
+      "loss": 0.0153,
+      "step": 13909
+    },
+    {
+      "epoch": 3.161363636363636,
+      "grad_norm": 1.9664536731307414,
+      "learning_rate": 3.727751164826822e-07,
+      "loss": 0.0077,
+      "step": 13910
+    },
+    {
+      "epoch": 3.161590909090909,
+      "grad_norm": 1.8628881255407677,
+      "learning_rate": 3.7269345966668135e-07,
+      "loss": 0.0127,
+      "step": 13911
+    },
+    {
+      "epoch": 3.1618181818181816,
+      "grad_norm": 1.1308282446279674,
+      "learning_rate": 3.726118079956513e-07,
+      "loss": 0.0078,
+      "step": 13912
+    },
+    {
+      "epoch": 3.1620454545454546,
+      "grad_norm": 0.775355939405654,
+      "learning_rate": 3.7253016147125664e-07,
+      "loss": 0.004,
+      "step": 13913
+    },
+    {
+      "epoch": 3.162272727272727,
+      "grad_norm": 1.2401890626147685,
+      "learning_rate": 3.724485200951623e-07,
+      "loss": 0.0104,
+      "step": 13914
+    },
+    {
+      "epoch": 3.1625,
+      "grad_norm": 1.0038327825616955,
+      "learning_rate": 3.7236688386903337e-07,
+      "loss": 0.0068,
+      "step": 13915
+    },
+    {
+      "epoch": 3.1627272727272726,
+      "grad_norm": 1.8007990442972597,
+      "learning_rate": 3.7228525279453447e-07,
+      "loss": 0.0201,
+      "step": 13916
+    },
+    {
+      "epoch": 3.1629545454545456,
+      "grad_norm": 0.6827018833069276,
+      "learning_rate": 3.722036268733301e-07,
+      "loss": 0.0077,
+      "step": 13917
+    },
+    {
+      "epoch": 3.163181818181818,
+      "grad_norm": 1.3305272960225847,
+      "learning_rate": 3.7212200610708476e-07,
+      "loss": 0.0041,
+      "step": 13918
+    },
+    {
+      "epoch": 3.163409090909091,
+      "grad_norm": 0.8087112762819502,
+      "learning_rate": 3.720403904974629e-07,
+      "loss": 0.0067,
+      "step": 13919
+    },
+    {
+      "epoch": 3.1636363636363636,
+      "grad_norm": 2.162480327343152,
+      "learning_rate": 3.7195878004612875e-07,
+      "loss": 0.0141,
+      "step": 13920
+    },
+    {
+      "epoch": 3.1638636363636365,
+      "grad_norm": 1.6511622122031502,
+      "learning_rate": 3.7187717475474673e-07,
+      "loss": 0.0072,
+      "step": 13921
+    },
+    {
+      "epoch": 3.164090909090909,
+      "grad_norm": 2.289512643541731,
+      "learning_rate": 3.717955746249805e-07,
+      "loss": 0.0114,
+      "step": 13922
+    },
+    {
+      "epoch": 3.1643181818181816,
+      "grad_norm": 3.3734917731112373,
+      "learning_rate": 3.717139796584943e-07,
+      "loss": 0.0229,
+      "step": 13923
+    },
+    {
+      "epoch": 3.1645454545454546,
+      "grad_norm": 1.2559634364769596,
+      "learning_rate": 3.716323898569519e-07,
+      "loss": 0.0123,
+      "step": 13924
+    },
+    {
+      "epoch": 3.164772727272727,
+      "grad_norm": 1.4698902157676363,
+      "learning_rate": 3.7155080522201736e-07,
+      "loss": 0.0117,
+      "step": 13925
+    },
+    {
+      "epoch": 3.165,
+      "grad_norm": 1.084798290884996,
+      "learning_rate": 3.7146922575535383e-07,
+      "loss": 0.0059,
+      "step": 13926
+    },
+    {
+      "epoch": 3.1652272727272726,
+      "grad_norm": 1.3526352314584138,
+      "learning_rate": 3.713876514586251e-07,
+      "loss": 0.0105,
+      "step": 13927
+    },
+    {
+      "epoch": 3.1654545454545455,
+      "grad_norm": 2.212883531517818,
+      "learning_rate": 3.7130608233349464e-07,
+      "loss": 0.0141,
+      "step": 13928
+    },
+    {
+      "epoch": 3.165681818181818,
+      "grad_norm": 1.1834793473764134,
+      "learning_rate": 3.7122451838162576e-07,
+      "loss": 0.0046,
+      "step": 13929
+    },
+    {
+      "epoch": 3.165909090909091,
+      "grad_norm": 1.4791560616202448,
+      "learning_rate": 3.7114295960468164e-07,
+      "loss": 0.0041,
+      "step": 13930
+    },
+    {
+      "epoch": 3.1661363636363635,
+      "grad_norm": 1.430286751347182,
+      "learning_rate": 3.710614060043255e-07,
+      "loss": 0.0066,
+      "step": 13931
+    },
+    {
+      "epoch": 3.1663636363636365,
+      "grad_norm": 1.1599384698659174,
+      "learning_rate": 3.7097985758222023e-07,
+      "loss": 0.0064,
+      "step": 13932
+    },
+    {
+      "epoch": 3.166590909090909,
+      "grad_norm": 1.6053769378797795,
+      "learning_rate": 3.708983143400288e-07,
+      "loss": 0.0091,
+      "step": 13933
+    },
+    {
+      "epoch": 3.166818181818182,
+      "grad_norm": 1.6430376848684458,
+      "learning_rate": 3.708167762794144e-07,
+      "loss": 0.013,
+      "step": 13934
+    },
+    {
+      "epoch": 3.1670454545454545,
+      "grad_norm": 1.4383771768587679,
+      "learning_rate": 3.7073524340203905e-07,
+      "loss": 0.009,
+      "step": 13935
+    },
+    {
+      "epoch": 3.1672727272727275,
+      "grad_norm": 2.405808485299479,
+      "learning_rate": 3.706537157095658e-07,
+      "loss": 0.0217,
+      "step": 13936
+    },
+    {
+      "epoch": 3.1675,
+      "grad_norm": 1.1591884170335875,
+      "learning_rate": 3.705721932036568e-07,
+      "loss": 0.0045,
+      "step": 13937
+    },
+    {
+      "epoch": 3.167727272727273,
+      "grad_norm": 0.7501357888485806,
+      "learning_rate": 3.7049067588597496e-07,
+      "loss": 0.0062,
+      "step": 13938
+    },
+    {
+      "epoch": 3.1679545454545455,
+      "grad_norm": 1.8799867047758203,
+      "learning_rate": 3.704091637581821e-07,
+      "loss": 0.0119,
+      "step": 13939
+    },
+    {
+      "epoch": 3.168181818181818,
+      "grad_norm": 0.9119259296988428,
+      "learning_rate": 3.7032765682194047e-07,
+      "loss": 0.0035,
+      "step": 13940
+    },
+    {
+      "epoch": 3.168409090909091,
+      "grad_norm": 0.3107798139111199,
+      "learning_rate": 3.702461550789123e-07,
+      "loss": 0.0019,
+      "step": 13941
+    },
+    {
+      "epoch": 3.1686363636363635,
+      "grad_norm": 2.211465552631953,
+      "learning_rate": 3.7016465853075946e-07,
+      "loss": 0.0178,
+      "step": 13942
+    },
+    {
+      "epoch": 3.1688636363636364,
+      "grad_norm": 1.155490295001904,
+      "learning_rate": 3.7008316717914396e-07,
+      "loss": 0.0129,
+      "step": 13943
+    },
+    {
+      "epoch": 3.169090909090909,
+      "grad_norm": 1.2850751925605692,
+      "learning_rate": 3.700016810257273e-07,
+      "loss": 0.0077,
+      "step": 13944
+    },
+    {
+      "epoch": 3.169318181818182,
+      "grad_norm": 2.217092649068937,
+      "learning_rate": 3.699202000721712e-07,
+      "loss": 0.0167,
+      "step": 13945
+    },
+    {
+      "epoch": 3.1695454545454544,
+      "grad_norm": 0.5472515296194284,
+      "learning_rate": 3.6983872432013707e-07,
+      "loss": 0.0037,
+      "step": 13946
+    },
+    {
+      "epoch": 3.1697727272727274,
+      "grad_norm": 2.283666568859524,
+      "learning_rate": 3.6975725377128685e-07,
+      "loss": 0.019,
+      "step": 13947
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 1.4521780031278495,
+      "learning_rate": 3.6967578842728136e-07,
+      "loss": 0.0117,
+      "step": 13948
+    },
+    {
+      "epoch": 3.170227272727273,
+      "grad_norm": 1.1801167237852668,
+      "learning_rate": 3.6959432828978193e-07,
+      "loss": 0.006,
+      "step": 13949
+    },
+    {
+      "epoch": 3.1704545454545454,
+      "grad_norm": 1.5649065713398072,
+      "learning_rate": 3.695128733604497e-07,
+      "loss": 0.0121,
+      "step": 13950
+    },
+    {
+      "epoch": 3.1706818181818184,
+      "grad_norm": 1.6124229663163452,
+      "learning_rate": 3.694314236409459e-07,
+      "loss": 0.0099,
+      "step": 13951
+    },
+    {
+      "epoch": 3.170909090909091,
+      "grad_norm": 1.7088744041842632,
+      "learning_rate": 3.6934997913293106e-07,
+      "loss": 0.0088,
+      "step": 13952
+    },
+    {
+      "epoch": 3.1711363636363634,
+      "grad_norm": 2.6154684765550362,
+      "learning_rate": 3.6926853983806617e-07,
+      "loss": 0.0195,
+      "step": 13953
+    },
+    {
+      "epoch": 3.1713636363636364,
+      "grad_norm": 1.222619021991547,
+      "learning_rate": 3.6918710575801187e-07,
+      "loss": 0.0081,
+      "step": 13954
+    },
+    {
+      "epoch": 3.171590909090909,
+      "grad_norm": 1.7355593912598613,
+      "learning_rate": 3.6910567689442877e-07,
+      "loss": 0.0104,
+      "step": 13955
+    },
+    {
+      "epoch": 3.171818181818182,
+      "grad_norm": 1.788627396058333,
+      "learning_rate": 3.690242532489777e-07,
+      "loss": 0.0114,
+      "step": 13956
+    },
+    {
+      "epoch": 3.1720454545454544,
+      "grad_norm": 1.9223528188922794,
+      "learning_rate": 3.689428348233183e-07,
+      "loss": 0.0106,
+      "step": 13957
+    },
+    {
+      "epoch": 3.1722727272727274,
+      "grad_norm": 2.8625889010086687,
+      "learning_rate": 3.688614216191112e-07,
+      "loss": 0.0214,
+      "step": 13958
+    },
+    {
+      "epoch": 3.1725,
+      "grad_norm": 0.905456614807084,
+      "learning_rate": 3.687800136380167e-07,
+      "loss": 0.0074,
+      "step": 13959
+    },
+    {
+      "epoch": 3.172727272727273,
+      "grad_norm": 1.4570739582845589,
+      "learning_rate": 3.686986108816948e-07,
+      "loss": 0.0102,
+      "step": 13960
+    },
+    {
+      "epoch": 3.1729545454545454,
+      "grad_norm": 0.8856614771999769,
+      "learning_rate": 3.6861721335180526e-07,
+      "loss": 0.0037,
+      "step": 13961
+    },
+    {
+      "epoch": 3.1731818181818183,
+      "grad_norm": 1.0594677243456414,
+      "learning_rate": 3.68535821050008e-07,
+      "loss": 0.0046,
+      "step": 13962
+    },
+    {
+      "epoch": 3.173409090909091,
+      "grad_norm": 1.4328582509857677,
+      "learning_rate": 3.684544339779628e-07,
+      "loss": 0.0126,
+      "step": 13963
+    },
+    {
+      "epoch": 3.173636363636364,
+      "grad_norm": 0.43466413222336403,
+      "learning_rate": 3.6837305213732915e-07,
+      "loss": 0.0017,
+      "step": 13964
+    },
+    {
+      "epoch": 3.1738636363636363,
+      "grad_norm": 0.4465265326477732,
+      "learning_rate": 3.6829167552976703e-07,
+      "loss": 0.0019,
+      "step": 13965
+    },
+    {
+      "epoch": 3.1740909090909093,
+      "grad_norm": 0.6961036943462516,
+      "learning_rate": 3.682103041569352e-07,
+      "loss": 0.0049,
+      "step": 13966
+    },
+    {
+      "epoch": 3.174318181818182,
+      "grad_norm": 1.4255272537820387,
+      "learning_rate": 3.6812893802049334e-07,
+      "loss": 0.0054,
+      "step": 13967
+    },
+    {
+      "epoch": 3.174545454545455,
+      "grad_norm": 1.6447341203180657,
+      "learning_rate": 3.680475771221005e-07,
+      "loss": 0.0082,
+      "step": 13968
+    },
+    {
+      "epoch": 3.1747727272727273,
+      "grad_norm": 1.386220325266305,
+      "learning_rate": 3.6796622146341594e-07,
+      "loss": 0.0063,
+      "step": 13969
+    },
+    {
+      "epoch": 3.175,
+      "grad_norm": 1.121271760806407,
+      "learning_rate": 3.6788487104609847e-07,
+      "loss": 0.0069,
+      "step": 13970
+    },
+    {
+      "epoch": 3.175227272727273,
+      "grad_norm": 0.5566696532893497,
+      "learning_rate": 3.6780352587180705e-07,
+      "loss": 0.0032,
+      "step": 13971
+    },
+    {
+      "epoch": 3.1754545454545453,
+      "grad_norm": 1.2276229457750434,
+      "learning_rate": 3.677221859422005e-07,
+      "loss": 0.0069,
+      "step": 13972
+    },
+    {
+      "epoch": 3.1756818181818183,
+      "grad_norm": 3.179393154790371,
+      "learning_rate": 3.676408512589376e-07,
+      "loss": 0.0369,
+      "step": 13973
+    },
+    {
+      "epoch": 3.175909090909091,
+      "grad_norm": 1.3644861684690703,
+      "learning_rate": 3.6755952182367644e-07,
+      "loss": 0.0155,
+      "step": 13974
+    },
+    {
+      "epoch": 3.1761363636363638,
+      "grad_norm": 0.88767635570413,
+      "learning_rate": 3.6747819763807577e-07,
+      "loss": 0.0082,
+      "step": 13975
+    },
+    {
+      "epoch": 3.1763636363636363,
+      "grad_norm": 3.2777407838756054,
+      "learning_rate": 3.673968787037938e-07,
+      "loss": 0.0256,
+      "step": 13976
+    },
+    {
+      "epoch": 3.1765909090909092,
+      "grad_norm": 1.2505261033719701,
+      "learning_rate": 3.673155650224891e-07,
+      "loss": 0.0083,
+      "step": 13977
+    },
+    {
+      "epoch": 3.1768181818181818,
+      "grad_norm": 0.7634311862394939,
+      "learning_rate": 3.672342565958198e-07,
+      "loss": 0.006,
+      "step": 13978
+    },
+    {
+      "epoch": 3.1770454545454547,
+      "grad_norm": 0.6772776247861789,
+      "learning_rate": 3.671529534254434e-07,
+      "loss": 0.0027,
+      "step": 13979
+    },
+    {
+      "epoch": 3.1772727272727272,
+      "grad_norm": 0.4836408686581746,
+      "learning_rate": 3.6707165551301814e-07,
+      "loss": 0.0037,
+      "step": 13980
+    },
+    {
+      "epoch": 3.1775,
+      "grad_norm": 1.0743275864206316,
+      "learning_rate": 3.669903628602018e-07,
+      "loss": 0.0048,
+      "step": 13981
+    },
+    {
+      "epoch": 3.1777272727272727,
+      "grad_norm": 8.550641646005047,
+      "learning_rate": 3.669090754686522e-07,
+      "loss": 0.0109,
+      "step": 13982
+    },
+    {
+      "epoch": 3.1779545454545453,
+      "grad_norm": 1.8485898075117368,
+      "learning_rate": 3.6682779334002676e-07,
+      "loss": 0.0132,
+      "step": 13983
+    },
+    {
+      "epoch": 3.178181818181818,
+      "grad_norm": 3.336968406315612,
+      "learning_rate": 3.6674651647598295e-07,
+      "loss": 0.0243,
+      "step": 13984
+    },
+    {
+      "epoch": 3.1784090909090907,
+      "grad_norm": 1.0494419856594397,
+      "learning_rate": 3.6666524487817824e-07,
+      "loss": 0.0082,
+      "step": 13985
+    },
+    {
+      "epoch": 3.1786363636363637,
+      "grad_norm": 2.250621746248187,
+      "learning_rate": 3.6658397854826984e-07,
+      "loss": 0.0228,
+      "step": 13986
+    },
+    {
+      "epoch": 3.1788636363636362,
+      "grad_norm": 3.6466332065444913,
+      "learning_rate": 3.6650271748791524e-07,
+      "loss": 0.0144,
+      "step": 13987
+    },
+    {
+      "epoch": 3.179090909090909,
+      "grad_norm": 2.5264392815683525,
+      "learning_rate": 3.66421461698771e-07,
+      "loss": 0.0202,
+      "step": 13988
+    },
+    {
+      "epoch": 3.1793181818181817,
+      "grad_norm": 1.7370695911654648,
+      "learning_rate": 3.6634021118249416e-07,
+      "loss": 0.0219,
+      "step": 13989
+    },
+    {
+      "epoch": 3.1795454545454547,
+      "grad_norm": 1.706049293417877,
+      "learning_rate": 3.6625896594074174e-07,
+      "loss": 0.0156,
+      "step": 13990
+    },
+    {
+      "epoch": 3.179772727272727,
+      "grad_norm": 1.9338008663555972,
+      "learning_rate": 3.661777259751705e-07,
+      "loss": 0.0285,
+      "step": 13991
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 1.01983660306618,
+      "learning_rate": 3.660964912874369e-07,
+      "loss": 0.0087,
+      "step": 13992
+    },
+    {
+      "epoch": 3.1802272727272727,
+      "grad_norm": 2.249249357835759,
+      "learning_rate": 3.660152618791975e-07,
+      "loss": 0.0145,
+      "step": 13993
+    },
+    {
+      "epoch": 3.1804545454545456,
+      "grad_norm": 0.8590495148009503,
+      "learning_rate": 3.659340377521087e-07,
+      "loss": 0.0034,
+      "step": 13994
+    },
+    {
+      "epoch": 3.180681818181818,
+      "grad_norm": 2.0481552718859017,
+      "learning_rate": 3.658528189078271e-07,
+      "loss": 0.0083,
+      "step": 13995
+    },
+    {
+      "epoch": 3.1809090909090907,
+      "grad_norm": 3.015160558059644,
+      "learning_rate": 3.657716053480084e-07,
+      "loss": 0.0091,
+      "step": 13996
+    },
+    {
+      "epoch": 3.1811363636363637,
+      "grad_norm": 1.3632745801276518,
+      "learning_rate": 3.656903970743089e-07,
+      "loss": 0.0083,
+      "step": 13997
+    },
+    {
+      "epoch": 3.181363636363636,
+      "grad_norm": 0.2452611438793574,
+      "learning_rate": 3.656091940883845e-07,
+      "loss": 0.0008,
+      "step": 13998
+    },
+    {
+      "epoch": 3.181590909090909,
+      "grad_norm": 1.6526099750617635,
+      "learning_rate": 3.6552799639189134e-07,
+      "loss": 0.012,
+      "step": 13999
+    },
+    {
+      "epoch": 3.1818181818181817,
+      "grad_norm": 1.8266044306726017,
+      "learning_rate": 3.6544680398648504e-07,
+      "loss": 0.0195,
+      "step": 14000
+    },
+    {
+      "epoch": 3.1820454545454546,
+      "grad_norm": 1.23698492337562,
+      "learning_rate": 3.653656168738211e-07,
+      "loss": 0.0073,
+      "step": 14001
+    },
+    {
+      "epoch": 3.182272727272727,
+      "grad_norm": 3.3933874887421163,
+      "learning_rate": 3.652844350555552e-07,
+      "loss": 0.0249,
+      "step": 14002
+    },
+    {
+      "epoch": 3.1825,
+      "grad_norm": 1.7542804379844643,
+      "learning_rate": 3.6520325853334274e-07,
+      "loss": 0.0123,
+      "step": 14003
+    },
+    {
+      "epoch": 3.1827272727272726,
+      "grad_norm": 1.4814477809470687,
+      "learning_rate": 3.651220873088393e-07,
+      "loss": 0.0095,
+      "step": 14004
+    },
+    {
+      "epoch": 3.1829545454545456,
+      "grad_norm": 4.459866751208436,
+      "learning_rate": 3.6504092138369957e-07,
+      "loss": 0.0519,
+      "step": 14005
+    },
+    {
+      "epoch": 3.183181818181818,
+      "grad_norm": 2.2066870087030654,
+      "learning_rate": 3.649597607595791e-07,
+      "loss": 0.0119,
+      "step": 14006
+    },
+    {
+      "epoch": 3.183409090909091,
+      "grad_norm": 1.436798998125274,
+      "learning_rate": 3.648786054381328e-07,
+      "loss": 0.0061,
+      "step": 14007
+    },
+    {
+      "epoch": 3.1836363636363636,
+      "grad_norm": 0.5104657940757688,
+      "learning_rate": 3.647974554210157e-07,
+      "loss": 0.0034,
+      "step": 14008
+    },
+    {
+      "epoch": 3.1838636363636366,
+      "grad_norm": 2.022955435065844,
+      "learning_rate": 3.6471631070988223e-07,
+      "loss": 0.0174,
+      "step": 14009
+    },
+    {
+      "epoch": 3.184090909090909,
+      "grad_norm": 1.0607091482535973,
+      "learning_rate": 3.6463517130638735e-07,
+      "loss": 0.0067,
+      "step": 14010
+    },
+    {
+      "epoch": 3.184318181818182,
+      "grad_norm": 1.2728445940367004,
+      "learning_rate": 3.6455403721218546e-07,
+      "loss": 0.0109,
+      "step": 14011
+    },
+    {
+      "epoch": 3.1845454545454546,
+      "grad_norm": 2.1404164599467586,
+      "learning_rate": 3.644729084289311e-07,
+      "loss": 0.0088,
+      "step": 14012
+    },
+    {
+      "epoch": 3.184772727272727,
+      "grad_norm": 0.8401808190161787,
+      "learning_rate": 3.64391784958279e-07,
+      "loss": 0.0054,
+      "step": 14013
+    },
+    {
+      "epoch": 3.185,
+      "grad_norm": 1.221459150587376,
+      "learning_rate": 3.643106668018828e-07,
+      "loss": 0.0082,
+      "step": 14014
+    },
+    {
+      "epoch": 3.1852272727272726,
+      "grad_norm": 2.5186726126014043,
+      "learning_rate": 3.6422955396139684e-07,
+      "loss": 0.0395,
+      "step": 14015
+    },
+    {
+      "epoch": 3.1854545454545455,
+      "grad_norm": 1.594940052177264,
+      "learning_rate": 3.641484464384752e-07,
+      "loss": 0.0141,
+      "step": 14016
+    },
+    {
+      "epoch": 3.185681818181818,
+      "grad_norm": 2.3059067155349124,
+      "learning_rate": 3.640673442347719e-07,
+      "loss": 0.0235,
+      "step": 14017
+    },
+    {
+      "epoch": 3.185909090909091,
+      "grad_norm": 3.3444522742191287,
+      "learning_rate": 3.6398624735194066e-07,
+      "loss": 0.0307,
+      "step": 14018
+    },
+    {
+      "epoch": 3.1861363636363635,
+      "grad_norm": 3.3217645593996354,
+      "learning_rate": 3.639051557916351e-07,
+      "loss": 0.0175,
+      "step": 14019
+    },
+    {
+      "epoch": 3.1863636363636365,
+      "grad_norm": 3.469931375192992,
+      "learning_rate": 3.6382406955550906e-07,
+      "loss": 0.0153,
+      "step": 14020
+    },
+    {
+      "epoch": 3.186590909090909,
+      "grad_norm": 0.7068279316613271,
+      "learning_rate": 3.637429886452158e-07,
+      "loss": 0.0031,
+      "step": 14021
+    },
+    {
+      "epoch": 3.186818181818182,
+      "grad_norm": 1.246914179912642,
+      "learning_rate": 3.6366191306240904e-07,
+      "loss": 0.0089,
+      "step": 14022
+    },
+    {
+      "epoch": 3.1870454545454545,
+      "grad_norm": 1.4614931867108338,
+      "learning_rate": 3.6358084280874157e-07,
+      "loss": 0.0084,
+      "step": 14023
+    },
+    {
+      "epoch": 3.1872727272727275,
+      "grad_norm": 1.1790561926887615,
+      "learning_rate": 3.634997778858668e-07,
+      "loss": 0.0054,
+      "step": 14024
+    },
+    {
+      "epoch": 3.1875,
+      "grad_norm": 3.671740895920797,
+      "learning_rate": 3.6341871829543775e-07,
+      "loss": 0.0071,
+      "step": 14025
+    },
+    {
+      "epoch": 3.1877272727272725,
+      "grad_norm": 2.841169055262012,
+      "learning_rate": 3.6333766403910764e-07,
+      "loss": 0.0193,
+      "step": 14026
+    },
+    {
+      "epoch": 3.1879545454545455,
+      "grad_norm": 1.9296564535665384,
+      "learning_rate": 3.632566151185288e-07,
+      "loss": 0.0122,
+      "step": 14027
+    },
+    {
+      "epoch": 3.188181818181818,
+      "grad_norm": 2.8280741644248577,
+      "learning_rate": 3.6317557153535423e-07,
+      "loss": 0.0231,
+      "step": 14028
+    },
+    {
+      "epoch": 3.188409090909091,
+      "grad_norm": 3.468979185516024,
+      "learning_rate": 3.6309453329123654e-07,
+      "loss": 0.0185,
+      "step": 14029
+    },
+    {
+      "epoch": 3.1886363636363635,
+      "grad_norm": 1.5098210254366633,
+      "learning_rate": 3.630135003878285e-07,
+      "loss": 0.0083,
+      "step": 14030
+    },
+    {
+      "epoch": 3.1888636363636365,
+      "grad_norm": 4.775824793090362,
+      "learning_rate": 3.6293247282678193e-07,
+      "loss": 0.0385,
+      "step": 14031
+    },
+    {
+      "epoch": 3.189090909090909,
+      "grad_norm": 1.4441725396295255,
+      "learning_rate": 3.628514506097493e-07,
+      "loss": 0.0074,
+      "step": 14032
+    },
+    {
+      "epoch": 3.189318181818182,
+      "grad_norm": 0.599966731373092,
+      "learning_rate": 3.6277043373838307e-07,
+      "loss": 0.0031,
+      "step": 14033
+    },
+    {
+      "epoch": 3.1895454545454545,
+      "grad_norm": 0.5094578736852562,
+      "learning_rate": 3.626894222143353e-07,
+      "loss": 0.003,
+      "step": 14034
+    },
+    {
+      "epoch": 3.1897727272727274,
+      "grad_norm": 0.7557296198412877,
+      "learning_rate": 3.6260841603925787e-07,
+      "loss": 0.0047,
+      "step": 14035
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 2.293007716077908,
+      "learning_rate": 3.625274152148026e-07,
+      "loss": 0.0177,
+      "step": 14036
+    },
+    {
+      "epoch": 3.190227272727273,
+      "grad_norm": 1.3023531141649742,
+      "learning_rate": 3.624464197426211e-07,
+      "loss": 0.0052,
+      "step": 14037
+    },
+    {
+      "epoch": 3.1904545454545454,
+      "grad_norm": 0.5204551714288095,
+      "learning_rate": 3.623654296243652e-07,
+      "loss": 0.0042,
+      "step": 14038
+    },
+    {
+      "epoch": 3.190681818181818,
+      "grad_norm": 1.409117961682039,
+      "learning_rate": 3.6228444486168663e-07,
+      "loss": 0.0236,
+      "step": 14039
+    },
+    {
+      "epoch": 3.190909090909091,
+      "grad_norm": 1.9549554855229199,
+      "learning_rate": 3.622034654562363e-07,
+      "loss": 0.0367,
+      "step": 14040
+    },
+    {
+      "epoch": 3.1911363636363634,
+      "grad_norm": 2.3311577143748763,
+      "learning_rate": 3.621224914096658e-07,
+      "loss": 0.0198,
+      "step": 14041
+    },
+    {
+      "epoch": 3.1913636363636364,
+      "grad_norm": 2.065659194299564,
+      "learning_rate": 3.620415227236262e-07,
+      "loss": 0.0268,
+      "step": 14042
+    },
+    {
+      "epoch": 3.191590909090909,
+      "grad_norm": 1.722718423131247,
+      "learning_rate": 3.619605593997689e-07,
+      "loss": 0.0094,
+      "step": 14043
+    },
+    {
+      "epoch": 3.191818181818182,
+      "grad_norm": 0.8653402585221314,
+      "learning_rate": 3.6187960143974477e-07,
+      "loss": 0.0043,
+      "step": 14044
+    },
+    {
+      "epoch": 3.1920454545454544,
+      "grad_norm": 1.440792868051795,
+      "learning_rate": 3.617986488452045e-07,
+      "loss": 0.0151,
+      "step": 14045
+    },
+    {
+      "epoch": 3.1922727272727274,
+      "grad_norm": 1.397955467367182,
+      "learning_rate": 3.617177016177989e-07,
+      "loss": 0.0095,
+      "step": 14046
+    },
+    {
+      "epoch": 3.1925,
+      "grad_norm": 1.293178228792711,
+      "learning_rate": 3.616367597591787e-07,
+      "loss": 0.0125,
+      "step": 14047
+    },
+    {
+      "epoch": 3.192727272727273,
+      "grad_norm": 1.0764235025268942,
+      "learning_rate": 3.615558232709944e-07,
+      "loss": 0.0097,
+      "step": 14048
+    },
+    {
+      "epoch": 3.1929545454545454,
+      "grad_norm": 1.6432455233654546,
+      "learning_rate": 3.6147489215489647e-07,
+      "loss": 0.0148,
+      "step": 14049
+    },
+    {
+      "epoch": 3.1931818181818183,
+      "grad_norm": 0.7399569946580774,
+      "learning_rate": 3.613939664125353e-07,
+      "loss": 0.0031,
+      "step": 14050
+    },
+    {
+      "epoch": 3.193409090909091,
+      "grad_norm": 2.4072422304511965,
+      "learning_rate": 3.6131304604556093e-07,
+      "loss": 0.0089,
+      "step": 14051
+    },
+    {
+      "epoch": 3.193636363636364,
+      "grad_norm": 2.3093464827222867,
+      "learning_rate": 3.6123213105562387e-07,
+      "loss": 0.0152,
+      "step": 14052
+    },
+    {
+      "epoch": 3.1938636363636363,
+      "grad_norm": 1.2632331754555113,
+      "learning_rate": 3.611512214443736e-07,
+      "loss": 0.006,
+      "step": 14053
+    },
+    {
+      "epoch": 3.1940909090909093,
+      "grad_norm": 0.9713398493799634,
+      "learning_rate": 3.6107031721346027e-07,
+      "loss": 0.0059,
+      "step": 14054
+    },
+    {
+      "epoch": 3.194318181818182,
+      "grad_norm": 1.9452987309538228,
+      "learning_rate": 3.609894183645335e-07,
+      "loss": 0.0161,
+      "step": 14055
+    },
+    {
+      "epoch": 3.1945454545454544,
+      "grad_norm": 1.2709701840775112,
+      "learning_rate": 3.6090852489924317e-07,
+      "loss": 0.0076,
+      "step": 14056
+    },
+    {
+      "epoch": 3.1947727272727273,
+      "grad_norm": 1.6118672760227664,
+      "learning_rate": 3.60827636819239e-07,
+      "loss": 0.0088,
+      "step": 14057
+    },
+    {
+      "epoch": 3.195,
+      "grad_norm": 2.664447963388216,
+      "learning_rate": 3.6074675412617e-07,
+      "loss": 0.0161,
+      "step": 14058
+    },
+    {
+      "epoch": 3.195227272727273,
+      "grad_norm": 1.6621493684031803,
+      "learning_rate": 3.606658768216856e-07,
+      "loss": 0.0146,
+      "step": 14059
+    },
+    {
+      "epoch": 3.1954545454545453,
+      "grad_norm": 0.5398853045437702,
+      "learning_rate": 3.6058500490743514e-07,
+      "loss": 0.0047,
+      "step": 14060
+    },
+    {
+      "epoch": 3.1956818181818183,
+      "grad_norm": 1.1417552450711652,
+      "learning_rate": 3.6050413838506796e-07,
+      "loss": 0.0061,
+      "step": 14061
+    },
+    {
+      "epoch": 3.195909090909091,
+      "grad_norm": 0.6737871959104917,
+      "learning_rate": 3.6042327725623236e-07,
+      "loss": 0.0042,
+      "step": 14062
+    },
+    {
+      "epoch": 3.1961363636363638,
+      "grad_norm": 0.7458248701993194,
+      "learning_rate": 3.60342421522578e-07,
+      "loss": 0.0033,
+      "step": 14063
+    },
+    {
+      "epoch": 3.1963636363636363,
+      "grad_norm": 3.300151588901975,
+      "learning_rate": 3.602615711857534e-07,
+      "loss": 0.0143,
+      "step": 14064
+    },
+    {
+      "epoch": 3.1965909090909093,
+      "grad_norm": 1.0560953311166559,
+      "learning_rate": 3.601807262474072e-07,
+      "loss": 0.0047,
+      "step": 14065
+    },
+    {
+      "epoch": 3.196818181818182,
+      "grad_norm": 2.1260571063477394,
+      "learning_rate": 3.6009988670918814e-07,
+      "loss": 0.0104,
+      "step": 14066
+    },
+    {
+      "epoch": 3.1970454545454547,
+      "grad_norm": 1.749753635365068,
+      "learning_rate": 3.6001905257274443e-07,
+      "loss": 0.0142,
+      "step": 14067
+    },
+    {
+      "epoch": 3.1972727272727273,
+      "grad_norm": 1.3270310477787184,
+      "learning_rate": 3.599382238397244e-07,
+      "loss": 0.0055,
+      "step": 14068
+    },
+    {
+      "epoch": 3.1975,
+      "grad_norm": 1.3925295999469,
+      "learning_rate": 3.598574005117765e-07,
+      "loss": 0.0071,
+      "step": 14069
+    },
+    {
+      "epoch": 3.1977272727272728,
+      "grad_norm": 1.7570555717895207,
+      "learning_rate": 3.59776582590549e-07,
+      "loss": 0.0168,
+      "step": 14070
+    },
+    {
+      "epoch": 3.1979545454545453,
+      "grad_norm": 1.9130513862684102,
+      "learning_rate": 3.596957700776894e-07,
+      "loss": 0.0065,
+      "step": 14071
+    },
+    {
+      "epoch": 3.1981818181818182,
+      "grad_norm": 1.533784297248602,
+      "learning_rate": 3.5961496297484596e-07,
+      "loss": 0.0093,
+      "step": 14072
+    },
+    {
+      "epoch": 3.1984090909090908,
+      "grad_norm": 0.5386770324575121,
+      "learning_rate": 3.595341612836664e-07,
+      "loss": 0.0056,
+      "step": 14073
+    },
+    {
+      "epoch": 3.1986363636363637,
+      "grad_norm": 1.1206692862003258,
+      "learning_rate": 3.5945336500579867e-07,
+      "loss": 0.0078,
+      "step": 14074
+    },
+    {
+      "epoch": 3.1988636363636362,
+      "grad_norm": 3.1960044223574013,
+      "learning_rate": 3.593725741428898e-07,
+      "loss": 0.0271,
+      "step": 14075
+    },
+    {
+      "epoch": 3.199090909090909,
+      "grad_norm": 0.4584256488505626,
+      "learning_rate": 3.5929178869658757e-07,
+      "loss": 0.0018,
+      "step": 14076
+    },
+    {
+      "epoch": 3.1993181818181817,
+      "grad_norm": 1.1378383955009221,
+      "learning_rate": 3.5921100866853934e-07,
+      "loss": 0.0087,
+      "step": 14077
+    },
+    {
+      "epoch": 3.1995454545454547,
+      "grad_norm": 1.842451450321353,
+      "learning_rate": 3.5913023406039235e-07,
+      "loss": 0.0095,
+      "step": 14078
+    },
+    {
+      "epoch": 3.199772727272727,
+      "grad_norm": 3.66744591121373,
+      "learning_rate": 3.590494648737938e-07,
+      "loss": 0.0104,
+      "step": 14079
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 1.8947457092936462,
+      "learning_rate": 3.589687011103905e-07,
+      "loss": 0.015,
+      "step": 14080
+    },
+    {
+      "epoch": 3.2002272727272727,
+      "grad_norm": 1.933227155287281,
+      "learning_rate": 3.588879427718296e-07,
+      "loss": 0.0142,
+      "step": 14081
+    },
+    {
+      "epoch": 3.200454545454545,
+      "grad_norm": 1.9664571632210346,
+      "learning_rate": 3.588071898597577e-07,
+      "loss": 0.0161,
+      "step": 14082
+    },
+    {
+      "epoch": 3.200681818181818,
+      "grad_norm": 2.4256276737679716,
+      "learning_rate": 3.587264423758219e-07,
+      "loss": 0.0199,
+      "step": 14083
+    },
+    {
+      "epoch": 3.2009090909090907,
+      "grad_norm": 1.3973726482263615,
+      "learning_rate": 3.5864570032166824e-07,
+      "loss": 0.0056,
+      "step": 14084
+    },
+    {
+      "epoch": 3.2011363636363637,
+      "grad_norm": 0.9667302013805353,
+      "learning_rate": 3.5856496369894343e-07,
+      "loss": 0.0081,
+      "step": 14085
+    },
+    {
+      "epoch": 3.201363636363636,
+      "grad_norm": 0.6194557641774832,
+      "learning_rate": 3.584842325092938e-07,
+      "loss": 0.0072,
+      "step": 14086
+    },
+    {
+      "epoch": 3.201590909090909,
+      "grad_norm": 1.1422690270663303,
+      "learning_rate": 3.584035067543657e-07,
+      "loss": 0.0044,
+      "step": 14087
+    },
+    {
+      "epoch": 3.2018181818181817,
+      "grad_norm": 1.0204035003805452,
+      "learning_rate": 3.5832278643580543e-07,
+      "loss": 0.0054,
+      "step": 14088
+    },
+    {
+      "epoch": 3.2020454545454546,
+      "grad_norm": 1.205464402977852,
+      "learning_rate": 3.582420715552586e-07,
+      "loss": 0.0087,
+      "step": 14089
+    },
+    {
+      "epoch": 3.202272727272727,
+      "grad_norm": 2.0849260145170376,
+      "learning_rate": 3.581613621143712e-07,
+      "loss": 0.0109,
+      "step": 14090
+    },
+    {
+      "epoch": 3.2025,
+      "grad_norm": 1.8492805461950261,
+      "learning_rate": 3.5808065811478916e-07,
+      "loss": 0.0142,
+      "step": 14091
+    },
+    {
+      "epoch": 3.2027272727272726,
+      "grad_norm": 1.205391983550896,
+      "learning_rate": 3.579999595581582e-07,
+      "loss": 0.0069,
+      "step": 14092
+    },
+    {
+      "epoch": 3.2029545454545456,
+      "grad_norm": 0.543569233266973,
+      "learning_rate": 3.5791926644612384e-07,
+      "loss": 0.0047,
+      "step": 14093
+    },
+    {
+      "epoch": 3.203181818181818,
+      "grad_norm": 2.559757324240475,
+      "learning_rate": 3.5783857878033165e-07,
+      "loss": 0.0267,
+      "step": 14094
+    },
+    {
+      "epoch": 3.203409090909091,
+      "grad_norm": 0.6619301227853126,
+      "learning_rate": 3.5775789656242685e-07,
+      "loss": 0.0062,
+      "step": 14095
+    },
+    {
+      "epoch": 3.2036363636363636,
+      "grad_norm": 1.2883630068830296,
+      "learning_rate": 3.5767721979405504e-07,
+      "loss": 0.0151,
+      "step": 14096
+    },
+    {
+      "epoch": 3.2038636363636366,
+      "grad_norm": 1.2761584843083034,
+      "learning_rate": 3.5759654847686086e-07,
+      "loss": 0.0088,
+      "step": 14097
+    },
+    {
+      "epoch": 3.204090909090909,
+      "grad_norm": 2.6320078458309664,
+      "learning_rate": 3.5751588261248956e-07,
+      "loss": 0.0212,
+      "step": 14098
+    },
+    {
+      "epoch": 3.2043181818181816,
+      "grad_norm": 1.6271079461607192,
+      "learning_rate": 3.5743522220258607e-07,
+      "loss": 0.0087,
+      "step": 14099
+    },
+    {
+      "epoch": 3.2045454545454546,
+      "grad_norm": 1.9504313383125873,
+      "learning_rate": 3.573545672487952e-07,
+      "loss": 0.015,
+      "step": 14100
+    },
+    {
+      "epoch": 3.204772727272727,
+      "grad_norm": 1.0367766327708097,
+      "learning_rate": 3.5727391775276193e-07,
+      "loss": 0.0069,
+      "step": 14101
+    },
+    {
+      "epoch": 3.205,
+      "grad_norm": 1.6956266034533514,
+      "learning_rate": 3.5719327371613014e-07,
+      "loss": 0.0101,
+      "step": 14102
+    },
+    {
+      "epoch": 3.2052272727272726,
+      "grad_norm": 0.9513298941278978,
+      "learning_rate": 3.5711263514054487e-07,
+      "loss": 0.0054,
+      "step": 14103
+    },
+    {
+      "epoch": 3.2054545454545456,
+      "grad_norm": 1.7433430274506807,
+      "learning_rate": 3.5703200202765025e-07,
+      "loss": 0.0088,
+      "step": 14104
+    },
+    {
+      "epoch": 3.205681818181818,
+      "grad_norm": 0.4722152743027184,
+      "learning_rate": 3.569513743790909e-07,
+      "loss": 0.0026,
+      "step": 14105
+    },
+    {
+      "epoch": 3.205909090909091,
+      "grad_norm": 1.6630117759881882,
+      "learning_rate": 3.5687075219651034e-07,
+      "loss": 0.0222,
+      "step": 14106
+    },
+    {
+      "epoch": 3.2061363636363636,
+      "grad_norm": 2.863814588721767,
+      "learning_rate": 3.56790135481553e-07,
+      "loss": 0.009,
+      "step": 14107
+    },
+    {
+      "epoch": 3.2063636363636365,
+      "grad_norm": 2.2125467187072196,
+      "learning_rate": 3.567095242358627e-07,
+      "loss": 0.007,
+      "step": 14108
+    },
+    {
+      "epoch": 3.206590909090909,
+      "grad_norm": 2.926177190280243,
+      "learning_rate": 3.5662891846108326e-07,
+      "loss": 0.0239,
+      "step": 14109
+    },
+    {
+      "epoch": 3.206818181818182,
+      "grad_norm": 1.2289457110852366,
+      "learning_rate": 3.565483181588583e-07,
+      "loss": 0.0054,
+      "step": 14110
+    },
+    {
+      "epoch": 3.2070454545454545,
+      "grad_norm": 1.3753312066853887,
+      "learning_rate": 3.5646772333083155e-07,
+      "loss": 0.0068,
+      "step": 14111
+    },
+    {
+      "epoch": 3.207272727272727,
+      "grad_norm": 2.890917050167514,
+      "learning_rate": 3.563871339786464e-07,
+      "loss": 0.0339,
+      "step": 14112
+    },
+    {
+      "epoch": 3.2075,
+      "grad_norm": 1.118775073678119,
+      "learning_rate": 3.563065501039462e-07,
+      "loss": 0.0061,
+      "step": 14113
+    },
+    {
+      "epoch": 3.2077272727272725,
+      "grad_norm": 0.4682804015207973,
+      "learning_rate": 3.562259717083745e-07,
+      "loss": 0.0032,
+      "step": 14114
+    },
+    {
+      "epoch": 3.2079545454545455,
+      "grad_norm": 2.739108545323276,
+      "learning_rate": 3.5614539879357384e-07,
+      "loss": 0.034,
+      "step": 14115
+    },
+    {
+      "epoch": 3.208181818181818,
+      "grad_norm": 1.3830754676882633,
+      "learning_rate": 3.5606483136118753e-07,
+      "loss": 0.015,
+      "step": 14116
+    },
+    {
+      "epoch": 3.208409090909091,
+      "grad_norm": 1.9057053218954128,
+      "learning_rate": 3.559842694128585e-07,
+      "loss": 0.0087,
+      "step": 14117
+    },
+    {
+      "epoch": 3.2086363636363635,
+      "grad_norm": 1.3325011881345101,
+      "learning_rate": 3.5590371295022975e-07,
+      "loss": 0.0096,
+      "step": 14118
+    },
+    {
+      "epoch": 3.2088636363636365,
+      "grad_norm": 3.8970850824526533,
+      "learning_rate": 3.558231619749436e-07,
+      "loss": 0.0283,
+      "step": 14119
+    },
+    {
+      "epoch": 3.209090909090909,
+      "grad_norm": 1.0979159296268257,
+      "learning_rate": 3.5574261648864266e-07,
+      "loss": 0.0066,
+      "step": 14120
+    },
+    {
+      "epoch": 3.209318181818182,
+      "grad_norm": 1.1437182811715123,
+      "learning_rate": 3.5566207649296953e-07,
+      "loss": 0.0064,
+      "step": 14121
+    },
+    {
+      "epoch": 3.2095454545454545,
+      "grad_norm": 0.8716204133293017,
+      "learning_rate": 3.5558154198956667e-07,
+      "loss": 0.0085,
+      "step": 14122
+    },
+    {
+      "epoch": 3.2097727272727274,
+      "grad_norm": 1.3874023332312095,
+      "learning_rate": 3.555010129800761e-07,
+      "loss": 0.0071,
+      "step": 14123
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 1.3150292625926816,
+      "learning_rate": 3.5542048946614003e-07,
+      "loss": 0.0101,
+      "step": 14124
+    },
+    {
+      "epoch": 3.210227272727273,
+      "grad_norm": 1.638209892964432,
+      "learning_rate": 3.553399714494006e-07,
+      "loss": 0.0093,
+      "step": 14125
+    },
+    {
+      "epoch": 3.2104545454545454,
+      "grad_norm": 1.1155241073134052,
+      "learning_rate": 3.552594589314995e-07,
+      "loss": 0.0084,
+      "step": 14126
+    },
+    {
+      "epoch": 3.2106818181818184,
+      "grad_norm": 1.3114229414880931,
+      "learning_rate": 3.551789519140789e-07,
+      "loss": 0.0134,
+      "step": 14127
+    },
+    {
+      "epoch": 3.210909090909091,
+      "grad_norm": 2.5908066339252915,
+      "learning_rate": 3.550984503987799e-07,
+      "loss": 0.0119,
+      "step": 14128
+    },
+    {
+      "epoch": 3.2111363636363635,
+      "grad_norm": 0.995187003766412,
+      "learning_rate": 3.5501795438724446e-07,
+      "loss": 0.0043,
+      "step": 14129
+    },
+    {
+      "epoch": 3.2113636363636364,
+      "grad_norm": 0.9517741041969363,
+      "learning_rate": 3.549374638811139e-07,
+      "loss": 0.0051,
+      "step": 14130
+    },
+    {
+      "epoch": 3.211590909090909,
+      "grad_norm": 1.2327657507867635,
+      "learning_rate": 3.5485697888202985e-07,
+      "loss": 0.0103,
+      "step": 14131
+    },
+    {
+      "epoch": 3.211818181818182,
+      "grad_norm": 0.7596259108429749,
+      "learning_rate": 3.54776499391633e-07,
+      "loss": 0.0062,
+      "step": 14132
+    },
+    {
+      "epoch": 3.2120454545454544,
+      "grad_norm": 1.557976238207869,
+      "learning_rate": 3.546960254115648e-07,
+      "loss": 0.0108,
+      "step": 14133
+    },
+    {
+      "epoch": 3.2122727272727274,
+      "grad_norm": 2.6474038374205913,
+      "learning_rate": 3.546155569434662e-07,
+      "loss": 0.0138,
+      "step": 14134
+    },
+    {
+      "epoch": 3.2125,
+      "grad_norm": 1.3309080369503072,
+      "learning_rate": 3.545350939889781e-07,
+      "loss": 0.0041,
+      "step": 14135
+    },
+    {
+      "epoch": 3.212727272727273,
+      "grad_norm": 1.2152959271192694,
+      "learning_rate": 3.5445463654974157e-07,
+      "loss": 0.0054,
+      "step": 14136
+    },
+    {
+      "epoch": 3.2129545454545454,
+      "grad_norm": 1.3090996295068904,
+      "learning_rate": 3.5437418462739667e-07,
+      "loss": 0.0063,
+      "step": 14137
+    },
+    {
+      "epoch": 3.2131818181818184,
+      "grad_norm": 1.2116522829903795,
+      "learning_rate": 3.5429373822358434e-07,
+      "loss": 0.0102,
+      "step": 14138
+    },
+    {
+      "epoch": 3.213409090909091,
+      "grad_norm": 2.5897458593111633,
+      "learning_rate": 3.5421329733994496e-07,
+      "loss": 0.0189,
+      "step": 14139
+    },
+    {
+      "epoch": 3.213636363636364,
+      "grad_norm": 0.9201816363418136,
+      "learning_rate": 3.541328619781188e-07,
+      "loss": 0.0051,
+      "step": 14140
+    },
+    {
+      "epoch": 3.2138636363636364,
+      "grad_norm": 1.008366406663488,
+      "learning_rate": 3.540524321397462e-07,
+      "loss": 0.0066,
+      "step": 14141
+    },
+    {
+      "epoch": 3.214090909090909,
+      "grad_norm": 1.5523228425073798,
+      "learning_rate": 3.5397200782646716e-07,
+      "loss": 0.0214,
+      "step": 14142
+    },
+    {
+      "epoch": 3.214318181818182,
+      "grad_norm": 2.141663063912377,
+      "learning_rate": 3.5389158903992177e-07,
+      "loss": 0.0145,
+      "step": 14143
+    },
+    {
+      "epoch": 3.2145454545454544,
+      "grad_norm": 1.345942962582293,
+      "learning_rate": 3.538111757817499e-07,
+      "loss": 0.0102,
+      "step": 14144
+    },
+    {
+      "epoch": 3.2147727272727273,
+      "grad_norm": 1.0396660221599117,
+      "learning_rate": 3.537307680535913e-07,
+      "loss": 0.0041,
+      "step": 14145
+    },
+    {
+      "epoch": 3.215,
+      "grad_norm": 1.3244757815860477,
+      "learning_rate": 3.536503658570856e-07,
+      "loss": 0.0108,
+      "step": 14146
+    },
+    {
+      "epoch": 3.215227272727273,
+      "grad_norm": 0.6525347456069986,
+      "learning_rate": 3.535699691938722e-07,
+      "loss": 0.0026,
+      "step": 14147
+    },
+    {
+      "epoch": 3.2154545454545453,
+      "grad_norm": 2.214824674801069,
+      "learning_rate": 3.5348957806559066e-07,
+      "loss": 0.014,
+      "step": 14148
+    },
+    {
+      "epoch": 3.2156818181818183,
+      "grad_norm": 0.5681811460149518,
+      "learning_rate": 3.5340919247388054e-07,
+      "loss": 0.003,
+      "step": 14149
+    },
+    {
+      "epoch": 3.215909090909091,
+      "grad_norm": 1.065677020196814,
+      "learning_rate": 3.5332881242038063e-07,
+      "loss": 0.0068,
+      "step": 14150
+    },
+    {
+      "epoch": 3.216136363636364,
+      "grad_norm": 0.8321973342352604,
+      "learning_rate": 3.532484379067301e-07,
+      "loss": 0.0064,
+      "step": 14151
+    },
+    {
+      "epoch": 3.2163636363636363,
+      "grad_norm": 2.6573313539435146,
+      "learning_rate": 3.531680689345681e-07,
+      "loss": 0.0211,
+      "step": 14152
+    },
+    {
+      "epoch": 3.2165909090909093,
+      "grad_norm": 1.8647737454204092,
+      "learning_rate": 3.5308770550553344e-07,
+      "loss": 0.0098,
+      "step": 14153
+    },
+    {
+      "epoch": 3.216818181818182,
+      "grad_norm": 3.178544146496893,
+      "learning_rate": 3.5300734762126475e-07,
+      "loss": 0.0306,
+      "step": 14154
+    },
+    {
+      "epoch": 3.2170454545454543,
+      "grad_norm": 1.4374096376526946,
+      "learning_rate": 3.529269952834008e-07,
+      "loss": 0.0096,
+      "step": 14155
+    },
+    {
+      "epoch": 3.2172727272727273,
+      "grad_norm": 1.375832248830977,
+      "learning_rate": 3.5284664849358e-07,
+      "loss": 0.0064,
+      "step": 14156
+    },
+    {
+      "epoch": 3.2175,
+      "grad_norm": 1.438630406343765,
+      "learning_rate": 3.5276630725344094e-07,
+      "loss": 0.0051,
+      "step": 14157
+    },
+    {
+      "epoch": 3.2177272727272728,
+      "grad_norm": 0.6154655837021,
+      "learning_rate": 3.5268597156462206e-07,
+      "loss": 0.0086,
+      "step": 14158
+    },
+    {
+      "epoch": 3.2179545454545453,
+      "grad_norm": 2.0142505565411857,
+      "learning_rate": 3.52605641428761e-07,
+      "loss": 0.0298,
+      "step": 14159
+    },
+    {
+      "epoch": 3.2181818181818183,
+      "grad_norm": 0.721462080271675,
+      "learning_rate": 3.525253168474961e-07,
+      "loss": 0.0082,
+      "step": 14160
+    },
+    {
+      "epoch": 3.2184090909090908,
+      "grad_norm": 1.3559261680554047,
+      "learning_rate": 3.524449978224654e-07,
+      "loss": 0.0153,
+      "step": 14161
+    },
+    {
+      "epoch": 3.2186363636363637,
+      "grad_norm": 1.6621618634841469,
+      "learning_rate": 3.5236468435530693e-07,
+      "loss": 0.0105,
+      "step": 14162
+    },
+    {
+      "epoch": 3.2188636363636363,
+      "grad_norm": 2.300447208037675,
+      "learning_rate": 3.5228437644765794e-07,
+      "loss": 0.0145,
+      "step": 14163
+    },
+    {
+      "epoch": 3.2190909090909092,
+      "grad_norm": 1.717687538720456,
+      "learning_rate": 3.5220407410115627e-07,
+      "loss": 0.0109,
+      "step": 14164
+    },
+    {
+      "epoch": 3.2193181818181817,
+      "grad_norm": 1.113893919433105,
+      "learning_rate": 3.521237773174395e-07,
+      "loss": 0.0117,
+      "step": 14165
+    },
+    {
+      "epoch": 3.2195454545454547,
+      "grad_norm": 2.0260011747914515,
+      "learning_rate": 3.520434860981449e-07,
+      "loss": 0.0093,
+      "step": 14166
+    },
+    {
+      "epoch": 3.2197727272727272,
+      "grad_norm": 2.4071554992289137,
+      "learning_rate": 3.5196320044491014e-07,
+      "loss": 0.0216,
+      "step": 14167
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 1.7819208894200804,
+      "learning_rate": 3.518829203593717e-07,
+      "loss": 0.0091,
+      "step": 14168
+    },
+    {
+      "epoch": 3.2202272727272727,
+      "grad_norm": 1.2098043632590567,
+      "learning_rate": 3.518026458431671e-07,
+      "loss": 0.0085,
+      "step": 14169
+    },
+    {
+      "epoch": 3.2204545454545457,
+      "grad_norm": 3.149068066776922,
+      "learning_rate": 3.517223768979332e-07,
+      "loss": 0.0155,
+      "step": 14170
+    },
+    {
+      "epoch": 3.220681818181818,
+      "grad_norm": 1.056538578462248,
+      "learning_rate": 3.516421135253066e-07,
+      "loss": 0.0036,
+      "step": 14171
+    },
+    {
+      "epoch": 3.2209090909090907,
+      "grad_norm": 4.18760505286769,
+      "learning_rate": 3.5156185572692443e-07,
+      "loss": 0.0144,
+      "step": 14172
+    },
+    {
+      "epoch": 3.2211363636363637,
+      "grad_norm": 1.9074134959472913,
+      "learning_rate": 3.5148160350442296e-07,
+      "loss": 0.0127,
+      "step": 14173
+    },
+    {
+      "epoch": 3.221363636363636,
+      "grad_norm": 0.6128748149408658,
+      "learning_rate": 3.514013568594388e-07,
+      "loss": 0.0026,
+      "step": 14174
+    },
+    {
+      "epoch": 3.221590909090909,
+      "grad_norm": 0.9287073597838392,
+      "learning_rate": 3.513211157936085e-07,
+      "loss": 0.0027,
+      "step": 14175
+    },
+    {
+      "epoch": 3.2218181818181817,
+      "grad_norm": 1.933014535415246,
+      "learning_rate": 3.5124088030856783e-07,
+      "loss": 0.0097,
+      "step": 14176
+    },
+    {
+      "epoch": 3.2220454545454547,
+      "grad_norm": 1.02705717862649,
+      "learning_rate": 3.5116065040595325e-07,
+      "loss": 0.0047,
+      "step": 14177
+    },
+    {
+      "epoch": 3.222272727272727,
+      "grad_norm": 2.255641224146622,
+      "learning_rate": 3.5108042608740083e-07,
+      "loss": 0.0091,
+      "step": 14178
+    },
+    {
+      "epoch": 3.2225,
+      "grad_norm": 3.0388753754970446,
+      "learning_rate": 3.5100020735454625e-07,
+      "loss": 0.0337,
+      "step": 14179
+    },
+    {
+      "epoch": 3.2227272727272727,
+      "grad_norm": 1.618466960995292,
+      "learning_rate": 3.5091999420902576e-07,
+      "loss": 0.0132,
+      "step": 14180
+    },
+    {
+      "epoch": 3.2229545454545456,
+      "grad_norm": 1.2405402659113403,
+      "learning_rate": 3.508397866524746e-07,
+      "loss": 0.0145,
+      "step": 14181
+    },
+    {
+      "epoch": 3.223181818181818,
+      "grad_norm": 1.480082375940356,
+      "learning_rate": 3.507595846865284e-07,
+      "loss": 0.0066,
+      "step": 14182
+    },
+    {
+      "epoch": 3.223409090909091,
+      "grad_norm": 0.8095394485122135,
+      "learning_rate": 3.5067938831282273e-07,
+      "loss": 0.0046,
+      "step": 14183
+    },
+    {
+      "epoch": 3.2236363636363636,
+      "grad_norm": 1.3023720046310985,
+      "learning_rate": 3.505991975329928e-07,
+      "loss": 0.0072,
+      "step": 14184
+    },
+    {
+      "epoch": 3.223863636363636,
+      "grad_norm": 4.2488820777269405,
+      "learning_rate": 3.505190123486742e-07,
+      "loss": 0.0464,
+      "step": 14185
+    },
+    {
+      "epoch": 3.224090909090909,
+      "grad_norm": 0.6734585395827893,
+      "learning_rate": 3.504388327615016e-07,
+      "loss": 0.0021,
+      "step": 14186
+    },
+    {
+      "epoch": 3.2243181818181816,
+      "grad_norm": 0.781544995358499,
+      "learning_rate": 3.503586587731102e-07,
+      "loss": 0.0096,
+      "step": 14187
+    },
+    {
+      "epoch": 3.2245454545454546,
+      "grad_norm": 0.37014766657284043,
+      "learning_rate": 3.5027849038513493e-07,
+      "loss": 0.0017,
+      "step": 14188
+    },
+    {
+      "epoch": 3.224772727272727,
+      "grad_norm": 1.4095927012234482,
+      "learning_rate": 3.5019832759921077e-07,
+      "loss": 0.0087,
+      "step": 14189
+    },
+    {
+      "epoch": 3.225,
+      "grad_norm": 2.0400439036705937,
+      "learning_rate": 3.501181704169719e-07,
+      "loss": 0.0089,
+      "step": 14190
+    },
+    {
+      "epoch": 3.2252272727272726,
+      "grad_norm": 2.5307505546614055,
+      "learning_rate": 3.50038018840053e-07,
+      "loss": 0.0237,
+      "step": 14191
+    },
+    {
+      "epoch": 3.2254545454545456,
+      "grad_norm": 3.034221649740033,
+      "learning_rate": 3.499578728700887e-07,
+      "loss": 0.0601,
+      "step": 14192
+    },
+    {
+      "epoch": 3.225681818181818,
+      "grad_norm": 0.5011908266472056,
+      "learning_rate": 3.498777325087134e-07,
+      "loss": 0.0015,
+      "step": 14193
+    },
+    {
+      "epoch": 3.225909090909091,
+      "grad_norm": 0.9705259957631822,
+      "learning_rate": 3.4979759775756095e-07,
+      "loss": 0.0028,
+      "step": 14194
+    },
+    {
+      "epoch": 3.2261363636363636,
+      "grad_norm": 2.2674968361470236,
+      "learning_rate": 3.4971746861826556e-07,
+      "loss": 0.0065,
+      "step": 14195
+    },
+    {
+      "epoch": 3.2263636363636365,
+      "grad_norm": 1.2934631522783397,
+      "learning_rate": 3.4963734509246117e-07,
+      "loss": 0.0058,
+      "step": 14196
+    },
+    {
+      "epoch": 3.226590909090909,
+      "grad_norm": 1.3731857038944626,
+      "learning_rate": 3.49557227181782e-07,
+      "loss": 0.0101,
+      "step": 14197
+    },
+    {
+      "epoch": 3.2268181818181816,
+      "grad_norm": 1.6938291449104081,
+      "learning_rate": 3.4947711488786125e-07,
+      "loss": 0.0082,
+      "step": 14198
+    },
+    {
+      "epoch": 3.2270454545454546,
+      "grad_norm": 2.8453604673623025,
+      "learning_rate": 3.4939700821233283e-07,
+      "loss": 0.0284,
+      "step": 14199
+    },
+    {
+      "epoch": 3.227272727272727,
+      "grad_norm": 1.6059177858815883,
+      "learning_rate": 3.4931690715682996e-07,
+      "loss": 0.0033,
+      "step": 14200
+    },
+    {
+      "epoch": 3.2275,
+      "grad_norm": 0.7086681060223918,
+      "learning_rate": 3.4923681172298656e-07,
+      "loss": 0.0039,
+      "step": 14201
+    },
+    {
+      "epoch": 3.2277272727272726,
+      "grad_norm": 2.325293959807072,
+      "learning_rate": 3.4915672191243586e-07,
+      "loss": 0.0142,
+      "step": 14202
+    },
+    {
+      "epoch": 3.2279545454545455,
+      "grad_norm": 0.6267659206808999,
+      "learning_rate": 3.490766377268107e-07,
+      "loss": 0.0033,
+      "step": 14203
+    },
+    {
+      "epoch": 3.228181818181818,
+      "grad_norm": 1.0295747480368955,
+      "learning_rate": 3.4899655916774426e-07,
+      "loss": 0.0063,
+      "step": 14204
+    },
+    {
+      "epoch": 3.228409090909091,
+      "grad_norm": 1.2647447244702457,
+      "learning_rate": 3.4891648623686936e-07,
+      "loss": 0.0077,
+      "step": 14205
+    },
+    {
+      "epoch": 3.2286363636363635,
+      "grad_norm": 1.3801132425727385,
+      "learning_rate": 3.4883641893581933e-07,
+      "loss": 0.0114,
+      "step": 14206
+    },
+    {
+      "epoch": 3.2288636363636365,
+      "grad_norm": 0.8616895416996976,
+      "learning_rate": 3.487563572662262e-07,
+      "loss": 0.0043,
+      "step": 14207
+    },
+    {
+      "epoch": 3.229090909090909,
+      "grad_norm": 0.5281154788681818,
+      "learning_rate": 3.486763012297229e-07,
+      "loss": 0.0039,
+      "step": 14208
+    },
+    {
+      "epoch": 3.229318181818182,
+      "grad_norm": 1.761143495336506,
+      "learning_rate": 3.48596250827942e-07,
+      "loss": 0.0064,
+      "step": 14209
+    },
+    {
+      "epoch": 3.2295454545454545,
+      "grad_norm": 2.165142216187855,
+      "learning_rate": 3.485162060625156e-07,
+      "loss": 0.0106,
+      "step": 14210
+    },
+    {
+      "epoch": 3.2297727272727275,
+      "grad_norm": 0.7762956124309197,
+      "learning_rate": 3.484361669350764e-07,
+      "loss": 0.0051,
+      "step": 14211
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 5.841342862326991,
+      "learning_rate": 3.48356133447256e-07,
+      "loss": 0.025,
+      "step": 14212
+    },
+    {
+      "epoch": 3.230227272727273,
+      "grad_norm": 1.2900448576473578,
+      "learning_rate": 3.482761056006867e-07,
+      "loss": 0.0098,
+      "step": 14213
+    },
+    {
+      "epoch": 3.2304545454545455,
+      "grad_norm": 1.6097755981298467,
+      "learning_rate": 3.4819608339700036e-07,
+      "loss": 0.0074,
+      "step": 14214
+    },
+    {
+      "epoch": 3.230681818181818,
+      "grad_norm": 1.6966565049994082,
+      "learning_rate": 3.4811606683782876e-07,
+      "loss": 0.0117,
+      "step": 14215
+    },
+    {
+      "epoch": 3.230909090909091,
+      "grad_norm": 1.5087599415695527,
+      "learning_rate": 3.480360559248036e-07,
+      "loss": 0.0143,
+      "step": 14216
+    },
+    {
+      "epoch": 3.2311363636363635,
+      "grad_norm": 1.752589819624512,
+      "learning_rate": 3.479560506595565e-07,
+      "loss": 0.0105,
+      "step": 14217
+    },
+    {
+      "epoch": 3.2313636363636364,
+      "grad_norm": 0.7672209015662881,
+      "learning_rate": 3.478760510437189e-07,
+      "loss": 0.0031,
+      "step": 14218
+    },
+    {
+      "epoch": 3.231590909090909,
+      "grad_norm": 1.33435192321868,
+      "learning_rate": 3.4779605707892226e-07,
+      "loss": 0.0071,
+      "step": 14219
+    },
+    {
+      "epoch": 3.231818181818182,
+      "grad_norm": 3.4197825394575436,
+      "learning_rate": 3.477160687667974e-07,
+      "loss": 0.02,
+      "step": 14220
+    },
+    {
+      "epoch": 3.2320454545454544,
+      "grad_norm": 1.6474922371854648,
+      "learning_rate": 3.4763608610897566e-07,
+      "loss": 0.0149,
+      "step": 14221
+    },
+    {
+      "epoch": 3.2322727272727274,
+      "grad_norm": 0.32450075618758367,
+      "learning_rate": 3.4755610910708793e-07,
+      "loss": 0.0015,
+      "step": 14222
+    },
+    {
+      "epoch": 3.2325,
+      "grad_norm": 0.8978125113340802,
+      "learning_rate": 3.474761377627653e-07,
+      "loss": 0.0029,
+      "step": 14223
+    },
+    {
+      "epoch": 3.232727272727273,
+      "grad_norm": 2.1404936474111875,
+      "learning_rate": 3.473961720776385e-07,
+      "loss": 0.018,
+      "step": 14224
+    },
+    {
+      "epoch": 3.2329545454545454,
+      "grad_norm": 1.247880381090479,
+      "learning_rate": 3.4731621205333797e-07,
+      "loss": 0.0099,
+      "step": 14225
+    },
+    {
+      "epoch": 3.2331818181818184,
+      "grad_norm": 0.6083454929776971,
+      "learning_rate": 3.472362576914942e-07,
+      "loss": 0.0088,
+      "step": 14226
+    },
+    {
+      "epoch": 3.233409090909091,
+      "grad_norm": 0.7348625822063598,
+      "learning_rate": 3.471563089937377e-07,
+      "loss": 0.0026,
+      "step": 14227
+    },
+    {
+      "epoch": 3.2336363636363634,
+      "grad_norm": 0.9229425456410133,
+      "learning_rate": 3.470763659616991e-07,
+      "loss": 0.0055,
+      "step": 14228
+    },
+    {
+      "epoch": 3.2338636363636364,
+      "grad_norm": 1.9238005618794412,
+      "learning_rate": 3.4699642859700776e-07,
+      "loss": 0.02,
+      "step": 14229
+    },
+    {
+      "epoch": 3.234090909090909,
+      "grad_norm": 1.998565622210094,
+      "learning_rate": 3.469164969012945e-07,
+      "loss": 0.0127,
+      "step": 14230
+    },
+    {
+      "epoch": 3.234318181818182,
+      "grad_norm": 1.2509009557315551,
+      "learning_rate": 3.46836570876189e-07,
+      "loss": 0.01,
+      "step": 14231
+    },
+    {
+      "epoch": 3.2345454545454544,
+      "grad_norm": 1.8420852712704363,
+      "learning_rate": 3.467566505233211e-07,
+      "loss": 0.0131,
+      "step": 14232
+    },
+    {
+      "epoch": 3.2347727272727274,
+      "grad_norm": 1.1079948343267851,
+      "learning_rate": 3.466767358443207e-07,
+      "loss": 0.0106,
+      "step": 14233
+    },
+    {
+      "epoch": 3.235,
+      "grad_norm": 0.459140214920412,
+      "learning_rate": 3.4659682684081707e-07,
+      "loss": 0.0025,
+      "step": 14234
+    },
+    {
+      "epoch": 3.235227272727273,
+      "grad_norm": 1.8767589340439563,
+      "learning_rate": 3.4651692351443987e-07,
+      "loss": 0.0122,
+      "step": 14235
+    },
+    {
+      "epoch": 3.2354545454545454,
+      "grad_norm": 0.8186312075461826,
+      "learning_rate": 3.464370258668183e-07,
+      "loss": 0.0049,
+      "step": 14236
+    },
+    {
+      "epoch": 3.2356818181818183,
+      "grad_norm": 1.118747844593326,
+      "learning_rate": 3.4635713389958204e-07,
+      "loss": 0.0057,
+      "step": 14237
+    },
+    {
+      "epoch": 3.235909090909091,
+      "grad_norm": 2.3029672713725904,
+      "learning_rate": 3.462772476143598e-07,
+      "loss": 0.0212,
+      "step": 14238
+    },
+    {
+      "epoch": 3.236136363636364,
+      "grad_norm": 3.4938177354486446,
+      "learning_rate": 3.4619736701278066e-07,
+      "loss": 0.0256,
+      "step": 14239
+    },
+    {
+      "epoch": 3.2363636363636363,
+      "grad_norm": 2.4177799794140227,
+      "learning_rate": 3.4611749209647355e-07,
+      "loss": 0.0159,
+      "step": 14240
+    },
+    {
+      "epoch": 3.2365909090909093,
+      "grad_norm": 0.5967041650161303,
+      "learning_rate": 3.460376228670677e-07,
+      "loss": 0.0038,
+      "step": 14241
+    },
+    {
+      "epoch": 3.236818181818182,
+      "grad_norm": 2.185110477566302,
+      "learning_rate": 3.4595775932619104e-07,
+      "loss": 0.015,
+      "step": 14242
+    },
+    {
+      "epoch": 3.237045454545455,
+      "grad_norm": 1.6399624350716249,
+      "learning_rate": 3.458779014754725e-07,
+      "loss": 0.0094,
+      "step": 14243
+    },
+    {
+      "epoch": 3.2372727272727273,
+      "grad_norm": 6.051996695646826,
+      "learning_rate": 3.4579804931654056e-07,
+      "loss": 0.0227,
+      "step": 14244
+    },
+    {
+      "epoch": 3.2375,
+      "grad_norm": 2.9884821850419727,
+      "learning_rate": 3.4571820285102336e-07,
+      "loss": 0.0146,
+      "step": 14245
+    },
+    {
+      "epoch": 3.237727272727273,
+      "grad_norm": 1.7371435602472114,
+      "learning_rate": 3.456383620805494e-07,
+      "loss": 0.0142,
+      "step": 14246
+    },
+    {
+      "epoch": 3.2379545454545453,
+      "grad_norm": 0.7160258483494174,
+      "learning_rate": 3.4555852700674644e-07,
+      "loss": 0.0051,
+      "step": 14247
+    },
+    {
+      "epoch": 3.2381818181818183,
+      "grad_norm": 1.5649492510359504,
+      "learning_rate": 3.4547869763124276e-07,
+      "loss": 0.0121,
+      "step": 14248
+    },
+    {
+      "epoch": 3.238409090909091,
+      "grad_norm": 1.6012563732319587,
+      "learning_rate": 3.4539887395566603e-07,
+      "loss": 0.01,
+      "step": 14249
+    },
+    {
+      "epoch": 3.2386363636363638,
+      "grad_norm": 1.1612065706779113,
+      "learning_rate": 3.453190559816443e-07,
+      "loss": 0.0096,
+      "step": 14250
+    },
+    {
+      "epoch": 3.2388636363636363,
+      "grad_norm": 0.9451560944035339,
+      "learning_rate": 3.4523924371080474e-07,
+      "loss": 0.003,
+      "step": 14251
+    },
+    {
+      "epoch": 3.2390909090909092,
+      "grad_norm": 1.3681540201643274,
+      "learning_rate": 3.45159437144775e-07,
+      "loss": 0.011,
+      "step": 14252
+    },
+    {
+      "epoch": 3.2393181818181818,
+      "grad_norm": 2.4155778152326266,
+      "learning_rate": 3.450796362851826e-07,
+      "loss": 0.0302,
+      "step": 14253
+    },
+    {
+      "epoch": 3.2395454545454547,
+      "grad_norm": 1.5907464098360276,
+      "learning_rate": 3.4499984113365506e-07,
+      "loss": 0.0127,
+      "step": 14254
+    },
+    {
+      "epoch": 3.2397727272727272,
+      "grad_norm": 0.7755195080006964,
+      "learning_rate": 3.4492005169181894e-07,
+      "loss": 0.0062,
+      "step": 14255
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 2.767320608675758,
+      "learning_rate": 3.448402679613016e-07,
+      "loss": 0.0093,
+      "step": 14256
+    },
+    {
+      "epoch": 3.2402272727272727,
+      "grad_norm": 1.1572129289040565,
+      "learning_rate": 3.4476048994372996e-07,
+      "loss": 0.0069,
+      "step": 14257
+    },
+    {
+      "epoch": 3.2404545454545453,
+      "grad_norm": 1.3435816163145828,
+      "learning_rate": 3.446807176407308e-07,
+      "loss": 0.0274,
+      "step": 14258
+    },
+    {
+      "epoch": 3.240681818181818,
+      "grad_norm": 1.8774252128626172,
+      "learning_rate": 3.446009510539309e-07,
+      "loss": 0.0243,
+      "step": 14259
+    },
+    {
+      "epoch": 3.2409090909090907,
+      "grad_norm": 1.039259190756074,
+      "learning_rate": 3.445211901849568e-07,
+      "loss": 0.0056,
+      "step": 14260
+    },
+    {
+      "epoch": 3.2411363636363637,
+      "grad_norm": 1.2211627452814042,
+      "learning_rate": 3.444414350354348e-07,
+      "loss": 0.01,
+      "step": 14261
+    },
+    {
+      "epoch": 3.2413636363636362,
+      "grad_norm": 1.1036147802279377,
+      "learning_rate": 3.443616856069916e-07,
+      "loss": 0.0094,
+      "step": 14262
+    },
+    {
+      "epoch": 3.241590909090909,
+      "grad_norm": 2.073969497022269,
+      "learning_rate": 3.4428194190125344e-07,
+      "loss": 0.0124,
+      "step": 14263
+    },
+    {
+      "epoch": 3.2418181818181817,
+      "grad_norm": 1.6622932134724715,
+      "learning_rate": 3.442022039198459e-07,
+      "loss": 0.0088,
+      "step": 14264
+    },
+    {
+      "epoch": 3.2420454545454547,
+      "grad_norm": 1.0007979747239077,
+      "learning_rate": 3.441224716643954e-07,
+      "loss": 0.0064,
+      "step": 14265
+    },
+    {
+      "epoch": 3.242272727272727,
+      "grad_norm": 1.080551345065435,
+      "learning_rate": 3.440427451365276e-07,
+      "loss": 0.0107,
+      "step": 14266
+    },
+    {
+      "epoch": 3.2425,
+      "grad_norm": 3.3806058421586855,
+      "learning_rate": 3.4396302433786845e-07,
+      "loss": 0.0235,
+      "step": 14267
+    },
+    {
+      "epoch": 3.2427272727272727,
+      "grad_norm": 2.0003126418963983,
+      "learning_rate": 3.438833092700438e-07,
+      "loss": 0.0174,
+      "step": 14268
+    },
+    {
+      "epoch": 3.2429545454545456,
+      "grad_norm": 2.260432108561024,
+      "learning_rate": 3.4380359993467865e-07,
+      "loss": 0.0277,
+      "step": 14269
+    },
+    {
+      "epoch": 3.243181818181818,
+      "grad_norm": 0.38998821795352306,
+      "learning_rate": 3.4372389633339863e-07,
+      "loss": 0.0021,
+      "step": 14270
+    },
+    {
+      "epoch": 3.2434090909090907,
+      "grad_norm": 1.956826771907653,
+      "learning_rate": 3.436441984678291e-07,
+      "loss": 0.0358,
+      "step": 14271
+    },
+    {
+      "epoch": 3.2436363636363637,
+      "grad_norm": 1.4410971715542618,
+      "learning_rate": 3.435645063395955e-07,
+      "loss": 0.0156,
+      "step": 14272
+    },
+    {
+      "epoch": 3.243863636363636,
+      "grad_norm": 1.6932585663789566,
+      "learning_rate": 3.434848199503223e-07,
+      "loss": 0.0098,
+      "step": 14273
+    },
+    {
+      "epoch": 3.244090909090909,
+      "grad_norm": 1.5298932696282082,
+      "learning_rate": 3.434051393016348e-07,
+      "loss": 0.0102,
+      "step": 14274
+    },
+    {
+      "epoch": 3.2443181818181817,
+      "grad_norm": 11.735314816356114,
+      "learning_rate": 3.433254643951578e-07,
+      "loss": 0.0208,
+      "step": 14275
+    },
+    {
+      "epoch": 3.2445454545454546,
+      "grad_norm": 0.8774015555209954,
+      "learning_rate": 3.432457952325159e-07,
+      "loss": 0.0042,
+      "step": 14276
+    },
+    {
+      "epoch": 3.244772727272727,
+      "grad_norm": 2.0670429936095136,
+      "learning_rate": 3.4316613181533387e-07,
+      "loss": 0.0246,
+      "step": 14277
+    },
+    {
+      "epoch": 3.245,
+      "grad_norm": 1.814919746839041,
+      "learning_rate": 3.43086474145236e-07,
+      "loss": 0.0145,
+      "step": 14278
+    },
+    {
+      "epoch": 3.2452272727272726,
+      "grad_norm": 1.2636858456509223,
+      "learning_rate": 3.4300682222384684e-07,
+      "loss": 0.0054,
+      "step": 14279
+    },
+    {
+      "epoch": 3.2454545454545456,
+      "grad_norm": 1.6062195788363227,
+      "learning_rate": 3.429271760527905e-07,
+      "loss": 0.0049,
+      "step": 14280
+    },
+    {
+      "epoch": 3.245681818181818,
+      "grad_norm": 3.058360813918412,
+      "learning_rate": 3.428475356336913e-07,
+      "loss": 0.0171,
+      "step": 14281
+    },
+    {
+      "epoch": 3.245909090909091,
+      "grad_norm": 1.657747415625622,
+      "learning_rate": 3.42767900968173e-07,
+      "loss": 0.0155,
+      "step": 14282
+    },
+    {
+      "epoch": 3.2461363636363636,
+      "grad_norm": 2.7309540754187522,
+      "learning_rate": 3.426882720578595e-07,
+      "loss": 0.0278,
+      "step": 14283
+    },
+    {
+      "epoch": 3.2463636363636366,
+      "grad_norm": 1.1766643377152448,
+      "learning_rate": 3.4260864890437463e-07,
+      "loss": 0.0083,
+      "step": 14284
+    },
+    {
+      "epoch": 3.246590909090909,
+      "grad_norm": 1.979911859972781,
+      "learning_rate": 3.425290315093424e-07,
+      "loss": 0.014,
+      "step": 14285
+    },
+    {
+      "epoch": 3.246818181818182,
+      "grad_norm": 0.6373889027275282,
+      "learning_rate": 3.424494198743857e-07,
+      "loss": 0.0031,
+      "step": 14286
+    },
+    {
+      "epoch": 3.2470454545454546,
+      "grad_norm": 1.217187264971257,
+      "learning_rate": 3.4236981400112833e-07,
+      "loss": 0.0135,
+      "step": 14287
+    },
+    {
+      "epoch": 3.247272727272727,
+      "grad_norm": 1.5362191251183217,
+      "learning_rate": 3.422902138911935e-07,
+      "loss": 0.0122,
+      "step": 14288
+    },
+    {
+      "epoch": 3.2475,
+      "grad_norm": 0.9610318115618955,
+      "learning_rate": 3.422106195462044e-07,
+      "loss": 0.0101,
+      "step": 14289
+    },
+    {
+      "epoch": 3.2477272727272726,
+      "grad_norm": 1.2318101533050823,
+      "learning_rate": 3.4213103096778415e-07,
+      "loss": 0.0079,
+      "step": 14290
+    },
+    {
+      "epoch": 3.2479545454545455,
+      "grad_norm": 1.5780063574751921,
+      "learning_rate": 3.420514481575556e-07,
+      "loss": 0.0175,
+      "step": 14291
+    },
+    {
+      "epoch": 3.248181818181818,
+      "grad_norm": 2.0270169784799417,
+      "learning_rate": 3.4197187111714174e-07,
+      "loss": 0.0174,
+      "step": 14292
+    },
+    {
+      "epoch": 3.248409090909091,
+      "grad_norm": 1.2183669413350946,
+      "learning_rate": 3.418922998481651e-07,
+      "loss": 0.01,
+      "step": 14293
+    },
+    {
+      "epoch": 3.2486363636363635,
+      "grad_norm": 1.1805404834298563,
+      "learning_rate": 3.418127343522487e-07,
+      "loss": 0.0064,
+      "step": 14294
+    },
+    {
+      "epoch": 3.2488636363636365,
+      "grad_norm": 1.6820935400844979,
+      "learning_rate": 3.4173317463101446e-07,
+      "loss": 0.0084,
+      "step": 14295
+    },
+    {
+      "epoch": 3.249090909090909,
+      "grad_norm": 2.7020903648788837,
+      "learning_rate": 3.4165362068608493e-07,
+      "loss": 0.0143,
+      "step": 14296
+    },
+    {
+      "epoch": 3.249318181818182,
+      "grad_norm": 1.3505871185851477,
+      "learning_rate": 3.415740725190824e-07,
+      "loss": 0.0078,
+      "step": 14297
+    },
+    {
+      "epoch": 3.2495454545454545,
+      "grad_norm": 0.7542405302174076,
+      "learning_rate": 3.414945301316293e-07,
+      "loss": 0.0048,
+      "step": 14298
+    },
+    {
+      "epoch": 3.2497727272727275,
+      "grad_norm": 1.153181649880806,
+      "learning_rate": 3.4141499352534703e-07,
+      "loss": 0.0093,
+      "step": 14299
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.5426215029490735,
+      "learning_rate": 3.413354627018578e-07,
+      "loss": 0.0027,
+      "step": 14300
+    },
+    {
+      "epoch": 3.2502272727272725,
+      "grad_norm": 0.7559857712187403,
+      "learning_rate": 3.4125593766278333e-07,
+      "loss": 0.0035,
+      "step": 14301
+    },
+    {
+      "epoch": 3.2504545454545455,
+      "grad_norm": 1.6282010875393003,
+      "learning_rate": 3.411764184097453e-07,
+      "loss": 0.0131,
+      "step": 14302
+    },
+    {
+      "epoch": 3.250681818181818,
+      "grad_norm": 2.0194871430677193,
+      "learning_rate": 3.410969049443656e-07,
+      "loss": 0.0148,
+      "step": 14303
+    },
+    {
+      "epoch": 3.250909090909091,
+      "grad_norm": 0.9808160089863555,
+      "learning_rate": 3.4101739726826497e-07,
+      "loss": 0.0099,
+      "step": 14304
+    },
+    {
+      "epoch": 3.2511363636363635,
+      "grad_norm": 1.355107571653822,
+      "learning_rate": 3.409378953830651e-07,
+      "loss": 0.0074,
+      "step": 14305
+    },
+    {
+      "epoch": 3.2513636363636365,
+      "grad_norm": 0.9622885157556503,
+      "learning_rate": 3.4085839929038705e-07,
+      "loss": 0.0088,
+      "step": 14306
+    },
+    {
+      "epoch": 3.251590909090909,
+      "grad_norm": 2.0394836901535136,
+      "learning_rate": 3.40778908991852e-07,
+      "loss": 0.0255,
+      "step": 14307
+    },
+    {
+      "epoch": 3.251818181818182,
+      "grad_norm": 0.6444246408495666,
+      "learning_rate": 3.4069942448908085e-07,
+      "loss": 0.003,
+      "step": 14308
+    },
+    {
+      "epoch": 3.2520454545454545,
+      "grad_norm": 2.465933016436926,
+      "learning_rate": 3.4061994578369437e-07,
+      "loss": 0.0187,
+      "step": 14309
+    },
+    {
+      "epoch": 3.2522727272727274,
+      "grad_norm": 2.9327368774165086,
+      "learning_rate": 3.4054047287731344e-07,
+      "loss": 0.0217,
+      "step": 14310
+    },
+    {
+      "epoch": 3.2525,
+      "grad_norm": 2.1453483465273195,
+      "learning_rate": 3.404610057715584e-07,
+      "loss": 0.0335,
+      "step": 14311
+    },
+    {
+      "epoch": 3.252727272727273,
+      "grad_norm": 2.436218618280224,
+      "learning_rate": 3.403815444680502e-07,
+      "loss": 0.0151,
+      "step": 14312
+    },
+    {
+      "epoch": 3.2529545454545454,
+      "grad_norm": 1.5850591703051156,
+      "learning_rate": 3.403020889684086e-07,
+      "loss": 0.0068,
+      "step": 14313
+    },
+    {
+      "epoch": 3.253181818181818,
+      "grad_norm": 1.4150372674604128,
+      "learning_rate": 3.40222639274254e-07,
+      "loss": 0.0234,
+      "step": 14314
+    },
+    {
+      "epoch": 3.253409090909091,
+      "grad_norm": 1.2317339163419492,
+      "learning_rate": 3.401431953872067e-07,
+      "loss": 0.0077,
+      "step": 14315
+    },
+    {
+      "epoch": 3.253636363636364,
+      "grad_norm": 0.8741476389275384,
+      "learning_rate": 3.400637573088868e-07,
+      "loss": 0.0051,
+      "step": 14316
+    },
+    {
+      "epoch": 3.2538636363636364,
+      "grad_norm": 0.3186849208826739,
+      "learning_rate": 3.399843250409138e-07,
+      "loss": 0.0017,
+      "step": 14317
+    },
+    {
+      "epoch": 3.254090909090909,
+      "grad_norm": 0.8178831083776859,
+      "learning_rate": 3.399048985849076e-07,
+      "loss": 0.0053,
+      "step": 14318
+    },
+    {
+      "epoch": 3.254318181818182,
+      "grad_norm": 0.8148723559801345,
+      "learning_rate": 3.398254779424879e-07,
+      "loss": 0.0046,
+      "step": 14319
+    },
+    {
+      "epoch": 3.2545454545454544,
+      "grad_norm": 3.3128776384624725,
+      "learning_rate": 3.397460631152742e-07,
+      "loss": 0.025,
+      "step": 14320
+    },
+    {
+      "epoch": 3.2547727272727274,
+      "grad_norm": 3.2285238671247893,
+      "learning_rate": 3.39666654104886e-07,
+      "loss": 0.0187,
+      "step": 14321
+    },
+    {
+      "epoch": 3.255,
+      "grad_norm": 2.585568886140622,
+      "learning_rate": 3.395872509129424e-07,
+      "loss": 0.0162,
+      "step": 14322
+    },
+    {
+      "epoch": 3.255227272727273,
+      "grad_norm": 1.952467925266296,
+      "learning_rate": 3.395078535410627e-07,
+      "loss": 0.0118,
+      "step": 14323
+    },
+    {
+      "epoch": 3.2554545454545454,
+      "grad_norm": 4.163837405033062,
+      "learning_rate": 3.39428461990866e-07,
+      "loss": 0.0278,
+      "step": 14324
+    },
+    {
+      "epoch": 3.2556818181818183,
+      "grad_norm": 1.3751626585173367,
+      "learning_rate": 3.393490762639715e-07,
+      "loss": 0.0266,
+      "step": 14325
+    },
+    {
+      "epoch": 3.255909090909091,
+      "grad_norm": 0.8308070784365134,
+      "learning_rate": 3.392696963619972e-07,
+      "loss": 0.0045,
+      "step": 14326
+    },
+    {
+      "epoch": 3.256136363636364,
+      "grad_norm": 1.2016789665665064,
+      "learning_rate": 3.3919032228656243e-07,
+      "loss": 0.0046,
+      "step": 14327
+    },
+    {
+      "epoch": 3.2563636363636363,
+      "grad_norm": 1.0731714471767402,
+      "learning_rate": 3.3911095403928564e-07,
+      "loss": 0.0149,
+      "step": 14328
+    },
+    {
+      "epoch": 3.2565909090909093,
+      "grad_norm": 1.3791054293033356,
+      "learning_rate": 3.390315916217854e-07,
+      "loss": 0.0104,
+      "step": 14329
+    },
+    {
+      "epoch": 3.256818181818182,
+      "grad_norm": 1.2379650899861872,
+      "learning_rate": 3.389522350356797e-07,
+      "loss": 0.0116,
+      "step": 14330
+    },
+    {
+      "epoch": 3.2570454545454544,
+      "grad_norm": 1.8315657003846868,
+      "learning_rate": 3.3887288428258697e-07,
+      "loss": 0.0065,
+      "step": 14331
+    },
+    {
+      "epoch": 3.2572727272727273,
+      "grad_norm": 0.6497534325072758,
+      "learning_rate": 3.3879353936412533e-07,
+      "loss": 0.0026,
+      "step": 14332
+    },
+    {
+      "epoch": 3.2575,
+      "grad_norm": 2.2821056772454797,
+      "learning_rate": 3.387142002819127e-07,
+      "loss": 0.0341,
+      "step": 14333
+    },
+    {
+      "epoch": 3.257727272727273,
+      "grad_norm": 1.7822827696227381,
+      "learning_rate": 3.386348670375672e-07,
+      "loss": 0.0125,
+      "step": 14334
+    },
+    {
+      "epoch": 3.2579545454545453,
+      "grad_norm": 2.328687944345858,
+      "learning_rate": 3.3855553963270614e-07,
+      "loss": 0.0131,
+      "step": 14335
+    },
+    {
+      "epoch": 3.2581818181818183,
+      "grad_norm": 1.1461136122374223,
+      "learning_rate": 3.384762180689474e-07,
+      "loss": 0.0064,
+      "step": 14336
+    },
+    {
+      "epoch": 3.258409090909091,
+      "grad_norm": 2.916569280198369,
+      "learning_rate": 3.3839690234790846e-07,
+      "loss": 0.0218,
+      "step": 14337
+    },
+    {
+      "epoch": 3.2586363636363638,
+      "grad_norm": 0.8317288672291724,
+      "learning_rate": 3.3831759247120663e-07,
+      "loss": 0.0081,
+      "step": 14338
+    },
+    {
+      "epoch": 3.2588636363636363,
+      "grad_norm": 1.5149734595569342,
+      "learning_rate": 3.3823828844045917e-07,
+      "loss": 0.0124,
+      "step": 14339
+    },
+    {
+      "epoch": 3.2590909090909093,
+      "grad_norm": 3.158331103969411,
+      "learning_rate": 3.3815899025728334e-07,
+      "loss": 0.0232,
+      "step": 14340
+    },
+    {
+      "epoch": 3.259318181818182,
+      "grad_norm": 1.2969252681625005,
+      "learning_rate": 3.380796979232961e-07,
+      "loss": 0.0104,
+      "step": 14341
+    },
+    {
+      "epoch": 3.2595454545454547,
+      "grad_norm": 1.2459115011327566,
+      "learning_rate": 3.3800041144011476e-07,
+      "loss": 0.0078,
+      "step": 14342
+    },
+    {
+      "epoch": 3.2597727272727273,
+      "grad_norm": 1.8367478422087888,
+      "learning_rate": 3.3792113080935535e-07,
+      "loss": 0.0125,
+      "step": 14343
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 1.5073180395623784,
+      "learning_rate": 3.37841856032635e-07,
+      "loss": 0.0126,
+      "step": 14344
+    },
+    {
+      "epoch": 3.2602272727272728,
+      "grad_norm": 1.7196945132339205,
+      "learning_rate": 3.3776258711157016e-07,
+      "loss": 0.0124,
+      "step": 14345
+    },
+    {
+      "epoch": 3.2604545454545453,
+      "grad_norm": 1.5817589421412477,
+      "learning_rate": 3.3768332404777724e-07,
+      "loss": 0.0049,
+      "step": 14346
+    },
+    {
+      "epoch": 3.2606818181818182,
+      "grad_norm": 1.682877323260659,
+      "learning_rate": 3.376040668428729e-07,
+      "loss": 0.0269,
+      "step": 14347
+    },
+    {
+      "epoch": 3.2609090909090908,
+      "grad_norm": 1.9046776493864042,
+      "learning_rate": 3.3752481549847266e-07,
+      "loss": 0.0257,
+      "step": 14348
+    },
+    {
+      "epoch": 3.2611363636363637,
+      "grad_norm": 1.7739841740853723,
+      "learning_rate": 3.3744557001619306e-07,
+      "loss": 0.0228,
+      "step": 14349
+    },
+    {
+      "epoch": 3.2613636363636362,
+      "grad_norm": 1.124311410944887,
+      "learning_rate": 3.373663303976499e-07,
+      "loss": 0.0097,
+      "step": 14350
+    },
+    {
+      "epoch": 3.261590909090909,
+      "grad_norm": 0.6462801475365952,
+      "learning_rate": 3.372870966444591e-07,
+      "loss": 0.0041,
+      "step": 14351
+    },
+    {
+      "epoch": 3.2618181818181817,
+      "grad_norm": 1.1907054253857894,
+      "learning_rate": 3.372078687582364e-07,
+      "loss": 0.0072,
+      "step": 14352
+    },
+    {
+      "epoch": 3.2620454545454547,
+      "grad_norm": 2.1507151639176256,
+      "learning_rate": 3.3712864674059724e-07,
+      "loss": 0.0149,
+      "step": 14353
+    },
+    {
+      "epoch": 3.262272727272727,
+      "grad_norm": 1.082923625404276,
+      "learning_rate": 3.3704943059315723e-07,
+      "loss": 0.0075,
+      "step": 14354
+    },
+    {
+      "epoch": 3.2625,
+      "grad_norm": 1.4982915349296309,
+      "learning_rate": 3.369702203175317e-07,
+      "loss": 0.0057,
+      "step": 14355
+    },
+    {
+      "epoch": 3.2627272727272727,
+      "grad_norm": 1.1373098849183394,
+      "learning_rate": 3.3689101591533617e-07,
+      "loss": 0.007,
+      "step": 14356
+    },
+    {
+      "epoch": 3.262954545454545,
+      "grad_norm": 0.9772786203660263,
+      "learning_rate": 3.368118173881851e-07,
+      "loss": 0.0072,
+      "step": 14357
+    },
+    {
+      "epoch": 3.263181818181818,
+      "grad_norm": 0.8723213876438637,
+      "learning_rate": 3.367326247376939e-07,
+      "loss": 0.0078,
+      "step": 14358
+    },
+    {
+      "epoch": 3.263409090909091,
+      "grad_norm": 0.7624959973889364,
+      "learning_rate": 3.366534379654774e-07,
+      "loss": 0.0047,
+      "step": 14359
+    },
+    {
+      "epoch": 3.2636363636363637,
+      "grad_norm": 2.8519779108386194,
+      "learning_rate": 3.365742570731506e-07,
+      "loss": 0.0202,
+      "step": 14360
+    },
+    {
+      "epoch": 3.263863636363636,
+      "grad_norm": 1.815301894129936,
+      "learning_rate": 3.364950820623276e-07,
+      "loss": 0.0126,
+      "step": 14361
+    },
+    {
+      "epoch": 3.264090909090909,
+      "grad_norm": 2.672406201302277,
+      "learning_rate": 3.3641591293462323e-07,
+      "loss": 0.0176,
+      "step": 14362
+    },
+    {
+      "epoch": 3.2643181818181817,
+      "grad_norm": 1.5631155128748395,
+      "learning_rate": 3.363367496916518e-07,
+      "loss": 0.0101,
+      "step": 14363
+    },
+    {
+      "epoch": 3.2645454545454546,
+      "grad_norm": 1.4841445117198482,
+      "learning_rate": 3.3625759233502797e-07,
+      "loss": 0.0144,
+      "step": 14364
+    },
+    {
+      "epoch": 3.264772727272727,
+      "grad_norm": 1.0207260209703382,
+      "learning_rate": 3.361784408663653e-07,
+      "loss": 0.0048,
+      "step": 14365
+    },
+    {
+      "epoch": 3.265,
+      "grad_norm": 2.1930122663779748,
+      "learning_rate": 3.360992952872781e-07,
+      "loss": 0.0228,
+      "step": 14366
+    },
+    {
+      "epoch": 3.2652272727272726,
+      "grad_norm": 2.352079804335192,
+      "learning_rate": 3.360201555993803e-07,
+      "loss": 0.0171,
+      "step": 14367
+    },
+    {
+      "epoch": 3.2654545454545456,
+      "grad_norm": 1.421001406154217,
+      "learning_rate": 3.359410218042854e-07,
+      "loss": 0.0062,
+      "step": 14368
+    },
+    {
+      "epoch": 3.265681818181818,
+      "grad_norm": 1.259471963866057,
+      "learning_rate": 3.358618939036079e-07,
+      "loss": 0.0104,
+      "step": 14369
+    },
+    {
+      "epoch": 3.265909090909091,
+      "grad_norm": 1.0992381223844356,
+      "learning_rate": 3.357827718989605e-07,
+      "loss": 0.0072,
+      "step": 14370
+    },
+    {
+      "epoch": 3.2661363636363636,
+      "grad_norm": 4.662538621888133,
+      "learning_rate": 3.35703655791957e-07,
+      "loss": 0.0286,
+      "step": 14371
+    },
+    {
+      "epoch": 3.2663636363636366,
+      "grad_norm": 1.9759527229468679,
+      "learning_rate": 3.356245455842106e-07,
+      "loss": 0.0224,
+      "step": 14372
+    },
+    {
+      "epoch": 3.266590909090909,
+      "grad_norm": 1.8816689412268464,
+      "learning_rate": 3.3554544127733483e-07,
+      "loss": 0.0079,
+      "step": 14373
+    },
+    {
+      "epoch": 3.2668181818181816,
+      "grad_norm": 1.2518118230672257,
+      "learning_rate": 3.3546634287294225e-07,
+      "loss": 0.0117,
+      "step": 14374
+    },
+    {
+      "epoch": 3.2670454545454546,
+      "grad_norm": 1.6477631136725954,
+      "learning_rate": 3.353872503726461e-07,
+      "loss": 0.0189,
+      "step": 14375
+    },
+    {
+      "epoch": 3.267272727272727,
+      "grad_norm": 2.499758321067624,
+      "learning_rate": 3.353081637780591e-07,
+      "loss": 0.0352,
+      "step": 14376
+    },
+    {
+      "epoch": 3.2675,
+      "grad_norm": 0.9669651689876253,
+      "learning_rate": 3.3522908309079434e-07,
+      "loss": 0.0106,
+      "step": 14377
+    },
+    {
+      "epoch": 3.2677272727272726,
+      "grad_norm": 1.0618383459952914,
+      "learning_rate": 3.3515000831246385e-07,
+      "loss": 0.0096,
+      "step": 14378
+    },
+    {
+      "epoch": 3.2679545454545456,
+      "grad_norm": 2.5236454769328702,
+      "learning_rate": 3.3507093944468035e-07,
+      "loss": 0.0286,
+      "step": 14379
+    },
+    {
+      "epoch": 3.268181818181818,
+      "grad_norm": 1.2849263691842816,
+      "learning_rate": 3.3499187648905627e-07,
+      "loss": 0.0061,
+      "step": 14380
+    },
+    {
+      "epoch": 3.268409090909091,
+      "grad_norm": 2.4296692903528387,
+      "learning_rate": 3.3491281944720373e-07,
+      "loss": 0.038,
+      "step": 14381
+    },
+    {
+      "epoch": 3.2686363636363636,
+      "grad_norm": 2.630199029218424,
+      "learning_rate": 3.348337683207348e-07,
+      "loss": 0.0212,
+      "step": 14382
+    },
+    {
+      "epoch": 3.2688636363636365,
+      "grad_norm": 1.5812369717970287,
+      "learning_rate": 3.347547231112616e-07,
+      "loss": 0.0121,
+      "step": 14383
+    },
+    {
+      "epoch": 3.269090909090909,
+      "grad_norm": 0.8389595993627444,
+      "learning_rate": 3.34675683820396e-07,
+      "loss": 0.0043,
+      "step": 14384
+    },
+    {
+      "epoch": 3.269318181818182,
+      "grad_norm": 2.2333212676417276,
+      "learning_rate": 3.345966504497497e-07,
+      "loss": 0.0308,
+      "step": 14385
+    },
+    {
+      "epoch": 3.2695454545454545,
+      "grad_norm": 1.1030930976602353,
+      "learning_rate": 3.345176230009346e-07,
+      "loss": 0.0061,
+      "step": 14386
+    },
+    {
+      "epoch": 3.269772727272727,
+      "grad_norm": 1.8584571303729065,
+      "learning_rate": 3.344386014755617e-07,
+      "loss": 0.0231,
+      "step": 14387
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 1.29805824107191,
+      "learning_rate": 3.3435958587524255e-07,
+      "loss": 0.0038,
+      "step": 14388
+    },
+    {
+      "epoch": 3.2702272727272725,
+      "grad_norm": 2.0755942130835954,
+      "learning_rate": 3.3428057620158857e-07,
+      "loss": 0.0197,
+      "step": 14389
+    },
+    {
+      "epoch": 3.2704545454545455,
+      "grad_norm": 1.319053943610463,
+      "learning_rate": 3.3420157245621073e-07,
+      "loss": 0.0108,
+      "step": 14390
+    },
+    {
+      "epoch": 3.270681818181818,
+      "grad_norm": 1.491927248038698,
+      "learning_rate": 3.341225746407205e-07,
+      "loss": 0.0039,
+      "step": 14391
+    },
+    {
+      "epoch": 3.270909090909091,
+      "grad_norm": 1.964875727750319,
+      "learning_rate": 3.3404358275672813e-07,
+      "loss": 0.0057,
+      "step": 14392
+    },
+    {
+      "epoch": 3.2711363636363635,
+      "grad_norm": 1.7118123719410872,
+      "learning_rate": 3.3396459680584476e-07,
+      "loss": 0.0121,
+      "step": 14393
+    },
+    {
+      "epoch": 3.2713636363636365,
+      "grad_norm": 2.0656376740857287,
+      "learning_rate": 3.3388561678968105e-07,
+      "loss": 0.0298,
+      "step": 14394
+    },
+    {
+      "epoch": 3.271590909090909,
+      "grad_norm": 2.448860809134318,
+      "learning_rate": 3.3380664270984754e-07,
+      "loss": 0.012,
+      "step": 14395
+    },
+    {
+      "epoch": 3.271818181818182,
+      "grad_norm": 2.245535337073364,
+      "learning_rate": 3.337276745679545e-07,
+      "loss": 0.014,
+      "step": 14396
+    },
+    {
+      "epoch": 3.2720454545454545,
+      "grad_norm": 1.2235928274507124,
+      "learning_rate": 3.336487123656121e-07,
+      "loss": 0.0137,
+      "step": 14397
+    },
+    {
+      "epoch": 3.2722727272727274,
+      "grad_norm": 3.6026654726317524,
+      "learning_rate": 3.3356975610443096e-07,
+      "loss": 0.0262,
+      "step": 14398
+    },
+    {
+      "epoch": 3.2725,
+      "grad_norm": 1.368027685658092,
+      "learning_rate": 3.334908057860212e-07,
+      "loss": 0.0231,
+      "step": 14399
+    },
+    {
+      "epoch": 3.2727272727272725,
+      "grad_norm": 0.9437615807331389,
+      "learning_rate": 3.3341186141199216e-07,
+      "loss": 0.002,
+      "step": 14400
+    },
+    {
+      "epoch": 3.2729545454545454,
+      "grad_norm": 1.6457097688244382,
+      "learning_rate": 3.3333292298395396e-07,
+      "loss": 0.0079,
+      "step": 14401
+    },
+    {
+      "epoch": 3.2731818181818184,
+      "grad_norm": 2.798536801817926,
+      "learning_rate": 3.332539905035163e-07,
+      "loss": 0.0218,
+      "step": 14402
+    },
+    {
+      "epoch": 3.273409090909091,
+      "grad_norm": 0.8678605568649652,
+      "learning_rate": 3.331750639722889e-07,
+      "loss": 0.0056,
+      "step": 14403
+    },
+    {
+      "epoch": 3.2736363636363635,
+      "grad_norm": 0.8697004535341715,
+      "learning_rate": 3.3309614339188114e-07,
+      "loss": 0.006,
+      "step": 14404
+    },
+    {
+      "epoch": 3.2738636363636364,
+      "grad_norm": 1.2866034133434898,
+      "learning_rate": 3.3301722876390207e-07,
+      "loss": 0.0086,
+      "step": 14405
+    },
+    {
+      "epoch": 3.274090909090909,
+      "grad_norm": 1.2746235969668405,
+      "learning_rate": 3.3293832008996105e-07,
+      "loss": 0.019,
+      "step": 14406
+    },
+    {
+      "epoch": 3.274318181818182,
+      "grad_norm": 1.5182538516193136,
+      "learning_rate": 3.3285941737166734e-07,
+      "loss": 0.0126,
+      "step": 14407
+    },
+    {
+      "epoch": 3.2745454545454544,
+      "grad_norm": 1.139473362275713,
+      "learning_rate": 3.3278052061062994e-07,
+      "loss": 0.0095,
+      "step": 14408
+    },
+    {
+      "epoch": 3.2747727272727274,
+      "grad_norm": 3.836303510901448,
+      "learning_rate": 3.327016298084573e-07,
+      "loss": 0.0318,
+      "step": 14409
+    },
+    {
+      "epoch": 3.275,
+      "grad_norm": 2.88012394304215,
+      "learning_rate": 3.326227449667584e-07,
+      "loss": 0.0152,
+      "step": 14410
+    },
+    {
+      "epoch": 3.275227272727273,
+      "grad_norm": 0.782737063798894,
+      "learning_rate": 3.325438660871418e-07,
+      "loss": 0.0039,
+      "step": 14411
+    },
+    {
+      "epoch": 3.2754545454545454,
+      "grad_norm": 2.224110150878614,
+      "learning_rate": 3.3246499317121596e-07,
+      "loss": 0.0155,
+      "step": 14412
+    },
+    {
+      "epoch": 3.2756818181818184,
+      "grad_norm": 3.982132838204702,
+      "learning_rate": 3.323861262205892e-07,
+      "loss": 0.025,
+      "step": 14413
+    },
+    {
+      "epoch": 3.275909090909091,
+      "grad_norm": 1.9581589338300043,
+      "learning_rate": 3.323072652368699e-07,
+      "loss": 0.0143,
+      "step": 14414
+    },
+    {
+      "epoch": 3.276136363636364,
+      "grad_norm": 2.5266729770306027,
+      "learning_rate": 3.322284102216661e-07,
+      "loss": 0.0339,
+      "step": 14415
+    },
+    {
+      "epoch": 3.2763636363636364,
+      "grad_norm": 1.400959621234091,
+      "learning_rate": 3.321495611765858e-07,
+      "loss": 0.0056,
+      "step": 14416
+    },
+    {
+      "epoch": 3.276590909090909,
+      "grad_norm": 1.540320504898263,
+      "learning_rate": 3.3207071810323706e-07,
+      "loss": 0.0084,
+      "step": 14417
+    },
+    {
+      "epoch": 3.276818181818182,
+      "grad_norm": 0.7037155773173867,
+      "learning_rate": 3.3199188100322713e-07,
+      "loss": 0.0038,
+      "step": 14418
+    },
+    {
+      "epoch": 3.2770454545454544,
+      "grad_norm": 2.5442695797751558,
+      "learning_rate": 3.3191304987816406e-07,
+      "loss": 0.0122,
+      "step": 14419
+    },
+    {
+      "epoch": 3.2772727272727273,
+      "grad_norm": 2.7361132759042706,
+      "learning_rate": 3.318342247296551e-07,
+      "loss": 0.0155,
+      "step": 14420
+    },
+    {
+      "epoch": 3.2775,
+      "grad_norm": 0.811476893689481,
+      "learning_rate": 3.317554055593081e-07,
+      "loss": 0.0021,
+      "step": 14421
+    },
+    {
+      "epoch": 3.277727272727273,
+      "grad_norm": 1.483095958360067,
+      "learning_rate": 3.316765923687296e-07,
+      "loss": 0.0139,
+      "step": 14422
+    },
+    {
+      "epoch": 3.2779545454545453,
+      "grad_norm": 0.659119369802437,
+      "learning_rate": 3.3159778515952716e-07,
+      "loss": 0.0042,
+      "step": 14423
+    },
+    {
+      "epoch": 3.2781818181818183,
+      "grad_norm": 2.923685609717366,
+      "learning_rate": 3.315189839333078e-07,
+      "loss": 0.0225,
+      "step": 14424
+    },
+    {
+      "epoch": 3.278409090909091,
+      "grad_norm": 0.8166650688223933,
+      "learning_rate": 3.3144018869167837e-07,
+      "loss": 0.0044,
+      "step": 14425
+    },
+    {
+      "epoch": 3.278636363636364,
+      "grad_norm": 1.0973859914372315,
+      "learning_rate": 3.313613994362456e-07,
+      "loss": 0.0059,
+      "step": 14426
+    },
+    {
+      "epoch": 3.2788636363636363,
+      "grad_norm": 3.102976441860849,
+      "learning_rate": 3.3128261616861615e-07,
+      "loss": 0.0189,
+      "step": 14427
+    },
+    {
+      "epoch": 3.2790909090909093,
+      "grad_norm": 2.110268391077711,
+      "learning_rate": 3.312038388903966e-07,
+      "loss": 0.0193,
+      "step": 14428
+    },
+    {
+      "epoch": 3.279318181818182,
+      "grad_norm": 2.390806801358352,
+      "learning_rate": 3.311250676031933e-07,
+      "loss": 0.0087,
+      "step": 14429
+    },
+    {
+      "epoch": 3.2795454545454543,
+      "grad_norm": 1.7289820249101608,
+      "learning_rate": 3.310463023086129e-07,
+      "loss": 0.0108,
+      "step": 14430
+    },
+    {
+      "epoch": 3.2797727272727273,
+      "grad_norm": 2.066678392550814,
+      "learning_rate": 3.309675430082609e-07,
+      "loss": 0.0268,
+      "step": 14431
+    },
+    {
+      "epoch": 3.2800000000000002,
+      "grad_norm": 0.37703246384085715,
+      "learning_rate": 3.308887897037437e-07,
+      "loss": 0.0019,
+      "step": 14432
+    },
+    {
+      "epoch": 3.2802272727272728,
+      "grad_norm": 0.8656568528533244,
+      "learning_rate": 3.3081004239666715e-07,
+      "loss": 0.0058,
+      "step": 14433
+    },
+    {
+      "epoch": 3.2804545454545453,
+      "grad_norm": 1.0096495882067826,
+      "learning_rate": 3.3073130108863703e-07,
+      "loss": 0.0047,
+      "step": 14434
+    },
+    {
+      "epoch": 3.2806818181818183,
+      "grad_norm": 1.5451438905342751,
+      "learning_rate": 3.306525657812594e-07,
+      "loss": 0.012,
+      "step": 14435
+    },
+    {
+      "epoch": 3.2809090909090908,
+      "grad_norm": 1.6064378832477004,
+      "learning_rate": 3.305738364761392e-07,
+      "loss": 0.0068,
+      "step": 14436
+    },
+    {
+      "epoch": 3.2811363636363637,
+      "grad_norm": 1.628214369902856,
+      "learning_rate": 3.3049511317488214e-07,
+      "loss": 0.0118,
+      "step": 14437
+    },
+    {
+      "epoch": 3.2813636363636363,
+      "grad_norm": 1.8502324603869131,
+      "learning_rate": 3.304163958790935e-07,
+      "loss": 0.0134,
+      "step": 14438
+    },
+    {
+      "epoch": 3.2815909090909092,
+      "grad_norm": 2.1783178092522064,
+      "learning_rate": 3.3033768459037873e-07,
+      "loss": 0.0196,
+      "step": 14439
+    },
+    {
+      "epoch": 3.2818181818181817,
+      "grad_norm": 4.14402383291845,
+      "learning_rate": 3.302589793103424e-07,
+      "loss": 0.035,
+      "step": 14440
+    },
+    {
+      "epoch": 3.2820454545454547,
+      "grad_norm": 1.715438190767269,
+      "learning_rate": 3.301802800405897e-07,
+      "loss": 0.0096,
+      "step": 14441
+    },
+    {
+      "epoch": 3.2822727272727272,
+      "grad_norm": 2.151916583989363,
+      "learning_rate": 3.3010158678272537e-07,
+      "loss": 0.0096,
+      "step": 14442
+    },
+    {
+      "epoch": 3.2824999999999998,
+      "grad_norm": 1.7002705541462662,
+      "learning_rate": 3.300228995383542e-07,
+      "loss": 0.0098,
+      "step": 14443
+    },
+    {
+      "epoch": 3.2827272727272727,
+      "grad_norm": 0.6276687222017701,
+      "learning_rate": 3.299442183090808e-07,
+      "loss": 0.0024,
+      "step": 14444
+    },
+    {
+      "epoch": 3.2829545454545457,
+      "grad_norm": 1.2708625363562356,
+      "learning_rate": 3.2986554309650935e-07,
+      "loss": 0.0076,
+      "step": 14445
+    },
+    {
+      "epoch": 3.283181818181818,
+      "grad_norm": 1.0540705526532292,
+      "learning_rate": 3.2978687390224454e-07,
+      "loss": 0.005,
+      "step": 14446
+    },
+    {
+      "epoch": 3.2834090909090907,
+      "grad_norm": 1.850320245185517,
+      "learning_rate": 3.297082107278903e-07,
+      "loss": 0.0137,
+      "step": 14447
+    },
+    {
+      "epoch": 3.2836363636363637,
+      "grad_norm": 1.3322549489228375,
+      "learning_rate": 3.2962955357505113e-07,
+      "loss": 0.0098,
+      "step": 14448
+    },
+    {
+      "epoch": 3.283863636363636,
+      "grad_norm": 1.2257929231902627,
+      "learning_rate": 3.2955090244533035e-07,
+      "loss": 0.0049,
+      "step": 14449
+    },
+    {
+      "epoch": 3.284090909090909,
+      "grad_norm": 1.1593579565183532,
+      "learning_rate": 3.2947225734033215e-07,
+      "loss": 0.0085,
+      "step": 14450
+    },
+    {
+      "epoch": 3.2843181818181817,
+      "grad_norm": 3.1026566402191316,
+      "learning_rate": 3.293936182616602e-07,
+      "loss": 0.0185,
+      "step": 14451
+    },
+    {
+      "epoch": 3.2845454545454547,
+      "grad_norm": 2.878073504368733,
+      "learning_rate": 3.293149852109183e-07,
+      "loss": 0.0222,
+      "step": 14452
+    },
+    {
+      "epoch": 3.284772727272727,
+      "grad_norm": 5.74160439606392,
+      "learning_rate": 3.2923635818970947e-07,
+      "loss": 0.0103,
+      "step": 14453
+    },
+    {
+      "epoch": 3.285,
+      "grad_norm": 3.073232525821986,
+      "learning_rate": 3.291577371996373e-07,
+      "loss": 0.0082,
+      "step": 14454
+    },
+    {
+      "epoch": 3.2852272727272727,
+      "grad_norm": 0.9198621921596396,
+      "learning_rate": 3.2907912224230497e-07,
+      "loss": 0.0031,
+      "step": 14455
+    },
+    {
+      "epoch": 3.2854545454545456,
+      "grad_norm": 1.9295190205139305,
+      "learning_rate": 3.290005133193157e-07,
+      "loss": 0.0143,
+      "step": 14456
+    },
+    {
+      "epoch": 3.285681818181818,
+      "grad_norm": 1.6283842842345382,
+      "learning_rate": 3.2892191043227226e-07,
+      "loss": 0.0148,
+      "step": 14457
+    },
+    {
+      "epoch": 3.285909090909091,
+      "grad_norm": 0.3946364881073521,
+      "learning_rate": 3.288433135827776e-07,
+      "loss": 0.0045,
+      "step": 14458
+    },
+    {
+      "epoch": 3.2861363636363636,
+      "grad_norm": 1.1922155902937248,
+      "learning_rate": 3.2876472277243445e-07,
+      "loss": 0.0097,
+      "step": 14459
+    },
+    {
+      "epoch": 3.286363636363636,
+      "grad_norm": 2.71812039681034,
+      "learning_rate": 3.2868613800284543e-07,
+      "loss": 0.0204,
+      "step": 14460
+    },
+    {
+      "epoch": 3.286590909090909,
+      "grad_norm": 1.441689826944961,
+      "learning_rate": 3.2860755927561326e-07,
+      "loss": 0.0116,
+      "step": 14461
+    },
+    {
+      "epoch": 3.2868181818181816,
+      "grad_norm": 4.304312371633138,
+      "learning_rate": 3.285289865923398e-07,
+      "loss": 0.0537,
+      "step": 14462
+    },
+    {
+      "epoch": 3.2870454545454546,
+      "grad_norm": 1.1065704642139316,
+      "learning_rate": 3.2845041995462746e-07,
+      "loss": 0.0071,
+      "step": 14463
+    },
+    {
+      "epoch": 3.287272727272727,
+      "grad_norm": 1.9742339180535204,
+      "learning_rate": 3.2837185936407857e-07,
+      "loss": 0.0202,
+      "step": 14464
+    },
+    {
+      "epoch": 3.2875,
+      "grad_norm": 0.9895053016799545,
+      "learning_rate": 3.2829330482229514e-07,
+      "loss": 0.0044,
+      "step": 14465
+    },
+    {
+      "epoch": 3.2877272727272726,
+      "grad_norm": 2.12832346845594,
+      "learning_rate": 3.2821475633087854e-07,
+      "loss": 0.0132,
+      "step": 14466
+    },
+    {
+      "epoch": 3.2879545454545456,
+      "grad_norm": 1.1465355902450889,
+      "learning_rate": 3.281362138914308e-07,
+      "loss": 0.007,
+      "step": 14467
+    },
+    {
+      "epoch": 3.288181818181818,
+      "grad_norm": 0.6281781718844035,
+      "learning_rate": 3.280576775055537e-07,
+      "loss": 0.0023,
+      "step": 14468
+    },
+    {
+      "epoch": 3.288409090909091,
+      "grad_norm": 2.1960700065749696,
+      "learning_rate": 3.279791471748485e-07,
+      "loss": 0.0093,
+      "step": 14469
+    },
+    {
+      "epoch": 3.2886363636363636,
+      "grad_norm": 2.394293442239701,
+      "learning_rate": 3.27900622900917e-07,
+      "loss": 0.0113,
+      "step": 14470
+    },
+    {
+      "epoch": 3.2888636363636365,
+      "grad_norm": 1.2776053856208867,
+      "learning_rate": 3.278221046853598e-07,
+      "loss": 0.012,
+      "step": 14471
+    },
+    {
+      "epoch": 3.289090909090909,
+      "grad_norm": 0.9533303463744962,
+      "learning_rate": 3.277435925297784e-07,
+      "loss": 0.0037,
+      "step": 14472
+    },
+    {
+      "epoch": 3.2893181818181816,
+      "grad_norm": 1.9287025236407196,
+      "learning_rate": 3.2766508643577377e-07,
+      "loss": 0.0132,
+      "step": 14473
+    },
+    {
+      "epoch": 3.2895454545454546,
+      "grad_norm": 0.9283851071136244,
+      "learning_rate": 3.2758658640494664e-07,
+      "loss": 0.0085,
+      "step": 14474
+    },
+    {
+      "epoch": 3.2897727272727275,
+      "grad_norm": 1.531400015559303,
+      "learning_rate": 3.275080924388979e-07,
+      "loss": 0.0152,
+      "step": 14475
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 0.6869762124083315,
+      "learning_rate": 3.274296045392282e-07,
+      "loss": 0.005,
+      "step": 14476
+    },
+    {
+      "epoch": 3.2902272727272726,
+      "grad_norm": 1.6239809613642853,
+      "learning_rate": 3.27351122707538e-07,
+      "loss": 0.012,
+      "step": 14477
+    },
+    {
+      "epoch": 3.2904545454545455,
+      "grad_norm": 1.671701730919624,
+      "learning_rate": 3.2727264694542764e-07,
+      "loss": 0.0107,
+      "step": 14478
+    },
+    {
+      "epoch": 3.290681818181818,
+      "grad_norm": 0.8382301254312401,
+      "learning_rate": 3.2719417725449774e-07,
+      "loss": 0.0085,
+      "step": 14479
+    },
+    {
+      "epoch": 3.290909090909091,
+      "grad_norm": 2.201064243120328,
+      "learning_rate": 3.2711571363634784e-07,
+      "loss": 0.0219,
+      "step": 14480
+    },
+    {
+      "epoch": 3.2911363636363635,
+      "grad_norm": 1.4807199979791907,
+      "learning_rate": 3.2703725609257826e-07,
+      "loss": 0.01,
+      "step": 14481
+    },
+    {
+      "epoch": 3.2913636363636365,
+      "grad_norm": 2.0620674932623277,
+      "learning_rate": 3.2695880462478883e-07,
+      "loss": 0.0061,
+      "step": 14482
+    },
+    {
+      "epoch": 3.291590909090909,
+      "grad_norm": 1.708986997769534,
+      "learning_rate": 3.2688035923457953e-07,
+      "loss": 0.0075,
+      "step": 14483
+    },
+    {
+      "epoch": 3.291818181818182,
+      "grad_norm": 1.9034668936931247,
+      "learning_rate": 3.2680191992354963e-07,
+      "loss": 0.0238,
+      "step": 14484
+    },
+    {
+      "epoch": 3.2920454545454545,
+      "grad_norm": 3.1126942681633136,
+      "learning_rate": 3.2672348669329874e-07,
+      "loss": 0.0175,
+      "step": 14485
+    },
+    {
+      "epoch": 3.2922727272727275,
+      "grad_norm": 0.4282275035797667,
+      "learning_rate": 3.266450595454264e-07,
+      "loss": 0.0026,
+      "step": 14486
+    },
+    {
+      "epoch": 3.2925,
+      "grad_norm": 4.586795002064447,
+      "learning_rate": 3.265666384815318e-07,
+      "loss": 0.0223,
+      "step": 14487
+    },
+    {
+      "epoch": 3.292727272727273,
+      "grad_norm": 2.149190239094213,
+      "learning_rate": 3.264882235032141e-07,
+      "loss": 0.0146,
+      "step": 14488
+    },
+    {
+      "epoch": 3.2929545454545455,
+      "grad_norm": 0.9846104954924065,
+      "learning_rate": 3.264098146120723e-07,
+      "loss": 0.0059,
+      "step": 14489
+    },
+    {
+      "epoch": 3.293181818181818,
+      "grad_norm": 2.3588928936791076,
+      "learning_rate": 3.2633141180970523e-07,
+      "loss": 0.0154,
+      "step": 14490
+    },
+    {
+      "epoch": 3.293409090909091,
+      "grad_norm": 3.07697063963749,
+      "learning_rate": 3.2625301509771174e-07,
+      "loss": 0.019,
+      "step": 14491
+    },
+    {
+      "epoch": 3.2936363636363635,
+      "grad_norm": 1.1367978612166656,
+      "learning_rate": 3.2617462447769075e-07,
+      "loss": 0.0118,
+      "step": 14492
+    },
+    {
+      "epoch": 3.2938636363636364,
+      "grad_norm": 1.461903066924712,
+      "learning_rate": 3.260962399512402e-07,
+      "loss": 0.0078,
+      "step": 14493
+    },
+    {
+      "epoch": 3.294090909090909,
+      "grad_norm": 2.1792602304103688,
+      "learning_rate": 3.260178615199588e-07,
+      "loss": 0.0155,
+      "step": 14494
+    },
+    {
+      "epoch": 3.294318181818182,
+      "grad_norm": 4.987606934369704,
+      "learning_rate": 3.259394891854449e-07,
+      "loss": 0.0383,
+      "step": 14495
+    },
+    {
+      "epoch": 3.2945454545454544,
+      "grad_norm": 1.978619124429341,
+      "learning_rate": 3.258611229492967e-07,
+      "loss": 0.0196,
+      "step": 14496
+    },
+    {
+      "epoch": 3.2947727272727274,
+      "grad_norm": 2.926621715756907,
+      "learning_rate": 3.257827628131119e-07,
+      "loss": 0.0199,
+      "step": 14497
+    },
+    {
+      "epoch": 3.295,
+      "grad_norm": 1.6637442921570522,
+      "learning_rate": 3.257044087784885e-07,
+      "loss": 0.0196,
+      "step": 14498
+    },
+    {
+      "epoch": 3.295227272727273,
+      "grad_norm": 3.6105607193693485,
+      "learning_rate": 3.256260608470244e-07,
+      "loss": 0.0162,
+      "step": 14499
+    },
+    {
+      "epoch": 3.2954545454545454,
+      "grad_norm": 1.2365621207696222,
+      "learning_rate": 3.2554771902031723e-07,
+      "loss": 0.0088,
+      "step": 14500
+    },
+    {
+      "epoch": 3.2956818181818184,
+      "grad_norm": 1.2902050564179675,
+      "learning_rate": 3.2546938329996475e-07,
+      "loss": 0.0116,
+      "step": 14501
+    },
+    {
+      "epoch": 3.295909090909091,
+      "grad_norm": 1.9623160689265322,
+      "learning_rate": 3.2539105368756373e-07,
+      "loss": 0.017,
+      "step": 14502
+    },
+    {
+      "epoch": 3.2961363636363634,
+      "grad_norm": 1.5946714292594293,
+      "learning_rate": 3.25312730184712e-07,
+      "loss": 0.0143,
+      "step": 14503
+    },
+    {
+      "epoch": 3.2963636363636364,
+      "grad_norm": 0.5368403683574414,
+      "learning_rate": 3.252344127930064e-07,
+      "loss": 0.0033,
+      "step": 14504
+    },
+    {
+      "epoch": 3.296590909090909,
+      "grad_norm": 1.2488692659038876,
+      "learning_rate": 3.251561015140442e-07,
+      "loss": 0.0056,
+      "step": 14505
+    },
+    {
+      "epoch": 3.296818181818182,
+      "grad_norm": 1.6333652286225668,
+      "learning_rate": 3.2507779634942216e-07,
+      "loss": 0.017,
+      "step": 14506
+    },
+    {
+      "epoch": 3.2970454545454544,
+      "grad_norm": 0.9714430269665639,
+      "learning_rate": 3.2499949730073697e-07,
+      "loss": 0.0021,
+      "step": 14507
+    },
+    {
+      "epoch": 3.2972727272727274,
+      "grad_norm": 3.7267223354211527,
+      "learning_rate": 3.2492120436958557e-07,
+      "loss": 0.0166,
+      "step": 14508
+    },
+    {
+      "epoch": 3.2975,
+      "grad_norm": 1.6447145023280287,
+      "learning_rate": 3.2484291755756445e-07,
+      "loss": 0.0056,
+      "step": 14509
+    },
+    {
+      "epoch": 3.297727272727273,
+      "grad_norm": 1.409314689122566,
+      "learning_rate": 3.247646368662698e-07,
+      "loss": 0.0072,
+      "step": 14510
+    },
+    {
+      "epoch": 3.2979545454545454,
+      "grad_norm": 3.8023797032621403,
+      "learning_rate": 3.246863622972979e-07,
+      "loss": 0.016,
+      "step": 14511
+    },
+    {
+      "epoch": 3.2981818181818183,
+      "grad_norm": 0.6766452176031923,
+      "learning_rate": 3.24608093852245e-07,
+      "loss": 0.0064,
+      "step": 14512
+    },
+    {
+      "epoch": 3.298409090909091,
+      "grad_norm": 1.4268267761712157,
+      "learning_rate": 3.245298315327071e-07,
+      "loss": 0.0067,
+      "step": 14513
+    },
+    {
+      "epoch": 3.298636363636364,
+      "grad_norm": 0.8236069769050756,
+      "learning_rate": 3.244515753402804e-07,
+      "loss": 0.0092,
+      "step": 14514
+    },
+    {
+      "epoch": 3.2988636363636363,
+      "grad_norm": 2.7495437747357494,
+      "learning_rate": 3.243733252765602e-07,
+      "loss": 0.0098,
+      "step": 14515
+    },
+    {
+      "epoch": 3.299090909090909,
+      "grad_norm": 2.9457851979513827,
+      "learning_rate": 3.2429508134314237e-07,
+      "loss": 0.0139,
+      "step": 14516
+    },
+    {
+      "epoch": 3.299318181818182,
+      "grad_norm": 1.3060138077897536,
+      "learning_rate": 3.242168435416224e-07,
+      "loss": 0.0061,
+      "step": 14517
+    },
+    {
+      "epoch": 3.299545454545455,
+      "grad_norm": 4.814940152250426,
+      "learning_rate": 3.241386118735958e-07,
+      "loss": 0.0212,
+      "step": 14518
+    },
+    {
+      "epoch": 3.2997727272727273,
+      "grad_norm": 1.3813700740441266,
+      "learning_rate": 3.2406038634065773e-07,
+      "loss": 0.0117,
+      "step": 14519
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.9114092446547166,
+      "learning_rate": 3.239821669444034e-07,
+      "loss": 0.0038,
+      "step": 14520
+    },
+    {
+      "epoch": 3.300227272727273,
+      "grad_norm": 1.2289789374798172,
+      "learning_rate": 3.239039536864279e-07,
+      "loss": 0.0077,
+      "step": 14521
+    },
+    {
+      "epoch": 3.3004545454545453,
+      "grad_norm": 2.1281866229971462,
+      "learning_rate": 3.238257465683263e-07,
+      "loss": 0.0193,
+      "step": 14522
+    },
+    {
+      "epoch": 3.3006818181818183,
+      "grad_norm": 2.110684980659572,
+      "learning_rate": 3.2374754559169287e-07,
+      "loss": 0.0221,
+      "step": 14523
+    },
+    {
+      "epoch": 3.300909090909091,
+      "grad_norm": 2.0302433101195185,
+      "learning_rate": 3.2366935075812256e-07,
+      "loss": 0.0095,
+      "step": 14524
+    },
+    {
+      "epoch": 3.3011363636363638,
+      "grad_norm": 1.4398967811529955,
+      "learning_rate": 3.235911620692099e-07,
+      "loss": 0.0095,
+      "step": 14525
+    },
+    {
+      "epoch": 3.3013636363636363,
+      "grad_norm": 1.353490052317498,
+      "learning_rate": 3.235129795265492e-07,
+      "loss": 0.0059,
+      "step": 14526
+    },
+    {
+      "epoch": 3.3015909090909092,
+      "grad_norm": 2.3168414070678707,
+      "learning_rate": 3.234348031317352e-07,
+      "loss": 0.0078,
+      "step": 14527
+    },
+    {
+      "epoch": 3.3018181818181818,
+      "grad_norm": 3.164752830657773,
+      "learning_rate": 3.2335663288636134e-07,
+      "loss": 0.0424,
+      "step": 14528
+    },
+    {
+      "epoch": 3.3020454545454547,
+      "grad_norm": 1.0753505174733178,
+      "learning_rate": 3.232784687920221e-07,
+      "loss": 0.0064,
+      "step": 14529
+    },
+    {
+      "epoch": 3.3022727272727272,
+      "grad_norm": 1.2849136425735996,
+      "learning_rate": 3.2320031085031107e-07,
+      "loss": 0.0106,
+      "step": 14530
+    },
+    {
+      "epoch": 3.3025,
+      "grad_norm": 0.5522206234295488,
+      "learning_rate": 3.231221590628226e-07,
+      "loss": 0.0033,
+      "step": 14531
+    },
+    {
+      "epoch": 3.3027272727272727,
+      "grad_norm": 1.5162707757912672,
+      "learning_rate": 3.230440134311497e-07,
+      "loss": 0.0136,
+      "step": 14532
+    },
+    {
+      "epoch": 3.3029545454545453,
+      "grad_norm": 0.635075606676454,
+      "learning_rate": 3.2296587395688613e-07,
+      "loss": 0.0065,
+      "step": 14533
+    },
+    {
+      "epoch": 3.303181818181818,
+      "grad_norm": 1.2678939028103213,
+      "learning_rate": 3.228877406416253e-07,
+      "loss": 0.0099,
+      "step": 14534
+    },
+    {
+      "epoch": 3.3034090909090907,
+      "grad_norm": 0.3857990290705267,
+      "learning_rate": 3.228096134869603e-07,
+      "loss": 0.0016,
+      "step": 14535
+    },
+    {
+      "epoch": 3.3036363636363637,
+      "grad_norm": 1.5286578263766541,
+      "learning_rate": 3.227314924944849e-07,
+      "loss": 0.0093,
+      "step": 14536
+    },
+    {
+      "epoch": 3.3038636363636362,
+      "grad_norm": 1.9257966434205158,
+      "learning_rate": 3.226533776657916e-07,
+      "loss": 0.0135,
+      "step": 14537
+    },
+    {
+      "epoch": 3.304090909090909,
+      "grad_norm": 1.7709182942561312,
+      "learning_rate": 3.2257526900247335e-07,
+      "loss": 0.0094,
+      "step": 14538
+    },
+    {
+      "epoch": 3.3043181818181817,
+      "grad_norm": 0.9841948788235706,
+      "learning_rate": 3.2249716650612286e-07,
+      "loss": 0.0049,
+      "step": 14539
+    },
+    {
+      "epoch": 3.3045454545454547,
+      "grad_norm": 1.8210028423326519,
+      "learning_rate": 3.2241907017833327e-07,
+      "loss": 0.022,
+      "step": 14540
+    },
+    {
+      "epoch": 3.304772727272727,
+      "grad_norm": 5.696989917694332,
+      "learning_rate": 3.223409800206964e-07,
+      "loss": 0.0148,
+      "step": 14541
+    },
+    {
+      "epoch": 3.305,
+      "grad_norm": 1.2476096100512526,
+      "learning_rate": 3.22262896034805e-07,
+      "loss": 0.0036,
+      "step": 14542
+    },
+    {
+      "epoch": 3.3052272727272727,
+      "grad_norm": 1.1098302336831614,
+      "learning_rate": 3.2218481822225133e-07,
+      "loss": 0.0066,
+      "step": 14543
+    },
+    {
+      "epoch": 3.3054545454545456,
+      "grad_norm": 1.6601668873455147,
+      "learning_rate": 3.2210674658462766e-07,
+      "loss": 0.0088,
+      "step": 14544
+    },
+    {
+      "epoch": 3.305681818181818,
+      "grad_norm": 1.0191873686365924,
+      "learning_rate": 3.220286811235257e-07,
+      "loss": 0.0045,
+      "step": 14545
+    },
+    {
+      "epoch": 3.3059090909090907,
+      "grad_norm": 1.5308351852838602,
+      "learning_rate": 3.2195062184053745e-07,
+      "loss": 0.008,
+      "step": 14546
+    },
+    {
+      "epoch": 3.3061363636363637,
+      "grad_norm": 1.098937648531083,
+      "learning_rate": 3.2187256873725473e-07,
+      "loss": 0.0008,
+      "step": 14547
+    },
+    {
+      "epoch": 3.306363636363636,
+      "grad_norm": 1.0290179342350767,
+      "learning_rate": 3.2179452181526917e-07,
+      "loss": 0.0067,
+      "step": 14548
+    },
+    {
+      "epoch": 3.306590909090909,
+      "grad_norm": 2.6696767107936896,
+      "learning_rate": 3.217164810761723e-07,
+      "loss": 0.0204,
+      "step": 14549
+    },
+    {
+      "epoch": 3.3068181818181817,
+      "grad_norm": 2.5200026211954425,
+      "learning_rate": 3.2163844652155544e-07,
+      "loss": 0.0116,
+      "step": 14550
+    },
+    {
+      "epoch": 3.3070454545454546,
+      "grad_norm": 1.0301368126582726,
+      "learning_rate": 3.2156041815300995e-07,
+      "loss": 0.0067,
+      "step": 14551
+    },
+    {
+      "epoch": 3.307272727272727,
+      "grad_norm": 0.5086716921635019,
+      "learning_rate": 3.2148239597212685e-07,
+      "loss": 0.0036,
+      "step": 14552
+    },
+    {
+      "epoch": 3.3075,
+      "grad_norm": 0.702352453872063,
+      "learning_rate": 3.214043799804975e-07,
+      "loss": 0.0043,
+      "step": 14553
+    },
+    {
+      "epoch": 3.3077272727272726,
+      "grad_norm": 2.4676521843883457,
+      "learning_rate": 3.213263701797122e-07,
+      "loss": 0.0175,
+      "step": 14554
+    },
+    {
+      "epoch": 3.3079545454545456,
+      "grad_norm": 1.8614298804705267,
+      "learning_rate": 3.2124836657136205e-07,
+      "loss": 0.0146,
+      "step": 14555
+    },
+    {
+      "epoch": 3.308181818181818,
+      "grad_norm": 3.0215810799725897,
+      "learning_rate": 3.2117036915703766e-07,
+      "loss": 0.0201,
+      "step": 14556
+    },
+    {
+      "epoch": 3.308409090909091,
+      "grad_norm": 2.084045995633674,
+      "learning_rate": 3.2109237793832946e-07,
+      "loss": 0.0203,
+      "step": 14557
+    },
+    {
+      "epoch": 3.3086363636363636,
+      "grad_norm": 1.6486896732198135,
+      "learning_rate": 3.2101439291682813e-07,
+      "loss": 0.0099,
+      "step": 14558
+    },
+    {
+      "epoch": 3.308863636363636,
+      "grad_norm": 2.1663483191394697,
+      "learning_rate": 3.209364140941234e-07,
+      "loss": 0.0222,
+      "step": 14559
+    },
+    {
+      "epoch": 3.309090909090909,
+      "grad_norm": 0.5276442868254555,
+      "learning_rate": 3.2085844147180565e-07,
+      "loss": 0.0074,
+      "step": 14560
+    },
+    {
+      "epoch": 3.309318181818182,
+      "grad_norm": 1.4447987440750838,
+      "learning_rate": 3.207804750514649e-07,
+      "loss": 0.0041,
+      "step": 14561
+    },
+    {
+      "epoch": 3.3095454545454546,
+      "grad_norm": 0.3819530278841556,
+      "learning_rate": 3.207025148346912e-07,
+      "loss": 0.001,
+      "step": 14562
+    },
+    {
+      "epoch": 3.309772727272727,
+      "grad_norm": 0.9869444899184336,
+      "learning_rate": 3.2062456082307387e-07,
+      "loss": 0.0053,
+      "step": 14563
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 2.265652765608752,
+      "learning_rate": 3.205466130182028e-07,
+      "loss": 0.016,
+      "step": 14564
+    },
+    {
+      "epoch": 3.3102272727272726,
+      "grad_norm": 2.879312726060716,
+      "learning_rate": 3.204686714216672e-07,
+      "loss": 0.0128,
+      "step": 14565
+    },
+    {
+      "epoch": 3.3104545454545455,
+      "grad_norm": 2.191792071846753,
+      "learning_rate": 3.2039073603505715e-07,
+      "loss": 0.0203,
+      "step": 14566
+    },
+    {
+      "epoch": 3.310681818181818,
+      "grad_norm": 1.192154796184981,
+      "learning_rate": 3.203128068599612e-07,
+      "loss": 0.0053,
+      "step": 14567
+    },
+    {
+      "epoch": 3.310909090909091,
+      "grad_norm": 2.80480101834089,
+      "learning_rate": 3.202348838979685e-07,
+      "loss": 0.0125,
+      "step": 14568
+    },
+    {
+      "epoch": 3.3111363636363635,
+      "grad_norm": 1.4497156848573483,
+      "learning_rate": 3.2015696715066835e-07,
+      "loss": 0.0188,
+      "step": 14569
+    },
+    {
+      "epoch": 3.3113636363636365,
+      "grad_norm": 1.2280792521892312,
+      "learning_rate": 3.2007905661964945e-07,
+      "loss": 0.0115,
+      "step": 14570
+    },
+    {
+      "epoch": 3.311590909090909,
+      "grad_norm": 1.1384524033594372,
+      "learning_rate": 3.200011523065009e-07,
+      "loss": 0.0114,
+      "step": 14571
+    },
+    {
+      "epoch": 3.311818181818182,
+      "grad_norm": 1.282524400076846,
+      "learning_rate": 3.199232542128106e-07,
+      "loss": 0.0117,
+      "step": 14572
+    },
+    {
+      "epoch": 3.3120454545454545,
+      "grad_norm": 0.9838912128358415,
+      "learning_rate": 3.198453623401673e-07,
+      "loss": 0.0094,
+      "step": 14573
+    },
+    {
+      "epoch": 3.3122727272727275,
+      "grad_norm": 1.3576989154233485,
+      "learning_rate": 3.1976747669015955e-07,
+      "loss": 0.012,
+      "step": 14574
+    },
+    {
+      "epoch": 3.3125,
+      "grad_norm": 1.2436458715305907,
+      "learning_rate": 3.196895972643757e-07,
+      "loss": 0.0109,
+      "step": 14575
+    },
+    {
+      "epoch": 3.3127272727272725,
+      "grad_norm": 1.6513353063273508,
+      "learning_rate": 3.196117240644034e-07,
+      "loss": 0.004,
+      "step": 14576
+    },
+    {
+      "epoch": 3.3129545454545455,
+      "grad_norm": 0.9490630623641084,
+      "learning_rate": 3.195338570918307e-07,
+      "loss": 0.0079,
+      "step": 14577
+    },
+    {
+      "epoch": 3.313181818181818,
+      "grad_norm": 0.9277927093632466,
+      "learning_rate": 3.1945599634824564e-07,
+      "loss": 0.005,
+      "step": 14578
+    },
+    {
+      "epoch": 3.313409090909091,
+      "grad_norm": 0.8029799824876325,
+      "learning_rate": 3.193781418352358e-07,
+      "loss": 0.0064,
+      "step": 14579
+    },
+    {
+      "epoch": 3.3136363636363635,
+      "grad_norm": 1.6062847772805846,
+      "learning_rate": 3.1930029355438895e-07,
+      "loss": 0.0058,
+      "step": 14580
+    },
+    {
+      "epoch": 3.3138636363636365,
+      "grad_norm": 1.4532460676073904,
+      "learning_rate": 3.192224515072924e-07,
+      "loss": 0.0079,
+      "step": 14581
+    },
+    {
+      "epoch": 3.314090909090909,
+      "grad_norm": 1.0505678581223152,
+      "learning_rate": 3.191446156955335e-07,
+      "loss": 0.0027,
+      "step": 14582
+    },
+    {
+      "epoch": 3.314318181818182,
+      "grad_norm": 0.761291699003442,
+      "learning_rate": 3.190667861206995e-07,
+      "loss": 0.0039,
+      "step": 14583
+    },
+    {
+      "epoch": 3.3145454545454545,
+      "grad_norm": 1.7645418347866078,
+      "learning_rate": 3.189889627843776e-07,
+      "loss": 0.0072,
+      "step": 14584
+    },
+    {
+      "epoch": 3.3147727272727274,
+      "grad_norm": 5.390257075509869,
+      "learning_rate": 3.189111456881545e-07,
+      "loss": 0.0294,
+      "step": 14585
+    },
+    {
+      "epoch": 3.315,
+      "grad_norm": 0.934599532516431,
+      "learning_rate": 3.1883333483361714e-07,
+      "loss": 0.0051,
+      "step": 14586
+    },
+    {
+      "epoch": 3.315227272727273,
+      "grad_norm": 1.7581629520073385,
+      "learning_rate": 3.187555302223522e-07,
+      "loss": 0.0076,
+      "step": 14587
+    },
+    {
+      "epoch": 3.3154545454545454,
+      "grad_norm": 2.117974543938185,
+      "learning_rate": 3.186777318559464e-07,
+      "loss": 0.0082,
+      "step": 14588
+    },
+    {
+      "epoch": 3.315681818181818,
+      "grad_norm": 0.3913267981599301,
+      "learning_rate": 3.1859993973598597e-07,
+      "loss": 0.002,
+      "step": 14589
+    },
+    {
+      "epoch": 3.315909090909091,
+      "grad_norm": 1.3916037943076764,
+      "learning_rate": 3.185221538640572e-07,
+      "loss": 0.0076,
+      "step": 14590
+    },
+    {
+      "epoch": 3.316136363636364,
+      "grad_norm": 3.853401824639004,
+      "learning_rate": 3.1844437424174645e-07,
+      "loss": 0.0329,
+      "step": 14591
+    },
+    {
+      "epoch": 3.3163636363636364,
+      "grad_norm": 1.5607681045237334,
+      "learning_rate": 3.183666008706397e-07,
+      "loss": 0.0054,
+      "step": 14592
+    },
+    {
+      "epoch": 3.316590909090909,
+      "grad_norm": 1.3465840889364382,
+      "learning_rate": 3.182888337523232e-07,
+      "loss": 0.0053,
+      "step": 14593
+    },
+    {
+      "epoch": 3.316818181818182,
+      "grad_norm": 2.257685045934358,
+      "learning_rate": 3.1821107288838193e-07,
+      "loss": 0.0127,
+      "step": 14594
+    },
+    {
+      "epoch": 3.3170454545454544,
+      "grad_norm": 1.7728698563370129,
+      "learning_rate": 3.1813331828040234e-07,
+      "loss": 0.0158,
+      "step": 14595
+    },
+    {
+      "epoch": 3.3172727272727274,
+      "grad_norm": 0.8834379549367293,
+      "learning_rate": 3.1805556992996975e-07,
+      "loss": 0.0027,
+      "step": 14596
+    },
+    {
+      "epoch": 3.3175,
+      "grad_norm": 0.6713611313949659,
+      "learning_rate": 3.179778278386698e-07,
+      "loss": 0.0055,
+      "step": 14597
+    },
+    {
+      "epoch": 3.317727272727273,
+      "grad_norm": 0.3307954009012571,
+      "learning_rate": 3.179000920080875e-07,
+      "loss": 0.0064,
+      "step": 14598
+    },
+    {
+      "epoch": 3.3179545454545454,
+      "grad_norm": 1.9257652969727481,
+      "learning_rate": 3.1782236243980795e-07,
+      "loss": 0.0102,
+      "step": 14599
+    },
+    {
+      "epoch": 3.3181818181818183,
+      "grad_norm": 3.1272946985527166,
+      "learning_rate": 3.1774463913541635e-07,
+      "loss": 0.0082,
+      "step": 14600
+    },
+    {
+      "epoch": 3.318409090909091,
+      "grad_norm": 2.7986370460630567,
+      "learning_rate": 3.1766692209649756e-07,
+      "loss": 0.0148,
+      "step": 14601
+    },
+    {
+      "epoch": 3.318636363636364,
+      "grad_norm": 1.353169884963484,
+      "learning_rate": 3.175892113246367e-07,
+      "loss": 0.0073,
+      "step": 14602
+    },
+    {
+      "epoch": 3.3188636363636363,
+      "grad_norm": 3.8749373176500694,
+      "learning_rate": 3.175115068214178e-07,
+      "loss": 0.022,
+      "step": 14603
+    },
+    {
+      "epoch": 3.3190909090909093,
+      "grad_norm": 4.03752200720067,
+      "learning_rate": 3.1743380858842585e-07,
+      "loss": 0.0241,
+      "step": 14604
+    },
+    {
+      "epoch": 3.319318181818182,
+      "grad_norm": 1.8184016338729183,
+      "learning_rate": 3.17356116627245e-07,
+      "loss": 0.0145,
+      "step": 14605
+    },
+    {
+      "epoch": 3.3195454545454544,
+      "grad_norm": 1.4264273966552143,
+      "learning_rate": 3.1727843093946e-07,
+      "loss": 0.0079,
+      "step": 14606
+    },
+    {
+      "epoch": 3.3197727272727273,
+      "grad_norm": 2.0458054791337332,
+      "learning_rate": 3.172007515266543e-07,
+      "loss": 0.0222,
+      "step": 14607
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 1.8433376132883204,
+      "learning_rate": 3.171230783904124e-07,
+      "loss": 0.0182,
+      "step": 14608
+    },
+    {
+      "epoch": 3.320227272727273,
+      "grad_norm": 1.5353260059755462,
+      "learning_rate": 3.170454115323179e-07,
+      "loss": 0.0038,
+      "step": 14609
+    },
+    {
+      "epoch": 3.3204545454545453,
+      "grad_norm": 1.917588564907027,
+      "learning_rate": 3.1696775095395477e-07,
+      "loss": 0.0101,
+      "step": 14610
+    },
+    {
+      "epoch": 3.3206818181818183,
+      "grad_norm": 1.978388899194453,
+      "learning_rate": 3.168900966569066e-07,
+      "loss": 0.0149,
+      "step": 14611
+    },
+    {
+      "epoch": 3.320909090909091,
+      "grad_norm": 1.497664827890105,
+      "learning_rate": 3.1681244864275685e-07,
+      "loss": 0.0108,
+      "step": 14612
+    },
+    {
+      "epoch": 3.3211363636363638,
+      "grad_norm": 1.4469109303107373,
+      "learning_rate": 3.1673480691308894e-07,
+      "loss": 0.0079,
+      "step": 14613
+    },
+    {
+      "epoch": 3.3213636363636363,
+      "grad_norm": 1.7885141094304369,
+      "learning_rate": 3.166571714694862e-07,
+      "loss": 0.0069,
+      "step": 14614
+    },
+    {
+      "epoch": 3.3215909090909093,
+      "grad_norm": 1.4488637623316762,
+      "learning_rate": 3.1657954231353176e-07,
+      "loss": 0.0213,
+      "step": 14615
+    },
+    {
+      "epoch": 3.321818181818182,
+      "grad_norm": 1.8301458951495349,
+      "learning_rate": 3.1650191944680843e-07,
+      "loss": 0.0244,
+      "step": 14616
+    },
+    {
+      "epoch": 3.3220454545454547,
+      "grad_norm": 0.6394318437703834,
+      "learning_rate": 3.1642430287089907e-07,
+      "loss": 0.0039,
+      "step": 14617
+    },
+    {
+      "epoch": 3.3222727272727273,
+      "grad_norm": 1.4877429119014178,
+      "learning_rate": 3.1634669258738655e-07,
+      "loss": 0.0074,
+      "step": 14618
+    },
+    {
+      "epoch": 3.3225,
+      "grad_norm": 3.0187835838177706,
+      "learning_rate": 3.162690885978537e-07,
+      "loss": 0.0237,
+      "step": 14619
+    },
+    {
+      "epoch": 3.3227272727272728,
+      "grad_norm": 2.2893931199793465,
+      "learning_rate": 3.161914909038825e-07,
+      "loss": 0.0224,
+      "step": 14620
+    },
+    {
+      "epoch": 3.3229545454545453,
+      "grad_norm": 1.1345139006105558,
+      "learning_rate": 3.1611389950705547e-07,
+      "loss": 0.0061,
+      "step": 14621
+    },
+    {
+      "epoch": 3.3231818181818182,
+      "grad_norm": 1.4361827946862902,
+      "learning_rate": 3.1603631440895496e-07,
+      "loss": 0.0068,
+      "step": 14622
+    },
+    {
+      "epoch": 3.3234090909090908,
+      "grad_norm": 2.0035889800925197,
+      "learning_rate": 3.1595873561116307e-07,
+      "loss": 0.0206,
+      "step": 14623
+    },
+    {
+      "epoch": 3.3236363636363637,
+      "grad_norm": 0.21090172251685518,
+      "learning_rate": 3.158811631152617e-07,
+      "loss": 0.001,
+      "step": 14624
+    },
+    {
+      "epoch": 3.3238636363636362,
+      "grad_norm": 1.4010714341631323,
+      "learning_rate": 3.1580359692283263e-07,
+      "loss": 0.0114,
+      "step": 14625
+    },
+    {
+      "epoch": 3.324090909090909,
+      "grad_norm": 1.2862207621857762,
+      "learning_rate": 3.157260370354577e-07,
+      "loss": 0.008,
+      "step": 14626
+    },
+    {
+      "epoch": 3.3243181818181817,
+      "grad_norm": 0.7995930518276104,
+      "learning_rate": 3.156484834547184e-07,
+      "loss": 0.0091,
+      "step": 14627
+    },
+    {
+      "epoch": 3.3245454545454547,
+      "grad_norm": 1.5905659636298708,
+      "learning_rate": 3.1557093618219644e-07,
+      "loss": 0.0085,
+      "step": 14628
+    },
+    {
+      "epoch": 3.324772727272727,
+      "grad_norm": 1.4715824307102277,
+      "learning_rate": 3.1549339521947276e-07,
+      "loss": 0.0088,
+      "step": 14629
+    },
+    {
+      "epoch": 3.325,
+      "grad_norm": 2.1618316426878414,
+      "learning_rate": 3.154158605681287e-07,
+      "loss": 0.0079,
+      "step": 14630
+    },
+    {
+      "epoch": 3.3252272727272727,
+      "grad_norm": 1.6765658648357253,
+      "learning_rate": 3.153383322297453e-07,
+      "loss": 0.0103,
+      "step": 14631
+    },
+    {
+      "epoch": 3.325454545454545,
+      "grad_norm": 2.029162835061917,
+      "learning_rate": 3.1526081020590376e-07,
+      "loss": 0.0112,
+      "step": 14632
+    },
+    {
+      "epoch": 3.325681818181818,
+      "grad_norm": 2.266886806121161,
+      "learning_rate": 3.1518329449818447e-07,
+      "loss": 0.0114,
+      "step": 14633
+    },
+    {
+      "epoch": 3.325909090909091,
+      "grad_norm": 1.6393601885033842,
+      "learning_rate": 3.1510578510816826e-07,
+      "loss": 0.0103,
+      "step": 14634
+    },
+    {
+      "epoch": 3.3261363636363637,
+      "grad_norm": 3.044886201579731,
+      "learning_rate": 3.150282820374358e-07,
+      "loss": 0.0217,
+      "step": 14635
+    },
+    {
+      "epoch": 3.326363636363636,
+      "grad_norm": 1.489057632501786,
+      "learning_rate": 3.1495078528756746e-07,
+      "loss": 0.0171,
+      "step": 14636
+    },
+    {
+      "epoch": 3.326590909090909,
+      "grad_norm": 0.7255157000736588,
+      "learning_rate": 3.1487329486014364e-07,
+      "loss": 0.007,
+      "step": 14637
+    },
+    {
+      "epoch": 3.3268181818181817,
+      "grad_norm": 1.3082262735662564,
+      "learning_rate": 3.147958107567442e-07,
+      "loss": 0.0049,
+      "step": 14638
+    },
+    {
+      "epoch": 3.3270454545454546,
+      "grad_norm": 1.1571438136570233,
+      "learning_rate": 3.1471833297894937e-07,
+      "loss": 0.0038,
+      "step": 14639
+    },
+    {
+      "epoch": 3.327272727272727,
+      "grad_norm": 0.8652632191604277,
+      "learning_rate": 3.146408615283391e-07,
+      "loss": 0.0047,
+      "step": 14640
+    },
+    {
+      "epoch": 3.3275,
+      "grad_norm": 1.8493895416952497,
+      "learning_rate": 3.145633964064931e-07,
+      "loss": 0.0121,
+      "step": 14641
+    },
+    {
+      "epoch": 3.3277272727272726,
+      "grad_norm": 3.309406881842238,
+      "learning_rate": 3.14485937614991e-07,
+      "loss": 0.0224,
+      "step": 14642
+    },
+    {
+      "epoch": 3.3279545454545456,
+      "grad_norm": 3.0332097835980356,
+      "learning_rate": 3.144084851554123e-07,
+      "loss": 0.0202,
+      "step": 14643
+    },
+    {
+      "epoch": 3.328181818181818,
+      "grad_norm": 2.2618592397903234,
+      "learning_rate": 3.143310390293365e-07,
+      "loss": 0.0136,
+      "step": 14644
+    },
+    {
+      "epoch": 3.328409090909091,
+      "grad_norm": 5.862507943150675,
+      "learning_rate": 3.1425359923834303e-07,
+      "loss": 0.0262,
+      "step": 14645
+    },
+    {
+      "epoch": 3.3286363636363636,
+      "grad_norm": 1.3479618004175071,
+      "learning_rate": 3.141761657840106e-07,
+      "loss": 0.0072,
+      "step": 14646
+    },
+    {
+      "epoch": 3.3288636363636366,
+      "grad_norm": 1.9755106767546546,
+      "learning_rate": 3.140987386679184e-07,
+      "loss": 0.0131,
+      "step": 14647
+    },
+    {
+      "epoch": 3.329090909090909,
+      "grad_norm": 2.6786329929227586,
+      "learning_rate": 3.140213178916453e-07,
+      "loss": 0.0451,
+      "step": 14648
+    },
+    {
+      "epoch": 3.3293181818181816,
+      "grad_norm": 5.5318933377221216,
+      "learning_rate": 3.1394390345677e-07,
+      "loss": 0.0232,
+      "step": 14649
+    },
+    {
+      "epoch": 3.3295454545454546,
+      "grad_norm": 0.7204221767691753,
+      "learning_rate": 3.138664953648715e-07,
+      "loss": 0.0042,
+      "step": 14650
+    },
+    {
+      "epoch": 3.329772727272727,
+      "grad_norm": 1.069508398484204,
+      "learning_rate": 3.1378909361752773e-07,
+      "loss": 0.0146,
+      "step": 14651
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 2.2787121635954506,
+      "learning_rate": 3.1371169821631727e-07,
+      "loss": 0.0133,
+      "step": 14652
+    },
+    {
+      "epoch": 3.3302272727272726,
+      "grad_norm": 0.5967338414469777,
+      "learning_rate": 3.1363430916281826e-07,
+      "loss": 0.0036,
+      "step": 14653
+    },
+    {
+      "epoch": 3.3304545454545456,
+      "grad_norm": 1.2040032573926218,
+      "learning_rate": 3.1355692645860897e-07,
+      "loss": 0.0041,
+      "step": 14654
+    },
+    {
+      "epoch": 3.330681818181818,
+      "grad_norm": 2.2697751236553176,
+      "learning_rate": 3.1347955010526733e-07,
+      "loss": 0.0134,
+      "step": 14655
+    },
+    {
+      "epoch": 3.330909090909091,
+      "grad_norm": 0.5006013488406281,
+      "learning_rate": 3.134021801043711e-07,
+      "loss": 0.0035,
+      "step": 14656
+    },
+    {
+      "epoch": 3.3311363636363636,
+      "grad_norm": 2.409486289508824,
+      "learning_rate": 3.13324816457498e-07,
+      "loss": 0.0121,
+      "step": 14657
+    },
+    {
+      "epoch": 3.3313636363636365,
+      "grad_norm": 1.4791705187569135,
+      "learning_rate": 3.1324745916622574e-07,
+      "loss": 0.0104,
+      "step": 14658
+    },
+    {
+      "epoch": 3.331590909090909,
+      "grad_norm": 5.086154666021941,
+      "learning_rate": 3.1317010823213183e-07,
+      "loss": 0.0311,
+      "step": 14659
+    },
+    {
+      "epoch": 3.331818181818182,
+      "grad_norm": 1.8643019221120967,
+      "learning_rate": 3.130927636567932e-07,
+      "loss": 0.0082,
+      "step": 14660
+    },
+    {
+      "epoch": 3.3320454545454545,
+      "grad_norm": 1.3194825470274179,
+      "learning_rate": 3.130154254417872e-07,
+      "loss": 0.0092,
+      "step": 14661
+    },
+    {
+      "epoch": 3.332272727272727,
+      "grad_norm": 1.7805731465897563,
+      "learning_rate": 3.129380935886911e-07,
+      "loss": 0.0156,
+      "step": 14662
+    },
+    {
+      "epoch": 3.3325,
+      "grad_norm": 0.2974918130887638,
+      "learning_rate": 3.1286076809908183e-07,
+      "loss": 0.0016,
+      "step": 14663
+    },
+    {
+      "epoch": 3.3327272727272725,
+      "grad_norm": 0.6370035809148882,
+      "learning_rate": 3.1278344897453585e-07,
+      "loss": 0.002,
+      "step": 14664
+    },
+    {
+      "epoch": 3.3329545454545455,
+      "grad_norm": 1.939657719226372,
+      "learning_rate": 3.1270613621663006e-07,
+      "loss": 0.0206,
+      "step": 14665
+    },
+    {
+      "epoch": 3.333181818181818,
+      "grad_norm": 1.902334628215835,
+      "learning_rate": 3.1262882982694085e-07,
+      "loss": 0.0125,
+      "step": 14666
+    },
+    {
+      "epoch": 3.333409090909091,
+      "grad_norm": 1.789086069656708,
+      "learning_rate": 3.125515298070451e-07,
+      "loss": 0.0103,
+      "step": 14667
+    },
+    {
+      "epoch": 3.3336363636363635,
+      "grad_norm": 2.8250830412214216,
+      "learning_rate": 3.124742361585186e-07,
+      "loss": 0.0296,
+      "step": 14668
+    },
+    {
+      "epoch": 3.3338636363636365,
+      "grad_norm": 1.1052010408800275,
+      "learning_rate": 3.1239694888293757e-07,
+      "loss": 0.006,
+      "step": 14669
+    },
+    {
+      "epoch": 3.334090909090909,
+      "grad_norm": 2.1674997128095357,
+      "learning_rate": 3.123196679818781e-07,
+      "loss": 0.0154,
+      "step": 14670
+    },
+    {
+      "epoch": 3.334318181818182,
+      "grad_norm": 1.0495521992687808,
+      "learning_rate": 3.122423934569162e-07,
+      "loss": 0.0055,
+      "step": 14671
+    },
+    {
+      "epoch": 3.3345454545454545,
+      "grad_norm": 3.889438303003471,
+      "learning_rate": 3.121651253096274e-07,
+      "loss": 0.0211,
+      "step": 14672
+    },
+    {
+      "epoch": 3.3347727272727274,
+      "grad_norm": 1.9330484358963065,
+      "learning_rate": 3.1208786354158753e-07,
+      "loss": 0.0127,
+      "step": 14673
+    },
+    {
+      "epoch": 3.335,
+      "grad_norm": 1.611122870502531,
+      "learning_rate": 3.12010608154372e-07,
+      "loss": 0.0102,
+      "step": 14674
+    },
+    {
+      "epoch": 3.3352272727272725,
+      "grad_norm": 2.284121806795122,
+      "learning_rate": 3.1193335914955624e-07,
+      "loss": 0.0164,
+      "step": 14675
+    },
+    {
+      "epoch": 3.3354545454545454,
+      "grad_norm": 1.3382417362615613,
+      "learning_rate": 3.118561165287157e-07,
+      "loss": 0.0092,
+      "step": 14676
+    },
+    {
+      "epoch": 3.3356818181818184,
+      "grad_norm": 1.5006095106202235,
+      "learning_rate": 3.1177888029342504e-07,
+      "loss": 0.0172,
+      "step": 14677
+    },
+    {
+      "epoch": 3.335909090909091,
+      "grad_norm": 2.6207454736520313,
+      "learning_rate": 3.117016504452594e-07,
+      "loss": 0.0057,
+      "step": 14678
+    },
+    {
+      "epoch": 3.3361363636363635,
+      "grad_norm": 1.5067828985313367,
+      "learning_rate": 3.1162442698579367e-07,
+      "loss": 0.007,
+      "step": 14679
+    },
+    {
+      "epoch": 3.3363636363636364,
+      "grad_norm": 2.730991875476565,
+      "learning_rate": 3.1154720991660267e-07,
+      "loss": 0.0164,
+      "step": 14680
+    },
+    {
+      "epoch": 3.336590909090909,
+      "grad_norm": 3.064934842193607,
+      "learning_rate": 3.1146999923926104e-07,
+      "loss": 0.0072,
+      "step": 14681
+    },
+    {
+      "epoch": 3.336818181818182,
+      "grad_norm": 1.350909226222192,
+      "learning_rate": 3.1139279495534295e-07,
+      "loss": 0.0087,
+      "step": 14682
+    },
+    {
+      "epoch": 3.3370454545454544,
+      "grad_norm": 1.9181460713592016,
+      "learning_rate": 3.113155970664229e-07,
+      "loss": 0.0078,
+      "step": 14683
+    },
+    {
+      "epoch": 3.3372727272727274,
+      "grad_norm": 1.964606126540366,
+      "learning_rate": 3.1123840557407506e-07,
+      "loss": 0.0108,
+      "step": 14684
+    },
+    {
+      "epoch": 3.3375,
+      "grad_norm": 1.5907791964556877,
+      "learning_rate": 3.111612204798735e-07,
+      "loss": 0.0126,
+      "step": 14685
+    },
+    {
+      "epoch": 3.337727272727273,
+      "grad_norm": 0.5457085113245932,
+      "learning_rate": 3.110840417853922e-07,
+      "loss": 0.0029,
+      "step": 14686
+    },
+    {
+      "epoch": 3.3379545454545454,
+      "grad_norm": 2.475320609990316,
+      "learning_rate": 3.110068694922049e-07,
+      "loss": 0.0069,
+      "step": 14687
+    },
+    {
+      "epoch": 3.3381818181818184,
+      "grad_norm": 0.7925091751221381,
+      "learning_rate": 3.1092970360188533e-07,
+      "loss": 0.0052,
+      "step": 14688
+    },
+    {
+      "epoch": 3.338409090909091,
+      "grad_norm": 2.973320793481258,
+      "learning_rate": 3.108525441160073e-07,
+      "loss": 0.0079,
+      "step": 14689
+    },
+    {
+      "epoch": 3.338636363636364,
+      "grad_norm": 1.3870113912543411,
+      "learning_rate": 3.1077539103614373e-07,
+      "loss": 0.0087,
+      "step": 14690
+    },
+    {
+      "epoch": 3.3388636363636364,
+      "grad_norm": 1.8581392202249523,
+      "learning_rate": 3.106982443638681e-07,
+      "loss": 0.0086,
+      "step": 14691
+    },
+    {
+      "epoch": 3.339090909090909,
+      "grad_norm": 2.8722662234142318,
+      "learning_rate": 3.106211041007536e-07,
+      "loss": 0.0438,
+      "step": 14692
+    },
+    {
+      "epoch": 3.339318181818182,
+      "grad_norm": 2.231780742699464,
+      "learning_rate": 3.105439702483733e-07,
+      "loss": 0.0161,
+      "step": 14693
+    },
+    {
+      "epoch": 3.3395454545454544,
+      "grad_norm": 0.40466727953292814,
+      "learning_rate": 3.104668428083002e-07,
+      "loss": 0.0025,
+      "step": 14694
+    },
+    {
+      "epoch": 3.3397727272727273,
+      "grad_norm": 1.298077278663543,
+      "learning_rate": 3.103897217821067e-07,
+      "loss": 0.0079,
+      "step": 14695
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 3.187659305590303,
+      "learning_rate": 3.103126071713657e-07,
+      "loss": 0.0194,
+      "step": 14696
+    },
+    {
+      "epoch": 3.340227272727273,
+      "grad_norm": 1.261245911428879,
+      "learning_rate": 3.1023549897764954e-07,
+      "loss": 0.0064,
+      "step": 14697
+    },
+    {
+      "epoch": 3.3404545454545453,
+      "grad_norm": 1.1156902262215314,
+      "learning_rate": 3.10158397202531e-07,
+      "loss": 0.0061,
+      "step": 14698
+    },
+    {
+      "epoch": 3.3406818181818183,
+      "grad_norm": 0.8746795119909072,
+      "learning_rate": 3.100813018475817e-07,
+      "loss": 0.0054,
+      "step": 14699
+    },
+    {
+      "epoch": 3.340909090909091,
+      "grad_norm": 2.309520497545161,
+      "learning_rate": 3.100042129143741e-07,
+      "loss": 0.0174,
+      "step": 14700
+    },
+    {
+      "epoch": 3.341136363636364,
+      "grad_norm": 1.344589015511832,
+      "learning_rate": 3.099271304044802e-07,
+      "loss": 0.0069,
+      "step": 14701
+    },
+    {
+      "epoch": 3.3413636363636363,
+      "grad_norm": 0.5873773490589581,
+      "learning_rate": 3.0985005431947173e-07,
+      "loss": 0.0038,
+      "step": 14702
+    },
+    {
+      "epoch": 3.3415909090909093,
+      "grad_norm": 1.9217952173765827,
+      "learning_rate": 3.097729846609205e-07,
+      "loss": 0.01,
+      "step": 14703
+    },
+    {
+      "epoch": 3.341818181818182,
+      "grad_norm": 1.6943093783874907,
+      "learning_rate": 3.0969592143039803e-07,
+      "loss": 0.0132,
+      "step": 14704
+    },
+    {
+      "epoch": 3.3420454545454543,
+      "grad_norm": 1.0071810723022634,
+      "learning_rate": 3.096188646294758e-07,
+      "loss": 0.0056,
+      "step": 14705
+    },
+    {
+      "epoch": 3.3422727272727273,
+      "grad_norm": 1.507135560465122,
+      "learning_rate": 3.0954181425972515e-07,
+      "loss": 0.0096,
+      "step": 14706
+    },
+    {
+      "epoch": 3.3425000000000002,
+      "grad_norm": 1.1943729060670751,
+      "learning_rate": 3.0946477032271745e-07,
+      "loss": 0.0117,
+      "step": 14707
+    },
+    {
+      "epoch": 3.3427272727272728,
+      "grad_norm": 1.0128334214457282,
+      "learning_rate": 3.0938773282002337e-07,
+      "loss": 0.003,
+      "step": 14708
+    },
+    {
+      "epoch": 3.3429545454545453,
+      "grad_norm": 1.7880974868822042,
+      "learning_rate": 3.0931070175321395e-07,
+      "loss": 0.0129,
+      "step": 14709
+    },
+    {
+      "epoch": 3.3431818181818183,
+      "grad_norm": 0.7616760245171101,
+      "learning_rate": 3.0923367712386017e-07,
+      "loss": 0.0028,
+      "step": 14710
+    },
+    {
+      "epoch": 3.3434090909090908,
+      "grad_norm": 2.325721510032304,
+      "learning_rate": 3.091566589335328e-07,
+      "loss": 0.0139,
+      "step": 14711
+    },
+    {
+      "epoch": 3.3436363636363637,
+      "grad_norm": 1.6140309375402768,
+      "learning_rate": 3.090796471838019e-07,
+      "loss": 0.0113,
+      "step": 14712
+    },
+    {
+      "epoch": 3.3438636363636363,
+      "grad_norm": 1.1216436940703465,
+      "learning_rate": 3.090026418762382e-07,
+      "loss": 0.0081,
+      "step": 14713
+    },
+    {
+      "epoch": 3.3440909090909092,
+      "grad_norm": 1.5126447611901042,
+      "learning_rate": 3.089256430124119e-07,
+      "loss": 0.007,
+      "step": 14714
+    },
+    {
+      "epoch": 3.3443181818181817,
+      "grad_norm": 3.0730491958929282,
+      "learning_rate": 3.088486505938932e-07,
+      "loss": 0.0128,
+      "step": 14715
+    },
+    {
+      "epoch": 3.3445454545454547,
+      "grad_norm": 1.7576711586527816,
+      "learning_rate": 3.0877166462225203e-07,
+      "loss": 0.0139,
+      "step": 14716
+    },
+    {
+      "epoch": 3.3447727272727272,
+      "grad_norm": 3.5670439339867204,
+      "learning_rate": 3.0869468509905833e-07,
+      "loss": 0.0157,
+      "step": 14717
+    },
+    {
+      "epoch": 3.3449999999999998,
+      "grad_norm": 1.2916439051936273,
+      "learning_rate": 3.086177120258818e-07,
+      "loss": 0.0113,
+      "step": 14718
+    },
+    {
+      "epoch": 3.3452272727272727,
+      "grad_norm": 1.088619845833293,
+      "learning_rate": 3.0854074540429204e-07,
+      "loss": 0.0076,
+      "step": 14719
+    },
+    {
+      "epoch": 3.3454545454545457,
+      "grad_norm": 0.6807105787974268,
+      "learning_rate": 3.0846378523585893e-07,
+      "loss": 0.0043,
+      "step": 14720
+    },
+    {
+      "epoch": 3.345681818181818,
+      "grad_norm": 2.623932372104795,
+      "learning_rate": 3.0838683152215115e-07,
+      "loss": 0.0132,
+      "step": 14721
+    },
+    {
+      "epoch": 3.3459090909090907,
+      "grad_norm": 2.9763810130030963,
+      "learning_rate": 3.0830988426473823e-07,
+      "loss": 0.0107,
+      "step": 14722
+    },
+    {
+      "epoch": 3.3461363636363637,
+      "grad_norm": 0.9961601540815102,
+      "learning_rate": 3.0823294346518926e-07,
+      "loss": 0.0061,
+      "step": 14723
+    },
+    {
+      "epoch": 3.346363636363636,
+      "grad_norm": 1.5147887129723263,
+      "learning_rate": 3.0815600912507326e-07,
+      "loss": 0.0088,
+      "step": 14724
+    },
+    {
+      "epoch": 3.346590909090909,
+      "grad_norm": 1.15319735276517,
+      "learning_rate": 3.080790812459592e-07,
+      "loss": 0.0029,
+      "step": 14725
+    },
+    {
+      "epoch": 3.3468181818181817,
+      "grad_norm": 2.0529767555811156,
+      "learning_rate": 3.0800215982941535e-07,
+      "loss": 0.0145,
+      "step": 14726
+    },
+    {
+      "epoch": 3.3470454545454547,
+      "grad_norm": 1.3883603139783018,
+      "learning_rate": 3.079252448770105e-07,
+      "loss": 0.0051,
+      "step": 14727
+    },
+    {
+      "epoch": 3.347272727272727,
+      "grad_norm": 1.3381737705580143,
+      "learning_rate": 3.078483363903131e-07,
+      "loss": 0.0084,
+      "step": 14728
+    },
+    {
+      "epoch": 3.3475,
+      "grad_norm": 1.176231902892993,
+      "learning_rate": 3.077714343708916e-07,
+      "loss": 0.0056,
+      "step": 14729
+    },
+    {
+      "epoch": 3.3477272727272727,
+      "grad_norm": 0.6007369789509198,
+      "learning_rate": 3.0769453882031374e-07,
+      "loss": 0.0033,
+      "step": 14730
+    },
+    {
+      "epoch": 3.3479545454545456,
+      "grad_norm": 5.225420120447926,
+      "learning_rate": 3.0761764974014786e-07,
+      "loss": 0.0278,
+      "step": 14731
+    },
+    {
+      "epoch": 3.348181818181818,
+      "grad_norm": 0.7164373696482718,
+      "learning_rate": 3.075407671319615e-07,
+      "loss": 0.0041,
+      "step": 14732
+    },
+    {
+      "epoch": 3.348409090909091,
+      "grad_norm": 1.1425017507902542,
+      "learning_rate": 3.074638909973233e-07,
+      "loss": 0.0073,
+      "step": 14733
+    },
+    {
+      "epoch": 3.3486363636363636,
+      "grad_norm": 2.9094548213496934,
+      "learning_rate": 3.073870213378e-07,
+      "loss": 0.0411,
+      "step": 14734
+    },
+    {
+      "epoch": 3.348863636363636,
+      "grad_norm": 1.206820815706784,
+      "learning_rate": 3.073101581549595e-07,
+      "loss": 0.007,
+      "step": 14735
+    },
+    {
+      "epoch": 3.349090909090909,
+      "grad_norm": 0.7075823798615369,
+      "learning_rate": 3.072333014503691e-07,
+      "loss": 0.0029,
+      "step": 14736
+    },
+    {
+      "epoch": 3.3493181818181816,
+      "grad_norm": 0.7350430327100151,
+      "learning_rate": 3.0715645122559605e-07,
+      "loss": 0.0044,
+      "step": 14737
+    },
+    {
+      "epoch": 3.3495454545454546,
+      "grad_norm": 2.4418871828757354,
+      "learning_rate": 3.0707960748220773e-07,
+      "loss": 0.021,
+      "step": 14738
+    },
+    {
+      "epoch": 3.349772727272727,
+      "grad_norm": 1.42005242451391,
+      "learning_rate": 3.070027702217706e-07,
+      "loss": 0.0095,
+      "step": 14739
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 1.8542947022934897,
+      "learning_rate": 3.0692593944585183e-07,
+      "loss": 0.0135,
+      "step": 14740
+    },
+    {
+      "epoch": 3.3502272727272726,
+      "grad_norm": 2.0128627990330283,
+      "learning_rate": 3.0684911515601795e-07,
+      "loss": 0.0103,
+      "step": 14741
+    },
+    {
+      "epoch": 3.3504545454545456,
+      "grad_norm": 0.9570684229194388,
+      "learning_rate": 3.0677229735383595e-07,
+      "loss": 0.0075,
+      "step": 14742
+    },
+    {
+      "epoch": 3.350681818181818,
+      "grad_norm": 1.2296003991854476,
+      "learning_rate": 3.0669548604087177e-07,
+      "loss": 0.0061,
+      "step": 14743
+    },
+    {
+      "epoch": 3.350909090909091,
+      "grad_norm": 1.6095619705640916,
+      "learning_rate": 3.0661868121869195e-07,
+      "loss": 0.0133,
+      "step": 14744
+    },
+    {
+      "epoch": 3.3511363636363636,
+      "grad_norm": 2.182761051061193,
+      "learning_rate": 3.0654188288886255e-07,
+      "loss": 0.0242,
+      "step": 14745
+    },
+    {
+      "epoch": 3.3513636363636365,
+      "grad_norm": 1.8849466560843242,
+      "learning_rate": 3.064650910529499e-07,
+      "loss": 0.0067,
+      "step": 14746
+    },
+    {
+      "epoch": 3.351590909090909,
+      "grad_norm": 0.5458905136532093,
+      "learning_rate": 3.0638830571251964e-07,
+      "loss": 0.0048,
+      "step": 14747
+    },
+    {
+      "epoch": 3.3518181818181816,
+      "grad_norm": 1.0557881036459325,
+      "learning_rate": 3.063115268691377e-07,
+      "loss": 0.0051,
+      "step": 14748
+    },
+    {
+      "epoch": 3.3520454545454546,
+      "grad_norm": 2.498407313739101,
+      "learning_rate": 3.0623475452436977e-07,
+      "loss": 0.011,
+      "step": 14749
+    },
+    {
+      "epoch": 3.3522727272727275,
+      "grad_norm": 1.2566030106215644,
+      "learning_rate": 3.061579886797812e-07,
+      "loss": 0.0057,
+      "step": 14750
+    },
+    {
+      "epoch": 3.3525,
+      "grad_norm": 1.5556407417392684,
+      "learning_rate": 3.0608122933693786e-07,
+      "loss": 0.0061,
+      "step": 14751
+    },
+    {
+      "epoch": 3.3527272727272726,
+      "grad_norm": 1.9339825542135305,
+      "learning_rate": 3.060044764974043e-07,
+      "loss": 0.0206,
+      "step": 14752
+    },
+    {
+      "epoch": 3.3529545454545455,
+      "grad_norm": 1.7191190461828982,
+      "learning_rate": 3.0592773016274606e-07,
+      "loss": 0.0131,
+      "step": 14753
+    },
+    {
+      "epoch": 3.353181818181818,
+      "grad_norm": 0.689225719919005,
+      "learning_rate": 3.0585099033452803e-07,
+      "loss": 0.0021,
+      "step": 14754
+    },
+    {
+      "epoch": 3.353409090909091,
+      "grad_norm": 3.5145145615078968,
+      "learning_rate": 3.0577425701431533e-07,
+      "loss": 0.0159,
+      "step": 14755
+    },
+    {
+      "epoch": 3.3536363636363635,
+      "grad_norm": 0.8610695657453881,
+      "learning_rate": 3.056975302036722e-07,
+      "loss": 0.0037,
+      "step": 14756
+    },
+    {
+      "epoch": 3.3538636363636365,
+      "grad_norm": 1.8421528615054747,
+      "learning_rate": 3.056208099041635e-07,
+      "loss": 0.0164,
+      "step": 14757
+    },
+    {
+      "epoch": 3.354090909090909,
+      "grad_norm": 2.359881178275831,
+      "learning_rate": 3.0554409611735356e-07,
+      "loss": 0.0133,
+      "step": 14758
+    },
+    {
+      "epoch": 3.354318181818182,
+      "grad_norm": 2.150915285952851,
+      "learning_rate": 3.0546738884480694e-07,
+      "loss": 0.0111,
+      "step": 14759
+    },
+    {
+      "epoch": 3.3545454545454545,
+      "grad_norm": 0.9152931753919175,
+      "learning_rate": 3.0539068808808795e-07,
+      "loss": 0.0033,
+      "step": 14760
+    },
+    {
+      "epoch": 3.3547727272727275,
+      "grad_norm": 2.5689215592799366,
+      "learning_rate": 3.0531399384876016e-07,
+      "loss": 0.0083,
+      "step": 14761
+    },
+    {
+      "epoch": 3.355,
+      "grad_norm": 0.27125609583069765,
+      "learning_rate": 3.0523730612838756e-07,
+      "loss": 0.0014,
+      "step": 14762
+    },
+    {
+      "epoch": 3.355227272727273,
+      "grad_norm": 2.298019885785823,
+      "learning_rate": 3.051606249285343e-07,
+      "loss": 0.0151,
+      "step": 14763
+    },
+    {
+      "epoch": 3.3554545454545455,
+      "grad_norm": 2.653052746048883,
+      "learning_rate": 3.0508395025076417e-07,
+      "loss": 0.0081,
+      "step": 14764
+    },
+    {
+      "epoch": 3.355681818181818,
+      "grad_norm": 1.453584882756855,
+      "learning_rate": 3.0500728209664015e-07,
+      "loss": 0.0125,
+      "step": 14765
+    },
+    {
+      "epoch": 3.355909090909091,
+      "grad_norm": 1.4219533794510435,
+      "learning_rate": 3.0493062046772597e-07,
+      "loss": 0.0051,
+      "step": 14766
+    },
+    {
+      "epoch": 3.3561363636363635,
+      "grad_norm": 3.4944275403230054,
+      "learning_rate": 3.048539653655848e-07,
+      "loss": 0.0513,
+      "step": 14767
+    },
+    {
+      "epoch": 3.3563636363636364,
+      "grad_norm": 1.4198954182267733,
+      "learning_rate": 3.0477731679178005e-07,
+      "loss": 0.0105,
+      "step": 14768
+    },
+    {
+      "epoch": 3.356590909090909,
+      "grad_norm": 1.457452757653818,
+      "learning_rate": 3.047006747478742e-07,
+      "loss": 0.0124,
+      "step": 14769
+    },
+    {
+      "epoch": 3.356818181818182,
+      "grad_norm": 2.944531284641324,
+      "learning_rate": 3.046240392354304e-07,
+      "loss": 0.0554,
+      "step": 14770
+    },
+    {
+      "epoch": 3.3570454545454544,
+      "grad_norm": 0.7330603107792545,
+      "learning_rate": 3.0454741025601135e-07,
+      "loss": 0.0038,
+      "step": 14771
+    },
+    {
+      "epoch": 3.3572727272727274,
+      "grad_norm": 1.9055775023001045,
+      "learning_rate": 3.0447078781117976e-07,
+      "loss": 0.022,
+      "step": 14772
+    },
+    {
+      "epoch": 3.3575,
+      "grad_norm": 1.5169962112374735,
+      "learning_rate": 3.043941719024981e-07,
+      "loss": 0.0132,
+      "step": 14773
+    },
+    {
+      "epoch": 3.357727272727273,
+      "grad_norm": 2.122943080734216,
+      "learning_rate": 3.0431756253152843e-07,
+      "loss": 0.0068,
+      "step": 14774
+    },
+    {
+      "epoch": 3.3579545454545454,
+      "grad_norm": 1.8296195983014822,
+      "learning_rate": 3.0424095969983314e-07,
+      "loss": 0.0138,
+      "step": 14775
+    },
+    {
+      "epoch": 3.3581818181818184,
+      "grad_norm": 0.7591207009624351,
+      "learning_rate": 3.0416436340897423e-07,
+      "loss": 0.0147,
+      "step": 14776
+    },
+    {
+      "epoch": 3.358409090909091,
+      "grad_norm": 2.550549928011734,
+      "learning_rate": 3.040877736605137e-07,
+      "loss": 0.0096,
+      "step": 14777
+    },
+    {
+      "epoch": 3.3586363636363634,
+      "grad_norm": 1.0834251121920626,
+      "learning_rate": 3.040111904560133e-07,
+      "loss": 0.0045,
+      "step": 14778
+    },
+    {
+      "epoch": 3.3588636363636364,
+      "grad_norm": 1.4821323169283778,
+      "learning_rate": 3.039346137970348e-07,
+      "loss": 0.0116,
+      "step": 14779
+    },
+    {
+      "epoch": 3.359090909090909,
+      "grad_norm": 1.0431736937660312,
+      "learning_rate": 3.038580436851396e-07,
+      "loss": 0.0086,
+      "step": 14780
+    },
+    {
+      "epoch": 3.359318181818182,
+      "grad_norm": 1.5772748443363394,
+      "learning_rate": 3.037814801218892e-07,
+      "loss": 0.0054,
+      "step": 14781
+    },
+    {
+      "epoch": 3.3595454545454544,
+      "grad_norm": 1.6168666396603926,
+      "learning_rate": 3.03704923108845e-07,
+      "loss": 0.0114,
+      "step": 14782
+    },
+    {
+      "epoch": 3.3597727272727274,
+      "grad_norm": 3.820815127253125,
+      "learning_rate": 3.036283726475677e-07,
+      "loss": 0.0237,
+      "step": 14783
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 1.1650674296297203,
+      "learning_rate": 3.035518287396186e-07,
+      "loss": 0.0048,
+      "step": 14784
+    },
+    {
+      "epoch": 3.360227272727273,
+      "grad_norm": 1.3040477889947677,
+      "learning_rate": 3.0347529138655845e-07,
+      "loss": 0.0082,
+      "step": 14785
+    },
+    {
+      "epoch": 3.3604545454545454,
+      "grad_norm": 5.492505552739246,
+      "learning_rate": 3.033987605899483e-07,
+      "loss": 0.0701,
+      "step": 14786
+    },
+    {
+      "epoch": 3.3606818181818183,
+      "grad_norm": 2.6511755652127715,
+      "learning_rate": 3.0332223635134826e-07,
+      "loss": 0.0205,
+      "step": 14787
+    },
+    {
+      "epoch": 3.360909090909091,
+      "grad_norm": 1.7377295177610605,
+      "learning_rate": 3.03245718672319e-07,
+      "loss": 0.0063,
+      "step": 14788
+    },
+    {
+      "epoch": 3.361136363636364,
+      "grad_norm": 2.007012829085465,
+      "learning_rate": 3.031692075544209e-07,
+      "loss": 0.0245,
+      "step": 14789
+    },
+    {
+      "epoch": 3.3613636363636363,
+      "grad_norm": 1.1532955402178078,
+      "learning_rate": 3.030927029992144e-07,
+      "loss": 0.0081,
+      "step": 14790
+    },
+    {
+      "epoch": 3.361590909090909,
+      "grad_norm": 1.7818271312649538,
+      "learning_rate": 3.030162050082587e-07,
+      "loss": 0.021,
+      "step": 14791
+    },
+    {
+      "epoch": 3.361818181818182,
+      "grad_norm": 0.8656161798544414,
+      "learning_rate": 3.0293971358311453e-07,
+      "loss": 0.0047,
+      "step": 14792
+    },
+    {
+      "epoch": 3.362045454545455,
+      "grad_norm": 2.6196371466268973,
+      "learning_rate": 3.0286322872534153e-07,
+      "loss": 0.0119,
+      "step": 14793
+    },
+    {
+      "epoch": 3.3622727272727273,
+      "grad_norm": 1.1854974591156056,
+      "learning_rate": 3.0278675043649936e-07,
+      "loss": 0.0089,
+      "step": 14794
+    },
+    {
+      "epoch": 3.3625,
+      "grad_norm": 1.8612174525149652,
+      "learning_rate": 3.0271027871814766e-07,
+      "loss": 0.013,
+      "step": 14795
+    },
+    {
+      "epoch": 3.362727272727273,
+      "grad_norm": 1.9873614327583773,
+      "learning_rate": 3.026338135718454e-07,
+      "loss": 0.0098,
+      "step": 14796
+    },
+    {
+      "epoch": 3.3629545454545453,
+      "grad_norm": 1.492397879735934,
+      "learning_rate": 3.025573549991521e-07,
+      "loss": 0.0091,
+      "step": 14797
+    },
+    {
+      "epoch": 3.3631818181818183,
+      "grad_norm": 2.70183514747012,
+      "learning_rate": 3.024809030016268e-07,
+      "loss": 0.0109,
+      "step": 14798
+    },
+    {
+      "epoch": 3.363409090909091,
+      "grad_norm": 5.24223669855548,
+      "learning_rate": 3.0240445758082884e-07,
+      "loss": 0.031,
+      "step": 14799
+    },
+    {
+      "epoch": 3.3636363636363638,
+      "grad_norm": 0.459151950395205,
+      "learning_rate": 3.023280187383165e-07,
+      "loss": 0.0022,
+      "step": 14800
+    },
+    {
+      "epoch": 3.3638636363636363,
+      "grad_norm": 1.4602874172610634,
+      "learning_rate": 3.022515864756489e-07,
+      "loss": 0.0056,
+      "step": 14801
+    },
+    {
+      "epoch": 3.3640909090909092,
+      "grad_norm": 1.1385497589515785,
+      "learning_rate": 3.0217516079438444e-07,
+      "loss": 0.008,
+      "step": 14802
+    },
+    {
+      "epoch": 3.3643181818181818,
+      "grad_norm": 0.8178291365363626,
+      "learning_rate": 3.020987416960817e-07,
+      "loss": 0.0062,
+      "step": 14803
+    },
+    {
+      "epoch": 3.3645454545454547,
+      "grad_norm": 2.1552969231604697,
+      "learning_rate": 3.0202232918229915e-07,
+      "loss": 0.0272,
+      "step": 14804
+    },
+    {
+      "epoch": 3.3647727272727272,
+      "grad_norm": 1.7052430060053738,
+      "learning_rate": 3.0194592325459455e-07,
+      "loss": 0.0095,
+      "step": 14805
+    },
+    {
+      "epoch": 3.365,
+      "grad_norm": 2.498975992811564,
+      "learning_rate": 3.0186952391452613e-07,
+      "loss": 0.0121,
+      "step": 14806
+    },
+    {
+      "epoch": 3.3652272727272727,
+      "grad_norm": 5.488897117562769,
+      "learning_rate": 3.0179313116365183e-07,
+      "loss": 0.0366,
+      "step": 14807
+    },
+    {
+      "epoch": 3.3654545454545453,
+      "grad_norm": 1.3721772825621472,
+      "learning_rate": 3.017167450035295e-07,
+      "loss": 0.011,
+      "step": 14808
+    },
+    {
+      "epoch": 3.365681818181818,
+      "grad_norm": 1.2039584985887077,
+      "learning_rate": 3.0164036543571676e-07,
+      "loss": 0.01,
+      "step": 14809
+    },
+    {
+      "epoch": 3.3659090909090907,
+      "grad_norm": 2.2118985854803426,
+      "learning_rate": 3.015639924617711e-07,
+      "loss": 0.0146,
+      "step": 14810
+    },
+    {
+      "epoch": 3.3661363636363637,
+      "grad_norm": 1.141533584397946,
+      "learning_rate": 3.0148762608324984e-07,
+      "loss": 0.0051,
+      "step": 14811
+    },
+    {
+      "epoch": 3.3663636363636362,
+      "grad_norm": 1.8647762895924882,
+      "learning_rate": 3.014112663017105e-07,
+      "loss": 0.0183,
+      "step": 14812
+    },
+    {
+      "epoch": 3.366590909090909,
+      "grad_norm": 5.648093321688051,
+      "learning_rate": 3.0133491311870986e-07,
+      "loss": 0.0231,
+      "step": 14813
+    },
+    {
+      "epoch": 3.3668181818181817,
+      "grad_norm": 0.5939501252743161,
+      "learning_rate": 3.0125856653580496e-07,
+      "loss": 0.0043,
+      "step": 14814
+    },
+    {
+      "epoch": 3.3670454545454547,
+      "grad_norm": 2.0988410229667718,
+      "learning_rate": 3.011822265545526e-07,
+      "loss": 0.0329,
+      "step": 14815
+    },
+    {
+      "epoch": 3.367272727272727,
+      "grad_norm": 0.7403454918439848,
+      "learning_rate": 3.011058931765097e-07,
+      "loss": 0.0033,
+      "step": 14816
+    },
+    {
+      "epoch": 3.3675,
+      "grad_norm": 0.9383557813152584,
+      "learning_rate": 3.0102956640323284e-07,
+      "loss": 0.0085,
+      "step": 14817
+    },
+    {
+      "epoch": 3.3677272727272727,
+      "grad_norm": 1.796633524029563,
+      "learning_rate": 3.009532462362781e-07,
+      "loss": 0.0076,
+      "step": 14818
+    },
+    {
+      "epoch": 3.3679545454545456,
+      "grad_norm": 1.353081147591766,
+      "learning_rate": 3.00876932677202e-07,
+      "loss": 0.0061,
+      "step": 14819
+    },
+    {
+      "epoch": 3.368181818181818,
+      "grad_norm": 2.6111719129386013,
+      "learning_rate": 3.008006257275608e-07,
+      "loss": 0.0164,
+      "step": 14820
+    },
+    {
+      "epoch": 3.3684090909090907,
+      "grad_norm": 7.227304293259686,
+      "learning_rate": 3.0072432538891033e-07,
+      "loss": 0.007,
+      "step": 14821
+    },
+    {
+      "epoch": 3.3686363636363637,
+      "grad_norm": 0.997012428502806,
+      "learning_rate": 3.006480316628066e-07,
+      "loss": 0.0046,
+      "step": 14822
+    },
+    {
+      "epoch": 3.368863636363636,
+      "grad_norm": 1.9229721966311981,
+      "learning_rate": 3.0057174455080543e-07,
+      "loss": 0.0144,
+      "step": 14823
+    },
+    {
+      "epoch": 3.369090909090909,
+      "grad_norm": 1.9694394945117233,
+      "learning_rate": 3.004954640544623e-07,
+      "loss": 0.0167,
+      "step": 14824
+    },
+    {
+      "epoch": 3.3693181818181817,
+      "grad_norm": 1.382450642199829,
+      "learning_rate": 3.004191901753328e-07,
+      "loss": 0.0032,
+      "step": 14825
+    },
+    {
+      "epoch": 3.3695454545454546,
+      "grad_norm": 0.9371076157749163,
+      "learning_rate": 3.003429229149726e-07,
+      "loss": 0.0059,
+      "step": 14826
+    },
+    {
+      "epoch": 3.369772727272727,
+      "grad_norm": 2.2746809808817954,
+      "learning_rate": 3.002666622749363e-07,
+      "loss": 0.0154,
+      "step": 14827
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 2.166778968790428,
+      "learning_rate": 3.001904082567793e-07,
+      "loss": 0.0123,
+      "step": 14828
+    },
+    {
+      "epoch": 3.3702272727272726,
+      "grad_norm": 1.635404270103046,
+      "learning_rate": 3.0011416086205657e-07,
+      "loss": 0.0086,
+      "step": 14829
+    },
+    {
+      "epoch": 3.3704545454545456,
+      "grad_norm": 0.5275470079556056,
+      "learning_rate": 3.0003792009232305e-07,
+      "loss": 0.0022,
+      "step": 14830
+    },
+    {
+      "epoch": 3.370681818181818,
+      "grad_norm": 1.7648461905657733,
+      "learning_rate": 2.9996168594913305e-07,
+      "loss": 0.0122,
+      "step": 14831
+    },
+    {
+      "epoch": 3.370909090909091,
+      "grad_norm": 2.331224123199841,
+      "learning_rate": 2.998854584340414e-07,
+      "loss": 0.0196,
+      "step": 14832
+    },
+    {
+      "epoch": 3.3711363636363636,
+      "grad_norm": 0.4309120256250248,
+      "learning_rate": 2.998092375486024e-07,
+      "loss": 0.0028,
+      "step": 14833
+    },
+    {
+      "epoch": 3.371363636363636,
+      "grad_norm": 2.158492616669733,
+      "learning_rate": 2.997330232943706e-07,
+      "loss": 0.0149,
+      "step": 14834
+    },
+    {
+      "epoch": 3.371590909090909,
+      "grad_norm": 2.832379899103242,
+      "learning_rate": 2.9965681567289966e-07,
+      "loss": 0.019,
+      "step": 14835
+    },
+    {
+      "epoch": 3.371818181818182,
+      "grad_norm": 0.777525369670447,
+      "learning_rate": 2.995806146857439e-07,
+      "loss": 0.0062,
+      "step": 14836
+    },
+    {
+      "epoch": 3.3720454545454546,
+      "grad_norm": 1.1841506466704397,
+      "learning_rate": 2.995044203344571e-07,
+      "loss": 0.0075,
+      "step": 14837
+    },
+    {
+      "epoch": 3.372272727272727,
+      "grad_norm": 0.6919915173405882,
+      "learning_rate": 2.99428232620593e-07,
+      "loss": 0.005,
+      "step": 14838
+    },
+    {
+      "epoch": 3.3725,
+      "grad_norm": 1.2671607877012618,
+      "learning_rate": 2.9935205154570523e-07,
+      "loss": 0.0056,
+      "step": 14839
+    },
+    {
+      "epoch": 3.3727272727272726,
+      "grad_norm": 2.7170973273761683,
+      "learning_rate": 2.992758771113473e-07,
+      "loss": 0.0121,
+      "step": 14840
+    },
+    {
+      "epoch": 3.3729545454545455,
+      "grad_norm": 1.3498283841619865,
+      "learning_rate": 2.991997093190725e-07,
+      "loss": 0.0103,
+      "step": 14841
+    },
+    {
+      "epoch": 3.373181818181818,
+      "grad_norm": 2.0524517369626665,
+      "learning_rate": 2.9912354817043393e-07,
+      "loss": 0.0115,
+      "step": 14842
+    },
+    {
+      "epoch": 3.373409090909091,
+      "grad_norm": 1.0637188894696101,
+      "learning_rate": 2.9904739366698507e-07,
+      "loss": 0.0076,
+      "step": 14843
+    },
+    {
+      "epoch": 3.3736363636363635,
+      "grad_norm": 2.500913147146442,
+      "learning_rate": 2.9897124581027823e-07,
+      "loss": 0.0192,
+      "step": 14844
+    },
+    {
+      "epoch": 3.3738636363636365,
+      "grad_norm": 2.425672704237152,
+      "learning_rate": 2.988951046018665e-07,
+      "loss": 0.021,
+      "step": 14845
+    },
+    {
+      "epoch": 3.374090909090909,
+      "grad_norm": 2.2401044341231438,
+      "learning_rate": 2.9881897004330244e-07,
+      "loss": 0.0206,
+      "step": 14846
+    },
+    {
+      "epoch": 3.374318181818182,
+      "grad_norm": 1.836062041487772,
+      "learning_rate": 2.987428421361387e-07,
+      "loss": 0.0217,
+      "step": 14847
+    },
+    {
+      "epoch": 3.3745454545454545,
+      "grad_norm": 1.100181012421458,
+      "learning_rate": 2.9866672088192777e-07,
+      "loss": 0.0031,
+      "step": 14848
+    },
+    {
+      "epoch": 3.3747727272727275,
+      "grad_norm": 1.4005618431190714,
+      "learning_rate": 2.985906062822216e-07,
+      "loss": 0.0083,
+      "step": 14849
+    },
+    {
+      "epoch": 3.375,
+      "grad_norm": 1.8821049981676117,
+      "learning_rate": 2.985144983385723e-07,
+      "loss": 0.0116,
+      "step": 14850
+    },
+    {
+      "epoch": 3.3752272727272725,
+      "grad_norm": 1.437666308862748,
+      "learning_rate": 2.98438397052532e-07,
+      "loss": 0.0037,
+      "step": 14851
+    },
+    {
+      "epoch": 3.3754545454545455,
+      "grad_norm": 1.6757452048954966,
+      "learning_rate": 2.9836230242565255e-07,
+      "loss": 0.0077,
+      "step": 14852
+    },
+    {
+      "epoch": 3.375681818181818,
+      "grad_norm": 1.5421782437233202,
+      "learning_rate": 2.9828621445948555e-07,
+      "loss": 0.011,
+      "step": 14853
+    },
+    {
+      "epoch": 3.375909090909091,
+      "grad_norm": 1.0097499665042453,
+      "learning_rate": 2.982101331555826e-07,
+      "loss": 0.0068,
+      "step": 14854
+    },
+    {
+      "epoch": 3.3761363636363635,
+      "grad_norm": 2.2457013501187144,
+      "learning_rate": 2.9813405851549524e-07,
+      "loss": 0.0206,
+      "step": 14855
+    },
+    {
+      "epoch": 3.3763636363636365,
+      "grad_norm": 0.9992291515766529,
+      "learning_rate": 2.9805799054077487e-07,
+      "loss": 0.0047,
+      "step": 14856
+    },
+    {
+      "epoch": 3.376590909090909,
+      "grad_norm": 1.3053618369270445,
+      "learning_rate": 2.979819292329722e-07,
+      "loss": 0.01,
+      "step": 14857
+    },
+    {
+      "epoch": 3.376818181818182,
+      "grad_norm": 2.49545842885851,
+      "learning_rate": 2.979058745936385e-07,
+      "loss": 0.0089,
+      "step": 14858
+    },
+    {
+      "epoch": 3.3770454545454545,
+      "grad_norm": 0.8546063418000921,
+      "learning_rate": 2.978298266243247e-07,
+      "loss": 0.0048,
+      "step": 14859
+    },
+    {
+      "epoch": 3.3772727272727274,
+      "grad_norm": 0.8878892697735246,
+      "learning_rate": 2.977537853265815e-07,
+      "loss": 0.0037,
+      "step": 14860
+    },
+    {
+      "epoch": 3.3775,
+      "grad_norm": 1.2199614483418764,
+      "learning_rate": 2.976777507019597e-07,
+      "loss": 0.006,
+      "step": 14861
+    },
+    {
+      "epoch": 3.377727272727273,
+      "grad_norm": 2.068311339316888,
+      "learning_rate": 2.976017227520095e-07,
+      "loss": 0.0425,
+      "step": 14862
+    },
+    {
+      "epoch": 3.3779545454545454,
+      "grad_norm": 2.570983364125201,
+      "learning_rate": 2.9752570147828124e-07,
+      "loss": 0.0101,
+      "step": 14863
+    },
+    {
+      "epoch": 3.378181818181818,
+      "grad_norm": 0.6283702005025336,
+      "learning_rate": 2.9744968688232535e-07,
+      "loss": 0.0056,
+      "step": 14864
+    },
+    {
+      "epoch": 3.378409090909091,
+      "grad_norm": 1.8576463289209257,
+      "learning_rate": 2.973736789656919e-07,
+      "loss": 0.0121,
+      "step": 14865
+    },
+    {
+      "epoch": 3.378636363636364,
+      "grad_norm": 0.31223982304809605,
+      "learning_rate": 2.972976777299306e-07,
+      "loss": 0.001,
+      "step": 14866
+    },
+    {
+      "epoch": 3.3788636363636364,
+      "grad_norm": 1.6235143565813326,
+      "learning_rate": 2.9722168317659126e-07,
+      "loss": 0.0084,
+      "step": 14867
+    },
+    {
+      "epoch": 3.379090909090909,
+      "grad_norm": 3.272027623810863,
+      "learning_rate": 2.9714569530722365e-07,
+      "loss": 0.0279,
+      "step": 14868
+    },
+    {
+      "epoch": 3.379318181818182,
+      "grad_norm": 2.2430057175928737,
+      "learning_rate": 2.9706971412337725e-07,
+      "loss": 0.013,
+      "step": 14869
+    },
+    {
+      "epoch": 3.3795454545454544,
+      "grad_norm": 5.946338340031226,
+      "learning_rate": 2.9699373962660154e-07,
+      "loss": 0.0218,
+      "step": 14870
+    },
+    {
+      "epoch": 3.3797727272727274,
+      "grad_norm": 1.6083654621639114,
+      "learning_rate": 2.969177718184458e-07,
+      "loss": 0.0087,
+      "step": 14871
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 1.1277058580191177,
+      "learning_rate": 2.96841810700459e-07,
+      "loss": 0.008,
+      "step": 14872
+    },
+    {
+      "epoch": 3.380227272727273,
+      "grad_norm": 1.7546926357867387,
+      "learning_rate": 2.9676585627419016e-07,
+      "loss": 0.009,
+      "step": 14873
+    },
+    {
+      "epoch": 3.3804545454545454,
+      "grad_norm": 4.577998261661245,
+      "learning_rate": 2.9668990854118833e-07,
+      "loss": 0.0225,
+      "step": 14874
+    },
+    {
+      "epoch": 3.3806818181818183,
+      "grad_norm": 1.454970633385817,
+      "learning_rate": 2.9661396750300183e-07,
+      "loss": 0.0069,
+      "step": 14875
+    },
+    {
+      "epoch": 3.380909090909091,
+      "grad_norm": 1.1167035341340776,
+      "learning_rate": 2.965380331611795e-07,
+      "loss": 0.0085,
+      "step": 14876
+    },
+    {
+      "epoch": 3.381136363636364,
+      "grad_norm": 1.3402208982948876,
+      "learning_rate": 2.9646210551726964e-07,
+      "loss": 0.0048,
+      "step": 14877
+    },
+    {
+      "epoch": 3.3813636363636363,
+      "grad_norm": 1.8922114316501955,
+      "learning_rate": 2.963861845728209e-07,
+      "loss": 0.0185,
+      "step": 14878
+    },
+    {
+      "epoch": 3.3815909090909093,
+      "grad_norm": 1.7700685900228554,
+      "learning_rate": 2.963102703293809e-07,
+      "loss": 0.0042,
+      "step": 14879
+    },
+    {
+      "epoch": 3.381818181818182,
+      "grad_norm": 2.869234876660909,
+      "learning_rate": 2.962343627884979e-07,
+      "loss": 0.0119,
+      "step": 14880
+    },
+    {
+      "epoch": 3.3820454545454544,
+      "grad_norm": 1.0751329075925014,
+      "learning_rate": 2.961584619517198e-07,
+      "loss": 0.0078,
+      "step": 14881
+    },
+    {
+      "epoch": 3.3822727272727273,
+      "grad_norm": 3.346413438905261,
+      "learning_rate": 2.960825678205944e-07,
+      "loss": 0.0165,
+      "step": 14882
+    },
+    {
+      "epoch": 3.3825,
+      "grad_norm": 7.855796589618396,
+      "learning_rate": 2.960066803966692e-07,
+      "loss": 0.0087,
+      "step": 14883
+    },
+    {
+      "epoch": 3.382727272727273,
+      "grad_norm": 1.1168658591403382,
+      "learning_rate": 2.959307996814918e-07,
+      "loss": 0.0086,
+      "step": 14884
+    },
+    {
+      "epoch": 3.3829545454545453,
+      "grad_norm": 0.8431464058032758,
+      "learning_rate": 2.958549256766095e-07,
+      "loss": 0.0047,
+      "step": 14885
+    },
+    {
+      "epoch": 3.3831818181818183,
+      "grad_norm": 1.350497329578761,
+      "learning_rate": 2.957790583835695e-07,
+      "loss": 0.0063,
+      "step": 14886
+    },
+    {
+      "epoch": 3.383409090909091,
+      "grad_norm": 0.9224523892023503,
+      "learning_rate": 2.9570319780391906e-07,
+      "loss": 0.0032,
+      "step": 14887
+    },
+    {
+      "epoch": 3.3836363636363638,
+      "grad_norm": 2.4988637941195804,
+      "learning_rate": 2.956273439392047e-07,
+      "loss": 0.0176,
+      "step": 14888
+    },
+    {
+      "epoch": 3.3838636363636363,
+      "grad_norm": 1.2256171282182944,
+      "learning_rate": 2.955514967909734e-07,
+      "loss": 0.0085,
+      "step": 14889
+    },
+    {
+      "epoch": 3.3840909090909093,
+      "grad_norm": 1.257915991372814,
+      "learning_rate": 2.954756563607718e-07,
+      "loss": 0.0053,
+      "step": 14890
+    },
+    {
+      "epoch": 3.384318181818182,
+      "grad_norm": 0.8740844188175856,
+      "learning_rate": 2.953998226501467e-07,
+      "loss": 0.0092,
+      "step": 14891
+    },
+    {
+      "epoch": 3.3845454545454547,
+      "grad_norm": 3.8163201190537883,
+      "learning_rate": 2.95323995660644e-07,
+      "loss": 0.0246,
+      "step": 14892
+    },
+    {
+      "epoch": 3.3847727272727273,
+      "grad_norm": 1.629581687022322,
+      "learning_rate": 2.9524817539381023e-07,
+      "loss": 0.0125,
+      "step": 14893
+    },
+    {
+      "epoch": 3.385,
+      "grad_norm": 0.567679082223827,
+      "learning_rate": 2.951723618511914e-07,
+      "loss": 0.0032,
+      "step": 14894
+    },
+    {
+      "epoch": 3.3852272727272728,
+      "grad_norm": 1.0766288856479949,
+      "learning_rate": 2.9509655503433357e-07,
+      "loss": 0.0066,
+      "step": 14895
+    },
+    {
+      "epoch": 3.3854545454545453,
+      "grad_norm": 1.554854833314714,
+      "learning_rate": 2.950207549447827e-07,
+      "loss": 0.0039,
+      "step": 14896
+    },
+    {
+      "epoch": 3.3856818181818182,
+      "grad_norm": 1.4107389917202968,
+      "learning_rate": 2.949449615840842e-07,
+      "loss": 0.0067,
+      "step": 14897
+    },
+    {
+      "epoch": 3.3859090909090908,
+      "grad_norm": 0.5295343568753611,
+      "learning_rate": 2.948691749537836e-07,
+      "loss": 0.0023,
+      "step": 14898
+    },
+    {
+      "epoch": 3.3861363636363637,
+      "grad_norm": 1.403697053465331,
+      "learning_rate": 2.9479339505542654e-07,
+      "loss": 0.0088,
+      "step": 14899
+    },
+    {
+      "epoch": 3.3863636363636362,
+      "grad_norm": 1.205566787671913,
+      "learning_rate": 2.947176218905582e-07,
+      "loss": 0.0104,
+      "step": 14900
+    },
+    {
+      "epoch": 3.386590909090909,
+      "grad_norm": 2.401783180608921,
+      "learning_rate": 2.9464185546072383e-07,
+      "loss": 0.01,
+      "step": 14901
+    },
+    {
+      "epoch": 3.3868181818181817,
+      "grad_norm": 3.0461952969493082,
+      "learning_rate": 2.945660957674683e-07,
+      "loss": 0.0068,
+      "step": 14902
+    },
+    {
+      "epoch": 3.3870454545454547,
+      "grad_norm": 0.8044366150116564,
+      "learning_rate": 2.944903428123366e-07,
+      "loss": 0.0062,
+      "step": 14903
+    },
+    {
+      "epoch": 3.387272727272727,
+      "grad_norm": 2.6076690505378974,
+      "learning_rate": 2.944145965968733e-07,
+      "loss": 0.0248,
+      "step": 14904
+    },
+    {
+      "epoch": 3.3875,
+      "grad_norm": 1.4290978880681384,
+      "learning_rate": 2.943388571226235e-07,
+      "loss": 0.0141,
+      "step": 14905
+    },
+    {
+      "epoch": 3.3877272727272727,
+      "grad_norm": 0.7269538294171102,
+      "learning_rate": 2.94263124391131e-07,
+      "loss": 0.0047,
+      "step": 14906
+    },
+    {
+      "epoch": 3.387954545454545,
+      "grad_norm": 0.626622328690851,
+      "learning_rate": 2.941873984039403e-07,
+      "loss": 0.0026,
+      "step": 14907
+    },
+    {
+      "epoch": 3.388181818181818,
+      "grad_norm": 3.4222858213148895,
+      "learning_rate": 2.941116791625958e-07,
+      "loss": 0.014,
+      "step": 14908
+    },
+    {
+      "epoch": 3.388409090909091,
+      "grad_norm": 1.7889287581952875,
+      "learning_rate": 2.940359666686416e-07,
+      "loss": 0.0137,
+      "step": 14909
+    },
+    {
+      "epoch": 3.3886363636363637,
+      "grad_norm": 0.3143139291222979,
+      "learning_rate": 2.939602609236212e-07,
+      "loss": 0.0032,
+      "step": 14910
+    },
+    {
+      "epoch": 3.388863636363636,
+      "grad_norm": 1.1488233385599969,
+      "learning_rate": 2.938845619290787e-07,
+      "loss": 0.0046,
+      "step": 14911
+    },
+    {
+      "epoch": 3.389090909090909,
+      "grad_norm": 1.3109308125187882,
+      "learning_rate": 2.938088696865575e-07,
+      "loss": 0.0072,
+      "step": 14912
+    },
+    {
+      "epoch": 3.3893181818181817,
+      "grad_norm": 1.8759611269961014,
+      "learning_rate": 2.9373318419760133e-07,
+      "loss": 0.0166,
+      "step": 14913
+    },
+    {
+      "epoch": 3.3895454545454546,
+      "grad_norm": 1.1656556588240057,
+      "learning_rate": 2.9365750546375345e-07,
+      "loss": 0.0082,
+      "step": 14914
+    },
+    {
+      "epoch": 3.389772727272727,
+      "grad_norm": 2.358717295330695,
+      "learning_rate": 2.9358183348655705e-07,
+      "loss": 0.0247,
+      "step": 14915
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 1.126523766112721,
+      "learning_rate": 2.935061682675553e-07,
+      "loss": 0.0088,
+      "step": 14916
+    },
+    {
+      "epoch": 3.3902272727272726,
+      "grad_norm": 0.9960333323338897,
+      "learning_rate": 2.934305098082911e-07,
+      "loss": 0.0026,
+      "step": 14917
+    },
+    {
+      "epoch": 3.3904545454545456,
+      "grad_norm": 1.2167143871338817,
+      "learning_rate": 2.933548581103075e-07,
+      "loss": 0.0069,
+      "step": 14918
+    },
+    {
+      "epoch": 3.390681818181818,
+      "grad_norm": 1.1465516540705725,
+      "learning_rate": 2.932792131751467e-07,
+      "loss": 0.0112,
+      "step": 14919
+    },
+    {
+      "epoch": 3.390909090909091,
+      "grad_norm": 4.193324375245331,
+      "learning_rate": 2.932035750043515e-07,
+      "loss": 0.0218,
+      "step": 14920
+    },
+    {
+      "epoch": 3.3911363636363636,
+      "grad_norm": 1.166278921970567,
+      "learning_rate": 2.931279435994642e-07,
+      "loss": 0.0063,
+      "step": 14921
+    },
+    {
+      "epoch": 3.3913636363636366,
+      "grad_norm": 3.357859005642428,
+      "learning_rate": 2.9305231896202734e-07,
+      "loss": 0.0177,
+      "step": 14922
+    },
+    {
+      "epoch": 3.391590909090909,
+      "grad_norm": 4.1698572400253475,
+      "learning_rate": 2.929767010935825e-07,
+      "loss": 0.0256,
+      "step": 14923
+    },
+    {
+      "epoch": 3.3918181818181816,
+      "grad_norm": 1.5477639307829991,
+      "learning_rate": 2.929010899956721e-07,
+      "loss": 0.0067,
+      "step": 14924
+    },
+    {
+      "epoch": 3.3920454545454546,
+      "grad_norm": 2.3901216935395446,
+      "learning_rate": 2.928254856698378e-07,
+      "loss": 0.0312,
+      "step": 14925
+    },
+    {
+      "epoch": 3.392272727272727,
+      "grad_norm": 1.1509385909888068,
+      "learning_rate": 2.927498881176214e-07,
+      "loss": 0.0091,
+      "step": 14926
+    },
+    {
+      "epoch": 3.3925,
+      "grad_norm": 1.0438782113964906,
+      "learning_rate": 2.9267429734056464e-07,
+      "loss": 0.0067,
+      "step": 14927
+    },
+    {
+      "epoch": 3.3927272727272726,
+      "grad_norm": 1.7394414437524506,
+      "learning_rate": 2.9259871334020846e-07,
+      "loss": 0.0082,
+      "step": 14928
+    },
+    {
+      "epoch": 3.3929545454545456,
+      "grad_norm": 0.904551598741827,
+      "learning_rate": 2.925231361180942e-07,
+      "loss": 0.0039,
+      "step": 14929
+    },
+    {
+      "epoch": 3.393181818181818,
+      "grad_norm": 2.2056504745592034,
+      "learning_rate": 2.924475656757635e-07,
+      "loss": 0.0195,
+      "step": 14930
+    },
+    {
+      "epoch": 3.393409090909091,
+      "grad_norm": 3.7336601104132883,
+      "learning_rate": 2.923720020147573e-07,
+      "loss": 0.0492,
+      "step": 14931
+    },
+    {
+      "epoch": 3.3936363636363636,
+      "grad_norm": 1.330033981603453,
+      "learning_rate": 2.922964451366161e-07,
+      "loss": 0.0064,
+      "step": 14932
+    },
+    {
+      "epoch": 3.3938636363636365,
+      "grad_norm": 1.9077666462989022,
+      "learning_rate": 2.922208950428807e-07,
+      "loss": 0.0134,
+      "step": 14933
+    },
+    {
+      "epoch": 3.394090909090909,
+      "grad_norm": 1.2868271165646996,
+      "learning_rate": 2.9214535173509193e-07,
+      "loss": 0.0068,
+      "step": 14934
+    },
+    {
+      "epoch": 3.394318181818182,
+      "grad_norm": 1.75770044264619,
+      "learning_rate": 2.9206981521479036e-07,
+      "loss": 0.0114,
+      "step": 14935
+    },
+    {
+      "epoch": 3.3945454545454545,
+      "grad_norm": 3.2358150951335016,
+      "learning_rate": 2.9199428548351583e-07,
+      "loss": 0.0153,
+      "step": 14936
+    },
+    {
+      "epoch": 3.394772727272727,
+      "grad_norm": 1.291830489358793,
+      "learning_rate": 2.919187625428088e-07,
+      "loss": 0.0114,
+      "step": 14937
+    },
+    {
+      "epoch": 3.395,
+      "grad_norm": 1.5256789192502649,
+      "learning_rate": 2.918432463942092e-07,
+      "loss": 0.0088,
+      "step": 14938
+    },
+    {
+      "epoch": 3.3952272727272725,
+      "grad_norm": 1.859176673895922,
+      "learning_rate": 2.9176773703925714e-07,
+      "loss": 0.0066,
+      "step": 14939
+    },
+    {
+      "epoch": 3.3954545454545455,
+      "grad_norm": 1.7080720441094412,
+      "learning_rate": 2.916922344794924e-07,
+      "loss": 0.0052,
+      "step": 14940
+    },
+    {
+      "epoch": 3.395681818181818,
+      "grad_norm": 1.5067754452625768,
+      "learning_rate": 2.916167387164544e-07,
+      "loss": 0.0066,
+      "step": 14941
+    },
+    {
+      "epoch": 3.395909090909091,
+      "grad_norm": 0.3588601695274511,
+      "learning_rate": 2.9154124975168256e-07,
+      "loss": 0.0018,
+      "step": 14942
+    },
+    {
+      "epoch": 3.3961363636363635,
+      "grad_norm": 2.0831661467766227,
+      "learning_rate": 2.914657675867165e-07,
+      "loss": 0.0087,
+      "step": 14943
+    },
+    {
+      "epoch": 3.3963636363636365,
+      "grad_norm": 1.7537371174012628,
+      "learning_rate": 2.913902922230953e-07,
+      "loss": 0.0176,
+      "step": 14944
+    },
+    {
+      "epoch": 3.396590909090909,
+      "grad_norm": 0.3633424336560802,
+      "learning_rate": 2.9131482366235806e-07,
+      "loss": 0.003,
+      "step": 14945
+    },
+    {
+      "epoch": 3.396818181818182,
+      "grad_norm": 2.608707075872999,
+      "learning_rate": 2.912393619060437e-07,
+      "loss": 0.0121,
+      "step": 14946
+    },
+    {
+      "epoch": 3.3970454545454545,
+      "grad_norm": 1.8076243578778421,
+      "learning_rate": 2.91163906955691e-07,
+      "loss": 0.0132,
+      "step": 14947
+    },
+    {
+      "epoch": 3.3972727272727274,
+      "grad_norm": 0.1979067695486261,
+      "learning_rate": 2.9108845881283866e-07,
+      "loss": 0.0005,
+      "step": 14948
+    },
+    {
+      "epoch": 3.3975,
+      "grad_norm": 1.7523438217694567,
+      "learning_rate": 2.9101301747902545e-07,
+      "loss": 0.0069,
+      "step": 14949
+    },
+    {
+      "epoch": 3.3977272727272725,
+      "grad_norm": 0.7691141055708954,
+      "learning_rate": 2.909375829557892e-07,
+      "loss": 0.0061,
+      "step": 14950
+    },
+    {
+      "epoch": 3.3979545454545454,
+      "grad_norm": 1.8220933632320333,
+      "learning_rate": 2.908621552446685e-07,
+      "loss": 0.0069,
+      "step": 14951
+    },
+    {
+      "epoch": 3.3981818181818184,
+      "grad_norm": 1.8259798326575976,
+      "learning_rate": 2.907867343472014e-07,
+      "loss": 0.0104,
+      "step": 14952
+    },
+    {
+      "epoch": 3.398409090909091,
+      "grad_norm": 2.860606611265623,
+      "learning_rate": 2.9071132026492603e-07,
+      "loss": 0.0249,
+      "step": 14953
+    },
+    {
+      "epoch": 3.3986363636363635,
+      "grad_norm": 1.1612816520502003,
+      "learning_rate": 2.9063591299937977e-07,
+      "loss": 0.0095,
+      "step": 14954
+    },
+    {
+      "epoch": 3.3988636363636364,
+      "grad_norm": 1.8154996034832898,
+      "learning_rate": 2.905605125521006e-07,
+      "loss": 0.0085,
+      "step": 14955
+    },
+    {
+      "epoch": 3.399090909090909,
+      "grad_norm": 0.6221902436907859,
+      "learning_rate": 2.904851189246261e-07,
+      "loss": 0.0068,
+      "step": 14956
+    },
+    {
+      "epoch": 3.399318181818182,
+      "grad_norm": 0.6611092848754762,
+      "learning_rate": 2.9040973211849387e-07,
+      "loss": 0.0031,
+      "step": 14957
+    },
+    {
+      "epoch": 3.3995454545454544,
+      "grad_norm": 1.4935600122937276,
+      "learning_rate": 2.903343521352406e-07,
+      "loss": 0.0112,
+      "step": 14958
+    },
+    {
+      "epoch": 3.3997727272727274,
+      "grad_norm": 1.7453684263925777,
+      "learning_rate": 2.902589789764036e-07,
+      "loss": 0.0168,
+      "step": 14959
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 2.2672803397776193,
+      "learning_rate": 2.9018361264352015e-07,
+      "loss": 0.0169,
+      "step": 14960
+    },
+    {
+      "epoch": 3.400227272727273,
+      "grad_norm": 1.995722612927249,
+      "learning_rate": 2.90108253138127e-07,
+      "loss": 0.0263,
+      "step": 14961
+    },
+    {
+      "epoch": 3.4004545454545454,
+      "grad_norm": 2.247013474112504,
+      "learning_rate": 2.9003290046176107e-07,
+      "loss": 0.0084,
+      "step": 14962
+    },
+    {
+      "epoch": 3.4006818181818184,
+      "grad_norm": 2.1231070955101985,
+      "learning_rate": 2.899575546159584e-07,
+      "loss": 0.0166,
+      "step": 14963
+    },
+    {
+      "epoch": 3.400909090909091,
+      "grad_norm": 1.3130103115034937,
+      "learning_rate": 2.898822156022557e-07,
+      "loss": 0.0115,
+      "step": 14964
+    },
+    {
+      "epoch": 3.401136363636364,
+      "grad_norm": 0.9691330332423749,
+      "learning_rate": 2.8980688342218933e-07,
+      "loss": 0.0049,
+      "step": 14965
+    },
+    {
+      "epoch": 3.4013636363636364,
+      "grad_norm": 2.930286845043086,
+      "learning_rate": 2.897315580772956e-07,
+      "loss": 0.0312,
+      "step": 14966
+    },
+    {
+      "epoch": 3.401590909090909,
+      "grad_norm": 1.9850431554597094,
+      "learning_rate": 2.896562395691101e-07,
+      "loss": 0.0123,
+      "step": 14967
+    },
+    {
+      "epoch": 3.401818181818182,
+      "grad_norm": 2.090708012573773,
+      "learning_rate": 2.895809278991689e-07,
+      "loss": 0.0186,
+      "step": 14968
+    },
+    {
+      "epoch": 3.4020454545454544,
+      "grad_norm": 0.9588122358299722,
+      "learning_rate": 2.895056230690077e-07,
+      "loss": 0.0049,
+      "step": 14969
+    },
+    {
+      "epoch": 3.4022727272727273,
+      "grad_norm": 2.0323371525600997,
+      "learning_rate": 2.8943032508016217e-07,
+      "loss": 0.0078,
+      "step": 14970
+    },
+    {
+      "epoch": 3.4025,
+      "grad_norm": 0.8499426115914086,
+      "learning_rate": 2.8935503393416805e-07,
+      "loss": 0.006,
+      "step": 14971
+    },
+    {
+      "epoch": 3.402727272727273,
+      "grad_norm": 2.203596461058399,
+      "learning_rate": 2.8927974963256e-07,
+      "loss": 0.0172,
+      "step": 14972
+    },
+    {
+      "epoch": 3.4029545454545453,
+      "grad_norm": 1.2795226674176996,
+      "learning_rate": 2.892044721768736e-07,
+      "loss": 0.0069,
+      "step": 14973
+    },
+    {
+      "epoch": 3.4031818181818183,
+      "grad_norm": 1.821039883831525,
+      "learning_rate": 2.891292015686439e-07,
+      "loss": 0.0137,
+      "step": 14974
+    },
+    {
+      "epoch": 3.403409090909091,
+      "grad_norm": 1.110706439680722,
+      "learning_rate": 2.890539378094057e-07,
+      "loss": 0.0087,
+      "step": 14975
+    },
+    {
+      "epoch": 3.403636363636364,
+      "grad_norm": 1.8357308221545536,
+      "learning_rate": 2.889786809006937e-07,
+      "loss": 0.0132,
+      "step": 14976
+    },
+    {
+      "epoch": 3.4038636363636363,
+      "grad_norm": 0.9290809198200921,
+      "learning_rate": 2.8890343084404276e-07,
+      "loss": 0.0068,
+      "step": 14977
+    },
+    {
+      "epoch": 3.4040909090909093,
+      "grad_norm": 2.2080212798163315,
+      "learning_rate": 2.8882818764098717e-07,
+      "loss": 0.0143,
+      "step": 14978
+    },
+    {
+      "epoch": 3.404318181818182,
+      "grad_norm": 1.3113165244542837,
+      "learning_rate": 2.8875295129306154e-07,
+      "loss": 0.0083,
+      "step": 14979
+    },
+    {
+      "epoch": 3.4045454545454543,
+      "grad_norm": 1.5983766819382883,
+      "learning_rate": 2.8867772180179954e-07,
+      "loss": 0.0103,
+      "step": 14980
+    },
+    {
+      "epoch": 3.4047727272727273,
+      "grad_norm": 0.713936199399816,
+      "learning_rate": 2.886024991687357e-07,
+      "loss": 0.0042,
+      "step": 14981
+    },
+    {
+      "epoch": 3.4050000000000002,
+      "grad_norm": 5.240508372052065,
+      "learning_rate": 2.8852728339540366e-07,
+      "loss": 0.0203,
+      "step": 14982
+    },
+    {
+      "epoch": 3.4052272727272728,
+      "grad_norm": 1.1162680845243025,
+      "learning_rate": 2.884520744833374e-07,
+      "loss": 0.0095,
+      "step": 14983
+    },
+    {
+      "epoch": 3.4054545454545453,
+      "grad_norm": 0.7430830248039652,
+      "learning_rate": 2.8837687243407064e-07,
+      "loss": 0.0041,
+      "step": 14984
+    },
+    {
+      "epoch": 3.4056818181818183,
+      "grad_norm": 2.4784333766705813,
+      "learning_rate": 2.883016772491365e-07,
+      "loss": 0.0312,
+      "step": 14985
+    },
+    {
+      "epoch": 3.4059090909090908,
+      "grad_norm": 1.2373954796928108,
+      "learning_rate": 2.882264889300686e-07,
+      "loss": 0.0147,
+      "step": 14986
+    },
+    {
+      "epoch": 3.4061363636363637,
+      "grad_norm": 1.5751578854700825,
+      "learning_rate": 2.881513074784001e-07,
+      "loss": 0.0147,
+      "step": 14987
+    },
+    {
+      "epoch": 3.4063636363636363,
+      "grad_norm": 2.804279415220475,
+      "learning_rate": 2.880761328956641e-07,
+      "loss": 0.0274,
+      "step": 14988
+    },
+    {
+      "epoch": 3.4065909090909092,
+      "grad_norm": 0.7173470035446863,
+      "learning_rate": 2.880009651833936e-07,
+      "loss": 0.0031,
+      "step": 14989
+    },
+    {
+      "epoch": 3.4068181818181817,
+      "grad_norm": 1.3814421792604026,
+      "learning_rate": 2.879258043431213e-07,
+      "loss": 0.0105,
+      "step": 14990
+    },
+    {
+      "epoch": 3.4070454545454547,
+      "grad_norm": 0.861347499530473,
+      "learning_rate": 2.878506503763799e-07,
+      "loss": 0.0033,
+      "step": 14991
+    },
+    {
+      "epoch": 3.4072727272727272,
+      "grad_norm": 1.566169841681974,
+      "learning_rate": 2.8777550328470195e-07,
+      "loss": 0.0087,
+      "step": 14992
+    },
+    {
+      "epoch": 3.4074999999999998,
+      "grad_norm": 0.8483827179874721,
+      "learning_rate": 2.877003630696201e-07,
+      "loss": 0.0039,
+      "step": 14993
+    },
+    {
+      "epoch": 3.4077272727272727,
+      "grad_norm": 4.953262990201939,
+      "learning_rate": 2.876252297326659e-07,
+      "loss": 0.0138,
+      "step": 14994
+    },
+    {
+      "epoch": 3.4079545454545457,
+      "grad_norm": 2.410766791612414,
+      "learning_rate": 2.87550103275372e-07,
+      "loss": 0.017,
+      "step": 14995
+    },
+    {
+      "epoch": 3.408181818181818,
+      "grad_norm": 1.9043896696043157,
+      "learning_rate": 2.874749836992701e-07,
+      "loss": 0.0106,
+      "step": 14996
+    },
+    {
+      "epoch": 3.4084090909090907,
+      "grad_norm": 1.4037545261647255,
+      "learning_rate": 2.8739987100589244e-07,
+      "loss": 0.0073,
+      "step": 14997
+    },
+    {
+      "epoch": 3.4086363636363637,
+      "grad_norm": 1.6653683005861728,
+      "learning_rate": 2.8732476519677013e-07,
+      "loss": 0.0147,
+      "step": 14998
+    },
+    {
+      "epoch": 3.408863636363636,
+      "grad_norm": 0.4717312373335682,
+      "learning_rate": 2.87249666273435e-07,
+      "loss": 0.0018,
+      "step": 14999
+    },
+    {
+      "epoch": 3.409090909090909,
+      "grad_norm": 1.3400643723953802,
+      "learning_rate": 2.871745742374184e-07,
+      "loss": 0.007,
+      "step": 15000
+    },
+    {
+      "epoch": 3.4093181818181817,
+      "grad_norm": 0.26600884965497756,
+      "learning_rate": 2.8709948909025175e-07,
+      "loss": 0.0013,
+      "step": 15001
+    },
+    {
+      "epoch": 3.4095454545454547,
+      "grad_norm": 0.7938366834652838,
+      "learning_rate": 2.8702441083346586e-07,
+      "loss": 0.0042,
+      "step": 15002
+    },
+    {
+      "epoch": 3.409772727272727,
+      "grad_norm": 1.8280853506478212,
+      "learning_rate": 2.869493394685918e-07,
+      "loss": 0.0082,
+      "step": 15003
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 1.1676186775869615,
+      "learning_rate": 2.8687427499716056e-07,
+      "loss": 0.0061,
+      "step": 15004
+    },
+    {
+      "epoch": 3.4102272727272727,
+      "grad_norm": 1.4583291903980342,
+      "learning_rate": 2.8679921742070265e-07,
+      "loss": 0.006,
+      "step": 15005
+    },
+    {
+      "epoch": 3.4104545454545456,
+      "grad_norm": 1.8397353058136305,
+      "learning_rate": 2.8672416674074883e-07,
+      "loss": 0.0115,
+      "step": 15006
+    },
+    {
+      "epoch": 3.410681818181818,
+      "grad_norm": 2.194801312153386,
+      "learning_rate": 2.866491229588293e-07,
+      "loss": 0.0183,
+      "step": 15007
+    },
+    {
+      "epoch": 3.410909090909091,
+      "grad_norm": 1.5598879293775925,
+      "learning_rate": 2.8657408607647453e-07,
+      "loss": 0.0155,
+      "step": 15008
+    },
+    {
+      "epoch": 3.4111363636363636,
+      "grad_norm": 0.8212369706013413,
+      "learning_rate": 2.8649905609521454e-07,
+      "loss": 0.0029,
+      "step": 15009
+    },
+    {
+      "epoch": 3.411363636363636,
+      "grad_norm": 1.8315991798931976,
+      "learning_rate": 2.8642403301657946e-07,
+      "loss": 0.0158,
+      "step": 15010
+    },
+    {
+      "epoch": 3.411590909090909,
+      "grad_norm": 0.9524711629275406,
+      "learning_rate": 2.86349016842099e-07,
+      "loss": 0.0046,
+      "step": 15011
+    },
+    {
+      "epoch": 3.4118181818181816,
+      "grad_norm": 0.8748008982584218,
+      "learning_rate": 2.862740075733027e-07,
+      "loss": 0.0046,
+      "step": 15012
+    },
+    {
+      "epoch": 3.4120454545454546,
+      "grad_norm": 1.7979115080520334,
+      "learning_rate": 2.8619900521172044e-07,
+      "loss": 0.0118,
+      "step": 15013
+    },
+    {
+      "epoch": 3.412272727272727,
+      "grad_norm": 2.681951165678067,
+      "learning_rate": 2.8612400975888176e-07,
+      "loss": 0.0185,
+      "step": 15014
+    },
+    {
+      "epoch": 3.4125,
+      "grad_norm": 1.6044218285129854,
+      "learning_rate": 2.860490212163154e-07,
+      "loss": 0.0142,
+      "step": 15015
+    },
+    {
+      "epoch": 3.4127272727272726,
+      "grad_norm": 1.1374448289585448,
+      "learning_rate": 2.859740395855509e-07,
+      "loss": 0.0061,
+      "step": 15016
+    },
+    {
+      "epoch": 3.4129545454545456,
+      "grad_norm": 1.2357352709270986,
+      "learning_rate": 2.858990648681172e-07,
+      "loss": 0.0084,
+      "step": 15017
+    },
+    {
+      "epoch": 3.413181818181818,
+      "grad_norm": 1.1604318249068422,
+      "learning_rate": 2.8582409706554313e-07,
+      "loss": 0.011,
+      "step": 15018
+    },
+    {
+      "epoch": 3.413409090909091,
+      "grad_norm": 1.3572939892979434,
+      "learning_rate": 2.8574913617935744e-07,
+      "loss": 0.0067,
+      "step": 15019
+    },
+    {
+      "epoch": 3.4136363636363636,
+      "grad_norm": 1.3962988466907422,
+      "learning_rate": 2.856741822110888e-07,
+      "loss": 0.0061,
+      "step": 15020
+    },
+    {
+      "epoch": 3.4138636363636365,
+      "grad_norm": 1.204023126905033,
+      "learning_rate": 2.855992351622655e-07,
+      "loss": 0.0052,
+      "step": 15021
+    },
+    {
+      "epoch": 3.414090909090909,
+      "grad_norm": 0.9368260825546881,
+      "learning_rate": 2.8552429503441595e-07,
+      "loss": 0.0019,
+      "step": 15022
+    },
+    {
+      "epoch": 3.4143181818181816,
+      "grad_norm": 0.7063481333297774,
+      "learning_rate": 2.8544936182906845e-07,
+      "loss": 0.0036,
+      "step": 15023
+    },
+    {
+      "epoch": 3.4145454545454546,
+      "grad_norm": 1.5129061328464648,
+      "learning_rate": 2.8537443554775065e-07,
+      "loss": 0.0196,
+      "step": 15024
+    },
+    {
+      "epoch": 3.4147727272727275,
+      "grad_norm": 1.6496275822560384,
+      "learning_rate": 2.852995161919906e-07,
+      "loss": 0.0094,
+      "step": 15025
+    },
+    {
+      "epoch": 3.415,
+      "grad_norm": 3.0339489363252055,
+      "learning_rate": 2.8522460376331605e-07,
+      "loss": 0.0249,
+      "step": 15026
+    },
+    {
+      "epoch": 3.4152272727272726,
+      "grad_norm": 1.509438551339292,
+      "learning_rate": 2.851496982632546e-07,
+      "loss": 0.0066,
+      "step": 15027
+    },
+    {
+      "epoch": 3.4154545454545455,
+      "grad_norm": 2.314067699587611,
+      "learning_rate": 2.8507479969333404e-07,
+      "loss": 0.0253,
+      "step": 15028
+    },
+    {
+      "epoch": 3.415681818181818,
+      "grad_norm": 2.9312584417079806,
+      "learning_rate": 2.8499990805508097e-07,
+      "loss": 0.0142,
+      "step": 15029
+    },
+    {
+      "epoch": 3.415909090909091,
+      "grad_norm": 0.7756089911511901,
+      "learning_rate": 2.84925023350023e-07,
+      "loss": 0.0028,
+      "step": 15030
+    },
+    {
+      "epoch": 3.4161363636363635,
+      "grad_norm": 1.1106876631758198,
+      "learning_rate": 2.848501455796871e-07,
+      "loss": 0.003,
+      "step": 15031
+    },
+    {
+      "epoch": 3.4163636363636365,
+      "grad_norm": 3.3022273848233343,
+      "learning_rate": 2.847752747456003e-07,
+      "loss": 0.0166,
+      "step": 15032
+    },
+    {
+      "epoch": 3.416590909090909,
+      "grad_norm": 1.5210871686388037,
+      "learning_rate": 2.8470041084928913e-07,
+      "loss": 0.0116,
+      "step": 15033
+    },
+    {
+      "epoch": 3.416818181818182,
+      "grad_norm": 1.5000116443197393,
+      "learning_rate": 2.846255538922802e-07,
+      "loss": 0.0153,
+      "step": 15034
+    },
+    {
+      "epoch": 3.4170454545454545,
+      "grad_norm": 1.2501194211393902,
+      "learning_rate": 2.845507038761e-07,
+      "loss": 0.01,
+      "step": 15035
+    },
+    {
+      "epoch": 3.4172727272727275,
+      "grad_norm": 3.357057892361584,
+      "learning_rate": 2.844758608022749e-07,
+      "loss": 0.052,
+      "step": 15036
+    },
+    {
+      "epoch": 3.4175,
+      "grad_norm": 2.6677483854472976,
+      "learning_rate": 2.844010246723311e-07,
+      "loss": 0.0265,
+      "step": 15037
+    },
+    {
+      "epoch": 3.417727272727273,
+      "grad_norm": 1.1836204196151816,
+      "learning_rate": 2.843261954877946e-07,
+      "loss": 0.007,
+      "step": 15038
+    },
+    {
+      "epoch": 3.4179545454545455,
+      "grad_norm": 2.4492450507490036,
+      "learning_rate": 2.842513732501914e-07,
+      "loss": 0.0135,
+      "step": 15039
+    },
+    {
+      "epoch": 3.418181818181818,
+      "grad_norm": 1.7191341738040964,
+      "learning_rate": 2.841765579610471e-07,
+      "loss": 0.0112,
+      "step": 15040
+    },
+    {
+      "epoch": 3.418409090909091,
+      "grad_norm": 1.0548235867179945,
+      "learning_rate": 2.8410174962188757e-07,
+      "loss": 0.0074,
+      "step": 15041
+    },
+    {
+      "epoch": 3.4186363636363635,
+      "grad_norm": 1.8723112990635775,
+      "learning_rate": 2.840269482342379e-07,
+      "loss": 0.0077,
+      "step": 15042
+    },
+    {
+      "epoch": 3.4188636363636364,
+      "grad_norm": 0.9136609812026688,
+      "learning_rate": 2.8395215379962364e-07,
+      "loss": 0.0045,
+      "step": 15043
+    },
+    {
+      "epoch": 3.419090909090909,
+      "grad_norm": 1.8932849807332304,
+      "learning_rate": 2.838773663195699e-07,
+      "loss": 0.01,
+      "step": 15044
+    },
+    {
+      "epoch": 3.419318181818182,
+      "grad_norm": 2.1836953253138356,
+      "learning_rate": 2.8380258579560205e-07,
+      "loss": 0.019,
+      "step": 15045
+    },
+    {
+      "epoch": 3.4195454545454544,
+      "grad_norm": 0.5367604172061377,
+      "learning_rate": 2.8372781222924446e-07,
+      "loss": 0.0024,
+      "step": 15046
+    },
+    {
+      "epoch": 3.4197727272727274,
+      "grad_norm": 0.8177190487587819,
+      "learning_rate": 2.836530456220222e-07,
+      "loss": 0.0043,
+      "step": 15047
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 2.2278047062186763,
+      "learning_rate": 2.835782859754598e-07,
+      "loss": 0.0164,
+      "step": 15048
+    },
+    {
+      "epoch": 3.420227272727273,
+      "grad_norm": 2.278679717243572,
+      "learning_rate": 2.835035332910819e-07,
+      "loss": 0.0131,
+      "step": 15049
+    },
+    {
+      "epoch": 3.4204545454545454,
+      "grad_norm": 2.008382536519693,
+      "learning_rate": 2.8342878757041267e-07,
+      "loss": 0.0062,
+      "step": 15050
+    },
+    {
+      "epoch": 3.4206818181818184,
+      "grad_norm": 2.9743620964460926,
+      "learning_rate": 2.833540488149764e-07,
+      "loss": 0.0234,
+      "step": 15051
+    },
+    {
+      "epoch": 3.420909090909091,
+      "grad_norm": 2.0088930980472957,
+      "learning_rate": 2.8327931702629703e-07,
+      "loss": 0.0121,
+      "step": 15052
+    },
+    {
+      "epoch": 3.4211363636363634,
+      "grad_norm": 1.1814551167819471,
+      "learning_rate": 2.8320459220589855e-07,
+      "loss": 0.0101,
+      "step": 15053
+    },
+    {
+      "epoch": 3.4213636363636364,
+      "grad_norm": 1.3473701040456818,
+      "learning_rate": 2.83129874355305e-07,
+      "loss": 0.0089,
+      "step": 15054
+    },
+    {
+      "epoch": 3.421590909090909,
+      "grad_norm": 1.764546406492666,
+      "learning_rate": 2.8305516347603953e-07,
+      "loss": 0.0101,
+      "step": 15055
+    },
+    {
+      "epoch": 3.421818181818182,
+      "grad_norm": 0.43309955140560547,
+      "learning_rate": 2.8298045956962577e-07,
+      "loss": 0.0015,
+      "step": 15056
+    },
+    {
+      "epoch": 3.4220454545454544,
+      "grad_norm": 2.1488898175172557,
+      "learning_rate": 2.829057626375872e-07,
+      "loss": 0.0162,
+      "step": 15057
+    },
+    {
+      "epoch": 3.4222727272727274,
+      "grad_norm": 0.8537987944247134,
+      "learning_rate": 2.8283107268144713e-07,
+      "loss": 0.0122,
+      "step": 15058
+    },
+    {
+      "epoch": 3.4225,
+      "grad_norm": 1.9584002337338349,
+      "learning_rate": 2.8275638970272823e-07,
+      "loss": 0.015,
+      "step": 15059
+    },
+    {
+      "epoch": 3.422727272727273,
+      "grad_norm": 4.405065074264488,
+      "learning_rate": 2.8268171370295355e-07,
+      "loss": 0.0369,
+      "step": 15060
+    },
+    {
+      "epoch": 3.4229545454545454,
+      "grad_norm": 2.06214083665106,
+      "learning_rate": 2.82607044683646e-07,
+      "loss": 0.0106,
+      "step": 15061
+    },
+    {
+      "epoch": 3.4231818181818183,
+      "grad_norm": 0.6791026477846699,
+      "learning_rate": 2.825323826463281e-07,
+      "loss": 0.0082,
+      "step": 15062
+    },
+    {
+      "epoch": 3.423409090909091,
+      "grad_norm": 1.5234003724899936,
+      "learning_rate": 2.824577275925227e-07,
+      "loss": 0.0063,
+      "step": 15063
+    },
+    {
+      "epoch": 3.423636363636364,
+      "grad_norm": 2.4816738092471247,
+      "learning_rate": 2.823830795237515e-07,
+      "loss": 0.0176,
+      "step": 15064
+    },
+    {
+      "epoch": 3.4238636363636363,
+      "grad_norm": 3.6344864419093548,
+      "learning_rate": 2.8230843844153714e-07,
+      "loss": 0.0317,
+      "step": 15065
+    },
+    {
+      "epoch": 3.424090909090909,
+      "grad_norm": 1.8828519593282549,
+      "learning_rate": 2.822338043474015e-07,
+      "loss": 0.0113,
+      "step": 15066
+    },
+    {
+      "epoch": 3.424318181818182,
+      "grad_norm": 3.7179628165391727,
+      "learning_rate": 2.821591772428666e-07,
+      "loss": 0.009,
+      "step": 15067
+    },
+    {
+      "epoch": 3.424545454545455,
+      "grad_norm": 1.5606180227703925,
+      "learning_rate": 2.820845571294542e-07,
+      "loss": 0.0051,
+      "step": 15068
+    },
+    {
+      "epoch": 3.4247727272727273,
+      "grad_norm": 0.9115448995098834,
+      "learning_rate": 2.8200994400868593e-07,
+      "loss": 0.0055,
+      "step": 15069
+    },
+    {
+      "epoch": 3.425,
+      "grad_norm": 0.9493059452208668,
+      "learning_rate": 2.8193533788208324e-07,
+      "loss": 0.0067,
+      "step": 15070
+    },
+    {
+      "epoch": 3.425227272727273,
+      "grad_norm": 2.8605704057463095,
+      "learning_rate": 2.818607387511676e-07,
+      "loss": 0.0148,
+      "step": 15071
+    },
+    {
+      "epoch": 3.4254545454545453,
+      "grad_norm": 3.482172316722107,
+      "learning_rate": 2.817861466174603e-07,
+      "loss": 0.0238,
+      "step": 15072
+    },
+    {
+      "epoch": 3.4256818181818183,
+      "grad_norm": 1.7199180546374089,
+      "learning_rate": 2.8171156148248203e-07,
+      "loss": 0.0131,
+      "step": 15073
+    },
+    {
+      "epoch": 3.425909090909091,
+      "grad_norm": 1.5634155514707608,
+      "learning_rate": 2.816369833477539e-07,
+      "loss": 0.0108,
+      "step": 15074
+    },
+    {
+      "epoch": 3.4261363636363638,
+      "grad_norm": 1.403275467935487,
+      "learning_rate": 2.8156241221479676e-07,
+      "loss": 0.0074,
+      "step": 15075
+    },
+    {
+      "epoch": 3.4263636363636363,
+      "grad_norm": 1.742787449220099,
+      "learning_rate": 2.8148784808513146e-07,
+      "loss": 0.0179,
+      "step": 15076
+    },
+    {
+      "epoch": 3.4265909090909092,
+      "grad_norm": 2.8533772111679228,
+      "learning_rate": 2.814132909602779e-07,
+      "loss": 0.0265,
+      "step": 15077
+    },
+    {
+      "epoch": 3.4268181818181818,
+      "grad_norm": 2.661818083606212,
+      "learning_rate": 2.8133874084175676e-07,
+      "loss": 0.0073,
+      "step": 15078
+    },
+    {
+      "epoch": 3.4270454545454547,
+      "grad_norm": 2.6675542129613867,
+      "learning_rate": 2.8126419773108827e-07,
+      "loss": 0.012,
+      "step": 15079
+    },
+    {
+      "epoch": 3.4272727272727272,
+      "grad_norm": 0.6436705154780652,
+      "learning_rate": 2.8118966162979253e-07,
+      "loss": 0.0048,
+      "step": 15080
+    },
+    {
+      "epoch": 3.4275,
+      "grad_norm": 2.342258870655518,
+      "learning_rate": 2.811151325393893e-07,
+      "loss": 0.0081,
+      "step": 15081
+    },
+    {
+      "epoch": 3.4277272727272727,
+      "grad_norm": 2.7893101564065965,
+      "learning_rate": 2.810406104613985e-07,
+      "loss": 0.0228,
+      "step": 15082
+    },
+    {
+      "epoch": 3.4279545454545453,
+      "grad_norm": 2.319316266983303,
+      "learning_rate": 2.8096609539733974e-07,
+      "loss": 0.0091,
+      "step": 15083
+    },
+    {
+      "epoch": 3.428181818181818,
+      "grad_norm": 1.5929685299682783,
+      "learning_rate": 2.808915873487325e-07,
+      "loss": 0.007,
+      "step": 15084
+    },
+    {
+      "epoch": 3.4284090909090907,
+      "grad_norm": 1.5399386202527166,
+      "learning_rate": 2.8081708631709627e-07,
+      "loss": 0.0079,
+      "step": 15085
+    },
+    {
+      "epoch": 3.4286363636363637,
+      "grad_norm": 1.8767072894042625,
+      "learning_rate": 2.8074259230395003e-07,
+      "loss": 0.0102,
+      "step": 15086
+    },
+    {
+      "epoch": 3.4288636363636362,
+      "grad_norm": 1.6491844195514114,
+      "learning_rate": 2.8066810531081275e-07,
+      "loss": 0.0126,
+      "step": 15087
+    },
+    {
+      "epoch": 3.429090909090909,
+      "grad_norm": 0.5513189567637762,
+      "learning_rate": 2.805936253392037e-07,
+      "loss": 0.0028,
+      "step": 15088
+    },
+    {
+      "epoch": 3.4293181818181817,
+      "grad_norm": 0.41548363506948516,
+      "learning_rate": 2.8051915239064156e-07,
+      "loss": 0.0039,
+      "step": 15089
+    },
+    {
+      "epoch": 3.4295454545454547,
+      "grad_norm": 0.9781426832961694,
+      "learning_rate": 2.804446864666447e-07,
+      "loss": 0.0046,
+      "step": 15090
+    },
+    {
+      "epoch": 3.429772727272727,
+      "grad_norm": 0.2505008424495744,
+      "learning_rate": 2.803702275687318e-07,
+      "loss": 0.0032,
+      "step": 15091
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 1.215952315493585,
+      "learning_rate": 2.802957756984212e-07,
+      "loss": 0.0128,
+      "step": 15092
+    },
+    {
+      "epoch": 3.4302272727272727,
+      "grad_norm": 1.062665189799689,
+      "learning_rate": 2.802213308572312e-07,
+      "loss": 0.0082,
+      "step": 15093
+    },
+    {
+      "epoch": 3.4304545454545456,
+      "grad_norm": 2.114307010970464,
+      "learning_rate": 2.801468930466798e-07,
+      "loss": 0.0119,
+      "step": 15094
+    },
+    {
+      "epoch": 3.430681818181818,
+      "grad_norm": 1.439254243882297,
+      "learning_rate": 2.800724622682848e-07,
+      "loss": 0.0069,
+      "step": 15095
+    },
+    {
+      "epoch": 3.4309090909090907,
+      "grad_norm": 0.5105482014179579,
+      "learning_rate": 2.79998038523564e-07,
+      "loss": 0.0027,
+      "step": 15096
+    },
+    {
+      "epoch": 3.4311363636363637,
+      "grad_norm": 6.707579606698787,
+      "learning_rate": 2.7992362181403486e-07,
+      "loss": 0.0222,
+      "step": 15097
+    },
+    {
+      "epoch": 3.431363636363636,
+      "grad_norm": 1.77455496734515,
+      "learning_rate": 2.7984921214121554e-07,
+      "loss": 0.015,
+      "step": 15098
+    },
+    {
+      "epoch": 3.431590909090909,
+      "grad_norm": 3.430027015704116,
+      "learning_rate": 2.7977480950662274e-07,
+      "loss": 0.0254,
+      "step": 15099
+    },
+    {
+      "epoch": 3.4318181818181817,
+      "grad_norm": 0.9646932255060904,
+      "learning_rate": 2.797004139117738e-07,
+      "loss": 0.0049,
+      "step": 15100
+    },
+    {
+      "epoch": 3.4320454545454546,
+      "grad_norm": 3.955061548138044,
+      "learning_rate": 2.7962602535818583e-07,
+      "loss": 0.0349,
+      "step": 15101
+    },
+    {
+      "epoch": 3.432272727272727,
+      "grad_norm": 0.3812923624968921,
+      "learning_rate": 2.7955164384737596e-07,
+      "loss": 0.0024,
+      "step": 15102
+    },
+    {
+      "epoch": 3.4325,
+      "grad_norm": 3.1581163197165116,
+      "learning_rate": 2.7947726938086053e-07,
+      "loss": 0.0301,
+      "step": 15103
+    },
+    {
+      "epoch": 3.4327272727272726,
+      "grad_norm": 0.9771023717492855,
+      "learning_rate": 2.7940290196015644e-07,
+      "loss": 0.0057,
+      "step": 15104
+    },
+    {
+      "epoch": 3.4329545454545456,
+      "grad_norm": 1.9886422315386747,
+      "learning_rate": 2.7932854158678e-07,
+      "loss": 0.0185,
+      "step": 15105
+    },
+    {
+      "epoch": 3.433181818181818,
+      "grad_norm": 1.6016468285517254,
+      "learning_rate": 2.792541882622477e-07,
+      "loss": 0.004,
+      "step": 15106
+    },
+    {
+      "epoch": 3.433409090909091,
+      "grad_norm": 1.8028240969864195,
+      "learning_rate": 2.79179841988076e-07,
+      "loss": 0.021,
+      "step": 15107
+    },
+    {
+      "epoch": 3.4336363636363636,
+      "grad_norm": 1.806830577238193,
+      "learning_rate": 2.791055027657803e-07,
+      "loss": 0.009,
+      "step": 15108
+    },
+    {
+      "epoch": 3.433863636363636,
+      "grad_norm": 1.400051508152951,
+      "learning_rate": 2.7903117059687684e-07,
+      "loss": 0.0088,
+      "step": 15109
+    },
+    {
+      "epoch": 3.434090909090909,
+      "grad_norm": 1.1898998689687679,
+      "learning_rate": 2.789568454828814e-07,
+      "loss": 0.0186,
+      "step": 15110
+    },
+    {
+      "epoch": 3.434318181818182,
+      "grad_norm": 0.978745604018448,
+      "learning_rate": 2.7888252742530967e-07,
+      "loss": 0.0082,
+      "step": 15111
+    },
+    {
+      "epoch": 3.4345454545454546,
+      "grad_norm": 1.517441999639122,
+      "learning_rate": 2.7880821642567697e-07,
+      "loss": 0.019,
+      "step": 15112
+    },
+    {
+      "epoch": 3.434772727272727,
+      "grad_norm": 2.329973373728543,
+      "learning_rate": 2.7873391248549876e-07,
+      "loss": 0.0118,
+      "step": 15113
+    },
+    {
+      "epoch": 3.435,
+      "grad_norm": 1.7881886009396364,
+      "learning_rate": 2.7865961560629005e-07,
+      "loss": 0.0119,
+      "step": 15114
+    },
+    {
+      "epoch": 3.4352272727272726,
+      "grad_norm": 2.6141433380876355,
+      "learning_rate": 2.7858532578956615e-07,
+      "loss": 0.0219,
+      "step": 15115
+    },
+    {
+      "epoch": 3.4354545454545455,
+      "grad_norm": 0.5419209618068501,
+      "learning_rate": 2.785110430368419e-07,
+      "loss": 0.0021,
+      "step": 15116
+    },
+    {
+      "epoch": 3.435681818181818,
+      "grad_norm": 1.0267943444180037,
+      "learning_rate": 2.784367673496318e-07,
+      "loss": 0.0061,
+      "step": 15117
+    },
+    {
+      "epoch": 3.435909090909091,
+      "grad_norm": 1.9614285271350453,
+      "learning_rate": 2.7836249872945066e-07,
+      "loss": 0.0087,
+      "step": 15118
+    },
+    {
+      "epoch": 3.4361363636363635,
+      "grad_norm": 2.8158559681926505,
+      "learning_rate": 2.7828823717781286e-07,
+      "loss": 0.0109,
+      "step": 15119
+    },
+    {
+      "epoch": 3.4363636363636365,
+      "grad_norm": 1.7687713031578425,
+      "learning_rate": 2.78213982696233e-07,
+      "loss": 0.0298,
+      "step": 15120
+    },
+    {
+      "epoch": 3.436590909090909,
+      "grad_norm": 1.8812425861403834,
+      "learning_rate": 2.781397352862249e-07,
+      "loss": 0.0122,
+      "step": 15121
+    },
+    {
+      "epoch": 3.436818181818182,
+      "grad_norm": 3.6883176504662982,
+      "learning_rate": 2.780654949493026e-07,
+      "loss": 0.0256,
+      "step": 15122
+    },
+    {
+      "epoch": 3.4370454545454545,
+      "grad_norm": 1.5194484634171752,
+      "learning_rate": 2.779912616869802e-07,
+      "loss": 0.0155,
+      "step": 15123
+    },
+    {
+      "epoch": 3.4372727272727275,
+      "grad_norm": 1.7788209625084819,
+      "learning_rate": 2.779170355007715e-07,
+      "loss": 0.0081,
+      "step": 15124
+    },
+    {
+      "epoch": 3.4375,
+      "grad_norm": 2.344075158179449,
+      "learning_rate": 2.7784281639218986e-07,
+      "loss": 0.0159,
+      "step": 15125
+    },
+    {
+      "epoch": 3.4377272727272725,
+      "grad_norm": 1.4339837278625243,
+      "learning_rate": 2.777686043627488e-07,
+      "loss": 0.0103,
+      "step": 15126
+    },
+    {
+      "epoch": 3.4379545454545455,
+      "grad_norm": 0.3083665249903847,
+      "learning_rate": 2.7769439941396143e-07,
+      "loss": 0.0014,
+      "step": 15127
+    },
+    {
+      "epoch": 3.438181818181818,
+      "grad_norm": 3.6860657877542424,
+      "learning_rate": 2.7762020154734137e-07,
+      "loss": 0.0293,
+      "step": 15128
+    },
+    {
+      "epoch": 3.438409090909091,
+      "grad_norm": 1.9096416501895903,
+      "learning_rate": 2.7754601076440173e-07,
+      "loss": 0.0217,
+      "step": 15129
+    },
+    {
+      "epoch": 3.4386363636363635,
+      "grad_norm": 1.3218627686080662,
+      "learning_rate": 2.774718270666548e-07,
+      "loss": 0.0046,
+      "step": 15130
+    },
+    {
+      "epoch": 3.4388636363636365,
+      "grad_norm": 0.7085398646550584,
+      "learning_rate": 2.773976504556137e-07,
+      "loss": 0.0045,
+      "step": 15131
+    },
+    {
+      "epoch": 3.439090909090909,
+      "grad_norm": 1.0335164450123961,
+      "learning_rate": 2.7732348093279087e-07,
+      "loss": 0.0045,
+      "step": 15132
+    },
+    {
+      "epoch": 3.439318181818182,
+      "grad_norm": 2.5076675862862334,
+      "learning_rate": 2.77249318499699e-07,
+      "loss": 0.0236,
+      "step": 15133
+    },
+    {
+      "epoch": 3.4395454545454545,
+      "grad_norm": 1.2374840252023862,
+      "learning_rate": 2.7717516315785007e-07,
+      "loss": 0.0063,
+      "step": 15134
+    },
+    {
+      "epoch": 3.4397727272727274,
+      "grad_norm": 1.7192077142722053,
+      "learning_rate": 2.7710101490875633e-07,
+      "loss": 0.0127,
+      "step": 15135
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 2.101679005167337,
+      "learning_rate": 2.7702687375392986e-07,
+      "loss": 0.0246,
+      "step": 15136
+    },
+    {
+      "epoch": 3.440227272727273,
+      "grad_norm": 2.5230240530989954,
+      "learning_rate": 2.769527396948828e-07,
+      "loss": 0.017,
+      "step": 15137
+    },
+    {
+      "epoch": 3.4404545454545454,
+      "grad_norm": 1.1629150240526993,
+      "learning_rate": 2.768786127331262e-07,
+      "loss": 0.0039,
+      "step": 15138
+    },
+    {
+      "epoch": 3.440681818181818,
+      "grad_norm": 1.8930501871839531,
+      "learning_rate": 2.768044928701721e-07,
+      "loss": 0.0351,
+      "step": 15139
+    },
+    {
+      "epoch": 3.440909090909091,
+      "grad_norm": 1.7489414479964223,
+      "learning_rate": 2.767303801075318e-07,
+      "loss": 0.0112,
+      "step": 15140
+    },
+    {
+      "epoch": 3.441136363636364,
+      "grad_norm": 0.6229697031265881,
+      "learning_rate": 2.7665627444671673e-07,
+      "loss": 0.0087,
+      "step": 15141
+    },
+    {
+      "epoch": 3.4413636363636364,
+      "grad_norm": 0.9836413267413662,
+      "learning_rate": 2.765821758892379e-07,
+      "loss": 0.0048,
+      "step": 15142
+    },
+    {
+      "epoch": 3.441590909090909,
+      "grad_norm": 2.3099990223762012,
+      "learning_rate": 2.765080844366064e-07,
+      "loss": 0.0228,
+      "step": 15143
+    },
+    {
+      "epoch": 3.441818181818182,
+      "grad_norm": 0.44256493156096094,
+      "learning_rate": 2.7643400009033294e-07,
+      "loss": 0.0027,
+      "step": 15144
+    },
+    {
+      "epoch": 3.4420454545454544,
+      "grad_norm": 1.4943186484538575,
+      "learning_rate": 2.763599228519283e-07,
+      "loss": 0.009,
+      "step": 15145
+    },
+    {
+      "epoch": 3.4422727272727274,
+      "grad_norm": 3.9667239417932194,
+      "learning_rate": 2.7628585272290343e-07,
+      "loss": 0.0394,
+      "step": 15146
+    },
+    {
+      "epoch": 3.4425,
+      "grad_norm": 1.4344086583636155,
+      "learning_rate": 2.762117897047681e-07,
+      "loss": 0.0152,
+      "step": 15147
+    },
+    {
+      "epoch": 3.442727272727273,
+      "grad_norm": 0.9062798627766027,
+      "learning_rate": 2.761377337990329e-07,
+      "loss": 0.0053,
+      "step": 15148
+    },
+    {
+      "epoch": 3.4429545454545454,
+      "grad_norm": 1.524696315528924,
+      "learning_rate": 2.7606368500720785e-07,
+      "loss": 0.0151,
+      "step": 15149
+    },
+    {
+      "epoch": 3.4431818181818183,
+      "grad_norm": 2.339587084957424,
+      "learning_rate": 2.759896433308031e-07,
+      "loss": 0.0235,
+      "step": 15150
+    },
+    {
+      "epoch": 3.443409090909091,
+      "grad_norm": 1.9153822672923317,
+      "learning_rate": 2.7591560877132864e-07,
+      "loss": 0.0106,
+      "step": 15151
+    },
+    {
+      "epoch": 3.443636363636364,
+      "grad_norm": 1.3234303347983996,
+      "learning_rate": 2.7584158133029365e-07,
+      "loss": 0.0092,
+      "step": 15152
+    },
+    {
+      "epoch": 3.4438636363636363,
+      "grad_norm": 1.6325897208924136,
+      "learning_rate": 2.7576756100920797e-07,
+      "loss": 0.0129,
+      "step": 15153
+    },
+    {
+      "epoch": 3.4440909090909093,
+      "grad_norm": 1.7154124706601348,
+      "learning_rate": 2.7569354780958097e-07,
+      "loss": 0.0257,
+      "step": 15154
+    },
+    {
+      "epoch": 3.444318181818182,
+      "grad_norm": 1.1551721292670247,
+      "learning_rate": 2.7561954173292216e-07,
+      "loss": 0.0089,
+      "step": 15155
+    },
+    {
+      "epoch": 3.4445454545454544,
+      "grad_norm": 0.7998945154065505,
+      "learning_rate": 2.7554554278074e-07,
+      "loss": 0.0043,
+      "step": 15156
+    },
+    {
+      "epoch": 3.4447727272727273,
+      "grad_norm": 0.45030747289269185,
+      "learning_rate": 2.7547155095454406e-07,
+      "loss": 0.0048,
+      "step": 15157
+    },
+    {
+      "epoch": 3.445,
+      "grad_norm": 1.7706665364034908,
+      "learning_rate": 2.75397566255843e-07,
+      "loss": 0.0146,
+      "step": 15158
+    },
+    {
+      "epoch": 3.445227272727273,
+      "grad_norm": 0.5415920039620339,
+      "learning_rate": 2.753235886861457e-07,
+      "loss": 0.0027,
+      "step": 15159
+    },
+    {
+      "epoch": 3.4454545454545453,
+      "grad_norm": 0.8730179736898658,
+      "learning_rate": 2.752496182469603e-07,
+      "loss": 0.0033,
+      "step": 15160
+    },
+    {
+      "epoch": 3.4456818181818183,
+      "grad_norm": 5.114356716485947,
+      "learning_rate": 2.7517565493979535e-07,
+      "loss": 0.0201,
+      "step": 15161
+    },
+    {
+      "epoch": 3.445909090909091,
+      "grad_norm": 0.32732026143648735,
+      "learning_rate": 2.75101698766159e-07,
+      "loss": 0.0013,
+      "step": 15162
+    },
+    {
+      "epoch": 3.4461363636363638,
+      "grad_norm": 0.8853516539511918,
+      "learning_rate": 2.7502774972755957e-07,
+      "loss": 0.0048,
+      "step": 15163
+    },
+    {
+      "epoch": 3.4463636363636363,
+      "grad_norm": 2.589846102914889,
+      "learning_rate": 2.749538078255051e-07,
+      "loss": 0.0569,
+      "step": 15164
+    },
+    {
+      "epoch": 3.4465909090909093,
+      "grad_norm": 1.0039288532153892,
+      "learning_rate": 2.748798730615028e-07,
+      "loss": 0.0122,
+      "step": 15165
+    },
+    {
+      "epoch": 3.446818181818182,
+      "grad_norm": 1.8842818971163924,
+      "learning_rate": 2.748059454370608e-07,
+      "loss": 0.0134,
+      "step": 15166
+    },
+    {
+      "epoch": 3.4470454545454547,
+      "grad_norm": 1.0035529313297835,
+      "learning_rate": 2.7473202495368653e-07,
+      "loss": 0.0122,
+      "step": 15167
+    },
+    {
+      "epoch": 3.4472727272727273,
+      "grad_norm": 0.45097056935156066,
+      "learning_rate": 2.746581116128876e-07,
+      "loss": 0.0025,
+      "step": 15168
+    },
+    {
+      "epoch": 3.4475,
+      "grad_norm": 0.897302446304034,
+      "learning_rate": 2.745842054161707e-07,
+      "loss": 0.0053,
+      "step": 15169
+    },
+    {
+      "epoch": 3.4477272727272728,
+      "grad_norm": 1.4731052578553225,
+      "learning_rate": 2.7451030636504323e-07,
+      "loss": 0.0131,
+      "step": 15170
+    },
+    {
+      "epoch": 3.4479545454545453,
+      "grad_norm": 1.0090759395069937,
+      "learning_rate": 2.7443641446101206e-07,
+      "loss": 0.0074,
+      "step": 15171
+    },
+    {
+      "epoch": 3.4481818181818182,
+      "grad_norm": 4.1590083703261715,
+      "learning_rate": 2.7436252970558394e-07,
+      "loss": 0.0208,
+      "step": 15172
+    },
+    {
+      "epoch": 3.4484090909090908,
+      "grad_norm": 2.374437584559937,
+      "learning_rate": 2.7428865210026566e-07,
+      "loss": 0.0095,
+      "step": 15173
+    },
+    {
+      "epoch": 3.4486363636363637,
+      "grad_norm": 0.8801799246235704,
+      "learning_rate": 2.7421478164656354e-07,
+      "loss": 0.0062,
+      "step": 15174
+    },
+    {
+      "epoch": 3.4488636363636362,
+      "grad_norm": 0.7361863897667754,
+      "learning_rate": 2.74140918345984e-07,
+      "loss": 0.005,
+      "step": 15175
+    },
+    {
+      "epoch": 3.449090909090909,
+      "grad_norm": 3.7177867058727045,
+      "learning_rate": 2.7406706220003324e-07,
+      "loss": 0.0173,
+      "step": 15176
+    },
+    {
+      "epoch": 3.4493181818181817,
+      "grad_norm": 2.6580401762777823,
+      "learning_rate": 2.739932132102176e-07,
+      "loss": 0.029,
+      "step": 15177
+    },
+    {
+      "epoch": 3.4495454545454547,
+      "grad_norm": 1.6758476473908481,
+      "learning_rate": 2.739193713780424e-07,
+      "loss": 0.012,
+      "step": 15178
+    },
+    {
+      "epoch": 3.449772727272727,
+      "grad_norm": 1.5341803595874788,
+      "learning_rate": 2.738455367050138e-07,
+      "loss": 0.0067,
+      "step": 15179
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 1.4671320387474736,
+      "learning_rate": 2.737717091926373e-07,
+      "loss": 0.0104,
+      "step": 15180
+    },
+    {
+      "epoch": 3.4502272727272727,
+      "grad_norm": 0.9227966116894034,
+      "learning_rate": 2.7369788884241856e-07,
+      "loss": 0.0054,
+      "step": 15181
+    },
+    {
+      "epoch": 3.450454545454545,
+      "grad_norm": 0.8557259870578243,
+      "learning_rate": 2.736240756558626e-07,
+      "loss": 0.0045,
+      "step": 15182
+    },
+    {
+      "epoch": 3.450681818181818,
+      "grad_norm": 1.6749875799293477,
+      "learning_rate": 2.735502696344747e-07,
+      "loss": 0.0045,
+      "step": 15183
+    },
+    {
+      "epoch": 3.450909090909091,
+      "grad_norm": 2.121781852225656,
+      "learning_rate": 2.7347647077976e-07,
+      "loss": 0.0165,
+      "step": 15184
+    },
+    {
+      "epoch": 3.4511363636363637,
+      "grad_norm": 2.2320377722824554,
+      "learning_rate": 2.734026790932232e-07,
+      "loss": 0.0224,
+      "step": 15185
+    },
+    {
+      "epoch": 3.451363636363636,
+      "grad_norm": 1.4806326821431062,
+      "learning_rate": 2.7332889457636927e-07,
+      "loss": 0.0104,
+      "step": 15186
+    },
+    {
+      "epoch": 3.451590909090909,
+      "grad_norm": 1.5805431275564767,
+      "learning_rate": 2.732551172307027e-07,
+      "loss": 0.0095,
+      "step": 15187
+    },
+    {
+      "epoch": 3.4518181818181817,
+      "grad_norm": 1.313131845990157,
+      "learning_rate": 2.7318134705772797e-07,
+      "loss": 0.0051,
+      "step": 15188
+    },
+    {
+      "epoch": 3.4520454545454546,
+      "grad_norm": 0.6907200377993076,
+      "learning_rate": 2.7310758405894927e-07,
+      "loss": 0.0036,
+      "step": 15189
+    },
+    {
+      "epoch": 3.452272727272727,
+      "grad_norm": 2.8287800794202385,
+      "learning_rate": 2.730338282358711e-07,
+      "loss": 0.0247,
+      "step": 15190
+    },
+    {
+      "epoch": 3.4525,
+      "grad_norm": 2.792479151413951,
+      "learning_rate": 2.7296007958999704e-07,
+      "loss": 0.0132,
+      "step": 15191
+    },
+    {
+      "epoch": 3.4527272727272726,
+      "grad_norm": 1.075983477293136,
+      "learning_rate": 2.728863381228311e-07,
+      "loss": 0.0065,
+      "step": 15192
+    },
+    {
+      "epoch": 3.4529545454545456,
+      "grad_norm": 1.6879573814943876,
+      "learning_rate": 2.72812603835877e-07,
+      "loss": 0.0149,
+      "step": 15193
+    },
+    {
+      "epoch": 3.453181818181818,
+      "grad_norm": 3.7356937712760843,
+      "learning_rate": 2.7273887673063833e-07,
+      "loss": 0.0157,
+      "step": 15194
+    },
+    {
+      "epoch": 3.453409090909091,
+      "grad_norm": 0.8409975080173838,
+      "learning_rate": 2.7266515680861876e-07,
+      "loss": 0.0066,
+      "step": 15195
+    },
+    {
+      "epoch": 3.4536363636363636,
+      "grad_norm": 1.0826145402783254,
+      "learning_rate": 2.7259144407132105e-07,
+      "loss": 0.0068,
+      "step": 15196
+    },
+    {
+      "epoch": 3.4538636363636366,
+      "grad_norm": 1.945678579470229,
+      "learning_rate": 2.7251773852024865e-07,
+      "loss": 0.0227,
+      "step": 15197
+    },
+    {
+      "epoch": 3.454090909090909,
+      "grad_norm": 0.9363673121608694,
+      "learning_rate": 2.724440401569045e-07,
+      "loss": 0.0043,
+      "step": 15198
+    },
+    {
+      "epoch": 3.4543181818181816,
+      "grad_norm": 2.084400870090571,
+      "learning_rate": 2.7237034898279167e-07,
+      "loss": 0.0188,
+      "step": 15199
+    },
+    {
+      "epoch": 3.4545454545454546,
+      "grad_norm": 1.2812248868173097,
+      "learning_rate": 2.7229666499941237e-07,
+      "loss": 0.0152,
+      "step": 15200
+    },
+    {
+      "epoch": 3.454772727272727,
+      "grad_norm": 1.3172582491952538,
+      "learning_rate": 2.7222298820826944e-07,
+      "loss": 0.0122,
+      "step": 15201
+    },
+    {
+      "epoch": 3.455,
+      "grad_norm": 0.6578364780365312,
+      "learning_rate": 2.7214931861086515e-07,
+      "loss": 0.0025,
+      "step": 15202
+    },
+    {
+      "epoch": 3.4552272727272726,
+      "grad_norm": 0.6704033833921734,
+      "learning_rate": 2.7207565620870193e-07,
+      "loss": 0.005,
+      "step": 15203
+    },
+    {
+      "epoch": 3.4554545454545456,
+      "grad_norm": 1.8519098493222248,
+      "learning_rate": 2.7200200100328173e-07,
+      "loss": 0.009,
+      "step": 15204
+    },
+    {
+      "epoch": 3.455681818181818,
+      "grad_norm": 0.4995950666388443,
+      "learning_rate": 2.7192835299610654e-07,
+      "loss": 0.002,
+      "step": 15205
+    },
+    {
+      "epoch": 3.455909090909091,
+      "grad_norm": 1.560864020727417,
+      "learning_rate": 2.718547121886783e-07,
+      "loss": 0.0108,
+      "step": 15206
+    },
+    {
+      "epoch": 3.4561363636363636,
+      "grad_norm": 0.831310986371504,
+      "learning_rate": 2.717810785824985e-07,
+      "loss": 0.0048,
+      "step": 15207
+    },
+    {
+      "epoch": 3.4563636363636365,
+      "grad_norm": 0.52797286300455,
+      "learning_rate": 2.7170745217906897e-07,
+      "loss": 0.0024,
+      "step": 15208
+    },
+    {
+      "epoch": 3.456590909090909,
+      "grad_norm": 0.7748145427587947,
+      "learning_rate": 2.7163383297989063e-07,
+      "loss": 0.0043,
+      "step": 15209
+    },
+    {
+      "epoch": 3.456818181818182,
+      "grad_norm": 4.081477277967676,
+      "learning_rate": 2.7156022098646493e-07,
+      "loss": 0.0112,
+      "step": 15210
+    },
+    {
+      "epoch": 3.4570454545454545,
+      "grad_norm": 2.8659782761279406,
+      "learning_rate": 2.714866162002929e-07,
+      "loss": 0.0318,
+      "step": 15211
+    },
+    {
+      "epoch": 3.457272727272727,
+      "grad_norm": 3.0564616541524248,
+      "learning_rate": 2.7141301862287575e-07,
+      "loss": 0.0414,
+      "step": 15212
+    },
+    {
+      "epoch": 3.4575,
+      "grad_norm": 0.7861693954261957,
+      "learning_rate": 2.713394282557138e-07,
+      "loss": 0.0032,
+      "step": 15213
+    },
+    {
+      "epoch": 3.4577272727272725,
+      "grad_norm": 0.7144991293172988,
+      "learning_rate": 2.7126584510030785e-07,
+      "loss": 0.0037,
+      "step": 15214
+    },
+    {
+      "epoch": 3.4579545454545455,
+      "grad_norm": 1.5410646545507927,
+      "learning_rate": 2.7119226915815853e-07,
+      "loss": 0.0115,
+      "step": 15215
+    },
+    {
+      "epoch": 3.458181818181818,
+      "grad_norm": 1.1775476562541822,
+      "learning_rate": 2.71118700430766e-07,
+      "loss": 0.0068,
+      "step": 15216
+    },
+    {
+      "epoch": 3.458409090909091,
+      "grad_norm": 1.0788974658449073,
+      "learning_rate": 2.710451389196306e-07,
+      "loss": 0.0045,
+      "step": 15217
+    },
+    {
+      "epoch": 3.4586363636363635,
+      "grad_norm": 1.5851248214252496,
+      "learning_rate": 2.709715846262523e-07,
+      "loss": 0.0099,
+      "step": 15218
+    },
+    {
+      "epoch": 3.4588636363636365,
+      "grad_norm": 2.202162690049556,
+      "learning_rate": 2.7089803755213106e-07,
+      "loss": 0.0178,
+      "step": 15219
+    },
+    {
+      "epoch": 3.459090909090909,
+      "grad_norm": 1.1852290188170116,
+      "learning_rate": 2.708244976987666e-07,
+      "loss": 0.0102,
+      "step": 15220
+    },
+    {
+      "epoch": 3.459318181818182,
+      "grad_norm": 1.469537399953894,
+      "learning_rate": 2.707509650676587e-07,
+      "loss": 0.0056,
+      "step": 15221
+    },
+    {
+      "epoch": 3.4595454545454545,
+      "grad_norm": 1.7719268685851834,
+      "learning_rate": 2.7067743966030645e-07,
+      "loss": 0.0063,
+      "step": 15222
+    },
+    {
+      "epoch": 3.4597727272727274,
+      "grad_norm": 2.4116546406823285,
+      "learning_rate": 2.706039214782093e-07,
+      "loss": 0.0202,
+      "step": 15223
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 1.0384973265072575,
+      "learning_rate": 2.705304105228666e-07,
+      "loss": 0.0067,
+      "step": 15224
+    },
+    {
+      "epoch": 3.4602272727272725,
+      "grad_norm": 1.7115811415670512,
+      "learning_rate": 2.7045690679577734e-07,
+      "loss": 0.0101,
+      "step": 15225
+    },
+    {
+      "epoch": 3.4604545454545454,
+      "grad_norm": 1.858219755009888,
+      "learning_rate": 2.703834102984401e-07,
+      "loss": 0.0142,
+      "step": 15226
+    },
+    {
+      "epoch": 3.4606818181818184,
+      "grad_norm": 1.640585074513842,
+      "learning_rate": 2.703099210323538e-07,
+      "loss": 0.0087,
+      "step": 15227
+    },
+    {
+      "epoch": 3.460909090909091,
+      "grad_norm": 1.3905410515897194,
+      "learning_rate": 2.7023643899901703e-07,
+      "loss": 0.0111,
+      "step": 15228
+    },
+    {
+      "epoch": 3.4611363636363635,
+      "grad_norm": 0.6467232037772791,
+      "learning_rate": 2.7016296419992813e-07,
+      "loss": 0.0025,
+      "step": 15229
+    },
+    {
+      "epoch": 3.4613636363636364,
+      "grad_norm": 2.112392588918475,
+      "learning_rate": 2.7008949663658563e-07,
+      "loss": 0.0214,
+      "step": 15230
+    },
+    {
+      "epoch": 3.461590909090909,
+      "grad_norm": 1.0882971005283872,
+      "learning_rate": 2.7001603631048734e-07,
+      "loss": 0.0069,
+      "step": 15231
+    },
+    {
+      "epoch": 3.461818181818182,
+      "grad_norm": 0.8911101385446596,
+      "learning_rate": 2.699425832231313e-07,
+      "loss": 0.0062,
+      "step": 15232
+    },
+    {
+      "epoch": 3.4620454545454544,
+      "grad_norm": 0.8489714137272328,
+      "learning_rate": 2.6986913737601546e-07,
+      "loss": 0.0039,
+      "step": 15233
+    },
+    {
+      "epoch": 3.4622727272727274,
+      "grad_norm": 0.6131994237714499,
+      "learning_rate": 2.6979569877063746e-07,
+      "loss": 0.02,
+      "step": 15234
+    },
+    {
+      "epoch": 3.4625,
+      "grad_norm": 2.0233463801380234,
+      "learning_rate": 2.6972226740849485e-07,
+      "loss": 0.0103,
+      "step": 15235
+    },
+    {
+      "epoch": 3.462727272727273,
+      "grad_norm": 0.547840801050556,
+      "learning_rate": 2.69648843291085e-07,
+      "loss": 0.0034,
+      "step": 15236
+    },
+    {
+      "epoch": 3.4629545454545454,
+      "grad_norm": 1.7449355070173216,
+      "learning_rate": 2.695754264199052e-07,
+      "loss": 0.0127,
+      "step": 15237
+    },
+    {
+      "epoch": 3.4631818181818184,
+      "grad_norm": 0.7435971469818843,
+      "learning_rate": 2.695020167964525e-07,
+      "loss": 0.0066,
+      "step": 15238
+    },
+    {
+      "epoch": 3.463409090909091,
+      "grad_norm": 2.04899631194605,
+      "learning_rate": 2.6942861442222417e-07,
+      "loss": 0.0139,
+      "step": 15239
+    },
+    {
+      "epoch": 3.463636363636364,
+      "grad_norm": 2.977974396228985,
+      "learning_rate": 2.6935521929871643e-07,
+      "loss": 0.028,
+      "step": 15240
+    },
+    {
+      "epoch": 3.4638636363636364,
+      "grad_norm": 1.643473986418778,
+      "learning_rate": 2.692818314274263e-07,
+      "loss": 0.0234,
+      "step": 15241
+    },
+    {
+      "epoch": 3.464090909090909,
+      "grad_norm": 2.9138176465917143,
+      "learning_rate": 2.6920845080985016e-07,
+      "loss": 0.0268,
+      "step": 15242
+    },
+    {
+      "epoch": 3.464318181818182,
+      "grad_norm": 1.4432308480929055,
+      "learning_rate": 2.691350774474846e-07,
+      "loss": 0.022,
+      "step": 15243
+    },
+    {
+      "epoch": 3.4645454545454544,
+      "grad_norm": 1.4708990393512493,
+      "learning_rate": 2.690617113418255e-07,
+      "loss": 0.0082,
+      "step": 15244
+    },
+    {
+      "epoch": 3.4647727272727273,
+      "grad_norm": 1.6311232804945859,
+      "learning_rate": 2.6898835249436907e-07,
+      "loss": 0.011,
+      "step": 15245
+    },
+    {
+      "epoch": 3.465,
+      "grad_norm": 2.0814173550658905,
+      "learning_rate": 2.689150009066112e-07,
+      "loss": 0.0109,
+      "step": 15246
+    },
+    {
+      "epoch": 3.465227272727273,
+      "grad_norm": 2.2543704091799857,
+      "learning_rate": 2.688416565800477e-07,
+      "loss": 0.025,
+      "step": 15247
+    },
+    {
+      "epoch": 3.4654545454545453,
+      "grad_norm": 0.9814542800113094,
+      "learning_rate": 2.687683195161742e-07,
+      "loss": 0.0062,
+      "step": 15248
+    },
+    {
+      "epoch": 3.4656818181818183,
+      "grad_norm": 1.0360382065257538,
+      "learning_rate": 2.686949897164861e-07,
+      "loss": 0.0104,
+      "step": 15249
+    },
+    {
+      "epoch": 3.465909090909091,
+      "grad_norm": 0.6432741965893727,
+      "learning_rate": 2.6862166718247873e-07,
+      "loss": 0.0031,
+      "step": 15250
+    },
+    {
+      "epoch": 3.466136363636364,
+      "grad_norm": 2.1031943016820587,
+      "learning_rate": 2.685483519156474e-07,
+      "loss": 0.0285,
+      "step": 15251
+    },
+    {
+      "epoch": 3.4663636363636363,
+      "grad_norm": 1.3430946058863644,
+      "learning_rate": 2.684750439174872e-07,
+      "loss": 0.0061,
+      "step": 15252
+    },
+    {
+      "epoch": 3.4665909090909093,
+      "grad_norm": 1.515791135819858,
+      "learning_rate": 2.684017431894926e-07,
+      "loss": 0.0105,
+      "step": 15253
+    },
+    {
+      "epoch": 3.466818181818182,
+      "grad_norm": 1.6144121108757254,
+      "learning_rate": 2.6832844973315864e-07,
+      "loss": 0.0152,
+      "step": 15254
+    },
+    {
+      "epoch": 3.4670454545454543,
+      "grad_norm": 2.198554793203509,
+      "learning_rate": 2.682551635499798e-07,
+      "loss": 0.0093,
+      "step": 15255
+    },
+    {
+      "epoch": 3.4672727272727273,
+      "grad_norm": 1.2639305200972224,
+      "learning_rate": 2.6818188464145076e-07,
+      "loss": 0.0149,
+      "step": 15256
+    },
+    {
+      "epoch": 3.4675000000000002,
+      "grad_norm": 1.3899039075015471,
+      "learning_rate": 2.6810861300906546e-07,
+      "loss": 0.0051,
+      "step": 15257
+    },
+    {
+      "epoch": 3.4677272727272728,
+      "grad_norm": 0.5516322178928612,
+      "learning_rate": 2.680353486543181e-07,
+      "loss": 0.0031,
+      "step": 15258
+    },
+    {
+      "epoch": 3.4679545454545453,
+      "grad_norm": 1.7915538780764144,
+      "learning_rate": 2.679620915787027e-07,
+      "loss": 0.011,
+      "step": 15259
+    },
+    {
+      "epoch": 3.4681818181818183,
+      "grad_norm": 1.340037571938558,
+      "learning_rate": 2.6788884178371345e-07,
+      "loss": 0.0084,
+      "step": 15260
+    },
+    {
+      "epoch": 3.4684090909090908,
+      "grad_norm": 1.029574828863245,
+      "learning_rate": 2.6781559927084344e-07,
+      "loss": 0.0052,
+      "step": 15261
+    },
+    {
+      "epoch": 3.4686363636363637,
+      "grad_norm": 0.7430135870407547,
+      "learning_rate": 2.6774236404158655e-07,
+      "loss": 0.0051,
+      "step": 15262
+    },
+    {
+      "epoch": 3.4688636363636363,
+      "grad_norm": 1.8414126100634283,
+      "learning_rate": 2.676691360974361e-07,
+      "loss": 0.0219,
+      "step": 15263
+    },
+    {
+      "epoch": 3.4690909090909092,
+      "grad_norm": 2.358670406019042,
+      "learning_rate": 2.675959154398854e-07,
+      "loss": 0.017,
+      "step": 15264
+    },
+    {
+      "epoch": 3.4693181818181817,
+      "grad_norm": 3.0772328028613454,
+      "learning_rate": 2.6752270207042746e-07,
+      "loss": 0.0262,
+      "step": 15265
+    },
+    {
+      "epoch": 3.4695454545454547,
+      "grad_norm": 1.2486305616003854,
+      "learning_rate": 2.6744949599055527e-07,
+      "loss": 0.0091,
+      "step": 15266
+    },
+    {
+      "epoch": 3.4697727272727272,
+      "grad_norm": 0.7743169589367747,
+      "learning_rate": 2.673762972017617e-07,
+      "loss": 0.0091,
+      "step": 15267
+    },
+    {
+      "epoch": 3.4699999999999998,
+      "grad_norm": 1.3863694060498533,
+      "learning_rate": 2.6730310570553933e-07,
+      "loss": 0.0105,
+      "step": 15268
+    },
+    {
+      "epoch": 3.4702272727272727,
+      "grad_norm": 4.210946027078112,
+      "learning_rate": 2.6722992150338085e-07,
+      "loss": 0.0424,
+      "step": 15269
+    },
+    {
+      "epoch": 3.4704545454545457,
+      "grad_norm": 2.4975458560439163,
+      "learning_rate": 2.671567445967782e-07,
+      "loss": 0.015,
+      "step": 15270
+    },
+    {
+      "epoch": 3.470681818181818,
+      "grad_norm": 3.0378768696188456,
+      "learning_rate": 2.6708357498722386e-07,
+      "loss": 0.0336,
+      "step": 15271
+    },
+    {
+      "epoch": 3.4709090909090907,
+      "grad_norm": 0.7296062536847562,
+      "learning_rate": 2.6701041267620987e-07,
+      "loss": 0.003,
+      "step": 15272
+    },
+    {
+      "epoch": 3.4711363636363637,
+      "grad_norm": 1.053545779415203,
+      "learning_rate": 2.6693725766522805e-07,
+      "loss": 0.0109,
+      "step": 15273
+    },
+    {
+      "epoch": 3.471363636363636,
+      "grad_norm": 0.4916887805256936,
+      "learning_rate": 2.6686410995577046e-07,
+      "loss": 0.0022,
+      "step": 15274
+    },
+    {
+      "epoch": 3.471590909090909,
+      "grad_norm": 5.1924346476493435,
+      "learning_rate": 2.667909695493283e-07,
+      "loss": 0.0203,
+      "step": 15275
+    },
+    {
+      "epoch": 3.4718181818181817,
+      "grad_norm": 1.6940299209147613,
+      "learning_rate": 2.6671783644739315e-07,
+      "loss": 0.012,
+      "step": 15276
+    },
+    {
+      "epoch": 3.4720454545454547,
+      "grad_norm": 2.257915359935807,
+      "learning_rate": 2.6664471065145637e-07,
+      "loss": 0.0119,
+      "step": 15277
+    },
+    {
+      "epoch": 3.472272727272727,
+      "grad_norm": 0.7583596393528377,
+      "learning_rate": 2.6657159216300915e-07,
+      "loss": 0.0031,
+      "step": 15278
+    },
+    {
+      "epoch": 3.4725,
+      "grad_norm": 1.1363312151785787,
+      "learning_rate": 2.664984809835425e-07,
+      "loss": 0.0075,
+      "step": 15279
+    },
+    {
+      "epoch": 3.4727272727272727,
+      "grad_norm": 1.3256568315499628,
+      "learning_rate": 2.6642537711454717e-07,
+      "loss": 0.008,
+      "step": 15280
+    },
+    {
+      "epoch": 3.4729545454545456,
+      "grad_norm": 1.287119366504297,
+      "learning_rate": 2.663522805575141e-07,
+      "loss": 0.0165,
+      "step": 15281
+    },
+    {
+      "epoch": 3.473181818181818,
+      "grad_norm": 1.3929023336909618,
+      "learning_rate": 2.6627919131393387e-07,
+      "loss": 0.0045,
+      "step": 15282
+    },
+    {
+      "epoch": 3.473409090909091,
+      "grad_norm": 1.9267636502090206,
+      "learning_rate": 2.6620610938529646e-07,
+      "loss": 0.0097,
+      "step": 15283
+    },
+    {
+      "epoch": 3.4736363636363636,
+      "grad_norm": 1.1119856205609226,
+      "learning_rate": 2.6613303477309254e-07,
+      "loss": 0.0072,
+      "step": 15284
+    },
+    {
+      "epoch": 3.473863636363636,
+      "grad_norm": 1.3818283896223806,
+      "learning_rate": 2.660599674788121e-07,
+      "loss": 0.0126,
+      "step": 15285
+    },
+    {
+      "epoch": 3.474090909090909,
+      "grad_norm": 2.4715019822431925,
+      "learning_rate": 2.6598690750394513e-07,
+      "loss": 0.0124,
+      "step": 15286
+    },
+    {
+      "epoch": 3.4743181818181816,
+      "grad_norm": 1.8003958243122387,
+      "learning_rate": 2.6591385484998166e-07,
+      "loss": 0.0134,
+      "step": 15287
+    },
+    {
+      "epoch": 3.4745454545454546,
+      "grad_norm": 0.9393434991536128,
+      "learning_rate": 2.658408095184109e-07,
+      "loss": 0.0063,
+      "step": 15288
+    },
+    {
+      "epoch": 3.474772727272727,
+      "grad_norm": 2.4494326327634175,
+      "learning_rate": 2.6576777151072264e-07,
+      "loss": 0.0202,
+      "step": 15289
+    },
+    {
+      "epoch": 3.475,
+      "grad_norm": 2.0408035236788082,
+      "learning_rate": 2.656947408284062e-07,
+      "loss": 0.0167,
+      "step": 15290
+    },
+    {
+      "epoch": 3.4752272727272726,
+      "grad_norm": 0.9481107555125021,
+      "learning_rate": 2.656217174729511e-07,
+      "loss": 0.0049,
+      "step": 15291
+    },
+    {
+      "epoch": 3.4754545454545456,
+      "grad_norm": 2.9569806419273985,
+      "learning_rate": 2.65548701445846e-07,
+      "loss": 0.027,
+      "step": 15292
+    },
+    {
+      "epoch": 3.475681818181818,
+      "grad_norm": 1.1970199554757495,
+      "learning_rate": 2.654756927485799e-07,
+      "loss": 0.0076,
+      "step": 15293
+    },
+    {
+      "epoch": 3.475909090909091,
+      "grad_norm": 2.0119115962922347,
+      "learning_rate": 2.6540269138264137e-07,
+      "loss": 0.0145,
+      "step": 15294
+    },
+    {
+      "epoch": 3.4761363636363636,
+      "grad_norm": 1.6584596747596756,
+      "learning_rate": 2.653296973495196e-07,
+      "loss": 0.0151,
+      "step": 15295
+    },
+    {
+      "epoch": 3.4763636363636365,
+      "grad_norm": 1.089288775990237,
+      "learning_rate": 2.65256710650703e-07,
+      "loss": 0.0045,
+      "step": 15296
+    },
+    {
+      "epoch": 3.476590909090909,
+      "grad_norm": 0.9485176312878887,
+      "learning_rate": 2.6518373128767947e-07,
+      "loss": 0.0065,
+      "step": 15297
+    },
+    {
+      "epoch": 3.4768181818181816,
+      "grad_norm": 1.8253778724346479,
+      "learning_rate": 2.651107592619374e-07,
+      "loss": 0.0072,
+      "step": 15298
+    },
+    {
+      "epoch": 3.4770454545454546,
+      "grad_norm": 2.2424216452419454,
+      "learning_rate": 2.6503779457496473e-07,
+      "loss": 0.0242,
+      "step": 15299
+    },
+    {
+      "epoch": 3.4772727272727275,
+      "grad_norm": 1.6681357917234998,
+      "learning_rate": 2.649648372282497e-07,
+      "loss": 0.0099,
+      "step": 15300
+    },
+    {
+      "epoch": 3.4775,
+      "grad_norm": 1.2063781973634753,
+      "learning_rate": 2.6489188722327953e-07,
+      "loss": 0.0053,
+      "step": 15301
+    },
+    {
+      "epoch": 3.4777272727272726,
+      "grad_norm": 1.0302233619275216,
+      "learning_rate": 2.6481894456154197e-07,
+      "loss": 0.0042,
+      "step": 15302
+    },
+    {
+      "epoch": 3.4779545454545455,
+      "grad_norm": 1.6186474525716998,
+      "learning_rate": 2.647460092445246e-07,
+      "loss": 0.0132,
+      "step": 15303
+    },
+    {
+      "epoch": 3.478181818181818,
+      "grad_norm": 2.201792284756348,
+      "learning_rate": 2.646730812737148e-07,
+      "loss": 0.0205,
+      "step": 15304
+    },
+    {
+      "epoch": 3.478409090909091,
+      "grad_norm": 1.7374550906062183,
+      "learning_rate": 2.646001606505992e-07,
+      "loss": 0.0049,
+      "step": 15305
+    },
+    {
+      "epoch": 3.4786363636363635,
+      "grad_norm": 1.529635940427012,
+      "learning_rate": 2.645272473766651e-07,
+      "loss": 0.005,
+      "step": 15306
+    },
+    {
+      "epoch": 3.4788636363636365,
+      "grad_norm": 1.0515486283836335,
+      "learning_rate": 2.6445434145339936e-07,
+      "loss": 0.0119,
+      "step": 15307
+    },
+    {
+      "epoch": 3.479090909090909,
+      "grad_norm": 1.5862989389120181,
+      "learning_rate": 2.6438144288228856e-07,
+      "loss": 0.0072,
+      "step": 15308
+    },
+    {
+      "epoch": 3.479318181818182,
+      "grad_norm": 1.3381610002559472,
+      "learning_rate": 2.643085516648193e-07,
+      "loss": 0.0107,
+      "step": 15309
+    },
+    {
+      "epoch": 3.4795454545454545,
+      "grad_norm": 0.5427108057833298,
+      "learning_rate": 2.642356678024779e-07,
+      "loss": 0.0025,
+      "step": 15310
+    },
+    {
+      "epoch": 3.4797727272727275,
+      "grad_norm": 0.6025860095280876,
+      "learning_rate": 2.6416279129675065e-07,
+      "loss": 0.0047,
+      "step": 15311
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 1.6383672425865539,
+      "learning_rate": 2.640899221491236e-07,
+      "loss": 0.0169,
+      "step": 15312
+    },
+    {
+      "epoch": 3.480227272727273,
+      "grad_norm": 0.6469799608070326,
+      "learning_rate": 2.6401706036108287e-07,
+      "loss": 0.0033,
+      "step": 15313
+    },
+    {
+      "epoch": 3.4804545454545455,
+      "grad_norm": 1.6946891876005772,
+      "learning_rate": 2.639442059341139e-07,
+      "loss": 0.012,
+      "step": 15314
+    },
+    {
+      "epoch": 3.480681818181818,
+      "grad_norm": 2.9743328617413045,
+      "learning_rate": 2.6387135886970235e-07,
+      "loss": 0.0307,
+      "step": 15315
+    },
+    {
+      "epoch": 3.480909090909091,
+      "grad_norm": 2.27114100958702,
+      "learning_rate": 2.637985191693339e-07,
+      "loss": 0.0068,
+      "step": 15316
+    },
+    {
+      "epoch": 3.4811363636363635,
+      "grad_norm": 1.0382767744065682,
+      "learning_rate": 2.6372568683449375e-07,
+      "loss": 0.0121,
+      "step": 15317
+    },
+    {
+      "epoch": 3.4813636363636364,
+      "grad_norm": 2.5852561357384793,
+      "learning_rate": 2.636528618666674e-07,
+      "loss": 0.0218,
+      "step": 15318
+    },
+    {
+      "epoch": 3.481590909090909,
+      "grad_norm": 1.4283816739398962,
+      "learning_rate": 2.635800442673393e-07,
+      "loss": 0.0096,
+      "step": 15319
+    },
+    {
+      "epoch": 3.481818181818182,
+      "grad_norm": 0.9496028942490071,
+      "learning_rate": 2.635072340379947e-07,
+      "loss": 0.0043,
+      "step": 15320
+    },
+    {
+      "epoch": 3.4820454545454544,
+      "grad_norm": 1.7416038567591463,
+      "learning_rate": 2.634344311801183e-07,
+      "loss": 0.011,
+      "step": 15321
+    },
+    {
+      "epoch": 3.4822727272727274,
+      "grad_norm": 2.1516460247639797,
+      "learning_rate": 2.633616356951947e-07,
+      "loss": 0.0117,
+      "step": 15322
+    },
+    {
+      "epoch": 3.4825,
+      "grad_norm": 0.36967358756760255,
+      "learning_rate": 2.632888475847082e-07,
+      "loss": 0.0019,
+      "step": 15323
+    },
+    {
+      "epoch": 3.482727272727273,
+      "grad_norm": 1.84630741423415,
+      "learning_rate": 2.632160668501428e-07,
+      "loss": 0.0164,
+      "step": 15324
+    },
+    {
+      "epoch": 3.4829545454545454,
+      "grad_norm": 0.926350266147572,
+      "learning_rate": 2.631432934929833e-07,
+      "loss": 0.0038,
+      "step": 15325
+    },
+    {
+      "epoch": 3.4831818181818184,
+      "grad_norm": 0.786969590459282,
+      "learning_rate": 2.6307052751471343e-07,
+      "loss": 0.0022,
+      "step": 15326
+    },
+    {
+      "epoch": 3.483409090909091,
+      "grad_norm": 1.3920883051452357,
+      "learning_rate": 2.629977689168168e-07,
+      "loss": 0.0014,
+      "step": 15327
+    },
+    {
+      "epoch": 3.4836363636363634,
+      "grad_norm": 1.1347762522524223,
+      "learning_rate": 2.6292501770077716e-07,
+      "loss": 0.0054,
+      "step": 15328
+    },
+    {
+      "epoch": 3.4838636363636364,
+      "grad_norm": 1.4759107750266143,
+      "learning_rate": 2.6285227386807805e-07,
+      "loss": 0.0177,
+      "step": 15329
+    },
+    {
+      "epoch": 3.484090909090909,
+      "grad_norm": 3.54470075751194,
+      "learning_rate": 2.627795374202029e-07,
+      "loss": 0.0256,
+      "step": 15330
+    },
+    {
+      "epoch": 3.484318181818182,
+      "grad_norm": 3.5115561630813494,
+      "learning_rate": 2.6270680835863525e-07,
+      "loss": 0.0419,
+      "step": 15331
+    },
+    {
+      "epoch": 3.4845454545454544,
+      "grad_norm": 2.179764173807924,
+      "learning_rate": 2.626340866848575e-07,
+      "loss": 0.034,
+      "step": 15332
+    },
+    {
+      "epoch": 3.4847727272727274,
+      "grad_norm": 0.8352247065610375,
+      "learning_rate": 2.625613724003529e-07,
+      "loss": 0.0128,
+      "step": 15333
+    },
+    {
+      "epoch": 3.485,
+      "grad_norm": 1.361612447574515,
+      "learning_rate": 2.6248866550660426e-07,
+      "loss": 0.0042,
+      "step": 15334
+    },
+    {
+      "epoch": 3.485227272727273,
+      "grad_norm": 1.027430513677489,
+      "learning_rate": 2.6241596600509434e-07,
+      "loss": 0.0068,
+      "step": 15335
+    },
+    {
+      "epoch": 3.4854545454545454,
+      "grad_norm": 1.3645675180504317,
+      "learning_rate": 2.6234327389730527e-07,
+      "loss": 0.0114,
+      "step": 15336
+    },
+    {
+      "epoch": 3.4856818181818183,
+      "grad_norm": 1.843219398422678,
+      "learning_rate": 2.622705891847195e-07,
+      "loss": 0.0151,
+      "step": 15337
+    },
+    {
+      "epoch": 3.485909090909091,
+      "grad_norm": 1.3634336681568446,
+      "learning_rate": 2.6219791186881927e-07,
+      "loss": 0.0036,
+      "step": 15338
+    },
+    {
+      "epoch": 3.486136363636364,
+      "grad_norm": 1.1057599104446993,
+      "learning_rate": 2.621252419510865e-07,
+      "loss": 0.0099,
+      "step": 15339
+    },
+    {
+      "epoch": 3.4863636363636363,
+      "grad_norm": 1.444694291039274,
+      "learning_rate": 2.6205257943300317e-07,
+      "loss": 0.0071,
+      "step": 15340
+    },
+    {
+      "epoch": 3.486590909090909,
+      "grad_norm": 1.1598588042375477,
+      "learning_rate": 2.619799243160509e-07,
+      "loss": 0.0067,
+      "step": 15341
+    },
+    {
+      "epoch": 3.486818181818182,
+      "grad_norm": 1.4644764072576302,
+      "learning_rate": 2.619072766017113e-07,
+      "loss": 0.0162,
+      "step": 15342
+    },
+    {
+      "epoch": 3.487045454545455,
+      "grad_norm": 1.2759125013318957,
+      "learning_rate": 2.618346362914659e-07,
+      "loss": 0.01,
+      "step": 15343
+    },
+    {
+      "epoch": 3.4872727272727273,
+      "grad_norm": 1.2768164973924185,
+      "learning_rate": 2.617620033867959e-07,
+      "loss": 0.0072,
+      "step": 15344
+    },
+    {
+      "epoch": 3.4875,
+      "grad_norm": 2.067466937010182,
+      "learning_rate": 2.6168937788918223e-07,
+      "loss": 0.0108,
+      "step": 15345
+    },
+    {
+      "epoch": 3.487727272727273,
+      "grad_norm": 1.2731855450206675,
+      "learning_rate": 2.616167598001059e-07,
+      "loss": 0.0029,
+      "step": 15346
+    },
+    {
+      "epoch": 3.4879545454545453,
+      "grad_norm": 1.2080535813439444,
+      "learning_rate": 2.6154414912104785e-07,
+      "loss": 0.0068,
+      "step": 15347
+    },
+    {
+      "epoch": 3.4881818181818183,
+      "grad_norm": 4.128066297709324,
+      "learning_rate": 2.614715458534889e-07,
+      "loss": 0.0409,
+      "step": 15348
+    },
+    {
+      "epoch": 3.488409090909091,
+      "grad_norm": 1.1586960013745635,
+      "learning_rate": 2.6139894999890913e-07,
+      "loss": 0.0105,
+      "step": 15349
+    },
+    {
+      "epoch": 3.4886363636363638,
+      "grad_norm": 0.5368451124709825,
+      "learning_rate": 2.613263615587891e-07,
+      "loss": 0.0029,
+      "step": 15350
+    },
+    {
+      "epoch": 3.4888636363636363,
+      "grad_norm": 0.623190581319586,
+      "learning_rate": 2.6125378053460896e-07,
+      "loss": 0.0092,
+      "step": 15351
+    },
+    {
+      "epoch": 3.4890909090909092,
+      "grad_norm": 0.9538646269838507,
+      "learning_rate": 2.611812069278489e-07,
+      "loss": 0.0146,
+      "step": 15352
+    },
+    {
+      "epoch": 3.4893181818181818,
+      "grad_norm": 1.1663791454921337,
+      "learning_rate": 2.611086407399887e-07,
+      "loss": 0.0079,
+      "step": 15353
+    },
+    {
+      "epoch": 3.4895454545454547,
+      "grad_norm": 1.915027631929644,
+      "learning_rate": 2.610360819725082e-07,
+      "loss": 0.0167,
+      "step": 15354
+    },
+    {
+      "epoch": 3.4897727272727272,
+      "grad_norm": 2.551004379727549,
+      "learning_rate": 2.6096353062688694e-07,
+      "loss": 0.0179,
+      "step": 15355
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 0.3992676624742516,
+      "learning_rate": 2.608909867046044e-07,
+      "loss": 0.0015,
+      "step": 15356
+    },
+    {
+      "epoch": 3.4902272727272727,
+      "grad_norm": 2.5391908656935893,
+      "learning_rate": 2.6081845020714007e-07,
+      "loss": 0.0121,
+      "step": 15357
+    },
+    {
+      "epoch": 3.4904545454545453,
+      "grad_norm": 2.976234330133564,
+      "learning_rate": 2.6074592113597264e-07,
+      "loss": 0.0278,
+      "step": 15358
+    },
+    {
+      "epoch": 3.490681818181818,
+      "grad_norm": 2.6385484553002523,
+      "learning_rate": 2.6067339949258143e-07,
+      "loss": 0.015,
+      "step": 15359
+    },
+    {
+      "epoch": 3.4909090909090907,
+      "grad_norm": 0.5510350073991371,
+      "learning_rate": 2.6060088527844517e-07,
+      "loss": 0.0028,
+      "step": 15360
+    },
+    {
+      "epoch": 3.4911363636363637,
+      "grad_norm": 1.1318237562612796,
+      "learning_rate": 2.6052837849504254e-07,
+      "loss": 0.0087,
+      "step": 15361
+    },
+    {
+      "epoch": 3.4913636363636362,
+      "grad_norm": 1.7382839427135546,
+      "learning_rate": 2.6045587914385237e-07,
+      "loss": 0.0122,
+      "step": 15362
+    },
+    {
+      "epoch": 3.491590909090909,
+      "grad_norm": 1.256131009324666,
+      "learning_rate": 2.603833872263526e-07,
+      "loss": 0.0088,
+      "step": 15363
+    },
+    {
+      "epoch": 3.4918181818181817,
+      "grad_norm": 2.6864323704798796,
+      "learning_rate": 2.603109027440217e-07,
+      "loss": 0.0254,
+      "step": 15364
+    },
+    {
+      "epoch": 3.4920454545454547,
+      "grad_norm": 3.0716846067531995,
+      "learning_rate": 2.602384256983377e-07,
+      "loss": 0.0124,
+      "step": 15365
+    },
+    {
+      "epoch": 3.492272727272727,
+      "grad_norm": 3.028154723035742,
+      "learning_rate": 2.6016595609077877e-07,
+      "loss": 0.0182,
+      "step": 15366
+    },
+    {
+      "epoch": 3.4925,
+      "grad_norm": 1.545667064922188,
+      "learning_rate": 2.600934939228223e-07,
+      "loss": 0.0071,
+      "step": 15367
+    },
+    {
+      "epoch": 3.4927272727272727,
+      "grad_norm": 1.9000582593970794,
+      "learning_rate": 2.6002103919594607e-07,
+      "loss": 0.0096,
+      "step": 15368
+    },
+    {
+      "epoch": 3.4929545454545456,
+      "grad_norm": 1.9193869813245077,
+      "learning_rate": 2.5994859191162765e-07,
+      "loss": 0.014,
+      "step": 15369
+    },
+    {
+      "epoch": 3.493181818181818,
+      "grad_norm": 2.403819923671975,
+      "learning_rate": 2.598761520713442e-07,
+      "loss": 0.0135,
+      "step": 15370
+    },
+    {
+      "epoch": 3.4934090909090907,
+      "grad_norm": 1.750663223047224,
+      "learning_rate": 2.5980371967657305e-07,
+      "loss": 0.0147,
+      "step": 15371
+    },
+    {
+      "epoch": 3.4936363636363637,
+      "grad_norm": 3.333337232038748,
+      "learning_rate": 2.5973129472879127e-07,
+      "loss": 0.0175,
+      "step": 15372
+    },
+    {
+      "epoch": 3.493863636363636,
+      "grad_norm": 1.9214676973471767,
+      "learning_rate": 2.596588772294755e-07,
+      "loss": 0.0258,
+      "step": 15373
+    },
+    {
+      "epoch": 3.494090909090909,
+      "grad_norm": 1.6005746364624474,
+      "learning_rate": 2.5958646718010266e-07,
+      "loss": 0.0152,
+      "step": 15374
+    },
+    {
+      "epoch": 3.4943181818181817,
+      "grad_norm": 2.5196333606175596,
+      "learning_rate": 2.5951406458214944e-07,
+      "loss": 0.0162,
+      "step": 15375
+    },
+    {
+      "epoch": 3.4945454545454546,
+      "grad_norm": 2.2762166146029803,
+      "learning_rate": 2.594416694370919e-07,
+      "loss": 0.004,
+      "step": 15376
+    },
+    {
+      "epoch": 3.494772727272727,
+      "grad_norm": 1.21405950993682,
+      "learning_rate": 2.593692817464064e-07,
+      "loss": 0.0128,
+      "step": 15377
+    },
+    {
+      "epoch": 3.495,
+      "grad_norm": 1.3701522677759355,
+      "learning_rate": 2.592969015115692e-07,
+      "loss": 0.0083,
+      "step": 15378
+    },
+    {
+      "epoch": 3.4952272727272726,
+      "grad_norm": 2.780153821602523,
+      "learning_rate": 2.592245287340564e-07,
+      "loss": 0.0247,
+      "step": 15379
+    },
+    {
+      "epoch": 3.4954545454545456,
+      "grad_norm": 2.412083923537745,
+      "learning_rate": 2.5915216341534335e-07,
+      "loss": 0.0163,
+      "step": 15380
+    },
+    {
+      "epoch": 3.495681818181818,
+      "grad_norm": 1.9940152818533845,
+      "learning_rate": 2.590798055569059e-07,
+      "loss": 0.0202,
+      "step": 15381
+    },
+    {
+      "epoch": 3.495909090909091,
+      "grad_norm": 1.5235099679738475,
+      "learning_rate": 2.5900745516021964e-07,
+      "loss": 0.0117,
+      "step": 15382
+    },
+    {
+      "epoch": 3.4961363636363636,
+      "grad_norm": 1.0477970941434984,
+      "learning_rate": 2.5893511222675984e-07,
+      "loss": 0.0097,
+      "step": 15383
+    },
+    {
+      "epoch": 3.496363636363636,
+      "grad_norm": 1.7124439465726777,
+      "learning_rate": 2.5886277675800183e-07,
+      "loss": 0.011,
+      "step": 15384
+    },
+    {
+      "epoch": 3.496590909090909,
+      "grad_norm": 0.8512248784771715,
+      "learning_rate": 2.587904487554205e-07,
+      "loss": 0.0044,
+      "step": 15385
+    },
+    {
+      "epoch": 3.496818181818182,
+      "grad_norm": 1.3215219877552389,
+      "learning_rate": 2.587181282204908e-07,
+      "loss": 0.0105,
+      "step": 15386
+    },
+    {
+      "epoch": 3.4970454545454546,
+      "grad_norm": 3.109118580771091,
+      "learning_rate": 2.5864581515468745e-07,
+      "loss": 0.0132,
+      "step": 15387
+    },
+    {
+      "epoch": 3.497272727272727,
+      "grad_norm": 2.6486001276155506,
+      "learning_rate": 2.585735095594853e-07,
+      "loss": 0.0241,
+      "step": 15388
+    },
+    {
+      "epoch": 3.4975,
+      "grad_norm": 1.491358281476106,
+      "learning_rate": 2.585012114363584e-07,
+      "loss": 0.0104,
+      "step": 15389
+    },
+    {
+      "epoch": 3.4977272727272726,
+      "grad_norm": 1.229948004348701,
+      "learning_rate": 2.5842892078678113e-07,
+      "loss": 0.0053,
+      "step": 15390
+    },
+    {
+      "epoch": 3.4979545454545455,
+      "grad_norm": 1.0652027925346137,
+      "learning_rate": 2.5835663761222776e-07,
+      "loss": 0.006,
+      "step": 15391
+    },
+    {
+      "epoch": 3.498181818181818,
+      "grad_norm": 1.582005357794189,
+      "learning_rate": 2.582843619141723e-07,
+      "loss": 0.0152,
+      "step": 15392
+    },
+    {
+      "epoch": 3.498409090909091,
+      "grad_norm": 4.198753451741911,
+      "learning_rate": 2.582120936940884e-07,
+      "loss": 0.0214,
+      "step": 15393
+    },
+    {
+      "epoch": 3.4986363636363635,
+      "grad_norm": 1.8022072905815703,
+      "learning_rate": 2.5813983295344965e-07,
+      "loss": 0.0045,
+      "step": 15394
+    },
+    {
+      "epoch": 3.4988636363636365,
+      "grad_norm": 0.8775827002562391,
+      "learning_rate": 2.580675796937298e-07,
+      "loss": 0.0056,
+      "step": 15395
+    },
+    {
+      "epoch": 3.499090909090909,
+      "grad_norm": 2.132260325374214,
+      "learning_rate": 2.5799533391640216e-07,
+      "loss": 0.0114,
+      "step": 15396
+    },
+    {
+      "epoch": 3.499318181818182,
+      "grad_norm": 0.44738507806514416,
+      "learning_rate": 2.5792309562294014e-07,
+      "loss": 0.0017,
+      "step": 15397
+    },
+    {
+      "epoch": 3.4995454545454545,
+      "grad_norm": 1.4736288538705073,
+      "learning_rate": 2.5785086481481643e-07,
+      "loss": 0.0178,
+      "step": 15398
+    },
+    {
+      "epoch": 3.4997727272727275,
+      "grad_norm": 1.6289680303392242,
+      "learning_rate": 2.57778641493504e-07,
+      "loss": 0.0072,
+      "step": 15399
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.6969600200716277,
+      "learning_rate": 2.5770642566047584e-07,
+      "loss": 0.0047,
+      "step": 15400
+    },
+    {
+      "epoch": 3.5002272727272725,
+      "grad_norm": 1.68405435090908,
+      "learning_rate": 2.5763421731720437e-07,
+      "loss": 0.0122,
+      "step": 15401
+    },
+    {
+      "epoch": 3.5004545454545455,
+      "grad_norm": 1.7965606278228976,
+      "learning_rate": 2.575620164651621e-07,
+      "loss": 0.0118,
+      "step": 15402
+    },
+    {
+      "epoch": 3.5006818181818184,
+      "grad_norm": 0.8536330352307427,
+      "learning_rate": 2.574898231058214e-07,
+      "loss": 0.0054,
+      "step": 15403
+    },
+    {
+      "epoch": 3.500909090909091,
+      "grad_norm": 1.0395365366362326,
+      "learning_rate": 2.5741763724065433e-07,
+      "loss": 0.0033,
+      "step": 15404
+    },
+    {
+      "epoch": 3.5011363636363635,
+      "grad_norm": 3.547070963188544,
+      "learning_rate": 2.5734545887113315e-07,
+      "loss": 0.013,
+      "step": 15405
+    },
+    {
+      "epoch": 3.5013636363636365,
+      "grad_norm": 2.4695379490860905,
+      "learning_rate": 2.572732879987293e-07,
+      "loss": 0.0075,
+      "step": 15406
+    },
+    {
+      "epoch": 3.501590909090909,
+      "grad_norm": 2.115376238461881,
+      "learning_rate": 2.5720112462491455e-07,
+      "loss": 0.0091,
+      "step": 15407
+    },
+    {
+      "epoch": 3.501818181818182,
+      "grad_norm": 0.8114839947851586,
+      "learning_rate": 2.5712896875116056e-07,
+      "loss": 0.0072,
+      "step": 15408
+    },
+    {
+      "epoch": 3.5020454545454545,
+      "grad_norm": 0.7546522668419867,
+      "learning_rate": 2.570568203789387e-07,
+      "loss": 0.0033,
+      "step": 15409
+    },
+    {
+      "epoch": 3.5022727272727274,
+      "grad_norm": 3.9171681578757402,
+      "learning_rate": 2.5698467950972044e-07,
+      "loss": 0.0202,
+      "step": 15410
+    },
+    {
+      "epoch": 3.5025,
+      "grad_norm": 2.1202559417567652,
+      "learning_rate": 2.569125461449763e-07,
+      "loss": 0.0288,
+      "step": 15411
+    },
+    {
+      "epoch": 3.502727272727273,
+      "grad_norm": 1.6315546844861601,
+      "learning_rate": 2.568404202861776e-07,
+      "loss": 0.0237,
+      "step": 15412
+    },
+    {
+      "epoch": 3.5029545454545454,
+      "grad_norm": 3.666612129715128,
+      "learning_rate": 2.5676830193479495e-07,
+      "loss": 0.0387,
+      "step": 15413
+    },
+    {
+      "epoch": 3.503181818181818,
+      "grad_norm": 2.5444260614325604,
+      "learning_rate": 2.56696191092299e-07,
+      "loss": 0.0207,
+      "step": 15414
+    },
+    {
+      "epoch": 3.503409090909091,
+      "grad_norm": 0.5384944369901511,
+      "learning_rate": 2.566240877601604e-07,
+      "loss": 0.0017,
+      "step": 15415
+    },
+    {
+      "epoch": 3.503636363636364,
+      "grad_norm": 1.0179540176337285,
+      "learning_rate": 2.5655199193984915e-07,
+      "loss": 0.0045,
+      "step": 15416
+    },
+    {
+      "epoch": 3.5038636363636364,
+      "grad_norm": 2.0933562602562277,
+      "learning_rate": 2.5647990363283573e-07,
+      "loss": 0.0165,
+      "step": 15417
+    },
+    {
+      "epoch": 3.504090909090909,
+      "grad_norm": 1.7034959843848705,
+      "learning_rate": 2.564078228405899e-07,
+      "loss": 0.0199,
+      "step": 15418
+    },
+    {
+      "epoch": 3.504318181818182,
+      "grad_norm": 4.33314266454507,
+      "learning_rate": 2.5633574956458183e-07,
+      "loss": 0.0205,
+      "step": 15419
+    },
+    {
+      "epoch": 3.5045454545454544,
+      "grad_norm": 1.813126032412574,
+      "learning_rate": 2.5626368380628083e-07,
+      "loss": 0.0299,
+      "step": 15420
+    },
+    {
+      "epoch": 3.5047727272727274,
+      "grad_norm": 2.519662156286633,
+      "learning_rate": 2.5619162556715657e-07,
+      "loss": 0.0101,
+      "step": 15421
+    },
+    {
+      "epoch": 3.505,
+      "grad_norm": 1.264217667605485,
+      "learning_rate": 2.5611957484867845e-07,
+      "loss": 0.0043,
+      "step": 15422
+    },
+    {
+      "epoch": 3.505227272727273,
+      "grad_norm": 0.6329684944360237,
+      "learning_rate": 2.5604753165231607e-07,
+      "loss": 0.0031,
+      "step": 15423
+    },
+    {
+      "epoch": 3.5054545454545454,
+      "grad_norm": 0.9584883065673007,
+      "learning_rate": 2.559754959795379e-07,
+      "loss": 0.0069,
+      "step": 15424
+    },
+    {
+      "epoch": 3.5056818181818183,
+      "grad_norm": 1.6414825272109983,
+      "learning_rate": 2.559034678318132e-07,
+      "loss": 0.0173,
+      "step": 15425
+    },
+    {
+      "epoch": 3.505909090909091,
+      "grad_norm": 0.8575752558604388,
+      "learning_rate": 2.558314472106107e-07,
+      "loss": 0.0045,
+      "step": 15426
+    },
+    {
+      "epoch": 3.5061363636363634,
+      "grad_norm": 1.646117466887481,
+      "learning_rate": 2.5575943411739926e-07,
+      "loss": 0.0129,
+      "step": 15427
+    },
+    {
+      "epoch": 3.5063636363636363,
+      "grad_norm": 2.3312198575964995,
+      "learning_rate": 2.556874285536469e-07,
+      "loss": 0.0137,
+      "step": 15428
+    },
+    {
+      "epoch": 3.5065909090909093,
+      "grad_norm": 1.0877285439624746,
+      "learning_rate": 2.5561543052082214e-07,
+      "loss": 0.0115,
+      "step": 15429
+    },
+    {
+      "epoch": 3.506818181818182,
+      "grad_norm": 1.1860816610211644,
+      "learning_rate": 2.555434400203933e-07,
+      "loss": 0.0123,
+      "step": 15430
+    },
+    {
+      "epoch": 3.5070454545454544,
+      "grad_norm": 1.523793427100534,
+      "learning_rate": 2.554714570538282e-07,
+      "loss": 0.0087,
+      "step": 15431
+    },
+    {
+      "epoch": 3.5072727272727273,
+      "grad_norm": 1.0736420069324442,
+      "learning_rate": 2.5539948162259473e-07,
+      "loss": 0.0044,
+      "step": 15432
+    },
+    {
+      "epoch": 3.5075,
+      "grad_norm": 1.134723266740821,
+      "learning_rate": 2.5532751372816064e-07,
+      "loss": 0.0072,
+      "step": 15433
+    },
+    {
+      "epoch": 3.507727272727273,
+      "grad_norm": 2.035599273210648,
+      "learning_rate": 2.552555533719935e-07,
+      "loss": 0.0174,
+      "step": 15434
+    },
+    {
+      "epoch": 3.5079545454545453,
+      "grad_norm": 2.2990975369043185,
+      "learning_rate": 2.5518360055556067e-07,
+      "loss": 0.0347,
+      "step": 15435
+    },
+    {
+      "epoch": 3.5081818181818183,
+      "grad_norm": 1.80563617036263,
+      "learning_rate": 2.551116552803296e-07,
+      "loss": 0.0085,
+      "step": 15436
+    },
+    {
+      "epoch": 3.508409090909091,
+      "grad_norm": 0.8636605191795617,
+      "learning_rate": 2.55039717547767e-07,
+      "loss": 0.0061,
+      "step": 15437
+    },
+    {
+      "epoch": 3.5086363636363638,
+      "grad_norm": 3.6191988906403325,
+      "learning_rate": 2.5496778735934e-07,
+      "loss": 0.0329,
+      "step": 15438
+    },
+    {
+      "epoch": 3.5088636363636363,
+      "grad_norm": 1.5365586464590866,
+      "learning_rate": 2.548958647165154e-07,
+      "loss": 0.0064,
+      "step": 15439
+    },
+    {
+      "epoch": 3.509090909090909,
+      "grad_norm": 0.8112232640191012,
+      "learning_rate": 2.548239496207598e-07,
+      "loss": 0.0048,
+      "step": 15440
+    },
+    {
+      "epoch": 3.509318181818182,
+      "grad_norm": 1.544706537318148,
+      "learning_rate": 2.5475204207353985e-07,
+      "loss": 0.0186,
+      "step": 15441
+    },
+    {
+      "epoch": 3.5095454545454547,
+      "grad_norm": 1.0298603088498752,
+      "learning_rate": 2.5468014207632147e-07,
+      "loss": 0.0055,
+      "step": 15442
+    },
+    {
+      "epoch": 3.5097727272727273,
+      "grad_norm": 1.4266773512299047,
+      "learning_rate": 2.54608249630571e-07,
+      "loss": 0.0091,
+      "step": 15443
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 2.1556792600241343,
+      "learning_rate": 2.5453636473775463e-07,
+      "loss": 0.0077,
+      "step": 15444
+    },
+    {
+      "epoch": 3.5102272727272728,
+      "grad_norm": 3.933915382126207,
+      "learning_rate": 2.54464487399338e-07,
+      "loss": 0.0338,
+      "step": 15445
+    },
+    {
+      "epoch": 3.5104545454545457,
+      "grad_norm": 0.888994871832733,
+      "learning_rate": 2.543926176167869e-07,
+      "loss": 0.0051,
+      "step": 15446
+    },
+    {
+      "epoch": 3.5106818181818182,
+      "grad_norm": 2.2857432401192126,
+      "learning_rate": 2.54320755391567e-07,
+      "loss": 0.0311,
+      "step": 15447
+    },
+    {
+      "epoch": 3.5109090909090908,
+      "grad_norm": 1.1349621658013802,
+      "learning_rate": 2.542489007251435e-07,
+      "loss": 0.0145,
+      "step": 15448
+    },
+    {
+      "epoch": 3.5111363636363637,
+      "grad_norm": 2.0597496353582914,
+      "learning_rate": 2.5417705361898195e-07,
+      "loss": 0.0268,
+      "step": 15449
+    },
+    {
+      "epoch": 3.5113636363636362,
+      "grad_norm": 1.7454639329945167,
+      "learning_rate": 2.5410521407454694e-07,
+      "loss": 0.0196,
+      "step": 15450
+    },
+    {
+      "epoch": 3.511590909090909,
+      "grad_norm": 1.7097258350995312,
+      "learning_rate": 2.540333820933038e-07,
+      "loss": 0.0098,
+      "step": 15451
+    },
+    {
+      "epoch": 3.5118181818181817,
+      "grad_norm": 1.0010105459413923,
+      "learning_rate": 2.539615576767171e-07,
+      "loss": 0.005,
+      "step": 15452
+    },
+    {
+      "epoch": 3.5120454545454547,
+      "grad_norm": 0.8931361320042841,
+      "learning_rate": 2.5388974082625154e-07,
+      "loss": 0.005,
+      "step": 15453
+    },
+    {
+      "epoch": 3.512272727272727,
+      "grad_norm": 1.039878209000329,
+      "learning_rate": 2.5381793154337184e-07,
+      "loss": 0.0078,
+      "step": 15454
+    },
+    {
+      "epoch": 3.5125,
+      "grad_norm": 2.9884612948392832,
+      "learning_rate": 2.537461298295419e-07,
+      "loss": 0.0165,
+      "step": 15455
+    },
+    {
+      "epoch": 3.5127272727272727,
+      "grad_norm": 1.1632812900826948,
+      "learning_rate": 2.53674335686226e-07,
+      "loss": 0.0056,
+      "step": 15456
+    },
+    {
+      "epoch": 3.512954545454545,
+      "grad_norm": 0.4594157208964846,
+      "learning_rate": 2.5360254911488824e-07,
+      "loss": 0.0021,
+      "step": 15457
+    },
+    {
+      "epoch": 3.513181818181818,
+      "grad_norm": 0.5426478355460331,
+      "learning_rate": 2.535307701169926e-07,
+      "loss": 0.0026,
+      "step": 15458
+    },
+    {
+      "epoch": 3.513409090909091,
+      "grad_norm": 1.3831331074557978,
+      "learning_rate": 2.5345899869400253e-07,
+      "loss": 0.0077,
+      "step": 15459
+    },
+    {
+      "epoch": 3.5136363636363637,
+      "grad_norm": 1.4108859591273732,
+      "learning_rate": 2.533872348473816e-07,
+      "loss": 0.0081,
+      "step": 15460
+    },
+    {
+      "epoch": 3.513863636363636,
+      "grad_norm": 1.7435546540293185,
+      "learning_rate": 2.5331547857859323e-07,
+      "loss": 0.0071,
+      "step": 15461
+    },
+    {
+      "epoch": 3.514090909090909,
+      "grad_norm": 2.3972497711120866,
+      "learning_rate": 2.532437298891005e-07,
+      "loss": 0.0153,
+      "step": 15462
+    },
+    {
+      "epoch": 3.5143181818181817,
+      "grad_norm": 0.8309296453655546,
+      "learning_rate": 2.5317198878036716e-07,
+      "loss": 0.0055,
+      "step": 15463
+    },
+    {
+      "epoch": 3.5145454545454546,
+      "grad_norm": 3.020037857269488,
+      "learning_rate": 2.5310025525385533e-07,
+      "loss": 0.019,
+      "step": 15464
+    },
+    {
+      "epoch": 3.514772727272727,
+      "grad_norm": 0.7838519152026951,
+      "learning_rate": 2.530285293110282e-07,
+      "loss": 0.0076,
+      "step": 15465
+    },
+    {
+      "epoch": 3.515,
+      "grad_norm": 1.1790296328686982,
+      "learning_rate": 2.529568109533483e-07,
+      "loss": 0.0079,
+      "step": 15466
+    },
+    {
+      "epoch": 3.5152272727272726,
+      "grad_norm": 0.871676851898223,
+      "learning_rate": 2.528851001822783e-07,
+      "loss": 0.0081,
+      "step": 15467
+    },
+    {
+      "epoch": 3.5154545454545456,
+      "grad_norm": 0.5023390253038601,
+      "learning_rate": 2.5281339699928006e-07,
+      "loss": 0.0019,
+      "step": 15468
+    },
+    {
+      "epoch": 3.515681818181818,
+      "grad_norm": 1.6132524996617246,
+      "learning_rate": 2.52741701405816e-07,
+      "loss": 0.0054,
+      "step": 15469
+    },
+    {
+      "epoch": 3.5159090909090907,
+      "grad_norm": 0.6198100886328362,
+      "learning_rate": 2.526700134033481e-07,
+      "loss": 0.0017,
+      "step": 15470
+    },
+    {
+      "epoch": 3.5161363636363636,
+      "grad_norm": 2.268893975134785,
+      "learning_rate": 2.525983329933383e-07,
+      "loss": 0.0106,
+      "step": 15471
+    },
+    {
+      "epoch": 3.5163636363636366,
+      "grad_norm": 1.1419853759428875,
+      "learning_rate": 2.525266601772481e-07,
+      "loss": 0.0081,
+      "step": 15472
+    },
+    {
+      "epoch": 3.516590909090909,
+      "grad_norm": 2.4518491254113624,
+      "learning_rate": 2.52454994956539e-07,
+      "loss": 0.0198,
+      "step": 15473
+    },
+    {
+      "epoch": 3.5168181818181816,
+      "grad_norm": 1.4234668389930536,
+      "learning_rate": 2.523833373326726e-07,
+      "loss": 0.0082,
+      "step": 15474
+    },
+    {
+      "epoch": 3.5170454545454546,
+      "grad_norm": 1.3007572123545104,
+      "learning_rate": 2.5231168730710995e-07,
+      "loss": 0.0074,
+      "step": 15475
+    },
+    {
+      "epoch": 3.517272727272727,
+      "grad_norm": 1.4178839062713975,
+      "learning_rate": 2.522400448813122e-07,
+      "loss": 0.0082,
+      "step": 15476
+    },
+    {
+      "epoch": 3.5175,
+      "grad_norm": 1.3884257830714057,
+      "learning_rate": 2.521684100567402e-07,
+      "loss": 0.0098,
+      "step": 15477
+    },
+    {
+      "epoch": 3.5177272727272726,
+      "grad_norm": 3.327941365381701,
+      "learning_rate": 2.520967828348548e-07,
+      "loss": 0.0153,
+      "step": 15478
+    },
+    {
+      "epoch": 3.5179545454545456,
+      "grad_norm": 1.9425863706658135,
+      "learning_rate": 2.520251632171166e-07,
+      "loss": 0.0137,
+      "step": 15479
+    },
+    {
+      "epoch": 3.518181818181818,
+      "grad_norm": 0.520755561980395,
+      "learning_rate": 2.519535512049862e-07,
+      "loss": 0.0029,
+      "step": 15480
+    },
+    {
+      "epoch": 3.518409090909091,
+      "grad_norm": 1.0289026650407855,
+      "learning_rate": 2.518819467999235e-07,
+      "loss": 0.0055,
+      "step": 15481
+    },
+    {
+      "epoch": 3.5186363636363636,
+      "grad_norm": 1.3670252025739997,
+      "learning_rate": 2.518103500033888e-07,
+      "loss": 0.0092,
+      "step": 15482
+    },
+    {
+      "epoch": 3.5188636363636365,
+      "grad_norm": 2.4233850980461007,
+      "learning_rate": 2.517387608168422e-07,
+      "loss": 0.0263,
+      "step": 15483
+    },
+    {
+      "epoch": 3.519090909090909,
+      "grad_norm": 2.2715726188514003,
+      "learning_rate": 2.5166717924174343e-07,
+      "loss": 0.0166,
+      "step": 15484
+    },
+    {
+      "epoch": 3.519318181818182,
+      "grad_norm": 1.2085897466271682,
+      "learning_rate": 2.515956052795524e-07,
+      "loss": 0.0085,
+      "step": 15485
+    },
+    {
+      "epoch": 3.5195454545454545,
+      "grad_norm": 1.0000508613397154,
+      "learning_rate": 2.515240389317282e-07,
+      "loss": 0.008,
+      "step": 15486
+    },
+    {
+      "epoch": 3.519772727272727,
+      "grad_norm": 1.4552957767032177,
+      "learning_rate": 2.514524801997304e-07,
+      "loss": 0.0126,
+      "step": 15487
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 1.1630478234751154,
+      "learning_rate": 2.513809290850182e-07,
+      "loss": 0.0069,
+      "step": 15488
+    },
+    {
+      "epoch": 3.520227272727273,
+      "grad_norm": 1.8032209465722813,
+      "learning_rate": 2.5130938558905094e-07,
+      "loss": 0.0185,
+      "step": 15489
+    },
+    {
+      "epoch": 3.5204545454545455,
+      "grad_norm": 1.143203212733499,
+      "learning_rate": 2.51237849713287e-07,
+      "loss": 0.0119,
+      "step": 15490
+    },
+    {
+      "epoch": 3.520681818181818,
+      "grad_norm": 1.441640048137532,
+      "learning_rate": 2.511663214591852e-07,
+      "loss": 0.0191,
+      "step": 15491
+    },
+    {
+      "epoch": 3.520909090909091,
+      "grad_norm": 2.943513556397138,
+      "learning_rate": 2.510948008282044e-07,
+      "loss": 0.0191,
+      "step": 15492
+    },
+    {
+      "epoch": 3.5211363636363635,
+      "grad_norm": 2.641270004289902,
+      "learning_rate": 2.510232878218032e-07,
+      "loss": 0.0096,
+      "step": 15493
+    },
+    {
+      "epoch": 3.5213636363636365,
+      "grad_norm": 1.9508053249324726,
+      "learning_rate": 2.509517824414394e-07,
+      "loss": 0.0161,
+      "step": 15494
+    },
+    {
+      "epoch": 3.521590909090909,
+      "grad_norm": 1.1577618268687122,
+      "learning_rate": 2.5088028468857135e-07,
+      "loss": 0.0054,
+      "step": 15495
+    },
+    {
+      "epoch": 3.521818181818182,
+      "grad_norm": 0.8598078906422414,
+      "learning_rate": 2.5080879456465696e-07,
+      "loss": 0.008,
+      "step": 15496
+    },
+    {
+      "epoch": 3.5220454545454545,
+      "grad_norm": 0.888731025320284,
+      "learning_rate": 2.5073731207115397e-07,
+      "loss": 0.0035,
+      "step": 15497
+    },
+    {
+      "epoch": 3.5222727272727274,
+      "grad_norm": 0.8351340072514363,
+      "learning_rate": 2.506658372095205e-07,
+      "loss": 0.007,
+      "step": 15498
+    },
+    {
+      "epoch": 3.5225,
+      "grad_norm": 0.785323916489356,
+      "learning_rate": 2.5059436998121333e-07,
+      "loss": 0.0064,
+      "step": 15499
+    },
+    {
+      "epoch": 3.5227272727272725,
+      "grad_norm": 1.957500908828941,
+      "learning_rate": 2.5052291038769017e-07,
+      "loss": 0.0212,
+      "step": 15500
+    },
+    {
+      "epoch": 3.5229545454545454,
+      "grad_norm": 1.3071290900468808,
+      "learning_rate": 2.5045145843040816e-07,
+      "loss": 0.01,
+      "step": 15501
+    },
+    {
+      "epoch": 3.5231818181818184,
+      "grad_norm": 1.3885443119773557,
+      "learning_rate": 2.5038001411082467e-07,
+      "loss": 0.0119,
+      "step": 15502
+    },
+    {
+      "epoch": 3.523409090909091,
+      "grad_norm": 1.1937928891458234,
+      "learning_rate": 2.503085774303959e-07,
+      "loss": 0.0078,
+      "step": 15503
+    },
+    {
+      "epoch": 3.5236363636363635,
+      "grad_norm": 2.2619480895685777,
+      "learning_rate": 2.50237148390579e-07,
+      "loss": 0.0159,
+      "step": 15504
+    },
+    {
+      "epoch": 3.5238636363636364,
+      "grad_norm": 3.4779521723879174,
+      "learning_rate": 2.5016572699283043e-07,
+      "loss": 0.0251,
+      "step": 15505
+    },
+    {
+      "epoch": 3.524090909090909,
+      "grad_norm": 1.3105491104041438,
+      "learning_rate": 2.500943132386066e-07,
+      "loss": 0.0065,
+      "step": 15506
+    },
+    {
+      "epoch": 3.524318181818182,
+      "grad_norm": 1.891078207697565,
+      "learning_rate": 2.500229071293638e-07,
+      "loss": 0.0086,
+      "step": 15507
+    },
+    {
+      "epoch": 3.5245454545454544,
+      "grad_norm": 3.369672439375092,
+      "learning_rate": 2.4995150866655813e-07,
+      "loss": 0.0089,
+      "step": 15508
+    },
+    {
+      "epoch": 3.5247727272727274,
+      "grad_norm": 2.70149894734578,
+      "learning_rate": 2.4988011785164553e-07,
+      "loss": 0.0419,
+      "step": 15509
+    },
+    {
+      "epoch": 3.525,
+      "grad_norm": 6.262566938201487,
+      "learning_rate": 2.498087346860817e-07,
+      "loss": 0.0199,
+      "step": 15510
+    },
+    {
+      "epoch": 3.525227272727273,
+      "grad_norm": 0.7741369173539343,
+      "learning_rate": 2.497373591713226e-07,
+      "loss": 0.0068,
+      "step": 15511
+    },
+    {
+      "epoch": 3.5254545454545454,
+      "grad_norm": 0.479827750172203,
+      "learning_rate": 2.4966599130882313e-07,
+      "loss": 0.0022,
+      "step": 15512
+    },
+    {
+      "epoch": 3.525681818181818,
+      "grad_norm": 0.9285782066998585,
+      "learning_rate": 2.49594631100039e-07,
+      "loss": 0.0086,
+      "step": 15513
+    },
+    {
+      "epoch": 3.525909090909091,
+      "grad_norm": 2.5264555993845486,
+      "learning_rate": 2.495232785464252e-07,
+      "loss": 0.011,
+      "step": 15514
+    },
+    {
+      "epoch": 3.526136363636364,
+      "grad_norm": 1.0196356489611513,
+      "learning_rate": 2.4945193364943707e-07,
+      "loss": 0.0124,
+      "step": 15515
+    },
+    {
+      "epoch": 3.5263636363636364,
+      "grad_norm": 1.5735843477599136,
+      "learning_rate": 2.4938059641052894e-07,
+      "loss": 0.0083,
+      "step": 15516
+    },
+    {
+      "epoch": 3.526590909090909,
+      "grad_norm": 2.6384429453719136,
+      "learning_rate": 2.493092668311557e-07,
+      "loss": 0.0145,
+      "step": 15517
+    },
+    {
+      "epoch": 3.526818181818182,
+      "grad_norm": 1.5894142971432141,
+      "learning_rate": 2.4923794491277205e-07,
+      "loss": 0.0099,
+      "step": 15518
+    },
+    {
+      "epoch": 3.527045454545455,
+      "grad_norm": 1.7336614238519803,
+      "learning_rate": 2.4916663065683223e-07,
+      "loss": 0.0133,
+      "step": 15519
+    },
+    {
+      "epoch": 3.5272727272727273,
+      "grad_norm": 1.8390803096599773,
+      "learning_rate": 2.4909532406479067e-07,
+      "loss": 0.012,
+      "step": 15520
+    },
+    {
+      "epoch": 3.5275,
+      "grad_norm": 0.7512551394232265,
+      "learning_rate": 2.490240251381008e-07,
+      "loss": 0.0028,
+      "step": 15521
+    },
+    {
+      "epoch": 3.527727272727273,
+      "grad_norm": 1.57452536763819,
+      "learning_rate": 2.4895273387821724e-07,
+      "loss": 0.0103,
+      "step": 15522
+    },
+    {
+      "epoch": 3.5279545454545453,
+      "grad_norm": 2.5938456264714076,
+      "learning_rate": 2.488814502865935e-07,
+      "loss": 0.0206,
+      "step": 15523
+    },
+    {
+      "epoch": 3.5281818181818183,
+      "grad_norm": 1.0881629237863315,
+      "learning_rate": 2.488101743646833e-07,
+      "loss": 0.0038,
+      "step": 15524
+    },
+    {
+      "epoch": 3.528409090909091,
+      "grad_norm": 1.2678450027315744,
+      "learning_rate": 2.487389061139398e-07,
+      "loss": 0.0086,
+      "step": 15525
+    },
+    {
+      "epoch": 3.528636363636364,
+      "grad_norm": 1.1792087245684322,
+      "learning_rate": 2.4866764553581647e-07,
+      "loss": 0.0115,
+      "step": 15526
+    },
+    {
+      "epoch": 3.5288636363636363,
+      "grad_norm": 1.4082998871776196,
+      "learning_rate": 2.4859639263176633e-07,
+      "loss": 0.0151,
+      "step": 15527
+    },
+    {
+      "epoch": 3.5290909090909093,
+      "grad_norm": 3.3714115682405508,
+      "learning_rate": 2.485251474032426e-07,
+      "loss": 0.0113,
+      "step": 15528
+    },
+    {
+      "epoch": 3.529318181818182,
+      "grad_norm": 1.0627514625588115,
+      "learning_rate": 2.4845390985169774e-07,
+      "loss": 0.0046,
+      "step": 15529
+    },
+    {
+      "epoch": 3.5295454545454543,
+      "grad_norm": 3.0388221210488333,
+      "learning_rate": 2.483826799785846e-07,
+      "loss": 0.0162,
+      "step": 15530
+    },
+    {
+      "epoch": 3.5297727272727273,
+      "grad_norm": 1.5326521308453254,
+      "learning_rate": 2.4831145778535563e-07,
+      "loss": 0.0095,
+      "step": 15531
+    },
+    {
+      "epoch": 3.5300000000000002,
+      "grad_norm": 2.286456722829759,
+      "learning_rate": 2.4824024327346325e-07,
+      "loss": 0.0094,
+      "step": 15532
+    },
+    {
+      "epoch": 3.5302272727272728,
+      "grad_norm": 1.1173960502398508,
+      "learning_rate": 2.481690364443597e-07,
+      "loss": 0.0093,
+      "step": 15533
+    },
+    {
+      "epoch": 3.5304545454545453,
+      "grad_norm": 1.890012850366083,
+      "learning_rate": 2.480978372994968e-07,
+      "loss": 0.0248,
+      "step": 15534
+    },
+    {
+      "epoch": 3.5306818181818183,
+      "grad_norm": 1.4560216678284523,
+      "learning_rate": 2.4802664584032655e-07,
+      "loss": 0.0098,
+      "step": 15535
+    },
+    {
+      "epoch": 3.5309090909090908,
+      "grad_norm": 1.9390972718895387,
+      "learning_rate": 2.479554620683006e-07,
+      "loss": 0.0091,
+      "step": 15536
+    },
+    {
+      "epoch": 3.5311363636363637,
+      "grad_norm": 1.9141793429108491,
+      "learning_rate": 2.478842859848706e-07,
+      "loss": 0.0204,
+      "step": 15537
+    },
+    {
+      "epoch": 3.5313636363636363,
+      "grad_norm": 1.3985768112991084,
+      "learning_rate": 2.4781311759148795e-07,
+      "loss": 0.0055,
+      "step": 15538
+    },
+    {
+      "epoch": 3.5315909090909092,
+      "grad_norm": 1.0425512328411888,
+      "learning_rate": 2.4774195688960384e-07,
+      "loss": 0.0049,
+      "step": 15539
+    },
+    {
+      "epoch": 3.5318181818181817,
+      "grad_norm": 0.9299129524942231,
+      "learning_rate": 2.4767080388066936e-07,
+      "loss": 0.0049,
+      "step": 15540
+    },
+    {
+      "epoch": 3.5320454545454547,
+      "grad_norm": 0.7759085631236868,
+      "learning_rate": 2.4759965856613554e-07,
+      "loss": 0.0038,
+      "step": 15541
+    },
+    {
+      "epoch": 3.5322727272727272,
+      "grad_norm": 1.6553368888806208,
+      "learning_rate": 2.4752852094745323e-07,
+      "loss": 0.0219,
+      "step": 15542
+    },
+    {
+      "epoch": 3.5324999999999998,
+      "grad_norm": 1.4004154559785402,
+      "learning_rate": 2.4745739102607277e-07,
+      "loss": 0.0054,
+      "step": 15543
+    },
+    {
+      "epoch": 3.5327272727272727,
+      "grad_norm": 3.740987678747039,
+      "learning_rate": 2.4738626880344475e-07,
+      "loss": 0.0444,
+      "step": 15544
+    },
+    {
+      "epoch": 3.5329545454545457,
+      "grad_norm": 2.218711397616507,
+      "learning_rate": 2.4731515428101945e-07,
+      "loss": 0.0146,
+      "step": 15545
+    },
+    {
+      "epoch": 3.533181818181818,
+      "grad_norm": 1.0492711266388128,
+      "learning_rate": 2.4724404746024737e-07,
+      "loss": 0.0097,
+      "step": 15546
+    },
+    {
+      "epoch": 3.5334090909090907,
+      "grad_norm": 2.3859037417781117,
+      "learning_rate": 2.4717294834257793e-07,
+      "loss": 0.0098,
+      "step": 15547
+    },
+    {
+      "epoch": 3.5336363636363637,
+      "grad_norm": 1.2838524627182957,
+      "learning_rate": 2.4710185692946116e-07,
+      "loss": 0.007,
+      "step": 15548
+    },
+    {
+      "epoch": 3.533863636363636,
+      "grad_norm": 1.3425018918714542,
+      "learning_rate": 2.470307732223469e-07,
+      "loss": 0.0073,
+      "step": 15549
+    },
+    {
+      "epoch": 3.534090909090909,
+      "grad_norm": 1.7525766376751675,
+      "learning_rate": 2.4695969722268453e-07,
+      "loss": 0.0134,
+      "step": 15550
+    },
+    {
+      "epoch": 3.5343181818181817,
+      "grad_norm": 1.8601037183727418,
+      "learning_rate": 2.468886289319235e-07,
+      "loss": 0.029,
+      "step": 15551
+    },
+    {
+      "epoch": 3.5345454545454547,
+      "grad_norm": 7.656971516700598,
+      "learning_rate": 2.468175683515129e-07,
+      "loss": 0.0178,
+      "step": 15552
+    },
+    {
+      "epoch": 3.534772727272727,
+      "grad_norm": 1.0196012894746351,
+      "learning_rate": 2.467465154829019e-07,
+      "loss": 0.0069,
+      "step": 15553
+    },
+    {
+      "epoch": 3.535,
+      "grad_norm": 1.7667682791513728,
+      "learning_rate": 2.466754703275393e-07,
+      "loss": 0.0161,
+      "step": 15554
+    },
+    {
+      "epoch": 3.5352272727272727,
+      "grad_norm": 1.582202963022989,
+      "learning_rate": 2.466044328868742e-07,
+      "loss": 0.0093,
+      "step": 15555
+    },
+    {
+      "epoch": 3.535454545454545,
+      "grad_norm": 1.3887723089555564,
+      "learning_rate": 2.465334031623546e-07,
+      "loss": 0.0085,
+      "step": 15556
+    },
+    {
+      "epoch": 3.535681818181818,
+      "grad_norm": 0.9591485795228105,
+      "learning_rate": 2.464623811554292e-07,
+      "loss": 0.0046,
+      "step": 15557
+    },
+    {
+      "epoch": 3.535909090909091,
+      "grad_norm": 2.140389681403438,
+      "learning_rate": 2.463913668675462e-07,
+      "loss": 0.0288,
+      "step": 15558
+    },
+    {
+      "epoch": 3.5361363636363636,
+      "grad_norm": 1.243280189003564,
+      "learning_rate": 2.46320360300154e-07,
+      "loss": 0.0065,
+      "step": 15559
+    },
+    {
+      "epoch": 3.536363636363636,
+      "grad_norm": 1.7575851247405463,
+      "learning_rate": 2.462493614547001e-07,
+      "loss": 0.0252,
+      "step": 15560
+    },
+    {
+      "epoch": 3.536590909090909,
+      "grad_norm": 1.0266356369775331,
+      "learning_rate": 2.461783703326324e-07,
+      "loss": 0.0065,
+      "step": 15561
+    },
+    {
+      "epoch": 3.536818181818182,
+      "grad_norm": 1.7648782128457525,
+      "learning_rate": 2.4610738693539867e-07,
+      "loss": 0.0158,
+      "step": 15562
+    },
+    {
+      "epoch": 3.5370454545454546,
+      "grad_norm": 3.134218817865125,
+      "learning_rate": 2.460364112644464e-07,
+      "loss": 0.0081,
+      "step": 15563
+    },
+    {
+      "epoch": 3.537272727272727,
+      "grad_norm": 1.9060748313516984,
+      "learning_rate": 2.4596544332122296e-07,
+      "loss": 0.01,
+      "step": 15564
+    },
+    {
+      "epoch": 3.5375,
+      "grad_norm": 1.328603046140892,
+      "learning_rate": 2.4589448310717523e-07,
+      "loss": 0.0046,
+      "step": 15565
+    },
+    {
+      "epoch": 3.5377272727272726,
+      "grad_norm": 0.947424054121937,
+      "learning_rate": 2.458235306237503e-07,
+      "loss": 0.0096,
+      "step": 15566
+    },
+    {
+      "epoch": 3.5379545454545456,
+      "grad_norm": 1.704333024099025,
+      "learning_rate": 2.4575258587239517e-07,
+      "loss": 0.0148,
+      "step": 15567
+    },
+    {
+      "epoch": 3.538181818181818,
+      "grad_norm": 0.9286543468895385,
+      "learning_rate": 2.4568164885455646e-07,
+      "loss": 0.009,
+      "step": 15568
+    },
+    {
+      "epoch": 3.538409090909091,
+      "grad_norm": 1.24610591519269,
+      "learning_rate": 2.456107195716806e-07,
+      "loss": 0.0059,
+      "step": 15569
+    },
+    {
+      "epoch": 3.5386363636363636,
+      "grad_norm": 1.3494012087005216,
+      "learning_rate": 2.455397980252141e-07,
+      "loss": 0.0072,
+      "step": 15570
+    },
+    {
+      "epoch": 3.5388636363636365,
+      "grad_norm": 0.861475072462516,
+      "learning_rate": 2.454688842166032e-07,
+      "loss": 0.0116,
+      "step": 15571
+    },
+    {
+      "epoch": 3.539090909090909,
+      "grad_norm": 2.362746979700265,
+      "learning_rate": 2.45397978147294e-07,
+      "loss": 0.0109,
+      "step": 15572
+    },
+    {
+      "epoch": 3.5393181818181816,
+      "grad_norm": 0.8037350931729879,
+      "learning_rate": 2.45327079818732e-07,
+      "loss": 0.0045,
+      "step": 15573
+    },
+    {
+      "epoch": 3.5395454545454546,
+      "grad_norm": 0.7495946169854265,
+      "learning_rate": 2.452561892323634e-07,
+      "loss": 0.0037,
+      "step": 15574
+    },
+    {
+      "epoch": 3.5397727272727275,
+      "grad_norm": 2.036168672041635,
+      "learning_rate": 2.4518530638963347e-07,
+      "loss": 0.0319,
+      "step": 15575
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.2810879021668344,
+      "learning_rate": 2.451144312919878e-07,
+      "loss": 0.0017,
+      "step": 15576
+    },
+    {
+      "epoch": 3.5402272727272726,
+      "grad_norm": 1.3381902574445852,
+      "learning_rate": 2.450435639408718e-07,
+      "loss": 0.0163,
+      "step": 15577
+    },
+    {
+      "epoch": 3.5404545454545455,
+      "grad_norm": 1.1985546642352534,
+      "learning_rate": 2.449727043377302e-07,
+      "loss": 0.0064,
+      "step": 15578
+    },
+    {
+      "epoch": 3.540681818181818,
+      "grad_norm": 0.8893778886981952,
+      "learning_rate": 2.449018524840082e-07,
+      "loss": 0.0064,
+      "step": 15579
+    },
+    {
+      "epoch": 3.540909090909091,
+      "grad_norm": 0.8189673107026124,
+      "learning_rate": 2.4483100838115054e-07,
+      "loss": 0.0044,
+      "step": 15580
+    },
+    {
+      "epoch": 3.5411363636363635,
+      "grad_norm": 1.8689437993286477,
+      "learning_rate": 2.4476017203060184e-07,
+      "loss": 0.0085,
+      "step": 15581
+    },
+    {
+      "epoch": 3.5413636363636365,
+      "grad_norm": 7.328902612154568,
+      "learning_rate": 2.4468934343380655e-07,
+      "loss": 0.0207,
+      "step": 15582
+    },
+    {
+      "epoch": 3.541590909090909,
+      "grad_norm": 1.6264901723868368,
+      "learning_rate": 2.446185225922091e-07,
+      "loss": 0.0146,
+      "step": 15583
+    },
+    {
+      "epoch": 3.541818181818182,
+      "grad_norm": 0.47991554702493056,
+      "learning_rate": 2.4454770950725357e-07,
+      "loss": 0.0016,
+      "step": 15584
+    },
+    {
+      "epoch": 3.5420454545454545,
+      "grad_norm": 1.693047206118526,
+      "learning_rate": 2.44476904180384e-07,
+      "loss": 0.0106,
+      "step": 15585
+    },
+    {
+      "epoch": 3.542272727272727,
+      "grad_norm": 1.3631179070625055,
+      "learning_rate": 2.444061066130444e-07,
+      "loss": 0.0065,
+      "step": 15586
+    },
+    {
+      "epoch": 3.5425,
+      "grad_norm": 2.451414978227188,
+      "learning_rate": 2.4433531680667805e-07,
+      "loss": 0.0194,
+      "step": 15587
+    },
+    {
+      "epoch": 3.542727272727273,
+      "grad_norm": 1.049633556025247,
+      "learning_rate": 2.442645347627287e-07,
+      "loss": 0.006,
+      "step": 15588
+    },
+    {
+      "epoch": 3.5429545454545455,
+      "grad_norm": 3.3379543831418244,
+      "learning_rate": 2.441937604826397e-07,
+      "loss": 0.0244,
+      "step": 15589
+    },
+    {
+      "epoch": 3.543181818181818,
+      "grad_norm": 2.83902429002377,
+      "learning_rate": 2.441229939678545e-07,
+      "loss": 0.0334,
+      "step": 15590
+    },
+    {
+      "epoch": 3.543409090909091,
+      "grad_norm": 0.7003828427409489,
+      "learning_rate": 2.4405223521981577e-07,
+      "loss": 0.0041,
+      "step": 15591
+    },
+    {
+      "epoch": 3.5436363636363635,
+      "grad_norm": 1.5722704928447415,
+      "learning_rate": 2.439814842399665e-07,
+      "loss": 0.0117,
+      "step": 15592
+    },
+    {
+      "epoch": 3.5438636363636364,
+      "grad_norm": 2.211101750443548,
+      "learning_rate": 2.439107410297495e-07,
+      "loss": 0.0127,
+      "step": 15593
+    },
+    {
+      "epoch": 3.544090909090909,
+      "grad_norm": 2.569465423465564,
+      "learning_rate": 2.4384000559060755e-07,
+      "loss": 0.0284,
+      "step": 15594
+    },
+    {
+      "epoch": 3.544318181818182,
+      "grad_norm": 1.4863029911022017,
+      "learning_rate": 2.4376927792398265e-07,
+      "loss": 0.0124,
+      "step": 15595
+    },
+    {
+      "epoch": 3.5445454545454544,
+      "grad_norm": 1.628050948538984,
+      "learning_rate": 2.4369855803131735e-07,
+      "loss": 0.0092,
+      "step": 15596
+    },
+    {
+      "epoch": 3.5447727272727274,
+      "grad_norm": 1.3313854648983756,
+      "learning_rate": 2.4362784591405357e-07,
+      "loss": 0.0095,
+      "step": 15597
+    },
+    {
+      "epoch": 3.545,
+      "grad_norm": 0.9879609887197077,
+      "learning_rate": 2.435571415736333e-07,
+      "loss": 0.0067,
+      "step": 15598
+    },
+    {
+      "epoch": 3.545227272727273,
+      "grad_norm": 1.6190706106222585,
+      "learning_rate": 2.4348644501149846e-07,
+      "loss": 0.0078,
+      "step": 15599
+    },
+    {
+      "epoch": 3.5454545454545454,
+      "grad_norm": 1.2840714626847873,
+      "learning_rate": 2.4341575622909054e-07,
+      "loss": 0.0113,
+      "step": 15600
+    },
+    {
+      "epoch": 3.5456818181818184,
+      "grad_norm": 0.9741130235038703,
+      "learning_rate": 2.433450752278511e-07,
+      "loss": 0.0107,
+      "step": 15601
+    },
+    {
+      "epoch": 3.545909090909091,
+      "grad_norm": 1.1705517206788105,
+      "learning_rate": 2.432744020092215e-07,
+      "loss": 0.0074,
+      "step": 15602
+    },
+    {
+      "epoch": 3.5461363636363634,
+      "grad_norm": 1.0304423833671728,
+      "learning_rate": 2.432037365746429e-07,
+      "loss": 0.0068,
+      "step": 15603
+    },
+    {
+      "epoch": 3.5463636363636364,
+      "grad_norm": 2.1687561359684606,
+      "learning_rate": 2.43133078925556e-07,
+      "loss": 0.0239,
+      "step": 15604
+    },
+    {
+      "epoch": 3.5465909090909093,
+      "grad_norm": 2.4937307185747994,
+      "learning_rate": 2.430624290634018e-07,
+      "loss": 0.0037,
+      "step": 15605
+    },
+    {
+      "epoch": 3.546818181818182,
+      "grad_norm": 0.6115296531228992,
+      "learning_rate": 2.4299178698962105e-07,
+      "loss": 0.0031,
+      "step": 15606
+    },
+    {
+      "epoch": 3.5470454545454544,
+      "grad_norm": 2.056187886045713,
+      "learning_rate": 2.4292115270565415e-07,
+      "loss": 0.0105,
+      "step": 15607
+    },
+    {
+      "epoch": 3.5472727272727274,
+      "grad_norm": 2.343895724471461,
+      "learning_rate": 2.428505262129418e-07,
+      "loss": 0.0148,
+      "step": 15608
+    },
+    {
+      "epoch": 3.5475,
+      "grad_norm": 1.1016940045397112,
+      "learning_rate": 2.427799075129237e-07,
+      "loss": 0.0059,
+      "step": 15609
+    },
+    {
+      "epoch": 3.547727272727273,
+      "grad_norm": 2.2477779936266096,
+      "learning_rate": 2.4270929660704005e-07,
+      "loss": 0.0125,
+      "step": 15610
+    },
+    {
+      "epoch": 3.5479545454545454,
+      "grad_norm": 0.7265107795350717,
+      "learning_rate": 2.4263869349673075e-07,
+      "loss": 0.0032,
+      "step": 15611
+    },
+    {
+      "epoch": 3.5481818181818183,
+      "grad_norm": 1.280269854993803,
+      "learning_rate": 2.4256809818343564e-07,
+      "loss": 0.0096,
+      "step": 15612
+    },
+    {
+      "epoch": 3.548409090909091,
+      "grad_norm": 2.116301614843995,
+      "learning_rate": 2.424975106685942e-07,
+      "loss": 0.0186,
+      "step": 15613
+    },
+    {
+      "epoch": 3.548636363636364,
+      "grad_norm": 1.3746628260850722,
+      "learning_rate": 2.424269309536458e-07,
+      "loss": 0.0066,
+      "step": 15614
+    },
+    {
+      "epoch": 3.5488636363636363,
+      "grad_norm": 1.0867863564262559,
+      "learning_rate": 2.423563590400297e-07,
+      "loss": 0.0089,
+      "step": 15615
+    },
+    {
+      "epoch": 3.549090909090909,
+      "grad_norm": 1.8983938201817259,
+      "learning_rate": 2.4228579492918527e-07,
+      "loss": 0.0087,
+      "step": 15616
+    },
+    {
+      "epoch": 3.549318181818182,
+      "grad_norm": 1.5266097116979938,
+      "learning_rate": 2.422152386225509e-07,
+      "loss": 0.0103,
+      "step": 15617
+    },
+    {
+      "epoch": 3.549545454545455,
+      "grad_norm": 1.2736211951582788,
+      "learning_rate": 2.4214469012156563e-07,
+      "loss": 0.0101,
+      "step": 15618
+    },
+    {
+      "epoch": 3.5497727272727273,
+      "grad_norm": 1.9840074248856896,
+      "learning_rate": 2.4207414942766806e-07,
+      "loss": 0.0096,
+      "step": 15619
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 1.2983401982482885,
+      "learning_rate": 2.420036165422966e-07,
+      "loss": 0.0066,
+      "step": 15620
+    },
+    {
+      "epoch": 3.550227272727273,
+      "grad_norm": 2.53002023436093,
+      "learning_rate": 2.4193309146688984e-07,
+      "loss": 0.0292,
+      "step": 15621
+    },
+    {
+      "epoch": 3.5504545454545453,
+      "grad_norm": 1.1715130808695344,
+      "learning_rate": 2.418625742028854e-07,
+      "loss": 0.0124,
+      "step": 15622
+    },
+    {
+      "epoch": 3.5506818181818183,
+      "grad_norm": 2.4369424352523446,
+      "learning_rate": 2.417920647517215e-07,
+      "loss": 0.0137,
+      "step": 15623
+    },
+    {
+      "epoch": 3.550909090909091,
+      "grad_norm": 1.5304219594205124,
+      "learning_rate": 2.417215631148359e-07,
+      "loss": 0.009,
+      "step": 15624
+    },
+    {
+      "epoch": 3.5511363636363638,
+      "grad_norm": 3.1966297477428034,
+      "learning_rate": 2.416510692936666e-07,
+      "loss": 0.015,
+      "step": 15625
+    },
+    {
+      "epoch": 3.5513636363636363,
+      "grad_norm": 1.3640626639326079,
+      "learning_rate": 2.4158058328965045e-07,
+      "loss": 0.0055,
+      "step": 15626
+    },
+    {
+      "epoch": 3.5515909090909092,
+      "grad_norm": 2.607651694491898,
+      "learning_rate": 2.4151010510422524e-07,
+      "loss": 0.0098,
+      "step": 15627
+    },
+    {
+      "epoch": 3.5518181818181818,
+      "grad_norm": 4.383569256859569,
+      "learning_rate": 2.4143963473882795e-07,
+      "loss": 0.0411,
+      "step": 15628
+    },
+    {
+      "epoch": 3.5520454545454543,
+      "grad_norm": 2.960622130517042,
+      "learning_rate": 2.413691721948957e-07,
+      "loss": 0.0228,
+      "step": 15629
+    },
+    {
+      "epoch": 3.5522727272727272,
+      "grad_norm": 2.277199114706291,
+      "learning_rate": 2.4129871747386525e-07,
+      "loss": 0.0195,
+      "step": 15630
+    },
+    {
+      "epoch": 3.5525,
+      "grad_norm": 1.9041871321257666,
+      "learning_rate": 2.4122827057717346e-07,
+      "loss": 0.0144,
+      "step": 15631
+    },
+    {
+      "epoch": 3.5527272727272727,
+      "grad_norm": 2.308417381582883,
+      "learning_rate": 2.411578315062567e-07,
+      "loss": 0.0189,
+      "step": 15632
+    },
+    {
+      "epoch": 3.5529545454545453,
+      "grad_norm": 2.1512990067244666,
+      "learning_rate": 2.4108740026255135e-07,
+      "loss": 0.0089,
+      "step": 15633
+    },
+    {
+      "epoch": 3.553181818181818,
+      "grad_norm": 1.6805519399597357,
+      "learning_rate": 2.4101697684749396e-07,
+      "loss": 0.0055,
+      "step": 15634
+    },
+    {
+      "epoch": 3.553409090909091,
+      "grad_norm": 1.8640759523338808,
+      "learning_rate": 2.409465612625201e-07,
+      "loss": 0.0126,
+      "step": 15635
+    },
+    {
+      "epoch": 3.5536363636363637,
+      "grad_norm": 0.698007077574725,
+      "learning_rate": 2.4087615350906584e-07,
+      "loss": 0.008,
+      "step": 15636
+    },
+    {
+      "epoch": 3.5538636363636362,
+      "grad_norm": 2.2608359166449334,
+      "learning_rate": 2.40805753588567e-07,
+      "loss": 0.0124,
+      "step": 15637
+    },
+    {
+      "epoch": 3.554090909090909,
+      "grad_norm": 1.6940737579634084,
+      "learning_rate": 2.407353615024593e-07,
+      "loss": 0.0174,
+      "step": 15638
+    },
+    {
+      "epoch": 3.5543181818181817,
+      "grad_norm": 6.0911662349133495,
+      "learning_rate": 2.406649772521777e-07,
+      "loss": 0.0301,
+      "step": 15639
+    },
+    {
+      "epoch": 3.5545454545454547,
+      "grad_norm": 1.6076712721319004,
+      "learning_rate": 2.405946008391578e-07,
+      "loss": 0.0076,
+      "step": 15640
+    },
+    {
+      "epoch": 3.554772727272727,
+      "grad_norm": 2.1638962239106445,
+      "learning_rate": 2.405242322648346e-07,
+      "loss": 0.0049,
+      "step": 15641
+    },
+    {
+      "epoch": 3.555,
+      "grad_norm": 1.2004878283281717,
+      "learning_rate": 2.404538715306429e-07,
+      "loss": 0.0059,
+      "step": 15642
+    },
+    {
+      "epoch": 3.5552272727272727,
+      "grad_norm": 1.934301774256975,
+      "learning_rate": 2.403835186380178e-07,
+      "loss": 0.0118,
+      "step": 15643
+    },
+    {
+      "epoch": 3.5554545454545456,
+      "grad_norm": 0.9374030944064513,
+      "learning_rate": 2.403131735883938e-07,
+      "loss": 0.004,
+      "step": 15644
+    },
+    {
+      "epoch": 3.555681818181818,
+      "grad_norm": 3.610543330270333,
+      "learning_rate": 2.402428363832052e-07,
+      "loss": 0.0367,
+      "step": 15645
+    },
+    {
+      "epoch": 3.5559090909090907,
+      "grad_norm": 3.834725882415928,
+      "learning_rate": 2.4017250702388644e-07,
+      "loss": 0.0248,
+      "step": 15646
+    },
+    {
+      "epoch": 3.5561363636363637,
+      "grad_norm": 1.794148112546381,
+      "learning_rate": 2.401021855118719e-07,
+      "loss": 0.0055,
+      "step": 15647
+    },
+    {
+      "epoch": 3.5563636363636366,
+      "grad_norm": 1.9038507183228421,
+      "learning_rate": 2.4003187184859506e-07,
+      "loss": 0.012,
+      "step": 15648
+    },
+    {
+      "epoch": 3.556590909090909,
+      "grad_norm": 0.790771429311891,
+      "learning_rate": 2.3996156603548994e-07,
+      "loss": 0.0038,
+      "step": 15649
+    },
+    {
+      "epoch": 3.5568181818181817,
+      "grad_norm": 1.9107906648991508,
+      "learning_rate": 2.3989126807399024e-07,
+      "loss": 0.0101,
+      "step": 15650
+    },
+    {
+      "epoch": 3.5570454545454546,
+      "grad_norm": 1.3514888646869425,
+      "learning_rate": 2.3982097796552966e-07,
+      "loss": 0.0119,
+      "step": 15651
+    },
+    {
+      "epoch": 3.557272727272727,
+      "grad_norm": 1.1726481437410343,
+      "learning_rate": 2.3975069571154107e-07,
+      "loss": 0.0049,
+      "step": 15652
+    },
+    {
+      "epoch": 3.5575,
+      "grad_norm": 1.2510304675868082,
+      "learning_rate": 2.396804213134579e-07,
+      "loss": 0.0096,
+      "step": 15653
+    },
+    {
+      "epoch": 3.5577272727272726,
+      "grad_norm": 0.8888935044830336,
+      "learning_rate": 2.3961015477271315e-07,
+      "loss": 0.0057,
+      "step": 15654
+    },
+    {
+      "epoch": 3.5579545454545456,
+      "grad_norm": 1.5087753013066283,
+      "learning_rate": 2.3953989609073976e-07,
+      "loss": 0.0061,
+      "step": 15655
+    },
+    {
+      "epoch": 3.558181818181818,
+      "grad_norm": 0.659216810978979,
+      "learning_rate": 2.3946964526897054e-07,
+      "loss": 0.0052,
+      "step": 15656
+    },
+    {
+      "epoch": 3.558409090909091,
+      "grad_norm": 1.3219274761930686,
+      "learning_rate": 2.393994023088377e-07,
+      "loss": 0.008,
+      "step": 15657
+    },
+    {
+      "epoch": 3.5586363636363636,
+      "grad_norm": 1.8297689490555133,
+      "learning_rate": 2.393291672117737e-07,
+      "loss": 0.0108,
+      "step": 15658
+    },
+    {
+      "epoch": 3.558863636363636,
+      "grad_norm": 0.9104264402617693,
+      "learning_rate": 2.392589399792106e-07,
+      "loss": 0.0067,
+      "step": 15659
+    },
+    {
+      "epoch": 3.559090909090909,
+      "grad_norm": 0.922947258749597,
+      "learning_rate": 2.3918872061258103e-07,
+      "loss": 0.0088,
+      "step": 15660
+    },
+    {
+      "epoch": 3.559318181818182,
+      "grad_norm": 0.5235837080515938,
+      "learning_rate": 2.391185091133164e-07,
+      "loss": 0.0061,
+      "step": 15661
+    },
+    {
+      "epoch": 3.5595454545454546,
+      "grad_norm": 1.7812970718066266,
+      "learning_rate": 2.3904830548284865e-07,
+      "loss": 0.0071,
+      "step": 15662
+    },
+    {
+      "epoch": 3.559772727272727,
+      "grad_norm": 2.8500246523747212,
+      "learning_rate": 2.3897810972260916e-07,
+      "loss": 0.0113,
+      "step": 15663
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 2.1613100963433554,
+      "learning_rate": 2.3890792183402955e-07,
+      "loss": 0.0174,
+      "step": 15664
+    },
+    {
+      "epoch": 3.5602272727272726,
+      "grad_norm": 3.103699748940164,
+      "learning_rate": 2.3883774181854114e-07,
+      "loss": 0.0225,
+      "step": 15665
+    },
+    {
+      "epoch": 3.5604545454545455,
+      "grad_norm": 1.331132349887661,
+      "learning_rate": 2.387675696775747e-07,
+      "loss": 0.0073,
+      "step": 15666
+    },
+    {
+      "epoch": 3.560681818181818,
+      "grad_norm": 1.360914258583074,
+      "learning_rate": 2.386974054125613e-07,
+      "loss": 0.0133,
+      "step": 15667
+    },
+    {
+      "epoch": 3.560909090909091,
+      "grad_norm": 1.3888171916771843,
+      "learning_rate": 2.3862724902493177e-07,
+      "loss": 0.0189,
+      "step": 15668
+    },
+    {
+      "epoch": 3.5611363636363635,
+      "grad_norm": 0.7723276005899102,
+      "learning_rate": 2.385571005161169e-07,
+      "loss": 0.0063,
+      "step": 15669
+    },
+    {
+      "epoch": 3.5613636363636365,
+      "grad_norm": 1.2143080806394233,
+      "learning_rate": 2.3848695988754674e-07,
+      "loss": 0.0091,
+      "step": 15670
+    },
+    {
+      "epoch": 3.561590909090909,
+      "grad_norm": 0.20086121206207733,
+      "learning_rate": 2.3841682714065172e-07,
+      "loss": 0.0008,
+      "step": 15671
+    },
+    {
+      "epoch": 3.5618181818181816,
+      "grad_norm": 1.3344249478720627,
+      "learning_rate": 2.3834670227686208e-07,
+      "loss": 0.0058,
+      "step": 15672
+    },
+    {
+      "epoch": 3.5620454545454545,
+      "grad_norm": 2.0843837175675675,
+      "learning_rate": 2.3827658529760774e-07,
+      "loss": 0.0206,
+      "step": 15673
+    },
+    {
+      "epoch": 3.5622727272727275,
+      "grad_norm": 1.1137600585762584,
+      "learning_rate": 2.3820647620431847e-07,
+      "loss": 0.0119,
+      "step": 15674
+    },
+    {
+      "epoch": 3.5625,
+      "grad_norm": 2.452695008214434,
+      "learning_rate": 2.3813637499842398e-07,
+      "loss": 0.0076,
+      "step": 15675
+    },
+    {
+      "epoch": 3.5627272727272725,
+      "grad_norm": 1.4213883743078903,
+      "learning_rate": 2.3806628168135372e-07,
+      "loss": 0.0206,
+      "step": 15676
+    },
+    {
+      "epoch": 3.5629545454545455,
+      "grad_norm": 2.0681184840556566,
+      "learning_rate": 2.3799619625453698e-07,
+      "loss": 0.0309,
+      "step": 15677
+    },
+    {
+      "epoch": 3.5631818181818184,
+      "grad_norm": 3.9808825860929056,
+      "learning_rate": 2.3792611871940325e-07,
+      "loss": 0.0135,
+      "step": 15678
+    },
+    {
+      "epoch": 3.563409090909091,
+      "grad_norm": 2.5649416579483324,
+      "learning_rate": 2.3785604907738096e-07,
+      "loss": 0.0072,
+      "step": 15679
+    },
+    {
+      "epoch": 3.5636363636363635,
+      "grad_norm": 2.258062767267574,
+      "learning_rate": 2.3778598732989926e-07,
+      "loss": 0.0091,
+      "step": 15680
+    },
+    {
+      "epoch": 3.5638636363636365,
+      "grad_norm": 1.957922751838959,
+      "learning_rate": 2.3771593347838685e-07,
+      "loss": 0.0148,
+      "step": 15681
+    },
+    {
+      "epoch": 3.564090909090909,
+      "grad_norm": 0.8814813066154183,
+      "learning_rate": 2.3764588752427236e-07,
+      "loss": 0.0021,
+      "step": 15682
+    },
+    {
+      "epoch": 3.564318181818182,
+      "grad_norm": 1.6913620275119476,
+      "learning_rate": 2.375758494689838e-07,
+      "loss": 0.0194,
+      "step": 15683
+    },
+    {
+      "epoch": 3.5645454545454545,
+      "grad_norm": 2.1397279251663015,
+      "learning_rate": 2.3750581931394963e-07,
+      "loss": 0.0209,
+      "step": 15684
+    },
+    {
+      "epoch": 3.5647727272727274,
+      "grad_norm": 0.6640129239897827,
+      "learning_rate": 2.3743579706059776e-07,
+      "loss": 0.0028,
+      "step": 15685
+    },
+    {
+      "epoch": 3.565,
+      "grad_norm": 1.2260473406170245,
+      "learning_rate": 2.3736578271035615e-07,
+      "loss": 0.0087,
+      "step": 15686
+    },
+    {
+      "epoch": 3.565227272727273,
+      "grad_norm": 1.6248592594806663,
+      "learning_rate": 2.372957762646527e-07,
+      "loss": 0.0141,
+      "step": 15687
+    },
+    {
+      "epoch": 3.5654545454545454,
+      "grad_norm": 0.4585073192681158,
+      "learning_rate": 2.3722577772491435e-07,
+      "loss": 0.0018,
+      "step": 15688
+    },
+    {
+      "epoch": 3.565681818181818,
+      "grad_norm": 0.6978905552828609,
+      "learning_rate": 2.3715578709256914e-07,
+      "loss": 0.0111,
+      "step": 15689
+    },
+    {
+      "epoch": 3.565909090909091,
+      "grad_norm": 1.5733324172614724,
+      "learning_rate": 2.3708580436904396e-07,
+      "loss": 0.0111,
+      "step": 15690
+    },
+    {
+      "epoch": 3.566136363636364,
+      "grad_norm": 0.7464688587326498,
+      "learning_rate": 2.3701582955576632e-07,
+      "loss": 0.0056,
+      "step": 15691
+    },
+    {
+      "epoch": 3.5663636363636364,
+      "grad_norm": 3.254494734593514,
+      "learning_rate": 2.369458626541625e-07,
+      "loss": 0.0205,
+      "step": 15692
+    },
+    {
+      "epoch": 3.566590909090909,
+      "grad_norm": 1.615766246556954,
+      "learning_rate": 2.3687590366565956e-07,
+      "loss": 0.0061,
+      "step": 15693
+    },
+    {
+      "epoch": 3.566818181818182,
+      "grad_norm": 1.4738160555101152,
+      "learning_rate": 2.3680595259168403e-07,
+      "loss": 0.0177,
+      "step": 15694
+    },
+    {
+      "epoch": 3.5670454545454544,
+      "grad_norm": 1.6767230850906594,
+      "learning_rate": 2.3673600943366265e-07,
+      "loss": 0.0155,
+      "step": 15695
+    },
+    {
+      "epoch": 3.5672727272727274,
+      "grad_norm": 1.4784007918840378,
+      "learning_rate": 2.3666607419302114e-07,
+      "loss": 0.0107,
+      "step": 15696
+    },
+    {
+      "epoch": 3.5675,
+      "grad_norm": 1.5540565971678113,
+      "learning_rate": 2.365961468711859e-07,
+      "loss": 0.01,
+      "step": 15697
+    },
+    {
+      "epoch": 3.567727272727273,
+      "grad_norm": 2.0402924909001077,
+      "learning_rate": 2.365262274695828e-07,
+      "loss": 0.0072,
+      "step": 15698
+    },
+    {
+      "epoch": 3.5679545454545454,
+      "grad_norm": 2.8216416316008908,
+      "learning_rate": 2.3645631598963763e-07,
+      "loss": 0.0284,
+      "step": 15699
+    },
+    {
+      "epoch": 3.5681818181818183,
+      "grad_norm": 2.132048075171253,
+      "learning_rate": 2.363864124327763e-07,
+      "loss": 0.0168,
+      "step": 15700
+    },
+    {
+      "epoch": 3.568409090909091,
+      "grad_norm": 2.617689946475249,
+      "learning_rate": 2.3631651680042374e-07,
+      "loss": 0.0264,
+      "step": 15701
+    },
+    {
+      "epoch": 3.5686363636363634,
+      "grad_norm": 0.9533755292553764,
+      "learning_rate": 2.3624662909400553e-07,
+      "loss": 0.0016,
+      "step": 15702
+    },
+    {
+      "epoch": 3.5688636363636363,
+      "grad_norm": 1.3744839025757163,
+      "learning_rate": 2.3617674931494673e-07,
+      "loss": 0.021,
+      "step": 15703
+    },
+    {
+      "epoch": 3.5690909090909093,
+      "grad_norm": 1.6743844070183993,
+      "learning_rate": 2.3610687746467236e-07,
+      "loss": 0.0061,
+      "step": 15704
+    },
+    {
+      "epoch": 3.569318181818182,
+      "grad_norm": 1.973171349832979,
+      "learning_rate": 2.3603701354460725e-07,
+      "loss": 0.0235,
+      "step": 15705
+    },
+    {
+      "epoch": 3.5695454545454544,
+      "grad_norm": 2.687449473723099,
+      "learning_rate": 2.3596715755617602e-07,
+      "loss": 0.0415,
+      "step": 15706
+    },
+    {
+      "epoch": 3.5697727272727273,
+      "grad_norm": 1.2570514881839587,
+      "learning_rate": 2.3589730950080312e-07,
+      "loss": 0.0052,
+      "step": 15707
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 2.4926462504839484,
+      "learning_rate": 2.3582746937991293e-07,
+      "loss": 0.0065,
+      "step": 15708
+    },
+    {
+      "epoch": 3.570227272727273,
+      "grad_norm": 1.1171949828919352,
+      "learning_rate": 2.357576371949298e-07,
+      "loss": 0.007,
+      "step": 15709
+    },
+    {
+      "epoch": 3.5704545454545453,
+      "grad_norm": 1.847583139149406,
+      "learning_rate": 2.3568781294727728e-07,
+      "loss": 0.0065,
+      "step": 15710
+    },
+    {
+      "epoch": 3.5706818181818183,
+      "grad_norm": 1.1084793318733632,
+      "learning_rate": 2.3561799663837945e-07,
+      "loss": 0.0085,
+      "step": 15711
+    },
+    {
+      "epoch": 3.570909090909091,
+      "grad_norm": 2.7240393166556496,
+      "learning_rate": 2.3554818826966004e-07,
+      "loss": 0.0137,
+      "step": 15712
+    },
+    {
+      "epoch": 3.5711363636363638,
+      "grad_norm": 1.5484508214763897,
+      "learning_rate": 2.3547838784254263e-07,
+      "loss": 0.0108,
+      "step": 15713
+    },
+    {
+      "epoch": 3.5713636363636363,
+      "grad_norm": 1.204217700159844,
+      "learning_rate": 2.3540859535845026e-07,
+      "loss": 0.007,
+      "step": 15714
+    },
+    {
+      "epoch": 3.571590909090909,
+      "grad_norm": 2.920662716996726,
+      "learning_rate": 2.3533881081880635e-07,
+      "loss": 0.0072,
+      "step": 15715
+    },
+    {
+      "epoch": 3.571818181818182,
+      "grad_norm": 2.920140894406062,
+      "learning_rate": 2.352690342250339e-07,
+      "loss": 0.0216,
+      "step": 15716
+    },
+    {
+      "epoch": 3.5720454545454547,
+      "grad_norm": 1.4399163744438057,
+      "learning_rate": 2.3519926557855592e-07,
+      "loss": 0.0068,
+      "step": 15717
+    },
+    {
+      "epoch": 3.5722727272727273,
+      "grad_norm": 1.2438776303382182,
+      "learning_rate": 2.3512950488079458e-07,
+      "loss": 0.0085,
+      "step": 15718
+    },
+    {
+      "epoch": 3.5725,
+      "grad_norm": 1.121430950756241,
+      "learning_rate": 2.3505975213317294e-07,
+      "loss": 0.0042,
+      "step": 15719
+    },
+    {
+      "epoch": 3.5727272727272728,
+      "grad_norm": 0.7418252955318385,
+      "learning_rate": 2.3499000733711336e-07,
+      "loss": 0.0054,
+      "step": 15720
+    },
+    {
+      "epoch": 3.5729545454545457,
+      "grad_norm": 1.6578312851027446,
+      "learning_rate": 2.3492027049403788e-07,
+      "loss": 0.0125,
+      "step": 15721
+    },
+    {
+      "epoch": 3.5731818181818182,
+      "grad_norm": 2.047387608627983,
+      "learning_rate": 2.3485054160536883e-07,
+      "loss": 0.0188,
+      "step": 15722
+    },
+    {
+      "epoch": 3.5734090909090908,
+      "grad_norm": 1.9496637414957756,
+      "learning_rate": 2.3478082067252764e-07,
+      "loss": 0.0129,
+      "step": 15723
+    },
+    {
+      "epoch": 3.5736363636363637,
+      "grad_norm": 1.3958989560955852,
+      "learning_rate": 2.3471110769693633e-07,
+      "loss": 0.0072,
+      "step": 15724
+    },
+    {
+      "epoch": 3.5738636363636362,
+      "grad_norm": 1.6961912092991631,
+      "learning_rate": 2.3464140268001644e-07,
+      "loss": 0.0143,
+      "step": 15725
+    },
+    {
+      "epoch": 3.574090909090909,
+      "grad_norm": 1.2727934553104456,
+      "learning_rate": 2.3457170562318957e-07,
+      "loss": 0.0054,
+      "step": 15726
+    },
+    {
+      "epoch": 3.5743181818181817,
+      "grad_norm": 1.0797368105350353,
+      "learning_rate": 2.345020165278765e-07,
+      "loss": 0.0036,
+      "step": 15727
+    },
+    {
+      "epoch": 3.5745454545454547,
+      "grad_norm": 2.5614894275866193,
+      "learning_rate": 2.3443233539549865e-07,
+      "loss": 0.0255,
+      "step": 15728
+    },
+    {
+      "epoch": 3.574772727272727,
+      "grad_norm": 1.244854234071212,
+      "learning_rate": 2.343626622274768e-07,
+      "loss": 0.0106,
+      "step": 15729
+    },
+    {
+      "epoch": 3.575,
+      "grad_norm": 1.594081604751603,
+      "learning_rate": 2.3429299702523182e-07,
+      "loss": 0.0084,
+      "step": 15730
+    },
+    {
+      "epoch": 3.5752272727272727,
+      "grad_norm": 1.9175818013815116,
+      "learning_rate": 2.342233397901845e-07,
+      "loss": 0.0187,
+      "step": 15731
+    },
+    {
+      "epoch": 3.575454545454545,
+      "grad_norm": 0.7513775580820059,
+      "learning_rate": 2.3415369052375473e-07,
+      "loss": 0.0062,
+      "step": 15732
+    },
+    {
+      "epoch": 3.575681818181818,
+      "grad_norm": 1.128110913401984,
+      "learning_rate": 2.3408404922736314e-07,
+      "loss": 0.0079,
+      "step": 15733
+    },
+    {
+      "epoch": 3.575909090909091,
+      "grad_norm": 2.141032845692598,
+      "learning_rate": 2.3401441590242969e-07,
+      "loss": 0.0143,
+      "step": 15734
+    },
+    {
+      "epoch": 3.5761363636363637,
+      "grad_norm": 0.908238335092314,
+      "learning_rate": 2.339447905503745e-07,
+      "loss": 0.0049,
+      "step": 15735
+    },
+    {
+      "epoch": 3.576363636363636,
+      "grad_norm": 0.9009404114698019,
+      "learning_rate": 2.3387517317261723e-07,
+      "loss": 0.0043,
+      "step": 15736
+    },
+    {
+      "epoch": 3.576590909090909,
+      "grad_norm": 1.2454149145175466,
+      "learning_rate": 2.3380556377057758e-07,
+      "loss": 0.0099,
+      "step": 15737
+    },
+    {
+      "epoch": 3.5768181818181817,
+      "grad_norm": 0.9179988754430087,
+      "learning_rate": 2.3373596234567488e-07,
+      "loss": 0.0052,
+      "step": 15738
+    },
+    {
+      "epoch": 3.5770454545454546,
+      "grad_norm": 1.4419579228910655,
+      "learning_rate": 2.3366636889932882e-07,
+      "loss": 0.0067,
+      "step": 15739
+    },
+    {
+      "epoch": 3.577272727272727,
+      "grad_norm": 1.949706258047337,
+      "learning_rate": 2.335967834329579e-07,
+      "loss": 0.0201,
+      "step": 15740
+    },
+    {
+      "epoch": 3.5775,
+      "grad_norm": 1.4519778600096496,
+      "learning_rate": 2.335272059479815e-07,
+      "loss": 0.0087,
+      "step": 15741
+    },
+    {
+      "epoch": 3.5777272727272726,
+      "grad_norm": 5.53325018771872,
+      "learning_rate": 2.3345763644581827e-07,
+      "loss": 0.0174,
+      "step": 15742
+    },
+    {
+      "epoch": 3.5779545454545456,
+      "grad_norm": 0.7925278735198444,
+      "learning_rate": 2.333880749278869e-07,
+      "loss": 0.006,
+      "step": 15743
+    },
+    {
+      "epoch": 3.578181818181818,
+      "grad_norm": 1.4660389164136745,
+      "learning_rate": 2.3331852139560614e-07,
+      "loss": 0.0097,
+      "step": 15744
+    },
+    {
+      "epoch": 3.5784090909090907,
+      "grad_norm": 2.851676772031119,
+      "learning_rate": 2.3324897585039383e-07,
+      "loss": 0.0123,
+      "step": 15745
+    },
+    {
+      "epoch": 3.5786363636363636,
+      "grad_norm": 0.37419569106038914,
+      "learning_rate": 2.3317943829366833e-07,
+      "loss": 0.0019,
+      "step": 15746
+    },
+    {
+      "epoch": 3.5788636363636366,
+      "grad_norm": 1.6724967333824612,
+      "learning_rate": 2.331099087268477e-07,
+      "loss": 0.0204,
+      "step": 15747
+    },
+    {
+      "epoch": 3.579090909090909,
+      "grad_norm": 0.9536786970486711,
+      "learning_rate": 2.3304038715134973e-07,
+      "loss": 0.0033,
+      "step": 15748
+    },
+    {
+      "epoch": 3.5793181818181816,
+      "grad_norm": 1.299300948483727,
+      "learning_rate": 2.3297087356859206e-07,
+      "loss": 0.0135,
+      "step": 15749
+    },
+    {
+      "epoch": 3.5795454545454546,
+      "grad_norm": 1.8896504365225508,
+      "learning_rate": 2.3290136797999214e-07,
+      "loss": 0.0256,
+      "step": 15750
+    },
+    {
+      "epoch": 3.579772727272727,
+      "grad_norm": 8.850236199630748,
+      "learning_rate": 2.3283187038696746e-07,
+      "loss": 0.0148,
+      "step": 15751
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 1.5953268333043444,
+      "learning_rate": 2.327623807909353e-07,
+      "loss": 0.0114,
+      "step": 15752
+    },
+    {
+      "epoch": 3.5802272727272726,
+      "grad_norm": 1.5354314581878539,
+      "learning_rate": 2.326928991933123e-07,
+      "loss": 0.006,
+      "step": 15753
+    },
+    {
+      "epoch": 3.5804545454545456,
+      "grad_norm": 1.234081003178225,
+      "learning_rate": 2.3262342559551544e-07,
+      "loss": 0.0135,
+      "step": 15754
+    },
+    {
+      "epoch": 3.580681818181818,
+      "grad_norm": 2.4172308745634234,
+      "learning_rate": 2.3255395999896153e-07,
+      "loss": 0.0108,
+      "step": 15755
+    },
+    {
+      "epoch": 3.580909090909091,
+      "grad_norm": 1.2582367013063809,
+      "learning_rate": 2.3248450240506704e-07,
+      "loss": 0.0076,
+      "step": 15756
+    },
+    {
+      "epoch": 3.5811363636363636,
+      "grad_norm": 1.5008305733876437,
+      "learning_rate": 2.324150528152485e-07,
+      "loss": 0.0113,
+      "step": 15757
+    },
+    {
+      "epoch": 3.5813636363636365,
+      "grad_norm": 1.9602950431101307,
+      "learning_rate": 2.3234561123092172e-07,
+      "loss": 0.0082,
+      "step": 15758
+    },
+    {
+      "epoch": 3.581590909090909,
+      "grad_norm": 1.0762921752155334,
+      "learning_rate": 2.3227617765350298e-07,
+      "loss": 0.0058,
+      "step": 15759
+    },
+    {
+      "epoch": 3.581818181818182,
+      "grad_norm": 1.3924580773885686,
+      "learning_rate": 2.3220675208440811e-07,
+      "loss": 0.017,
+      "step": 15760
+    },
+    {
+      "epoch": 3.5820454545454545,
+      "grad_norm": 0.7712270300066031,
+      "learning_rate": 2.3213733452505307e-07,
+      "loss": 0.0084,
+      "step": 15761
+    },
+    {
+      "epoch": 3.582272727272727,
+      "grad_norm": 0.6979211693992453,
+      "learning_rate": 2.3206792497685295e-07,
+      "loss": 0.0034,
+      "step": 15762
+    },
+    {
+      "epoch": 3.5825,
+      "grad_norm": 2.681666795843775,
+      "learning_rate": 2.319985234412233e-07,
+      "loss": 0.0159,
+      "step": 15763
+    },
+    {
+      "epoch": 3.582727272727273,
+      "grad_norm": 0.8076061784482425,
+      "learning_rate": 2.3192912991957944e-07,
+      "loss": 0.0074,
+      "step": 15764
+    },
+    {
+      "epoch": 3.5829545454545455,
+      "grad_norm": 1.4184265224784158,
+      "learning_rate": 2.3185974441333629e-07,
+      "loss": 0.0033,
+      "step": 15765
+    },
+    {
+      "epoch": 3.583181818181818,
+      "grad_norm": 3.605605752454988,
+      "learning_rate": 2.317903669239089e-07,
+      "loss": 0.0266,
+      "step": 15766
+    },
+    {
+      "epoch": 3.583409090909091,
+      "grad_norm": 2.139845257902514,
+      "learning_rate": 2.3172099745271188e-07,
+      "loss": 0.0214,
+      "step": 15767
+    },
+    {
+      "epoch": 3.5836363636363635,
+      "grad_norm": 1.9485979714380008,
+      "learning_rate": 2.316516360011599e-07,
+      "loss": 0.006,
+      "step": 15768
+    },
+    {
+      "epoch": 3.5838636363636365,
+      "grad_norm": 1.8596947156813655,
+      "learning_rate": 2.315822825706672e-07,
+      "loss": 0.0129,
+      "step": 15769
+    },
+    {
+      "epoch": 3.584090909090909,
+      "grad_norm": 1.713515403906045,
+      "learning_rate": 2.3151293716264838e-07,
+      "loss": 0.0143,
+      "step": 15770
+    },
+    {
+      "epoch": 3.584318181818182,
+      "grad_norm": 2.0065779529970227,
+      "learning_rate": 2.3144359977851708e-07,
+      "loss": 0.0043,
+      "step": 15771
+    },
+    {
+      "epoch": 3.5845454545454545,
+      "grad_norm": 0.46757227034818416,
+      "learning_rate": 2.3137427041968732e-07,
+      "loss": 0.0025,
+      "step": 15772
+    },
+    {
+      "epoch": 3.5847727272727274,
+      "grad_norm": 1.6920722420819343,
+      "learning_rate": 2.3130494908757288e-07,
+      "loss": 0.0112,
+      "step": 15773
+    },
+    {
+      "epoch": 3.585,
+      "grad_norm": 1.5193762836891673,
+      "learning_rate": 2.3123563578358758e-07,
+      "loss": 0.0127,
+      "step": 15774
+    },
+    {
+      "epoch": 3.5852272727272725,
+      "grad_norm": 0.7054472984792006,
+      "learning_rate": 2.3116633050914446e-07,
+      "loss": 0.0036,
+      "step": 15775
+    },
+    {
+      "epoch": 3.5854545454545454,
+      "grad_norm": 3.541343715557223,
+      "learning_rate": 2.3109703326565696e-07,
+      "loss": 0.009,
+      "step": 15776
+    },
+    {
+      "epoch": 3.5856818181818184,
+      "grad_norm": 0.6888608406016684,
+      "learning_rate": 2.3102774405453812e-07,
+      "loss": 0.0058,
+      "step": 15777
+    },
+    {
+      "epoch": 3.585909090909091,
+      "grad_norm": 1.4599454674376757,
+      "learning_rate": 2.3095846287720088e-07,
+      "loss": 0.0063,
+      "step": 15778
+    },
+    {
+      "epoch": 3.5861363636363635,
+      "grad_norm": 1.2133659075823418,
+      "learning_rate": 2.3088918973505804e-07,
+      "loss": 0.0057,
+      "step": 15779
+    },
+    {
+      "epoch": 3.5863636363636364,
+      "grad_norm": 2.719700303567712,
+      "learning_rate": 2.3081992462952215e-07,
+      "loss": 0.0125,
+      "step": 15780
+    },
+    {
+      "epoch": 3.586590909090909,
+      "grad_norm": 1.5912771067799352,
+      "learning_rate": 2.3075066756200573e-07,
+      "loss": 0.0052,
+      "step": 15781
+    },
+    {
+      "epoch": 3.586818181818182,
+      "grad_norm": 2.9789163364164875,
+      "learning_rate": 2.3068141853392103e-07,
+      "loss": 0.0264,
+      "step": 15782
+    },
+    {
+      "epoch": 3.5870454545454544,
+      "grad_norm": 2.0430603264511387,
+      "learning_rate": 2.3061217754668026e-07,
+      "loss": 0.0265,
+      "step": 15783
+    },
+    {
+      "epoch": 3.5872727272727274,
+      "grad_norm": 0.7101082600032353,
+      "learning_rate": 2.305429446016951e-07,
+      "loss": 0.0038,
+      "step": 15784
+    },
+    {
+      "epoch": 3.5875,
+      "grad_norm": 2.4934976010730985,
+      "learning_rate": 2.3047371970037744e-07,
+      "loss": 0.0086,
+      "step": 15785
+    },
+    {
+      "epoch": 3.587727272727273,
+      "grad_norm": 0.7974522728268356,
+      "learning_rate": 2.3040450284413895e-07,
+      "loss": 0.0037,
+      "step": 15786
+    },
+    {
+      "epoch": 3.5879545454545454,
+      "grad_norm": 1.3093003531816265,
+      "learning_rate": 2.30335294034391e-07,
+      "loss": 0.0082,
+      "step": 15787
+    },
+    {
+      "epoch": 3.588181818181818,
+      "grad_norm": 0.6633263824164897,
+      "learning_rate": 2.302660932725452e-07,
+      "loss": 0.0024,
+      "step": 15788
+    },
+    {
+      "epoch": 3.588409090909091,
+      "grad_norm": 2.053195992263889,
+      "learning_rate": 2.3019690056001214e-07,
+      "loss": 0.0073,
+      "step": 15789
+    },
+    {
+      "epoch": 3.588636363636364,
+      "grad_norm": 0.8897589991754616,
+      "learning_rate": 2.3012771589820314e-07,
+      "loss": 0.0064,
+      "step": 15790
+    },
+    {
+      "epoch": 3.5888636363636364,
+      "grad_norm": 2.371726973128599,
+      "learning_rate": 2.3005853928852886e-07,
+      "loss": 0.0092,
+      "step": 15791
+    },
+    {
+      "epoch": 3.589090909090909,
+      "grad_norm": 0.7252473792363558,
+      "learning_rate": 2.2998937073240017e-07,
+      "loss": 0.0039,
+      "step": 15792
+    },
+    {
+      "epoch": 3.589318181818182,
+      "grad_norm": 0.9638366521763749,
+      "learning_rate": 2.2992021023122716e-07,
+      "loss": 0.0038,
+      "step": 15793
+    },
+    {
+      "epoch": 3.589545454545455,
+      "grad_norm": 0.5426316985621782,
+      "learning_rate": 2.2985105778642026e-07,
+      "loss": 0.0023,
+      "step": 15794
+    },
+    {
+      "epoch": 3.5897727272727273,
+      "grad_norm": 1.5885278688698738,
+      "learning_rate": 2.297819133993897e-07,
+      "loss": 0.0072,
+      "step": 15795
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 2.035080902140148,
+      "learning_rate": 2.2971277707154543e-07,
+      "loss": 0.0205,
+      "step": 15796
+    },
+    {
+      "epoch": 3.590227272727273,
+      "grad_norm": 1.9950389500789885,
+      "learning_rate": 2.2964364880429726e-07,
+      "loss": 0.0071,
+      "step": 15797
+    },
+    {
+      "epoch": 3.5904545454545453,
+      "grad_norm": 2.490294715360187,
+      "learning_rate": 2.2957452859905479e-07,
+      "loss": 0.0202,
+      "step": 15798
+    },
+    {
+      "epoch": 3.5906818181818183,
+      "grad_norm": 2.302945388132369,
+      "learning_rate": 2.295054164572276e-07,
+      "loss": 0.0293,
+      "step": 15799
+    },
+    {
+      "epoch": 3.590909090909091,
+      "grad_norm": 1.4284108853941404,
+      "learning_rate": 2.2943631238022492e-07,
+      "loss": 0.0061,
+      "step": 15800
+    },
+    {
+      "epoch": 3.591136363636364,
+      "grad_norm": 2.3555855402966506,
+      "learning_rate": 2.2936721636945609e-07,
+      "loss": 0.0142,
+      "step": 15801
+    },
+    {
+      "epoch": 3.5913636363636363,
+      "grad_norm": 2.679263101592855,
+      "learning_rate": 2.292981284263298e-07,
+      "loss": 0.0136,
+      "step": 15802
+    },
+    {
+      "epoch": 3.5915909090909093,
+      "grad_norm": 0.6692731135127629,
+      "learning_rate": 2.2922904855225498e-07,
+      "loss": 0.0025,
+      "step": 15803
+    },
+    {
+      "epoch": 3.591818181818182,
+      "grad_norm": 0.7085498096939964,
+      "learning_rate": 2.2915997674864034e-07,
+      "loss": 0.0033,
+      "step": 15804
+    },
+    {
+      "epoch": 3.5920454545454543,
+      "grad_norm": 1.470879182581817,
+      "learning_rate": 2.2909091301689452e-07,
+      "loss": 0.0062,
+      "step": 15805
+    },
+    {
+      "epoch": 3.5922727272727273,
+      "grad_norm": 1.002929964945297,
+      "learning_rate": 2.2902185735842555e-07,
+      "loss": 0.0047,
+      "step": 15806
+    },
+    {
+      "epoch": 3.5925000000000002,
+      "grad_norm": 3.6558298033005436,
+      "learning_rate": 2.2895280977464164e-07,
+      "loss": 0.0412,
+      "step": 15807
+    },
+    {
+      "epoch": 3.5927272727272728,
+      "grad_norm": 3.757381648603628,
+      "learning_rate": 2.2888377026695094e-07,
+      "loss": 0.0383,
+      "step": 15808
+    },
+    {
+      "epoch": 3.5929545454545453,
+      "grad_norm": 2.216462459311967,
+      "learning_rate": 2.288147388367612e-07,
+      "loss": 0.0093,
+      "step": 15809
+    },
+    {
+      "epoch": 3.5931818181818183,
+      "grad_norm": 1.9617513638464115,
+      "learning_rate": 2.287457154854801e-07,
+      "loss": 0.012,
+      "step": 15810
+    },
+    {
+      "epoch": 3.5934090909090908,
+      "grad_norm": 1.8188375827482195,
+      "learning_rate": 2.2867670021451522e-07,
+      "loss": 0.0187,
+      "step": 15811
+    },
+    {
+      "epoch": 3.5936363636363637,
+      "grad_norm": 1.5331815057712057,
+      "learning_rate": 2.2860769302527382e-07,
+      "loss": 0.0136,
+      "step": 15812
+    },
+    {
+      "epoch": 3.5938636363636363,
+      "grad_norm": 2.165508678522829,
+      "learning_rate": 2.2853869391916313e-07,
+      "loss": 0.0216,
+      "step": 15813
+    },
+    {
+      "epoch": 3.5940909090909092,
+      "grad_norm": 1.7642144334570136,
+      "learning_rate": 2.2846970289759032e-07,
+      "loss": 0.0163,
+      "step": 15814
+    },
+    {
+      "epoch": 3.5943181818181817,
+      "grad_norm": 0.6330196976770874,
+      "learning_rate": 2.2840071996196187e-07,
+      "loss": 0.0028,
+      "step": 15815
+    },
+    {
+      "epoch": 3.5945454545454547,
+      "grad_norm": 1.292544439247858,
+      "learning_rate": 2.2833174511368462e-07,
+      "loss": 0.0096,
+      "step": 15816
+    },
+    {
+      "epoch": 3.5947727272727272,
+      "grad_norm": 1.184055156847822,
+      "learning_rate": 2.2826277835416507e-07,
+      "loss": 0.0064,
+      "step": 15817
+    },
+    {
+      "epoch": 3.5949999999999998,
+      "grad_norm": 0.7091924819612236,
+      "learning_rate": 2.2819381968480985e-07,
+      "loss": 0.0038,
+      "step": 15818
+    },
+    {
+      "epoch": 3.5952272727272727,
+      "grad_norm": 1.0316076728466133,
+      "learning_rate": 2.2812486910702474e-07,
+      "loss": 0.0044,
+      "step": 15819
+    },
+    {
+      "epoch": 3.5954545454545457,
+      "grad_norm": 0.8476143162697091,
+      "learning_rate": 2.280559266222159e-07,
+      "loss": 0.0053,
+      "step": 15820
+    },
+    {
+      "epoch": 3.595681818181818,
+      "grad_norm": 2.3377249713033086,
+      "learning_rate": 2.2798699223178928e-07,
+      "loss": 0.0143,
+      "step": 15821
+    },
+    {
+      "epoch": 3.5959090909090907,
+      "grad_norm": 1.2832909293190466,
+      "learning_rate": 2.279180659371504e-07,
+      "loss": 0.005,
+      "step": 15822
+    },
+    {
+      "epoch": 3.5961363636363637,
+      "grad_norm": 1.4493708500544207,
+      "learning_rate": 2.2784914773970524e-07,
+      "loss": 0.0051,
+      "step": 15823
+    },
+    {
+      "epoch": 3.596363636363636,
+      "grad_norm": 2.1293011378500992,
+      "learning_rate": 2.2778023764085854e-07,
+      "loss": 0.0119,
+      "step": 15824
+    },
+    {
+      "epoch": 3.596590909090909,
+      "grad_norm": 0.9527289922399167,
+      "learning_rate": 2.277113356420158e-07,
+      "loss": 0.0047,
+      "step": 15825
+    },
+    {
+      "epoch": 3.5968181818181817,
+      "grad_norm": 3.1235411879233776,
+      "learning_rate": 2.2764244174458202e-07,
+      "loss": 0.0384,
+      "step": 15826
+    },
+    {
+      "epoch": 3.5970454545454547,
+      "grad_norm": 1.6980580398508027,
+      "learning_rate": 2.2757355594996214e-07,
+      "loss": 0.0153,
+      "step": 15827
+    },
+    {
+      "epoch": 3.597272727272727,
+      "grad_norm": 1.021852945020311,
+      "learning_rate": 2.2750467825956072e-07,
+      "loss": 0.0071,
+      "step": 15828
+    },
+    {
+      "epoch": 3.5975,
+      "grad_norm": 1.2384996116579179,
+      "learning_rate": 2.2743580867478244e-07,
+      "loss": 0.0055,
+      "step": 15829
+    },
+    {
+      "epoch": 3.5977272727272727,
+      "grad_norm": 1.5122831959891976,
+      "learning_rate": 2.273669471970316e-07,
+      "loss": 0.0069,
+      "step": 15830
+    },
+    {
+      "epoch": 3.597954545454545,
+      "grad_norm": 0.8359024315107285,
+      "learning_rate": 2.2729809382771244e-07,
+      "loss": 0.0041,
+      "step": 15831
+    },
+    {
+      "epoch": 3.598181818181818,
+      "grad_norm": 0.9392203514809639,
+      "learning_rate": 2.2722924856822912e-07,
+      "loss": 0.0026,
+      "step": 15832
+    },
+    {
+      "epoch": 3.598409090909091,
+      "grad_norm": 3.478899184929875,
+      "learning_rate": 2.2716041141998518e-07,
+      "loss": 0.0232,
+      "step": 15833
+    },
+    {
+      "epoch": 3.5986363636363636,
+      "grad_norm": 2.1448109908554014,
+      "learning_rate": 2.2709158238438458e-07,
+      "loss": 0.0173,
+      "step": 15834
+    },
+    {
+      "epoch": 3.598863636363636,
+      "grad_norm": 1.303274764724073,
+      "learning_rate": 2.270227614628307e-07,
+      "loss": 0.007,
+      "step": 15835
+    },
+    {
+      "epoch": 3.599090909090909,
+      "grad_norm": 1.2334606549571463,
+      "learning_rate": 2.269539486567273e-07,
+      "loss": 0.0064,
+      "step": 15836
+    },
+    {
+      "epoch": 3.599318181818182,
+      "grad_norm": 1.9886115243637346,
+      "learning_rate": 2.2688514396747714e-07,
+      "loss": 0.0092,
+      "step": 15837
+    },
+    {
+      "epoch": 3.5995454545454546,
+      "grad_norm": 1.9372480636569132,
+      "learning_rate": 2.2681634739648336e-07,
+      "loss": 0.0103,
+      "step": 15838
+    },
+    {
+      "epoch": 3.599772727272727,
+      "grad_norm": 0.466746534871408,
+      "learning_rate": 2.2674755894514898e-07,
+      "loss": 0.0033,
+      "step": 15839
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 2.4417462345166583,
+      "learning_rate": 2.2667877861487662e-07,
+      "loss": 0.0155,
+      "step": 15840
+    },
+    {
+      "epoch": 3.6002272727272726,
+      "grad_norm": 0.9681045316199849,
+      "learning_rate": 2.2661000640706893e-07,
+      "loss": 0.0018,
+      "step": 15841
+    },
+    {
+      "epoch": 3.6004545454545456,
+      "grad_norm": 0.6766150096784805,
+      "learning_rate": 2.2654124232312821e-07,
+      "loss": 0.0025,
+      "step": 15842
+    },
+    {
+      "epoch": 3.600681818181818,
+      "grad_norm": 2.923177034097511,
+      "learning_rate": 2.264724863644567e-07,
+      "loss": 0.0139,
+      "step": 15843
+    },
+    {
+      "epoch": 3.600909090909091,
+      "grad_norm": 2.491924011191986,
+      "learning_rate": 2.2640373853245647e-07,
+      "loss": 0.0113,
+      "step": 15844
+    },
+    {
+      "epoch": 3.6011363636363636,
+      "grad_norm": 1.0784497892263634,
+      "learning_rate": 2.2633499882852969e-07,
+      "loss": 0.0104,
+      "step": 15845
+    },
+    {
+      "epoch": 3.6013636363636365,
+      "grad_norm": 1.6706679406561833,
+      "learning_rate": 2.2626626725407747e-07,
+      "loss": 0.0141,
+      "step": 15846
+    },
+    {
+      "epoch": 3.601590909090909,
+      "grad_norm": 1.6976148057297866,
+      "learning_rate": 2.2619754381050172e-07,
+      "loss": 0.0109,
+      "step": 15847
+    },
+    {
+      "epoch": 3.6018181818181816,
+      "grad_norm": 1.4427546180102766,
+      "learning_rate": 2.261288284992038e-07,
+      "loss": 0.0122,
+      "step": 15848
+    },
+    {
+      "epoch": 3.6020454545454546,
+      "grad_norm": 1.7678207905762509,
+      "learning_rate": 2.2606012132158518e-07,
+      "loss": 0.0114,
+      "step": 15849
+    },
+    {
+      "epoch": 3.6022727272727275,
+      "grad_norm": 1.7684881575251352,
+      "learning_rate": 2.2599142227904645e-07,
+      "loss": 0.0038,
+      "step": 15850
+    },
+    {
+      "epoch": 3.6025,
+      "grad_norm": 2.9782487503368946,
+      "learning_rate": 2.2592273137298872e-07,
+      "loss": 0.0222,
+      "step": 15851
+    },
+    {
+      "epoch": 3.6027272727272726,
+      "grad_norm": 1.1485390197291723,
+      "learning_rate": 2.2585404860481267e-07,
+      "loss": 0.0069,
+      "step": 15852
+    },
+    {
+      "epoch": 3.6029545454545455,
+      "grad_norm": 1.5115836309174326,
+      "learning_rate": 2.2578537397591903e-07,
+      "loss": 0.018,
+      "step": 15853
+    },
+    {
+      "epoch": 3.603181818181818,
+      "grad_norm": 1.974526247656551,
+      "learning_rate": 2.257167074877082e-07,
+      "loss": 0.0135,
+      "step": 15854
+    },
+    {
+      "epoch": 3.603409090909091,
+      "grad_norm": 0.9667395487095175,
+      "learning_rate": 2.2564804914158008e-07,
+      "loss": 0.0055,
+      "step": 15855
+    },
+    {
+      "epoch": 3.6036363636363635,
+      "grad_norm": 2.4022071015518525,
+      "learning_rate": 2.255793989389348e-07,
+      "loss": 0.0135,
+      "step": 15856
+    },
+    {
+      "epoch": 3.6038636363636365,
+      "grad_norm": 0.7547124216581474,
+      "learning_rate": 2.2551075688117256e-07,
+      "loss": 0.0029,
+      "step": 15857
+    },
+    {
+      "epoch": 3.604090909090909,
+      "grad_norm": 2.2721165939022345,
+      "learning_rate": 2.254421229696932e-07,
+      "loss": 0.0247,
+      "step": 15858
+    },
+    {
+      "epoch": 3.604318181818182,
+      "grad_norm": 0.32566047736769693,
+      "learning_rate": 2.253734972058958e-07,
+      "loss": 0.0014,
+      "step": 15859
+    },
+    {
+      "epoch": 3.6045454545454545,
+      "grad_norm": 1.0729275060941454,
+      "learning_rate": 2.2530487959117998e-07,
+      "loss": 0.0046,
+      "step": 15860
+    },
+    {
+      "epoch": 3.604772727272727,
+      "grad_norm": 0.4813842010646856,
+      "learning_rate": 2.252362701269449e-07,
+      "loss": 0.0023,
+      "step": 15861
+    },
+    {
+      "epoch": 3.605,
+      "grad_norm": 5.459386092114649,
+      "learning_rate": 2.2516766881458998e-07,
+      "loss": 0.0195,
+      "step": 15862
+    },
+    {
+      "epoch": 3.605227272727273,
+      "grad_norm": 2.3574840779103767,
+      "learning_rate": 2.2509907565551361e-07,
+      "loss": 0.0254,
+      "step": 15863
+    },
+    {
+      "epoch": 3.6054545454545455,
+      "grad_norm": 2.0204272900831795,
+      "learning_rate": 2.250304906511147e-07,
+      "loss": 0.0132,
+      "step": 15864
+    },
+    {
+      "epoch": 3.605681818181818,
+      "grad_norm": 2.254894351066731,
+      "learning_rate": 2.2496191380279197e-07,
+      "loss": 0.0158,
+      "step": 15865
+    },
+    {
+      "epoch": 3.605909090909091,
+      "grad_norm": 1.5781098819849328,
+      "learning_rate": 2.2489334511194367e-07,
+      "loss": 0.0078,
+      "step": 15866
+    },
+    {
+      "epoch": 3.6061363636363635,
+      "grad_norm": 1.212270370285254,
+      "learning_rate": 2.248247845799683e-07,
+      "loss": 0.0066,
+      "step": 15867
+    },
+    {
+      "epoch": 3.6063636363636364,
+      "grad_norm": 2.9518528988326715,
+      "learning_rate": 2.2475623220826346e-07,
+      "loss": 0.0099,
+      "step": 15868
+    },
+    {
+      "epoch": 3.606590909090909,
+      "grad_norm": 1.2714647295169168,
+      "learning_rate": 2.2468768799822738e-07,
+      "loss": 0.0077,
+      "step": 15869
+    },
+    {
+      "epoch": 3.606818181818182,
+      "grad_norm": 2.7243046863002505,
+      "learning_rate": 2.246191519512577e-07,
+      "loss": 0.0253,
+      "step": 15870
+    },
+    {
+      "epoch": 3.6070454545454544,
+      "grad_norm": 1.4257586123783865,
+      "learning_rate": 2.24550624068752e-07,
+      "loss": 0.0151,
+      "step": 15871
+    },
+    {
+      "epoch": 3.6072727272727274,
+      "grad_norm": 1.157588689157874,
+      "learning_rate": 2.2448210435210767e-07,
+      "loss": 0.0069,
+      "step": 15872
+    },
+    {
+      "epoch": 3.6075,
+      "grad_norm": 1.6697918062075972,
+      "learning_rate": 2.2441359280272205e-07,
+      "loss": 0.0086,
+      "step": 15873
+    },
+    {
+      "epoch": 3.607727272727273,
+      "grad_norm": 3.066530662087564,
+      "learning_rate": 2.2434508942199207e-07,
+      "loss": 0.018,
+      "step": 15874
+    },
+    {
+      "epoch": 3.6079545454545454,
+      "grad_norm": 2.5196703644548375,
+      "learning_rate": 2.2427659421131474e-07,
+      "loss": 0.0427,
+      "step": 15875
+    },
+    {
+      "epoch": 3.6081818181818184,
+      "grad_norm": 1.0571337291991776,
+      "learning_rate": 2.2420810717208696e-07,
+      "loss": 0.0059,
+      "step": 15876
+    },
+    {
+      "epoch": 3.608409090909091,
+      "grad_norm": 0.36111715663461114,
+      "learning_rate": 2.241396283057049e-07,
+      "loss": 0.0016,
+      "step": 15877
+    },
+    {
+      "epoch": 3.6086363636363634,
+      "grad_norm": 1.0820877942827576,
+      "learning_rate": 2.2407115761356513e-07,
+      "loss": 0.0048,
+      "step": 15878
+    },
+    {
+      "epoch": 3.6088636363636364,
+      "grad_norm": 1.7908394311082858,
+      "learning_rate": 2.2400269509706392e-07,
+      "loss": 0.0136,
+      "step": 15879
+    },
+    {
+      "epoch": 3.6090909090909093,
+      "grad_norm": 0.8824019872649201,
+      "learning_rate": 2.2393424075759757e-07,
+      "loss": 0.0075,
+      "step": 15880
+    },
+    {
+      "epoch": 3.609318181818182,
+      "grad_norm": 1.6137389539230242,
+      "learning_rate": 2.2386579459656146e-07,
+      "loss": 0.0101,
+      "step": 15881
+    },
+    {
+      "epoch": 3.6095454545454544,
+      "grad_norm": 1.4429544216694103,
+      "learning_rate": 2.2379735661535167e-07,
+      "loss": 0.0086,
+      "step": 15882
+    },
+    {
+      "epoch": 3.6097727272727274,
+      "grad_norm": 1.7870015886956643,
+      "learning_rate": 2.2372892681536368e-07,
+      "loss": 0.0101,
+      "step": 15883
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 3.7368705654664507,
+      "learning_rate": 2.2366050519799322e-07,
+      "loss": 0.0157,
+      "step": 15884
+    },
+    {
+      "epoch": 3.610227272727273,
+      "grad_norm": 1.3343618008971623,
+      "learning_rate": 2.2359209176463477e-07,
+      "loss": 0.0046,
+      "step": 15885
+    },
+    {
+      "epoch": 3.6104545454545454,
+      "grad_norm": 1.5757419827105066,
+      "learning_rate": 2.2352368651668407e-07,
+      "loss": 0.0157,
+      "step": 15886
+    },
+    {
+      "epoch": 3.6106818181818183,
+      "grad_norm": 2.423003390078847,
+      "learning_rate": 2.2345528945553578e-07,
+      "loss": 0.0351,
+      "step": 15887
+    },
+    {
+      "epoch": 3.610909090909091,
+      "grad_norm": 2.026336325607938,
+      "learning_rate": 2.233869005825847e-07,
+      "loss": 0.0179,
+      "step": 15888
+    },
+    {
+      "epoch": 3.611136363636364,
+      "grad_norm": 2.2640300411291796,
+      "learning_rate": 2.2331851989922563e-07,
+      "loss": 0.0075,
+      "step": 15889
+    },
+    {
+      "epoch": 3.6113636363636363,
+      "grad_norm": 3.46841529781796,
+      "learning_rate": 2.2325014740685242e-07,
+      "loss": 0.025,
+      "step": 15890
+    },
+    {
+      "epoch": 3.611590909090909,
+      "grad_norm": 1.0934847871333149,
+      "learning_rate": 2.2318178310685966e-07,
+      "loss": 0.0051,
+      "step": 15891
+    },
+    {
+      "epoch": 3.611818181818182,
+      "grad_norm": 2.209269818699116,
+      "learning_rate": 2.2311342700064132e-07,
+      "loss": 0.0111,
+      "step": 15892
+    },
+    {
+      "epoch": 3.612045454545455,
+      "grad_norm": 1.6673953414002232,
+      "learning_rate": 2.2304507908959154e-07,
+      "loss": 0.0087,
+      "step": 15893
+    },
+    {
+      "epoch": 3.6122727272727273,
+      "grad_norm": 1.7759207659004372,
+      "learning_rate": 2.2297673937510368e-07,
+      "loss": 0.0142,
+      "step": 15894
+    },
+    {
+      "epoch": 3.6125,
+      "grad_norm": 1.3021064881567976,
+      "learning_rate": 2.2290840785857138e-07,
+      "loss": 0.0072,
+      "step": 15895
+    },
+    {
+      "epoch": 3.612727272727273,
+      "grad_norm": 1.3608971127119258,
+      "learning_rate": 2.228400845413882e-07,
+      "loss": 0.0038,
+      "step": 15896
+    },
+    {
+      "epoch": 3.6129545454545453,
+      "grad_norm": 0.8056869270623577,
+      "learning_rate": 2.2277176942494747e-07,
+      "loss": 0.0029,
+      "step": 15897
+    },
+    {
+      "epoch": 3.6131818181818183,
+      "grad_norm": 1.5121280275632834,
+      "learning_rate": 2.2270346251064195e-07,
+      "loss": 0.0038,
+      "step": 15898
+    },
+    {
+      "epoch": 3.613409090909091,
+      "grad_norm": 1.181165960944288,
+      "learning_rate": 2.2263516379986464e-07,
+      "loss": 0.0067,
+      "step": 15899
+    },
+    {
+      "epoch": 3.6136363636363638,
+      "grad_norm": 1.1187601521144537,
+      "learning_rate": 2.2256687329400826e-07,
+      "loss": 0.0052,
+      "step": 15900
+    },
+    {
+      "epoch": 3.6138636363636363,
+      "grad_norm": 1.9281851107642185,
+      "learning_rate": 2.2249859099446547e-07,
+      "loss": 0.0063,
+      "step": 15901
+    },
+    {
+      "epoch": 3.6140909090909092,
+      "grad_norm": 1.891626880757846,
+      "learning_rate": 2.2243031690262859e-07,
+      "loss": 0.007,
+      "step": 15902
+    },
+    {
+      "epoch": 3.6143181818181818,
+      "grad_norm": 0.8383577528675731,
+      "learning_rate": 2.2236205101988983e-07,
+      "loss": 0.0075,
+      "step": 15903
+    },
+    {
+      "epoch": 3.6145454545454543,
+      "grad_norm": 1.176945520238281,
+      "learning_rate": 2.2229379334764136e-07,
+      "loss": 0.0113,
+      "step": 15904
+    },
+    {
+      "epoch": 3.6147727272727272,
+      "grad_norm": 4.454684119842563,
+      "learning_rate": 2.2222554388727496e-07,
+      "loss": 0.0209,
+      "step": 15905
+    },
+    {
+      "epoch": 3.615,
+      "grad_norm": 1.3713538321397516,
+      "learning_rate": 2.2215730264018264e-07,
+      "loss": 0.009,
+      "step": 15906
+    },
+    {
+      "epoch": 3.6152272727272727,
+      "grad_norm": 1.564106109826806,
+      "learning_rate": 2.2208906960775552e-07,
+      "loss": 0.0054,
+      "step": 15907
+    },
+    {
+      "epoch": 3.6154545454545453,
+      "grad_norm": 0.6319868255564415,
+      "learning_rate": 2.2202084479138512e-07,
+      "loss": 0.0042,
+      "step": 15908
+    },
+    {
+      "epoch": 3.615681818181818,
+      "grad_norm": 2.8197633971787277,
+      "learning_rate": 2.2195262819246278e-07,
+      "loss": 0.0202,
+      "step": 15909
+    },
+    {
+      "epoch": 3.615909090909091,
+      "grad_norm": 0.7925688672260471,
+      "learning_rate": 2.218844198123795e-07,
+      "loss": 0.0032,
+      "step": 15910
+    },
+    {
+      "epoch": 3.6161363636363637,
+      "grad_norm": 1.190638681751303,
+      "learning_rate": 2.2181621965252634e-07,
+      "loss": 0.0181,
+      "step": 15911
+    },
+    {
+      "epoch": 3.6163636363636362,
+      "grad_norm": 2.413315930121464,
+      "learning_rate": 2.2174802771429375e-07,
+      "loss": 0.0088,
+      "step": 15912
+    },
+    {
+      "epoch": 3.616590909090909,
+      "grad_norm": 3.8819175959317276,
+      "learning_rate": 2.216798439990723e-07,
+      "loss": 0.0352,
+      "step": 15913
+    },
+    {
+      "epoch": 3.6168181818181817,
+      "grad_norm": 1.9829038861669062,
+      "learning_rate": 2.2161166850825248e-07,
+      "loss": 0.0085,
+      "step": 15914
+    },
+    {
+      "epoch": 3.6170454545454547,
+      "grad_norm": 2.52132156052364,
+      "learning_rate": 2.2154350124322446e-07,
+      "loss": 0.0146,
+      "step": 15915
+    },
+    {
+      "epoch": 3.617272727272727,
+      "grad_norm": 2.275303042499135,
+      "learning_rate": 2.2147534220537837e-07,
+      "loss": 0.0159,
+      "step": 15916
+    },
+    {
+      "epoch": 3.6175,
+      "grad_norm": 1.2757064955493411,
+      "learning_rate": 2.2140719139610397e-07,
+      "loss": 0.0089,
+      "step": 15917
+    },
+    {
+      "epoch": 3.6177272727272727,
+      "grad_norm": 1.223576330675856,
+      "learning_rate": 2.2133904881679115e-07,
+      "loss": 0.0088,
+      "step": 15918
+    },
+    {
+      "epoch": 3.6179545454545456,
+      "grad_norm": 1.2531586348623738,
+      "learning_rate": 2.212709144688295e-07,
+      "loss": 0.0083,
+      "step": 15919
+    },
+    {
+      "epoch": 3.618181818181818,
+      "grad_norm": 0.8107504420982674,
+      "learning_rate": 2.2120278835360804e-07,
+      "loss": 0.0042,
+      "step": 15920
+    },
+    {
+      "epoch": 3.6184090909090907,
+      "grad_norm": 1.110461235489247,
+      "learning_rate": 2.2113467047251619e-07,
+      "loss": 0.0073,
+      "step": 15921
+    },
+    {
+      "epoch": 3.6186363636363637,
+      "grad_norm": 0.7439445898785875,
+      "learning_rate": 2.2106656082694294e-07,
+      "loss": 0.0145,
+      "step": 15922
+    },
+    {
+      "epoch": 3.6188636363636366,
+      "grad_norm": 1.3322329867635856,
+      "learning_rate": 2.2099845941827729e-07,
+      "loss": 0.008,
+      "step": 15923
+    },
+    {
+      "epoch": 3.619090909090909,
+      "grad_norm": 1.1361962339915506,
+      "learning_rate": 2.2093036624790808e-07,
+      "loss": 0.0069,
+      "step": 15924
+    },
+    {
+      "epoch": 3.6193181818181817,
+      "grad_norm": 2.4238821736025744,
+      "learning_rate": 2.2086228131722336e-07,
+      "loss": 0.0154,
+      "step": 15925
+    },
+    {
+      "epoch": 3.6195454545454546,
+      "grad_norm": 1.4632528359070862,
+      "learning_rate": 2.2079420462761188e-07,
+      "loss": 0.0073,
+      "step": 15926
+    },
+    {
+      "epoch": 3.619772727272727,
+      "grad_norm": 1.2724033017722325,
+      "learning_rate": 2.2072613618046162e-07,
+      "loss": 0.0094,
+      "step": 15927
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 1.7410021760167025,
+      "learning_rate": 2.2065807597716108e-07,
+      "loss": 0.0058,
+      "step": 15928
+    },
+    {
+      "epoch": 3.6202272727272726,
+      "grad_norm": 1.7773161016981365,
+      "learning_rate": 2.2059002401909752e-07,
+      "loss": 0.0159,
+      "step": 15929
+    },
+    {
+      "epoch": 3.6204545454545456,
+      "grad_norm": 1.9046090375786269,
+      "learning_rate": 2.2052198030765886e-07,
+      "loss": 0.0193,
+      "step": 15930
+    },
+    {
+      "epoch": 3.620681818181818,
+      "grad_norm": 1.4316615368928134,
+      "learning_rate": 2.204539448442327e-07,
+      "loss": 0.0057,
+      "step": 15931
+    },
+    {
+      "epoch": 3.620909090909091,
+      "grad_norm": 1.1257637419957176,
+      "learning_rate": 2.2038591763020637e-07,
+      "loss": 0.0064,
+      "step": 15932
+    },
+    {
+      "epoch": 3.6211363636363636,
+      "grad_norm": 2.0052742234637884,
+      "learning_rate": 2.2031789866696706e-07,
+      "loss": 0.014,
+      "step": 15933
+    },
+    {
+      "epoch": 3.621363636363636,
+      "grad_norm": 2.485192918756668,
+      "learning_rate": 2.2024988795590182e-07,
+      "loss": 0.0169,
+      "step": 15934
+    },
+    {
+      "epoch": 3.621590909090909,
+      "grad_norm": 1.0473328367264265,
+      "learning_rate": 2.2018188549839753e-07,
+      "loss": 0.0027,
+      "step": 15935
+    },
+    {
+      "epoch": 3.621818181818182,
+      "grad_norm": 2.0983362717086806,
+      "learning_rate": 2.201138912958408e-07,
+      "loss": 0.0123,
+      "step": 15936
+    },
+    {
+      "epoch": 3.6220454545454546,
+      "grad_norm": 1.1524757507689674,
+      "learning_rate": 2.200459053496184e-07,
+      "loss": 0.0174,
+      "step": 15937
+    },
+    {
+      "epoch": 3.622272727272727,
+      "grad_norm": 1.2878811808592494,
+      "learning_rate": 2.1997792766111622e-07,
+      "loss": 0.0099,
+      "step": 15938
+    },
+    {
+      "epoch": 3.6225,
+      "grad_norm": 1.1580397604783772,
+      "learning_rate": 2.1990995823172073e-07,
+      "loss": 0.0033,
+      "step": 15939
+    },
+    {
+      "epoch": 3.6227272727272726,
+      "grad_norm": 1.4884085053068965,
+      "learning_rate": 2.1984199706281796e-07,
+      "loss": 0.0157,
+      "step": 15940
+    },
+    {
+      "epoch": 3.6229545454545455,
+      "grad_norm": 0.9124078866560227,
+      "learning_rate": 2.1977404415579384e-07,
+      "loss": 0.0051,
+      "step": 15941
+    },
+    {
+      "epoch": 3.623181818181818,
+      "grad_norm": 2.125614695150418,
+      "learning_rate": 2.1970609951203376e-07,
+      "loss": 0.0141,
+      "step": 15942
+    },
+    {
+      "epoch": 3.623409090909091,
+      "grad_norm": 3.0019335501726827,
+      "learning_rate": 2.1963816313292335e-07,
+      "loss": 0.0506,
+      "step": 15943
+    },
+    {
+      "epoch": 3.6236363636363635,
+      "grad_norm": 1.7783848026040485,
+      "learning_rate": 2.1957023501984803e-07,
+      "loss": 0.0125,
+      "step": 15944
+    },
+    {
+      "epoch": 3.6238636363636365,
+      "grad_norm": 1.2397802337740074,
+      "learning_rate": 2.1950231517419284e-07,
+      "loss": 0.0065,
+      "step": 15945
+    },
+    {
+      "epoch": 3.624090909090909,
+      "grad_norm": 1.9596160929308115,
+      "learning_rate": 2.1943440359734294e-07,
+      "loss": 0.0131,
+      "step": 15946
+    },
+    {
+      "epoch": 3.6243181818181816,
+      "grad_norm": 1.0561374390334233,
+      "learning_rate": 2.1936650029068307e-07,
+      "loss": 0.0114,
+      "step": 15947
+    },
+    {
+      "epoch": 3.6245454545454545,
+      "grad_norm": 1.5290116663593507,
+      "learning_rate": 2.1929860525559794e-07,
+      "loss": 0.0173,
+      "step": 15948
+    },
+    {
+      "epoch": 3.6247727272727275,
+      "grad_norm": 1.7955975969020506,
+      "learning_rate": 2.1923071849347207e-07,
+      "loss": 0.0132,
+      "step": 15949
+    },
+    {
+      "epoch": 3.625,
+      "grad_norm": 1.2983175859100151,
+      "learning_rate": 2.1916284000568993e-07,
+      "loss": 0.0094,
+      "step": 15950
+    },
+    {
+      "epoch": 3.6252272727272725,
+      "grad_norm": 1.0904373289784741,
+      "learning_rate": 2.190949697936353e-07,
+      "loss": 0.0057,
+      "step": 15951
+    },
+    {
+      "epoch": 3.6254545454545455,
+      "grad_norm": 1.313995395830452,
+      "learning_rate": 2.1902710785869231e-07,
+      "loss": 0.0089,
+      "step": 15952
+    },
+    {
+      "epoch": 3.6256818181818184,
+      "grad_norm": 1.2824161149609266,
+      "learning_rate": 2.1895925420224494e-07,
+      "loss": 0.0135,
+      "step": 15953
+    },
+    {
+      "epoch": 3.625909090909091,
+      "grad_norm": 1.5746512340695367,
+      "learning_rate": 2.1889140882567675e-07,
+      "loss": 0.0133,
+      "step": 15954
+    },
+    {
+      "epoch": 3.6261363636363635,
+      "grad_norm": 2.4151966959932647,
+      "learning_rate": 2.188235717303714e-07,
+      "loss": 0.0029,
+      "step": 15955
+    },
+    {
+      "epoch": 3.6263636363636365,
+      "grad_norm": 0.9481572344837067,
+      "learning_rate": 2.187557429177118e-07,
+      "loss": 0.0058,
+      "step": 15956
+    },
+    {
+      "epoch": 3.626590909090909,
+      "grad_norm": 0.9999917542663722,
+      "learning_rate": 2.1868792238908142e-07,
+      "loss": 0.0041,
+      "step": 15957
+    },
+    {
+      "epoch": 3.626818181818182,
+      "grad_norm": 0.6156717476073235,
+      "learning_rate": 2.1862011014586308e-07,
+      "loss": 0.0038,
+      "step": 15958
+    },
+    {
+      "epoch": 3.6270454545454545,
+      "grad_norm": 4.026479668412748,
+      "learning_rate": 2.185523061894399e-07,
+      "loss": 0.0099,
+      "step": 15959
+    },
+    {
+      "epoch": 3.6272727272727274,
+      "grad_norm": 0.7544385387000735,
+      "learning_rate": 2.1848451052119405e-07,
+      "loss": 0.0075,
+      "step": 15960
+    },
+    {
+      "epoch": 3.6275,
+      "grad_norm": 1.904482073312342,
+      "learning_rate": 2.1841672314250828e-07,
+      "loss": 0.0233,
+      "step": 15961
+    },
+    {
+      "epoch": 3.627727272727273,
+      "grad_norm": 2.5032185380400427,
+      "learning_rate": 2.183489440547648e-07,
+      "loss": 0.0208,
+      "step": 15962
+    },
+    {
+      "epoch": 3.6279545454545454,
+      "grad_norm": 5.494285374162116,
+      "learning_rate": 2.182811732593458e-07,
+      "loss": 0.0208,
+      "step": 15963
+    },
+    {
+      "epoch": 3.628181818181818,
+      "grad_norm": 3.1717043595876393,
+      "learning_rate": 2.1821341075763328e-07,
+      "loss": 0.0272,
+      "step": 15964
+    },
+    {
+      "epoch": 3.628409090909091,
+      "grad_norm": 0.7831867786831634,
+      "learning_rate": 2.18145656551009e-07,
+      "loss": 0.0037,
+      "step": 15965
+    },
+    {
+      "epoch": 3.628636363636364,
+      "grad_norm": 1.582183848591283,
+      "learning_rate": 2.1807791064085453e-07,
+      "loss": 0.007,
+      "step": 15966
+    },
+    {
+      "epoch": 3.6288636363636364,
+      "grad_norm": 1.6244649756550646,
+      "learning_rate": 2.1801017302855139e-07,
+      "loss": 0.0235,
+      "step": 15967
+    },
+    {
+      "epoch": 3.629090909090909,
+      "grad_norm": 2.8784546070365016,
+      "learning_rate": 2.1794244371548103e-07,
+      "loss": 0.0262,
+      "step": 15968
+    },
+    {
+      "epoch": 3.629318181818182,
+      "grad_norm": 1.19045109646666,
+      "learning_rate": 2.1787472270302423e-07,
+      "loss": 0.0053,
+      "step": 15969
+    },
+    {
+      "epoch": 3.6295454545454544,
+      "grad_norm": 1.3671029697742774,
+      "learning_rate": 2.1780700999256215e-07,
+      "loss": 0.0174,
+      "step": 15970
+    },
+    {
+      "epoch": 3.6297727272727274,
+      "grad_norm": 0.8018282568268885,
+      "learning_rate": 2.1773930558547548e-07,
+      "loss": 0.0043,
+      "step": 15971
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 1.4234465338573234,
+      "learning_rate": 2.1767160948314505e-07,
+      "loss": 0.0084,
+      "step": 15972
+    },
+    {
+      "epoch": 3.630227272727273,
+      "grad_norm": 2.2338119973032113,
+      "learning_rate": 2.1760392168695092e-07,
+      "loss": 0.0218,
+      "step": 15973
+    },
+    {
+      "epoch": 3.6304545454545454,
+      "grad_norm": 0.4995376185926977,
+      "learning_rate": 2.1753624219827358e-07,
+      "loss": 0.0021,
+      "step": 15974
+    },
+    {
+      "epoch": 3.6306818181818183,
+      "grad_norm": 1.4150298608109722,
+      "learning_rate": 2.1746857101849317e-07,
+      "loss": 0.0113,
+      "step": 15975
+    },
+    {
+      "epoch": 3.630909090909091,
+      "grad_norm": 1.5131144468421918,
+      "learning_rate": 2.1740090814898953e-07,
+      "loss": 0.0162,
+      "step": 15976
+    },
+    {
+      "epoch": 3.6311363636363634,
+      "grad_norm": 0.985683547333267,
+      "learning_rate": 2.1733325359114244e-07,
+      "loss": 0.0107,
+      "step": 15977
+    },
+    {
+      "epoch": 3.6313636363636363,
+      "grad_norm": 2.0622015283930732,
+      "learning_rate": 2.1726560734633153e-07,
+      "loss": 0.006,
+      "step": 15978
+    },
+    {
+      "epoch": 3.6315909090909093,
+      "grad_norm": 0.9889963549689883,
+      "learning_rate": 2.1719796941593625e-07,
+      "loss": 0.0034,
+      "step": 15979
+    },
+    {
+      "epoch": 3.631818181818182,
+      "grad_norm": 0.7449848359143835,
+      "learning_rate": 2.1713033980133586e-07,
+      "loss": 0.0041,
+      "step": 15980
+    },
+    {
+      "epoch": 3.6320454545454544,
+      "grad_norm": 5.743819629011239,
+      "learning_rate": 2.1706271850390953e-07,
+      "loss": 0.0326,
+      "step": 15981
+    },
+    {
+      "epoch": 3.6322727272727273,
+      "grad_norm": 3.4564181147949378,
+      "learning_rate": 2.1699510552503592e-07,
+      "loss": 0.0224,
+      "step": 15982
+    },
+    {
+      "epoch": 3.6325,
+      "grad_norm": 1.7026860837330995,
+      "learning_rate": 2.1692750086609384e-07,
+      "loss": 0.0113,
+      "step": 15983
+    },
+    {
+      "epoch": 3.632727272727273,
+      "grad_norm": 0.8620333821358508,
+      "learning_rate": 2.1685990452846202e-07,
+      "loss": 0.0038,
+      "step": 15984
+    },
+    {
+      "epoch": 3.6329545454545453,
+      "grad_norm": 2.67037791407825,
+      "learning_rate": 2.1679231651351892e-07,
+      "loss": 0.0302,
+      "step": 15985
+    },
+    {
+      "epoch": 3.6331818181818183,
+      "grad_norm": 5.602806355416981,
+      "learning_rate": 2.1672473682264253e-07,
+      "loss": 0.0259,
+      "step": 15986
+    },
+    {
+      "epoch": 3.633409090909091,
+      "grad_norm": 0.9375596680444922,
+      "learning_rate": 2.1665716545721096e-07,
+      "loss": 0.0041,
+      "step": 15987
+    },
+    {
+      "epoch": 3.6336363636363638,
+      "grad_norm": 2.94373011845969,
+      "learning_rate": 2.165896024186022e-07,
+      "loss": 0.0164,
+      "step": 15988
+    },
+    {
+      "epoch": 3.6338636363636363,
+      "grad_norm": 2.1839808720435716,
+      "learning_rate": 2.165220477081939e-07,
+      "loss": 0.0136,
+      "step": 15989
+    },
+    {
+      "epoch": 3.634090909090909,
+      "grad_norm": 1.63167392147309,
+      "learning_rate": 2.1645450132736397e-07,
+      "loss": 0.0147,
+      "step": 15990
+    },
+    {
+      "epoch": 3.634318181818182,
+      "grad_norm": 4.754489149614623,
+      "learning_rate": 2.1638696327748925e-07,
+      "loss": 0.0103,
+      "step": 15991
+    },
+    {
+      "epoch": 3.6345454545454547,
+      "grad_norm": 4.520726256771099,
+      "learning_rate": 2.1631943355994723e-07,
+      "loss": 0.0218,
+      "step": 15992
+    },
+    {
+      "epoch": 3.6347727272727273,
+      "grad_norm": 2.849111278569556,
+      "learning_rate": 2.1625191217611495e-07,
+      "loss": 0.0126,
+      "step": 15993
+    },
+    {
+      "epoch": 3.635,
+      "grad_norm": 1.694910186575933,
+      "learning_rate": 2.1618439912736926e-07,
+      "loss": 0.0241,
+      "step": 15994
+    },
+    {
+      "epoch": 3.6352272727272728,
+      "grad_norm": 2.5672191313466555,
+      "learning_rate": 2.161168944150869e-07,
+      "loss": 0.0254,
+      "step": 15995
+    },
+    {
+      "epoch": 3.6354545454545457,
+      "grad_norm": 3.1065916598367496,
+      "learning_rate": 2.160493980406444e-07,
+      "loss": 0.0307,
+      "step": 15996
+    },
+    {
+      "epoch": 3.6356818181818182,
+      "grad_norm": 1.847158787573095,
+      "learning_rate": 2.1598191000541807e-07,
+      "loss": 0.0117,
+      "step": 15997
+    },
+    {
+      "epoch": 3.6359090909090908,
+      "grad_norm": 2.098100707651361,
+      "learning_rate": 2.159144303107842e-07,
+      "loss": 0.0202,
+      "step": 15998
+    },
+    {
+      "epoch": 3.6361363636363637,
+      "grad_norm": 2.146253646297556,
+      "learning_rate": 2.1584695895811906e-07,
+      "loss": 0.0203,
+      "step": 15999
+    },
+    {
+      "epoch": 3.6363636363636362,
+      "grad_norm": 0.6726663455363552,
+      "learning_rate": 2.1577949594879799e-07,
+      "loss": 0.007,
+      "step": 16000
+    },
+    {
+      "epoch": 3.636590909090909,
+      "grad_norm": 0.723458425141288,
+      "learning_rate": 2.157120412841969e-07,
+      "loss": 0.0052,
+      "step": 16001
+    },
+    {
+      "epoch": 3.6368181818181817,
+      "grad_norm": 2.1703404598901845,
+      "learning_rate": 2.1564459496569128e-07,
+      "loss": 0.008,
+      "step": 16002
+    },
+    {
+      "epoch": 3.6370454545454547,
+      "grad_norm": 1.2970611510826544,
+      "learning_rate": 2.155771569946568e-07,
+      "loss": 0.0059,
+      "step": 16003
+    },
+    {
+      "epoch": 3.637272727272727,
+      "grad_norm": 1.6796055323203425,
+      "learning_rate": 2.1550972737246814e-07,
+      "loss": 0.0083,
+      "step": 16004
+    },
+    {
+      "epoch": 3.6375,
+      "grad_norm": 1.067972615583939,
+      "learning_rate": 2.154423061005006e-07,
+      "loss": 0.0074,
+      "step": 16005
+    },
+    {
+      "epoch": 3.6377272727272727,
+      "grad_norm": 0.8492371826043577,
+      "learning_rate": 2.1537489318012888e-07,
+      "loss": 0.0035,
+      "step": 16006
+    },
+    {
+      "epoch": 3.637954545454545,
+      "grad_norm": 0.563849353745603,
+      "learning_rate": 2.1530748861272773e-07,
+      "loss": 0.0033,
+      "step": 16007
+    },
+    {
+      "epoch": 3.638181818181818,
+      "grad_norm": 1.9084366291448136,
+      "learning_rate": 2.1524009239967163e-07,
+      "loss": 0.0122,
+      "step": 16008
+    },
+    {
+      "epoch": 3.638409090909091,
+      "grad_norm": 1.7282796059562124,
+      "learning_rate": 2.1517270454233498e-07,
+      "loss": 0.0164,
+      "step": 16009
+    },
+    {
+      "epoch": 3.6386363636363637,
+      "grad_norm": 2.3488608055508444,
+      "learning_rate": 2.1510532504209175e-07,
+      "loss": 0.0105,
+      "step": 16010
+    },
+    {
+      "epoch": 3.638863636363636,
+      "grad_norm": 2.4171193120391106,
+      "learning_rate": 2.1503795390031612e-07,
+      "loss": 0.015,
+      "step": 16011
+    },
+    {
+      "epoch": 3.639090909090909,
+      "grad_norm": 0.7666902884792639,
+      "learning_rate": 2.1497059111838206e-07,
+      "loss": 0.0023,
+      "step": 16012
+    },
+    {
+      "epoch": 3.6393181818181817,
+      "grad_norm": 0.8378575863007836,
+      "learning_rate": 2.149032366976627e-07,
+      "loss": 0.0048,
+      "step": 16013
+    },
+    {
+      "epoch": 3.6395454545454546,
+      "grad_norm": 1.6466518979765552,
+      "learning_rate": 2.1483589063953182e-07,
+      "loss": 0.0259,
+      "step": 16014
+    },
+    {
+      "epoch": 3.639772727272727,
+      "grad_norm": 3.454863173422985,
+      "learning_rate": 2.1476855294536275e-07,
+      "loss": 0.0341,
+      "step": 16015
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 2.097559441564089,
+      "learning_rate": 2.147012236165287e-07,
+      "loss": 0.0336,
+      "step": 16016
+    },
+    {
+      "epoch": 3.6402272727272726,
+      "grad_norm": 1.9081828845747713,
+      "learning_rate": 2.1463390265440233e-07,
+      "loss": 0.016,
+      "step": 16017
+    },
+    {
+      "epoch": 3.6404545454545456,
+      "grad_norm": 2.587543170034961,
+      "learning_rate": 2.1456659006035662e-07,
+      "loss": 0.0119,
+      "step": 16018
+    },
+    {
+      "epoch": 3.640681818181818,
+      "grad_norm": 1.354454874748149,
+      "learning_rate": 2.1449928583576418e-07,
+      "loss": 0.0129,
+      "step": 16019
+    },
+    {
+      "epoch": 3.6409090909090907,
+      "grad_norm": 1.0903943151268276,
+      "learning_rate": 2.144319899819977e-07,
+      "loss": 0.0067,
+      "step": 16020
+    },
+    {
+      "epoch": 3.6411363636363636,
+      "grad_norm": 1.814257609626117,
+      "learning_rate": 2.1436470250042898e-07,
+      "loss": 0.0133,
+      "step": 16021
+    },
+    {
+      "epoch": 3.6413636363636366,
+      "grad_norm": 2.410925452149287,
+      "learning_rate": 2.1429742339243034e-07,
+      "loss": 0.0211,
+      "step": 16022
+    },
+    {
+      "epoch": 3.641590909090909,
+      "grad_norm": 0.6143336753516889,
+      "learning_rate": 2.1423015265937375e-07,
+      "loss": 0.0085,
+      "step": 16023
+    },
+    {
+      "epoch": 3.6418181818181816,
+      "grad_norm": 1.4016654642535114,
+      "learning_rate": 2.1416289030263082e-07,
+      "loss": 0.0043,
+      "step": 16024
+    },
+    {
+      "epoch": 3.6420454545454546,
+      "grad_norm": 1.8050036057273868,
+      "learning_rate": 2.140956363235737e-07,
+      "loss": 0.0175,
+      "step": 16025
+    },
+    {
+      "epoch": 3.642272727272727,
+      "grad_norm": 1.8783225002314279,
+      "learning_rate": 2.1402839072357316e-07,
+      "loss": 0.0148,
+      "step": 16026
+    },
+    {
+      "epoch": 3.6425,
+      "grad_norm": 0.9583354837507704,
+      "learning_rate": 2.139611535040008e-07,
+      "loss": 0.0054,
+      "step": 16027
+    },
+    {
+      "epoch": 3.6427272727272726,
+      "grad_norm": 0.6167996543434537,
+      "learning_rate": 2.138939246662275e-07,
+      "loss": 0.0052,
+      "step": 16028
+    },
+    {
+      "epoch": 3.6429545454545456,
+      "grad_norm": 2.2315954798611455,
+      "learning_rate": 2.138267042116246e-07,
+      "loss": 0.0151,
+      "step": 16029
+    },
+    {
+      "epoch": 3.643181818181818,
+      "grad_norm": 1.5034378932634256,
+      "learning_rate": 2.1375949214156227e-07,
+      "loss": 0.0249,
+      "step": 16030
+    },
+    {
+      "epoch": 3.643409090909091,
+      "grad_norm": 1.862917476899397,
+      "learning_rate": 2.1369228845741136e-07,
+      "loss": 0.0143,
+      "step": 16031
+    },
+    {
+      "epoch": 3.6436363636363636,
+      "grad_norm": 2.2477874706659806,
+      "learning_rate": 2.1362509316054229e-07,
+      "loss": 0.0138,
+      "step": 16032
+    },
+    {
+      "epoch": 3.6438636363636365,
+      "grad_norm": 0.9907805441093067,
+      "learning_rate": 2.135579062523252e-07,
+      "loss": 0.0048,
+      "step": 16033
+    },
+    {
+      "epoch": 3.644090909090909,
+      "grad_norm": 2.621086192352543,
+      "learning_rate": 2.134907277341304e-07,
+      "loss": 0.0203,
+      "step": 16034
+    },
+    {
+      "epoch": 3.644318181818182,
+      "grad_norm": 1.213966745590946,
+      "learning_rate": 2.1342355760732744e-07,
+      "loss": 0.0054,
+      "step": 16035
+    },
+    {
+      "epoch": 3.6445454545454545,
+      "grad_norm": 1.266789074943036,
+      "learning_rate": 2.1335639587328607e-07,
+      "loss": 0.0087,
+      "step": 16036
+    },
+    {
+      "epoch": 3.644772727272727,
+      "grad_norm": 1.6404821142556232,
+      "learning_rate": 2.1328924253337601e-07,
+      "loss": 0.0053,
+      "step": 16037
+    },
+    {
+      "epoch": 3.645,
+      "grad_norm": 1.0663868358512056,
+      "learning_rate": 2.1322209758896658e-07,
+      "loss": 0.0058,
+      "step": 16038
+    },
+    {
+      "epoch": 3.645227272727273,
+      "grad_norm": 1.3737500979712924,
+      "learning_rate": 2.1315496104142692e-07,
+      "loss": 0.0074,
+      "step": 16039
+    },
+    {
+      "epoch": 3.6454545454545455,
+      "grad_norm": 0.7456082646444343,
+      "learning_rate": 2.1308783289212613e-07,
+      "loss": 0.0039,
+      "step": 16040
+    },
+    {
+      "epoch": 3.645681818181818,
+      "grad_norm": 2.820346587581378,
+      "learning_rate": 2.1302071314243307e-07,
+      "loss": 0.0151,
+      "step": 16041
+    },
+    {
+      "epoch": 3.645909090909091,
+      "grad_norm": 1.5074757978916793,
+      "learning_rate": 2.1295360179371656e-07,
+      "loss": 0.0107,
+      "step": 16042
+    },
+    {
+      "epoch": 3.6461363636363635,
+      "grad_norm": 0.8748656301736942,
+      "learning_rate": 2.128864988473448e-07,
+      "loss": 0.005,
+      "step": 16043
+    },
+    {
+      "epoch": 3.6463636363636365,
+      "grad_norm": 1.585222725294941,
+      "learning_rate": 2.1281940430468628e-07,
+      "loss": 0.0129,
+      "step": 16044
+    },
+    {
+      "epoch": 3.646590909090909,
+      "grad_norm": 1.479297152037242,
+      "learning_rate": 2.127523181671092e-07,
+      "loss": 0.0098,
+      "step": 16045
+    },
+    {
+      "epoch": 3.646818181818182,
+      "grad_norm": 0.7853262144618424,
+      "learning_rate": 2.126852404359816e-07,
+      "loss": 0.0057,
+      "step": 16046
+    },
+    {
+      "epoch": 3.6470454545454545,
+      "grad_norm": 1.897517459601675,
+      "learning_rate": 2.1261817111267139e-07,
+      "loss": 0.0203,
+      "step": 16047
+    },
+    {
+      "epoch": 3.6472727272727274,
+      "grad_norm": 1.1989444624849226,
+      "learning_rate": 2.12551110198546e-07,
+      "loss": 0.0097,
+      "step": 16048
+    },
+    {
+      "epoch": 3.6475,
+      "grad_norm": 1.3282779895193513,
+      "learning_rate": 2.12484057694973e-07,
+      "loss": 0.0237,
+      "step": 16049
+    },
+    {
+      "epoch": 3.6477272727272725,
+      "grad_norm": 1.2661194009752463,
+      "learning_rate": 2.124170136033197e-07,
+      "loss": 0.0117,
+      "step": 16050
+    },
+    {
+      "epoch": 3.6479545454545454,
+      "grad_norm": 0.8602187224881888,
+      "learning_rate": 2.1234997792495348e-07,
+      "loss": 0.0091,
+      "step": 16051
+    },
+    {
+      "epoch": 3.6481818181818184,
+      "grad_norm": 0.6263146384378763,
+      "learning_rate": 2.1228295066124096e-07,
+      "loss": 0.0039,
+      "step": 16052
+    },
+    {
+      "epoch": 3.648409090909091,
+      "grad_norm": 0.6720481708523653,
+      "learning_rate": 2.1221593181354887e-07,
+      "loss": 0.0055,
+      "step": 16053
+    },
+    {
+      "epoch": 3.6486363636363635,
+      "grad_norm": 1.173959118793135,
+      "learning_rate": 2.121489213832443e-07,
+      "loss": 0.0068,
+      "step": 16054
+    },
+    {
+      "epoch": 3.6488636363636364,
+      "grad_norm": 0.6428973557005778,
+      "learning_rate": 2.1208191937169344e-07,
+      "loss": 0.0034,
+      "step": 16055
+    },
+    {
+      "epoch": 3.649090909090909,
+      "grad_norm": 1.7218340651287203,
+      "learning_rate": 2.1201492578026282e-07,
+      "loss": 0.0133,
+      "step": 16056
+    },
+    {
+      "epoch": 3.649318181818182,
+      "grad_norm": 1.8992278185857296,
+      "learning_rate": 2.1194794061031816e-07,
+      "loss": 0.0101,
+      "step": 16057
+    },
+    {
+      "epoch": 3.6495454545454544,
+      "grad_norm": 1.9760795917409617,
+      "learning_rate": 2.1188096386322564e-07,
+      "loss": 0.0066,
+      "step": 16058
+    },
+    {
+      "epoch": 3.6497727272727274,
+      "grad_norm": 2.2202604691888235,
+      "learning_rate": 2.1181399554035093e-07,
+      "loss": 0.0232,
+      "step": 16059
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 2.2262788000284464,
+      "learning_rate": 2.1174703564305997e-07,
+      "loss": 0.0232,
+      "step": 16060
+    },
+    {
+      "epoch": 3.650227272727273,
+      "grad_norm": 2.9717509072605277,
+      "learning_rate": 2.116800841727177e-07,
+      "loss": 0.0122,
+      "step": 16061
+    },
+    {
+      "epoch": 3.6504545454545454,
+      "grad_norm": 1.4753719446900835,
+      "learning_rate": 2.1161314113068952e-07,
+      "loss": 0.0088,
+      "step": 16062
+    },
+    {
+      "epoch": 3.650681818181818,
+      "grad_norm": 2.5107749471262313,
+      "learning_rate": 2.1154620651834066e-07,
+      "loss": 0.0051,
+      "step": 16063
+    },
+    {
+      "epoch": 3.650909090909091,
+      "grad_norm": 1.8426057518762422,
+      "learning_rate": 2.1147928033703615e-07,
+      "loss": 0.0066,
+      "step": 16064
+    },
+    {
+      "epoch": 3.651136363636364,
+      "grad_norm": 1.7310941861537315,
+      "learning_rate": 2.114123625881403e-07,
+      "loss": 0.0125,
+      "step": 16065
+    },
+    {
+      "epoch": 3.6513636363636364,
+      "grad_norm": 1.7977216295552185,
+      "learning_rate": 2.1134545327301796e-07,
+      "loss": 0.0155,
+      "step": 16066
+    },
+    {
+      "epoch": 3.651590909090909,
+      "grad_norm": 0.9403578805010635,
+      "learning_rate": 2.1127855239303344e-07,
+      "loss": 0.0054,
+      "step": 16067
+    },
+    {
+      "epoch": 3.651818181818182,
+      "grad_norm": 1.5006147685450912,
+      "learning_rate": 2.1121165994955104e-07,
+      "loss": 0.008,
+      "step": 16068
+    },
+    {
+      "epoch": 3.652045454545455,
+      "grad_norm": 3.0019078425145875,
+      "learning_rate": 2.111447759439348e-07,
+      "loss": 0.0501,
+      "step": 16069
+    },
+    {
+      "epoch": 3.6522727272727273,
+      "grad_norm": 1.0246720425682125,
+      "learning_rate": 2.1107790037754857e-07,
+      "loss": 0.0069,
+      "step": 16070
+    },
+    {
+      "epoch": 3.6525,
+      "grad_norm": 0.2905946981841879,
+      "learning_rate": 2.1101103325175606e-07,
+      "loss": 0.0011,
+      "step": 16071
+    },
+    {
+      "epoch": 3.652727272727273,
+      "grad_norm": 0.8044671175337612,
+      "learning_rate": 2.1094417456792088e-07,
+      "loss": 0.0042,
+      "step": 16072
+    },
+    {
+      "epoch": 3.6529545454545453,
+      "grad_norm": 1.1606635814067467,
+      "learning_rate": 2.1087732432740648e-07,
+      "loss": 0.0055,
+      "step": 16073
+    },
+    {
+      "epoch": 3.6531818181818183,
+      "grad_norm": 2.0042839667308905,
+      "learning_rate": 2.1081048253157577e-07,
+      "loss": 0.0251,
+      "step": 16074
+    },
+    {
+      "epoch": 3.653409090909091,
+      "grad_norm": 0.7730681009997864,
+      "learning_rate": 2.1074364918179185e-07,
+      "loss": 0.0053,
+      "step": 16075
+    },
+    {
+      "epoch": 3.653636363636364,
+      "grad_norm": 1.6681131406853433,
+      "learning_rate": 2.1067682427941772e-07,
+      "loss": 0.011,
+      "step": 16076
+    },
+    {
+      "epoch": 3.6538636363636363,
+      "grad_norm": 0.7224774049516701,
+      "learning_rate": 2.106100078258159e-07,
+      "loss": 0.0071,
+      "step": 16077
+    },
+    {
+      "epoch": 3.6540909090909093,
+      "grad_norm": 1.7914584095705826,
+      "learning_rate": 2.1054319982234915e-07,
+      "loss": 0.0116,
+      "step": 16078
+    },
+    {
+      "epoch": 3.654318181818182,
+      "grad_norm": 1.123007642730181,
+      "learning_rate": 2.104764002703795e-07,
+      "loss": 0.0047,
+      "step": 16079
+    },
+    {
+      "epoch": 3.6545454545454543,
+      "grad_norm": 0.6257938578325546,
+      "learning_rate": 2.1040960917126916e-07,
+      "loss": 0.0022,
+      "step": 16080
+    },
+    {
+      "epoch": 3.6547727272727273,
+      "grad_norm": 1.4165505435723251,
+      "learning_rate": 2.1034282652638015e-07,
+      "loss": 0.0122,
+      "step": 16081
+    },
+    {
+      "epoch": 3.6550000000000002,
+      "grad_norm": 0.7710988544846227,
+      "learning_rate": 2.102760523370744e-07,
+      "loss": 0.0057,
+      "step": 16082
+    },
+    {
+      "epoch": 3.6552272727272728,
+      "grad_norm": 2.1780694634065285,
+      "learning_rate": 2.102092866047134e-07,
+      "loss": 0.0189,
+      "step": 16083
+    },
+    {
+      "epoch": 3.6554545454545453,
+      "grad_norm": 0.9732260767065213,
+      "learning_rate": 2.1014252933065873e-07,
+      "loss": 0.0076,
+      "step": 16084
+    },
+    {
+      "epoch": 3.6556818181818183,
+      "grad_norm": 0.3622236240519091,
+      "learning_rate": 2.1007578051627161e-07,
+      "loss": 0.0025,
+      "step": 16085
+    },
+    {
+      "epoch": 3.6559090909090908,
+      "grad_norm": 1.4803631976007352,
+      "learning_rate": 2.1000904016291347e-07,
+      "loss": 0.0105,
+      "step": 16086
+    },
+    {
+      "epoch": 3.6561363636363637,
+      "grad_norm": 1.8673541079056937,
+      "learning_rate": 2.0994230827194477e-07,
+      "loss": 0.0186,
+      "step": 16087
+    },
+    {
+      "epoch": 3.6563636363636363,
+      "grad_norm": 2.92451932784974,
+      "learning_rate": 2.0987558484472647e-07,
+      "loss": 0.0215,
+      "step": 16088
+    },
+    {
+      "epoch": 3.6565909090909092,
+      "grad_norm": 2.6387002491066998,
+      "learning_rate": 2.0980886988261923e-07,
+      "loss": 0.0142,
+      "step": 16089
+    },
+    {
+      "epoch": 3.6568181818181817,
+      "grad_norm": 1.5818528280085677,
+      "learning_rate": 2.0974216338698345e-07,
+      "loss": 0.0075,
+      "step": 16090
+    },
+    {
+      "epoch": 3.6570454545454547,
+      "grad_norm": 1.6322247486686503,
+      "learning_rate": 2.0967546535917968e-07,
+      "loss": 0.0122,
+      "step": 16091
+    },
+    {
+      "epoch": 3.6572727272727272,
+      "grad_norm": 2.2895213899475504,
+      "learning_rate": 2.096087758005675e-07,
+      "loss": 0.0122,
+      "step": 16092
+    },
+    {
+      "epoch": 3.6574999999999998,
+      "grad_norm": 1.676798324859186,
+      "learning_rate": 2.0954209471250708e-07,
+      "loss": 0.0129,
+      "step": 16093
+    },
+    {
+      "epoch": 3.6577272727272727,
+      "grad_norm": 2.0036318463485996,
+      "learning_rate": 2.0947542209635816e-07,
+      "loss": 0.01,
+      "step": 16094
+    },
+    {
+      "epoch": 3.6579545454545457,
+      "grad_norm": 0.9027242236208962,
+      "learning_rate": 2.0940875795348052e-07,
+      "loss": 0.0053,
+      "step": 16095
+    },
+    {
+      "epoch": 3.658181818181818,
+      "grad_norm": 0.8833581282601377,
+      "learning_rate": 2.0934210228523315e-07,
+      "loss": 0.0052,
+      "step": 16096
+    },
+    {
+      "epoch": 3.6584090909090907,
+      "grad_norm": 1.983744778383342,
+      "learning_rate": 2.0927545509297543e-07,
+      "loss": 0.0072,
+      "step": 16097
+    },
+    {
+      "epoch": 3.6586363636363637,
+      "grad_norm": 2.1903732396039506,
+      "learning_rate": 2.0920881637806646e-07,
+      "loss": 0.01,
+      "step": 16098
+    },
+    {
+      "epoch": 3.658863636363636,
+      "grad_norm": 2.5476677917538404,
+      "learning_rate": 2.0914218614186515e-07,
+      "loss": 0.0194,
+      "step": 16099
+    },
+    {
+      "epoch": 3.659090909090909,
+      "grad_norm": 0.5058733974398619,
+      "learning_rate": 2.0907556438573014e-07,
+      "loss": 0.0024,
+      "step": 16100
+    },
+    {
+      "epoch": 3.6593181818181817,
+      "grad_norm": 1.6648625096355631,
+      "learning_rate": 2.0900895111101994e-07,
+      "loss": 0.0196,
+      "step": 16101
+    },
+    {
+      "epoch": 3.6595454545454547,
+      "grad_norm": 5.548807413836822,
+      "learning_rate": 2.08942346319093e-07,
+      "loss": 0.0357,
+      "step": 16102
+    },
+    {
+      "epoch": 3.659772727272727,
+      "grad_norm": 1.4351227134422302,
+      "learning_rate": 2.0887575001130748e-07,
+      "loss": 0.0058,
+      "step": 16103
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 2.963439153914582,
+      "learning_rate": 2.088091621890216e-07,
+      "loss": 0.011,
+      "step": 16104
+    },
+    {
+      "epoch": 3.6602272727272727,
+      "grad_norm": 2.0148345696727605,
+      "learning_rate": 2.087425828535928e-07,
+      "loss": 0.0106,
+      "step": 16105
+    },
+    {
+      "epoch": 3.660454545454545,
+      "grad_norm": 2.893814161221206,
+      "learning_rate": 2.086760120063789e-07,
+      "loss": 0.0127,
+      "step": 16106
+    },
+    {
+      "epoch": 3.660681818181818,
+      "grad_norm": 2.5061182748271493,
+      "learning_rate": 2.0860944964873745e-07,
+      "loss": 0.0135,
+      "step": 16107
+    },
+    {
+      "epoch": 3.660909090909091,
+      "grad_norm": 1.9017117809703719,
+      "learning_rate": 2.0854289578202595e-07,
+      "loss": 0.0143,
+      "step": 16108
+    },
+    {
+      "epoch": 3.6611363636363636,
+      "grad_norm": 5.510614492487833,
+      "learning_rate": 2.0847635040760117e-07,
+      "loss": 0.0194,
+      "step": 16109
+    },
+    {
+      "epoch": 3.661363636363636,
+      "grad_norm": 0.8026791208250851,
+      "learning_rate": 2.0840981352682025e-07,
+      "loss": 0.0058,
+      "step": 16110
+    },
+    {
+      "epoch": 3.661590909090909,
+      "grad_norm": 2.3947369171133457,
+      "learning_rate": 2.0834328514104004e-07,
+      "loss": 0.0234,
+      "step": 16111
+    },
+    {
+      "epoch": 3.661818181818182,
+      "grad_norm": 1.8138180609349868,
+      "learning_rate": 2.0827676525161712e-07,
+      "loss": 0.0083,
+      "step": 16112
+    },
+    {
+      "epoch": 3.6620454545454546,
+      "grad_norm": 1.3967465387793816,
+      "learning_rate": 2.0821025385990795e-07,
+      "loss": 0.0084,
+      "step": 16113
+    },
+    {
+      "epoch": 3.662272727272727,
+      "grad_norm": 1.2711614124735324,
+      "learning_rate": 2.0814375096726888e-07,
+      "loss": 0.0067,
+      "step": 16114
+    },
+    {
+      "epoch": 3.6625,
+      "grad_norm": 1.780144753782806,
+      "learning_rate": 2.0807725657505593e-07,
+      "loss": 0.01,
+      "step": 16115
+    },
+    {
+      "epoch": 3.6627272727272726,
+      "grad_norm": 6.36812687481376,
+      "learning_rate": 2.0801077068462508e-07,
+      "loss": 0.0409,
+      "step": 16116
+    },
+    {
+      "epoch": 3.6629545454545456,
+      "grad_norm": 1.1408838492040387,
+      "learning_rate": 2.079442932973324e-07,
+      "loss": 0.0068,
+      "step": 16117
+    },
+    {
+      "epoch": 3.663181818181818,
+      "grad_norm": 3.010392690177395,
+      "learning_rate": 2.0787782441453284e-07,
+      "loss": 0.013,
+      "step": 16118
+    },
+    {
+      "epoch": 3.663409090909091,
+      "grad_norm": 0.8272440533612032,
+      "learning_rate": 2.0781136403758222e-07,
+      "loss": 0.0109,
+      "step": 16119
+    },
+    {
+      "epoch": 3.6636363636363636,
+      "grad_norm": 1.4629930826880082,
+      "learning_rate": 2.077449121678357e-07,
+      "loss": 0.0167,
+      "step": 16120
+    },
+    {
+      "epoch": 3.6638636363636365,
+      "grad_norm": 0.4962669398623692,
+      "learning_rate": 2.0767846880664834e-07,
+      "loss": 0.0024,
+      "step": 16121
+    },
+    {
+      "epoch": 3.664090909090909,
+      "grad_norm": 1.265111656987057,
+      "learning_rate": 2.0761203395537527e-07,
+      "loss": 0.0065,
+      "step": 16122
+    },
+    {
+      "epoch": 3.6643181818181816,
+      "grad_norm": 2.7139460023703053,
+      "learning_rate": 2.0754560761537076e-07,
+      "loss": 0.0083,
+      "step": 16123
+    },
+    {
+      "epoch": 3.6645454545454546,
+      "grad_norm": 2.7646836348209973,
+      "learning_rate": 2.0747918978798972e-07,
+      "loss": 0.0185,
+      "step": 16124
+    },
+    {
+      "epoch": 3.6647727272727275,
+      "grad_norm": 2.062289933704907,
+      "learning_rate": 2.0741278047458632e-07,
+      "loss": 0.0282,
+      "step": 16125
+    },
+    {
+      "epoch": 3.665,
+      "grad_norm": 0.76749345630099,
+      "learning_rate": 2.0734637967651503e-07,
+      "loss": 0.0042,
+      "step": 16126
+    },
+    {
+      "epoch": 3.6652272727272726,
+      "grad_norm": 1.8298261601865884,
+      "learning_rate": 2.072799873951295e-07,
+      "loss": 0.0117,
+      "step": 16127
+    },
+    {
+      "epoch": 3.6654545454545455,
+      "grad_norm": 1.5360918839333007,
+      "learning_rate": 2.072136036317838e-07,
+      "loss": 0.0103,
+      "step": 16128
+    },
+    {
+      "epoch": 3.665681818181818,
+      "grad_norm": 1.7600303727142752,
+      "learning_rate": 2.071472283878316e-07,
+      "loss": 0.013,
+      "step": 16129
+    },
+    {
+      "epoch": 3.665909090909091,
+      "grad_norm": 1.3268799098062722,
+      "learning_rate": 2.0708086166462639e-07,
+      "loss": 0.0094,
+      "step": 16130
+    },
+    {
+      "epoch": 3.6661363636363635,
+      "grad_norm": 1.0028328574138388,
+      "learning_rate": 2.0701450346352152e-07,
+      "loss": 0.0061,
+      "step": 16131
+    },
+    {
+      "epoch": 3.6663636363636365,
+      "grad_norm": 1.5451930263141223,
+      "learning_rate": 2.0694815378587008e-07,
+      "loss": 0.0095,
+      "step": 16132
+    },
+    {
+      "epoch": 3.666590909090909,
+      "grad_norm": 1.1423464221523614,
+      "learning_rate": 2.068818126330252e-07,
+      "loss": 0.0107,
+      "step": 16133
+    },
+    {
+      "epoch": 3.666818181818182,
+      "grad_norm": 2.009100752946933,
+      "learning_rate": 2.0681548000633955e-07,
+      "loss": 0.0122,
+      "step": 16134
+    },
+    {
+      "epoch": 3.6670454545454545,
+      "grad_norm": 2.170485164486619,
+      "learning_rate": 2.0674915590716604e-07,
+      "loss": 0.0117,
+      "step": 16135
+    },
+    {
+      "epoch": 3.667272727272727,
+      "grad_norm": 0.977726141094126,
+      "learning_rate": 2.0668284033685677e-07,
+      "loss": 0.007,
+      "step": 16136
+    },
+    {
+      "epoch": 3.6675,
+      "grad_norm": 1.5790609291333497,
+      "learning_rate": 2.0661653329676408e-07,
+      "loss": 0.0133,
+      "step": 16137
+    },
+    {
+      "epoch": 3.667727272727273,
+      "grad_norm": 1.128567019315453,
+      "learning_rate": 2.0655023478824024e-07,
+      "loss": 0.0111,
+      "step": 16138
+    },
+    {
+      "epoch": 3.6679545454545455,
+      "grad_norm": 3.4419016901203903,
+      "learning_rate": 2.064839448126374e-07,
+      "loss": 0.0259,
+      "step": 16139
+    },
+    {
+      "epoch": 3.668181818181818,
+      "grad_norm": 0.7280827966012733,
+      "learning_rate": 2.0641766337130678e-07,
+      "loss": 0.0054,
+      "step": 16140
+    },
+    {
+      "epoch": 3.668409090909091,
+      "grad_norm": 1.8264763745239934,
+      "learning_rate": 2.0635139046560025e-07,
+      "loss": 0.0107,
+      "step": 16141
+    },
+    {
+      "epoch": 3.6686363636363635,
+      "grad_norm": 2.139122098235541,
+      "learning_rate": 2.0628512609686929e-07,
+      "loss": 0.0186,
+      "step": 16142
+    },
+    {
+      "epoch": 3.6688636363636364,
+      "grad_norm": 0.9010145659691177,
+      "learning_rate": 2.062188702664651e-07,
+      "loss": 0.0059,
+      "step": 16143
+    },
+    {
+      "epoch": 3.669090909090909,
+      "grad_norm": 1.5215692736480078,
+      "learning_rate": 2.0615262297573872e-07,
+      "loss": 0.0164,
+      "step": 16144
+    },
+    {
+      "epoch": 3.669318181818182,
+      "grad_norm": 0.7260007232588884,
+      "learning_rate": 2.0608638422604112e-07,
+      "loss": 0.0023,
+      "step": 16145
+    },
+    {
+      "epoch": 3.6695454545454544,
+      "grad_norm": 1.2277729981860814,
+      "learning_rate": 2.0602015401872297e-07,
+      "loss": 0.0107,
+      "step": 16146
+    },
+    {
+      "epoch": 3.6697727272727274,
+      "grad_norm": 1.4687893997827899,
+      "learning_rate": 2.0595393235513475e-07,
+      "loss": 0.0076,
+      "step": 16147
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 1.5777381632828809,
+      "learning_rate": 2.0588771923662724e-07,
+      "loss": 0.0063,
+      "step": 16148
+    },
+    {
+      "epoch": 3.670227272727273,
+      "grad_norm": 0.5970510470384669,
+      "learning_rate": 2.0582151466455e-07,
+      "loss": 0.0015,
+      "step": 16149
+    },
+    {
+      "epoch": 3.6704545454545454,
+      "grad_norm": 0.6160632722360743,
+      "learning_rate": 2.0575531864025346e-07,
+      "loss": 0.0021,
+      "step": 16150
+    },
+    {
+      "epoch": 3.6706818181818184,
+      "grad_norm": 0.6042176105556736,
+      "learning_rate": 2.0568913116508737e-07,
+      "loss": 0.005,
+      "step": 16151
+    },
+    {
+      "epoch": 3.670909090909091,
+      "grad_norm": 1.4824113755764412,
+      "learning_rate": 2.0562295224040158e-07,
+      "loss": 0.0133,
+      "step": 16152
+    },
+    {
+      "epoch": 3.6711363636363634,
+      "grad_norm": 0.8156263745066761,
+      "learning_rate": 2.0555678186754524e-07,
+      "loss": 0.0067,
+      "step": 16153
+    },
+    {
+      "epoch": 3.6713636363636364,
+      "grad_norm": 0.6330380056787052,
+      "learning_rate": 2.0549062004786792e-07,
+      "loss": 0.0018,
+      "step": 16154
+    },
+    {
+      "epoch": 3.6715909090909093,
+      "grad_norm": 3.2125874509432264,
+      "learning_rate": 2.0542446678271869e-07,
+      "loss": 0.0556,
+      "step": 16155
+    },
+    {
+      "epoch": 3.671818181818182,
+      "grad_norm": 1.1715304002010825,
+      "learning_rate": 2.053583220734466e-07,
+      "loss": 0.0068,
+      "step": 16156
+    },
+    {
+      "epoch": 3.6720454545454544,
+      "grad_norm": 0.9839667032752457,
+      "learning_rate": 2.0529218592140063e-07,
+      "loss": 0.0064,
+      "step": 16157
+    },
+    {
+      "epoch": 3.6722727272727274,
+      "grad_norm": 0.7158771452670245,
+      "learning_rate": 2.0522605832792898e-07,
+      "loss": 0.0041,
+      "step": 16158
+    },
+    {
+      "epoch": 3.6725,
+      "grad_norm": 2.801900334875112,
+      "learning_rate": 2.0515993929438038e-07,
+      "loss": 0.0104,
+      "step": 16159
+    },
+    {
+      "epoch": 3.672727272727273,
+      "grad_norm": 1.797088371305229,
+      "learning_rate": 2.0509382882210306e-07,
+      "loss": 0.008,
+      "step": 16160
+    },
+    {
+      "epoch": 3.6729545454545454,
+      "grad_norm": 1.2969044309282944,
+      "learning_rate": 2.050277269124451e-07,
+      "loss": 0.0112,
+      "step": 16161
+    },
+    {
+      "epoch": 3.6731818181818183,
+      "grad_norm": 1.881485233223364,
+      "learning_rate": 2.0496163356675453e-07,
+      "loss": 0.0103,
+      "step": 16162
+    },
+    {
+      "epoch": 3.673409090909091,
+      "grad_norm": 2.7387752963910055,
+      "learning_rate": 2.04895548786379e-07,
+      "loss": 0.0241,
+      "step": 16163
+    },
+    {
+      "epoch": 3.673636363636364,
+      "grad_norm": 1.4381975149884587,
+      "learning_rate": 2.0482947257266614e-07,
+      "loss": 0.012,
+      "step": 16164
+    },
+    {
+      "epoch": 3.6738636363636363,
+      "grad_norm": 1.383979803291462,
+      "learning_rate": 2.047634049269636e-07,
+      "loss": 0.0049,
+      "step": 16165
+    },
+    {
+      "epoch": 3.674090909090909,
+      "grad_norm": 2.859478810457569,
+      "learning_rate": 2.046973458506181e-07,
+      "loss": 0.0122,
+      "step": 16166
+    },
+    {
+      "epoch": 3.674318181818182,
+      "grad_norm": 2.2708128805056327,
+      "learning_rate": 2.0463129534497702e-07,
+      "loss": 0.0151,
+      "step": 16167
+    },
+    {
+      "epoch": 3.674545454545455,
+      "grad_norm": 0.8977846103909951,
+      "learning_rate": 2.0456525341138725e-07,
+      "loss": 0.005,
+      "step": 16168
+    },
+    {
+      "epoch": 3.6747727272727273,
+      "grad_norm": 1.4300324408119696,
+      "learning_rate": 2.044992200511954e-07,
+      "loss": 0.007,
+      "step": 16169
+    },
+    {
+      "epoch": 3.675,
+      "grad_norm": 1.1823665631802973,
+      "learning_rate": 2.0443319526574828e-07,
+      "loss": 0.0076,
+      "step": 16170
+    },
+    {
+      "epoch": 3.675227272727273,
+      "grad_norm": 2.9995361339919775,
+      "learning_rate": 2.0436717905639183e-07,
+      "loss": 0.0085,
+      "step": 16171
+    },
+    {
+      "epoch": 3.6754545454545453,
+      "grad_norm": 2.3202254373992517,
+      "learning_rate": 2.043011714244724e-07,
+      "loss": 0.0105,
+      "step": 16172
+    },
+    {
+      "epoch": 3.6756818181818183,
+      "grad_norm": 0.6776895565952681,
+      "learning_rate": 2.04235172371336e-07,
+      "loss": 0.0038,
+      "step": 16173
+    },
+    {
+      "epoch": 3.675909090909091,
+      "grad_norm": 2.41041392058561,
+      "learning_rate": 2.0416918189832858e-07,
+      "loss": 0.0193,
+      "step": 16174
+    },
+    {
+      "epoch": 3.6761363636363638,
+      "grad_norm": 2.400019151794406,
+      "learning_rate": 2.0410320000679571e-07,
+      "loss": 0.0094,
+      "step": 16175
+    },
+    {
+      "epoch": 3.6763636363636363,
+      "grad_norm": 2.109124864216686,
+      "learning_rate": 2.0403722669808285e-07,
+      "loss": 0.0076,
+      "step": 16176
+    },
+    {
+      "epoch": 3.6765909090909092,
+      "grad_norm": 1.8297873368186,
+      "learning_rate": 2.039712619735354e-07,
+      "loss": 0.0238,
+      "step": 16177
+    },
+    {
+      "epoch": 3.6768181818181818,
+      "grad_norm": 0.41741763015996214,
+      "learning_rate": 2.0390530583449841e-07,
+      "loss": 0.0008,
+      "step": 16178
+    },
+    {
+      "epoch": 3.6770454545454543,
+      "grad_norm": 1.4267360219901315,
+      "learning_rate": 2.0383935828231707e-07,
+      "loss": 0.0132,
+      "step": 16179
+    },
+    {
+      "epoch": 3.6772727272727272,
+      "grad_norm": 1.0374229735048974,
+      "learning_rate": 2.0377341931833578e-07,
+      "loss": 0.0083,
+      "step": 16180
+    },
+    {
+      "epoch": 3.6775,
+      "grad_norm": 1.5713536528989624,
+      "learning_rate": 2.0370748894389933e-07,
+      "loss": 0.0185,
+      "step": 16181
+    },
+    {
+      "epoch": 3.6777272727272727,
+      "grad_norm": 1.1053488328765606,
+      "learning_rate": 2.0364156716035213e-07,
+      "loss": 0.0049,
+      "step": 16182
+    },
+    {
+      "epoch": 3.6779545454545453,
+      "grad_norm": 0.8238016300861086,
+      "learning_rate": 2.0357565396903873e-07,
+      "loss": 0.0042,
+      "step": 16183
+    },
+    {
+      "epoch": 3.678181818181818,
+      "grad_norm": 2.1981134526349297,
+      "learning_rate": 2.0350974937130279e-07,
+      "loss": 0.0259,
+      "step": 16184
+    },
+    {
+      "epoch": 3.678409090909091,
+      "grad_norm": 1.8252604394238041,
+      "learning_rate": 2.0344385336848834e-07,
+      "loss": 0.0081,
+      "step": 16185
+    },
+    {
+      "epoch": 3.6786363636363637,
+      "grad_norm": 1.1712184222428654,
+      "learning_rate": 2.0337796596193915e-07,
+      "loss": 0.0085,
+      "step": 16186
+    },
+    {
+      "epoch": 3.6788636363636362,
+      "grad_norm": 2.706477399537676,
+      "learning_rate": 2.0331208715299902e-07,
+      "loss": 0.015,
+      "step": 16187
+    },
+    {
+      "epoch": 3.679090909090909,
+      "grad_norm": 1.5858157025881772,
+      "learning_rate": 2.0324621694301092e-07,
+      "loss": 0.0145,
+      "step": 16188
+    },
+    {
+      "epoch": 3.6793181818181817,
+      "grad_norm": 1.9432781998637048,
+      "learning_rate": 2.0318035533331823e-07,
+      "loss": 0.0117,
+      "step": 16189
+    },
+    {
+      "epoch": 3.6795454545454547,
+      "grad_norm": 2.1222721126569697,
+      "learning_rate": 2.0311450232526398e-07,
+      "loss": 0.0119,
+      "step": 16190
+    },
+    {
+      "epoch": 3.679772727272727,
+      "grad_norm": 2.1864779943233192,
+      "learning_rate": 2.030486579201909e-07,
+      "loss": 0.0127,
+      "step": 16191
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 1.8923131659652013,
+      "learning_rate": 2.0298282211944216e-07,
+      "loss": 0.014,
+      "step": 16192
+    },
+    {
+      "epoch": 3.6802272727272727,
+      "grad_norm": 2.219365774112196,
+      "learning_rate": 2.0291699492435972e-07,
+      "loss": 0.0185,
+      "step": 16193
+    },
+    {
+      "epoch": 3.6804545454545456,
+      "grad_norm": 2.498968005715419,
+      "learning_rate": 2.0285117633628616e-07,
+      "loss": 0.015,
+      "step": 16194
+    },
+    {
+      "epoch": 3.680681818181818,
+      "grad_norm": 1.4706929455917617,
+      "learning_rate": 2.0278536635656363e-07,
+      "loss": 0.0118,
+      "step": 16195
+    },
+    {
+      "epoch": 3.6809090909090907,
+      "grad_norm": 1.739751977302912,
+      "learning_rate": 2.027195649865342e-07,
+      "loss": 0.011,
+      "step": 16196
+    },
+    {
+      "epoch": 3.6811363636363637,
+      "grad_norm": 0.8859734661806954,
+      "learning_rate": 2.0265377222753936e-07,
+      "loss": 0.0074,
+      "step": 16197
+    },
+    {
+      "epoch": 3.6813636363636366,
+      "grad_norm": 3.1487940839583817,
+      "learning_rate": 2.0258798808092097e-07,
+      "loss": 0.0239,
+      "step": 16198
+    },
+    {
+      "epoch": 3.681590909090909,
+      "grad_norm": 2.4134264389133793,
+      "learning_rate": 2.0252221254802044e-07,
+      "loss": 0.0126,
+      "step": 16199
+    },
+    {
+      "epoch": 3.6818181818181817,
+      "grad_norm": 0.8956279371934251,
+      "learning_rate": 2.0245644563017898e-07,
+      "loss": 0.0052,
+      "step": 16200
+    },
+    {
+      "epoch": 3.6820454545454546,
+      "grad_norm": 1.724897710999318,
+      "learning_rate": 2.0239068732873804e-07,
+      "loss": 0.0063,
+      "step": 16201
+    },
+    {
+      "epoch": 3.682272727272727,
+      "grad_norm": 1.6350692763176666,
+      "learning_rate": 2.02324937645038e-07,
+      "loss": 0.0113,
+      "step": 16202
+    },
+    {
+      "epoch": 3.6825,
+      "grad_norm": 1.6652230195706412,
+      "learning_rate": 2.0225919658041994e-07,
+      "loss": 0.0085,
+      "step": 16203
+    },
+    {
+      "epoch": 3.6827272727272726,
+      "grad_norm": 1.2630763718408171,
+      "learning_rate": 2.0219346413622437e-07,
+      "loss": 0.0131,
+      "step": 16204
+    },
+    {
+      "epoch": 3.6829545454545456,
+      "grad_norm": 1.9443444473215796,
+      "learning_rate": 2.0212774031379162e-07,
+      "loss": 0.022,
+      "step": 16205
+    },
+    {
+      "epoch": 3.683181818181818,
+      "grad_norm": 1.584817402331907,
+      "learning_rate": 2.0206202511446206e-07,
+      "loss": 0.0098,
+      "step": 16206
+    },
+    {
+      "epoch": 3.683409090909091,
+      "grad_norm": 1.3006194816483714,
+      "learning_rate": 2.0199631853957562e-07,
+      "loss": 0.0064,
+      "step": 16207
+    },
+    {
+      "epoch": 3.6836363636363636,
+      "grad_norm": 0.534582347821083,
+      "learning_rate": 2.0193062059047226e-07,
+      "loss": 0.0037,
+      "step": 16208
+    },
+    {
+      "epoch": 3.683863636363636,
+      "grad_norm": 1.2933625583238484,
+      "learning_rate": 2.018649312684918e-07,
+      "loss": 0.0063,
+      "step": 16209
+    },
+    {
+      "epoch": 3.684090909090909,
+      "grad_norm": 2.3486689935537055,
+      "learning_rate": 2.0179925057497338e-07,
+      "loss": 0.0082,
+      "step": 16210
+    },
+    {
+      "epoch": 3.684318181818182,
+      "grad_norm": 0.25225107009194897,
+      "learning_rate": 2.0173357851125654e-07,
+      "loss": 0.001,
+      "step": 16211
+    },
+    {
+      "epoch": 3.6845454545454546,
+      "grad_norm": 1.6323690122404622,
+      "learning_rate": 2.016679150786805e-07,
+      "loss": 0.0088,
+      "step": 16212
+    },
+    {
+      "epoch": 3.684772727272727,
+      "grad_norm": 1.4460796918157486,
+      "learning_rate": 2.016022602785842e-07,
+      "loss": 0.0064,
+      "step": 16213
+    },
+    {
+      "epoch": 3.685,
+      "grad_norm": 1.1298303728357701,
+      "learning_rate": 2.0153661411230674e-07,
+      "loss": 0.0051,
+      "step": 16214
+    },
+    {
+      "epoch": 3.6852272727272726,
+      "grad_norm": 3.88161167502186,
+      "learning_rate": 2.014709765811862e-07,
+      "loss": 0.0342,
+      "step": 16215
+    },
+    {
+      "epoch": 3.6854545454545455,
+      "grad_norm": 1.6526355714339573,
+      "learning_rate": 2.0140534768656135e-07,
+      "loss": 0.0074,
+      "step": 16216
+    },
+    {
+      "epoch": 3.685681818181818,
+      "grad_norm": 1.9869558434547931,
+      "learning_rate": 2.013397274297705e-07,
+      "loss": 0.0148,
+      "step": 16217
+    },
+    {
+      "epoch": 3.685909090909091,
+      "grad_norm": 0.6649577910622456,
+      "learning_rate": 2.0127411581215187e-07,
+      "loss": 0.0062,
+      "step": 16218
+    },
+    {
+      "epoch": 3.6861363636363635,
+      "grad_norm": 1.5935784546926253,
+      "learning_rate": 2.0120851283504308e-07,
+      "loss": 0.0111,
+      "step": 16219
+    },
+    {
+      "epoch": 3.6863636363636365,
+      "grad_norm": 1.9131398085375266,
+      "learning_rate": 2.0114291849978206e-07,
+      "loss": 0.0081,
+      "step": 16220
+    },
+    {
+      "epoch": 3.686590909090909,
+      "grad_norm": 0.8879542322234147,
+      "learning_rate": 2.010773328077062e-07,
+      "loss": 0.0076,
+      "step": 16221
+    },
+    {
+      "epoch": 3.6868181818181816,
+      "grad_norm": 2.028590975815924,
+      "learning_rate": 2.0101175576015327e-07,
+      "loss": 0.0128,
+      "step": 16222
+    },
+    {
+      "epoch": 3.6870454545454545,
+      "grad_norm": 0.8946942555524833,
+      "learning_rate": 2.0094618735846055e-07,
+      "loss": 0.0075,
+      "step": 16223
+    },
+    {
+      "epoch": 3.6872727272727275,
+      "grad_norm": 1.8912644095699689,
+      "learning_rate": 2.0088062760396467e-07,
+      "loss": 0.0084,
+      "step": 16224
+    },
+    {
+      "epoch": 3.6875,
+      "grad_norm": 0.2218691679969775,
+      "learning_rate": 2.008150764980027e-07,
+      "loss": 0.0015,
+      "step": 16225
+    },
+    {
+      "epoch": 3.6877272727272725,
+      "grad_norm": 1.622402520154276,
+      "learning_rate": 2.007495340419114e-07,
+      "loss": 0.0138,
+      "step": 16226
+    },
+    {
+      "epoch": 3.6879545454545455,
+      "grad_norm": 1.136302975493611,
+      "learning_rate": 2.0068400023702737e-07,
+      "loss": 0.0142,
+      "step": 16227
+    },
+    {
+      "epoch": 3.6881818181818184,
+      "grad_norm": 0.38783006449447777,
+      "learning_rate": 2.0061847508468665e-07,
+      "loss": 0.0008,
+      "step": 16228
+    },
+    {
+      "epoch": 3.688409090909091,
+      "grad_norm": 1.4412920146626602,
+      "learning_rate": 2.0055295858622564e-07,
+      "loss": 0.0037,
+      "step": 16229
+    },
+    {
+      "epoch": 3.6886363636363635,
+      "grad_norm": 2.433686045137656,
+      "learning_rate": 2.0048745074298028e-07,
+      "loss": 0.014,
+      "step": 16230
+    },
+    {
+      "epoch": 3.6888636363636365,
+      "grad_norm": 10.656425501848481,
+      "learning_rate": 2.0042195155628656e-07,
+      "loss": 0.0241,
+      "step": 16231
+    },
+    {
+      "epoch": 3.689090909090909,
+      "grad_norm": 1.321913732853815,
+      "learning_rate": 2.0035646102747982e-07,
+      "loss": 0.008,
+      "step": 16232
+    },
+    {
+      "epoch": 3.689318181818182,
+      "grad_norm": 1.0761733859170997,
+      "learning_rate": 2.0029097915789566e-07,
+      "loss": 0.0063,
+      "step": 16233
+    },
+    {
+      "epoch": 3.6895454545454545,
+      "grad_norm": 1.7564519904551705,
+      "learning_rate": 2.0022550594886932e-07,
+      "loss": 0.0129,
+      "step": 16234
+    },
+    {
+      "epoch": 3.6897727272727274,
+      "grad_norm": 0.885718482033691,
+      "learning_rate": 2.0016004140173608e-07,
+      "loss": 0.0049,
+      "step": 16235
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.42513778249416107,
+      "learning_rate": 2.0009458551783065e-07,
+      "loss": 0.0018,
+      "step": 16236
+    },
+    {
+      "epoch": 3.690227272727273,
+      "grad_norm": 1.9490711806554157,
+      "learning_rate": 2.00029138298488e-07,
+      "loss": 0.0213,
+      "step": 16237
+    },
+    {
+      "epoch": 3.6904545454545454,
+      "grad_norm": 1.953484579683175,
+      "learning_rate": 1.9996369974504255e-07,
+      "loss": 0.0121,
+      "step": 16238
+    },
+    {
+      "epoch": 3.690681818181818,
+      "grad_norm": 1.9258380849508794,
+      "learning_rate": 1.9989826985882886e-07,
+      "loss": 0.0071,
+      "step": 16239
+    },
+    {
+      "epoch": 3.690909090909091,
+      "grad_norm": 0.9665174480416665,
+      "learning_rate": 1.9983284864118122e-07,
+      "loss": 0.0079,
+      "step": 16240
+    },
+    {
+      "epoch": 3.691136363636364,
+      "grad_norm": 1.7919443930150405,
+      "learning_rate": 1.9976743609343338e-07,
+      "loss": 0.0154,
+      "step": 16241
+    },
+    {
+      "epoch": 3.6913636363636364,
+      "grad_norm": 0.9572556103256942,
+      "learning_rate": 1.9970203221691935e-07,
+      "loss": 0.0098,
+      "step": 16242
+    },
+    {
+      "epoch": 3.691590909090909,
+      "grad_norm": 1.0242749065819612,
+      "learning_rate": 1.9963663701297285e-07,
+      "loss": 0.0056,
+      "step": 16243
+    },
+    {
+      "epoch": 3.691818181818182,
+      "grad_norm": 2.7660930899437046,
+      "learning_rate": 1.9957125048292746e-07,
+      "loss": 0.006,
+      "step": 16244
+    },
+    {
+      "epoch": 3.6920454545454544,
+      "grad_norm": 0.9676698058102915,
+      "learning_rate": 1.9950587262811662e-07,
+      "loss": 0.013,
+      "step": 16245
+    },
+    {
+      "epoch": 3.6922727272727274,
+      "grad_norm": 2.0983396378683428,
+      "learning_rate": 1.9944050344987316e-07,
+      "loss": 0.0433,
+      "step": 16246
+    },
+    {
+      "epoch": 3.6925,
+      "grad_norm": 0.9950903568490531,
+      "learning_rate": 1.9937514294953024e-07,
+      "loss": 0.0035,
+      "step": 16247
+    },
+    {
+      "epoch": 3.692727272727273,
+      "grad_norm": 3.9691171308609006,
+      "learning_rate": 1.993097911284208e-07,
+      "loss": 0.0183,
+      "step": 16248
+    },
+    {
+      "epoch": 3.6929545454545454,
+      "grad_norm": 1.391907573165208,
+      "learning_rate": 1.9924444798787744e-07,
+      "loss": 0.0037,
+      "step": 16249
+    },
+    {
+      "epoch": 3.6931818181818183,
+      "grad_norm": 1.174130606030493,
+      "learning_rate": 1.9917911352923225e-07,
+      "loss": 0.0091,
+      "step": 16250
+    },
+    {
+      "epoch": 3.693409090909091,
+      "grad_norm": 2.2463359611303044,
+      "learning_rate": 1.9911378775381802e-07,
+      "loss": 0.0219,
+      "step": 16251
+    },
+    {
+      "epoch": 3.6936363636363634,
+      "grad_norm": 1.1607411876064542,
+      "learning_rate": 1.990484706629666e-07,
+      "loss": 0.0073,
+      "step": 16252
+    },
+    {
+      "epoch": 3.6938636363636363,
+      "grad_norm": 2.1230019272030556,
+      "learning_rate": 1.9898316225801024e-07,
+      "loss": 0.0164,
+      "step": 16253
+    },
+    {
+      "epoch": 3.6940909090909093,
+      "grad_norm": 0.37989648982625346,
+      "learning_rate": 1.9891786254028016e-07,
+      "loss": 0.0038,
+      "step": 16254
+    },
+    {
+      "epoch": 3.694318181818182,
+      "grad_norm": 2.1355655035402354,
+      "learning_rate": 1.9885257151110822e-07,
+      "loss": 0.0148,
+      "step": 16255
+    },
+    {
+      "epoch": 3.6945454545454544,
+      "grad_norm": 2.226004452285899,
+      "learning_rate": 1.9878728917182586e-07,
+      "loss": 0.0155,
+      "step": 16256
+    },
+    {
+      "epoch": 3.6947727272727273,
+      "grad_norm": 1.7058163166594806,
+      "learning_rate": 1.9872201552376422e-07,
+      "loss": 0.0122,
+      "step": 16257
+    },
+    {
+      "epoch": 3.695,
+      "grad_norm": 1.2720203404165462,
+      "learning_rate": 1.9865675056825455e-07,
+      "loss": 0.0075,
+      "step": 16258
+    },
+    {
+      "epoch": 3.695227272727273,
+      "grad_norm": 1.8760484703405071,
+      "learning_rate": 1.9859149430662728e-07,
+      "loss": 0.0102,
+      "step": 16259
+    },
+    {
+      "epoch": 3.6954545454545453,
+      "grad_norm": 2.1180174508068195,
+      "learning_rate": 1.9852624674021347e-07,
+      "loss": 0.0139,
+      "step": 16260
+    },
+    {
+      "epoch": 3.6956818181818183,
+      "grad_norm": 2.553910139029285,
+      "learning_rate": 1.9846100787034342e-07,
+      "loss": 0.0175,
+      "step": 16261
+    },
+    {
+      "epoch": 3.695909090909091,
+      "grad_norm": 2.8960966253081795,
+      "learning_rate": 1.9839577769834778e-07,
+      "loss": 0.0398,
+      "step": 16262
+    },
+    {
+      "epoch": 3.6961363636363638,
+      "grad_norm": 2.3232603631925324,
+      "learning_rate": 1.9833055622555628e-07,
+      "loss": 0.0114,
+      "step": 16263
+    },
+    {
+      "epoch": 3.6963636363636363,
+      "grad_norm": 5.138703811738994,
+      "learning_rate": 1.9826534345329911e-07,
+      "loss": 0.0284,
+      "step": 16264
+    },
+    {
+      "epoch": 3.696590909090909,
+      "grad_norm": 1.0524150890545068,
+      "learning_rate": 1.982001393829061e-07,
+      "loss": 0.0098,
+      "step": 16265
+    },
+    {
+      "epoch": 3.696818181818182,
+      "grad_norm": 0.5422492850654802,
+      "learning_rate": 1.9813494401570683e-07,
+      "loss": 0.0019,
+      "step": 16266
+    },
+    {
+      "epoch": 3.6970454545454547,
+      "grad_norm": 0.872697971501551,
+      "learning_rate": 1.9806975735303075e-07,
+      "loss": 0.0042,
+      "step": 16267
+    },
+    {
+      "epoch": 3.6972727272727273,
+      "grad_norm": 2.734201147492588,
+      "learning_rate": 1.9800457939620709e-07,
+      "loss": 0.0133,
+      "step": 16268
+    },
+    {
+      "epoch": 3.6975,
+      "grad_norm": 1.5797636189354987,
+      "learning_rate": 1.97939410146565e-07,
+      "loss": 0.0103,
+      "step": 16269
+    },
+    {
+      "epoch": 3.6977272727272728,
+      "grad_norm": 1.3808713378871744,
+      "learning_rate": 1.9787424960543347e-07,
+      "loss": 0.0063,
+      "step": 16270
+    },
+    {
+      "epoch": 3.6979545454545457,
+      "grad_norm": 4.844579837633733,
+      "learning_rate": 1.9780909777414126e-07,
+      "loss": 0.0335,
+      "step": 16271
+    },
+    {
+      "epoch": 3.6981818181818182,
+      "grad_norm": 2.0827521974627157,
+      "learning_rate": 1.977439546540167e-07,
+      "loss": 0.0051,
+      "step": 16272
+    },
+    {
+      "epoch": 3.6984090909090908,
+      "grad_norm": 2.5290190465566575,
+      "learning_rate": 1.9767882024638821e-07,
+      "loss": 0.0173,
+      "step": 16273
+    },
+    {
+      "epoch": 3.6986363636363637,
+      "grad_norm": 2.1157395017713085,
+      "learning_rate": 1.9761369455258416e-07,
+      "loss": 0.0309,
+      "step": 16274
+    },
+    {
+      "epoch": 3.6988636363636362,
+      "grad_norm": 0.9300809837203798,
+      "learning_rate": 1.975485775739326e-07,
+      "loss": 0.0109,
+      "step": 16275
+    },
+    {
+      "epoch": 3.699090909090909,
+      "grad_norm": 1.0625967627813824,
+      "learning_rate": 1.9748346931176117e-07,
+      "loss": 0.0061,
+      "step": 16276
+    },
+    {
+      "epoch": 3.6993181818181817,
+      "grad_norm": 2.400893363182968,
+      "learning_rate": 1.974183697673976e-07,
+      "loss": 0.0158,
+      "step": 16277
+    },
+    {
+      "epoch": 3.6995454545454547,
+      "grad_norm": 1.6148818524567448,
+      "learning_rate": 1.9735327894216942e-07,
+      "loss": 0.0077,
+      "step": 16278
+    },
+    {
+      "epoch": 3.699772727272727,
+      "grad_norm": 1.3411311637986816,
+      "learning_rate": 1.9728819683740394e-07,
+      "loss": 0.0104,
+      "step": 16279
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 5.298813389556742,
+      "learning_rate": 1.9722312345442835e-07,
+      "loss": 0.0289,
+      "step": 16280
+    },
+    {
+      "epoch": 3.7002272727272727,
+      "grad_norm": 1.7442611183245467,
+      "learning_rate": 1.9715805879456955e-07,
+      "loss": 0.0139,
+      "step": 16281
+    },
+    {
+      "epoch": 3.700454545454545,
+      "grad_norm": 2.652718032235274,
+      "learning_rate": 1.9709300285915437e-07,
+      "loss": 0.019,
+      "step": 16282
+    },
+    {
+      "epoch": 3.700681818181818,
+      "grad_norm": 2.211959591028043,
+      "learning_rate": 1.9702795564950937e-07,
+      "loss": 0.0162,
+      "step": 16283
+    },
+    {
+      "epoch": 3.700909090909091,
+      "grad_norm": 1.1265685026081578,
+      "learning_rate": 1.9696291716696118e-07,
+      "loss": 0.0123,
+      "step": 16284
+    },
+    {
+      "epoch": 3.7011363636363637,
+      "grad_norm": 2.261125685447353,
+      "learning_rate": 1.9689788741283563e-07,
+      "loss": 0.0204,
+      "step": 16285
+    },
+    {
+      "epoch": 3.701363636363636,
+      "grad_norm": 1.4390677982081028,
+      "learning_rate": 1.968328663884591e-07,
+      "loss": 0.013,
+      "step": 16286
+    },
+    {
+      "epoch": 3.701590909090909,
+      "grad_norm": 1.1858981425026545,
+      "learning_rate": 1.967678540951573e-07,
+      "loss": 0.007,
+      "step": 16287
+    },
+    {
+      "epoch": 3.7018181818181817,
+      "grad_norm": 0.6147174735450751,
+      "learning_rate": 1.9670285053425637e-07,
+      "loss": 0.0038,
+      "step": 16288
+    },
+    {
+      "epoch": 3.7020454545454546,
+      "grad_norm": 2.0144421437132722,
+      "learning_rate": 1.9663785570708123e-07,
+      "loss": 0.0126,
+      "step": 16289
+    },
+    {
+      "epoch": 3.702272727272727,
+      "grad_norm": 3.1074672884619194,
+      "learning_rate": 1.9657286961495756e-07,
+      "loss": 0.0239,
+      "step": 16290
+    },
+    {
+      "epoch": 3.7025,
+      "grad_norm": 0.29668823958978485,
+      "learning_rate": 1.9650789225921049e-07,
+      "loss": 0.001,
+      "step": 16291
+    },
+    {
+      "epoch": 3.7027272727272726,
+      "grad_norm": 0.9232190496297366,
+      "learning_rate": 1.964429236411651e-07,
+      "loss": 0.0052,
+      "step": 16292
+    },
+    {
+      "epoch": 3.7029545454545456,
+      "grad_norm": 0.903702931621856,
+      "learning_rate": 1.963779637621463e-07,
+      "loss": 0.0048,
+      "step": 16293
+    },
+    {
+      "epoch": 3.703181818181818,
+      "grad_norm": 0.7272664511488273,
+      "learning_rate": 1.9631301262347844e-07,
+      "loss": 0.004,
+      "step": 16294
+    },
+    {
+      "epoch": 3.7034090909090907,
+      "grad_norm": 1.0864552418680944,
+      "learning_rate": 1.962480702264861e-07,
+      "loss": 0.007,
+      "step": 16295
+    },
+    {
+      "epoch": 3.7036363636363636,
+      "grad_norm": 0.9104561232083861,
+      "learning_rate": 1.961831365724936e-07,
+      "loss": 0.0063,
+      "step": 16296
+    },
+    {
+      "epoch": 3.7038636363636366,
+      "grad_norm": 0.8750400859532774,
+      "learning_rate": 1.9611821166282505e-07,
+      "loss": 0.0039,
+      "step": 16297
+    },
+    {
+      "epoch": 3.704090909090909,
+      "grad_norm": 3.729845162782018,
+      "learning_rate": 1.960532954988044e-07,
+      "loss": 0.0273,
+      "step": 16298
+    },
+    {
+      "epoch": 3.7043181818181816,
+      "grad_norm": 1.0039104711679399,
+      "learning_rate": 1.9598838808175543e-07,
+      "loss": 0.0036,
+      "step": 16299
+    },
+    {
+      "epoch": 3.7045454545454546,
+      "grad_norm": 2.4442639710286236,
+      "learning_rate": 1.9592348941300167e-07,
+      "loss": 0.0194,
+      "step": 16300
+    },
+    {
+      "epoch": 3.704772727272727,
+      "grad_norm": 1.11412600656757,
+      "learning_rate": 1.958585994938665e-07,
+      "loss": 0.0063,
+      "step": 16301
+    },
+    {
+      "epoch": 3.705,
+      "grad_norm": 1.4294591363381726,
+      "learning_rate": 1.9579371832567334e-07,
+      "loss": 0.0083,
+      "step": 16302
+    },
+    {
+      "epoch": 3.7052272727272726,
+      "grad_norm": 1.2607475110968385,
+      "learning_rate": 1.9572884590974492e-07,
+      "loss": 0.0048,
+      "step": 16303
+    },
+    {
+      "epoch": 3.7054545454545456,
+      "grad_norm": 1.201460645482506,
+      "learning_rate": 1.956639822474042e-07,
+      "loss": 0.0074,
+      "step": 16304
+    },
+    {
+      "epoch": 3.705681818181818,
+      "grad_norm": 2.461517772980656,
+      "learning_rate": 1.9559912733997392e-07,
+      "loss": 0.0161,
+      "step": 16305
+    },
+    {
+      "epoch": 3.705909090909091,
+      "grad_norm": 0.8870825507371964,
+      "learning_rate": 1.9553428118877672e-07,
+      "loss": 0.0047,
+      "step": 16306
+    },
+    {
+      "epoch": 3.7061363636363636,
+      "grad_norm": 4.9047519545494085,
+      "learning_rate": 1.954694437951346e-07,
+      "loss": 0.0304,
+      "step": 16307
+    },
+    {
+      "epoch": 3.7063636363636365,
+      "grad_norm": 1.8415199735143117,
+      "learning_rate": 1.954046151603699e-07,
+      "loss": 0.0106,
+      "step": 16308
+    },
+    {
+      "epoch": 3.706590909090909,
+      "grad_norm": 0.779357659289422,
+      "learning_rate": 1.953397952858045e-07,
+      "loss": 0.0115,
+      "step": 16309
+    },
+    {
+      "epoch": 3.706818181818182,
+      "grad_norm": 1.4864169151457547,
+      "learning_rate": 1.952749841727603e-07,
+      "loss": 0.0143,
+      "step": 16310
+    },
+    {
+      "epoch": 3.7070454545454545,
+      "grad_norm": 1.2284208400659833,
+      "learning_rate": 1.9521018182255894e-07,
+      "loss": 0.0062,
+      "step": 16311
+    },
+    {
+      "epoch": 3.707272727272727,
+      "grad_norm": 1.3486714360343388,
+      "learning_rate": 1.9514538823652166e-07,
+      "loss": 0.0092,
+      "step": 16312
+    },
+    {
+      "epoch": 3.7075,
+      "grad_norm": 2.8177572857653836,
+      "learning_rate": 1.950806034159699e-07,
+      "loss": 0.0313,
+      "step": 16313
+    },
+    {
+      "epoch": 3.707727272727273,
+      "grad_norm": 2.5157954068533357,
+      "learning_rate": 1.950158273622247e-07,
+      "loss": 0.0207,
+      "step": 16314
+    },
+    {
+      "epoch": 3.7079545454545455,
+      "grad_norm": 1.5886095945320746,
+      "learning_rate": 1.9495106007660705e-07,
+      "loss": 0.0158,
+      "step": 16315
+    },
+    {
+      "epoch": 3.708181818181818,
+      "grad_norm": 2.127669844609934,
+      "learning_rate": 1.948863015604374e-07,
+      "loss": 0.0151,
+      "step": 16316
+    },
+    {
+      "epoch": 3.708409090909091,
+      "grad_norm": 1.1194548973682483,
+      "learning_rate": 1.9482155181503642e-07,
+      "loss": 0.0124,
+      "step": 16317
+    },
+    {
+      "epoch": 3.7086363636363635,
+      "grad_norm": 0.5222208082329594,
+      "learning_rate": 1.947568108417245e-07,
+      "loss": 0.0034,
+      "step": 16318
+    },
+    {
+      "epoch": 3.7088636363636365,
+      "grad_norm": 1.1670642087979846,
+      "learning_rate": 1.94692078641822e-07,
+      "loss": 0.0061,
+      "step": 16319
+    },
+    {
+      "epoch": 3.709090909090909,
+      "grad_norm": 2.3944061488727013,
+      "learning_rate": 1.9462735521664853e-07,
+      "loss": 0.0203,
+      "step": 16320
+    },
+    {
+      "epoch": 3.709318181818182,
+      "grad_norm": 1.6312992194897213,
+      "learning_rate": 1.945626405675241e-07,
+      "loss": 0.0122,
+      "step": 16321
+    },
+    {
+      "epoch": 3.7095454545454545,
+      "grad_norm": 1.4925873663556268,
+      "learning_rate": 1.9449793469576837e-07,
+      "loss": 0.0165,
+      "step": 16322
+    },
+    {
+      "epoch": 3.7097727272727274,
+      "grad_norm": 1.8115273696655037,
+      "learning_rate": 1.9443323760270083e-07,
+      "loss": 0.0196,
+      "step": 16323
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 1.1874708114184058,
+      "learning_rate": 1.9436854928964092e-07,
+      "loss": 0.0098,
+      "step": 16324
+    },
+    {
+      "epoch": 3.7102272727272725,
+      "grad_norm": 1.172844157484334,
+      "learning_rate": 1.9430386975790736e-07,
+      "loss": 0.004,
+      "step": 16325
+    },
+    {
+      "epoch": 3.7104545454545454,
+      "grad_norm": 0.8166114206927868,
+      "learning_rate": 1.942391990088193e-07,
+      "loss": 0.0075,
+      "step": 16326
+    },
+    {
+      "epoch": 3.7106818181818184,
+      "grad_norm": 2.484724338539014,
+      "learning_rate": 1.9417453704369546e-07,
+      "loss": 0.0235,
+      "step": 16327
+    },
+    {
+      "epoch": 3.710909090909091,
+      "grad_norm": 1.0760406723885552,
+      "learning_rate": 1.9410988386385442e-07,
+      "loss": 0.0058,
+      "step": 16328
+    },
+    {
+      "epoch": 3.7111363636363635,
+      "grad_norm": 1.9020336774652793,
+      "learning_rate": 1.9404523947061458e-07,
+      "loss": 0.0137,
+      "step": 16329
+    },
+    {
+      "epoch": 3.7113636363636364,
+      "grad_norm": 1.415263771473516,
+      "learning_rate": 1.9398060386529422e-07,
+      "loss": 0.0094,
+      "step": 16330
+    },
+    {
+      "epoch": 3.711590909090909,
+      "grad_norm": 2.816192290499309,
+      "learning_rate": 1.939159770492112e-07,
+      "loss": 0.0085,
+      "step": 16331
+    },
+    {
+      "epoch": 3.711818181818182,
+      "grad_norm": 0.42176118793452666,
+      "learning_rate": 1.9385135902368374e-07,
+      "loss": 0.0016,
+      "step": 16332
+    },
+    {
+      "epoch": 3.7120454545454544,
+      "grad_norm": 0.3738519035501332,
+      "learning_rate": 1.9378674979002906e-07,
+      "loss": 0.002,
+      "step": 16333
+    },
+    {
+      "epoch": 3.7122727272727274,
+      "grad_norm": 1.8251674040586503,
+      "learning_rate": 1.937221493495648e-07,
+      "loss": 0.0091,
+      "step": 16334
+    },
+    {
+      "epoch": 3.7125,
+      "grad_norm": 2.367328417812553,
+      "learning_rate": 1.936575577036083e-07,
+      "loss": 0.0242,
+      "step": 16335
+    },
+    {
+      "epoch": 3.712727272727273,
+      "grad_norm": 1.621597383858875,
+      "learning_rate": 1.9359297485347677e-07,
+      "loss": 0.0064,
+      "step": 16336
+    },
+    {
+      "epoch": 3.7129545454545454,
+      "grad_norm": 3.502093441014752,
+      "learning_rate": 1.935284008004873e-07,
+      "loss": 0.0226,
+      "step": 16337
+    },
+    {
+      "epoch": 3.713181818181818,
+      "grad_norm": 0.3450031578445032,
+      "learning_rate": 1.934638355459563e-07,
+      "loss": 0.0017,
+      "step": 16338
+    },
+    {
+      "epoch": 3.713409090909091,
+      "grad_norm": 1.97274407026737,
+      "learning_rate": 1.9339927909120052e-07,
+      "loss": 0.0124,
+      "step": 16339
+    },
+    {
+      "epoch": 3.713636363636364,
+      "grad_norm": 3.9064847565362872,
+      "learning_rate": 1.9333473143753638e-07,
+      "loss": 0.0082,
+      "step": 16340
+    },
+    {
+      "epoch": 3.7138636363636364,
+      "grad_norm": 0.8197637134737783,
+      "learning_rate": 1.9327019258628017e-07,
+      "loss": 0.0053,
+      "step": 16341
+    },
+    {
+      "epoch": 3.714090909090909,
+      "grad_norm": 1.7220028706419457,
+      "learning_rate": 1.9320566253874796e-07,
+      "loss": 0.0136,
+      "step": 16342
+    },
+    {
+      "epoch": 3.714318181818182,
+      "grad_norm": 1.3978176531098312,
+      "learning_rate": 1.9314114129625555e-07,
+      "loss": 0.003,
+      "step": 16343
+    },
+    {
+      "epoch": 3.714545454545455,
+      "grad_norm": 2.255357035362982,
+      "learning_rate": 1.9307662886011874e-07,
+      "loss": 0.0115,
+      "step": 16344
+    },
+    {
+      "epoch": 3.7147727272727273,
+      "grad_norm": 2.860112934011573,
+      "learning_rate": 1.9301212523165296e-07,
+      "loss": 0.0182,
+      "step": 16345
+    },
+    {
+      "epoch": 3.715,
+      "grad_norm": 1.9835555916979832,
+      "learning_rate": 1.9294763041217385e-07,
+      "loss": 0.0136,
+      "step": 16346
+    },
+    {
+      "epoch": 3.715227272727273,
+      "grad_norm": 0.9089508961258654,
+      "learning_rate": 1.9288314440299606e-07,
+      "loss": 0.0075,
+      "step": 16347
+    },
+    {
+      "epoch": 3.7154545454545453,
+      "grad_norm": 0.9383123438425012,
+      "learning_rate": 1.9281866720543483e-07,
+      "loss": 0.0048,
+      "step": 16348
+    },
+    {
+      "epoch": 3.7156818181818183,
+      "grad_norm": 0.6439260005911761,
+      "learning_rate": 1.9275419882080495e-07,
+      "loss": 0.004,
+      "step": 16349
+    },
+    {
+      "epoch": 3.715909090909091,
+      "grad_norm": 0.9496094739651834,
+      "learning_rate": 1.9268973925042123e-07,
+      "loss": 0.008,
+      "step": 16350
+    },
+    {
+      "epoch": 3.716136363636364,
+      "grad_norm": 0.8012808798620645,
+      "learning_rate": 1.926252884955978e-07,
+      "loss": 0.0046,
+      "step": 16351
+    },
+    {
+      "epoch": 3.7163636363636363,
+      "grad_norm": 1.739801182671284,
+      "learning_rate": 1.92560846557649e-07,
+      "loss": 0.0185,
+      "step": 16352
+    },
+    {
+      "epoch": 3.7165909090909093,
+      "grad_norm": 1.987337491624338,
+      "learning_rate": 1.92496413437889e-07,
+      "loss": 0.0146,
+      "step": 16353
+    },
+    {
+      "epoch": 3.716818181818182,
+      "grad_norm": 2.1442919183883355,
+      "learning_rate": 1.9243198913763187e-07,
+      "loss": 0.0163,
+      "step": 16354
+    },
+    {
+      "epoch": 3.7170454545454543,
+      "grad_norm": 0.930101962966628,
+      "learning_rate": 1.9236757365819093e-07,
+      "loss": 0.0061,
+      "step": 16355
+    },
+    {
+      "epoch": 3.7172727272727273,
+      "grad_norm": 0.7220321382772326,
+      "learning_rate": 1.9230316700087993e-07,
+      "loss": 0.0037,
+      "step": 16356
+    },
+    {
+      "epoch": 3.7175000000000002,
+      "grad_norm": 1.2440274967898581,
+      "learning_rate": 1.9223876916701225e-07,
+      "loss": 0.0129,
+      "step": 16357
+    },
+    {
+      "epoch": 3.7177272727272728,
+      "grad_norm": 0.9378218113810937,
+      "learning_rate": 1.9217438015790104e-07,
+      "loss": 0.0062,
+      "step": 16358
+    },
+    {
+      "epoch": 3.7179545454545453,
+      "grad_norm": 0.9314460820782963,
+      "learning_rate": 1.9210999997485929e-07,
+      "loss": 0.0035,
+      "step": 16359
+    },
+    {
+      "epoch": 3.7181818181818183,
+      "grad_norm": 1.9396458595507402,
+      "learning_rate": 1.920456286191999e-07,
+      "loss": 0.0243,
+      "step": 16360
+    },
+    {
+      "epoch": 3.7184090909090908,
+      "grad_norm": 1.5368751132814473,
+      "learning_rate": 1.919812660922355e-07,
+      "loss": 0.0181,
+      "step": 16361
+    },
+    {
+      "epoch": 3.7186363636363637,
+      "grad_norm": 0.7322046000916724,
+      "learning_rate": 1.9191691239527847e-07,
+      "loss": 0.0056,
+      "step": 16362
+    },
+    {
+      "epoch": 3.7188636363636363,
+      "grad_norm": 1.348448032944405,
+      "learning_rate": 1.9185256752964143e-07,
+      "loss": 0.0062,
+      "step": 16363
+    },
+    {
+      "epoch": 3.7190909090909092,
+      "grad_norm": 1.8597579223667426,
+      "learning_rate": 1.9178823149663595e-07,
+      "loss": 0.0164,
+      "step": 16364
+    },
+    {
+      "epoch": 3.7193181818181817,
+      "grad_norm": 2.3911375949867684,
+      "learning_rate": 1.9172390429757416e-07,
+      "loss": 0.0074,
+      "step": 16365
+    },
+    {
+      "epoch": 3.7195454545454547,
+      "grad_norm": 0.589987203837171,
+      "learning_rate": 1.9165958593376797e-07,
+      "loss": 0.01,
+      "step": 16366
+    },
+    {
+      "epoch": 3.7197727272727272,
+      "grad_norm": 2.3556194206963745,
+      "learning_rate": 1.9159527640652876e-07,
+      "loss": 0.0195,
+      "step": 16367
+    },
+    {
+      "epoch": 3.7199999999999998,
+      "grad_norm": 1.0538670745018415,
+      "learning_rate": 1.9153097571716814e-07,
+      "loss": 0.0077,
+      "step": 16368
+    },
+    {
+      "epoch": 3.7202272727272727,
+      "grad_norm": 1.0944478123100425,
+      "learning_rate": 1.91466683866997e-07,
+      "loss": 0.0047,
+      "step": 16369
+    },
+    {
+      "epoch": 3.7204545454545457,
+      "grad_norm": 2.801930095619496,
+      "learning_rate": 1.9140240085732655e-07,
+      "loss": 0.0183,
+      "step": 16370
+    },
+    {
+      "epoch": 3.720681818181818,
+      "grad_norm": 1.4456423483691974,
+      "learning_rate": 1.9133812668946753e-07,
+      "loss": 0.0075,
+      "step": 16371
+    },
+    {
+      "epoch": 3.7209090909090907,
+      "grad_norm": 3.267740411070446,
+      "learning_rate": 1.9127386136473074e-07,
+      "loss": 0.0212,
+      "step": 16372
+    },
+    {
+      "epoch": 3.7211363636363637,
+      "grad_norm": 3.4609001738986254,
+      "learning_rate": 1.9120960488442656e-07,
+      "loss": 0.0411,
+      "step": 16373
+    },
+    {
+      "epoch": 3.721363636363636,
+      "grad_norm": 1.043208023317867,
+      "learning_rate": 1.9114535724986533e-07,
+      "loss": 0.0036,
+      "step": 16374
+    },
+    {
+      "epoch": 3.721590909090909,
+      "grad_norm": 1.2637381025115213,
+      "learning_rate": 1.910811184623571e-07,
+      "loss": 0.0035,
+      "step": 16375
+    },
+    {
+      "epoch": 3.7218181818181817,
+      "grad_norm": 0.8215304986887042,
+      "learning_rate": 1.910168885232122e-07,
+      "loss": 0.0065,
+      "step": 16376
+    },
+    {
+      "epoch": 3.7220454545454547,
+      "grad_norm": 1.819423106399886,
+      "learning_rate": 1.909526674337398e-07,
+      "loss": 0.015,
+      "step": 16377
+    },
+    {
+      "epoch": 3.722272727272727,
+      "grad_norm": 3.6712589992692326,
+      "learning_rate": 1.9088845519524974e-07,
+      "loss": 0.0118,
+      "step": 16378
+    },
+    {
+      "epoch": 3.7225,
+      "grad_norm": 0.41556179603241306,
+      "learning_rate": 1.9082425180905146e-07,
+      "loss": 0.0026,
+      "step": 16379
+    },
+    {
+      "epoch": 3.7227272727272727,
+      "grad_norm": 1.085559822503538,
+      "learning_rate": 1.907600572764541e-07,
+      "loss": 0.0121,
+      "step": 16380
+    },
+    {
+      "epoch": 3.722954545454545,
+      "grad_norm": 1.151444871778071,
+      "learning_rate": 1.9069587159876702e-07,
+      "loss": 0.0067,
+      "step": 16381
+    },
+    {
+      "epoch": 3.723181818181818,
+      "grad_norm": 1.5474878595702162,
+      "learning_rate": 1.9063169477729856e-07,
+      "loss": 0.0101,
+      "step": 16382
+    },
+    {
+      "epoch": 3.723409090909091,
+      "grad_norm": 2.0925790476078094,
+      "learning_rate": 1.9056752681335763e-07,
+      "loss": 0.0076,
+      "step": 16383
+    },
+    {
+      "epoch": 3.7236363636363636,
+      "grad_norm": 2.7649375153896742,
+      "learning_rate": 1.9050336770825274e-07,
+      "loss": 0.0119,
+      "step": 16384
+    },
+    {
+      "epoch": 3.723863636363636,
+      "grad_norm": 2.0415082532248023,
+      "learning_rate": 1.9043921746329242e-07,
+      "loss": 0.01,
+      "step": 16385
+    },
+    {
+      "epoch": 3.724090909090909,
+      "grad_norm": 2.159825083487805,
+      "learning_rate": 1.9037507607978436e-07,
+      "loss": 0.033,
+      "step": 16386
+    },
+    {
+      "epoch": 3.724318181818182,
+      "grad_norm": 1.5938815156816553,
+      "learning_rate": 1.9031094355903674e-07,
+      "loss": 0.0104,
+      "step": 16387
+    },
+    {
+      "epoch": 3.7245454545454546,
+      "grad_norm": 1.1182092337801666,
+      "learning_rate": 1.9024681990235717e-07,
+      "loss": 0.007,
+      "step": 16388
+    },
+    {
+      "epoch": 3.724772727272727,
+      "grad_norm": 0.2900922587912958,
+      "learning_rate": 1.901827051110538e-07,
+      "loss": 0.0013,
+      "step": 16389
+    },
+    {
+      "epoch": 3.725,
+      "grad_norm": 1.1492741112268805,
+      "learning_rate": 1.9011859918643342e-07,
+      "loss": 0.01,
+      "step": 16390
+    },
+    {
+      "epoch": 3.7252272727272726,
+      "grad_norm": 2.0439456193309686,
+      "learning_rate": 1.9005450212980357e-07,
+      "loss": 0.0146,
+      "step": 16391
+    },
+    {
+      "epoch": 3.7254545454545456,
+      "grad_norm": 1.5942177528055153,
+      "learning_rate": 1.8999041394247118e-07,
+      "loss": 0.0128,
+      "step": 16392
+    },
+    {
+      "epoch": 3.725681818181818,
+      "grad_norm": 0.6957731032109943,
+      "learning_rate": 1.8992633462574315e-07,
+      "loss": 0.004,
+      "step": 16393
+    },
+    {
+      "epoch": 3.725909090909091,
+      "grad_norm": 3.5564393488463844,
+      "learning_rate": 1.898622641809264e-07,
+      "loss": 0.0479,
+      "step": 16394
+    },
+    {
+      "epoch": 3.7261363636363636,
+      "grad_norm": 1.1798036846606421,
+      "learning_rate": 1.8979820260932703e-07,
+      "loss": 0.0058,
+      "step": 16395
+    },
+    {
+      "epoch": 3.7263636363636365,
+      "grad_norm": 1.5596362264731056,
+      "learning_rate": 1.897341499122516e-07,
+      "loss": 0.0082,
+      "step": 16396
+    },
+    {
+      "epoch": 3.726590909090909,
+      "grad_norm": 1.2424443138320456,
+      "learning_rate": 1.896701060910061e-07,
+      "loss": 0.0054,
+      "step": 16397
+    },
+    {
+      "epoch": 3.7268181818181816,
+      "grad_norm": 0.977033582603996,
+      "learning_rate": 1.8960607114689693e-07,
+      "loss": 0.0063,
+      "step": 16398
+    },
+    {
+      "epoch": 3.7270454545454546,
+      "grad_norm": 1.8086333738275715,
+      "learning_rate": 1.895420450812293e-07,
+      "loss": 0.0158,
+      "step": 16399
+    },
+    {
+      "epoch": 3.7272727272727275,
+      "grad_norm": 2.442020403615476,
+      "learning_rate": 1.8947802789530908e-07,
+      "loss": 0.0206,
+      "step": 16400
+    },
+    {
+      "epoch": 3.7275,
+      "grad_norm": 2.924106412628365,
+      "learning_rate": 1.8941401959044167e-07,
+      "loss": 0.0189,
+      "step": 16401
+    },
+    {
+      "epoch": 3.7277272727272726,
+      "grad_norm": 1.3176307832598486,
+      "learning_rate": 1.8935002016793233e-07,
+      "loss": 0.0077,
+      "step": 16402
+    },
+    {
+      "epoch": 3.7279545454545455,
+      "grad_norm": 1.439608942364547,
+      "learning_rate": 1.8928602962908612e-07,
+      "loss": 0.0117,
+      "step": 16403
+    },
+    {
+      "epoch": 3.728181818181818,
+      "grad_norm": 5.341126253055345,
+      "learning_rate": 1.8922204797520792e-07,
+      "loss": 0.0122,
+      "step": 16404
+    },
+    {
+      "epoch": 3.728409090909091,
+      "grad_norm": 1.6533520525750356,
+      "learning_rate": 1.8915807520760241e-07,
+      "loss": 0.0116,
+      "step": 16405
+    },
+    {
+      "epoch": 3.7286363636363635,
+      "grad_norm": 3.7122703745077827,
+      "learning_rate": 1.8909411132757413e-07,
+      "loss": 0.0109,
+      "step": 16406
+    },
+    {
+      "epoch": 3.7288636363636365,
+      "grad_norm": 1.8135728522677335,
+      "learning_rate": 1.8903015633642754e-07,
+      "loss": 0.0158,
+      "step": 16407
+    },
+    {
+      "epoch": 3.729090909090909,
+      "grad_norm": 1.5125473280576238,
+      "learning_rate": 1.8896621023546647e-07,
+      "loss": 0.0062,
+      "step": 16408
+    },
+    {
+      "epoch": 3.729318181818182,
+      "grad_norm": 1.0788664274661408,
+      "learning_rate": 1.8890227302599511e-07,
+      "loss": 0.0065,
+      "step": 16409
+    },
+    {
+      "epoch": 3.7295454545454545,
+      "grad_norm": 1.8261408719409766,
+      "learning_rate": 1.888383447093172e-07,
+      "loss": 0.0188,
+      "step": 16410
+    },
+    {
+      "epoch": 3.729772727272727,
+      "grad_norm": 1.7840168867621575,
+      "learning_rate": 1.8877442528673656e-07,
+      "loss": 0.007,
+      "step": 16411
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 1.920867280646616,
+      "learning_rate": 1.8871051475955624e-07,
+      "loss": 0.0093,
+      "step": 16412
+    },
+    {
+      "epoch": 3.730227272727273,
+      "grad_norm": 1.7087786430835141,
+      "learning_rate": 1.886466131290797e-07,
+      "loss": 0.0093,
+      "step": 16413
+    },
+    {
+      "epoch": 3.7304545454545455,
+      "grad_norm": 1.183208245077349,
+      "learning_rate": 1.8858272039660995e-07,
+      "loss": 0.0072,
+      "step": 16414
+    },
+    {
+      "epoch": 3.730681818181818,
+      "grad_norm": 2.073871356551547,
+      "learning_rate": 1.8851883656344987e-07,
+      "loss": 0.0186,
+      "step": 16415
+    },
+    {
+      "epoch": 3.730909090909091,
+      "grad_norm": 0.6570591731206863,
+      "learning_rate": 1.884549616309024e-07,
+      "loss": 0.0042,
+      "step": 16416
+    },
+    {
+      "epoch": 3.7311363636363635,
+      "grad_norm": 6.4145010383289405,
+      "learning_rate": 1.883910956002697e-07,
+      "loss": 0.0434,
+      "step": 16417
+    },
+    {
+      "epoch": 3.7313636363636364,
+      "grad_norm": 3.1403748880485884,
+      "learning_rate": 1.8832723847285406e-07,
+      "loss": 0.0284,
+      "step": 16418
+    },
+    {
+      "epoch": 3.731590909090909,
+      "grad_norm": 2.6050015494116328,
+      "learning_rate": 1.8826339024995803e-07,
+      "loss": 0.0218,
+      "step": 16419
+    },
+    {
+      "epoch": 3.731818181818182,
+      "grad_norm": 2.7078137955191517,
+      "learning_rate": 1.8819955093288364e-07,
+      "loss": 0.0187,
+      "step": 16420
+    },
+    {
+      "epoch": 3.7320454545454544,
+      "grad_norm": 1.0582541193304165,
+      "learning_rate": 1.881357205229322e-07,
+      "loss": 0.0125,
+      "step": 16421
+    },
+    {
+      "epoch": 3.7322727272727274,
+      "grad_norm": 1.8721464192601405,
+      "learning_rate": 1.880718990214056e-07,
+      "loss": 0.0177,
+      "step": 16422
+    },
+    {
+      "epoch": 3.7325,
+      "grad_norm": 2.807996220940123,
+      "learning_rate": 1.8800808642960523e-07,
+      "loss": 0.0233,
+      "step": 16423
+    },
+    {
+      "epoch": 3.732727272727273,
+      "grad_norm": 1.3829903282106328,
+      "learning_rate": 1.879442827488323e-07,
+      "loss": 0.0064,
+      "step": 16424
+    },
+    {
+      "epoch": 3.7329545454545454,
+      "grad_norm": 0.9228991598522555,
+      "learning_rate": 1.8788048798038818e-07,
+      "loss": 0.0058,
+      "step": 16425
+    },
+    {
+      "epoch": 3.7331818181818184,
+      "grad_norm": 1.0484043465339168,
+      "learning_rate": 1.8781670212557327e-07,
+      "loss": 0.0093,
+      "step": 16426
+    },
+    {
+      "epoch": 3.733409090909091,
+      "grad_norm": 1.227885322474756,
+      "learning_rate": 1.8775292518568854e-07,
+      "loss": 0.0079,
+      "step": 16427
+    },
+    {
+      "epoch": 3.7336363636363634,
+      "grad_norm": 1.859118774853044,
+      "learning_rate": 1.8768915716203457e-07,
+      "loss": 0.0081,
+      "step": 16428
+    },
+    {
+      "epoch": 3.7338636363636364,
+      "grad_norm": 4.062330335758832,
+      "learning_rate": 1.8762539805591167e-07,
+      "loss": 0.0363,
+      "step": 16429
+    },
+    {
+      "epoch": 3.7340909090909093,
+      "grad_norm": 0.9035382599515397,
+      "learning_rate": 1.8756164786861986e-07,
+      "loss": 0.0064,
+      "step": 16430
+    },
+    {
+      "epoch": 3.734318181818182,
+      "grad_norm": 1.4530351466682072,
+      "learning_rate": 1.8749790660145915e-07,
+      "loss": 0.0089,
+      "step": 16431
+    },
+    {
+      "epoch": 3.7345454545454544,
+      "grad_norm": 2.913878955171933,
+      "learning_rate": 1.874341742557294e-07,
+      "loss": 0.0289,
+      "step": 16432
+    },
+    {
+      "epoch": 3.7347727272727274,
+      "grad_norm": 1.956816299882883,
+      "learning_rate": 1.8737045083273017e-07,
+      "loss": 0.0178,
+      "step": 16433
+    },
+    {
+      "epoch": 3.735,
+      "grad_norm": 1.6181479820648503,
+      "learning_rate": 1.87306736333761e-07,
+      "loss": 0.0116,
+      "step": 16434
+    },
+    {
+      "epoch": 3.735227272727273,
+      "grad_norm": 2.54412660196773,
+      "learning_rate": 1.8724303076012105e-07,
+      "loss": 0.0217,
+      "step": 16435
+    },
+    {
+      "epoch": 3.7354545454545454,
+      "grad_norm": 2.936312777319073,
+      "learning_rate": 1.8717933411310937e-07,
+      "loss": 0.024,
+      "step": 16436
+    },
+    {
+      "epoch": 3.7356818181818183,
+      "grad_norm": 0.47286788731558055,
+      "learning_rate": 1.871156463940249e-07,
+      "loss": 0.0029,
+      "step": 16437
+    },
+    {
+      "epoch": 3.735909090909091,
+      "grad_norm": 2.5975044573924353,
+      "learning_rate": 1.8705196760416662e-07,
+      "loss": 0.0152,
+      "step": 16438
+    },
+    {
+      "epoch": 3.736136363636364,
+      "grad_norm": 1.355036622645436,
+      "learning_rate": 1.8698829774483244e-07,
+      "loss": 0.01,
+      "step": 16439
+    },
+    {
+      "epoch": 3.7363636363636363,
+      "grad_norm": 1.9327885162951552,
+      "learning_rate": 1.869246368173211e-07,
+      "loss": 0.0105,
+      "step": 16440
+    },
+    {
+      "epoch": 3.736590909090909,
+      "grad_norm": 5.676718521335026,
+      "learning_rate": 1.8686098482293067e-07,
+      "loss": 0.0247,
+      "step": 16441
+    },
+    {
+      "epoch": 3.736818181818182,
+      "grad_norm": 1.5877078472068455,
+      "learning_rate": 1.8679734176295934e-07,
+      "loss": 0.0135,
+      "step": 16442
+    },
+    {
+      "epoch": 3.737045454545455,
+      "grad_norm": 0.7445725042817858,
+      "learning_rate": 1.8673370763870443e-07,
+      "loss": 0.004,
+      "step": 16443
+    },
+    {
+      "epoch": 3.7372727272727273,
+      "grad_norm": 1.4194077045339342,
+      "learning_rate": 1.866700824514639e-07,
+      "loss": 0.0052,
+      "step": 16444
+    },
+    {
+      "epoch": 3.7375,
+      "grad_norm": 23.02489488501798,
+      "learning_rate": 1.8660646620253505e-07,
+      "loss": 0.0092,
+      "step": 16445
+    },
+    {
+      "epoch": 3.737727272727273,
+      "grad_norm": 4.3608842574178865,
+      "learning_rate": 1.8654285889321521e-07,
+      "loss": 0.0328,
+      "step": 16446
+    },
+    {
+      "epoch": 3.7379545454545453,
+      "grad_norm": 1.3608090766019887,
+      "learning_rate": 1.8647926052480134e-07,
+      "loss": 0.004,
+      "step": 16447
+    },
+    {
+      "epoch": 3.7381818181818183,
+      "grad_norm": 1.2961382327972442,
+      "learning_rate": 1.8641567109859052e-07,
+      "loss": 0.0059,
+      "step": 16448
+    },
+    {
+      "epoch": 3.738409090909091,
+      "grad_norm": 1.7364994482192835,
+      "learning_rate": 1.8635209061587917e-07,
+      "loss": 0.0145,
+      "step": 16449
+    },
+    {
+      "epoch": 3.7386363636363638,
+      "grad_norm": 0.7718594918179218,
+      "learning_rate": 1.8628851907796405e-07,
+      "loss": 0.0033,
+      "step": 16450
+    },
+    {
+      "epoch": 3.7388636363636363,
+      "grad_norm": 1.2153461885055699,
+      "learning_rate": 1.862249564861416e-07,
+      "loss": 0.0058,
+      "step": 16451
+    },
+    {
+      "epoch": 3.7390909090909092,
+      "grad_norm": 1.4533069584333813,
+      "learning_rate": 1.8616140284170754e-07,
+      "loss": 0.0095,
+      "step": 16452
+    },
+    {
+      "epoch": 3.7393181818181818,
+      "grad_norm": 1.2633085334303256,
+      "learning_rate": 1.8609785814595806e-07,
+      "loss": 0.0077,
+      "step": 16453
+    },
+    {
+      "epoch": 3.7395454545454543,
+      "grad_norm": 2.434038396153352,
+      "learning_rate": 1.86034322400189e-07,
+      "loss": 0.0302,
+      "step": 16454
+    },
+    {
+      "epoch": 3.7397727272727272,
+      "grad_norm": 2.9414775262071338,
+      "learning_rate": 1.859707956056961e-07,
+      "loss": 0.0205,
+      "step": 16455
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 3.7643714760997247,
+      "learning_rate": 1.8590727776377437e-07,
+      "loss": 0.018,
+      "step": 16456
+    },
+    {
+      "epoch": 3.7402272727272727,
+      "grad_norm": 0.5976960221803377,
+      "learning_rate": 1.8584376887571934e-07,
+      "loss": 0.004,
+      "step": 16457
+    },
+    {
+      "epoch": 3.7404545454545453,
+      "grad_norm": 1.5564265001363207,
+      "learning_rate": 1.8578026894282597e-07,
+      "loss": 0.0098,
+      "step": 16458
+    },
+    {
+      "epoch": 3.740681818181818,
+      "grad_norm": 1.0690501069825724,
+      "learning_rate": 1.8571677796638924e-07,
+      "loss": 0.0064,
+      "step": 16459
+    },
+    {
+      "epoch": 3.740909090909091,
+      "grad_norm": 1.4801767077941672,
+      "learning_rate": 1.8565329594770394e-07,
+      "loss": 0.0071,
+      "step": 16460
+    },
+    {
+      "epoch": 3.7411363636363637,
+      "grad_norm": 1.101160551402222,
+      "learning_rate": 1.855898228880642e-07,
+      "loss": 0.0123,
+      "step": 16461
+    },
+    {
+      "epoch": 3.7413636363636362,
+      "grad_norm": 3.3286803335887782,
+      "learning_rate": 1.8552635878876456e-07,
+      "loss": 0.0241,
+      "step": 16462
+    },
+    {
+      "epoch": 3.741590909090909,
+      "grad_norm": 0.7826086755062408,
+      "learning_rate": 1.8546290365109923e-07,
+      "loss": 0.0039,
+      "step": 16463
+    },
+    {
+      "epoch": 3.7418181818181817,
+      "grad_norm": 2.4790101465594963,
+      "learning_rate": 1.8539945747636206e-07,
+      "loss": 0.0192,
+      "step": 16464
+    },
+    {
+      "epoch": 3.7420454545454547,
+      "grad_norm": 1.1290599055911565,
+      "learning_rate": 1.853360202658469e-07,
+      "loss": 0.0053,
+      "step": 16465
+    },
+    {
+      "epoch": 3.742272727272727,
+      "grad_norm": 2.607801460711981,
+      "learning_rate": 1.852725920208473e-07,
+      "loss": 0.0308,
+      "step": 16466
+    },
+    {
+      "epoch": 3.7425,
+      "grad_norm": 3.352318102576281,
+      "learning_rate": 1.8520917274265672e-07,
+      "loss": 0.0393,
+      "step": 16467
+    },
+    {
+      "epoch": 3.7427272727272727,
+      "grad_norm": 0.46623649842626413,
+      "learning_rate": 1.851457624325683e-07,
+      "loss": 0.0054,
+      "step": 16468
+    },
+    {
+      "epoch": 3.7429545454545456,
+      "grad_norm": 1.1598275721428546,
+      "learning_rate": 1.850823610918754e-07,
+      "loss": 0.009,
+      "step": 16469
+    },
+    {
+      "epoch": 3.743181818181818,
+      "grad_norm": 2.034928828086659,
+      "learning_rate": 1.8501896872187042e-07,
+      "loss": 0.012,
+      "step": 16470
+    },
+    {
+      "epoch": 3.7434090909090907,
+      "grad_norm": 2.5701716528241607,
+      "learning_rate": 1.8495558532384633e-07,
+      "loss": 0.0187,
+      "step": 16471
+    },
+    {
+      "epoch": 3.7436363636363637,
+      "grad_norm": 0.8874561963765548,
+      "learning_rate": 1.8489221089909546e-07,
+      "loss": 0.0039,
+      "step": 16472
+    },
+    {
+      "epoch": 3.7438636363636366,
+      "grad_norm": 1.351785676129951,
+      "learning_rate": 1.8482884544891048e-07,
+      "loss": 0.0116,
+      "step": 16473
+    },
+    {
+      "epoch": 3.744090909090909,
+      "grad_norm": 1.8853871528680217,
+      "learning_rate": 1.8476548897458306e-07,
+      "loss": 0.0149,
+      "step": 16474
+    },
+    {
+      "epoch": 3.7443181818181817,
+      "grad_norm": 4.163349232975903,
+      "learning_rate": 1.8470214147740528e-07,
+      "loss": 0.007,
+      "step": 16475
+    },
+    {
+      "epoch": 3.7445454545454546,
+      "grad_norm": 0.7279536404882234,
+      "learning_rate": 1.8463880295866904e-07,
+      "loss": 0.0028,
+      "step": 16476
+    },
+    {
+      "epoch": 3.744772727272727,
+      "grad_norm": 2.1026113210839523,
+      "learning_rate": 1.8457547341966585e-07,
+      "loss": 0.0116,
+      "step": 16477
+    },
+    {
+      "epoch": 3.745,
+      "grad_norm": 1.9104815114720117,
+      "learning_rate": 1.8451215286168715e-07,
+      "loss": 0.0137,
+      "step": 16478
+    },
+    {
+      "epoch": 3.7452272727272726,
+      "grad_norm": 1.4836361744315631,
+      "learning_rate": 1.844488412860241e-07,
+      "loss": 0.0117,
+      "step": 16479
+    },
+    {
+      "epoch": 3.7454545454545456,
+      "grad_norm": 1.3421468548020907,
+      "learning_rate": 1.8438553869396775e-07,
+      "loss": 0.0083,
+      "step": 16480
+    },
+    {
+      "epoch": 3.745681818181818,
+      "grad_norm": 1.3292691832773411,
+      "learning_rate": 1.8432224508680896e-07,
+      "loss": 0.0059,
+      "step": 16481
+    },
+    {
+      "epoch": 3.745909090909091,
+      "grad_norm": 0.7216186165823835,
+      "learning_rate": 1.842589604658386e-07,
+      "loss": 0.0036,
+      "step": 16482
+    },
+    {
+      "epoch": 3.7461363636363636,
+      "grad_norm": 1.7217103641546734,
+      "learning_rate": 1.8419568483234673e-07,
+      "loss": 0.0106,
+      "step": 16483
+    },
+    {
+      "epoch": 3.746363636363636,
+      "grad_norm": 1.8474792921666667,
+      "learning_rate": 1.841324181876239e-07,
+      "loss": 0.0181,
+      "step": 16484
+    },
+    {
+      "epoch": 3.746590909090909,
+      "grad_norm": 1.4564525746591632,
+      "learning_rate": 1.8406916053296025e-07,
+      "loss": 0.0136,
+      "step": 16485
+    },
+    {
+      "epoch": 3.746818181818182,
+      "grad_norm": 2.4929925195158886,
+      "learning_rate": 1.8400591186964577e-07,
+      "loss": 0.0343,
+      "step": 16486
+    },
+    {
+      "epoch": 3.7470454545454546,
+      "grad_norm": 4.1495874247458175,
+      "learning_rate": 1.8394267219896997e-07,
+      "loss": 0.0176,
+      "step": 16487
+    },
+    {
+      "epoch": 3.747272727272727,
+      "grad_norm": 1.1715752426125194,
+      "learning_rate": 1.8387944152222245e-07,
+      "loss": 0.0026,
+      "step": 16488
+    },
+    {
+      "epoch": 3.7475,
+      "grad_norm": 1.8714348860738363,
+      "learning_rate": 1.838162198406928e-07,
+      "loss": 0.0174,
+      "step": 16489
+    },
+    {
+      "epoch": 3.7477272727272726,
+      "grad_norm": 2.9677404697330103,
+      "learning_rate": 1.8375300715567006e-07,
+      "loss": 0.0303,
+      "step": 16490
+    },
+    {
+      "epoch": 3.7479545454545455,
+      "grad_norm": 1.3459797358194199,
+      "learning_rate": 1.8368980346844352e-07,
+      "loss": 0.0173,
+      "step": 16491
+    },
+    {
+      "epoch": 3.748181818181818,
+      "grad_norm": 2.573875187690485,
+      "learning_rate": 1.836266087803016e-07,
+      "loss": 0.0139,
+      "step": 16492
+    },
+    {
+      "epoch": 3.748409090909091,
+      "grad_norm": 1.6019999513594099,
+      "learning_rate": 1.8356342309253317e-07,
+      "loss": 0.0109,
+      "step": 16493
+    },
+    {
+      "epoch": 3.7486363636363635,
+      "grad_norm": 1.543137108883014,
+      "learning_rate": 1.8350024640642665e-07,
+      "loss": 0.0207,
+      "step": 16494
+    },
+    {
+      "epoch": 3.7488636363636365,
+      "grad_norm": 1.7985295175090426,
+      "learning_rate": 1.834370787232703e-07,
+      "loss": 0.0076,
+      "step": 16495
+    },
+    {
+      "epoch": 3.749090909090909,
+      "grad_norm": 0.6918881474119905,
+      "learning_rate": 1.8337392004435233e-07,
+      "loss": 0.0034,
+      "step": 16496
+    },
+    {
+      "epoch": 3.7493181818181816,
+      "grad_norm": 2.3923264054035673,
+      "learning_rate": 1.8331077037096057e-07,
+      "loss": 0.0343,
+      "step": 16497
+    },
+    {
+      "epoch": 3.7495454545454545,
+      "grad_norm": 2.7686602123513,
+      "learning_rate": 1.8324762970438273e-07,
+      "loss": 0.0169,
+      "step": 16498
+    },
+    {
+      "epoch": 3.7497727272727275,
+      "grad_norm": 1.5569127019147886,
+      "learning_rate": 1.8318449804590666e-07,
+      "loss": 0.0164,
+      "step": 16499
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 3.1114525724624382,
+      "learning_rate": 1.8312137539681924e-07,
+      "loss": 0.0123,
+      "step": 16500
+    },
+    {
+      "epoch": 3.7502272727272725,
+      "grad_norm": 1.3181502664891716,
+      "learning_rate": 1.8305826175840785e-07,
+      "loss": 0.0129,
+      "step": 16501
+    },
+    {
+      "epoch": 3.7504545454545455,
+      "grad_norm": 0.7463554966141465,
+      "learning_rate": 1.829951571319595e-07,
+      "loss": 0.0043,
+      "step": 16502
+    },
+    {
+      "epoch": 3.7506818181818184,
+      "grad_norm": 1.1145700644474577,
+      "learning_rate": 1.829320615187611e-07,
+      "loss": 0.0093,
+      "step": 16503
+    },
+    {
+      "epoch": 3.750909090909091,
+      "grad_norm": 1.83699092910409,
+      "learning_rate": 1.828689749200993e-07,
+      "loss": 0.0162,
+      "step": 16504
+    },
+    {
+      "epoch": 3.7511363636363635,
+      "grad_norm": 0.6970039499821684,
+      "learning_rate": 1.8280589733726033e-07,
+      "loss": 0.0039,
+      "step": 16505
+    },
+    {
+      "epoch": 3.7513636363636365,
+      "grad_norm": 3.5615609036247378,
+      "learning_rate": 1.827428287715305e-07,
+      "loss": 0.0114,
+      "step": 16506
+    },
+    {
+      "epoch": 3.751590909090909,
+      "grad_norm": 0.7820464226837451,
+      "learning_rate": 1.82679769224196e-07,
+      "loss": 0.0043,
+      "step": 16507
+    },
+    {
+      "epoch": 3.751818181818182,
+      "grad_norm": 1.6981197369920937,
+      "learning_rate": 1.826167186965426e-07,
+      "loss": 0.011,
+      "step": 16508
+    },
+    {
+      "epoch": 3.7520454545454545,
+      "grad_norm": 3.337347252955411,
+      "learning_rate": 1.8255367718985616e-07,
+      "loss": 0.0238,
+      "step": 16509
+    },
+    {
+      "epoch": 3.7522727272727274,
+      "grad_norm": 0.9441742207023913,
+      "learning_rate": 1.8249064470542206e-07,
+      "loss": 0.0042,
+      "step": 16510
+    },
+    {
+      "epoch": 3.7525,
+      "grad_norm": 1.5407260270281833,
+      "learning_rate": 1.8242762124452578e-07,
+      "loss": 0.0113,
+      "step": 16511
+    },
+    {
+      "epoch": 3.752727272727273,
+      "grad_norm": 1.262213637907058,
+      "learning_rate": 1.823646068084526e-07,
+      "loss": 0.0052,
+      "step": 16512
+    },
+    {
+      "epoch": 3.7529545454545454,
+      "grad_norm": 1.5389172924200865,
+      "learning_rate": 1.8230160139848712e-07,
+      "loss": 0.0122,
+      "step": 16513
+    },
+    {
+      "epoch": 3.753181818181818,
+      "grad_norm": 2.0018922453902133,
+      "learning_rate": 1.8223860501591423e-07,
+      "loss": 0.0116,
+      "step": 16514
+    },
+    {
+      "epoch": 3.753409090909091,
+      "grad_norm": 0.47992389448392353,
+      "learning_rate": 1.8217561766201873e-07,
+      "loss": 0.0033,
+      "step": 16515
+    },
+    {
+      "epoch": 3.753636363636364,
+      "grad_norm": 1.361042983925254,
+      "learning_rate": 1.821126393380849e-07,
+      "loss": 0.0061,
+      "step": 16516
+    },
+    {
+      "epoch": 3.7538636363636364,
+      "grad_norm": 0.9463546510687902,
+      "learning_rate": 1.820496700453972e-07,
+      "loss": 0.0068,
+      "step": 16517
+    },
+    {
+      "epoch": 3.754090909090909,
+      "grad_norm": 1.8083196221323332,
+      "learning_rate": 1.819867097852393e-07,
+      "loss": 0.0088,
+      "step": 16518
+    },
+    {
+      "epoch": 3.754318181818182,
+      "grad_norm": 0.8873714984762004,
+      "learning_rate": 1.8192375855889532e-07,
+      "loss": 0.0028,
+      "step": 16519
+    },
+    {
+      "epoch": 3.7545454545454544,
+      "grad_norm": 3.606456443854446,
+      "learning_rate": 1.8186081636764879e-07,
+      "loss": 0.0223,
+      "step": 16520
+    },
+    {
+      "epoch": 3.7547727272727274,
+      "grad_norm": 2.1766258462906536,
+      "learning_rate": 1.8179788321278354e-07,
+      "loss": 0.0192,
+      "step": 16521
+    },
+    {
+      "epoch": 3.755,
+      "grad_norm": 2.0128587317499926,
+      "learning_rate": 1.8173495909558255e-07,
+      "loss": 0.0091,
+      "step": 16522
+    },
+    {
+      "epoch": 3.755227272727273,
+      "grad_norm": 5.282448745770988,
+      "learning_rate": 1.8167204401732899e-07,
+      "loss": 0.0164,
+      "step": 16523
+    },
+    {
+      "epoch": 3.7554545454545454,
+      "grad_norm": 1.7319147026677124,
+      "learning_rate": 1.8160913797930584e-07,
+      "loss": 0.024,
+      "step": 16524
+    },
+    {
+      "epoch": 3.7556818181818183,
+      "grad_norm": 0.7747017465689512,
+      "learning_rate": 1.8154624098279594e-07,
+      "loss": 0.0043,
+      "step": 16525
+    },
+    {
+      "epoch": 3.755909090909091,
+      "grad_norm": 0.47088145497856504,
+      "learning_rate": 1.8148335302908187e-07,
+      "loss": 0.0015,
+      "step": 16526
+    },
+    {
+      "epoch": 3.7561363636363634,
+      "grad_norm": 1.8331607349380787,
+      "learning_rate": 1.8142047411944598e-07,
+      "loss": 0.01,
+      "step": 16527
+    },
+    {
+      "epoch": 3.7563636363636363,
+      "grad_norm": 1.0539370918319775,
+      "learning_rate": 1.8135760425517047e-07,
+      "loss": 0.0062,
+      "step": 16528
+    },
+    {
+      "epoch": 3.7565909090909093,
+      "grad_norm": 1.3171649898641593,
+      "learning_rate": 1.8129474343753741e-07,
+      "loss": 0.0144,
+      "step": 16529
+    },
+    {
+      "epoch": 3.756818181818182,
+      "grad_norm": 2.007047477091049,
+      "learning_rate": 1.8123189166782877e-07,
+      "loss": 0.0105,
+      "step": 16530
+    },
+    {
+      "epoch": 3.7570454545454544,
+      "grad_norm": 1.7775390520628958,
+      "learning_rate": 1.8116904894732592e-07,
+      "loss": 0.0149,
+      "step": 16531
+    },
+    {
+      "epoch": 3.7572727272727273,
+      "grad_norm": 1.5301173962767591,
+      "learning_rate": 1.811062152773104e-07,
+      "loss": 0.0073,
+      "step": 16532
+    },
+    {
+      "epoch": 3.7575,
+      "grad_norm": 1.8317080844070686,
+      "learning_rate": 1.8104339065906363e-07,
+      "loss": 0.0102,
+      "step": 16533
+    },
+    {
+      "epoch": 3.757727272727273,
+      "grad_norm": 2.2390008000145234,
+      "learning_rate": 1.809805750938668e-07,
+      "loss": 0.0232,
+      "step": 16534
+    },
+    {
+      "epoch": 3.7579545454545453,
+      "grad_norm": 1.1977644911224201,
+      "learning_rate": 1.8091776858300048e-07,
+      "loss": 0.0086,
+      "step": 16535
+    },
+    {
+      "epoch": 3.7581818181818183,
+      "grad_norm": 2.6987169086535148,
+      "learning_rate": 1.808549711277456e-07,
+      "loss": 0.022,
+      "step": 16536
+    },
+    {
+      "epoch": 3.758409090909091,
+      "grad_norm": 0.7854046860443246,
+      "learning_rate": 1.807921827293827e-07,
+      "loss": 0.0038,
+      "step": 16537
+    },
+    {
+      "epoch": 3.7586363636363638,
+      "grad_norm": 1.83975421333064,
+      "learning_rate": 1.8072940338919222e-07,
+      "loss": 0.0084,
+      "step": 16538
+    },
+    {
+      "epoch": 3.7588636363636363,
+      "grad_norm": 1.5639486922024686,
+      "learning_rate": 1.8066663310845424e-07,
+      "loss": 0.0103,
+      "step": 16539
+    },
+    {
+      "epoch": 3.759090909090909,
+      "grad_norm": 0.7348967872365237,
+      "learning_rate": 1.8060387188844884e-07,
+      "loss": 0.0029,
+      "step": 16540
+    },
+    {
+      "epoch": 3.759318181818182,
+      "grad_norm": 0.7971074256715296,
+      "learning_rate": 1.8054111973045573e-07,
+      "loss": 0.0048,
+      "step": 16541
+    },
+    {
+      "epoch": 3.7595454545454547,
+      "grad_norm": 2.240593386844313,
+      "learning_rate": 1.8047837663575458e-07,
+      "loss": 0.0189,
+      "step": 16542
+    },
+    {
+      "epoch": 3.7597727272727273,
+      "grad_norm": 0.5578646656087116,
+      "learning_rate": 1.8041564260562505e-07,
+      "loss": 0.0033,
+      "step": 16543
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 3.7570491710397858,
+      "learning_rate": 1.80352917641346e-07,
+      "loss": 0.0232,
+      "step": 16544
+    },
+    {
+      "epoch": 3.7602272727272728,
+      "grad_norm": 1.506278554052099,
+      "learning_rate": 1.8029020174419669e-07,
+      "loss": 0.0162,
+      "step": 16545
+    },
+    {
+      "epoch": 3.7604545454545457,
+      "grad_norm": 1.4157320517769754,
+      "learning_rate": 1.80227494915456e-07,
+      "loss": 0.0083,
+      "step": 16546
+    },
+    {
+      "epoch": 3.7606818181818182,
+      "grad_norm": 1.2695796676740905,
+      "learning_rate": 1.8016479715640265e-07,
+      "loss": 0.0051,
+      "step": 16547
+    },
+    {
+      "epoch": 3.7609090909090908,
+      "grad_norm": 2.546710207548228,
+      "learning_rate": 1.8010210846831536e-07,
+      "loss": 0.0107,
+      "step": 16548
+    },
+    {
+      "epoch": 3.7611363636363637,
+      "grad_norm": 0.8455379647829657,
+      "learning_rate": 1.80039428852472e-07,
+      "loss": 0.0034,
+      "step": 16549
+    },
+    {
+      "epoch": 3.7613636363636362,
+      "grad_norm": 1.6685220852569058,
+      "learning_rate": 1.7997675831015095e-07,
+      "loss": 0.0121,
+      "step": 16550
+    },
+    {
+      "epoch": 3.761590909090909,
+      "grad_norm": 1.9008787736568005,
+      "learning_rate": 1.7991409684263024e-07,
+      "loss": 0.0166,
+      "step": 16551
+    },
+    {
+      "epoch": 3.7618181818181817,
+      "grad_norm": 1.8032722117267301,
+      "learning_rate": 1.798514444511877e-07,
+      "loss": 0.0089,
+      "step": 16552
+    },
+    {
+      "epoch": 3.7620454545454547,
+      "grad_norm": 3.166474663725073,
+      "learning_rate": 1.7978880113710074e-07,
+      "loss": 0.0172,
+      "step": 16553
+    },
+    {
+      "epoch": 3.762272727272727,
+      "grad_norm": 0.9547582031210989,
+      "learning_rate": 1.7972616690164674e-07,
+      "loss": 0.0053,
+      "step": 16554
+    },
+    {
+      "epoch": 3.7625,
+      "grad_norm": 2.0671890208225934,
+      "learning_rate": 1.7966354174610303e-07,
+      "loss": 0.0151,
+      "step": 16555
+    },
+    {
+      "epoch": 3.7627272727272727,
+      "grad_norm": 1.2462314829538657,
+      "learning_rate": 1.7960092567174667e-07,
+      "loss": 0.0047,
+      "step": 16556
+    },
+    {
+      "epoch": 3.762954545454545,
+      "grad_norm": 2.4180279820401984,
+      "learning_rate": 1.7953831867985452e-07,
+      "loss": 0.0282,
+      "step": 16557
+    },
+    {
+      "epoch": 3.763181818181818,
+      "grad_norm": 2.8725452813682866,
+      "learning_rate": 1.7947572077170316e-07,
+      "loss": 0.0142,
+      "step": 16558
+    },
+    {
+      "epoch": 3.763409090909091,
+      "grad_norm": 0.8475323044865729,
+      "learning_rate": 1.7941313194856903e-07,
+      "loss": 0.0046,
+      "step": 16559
+    },
+    {
+      "epoch": 3.7636363636363637,
+      "grad_norm": 0.9306746862853505,
+      "learning_rate": 1.7935055221172864e-07,
+      "loss": 0.0076,
+      "step": 16560
+    },
+    {
+      "epoch": 3.763863636363636,
+      "grad_norm": 1.0610221460322473,
+      "learning_rate": 1.7928798156245806e-07,
+      "loss": 0.0057,
+      "step": 16561
+    },
+    {
+      "epoch": 3.764090909090909,
+      "grad_norm": 1.3764531092882364,
+      "learning_rate": 1.7922542000203297e-07,
+      "loss": 0.0128,
+      "step": 16562
+    },
+    {
+      "epoch": 3.7643181818181817,
+      "grad_norm": 0.9517225504452502,
+      "learning_rate": 1.7916286753172932e-07,
+      "loss": 0.0124,
+      "step": 16563
+    },
+    {
+      "epoch": 3.7645454545454546,
+      "grad_norm": 0.875936891057738,
+      "learning_rate": 1.7910032415282255e-07,
+      "loss": 0.0079,
+      "step": 16564
+    },
+    {
+      "epoch": 3.764772727272727,
+      "grad_norm": 1.611129613515252,
+      "learning_rate": 1.7903778986658828e-07,
+      "loss": 0.014,
+      "step": 16565
+    },
+    {
+      "epoch": 3.765,
+      "grad_norm": 1.015653629404247,
+      "learning_rate": 1.7897526467430128e-07,
+      "loss": 0.0059,
+      "step": 16566
+    },
+    {
+      "epoch": 3.7652272727272726,
+      "grad_norm": 2.9854489429568107,
+      "learning_rate": 1.7891274857723678e-07,
+      "loss": 0.022,
+      "step": 16567
+    },
+    {
+      "epoch": 3.7654545454545456,
+      "grad_norm": 1.900778374352806,
+      "learning_rate": 1.7885024157666963e-07,
+      "loss": 0.0092,
+      "step": 16568
+    },
+    {
+      "epoch": 3.765681818181818,
+      "grad_norm": 1.0287605308186394,
+      "learning_rate": 1.7878774367387434e-07,
+      "loss": 0.0032,
+      "step": 16569
+    },
+    {
+      "epoch": 3.7659090909090907,
+      "grad_norm": 2.4177740082500625,
+      "learning_rate": 1.7872525487012542e-07,
+      "loss": 0.0114,
+      "step": 16570
+    },
+    {
+      "epoch": 3.7661363636363636,
+      "grad_norm": 0.7580410406271959,
+      "learning_rate": 1.7866277516669716e-07,
+      "loss": 0.0052,
+      "step": 16571
+    },
+    {
+      "epoch": 3.7663636363636366,
+      "grad_norm": 1.8326427995286405,
+      "learning_rate": 1.7860030456486355e-07,
+      "loss": 0.0142,
+      "step": 16572
+    },
+    {
+      "epoch": 3.766590909090909,
+      "grad_norm": 2.2752302591527234,
+      "learning_rate": 1.7853784306589857e-07,
+      "loss": 0.0248,
+      "step": 16573
+    },
+    {
+      "epoch": 3.7668181818181816,
+      "grad_norm": 2.9989418038364644,
+      "learning_rate": 1.7847539067107601e-07,
+      "loss": 0.03,
+      "step": 16574
+    },
+    {
+      "epoch": 3.7670454545454546,
+      "grad_norm": 0.794317577243393,
+      "learning_rate": 1.7841294738166912e-07,
+      "loss": 0.0042,
+      "step": 16575
+    },
+    {
+      "epoch": 3.767272727272727,
+      "grad_norm": 2.295385944843749,
+      "learning_rate": 1.783505131989513e-07,
+      "loss": 0.0141,
+      "step": 16576
+    },
+    {
+      "epoch": 3.7675,
+      "grad_norm": 1.7708153429565896,
+      "learning_rate": 1.7828808812419572e-07,
+      "loss": 0.0126,
+      "step": 16577
+    },
+    {
+      "epoch": 3.7677272727272726,
+      "grad_norm": 1.3536236041525143,
+      "learning_rate": 1.7822567215867553e-07,
+      "loss": 0.0137,
+      "step": 16578
+    },
+    {
+      "epoch": 3.7679545454545456,
+      "grad_norm": 2.103966513572833,
+      "learning_rate": 1.7816326530366315e-07,
+      "loss": 0.0098,
+      "step": 16579
+    },
+    {
+      "epoch": 3.768181818181818,
+      "grad_norm": 1.0478482479297135,
+      "learning_rate": 1.7810086756043142e-07,
+      "loss": 0.0034,
+      "step": 16580
+    },
+    {
+      "epoch": 3.768409090909091,
+      "grad_norm": 1.476786056122564,
+      "learning_rate": 1.7803847893025257e-07,
+      "loss": 0.016,
+      "step": 16581
+    },
+    {
+      "epoch": 3.7686363636363636,
+      "grad_norm": 1.5640123565271855,
+      "learning_rate": 1.7797609941439893e-07,
+      "loss": 0.0099,
+      "step": 16582
+    },
+    {
+      "epoch": 3.7688636363636365,
+      "grad_norm": 0.8234866499546045,
+      "learning_rate": 1.7791372901414263e-07,
+      "loss": 0.0048,
+      "step": 16583
+    },
+    {
+      "epoch": 3.769090909090909,
+      "grad_norm": 1.2365327064835716,
+      "learning_rate": 1.7785136773075529e-07,
+      "loss": 0.008,
+      "step": 16584
+    },
+    {
+      "epoch": 3.769318181818182,
+      "grad_norm": 1.2157502987004476,
+      "learning_rate": 1.7778901556550842e-07,
+      "loss": 0.0064,
+      "step": 16585
+    },
+    {
+      "epoch": 3.7695454545454545,
+      "grad_norm": 0.36430744978955737,
+      "learning_rate": 1.7772667251967392e-07,
+      "loss": 0.0013,
+      "step": 16586
+    },
+    {
+      "epoch": 3.769772727272727,
+      "grad_norm": 0.9458203317066357,
+      "learning_rate": 1.7766433859452309e-07,
+      "loss": 0.0045,
+      "step": 16587
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 1.69677971910505,
+      "learning_rate": 1.7760201379132653e-07,
+      "loss": 0.007,
+      "step": 16588
+    },
+    {
+      "epoch": 3.770227272727273,
+      "grad_norm": 1.8921188651157244,
+      "learning_rate": 1.775396981113555e-07,
+      "loss": 0.0176,
+      "step": 16589
+    },
+    {
+      "epoch": 3.7704545454545455,
+      "grad_norm": 2.1311234303229165,
+      "learning_rate": 1.774773915558807e-07,
+      "loss": 0.0152,
+      "step": 16590
+    },
+    {
+      "epoch": 3.770681818181818,
+      "grad_norm": 1.6023593056623542,
+      "learning_rate": 1.7741509412617257e-07,
+      "loss": 0.0104,
+      "step": 16591
+    },
+    {
+      "epoch": 3.770909090909091,
+      "grad_norm": 1.2978161274506128,
+      "learning_rate": 1.7735280582350176e-07,
+      "loss": 0.006,
+      "step": 16592
+    },
+    {
+      "epoch": 3.7711363636363635,
+      "grad_norm": 2.190347444835615,
+      "learning_rate": 1.7729052664913802e-07,
+      "loss": 0.0055,
+      "step": 16593
+    },
+    {
+      "epoch": 3.7713636363636365,
+      "grad_norm": 0.8390255039732485,
+      "learning_rate": 1.772282566043515e-07,
+      "loss": 0.006,
+      "step": 16594
+    },
+    {
+      "epoch": 3.771590909090909,
+      "grad_norm": 2.093730784780625,
+      "learning_rate": 1.7716599569041197e-07,
+      "loss": 0.0226,
+      "step": 16595
+    },
+    {
+      "epoch": 3.771818181818182,
+      "grad_norm": 1.3961787969708095,
+      "learning_rate": 1.771037439085893e-07,
+      "loss": 0.0139,
+      "step": 16596
+    },
+    {
+      "epoch": 3.7720454545454545,
+      "grad_norm": 0.9734710048486637,
+      "learning_rate": 1.770415012601525e-07,
+      "loss": 0.0077,
+      "step": 16597
+    },
+    {
+      "epoch": 3.7722727272727274,
+      "grad_norm": 1.60712215027075,
+      "learning_rate": 1.7697926774637102e-07,
+      "loss": 0.0083,
+      "step": 16598
+    },
+    {
+      "epoch": 3.7725,
+      "grad_norm": 1.196397177329153,
+      "learning_rate": 1.7691704336851385e-07,
+      "loss": 0.0055,
+      "step": 16599
+    },
+    {
+      "epoch": 3.7727272727272725,
+      "grad_norm": 3.624572028997969,
+      "learning_rate": 1.768548281278499e-07,
+      "loss": 0.0175,
+      "step": 16600
+    },
+    {
+      "epoch": 3.7729545454545454,
+      "grad_norm": 1.9501265605795643,
+      "learning_rate": 1.7679262202564786e-07,
+      "loss": 0.0123,
+      "step": 16601
+    },
+    {
+      "epoch": 3.7731818181818184,
+      "grad_norm": 0.9200525241065673,
+      "learning_rate": 1.7673042506317614e-07,
+      "loss": 0.0047,
+      "step": 16602
+    },
+    {
+      "epoch": 3.773409090909091,
+      "grad_norm": 0.7068505453709204,
+      "learning_rate": 1.7666823724170314e-07,
+      "loss": 0.0046,
+      "step": 16603
+    },
+    {
+      "epoch": 3.7736363636363635,
+      "grad_norm": 1.9054484976087083,
+      "learning_rate": 1.7660605856249694e-07,
+      "loss": 0.0076,
+      "step": 16604
+    },
+    {
+      "epoch": 3.7738636363636364,
+      "grad_norm": 1.7128708215957014,
+      "learning_rate": 1.7654388902682568e-07,
+      "loss": 0.0082,
+      "step": 16605
+    },
+    {
+      "epoch": 3.774090909090909,
+      "grad_norm": 3.4410784751770653,
+      "learning_rate": 1.7648172863595665e-07,
+      "loss": 0.0139,
+      "step": 16606
+    },
+    {
+      "epoch": 3.774318181818182,
+      "grad_norm": 1.6445712589875492,
+      "learning_rate": 1.7641957739115769e-07,
+      "loss": 0.0082,
+      "step": 16607
+    },
+    {
+      "epoch": 3.7745454545454544,
+      "grad_norm": 2.2763763736261033,
+      "learning_rate": 1.7635743529369614e-07,
+      "loss": 0.0134,
+      "step": 16608
+    },
+    {
+      "epoch": 3.7747727272727274,
+      "grad_norm": 13.118373897248619,
+      "learning_rate": 1.7629530234483938e-07,
+      "loss": 0.0082,
+      "step": 16609
+    },
+    {
+      "epoch": 3.775,
+      "grad_norm": 0.37872439590268336,
+      "learning_rate": 1.7623317854585402e-07,
+      "loss": 0.0018,
+      "step": 16610
+    },
+    {
+      "epoch": 3.775227272727273,
+      "grad_norm": 0.6213862298433244,
+      "learning_rate": 1.7617106389800709e-07,
+      "loss": 0.0029,
+      "step": 16611
+    },
+    {
+      "epoch": 3.7754545454545454,
+      "grad_norm": 0.9173841027227073,
+      "learning_rate": 1.7610895840256513e-07,
+      "loss": 0.0047,
+      "step": 16612
+    },
+    {
+      "epoch": 3.775681818181818,
+      "grad_norm": 2.0991966030935685,
+      "learning_rate": 1.760468620607947e-07,
+      "loss": 0.0111,
+      "step": 16613
+    },
+    {
+      "epoch": 3.775909090909091,
+      "grad_norm": 0.42930008858529667,
+      "learning_rate": 1.7598477487396216e-07,
+      "loss": 0.0078,
+      "step": 16614
+    },
+    {
+      "epoch": 3.776136363636364,
+      "grad_norm": 2.2371499405062463,
+      "learning_rate": 1.7592269684333302e-07,
+      "loss": 0.0083,
+      "step": 16615
+    },
+    {
+      "epoch": 3.7763636363636364,
+      "grad_norm": 2.0944912621605583,
+      "learning_rate": 1.7586062797017376e-07,
+      "loss": 0.0316,
+      "step": 16616
+    },
+    {
+      "epoch": 3.776590909090909,
+      "grad_norm": 1.2816989629623947,
+      "learning_rate": 1.7579856825574984e-07,
+      "loss": 0.0083,
+      "step": 16617
+    },
+    {
+      "epoch": 3.776818181818182,
+      "grad_norm": 1.5318050063811668,
+      "learning_rate": 1.7573651770132698e-07,
+      "loss": 0.0205,
+      "step": 16618
+    },
+    {
+      "epoch": 3.777045454545455,
+      "grad_norm": 1.5355958523374968,
+      "learning_rate": 1.7567447630817017e-07,
+      "loss": 0.0066,
+      "step": 16619
+    },
+    {
+      "epoch": 3.7772727272727273,
+      "grad_norm": 1.2020794227368667,
+      "learning_rate": 1.756124440775446e-07,
+      "loss": 0.007,
+      "step": 16620
+    },
+    {
+      "epoch": 3.7775,
+      "grad_norm": 1.4253889775155286,
+      "learning_rate": 1.7555042101071534e-07,
+      "loss": 0.0085,
+      "step": 16621
+    },
+    {
+      "epoch": 3.777727272727273,
+      "grad_norm": 1.9045728440098328,
+      "learning_rate": 1.754884071089473e-07,
+      "loss": 0.0165,
+      "step": 16622
+    },
+    {
+      "epoch": 3.7779545454545453,
+      "grad_norm": 3.1576954812050486,
+      "learning_rate": 1.7542640237350466e-07,
+      "loss": 0.0235,
+      "step": 16623
+    },
+    {
+      "epoch": 3.7781818181818183,
+      "grad_norm": 1.9350996246181618,
+      "learning_rate": 1.7536440680565197e-07,
+      "loss": 0.0164,
+      "step": 16624
+    },
+    {
+      "epoch": 3.778409090909091,
+      "grad_norm": 2.5924371440155807,
+      "learning_rate": 1.7530242040665348e-07,
+      "loss": 0.0201,
+      "step": 16625
+    },
+    {
+      "epoch": 3.778636363636364,
+      "grad_norm": 1.494266942827218,
+      "learning_rate": 1.7524044317777323e-07,
+      "loss": 0.0041,
+      "step": 16626
+    },
+    {
+      "epoch": 3.7788636363636363,
+      "grad_norm": 0.4749514470528812,
+      "learning_rate": 1.751784751202751e-07,
+      "loss": 0.0042,
+      "step": 16627
+    },
+    {
+      "epoch": 3.7790909090909093,
+      "grad_norm": 3.826872420075198,
+      "learning_rate": 1.7511651623542244e-07,
+      "loss": 0.0265,
+      "step": 16628
+    },
+    {
+      "epoch": 3.779318181818182,
+      "grad_norm": 2.1817462287077896,
+      "learning_rate": 1.7505456652447893e-07,
+      "loss": 0.0096,
+      "step": 16629
+    },
+    {
+      "epoch": 3.7795454545454543,
+      "grad_norm": 2.3288146898155215,
+      "learning_rate": 1.7499262598870776e-07,
+      "loss": 0.0183,
+      "step": 16630
+    },
+    {
+      "epoch": 3.7797727272727273,
+      "grad_norm": 3.3369223488267017,
+      "learning_rate": 1.7493069462937197e-07,
+      "loss": 0.0082,
+      "step": 16631
+    },
+    {
+      "epoch": 3.7800000000000002,
+      "grad_norm": 1.5401740139254425,
+      "learning_rate": 1.7486877244773457e-07,
+      "loss": 0.0071,
+      "step": 16632
+    },
+    {
+      "epoch": 3.7802272727272728,
+      "grad_norm": 1.6513331986908648,
+      "learning_rate": 1.748068594450582e-07,
+      "loss": 0.0068,
+      "step": 16633
+    },
+    {
+      "epoch": 3.7804545454545453,
+      "grad_norm": 2.6277101873930926,
+      "learning_rate": 1.747449556226053e-07,
+      "loss": 0.0316,
+      "step": 16634
+    },
+    {
+      "epoch": 3.7806818181818183,
+      "grad_norm": 2.9862789675273733,
+      "learning_rate": 1.7468306098163856e-07,
+      "loss": 0.0207,
+      "step": 16635
+    },
+    {
+      "epoch": 3.7809090909090908,
+      "grad_norm": 1.2363817152196135,
+      "learning_rate": 1.7462117552341957e-07,
+      "loss": 0.011,
+      "step": 16636
+    },
+    {
+      "epoch": 3.7811363636363637,
+      "grad_norm": 1.5553156422893308,
+      "learning_rate": 1.7455929924921053e-07,
+      "loss": 0.0144,
+      "step": 16637
+    },
+    {
+      "epoch": 3.7813636363636363,
+      "grad_norm": 2.0756065892056026,
+      "learning_rate": 1.7449743216027325e-07,
+      "loss": 0.0179,
+      "step": 16638
+    },
+    {
+      "epoch": 3.7815909090909092,
+      "grad_norm": 1.6475557215251921,
+      "learning_rate": 1.7443557425786926e-07,
+      "loss": 0.01,
+      "step": 16639
+    },
+    {
+      "epoch": 3.7818181818181817,
+      "grad_norm": 1.8902333051628697,
+      "learning_rate": 1.7437372554326018e-07,
+      "loss": 0.0112,
+      "step": 16640
+    },
+    {
+      "epoch": 3.7820454545454547,
+      "grad_norm": 0.6103501337594314,
+      "learning_rate": 1.7431188601770676e-07,
+      "loss": 0.0027,
+      "step": 16641
+    },
+    {
+      "epoch": 3.7822727272727272,
+      "grad_norm": 2.073240037320524,
+      "learning_rate": 1.7425005568247028e-07,
+      "loss": 0.0245,
+      "step": 16642
+    },
+    {
+      "epoch": 3.7824999999999998,
+      "grad_norm": 1.1605462676160467,
+      "learning_rate": 1.7418823453881155e-07,
+      "loss": 0.0143,
+      "step": 16643
+    },
+    {
+      "epoch": 3.7827272727272727,
+      "grad_norm": 2.0445029088662023,
+      "learning_rate": 1.7412642258799122e-07,
+      "loss": 0.0057,
+      "step": 16644
+    },
+    {
+      "epoch": 3.7829545454545457,
+      "grad_norm": 0.31601626043798114,
+      "learning_rate": 1.7406461983126968e-07,
+      "loss": 0.0013,
+      "step": 16645
+    },
+    {
+      "epoch": 3.783181818181818,
+      "grad_norm": 1.3944013204306107,
+      "learning_rate": 1.740028262699072e-07,
+      "loss": 0.008,
+      "step": 16646
+    },
+    {
+      "epoch": 3.7834090909090907,
+      "grad_norm": 0.534923388919015,
+      "learning_rate": 1.7394104190516397e-07,
+      "loss": 0.0021,
+      "step": 16647
+    },
+    {
+      "epoch": 3.7836363636363637,
+      "grad_norm": 0.9751140056962009,
+      "learning_rate": 1.7387926673829983e-07,
+      "loss": 0.0054,
+      "step": 16648
+    },
+    {
+      "epoch": 3.783863636363636,
+      "grad_norm": 1.3250680276727012,
+      "learning_rate": 1.738175007705746e-07,
+      "loss": 0.0103,
+      "step": 16649
+    },
+    {
+      "epoch": 3.784090909090909,
+      "grad_norm": 3.121709977284907,
+      "learning_rate": 1.737557440032475e-07,
+      "loss": 0.0243,
+      "step": 16650
+    },
+    {
+      "epoch": 3.7843181818181817,
+      "grad_norm": 0.9926859488712199,
+      "learning_rate": 1.73693996437578e-07,
+      "loss": 0.0059,
+      "step": 16651
+    },
+    {
+      "epoch": 3.7845454545454547,
+      "grad_norm": 1.9177638747166796,
+      "learning_rate": 1.736322580748253e-07,
+      "loss": 0.0146,
+      "step": 16652
+    },
+    {
+      "epoch": 3.784772727272727,
+      "grad_norm": 1.0030274813891393,
+      "learning_rate": 1.7357052891624842e-07,
+      "loss": 0.0025,
+      "step": 16653
+    },
+    {
+      "epoch": 3.785,
+      "grad_norm": 4.391553278113622,
+      "learning_rate": 1.7350880896310588e-07,
+      "loss": 0.0204,
+      "step": 16654
+    },
+    {
+      "epoch": 3.7852272727272727,
+      "grad_norm": 0.914243950473224,
+      "learning_rate": 1.734470982166564e-07,
+      "loss": 0.002,
+      "step": 16655
+    },
+    {
+      "epoch": 3.785454545454545,
+      "grad_norm": 1.9381530505292834,
+      "learning_rate": 1.733853966781584e-07,
+      "loss": 0.0089,
+      "step": 16656
+    },
+    {
+      "epoch": 3.785681818181818,
+      "grad_norm": 3.2567262735236007,
+      "learning_rate": 1.7332370434887017e-07,
+      "loss": 0.0226,
+      "step": 16657
+    },
+    {
+      "epoch": 3.785909090909091,
+      "grad_norm": 1.2976107013768323,
+      "learning_rate": 1.7326202123004947e-07,
+      "loss": 0.0044,
+      "step": 16658
+    },
+    {
+      "epoch": 3.7861363636363636,
+      "grad_norm": 1.3122669026440212,
+      "learning_rate": 1.7320034732295418e-07,
+      "loss": 0.0102,
+      "step": 16659
+    },
+    {
+      "epoch": 3.786363636363636,
+      "grad_norm": 5.604500312233522,
+      "learning_rate": 1.7313868262884206e-07,
+      "loss": 0.0343,
+      "step": 16660
+    },
+    {
+      "epoch": 3.786590909090909,
+      "grad_norm": 0.8376231309249218,
+      "learning_rate": 1.7307702714897054e-07,
+      "loss": 0.0093,
+      "step": 16661
+    },
+    {
+      "epoch": 3.786818181818182,
+      "grad_norm": 2.600970264195105,
+      "learning_rate": 1.7301538088459684e-07,
+      "loss": 0.0193,
+      "step": 16662
+    },
+    {
+      "epoch": 3.7870454545454546,
+      "grad_norm": 1.5426667356573218,
+      "learning_rate": 1.7295374383697805e-07,
+      "loss": 0.0117,
+      "step": 16663
+    },
+    {
+      "epoch": 3.787272727272727,
+      "grad_norm": 1.029920372971431,
+      "learning_rate": 1.7289211600737107e-07,
+      "loss": 0.0067,
+      "step": 16664
+    },
+    {
+      "epoch": 3.7875,
+      "grad_norm": 3.7886240612828104,
+      "learning_rate": 1.7283049739703253e-07,
+      "loss": 0.0302,
+      "step": 16665
+    },
+    {
+      "epoch": 3.7877272727272726,
+      "grad_norm": 1.2306655399065252,
+      "learning_rate": 1.7276888800721926e-07,
+      "loss": 0.0022,
+      "step": 16666
+    },
+    {
+      "epoch": 3.7879545454545456,
+      "grad_norm": 1.5598430765775069,
+      "learning_rate": 1.7270728783918714e-07,
+      "loss": 0.0088,
+      "step": 16667
+    },
+    {
+      "epoch": 3.788181818181818,
+      "grad_norm": 1.0874111609854766,
+      "learning_rate": 1.7264569689419246e-07,
+      "loss": 0.0091,
+      "step": 16668
+    },
+    {
+      "epoch": 3.788409090909091,
+      "grad_norm": 1.3669727309708437,
+      "learning_rate": 1.7258411517349117e-07,
+      "loss": 0.0054,
+      "step": 16669
+    },
+    {
+      "epoch": 3.7886363636363636,
+      "grad_norm": 2.1865099073813288,
+      "learning_rate": 1.725225426783391e-07,
+      "loss": 0.0112,
+      "step": 16670
+    },
+    {
+      "epoch": 3.7888636363636365,
+      "grad_norm": 1.2423500056117076,
+      "learning_rate": 1.7246097940999198e-07,
+      "loss": 0.0076,
+      "step": 16671
+    },
+    {
+      "epoch": 3.789090909090909,
+      "grad_norm": 1.0026518363356633,
+      "learning_rate": 1.7239942536970478e-07,
+      "loss": 0.0064,
+      "step": 16672
+    },
+    {
+      "epoch": 3.7893181818181816,
+      "grad_norm": 1.1418074925216464,
+      "learning_rate": 1.7233788055873288e-07,
+      "loss": 0.006,
+      "step": 16673
+    },
+    {
+      "epoch": 3.7895454545454546,
+      "grad_norm": 3.531882106237678,
+      "learning_rate": 1.722763449783313e-07,
+      "loss": 0.01,
+      "step": 16674
+    },
+    {
+      "epoch": 3.7897727272727275,
+      "grad_norm": 0.9309858344572668,
+      "learning_rate": 1.7221481862975487e-07,
+      "loss": 0.0057,
+      "step": 16675
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 3.9306960064212646,
+      "learning_rate": 1.7215330151425827e-07,
+      "loss": 0.0278,
+      "step": 16676
+    },
+    {
+      "epoch": 3.7902272727272726,
+      "grad_norm": 2.0420407840112595,
+      "learning_rate": 1.7209179363309584e-07,
+      "loss": 0.0075,
+      "step": 16677
+    },
+    {
+      "epoch": 3.7904545454545455,
+      "grad_norm": 1.1197198517539355,
+      "learning_rate": 1.7203029498752188e-07,
+      "loss": 0.0111,
+      "step": 16678
+    },
+    {
+      "epoch": 3.790681818181818,
+      "grad_norm": 1.420937908330814,
+      "learning_rate": 1.719688055787906e-07,
+      "loss": 0.013,
+      "step": 16679
+    },
+    {
+      "epoch": 3.790909090909091,
+      "grad_norm": 3.0476748230091846,
+      "learning_rate": 1.719073254081556e-07,
+      "loss": 0.0093,
+      "step": 16680
+    },
+    {
+      "epoch": 3.7911363636363635,
+      "grad_norm": 1.3120914860344661,
+      "learning_rate": 1.7184585447687072e-07,
+      "loss": 0.0139,
+      "step": 16681
+    },
+    {
+      "epoch": 3.7913636363636365,
+      "grad_norm": 1.4800115266854466,
+      "learning_rate": 1.717843927861894e-07,
+      "loss": 0.0077,
+      "step": 16682
+    },
+    {
+      "epoch": 3.791590909090909,
+      "grad_norm": 0.9514985598306533,
+      "learning_rate": 1.71722940337365e-07,
+      "loss": 0.0052,
+      "step": 16683
+    },
+    {
+      "epoch": 3.791818181818182,
+      "grad_norm": 0.906205420067718,
+      "learning_rate": 1.7166149713165075e-07,
+      "loss": 0.0086,
+      "step": 16684
+    },
+    {
+      "epoch": 3.7920454545454545,
+      "grad_norm": 2.37014018581346,
+      "learning_rate": 1.7160006317029938e-07,
+      "loss": 0.0122,
+      "step": 16685
+    },
+    {
+      "epoch": 3.792272727272727,
+      "grad_norm": 1.472727062383276,
+      "learning_rate": 1.7153863845456372e-07,
+      "loss": 0.0101,
+      "step": 16686
+    },
+    {
+      "epoch": 3.7925,
+      "grad_norm": 1.160365524487411,
+      "learning_rate": 1.7147722298569627e-07,
+      "loss": 0.0034,
+      "step": 16687
+    },
+    {
+      "epoch": 3.792727272727273,
+      "grad_norm": 1.3129885205762593,
+      "learning_rate": 1.7141581676494968e-07,
+      "loss": 0.0077,
+      "step": 16688
+    },
+    {
+      "epoch": 3.7929545454545455,
+      "grad_norm": 2.3175095241260597,
+      "learning_rate": 1.713544197935757e-07,
+      "loss": 0.0062,
+      "step": 16689
+    },
+    {
+      "epoch": 3.793181818181818,
+      "grad_norm": 1.3523964604140346,
+      "learning_rate": 1.712930320728265e-07,
+      "loss": 0.0184,
+      "step": 16690
+    },
+    {
+      "epoch": 3.793409090909091,
+      "grad_norm": 2.055529294194914,
+      "learning_rate": 1.71231653603954e-07,
+      "loss": 0.0152,
+      "step": 16691
+    },
+    {
+      "epoch": 3.7936363636363635,
+      "grad_norm": 3.268556339754942,
+      "learning_rate": 1.7117028438820967e-07,
+      "loss": 0.011,
+      "step": 16692
+    },
+    {
+      "epoch": 3.7938636363636364,
+      "grad_norm": 0.5630207801528018,
+      "learning_rate": 1.7110892442684498e-07,
+      "loss": 0.0046,
+      "step": 16693
+    },
+    {
+      "epoch": 3.794090909090909,
+      "grad_norm": 1.5153198172007853,
+      "learning_rate": 1.710475737211112e-07,
+      "loss": 0.0068,
+      "step": 16694
+    },
+    {
+      "epoch": 3.794318181818182,
+      "grad_norm": 1.9018586500943027,
+      "learning_rate": 1.7098623227225927e-07,
+      "loss": 0.0153,
+      "step": 16695
+    },
+    {
+      "epoch": 3.7945454545454544,
+      "grad_norm": 0.363934191375704,
+      "learning_rate": 1.7092490008154022e-07,
+      "loss": 0.0013,
+      "step": 16696
+    },
+    {
+      "epoch": 3.7947727272727274,
+      "grad_norm": 0.770811680331664,
+      "learning_rate": 1.7086357715020484e-07,
+      "loss": 0.0048,
+      "step": 16697
+    },
+    {
+      "epoch": 3.795,
+      "grad_norm": 3.498477484359102,
+      "learning_rate": 1.708022634795032e-07,
+      "loss": 0.0278,
+      "step": 16698
+    },
+    {
+      "epoch": 3.795227272727273,
+      "grad_norm": 1.1144282592223795,
+      "learning_rate": 1.707409590706858e-07,
+      "loss": 0.0055,
+      "step": 16699
+    },
+    {
+      "epoch": 3.7954545454545454,
+      "grad_norm": 1.5002390982351956,
+      "learning_rate": 1.7067966392500274e-07,
+      "loss": 0.012,
+      "step": 16700
+    },
+    {
+      "epoch": 3.7956818181818184,
+      "grad_norm": 2.277573784997163,
+      "learning_rate": 1.706183780437041e-07,
+      "loss": 0.0158,
+      "step": 16701
+    },
+    {
+      "epoch": 3.795909090909091,
+      "grad_norm": 2.1459704570143385,
+      "learning_rate": 1.7055710142803924e-07,
+      "loss": 0.017,
+      "step": 16702
+    },
+    {
+      "epoch": 3.7961363636363634,
+      "grad_norm": 1.8698679707601193,
+      "learning_rate": 1.7049583407925798e-07,
+      "loss": 0.0149,
+      "step": 16703
+    },
+    {
+      "epoch": 3.7963636363636364,
+      "grad_norm": 2.2971872023789826,
+      "learning_rate": 1.7043457599860952e-07,
+      "loss": 0.0251,
+      "step": 16704
+    },
+    {
+      "epoch": 3.7965909090909093,
+      "grad_norm": 1.743422931670618,
+      "learning_rate": 1.7037332718734313e-07,
+      "loss": 0.0179,
+      "step": 16705
+    },
+    {
+      "epoch": 3.796818181818182,
+      "grad_norm": 1.3111506561493957,
+      "learning_rate": 1.7031208764670772e-07,
+      "loss": 0.015,
+      "step": 16706
+    },
+    {
+      "epoch": 3.7970454545454544,
+      "grad_norm": 8.44870698353076,
+      "learning_rate": 1.702508573779521e-07,
+      "loss": 0.0177,
+      "step": 16707
+    },
+    {
+      "epoch": 3.7972727272727274,
+      "grad_norm": 0.6506369550430326,
+      "learning_rate": 1.701896363823248e-07,
+      "loss": 0.002,
+      "step": 16708
+    },
+    {
+      "epoch": 3.7975,
+      "grad_norm": 1.4546004445001848,
+      "learning_rate": 1.7012842466107433e-07,
+      "loss": 0.0048,
+      "step": 16709
+    },
+    {
+      "epoch": 3.797727272727273,
+      "grad_norm": 1.6452617241409353,
+      "learning_rate": 1.7006722221544905e-07,
+      "loss": 0.0057,
+      "step": 16710
+    },
+    {
+      "epoch": 3.7979545454545454,
+      "grad_norm": 3.0158008579965436,
+      "learning_rate": 1.7000602904669658e-07,
+      "loss": 0.0132,
+      "step": 16711
+    },
+    {
+      "epoch": 3.7981818181818183,
+      "grad_norm": 1.8226451771293648,
+      "learning_rate": 1.6994484515606494e-07,
+      "loss": 0.014,
+      "step": 16712
+    },
+    {
+      "epoch": 3.798409090909091,
+      "grad_norm": 1.9292366730981518,
+      "learning_rate": 1.6988367054480175e-07,
+      "loss": 0.0138,
+      "step": 16713
+    },
+    {
+      "epoch": 3.798636363636364,
+      "grad_norm": 1.053763047866718,
+      "learning_rate": 1.6982250521415457e-07,
+      "loss": 0.0067,
+      "step": 16714
+    },
+    {
+      "epoch": 3.7988636363636363,
+      "grad_norm": 2.022946155800708,
+      "learning_rate": 1.697613491653708e-07,
+      "loss": 0.0233,
+      "step": 16715
+    },
+    {
+      "epoch": 3.799090909090909,
+      "grad_norm": 2.5916137593314734,
+      "learning_rate": 1.6970020239969715e-07,
+      "loss": 0.0161,
+      "step": 16716
+    },
+    {
+      "epoch": 3.799318181818182,
+      "grad_norm": 0.8249493500162869,
+      "learning_rate": 1.6963906491838065e-07,
+      "loss": 0.004,
+      "step": 16717
+    },
+    {
+      "epoch": 3.799545454545455,
+      "grad_norm": 0.23004562945706455,
+      "learning_rate": 1.6957793672266804e-07,
+      "loss": 0.0009,
+      "step": 16718
+    },
+    {
+      "epoch": 3.7997727272727273,
+      "grad_norm": 0.6230877735024023,
+      "learning_rate": 1.6951681781380603e-07,
+      "loss": 0.0038,
+      "step": 16719
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 2.857724902137792,
+      "learning_rate": 1.6945570819304056e-07,
+      "loss": 0.076,
+      "step": 16720
+    },
+    {
+      "epoch": 3.800227272727273,
+      "grad_norm": 2.3349721434624606,
+      "learning_rate": 1.6939460786161792e-07,
+      "loss": 0.0134,
+      "step": 16721
+    },
+    {
+      "epoch": 3.8004545454545453,
+      "grad_norm": 0.8069920601353238,
+      "learning_rate": 1.6933351682078408e-07,
+      "loss": 0.0066,
+      "step": 16722
+    },
+    {
+      "epoch": 3.8006818181818183,
+      "grad_norm": 1.3452957533497292,
+      "learning_rate": 1.6927243507178482e-07,
+      "loss": 0.007,
+      "step": 16723
+    },
+    {
+      "epoch": 3.800909090909091,
+      "grad_norm": 1.5197925193298398,
+      "learning_rate": 1.6921136261586562e-07,
+      "loss": 0.0114,
+      "step": 16724
+    },
+    {
+      "epoch": 3.8011363636363638,
+      "grad_norm": 1.8866426354617447,
+      "learning_rate": 1.6915029945427195e-07,
+      "loss": 0.0092,
+      "step": 16725
+    },
+    {
+      "epoch": 3.8013636363636363,
+      "grad_norm": 1.6633354791855113,
+      "learning_rate": 1.690892455882489e-07,
+      "loss": 0.0099,
+      "step": 16726
+    },
+    {
+      "epoch": 3.8015909090909092,
+      "grad_norm": 1.7039011316130086,
+      "learning_rate": 1.6902820101904152e-07,
+      "loss": 0.0038,
+      "step": 16727
+    },
+    {
+      "epoch": 3.8018181818181818,
+      "grad_norm": 2.3263877176565835,
+      "learning_rate": 1.689671657478948e-07,
+      "loss": 0.0155,
+      "step": 16728
+    },
+    {
+      "epoch": 3.8020454545454543,
+      "grad_norm": 0.6288229560552222,
+      "learning_rate": 1.6890613977605295e-07,
+      "loss": 0.0061,
+      "step": 16729
+    },
+    {
+      "epoch": 3.8022727272727272,
+      "grad_norm": 2.0240428055342097,
+      "learning_rate": 1.6884512310476066e-07,
+      "loss": 0.0093,
+      "step": 16730
+    },
+    {
+      "epoch": 3.8025,
+      "grad_norm": 1.421582087817697,
+      "learning_rate": 1.687841157352621e-07,
+      "loss": 0.0165,
+      "step": 16731
+    },
+    {
+      "epoch": 3.8027272727272727,
+      "grad_norm": 0.34858728922164367,
+      "learning_rate": 1.6872311766880151e-07,
+      "loss": 0.0037,
+      "step": 16732
+    },
+    {
+      "epoch": 3.8029545454545453,
+      "grad_norm": 0.8549170062700724,
+      "learning_rate": 1.6866212890662237e-07,
+      "loss": 0.0052,
+      "step": 16733
+    },
+    {
+      "epoch": 3.803181818181818,
+      "grad_norm": 2.83846439362155,
+      "learning_rate": 1.6860114944996862e-07,
+      "loss": 0.0276,
+      "step": 16734
+    },
+    {
+      "epoch": 3.803409090909091,
+      "grad_norm": 1.0587244592718343,
+      "learning_rate": 1.6854017930008362e-07,
+      "loss": 0.0062,
+      "step": 16735
+    },
+    {
+      "epoch": 3.8036363636363637,
+      "grad_norm": 1.096640709843035,
+      "learning_rate": 1.6847921845821068e-07,
+      "loss": 0.0057,
+      "step": 16736
+    },
+    {
+      "epoch": 3.8038636363636362,
+      "grad_norm": 1.3108482112471294,
+      "learning_rate": 1.6841826692559293e-07,
+      "loss": 0.0067,
+      "step": 16737
+    },
+    {
+      "epoch": 3.804090909090909,
+      "grad_norm": 0.8062551008032331,
+      "learning_rate": 1.6835732470347324e-07,
+      "loss": 0.0058,
+      "step": 16738
+    },
+    {
+      "epoch": 3.8043181818181817,
+      "grad_norm": 2.068008076099901,
+      "learning_rate": 1.6829639179309437e-07,
+      "loss": 0.0139,
+      "step": 16739
+    },
+    {
+      "epoch": 3.8045454545454547,
+      "grad_norm": 1.6370523025009727,
+      "learning_rate": 1.6823546819569886e-07,
+      "loss": 0.007,
+      "step": 16740
+    },
+    {
+      "epoch": 3.804772727272727,
+      "grad_norm": 0.7355257308214167,
+      "learning_rate": 1.6817455391252922e-07,
+      "loss": 0.0042,
+      "step": 16741
+    },
+    {
+      "epoch": 3.805,
+      "grad_norm": 0.8648324954344,
+      "learning_rate": 1.681136489448272e-07,
+      "loss": 0.0045,
+      "step": 16742
+    },
+    {
+      "epoch": 3.8052272727272727,
+      "grad_norm": 1.2691368153041118,
+      "learning_rate": 1.6805275329383497e-07,
+      "loss": 0.0086,
+      "step": 16743
+    },
+    {
+      "epoch": 3.8054545454545456,
+      "grad_norm": 3.0019133978641532,
+      "learning_rate": 1.6799186696079424e-07,
+      "loss": 0.0131,
+      "step": 16744
+    },
+    {
+      "epoch": 3.805681818181818,
+      "grad_norm": 0.7345733018484237,
+      "learning_rate": 1.6793098994694687e-07,
+      "loss": 0.0076,
+      "step": 16745
+    },
+    {
+      "epoch": 3.8059090909090907,
+      "grad_norm": 1.3986890747421177,
+      "learning_rate": 1.678701222535338e-07,
+      "loss": 0.0113,
+      "step": 16746
+    },
+    {
+      "epoch": 3.8061363636363637,
+      "grad_norm": 1.5791044942038182,
+      "learning_rate": 1.6780926388179652e-07,
+      "loss": 0.0122,
+      "step": 16747
+    },
+    {
+      "epoch": 3.8063636363636366,
+      "grad_norm": 2.975724121404483,
+      "learning_rate": 1.677484148329759e-07,
+      "loss": 0.0065,
+      "step": 16748
+    },
+    {
+      "epoch": 3.806590909090909,
+      "grad_norm": 1.7221293719724238,
+      "learning_rate": 1.6768757510831276e-07,
+      "loss": 0.0107,
+      "step": 16749
+    },
+    {
+      "epoch": 3.8068181818181817,
+      "grad_norm": 2.683056702839264,
+      "learning_rate": 1.6762674470904807e-07,
+      "loss": 0.0157,
+      "step": 16750
+    },
+    {
+      "epoch": 3.8070454545454546,
+      "grad_norm": 2.3661913781450314,
+      "learning_rate": 1.6756592363642174e-07,
+      "loss": 0.0212,
+      "step": 16751
+    },
+    {
+      "epoch": 3.807272727272727,
+      "grad_norm": 1.2484926257910582,
+      "learning_rate": 1.6750511189167428e-07,
+      "loss": 0.0041,
+      "step": 16752
+    },
+    {
+      "epoch": 3.8075,
+      "grad_norm": 1.000087052745099,
+      "learning_rate": 1.6744430947604558e-07,
+      "loss": 0.0102,
+      "step": 16753
+    },
+    {
+      "epoch": 3.8077272727272726,
+      "grad_norm": 2.1768688116539052,
+      "learning_rate": 1.6738351639077603e-07,
+      "loss": 0.0118,
+      "step": 16754
+    },
+    {
+      "epoch": 3.8079545454545456,
+      "grad_norm": 1.808360440429818,
+      "learning_rate": 1.673227326371047e-07,
+      "loss": 0.0212,
+      "step": 16755
+    },
+    {
+      "epoch": 3.808181818181818,
+      "grad_norm": 1.22790301728936,
+      "learning_rate": 1.6726195821627135e-07,
+      "loss": 0.0121,
+      "step": 16756
+    },
+    {
+      "epoch": 3.808409090909091,
+      "grad_norm": 1.492908438440112,
+      "learning_rate": 1.6720119312951526e-07,
+      "loss": 0.0099,
+      "step": 16757
+    },
+    {
+      "epoch": 3.8086363636363636,
+      "grad_norm": 1.324428094465958,
+      "learning_rate": 1.6714043737807564e-07,
+      "loss": 0.0064,
+      "step": 16758
+    },
+    {
+      "epoch": 3.808863636363636,
+      "grad_norm": 0.8689231705880611,
+      "learning_rate": 1.6707969096319118e-07,
+      "loss": 0.0045,
+      "step": 16759
+    },
+    {
+      "epoch": 3.809090909090909,
+      "grad_norm": 3.6360905825494507,
+      "learning_rate": 1.6701895388610063e-07,
+      "loss": 0.0566,
+      "step": 16760
+    },
+    {
+      "epoch": 3.809318181818182,
+      "grad_norm": 0.7478183699578981,
+      "learning_rate": 1.6695822614804262e-07,
+      "loss": 0.0079,
+      "step": 16761
+    },
+    {
+      "epoch": 3.8095454545454546,
+      "grad_norm": 0.5254781169135814,
+      "learning_rate": 1.668975077502555e-07,
+      "loss": 0.0055,
+      "step": 16762
+    },
+    {
+      "epoch": 3.809772727272727,
+      "grad_norm": 1.0993588449220024,
+      "learning_rate": 1.6683679869397757e-07,
+      "loss": 0.0062,
+      "step": 16763
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.3238570915584381,
+      "learning_rate": 1.667760989804465e-07,
+      "loss": 0.0026,
+      "step": 16764
+    },
+    {
+      "epoch": 3.8102272727272726,
+      "grad_norm": 3.6465483977264617,
+      "learning_rate": 1.667154086109001e-07,
+      "loss": 0.0496,
+      "step": 16765
+    },
+    {
+      "epoch": 3.8104545454545455,
+      "grad_norm": 0.9511684080051599,
+      "learning_rate": 1.6665472758657603e-07,
+      "loss": 0.0074,
+      "step": 16766
+    },
+    {
+      "epoch": 3.810681818181818,
+      "grad_norm": 1.3639833093541711,
+      "learning_rate": 1.6659405590871175e-07,
+      "loss": 0.0083,
+      "step": 16767
+    },
+    {
+      "epoch": 3.810909090909091,
+      "grad_norm": 1.0203903346037517,
+      "learning_rate": 1.6653339357854438e-07,
+      "loss": 0.0032,
+      "step": 16768
+    },
+    {
+      "epoch": 3.8111363636363635,
+      "grad_norm": 2.1865946703074304,
+      "learning_rate": 1.66472740597311e-07,
+      "loss": 0.0118,
+      "step": 16769
+    },
+    {
+      "epoch": 3.8113636363636365,
+      "grad_norm": 3.8171336826457267,
+      "learning_rate": 1.6641209696624833e-07,
+      "loss": 0.0179,
+      "step": 16770
+    },
+    {
+      "epoch": 3.811590909090909,
+      "grad_norm": 1.1857124616642347,
+      "learning_rate": 1.663514626865931e-07,
+      "loss": 0.015,
+      "step": 16771
+    },
+    {
+      "epoch": 3.8118181818181816,
+      "grad_norm": 3.5883328961645855,
+      "learning_rate": 1.6629083775958183e-07,
+      "loss": 0.0084,
+      "step": 16772
+    },
+    {
+      "epoch": 3.8120454545454545,
+      "grad_norm": 1.8300572891641051,
+      "learning_rate": 1.662302221864505e-07,
+      "loss": 0.0178,
+      "step": 16773
+    },
+    {
+      "epoch": 3.8122727272727275,
+      "grad_norm": 0.7998137370435652,
+      "learning_rate": 1.6616961596843526e-07,
+      "loss": 0.0112,
+      "step": 16774
+    },
+    {
+      "epoch": 3.8125,
+      "grad_norm": 1.4614312758952057,
+      "learning_rate": 1.6610901910677206e-07,
+      "loss": 0.0148,
+      "step": 16775
+    },
+    {
+      "epoch": 3.8127272727272725,
+      "grad_norm": 0.8165972142024985,
+      "learning_rate": 1.660484316026967e-07,
+      "loss": 0.0063,
+      "step": 16776
+    },
+    {
+      "epoch": 3.8129545454545455,
+      "grad_norm": 1.0154959185461616,
+      "learning_rate": 1.659878534574443e-07,
+      "loss": 0.0047,
+      "step": 16777
+    },
+    {
+      "epoch": 3.8131818181818184,
+      "grad_norm": 0.6166974626419377,
+      "learning_rate": 1.659272846722504e-07,
+      "loss": 0.0063,
+      "step": 16778
+    },
+    {
+      "epoch": 3.813409090909091,
+      "grad_norm": 0.8868250464554329,
+      "learning_rate": 1.6586672524835008e-07,
+      "loss": 0.0056,
+      "step": 16779
+    },
+    {
+      "epoch": 3.8136363636363635,
+      "grad_norm": 1.9527781215007098,
+      "learning_rate": 1.6580617518697834e-07,
+      "loss": 0.0116,
+      "step": 16780
+    },
+    {
+      "epoch": 3.8138636363636365,
+      "grad_norm": 1.7775514873065519,
+      "learning_rate": 1.6574563448936971e-07,
+      "loss": 0.0094,
+      "step": 16781
+    },
+    {
+      "epoch": 3.814090909090909,
+      "grad_norm": 2.0904966064495407,
+      "learning_rate": 1.6568510315675856e-07,
+      "loss": 0.0114,
+      "step": 16782
+    },
+    {
+      "epoch": 3.814318181818182,
+      "grad_norm": 0.5372186857356389,
+      "learning_rate": 1.656245811903797e-07,
+      "loss": 0.0026,
+      "step": 16783
+    },
+    {
+      "epoch": 3.8145454545454545,
+      "grad_norm": 0.735261103832733,
+      "learning_rate": 1.6556406859146703e-07,
+      "loss": 0.0056,
+      "step": 16784
+    },
+    {
+      "epoch": 3.8147727272727274,
+      "grad_norm": 0.8872230832458161,
+      "learning_rate": 1.6550356536125466e-07,
+      "loss": 0.007,
+      "step": 16785
+    },
+    {
+      "epoch": 3.815,
+      "grad_norm": 1.4205161703296338,
+      "learning_rate": 1.654430715009761e-07,
+      "loss": 0.0123,
+      "step": 16786
+    },
+    {
+      "epoch": 3.815227272727273,
+      "grad_norm": 1.3663450443192833,
+      "learning_rate": 1.6538258701186506e-07,
+      "loss": 0.0086,
+      "step": 16787
+    },
+    {
+      "epoch": 3.8154545454545454,
+      "grad_norm": 1.754634133201407,
+      "learning_rate": 1.6532211189515494e-07,
+      "loss": 0.0178,
+      "step": 16788
+    },
+    {
+      "epoch": 3.815681818181818,
+      "grad_norm": 1.3114433368581342,
+      "learning_rate": 1.6526164615207902e-07,
+      "loss": 0.0087,
+      "step": 16789
+    },
+    {
+      "epoch": 3.815909090909091,
+      "grad_norm": 1.3514086758099508,
+      "learning_rate": 1.6520118978387012e-07,
+      "loss": 0.011,
+      "step": 16790
+    },
+    {
+      "epoch": 3.816136363636364,
+      "grad_norm": 1.6473947299842875,
+      "learning_rate": 1.6514074279176103e-07,
+      "loss": 0.0107,
+      "step": 16791
+    },
+    {
+      "epoch": 3.8163636363636364,
+      "grad_norm": 2.3520689107905572,
+      "learning_rate": 1.6508030517698448e-07,
+      "loss": 0.0135,
+      "step": 16792
+    },
+    {
+      "epoch": 3.816590909090909,
+      "grad_norm": 2.189870159760885,
+      "learning_rate": 1.6501987694077287e-07,
+      "loss": 0.0141,
+      "step": 16793
+    },
+    {
+      "epoch": 3.816818181818182,
+      "grad_norm": 1.3222368037243422,
+      "learning_rate": 1.6495945808435867e-07,
+      "loss": 0.008,
+      "step": 16794
+    },
+    {
+      "epoch": 3.8170454545454544,
+      "grad_norm": 6.275755203128111,
+      "learning_rate": 1.6489904860897348e-07,
+      "loss": 0.0234,
+      "step": 16795
+    },
+    {
+      "epoch": 3.8172727272727274,
+      "grad_norm": 0.772442058785164,
+      "learning_rate": 1.6483864851584945e-07,
+      "loss": 0.0025,
+      "step": 16796
+    },
+    {
+      "epoch": 3.8175,
+      "grad_norm": 4.001092495612939,
+      "learning_rate": 1.6477825780621814e-07,
+      "loss": 0.0137,
+      "step": 16797
+    },
+    {
+      "epoch": 3.817727272727273,
+      "grad_norm": 1.3037554835447263,
+      "learning_rate": 1.64717876481311e-07,
+      "loss": 0.0045,
+      "step": 16798
+    },
+    {
+      "epoch": 3.8179545454545454,
+      "grad_norm": 2.569978798006225,
+      "learning_rate": 1.6465750454235942e-07,
+      "loss": 0.0355,
+      "step": 16799
+    },
+    {
+      "epoch": 3.8181818181818183,
+      "grad_norm": 1.1225295461945104,
+      "learning_rate": 1.6459714199059437e-07,
+      "loss": 0.008,
+      "step": 16800
+    },
+    {
+      "epoch": 3.818409090909091,
+      "grad_norm": 1.0608673161344535,
+      "learning_rate": 1.6453678882724685e-07,
+      "loss": 0.0053,
+      "step": 16801
+    },
+    {
+      "epoch": 3.8186363636363634,
+      "grad_norm": 1.3886740989824542,
+      "learning_rate": 1.6447644505354778e-07,
+      "loss": 0.008,
+      "step": 16802
+    },
+    {
+      "epoch": 3.8188636363636363,
+      "grad_norm": 2.2613697243632855,
+      "learning_rate": 1.6441611067072718e-07,
+      "loss": 0.0239,
+      "step": 16803
+    },
+    {
+      "epoch": 3.8190909090909093,
+      "grad_norm": 0.6487234938353239,
+      "learning_rate": 1.6435578568001555e-07,
+      "loss": 0.0026,
+      "step": 16804
+    },
+    {
+      "epoch": 3.819318181818182,
+      "grad_norm": 4.702591203047617,
+      "learning_rate": 1.6429547008264317e-07,
+      "loss": 0.047,
+      "step": 16805
+    },
+    {
+      "epoch": 3.8195454545454544,
+      "grad_norm": 1.981321114667569,
+      "learning_rate": 1.6423516387983993e-07,
+      "loss": 0.0117,
+      "step": 16806
+    },
+    {
+      "epoch": 3.8197727272727273,
+      "grad_norm": 0.4001093543647472,
+      "learning_rate": 1.641748670728357e-07,
+      "loss": 0.0019,
+      "step": 16807
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 0.5826147224606816,
+      "learning_rate": 1.6411457966285965e-07,
+      "loss": 0.0022,
+      "step": 16808
+    },
+    {
+      "epoch": 3.820227272727273,
+      "grad_norm": 2.93743482725164,
+      "learning_rate": 1.6405430165114139e-07,
+      "loss": 0.0138,
+      "step": 16809
+    },
+    {
+      "epoch": 3.8204545454545453,
+      "grad_norm": 1.8344511282626956,
+      "learning_rate": 1.6399403303891012e-07,
+      "loss": 0.0128,
+      "step": 16810
+    },
+    {
+      "epoch": 3.8206818181818183,
+      "grad_norm": 1.231060319115169,
+      "learning_rate": 1.6393377382739495e-07,
+      "loss": 0.0044,
+      "step": 16811
+    },
+    {
+      "epoch": 3.820909090909091,
+      "grad_norm": 2.0255828972329506,
+      "learning_rate": 1.6387352401782416e-07,
+      "loss": 0.0079,
+      "step": 16812
+    },
+    {
+      "epoch": 3.8211363636363638,
+      "grad_norm": 1.112669629728121,
+      "learning_rate": 1.638132836114268e-07,
+      "loss": 0.0067,
+      "step": 16813
+    },
+    {
+      "epoch": 3.8213636363636363,
+      "grad_norm": 2.5646852685550763,
+      "learning_rate": 1.6375305260943129e-07,
+      "loss": 0.0091,
+      "step": 16814
+    },
+    {
+      "epoch": 3.821590909090909,
+      "grad_norm": 1.782274230878372,
+      "learning_rate": 1.6369283101306567e-07,
+      "loss": 0.011,
+      "step": 16815
+    },
+    {
+      "epoch": 3.821818181818182,
+      "grad_norm": 3.763688311309177,
+      "learning_rate": 1.6363261882355818e-07,
+      "loss": 0.0105,
+      "step": 16816
+    },
+    {
+      "epoch": 3.8220454545454547,
+      "grad_norm": 1.3664081238288441,
+      "learning_rate": 1.6357241604213635e-07,
+      "loss": 0.0062,
+      "step": 16817
+    },
+    {
+      "epoch": 3.8222727272727273,
+      "grad_norm": 1.1629363468050056,
+      "learning_rate": 1.6351222267002794e-07,
+      "loss": 0.0066,
+      "step": 16818
+    },
+    {
+      "epoch": 3.8225,
+      "grad_norm": 0.511701397838011,
+      "learning_rate": 1.6345203870846037e-07,
+      "loss": 0.0033,
+      "step": 16819
+    },
+    {
+      "epoch": 3.8227272727272728,
+      "grad_norm": 1.0112544332611135,
+      "learning_rate": 1.633918641586612e-07,
+      "loss": 0.0052,
+      "step": 16820
+    },
+    {
+      "epoch": 3.8229545454545457,
+      "grad_norm": 1.9967100929520698,
+      "learning_rate": 1.6333169902185704e-07,
+      "loss": 0.0114,
+      "step": 16821
+    },
+    {
+      "epoch": 3.8231818181818182,
+      "grad_norm": 2.5888879996222087,
+      "learning_rate": 1.6327154329927492e-07,
+      "loss": 0.0164,
+      "step": 16822
+    },
+    {
+      "epoch": 3.8234090909090908,
+      "grad_norm": 1.9223072524306073,
+      "learning_rate": 1.6321139699214165e-07,
+      "loss": 0.01,
+      "step": 16823
+    },
+    {
+      "epoch": 3.8236363636363637,
+      "grad_norm": 0.6484381362522641,
+      "learning_rate": 1.6315126010168373e-07,
+      "loss": 0.0038,
+      "step": 16824
+    },
+    {
+      "epoch": 3.8238636363636362,
+      "grad_norm": 1.0521016375611556,
+      "learning_rate": 1.630911326291272e-07,
+      "loss": 0.0056,
+      "step": 16825
+    },
+    {
+      "epoch": 3.824090909090909,
+      "grad_norm": 0.978223614411185,
+      "learning_rate": 1.6303101457569838e-07,
+      "loss": 0.0046,
+      "step": 16826
+    },
+    {
+      "epoch": 3.8243181818181817,
+      "grad_norm": 4.32984461956883,
+      "learning_rate": 1.62970905942623e-07,
+      "loss": 0.0101,
+      "step": 16827
+    },
+    {
+      "epoch": 3.8245454545454547,
+      "grad_norm": 0.9932042602413569,
+      "learning_rate": 1.6291080673112698e-07,
+      "loss": 0.0058,
+      "step": 16828
+    },
+    {
+      "epoch": 3.824772727272727,
+      "grad_norm": 1.1570459827584796,
+      "learning_rate": 1.6285071694243579e-07,
+      "loss": 0.0072,
+      "step": 16829
+    },
+    {
+      "epoch": 3.825,
+      "grad_norm": 0.6252414460104033,
+      "learning_rate": 1.627906365777747e-07,
+      "loss": 0.0044,
+      "step": 16830
+    },
+    {
+      "epoch": 3.8252272727272727,
+      "grad_norm": 1.304594385909255,
+      "learning_rate": 1.6273056563836888e-07,
+      "loss": 0.0097,
+      "step": 16831
+    },
+    {
+      "epoch": 3.825454545454545,
+      "grad_norm": 1.3036379986294164,
+      "learning_rate": 1.626705041254434e-07,
+      "loss": 0.0217,
+      "step": 16832
+    },
+    {
+      "epoch": 3.825681818181818,
+      "grad_norm": 1.220052818499068,
+      "learning_rate": 1.6261045204022302e-07,
+      "loss": 0.0149,
+      "step": 16833
+    },
+    {
+      "epoch": 3.825909090909091,
+      "grad_norm": 1.8427372719344988,
+      "learning_rate": 1.6255040938393204e-07,
+      "loss": 0.0319,
+      "step": 16834
+    },
+    {
+      "epoch": 3.8261363636363637,
+      "grad_norm": 0.6438811540745386,
+      "learning_rate": 1.62490376157795e-07,
+      "loss": 0.0041,
+      "step": 16835
+    },
+    {
+      "epoch": 3.826363636363636,
+      "grad_norm": 0.874631724634645,
+      "learning_rate": 1.6243035236303606e-07,
+      "loss": 0.0036,
+      "step": 16836
+    },
+    {
+      "epoch": 3.826590909090909,
+      "grad_norm": 1.3083600537635973,
+      "learning_rate": 1.6237033800087925e-07,
+      "loss": 0.0089,
+      "step": 16837
+    },
+    {
+      "epoch": 3.8268181818181817,
+      "grad_norm": 1.4177436949145825,
+      "learning_rate": 1.6231033307254855e-07,
+      "loss": 0.0079,
+      "step": 16838
+    },
+    {
+      "epoch": 3.8270454545454546,
+      "grad_norm": 1.3602092665401053,
+      "learning_rate": 1.6225033757926716e-07,
+      "loss": 0.0085,
+      "step": 16839
+    },
+    {
+      "epoch": 3.827272727272727,
+      "grad_norm": 2.8875375501772176,
+      "learning_rate": 1.621903515222587e-07,
+      "loss": 0.0184,
+      "step": 16840
+    },
+    {
+      "epoch": 3.8275,
+      "grad_norm": 1.0798844592743801,
+      "learning_rate": 1.6213037490274642e-07,
+      "loss": 0.0125,
+      "step": 16841
+    },
+    {
+      "epoch": 3.8277272727272726,
+      "grad_norm": 1.5309131130527507,
+      "learning_rate": 1.6207040772195322e-07,
+      "loss": 0.0171,
+      "step": 16842
+    },
+    {
+      "epoch": 3.8279545454545456,
+      "grad_norm": 2.166637860762877,
+      "learning_rate": 1.620104499811021e-07,
+      "loss": 0.0119,
+      "step": 16843
+    },
+    {
+      "epoch": 3.828181818181818,
+      "grad_norm": 1.3521838185444353,
+      "learning_rate": 1.6195050168141567e-07,
+      "loss": 0.006,
+      "step": 16844
+    },
+    {
+      "epoch": 3.8284090909090907,
+      "grad_norm": 1.9233434252565171,
+      "learning_rate": 1.6189056282411631e-07,
+      "loss": 0.0154,
+      "step": 16845
+    },
+    {
+      "epoch": 3.8286363636363636,
+      "grad_norm": 0.9462992320879943,
+      "learning_rate": 1.618306334104265e-07,
+      "loss": 0.0069,
+      "step": 16846
+    },
+    {
+      "epoch": 3.8288636363636366,
+      "grad_norm": 2.4924672550578824,
+      "learning_rate": 1.617707134415679e-07,
+      "loss": 0.03,
+      "step": 16847
+    },
+    {
+      "epoch": 3.829090909090909,
+      "grad_norm": 1.672733995907141,
+      "learning_rate": 1.6171080291876264e-07,
+      "loss": 0.0077,
+      "step": 16848
+    },
+    {
+      "epoch": 3.8293181818181816,
+      "grad_norm": 1.0850344900452735,
+      "learning_rate": 1.616509018432324e-07,
+      "loss": 0.0074,
+      "step": 16849
+    },
+    {
+      "epoch": 3.8295454545454546,
+      "grad_norm": 2.8825096441149247,
+      "learning_rate": 1.6159101021619856e-07,
+      "loss": 0.027,
+      "step": 16850
+    },
+    {
+      "epoch": 3.829772727272727,
+      "grad_norm": 1.7823873273205877,
+      "learning_rate": 1.6153112803888266e-07,
+      "loss": 0.0225,
+      "step": 16851
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 0.9924635282174569,
+      "learning_rate": 1.614712553125055e-07,
+      "loss": 0.0073,
+      "step": 16852
+    },
+    {
+      "epoch": 3.8302272727272726,
+      "grad_norm": 1.7973479396675325,
+      "learning_rate": 1.614113920382881e-07,
+      "loss": 0.0116,
+      "step": 16853
+    },
+    {
+      "epoch": 3.8304545454545456,
+      "grad_norm": 2.7105048517165122,
+      "learning_rate": 1.613515382174511e-07,
+      "loss": 0.019,
+      "step": 16854
+    },
+    {
+      "epoch": 3.830681818181818,
+      "grad_norm": 0.9398645325389346,
+      "learning_rate": 1.612916938512154e-07,
+      "loss": 0.0083,
+      "step": 16855
+    },
+    {
+      "epoch": 3.830909090909091,
+      "grad_norm": 3.6456708430967333,
+      "learning_rate": 1.612318589408008e-07,
+      "loss": 0.0201,
+      "step": 16856
+    },
+    {
+      "epoch": 3.8311363636363636,
+      "grad_norm": 1.491462952496836,
+      "learning_rate": 1.6117203348742775e-07,
+      "loss": 0.0167,
+      "step": 16857
+    },
+    {
+      "epoch": 3.8313636363636365,
+      "grad_norm": 0.7161550927912246,
+      "learning_rate": 1.611122174923161e-07,
+      "loss": 0.0041,
+      "step": 16858
+    },
+    {
+      "epoch": 3.831590909090909,
+      "grad_norm": 1.014090480246753,
+      "learning_rate": 1.6105241095668563e-07,
+      "loss": 0.005,
+      "step": 16859
+    },
+    {
+      "epoch": 3.831818181818182,
+      "grad_norm": 1.3725646271450462,
+      "learning_rate": 1.6099261388175586e-07,
+      "loss": 0.0089,
+      "step": 16860
+    },
+    {
+      "epoch": 3.8320454545454545,
+      "grad_norm": 2.51557044397012,
+      "learning_rate": 1.609328262687463e-07,
+      "loss": 0.0209,
+      "step": 16861
+    },
+    {
+      "epoch": 3.832272727272727,
+      "grad_norm": 0.7446317583225351,
+      "learning_rate": 1.6087304811887597e-07,
+      "loss": 0.0068,
+      "step": 16862
+    },
+    {
+      "epoch": 3.8325,
+      "grad_norm": 1.8243434940746568,
+      "learning_rate": 1.608132794333639e-07,
+      "loss": 0.0122,
+      "step": 16863
+    },
+    {
+      "epoch": 3.832727272727273,
+      "grad_norm": 1.4668132525263653,
+      "learning_rate": 1.6075352021342903e-07,
+      "loss": 0.0163,
+      "step": 16864
+    },
+    {
+      "epoch": 3.8329545454545455,
+      "grad_norm": 1.6863878192269257,
+      "learning_rate": 1.6069377046028962e-07,
+      "loss": 0.0067,
+      "step": 16865
+    },
+    {
+      "epoch": 3.833181818181818,
+      "grad_norm": 0.890011426383351,
+      "learning_rate": 1.606340301751643e-07,
+      "loss": 0.0062,
+      "step": 16866
+    },
+    {
+      "epoch": 3.833409090909091,
+      "grad_norm": 0.9689827035203237,
+      "learning_rate": 1.605742993592712e-07,
+      "loss": 0.0041,
+      "step": 16867
+    },
+    {
+      "epoch": 3.8336363636363635,
+      "grad_norm": 1.5205864431441793,
+      "learning_rate": 1.6051457801382862e-07,
+      "loss": 0.0117,
+      "step": 16868
+    },
+    {
+      "epoch": 3.8338636363636365,
+      "grad_norm": 2.790018340165025,
+      "learning_rate": 1.6045486614005386e-07,
+      "loss": 0.0044,
+      "step": 16869
+    },
+    {
+      "epoch": 3.834090909090909,
+      "grad_norm": 1.1738366228170471,
+      "learning_rate": 1.6039516373916493e-07,
+      "loss": 0.0057,
+      "step": 16870
+    },
+    {
+      "epoch": 3.834318181818182,
+      "grad_norm": 1.7395548385647068,
+      "learning_rate": 1.6033547081237908e-07,
+      "loss": 0.0115,
+      "step": 16871
+    },
+    {
+      "epoch": 3.8345454545454545,
+      "grad_norm": 1.7318267912746002,
+      "learning_rate": 1.6027578736091364e-07,
+      "loss": 0.0096,
+      "step": 16872
+    },
+    {
+      "epoch": 3.8347727272727274,
+      "grad_norm": 2.777007840042001,
+      "learning_rate": 1.6021611338598566e-07,
+      "loss": 0.0138,
+      "step": 16873
+    },
+    {
+      "epoch": 3.835,
+      "grad_norm": 1.2723345177424519,
+      "learning_rate": 1.60156448888812e-07,
+      "loss": 0.0069,
+      "step": 16874
+    },
+    {
+      "epoch": 3.8352272727272725,
+      "grad_norm": 1.6445058157831542,
+      "learning_rate": 1.600967938706093e-07,
+      "loss": 0.0129,
+      "step": 16875
+    },
+    {
+      "epoch": 3.8354545454545454,
+      "grad_norm": 3.7426758336548773,
+      "learning_rate": 1.6003714833259402e-07,
+      "loss": 0.0142,
+      "step": 16876
+    },
+    {
+      "epoch": 3.8356818181818184,
+      "grad_norm": 1.2167321727574492,
+      "learning_rate": 1.5997751227598267e-07,
+      "loss": 0.0062,
+      "step": 16877
+    },
+    {
+      "epoch": 3.835909090909091,
+      "grad_norm": 2.3821966926136637,
+      "learning_rate": 1.5991788570199097e-07,
+      "loss": 0.0138,
+      "step": 16878
+    },
+    {
+      "epoch": 3.8361363636363635,
+      "grad_norm": 2.003290682439299,
+      "learning_rate": 1.598582686118349e-07,
+      "loss": 0.0066,
+      "step": 16879
+    },
+    {
+      "epoch": 3.8363636363636364,
+      "grad_norm": 1.6157829929159304,
+      "learning_rate": 1.5979866100673023e-07,
+      "loss": 0.0105,
+      "step": 16880
+    },
+    {
+      "epoch": 3.836590909090909,
+      "grad_norm": 1.8974941260963136,
+      "learning_rate": 1.5973906288789266e-07,
+      "loss": 0.0182,
+      "step": 16881
+    },
+    {
+      "epoch": 3.836818181818182,
+      "grad_norm": 2.3816970253948195,
+      "learning_rate": 1.5967947425653712e-07,
+      "loss": 0.0149,
+      "step": 16882
+    },
+    {
+      "epoch": 3.8370454545454544,
+      "grad_norm": 0.5779562386487758,
+      "learning_rate": 1.596198951138788e-07,
+      "loss": 0.0019,
+      "step": 16883
+    },
+    {
+      "epoch": 3.8372727272727274,
+      "grad_norm": 0.584992000408535,
+      "learning_rate": 1.5956032546113283e-07,
+      "loss": 0.0058,
+      "step": 16884
+    },
+    {
+      "epoch": 3.8375,
+      "grad_norm": 0.9367628411798031,
+      "learning_rate": 1.5950076529951375e-07,
+      "loss": 0.0036,
+      "step": 16885
+    },
+    {
+      "epoch": 3.837727272727273,
+      "grad_norm": 1.6168641520247733,
+      "learning_rate": 1.594412146302364e-07,
+      "loss": 0.0089,
+      "step": 16886
+    },
+    {
+      "epoch": 3.8379545454545454,
+      "grad_norm": 0.916987981577565,
+      "learning_rate": 1.593816734545147e-07,
+      "loss": 0.0055,
+      "step": 16887
+    },
+    {
+      "epoch": 3.838181818181818,
+      "grad_norm": 1.8315331670919868,
+      "learning_rate": 1.59322141773563e-07,
+      "loss": 0.0108,
+      "step": 16888
+    },
+    {
+      "epoch": 3.838409090909091,
+      "grad_norm": 2.3472618459096015,
+      "learning_rate": 1.5926261958859524e-07,
+      "loss": 0.0146,
+      "step": 16889
+    },
+    {
+      "epoch": 3.838636363636364,
+      "grad_norm": 1.5208181022343015,
+      "learning_rate": 1.592031069008252e-07,
+      "loss": 0.006,
+      "step": 16890
+    },
+    {
+      "epoch": 3.8388636363636364,
+      "grad_norm": 2.2818616884761505,
+      "learning_rate": 1.5914360371146644e-07,
+      "loss": 0.0072,
+      "step": 16891
+    },
+    {
+      "epoch": 3.839090909090909,
+      "grad_norm": 0.7606111044446355,
+      "learning_rate": 1.5908411002173232e-07,
+      "loss": 0.0031,
+      "step": 16892
+    },
+    {
+      "epoch": 3.839318181818182,
+      "grad_norm": 2.478987744468982,
+      "learning_rate": 1.59024625832836e-07,
+      "loss": 0.0153,
+      "step": 16893
+    },
+    {
+      "epoch": 3.839545454545455,
+      "grad_norm": 2.0706375717852787,
+      "learning_rate": 1.5896515114599054e-07,
+      "loss": 0.0083,
+      "step": 16894
+    },
+    {
+      "epoch": 3.8397727272727273,
+      "grad_norm": 3.0856790522216033,
+      "learning_rate": 1.5890568596240884e-07,
+      "loss": 0.0185,
+      "step": 16895
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 1.6015569228250697,
+      "learning_rate": 1.5884623028330325e-07,
+      "loss": 0.011,
+      "step": 16896
+    },
+    {
+      "epoch": 3.840227272727273,
+      "grad_norm": 1.849184481264805,
+      "learning_rate": 1.587867841098862e-07,
+      "loss": 0.0122,
+      "step": 16897
+    },
+    {
+      "epoch": 3.8404545454545453,
+      "grad_norm": 2.571010906641081,
+      "learning_rate": 1.5872734744337003e-07,
+      "loss": 0.0184,
+      "step": 16898
+    },
+    {
+      "epoch": 3.8406818181818183,
+      "grad_norm": 1.6203432434616831,
+      "learning_rate": 1.5866792028496682e-07,
+      "loss": 0.0082,
+      "step": 16899
+    },
+    {
+      "epoch": 3.840909090909091,
+      "grad_norm": 1.230176801414878,
+      "learning_rate": 1.5860850263588816e-07,
+      "loss": 0.0102,
+      "step": 16900
+    },
+    {
+      "epoch": 3.841136363636364,
+      "grad_norm": 2.8255869762143893,
+      "learning_rate": 1.5854909449734571e-07,
+      "loss": 0.0195,
+      "step": 16901
+    },
+    {
+      "epoch": 3.8413636363636363,
+      "grad_norm": 3.243524147959681,
+      "learning_rate": 1.5848969587055105e-07,
+      "loss": 0.0131,
+      "step": 16902
+    },
+    {
+      "epoch": 3.8415909090909093,
+      "grad_norm": 0.9268978443204294,
+      "learning_rate": 1.5843030675671535e-07,
+      "loss": 0.0046,
+      "step": 16903
+    },
+    {
+      "epoch": 3.841818181818182,
+      "grad_norm": 1.2918471534083449,
+      "learning_rate": 1.5837092715704966e-07,
+      "loss": 0.0097,
+      "step": 16904
+    },
+    {
+      "epoch": 3.8420454545454543,
+      "grad_norm": 1.152689057250076,
+      "learning_rate": 1.5831155707276487e-07,
+      "loss": 0.0079,
+      "step": 16905
+    },
+    {
+      "epoch": 3.8422727272727273,
+      "grad_norm": 1.182890999744895,
+      "learning_rate": 1.582521965050716e-07,
+      "loss": 0.0081,
+      "step": 16906
+    },
+    {
+      "epoch": 3.8425000000000002,
+      "grad_norm": 1.7172636045056702,
+      "learning_rate": 1.581928454551803e-07,
+      "loss": 0.01,
+      "step": 16907
+    },
+    {
+      "epoch": 3.8427272727272728,
+      "grad_norm": 1.7689996479782892,
+      "learning_rate": 1.581335039243015e-07,
+      "loss": 0.0079,
+      "step": 16908
+    },
+    {
+      "epoch": 3.8429545454545453,
+      "grad_norm": 1.058898819206668,
+      "learning_rate": 1.5807417191364486e-07,
+      "loss": 0.0046,
+      "step": 16909
+    },
+    {
+      "epoch": 3.8431818181818183,
+      "grad_norm": 1.4417480000398784,
+      "learning_rate": 1.5801484942442048e-07,
+      "loss": 0.0058,
+      "step": 16910
+    },
+    {
+      "epoch": 3.8434090909090908,
+      "grad_norm": 2.2893309852814836,
+      "learning_rate": 1.5795553645783793e-07,
+      "loss": 0.0156,
+      "step": 16911
+    },
+    {
+      "epoch": 3.8436363636363637,
+      "grad_norm": 1.7824841473974224,
+      "learning_rate": 1.5789623301510702e-07,
+      "loss": 0.0159,
+      "step": 16912
+    },
+    {
+      "epoch": 3.8438636363636363,
+      "grad_norm": 1.9549999931892323,
+      "learning_rate": 1.5783693909743664e-07,
+      "loss": 0.0122,
+      "step": 16913
+    },
+    {
+      "epoch": 3.8440909090909092,
+      "grad_norm": 1.2814687597362608,
+      "learning_rate": 1.5777765470603612e-07,
+      "loss": 0.0051,
+      "step": 16914
+    },
+    {
+      "epoch": 3.8443181818181817,
+      "grad_norm": 2.4942213000073723,
+      "learning_rate": 1.5771837984211431e-07,
+      "loss": 0.0252,
+      "step": 16915
+    },
+    {
+      "epoch": 3.8445454545454547,
+      "grad_norm": 1.4881883194973458,
+      "learning_rate": 1.5765911450687996e-07,
+      "loss": 0.0101,
+      "step": 16916
+    },
+    {
+      "epoch": 3.8447727272727272,
+      "grad_norm": 3.1302631224999726,
+      "learning_rate": 1.5759985870154177e-07,
+      "loss": 0.0106,
+      "step": 16917
+    },
+    {
+      "epoch": 3.8449999999999998,
+      "grad_norm": 4.53758023020272,
+      "learning_rate": 1.5754061242730768e-07,
+      "loss": 0.0115,
+      "step": 16918
+    },
+    {
+      "epoch": 3.8452272727272727,
+      "grad_norm": 1.172054425861869,
+      "learning_rate": 1.574813756853861e-07,
+      "loss": 0.0048,
+      "step": 16919
+    },
+    {
+      "epoch": 3.8454545454545457,
+      "grad_norm": 1.0887517199526604,
+      "learning_rate": 1.574221484769848e-07,
+      "loss": 0.0047,
+      "step": 16920
+    },
+    {
+      "epoch": 3.845681818181818,
+      "grad_norm": 1.1671578208111058,
+      "learning_rate": 1.5736293080331166e-07,
+      "loss": 0.0056,
+      "step": 16921
+    },
+    {
+      "epoch": 3.8459090909090907,
+      "grad_norm": 1.2309800788133933,
+      "learning_rate": 1.5730372266557423e-07,
+      "loss": 0.0065,
+      "step": 16922
+    },
+    {
+      "epoch": 3.8461363636363637,
+      "grad_norm": 2.0629184526480286,
+      "learning_rate": 1.572445240649798e-07,
+      "loss": 0.0098,
+      "step": 16923
+    },
+    {
+      "epoch": 3.846363636363636,
+      "grad_norm": 1.8207311400397557,
+      "learning_rate": 1.571853350027356e-07,
+      "loss": 0.0149,
+      "step": 16924
+    },
+    {
+      "epoch": 3.846590909090909,
+      "grad_norm": 1.5041334120381562,
+      "learning_rate": 1.571261554800487e-07,
+      "loss": 0.02,
+      "step": 16925
+    },
+    {
+      "epoch": 3.8468181818181817,
+      "grad_norm": 1.36382176348304,
+      "learning_rate": 1.5706698549812564e-07,
+      "loss": 0.0205,
+      "step": 16926
+    },
+    {
+      "epoch": 3.8470454545454547,
+      "grad_norm": 2.3865573437774388,
+      "learning_rate": 1.5700782505817303e-07,
+      "loss": 0.0118,
+      "step": 16927
+    },
+    {
+      "epoch": 3.847272727272727,
+      "grad_norm": 2.4526116430562594,
+      "learning_rate": 1.5694867416139736e-07,
+      "loss": 0.0131,
+      "step": 16928
+    },
+    {
+      "epoch": 3.8475,
+      "grad_norm": 2.560740988328935,
+      "learning_rate": 1.5688953280900476e-07,
+      "loss": 0.0305,
+      "step": 16929
+    },
+    {
+      "epoch": 3.8477272727272727,
+      "grad_norm": 0.9362272852490529,
+      "learning_rate": 1.5683040100220142e-07,
+      "loss": 0.0022,
+      "step": 16930
+    },
+    {
+      "epoch": 3.847954545454545,
+      "grad_norm": 0.41257020116805637,
+      "learning_rate": 1.5677127874219282e-07,
+      "loss": 0.002,
+      "step": 16931
+    },
+    {
+      "epoch": 3.848181818181818,
+      "grad_norm": 1.19338498790489,
+      "learning_rate": 1.567121660301847e-07,
+      "loss": 0.0045,
+      "step": 16932
+    },
+    {
+      "epoch": 3.848409090909091,
+      "grad_norm": 1.451811241254136,
+      "learning_rate": 1.566530628673825e-07,
+      "loss": 0.0191,
+      "step": 16933
+    },
+    {
+      "epoch": 3.8486363636363636,
+      "grad_norm": 2.5131851173999404,
+      "learning_rate": 1.5659396925499143e-07,
+      "loss": 0.0217,
+      "step": 16934
+    },
+    {
+      "epoch": 3.848863636363636,
+      "grad_norm": 0.8714177690132249,
+      "learning_rate": 1.5653488519421645e-07,
+      "loss": 0.0033,
+      "step": 16935
+    },
+    {
+      "epoch": 3.849090909090909,
+      "grad_norm": 2.6544016258647147,
+      "learning_rate": 1.5647581068626252e-07,
+      "loss": 0.0144,
+      "step": 16936
+    },
+    {
+      "epoch": 3.849318181818182,
+      "grad_norm": 0.8230335453319589,
+      "learning_rate": 1.5641674573233414e-07,
+      "loss": 0.0054,
+      "step": 16937
+    },
+    {
+      "epoch": 3.8495454545454546,
+      "grad_norm": 1.3400094790970385,
+      "learning_rate": 1.5635769033363584e-07,
+      "loss": 0.0084,
+      "step": 16938
+    },
+    {
+      "epoch": 3.849772727272727,
+      "grad_norm": 1.3360627076614457,
+      "learning_rate": 1.56298644491372e-07,
+      "loss": 0.005,
+      "step": 16939
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.900868570028056,
+      "learning_rate": 1.5623960820674633e-07,
+      "loss": 0.0051,
+      "step": 16940
+    },
+    {
+      "epoch": 3.8502272727272726,
+      "grad_norm": 1.5282540591802198,
+      "learning_rate": 1.5618058148096278e-07,
+      "loss": 0.0118,
+      "step": 16941
+    },
+    {
+      "epoch": 3.8504545454545456,
+      "grad_norm": 0.8574773694619406,
+      "learning_rate": 1.5612156431522514e-07,
+      "loss": 0.0035,
+      "step": 16942
+    },
+    {
+      "epoch": 3.850681818181818,
+      "grad_norm": 1.9137185608443568,
+      "learning_rate": 1.5606255671073697e-07,
+      "loss": 0.0123,
+      "step": 16943
+    },
+    {
+      "epoch": 3.850909090909091,
+      "grad_norm": 2.907231851884999,
+      "learning_rate": 1.5600355866870116e-07,
+      "loss": 0.0252,
+      "step": 16944
+    },
+    {
+      "epoch": 3.8511363636363636,
+      "grad_norm": 0.9056449619213189,
+      "learning_rate": 1.5594457019032104e-07,
+      "loss": 0.0034,
+      "step": 16945
+    },
+    {
+      "epoch": 3.8513636363636365,
+      "grad_norm": 0.8594726775082003,
+      "learning_rate": 1.5588559127679946e-07,
+      "loss": 0.0027,
+      "step": 16946
+    },
+    {
+      "epoch": 3.851590909090909,
+      "grad_norm": 4.0992425538926796,
+      "learning_rate": 1.5582662192933923e-07,
+      "loss": 0.0241,
+      "step": 16947
+    },
+    {
+      "epoch": 3.8518181818181816,
+      "grad_norm": 3.99941644296803,
+      "learning_rate": 1.5576766214914253e-07,
+      "loss": 0.0334,
+      "step": 16948
+    },
+    {
+      "epoch": 3.8520454545454546,
+      "grad_norm": 0.45021194406406567,
+      "learning_rate": 1.5570871193741178e-07,
+      "loss": 0.0019,
+      "step": 16949
+    },
+    {
+      "epoch": 3.8522727272727275,
+      "grad_norm": 1.994213923506353,
+      "learning_rate": 1.55649771295349e-07,
+      "loss": 0.0181,
+      "step": 16950
+    },
+    {
+      "epoch": 3.8525,
+      "grad_norm": 1.7484629196654384,
+      "learning_rate": 1.5559084022415635e-07,
+      "loss": 0.0085,
+      "step": 16951
+    },
+    {
+      "epoch": 3.8527272727272726,
+      "grad_norm": 1.5911546902786025,
+      "learning_rate": 1.5553191872503557e-07,
+      "loss": 0.0075,
+      "step": 16952
+    },
+    {
+      "epoch": 3.8529545454545455,
+      "grad_norm": 1.6934021371178214,
+      "learning_rate": 1.5547300679918782e-07,
+      "loss": 0.0142,
+      "step": 16953
+    },
+    {
+      "epoch": 3.853181818181818,
+      "grad_norm": 3.4767081870802055,
+      "learning_rate": 1.5541410444781462e-07,
+      "loss": 0.0205,
+      "step": 16954
+    },
+    {
+      "epoch": 3.853409090909091,
+      "grad_norm": 1.1916203706394675,
+      "learning_rate": 1.5535521167211707e-07,
+      "loss": 0.0091,
+      "step": 16955
+    },
+    {
+      "epoch": 3.8536363636363635,
+      "grad_norm": 1.410739530377278,
+      "learning_rate": 1.5529632847329627e-07,
+      "loss": 0.0106,
+      "step": 16956
+    },
+    {
+      "epoch": 3.8538636363636365,
+      "grad_norm": 0.9362446811168009,
+      "learning_rate": 1.552374548525526e-07,
+      "loss": 0.006,
+      "step": 16957
+    },
+    {
+      "epoch": 3.854090909090909,
+      "grad_norm": 0.7873201413252934,
+      "learning_rate": 1.5517859081108679e-07,
+      "loss": 0.0083,
+      "step": 16958
+    },
+    {
+      "epoch": 3.854318181818182,
+      "grad_norm": 1.3339747953955063,
+      "learning_rate": 1.5511973635009912e-07,
+      "loss": 0.0074,
+      "step": 16959
+    },
+    {
+      "epoch": 3.8545454545454545,
+      "grad_norm": 1.297560144536101,
+      "learning_rate": 1.5506089147078977e-07,
+      "loss": 0.0131,
+      "step": 16960
+    },
+    {
+      "epoch": 3.854772727272727,
+      "grad_norm": 3.6424427058660322,
+      "learning_rate": 1.5500205617435892e-07,
+      "loss": 0.0118,
+      "step": 16961
+    },
+    {
+      "epoch": 3.855,
+      "grad_norm": 0.44907988256869097,
+      "learning_rate": 1.5494323046200592e-07,
+      "loss": 0.0038,
+      "step": 16962
+    },
+    {
+      "epoch": 3.855227272727273,
+      "grad_norm": 0.9420505891022699,
+      "learning_rate": 1.5488441433493048e-07,
+      "loss": 0.0055,
+      "step": 16963
+    },
+    {
+      "epoch": 3.8554545454545455,
+      "grad_norm": 2.303979170555527,
+      "learning_rate": 1.54825607794332e-07,
+      "loss": 0.0157,
+      "step": 16964
+    },
+    {
+      "epoch": 3.855681818181818,
+      "grad_norm": 2.7948043260604702,
+      "learning_rate": 1.5476681084140965e-07,
+      "loss": 0.0156,
+      "step": 16965
+    },
+    {
+      "epoch": 3.855909090909091,
+      "grad_norm": 1.0632563447150831,
+      "learning_rate": 1.5470802347736235e-07,
+      "loss": 0.0098,
+      "step": 16966
+    },
+    {
+      "epoch": 3.8561363636363635,
+      "grad_norm": 1.9747914826534116,
+      "learning_rate": 1.54649245703389e-07,
+      "loss": 0.0232,
+      "step": 16967
+    },
+    {
+      "epoch": 3.8563636363636364,
+      "grad_norm": 2.5700484917651045,
+      "learning_rate": 1.54590477520688e-07,
+      "loss": 0.0312,
+      "step": 16968
+    },
+    {
+      "epoch": 3.856590909090909,
+      "grad_norm": 0.9571014958293057,
+      "learning_rate": 1.5453171893045813e-07,
+      "loss": 0.0041,
+      "step": 16969
+    },
+    {
+      "epoch": 3.856818181818182,
+      "grad_norm": 3.1959605743568353,
+      "learning_rate": 1.5447296993389705e-07,
+      "loss": 0.0235,
+      "step": 16970
+    },
+    {
+      "epoch": 3.8570454545454544,
+      "grad_norm": 1.08201173099075,
+      "learning_rate": 1.54414230532203e-07,
+      "loss": 0.0068,
+      "step": 16971
+    },
+    {
+      "epoch": 3.8572727272727274,
+      "grad_norm": 1.7767240424542894,
+      "learning_rate": 1.5435550072657374e-07,
+      "loss": 0.0162,
+      "step": 16972
+    },
+    {
+      "epoch": 3.8575,
+      "grad_norm": 0.6205230359767526,
+      "learning_rate": 1.5429678051820697e-07,
+      "loss": 0.0052,
+      "step": 16973
+    },
+    {
+      "epoch": 3.857727272727273,
+      "grad_norm": 0.5645981635464294,
+      "learning_rate": 1.5423806990830016e-07,
+      "loss": 0.0035,
+      "step": 16974
+    },
+    {
+      "epoch": 3.8579545454545454,
+      "grad_norm": 0.7278498147340251,
+      "learning_rate": 1.5417936889805022e-07,
+      "loss": 0.004,
+      "step": 16975
+    },
+    {
+      "epoch": 3.8581818181818184,
+      "grad_norm": 1.1574422937396505,
+      "learning_rate": 1.5412067748865437e-07,
+      "loss": 0.0091,
+      "step": 16976
+    },
+    {
+      "epoch": 3.858409090909091,
+      "grad_norm": 2.151429493423349,
+      "learning_rate": 1.5406199568130934e-07,
+      "loss": 0.0091,
+      "step": 16977
+    },
+    {
+      "epoch": 3.8586363636363634,
+      "grad_norm": 2.3261451721528643,
+      "learning_rate": 1.5400332347721202e-07,
+      "loss": 0.0155,
+      "step": 16978
+    },
+    {
+      "epoch": 3.8588636363636364,
+      "grad_norm": 1.1092755302969421,
+      "learning_rate": 1.5394466087755835e-07,
+      "loss": 0.0066,
+      "step": 16979
+    },
+    {
+      "epoch": 3.8590909090909093,
+      "grad_norm": 0.6346787151043526,
+      "learning_rate": 1.5388600788354497e-07,
+      "loss": 0.0038,
+      "step": 16980
+    },
+    {
+      "epoch": 3.859318181818182,
+      "grad_norm": 1.361050000631691,
+      "learning_rate": 1.5382736449636775e-07,
+      "loss": 0.0077,
+      "step": 16981
+    },
+    {
+      "epoch": 3.8595454545454544,
+      "grad_norm": 0.9224503806960378,
+      "learning_rate": 1.537687307172226e-07,
+      "loss": 0.0053,
+      "step": 16982
+    },
+    {
+      "epoch": 3.8597727272727274,
+      "grad_norm": 1.6067778841483658,
+      "learning_rate": 1.5371010654730531e-07,
+      "loss": 0.012,
+      "step": 16983
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 1.1593641767282763,
+      "learning_rate": 1.5365149198781105e-07,
+      "loss": 0.008,
+      "step": 16984
+    },
+    {
+      "epoch": 3.860227272727273,
+      "grad_norm": 1.7621692884705105,
+      "learning_rate": 1.535928870399351e-07,
+      "loss": 0.0211,
+      "step": 16985
+    },
+    {
+      "epoch": 3.8604545454545454,
+      "grad_norm": 2.4447922504872466,
+      "learning_rate": 1.5353429170487266e-07,
+      "loss": 0.0205,
+      "step": 16986
+    },
+    {
+      "epoch": 3.8606818181818183,
+      "grad_norm": 1.397799477401288,
+      "learning_rate": 1.5347570598381864e-07,
+      "loss": 0.0075,
+      "step": 16987
+    },
+    {
+      "epoch": 3.860909090909091,
+      "grad_norm": 0.6265091899453702,
+      "learning_rate": 1.5341712987796749e-07,
+      "loss": 0.0037,
+      "step": 16988
+    },
+    {
+      "epoch": 3.861136363636364,
+      "grad_norm": 2.2804407941927898,
+      "learning_rate": 1.5335856338851373e-07,
+      "loss": 0.0248,
+      "step": 16989
+    },
+    {
+      "epoch": 3.8613636363636363,
+      "grad_norm": 2.9452240759439126,
+      "learning_rate": 1.5330000651665171e-07,
+      "loss": 0.0375,
+      "step": 16990
+    },
+    {
+      "epoch": 3.861590909090909,
+      "grad_norm": 0.7428611721465922,
+      "learning_rate": 1.5324145926357565e-07,
+      "loss": 0.0059,
+      "step": 16991
+    },
+    {
+      "epoch": 3.861818181818182,
+      "grad_norm": 0.42532546241658264,
+      "learning_rate": 1.5318292163047908e-07,
+      "loss": 0.0048,
+      "step": 16992
+    },
+    {
+      "epoch": 3.862045454545455,
+      "grad_norm": 1.0487055924972228,
+      "learning_rate": 1.5312439361855585e-07,
+      "loss": 0.0042,
+      "step": 16993
+    },
+    {
+      "epoch": 3.8622727272727273,
+      "grad_norm": 2.3612807272324323,
+      "learning_rate": 1.530658752289995e-07,
+      "loss": 0.0122,
+      "step": 16994
+    },
+    {
+      "epoch": 3.8625,
+      "grad_norm": 0.8786376526929676,
+      "learning_rate": 1.530073664630033e-07,
+      "loss": 0.0053,
+      "step": 16995
+    },
+    {
+      "epoch": 3.862727272727273,
+      "grad_norm": 2.076521666068703,
+      "learning_rate": 1.529488673217603e-07,
+      "loss": 0.0257,
+      "step": 16996
+    },
+    {
+      "epoch": 3.8629545454545453,
+      "grad_norm": 1.1993851840045213,
+      "learning_rate": 1.5289037780646338e-07,
+      "loss": 0.0066,
+      "step": 16997
+    },
+    {
+      "epoch": 3.8631818181818183,
+      "grad_norm": 1.480357805008943,
+      "learning_rate": 1.5283189791830542e-07,
+      "loss": 0.0088,
+      "step": 16998
+    },
+    {
+      "epoch": 3.863409090909091,
+      "grad_norm": 2.2642370803064678,
+      "learning_rate": 1.5277342765847873e-07,
+      "loss": 0.0083,
+      "step": 16999
+    },
+    {
+      "epoch": 3.8636363636363638,
+      "grad_norm": 1.7557692177433604,
+      "learning_rate": 1.5271496702817593e-07,
+      "loss": 0.0118,
+      "step": 17000
+    },
+    {
+      "epoch": 3.8638636363636363,
+      "grad_norm": 1.5103395200319238,
+      "learning_rate": 1.5265651602858867e-07,
+      "loss": 0.0089,
+      "step": 17001
+    },
+    {
+      "epoch": 3.8640909090909092,
+      "grad_norm": 1.6756938648750117,
+      "learning_rate": 1.525980746609091e-07,
+      "loss": 0.014,
+      "step": 17002
+    },
+    {
+      "epoch": 3.8643181818181818,
+      "grad_norm": 1.3321456146563595,
+      "learning_rate": 1.5253964292632893e-07,
+      "loss": 0.0042,
+      "step": 17003
+    },
+    {
+      "epoch": 3.8645454545454543,
+      "grad_norm": 1.792196938711082,
+      "learning_rate": 1.524812208260399e-07,
+      "loss": 0.0155,
+      "step": 17004
+    },
+    {
+      "epoch": 3.8647727272727272,
+      "grad_norm": 1.1348476907306342,
+      "learning_rate": 1.5242280836123293e-07,
+      "loss": 0.0125,
+      "step": 17005
+    },
+    {
+      "epoch": 3.865,
+      "grad_norm": 1.2187249583932036,
+      "learning_rate": 1.523644055330994e-07,
+      "loss": 0.0088,
+      "step": 17006
+    },
+    {
+      "epoch": 3.8652272727272727,
+      "grad_norm": 1.074967027283724,
+      "learning_rate": 1.5230601234283013e-07,
+      "loss": 0.0073,
+      "step": 17007
+    },
+    {
+      "epoch": 3.8654545454545453,
+      "grad_norm": 1.2797827713183778,
+      "learning_rate": 1.5224762879161595e-07,
+      "loss": 0.0124,
+      "step": 17008
+    },
+    {
+      "epoch": 3.865681818181818,
+      "grad_norm": 2.7665589065770146,
+      "learning_rate": 1.5218925488064738e-07,
+      "loss": 0.017,
+      "step": 17009
+    },
+    {
+      "epoch": 3.865909090909091,
+      "grad_norm": 1.081234208220015,
+      "learning_rate": 1.5213089061111476e-07,
+      "loss": 0.0077,
+      "step": 17010
+    },
+    {
+      "epoch": 3.8661363636363637,
+      "grad_norm": 1.7349918816600598,
+      "learning_rate": 1.5207253598420826e-07,
+      "loss": 0.0105,
+      "step": 17011
+    },
+    {
+      "epoch": 3.8663636363636362,
+      "grad_norm": 2.7914008703870477,
+      "learning_rate": 1.5201419100111782e-07,
+      "loss": 0.015,
+      "step": 17012
+    },
+    {
+      "epoch": 3.866590909090909,
+      "grad_norm": 1.424496129924393,
+      "learning_rate": 1.5195585566303337e-07,
+      "loss": 0.005,
+      "step": 17013
+    },
+    {
+      "epoch": 3.8668181818181817,
+      "grad_norm": 6.89695155435266,
+      "learning_rate": 1.5189752997114415e-07,
+      "loss": 0.021,
+      "step": 17014
+    },
+    {
+      "epoch": 3.8670454545454547,
+      "grad_norm": 2.168633818863624,
+      "learning_rate": 1.518392139266396e-07,
+      "loss": 0.0145,
+      "step": 17015
+    },
+    {
+      "epoch": 3.867272727272727,
+      "grad_norm": 2.6492811326844237,
+      "learning_rate": 1.5178090753070903e-07,
+      "loss": 0.0091,
+      "step": 17016
+    },
+    {
+      "epoch": 3.8675,
+      "grad_norm": 2.349890616388128,
+      "learning_rate": 1.5172261078454127e-07,
+      "loss": 0.0213,
+      "step": 17017
+    },
+    {
+      "epoch": 3.8677272727272727,
+      "grad_norm": 0.9667870208228246,
+      "learning_rate": 1.516643236893254e-07,
+      "loss": 0.004,
+      "step": 17018
+    },
+    {
+      "epoch": 3.8679545454545456,
+      "grad_norm": 1.0216310773782729,
+      "learning_rate": 1.5160604624624952e-07,
+      "loss": 0.0077,
+      "step": 17019
+    },
+    {
+      "epoch": 3.868181818181818,
+      "grad_norm": 0.3723141174190096,
+      "learning_rate": 1.5154777845650218e-07,
+      "loss": 0.0017,
+      "step": 17020
+    },
+    {
+      "epoch": 3.8684090909090907,
+      "grad_norm": 2.1695984167582822,
+      "learning_rate": 1.514895203212717e-07,
+      "loss": 0.0285,
+      "step": 17021
+    },
+    {
+      "epoch": 3.8686363636363637,
+      "grad_norm": 2.0900584313685666,
+      "learning_rate": 1.5143127184174614e-07,
+      "loss": 0.0104,
+      "step": 17022
+    },
+    {
+      "epoch": 3.8688636363636366,
+      "grad_norm": 1.0218727547873685,
+      "learning_rate": 1.5137303301911293e-07,
+      "loss": 0.0086,
+      "step": 17023
+    },
+    {
+      "epoch": 3.869090909090909,
+      "grad_norm": 0.953269428674404,
+      "learning_rate": 1.513148038545599e-07,
+      "loss": 0.0064,
+      "step": 17024
+    },
+    {
+      "epoch": 3.8693181818181817,
+      "grad_norm": 2.0729270570769542,
+      "learning_rate": 1.5125658434927442e-07,
+      "loss": 0.0124,
+      "step": 17025
+    },
+    {
+      "epoch": 3.8695454545454546,
+      "grad_norm": 0.991455002657366,
+      "learning_rate": 1.5119837450444368e-07,
+      "loss": 0.0081,
+      "step": 17026
+    },
+    {
+      "epoch": 3.869772727272727,
+      "grad_norm": 1.4474197006251774,
+      "learning_rate": 1.5114017432125462e-07,
+      "loss": 0.0074,
+      "step": 17027
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 1.2759523084870896,
+      "learning_rate": 1.5108198380089416e-07,
+      "loss": 0.0112,
+      "step": 17028
+    },
+    {
+      "epoch": 3.8702272727272726,
+      "grad_norm": 1.4443875973688596,
+      "learning_rate": 1.5102380294454885e-07,
+      "loss": 0.0207,
+      "step": 17029
+    },
+    {
+      "epoch": 3.8704545454545456,
+      "grad_norm": 2.4516931327229963,
+      "learning_rate": 1.50965631753405e-07,
+      "loss": 0.0141,
+      "step": 17030
+    },
+    {
+      "epoch": 3.870681818181818,
+      "grad_norm": 1.731056043282299,
+      "learning_rate": 1.509074702286492e-07,
+      "loss": 0.0654,
+      "step": 17031
+    },
+    {
+      "epoch": 3.870909090909091,
+      "grad_norm": 2.775566628217929,
+      "learning_rate": 1.5084931837146695e-07,
+      "loss": 0.0193,
+      "step": 17032
+    },
+    {
+      "epoch": 3.8711363636363636,
+      "grad_norm": 1.1398137707433178,
+      "learning_rate": 1.5079117618304432e-07,
+      "loss": 0.0111,
+      "step": 17033
+    },
+    {
+      "epoch": 3.871363636363636,
+      "grad_norm": 2.189147314050039,
+      "learning_rate": 1.5073304366456688e-07,
+      "loss": 0.0101,
+      "step": 17034
+    },
+    {
+      "epoch": 3.871590909090909,
+      "grad_norm": 0.9754673817809737,
+      "learning_rate": 1.506749208172203e-07,
+      "loss": 0.0051,
+      "step": 17035
+    },
+    {
+      "epoch": 3.871818181818182,
+      "grad_norm": 0.762857297411029,
+      "learning_rate": 1.506168076421894e-07,
+      "loss": 0.0046,
+      "step": 17036
+    },
+    {
+      "epoch": 3.8720454545454546,
+      "grad_norm": 1.225232602447853,
+      "learning_rate": 1.505587041406594e-07,
+      "loss": 0.006,
+      "step": 17037
+    },
+    {
+      "epoch": 3.872272727272727,
+      "grad_norm": 0.4678345955909657,
+      "learning_rate": 1.5050061031381514e-07,
+      "loss": 0.0049,
+      "step": 17038
+    },
+    {
+      "epoch": 3.8725,
+      "grad_norm": 1.6511159873945096,
+      "learning_rate": 1.5044252616284123e-07,
+      "loss": 0.0095,
+      "step": 17039
+    },
+    {
+      "epoch": 3.8727272727272726,
+      "grad_norm": 1.1942582907511994,
+      "learning_rate": 1.5038445168892212e-07,
+      "loss": 0.0121,
+      "step": 17040
+    },
+    {
+      "epoch": 3.8729545454545455,
+      "grad_norm": 1.1566246609284514,
+      "learning_rate": 1.5032638689324205e-07,
+      "loss": 0.0088,
+      "step": 17041
+    },
+    {
+      "epoch": 3.873181818181818,
+      "grad_norm": 1.535854824501893,
+      "learning_rate": 1.5026833177698506e-07,
+      "loss": 0.0073,
+      "step": 17042
+    },
+    {
+      "epoch": 3.873409090909091,
+      "grad_norm": 1.1652676225570209,
+      "learning_rate": 1.5021028634133503e-07,
+      "loss": 0.0094,
+      "step": 17043
+    },
+    {
+      "epoch": 3.8736363636363635,
+      "grad_norm": 1.2165830988519102,
+      "learning_rate": 1.501522505874757e-07,
+      "loss": 0.0089,
+      "step": 17044
+    },
+    {
+      "epoch": 3.8738636363636365,
+      "grad_norm": 0.7747382865886653,
+      "learning_rate": 1.5009422451659027e-07,
+      "loss": 0.0043,
+      "step": 17045
+    },
+    {
+      "epoch": 3.874090909090909,
+      "grad_norm": 0.7363680810860781,
+      "learning_rate": 1.5003620812986208e-07,
+      "loss": 0.0049,
+      "step": 17046
+    },
+    {
+      "epoch": 3.8743181818181816,
+      "grad_norm": 2.8887976944854508,
+      "learning_rate": 1.4997820142847433e-07,
+      "loss": 0.0194,
+      "step": 17047
+    },
+    {
+      "epoch": 3.8745454545454545,
+      "grad_norm": 0.9853344857425337,
+      "learning_rate": 1.4992020441360985e-07,
+      "loss": 0.0074,
+      "step": 17048
+    },
+    {
+      "epoch": 3.8747727272727275,
+      "grad_norm": 1.5485394669003374,
+      "learning_rate": 1.4986221708645103e-07,
+      "loss": 0.0162,
+      "step": 17049
+    },
+    {
+      "epoch": 3.875,
+      "grad_norm": 1.6828792183610872,
+      "learning_rate": 1.4980423944818057e-07,
+      "loss": 0.0061,
+      "step": 17050
+    },
+    {
+      "epoch": 3.8752272727272725,
+      "grad_norm": 2.7826207905443994,
+      "learning_rate": 1.4974627149998068e-07,
+      "loss": 0.0131,
+      "step": 17051
+    },
+    {
+      "epoch": 3.8754545454545455,
+      "grad_norm": 1.2817907629695198,
+      "learning_rate": 1.496883132430335e-07,
+      "loss": 0.0045,
+      "step": 17052
+    },
+    {
+      "epoch": 3.8756818181818184,
+      "grad_norm": 1.1754085265938747,
+      "learning_rate": 1.4963036467852098e-07,
+      "loss": 0.0094,
+      "step": 17053
+    },
+    {
+      "epoch": 3.875909090909091,
+      "grad_norm": 1.5493347934019808,
+      "learning_rate": 1.495724258076244e-07,
+      "loss": 0.005,
+      "step": 17054
+    },
+    {
+      "epoch": 3.8761363636363635,
+      "grad_norm": 0.29501350821771155,
+      "learning_rate": 1.495144966315256e-07,
+      "loss": 0.0018,
+      "step": 17055
+    },
+    {
+      "epoch": 3.8763636363636365,
+      "grad_norm": 1.0550303732192723,
+      "learning_rate": 1.4945657715140567e-07,
+      "loss": 0.0042,
+      "step": 17056
+    },
+    {
+      "epoch": 3.876590909090909,
+      "grad_norm": 0.4244542554147354,
+      "learning_rate": 1.4939866736844576e-07,
+      "loss": 0.0024,
+      "step": 17057
+    },
+    {
+      "epoch": 3.876818181818182,
+      "grad_norm": 1.5576351190252153,
+      "learning_rate": 1.4934076728382678e-07,
+      "loss": 0.0152,
+      "step": 17058
+    },
+    {
+      "epoch": 3.8770454545454545,
+      "grad_norm": 1.371178471107338,
+      "learning_rate": 1.4928287689872943e-07,
+      "loss": 0.0073,
+      "step": 17059
+    },
+    {
+      "epoch": 3.8772727272727274,
+      "grad_norm": 1.3535914128197168,
+      "learning_rate": 1.4922499621433412e-07,
+      "loss": 0.0075,
+      "step": 17060
+    },
+    {
+      "epoch": 3.8775,
+      "grad_norm": 2.4605348269548295,
+      "learning_rate": 1.4916712523182114e-07,
+      "loss": 0.0171,
+      "step": 17061
+    },
+    {
+      "epoch": 3.877727272727273,
+      "grad_norm": 1.06874197620838,
+      "learning_rate": 1.4910926395237077e-07,
+      "loss": 0.0073,
+      "step": 17062
+    },
+    {
+      "epoch": 3.8779545454545454,
+      "grad_norm": 1.2552205493398214,
+      "learning_rate": 1.490514123771627e-07,
+      "loss": 0.007,
+      "step": 17063
+    },
+    {
+      "epoch": 3.878181818181818,
+      "grad_norm": 1.7229608912367838,
+      "learning_rate": 1.4899357050737654e-07,
+      "loss": 0.0119,
+      "step": 17064
+    },
+    {
+      "epoch": 3.878409090909091,
+      "grad_norm": 1.1927870653758106,
+      "learning_rate": 1.4893573834419195e-07,
+      "loss": 0.0059,
+      "step": 17065
+    },
+    {
+      "epoch": 3.878636363636364,
+      "grad_norm": 1.8890432739291545,
+      "learning_rate": 1.488779158887884e-07,
+      "loss": 0.0115,
+      "step": 17066
+    },
+    {
+      "epoch": 3.8788636363636364,
+      "grad_norm": 1.6020551218544905,
+      "learning_rate": 1.488201031423446e-07,
+      "loss": 0.0179,
+      "step": 17067
+    },
+    {
+      "epoch": 3.879090909090909,
+      "grad_norm": 3.649700327106543,
+      "learning_rate": 1.487623001060396e-07,
+      "loss": 0.0245,
+      "step": 17068
+    },
+    {
+      "epoch": 3.879318181818182,
+      "grad_norm": 1.8156180726386146,
+      "learning_rate": 1.487045067810522e-07,
+      "loss": 0.0097,
+      "step": 17069
+    },
+    {
+      "epoch": 3.8795454545454544,
+      "grad_norm": 2.3422448594092082,
+      "learning_rate": 1.486467231685608e-07,
+      "loss": 0.0207,
+      "step": 17070
+    },
+    {
+      "epoch": 3.8797727272727274,
+      "grad_norm": 2.7240500920180177,
+      "learning_rate": 1.485889492697438e-07,
+      "loss": 0.0285,
+      "step": 17071
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 4.595356570371572,
+      "learning_rate": 1.4853118508577922e-07,
+      "loss": 0.0195,
+      "step": 17072
+    },
+    {
+      "epoch": 3.880227272727273,
+      "grad_norm": 1.6513800458941594,
+      "learning_rate": 1.48473430617845e-07,
+      "loss": 0.0098,
+      "step": 17073
+    },
+    {
+      "epoch": 3.8804545454545454,
+      "grad_norm": 1.872694059439454,
+      "learning_rate": 1.4841568586711895e-07,
+      "loss": 0.0084,
+      "step": 17074
+    },
+    {
+      "epoch": 3.8806818181818183,
+      "grad_norm": 0.6000649615805014,
+      "learning_rate": 1.483579508347787e-07,
+      "loss": 0.0041,
+      "step": 17075
+    },
+    {
+      "epoch": 3.880909090909091,
+      "grad_norm": 1.2270281898935151,
+      "learning_rate": 1.4830022552200112e-07,
+      "loss": 0.0028,
+      "step": 17076
+    },
+    {
+      "epoch": 3.8811363636363634,
+      "grad_norm": 1.5645852666308602,
+      "learning_rate": 1.4824250992996365e-07,
+      "loss": 0.0051,
+      "step": 17077
+    },
+    {
+      "epoch": 3.8813636363636363,
+      "grad_norm": 0.4767957307932769,
+      "learning_rate": 1.4818480405984312e-07,
+      "loss": 0.0016,
+      "step": 17078
+    },
+    {
+      "epoch": 3.8815909090909093,
+      "grad_norm": 1.0884467136508156,
+      "learning_rate": 1.4812710791281652e-07,
+      "loss": 0.0101,
+      "step": 17079
+    },
+    {
+      "epoch": 3.881818181818182,
+      "grad_norm": 1.9926461647867963,
+      "learning_rate": 1.4806942149005993e-07,
+      "loss": 0.0188,
+      "step": 17080
+    },
+    {
+      "epoch": 3.8820454545454544,
+      "grad_norm": 6.5488652616684995,
+      "learning_rate": 1.4801174479274982e-07,
+      "loss": 0.0099,
+      "step": 17081
+    },
+    {
+      "epoch": 3.8822727272727273,
+      "grad_norm": 1.1825113835614687,
+      "learning_rate": 1.4795407782206249e-07,
+      "loss": 0.0046,
+      "step": 17082
+    },
+    {
+      "epoch": 3.8825,
+      "grad_norm": 0.5706484916436794,
+      "learning_rate": 1.478964205791737e-07,
+      "loss": 0.0022,
+      "step": 17083
+    },
+    {
+      "epoch": 3.882727272727273,
+      "grad_norm": 1.5745937329648199,
+      "learning_rate": 1.4783877306525944e-07,
+      "loss": 0.0108,
+      "step": 17084
+    },
+    {
+      "epoch": 3.8829545454545453,
+      "grad_norm": 1.0834481653537322,
+      "learning_rate": 1.4778113528149486e-07,
+      "loss": 0.007,
+      "step": 17085
+    },
+    {
+      "epoch": 3.8831818181818183,
+      "grad_norm": 0.6151928560130533,
+      "learning_rate": 1.4772350722905551e-07,
+      "loss": 0.0031,
+      "step": 17086
+    },
+    {
+      "epoch": 3.883409090909091,
+      "grad_norm": 0.9980556911906243,
+      "learning_rate": 1.4766588890911648e-07,
+      "loss": 0.0038,
+      "step": 17087
+    },
+    {
+      "epoch": 3.8836363636363638,
+      "grad_norm": 3.307367194665117,
+      "learning_rate": 1.4760828032285277e-07,
+      "loss": 0.0176,
+      "step": 17088
+    },
+    {
+      "epoch": 3.8838636363636363,
+      "grad_norm": 0.8322885325565741,
+      "learning_rate": 1.4755068147143906e-07,
+      "loss": 0.0044,
+      "step": 17089
+    },
+    {
+      "epoch": 3.884090909090909,
+      "grad_norm": 2.533171692116893,
+      "learning_rate": 1.474930923560499e-07,
+      "loss": 0.0262,
+      "step": 17090
+    },
+    {
+      "epoch": 3.884318181818182,
+      "grad_norm": 0.9944436823927301,
+      "learning_rate": 1.4743551297785965e-07,
+      "loss": 0.0065,
+      "step": 17091
+    },
+    {
+      "epoch": 3.8845454545454547,
+      "grad_norm": 1.3759406116752044,
+      "learning_rate": 1.473779433380426e-07,
+      "loss": 0.0072,
+      "step": 17092
+    },
+    {
+      "epoch": 3.8847727272727273,
+      "grad_norm": 2.8022087182427455,
+      "learning_rate": 1.473203834377724e-07,
+      "loss": 0.021,
+      "step": 17093
+    },
+    {
+      "epoch": 3.885,
+      "grad_norm": 2.1271284121365186,
+      "learning_rate": 1.4726283327822297e-07,
+      "loss": 0.0168,
+      "step": 17094
+    },
+    {
+      "epoch": 3.8852272727272728,
+      "grad_norm": 0.9215306027430813,
+      "learning_rate": 1.472052928605678e-07,
+      "loss": 0.0086,
+      "step": 17095
+    },
+    {
+      "epoch": 3.8854545454545457,
+      "grad_norm": 1.6665693676972735,
+      "learning_rate": 1.4714776218598022e-07,
+      "loss": 0.0062,
+      "step": 17096
+    },
+    {
+      "epoch": 3.8856818181818182,
+      "grad_norm": 2.900358216273082,
+      "learning_rate": 1.4709024125563367e-07,
+      "loss": 0.0269,
+      "step": 17097
+    },
+    {
+      "epoch": 3.8859090909090908,
+      "grad_norm": 2.264861177927651,
+      "learning_rate": 1.4703273007070064e-07,
+      "loss": 0.0157,
+      "step": 17098
+    },
+    {
+      "epoch": 3.8861363636363637,
+      "grad_norm": 1.3527223091613056,
+      "learning_rate": 1.4697522863235413e-07,
+      "loss": 0.0108,
+      "step": 17099
+    },
+    {
+      "epoch": 3.8863636363636362,
+      "grad_norm": 0.7297038637883773,
+      "learning_rate": 1.469177369417666e-07,
+      "loss": 0.0052,
+      "step": 17100
+    },
+    {
+      "epoch": 3.886590909090909,
+      "grad_norm": 0.8670888195017582,
+      "learning_rate": 1.4686025500011054e-07,
+      "loss": 0.0036,
+      "step": 17101
+    },
+    {
+      "epoch": 3.8868181818181817,
+      "grad_norm": 1.8678685267122432,
+      "learning_rate": 1.4680278280855804e-07,
+      "loss": 0.0123,
+      "step": 17102
+    },
+    {
+      "epoch": 3.8870454545454547,
+      "grad_norm": 0.9915058063841805,
+      "learning_rate": 1.467453203682811e-07,
+      "loss": 0.0048,
+      "step": 17103
+    },
+    {
+      "epoch": 3.887272727272727,
+      "grad_norm": 1.9552296134886817,
+      "learning_rate": 1.466878676804513e-07,
+      "loss": 0.0156,
+      "step": 17104
+    },
+    {
+      "epoch": 3.8875,
+      "grad_norm": 2.303951968177325,
+      "learning_rate": 1.4663042474624043e-07,
+      "loss": 0.0138,
+      "step": 17105
+    },
+    {
+      "epoch": 3.8877272727272727,
+      "grad_norm": 1.684947473735316,
+      "learning_rate": 1.4657299156682001e-07,
+      "loss": 0.008,
+      "step": 17106
+    },
+    {
+      "epoch": 3.887954545454545,
+      "grad_norm": 0.9842624176647061,
+      "learning_rate": 1.4651556814336068e-07,
+      "loss": 0.0061,
+      "step": 17107
+    },
+    {
+      "epoch": 3.888181818181818,
+      "grad_norm": 1.4671893510778775,
+      "learning_rate": 1.4645815447703369e-07,
+      "loss": 0.01,
+      "step": 17108
+    },
+    {
+      "epoch": 3.888409090909091,
+      "grad_norm": 0.5268346042098571,
+      "learning_rate": 1.464007505690098e-07,
+      "loss": 0.0018,
+      "step": 17109
+    },
+    {
+      "epoch": 3.8886363636363637,
+      "grad_norm": 0.4149029595525852,
+      "learning_rate": 1.4634335642045971e-07,
+      "loss": 0.0023,
+      "step": 17110
+    },
+    {
+      "epoch": 3.888863636363636,
+      "grad_norm": 4.590410024375908,
+      "learning_rate": 1.462859720325535e-07,
+      "loss": 0.0289,
+      "step": 17111
+    },
+    {
+      "epoch": 3.889090909090909,
+      "grad_norm": 1.8424800932438625,
+      "learning_rate": 1.4622859740646152e-07,
+      "loss": 0.0134,
+      "step": 17112
+    },
+    {
+      "epoch": 3.8893181818181817,
+      "grad_norm": 2.330543132711256,
+      "learning_rate": 1.461712325433537e-07,
+      "loss": 0.0283,
+      "step": 17113
+    },
+    {
+      "epoch": 3.8895454545454546,
+      "grad_norm": 2.760735341536348,
+      "learning_rate": 1.4611387744439992e-07,
+      "loss": 0.0137,
+      "step": 17114
+    },
+    {
+      "epoch": 3.889772727272727,
+      "grad_norm": 1.2196281341406507,
+      "learning_rate": 1.460565321107695e-07,
+      "loss": 0.0079,
+      "step": 17115
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 1.1487817993529421,
+      "learning_rate": 1.4599919654363192e-07,
+      "loss": 0.0052,
+      "step": 17116
+    },
+    {
+      "epoch": 3.8902272727272726,
+      "grad_norm": 0.37830909083142694,
+      "learning_rate": 1.4594187074415643e-07,
+      "loss": 0.0016,
+      "step": 17117
+    },
+    {
+      "epoch": 3.8904545454545456,
+      "grad_norm": 1.104368060491311,
+      "learning_rate": 1.4588455471351182e-07,
+      "loss": 0.0039,
+      "step": 17118
+    },
+    {
+      "epoch": 3.890681818181818,
+      "grad_norm": 2.2817713678455265,
+      "learning_rate": 1.458272484528673e-07,
+      "loss": 0.0131,
+      "step": 17119
+    },
+    {
+      "epoch": 3.8909090909090907,
+      "grad_norm": 0.3867575477531098,
+      "learning_rate": 1.4576995196339094e-07,
+      "loss": 0.0022,
+      "step": 17120
+    },
+    {
+      "epoch": 3.8911363636363636,
+      "grad_norm": 1.2256516756277867,
+      "learning_rate": 1.4571266524625136e-07,
+      "loss": 0.0067,
+      "step": 17121
+    },
+    {
+      "epoch": 3.8913636363636366,
+      "grad_norm": 0.6206756184608704,
+      "learning_rate": 1.456553883026167e-07,
+      "loss": 0.0018,
+      "step": 17122
+    },
+    {
+      "epoch": 3.891590909090909,
+      "grad_norm": 4.019197363070231,
+      "learning_rate": 1.4559812113365512e-07,
+      "loss": 0.036,
+      "step": 17123
+    },
+    {
+      "epoch": 3.8918181818181816,
+      "grad_norm": 1.8494903151764137,
+      "learning_rate": 1.45540863740534e-07,
+      "loss": 0.0127,
+      "step": 17124
+    },
+    {
+      "epoch": 3.8920454545454546,
+      "grad_norm": 4.025379038334136,
+      "learning_rate": 1.4548361612442114e-07,
+      "loss": 0.0298,
+      "step": 17125
+    },
+    {
+      "epoch": 3.892272727272727,
+      "grad_norm": 0.6550397315085558,
+      "learning_rate": 1.4542637828648398e-07,
+      "loss": 0.0031,
+      "step": 17126
+    },
+    {
+      "epoch": 3.8925,
+      "grad_norm": 1.380236193760445,
+      "learning_rate": 1.453691502278897e-07,
+      "loss": 0.0083,
+      "step": 17127
+    },
+    {
+      "epoch": 3.8927272727272726,
+      "grad_norm": 1.0723456657478243,
+      "learning_rate": 1.453119319498051e-07,
+      "loss": 0.005,
+      "step": 17128
+    },
+    {
+      "epoch": 3.8929545454545456,
+      "grad_norm": 1.203744379114212,
+      "learning_rate": 1.4525472345339708e-07,
+      "loss": 0.0117,
+      "step": 17129
+    },
+    {
+      "epoch": 3.893181818181818,
+      "grad_norm": 1.4851599969473241,
+      "learning_rate": 1.451975247398322e-07,
+      "loss": 0.0055,
+      "step": 17130
+    },
+    {
+      "epoch": 3.893409090909091,
+      "grad_norm": 2.2227310263278435,
+      "learning_rate": 1.451403358102768e-07,
+      "loss": 0.0146,
+      "step": 17131
+    },
+    {
+      "epoch": 3.8936363636363636,
+      "grad_norm": 1.8253562357883806,
+      "learning_rate": 1.4508315666589716e-07,
+      "loss": 0.0208,
+      "step": 17132
+    },
+    {
+      "epoch": 3.8938636363636365,
+      "grad_norm": 1.3250071560096603,
+      "learning_rate": 1.4502598730785918e-07,
+      "loss": 0.0128,
+      "step": 17133
+    },
+    {
+      "epoch": 3.894090909090909,
+      "grad_norm": 1.6231258685374879,
+      "learning_rate": 1.4496882773732872e-07,
+      "loss": 0.0078,
+      "step": 17134
+    },
+    {
+      "epoch": 3.894318181818182,
+      "grad_norm": 1.656875172270521,
+      "learning_rate": 1.4491167795547128e-07,
+      "loss": 0.0057,
+      "step": 17135
+    },
+    {
+      "epoch": 3.8945454545454545,
+      "grad_norm": 1.4303473758978835,
+      "learning_rate": 1.4485453796345251e-07,
+      "loss": 0.0064,
+      "step": 17136
+    },
+    {
+      "epoch": 3.894772727272727,
+      "grad_norm": 0.6899476092034245,
+      "learning_rate": 1.4479740776243714e-07,
+      "loss": 0.0062,
+      "step": 17137
+    },
+    {
+      "epoch": 3.895,
+      "grad_norm": 2.8710034506225033,
+      "learning_rate": 1.4474028735359046e-07,
+      "loss": 0.0204,
+      "step": 17138
+    },
+    {
+      "epoch": 3.895227272727273,
+      "grad_norm": 1.810577947858718,
+      "learning_rate": 1.4468317673807714e-07,
+      "loss": 0.0075,
+      "step": 17139
+    },
+    {
+      "epoch": 3.8954545454545455,
+      "grad_norm": 2.348324310051794,
+      "learning_rate": 1.4462607591706181e-07,
+      "loss": 0.0132,
+      "step": 17140
+    },
+    {
+      "epoch": 3.895681818181818,
+      "grad_norm": 1.0218006723930355,
+      "learning_rate": 1.44568984891709e-07,
+      "loss": 0.0082,
+      "step": 17141
+    },
+    {
+      "epoch": 3.895909090909091,
+      "grad_norm": 2.2401128900889318,
+      "learning_rate": 1.4451190366318258e-07,
+      "loss": 0.015,
+      "step": 17142
+    },
+    {
+      "epoch": 3.8961363636363635,
+      "grad_norm": 4.647680779436196,
+      "learning_rate": 1.4445483223264674e-07,
+      "loss": 0.0204,
+      "step": 17143
+    },
+    {
+      "epoch": 3.8963636363636365,
+      "grad_norm": 3.482105752443717,
+      "learning_rate": 1.4439777060126518e-07,
+      "loss": 0.0324,
+      "step": 17144
+    },
+    {
+      "epoch": 3.896590909090909,
+      "grad_norm": 1.266785429153159,
+      "learning_rate": 1.4434071877020173e-07,
+      "loss": 0.0054,
+      "step": 17145
+    },
+    {
+      "epoch": 3.896818181818182,
+      "grad_norm": 2.3764519338906105,
+      "learning_rate": 1.4428367674061936e-07,
+      "loss": 0.0063,
+      "step": 17146
+    },
+    {
+      "epoch": 3.8970454545454545,
+      "grad_norm": 0.3807346799937595,
+      "learning_rate": 1.442266445136814e-07,
+      "loss": 0.0017,
+      "step": 17147
+    },
+    {
+      "epoch": 3.8972727272727274,
+      "grad_norm": 5.392006888413386,
+      "learning_rate": 1.44169622090551e-07,
+      "loss": 0.026,
+      "step": 17148
+    },
+    {
+      "epoch": 3.8975,
+      "grad_norm": 0.7576231758276549,
+      "learning_rate": 1.441126094723911e-07,
+      "loss": 0.0063,
+      "step": 17149
+    },
+    {
+      "epoch": 3.8977272727272725,
+      "grad_norm": 2.960647966408755,
+      "learning_rate": 1.4405560666036386e-07,
+      "loss": 0.0173,
+      "step": 17150
+    },
+    {
+      "epoch": 3.8979545454545454,
+      "grad_norm": 0.7724457836885249,
+      "learning_rate": 1.439986136556318e-07,
+      "loss": 0.0045,
+      "step": 17151
+    },
+    {
+      "epoch": 3.8981818181818184,
+      "grad_norm": 1.4211884465904467,
+      "learning_rate": 1.4394163045935715e-07,
+      "loss": 0.0115,
+      "step": 17152
+    },
+    {
+      "epoch": 3.898409090909091,
+      "grad_norm": 1.6187607330669247,
+      "learning_rate": 1.4388465707270197e-07,
+      "loss": 0.013,
+      "step": 17153
+    },
+    {
+      "epoch": 3.8986363636363635,
+      "grad_norm": 1.0792000470869447,
+      "learning_rate": 1.4382769349682816e-07,
+      "loss": 0.0119,
+      "step": 17154
+    },
+    {
+      "epoch": 3.8988636363636364,
+      "grad_norm": 1.3623199779644002,
+      "learning_rate": 1.4377073973289698e-07,
+      "loss": 0.0061,
+      "step": 17155
+    },
+    {
+      "epoch": 3.899090909090909,
+      "grad_norm": 1.3967631405818681,
+      "learning_rate": 1.4371379578206987e-07,
+      "loss": 0.0149,
+      "step": 17156
+    },
+    {
+      "epoch": 3.899318181818182,
+      "grad_norm": 1.5612481969209329,
+      "learning_rate": 1.4365686164550822e-07,
+      "loss": 0.0097,
+      "step": 17157
+    },
+    {
+      "epoch": 3.8995454545454544,
+      "grad_norm": 1.6376020599004442,
+      "learning_rate": 1.43599937324373e-07,
+      "loss": 0.0158,
+      "step": 17158
+    },
+    {
+      "epoch": 3.8997727272727274,
+      "grad_norm": 3.9382018518266992,
+      "learning_rate": 1.435430228198248e-07,
+      "loss": 0.0078,
+      "step": 17159
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 1.089063196831999,
+      "learning_rate": 1.4348611813302424e-07,
+      "loss": 0.0063,
+      "step": 17160
+    },
+    {
+      "epoch": 3.900227272727273,
+      "grad_norm": 1.530026307501932,
+      "learning_rate": 1.434292232651318e-07,
+      "loss": 0.012,
+      "step": 17161
+    },
+    {
+      "epoch": 3.9004545454545454,
+      "grad_norm": 1.9177281632369836,
+      "learning_rate": 1.4337233821730766e-07,
+      "loss": 0.0183,
+      "step": 17162
+    },
+    {
+      "epoch": 3.900681818181818,
+      "grad_norm": 1.6873008504153022,
+      "learning_rate": 1.433154629907118e-07,
+      "loss": 0.0084,
+      "step": 17163
+    },
+    {
+      "epoch": 3.900909090909091,
+      "grad_norm": 0.9302799753837485,
+      "learning_rate": 1.4325859758650398e-07,
+      "loss": 0.0072,
+      "step": 17164
+    },
+    {
+      "epoch": 3.901136363636364,
+      "grad_norm": 1.280776992952882,
+      "learning_rate": 1.4320174200584382e-07,
+      "loss": 0.0062,
+      "step": 17165
+    },
+    {
+      "epoch": 3.9013636363636364,
+      "grad_norm": 0.8857069723727525,
+      "learning_rate": 1.4314489624989068e-07,
+      "loss": 0.0058,
+      "step": 17166
+    },
+    {
+      "epoch": 3.901590909090909,
+      "grad_norm": 0.8322865332089866,
+      "learning_rate": 1.4308806031980385e-07,
+      "loss": 0.0138,
+      "step": 17167
+    },
+    {
+      "epoch": 3.901818181818182,
+      "grad_norm": 1.91438595537973,
+      "learning_rate": 1.430312342167421e-07,
+      "loss": 0.0151,
+      "step": 17168
+    },
+    {
+      "epoch": 3.902045454545455,
+      "grad_norm": 1.4410987221801086,
+      "learning_rate": 1.4297441794186434e-07,
+      "loss": 0.009,
+      "step": 17169
+    },
+    {
+      "epoch": 3.9022727272727273,
+      "grad_norm": 1.7710157302681069,
+      "learning_rate": 1.4291761149632912e-07,
+      "loss": 0.0114,
+      "step": 17170
+    },
+    {
+      "epoch": 3.9025,
+      "grad_norm": 1.1098351713351042,
+      "learning_rate": 1.4286081488129502e-07,
+      "loss": 0.0051,
+      "step": 17171
+    },
+    {
+      "epoch": 3.902727272727273,
+      "grad_norm": 1.8006365984594341,
+      "learning_rate": 1.428040280979199e-07,
+      "loss": 0.0127,
+      "step": 17172
+    },
+    {
+      "epoch": 3.9029545454545453,
+      "grad_norm": 2.0314523709159724,
+      "learning_rate": 1.4274725114736183e-07,
+      "loss": 0.0084,
+      "step": 17173
+    },
+    {
+      "epoch": 3.9031818181818183,
+      "grad_norm": 0.2902996107496488,
+      "learning_rate": 1.4269048403077863e-07,
+      "loss": 0.001,
+      "step": 17174
+    },
+    {
+      "epoch": 3.903409090909091,
+      "grad_norm": 2.4567552468747405,
+      "learning_rate": 1.4263372674932795e-07,
+      "loss": 0.0123,
+      "step": 17175
+    },
+    {
+      "epoch": 3.903636363636364,
+      "grad_norm": 2.341097423007469,
+      "learning_rate": 1.425769793041671e-07,
+      "loss": 0.0154,
+      "step": 17176
+    },
+    {
+      "epoch": 3.9038636363636363,
+      "grad_norm": 1.6005717032332327,
+      "learning_rate": 1.425202416964532e-07,
+      "loss": 0.0142,
+      "step": 17177
+    },
+    {
+      "epoch": 3.9040909090909093,
+      "grad_norm": 1.264233217811087,
+      "learning_rate": 1.4246351392734343e-07,
+      "loss": 0.0047,
+      "step": 17178
+    },
+    {
+      "epoch": 3.904318181818182,
+      "grad_norm": 1.771031878948342,
+      "learning_rate": 1.424067959979944e-07,
+      "loss": 0.0057,
+      "step": 17179
+    },
+    {
+      "epoch": 3.9045454545454543,
+      "grad_norm": 0.7384866045478093,
+      "learning_rate": 1.4235008790956288e-07,
+      "loss": 0.0027,
+      "step": 17180
+    },
+    {
+      "epoch": 3.9047727272727273,
+      "grad_norm": 2.8799748840494863,
+      "learning_rate": 1.422933896632049e-07,
+      "loss": 0.043,
+      "step": 17181
+    },
+    {
+      "epoch": 3.9050000000000002,
+      "grad_norm": 0.5757137877088213,
+      "learning_rate": 1.4223670126007687e-07,
+      "loss": 0.0044,
+      "step": 17182
+    },
+    {
+      "epoch": 3.9052272727272728,
+      "grad_norm": 1.224231019680451,
+      "learning_rate": 1.4218002270133472e-07,
+      "loss": 0.007,
+      "step": 17183
+    },
+    {
+      "epoch": 3.9054545454545453,
+      "grad_norm": 1.37440081121208,
+      "learning_rate": 1.421233539881342e-07,
+      "loss": 0.0072,
+      "step": 17184
+    },
+    {
+      "epoch": 3.9056818181818183,
+      "grad_norm": 0.6674273373191462,
+      "learning_rate": 1.4206669512163115e-07,
+      "loss": 0.0024,
+      "step": 17185
+    },
+    {
+      "epoch": 3.9059090909090908,
+      "grad_norm": 2.6289365308046064,
+      "learning_rate": 1.420100461029805e-07,
+      "loss": 0.0295,
+      "step": 17186
+    },
+    {
+      "epoch": 3.9061363636363637,
+      "grad_norm": 2.082556286567036,
+      "learning_rate": 1.4195340693333768e-07,
+      "loss": 0.019,
+      "step": 17187
+    },
+    {
+      "epoch": 3.9063636363636363,
+      "grad_norm": 0.8060144326178493,
+      "learning_rate": 1.418967776138575e-07,
+      "loss": 0.0052,
+      "step": 17188
+    },
+    {
+      "epoch": 3.9065909090909092,
+      "grad_norm": 1.9634119738029057,
+      "learning_rate": 1.4184015814569512e-07,
+      "loss": 0.0203,
+      "step": 17189
+    },
+    {
+      "epoch": 3.9068181818181817,
+      "grad_norm": 1.4406427557639383,
+      "learning_rate": 1.4178354853000468e-07,
+      "loss": 0.011,
+      "step": 17190
+    },
+    {
+      "epoch": 3.9070454545454547,
+      "grad_norm": 3.172673881160778,
+      "learning_rate": 1.4172694876794062e-07,
+      "loss": 0.0161,
+      "step": 17191
+    },
+    {
+      "epoch": 3.9072727272727272,
+      "grad_norm": 2.756637983750958,
+      "learning_rate": 1.4167035886065723e-07,
+      "loss": 0.0217,
+      "step": 17192
+    },
+    {
+      "epoch": 3.9074999999999998,
+      "grad_norm": 0.6041233716319404,
+      "learning_rate": 1.4161377880930846e-07,
+      "loss": 0.0056,
+      "step": 17193
+    },
+    {
+      "epoch": 3.9077272727272727,
+      "grad_norm": 2.041891921174849,
+      "learning_rate": 1.4155720861504803e-07,
+      "loss": 0.0093,
+      "step": 17194
+    },
+    {
+      "epoch": 3.9079545454545457,
+      "grad_norm": 1.8744128652216163,
+      "learning_rate": 1.4150064827902953e-07,
+      "loss": 0.0111,
+      "step": 17195
+    },
+    {
+      "epoch": 3.908181818181818,
+      "grad_norm": 1.1308332342949068,
+      "learning_rate": 1.4144409780240626e-07,
+      "loss": 0.0129,
+      "step": 17196
+    },
+    {
+      "epoch": 3.9084090909090907,
+      "grad_norm": 2.279196181559447,
+      "learning_rate": 1.4138755718633158e-07,
+      "loss": 0.009,
+      "step": 17197
+    },
+    {
+      "epoch": 3.9086363636363637,
+      "grad_norm": 1.8268184538043948,
+      "learning_rate": 1.413310264319584e-07,
+      "loss": 0.0099,
+      "step": 17198
+    },
+    {
+      "epoch": 3.908863636363636,
+      "grad_norm": 3.1703554976698287,
+      "learning_rate": 1.4127450554043924e-07,
+      "loss": 0.0297,
+      "step": 17199
+    },
+    {
+      "epoch": 3.909090909090909,
+      "grad_norm": 2.6913709527263108,
+      "learning_rate": 1.4121799451292683e-07,
+      "loss": 0.0252,
+      "step": 17200
+    },
+    {
+      "epoch": 3.9093181818181817,
+      "grad_norm": 1.1077147399349176,
+      "learning_rate": 1.4116149335057345e-07,
+      "loss": 0.0112,
+      "step": 17201
+    },
+    {
+      "epoch": 3.9095454545454547,
+      "grad_norm": 2.4675469446445866,
+      "learning_rate": 1.411050020545316e-07,
+      "loss": 0.0247,
+      "step": 17202
+    },
+    {
+      "epoch": 3.909772727272727,
+      "grad_norm": 0.6464598324176936,
+      "learning_rate": 1.410485206259527e-07,
+      "loss": 0.0028,
+      "step": 17203
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 1.791980405790831,
+      "learning_rate": 1.4099204906598878e-07,
+      "loss": 0.0107,
+      "step": 17204
+    },
+    {
+      "epoch": 3.9102272727272727,
+      "grad_norm": 1.0387813447620846,
+      "learning_rate": 1.4093558737579143e-07,
+      "loss": 0.0076,
+      "step": 17205
+    },
+    {
+      "epoch": 3.910454545454545,
+      "grad_norm": 1.598341395570942,
+      "learning_rate": 1.4087913555651194e-07,
+      "loss": 0.0082,
+      "step": 17206
+    },
+    {
+      "epoch": 3.910681818181818,
+      "grad_norm": 1.3997305281198604,
+      "learning_rate": 1.408226936093014e-07,
+      "loss": 0.0126,
+      "step": 17207
+    },
+    {
+      "epoch": 3.910909090909091,
+      "grad_norm": 1.0973606699534564,
+      "learning_rate": 1.4076626153531087e-07,
+      "loss": 0.0089,
+      "step": 17208
+    },
+    {
+      "epoch": 3.9111363636363636,
+      "grad_norm": 1.1147109804390969,
+      "learning_rate": 1.4070983933569106e-07,
+      "loss": 0.0052,
+      "step": 17209
+    },
+    {
+      "epoch": 3.911363636363636,
+      "grad_norm": 1.8135743913122058,
+      "learning_rate": 1.406534270115926e-07,
+      "loss": 0.0032,
+      "step": 17210
+    },
+    {
+      "epoch": 3.911590909090909,
+      "grad_norm": 1.2044109423626437,
+      "learning_rate": 1.405970245641658e-07,
+      "loss": 0.0065,
+      "step": 17211
+    },
+    {
+      "epoch": 3.911818181818182,
+      "grad_norm": 0.5886727092957017,
+      "learning_rate": 1.4054063199456062e-07,
+      "loss": 0.0034,
+      "step": 17212
+    },
+    {
+      "epoch": 3.9120454545454546,
+      "grad_norm": 1.5423419056511183,
+      "learning_rate": 1.4048424930392718e-07,
+      "loss": 0.0089,
+      "step": 17213
+    },
+    {
+      "epoch": 3.912272727272727,
+      "grad_norm": 2.348664581911293,
+      "learning_rate": 1.4042787649341513e-07,
+      "loss": 0.0123,
+      "step": 17214
+    },
+    {
+      "epoch": 3.9125,
+      "grad_norm": 1.4835705263130523,
+      "learning_rate": 1.403715135641743e-07,
+      "loss": 0.0061,
+      "step": 17215
+    },
+    {
+      "epoch": 3.9127272727272726,
+      "grad_norm": 1.0529028293156963,
+      "learning_rate": 1.4031516051735353e-07,
+      "loss": 0.0065,
+      "step": 17216
+    },
+    {
+      "epoch": 3.9129545454545456,
+      "grad_norm": 1.049837621479144,
+      "learning_rate": 1.4025881735410232e-07,
+      "loss": 0.0058,
+      "step": 17217
+    },
+    {
+      "epoch": 3.913181818181818,
+      "grad_norm": 1.5490614044269735,
+      "learning_rate": 1.402024840755694e-07,
+      "loss": 0.0115,
+      "step": 17218
+    },
+    {
+      "epoch": 3.913409090909091,
+      "grad_norm": 0.6545038854512946,
+      "learning_rate": 1.4014616068290368e-07,
+      "loss": 0.0024,
+      "step": 17219
+    },
+    {
+      "epoch": 3.9136363636363636,
+      "grad_norm": 1.6673054875972566,
+      "learning_rate": 1.4008984717725375e-07,
+      "loss": 0.0102,
+      "step": 17220
+    },
+    {
+      "epoch": 3.9138636363636365,
+      "grad_norm": 1.9111767379643028,
+      "learning_rate": 1.4003354355976766e-07,
+      "loss": 0.0065,
+      "step": 17221
+    },
+    {
+      "epoch": 3.914090909090909,
+      "grad_norm": 2.7224760073841625,
+      "learning_rate": 1.3997724983159364e-07,
+      "loss": 0.0273,
+      "step": 17222
+    },
+    {
+      "epoch": 3.9143181818181816,
+      "grad_norm": 0.6267072506510737,
+      "learning_rate": 1.3992096599387967e-07,
+      "loss": 0.0022,
+      "step": 17223
+    },
+    {
+      "epoch": 3.9145454545454546,
+      "grad_norm": 3.1815698545819866,
+      "learning_rate": 1.3986469204777352e-07,
+      "loss": 0.0283,
+      "step": 17224
+    },
+    {
+      "epoch": 3.9147727272727275,
+      "grad_norm": 1.5653900412776558,
+      "learning_rate": 1.3980842799442263e-07,
+      "loss": 0.0083,
+      "step": 17225
+    },
+    {
+      "epoch": 3.915,
+      "grad_norm": 2.0012221132797214,
+      "learning_rate": 1.3975217383497433e-07,
+      "loss": 0.0188,
+      "step": 17226
+    },
+    {
+      "epoch": 3.9152272727272726,
+      "grad_norm": 1.5782371806863953,
+      "learning_rate": 1.3969592957057584e-07,
+      "loss": 0.0114,
+      "step": 17227
+    },
+    {
+      "epoch": 3.9154545454545455,
+      "grad_norm": 0.5851737465393586,
+      "learning_rate": 1.3963969520237396e-07,
+      "loss": 0.0018,
+      "step": 17228
+    },
+    {
+      "epoch": 3.915681818181818,
+      "grad_norm": 3.0928989584166784,
+      "learning_rate": 1.3958347073151565e-07,
+      "loss": 0.0154,
+      "step": 17229
+    },
+    {
+      "epoch": 3.915909090909091,
+      "grad_norm": 0.8869502156340455,
+      "learning_rate": 1.3952725615914701e-07,
+      "loss": 0.0073,
+      "step": 17230
+    },
+    {
+      "epoch": 3.9161363636363635,
+      "grad_norm": 0.6462283249749554,
+      "learning_rate": 1.3947105148641462e-07,
+      "loss": 0.0033,
+      "step": 17231
+    },
+    {
+      "epoch": 3.9163636363636365,
+      "grad_norm": 4.078396888321773,
+      "learning_rate": 1.3941485671446452e-07,
+      "loss": 0.0239,
+      "step": 17232
+    },
+    {
+      "epoch": 3.916590909090909,
+      "grad_norm": 1.038004815704539,
+      "learning_rate": 1.3935867184444285e-07,
+      "loss": 0.0072,
+      "step": 17233
+    },
+    {
+      "epoch": 3.916818181818182,
+      "grad_norm": 0.5978045117252839,
+      "learning_rate": 1.3930249687749488e-07,
+      "loss": 0.0024,
+      "step": 17234
+    },
+    {
+      "epoch": 3.9170454545454545,
+      "grad_norm": 3.8196527155387217,
+      "learning_rate": 1.3924633181476632e-07,
+      "loss": 0.0581,
+      "step": 17235
+    },
+    {
+      "epoch": 3.917272727272727,
+      "grad_norm": 1.6761073881163284,
+      "learning_rate": 1.3919017665740254e-07,
+      "loss": 0.0183,
+      "step": 17236
+    },
+    {
+      "epoch": 3.9175,
+      "grad_norm": 0.8991235707116042,
+      "learning_rate": 1.3913403140654862e-07,
+      "loss": 0.0089,
+      "step": 17237
+    },
+    {
+      "epoch": 3.917727272727273,
+      "grad_norm": 1.4743836772728038,
+      "learning_rate": 1.3907789606334945e-07,
+      "loss": 0.0154,
+      "step": 17238
+    },
+    {
+      "epoch": 3.9179545454545455,
+      "grad_norm": 1.3654543921430924,
+      "learning_rate": 1.390217706289497e-07,
+      "loss": 0.0051,
+      "step": 17239
+    },
+    {
+      "epoch": 3.918181818181818,
+      "grad_norm": 3.4153004757485985,
+      "learning_rate": 1.3896565510449384e-07,
+      "loss": 0.0512,
+      "step": 17240
+    },
+    {
+      "epoch": 3.918409090909091,
+      "grad_norm": 2.8837422270263007,
+      "learning_rate": 1.3890954949112627e-07,
+      "loss": 0.0297,
+      "step": 17241
+    },
+    {
+      "epoch": 3.9186363636363635,
+      "grad_norm": 1.9510809826409763,
+      "learning_rate": 1.388534537899911e-07,
+      "loss": 0.0164,
+      "step": 17242
+    },
+    {
+      "epoch": 3.9188636363636364,
+      "grad_norm": 0.41990366449506544,
+      "learning_rate": 1.3879736800223204e-07,
+      "loss": 0.002,
+      "step": 17243
+    },
+    {
+      "epoch": 3.919090909090909,
+      "grad_norm": 0.4500128807834746,
+      "learning_rate": 1.3874129212899289e-07,
+      "loss": 0.0047,
+      "step": 17244
+    },
+    {
+      "epoch": 3.919318181818182,
+      "grad_norm": 0.7901150981845673,
+      "learning_rate": 1.3868522617141708e-07,
+      "loss": 0.0036,
+      "step": 17245
+    },
+    {
+      "epoch": 3.9195454545454544,
+      "grad_norm": 1.2985394272961372,
+      "learning_rate": 1.386291701306481e-07,
+      "loss": 0.0055,
+      "step": 17246
+    },
+    {
+      "epoch": 3.9197727272727274,
+      "grad_norm": 2.508017800481271,
+      "learning_rate": 1.385731240078287e-07,
+      "loss": 0.0114,
+      "step": 17247
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 0.5003178363073534,
+      "learning_rate": 1.3851708780410187e-07,
+      "loss": 0.0039,
+      "step": 17248
+    },
+    {
+      "epoch": 3.920227272727273,
+      "grad_norm": 2.144189528984042,
+      "learning_rate": 1.3846106152061043e-07,
+      "loss": 0.0175,
+      "step": 17249
+    },
+    {
+      "epoch": 3.9204545454545454,
+      "grad_norm": 0.7226913755318007,
+      "learning_rate": 1.3840504515849665e-07,
+      "loss": 0.0042,
+      "step": 17250
+    },
+    {
+      "epoch": 3.9206818181818184,
+      "grad_norm": 0.3878682681206709,
+      "learning_rate": 1.383490387189032e-07,
+      "loss": 0.0023,
+      "step": 17251
+    },
+    {
+      "epoch": 3.920909090909091,
+      "grad_norm": 1.3177470174037411,
+      "learning_rate": 1.3829304220297157e-07,
+      "loss": 0.0158,
+      "step": 17252
+    },
+    {
+      "epoch": 3.9211363636363634,
+      "grad_norm": 0.9862907825377537,
+      "learning_rate": 1.3823705561184404e-07,
+      "loss": 0.0084,
+      "step": 17253
+    },
+    {
+      "epoch": 3.9213636363636364,
+      "grad_norm": 1.4087352561868112,
+      "learning_rate": 1.3818107894666208e-07,
+      "loss": 0.0049,
+      "step": 17254
+    },
+    {
+      "epoch": 3.9215909090909093,
+      "grad_norm": 2.120666386766986,
+      "learning_rate": 1.3812511220856724e-07,
+      "loss": 0.0135,
+      "step": 17255
+    },
+    {
+      "epoch": 3.921818181818182,
+      "grad_norm": 2.047548891015841,
+      "learning_rate": 1.3806915539870068e-07,
+      "loss": 0.0197,
+      "step": 17256
+    },
+    {
+      "epoch": 3.9220454545454544,
+      "grad_norm": 1.6508720494415923,
+      "learning_rate": 1.3801320851820363e-07,
+      "loss": 0.0116,
+      "step": 17257
+    },
+    {
+      "epoch": 3.9222727272727274,
+      "grad_norm": 2.0637595150408266,
+      "learning_rate": 1.3795727156821683e-07,
+      "loss": 0.0062,
+      "step": 17258
+    },
+    {
+      "epoch": 3.9225,
+      "grad_norm": 1.1387966412713144,
+      "learning_rate": 1.3790134454988113e-07,
+      "loss": 0.0056,
+      "step": 17259
+    },
+    {
+      "epoch": 3.922727272727273,
+      "grad_norm": 0.8795473094766364,
+      "learning_rate": 1.3784542746433662e-07,
+      "loss": 0.0067,
+      "step": 17260
+    },
+    {
+      "epoch": 3.9229545454545454,
+      "grad_norm": 1.100539931149628,
+      "learning_rate": 1.377895203127238e-07,
+      "loss": 0.0091,
+      "step": 17261
+    },
+    {
+      "epoch": 3.9231818181818183,
+      "grad_norm": 0.9584123120804074,
+      "learning_rate": 1.377336230961826e-07,
+      "loss": 0.0042,
+      "step": 17262
+    },
+    {
+      "epoch": 3.923409090909091,
+      "grad_norm": 2.5174100655446456,
+      "learning_rate": 1.3767773581585286e-07,
+      "loss": 0.0372,
+      "step": 17263
+    },
+    {
+      "epoch": 3.923636363636364,
+      "grad_norm": 0.9044194155184898,
+      "learning_rate": 1.3762185847287446e-07,
+      "loss": 0.0088,
+      "step": 17264
+    },
+    {
+      "epoch": 3.9238636363636363,
+      "grad_norm": 1.8846971788092666,
+      "learning_rate": 1.3756599106838652e-07,
+      "loss": 0.0158,
+      "step": 17265
+    },
+    {
+      "epoch": 3.924090909090909,
+      "grad_norm": 1.7225117127348706,
+      "learning_rate": 1.375101336035284e-07,
+      "loss": 0.0178,
+      "step": 17266
+    },
+    {
+      "epoch": 3.924318181818182,
+      "grad_norm": 1.085785790988991,
+      "learning_rate": 1.3745428607943908e-07,
+      "loss": 0.0071,
+      "step": 17267
+    },
+    {
+      "epoch": 3.924545454545455,
+      "grad_norm": 1.5370483579064707,
+      "learning_rate": 1.3739844849725742e-07,
+      "loss": 0.0029,
+      "step": 17268
+    },
+    {
+      "epoch": 3.9247727272727273,
+      "grad_norm": 2.200850234931457,
+      "learning_rate": 1.373426208581221e-07,
+      "loss": 0.0147,
+      "step": 17269
+    },
+    {
+      "epoch": 3.925,
+      "grad_norm": 0.5584214029396993,
+      "learning_rate": 1.3728680316317153e-07,
+      "loss": 0.0089,
+      "step": 17270
+    },
+    {
+      "epoch": 3.925227272727273,
+      "grad_norm": 1.342800550386445,
+      "learning_rate": 1.3723099541354388e-07,
+      "loss": 0.0068,
+      "step": 17271
+    },
+    {
+      "epoch": 3.9254545454545453,
+      "grad_norm": 1.106321680645541,
+      "learning_rate": 1.3717519761037732e-07,
+      "loss": 0.0059,
+      "step": 17272
+    },
+    {
+      "epoch": 3.9256818181818183,
+      "grad_norm": 1.2004928348650117,
+      "learning_rate": 1.3711940975480946e-07,
+      "loss": 0.0081,
+      "step": 17273
+    },
+    {
+      "epoch": 3.925909090909091,
+      "grad_norm": 2.071440360168159,
+      "learning_rate": 1.370636318479779e-07,
+      "loss": 0.0104,
+      "step": 17274
+    },
+    {
+      "epoch": 3.9261363636363638,
+      "grad_norm": 1.3502917294707721,
+      "learning_rate": 1.3700786389102015e-07,
+      "loss": 0.0085,
+      "step": 17275
+    },
+    {
+      "epoch": 3.9263636363636363,
+      "grad_norm": 1.0927244808308796,
+      "learning_rate": 1.3695210588507347e-07,
+      "loss": 0.0135,
+      "step": 17276
+    },
+    {
+      "epoch": 3.9265909090909092,
+      "grad_norm": 1.6275024956287187,
+      "learning_rate": 1.3689635783127494e-07,
+      "loss": 0.0079,
+      "step": 17277
+    },
+    {
+      "epoch": 3.9268181818181818,
+      "grad_norm": 1.3068137946564733,
+      "learning_rate": 1.3684061973076105e-07,
+      "loss": 0.0065,
+      "step": 17278
+    },
+    {
+      "epoch": 3.9270454545454543,
+      "grad_norm": 0.7270110708941594,
+      "learning_rate": 1.3678489158466853e-07,
+      "loss": 0.0037,
+      "step": 17279
+    },
+    {
+      "epoch": 3.9272727272727272,
+      "grad_norm": 2.705845974232974,
+      "learning_rate": 1.367291733941338e-07,
+      "loss": 0.0537,
+      "step": 17280
+    },
+    {
+      "epoch": 3.9275,
+      "grad_norm": 1.5906686739360318,
+      "learning_rate": 1.3667346516029327e-07,
+      "loss": 0.0076,
+      "step": 17281
+    },
+    {
+      "epoch": 3.9277272727272727,
+      "grad_norm": 1.1117591078380518,
+      "learning_rate": 1.3661776688428253e-07,
+      "loss": 0.0063,
+      "step": 17282
+    },
+    {
+      "epoch": 3.9279545454545453,
+      "grad_norm": 2.056719664269329,
+      "learning_rate": 1.365620785672376e-07,
+      "loss": 0.0077,
+      "step": 17283
+    },
+    {
+      "epoch": 3.928181818181818,
+      "grad_norm": 0.723973389857322,
+      "learning_rate": 1.3650640021029394e-07,
+      "loss": 0.0029,
+      "step": 17284
+    },
+    {
+      "epoch": 3.928409090909091,
+      "grad_norm": 1.400685777599219,
+      "learning_rate": 1.3645073181458683e-07,
+      "loss": 0.0073,
+      "step": 17285
+    },
+    {
+      "epoch": 3.9286363636363637,
+      "grad_norm": 1.3118768334805053,
+      "learning_rate": 1.3639507338125204e-07,
+      "loss": 0.0098,
+      "step": 17286
+    },
+    {
+      "epoch": 3.9288636363636362,
+      "grad_norm": 0.19191741112617444,
+      "learning_rate": 1.3633942491142388e-07,
+      "loss": 0.001,
+      "step": 17287
+    },
+    {
+      "epoch": 3.929090909090909,
+      "grad_norm": 0.4914963805465359,
+      "learning_rate": 1.362837864062374e-07,
+      "loss": 0.0023,
+      "step": 17288
+    },
+    {
+      "epoch": 3.9293181818181817,
+      "grad_norm": 2.201027406824066,
+      "learning_rate": 1.362281578668271e-07,
+      "loss": 0.0201,
+      "step": 17289
+    },
+    {
+      "epoch": 3.9295454545454547,
+      "grad_norm": 1.0768827822466835,
+      "learning_rate": 1.3617253929432757e-07,
+      "loss": 0.0111,
+      "step": 17290
+    },
+    {
+      "epoch": 3.929772727272727,
+      "grad_norm": 1.2433245029745958,
+      "learning_rate": 1.361169306898726e-07,
+      "loss": 0.0055,
+      "step": 17291
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 0.7983693538161519,
+      "learning_rate": 1.3606133205459628e-07,
+      "loss": 0.0092,
+      "step": 17292
+    },
+    {
+      "epoch": 3.9302272727272727,
+      "grad_norm": 1.4033867328518235,
+      "learning_rate": 1.3600574338963246e-07,
+      "loss": 0.0176,
+      "step": 17293
+    },
+    {
+      "epoch": 3.9304545454545456,
+      "grad_norm": 1.9175913434111613,
+      "learning_rate": 1.3595016469611477e-07,
+      "loss": 0.017,
+      "step": 17294
+    },
+    {
+      "epoch": 3.930681818181818,
+      "grad_norm": 1.1649698506219999,
+      "learning_rate": 1.3589459597517625e-07,
+      "loss": 0.0069,
+      "step": 17295
+    },
+    {
+      "epoch": 3.9309090909090907,
+      "grad_norm": 1.5168311148507367,
+      "learning_rate": 1.3583903722795015e-07,
+      "loss": 0.0102,
+      "step": 17296
+    },
+    {
+      "epoch": 3.9311363636363637,
+      "grad_norm": 1.1037091687575813,
+      "learning_rate": 1.3578348845556957e-07,
+      "loss": 0.0081,
+      "step": 17297
+    },
+    {
+      "epoch": 3.9313636363636366,
+      "grad_norm": 0.2849576321349123,
+      "learning_rate": 1.3572794965916708e-07,
+      "loss": 0.0026,
+      "step": 17298
+    },
+    {
+      "epoch": 3.931590909090909,
+      "grad_norm": 0.8482656804321271,
+      "learning_rate": 1.3567242083987527e-07,
+      "loss": 0.0108,
+      "step": 17299
+    },
+    {
+      "epoch": 3.9318181818181817,
+      "grad_norm": 1.7780714055855142,
+      "learning_rate": 1.3561690199882647e-07,
+      "loss": 0.0133,
+      "step": 17300
+    },
+    {
+      "epoch": 3.9320454545454546,
+      "grad_norm": 1.454537263529175,
+      "learning_rate": 1.3556139313715285e-07,
+      "loss": 0.005,
+      "step": 17301
+    },
+    {
+      "epoch": 3.932272727272727,
+      "grad_norm": 1.2409928433442103,
+      "learning_rate": 1.3550589425598628e-07,
+      "loss": 0.007,
+      "step": 17302
+    },
+    {
+      "epoch": 3.9325,
+      "grad_norm": 1.0759008511451318,
+      "learning_rate": 1.354504053564587e-07,
+      "loss": 0.01,
+      "step": 17303
+    },
+    {
+      "epoch": 3.9327272727272726,
+      "grad_norm": 4.959222233451715,
+      "learning_rate": 1.353949264397012e-07,
+      "loss": 0.0192,
+      "step": 17304
+    },
+    {
+      "epoch": 3.9329545454545456,
+      "grad_norm": 1.265201872851688,
+      "learning_rate": 1.3533945750684536e-07,
+      "loss": 0.0124,
+      "step": 17305
+    },
+    {
+      "epoch": 3.933181818181818,
+      "grad_norm": 2.0508390187400574,
+      "learning_rate": 1.352839985590222e-07,
+      "loss": 0.0128,
+      "step": 17306
+    },
+    {
+      "epoch": 3.933409090909091,
+      "grad_norm": 2.3059937315520447,
+      "learning_rate": 1.352285495973627e-07,
+      "loss": 0.0178,
+      "step": 17307
+    },
+    {
+      "epoch": 3.9336363636363636,
+      "grad_norm": 1.6131666656795147,
+      "learning_rate": 1.3517311062299767e-07,
+      "loss": 0.0053,
+      "step": 17308
+    },
+    {
+      "epoch": 3.933863636363636,
+      "grad_norm": 4.306147685918921,
+      "learning_rate": 1.351176816370573e-07,
+      "loss": 0.0425,
+      "step": 17309
+    },
+    {
+      "epoch": 3.934090909090909,
+      "grad_norm": 1.5887644452932994,
+      "learning_rate": 1.350622626406721e-07,
+      "loss": 0.0152,
+      "step": 17310
+    },
+    {
+      "epoch": 3.934318181818182,
+      "grad_norm": 0.5994719588252106,
+      "learning_rate": 1.3500685363497204e-07,
+      "loss": 0.0026,
+      "step": 17311
+    },
+    {
+      "epoch": 3.9345454545454546,
+      "grad_norm": 0.7398419670238718,
+      "learning_rate": 1.3495145462108729e-07,
+      "loss": 0.0063,
+      "step": 17312
+    },
+    {
+      "epoch": 3.934772727272727,
+      "grad_norm": 2.2096976742324914,
+      "learning_rate": 1.348960656001471e-07,
+      "loss": 0.0234,
+      "step": 17313
+    },
+    {
+      "epoch": 3.935,
+      "grad_norm": 0.44257496445842043,
+      "learning_rate": 1.3484068657328117e-07,
+      "loss": 0.0017,
+      "step": 17314
+    },
+    {
+      "epoch": 3.9352272727272726,
+      "grad_norm": 3.0163470908240892,
+      "learning_rate": 1.347853175416186e-07,
+      "loss": 0.0126,
+      "step": 17315
+    },
+    {
+      "epoch": 3.9354545454545455,
+      "grad_norm": 2.0752449646873625,
+      "learning_rate": 1.3472995850628903e-07,
+      "loss": 0.0198,
+      "step": 17316
+    },
+    {
+      "epoch": 3.935681818181818,
+      "grad_norm": 1.9934665077857385,
+      "learning_rate": 1.3467460946842073e-07,
+      "loss": 0.0169,
+      "step": 17317
+    },
+    {
+      "epoch": 3.935909090909091,
+      "grad_norm": 3.111988088652659,
+      "learning_rate": 1.3461927042914254e-07,
+      "loss": 0.0246,
+      "step": 17318
+    },
+    {
+      "epoch": 3.9361363636363635,
+      "grad_norm": 1.1295812578126811,
+      "learning_rate": 1.3456394138958294e-07,
+      "loss": 0.0078,
+      "step": 17319
+    },
+    {
+      "epoch": 3.9363636363636365,
+      "grad_norm": 2.6939948763557653,
+      "learning_rate": 1.3450862235087023e-07,
+      "loss": 0.0157,
+      "step": 17320
+    },
+    {
+      "epoch": 3.936590909090909,
+      "grad_norm": 3.503047684540802,
+      "learning_rate": 1.3445331331413256e-07,
+      "loss": 0.0154,
+      "step": 17321
+    },
+    {
+      "epoch": 3.9368181818181816,
+      "grad_norm": 4.6034707563189485,
+      "learning_rate": 1.3439801428049743e-07,
+      "loss": 0.0375,
+      "step": 17322
+    },
+    {
+      "epoch": 3.9370454545454545,
+      "grad_norm": 0.7647667221292935,
+      "learning_rate": 1.3434272525109274e-07,
+      "loss": 0.0019,
+      "step": 17323
+    },
+    {
+      "epoch": 3.9372727272727275,
+      "grad_norm": 2.0087807954436157,
+      "learning_rate": 1.342874462270459e-07,
+      "loss": 0.0189,
+      "step": 17324
+    },
+    {
+      "epoch": 3.9375,
+      "grad_norm": 1.6788857370442745,
+      "learning_rate": 1.3423217720948428e-07,
+      "loss": 0.0128,
+      "step": 17325
+    },
+    {
+      "epoch": 3.9377272727272725,
+      "grad_norm": 1.422139235243473,
+      "learning_rate": 1.3417691819953456e-07,
+      "loss": 0.0124,
+      "step": 17326
+    },
+    {
+      "epoch": 3.9379545454545455,
+      "grad_norm": 2.1801910526940658,
+      "learning_rate": 1.3412166919832374e-07,
+      "loss": 0.0146,
+      "step": 17327
+    },
+    {
+      "epoch": 3.9381818181818184,
+      "grad_norm": 3.4266626001079605,
+      "learning_rate": 1.3406643020697856e-07,
+      "loss": 0.0184,
+      "step": 17328
+    },
+    {
+      "epoch": 3.938409090909091,
+      "grad_norm": 0.9498800569533413,
+      "learning_rate": 1.340112012266253e-07,
+      "loss": 0.0071,
+      "step": 17329
+    },
+    {
+      "epoch": 3.9386363636363635,
+      "grad_norm": 2.2104318723669665,
+      "learning_rate": 1.3395598225839023e-07,
+      "loss": 0.0108,
+      "step": 17330
+    },
+    {
+      "epoch": 3.9388636363636365,
+      "grad_norm": 0.6941967754881797,
+      "learning_rate": 1.339007733033993e-07,
+      "loss": 0.0024,
+      "step": 17331
+    },
+    {
+      "epoch": 3.939090909090909,
+      "grad_norm": 1.9664883300992924,
+      "learning_rate": 1.3384557436277848e-07,
+      "loss": 0.0188,
+      "step": 17332
+    },
+    {
+      "epoch": 3.939318181818182,
+      "grad_norm": 2.7007264096229084,
+      "learning_rate": 1.3379038543765314e-07,
+      "loss": 0.0143,
+      "step": 17333
+    },
+    {
+      "epoch": 3.9395454545454545,
+      "grad_norm": 2.9542242220900277,
+      "learning_rate": 1.3373520652914903e-07,
+      "loss": 0.0112,
+      "step": 17334
+    },
+    {
+      "epoch": 3.9397727272727274,
+      "grad_norm": 0.4935884321030353,
+      "learning_rate": 1.3368003763839102e-07,
+      "loss": 0.0031,
+      "step": 17335
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 1.6943114451145695,
+      "learning_rate": 1.3362487876650414e-07,
+      "loss": 0.0133,
+      "step": 17336
+    },
+    {
+      "epoch": 3.940227272727273,
+      "grad_norm": 1.7612442483691702,
+      "learning_rate": 1.3356972991461317e-07,
+      "loss": 0.0177,
+      "step": 17337
+    },
+    {
+      "epoch": 3.9404545454545454,
+      "grad_norm": 0.7102501836948099,
+      "learning_rate": 1.33514591083843e-07,
+      "loss": 0.0063,
+      "step": 17338
+    },
+    {
+      "epoch": 3.940681818181818,
+      "grad_norm": 1.2515073114661537,
+      "learning_rate": 1.3345946227531756e-07,
+      "loss": 0.0143,
+      "step": 17339
+    },
+    {
+      "epoch": 3.940909090909091,
+      "grad_norm": 2.216834552415995,
+      "learning_rate": 1.3340434349016129e-07,
+      "loss": 0.0156,
+      "step": 17340
+    },
+    {
+      "epoch": 3.941136363636364,
+      "grad_norm": 1.152668402622695,
+      "learning_rate": 1.3334923472949798e-07,
+      "loss": 0.0065,
+      "step": 17341
+    },
+    {
+      "epoch": 3.9413636363636364,
+      "grad_norm": 2.546153067441664,
+      "learning_rate": 1.3329413599445158e-07,
+      "loss": 0.013,
+      "step": 17342
+    },
+    {
+      "epoch": 3.941590909090909,
+      "grad_norm": 0.9907009692597812,
+      "learning_rate": 1.332390472861457e-07,
+      "loss": 0.0051,
+      "step": 17343
+    },
+    {
+      "epoch": 3.941818181818182,
+      "grad_norm": 1.8127059464124655,
+      "learning_rate": 1.3318396860570328e-07,
+      "loss": 0.0152,
+      "step": 17344
+    },
+    {
+      "epoch": 3.9420454545454544,
+      "grad_norm": 1.9269513384063808,
+      "learning_rate": 1.3312889995424787e-07,
+      "loss": 0.0068,
+      "step": 17345
+    },
+    {
+      "epoch": 3.9422727272727274,
+      "grad_norm": 0.7897696944742685,
+      "learning_rate": 1.330738413329024e-07,
+      "loss": 0.0041,
+      "step": 17346
+    },
+    {
+      "epoch": 3.9425,
+      "grad_norm": 2.3233455108285783,
+      "learning_rate": 1.3301879274278958e-07,
+      "loss": 0.0259,
+      "step": 17347
+    },
+    {
+      "epoch": 3.942727272727273,
+      "grad_norm": 3.072912258525486,
+      "learning_rate": 1.3296375418503176e-07,
+      "loss": 0.0255,
+      "step": 17348
+    },
+    {
+      "epoch": 3.9429545454545454,
+      "grad_norm": 0.8822945586600236,
+      "learning_rate": 1.3290872566075142e-07,
+      "loss": 0.0059,
+      "step": 17349
+    },
+    {
+      "epoch": 3.9431818181818183,
+      "grad_norm": 2.79069891379271,
+      "learning_rate": 1.3285370717107067e-07,
+      "loss": 0.0207,
+      "step": 17350
+    },
+    {
+      "epoch": 3.943409090909091,
+      "grad_norm": 0.7439337794994968,
+      "learning_rate": 1.3279869871711142e-07,
+      "loss": 0.0051,
+      "step": 17351
+    },
+    {
+      "epoch": 3.9436363636363634,
+      "grad_norm": 1.2333539650617231,
+      "learning_rate": 1.3274370029999552e-07,
+      "loss": 0.0068,
+      "step": 17352
+    },
+    {
+      "epoch": 3.9438636363636363,
+      "grad_norm": 1.5866386149738698,
+      "learning_rate": 1.326887119208442e-07,
+      "loss": 0.0078,
+      "step": 17353
+    },
+    {
+      "epoch": 3.9440909090909093,
+      "grad_norm": 1.241388843921308,
+      "learning_rate": 1.3263373358077897e-07,
+      "loss": 0.0043,
+      "step": 17354
+    },
+    {
+      "epoch": 3.944318181818182,
+      "grad_norm": 1.7645353937752686,
+      "learning_rate": 1.3257876528092087e-07,
+      "loss": 0.0124,
+      "step": 17355
+    },
+    {
+      "epoch": 3.9445454545454544,
+      "grad_norm": 1.274443459696372,
+      "learning_rate": 1.3252380702239096e-07,
+      "loss": 0.0044,
+      "step": 17356
+    },
+    {
+      "epoch": 3.9447727272727273,
+      "grad_norm": 0.9132426255137082,
+      "learning_rate": 1.3246885880630965e-07,
+      "loss": 0.0044,
+      "step": 17357
+    },
+    {
+      "epoch": 3.945,
+      "grad_norm": 0.24616503734400672,
+      "learning_rate": 1.3241392063379758e-07,
+      "loss": 0.0009,
+      "step": 17358
+    },
+    {
+      "epoch": 3.945227272727273,
+      "grad_norm": 1.5511151980299298,
+      "learning_rate": 1.3235899250597505e-07,
+      "loss": 0.0067,
+      "step": 17359
+    },
+    {
+      "epoch": 3.9454545454545453,
+      "grad_norm": 1.1762307848330955,
+      "learning_rate": 1.3230407442396211e-07,
+      "loss": 0.0155,
+      "step": 17360
+    },
+    {
+      "epoch": 3.9456818181818183,
+      "grad_norm": 1.1420490228924922,
+      "learning_rate": 1.3224916638887862e-07,
+      "loss": 0.0057,
+      "step": 17361
+    },
+    {
+      "epoch": 3.945909090909091,
+      "grad_norm": 3.171071151621929,
+      "learning_rate": 1.3219426840184436e-07,
+      "loss": 0.0175,
+      "step": 17362
+    },
+    {
+      "epoch": 3.9461363636363638,
+      "grad_norm": 2.6118931552463405,
+      "learning_rate": 1.321393804639786e-07,
+      "loss": 0.0222,
+      "step": 17363
+    },
+    {
+      "epoch": 3.9463636363636363,
+      "grad_norm": 1.0057652233537133,
+      "learning_rate": 1.320845025764008e-07,
+      "loss": 0.005,
+      "step": 17364
+    },
+    {
+      "epoch": 3.946590909090909,
+      "grad_norm": 1.4554578770262034,
+      "learning_rate": 1.3202963474023008e-07,
+      "loss": 0.0114,
+      "step": 17365
+    },
+    {
+      "epoch": 3.946818181818182,
+      "grad_norm": 0.623645663009206,
+      "learning_rate": 1.3197477695658496e-07,
+      "loss": 0.0033,
+      "step": 17366
+    },
+    {
+      "epoch": 3.9470454545454547,
+      "grad_norm": 1.6557145975061534,
+      "learning_rate": 1.319199292265843e-07,
+      "loss": 0.0128,
+      "step": 17367
+    },
+    {
+      "epoch": 3.9472727272727273,
+      "grad_norm": 1.1636050397643534,
+      "learning_rate": 1.318650915513464e-07,
+      "loss": 0.0095,
+      "step": 17368
+    },
+    {
+      "epoch": 3.9475,
+      "grad_norm": 4.1215536888576665,
+      "learning_rate": 1.3181026393198986e-07,
+      "loss": 0.0179,
+      "step": 17369
+    },
+    {
+      "epoch": 3.9477272727272728,
+      "grad_norm": 0.9835184798619734,
+      "learning_rate": 1.317554463696323e-07,
+      "loss": 0.0054,
+      "step": 17370
+    },
+    {
+      "epoch": 3.9479545454545457,
+      "grad_norm": 1.4693023343786442,
+      "learning_rate": 1.3170063886539173e-07,
+      "loss": 0.0043,
+      "step": 17371
+    },
+    {
+      "epoch": 3.9481818181818182,
+      "grad_norm": 1.058735699463343,
+      "learning_rate": 1.316458414203857e-07,
+      "loss": 0.0059,
+      "step": 17372
+    },
+    {
+      "epoch": 3.9484090909090908,
+      "grad_norm": 7.245471948165919,
+      "learning_rate": 1.315910540357316e-07,
+      "loss": 0.0129,
+      "step": 17373
+    },
+    {
+      "epoch": 3.9486363636363637,
+      "grad_norm": 2.278563552918246,
+      "learning_rate": 1.315362767125468e-07,
+      "loss": 0.0115,
+      "step": 17374
+    },
+    {
+      "epoch": 3.9488636363636362,
+      "grad_norm": 1.5043008186876416,
+      "learning_rate": 1.3148150945194817e-07,
+      "loss": 0.008,
+      "step": 17375
+    },
+    {
+      "epoch": 3.949090909090909,
+      "grad_norm": 2.629037948893517,
+      "learning_rate": 1.3142675225505253e-07,
+      "loss": 0.0212,
+      "step": 17376
+    },
+    {
+      "epoch": 3.9493181818181817,
+      "grad_norm": 0.7882751870799001,
+      "learning_rate": 1.313720051229766e-07,
+      "loss": 0.0044,
+      "step": 17377
+    },
+    {
+      "epoch": 3.9495454545454547,
+      "grad_norm": 1.7600957799050694,
+      "learning_rate": 1.313172680568367e-07,
+      "loss": 0.0156,
+      "step": 17378
+    },
+    {
+      "epoch": 3.949772727272727,
+      "grad_norm": 2.1171982693766243,
+      "learning_rate": 1.3126254105774886e-07,
+      "loss": 0.0352,
+      "step": 17379
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 1.6493537955748563,
+      "learning_rate": 1.3120782412682922e-07,
+      "loss": 0.0169,
+      "step": 17380
+    },
+    {
+      "epoch": 3.9502272727272727,
+      "grad_norm": 1.2177650065781074,
+      "learning_rate": 1.311531172651935e-07,
+      "loss": 0.0071,
+      "step": 17381
+    },
+    {
+      "epoch": 3.950454545454545,
+      "grad_norm": 2.3312492197079786,
+      "learning_rate": 1.3109842047395746e-07,
+      "loss": 0.0096,
+      "step": 17382
+    },
+    {
+      "epoch": 3.950681818181818,
+      "grad_norm": 3.5517673918781534,
+      "learning_rate": 1.3104373375423604e-07,
+      "loss": 0.0199,
+      "step": 17383
+    },
+    {
+      "epoch": 3.950909090909091,
+      "grad_norm": 1.2261000491906613,
+      "learning_rate": 1.3098905710714475e-07,
+      "loss": 0.0102,
+      "step": 17384
+    },
+    {
+      "epoch": 3.9511363636363637,
+      "grad_norm": 1.2239333297926298,
+      "learning_rate": 1.3093439053379835e-07,
+      "loss": 0.0068,
+      "step": 17385
+    },
+    {
+      "epoch": 3.951363636363636,
+      "grad_norm": 1.0651672760222992,
+      "learning_rate": 1.308797340353117e-07,
+      "loss": 0.0098,
+      "step": 17386
+    },
+    {
+      "epoch": 3.951590909090909,
+      "grad_norm": 1.4751777609737136,
+      "learning_rate": 1.308250876127995e-07,
+      "loss": 0.0173,
+      "step": 17387
+    },
+    {
+      "epoch": 3.9518181818181817,
+      "grad_norm": 2.128848879020357,
+      "learning_rate": 1.3077045126737576e-07,
+      "loss": 0.0169,
+      "step": 17388
+    },
+    {
+      "epoch": 3.9520454545454546,
+      "grad_norm": 2.3758874278993876,
+      "learning_rate": 1.3071582500015465e-07,
+      "loss": 0.0141,
+      "step": 17389
+    },
+    {
+      "epoch": 3.952272727272727,
+      "grad_norm": 2.3177112979695447,
+      "learning_rate": 1.3066120881225033e-07,
+      "loss": 0.0132,
+      "step": 17390
+    },
+    {
+      "epoch": 3.9525,
+      "grad_norm": 2.1934975684686684,
+      "learning_rate": 1.306066027047763e-07,
+      "loss": 0.0075,
+      "step": 17391
+    },
+    {
+      "epoch": 3.9527272727272726,
+      "grad_norm": 1.9546657911552323,
+      "learning_rate": 1.3055200667884612e-07,
+      "loss": 0.015,
+      "step": 17392
+    },
+    {
+      "epoch": 3.9529545454545456,
+      "grad_norm": 0.3794060252782392,
+      "learning_rate": 1.3049742073557317e-07,
+      "loss": 0.0023,
+      "step": 17393
+    },
+    {
+      "epoch": 3.953181818181818,
+      "grad_norm": 1.347399679338033,
+      "learning_rate": 1.3044284487607058e-07,
+      "loss": 0.0085,
+      "step": 17394
+    },
+    {
+      "epoch": 3.9534090909090907,
+      "grad_norm": 1.0680113948638437,
+      "learning_rate": 1.303882791014513e-07,
+      "loss": 0.0056,
+      "step": 17395
+    },
+    {
+      "epoch": 3.9536363636363636,
+      "grad_norm": 1.9957734755080194,
+      "learning_rate": 1.303337234128277e-07,
+      "loss": 0.0168,
+      "step": 17396
+    },
+    {
+      "epoch": 3.9538636363636366,
+      "grad_norm": 4.121615497470089,
+      "learning_rate": 1.3027917781131251e-07,
+      "loss": 0.0539,
+      "step": 17397
+    },
+    {
+      "epoch": 3.954090909090909,
+      "grad_norm": 0.7488187426798657,
+      "learning_rate": 1.3022464229801802e-07,
+      "loss": 0.0028,
+      "step": 17398
+    },
+    {
+      "epoch": 3.9543181818181816,
+      "grad_norm": 1.0901425969601097,
+      "learning_rate": 1.3017011687405613e-07,
+      "loss": 0.0116,
+      "step": 17399
+    },
+    {
+      "epoch": 3.9545454545454546,
+      "grad_norm": 1.8284817963289868,
+      "learning_rate": 1.3011560154053914e-07,
+      "loss": 0.0065,
+      "step": 17400
+    },
+    {
+      "epoch": 3.954772727272727,
+      "grad_norm": 2.2676095225166653,
+      "learning_rate": 1.3006109629857815e-07,
+      "loss": 0.0332,
+      "step": 17401
+    },
+    {
+      "epoch": 3.955,
+      "grad_norm": 0.4555256941807612,
+      "learning_rate": 1.3000660114928497e-07,
+      "loss": 0.0018,
+      "step": 17402
+    },
+    {
+      "epoch": 3.9552272727272726,
+      "grad_norm": 0.741830294125309,
+      "learning_rate": 1.2995211609377065e-07,
+      "loss": 0.0044,
+      "step": 17403
+    },
+    {
+      "epoch": 3.9554545454545456,
+      "grad_norm": 1.3849933017095155,
+      "learning_rate": 1.2989764113314636e-07,
+      "loss": 0.0156,
+      "step": 17404
+    },
+    {
+      "epoch": 3.955681818181818,
+      "grad_norm": 2.1113606359995574,
+      "learning_rate": 1.2984317626852295e-07,
+      "loss": 0.014,
+      "step": 17405
+    },
+    {
+      "epoch": 3.955909090909091,
+      "grad_norm": 1.846750144576529,
+      "learning_rate": 1.29788721501011e-07,
+      "loss": 0.0109,
+      "step": 17406
+    },
+    {
+      "epoch": 3.9561363636363636,
+      "grad_norm": 0.477476389105197,
+      "learning_rate": 1.2973427683172106e-07,
+      "loss": 0.0052,
+      "step": 17407
+    },
+    {
+      "epoch": 3.9563636363636365,
+      "grad_norm": 0.9555335868764708,
+      "learning_rate": 1.2967984226176316e-07,
+      "loss": 0.0109,
+      "step": 17408
+    },
+    {
+      "epoch": 3.956590909090909,
+      "grad_norm": 0.7206750254128398,
+      "learning_rate": 1.2962541779224757e-07,
+      "loss": 0.0042,
+      "step": 17409
+    },
+    {
+      "epoch": 3.956818181818182,
+      "grad_norm": 1.3816157378512468,
+      "learning_rate": 1.2957100342428378e-07,
+      "loss": 0.007,
+      "step": 17410
+    },
+    {
+      "epoch": 3.9570454545454545,
+      "grad_norm": 1.6790340997455282,
+      "learning_rate": 1.295165991589816e-07,
+      "loss": 0.0051,
+      "step": 17411
+    },
+    {
+      "epoch": 3.957272727272727,
+      "grad_norm": 1.2950528664071506,
+      "learning_rate": 1.2946220499745033e-07,
+      "loss": 0.0061,
+      "step": 17412
+    },
+    {
+      "epoch": 3.9575,
+      "grad_norm": 2.3067021027035883,
+      "learning_rate": 1.2940782094079944e-07,
+      "loss": 0.0143,
+      "step": 17413
+    },
+    {
+      "epoch": 3.957727272727273,
+      "grad_norm": 1.595659866400852,
+      "learning_rate": 1.2935344699013745e-07,
+      "loss": 0.011,
+      "step": 17414
+    },
+    {
+      "epoch": 3.9579545454545455,
+      "grad_norm": 1.3570738027813773,
+      "learning_rate": 1.292990831465734e-07,
+      "loss": 0.0087,
+      "step": 17415
+    },
+    {
+      "epoch": 3.958181818181818,
+      "grad_norm": 1.165161392364925,
+      "learning_rate": 1.292447294112159e-07,
+      "loss": 0.0141,
+      "step": 17416
+    },
+    {
+      "epoch": 3.958409090909091,
+      "grad_norm": 1.5596970025844494,
+      "learning_rate": 1.2919038578517334e-07,
+      "loss": 0.019,
+      "step": 17417
+    },
+    {
+      "epoch": 3.9586363636363635,
+      "grad_norm": 1.9043311053430618,
+      "learning_rate": 1.2913605226955368e-07,
+      "loss": 0.0138,
+      "step": 17418
+    },
+    {
+      "epoch": 3.9588636363636365,
+      "grad_norm": 1.6474184325523265,
+      "learning_rate": 1.29081728865465e-07,
+      "loss": 0.0145,
+      "step": 17419
+    },
+    {
+      "epoch": 3.959090909090909,
+      "grad_norm": 0.9817826892715606,
+      "learning_rate": 1.2902741557401504e-07,
+      "loss": 0.0056,
+      "step": 17420
+    },
+    {
+      "epoch": 3.959318181818182,
+      "grad_norm": 1.233900595103934,
+      "learning_rate": 1.2897311239631137e-07,
+      "loss": 0.0068,
+      "step": 17421
+    },
+    {
+      "epoch": 3.9595454545454545,
+      "grad_norm": 2.7832760549207065,
+      "learning_rate": 1.2891881933346125e-07,
+      "loss": 0.0307,
+      "step": 17422
+    },
+    {
+      "epoch": 3.9597727272727274,
+      "grad_norm": 1.1537366594998335,
+      "learning_rate": 1.288645363865719e-07,
+      "loss": 0.0099,
+      "step": 17423
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 1.5545346085024543,
+      "learning_rate": 1.2881026355675023e-07,
+      "loss": 0.0023,
+      "step": 17424
+    },
+    {
+      "epoch": 3.9602272727272725,
+      "grad_norm": 1.9523384298669395,
+      "learning_rate": 1.287560008451029e-07,
+      "loss": 0.0065,
+      "step": 17425
+    },
+    {
+      "epoch": 3.9604545454545454,
+      "grad_norm": 0.7674768955988482,
+      "learning_rate": 1.287017482527366e-07,
+      "loss": 0.0084,
+      "step": 17426
+    },
+    {
+      "epoch": 3.9606818181818184,
+      "grad_norm": 0.579537026885014,
+      "learning_rate": 1.2864750578075736e-07,
+      "loss": 0.0027,
+      "step": 17427
+    },
+    {
+      "epoch": 3.960909090909091,
+      "grad_norm": 1.4543025740426307,
+      "learning_rate": 1.285932734302714e-07,
+      "loss": 0.0072,
+      "step": 17428
+    },
+    {
+      "epoch": 3.9611363636363635,
+      "grad_norm": 0.603746766019446,
+      "learning_rate": 1.2853905120238464e-07,
+      "loss": 0.0025,
+      "step": 17429
+    },
+    {
+      "epoch": 3.9613636363636364,
+      "grad_norm": 0.7427725019699903,
+      "learning_rate": 1.2848483909820278e-07,
+      "loss": 0.0054,
+      "step": 17430
+    },
+    {
+      "epoch": 3.961590909090909,
+      "grad_norm": 0.7571419910511248,
+      "learning_rate": 1.284306371188314e-07,
+      "loss": 0.0039,
+      "step": 17431
+    },
+    {
+      "epoch": 3.961818181818182,
+      "grad_norm": 1.479999892103677,
+      "learning_rate": 1.2837644526537546e-07,
+      "loss": 0.0036,
+      "step": 17432
+    },
+    {
+      "epoch": 3.9620454545454544,
+      "grad_norm": 3.590713462068662,
+      "learning_rate": 1.283222635389402e-07,
+      "loss": 0.0186,
+      "step": 17433
+    },
+    {
+      "epoch": 3.9622727272727274,
+      "grad_norm": 0.6405096755254749,
+      "learning_rate": 1.2826809194063052e-07,
+      "loss": 0.0041,
+      "step": 17434
+    },
+    {
+      "epoch": 3.9625,
+      "grad_norm": 1.1728193925117323,
+      "learning_rate": 1.2821393047155107e-07,
+      "loss": 0.0035,
+      "step": 17435
+    },
+    {
+      "epoch": 3.962727272727273,
+      "grad_norm": 3.2550184219578893,
+      "learning_rate": 1.2815977913280624e-07,
+      "loss": 0.0145,
+      "step": 17436
+    },
+    {
+      "epoch": 3.9629545454545454,
+      "grad_norm": 0.8902744515805688,
+      "learning_rate": 1.281056379255003e-07,
+      "loss": 0.0057,
+      "step": 17437
+    },
+    {
+      "epoch": 3.963181818181818,
+      "grad_norm": 0.9353270453227355,
+      "learning_rate": 1.2805150685073723e-07,
+      "loss": 0.0054,
+      "step": 17438
+    },
+    {
+      "epoch": 3.963409090909091,
+      "grad_norm": 1.9168216386869625,
+      "learning_rate": 1.279973859096211e-07,
+      "loss": 0.0151,
+      "step": 17439
+    },
+    {
+      "epoch": 3.963636363636364,
+      "grad_norm": 1.7046352349894949,
+      "learning_rate": 1.279432751032552e-07,
+      "loss": 0.0105,
+      "step": 17440
+    },
+    {
+      "epoch": 3.9638636363636364,
+      "grad_norm": 1.4656510866034613,
+      "learning_rate": 1.278891744327431e-07,
+      "loss": 0.0154,
+      "step": 17441
+    },
+    {
+      "epoch": 3.964090909090909,
+      "grad_norm": 2.0181906885865306,
+      "learning_rate": 1.2783508389918798e-07,
+      "loss": 0.0119,
+      "step": 17442
+    },
+    {
+      "epoch": 3.964318181818182,
+      "grad_norm": 1.0705363007975721,
+      "learning_rate": 1.2778100350369284e-07,
+      "loss": 0.0104,
+      "step": 17443
+    },
+    {
+      "epoch": 3.964545454545455,
+      "grad_norm": 2.317741992800124,
+      "learning_rate": 1.2772693324736068e-07,
+      "loss": 0.0078,
+      "step": 17444
+    },
+    {
+      "epoch": 3.9647727272727273,
+      "grad_norm": 1.1829872793495488,
+      "learning_rate": 1.2767287313129365e-07,
+      "loss": 0.0063,
+      "step": 17445
+    },
+    {
+      "epoch": 3.965,
+      "grad_norm": 1.1491835946914761,
+      "learning_rate": 1.2761882315659447e-07,
+      "loss": 0.0052,
+      "step": 17446
+    },
+    {
+      "epoch": 3.965227272727273,
+      "grad_norm": 4.655482901498248,
+      "learning_rate": 1.2756478332436512e-07,
+      "loss": 0.0135,
+      "step": 17447
+    },
+    {
+      "epoch": 3.9654545454545453,
+      "grad_norm": 3.1982221981428305,
+      "learning_rate": 1.2751075363570788e-07,
+      "loss": 0.0188,
+      "step": 17448
+    },
+    {
+      "epoch": 3.9656818181818183,
+      "grad_norm": 0.972063498558888,
+      "learning_rate": 1.274567340917241e-07,
+      "loss": 0.0047,
+      "step": 17449
+    },
+    {
+      "epoch": 3.965909090909091,
+      "grad_norm": 2.7326230072130153,
+      "learning_rate": 1.2740272469351556e-07,
+      "loss": 0.0035,
+      "step": 17450
+    },
+    {
+      "epoch": 3.966136363636364,
+      "grad_norm": 1.1279220925301776,
+      "learning_rate": 1.2734872544218354e-07,
+      "loss": 0.0056,
+      "step": 17451
+    },
+    {
+      "epoch": 3.9663636363636363,
+      "grad_norm": 1.5275807796126086,
+      "learning_rate": 1.2729473633882916e-07,
+      "loss": 0.0072,
+      "step": 17452
+    },
+    {
+      "epoch": 3.9665909090909093,
+      "grad_norm": 1.814607279737188,
+      "learning_rate": 1.2724075738455347e-07,
+      "loss": 0.0055,
+      "step": 17453
+    },
+    {
+      "epoch": 3.966818181818182,
+      "grad_norm": 0.8369218671927057,
+      "learning_rate": 1.2718678858045716e-07,
+      "loss": 0.0035,
+      "step": 17454
+    },
+    {
+      "epoch": 3.9670454545454543,
+      "grad_norm": 0.9953783933337945,
+      "learning_rate": 1.2713282992764065e-07,
+      "loss": 0.0049,
+      "step": 17455
+    },
+    {
+      "epoch": 3.9672727272727273,
+      "grad_norm": 1.7283094434045798,
+      "learning_rate": 1.2707888142720434e-07,
+      "loss": 0.0149,
+      "step": 17456
+    },
+    {
+      "epoch": 3.9675000000000002,
+      "grad_norm": 1.8602337590126639,
+      "learning_rate": 1.2702494308024845e-07,
+      "loss": 0.0188,
+      "step": 17457
+    },
+    {
+      "epoch": 3.9677272727272728,
+      "grad_norm": 1.5281724101329746,
+      "learning_rate": 1.2697101488787257e-07,
+      "loss": 0.0097,
+      "step": 17458
+    },
+    {
+      "epoch": 3.9679545454545453,
+      "grad_norm": 5.740113605902906,
+      "learning_rate": 1.269170968511766e-07,
+      "loss": 0.01,
+      "step": 17459
+    },
+    {
+      "epoch": 3.9681818181818183,
+      "grad_norm": 1.2664134007581298,
+      "learning_rate": 1.2686318897125995e-07,
+      "loss": 0.0056,
+      "step": 17460
+    },
+    {
+      "epoch": 3.9684090909090908,
+      "grad_norm": 0.6996650860513053,
+      "learning_rate": 1.2680929124922206e-07,
+      "loss": 0.004,
+      "step": 17461
+    },
+    {
+      "epoch": 3.9686363636363637,
+      "grad_norm": 1.2177815936633873,
+      "learning_rate": 1.2675540368616173e-07,
+      "loss": 0.0093,
+      "step": 17462
+    },
+    {
+      "epoch": 3.9688636363636363,
+      "grad_norm": 0.8229501226718872,
+      "learning_rate": 1.2670152628317798e-07,
+      "loss": 0.0033,
+      "step": 17463
+    },
+    {
+      "epoch": 3.9690909090909092,
+      "grad_norm": 1.0002753206393231,
+      "learning_rate": 1.2664765904136942e-07,
+      "loss": 0.0029,
+      "step": 17464
+    },
+    {
+      "epoch": 3.9693181818181817,
+      "grad_norm": 1.8966880072407017,
+      "learning_rate": 1.265938019618345e-07,
+      "loss": 0.0152,
+      "step": 17465
+    },
+    {
+      "epoch": 3.9695454545454547,
+      "grad_norm": 1.2861121110471683,
+      "learning_rate": 1.2653995504567148e-07,
+      "loss": 0.009,
+      "step": 17466
+    },
+    {
+      "epoch": 3.9697727272727272,
+      "grad_norm": 1.257380352783398,
+      "learning_rate": 1.264861182939784e-07,
+      "loss": 0.0041,
+      "step": 17467
+    },
+    {
+      "epoch": 3.9699999999999998,
+      "grad_norm": 0.821711971699289,
+      "learning_rate": 1.2643229170785302e-07,
+      "loss": 0.004,
+      "step": 17468
+    },
+    {
+      "epoch": 3.9702272727272727,
+      "grad_norm": 0.40078646910017063,
+      "learning_rate": 1.2637847528839306e-07,
+      "loss": 0.0018,
+      "step": 17469
+    },
+    {
+      "epoch": 3.9704545454545457,
+      "grad_norm": 2.812287566472833,
+      "learning_rate": 1.2632466903669607e-07,
+      "loss": 0.0252,
+      "step": 17470
+    },
+    {
+      "epoch": 3.970681818181818,
+      "grad_norm": 0.4230697209308605,
+      "learning_rate": 1.2627087295385886e-07,
+      "loss": 0.0013,
+      "step": 17471
+    },
+    {
+      "epoch": 3.9709090909090907,
+      "grad_norm": 2.897643337402491,
+      "learning_rate": 1.262170870409786e-07,
+      "loss": 0.0399,
+      "step": 17472
+    },
+    {
+      "epoch": 3.9711363636363637,
+      "grad_norm": 1.294378514868576,
+      "learning_rate": 1.261633112991522e-07,
+      "loss": 0.007,
+      "step": 17473
+    },
+    {
+      "epoch": 3.971363636363636,
+      "grad_norm": 0.9900307629926022,
+      "learning_rate": 1.2610954572947608e-07,
+      "loss": 0.005,
+      "step": 17474
+    },
+    {
+      "epoch": 3.971590909090909,
+      "grad_norm": 1.2490071569035754,
+      "learning_rate": 1.2605579033304685e-07,
+      "loss": 0.0036,
+      "step": 17475
+    },
+    {
+      "epoch": 3.9718181818181817,
+      "grad_norm": 1.0022262766919352,
+      "learning_rate": 1.2600204511096039e-07,
+      "loss": 0.0059,
+      "step": 17476
+    },
+    {
+      "epoch": 3.9720454545454547,
+      "grad_norm": 1.1466124031895333,
+      "learning_rate": 1.2594831006431275e-07,
+      "loss": 0.0048,
+      "step": 17477
+    },
+    {
+      "epoch": 3.972272727272727,
+      "grad_norm": 1.9357934482406611,
+      "learning_rate": 1.258945851941997e-07,
+      "loss": 0.0117,
+      "step": 17478
+    },
+    {
+      "epoch": 3.9725,
+      "grad_norm": 1.9773133553165803,
+      "learning_rate": 1.2584087050171702e-07,
+      "loss": 0.0118,
+      "step": 17479
+    },
+    {
+      "epoch": 3.9727272727272727,
+      "grad_norm": 4.181582675474295,
+      "learning_rate": 1.2578716598795962e-07,
+      "loss": 0.0531,
+      "step": 17480
+    },
+    {
+      "epoch": 3.972954545454545,
+      "grad_norm": 0.797465893029511,
+      "learning_rate": 1.2573347165402282e-07,
+      "loss": 0.005,
+      "step": 17481
+    },
+    {
+      "epoch": 3.973181818181818,
+      "grad_norm": 1.5338426274773418,
+      "learning_rate": 1.2567978750100137e-07,
+      "loss": 0.0138,
+      "step": 17482
+    },
+    {
+      "epoch": 3.973409090909091,
+      "grad_norm": 1.4146208218993672,
+      "learning_rate": 1.256261135299905e-07,
+      "loss": 0.0094,
+      "step": 17483
+    },
+    {
+      "epoch": 3.9736363636363636,
+      "grad_norm": 1.6331935007084244,
+      "learning_rate": 1.2557244974208426e-07,
+      "loss": 0.0102,
+      "step": 17484
+    },
+    {
+      "epoch": 3.973863636363636,
+      "grad_norm": 1.3821369334607831,
+      "learning_rate": 1.25518796138377e-07,
+      "loss": 0.0064,
+      "step": 17485
+    },
+    {
+      "epoch": 3.974090909090909,
+      "grad_norm": 1.7815991813188485,
+      "learning_rate": 1.2546515271996296e-07,
+      "loss": 0.0062,
+      "step": 17486
+    },
+    {
+      "epoch": 3.974318181818182,
+      "grad_norm": 2.1583796660709984,
+      "learning_rate": 1.2541151948793586e-07,
+      "loss": 0.0195,
+      "step": 17487
+    },
+    {
+      "epoch": 3.9745454545454546,
+      "grad_norm": 0.9536660467001098,
+      "learning_rate": 1.2535789644338972e-07,
+      "loss": 0.0033,
+      "step": 17488
+    },
+    {
+      "epoch": 3.974772727272727,
+      "grad_norm": 2.7706584134931633,
+      "learning_rate": 1.253042835874175e-07,
+      "loss": 0.015,
+      "step": 17489
+    },
+    {
+      "epoch": 3.975,
+      "grad_norm": 1.3655047514840717,
+      "learning_rate": 1.2525068092111272e-07,
+      "loss": 0.0141,
+      "step": 17490
+    },
+    {
+      "epoch": 3.9752272727272726,
+      "grad_norm": 4.159911124067496,
+      "learning_rate": 1.251970884455684e-07,
+      "loss": 0.0168,
+      "step": 17491
+    },
+    {
+      "epoch": 3.9754545454545456,
+      "grad_norm": 1.9431683552248316,
+      "learning_rate": 1.251435061618776e-07,
+      "loss": 0.0167,
+      "step": 17492
+    },
+    {
+      "epoch": 3.975681818181818,
+      "grad_norm": 1.6226512874722177,
+      "learning_rate": 1.2508993407113257e-07,
+      "loss": 0.014,
+      "step": 17493
+    },
+    {
+      "epoch": 3.975909090909091,
+      "grad_norm": 1.139741173962305,
+      "learning_rate": 1.2503637217442598e-07,
+      "loss": 0.0041,
+      "step": 17494
+    },
+    {
+      "epoch": 3.9761363636363636,
+      "grad_norm": 1.8483704506331438,
+      "learning_rate": 1.2498282047284996e-07,
+      "loss": 0.0071,
+      "step": 17495
+    },
+    {
+      "epoch": 3.9763636363636365,
+      "grad_norm": 3.527935423780023,
+      "learning_rate": 1.2492927896749657e-07,
+      "loss": 0.0253,
+      "step": 17496
+    },
+    {
+      "epoch": 3.976590909090909,
+      "grad_norm": 1.0634680654682513,
+      "learning_rate": 1.2487574765945755e-07,
+      "loss": 0.0032,
+      "step": 17497
+    },
+    {
+      "epoch": 3.9768181818181816,
+      "grad_norm": 7.500355681567021,
+      "learning_rate": 1.248222265498246e-07,
+      "loss": 0.0229,
+      "step": 17498
+    },
+    {
+      "epoch": 3.9770454545454546,
+      "grad_norm": 1.0834620258557386,
+      "learning_rate": 1.2476871563968902e-07,
+      "loss": 0.0056,
+      "step": 17499
+    },
+    {
+      "epoch": 3.9772727272727275,
+      "grad_norm": 0.6621622313363543,
+      "learning_rate": 1.2471521493014207e-07,
+      "loss": 0.0155,
+      "step": 17500
+    },
+    {
+      "epoch": 3.9775,
+      "grad_norm": 0.8336737529402571,
+      "learning_rate": 1.2466172442227482e-07,
+      "loss": 0.0063,
+      "step": 17501
+    },
+    {
+      "epoch": 3.9777272727272726,
+      "grad_norm": 1.3857158213744944,
+      "learning_rate": 1.2460824411717776e-07,
+      "loss": 0.0138,
+      "step": 17502
+    },
+    {
+      "epoch": 3.9779545454545455,
+      "grad_norm": 1.3203759438415263,
+      "learning_rate": 1.2455477401594155e-07,
+      "loss": 0.0093,
+      "step": 17503
+    },
+    {
+      "epoch": 3.978181818181818,
+      "grad_norm": 1.1137860272091666,
+      "learning_rate": 1.245013141196566e-07,
+      "loss": 0.0128,
+      "step": 17504
+    },
+    {
+      "epoch": 3.978409090909091,
+      "grad_norm": 0.5513977913703771,
+      "learning_rate": 1.2444786442941316e-07,
+      "loss": 0.0026,
+      "step": 17505
+    },
+    {
+      "epoch": 3.9786363636363635,
+      "grad_norm": 1.2834981616217513,
+      "learning_rate": 1.243944249463009e-07,
+      "loss": 0.0048,
+      "step": 17506
+    },
+    {
+      "epoch": 3.9788636363636365,
+      "grad_norm": 3.2760161432203034,
+      "learning_rate": 1.2434099567140964e-07,
+      "loss": 0.0424,
+      "step": 17507
+    },
+    {
+      "epoch": 3.979090909090909,
+      "grad_norm": 1.472044316766287,
+      "learning_rate": 1.2428757660582894e-07,
+      "loss": 0.016,
+      "step": 17508
+    },
+    {
+      "epoch": 3.979318181818182,
+      "grad_norm": 1.7431453846034897,
+      "learning_rate": 1.2423416775064806e-07,
+      "loss": 0.0091,
+      "step": 17509
+    },
+    {
+      "epoch": 3.9795454545454545,
+      "grad_norm": 4.642415865214611,
+      "learning_rate": 1.241807691069563e-07,
+      "loss": 0.0161,
+      "step": 17510
+    },
+    {
+      "epoch": 3.979772727272727,
+      "grad_norm": 0.5673125223036902,
+      "learning_rate": 1.241273806758423e-07,
+      "loss": 0.0028,
+      "step": 17511
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 2.9115821086848186,
+      "learning_rate": 1.240740024583946e-07,
+      "loss": 0.0193,
+      "step": 17512
+    },
+    {
+      "epoch": 3.980227272727273,
+      "grad_norm": 2.9839339097071385,
+      "learning_rate": 1.2402063445570205e-07,
+      "loss": 0.012,
+      "step": 17513
+    },
+    {
+      "epoch": 3.9804545454545455,
+      "grad_norm": 1.7492069500011178,
+      "learning_rate": 1.2396727666885293e-07,
+      "loss": 0.0149,
+      "step": 17514
+    },
+    {
+      "epoch": 3.980681818181818,
+      "grad_norm": 2.207033485520136,
+      "learning_rate": 1.2391392909893502e-07,
+      "loss": 0.0113,
+      "step": 17515
+    },
+    {
+      "epoch": 3.980909090909091,
+      "grad_norm": 2.0342301493340034,
+      "learning_rate": 1.238605917470363e-07,
+      "loss": 0.0096,
+      "step": 17516
+    },
+    {
+      "epoch": 3.9811363636363635,
+      "grad_norm": 0.9162804276614636,
+      "learning_rate": 1.2380726461424436e-07,
+      "loss": 0.0045,
+      "step": 17517
+    },
+    {
+      "epoch": 3.9813636363636364,
+      "grad_norm": 1.6631821530449287,
+      "learning_rate": 1.2375394770164687e-07,
+      "loss": 0.0136,
+      "step": 17518
+    },
+    {
+      "epoch": 3.981590909090909,
+      "grad_norm": 0.5721037247442322,
+      "learning_rate": 1.2370064101033065e-07,
+      "loss": 0.0028,
+      "step": 17519
+    },
+    {
+      "epoch": 3.981818181818182,
+      "grad_norm": 2.7395189247470535,
+      "learning_rate": 1.23647344541383e-07,
+      "loss": 0.0254,
+      "step": 17520
+    },
+    {
+      "epoch": 3.9820454545454544,
+      "grad_norm": 0.9176834104729406,
+      "learning_rate": 1.2359405829589054e-07,
+      "loss": 0.0025,
+      "step": 17521
+    },
+    {
+      "epoch": 3.9822727272727274,
+      "grad_norm": 1.1716137189461715,
+      "learning_rate": 1.2354078227494008e-07,
+      "loss": 0.0065,
+      "step": 17522
+    },
+    {
+      "epoch": 3.9825,
+      "grad_norm": 0.7382805305211269,
+      "learning_rate": 1.2348751647961802e-07,
+      "loss": 0.0038,
+      "step": 17523
+    },
+    {
+      "epoch": 3.982727272727273,
+      "grad_norm": 2.805814393297994,
+      "learning_rate": 1.2343426091101028e-07,
+      "loss": 0.0129,
+      "step": 17524
+    },
+    {
+      "epoch": 3.9829545454545454,
+      "grad_norm": 1.32962024898894,
+      "learning_rate": 1.2338101557020303e-07,
+      "loss": 0.0107,
+      "step": 17525
+    },
+    {
+      "epoch": 3.9831818181818184,
+      "grad_norm": 1.510437222843201,
+      "learning_rate": 1.23327780458282e-07,
+      "loss": 0.0127,
+      "step": 17526
+    },
+    {
+      "epoch": 3.983409090909091,
+      "grad_norm": 1.5324783985994865,
+      "learning_rate": 1.2327455557633268e-07,
+      "loss": 0.0075,
+      "step": 17527
+    },
+    {
+      "epoch": 3.9836363636363634,
+      "grad_norm": 2.3888248784980655,
+      "learning_rate": 1.232213409254405e-07,
+      "loss": 0.0134,
+      "step": 17528
+    },
+    {
+      "epoch": 3.9838636363636364,
+      "grad_norm": 1.9982990685676854,
+      "learning_rate": 1.2316813650669058e-07,
+      "loss": 0.0184,
+      "step": 17529
+    },
+    {
+      "epoch": 3.9840909090909093,
+      "grad_norm": 0.72005974851753,
+      "learning_rate": 1.2311494232116784e-07,
+      "loss": 0.0053,
+      "step": 17530
+    },
+    {
+      "epoch": 3.984318181818182,
+      "grad_norm": 4.243457671148689,
+      "learning_rate": 1.2306175836995706e-07,
+      "loss": 0.0272,
+      "step": 17531
+    },
+    {
+      "epoch": 3.9845454545454544,
+      "grad_norm": 3.880158205527905,
+      "learning_rate": 1.2300858465414283e-07,
+      "loss": 0.0311,
+      "step": 17532
+    },
+    {
+      "epoch": 3.9847727272727274,
+      "grad_norm": 0.9028192902268222,
+      "learning_rate": 1.2295542117480915e-07,
+      "loss": 0.006,
+      "step": 17533
+    },
+    {
+      "epoch": 3.985,
+      "grad_norm": 4.5110184558927,
+      "learning_rate": 1.2290226793304037e-07,
+      "loss": 0.025,
+      "step": 17534
+    },
+    {
+      "epoch": 3.985227272727273,
+      "grad_norm": 2.292212799050257,
+      "learning_rate": 1.2284912492992025e-07,
+      "loss": 0.0109,
+      "step": 17535
+    },
+    {
+      "epoch": 3.9854545454545454,
+      "grad_norm": 1.7011743792754488,
+      "learning_rate": 1.2279599216653267e-07,
+      "loss": 0.0116,
+      "step": 17536
+    },
+    {
+      "epoch": 3.9856818181818183,
+      "grad_norm": 4.079851188294426,
+      "learning_rate": 1.227428696439608e-07,
+      "loss": 0.0299,
+      "step": 17537
+    },
+    {
+      "epoch": 3.985909090909091,
+      "grad_norm": 21.983717897497307,
+      "learning_rate": 1.2268975736328813e-07,
+      "loss": 0.0287,
+      "step": 17538
+    },
+    {
+      "epoch": 3.986136363636364,
+      "grad_norm": 3.611597686123244,
+      "learning_rate": 1.2263665532559754e-07,
+      "loss": 0.0251,
+      "step": 17539
+    },
+    {
+      "epoch": 3.9863636363636363,
+      "grad_norm": 1.1959828742806917,
+      "learning_rate": 1.2258356353197214e-07,
+      "loss": 0.0066,
+      "step": 17540
+    },
+    {
+      "epoch": 3.986590909090909,
+      "grad_norm": 1.928956695834107,
+      "learning_rate": 1.225304819834941e-07,
+      "loss": 0.0116,
+      "step": 17541
+    },
+    {
+      "epoch": 3.986818181818182,
+      "grad_norm": 1.4591506564775159,
+      "learning_rate": 1.224774106812463e-07,
+      "loss": 0.0071,
+      "step": 17542
+    },
+    {
+      "epoch": 3.987045454545455,
+      "grad_norm": 0.5920629251118507,
+      "learning_rate": 1.2242434962631075e-07,
+      "loss": 0.0036,
+      "step": 17543
+    },
+    {
+      "epoch": 3.9872727272727273,
+      "grad_norm": 1.7480238398629484,
+      "learning_rate": 1.2237129881976958e-07,
+      "loss": 0.0226,
+      "step": 17544
+    },
+    {
+      "epoch": 3.9875,
+      "grad_norm": 1.0563308869410268,
+      "learning_rate": 1.2231825826270465e-07,
+      "loss": 0.0055,
+      "step": 17545
+    },
+    {
+      "epoch": 3.987727272727273,
+      "grad_norm": 1.3955310505143925,
+      "learning_rate": 1.222652279561972e-07,
+      "loss": 0.0067,
+      "step": 17546
+    },
+    {
+      "epoch": 3.9879545454545453,
+      "grad_norm": 0.8366342083766453,
+      "learning_rate": 1.222122079013289e-07,
+      "loss": 0.0038,
+      "step": 17547
+    },
+    {
+      "epoch": 3.9881818181818183,
+      "grad_norm": 0.8269443832372365,
+      "learning_rate": 1.2215919809918076e-07,
+      "loss": 0.004,
+      "step": 17548
+    },
+    {
+      "epoch": 3.988409090909091,
+      "grad_norm": 1.5108523655215016,
+      "learning_rate": 1.2210619855083407e-07,
+      "loss": 0.0112,
+      "step": 17549
+    },
+    {
+      "epoch": 3.9886363636363638,
+      "grad_norm": 3.9928446425881745,
+      "learning_rate": 1.220532092573691e-07,
+      "loss": 0.0346,
+      "step": 17550
+    },
+    {
+      "epoch": 3.9888636363636363,
+      "grad_norm": 2.067906782364483,
+      "learning_rate": 1.2200023021986667e-07,
+      "loss": 0.0131,
+      "step": 17551
+    },
+    {
+      "epoch": 3.9890909090909092,
+      "grad_norm": 1.3514608782297097,
+      "learning_rate": 1.2194726143940704e-07,
+      "loss": 0.0067,
+      "step": 17552
+    },
+    {
+      "epoch": 3.9893181818181818,
+      "grad_norm": 1.3852752513454998,
+      "learning_rate": 1.2189430291707043e-07,
+      "loss": 0.0134,
+      "step": 17553
+    },
+    {
+      "epoch": 3.9895454545454543,
+      "grad_norm": 1.0453866181790918,
+      "learning_rate": 1.218413546539368e-07,
+      "loss": 0.0063,
+      "step": 17554
+    },
+    {
+      "epoch": 3.9897727272727272,
+      "grad_norm": 4.5873603506346985,
+      "learning_rate": 1.2178841665108562e-07,
+      "loss": 0.0111,
+      "step": 17555
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.922514608214128,
+      "learning_rate": 1.2173548890959654e-07,
+      "loss": 0.0037,
+      "step": 17556
+    },
+    {
+      "epoch": 3.9902272727272727,
+      "grad_norm": 1.8141656967398945,
+      "learning_rate": 1.2168257143054883e-07,
+      "loss": 0.014,
+      "step": 17557
+    },
+    {
+      "epoch": 3.9904545454545453,
+      "grad_norm": 1.939858369540016,
+      "learning_rate": 1.2162966421502157e-07,
+      "loss": 0.0116,
+      "step": 17558
+    },
+    {
+      "epoch": 3.990681818181818,
+      "grad_norm": 1.8660288283477378,
+      "learning_rate": 1.2157676726409363e-07,
+      "loss": 0.0095,
+      "step": 17559
+    },
+    {
+      "epoch": 3.990909090909091,
+      "grad_norm": 0.8681445506251966,
+      "learning_rate": 1.2152388057884372e-07,
+      "loss": 0.0055,
+      "step": 17560
+    },
+    {
+      "epoch": 3.9911363636363637,
+      "grad_norm": 0.5815947725689666,
+      "learning_rate": 1.214710041603502e-07,
+      "loss": 0.002,
+      "step": 17561
+    },
+    {
+      "epoch": 3.9913636363636362,
+      "grad_norm": 3.5076962372121647,
+      "learning_rate": 1.2141813800969155e-07,
+      "loss": 0.0133,
+      "step": 17562
+    },
+    {
+      "epoch": 3.991590909090909,
+      "grad_norm": 3.300703171473917,
+      "learning_rate": 1.213652821279454e-07,
+      "loss": 0.0186,
+      "step": 17563
+    },
+    {
+      "epoch": 3.9918181818181817,
+      "grad_norm": 2.4003786450971853,
+      "learning_rate": 1.213124365161898e-07,
+      "loss": 0.0116,
+      "step": 17564
+    },
+    {
+      "epoch": 3.9920454545454547,
+      "grad_norm": 1.3172609590525912,
+      "learning_rate": 1.212596011755024e-07,
+      "loss": 0.0113,
+      "step": 17565
+    },
+    {
+      "epoch": 3.992272727272727,
+      "grad_norm": 2.3769870618431774,
+      "learning_rate": 1.2120677610696053e-07,
+      "loss": 0.0217,
+      "step": 17566
+    },
+    {
+      "epoch": 3.9925,
+      "grad_norm": 1.1190957604942364,
+      "learning_rate": 1.2115396131164157e-07,
+      "loss": 0.0112,
+      "step": 17567
+    },
+    {
+      "epoch": 3.9927272727272727,
+      "grad_norm": 1.2475359959951118,
+      "learning_rate": 1.2110115679062214e-07,
+      "loss": 0.0061,
+      "step": 17568
+    },
+    {
+      "epoch": 3.9929545454545456,
+      "grad_norm": 1.4579447922680244,
+      "learning_rate": 1.2104836254497925e-07,
+      "loss": 0.0103,
+      "step": 17569
+    },
+    {
+      "epoch": 3.993181818181818,
+      "grad_norm": 2.782190647857759,
+      "learning_rate": 1.209955785757895e-07,
+      "loss": 0.0149,
+      "step": 17570
+    },
+    {
+      "epoch": 3.9934090909090907,
+      "grad_norm": 1.3538768801914742,
+      "learning_rate": 1.2094280488412913e-07,
+      "loss": 0.0087,
+      "step": 17571
+    },
+    {
+      "epoch": 3.9936363636363637,
+      "grad_norm": 1.7846192951115842,
+      "learning_rate": 1.208900414710743e-07,
+      "loss": 0.0243,
+      "step": 17572
+    },
+    {
+      "epoch": 3.9938636363636366,
+      "grad_norm": 2.206143603142302,
+      "learning_rate": 1.2083728833770105e-07,
+      "loss": 0.0128,
+      "step": 17573
+    },
+    {
+      "epoch": 3.994090909090909,
+      "grad_norm": 1.4223445189128827,
+      "learning_rate": 1.2078454548508507e-07,
+      "loss": 0.0094,
+      "step": 17574
+    },
+    {
+      "epoch": 3.9943181818181817,
+      "grad_norm": 0.8494543704331337,
+      "learning_rate": 1.2073181291430182e-07,
+      "loss": 0.0042,
+      "step": 17575
+    },
+    {
+      "epoch": 3.9945454545454546,
+      "grad_norm": 1.9966435880789069,
+      "learning_rate": 1.2067909062642684e-07,
+      "loss": 0.0109,
+      "step": 17576
+    },
+    {
+      "epoch": 3.994772727272727,
+      "grad_norm": 2.197335854630203,
+      "learning_rate": 1.2062637862253486e-07,
+      "loss": 0.0062,
+      "step": 17577
+    },
+    {
+      "epoch": 3.995,
+      "grad_norm": 1.2585131742378373,
+      "learning_rate": 1.2057367690370095e-07,
+      "loss": 0.0053,
+      "step": 17578
+    },
+    {
+      "epoch": 3.9952272727272726,
+      "grad_norm": 1.042201456391065,
+      "learning_rate": 1.2052098547099981e-07,
+      "loss": 0.0032,
+      "step": 17579
+    },
+    {
+      "epoch": 3.9954545454545456,
+      "grad_norm": 1.3242771186321785,
+      "learning_rate": 1.20468304325506e-07,
+      "loss": 0.0084,
+      "step": 17580
+    },
+    {
+      "epoch": 3.995681818181818,
+      "grad_norm": 1.7719811760082662,
+      "learning_rate": 1.2041563346829355e-07,
+      "loss": 0.0096,
+      "step": 17581
+    },
+    {
+      "epoch": 3.995909090909091,
+      "grad_norm": 2.173005611336799,
+      "learning_rate": 1.2036297290043666e-07,
+      "loss": 0.0144,
+      "step": 17582
+    },
+    {
+      "epoch": 3.9961363636363636,
+      "grad_norm": 2.316731388467128,
+      "learning_rate": 1.203103226230091e-07,
+      "loss": 0.011,
+      "step": 17583
+    },
+    {
+      "epoch": 3.996363636363636,
+      "grad_norm": 0.7916915627642885,
+      "learning_rate": 1.2025768263708464e-07,
+      "loss": 0.0062,
+      "step": 17584
+    },
+    {
+      "epoch": 3.996590909090909,
+      "grad_norm": 2.5676369739180367,
+      "learning_rate": 1.2020505294373652e-07,
+      "loss": 0.0127,
+      "step": 17585
+    },
+    {
+      "epoch": 3.996818181818182,
+      "grad_norm": 2.5822496447187318,
+      "learning_rate": 1.20152433544038e-07,
+      "loss": 0.0246,
+      "step": 17586
+    },
+    {
+      "epoch": 3.9970454545454546,
+      "grad_norm": 1.790001241648785,
+      "learning_rate": 1.2009982443906214e-07,
+      "loss": 0.0085,
+      "step": 17587
+    },
+    {
+      "epoch": 3.997272727272727,
+      "grad_norm": 2.1551385192202543,
+      "learning_rate": 1.2004722562988166e-07,
+      "loss": 0.0174,
+      "step": 17588
+    },
+    {
+      "epoch": 3.9975,
+      "grad_norm": 1.3230586878090915,
+      "learning_rate": 1.1999463711756928e-07,
+      "loss": 0.0109,
+      "step": 17589
+    },
+    {
+      "epoch": 3.9977272727272726,
+      "grad_norm": 1.7782999538335524,
+      "learning_rate": 1.1994205890319716e-07,
+      "loss": 0.0076,
+      "step": 17590
+    },
+    {
+      "epoch": 3.9979545454545455,
+      "grad_norm": 1.7447017745297042,
+      "learning_rate": 1.1988949098783767e-07,
+      "loss": 0.0125,
+      "step": 17591
+    },
+    {
+      "epoch": 3.998181818181818,
+      "grad_norm": 2.0333927533904324,
+      "learning_rate": 1.198369333725626e-07,
+      "loss": 0.0077,
+      "step": 17592
+    },
+    {
+      "epoch": 3.998409090909091,
+      "grad_norm": 2.372076854316426,
+      "learning_rate": 1.1978438605844404e-07,
+      "loss": 0.016,
+      "step": 17593
+    },
+    {
+      "epoch": 3.9986363636363635,
+      "grad_norm": 0.5476863742548845,
+      "learning_rate": 1.1973184904655302e-07,
+      "loss": 0.0026,
+      "step": 17594
+    },
+    {
+      "epoch": 3.9988636363636365,
+      "grad_norm": 1.3206149666379041,
+      "learning_rate": 1.19679322337961e-07,
+      "loss": 0.0124,
+      "step": 17595
+    },
+    {
+      "epoch": 3.999090909090909,
+      "grad_norm": 0.9523709039909886,
+      "learning_rate": 1.1962680593373931e-07,
+      "loss": 0.0091,
+      "step": 17596
+    },
+    {
+      "epoch": 3.9993181818181816,
+      "grad_norm": 1.5912794227027236,
+      "learning_rate": 1.1957429983495869e-07,
+      "loss": 0.0143,
+      "step": 17597
+    },
+    {
+      "epoch": 3.9995454545454545,
+      "grad_norm": 1.0119987238159431,
+      "learning_rate": 1.1952180404268994e-07,
+      "loss": 0.0074,
+      "step": 17598
+    },
+    {
+      "epoch": 3.9997727272727275,
+      "grad_norm": 0.8051518759856452,
+      "learning_rate": 1.1946931855800343e-07,
+      "loss": 0.003,
+      "step": 17599
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 1.2944847462853444,
+      "learning_rate": 1.1941684338196938e-07,
+      "loss": 0.0093,
+      "step": 17600
+    },
+    {
+      "epoch": 4.0002272727272725,
+      "grad_norm": 1.1678489100983227,
+      "learning_rate": 1.193643785156579e-07,
+      "loss": 0.0079,
+      "step": 17601
+    },
+    {
+      "epoch": 4.000454545454545,
+      "grad_norm": 0.9569381493828533,
+      "learning_rate": 1.1931192396013889e-07,
+      "loss": 0.0031,
+      "step": 17602
+    },
+    {
+      "epoch": 4.0006818181818184,
+      "grad_norm": 0.9458242039565967,
+      "learning_rate": 1.19259479716482e-07,
+      "loss": 0.0065,
+      "step": 17603
+    },
+    {
+      "epoch": 4.000909090909091,
+      "grad_norm": 0.9350714409753974,
+      "learning_rate": 1.1920704578575665e-07,
+      "loss": 0.0047,
+      "step": 17604
+    },
+    {
+      "epoch": 4.0011363636363635,
+      "grad_norm": 1.2494462751962139,
+      "learning_rate": 1.1915462216903199e-07,
+      "loss": 0.0102,
+      "step": 17605
+    },
+    {
+      "epoch": 4.001363636363636,
+      "grad_norm": 1.4469884489066958,
+      "learning_rate": 1.1910220886737732e-07,
+      "loss": 0.0111,
+      "step": 17606
+    },
+    {
+      "epoch": 4.001590909090909,
+      "grad_norm": 1.162672685276224,
+      "learning_rate": 1.1904980588186093e-07,
+      "loss": 0.0059,
+      "step": 17607
+    },
+    {
+      "epoch": 4.001818181818182,
+      "grad_norm": 0.3698439718214548,
+      "learning_rate": 1.1899741321355178e-07,
+      "loss": 0.0023,
+      "step": 17608
+    },
+    {
+      "epoch": 4.0020454545454545,
+      "grad_norm": 0.9347287397357668,
+      "learning_rate": 1.1894503086351809e-07,
+      "loss": 0.0122,
+      "step": 17609
+    },
+    {
+      "epoch": 4.002272727272727,
+      "grad_norm": 0.784527854103546,
+      "learning_rate": 1.1889265883282805e-07,
+      "loss": 0.0044,
+      "step": 17610
+    },
+    {
+      "epoch": 4.0025,
+      "grad_norm": 0.8687366348740754,
+      "learning_rate": 1.1884029712254983e-07,
+      "loss": 0.0061,
+      "step": 17611
+    },
+    {
+      "epoch": 4.002727272727273,
+      "grad_norm": 0.9551458549806066,
+      "learning_rate": 1.187879457337509e-07,
+      "loss": 0.0069,
+      "step": 17612
+    },
+    {
+      "epoch": 4.002954545454545,
+      "grad_norm": 1.527996254114365,
+      "learning_rate": 1.1873560466749881e-07,
+      "loss": 0.0063,
+      "step": 17613
+    },
+    {
+      "epoch": 4.003181818181818,
+      "grad_norm": 1.6749553936194101,
+      "learning_rate": 1.1868327392486098e-07,
+      "loss": 0.0159,
+      "step": 17614
+    },
+    {
+      "epoch": 4.0034090909090905,
+      "grad_norm": 0.34856708685899207,
+      "learning_rate": 1.1863095350690475e-07,
+      "loss": 0.0021,
+      "step": 17615
+    },
+    {
+      "epoch": 4.003636363636364,
+      "grad_norm": 1.3907514478072671,
+      "learning_rate": 1.1857864341469655e-07,
+      "loss": 0.0043,
+      "step": 17616
+    },
+    {
+      "epoch": 4.003863636363636,
+      "grad_norm": 0.9857217367468774,
+      "learning_rate": 1.1852634364930331e-07,
+      "loss": 0.0046,
+      "step": 17617
+    },
+    {
+      "epoch": 4.004090909090909,
+      "grad_norm": 1.5538654444175921,
+      "learning_rate": 1.1847405421179154e-07,
+      "loss": 0.0141,
+      "step": 17618
+    },
+    {
+      "epoch": 4.004318181818181,
+      "grad_norm": 0.6415905416054494,
+      "learning_rate": 1.1842177510322753e-07,
+      "loss": 0.003,
+      "step": 17619
+    },
+    {
+      "epoch": 4.004545454545455,
+      "grad_norm": 1.0108845748420068,
+      "learning_rate": 1.1836950632467723e-07,
+      "loss": 0.0062,
+      "step": 17620
+    },
+    {
+      "epoch": 4.004772727272727,
+      "grad_norm": 0.7490596941451653,
+      "learning_rate": 1.1831724787720656e-07,
+      "loss": 0.0048,
+      "step": 17621
+    },
+    {
+      "epoch": 4.005,
+      "grad_norm": 0.1879030765213133,
+      "learning_rate": 1.1826499976188122e-07,
+      "loss": 0.0007,
+      "step": 17622
+    },
+    {
+      "epoch": 4.005227272727272,
+      "grad_norm": 1.0300414604584125,
+      "learning_rate": 1.1821276197976655e-07,
+      "loss": 0.0038,
+      "step": 17623
+    },
+    {
+      "epoch": 4.005454545454546,
+      "grad_norm": 0.7192985077220779,
+      "learning_rate": 1.1816053453192797e-07,
+      "loss": 0.0038,
+      "step": 17624
+    },
+    {
+      "epoch": 4.005681818181818,
+      "grad_norm": 0.30893545472354184,
+      "learning_rate": 1.1810831741943016e-07,
+      "loss": 0.0013,
+      "step": 17625
+    },
+    {
+      "epoch": 4.005909090909091,
+      "grad_norm": 1.4315866975480513,
+      "learning_rate": 1.1805611064333804e-07,
+      "loss": 0.0116,
+      "step": 17626
+    },
+    {
+      "epoch": 4.006136363636363,
+      "grad_norm": 0.6399180198990636,
+      "learning_rate": 1.180039142047163e-07,
+      "loss": 0.0037,
+      "step": 17627
+    },
+    {
+      "epoch": 4.006363636363637,
+      "grad_norm": 1.2098478579714993,
+      "learning_rate": 1.1795172810462934e-07,
+      "loss": 0.0102,
+      "step": 17628
+    },
+    {
+      "epoch": 4.006590909090909,
+      "grad_norm": 0.6178249485619787,
+      "learning_rate": 1.1789955234414112e-07,
+      "loss": 0.0028,
+      "step": 17629
+    },
+    {
+      "epoch": 4.006818181818182,
+      "grad_norm": 0.6933167141666219,
+      "learning_rate": 1.1784738692431574e-07,
+      "loss": 0.0025,
+      "step": 17630
+    },
+    {
+      "epoch": 4.007045454545454,
+      "grad_norm": 0.41286942792134146,
+      "learning_rate": 1.1779523184621687e-07,
+      "loss": 0.0024,
+      "step": 17631
+    },
+    {
+      "epoch": 4.007272727272727,
+      "grad_norm": 0.8270245386797951,
+      "learning_rate": 1.177430871109081e-07,
+      "loss": 0.0039,
+      "step": 17632
+    },
+    {
+      "epoch": 4.0075,
+      "grad_norm": 0.2592237063565985,
+      "learning_rate": 1.176909527194528e-07,
+      "loss": 0.0013,
+      "step": 17633
+    },
+    {
+      "epoch": 4.007727272727273,
+      "grad_norm": 1.283296834089485,
+      "learning_rate": 1.1763882867291395e-07,
+      "loss": 0.0128,
+      "step": 17634
+    },
+    {
+      "epoch": 4.007954545454545,
+      "grad_norm": 0.27563816385685325,
+      "learning_rate": 1.1758671497235453e-07,
+      "loss": 0.001,
+      "step": 17635
+    },
+    {
+      "epoch": 4.008181818181818,
+      "grad_norm": 1.8605348664167896,
+      "learning_rate": 1.1753461161883722e-07,
+      "loss": 0.011,
+      "step": 17636
+    },
+    {
+      "epoch": 4.008409090909091,
+      "grad_norm": 0.34196075426192896,
+      "learning_rate": 1.1748251861342472e-07,
+      "loss": 0.0017,
+      "step": 17637
+    },
+    {
+      "epoch": 4.008636363636364,
+      "grad_norm": 1.148909592040871,
+      "learning_rate": 1.1743043595717888e-07,
+      "loss": 0.0072,
+      "step": 17638
+    },
+    {
+      "epoch": 4.008863636363636,
+      "grad_norm": 1.4052010415102316,
+      "learning_rate": 1.1737836365116197e-07,
+      "loss": 0.0173,
+      "step": 17639
+    },
+    {
+      "epoch": 4.009090909090909,
+      "grad_norm": 0.546308535662005,
+      "learning_rate": 1.1732630169643579e-07,
+      "loss": 0.002,
+      "step": 17640
+    },
+    {
+      "epoch": 4.009318181818182,
+      "grad_norm": 0.9492348776021341,
+      "learning_rate": 1.1727425009406219e-07,
+      "loss": 0.0026,
+      "step": 17641
+    },
+    {
+      "epoch": 4.009545454545455,
+      "grad_norm": 0.4606982244499528,
+      "learning_rate": 1.1722220884510222e-07,
+      "loss": 0.0018,
+      "step": 17642
+    },
+    {
+      "epoch": 4.009772727272727,
+      "grad_norm": 0.429514597739091,
+      "learning_rate": 1.1717017795061728e-07,
+      "loss": 0.0019,
+      "step": 17643
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 0.44910727553781365,
+      "learning_rate": 1.1711815741166838e-07,
+      "loss": 0.0028,
+      "step": 17644
+    },
+    {
+      "epoch": 4.010227272727272,
+      "grad_norm": 0.3110478478266627,
+      "learning_rate": 1.1706614722931629e-07,
+      "loss": 0.002,
+      "step": 17645
+    },
+    {
+      "epoch": 4.010454545454546,
+      "grad_norm": 1.4082069224035598,
+      "learning_rate": 1.1701414740462178e-07,
+      "loss": 0.0097,
+      "step": 17646
+    },
+    {
+      "epoch": 4.010681818181818,
+      "grad_norm": 0.9918322317975943,
+      "learning_rate": 1.1696215793864485e-07,
+      "loss": 0.0052,
+      "step": 17647
+    },
+    {
+      "epoch": 4.010909090909091,
+      "grad_norm": 0.31601855671975354,
+      "learning_rate": 1.1691017883244588e-07,
+      "loss": 0.0018,
+      "step": 17648
+    },
+    {
+      "epoch": 4.011136363636363,
+      "grad_norm": 0.5942542384202198,
+      "learning_rate": 1.1685821008708471e-07,
+      "loss": 0.0028,
+      "step": 17649
+    },
+    {
+      "epoch": 4.011363636363637,
+      "grad_norm": 0.6473157474991138,
+      "learning_rate": 1.1680625170362121e-07,
+      "loss": 0.0037,
+      "step": 17650
+    },
+    {
+      "epoch": 4.011590909090909,
+      "grad_norm": 1.082991714104131,
+      "learning_rate": 1.167543036831148e-07,
+      "loss": 0.0054,
+      "step": 17651
+    },
+    {
+      "epoch": 4.011818181818182,
+      "grad_norm": 0.5442031881276391,
+      "learning_rate": 1.1670236602662484e-07,
+      "loss": 0.0018,
+      "step": 17652
+    },
+    {
+      "epoch": 4.012045454545454,
+      "grad_norm": 0.6334097862350334,
+      "learning_rate": 1.1665043873521035e-07,
+      "loss": 0.0043,
+      "step": 17653
+    },
+    {
+      "epoch": 4.012272727272728,
+      "grad_norm": 2.4386719828753365,
+      "learning_rate": 1.1659852180993032e-07,
+      "loss": 0.009,
+      "step": 17654
+    },
+    {
+      "epoch": 4.0125,
+      "grad_norm": 1.2358118253917352,
+      "learning_rate": 1.1654661525184353e-07,
+      "loss": 0.0124,
+      "step": 17655
+    },
+    {
+      "epoch": 4.012727272727273,
+      "grad_norm": 1.1874776061322818,
+      "learning_rate": 1.1649471906200818e-07,
+      "loss": 0.0037,
+      "step": 17656
+    },
+    {
+      "epoch": 4.012954545454545,
+      "grad_norm": 0.8458859135113784,
+      "learning_rate": 1.1644283324148263e-07,
+      "loss": 0.0071,
+      "step": 17657
+    },
+    {
+      "epoch": 4.013181818181818,
+      "grad_norm": 1.3851989713563286,
+      "learning_rate": 1.163909577913249e-07,
+      "loss": 0.0065,
+      "step": 17658
+    },
+    {
+      "epoch": 4.013409090909091,
+      "grad_norm": 0.8788163749360528,
+      "learning_rate": 1.1633909271259305e-07,
+      "loss": 0.0039,
+      "step": 17659
+    },
+    {
+      "epoch": 4.013636363636364,
+      "grad_norm": 0.39897280631261817,
+      "learning_rate": 1.1628723800634435e-07,
+      "loss": 0.0022,
+      "step": 17660
+    },
+    {
+      "epoch": 4.013863636363636,
+      "grad_norm": 1.556424799503378,
+      "learning_rate": 1.1623539367363636e-07,
+      "loss": 0.009,
+      "step": 17661
+    },
+    {
+      "epoch": 4.014090909090909,
+      "grad_norm": 0.7490785430740357,
+      "learning_rate": 1.1618355971552628e-07,
+      "loss": 0.0059,
+      "step": 17662
+    },
+    {
+      "epoch": 4.014318181818182,
+      "grad_norm": 1.4731193339449407,
+      "learning_rate": 1.1613173613307105e-07,
+      "loss": 0.011,
+      "step": 17663
+    },
+    {
+      "epoch": 4.014545454545455,
+      "grad_norm": 0.5237278944171032,
+      "learning_rate": 1.1607992292732753e-07,
+      "loss": 0.002,
+      "step": 17664
+    },
+    {
+      "epoch": 4.014772727272727,
+      "grad_norm": 2.3660193866925114,
+      "learning_rate": 1.1602812009935224e-07,
+      "loss": 0.0185,
+      "step": 17665
+    },
+    {
+      "epoch": 4.015,
+      "grad_norm": 0.6416165021797933,
+      "learning_rate": 1.1597632765020156e-07,
+      "loss": 0.0073,
+      "step": 17666
+    },
+    {
+      "epoch": 4.015227272727273,
+      "grad_norm": 0.9388764159276183,
+      "learning_rate": 1.1592454558093158e-07,
+      "loss": 0.004,
+      "step": 17667
+    },
+    {
+      "epoch": 4.015454545454546,
+      "grad_norm": 0.3447101272613147,
+      "learning_rate": 1.158727738925984e-07,
+      "loss": 0.0056,
+      "step": 17668
+    },
+    {
+      "epoch": 4.015681818181818,
+      "grad_norm": 1.1447364261874824,
+      "learning_rate": 1.1582101258625744e-07,
+      "loss": 0.0038,
+      "step": 17669
+    },
+    {
+      "epoch": 4.015909090909091,
+      "grad_norm": 0.5116277013468374,
+      "learning_rate": 1.157692616629643e-07,
+      "loss": 0.0016,
+      "step": 17670
+    },
+    {
+      "epoch": 4.016136363636364,
+      "grad_norm": 0.8268598610067135,
+      "learning_rate": 1.1571752112377432e-07,
+      "loss": 0.0051,
+      "step": 17671
+    },
+    {
+      "epoch": 4.016363636363637,
+      "grad_norm": 0.6682926467349403,
+      "learning_rate": 1.1566579096974277e-07,
+      "loss": 0.0034,
+      "step": 17672
+    },
+    {
+      "epoch": 4.016590909090909,
+      "grad_norm": 0.6331992179197152,
+      "learning_rate": 1.1561407120192416e-07,
+      "loss": 0.0047,
+      "step": 17673
+    },
+    {
+      "epoch": 4.016818181818182,
+      "grad_norm": 0.24065078613594326,
+      "learning_rate": 1.1556236182137331e-07,
+      "loss": 0.0011,
+      "step": 17674
+    },
+    {
+      "epoch": 4.017045454545454,
+      "grad_norm": 1.100082775839527,
+      "learning_rate": 1.155106628291447e-07,
+      "loss": 0.0082,
+      "step": 17675
+    },
+    {
+      "epoch": 4.0172727272727276,
+      "grad_norm": 1.370352054628736,
+      "learning_rate": 1.1545897422629245e-07,
+      "loss": 0.0039,
+      "step": 17676
+    },
+    {
+      "epoch": 4.0175,
+      "grad_norm": 0.7594366135885279,
+      "learning_rate": 1.1540729601387086e-07,
+      "loss": 0.0022,
+      "step": 17677
+    },
+    {
+      "epoch": 4.017727272727273,
+      "grad_norm": 0.48850862989841765,
+      "learning_rate": 1.1535562819293334e-07,
+      "loss": 0.0024,
+      "step": 17678
+    },
+    {
+      "epoch": 4.017954545454545,
+      "grad_norm": 0.36512024190323744,
+      "learning_rate": 1.1530397076453356e-07,
+      "loss": 0.0009,
+      "step": 17679
+    },
+    {
+      "epoch": 4.0181818181818185,
+      "grad_norm": 0.29760468805226753,
+      "learning_rate": 1.1525232372972525e-07,
+      "loss": 0.0008,
+      "step": 17680
+    },
+    {
+      "epoch": 4.018409090909091,
+      "grad_norm": 2.2625852799582744,
+      "learning_rate": 1.1520068708956146e-07,
+      "loss": 0.0071,
+      "step": 17681
+    },
+    {
+      "epoch": 4.018636363636364,
+      "grad_norm": 3.7363689932742212,
+      "learning_rate": 1.151490608450949e-07,
+      "loss": 0.0163,
+      "step": 17682
+    },
+    {
+      "epoch": 4.018863636363636,
+      "grad_norm": 0.7619007526380769,
+      "learning_rate": 1.1509744499737845e-07,
+      "loss": 0.0024,
+      "step": 17683
+    },
+    {
+      "epoch": 4.0190909090909095,
+      "grad_norm": 1.2870523490867618,
+      "learning_rate": 1.1504583954746469e-07,
+      "loss": 0.006,
+      "step": 17684
+    },
+    {
+      "epoch": 4.019318181818182,
+      "grad_norm": 0.25167620487879516,
+      "learning_rate": 1.1499424449640604e-07,
+      "loss": 0.0014,
+      "step": 17685
+    },
+    {
+      "epoch": 4.0195454545454545,
+      "grad_norm": 0.26138223830686813,
+      "learning_rate": 1.1494265984525438e-07,
+      "loss": 0.0013,
+      "step": 17686
+    },
+    {
+      "epoch": 4.019772727272727,
+      "grad_norm": 1.8388291977685403,
+      "learning_rate": 1.1489108559506177e-07,
+      "loss": 0.012,
+      "step": 17687
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 0.6303759648202752,
+      "learning_rate": 1.148395217468798e-07,
+      "loss": 0.0024,
+      "step": 17688
+    },
+    {
+      "epoch": 4.020227272727273,
+      "grad_norm": 0.6093609179327967,
+      "learning_rate": 1.1478796830176005e-07,
+      "loss": 0.0037,
+      "step": 17689
+    },
+    {
+      "epoch": 4.0204545454545455,
+      "grad_norm": 1.1144295067892105,
+      "learning_rate": 1.1473642526075385e-07,
+      "loss": 0.0055,
+      "step": 17690
+    },
+    {
+      "epoch": 4.020681818181818,
+      "grad_norm": 3.09903044760843,
+      "learning_rate": 1.1468489262491202e-07,
+      "loss": 0.0169,
+      "step": 17691
+    },
+    {
+      "epoch": 4.0209090909090905,
+      "grad_norm": 0.6495992820446029,
+      "learning_rate": 1.1463337039528552e-07,
+      "loss": 0.0032,
+      "step": 17692
+    },
+    {
+      "epoch": 4.021136363636364,
+      "grad_norm": 1.185386117928571,
+      "learning_rate": 1.1458185857292497e-07,
+      "loss": 0.0051,
+      "step": 17693
+    },
+    {
+      "epoch": 4.0213636363636365,
+      "grad_norm": 0.8377430630659504,
+      "learning_rate": 1.145303571588808e-07,
+      "loss": 0.0087,
+      "step": 17694
+    },
+    {
+      "epoch": 4.021590909090909,
+      "grad_norm": 0.5547984631683116,
+      "learning_rate": 1.1447886615420317e-07,
+      "loss": 0.0028,
+      "step": 17695
+    },
+    {
+      "epoch": 4.0218181818181815,
+      "grad_norm": 0.9248604678462767,
+      "learning_rate": 1.1442738555994218e-07,
+      "loss": 0.0034,
+      "step": 17696
+    },
+    {
+      "epoch": 4.022045454545455,
+      "grad_norm": 0.7542616875618249,
+      "learning_rate": 1.1437591537714752e-07,
+      "loss": 0.0033,
+      "step": 17697
+    },
+    {
+      "epoch": 4.022272727272727,
+      "grad_norm": 0.8619206331801695,
+      "learning_rate": 1.143244556068688e-07,
+      "loss": 0.0026,
+      "step": 17698
+    },
+    {
+      "epoch": 4.0225,
+      "grad_norm": 0.38009345568257985,
+      "learning_rate": 1.1427300625015549e-07,
+      "loss": 0.0016,
+      "step": 17699
+    },
+    {
+      "epoch": 4.0227272727272725,
+      "grad_norm": 0.25154373163441635,
+      "learning_rate": 1.1422156730805641e-07,
+      "loss": 0.0013,
+      "step": 17700
+    },
+    {
+      "epoch": 4.022954545454546,
+      "grad_norm": 0.7726716372348269,
+      "learning_rate": 1.1417013878162069e-07,
+      "loss": 0.0037,
+      "step": 17701
+    },
+    {
+      "epoch": 4.023181818181818,
+      "grad_norm": 1.1496087913684423,
+      "learning_rate": 1.1411872067189702e-07,
+      "loss": 0.0115,
+      "step": 17702
+    },
+    {
+      "epoch": 4.023409090909091,
+      "grad_norm": 1.1124797142812757,
+      "learning_rate": 1.1406731297993412e-07,
+      "loss": 0.012,
+      "step": 17703
+    },
+    {
+      "epoch": 4.0236363636363635,
+      "grad_norm": 1.4072018910392983,
+      "learning_rate": 1.1401591570677987e-07,
+      "loss": 0.0054,
+      "step": 17704
+    },
+    {
+      "epoch": 4.023863636363636,
+      "grad_norm": 0.6875978404097394,
+      "learning_rate": 1.1396452885348256e-07,
+      "loss": 0.0041,
+      "step": 17705
+    },
+    {
+      "epoch": 4.024090909090909,
+      "grad_norm": 1.1221116417171055,
+      "learning_rate": 1.1391315242109006e-07,
+      "loss": 0.0023,
+      "step": 17706
+    },
+    {
+      "epoch": 4.024318181818182,
+      "grad_norm": 0.5627996166729486,
+      "learning_rate": 1.1386178641065025e-07,
+      "loss": 0.0028,
+      "step": 17707
+    },
+    {
+      "epoch": 4.024545454545454,
+      "grad_norm": 1.0287288536962618,
+      "learning_rate": 1.1381043082321013e-07,
+      "loss": 0.0046,
+      "step": 17708
+    },
+    {
+      "epoch": 4.024772727272727,
+      "grad_norm": 0.7783928615530015,
+      "learning_rate": 1.1375908565981697e-07,
+      "loss": 0.0043,
+      "step": 17709
+    },
+    {
+      "epoch": 4.025,
+      "grad_norm": 0.4858823790866675,
+      "learning_rate": 1.1370775092151815e-07,
+      "loss": 0.0024,
+      "step": 17710
+    },
+    {
+      "epoch": 4.025227272727273,
+      "grad_norm": 0.9644677675740115,
+      "learning_rate": 1.1365642660936025e-07,
+      "loss": 0.0049,
+      "step": 17711
+    },
+    {
+      "epoch": 4.025454545454545,
+      "grad_norm": 0.7021973480043195,
+      "learning_rate": 1.1360511272439013e-07,
+      "loss": 0.0058,
+      "step": 17712
+    },
+    {
+      "epoch": 4.025681818181818,
+      "grad_norm": 1.0077243762344457,
+      "learning_rate": 1.1355380926765372e-07,
+      "loss": 0.0056,
+      "step": 17713
+    },
+    {
+      "epoch": 4.025909090909091,
+      "grad_norm": 1.0858299135768823,
+      "learning_rate": 1.1350251624019738e-07,
+      "loss": 0.0039,
+      "step": 17714
+    },
+    {
+      "epoch": 4.026136363636364,
+      "grad_norm": 0.8084002110819889,
+      "learning_rate": 1.1345123364306716e-07,
+      "loss": 0.0058,
+      "step": 17715
+    },
+    {
+      "epoch": 4.026363636363636,
+      "grad_norm": 1.78936128222691,
+      "learning_rate": 1.1339996147730889e-07,
+      "loss": 0.0254,
+      "step": 17716
+    },
+    {
+      "epoch": 4.026590909090909,
+      "grad_norm": 0.41464243377503784,
+      "learning_rate": 1.1334869974396776e-07,
+      "loss": 0.0025,
+      "step": 17717
+    },
+    {
+      "epoch": 4.026818181818181,
+      "grad_norm": 1.4881241078537903,
+      "learning_rate": 1.1329744844408922e-07,
+      "loss": 0.0104,
+      "step": 17718
+    },
+    {
+      "epoch": 4.027045454545455,
+      "grad_norm": 0.7988729175869624,
+      "learning_rate": 1.132462075787185e-07,
+      "loss": 0.0049,
+      "step": 17719
+    },
+    {
+      "epoch": 4.027272727272727,
+      "grad_norm": 0.6484788826926136,
+      "learning_rate": 1.1319497714890034e-07,
+      "loss": 0.0037,
+      "step": 17720
+    },
+    {
+      "epoch": 4.0275,
+      "grad_norm": 0.6217463443666504,
+      "learning_rate": 1.1314375715567968e-07,
+      "loss": 0.003,
+      "step": 17721
+    },
+    {
+      "epoch": 4.027727272727272,
+      "grad_norm": 1.1113202762997612,
+      "learning_rate": 1.1309254760010068e-07,
+      "loss": 0.0044,
+      "step": 17722
+    },
+    {
+      "epoch": 4.027954545454546,
+      "grad_norm": 0.6776016660383757,
+      "learning_rate": 1.1304134848320762e-07,
+      "loss": 0.0047,
+      "step": 17723
+    },
+    {
+      "epoch": 4.028181818181818,
+      "grad_norm": 0.9783376671537147,
+      "learning_rate": 1.129901598060447e-07,
+      "loss": 0.0034,
+      "step": 17724
+    },
+    {
+      "epoch": 4.028409090909091,
+      "grad_norm": 0.6015295698803323,
+      "learning_rate": 1.1293898156965564e-07,
+      "loss": 0.0034,
+      "step": 17725
+    },
+    {
+      "epoch": 4.028636363636363,
+      "grad_norm": 0.656039815328359,
+      "learning_rate": 1.1288781377508406e-07,
+      "loss": 0.0048,
+      "step": 17726
+    },
+    {
+      "epoch": 4.028863636363637,
+      "grad_norm": 0.7938821018209584,
+      "learning_rate": 1.1283665642337342e-07,
+      "loss": 0.0049,
+      "step": 17727
+    },
+    {
+      "epoch": 4.029090909090909,
+      "grad_norm": 0.7241710120482744,
+      "learning_rate": 1.1278550951556687e-07,
+      "loss": 0.0033,
+      "step": 17728
+    },
+    {
+      "epoch": 4.029318181818182,
+      "grad_norm": 0.33918316843874957,
+      "learning_rate": 1.1273437305270751e-07,
+      "loss": 0.0016,
+      "step": 17729
+    },
+    {
+      "epoch": 4.029545454545454,
+      "grad_norm": 0.39575222262751786,
+      "learning_rate": 1.1268324703583786e-07,
+      "loss": 0.0014,
+      "step": 17730
+    },
+    {
+      "epoch": 4.029772727272727,
+      "grad_norm": 0.5529970064174202,
+      "learning_rate": 1.126321314660006e-07,
+      "loss": 0.0036,
+      "step": 17731
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 1.5294954653837471,
+      "learning_rate": 1.1258102634423798e-07,
+      "loss": 0.0167,
+      "step": 17732
+    },
+    {
+      "epoch": 4.030227272727273,
+      "grad_norm": 0.3296401121811765,
+      "learning_rate": 1.1252993167159227e-07,
+      "loss": 0.0012,
+      "step": 17733
+    },
+    {
+      "epoch": 4.030454545454545,
+      "grad_norm": 6.976674201887916,
+      "learning_rate": 1.1247884744910543e-07,
+      "loss": 0.0232,
+      "step": 17734
+    },
+    {
+      "epoch": 4.030681818181818,
+      "grad_norm": 0.8721790488487302,
+      "learning_rate": 1.1242777367781892e-07,
+      "loss": 0.005,
+      "step": 17735
+    },
+    {
+      "epoch": 4.030909090909091,
+      "grad_norm": 0.2820813132039552,
+      "learning_rate": 1.1237671035877431e-07,
+      "loss": 0.0011,
+      "step": 17736
+    },
+    {
+      "epoch": 4.031136363636364,
+      "grad_norm": 1.4881573038157632,
+      "learning_rate": 1.1232565749301289e-07,
+      "loss": 0.0064,
+      "step": 17737
+    },
+    {
+      "epoch": 4.031363636363636,
+      "grad_norm": 0.5586051300638102,
+      "learning_rate": 1.1227461508157574e-07,
+      "loss": 0.0029,
+      "step": 17738
+    },
+    {
+      "epoch": 4.031590909090909,
+      "grad_norm": 0.5616873145929867,
+      "learning_rate": 1.1222358312550364e-07,
+      "loss": 0.0025,
+      "step": 17739
+    },
+    {
+      "epoch": 4.031818181818182,
+      "grad_norm": 1.1665738046459881,
+      "learning_rate": 1.1217256162583734e-07,
+      "loss": 0.0048,
+      "step": 17740
+    },
+    {
+      "epoch": 4.032045454545455,
+      "grad_norm": 1.9321690116133106,
+      "learning_rate": 1.1212155058361713e-07,
+      "loss": 0.0123,
+      "step": 17741
+    },
+    {
+      "epoch": 4.032272727272727,
+      "grad_norm": 0.34019757009187246,
+      "learning_rate": 1.1207054999988324e-07,
+      "loss": 0.0012,
+      "step": 17742
+    },
+    {
+      "epoch": 4.0325,
+      "grad_norm": 2.103972724542233,
+      "learning_rate": 1.120195598756759e-07,
+      "loss": 0.0158,
+      "step": 17743
+    },
+    {
+      "epoch": 4.032727272727273,
+      "grad_norm": 1.071111019094846,
+      "learning_rate": 1.1196858021203454e-07,
+      "loss": 0.0077,
+      "step": 17744
+    },
+    {
+      "epoch": 4.032954545454546,
+      "grad_norm": 0.2707185185477458,
+      "learning_rate": 1.1191761100999884e-07,
+      "loss": 0.0014,
+      "step": 17745
+    },
+    {
+      "epoch": 4.033181818181818,
+      "grad_norm": 1.8449658778590359,
+      "learning_rate": 1.1186665227060814e-07,
+      "loss": 0.0092,
+      "step": 17746
+    },
+    {
+      "epoch": 4.033409090909091,
+      "grad_norm": 1.904249818620109,
+      "learning_rate": 1.118157039949018e-07,
+      "loss": 0.0109,
+      "step": 17747
+    },
+    {
+      "epoch": 4.033636363636363,
+      "grad_norm": 1.4557340871008244,
+      "learning_rate": 1.1176476618391836e-07,
+      "loss": 0.0064,
+      "step": 17748
+    },
+    {
+      "epoch": 4.033863636363637,
+      "grad_norm": 1.2556441668924947,
+      "learning_rate": 1.1171383883869679e-07,
+      "loss": 0.0062,
+      "step": 17749
+    },
+    {
+      "epoch": 4.034090909090909,
+      "grad_norm": 1.7379645604865444,
+      "learning_rate": 1.1166292196027551e-07,
+      "loss": 0.0151,
+      "step": 17750
+    },
+    {
+      "epoch": 4.034318181818182,
+      "grad_norm": 0.8737433164116236,
+      "learning_rate": 1.1161201554969296e-07,
+      "loss": 0.0032,
+      "step": 17751
+    },
+    {
+      "epoch": 4.034545454545454,
+      "grad_norm": 0.585879622234002,
+      "learning_rate": 1.1156111960798688e-07,
+      "loss": 0.0018,
+      "step": 17752
+    },
+    {
+      "epoch": 4.034772727272728,
+      "grad_norm": 0.4418156434985465,
+      "learning_rate": 1.1151023413619531e-07,
+      "loss": 0.0032,
+      "step": 17753
+    },
+    {
+      "epoch": 4.035,
+      "grad_norm": 0.8382519034195322,
+      "learning_rate": 1.1145935913535598e-07,
+      "loss": 0.0031,
+      "step": 17754
+    },
+    {
+      "epoch": 4.035227272727273,
+      "grad_norm": 1.0655154715465285,
+      "learning_rate": 1.1140849460650617e-07,
+      "loss": 0.0049,
+      "step": 17755
+    },
+    {
+      "epoch": 4.035454545454545,
+      "grad_norm": 0.6527621445685906,
+      "learning_rate": 1.1135764055068319e-07,
+      "loss": 0.0015,
+      "step": 17756
+    },
+    {
+      "epoch": 4.035681818181819,
+      "grad_norm": 0.40950218633545804,
+      "learning_rate": 1.1130679696892396e-07,
+      "loss": 0.0028,
+      "step": 17757
+    },
+    {
+      "epoch": 4.035909090909091,
+      "grad_norm": 2.482460734506877,
+      "learning_rate": 1.1125596386226539e-07,
+      "loss": 0.0141,
+      "step": 17758
+    },
+    {
+      "epoch": 4.036136363636364,
+      "grad_norm": 0.4518690383580129,
+      "learning_rate": 1.11205141231744e-07,
+      "loss": 0.0016,
+      "step": 17759
+    },
+    {
+      "epoch": 4.036363636363636,
+      "grad_norm": 0.9787013105295964,
+      "learning_rate": 1.1115432907839625e-07,
+      "loss": 0.0048,
+      "step": 17760
+    },
+    {
+      "epoch": 4.036590909090909,
+      "grad_norm": 0.44934227731478577,
+      "learning_rate": 1.1110352740325806e-07,
+      "loss": 0.0023,
+      "step": 17761
+    },
+    {
+      "epoch": 4.036818181818182,
+      "grad_norm": 0.7468847326968647,
+      "learning_rate": 1.1105273620736545e-07,
+      "loss": 0.0096,
+      "step": 17762
+    },
+    {
+      "epoch": 4.037045454545455,
+      "grad_norm": 1.4212962638319397,
+      "learning_rate": 1.1100195549175421e-07,
+      "loss": 0.0065,
+      "step": 17763
+    },
+    {
+      "epoch": 4.037272727272727,
+      "grad_norm": 0.42469652046025874,
+      "learning_rate": 1.1095118525745997e-07,
+      "loss": 0.0035,
+      "step": 17764
+    },
+    {
+      "epoch": 4.0375,
+      "grad_norm": 1.0742033308641101,
+      "learning_rate": 1.1090042550551773e-07,
+      "loss": 0.0096,
+      "step": 17765
+    },
+    {
+      "epoch": 4.037727272727273,
+      "grad_norm": 0.9155663389468479,
+      "learning_rate": 1.108496762369627e-07,
+      "loss": 0.0051,
+      "step": 17766
+    },
+    {
+      "epoch": 4.037954545454546,
+      "grad_norm": 0.18288163640739005,
+      "learning_rate": 1.1079893745282975e-07,
+      "loss": 0.0027,
+      "step": 17767
+    },
+    {
+      "epoch": 4.038181818181818,
+      "grad_norm": 0.6960720546659964,
+      "learning_rate": 1.1074820915415355e-07,
+      "loss": 0.0034,
+      "step": 17768
+    },
+    {
+      "epoch": 4.038409090909091,
+      "grad_norm": 1.5016963479227923,
+      "learning_rate": 1.1069749134196853e-07,
+      "loss": 0.0106,
+      "step": 17769
+    },
+    {
+      "epoch": 4.038636363636364,
+      "grad_norm": 0.33703285089359175,
+      "learning_rate": 1.1064678401730888e-07,
+      "loss": 0.0012,
+      "step": 17770
+    },
+    {
+      "epoch": 4.0388636363636365,
+      "grad_norm": 0.20797159262927678,
+      "learning_rate": 1.1059608718120867e-07,
+      "loss": 0.0007,
+      "step": 17771
+    },
+    {
+      "epoch": 4.039090909090909,
+      "grad_norm": 1.0973044027725425,
+      "learning_rate": 1.1054540083470164e-07,
+      "loss": 0.0044,
+      "step": 17772
+    },
+    {
+      "epoch": 4.039318181818182,
+      "grad_norm": 0.47666998005884814,
+      "learning_rate": 1.1049472497882158e-07,
+      "loss": 0.0022,
+      "step": 17773
+    },
+    {
+      "epoch": 4.039545454545454,
+      "grad_norm": 1.2218982074873141,
+      "learning_rate": 1.1044405961460156e-07,
+      "loss": 0.0062,
+      "step": 17774
+    },
+    {
+      "epoch": 4.0397727272727275,
+      "grad_norm": 0.43351252063263757,
+      "learning_rate": 1.1039340474307478e-07,
+      "loss": 0.0024,
+      "step": 17775
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 1.2350447408643765,
+      "learning_rate": 1.103427603652743e-07,
+      "loss": 0.0078,
+      "step": 17776
+    },
+    {
+      "epoch": 4.040227272727273,
+      "grad_norm": 1.1531957963072748,
+      "learning_rate": 1.1029212648223285e-07,
+      "loss": 0.0109,
+      "step": 17777
+    },
+    {
+      "epoch": 4.040454545454545,
+      "grad_norm": 1.4960671935075498,
+      "learning_rate": 1.1024150309498299e-07,
+      "loss": 0.015,
+      "step": 17778
+    },
+    {
+      "epoch": 4.0406818181818185,
+      "grad_norm": 0.5775852763413386,
+      "learning_rate": 1.1019089020455676e-07,
+      "loss": 0.0072,
+      "step": 17779
+    },
+    {
+      "epoch": 4.040909090909091,
+      "grad_norm": 0.9248103274197459,
+      "learning_rate": 1.1014028781198643e-07,
+      "loss": 0.0065,
+      "step": 17780
+    },
+    {
+      "epoch": 4.0411363636363635,
+      "grad_norm": 0.73130336529686,
+      "learning_rate": 1.1008969591830392e-07,
+      "loss": 0.002,
+      "step": 17781
+    },
+    {
+      "epoch": 4.041363636363636,
+      "grad_norm": 1.003436200743126,
+      "learning_rate": 1.1003911452454088e-07,
+      "loss": 0.0044,
+      "step": 17782
+    },
+    {
+      "epoch": 4.0415909090909095,
+      "grad_norm": 0.5501051585642937,
+      "learning_rate": 1.0998854363172864e-07,
+      "loss": 0.0021,
+      "step": 17783
+    },
+    {
+      "epoch": 4.041818181818182,
+      "grad_norm": 0.5531868471387533,
+      "learning_rate": 1.0993798324089839e-07,
+      "loss": 0.0013,
+      "step": 17784
+    },
+    {
+      "epoch": 4.0420454545454545,
+      "grad_norm": 0.7394886600567813,
+      "learning_rate": 1.0988743335308124e-07,
+      "loss": 0.0054,
+      "step": 17785
+    },
+    {
+      "epoch": 4.042272727272727,
+      "grad_norm": 0.3409743387952965,
+      "learning_rate": 1.0983689396930807e-07,
+      "loss": 0.001,
+      "step": 17786
+    },
+    {
+      "epoch": 4.0425,
+      "grad_norm": 0.18517938785722252,
+      "learning_rate": 1.0978636509060926e-07,
+      "loss": 0.0007,
+      "step": 17787
+    },
+    {
+      "epoch": 4.042727272727273,
+      "grad_norm": 0.812322186566946,
+      "learning_rate": 1.097358467180154e-07,
+      "loss": 0.0033,
+      "step": 17788
+    },
+    {
+      "epoch": 4.0429545454545455,
+      "grad_norm": 1.0430631333601512,
+      "learning_rate": 1.096853388525565e-07,
+      "loss": 0.0039,
+      "step": 17789
+    },
+    {
+      "epoch": 4.043181818181818,
+      "grad_norm": 1.3821342683054103,
+      "learning_rate": 1.0963484149526263e-07,
+      "loss": 0.0092,
+      "step": 17790
+    },
+    {
+      "epoch": 4.0434090909090905,
+      "grad_norm": 0.6096716680066313,
+      "learning_rate": 1.0958435464716352e-07,
+      "loss": 0.0019,
+      "step": 17791
+    },
+    {
+      "epoch": 4.043636363636364,
+      "grad_norm": 0.5784302329146593,
+      "learning_rate": 1.095338783092885e-07,
+      "loss": 0.0021,
+      "step": 17792
+    },
+    {
+      "epoch": 4.043863636363636,
+      "grad_norm": 1.7754879285636949,
+      "learning_rate": 1.09483412482667e-07,
+      "loss": 0.0062,
+      "step": 17793
+    },
+    {
+      "epoch": 4.044090909090909,
+      "grad_norm": 1.233000056866341,
+      "learning_rate": 1.0943295716832801e-07,
+      "loss": 0.0037,
+      "step": 17794
+    },
+    {
+      "epoch": 4.0443181818181815,
+      "grad_norm": 0.5542615332113369,
+      "learning_rate": 1.093825123673007e-07,
+      "loss": 0.0019,
+      "step": 17795
+    },
+    {
+      "epoch": 4.044545454545455,
+      "grad_norm": 0.09587706593309385,
+      "learning_rate": 1.0933207808061334e-07,
+      "loss": 0.0004,
+      "step": 17796
+    },
+    {
+      "epoch": 4.044772727272727,
+      "grad_norm": 1.1069977620530342,
+      "learning_rate": 1.0928165430929456e-07,
+      "loss": 0.0064,
+      "step": 17797
+    },
+    {
+      "epoch": 4.045,
+      "grad_norm": 1.2505799582495027,
+      "learning_rate": 1.092312410543725e-07,
+      "loss": 0.0041,
+      "step": 17798
+    },
+    {
+      "epoch": 4.0452272727272724,
+      "grad_norm": 1.0949726124276398,
+      "learning_rate": 1.0918083831687528e-07,
+      "loss": 0.004,
+      "step": 17799
+    },
+    {
+      "epoch": 4.045454545454546,
+      "grad_norm": 0.5599764978053116,
+      "learning_rate": 1.0913044609783069e-07,
+      "loss": 0.0036,
+      "step": 17800
+    },
+    {
+      "epoch": 4.045681818181818,
+      "grad_norm": 0.6752245692015577,
+      "learning_rate": 1.0908006439826625e-07,
+      "loss": 0.0039,
+      "step": 17801
+    },
+    {
+      "epoch": 4.045909090909091,
+      "grad_norm": 2.104640480529926,
+      "learning_rate": 1.090296932192094e-07,
+      "loss": 0.0096,
+      "step": 17802
+    },
+    {
+      "epoch": 4.046136363636363,
+      "grad_norm": 0.8993456196864913,
+      "learning_rate": 1.0897933256168724e-07,
+      "loss": 0.0075,
+      "step": 17803
+    },
+    {
+      "epoch": 4.046363636363636,
+      "grad_norm": 0.8286356091973804,
+      "learning_rate": 1.0892898242672694e-07,
+      "loss": 0.0037,
+      "step": 17804
+    },
+    {
+      "epoch": 4.046590909090909,
+      "grad_norm": 1.1753328599078465,
+      "learning_rate": 1.088786428153548e-07,
+      "loss": 0.0119,
+      "step": 17805
+    },
+    {
+      "epoch": 4.046818181818182,
+      "grad_norm": 0.5579100073204175,
+      "learning_rate": 1.0882831372859761e-07,
+      "loss": 0.0026,
+      "step": 17806
+    },
+    {
+      "epoch": 4.047045454545454,
+      "grad_norm": 0.548615798478767,
+      "learning_rate": 1.0877799516748159e-07,
+      "loss": 0.0021,
+      "step": 17807
+    },
+    {
+      "epoch": 4.047272727272727,
+      "grad_norm": 1.289400546154434,
+      "learning_rate": 1.0872768713303298e-07,
+      "loss": 0.0051,
+      "step": 17808
+    },
+    {
+      "epoch": 4.0475,
+      "grad_norm": 0.15561431441597684,
+      "learning_rate": 1.0867738962627738e-07,
+      "loss": 0.0003,
+      "step": 17809
+    },
+    {
+      "epoch": 4.047727272727273,
+      "grad_norm": 0.6449515076491389,
+      "learning_rate": 1.0862710264824057e-07,
+      "loss": 0.0048,
+      "step": 17810
+    },
+    {
+      "epoch": 4.047954545454545,
+      "grad_norm": 0.7299699653404326,
+      "learning_rate": 1.08576826199948e-07,
+      "loss": 0.0027,
+      "step": 17811
+    },
+    {
+      "epoch": 4.048181818181818,
+      "grad_norm": 2.064583576335398,
+      "learning_rate": 1.0852656028242495e-07,
+      "loss": 0.0077,
+      "step": 17812
+    },
+    {
+      "epoch": 4.048409090909091,
+      "grad_norm": 0.9041050922456804,
+      "learning_rate": 1.0847630489669642e-07,
+      "loss": 0.0046,
+      "step": 17813
+    },
+    {
+      "epoch": 4.048636363636364,
+      "grad_norm": 1.1139034845016387,
+      "learning_rate": 1.0842606004378712e-07,
+      "loss": 0.006,
+      "step": 17814
+    },
+    {
+      "epoch": 4.048863636363636,
+      "grad_norm": 1.1944093396074011,
+      "learning_rate": 1.0837582572472159e-07,
+      "loss": 0.0052,
+      "step": 17815
+    },
+    {
+      "epoch": 4.049090909090909,
+      "grad_norm": 0.22832905790733968,
+      "learning_rate": 1.0832560194052432e-07,
+      "loss": 0.0008,
+      "step": 17816
+    },
+    {
+      "epoch": 4.049318181818181,
+      "grad_norm": 1.3388484410688406,
+      "learning_rate": 1.0827538869221936e-07,
+      "loss": 0.0056,
+      "step": 17817
+    },
+    {
+      "epoch": 4.049545454545455,
+      "grad_norm": 1.8966128586915405,
+      "learning_rate": 1.0822518598083074e-07,
+      "loss": 0.0285,
+      "step": 17818
+    },
+    {
+      "epoch": 4.049772727272727,
+      "grad_norm": 0.8148563476826441,
+      "learning_rate": 1.0817499380738215e-07,
+      "loss": 0.0065,
+      "step": 17819
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 0.27489105931302804,
+      "learning_rate": 1.0812481217289707e-07,
+      "loss": 0.0014,
+      "step": 17820
+    },
+    {
+      "epoch": 4.050227272727272,
+      "grad_norm": 0.31999583205449966,
+      "learning_rate": 1.0807464107839879e-07,
+      "loss": 0.0012,
+      "step": 17821
+    },
+    {
+      "epoch": 4.050454545454546,
+      "grad_norm": 7.567920231416306,
+      "learning_rate": 1.080244805249106e-07,
+      "loss": 0.0091,
+      "step": 17822
+    },
+    {
+      "epoch": 4.050681818181818,
+      "grad_norm": 0.8230590401505782,
+      "learning_rate": 1.07974330513455e-07,
+      "loss": 0.0061,
+      "step": 17823
+    },
+    {
+      "epoch": 4.050909090909091,
+      "grad_norm": 1.0600299603697316,
+      "learning_rate": 1.0792419104505479e-07,
+      "loss": 0.0019,
+      "step": 17824
+    },
+    {
+      "epoch": 4.051136363636363,
+      "grad_norm": 1.198848156311976,
+      "learning_rate": 1.0787406212073241e-07,
+      "loss": 0.0021,
+      "step": 17825
+    },
+    {
+      "epoch": 4.051363636363637,
+      "grad_norm": 0.193195662669385,
+      "learning_rate": 1.0782394374151024e-07,
+      "loss": 0.0006,
+      "step": 17826
+    },
+    {
+      "epoch": 4.051590909090909,
+      "grad_norm": 0.4172278354034069,
+      "learning_rate": 1.0777383590840997e-07,
+      "loss": 0.0016,
+      "step": 17827
+    },
+    {
+      "epoch": 4.051818181818182,
+      "grad_norm": 0.549098337208047,
+      "learning_rate": 1.0772373862245347e-07,
+      "loss": 0.0024,
+      "step": 17828
+    },
+    {
+      "epoch": 4.052045454545454,
+      "grad_norm": 0.7963873477524867,
+      "learning_rate": 1.0767365188466245e-07,
+      "loss": 0.0068,
+      "step": 17829
+    },
+    {
+      "epoch": 4.052272727272728,
+      "grad_norm": 0.7583115966381276,
+      "learning_rate": 1.0762357569605822e-07,
+      "loss": 0.0027,
+      "step": 17830
+    },
+    {
+      "epoch": 4.0525,
+      "grad_norm": 0.9692424316609421,
+      "learning_rate": 1.0757351005766185e-07,
+      "loss": 0.0081,
+      "step": 17831
+    },
+    {
+      "epoch": 4.052727272727273,
+      "grad_norm": 0.8374217823885601,
+      "learning_rate": 1.0752345497049427e-07,
+      "loss": 0.0036,
+      "step": 17832
+    },
+    {
+      "epoch": 4.052954545454545,
+      "grad_norm": 0.5614503312050786,
+      "learning_rate": 1.0747341043557631e-07,
+      "loss": 0.0019,
+      "step": 17833
+    },
+    {
+      "epoch": 4.053181818181818,
+      "grad_norm": 0.7952518884815734,
+      "learning_rate": 1.0742337645392835e-07,
+      "loss": 0.0042,
+      "step": 17834
+    },
+    {
+      "epoch": 4.053409090909091,
+      "grad_norm": 0.3390612224139622,
+      "learning_rate": 1.0737335302657088e-07,
+      "loss": 0.002,
+      "step": 17835
+    },
+    {
+      "epoch": 4.053636363636364,
+      "grad_norm": 0.34769903417183434,
+      "learning_rate": 1.0732334015452359e-07,
+      "loss": 0.0015,
+      "step": 17836
+    },
+    {
+      "epoch": 4.053863636363636,
+      "grad_norm": 1.0356050935715961,
+      "learning_rate": 1.0727333783880654e-07,
+      "loss": 0.0036,
+      "step": 17837
+    },
+    {
+      "epoch": 4.054090909090909,
+      "grad_norm": 1.1409790438274268,
+      "learning_rate": 1.0722334608043936e-07,
+      "loss": 0.0042,
+      "step": 17838
+    },
+    {
+      "epoch": 4.054318181818182,
+      "grad_norm": 1.129751837694077,
+      "learning_rate": 1.071733648804417e-07,
+      "loss": 0.0057,
+      "step": 17839
+    },
+    {
+      "epoch": 4.054545454545455,
+      "grad_norm": 0.7753621965395049,
+      "learning_rate": 1.0712339423983226e-07,
+      "loss": 0.0041,
+      "step": 17840
+    },
+    {
+      "epoch": 4.054772727272727,
+      "grad_norm": 1.9801652049559784,
+      "learning_rate": 1.0707343415963037e-07,
+      "loss": 0.0125,
+      "step": 17841
+    },
+    {
+      "epoch": 4.055,
+      "grad_norm": 1.201893095994023,
+      "learning_rate": 1.0702348464085466e-07,
+      "loss": 0.0053,
+      "step": 17842
+    },
+    {
+      "epoch": 4.055227272727273,
+      "grad_norm": 1.6360581035609951,
+      "learning_rate": 1.0697354568452382e-07,
+      "loss": 0.0086,
+      "step": 17843
+    },
+    {
+      "epoch": 4.055454545454546,
+      "grad_norm": 0.7885260201331235,
+      "learning_rate": 1.0692361729165629e-07,
+      "loss": 0.004,
+      "step": 17844
+    },
+    {
+      "epoch": 4.055681818181818,
+      "grad_norm": 1.1935171074435489,
+      "learning_rate": 1.0687369946326985e-07,
+      "loss": 0.0127,
+      "step": 17845
+    },
+    {
+      "epoch": 4.055909090909091,
+      "grad_norm": 0.3289865976627132,
+      "learning_rate": 1.0682379220038258e-07,
+      "loss": 0.001,
+      "step": 17846
+    },
+    {
+      "epoch": 4.056136363636363,
+      "grad_norm": 0.7949273721609029,
+      "learning_rate": 1.0677389550401207e-07,
+      "loss": 0.0045,
+      "step": 17847
+    },
+    {
+      "epoch": 4.056363636363637,
+      "grad_norm": 0.3346921268126483,
+      "learning_rate": 1.0672400937517625e-07,
+      "loss": 0.0007,
+      "step": 17848
+    },
+    {
+      "epoch": 4.056590909090909,
+      "grad_norm": 0.6808170564894928,
+      "learning_rate": 1.0667413381489188e-07,
+      "loss": 0.005,
+      "step": 17849
+    },
+    {
+      "epoch": 4.056818181818182,
+      "grad_norm": 0.876882736803762,
+      "learning_rate": 1.0662426882417622e-07,
+      "loss": 0.0038,
+      "step": 17850
+    },
+    {
+      "epoch": 4.057045454545454,
+      "grad_norm": 1.5046422689544496,
+      "learning_rate": 1.0657441440404606e-07,
+      "loss": 0.0113,
+      "step": 17851
+    },
+    {
+      "epoch": 4.057272727272728,
+      "grad_norm": 0.8705430143249582,
+      "learning_rate": 1.0652457055551818e-07,
+      "loss": 0.0037,
+      "step": 17852
+    },
+    {
+      "epoch": 4.0575,
+      "grad_norm": 0.5801464180699935,
+      "learning_rate": 1.0647473727960864e-07,
+      "loss": 0.0041,
+      "step": 17853
+    },
+    {
+      "epoch": 4.057727272727273,
+      "grad_norm": 0.7391458102381611,
+      "learning_rate": 1.064249145773339e-07,
+      "loss": 0.0031,
+      "step": 17854
+    },
+    {
+      "epoch": 4.057954545454545,
+      "grad_norm": 0.5961641131103154,
+      "learning_rate": 1.0637510244970977e-07,
+      "loss": 0.0021,
+      "step": 17855
+    },
+    {
+      "epoch": 4.058181818181819,
+      "grad_norm": 0.6106541303076155,
+      "learning_rate": 1.0632530089775217e-07,
+      "loss": 0.0024,
+      "step": 17856
+    },
+    {
+      "epoch": 4.058409090909091,
+      "grad_norm": 2.364828028437014,
+      "learning_rate": 1.0627550992247666e-07,
+      "loss": 0.0077,
+      "step": 17857
+    },
+    {
+      "epoch": 4.058636363636364,
+      "grad_norm": 1.063641408499583,
+      "learning_rate": 1.0622572952489835e-07,
+      "loss": 0.0064,
+      "step": 17858
+    },
+    {
+      "epoch": 4.058863636363636,
+      "grad_norm": 0.6852641473862352,
+      "learning_rate": 1.0617595970603252e-07,
+      "loss": 0.0033,
+      "step": 17859
+    },
+    {
+      "epoch": 4.0590909090909095,
+      "grad_norm": 0.7580860881832775,
+      "learning_rate": 1.0612620046689392e-07,
+      "loss": 0.0041,
+      "step": 17860
+    },
+    {
+      "epoch": 4.059318181818182,
+      "grad_norm": 0.60699656007183,
+      "learning_rate": 1.0607645180849738e-07,
+      "loss": 0.0029,
+      "step": 17861
+    },
+    {
+      "epoch": 4.059545454545455,
+      "grad_norm": 0.7504974489270367,
+      "learning_rate": 1.060267137318572e-07,
+      "loss": 0.0022,
+      "step": 17862
+    },
+    {
+      "epoch": 4.059772727272727,
+      "grad_norm": 0.32878365545126403,
+      "learning_rate": 1.059769862379878e-07,
+      "loss": 0.0018,
+      "step": 17863
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 0.32611742450257886,
+      "learning_rate": 1.0592726932790315e-07,
+      "loss": 0.0014,
+      "step": 17864
+    },
+    {
+      "epoch": 4.060227272727273,
+      "grad_norm": 1.3484930918155822,
+      "learning_rate": 1.0587756300261702e-07,
+      "loss": 0.0077,
+      "step": 17865
+    },
+    {
+      "epoch": 4.0604545454545455,
+      "grad_norm": 2.33136875002568,
+      "learning_rate": 1.0582786726314319e-07,
+      "loss": 0.006,
+      "step": 17866
+    },
+    {
+      "epoch": 4.060681818181818,
+      "grad_norm": 0.643148190548915,
+      "learning_rate": 1.0577818211049481e-07,
+      "loss": 0.0016,
+      "step": 17867
+    },
+    {
+      "epoch": 4.060909090909091,
+      "grad_norm": 0.33553012933936055,
+      "learning_rate": 1.0572850754568501e-07,
+      "loss": 0.0014,
+      "step": 17868
+    },
+    {
+      "epoch": 4.061136363636364,
+      "grad_norm": 0.6782532631980546,
+      "learning_rate": 1.0567884356972695e-07,
+      "loss": 0.0029,
+      "step": 17869
+    },
+    {
+      "epoch": 4.0613636363636365,
+      "grad_norm": 0.8249900159888562,
+      "learning_rate": 1.0562919018363337e-07,
+      "loss": 0.0022,
+      "step": 17870
+    },
+    {
+      "epoch": 4.061590909090909,
+      "grad_norm": 0.7293165556063702,
+      "learning_rate": 1.0557954738841656e-07,
+      "loss": 0.0038,
+      "step": 17871
+    },
+    {
+      "epoch": 4.0618181818181816,
+      "grad_norm": 0.4437331119179323,
+      "learning_rate": 1.0552991518508901e-07,
+      "loss": 0.0016,
+      "step": 17872
+    },
+    {
+      "epoch": 4.062045454545455,
+      "grad_norm": 0.7408149365738019,
+      "learning_rate": 1.0548029357466267e-07,
+      "loss": 0.0043,
+      "step": 17873
+    },
+    {
+      "epoch": 4.0622727272727275,
+      "grad_norm": 1.3323716366688672,
+      "learning_rate": 1.054306825581497e-07,
+      "loss": 0.0126,
+      "step": 17874
+    },
+    {
+      "epoch": 4.0625,
+      "grad_norm": 0.48251685778834025,
+      "learning_rate": 1.053810821365614e-07,
+      "loss": 0.0012,
+      "step": 17875
+    },
+    {
+      "epoch": 4.0627272727272725,
+      "grad_norm": 2.0197252464940356,
+      "learning_rate": 1.0533149231090916e-07,
+      "loss": 0.0083,
+      "step": 17876
+    },
+    {
+      "epoch": 4.062954545454545,
+      "grad_norm": 0.45628292446782864,
+      "learning_rate": 1.0528191308220463e-07,
+      "loss": 0.0025,
+      "step": 17877
+    },
+    {
+      "epoch": 4.0631818181818184,
+      "grad_norm": 0.5307242962276323,
+      "learning_rate": 1.052323444514585e-07,
+      "loss": 0.0018,
+      "step": 17878
+    },
+    {
+      "epoch": 4.063409090909091,
+      "grad_norm": 0.5608428528157072,
+      "learning_rate": 1.0518278641968186e-07,
+      "loss": 0.0033,
+      "step": 17879
+    },
+    {
+      "epoch": 4.0636363636363635,
+      "grad_norm": 1.8217232941336479,
+      "learning_rate": 1.0513323898788494e-07,
+      "loss": 0.0049,
+      "step": 17880
+    },
+    {
+      "epoch": 4.063863636363636,
+      "grad_norm": 1.385649919424245,
+      "learning_rate": 1.0508370215707823e-07,
+      "loss": 0.007,
+      "step": 17881
+    },
+    {
+      "epoch": 4.064090909090909,
+      "grad_norm": 0.5221699058967534,
+      "learning_rate": 1.0503417592827185e-07,
+      "loss": 0.0013,
+      "step": 17882
+    },
+    {
+      "epoch": 4.064318181818182,
+      "grad_norm": 0.5381063928180095,
+      "learning_rate": 1.0498466030247595e-07,
+      "loss": 0.0026,
+      "step": 17883
+    },
+    {
+      "epoch": 4.0645454545454545,
+      "grad_norm": 1.1807603328491216,
+      "learning_rate": 1.049351552806998e-07,
+      "loss": 0.0078,
+      "step": 17884
+    },
+    {
+      "epoch": 4.064772727272727,
+      "grad_norm": 0.3134071694994194,
+      "learning_rate": 1.048856608639532e-07,
+      "loss": 0.0008,
+      "step": 17885
+    },
+    {
+      "epoch": 4.065,
+      "grad_norm": 0.6322000343247481,
+      "learning_rate": 1.0483617705324536e-07,
+      "loss": 0.0026,
+      "step": 17886
+    },
+    {
+      "epoch": 4.065227272727273,
+      "grad_norm": 1.5512076657903853,
+      "learning_rate": 1.0478670384958547e-07,
+      "loss": 0.0082,
+      "step": 17887
+    },
+    {
+      "epoch": 4.065454545454545,
+      "grad_norm": 0.3526457741400065,
+      "learning_rate": 1.0473724125398217e-07,
+      "loss": 0.0018,
+      "step": 17888
+    },
+    {
+      "epoch": 4.065681818181818,
+      "grad_norm": 0.10873916465427154,
+      "learning_rate": 1.046877892674441e-07,
+      "loss": 0.0004,
+      "step": 17889
+    },
+    {
+      "epoch": 4.0659090909090905,
+      "grad_norm": 2.1101564144792397,
+      "learning_rate": 1.0463834789097981e-07,
+      "loss": 0.0087,
+      "step": 17890
+    },
+    {
+      "epoch": 4.066136363636364,
+      "grad_norm": 0.7535553164034671,
+      "learning_rate": 1.0458891712559732e-07,
+      "loss": 0.0035,
+      "step": 17891
+    },
+    {
+      "epoch": 4.066363636363636,
+      "grad_norm": 0.21442685535536354,
+      "learning_rate": 1.045394969723048e-07,
+      "loss": 0.0004,
+      "step": 17892
+    },
+    {
+      "epoch": 4.066590909090909,
+      "grad_norm": 2.0248896266460688,
+      "learning_rate": 1.0449008743210989e-07,
+      "loss": 0.0056,
+      "step": 17893
+    },
+    {
+      "epoch": 4.066818181818181,
+      "grad_norm": 0.9951690928230622,
+      "learning_rate": 1.044406885060202e-07,
+      "loss": 0.0064,
+      "step": 17894
+    },
+    {
+      "epoch": 4.067045454545455,
+      "grad_norm": 1.0109388544404143,
+      "learning_rate": 1.0439130019504306e-07,
+      "loss": 0.01,
+      "step": 17895
+    },
+    {
+      "epoch": 4.067272727272727,
+      "grad_norm": 1.060272671200779,
+      "learning_rate": 1.0434192250018564e-07,
+      "loss": 0.0046,
+      "step": 17896
+    },
+    {
+      "epoch": 4.0675,
+      "grad_norm": 1.5283360250271807,
+      "learning_rate": 1.0429255542245463e-07,
+      "loss": 0.006,
+      "step": 17897
+    },
+    {
+      "epoch": 4.067727272727272,
+      "grad_norm": 1.0307807524056525,
+      "learning_rate": 1.0424319896285687e-07,
+      "loss": 0.0079,
+      "step": 17898
+    },
+    {
+      "epoch": 4.067954545454546,
+      "grad_norm": 0.44544304047632405,
+      "learning_rate": 1.0419385312239877e-07,
+      "loss": 0.0016,
+      "step": 17899
+    },
+    {
+      "epoch": 4.068181818181818,
+      "grad_norm": 2.29920466060641,
+      "learning_rate": 1.0414451790208661e-07,
+      "loss": 0.0134,
+      "step": 17900
+    },
+    {
+      "epoch": 4.068409090909091,
+      "grad_norm": 2.4855696106744603,
+      "learning_rate": 1.0409519330292661e-07,
+      "loss": 0.0165,
+      "step": 17901
+    },
+    {
+      "epoch": 4.068636363636363,
+      "grad_norm": 0.9131339957527329,
+      "learning_rate": 1.0404587932592421e-07,
+      "loss": 0.0049,
+      "step": 17902
+    },
+    {
+      "epoch": 4.068863636363637,
+      "grad_norm": 0.6919298512750645,
+      "learning_rate": 1.0399657597208513e-07,
+      "loss": 0.0025,
+      "step": 17903
+    },
+    {
+      "epoch": 4.069090909090909,
+      "grad_norm": 0.7967462371747374,
+      "learning_rate": 1.0394728324241488e-07,
+      "loss": 0.0025,
+      "step": 17904
+    },
+    {
+      "epoch": 4.069318181818182,
+      "grad_norm": 1.1220338718028797,
+      "learning_rate": 1.0389800113791863e-07,
+      "loss": 0.0038,
+      "step": 17905
+    },
+    {
+      "epoch": 4.069545454545454,
+      "grad_norm": 2.297841139867187,
+      "learning_rate": 1.0384872965960101e-07,
+      "loss": 0.0106,
+      "step": 17906
+    },
+    {
+      "epoch": 4.069772727272727,
+      "grad_norm": 0.6136366884750393,
+      "learning_rate": 1.0379946880846709e-07,
+      "loss": 0.0019,
+      "step": 17907
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 0.7545441462328147,
+      "learning_rate": 1.0375021858552131e-07,
+      "loss": 0.0036,
+      "step": 17908
+    },
+    {
+      "epoch": 4.070227272727273,
+      "grad_norm": 0.5618630719328621,
+      "learning_rate": 1.0370097899176804e-07,
+      "loss": 0.0019,
+      "step": 17909
+    },
+    {
+      "epoch": 4.070454545454545,
+      "grad_norm": 1.3974175118355088,
+      "learning_rate": 1.0365175002821124e-07,
+      "loss": 0.0038,
+      "step": 17910
+    },
+    {
+      "epoch": 4.070681818181818,
+      "grad_norm": 1.2544603778057621,
+      "learning_rate": 1.0360253169585472e-07,
+      "loss": 0.0067,
+      "step": 17911
+    },
+    {
+      "epoch": 4.070909090909091,
+      "grad_norm": 1.404236445085672,
+      "learning_rate": 1.0355332399570219e-07,
+      "loss": 0.0132,
+      "step": 17912
+    },
+    {
+      "epoch": 4.071136363636364,
+      "grad_norm": 0.639949964126301,
+      "learning_rate": 1.0350412692875713e-07,
+      "loss": 0.0019,
+      "step": 17913
+    },
+    {
+      "epoch": 4.071363636363636,
+      "grad_norm": 0.5143362339764432,
+      "learning_rate": 1.0345494049602285e-07,
+      "loss": 0.0018,
+      "step": 17914
+    },
+    {
+      "epoch": 4.071590909090909,
+      "grad_norm": 0.8735959713046568,
+      "learning_rate": 1.0340576469850205e-07,
+      "loss": 0.0036,
+      "step": 17915
+    },
+    {
+      "epoch": 4.071818181818182,
+      "grad_norm": 0.5924952843408161,
+      "learning_rate": 1.0335659953719775e-07,
+      "loss": 0.0029,
+      "step": 17916
+    },
+    {
+      "epoch": 4.072045454545455,
+      "grad_norm": 1.131175337769912,
+      "learning_rate": 1.0330744501311239e-07,
+      "loss": 0.0091,
+      "step": 17917
+    },
+    {
+      "epoch": 4.072272727272727,
+      "grad_norm": 0.47824361866885795,
+      "learning_rate": 1.0325830112724852e-07,
+      "loss": 0.0011,
+      "step": 17918
+    },
+    {
+      "epoch": 4.0725,
+      "grad_norm": 1.7813325897691716,
+      "learning_rate": 1.0320916788060797e-07,
+      "loss": 0.0116,
+      "step": 17919
+    },
+    {
+      "epoch": 4.072727272727272,
+      "grad_norm": 1.3868248606568991,
+      "learning_rate": 1.031600452741928e-07,
+      "loss": 0.0048,
+      "step": 17920
+    },
+    {
+      "epoch": 4.072954545454546,
+      "grad_norm": 1.1715665650562819,
+      "learning_rate": 1.0311093330900468e-07,
+      "loss": 0.0084,
+      "step": 17921
+    },
+    {
+      "epoch": 4.073181818181818,
+      "grad_norm": 1.357792076974556,
+      "learning_rate": 1.0306183198604518e-07,
+      "loss": 0.0122,
+      "step": 17922
+    },
+    {
+      "epoch": 4.073409090909091,
+      "grad_norm": 0.8176798327050182,
+      "learning_rate": 1.0301274130631547e-07,
+      "loss": 0.0095,
+      "step": 17923
+    },
+    {
+      "epoch": 4.073636363636363,
+      "grad_norm": 1.6699594019373762,
+      "learning_rate": 1.029636612708166e-07,
+      "loss": 0.0071,
+      "step": 17924
+    },
+    {
+      "epoch": 4.073863636363637,
+      "grad_norm": 0.6036955341464298,
+      "learning_rate": 1.0291459188054944e-07,
+      "loss": 0.0022,
+      "step": 17925
+    },
+    {
+      "epoch": 4.074090909090909,
+      "grad_norm": 0.28510171589500566,
+      "learning_rate": 1.0286553313651457e-07,
+      "loss": 0.0014,
+      "step": 17926
+    },
+    {
+      "epoch": 4.074318181818182,
+      "grad_norm": 0.4725438060507432,
+      "learning_rate": 1.0281648503971258e-07,
+      "loss": 0.0019,
+      "step": 17927
+    },
+    {
+      "epoch": 4.074545454545454,
+      "grad_norm": 1.721095916021691,
+      "learning_rate": 1.0276744759114325e-07,
+      "loss": 0.0136,
+      "step": 17928
+    },
+    {
+      "epoch": 4.074772727272728,
+      "grad_norm": 0.5568913022688995,
+      "learning_rate": 1.027184207918068e-07,
+      "loss": 0.0012,
+      "step": 17929
+    },
+    {
+      "epoch": 4.075,
+      "grad_norm": 0.5224249163839856,
+      "learning_rate": 1.0266940464270285e-07,
+      "loss": 0.0024,
+      "step": 17930
+    },
+    {
+      "epoch": 4.075227272727273,
+      "grad_norm": 0.7997673555218927,
+      "learning_rate": 1.0262039914483123e-07,
+      "loss": 0.0047,
+      "step": 17931
+    },
+    {
+      "epoch": 4.075454545454545,
+      "grad_norm": 1.4818483257180792,
+      "learning_rate": 1.0257140429919079e-07,
+      "loss": 0.0033,
+      "step": 17932
+    },
+    {
+      "epoch": 4.075681818181819,
+      "grad_norm": 0.5845758258227483,
+      "learning_rate": 1.0252242010678093e-07,
+      "loss": 0.0024,
+      "step": 17933
+    },
+    {
+      "epoch": 4.075909090909091,
+      "grad_norm": 1.7057655234162834,
+      "learning_rate": 1.0247344656860038e-07,
+      "loss": 0.0064,
+      "step": 17934
+    },
+    {
+      "epoch": 4.076136363636364,
+      "grad_norm": 0.44716849075996595,
+      "learning_rate": 1.0242448368564785e-07,
+      "loss": 0.0048,
+      "step": 17935
+    },
+    {
+      "epoch": 4.076363636363636,
+      "grad_norm": 1.536338214143035,
+      "learning_rate": 1.0237553145892177e-07,
+      "loss": 0.0095,
+      "step": 17936
+    },
+    {
+      "epoch": 4.076590909090909,
+      "grad_norm": 1.1788578063571304,
+      "learning_rate": 1.0232658988942039e-07,
+      "loss": 0.0027,
+      "step": 17937
+    },
+    {
+      "epoch": 4.076818181818182,
+      "grad_norm": 1.0630771463859594,
+      "learning_rate": 1.0227765897814171e-07,
+      "loss": 0.0088,
+      "step": 17938
+    },
+    {
+      "epoch": 4.077045454545455,
+      "grad_norm": 0.39464663945556344,
+      "learning_rate": 1.0222873872608348e-07,
+      "loss": 0.0018,
+      "step": 17939
+    },
+    {
+      "epoch": 4.077272727272727,
+      "grad_norm": 0.7371484331043383,
+      "learning_rate": 1.0217982913424344e-07,
+      "loss": 0.0107,
+      "step": 17940
+    },
+    {
+      "epoch": 4.0775,
+      "grad_norm": 1.3150441065354692,
+      "learning_rate": 1.0213093020361865e-07,
+      "loss": 0.0031,
+      "step": 17941
+    },
+    {
+      "epoch": 4.077727272727273,
+      "grad_norm": 0.9494384739851571,
+      "learning_rate": 1.0208204193520637e-07,
+      "loss": 0.0033,
+      "step": 17942
+    },
+    {
+      "epoch": 4.077954545454546,
+      "grad_norm": 0.285023146082904,
+      "learning_rate": 1.0203316433000356e-07,
+      "loss": 0.0011,
+      "step": 17943
+    },
+    {
+      "epoch": 4.078181818181818,
+      "grad_norm": 1.8239060445619866,
+      "learning_rate": 1.0198429738900687e-07,
+      "loss": 0.0109,
+      "step": 17944
+    },
+    {
+      "epoch": 4.078409090909091,
+      "grad_norm": 2.631365540697531,
+      "learning_rate": 1.0193544111321299e-07,
+      "loss": 0.0395,
+      "step": 17945
+    },
+    {
+      "epoch": 4.078636363636364,
+      "grad_norm": 1.3345722085921679,
+      "learning_rate": 1.0188659550361787e-07,
+      "loss": 0.0067,
+      "step": 17946
+    },
+    {
+      "epoch": 4.078863636363637,
+      "grad_norm": 0.541000670575789,
+      "learning_rate": 1.0183776056121766e-07,
+      "loss": 0.0022,
+      "step": 17947
+    },
+    {
+      "epoch": 4.079090909090909,
+      "grad_norm": 0.6238334703412757,
+      "learning_rate": 1.0178893628700823e-07,
+      "loss": 0.0018,
+      "step": 17948
+    },
+    {
+      "epoch": 4.079318181818182,
+      "grad_norm": 0.32526454089515405,
+      "learning_rate": 1.017401226819853e-07,
+      "loss": 0.0015,
+      "step": 17949
+    },
+    {
+      "epoch": 4.079545454545454,
+      "grad_norm": 2.213948632032243,
+      "learning_rate": 1.0169131974714398e-07,
+      "loss": 0.0176,
+      "step": 17950
+    },
+    {
+      "epoch": 4.0797727272727276,
+      "grad_norm": 0.652539645916412,
+      "learning_rate": 1.0164252748347967e-07,
+      "loss": 0.0028,
+      "step": 17951
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 1.4642778807398376,
+      "learning_rate": 1.0159374589198724e-07,
+      "loss": 0.0065,
+      "step": 17952
+    },
+    {
+      "epoch": 4.080227272727273,
+      "grad_norm": 0.9275597643731088,
+      "learning_rate": 1.0154497497366147e-07,
+      "loss": 0.0114,
+      "step": 17953
+    },
+    {
+      "epoch": 4.080454545454545,
+      "grad_norm": 1.058627607558428,
+      "learning_rate": 1.0149621472949691e-07,
+      "loss": 0.0027,
+      "step": 17954
+    },
+    {
+      "epoch": 4.0806818181818185,
+      "grad_norm": 1.6000930080323315,
+      "learning_rate": 1.0144746516048777e-07,
+      "loss": 0.0153,
+      "step": 17955
+    },
+    {
+      "epoch": 4.080909090909091,
+      "grad_norm": 0.74538241404769,
+      "learning_rate": 1.0139872626762824e-07,
+      "loss": 0.003,
+      "step": 17956
+    },
+    {
+      "epoch": 4.081136363636364,
+      "grad_norm": 0.731735151683017,
+      "learning_rate": 1.0134999805191218e-07,
+      "loss": 0.0022,
+      "step": 17957
+    },
+    {
+      "epoch": 4.081363636363636,
+      "grad_norm": 1.1272829676013705,
+      "learning_rate": 1.0130128051433333e-07,
+      "loss": 0.0128,
+      "step": 17958
+    },
+    {
+      "epoch": 4.0815909090909095,
+      "grad_norm": 2.9575306119107414,
+      "learning_rate": 1.0125257365588482e-07,
+      "loss": 0.0068,
+      "step": 17959
+    },
+    {
+      "epoch": 4.081818181818182,
+      "grad_norm": 1.315876374098006,
+      "learning_rate": 1.0120387747756013e-07,
+      "loss": 0.0035,
+      "step": 17960
+    },
+    {
+      "epoch": 4.0820454545454545,
+      "grad_norm": 1.407418087161961,
+      "learning_rate": 1.0115519198035217e-07,
+      "loss": 0.0041,
+      "step": 17961
+    },
+    {
+      "epoch": 4.082272727272727,
+      "grad_norm": 2.3295203815689596,
+      "learning_rate": 1.0110651716525393e-07,
+      "loss": 0.0121,
+      "step": 17962
+    },
+    {
+      "epoch": 4.0825,
+      "grad_norm": 0.2627319093316562,
+      "learning_rate": 1.010578530332576e-07,
+      "loss": 0.0011,
+      "step": 17963
+    },
+    {
+      "epoch": 4.082727272727273,
+      "grad_norm": 0.4806183275246989,
+      "learning_rate": 1.010091995853557e-07,
+      "loss": 0.0051,
+      "step": 17964
+    },
+    {
+      "epoch": 4.0829545454545455,
+      "grad_norm": 1.6492942249500784,
+      "learning_rate": 1.0096055682254038e-07,
+      "loss": 0.0116,
+      "step": 17965
+    },
+    {
+      "epoch": 4.083181818181818,
+      "grad_norm": 0.7803004701106848,
+      "learning_rate": 1.0091192474580354e-07,
+      "loss": 0.0021,
+      "step": 17966
+    },
+    {
+      "epoch": 4.0834090909090905,
+      "grad_norm": 1.7803009368789178,
+      "learning_rate": 1.0086330335613688e-07,
+      "loss": 0.01,
+      "step": 17967
+    },
+    {
+      "epoch": 4.083636363636364,
+      "grad_norm": 2.2885322937398986,
+      "learning_rate": 1.0081469265453191e-07,
+      "loss": 0.0099,
+      "step": 17968
+    },
+    {
+      "epoch": 4.0838636363636365,
+      "grad_norm": 0.6284488009094223,
+      "learning_rate": 1.0076609264197984e-07,
+      "loss": 0.0032,
+      "step": 17969
+    },
+    {
+      "epoch": 4.084090909090909,
+      "grad_norm": 0.29334879369105377,
+      "learning_rate": 1.007175033194717e-07,
+      "loss": 0.001,
+      "step": 17970
+    },
+    {
+      "epoch": 4.0843181818181815,
+      "grad_norm": 0.8065473636023333,
+      "learning_rate": 1.0066892468799841e-07,
+      "loss": 0.0022,
+      "step": 17971
+    },
+    {
+      "epoch": 4.084545454545455,
+      "grad_norm": 1.2313861168937648,
+      "learning_rate": 1.0062035674855044e-07,
+      "loss": 0.0058,
+      "step": 17972
+    },
+    {
+      "epoch": 4.084772727272727,
+      "grad_norm": 0.30666128750751764,
+      "learning_rate": 1.0057179950211819e-07,
+      "loss": 0.0016,
+      "step": 17973
+    },
+    {
+      "epoch": 4.085,
+      "grad_norm": 0.7669736997381614,
+      "learning_rate": 1.0052325294969183e-07,
+      "loss": 0.0031,
+      "step": 17974
+    },
+    {
+      "epoch": 4.0852272727272725,
+      "grad_norm": 0.6775848144101875,
+      "learning_rate": 1.0047471709226153e-07,
+      "loss": 0.0018,
+      "step": 17975
+    },
+    {
+      "epoch": 4.085454545454546,
+      "grad_norm": 0.5702543643693659,
+      "learning_rate": 1.0042619193081672e-07,
+      "loss": 0.0027,
+      "step": 17976
+    },
+    {
+      "epoch": 4.085681818181818,
+      "grad_norm": 0.4033292331871727,
+      "learning_rate": 1.0037767746634695e-07,
+      "loss": 0.0018,
+      "step": 17977
+    },
+    {
+      "epoch": 4.085909090909091,
+      "grad_norm": 1.3125222745681824,
+      "learning_rate": 1.0032917369984164e-07,
+      "loss": 0.0102,
+      "step": 17978
+    },
+    {
+      "epoch": 4.0861363636363635,
+      "grad_norm": 1.731669002516954,
+      "learning_rate": 1.0028068063228979e-07,
+      "loss": 0.0074,
+      "step": 17979
+    },
+    {
+      "epoch": 4.086363636363636,
+      "grad_norm": 0.6429250532484879,
+      "learning_rate": 1.0023219826468047e-07,
+      "loss": 0.0029,
+      "step": 17980
+    },
+    {
+      "epoch": 4.086590909090909,
+      "grad_norm": 1.5875479102006245,
+      "learning_rate": 1.0018372659800192e-07,
+      "loss": 0.0075,
+      "step": 17981
+    },
+    {
+      "epoch": 4.086818181818182,
+      "grad_norm": 0.5591615467163962,
+      "learning_rate": 1.0013526563324286e-07,
+      "loss": 0.0017,
+      "step": 17982
+    },
+    {
+      "epoch": 4.087045454545454,
+      "grad_norm": 4.768961932770657,
+      "learning_rate": 1.0008681537139138e-07,
+      "loss": 0.0066,
+      "step": 17983
+    },
+    {
+      "epoch": 4.087272727272727,
+      "grad_norm": 0.6857541572323747,
+      "learning_rate": 1.0003837581343545e-07,
+      "loss": 0.0051,
+      "step": 17984
+    },
+    {
+      "epoch": 4.0875,
+      "grad_norm": 0.35442471366857725,
+      "learning_rate": 9.998994696036296e-08,
+      "loss": 0.0008,
+      "step": 17985
+    },
+    {
+      "epoch": 4.087727272727273,
+      "grad_norm": 0.7948810189718803,
+      "learning_rate": 9.994152881316131e-08,
+      "loss": 0.0045,
+      "step": 17986
+    },
+    {
+      "epoch": 4.087954545454545,
+      "grad_norm": 0.8440168004031237,
+      "learning_rate": 9.989312137281791e-08,
+      "loss": 0.0032,
+      "step": 17987
+    },
+    {
+      "epoch": 4.088181818181818,
+      "grad_norm": 2.077449186336835,
+      "learning_rate": 9.984472464031989e-08,
+      "loss": 0.0258,
+      "step": 17988
+    },
+    {
+      "epoch": 4.088409090909091,
+      "grad_norm": 0.9577670781077505,
+      "learning_rate": 9.979633861665429e-08,
+      "loss": 0.0068,
+      "step": 17989
+    },
+    {
+      "epoch": 4.088636363636364,
+      "grad_norm": 1.7337393874570721,
+      "learning_rate": 9.974796330280742e-08,
+      "loss": 0.0086,
+      "step": 17990
+    },
+    {
+      "epoch": 4.088863636363636,
+      "grad_norm": 1.867305100966593,
+      "learning_rate": 9.9699598699766e-08,
+      "loss": 0.0122,
+      "step": 17991
+    },
+    {
+      "epoch": 4.089090909090909,
+      "grad_norm": 1.5101470509272876,
+      "learning_rate": 9.965124480851608e-08,
+      "loss": 0.0033,
+      "step": 17992
+    },
+    {
+      "epoch": 4.089318181818181,
+      "grad_norm": 1.441731761959355,
+      "learning_rate": 9.960290163004406e-08,
+      "loss": 0.0051,
+      "step": 17993
+    },
+    {
+      "epoch": 4.089545454545455,
+      "grad_norm": 1.4501443800619231,
+      "learning_rate": 9.955456916533525e-08,
+      "loss": 0.009,
+      "step": 17994
+    },
+    {
+      "epoch": 4.089772727272727,
+      "grad_norm": 1.016264434980308,
+      "learning_rate": 9.950624741537552e-08,
+      "loss": 0.0074,
+      "step": 17995
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 1.6991100714810061,
+      "learning_rate": 9.945793638115014e-08,
+      "loss": 0.0039,
+      "step": 17996
+    },
+    {
+      "epoch": 4.090227272727272,
+      "grad_norm": 0.9848243860859842,
+      "learning_rate": 9.940963606364435e-08,
+      "loss": 0.0037,
+      "step": 17997
+    },
+    {
+      "epoch": 4.090454545454546,
+      "grad_norm": 2.137804558282999,
+      "learning_rate": 9.936134646384301e-08,
+      "loss": 0.0096,
+      "step": 17998
+    },
+    {
+      "epoch": 4.090681818181818,
+      "grad_norm": 0.7452832774633853,
+      "learning_rate": 9.931306758273078e-08,
+      "loss": 0.0012,
+      "step": 17999
+    },
+    {
+      "epoch": 4.090909090909091,
+      "grad_norm": 1.1796761904337933,
+      "learning_rate": 9.926479942129232e-08,
+      "loss": 0.0087,
+      "step": 18000
+    },
+    {
+      "epoch": 4.091136363636363,
+      "grad_norm": 0.47791491524668145,
+      "learning_rate": 9.921654198051169e-08,
+      "loss": 0.0021,
+      "step": 18001
+    },
+    {
+      "epoch": 4.091363636363637,
+      "grad_norm": 0.20989110853623497,
+      "learning_rate": 9.916829526137323e-08,
+      "loss": 0.0007,
+      "step": 18002
+    },
+    {
+      "epoch": 4.091590909090909,
+      "grad_norm": 0.45446635465620927,
+      "learning_rate": 9.91200592648605e-08,
+      "loss": 0.0014,
+      "step": 18003
+    },
+    {
+      "epoch": 4.091818181818182,
+      "grad_norm": 2.821653581921011,
+      "learning_rate": 9.907183399195713e-08,
+      "loss": 0.0285,
+      "step": 18004
+    },
+    {
+      "epoch": 4.092045454545454,
+      "grad_norm": 1.6882320491540161,
+      "learning_rate": 9.902361944364662e-08,
+      "loss": 0.0085,
+      "step": 18005
+    },
+    {
+      "epoch": 4.092272727272727,
+      "grad_norm": 0.6259443708906103,
+      "learning_rate": 9.897541562091231e-08,
+      "loss": 0.0035,
+      "step": 18006
+    },
+    {
+      "epoch": 4.0925,
+      "grad_norm": 0.654924205734125,
+      "learning_rate": 9.892722252473681e-08,
+      "loss": 0.0065,
+      "step": 18007
+    },
+    {
+      "epoch": 4.092727272727273,
+      "grad_norm": 1.9912764301915382,
+      "learning_rate": 9.887904015610302e-08,
+      "loss": 0.0059,
+      "step": 18008
+    },
+    {
+      "epoch": 4.092954545454545,
+      "grad_norm": 0.5437519827560849,
+      "learning_rate": 9.883086851599347e-08,
+      "loss": 0.0018,
+      "step": 18009
+    },
+    {
+      "epoch": 4.093181818181818,
+      "grad_norm": 0.8642535928977502,
+      "learning_rate": 9.878270760539058e-08,
+      "loss": 0.0041,
+      "step": 18010
+    },
+    {
+      "epoch": 4.093409090909091,
+      "grad_norm": 1.41768909490534,
+      "learning_rate": 9.87345574252762e-08,
+      "loss": 0.0108,
+      "step": 18011
+    },
+    {
+      "epoch": 4.093636363636364,
+      "grad_norm": 1.0224456840291412,
+      "learning_rate": 9.868641797663233e-08,
+      "loss": 0.0055,
+      "step": 18012
+    },
+    {
+      "epoch": 4.093863636363636,
+      "grad_norm": 2.4765386272698606,
+      "learning_rate": 9.86382892604406e-08,
+      "loss": 0.007,
+      "step": 18013
+    },
+    {
+      "epoch": 4.094090909090909,
+      "grad_norm": 0.4645597215807599,
+      "learning_rate": 9.859017127768246e-08,
+      "loss": 0.0014,
+      "step": 18014
+    },
+    {
+      "epoch": 4.094318181818182,
+      "grad_norm": 1.0648411113982308,
+      "learning_rate": 9.854206402933907e-08,
+      "loss": 0.0114,
+      "step": 18015
+    },
+    {
+      "epoch": 4.094545454545455,
+      "grad_norm": 2.074055218930864,
+      "learning_rate": 9.849396751639143e-08,
+      "loss": 0.0147,
+      "step": 18016
+    },
+    {
+      "epoch": 4.094772727272727,
+      "grad_norm": 0.3036885230075695,
+      "learning_rate": 9.844588173982032e-08,
+      "loss": 0.0011,
+      "step": 18017
+    },
+    {
+      "epoch": 4.095,
+      "grad_norm": 1.4559215345926464,
+      "learning_rate": 9.839780670060633e-08,
+      "loss": 0.0013,
+      "step": 18018
+    },
+    {
+      "epoch": 4.095227272727273,
+      "grad_norm": 1.147288292573438,
+      "learning_rate": 9.834974239972997e-08,
+      "loss": 0.0036,
+      "step": 18019
+    },
+    {
+      "epoch": 4.095454545454546,
+      "grad_norm": 0.9584689413795143,
+      "learning_rate": 9.83016888381709e-08,
+      "loss": 0.0078,
+      "step": 18020
+    },
+    {
+      "epoch": 4.095681818181818,
+      "grad_norm": 1.395063388605479,
+      "learning_rate": 9.825364601690937e-08,
+      "loss": 0.0058,
+      "step": 18021
+    },
+    {
+      "epoch": 4.095909090909091,
+      "grad_norm": 0.261895335240738,
+      "learning_rate": 9.820561393692487e-08,
+      "loss": 0.0011,
+      "step": 18022
+    },
+    {
+      "epoch": 4.096136363636363,
+      "grad_norm": 0.7437123632478521,
+      "learning_rate": 9.815759259919705e-08,
+      "loss": 0.004,
+      "step": 18023
+    },
+    {
+      "epoch": 4.096363636363637,
+      "grad_norm": 0.539931531385608,
+      "learning_rate": 9.810958200470511e-08,
+      "loss": 0.0022,
+      "step": 18024
+    },
+    {
+      "epoch": 4.096590909090909,
+      "grad_norm": 0.7405439022388539,
+      "learning_rate": 9.806158215442791e-08,
+      "loss": 0.0022,
+      "step": 18025
+    },
+    {
+      "epoch": 4.096818181818182,
+      "grad_norm": 0.28927858538026013,
+      "learning_rate": 9.801359304934429e-08,
+      "loss": 0.0008,
+      "step": 18026
+    },
+    {
+      "epoch": 4.097045454545454,
+      "grad_norm": 0.9090060897690597,
+      "learning_rate": 9.796561469043293e-08,
+      "loss": 0.0022,
+      "step": 18027
+    },
+    {
+      "epoch": 4.097272727272728,
+      "grad_norm": 0.9495751518714272,
+      "learning_rate": 9.791764707867211e-08,
+      "loss": 0.0045,
+      "step": 18028
+    },
+    {
+      "epoch": 4.0975,
+      "grad_norm": 2.1639057911098334,
+      "learning_rate": 9.786969021504e-08,
+      "loss": 0.026,
+      "step": 18029
+    },
+    {
+      "epoch": 4.097727272727273,
+      "grad_norm": 1.4460076837626366,
+      "learning_rate": 9.782174410051457e-08,
+      "loss": 0.0091,
+      "step": 18030
+    },
+    {
+      "epoch": 4.097954545454545,
+      "grad_norm": 0.9452642158879645,
+      "learning_rate": 9.777380873607349e-08,
+      "loss": 0.0061,
+      "step": 18031
+    },
+    {
+      "epoch": 4.098181818181819,
+      "grad_norm": 0.999181977287375,
+      "learning_rate": 9.77258841226944e-08,
+      "loss": 0.0049,
+      "step": 18032
+    },
+    {
+      "epoch": 4.098409090909091,
+      "grad_norm": 1.0802815311082983,
+      "learning_rate": 9.76779702613543e-08,
+      "loss": 0.0033,
+      "step": 18033
+    },
+    {
+      "epoch": 4.098636363636364,
+      "grad_norm": 1.9302470781203045,
+      "learning_rate": 9.76300671530303e-08,
+      "loss": 0.009,
+      "step": 18034
+    },
+    {
+      "epoch": 4.098863636363636,
+      "grad_norm": 0.8170866376234004,
+      "learning_rate": 9.758217479869929e-08,
+      "loss": 0.0093,
+      "step": 18035
+    },
+    {
+      "epoch": 4.099090909090909,
+      "grad_norm": 0.23205788129449365,
+      "learning_rate": 9.753429319933783e-08,
+      "loss": 0.0011,
+      "step": 18036
+    },
+    {
+      "epoch": 4.099318181818182,
+      "grad_norm": 1.4063991460790524,
+      "learning_rate": 9.74864223559225e-08,
+      "loss": 0.0034,
+      "step": 18037
+    },
+    {
+      "epoch": 4.099545454545455,
+      "grad_norm": 1.0618627271860297,
+      "learning_rate": 9.743856226942919e-08,
+      "loss": 0.0107,
+      "step": 18038
+    },
+    {
+      "epoch": 4.099772727272727,
+      "grad_norm": 0.61184788583284,
+      "learning_rate": 9.739071294083399e-08,
+      "loss": 0.002,
+      "step": 18039
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 4.061933766443455,
+      "learning_rate": 9.734287437111266e-08,
+      "loss": 0.0234,
+      "step": 18040
+    },
+    {
+      "epoch": 4.100227272727273,
+      "grad_norm": 0.7456211788109854,
+      "learning_rate": 9.729504656124073e-08,
+      "loss": 0.0049,
+      "step": 18041
+    },
+    {
+      "epoch": 4.100454545454546,
+      "grad_norm": 0.6889717720053684,
+      "learning_rate": 9.724722951219331e-08,
+      "loss": 0.0039,
+      "step": 18042
+    },
+    {
+      "epoch": 4.100681818181818,
+      "grad_norm": 0.6375962381573081,
+      "learning_rate": 9.719942322494562e-08,
+      "loss": 0.002,
+      "step": 18043
+    },
+    {
+      "epoch": 4.100909090909091,
+      "grad_norm": 0.9615945837460185,
+      "learning_rate": 9.715162770047235e-08,
+      "loss": 0.0035,
+      "step": 18044
+    },
+    {
+      "epoch": 4.101136363636364,
+      "grad_norm": 0.7226280287024265,
+      "learning_rate": 9.710384293974841e-08,
+      "loss": 0.0015,
+      "step": 18045
+    },
+    {
+      "epoch": 4.1013636363636365,
+      "grad_norm": 2.6017272729866723,
+      "learning_rate": 9.705606894374816e-08,
+      "loss": 0.0073,
+      "step": 18046
+    },
+    {
+      "epoch": 4.101590909090909,
+      "grad_norm": 1.5175058379978552,
+      "learning_rate": 9.700830571344563e-08,
+      "loss": 0.0146,
+      "step": 18047
+    },
+    {
+      "epoch": 4.101818181818182,
+      "grad_norm": 1.4596941559169925,
+      "learning_rate": 9.69605532498149e-08,
+      "loss": 0.0066,
+      "step": 18048
+    },
+    {
+      "epoch": 4.102045454545454,
+      "grad_norm": 1.024475966602463,
+      "learning_rate": 9.691281155382968e-08,
+      "loss": 0.0043,
+      "step": 18049
+    },
+    {
+      "epoch": 4.1022727272727275,
+      "grad_norm": 0.8456273182563461,
+      "learning_rate": 9.686508062646365e-08,
+      "loss": 0.0055,
+      "step": 18050
+    },
+    {
+      "epoch": 4.1025,
+      "grad_norm": 1.6990249234202408,
+      "learning_rate": 9.681736046868985e-08,
+      "loss": 0.0078,
+      "step": 18051
+    },
+    {
+      "epoch": 4.102727272727273,
+      "grad_norm": 0.9317147924889776,
+      "learning_rate": 9.676965108148161e-08,
+      "loss": 0.0049,
+      "step": 18052
+    },
+    {
+      "epoch": 4.102954545454545,
+      "grad_norm": 0.753895963009531,
+      "learning_rate": 9.672195246581167e-08,
+      "loss": 0.0021,
+      "step": 18053
+    },
+    {
+      "epoch": 4.1031818181818185,
+      "grad_norm": 0.4760788392846025,
+      "learning_rate": 9.667426462265289e-08,
+      "loss": 0.002,
+      "step": 18054
+    },
+    {
+      "epoch": 4.103409090909091,
+      "grad_norm": 0.8650443552414352,
+      "learning_rate": 9.662658755297746e-08,
+      "loss": 0.0033,
+      "step": 18055
+    },
+    {
+      "epoch": 4.1036363636363635,
+      "grad_norm": 0.2510199406870211,
+      "learning_rate": 9.657892125775774e-08,
+      "loss": 0.0018,
+      "step": 18056
+    },
+    {
+      "epoch": 4.103863636363636,
+      "grad_norm": 1.173831159858513,
+      "learning_rate": 9.653126573796565e-08,
+      "loss": 0.0035,
+      "step": 18057
+    },
+    {
+      "epoch": 4.1040909090909095,
+      "grad_norm": 1.3838103681209495,
+      "learning_rate": 9.648362099457297e-08,
+      "loss": 0.0035,
+      "step": 18058
+    },
+    {
+      "epoch": 4.104318181818182,
+      "grad_norm": 0.7727753641800016,
+      "learning_rate": 9.643598702855139e-08,
+      "loss": 0.0052,
+      "step": 18059
+    },
+    {
+      "epoch": 4.1045454545454545,
+      "grad_norm": 1.9598428542744164,
+      "learning_rate": 9.638836384087212e-08,
+      "loss": 0.0089,
+      "step": 18060
+    },
+    {
+      "epoch": 4.104772727272727,
+      "grad_norm": 0.26146177176094,
+      "learning_rate": 9.634075143250634e-08,
+      "loss": 0.0006,
+      "step": 18061
+    },
+    {
+      "epoch": 4.105,
+      "grad_norm": 1.6031073798926476,
+      "learning_rate": 9.629314980442491e-08,
+      "loss": 0.0127,
+      "step": 18062
+    },
+    {
+      "epoch": 4.105227272727273,
+      "grad_norm": 1.928717439227056,
+      "learning_rate": 9.62455589575987e-08,
+      "loss": 0.0081,
+      "step": 18063
+    },
+    {
+      "epoch": 4.1054545454545455,
+      "grad_norm": 1.6332409503407785,
+      "learning_rate": 9.619797889299786e-08,
+      "loss": 0.0053,
+      "step": 18064
+    },
+    {
+      "epoch": 4.105681818181818,
+      "grad_norm": 1.8679121904753997,
+      "learning_rate": 9.615040961159277e-08,
+      "loss": 0.0082,
+      "step": 18065
+    },
+    {
+      "epoch": 4.1059090909090905,
+      "grad_norm": 0.23983356268140985,
+      "learning_rate": 9.610285111435351e-08,
+      "loss": 0.0026,
+      "step": 18066
+    },
+    {
+      "epoch": 4.106136363636364,
+      "grad_norm": 0.5588694741909283,
+      "learning_rate": 9.605530340224975e-08,
+      "loss": 0.0009,
+      "step": 18067
+    },
+    {
+      "epoch": 4.106363636363636,
+      "grad_norm": 1.5692728257082402,
+      "learning_rate": 9.600776647625134e-08,
+      "loss": 0.0156,
+      "step": 18068
+    },
+    {
+      "epoch": 4.106590909090909,
+      "grad_norm": 3.219003422157341,
+      "learning_rate": 9.596024033732728e-08,
+      "loss": 0.0127,
+      "step": 18069
+    },
+    {
+      "epoch": 4.1068181818181815,
+      "grad_norm": 0.8801732017196733,
+      "learning_rate": 9.591272498644687e-08,
+      "loss": 0.0063,
+      "step": 18070
+    },
+    {
+      "epoch": 4.107045454545455,
+      "grad_norm": 1.042455556036565,
+      "learning_rate": 9.586522042457906e-08,
+      "loss": 0.0039,
+      "step": 18071
+    },
+    {
+      "epoch": 4.107272727272727,
+      "grad_norm": 0.5180939036413501,
+      "learning_rate": 9.581772665269266e-08,
+      "loss": 0.0021,
+      "step": 18072
+    },
+    {
+      "epoch": 4.1075,
+      "grad_norm": 1.4246514519832338,
+      "learning_rate": 9.577024367175593e-08,
+      "loss": 0.0187,
+      "step": 18073
+    },
+    {
+      "epoch": 4.1077272727272724,
+      "grad_norm": 0.48756420451952803,
+      "learning_rate": 9.572277148273704e-08,
+      "loss": 0.0025,
+      "step": 18074
+    },
+    {
+      "epoch": 4.107954545454546,
+      "grad_norm": 1.8158989902336027,
+      "learning_rate": 9.567531008660438e-08,
+      "loss": 0.0097,
+      "step": 18075
+    },
+    {
+      "epoch": 4.108181818181818,
+      "grad_norm": 1.2000563678279186,
+      "learning_rate": 9.562785948432574e-08,
+      "loss": 0.0046,
+      "step": 18076
+    },
+    {
+      "epoch": 4.108409090909091,
+      "grad_norm": 0.34601463997434845,
+      "learning_rate": 9.558041967686838e-08,
+      "loss": 0.0012,
+      "step": 18077
+    },
+    {
+      "epoch": 4.108636363636363,
+      "grad_norm": 1.0527451894517428,
+      "learning_rate": 9.553299066519992e-08,
+      "loss": 0.0077,
+      "step": 18078
+    },
+    {
+      "epoch": 4.108863636363636,
+      "grad_norm": 1.3861836041941236,
+      "learning_rate": 9.548557245028754e-08,
+      "loss": 0.0088,
+      "step": 18079
+    },
+    {
+      "epoch": 4.109090909090909,
+      "grad_norm": 1.11610995415511,
+      "learning_rate": 9.543816503309808e-08,
+      "loss": 0.0041,
+      "step": 18080
+    },
+    {
+      "epoch": 4.109318181818182,
+      "grad_norm": 0.9653640169316297,
+      "learning_rate": 9.539076841459848e-08,
+      "loss": 0.0064,
+      "step": 18081
+    },
+    {
+      "epoch": 4.109545454545454,
+      "grad_norm": 1.323382061613617,
+      "learning_rate": 9.53433825957549e-08,
+      "loss": 0.006,
+      "step": 18082
+    },
+    {
+      "epoch": 4.109772727272727,
+      "grad_norm": 0.6169517744768069,
+      "learning_rate": 9.52960075775338e-08,
+      "loss": 0.0031,
+      "step": 18083
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 1.128161215506532,
+      "learning_rate": 9.524864336090123e-08,
+      "loss": 0.0062,
+      "step": 18084
+    },
+    {
+      "epoch": 4.110227272727273,
+      "grad_norm": 0.2641326447058374,
+      "learning_rate": 9.520128994682313e-08,
+      "loss": 0.0008,
+      "step": 18085
+    },
+    {
+      "epoch": 4.110454545454545,
+      "grad_norm": 2.1198401660412496,
+      "learning_rate": 9.51539473362649e-08,
+      "loss": 0.0126,
+      "step": 18086
+    },
+    {
+      "epoch": 4.110681818181818,
+      "grad_norm": 1.2280215747962822,
+      "learning_rate": 9.510661553019204e-08,
+      "loss": 0.0079,
+      "step": 18087
+    },
+    {
+      "epoch": 4.110909090909091,
+      "grad_norm": 1.7884299282971166,
+      "learning_rate": 9.505929452956975e-08,
+      "loss": 0.009,
+      "step": 18088
+    },
+    {
+      "epoch": 4.111136363636364,
+      "grad_norm": 0.34599162830132707,
+      "learning_rate": 9.501198433536293e-08,
+      "loss": 0.0014,
+      "step": 18089
+    },
+    {
+      "epoch": 4.111363636363636,
+      "grad_norm": 0.9558074144123128,
+      "learning_rate": 9.49646849485364e-08,
+      "loss": 0.0057,
+      "step": 18090
+    },
+    {
+      "epoch": 4.111590909090909,
+      "grad_norm": 0.6117844562679678,
+      "learning_rate": 9.491739637005462e-08,
+      "loss": 0.001,
+      "step": 18091
+    },
+    {
+      "epoch": 4.111818181818181,
+      "grad_norm": 1.1812160693294231,
+      "learning_rate": 9.487011860088192e-08,
+      "loss": 0.0046,
+      "step": 18092
+    },
+    {
+      "epoch": 4.112045454545455,
+      "grad_norm": 0.8212488005902563,
+      "learning_rate": 9.48228516419824e-08,
+      "loss": 0.0021,
+      "step": 18093
+    },
+    {
+      "epoch": 4.112272727272727,
+      "grad_norm": 0.5239232550874514,
+      "learning_rate": 9.477559549432004e-08,
+      "loss": 0.0023,
+      "step": 18094
+    },
+    {
+      "epoch": 4.1125,
+      "grad_norm": 0.6023524297130473,
+      "learning_rate": 9.472835015885806e-08,
+      "loss": 0.0014,
+      "step": 18095
+    },
+    {
+      "epoch": 4.112727272727272,
+      "grad_norm": 3.6107305057452543,
+      "learning_rate": 9.468111563656026e-08,
+      "loss": 0.0075,
+      "step": 18096
+    },
+    {
+      "epoch": 4.112954545454546,
+      "grad_norm": 0.3707095981861944,
+      "learning_rate": 9.463389192838963e-08,
+      "loss": 0.0016,
+      "step": 18097
+    },
+    {
+      "epoch": 4.113181818181818,
+      "grad_norm": 1.4229551207165452,
+      "learning_rate": 9.458667903530935e-08,
+      "loss": 0.0152,
+      "step": 18098
+    },
+    {
+      "epoch": 4.113409090909091,
+      "grad_norm": 1.5372210641424142,
+      "learning_rate": 9.453947695828191e-08,
+      "loss": 0.0031,
+      "step": 18099
+    },
+    {
+      "epoch": 4.113636363636363,
+      "grad_norm": 1.280759423490898,
+      "learning_rate": 9.449228569827003e-08,
+      "loss": 0.0033,
+      "step": 18100
+    },
+    {
+      "epoch": 4.113863636363637,
+      "grad_norm": 0.3221688508460509,
+      "learning_rate": 9.44451052562359e-08,
+      "loss": 0.0011,
+      "step": 18101
+    },
+    {
+      "epoch": 4.114090909090909,
+      "grad_norm": 3.7094184370528436,
+      "learning_rate": 9.439793563314168e-08,
+      "loss": 0.0444,
+      "step": 18102
+    },
+    {
+      "epoch": 4.114318181818182,
+      "grad_norm": 0.6231383418268767,
+      "learning_rate": 9.435077682994925e-08,
+      "loss": 0.002,
+      "step": 18103
+    },
+    {
+      "epoch": 4.114545454545454,
+      "grad_norm": 1.367791667072347,
+      "learning_rate": 9.430362884762025e-08,
+      "loss": 0.008,
+      "step": 18104
+    },
+    {
+      "epoch": 4.114772727272728,
+      "grad_norm": 0.7674767491325493,
+      "learning_rate": 9.425649168711616e-08,
+      "loss": 0.004,
+      "step": 18105
+    },
+    {
+      "epoch": 4.115,
+      "grad_norm": 0.665956228235945,
+      "learning_rate": 9.420936534939807e-08,
+      "loss": 0.0033,
+      "step": 18106
+    },
+    {
+      "epoch": 4.115227272727273,
+      "grad_norm": 0.9045760459072847,
+      "learning_rate": 9.416224983542722e-08,
+      "loss": 0.0044,
+      "step": 18107
+    },
+    {
+      "epoch": 4.115454545454545,
+      "grad_norm": 0.3087179940660742,
+      "learning_rate": 9.411514514616402e-08,
+      "loss": 0.001,
+      "step": 18108
+    },
+    {
+      "epoch": 4.115681818181818,
+      "grad_norm": 0.3288913112490595,
+      "learning_rate": 9.406805128256924e-08,
+      "loss": 0.0011,
+      "step": 18109
+    },
+    {
+      "epoch": 4.115909090909091,
+      "grad_norm": 0.26744187943745606,
+      "learning_rate": 9.402096824560309e-08,
+      "loss": 0.0013,
+      "step": 18110
+    },
+    {
+      "epoch": 4.116136363636364,
+      "grad_norm": 0.7089947284560215,
+      "learning_rate": 9.397389603622576e-08,
+      "loss": 0.0026,
+      "step": 18111
+    },
+    {
+      "epoch": 4.116363636363636,
+      "grad_norm": 0.698081886402863,
+      "learning_rate": 9.392683465539725e-08,
+      "loss": 0.0042,
+      "step": 18112
+    },
+    {
+      "epoch": 4.116590909090909,
+      "grad_norm": 1.211069587725626,
+      "learning_rate": 9.3879784104077e-08,
+      "loss": 0.0054,
+      "step": 18113
+    },
+    {
+      "epoch": 4.116818181818182,
+      "grad_norm": 2.0207123196615653,
+      "learning_rate": 9.383274438322452e-08,
+      "loss": 0.0134,
+      "step": 18114
+    },
+    {
+      "epoch": 4.117045454545455,
+      "grad_norm": 1.1567614239264572,
+      "learning_rate": 9.378571549379898e-08,
+      "loss": 0.0075,
+      "step": 18115
+    },
+    {
+      "epoch": 4.117272727272727,
+      "grad_norm": 0.3853895023134169,
+      "learning_rate": 9.373869743675969e-08,
+      "loss": 0.0011,
+      "step": 18116
+    },
+    {
+      "epoch": 4.1175,
+      "grad_norm": 3.363614670553215,
+      "learning_rate": 9.369169021306503e-08,
+      "loss": 0.0319,
+      "step": 18117
+    },
+    {
+      "epoch": 4.117727272727273,
+      "grad_norm": 0.5311601943336157,
+      "learning_rate": 9.364469382367368e-08,
+      "loss": 0.0019,
+      "step": 18118
+    },
+    {
+      "epoch": 4.117954545454546,
+      "grad_norm": 1.1358203451752662,
+      "learning_rate": 9.359770826954404e-08,
+      "loss": 0.0029,
+      "step": 18119
+    },
+    {
+      "epoch": 4.118181818181818,
+      "grad_norm": 0.6584240102383777,
+      "learning_rate": 9.355073355163426e-08,
+      "loss": 0.0036,
+      "step": 18120
+    },
+    {
+      "epoch": 4.118409090909091,
+      "grad_norm": 1.5158480289332998,
+      "learning_rate": 9.350376967090214e-08,
+      "loss": 0.0114,
+      "step": 18121
+    },
+    {
+      "epoch": 4.118636363636363,
+      "grad_norm": 0.848883861175993,
+      "learning_rate": 9.345681662830542e-08,
+      "loss": 0.0051,
+      "step": 18122
+    },
+    {
+      "epoch": 4.118863636363637,
+      "grad_norm": 1.14873599627218,
+      "learning_rate": 9.340987442480159e-08,
+      "loss": 0.0043,
+      "step": 18123
+    },
+    {
+      "epoch": 4.119090909090909,
+      "grad_norm": 0.9124378762728503,
+      "learning_rate": 9.336294306134773e-08,
+      "loss": 0.003,
+      "step": 18124
+    },
+    {
+      "epoch": 4.119318181818182,
+      "grad_norm": 1.1425904691249447,
+      "learning_rate": 9.331602253890113e-08,
+      "loss": 0.0042,
+      "step": 18125
+    },
+    {
+      "epoch": 4.119545454545454,
+      "grad_norm": 0.6513887664933558,
+      "learning_rate": 9.326911285841831e-08,
+      "loss": 0.0023,
+      "step": 18126
+    },
+    {
+      "epoch": 4.119772727272728,
+      "grad_norm": 0.42223281145338903,
+      "learning_rate": 9.322221402085586e-08,
+      "loss": 0.0014,
+      "step": 18127
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 1.3670366835671246,
+      "learning_rate": 9.317532602717027e-08,
+      "loss": 0.0036,
+      "step": 18128
+    },
+    {
+      "epoch": 4.120227272727273,
+      "grad_norm": 0.22025709504127036,
+      "learning_rate": 9.312844887831768e-08,
+      "loss": 0.0006,
+      "step": 18129
+    },
+    {
+      "epoch": 4.120454545454545,
+      "grad_norm": 1.6429646555297408,
+      "learning_rate": 9.308158257525388e-08,
+      "loss": 0.0052,
+      "step": 18130
+    },
+    {
+      "epoch": 4.120681818181819,
+      "grad_norm": 0.5295555154252531,
+      "learning_rate": 9.303472711893448e-08,
+      "loss": 0.0019,
+      "step": 18131
+    },
+    {
+      "epoch": 4.120909090909091,
+      "grad_norm": 0.35748111635333707,
+      "learning_rate": 9.298788251031511e-08,
+      "loss": 0.0014,
+      "step": 18132
+    },
+    {
+      "epoch": 4.121136363636364,
+      "grad_norm": 0.4643441111806272,
+      "learning_rate": 9.294104875035098e-08,
+      "loss": 0.0024,
+      "step": 18133
+    },
+    {
+      "epoch": 4.121363636363636,
+      "grad_norm": 1.2371105241452696,
+      "learning_rate": 9.289422583999702e-08,
+      "loss": 0.0067,
+      "step": 18134
+    },
+    {
+      "epoch": 4.1215909090909095,
+      "grad_norm": 1.2789459933894172,
+      "learning_rate": 9.284741378020816e-08,
+      "loss": 0.0057,
+      "step": 18135
+    },
+    {
+      "epoch": 4.121818181818182,
+      "grad_norm": 0.3880407182614436,
+      "learning_rate": 9.280061257193891e-08,
+      "loss": 0.0019,
+      "step": 18136
+    },
+    {
+      "epoch": 4.122045454545455,
+      "grad_norm": 1.5365960720684027,
+      "learning_rate": 9.275382221614366e-08,
+      "loss": 0.008,
+      "step": 18137
+    },
+    {
+      "epoch": 4.122272727272727,
+      "grad_norm": 0.5655059475976345,
+      "learning_rate": 9.270704271377668e-08,
+      "loss": 0.0023,
+      "step": 18138
+    },
+    {
+      "epoch": 4.1225,
+      "grad_norm": 1.1912952319672863,
+      "learning_rate": 9.266027406579156e-08,
+      "loss": 0.007,
+      "step": 18139
+    },
+    {
+      "epoch": 4.122727272727273,
+      "grad_norm": 1.4411718337579744,
+      "learning_rate": 9.261351627314216e-08,
+      "loss": 0.0115,
+      "step": 18140
+    },
+    {
+      "epoch": 4.1229545454545455,
+      "grad_norm": 2.467068850572729,
+      "learning_rate": 9.256676933678203e-08,
+      "loss": 0.0091,
+      "step": 18141
+    },
+    {
+      "epoch": 4.123181818181818,
+      "grad_norm": 0.2512024216785508,
+      "learning_rate": 9.252003325766444e-08,
+      "loss": 0.0007,
+      "step": 18142
+    },
+    {
+      "epoch": 4.123409090909091,
+      "grad_norm": 1.7098354899481907,
+      "learning_rate": 9.247330803674216e-08,
+      "loss": 0.0069,
+      "step": 18143
+    },
+    {
+      "epoch": 4.123636363636364,
+      "grad_norm": 1.017203789269882,
+      "learning_rate": 9.242659367496825e-08,
+      "loss": 0.0042,
+      "step": 18144
+    },
+    {
+      "epoch": 4.1238636363636365,
+      "grad_norm": 4.5884886790196155,
+      "learning_rate": 9.237989017329515e-08,
+      "loss": 0.0106,
+      "step": 18145
+    },
+    {
+      "epoch": 4.124090909090909,
+      "grad_norm": 0.2033647559973332,
+      "learning_rate": 9.233319753267537e-08,
+      "loss": 0.0005,
+      "step": 18146
+    },
+    {
+      "epoch": 4.1243181818181816,
+      "grad_norm": 0.9579269979627637,
+      "learning_rate": 9.22865157540611e-08,
+      "loss": 0.003,
+      "step": 18147
+    },
+    {
+      "epoch": 4.124545454545455,
+      "grad_norm": 1.276224618586455,
+      "learning_rate": 9.223984483840397e-08,
+      "loss": 0.0025,
+      "step": 18148
+    },
+    {
+      "epoch": 4.1247727272727275,
+      "grad_norm": 0.47101395729489537,
+      "learning_rate": 9.219318478665585e-08,
+      "loss": 0.0034,
+      "step": 18149
+    },
+    {
+      "epoch": 4.125,
+      "grad_norm": 1.2181058587855715,
+      "learning_rate": 9.214653559976826e-08,
+      "loss": 0.0047,
+      "step": 18150
+    },
+    {
+      "epoch": 4.1252272727272725,
+      "grad_norm": 0.4214150108782909,
+      "learning_rate": 9.20998972786924e-08,
+      "loss": 0.0026,
+      "step": 18151
+    },
+    {
+      "epoch": 4.125454545454545,
+      "grad_norm": 0.600773676640115,
+      "learning_rate": 9.205326982437938e-08,
+      "loss": 0.0037,
+      "step": 18152
+    },
+    {
+      "epoch": 4.1256818181818184,
+      "grad_norm": 0.22623573111859965,
+      "learning_rate": 9.200665323777988e-08,
+      "loss": 0.0004,
+      "step": 18153
+    },
+    {
+      "epoch": 4.125909090909091,
+      "grad_norm": 0.2945354028784642,
+      "learning_rate": 9.196004751984462e-08,
+      "loss": 0.0009,
+      "step": 18154
+    },
+    {
+      "epoch": 4.1261363636363635,
+      "grad_norm": 0.39855112904878875,
+      "learning_rate": 9.191345267152409e-08,
+      "loss": 0.0009,
+      "step": 18155
+    },
+    {
+      "epoch": 4.126363636363636,
+      "grad_norm": 1.9074792105021179,
+      "learning_rate": 9.186686869376814e-08,
+      "loss": 0.0138,
+      "step": 18156
+    },
+    {
+      "epoch": 4.126590909090909,
+      "grad_norm": 1.3891582000854,
+      "learning_rate": 9.182029558752678e-08,
+      "loss": 0.0055,
+      "step": 18157
+    },
+    {
+      "epoch": 4.126818181818182,
+      "grad_norm": 0.6268436599894208,
+      "learning_rate": 9.177373335374981e-08,
+      "loss": 0.0021,
+      "step": 18158
+    },
+    {
+      "epoch": 4.1270454545454545,
+      "grad_norm": 2.343584289226524,
+      "learning_rate": 9.172718199338668e-08,
+      "loss": 0.0119,
+      "step": 18159
+    },
+    {
+      "epoch": 4.127272727272727,
+      "grad_norm": 0.6491012017235916,
+      "learning_rate": 9.168064150738682e-08,
+      "loss": 0.0031,
+      "step": 18160
+    },
+    {
+      "epoch": 4.1275,
+      "grad_norm": 0.4833472952486554,
+      "learning_rate": 9.163411189669895e-08,
+      "loss": 0.0018,
+      "step": 18161
+    },
+    {
+      "epoch": 4.127727272727273,
+      "grad_norm": 0.7415345900099883,
+      "learning_rate": 9.158759316227208e-08,
+      "loss": 0.0034,
+      "step": 18162
+    },
+    {
+      "epoch": 4.127954545454545,
+      "grad_norm": 1.4301599205218674,
+      "learning_rate": 9.154108530505475e-08,
+      "loss": 0.0046,
+      "step": 18163
+    },
+    {
+      "epoch": 4.128181818181818,
+      "grad_norm": 0.5836475868582874,
+      "learning_rate": 9.149458832599531e-08,
+      "loss": 0.0016,
+      "step": 18164
+    },
+    {
+      "epoch": 4.128409090909091,
+      "grad_norm": 0.6455299867498668,
+      "learning_rate": 9.144810222604203e-08,
+      "loss": 0.0028,
+      "step": 18165
+    },
+    {
+      "epoch": 4.128636363636364,
+      "grad_norm": 0.3931720865090594,
+      "learning_rate": 9.140162700614276e-08,
+      "loss": 0.0016,
+      "step": 18166
+    },
+    {
+      "epoch": 4.128863636363636,
+      "grad_norm": 0.3771196105598037,
+      "learning_rate": 9.135516266724522e-08,
+      "loss": 0.0016,
+      "step": 18167
+    },
+    {
+      "epoch": 4.129090909090909,
+      "grad_norm": 1.4718082707618643,
+      "learning_rate": 9.13087092102969e-08,
+      "loss": 0.0115,
+      "step": 18168
+    },
+    {
+      "epoch": 4.129318181818181,
+      "grad_norm": 0.5135073222280502,
+      "learning_rate": 9.126226663624518e-08,
+      "loss": 0.0011,
+      "step": 18169
+    },
+    {
+      "epoch": 4.129545454545455,
+      "grad_norm": 0.35587903693018696,
+      "learning_rate": 9.121583494603695e-08,
+      "loss": 0.0012,
+      "step": 18170
+    },
+    {
+      "epoch": 4.129772727272727,
+      "grad_norm": 0.7376540136560198,
+      "learning_rate": 9.1169414140619e-08,
+      "loss": 0.0076,
+      "step": 18171
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 0.1778348322839295,
+      "learning_rate": 9.112300422093809e-08,
+      "loss": 0.0004,
+      "step": 18172
+    },
+    {
+      "epoch": 4.130227272727272,
+      "grad_norm": 0.7323991062154235,
+      "learning_rate": 9.107660518794061e-08,
+      "loss": 0.0016,
+      "step": 18173
+    },
+    {
+      "epoch": 4.130454545454546,
+      "grad_norm": 1.0950754976686068,
+      "learning_rate": 9.103021704257253e-08,
+      "loss": 0.0091,
+      "step": 18174
+    },
+    {
+      "epoch": 4.130681818181818,
+      "grad_norm": 0.7607651026881305,
+      "learning_rate": 9.098383978577985e-08,
+      "loss": 0.0036,
+      "step": 18175
+    },
+    {
+      "epoch": 4.130909090909091,
+      "grad_norm": 0.437339338068641,
+      "learning_rate": 9.093747341850832e-08,
+      "loss": 0.0018,
+      "step": 18176
+    },
+    {
+      "epoch": 4.131136363636363,
+      "grad_norm": 0.3965022379267742,
+      "learning_rate": 9.089111794170359e-08,
+      "loss": 0.002,
+      "step": 18177
+    },
+    {
+      "epoch": 4.131363636363637,
+      "grad_norm": 1.898572379694259,
+      "learning_rate": 9.084477335631066e-08,
+      "loss": 0.0094,
+      "step": 18178
+    },
+    {
+      "epoch": 4.131590909090909,
+      "grad_norm": 0.8957195154040862,
+      "learning_rate": 9.079843966327461e-08,
+      "loss": 0.0056,
+      "step": 18179
+    },
+    {
+      "epoch": 4.131818181818182,
+      "grad_norm": 1.9191153804201135,
+      "learning_rate": 9.075211686354036e-08,
+      "loss": 0.0105,
+      "step": 18180
+    },
+    {
+      "epoch": 4.132045454545454,
+      "grad_norm": 0.9360392466606017,
+      "learning_rate": 9.070580495805253e-08,
+      "loss": 0.0022,
+      "step": 18181
+    },
+    {
+      "epoch": 4.132272727272727,
+      "grad_norm": 0.4656523294739819,
+      "learning_rate": 9.065950394775542e-08,
+      "loss": 0.0015,
+      "step": 18182
+    },
+    {
+      "epoch": 4.1325,
+      "grad_norm": 0.5285052509555106,
+      "learning_rate": 9.061321383359328e-08,
+      "loss": 0.0035,
+      "step": 18183
+    },
+    {
+      "epoch": 4.132727272727273,
+      "grad_norm": 0.7828607504668219,
+      "learning_rate": 9.056693461651001e-08,
+      "loss": 0.0026,
+      "step": 18184
+    },
+    {
+      "epoch": 4.132954545454545,
+      "grad_norm": 2.646036887414386,
+      "learning_rate": 9.052066629744928e-08,
+      "loss": 0.0632,
+      "step": 18185
+    },
+    {
+      "epoch": 4.133181818181818,
+      "grad_norm": 0.4717389090396743,
+      "learning_rate": 9.047440887735473e-08,
+      "loss": 0.0008,
+      "step": 18186
+    },
+    {
+      "epoch": 4.133409090909091,
+      "grad_norm": 0.3306670712594922,
+      "learning_rate": 9.042816235716941e-08,
+      "loss": 0.0018,
+      "step": 18187
+    },
+    {
+      "epoch": 4.133636363636364,
+      "grad_norm": 0.18626025551352882,
+      "learning_rate": 9.038192673783648e-08,
+      "loss": 0.0004,
+      "step": 18188
+    },
+    {
+      "epoch": 4.133863636363636,
+      "grad_norm": 1.396624458902258,
+      "learning_rate": 9.033570202029871e-08,
+      "loss": 0.0082,
+      "step": 18189
+    },
+    {
+      "epoch": 4.134090909090909,
+      "grad_norm": 2.219817156094807,
+      "learning_rate": 9.028948820549873e-08,
+      "loss": 0.011,
+      "step": 18190
+    },
+    {
+      "epoch": 4.134318181818182,
+      "grad_norm": 0.7701869084579412,
+      "learning_rate": 9.024328529437916e-08,
+      "loss": 0.0061,
+      "step": 18191
+    },
+    {
+      "epoch": 4.134545454545455,
+      "grad_norm": 0.8059203665262773,
+      "learning_rate": 9.019709328788179e-08,
+      "loss": 0.0033,
+      "step": 18192
+    },
+    {
+      "epoch": 4.134772727272727,
+      "grad_norm": 0.597479016854597,
+      "learning_rate": 9.015091218694864e-08,
+      "loss": 0.001,
+      "step": 18193
+    },
+    {
+      "epoch": 4.135,
+      "grad_norm": 0.2901985260852377,
+      "learning_rate": 9.010474199252155e-08,
+      "loss": 0.001,
+      "step": 18194
+    },
+    {
+      "epoch": 4.135227272727272,
+      "grad_norm": 0.5274680187375944,
+      "learning_rate": 9.00585827055419e-08,
+      "loss": 0.0041,
+      "step": 18195
+    },
+    {
+      "epoch": 4.135454545454546,
+      "grad_norm": 1.2156936814968264,
+      "learning_rate": 9.001243432695102e-08,
+      "loss": 0.0033,
+      "step": 18196
+    },
+    {
+      "epoch": 4.135681818181818,
+      "grad_norm": 1.669124588041084,
+      "learning_rate": 8.996629685768997e-08,
+      "loss": 0.0066,
+      "step": 18197
+    },
+    {
+      "epoch": 4.135909090909091,
+      "grad_norm": 0.4994280957062713,
+      "learning_rate": 8.992017029869945e-08,
+      "loss": 0.0016,
+      "step": 18198
+    },
+    {
+      "epoch": 4.136136363636363,
+      "grad_norm": 1.041946894304795,
+      "learning_rate": 8.987405465092039e-08,
+      "loss": 0.0024,
+      "step": 18199
+    },
+    {
+      "epoch": 4.136363636363637,
+      "grad_norm": 0.4170215334288822,
+      "learning_rate": 8.982794991529278e-08,
+      "loss": 0.0011,
+      "step": 18200
+    },
+    {
+      "epoch": 4.136590909090909,
+      "grad_norm": 1.1733234625501605,
+      "learning_rate": 8.978185609275693e-08,
+      "loss": 0.0015,
+      "step": 18201
+    },
+    {
+      "epoch": 4.136818181818182,
+      "grad_norm": 0.3072334347673726,
+      "learning_rate": 8.973577318425276e-08,
+      "loss": 0.0009,
+      "step": 18202
+    },
+    {
+      "epoch": 4.137045454545454,
+      "grad_norm": 0.609986558231291,
+      "learning_rate": 8.968970119071996e-08,
+      "loss": 0.002,
+      "step": 18203
+    },
+    {
+      "epoch": 4.137272727272728,
+      "grad_norm": 1.8594615020480638,
+      "learning_rate": 8.964364011309824e-08,
+      "loss": 0.0067,
+      "step": 18204
+    },
+    {
+      "epoch": 4.1375,
+      "grad_norm": 2.0121847950569087,
+      "learning_rate": 8.959758995232653e-08,
+      "loss": 0.0081,
+      "step": 18205
+    },
+    {
+      "epoch": 4.137727272727273,
+      "grad_norm": 0.9128676091877779,
+      "learning_rate": 8.955155070934404e-08,
+      "loss": 0.002,
+      "step": 18206
+    },
+    {
+      "epoch": 4.137954545454545,
+      "grad_norm": 0.8920280838897586,
+      "learning_rate": 8.950552238508957e-08,
+      "loss": 0.0039,
+      "step": 18207
+    },
+    {
+      "epoch": 4.138181818181819,
+      "grad_norm": 0.7217971271563769,
+      "learning_rate": 8.945950498050179e-08,
+      "loss": 0.0047,
+      "step": 18208
+    },
+    {
+      "epoch": 4.138409090909091,
+      "grad_norm": 3.635285209757989,
+      "learning_rate": 8.941349849651892e-08,
+      "loss": 0.0055,
+      "step": 18209
+    },
+    {
+      "epoch": 4.138636363636364,
+      "grad_norm": 0.8571872399606075,
+      "learning_rate": 8.936750293407924e-08,
+      "loss": 0.0033,
+      "step": 18210
+    },
+    {
+      "epoch": 4.138863636363636,
+      "grad_norm": 0.4191463991272944,
+      "learning_rate": 8.932151829412053e-08,
+      "loss": 0.0016,
+      "step": 18211
+    },
+    {
+      "epoch": 4.139090909090909,
+      "grad_norm": 1.624473095345269,
+      "learning_rate": 8.927554457758052e-08,
+      "loss": 0.0076,
+      "step": 18212
+    },
+    {
+      "epoch": 4.139318181818182,
+      "grad_norm": 0.37683555092227977,
+      "learning_rate": 8.922958178539706e-08,
+      "loss": 0.0013,
+      "step": 18213
+    },
+    {
+      "epoch": 4.139545454545455,
+      "grad_norm": 0.8379911479612931,
+      "learning_rate": 8.918362991850706e-08,
+      "loss": 0.007,
+      "step": 18214
+    },
+    {
+      "epoch": 4.139772727272727,
+      "grad_norm": 1.1215799494083412,
+      "learning_rate": 8.913768897784753e-08,
+      "loss": 0.0057,
+      "step": 18215
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 0.6692358295227597,
+      "learning_rate": 8.909175896435542e-08,
+      "loss": 0.0013,
+      "step": 18216
+    },
+    {
+      "epoch": 4.140227272727273,
+      "grad_norm": 0.5574515510644846,
+      "learning_rate": 8.904583987896745e-08,
+      "loss": 0.0023,
+      "step": 18217
+    },
+    {
+      "epoch": 4.140454545454546,
+      "grad_norm": 0.772935659029236,
+      "learning_rate": 8.899993172261964e-08,
+      "loss": 0.0029,
+      "step": 18218
+    },
+    {
+      "epoch": 4.140681818181818,
+      "grad_norm": 0.5260154240805196,
+      "learning_rate": 8.895403449624834e-08,
+      "loss": 0.0013,
+      "step": 18219
+    },
+    {
+      "epoch": 4.140909090909091,
+      "grad_norm": 1.6557030058234676,
+      "learning_rate": 8.890814820078946e-08,
+      "loss": 0.0071,
+      "step": 18220
+    },
+    {
+      "epoch": 4.141136363636364,
+      "grad_norm": 0.46450959803428615,
+      "learning_rate": 8.886227283717887e-08,
+      "loss": 0.0016,
+      "step": 18221
+    },
+    {
+      "epoch": 4.141363636363637,
+      "grad_norm": 0.6298166059787161,
+      "learning_rate": 8.88164084063517e-08,
+      "loss": 0.0027,
+      "step": 18222
+    },
+    {
+      "epoch": 4.141590909090909,
+      "grad_norm": 1.1230858230792984,
+      "learning_rate": 8.877055490924346e-08,
+      "loss": 0.0049,
+      "step": 18223
+    },
+    {
+      "epoch": 4.141818181818182,
+      "grad_norm": 0.8826203247589195,
+      "learning_rate": 8.872471234678902e-08,
+      "loss": 0.0058,
+      "step": 18224
+    },
+    {
+      "epoch": 4.142045454545454,
+      "grad_norm": 1.3032761802121153,
+      "learning_rate": 8.867888071992334e-08,
+      "loss": 0.0067,
+      "step": 18225
+    },
+    {
+      "epoch": 4.1422727272727276,
+      "grad_norm": 0.350657120210943,
+      "learning_rate": 8.863306002958087e-08,
+      "loss": 0.0012,
+      "step": 18226
+    },
+    {
+      "epoch": 4.1425,
+      "grad_norm": 0.7392141174253399,
+      "learning_rate": 8.858725027669615e-08,
+      "loss": 0.0065,
+      "step": 18227
+    },
+    {
+      "epoch": 4.142727272727273,
+      "grad_norm": 1.4532920774802807,
+      "learning_rate": 8.854145146220315e-08,
+      "loss": 0.0082,
+      "step": 18228
+    },
+    {
+      "epoch": 4.142954545454545,
+      "grad_norm": 0.5604635963097679,
+      "learning_rate": 8.849566358703585e-08,
+      "loss": 0.0045,
+      "step": 18229
+    },
+    {
+      "epoch": 4.1431818181818185,
+      "grad_norm": 1.2051535079608808,
+      "learning_rate": 8.844988665212814e-08,
+      "loss": 0.0137,
+      "step": 18230
+    },
+    {
+      "epoch": 4.143409090909091,
+      "grad_norm": 1.2820091244446168,
+      "learning_rate": 8.840412065841318e-08,
+      "loss": 0.0063,
+      "step": 18231
+    },
+    {
+      "epoch": 4.143636363636364,
+      "grad_norm": 0.47880993919063264,
+      "learning_rate": 8.835836560682432e-08,
+      "loss": 0.0044,
+      "step": 18232
+    },
+    {
+      "epoch": 4.143863636363636,
+      "grad_norm": 0.7020237839423978,
+      "learning_rate": 8.831262149829455e-08,
+      "loss": 0.0033,
+      "step": 18233
+    },
+    {
+      "epoch": 4.1440909090909095,
+      "grad_norm": 1.7808608420556151,
+      "learning_rate": 8.826688833375681e-08,
+      "loss": 0.0108,
+      "step": 18234
+    },
+    {
+      "epoch": 4.144318181818182,
+      "grad_norm": 3.269810644771861,
+      "learning_rate": 8.82211661141437e-08,
+      "loss": 0.0026,
+      "step": 18235
+    },
+    {
+      "epoch": 4.1445454545454545,
+      "grad_norm": 2.2689886429866664,
+      "learning_rate": 8.81754548403873e-08,
+      "loss": 0.0032,
+      "step": 18236
+    },
+    {
+      "epoch": 4.144772727272727,
+      "grad_norm": 0.3098938507348553,
+      "learning_rate": 8.812975451342e-08,
+      "loss": 0.0011,
+      "step": 18237
+    },
+    {
+      "epoch": 4.145,
+      "grad_norm": 1.7330866136663479,
+      "learning_rate": 8.808406513417356e-08,
+      "loss": 0.0209,
+      "step": 18238
+    },
+    {
+      "epoch": 4.145227272727273,
+      "grad_norm": 0.6391267070669747,
+      "learning_rate": 8.803838670357993e-08,
+      "loss": 0.0021,
+      "step": 18239
+    },
+    {
+      "epoch": 4.1454545454545455,
+      "grad_norm": 0.6168754583966284,
+      "learning_rate": 8.799271922257019e-08,
+      "loss": 0.0028,
+      "step": 18240
+    },
+    {
+      "epoch": 4.145681818181818,
+      "grad_norm": 2.1881377082459696,
+      "learning_rate": 8.794706269207559e-08,
+      "loss": 0.0073,
+      "step": 18241
+    },
+    {
+      "epoch": 4.1459090909090905,
+      "grad_norm": 0.8345243036069497,
+      "learning_rate": 8.790141711302755e-08,
+      "loss": 0.0018,
+      "step": 18242
+    },
+    {
+      "epoch": 4.146136363636364,
+      "grad_norm": 1.076684384088468,
+      "learning_rate": 8.78557824863567e-08,
+      "loss": 0.0063,
+      "step": 18243
+    },
+    {
+      "epoch": 4.1463636363636365,
+      "grad_norm": 1.1641267877271966,
+      "learning_rate": 8.781015881299349e-08,
+      "loss": 0.0021,
+      "step": 18244
+    },
+    {
+      "epoch": 4.146590909090909,
+      "grad_norm": 0.5659894552549696,
+      "learning_rate": 8.776454609386823e-08,
+      "loss": 0.0051,
+      "step": 18245
+    },
+    {
+      "epoch": 4.1468181818181815,
+      "grad_norm": 1.669397530878632,
+      "learning_rate": 8.771894432991118e-08,
+      "loss": 0.0071,
+      "step": 18246
+    },
+    {
+      "epoch": 4.147045454545455,
+      "grad_norm": 0.5204232538811322,
+      "learning_rate": 8.767335352205218e-08,
+      "loss": 0.002,
+      "step": 18247
+    },
+    {
+      "epoch": 4.147272727272727,
+      "grad_norm": 1.0727599584255059,
+      "learning_rate": 8.76277736712211e-08,
+      "loss": 0.0026,
+      "step": 18248
+    },
+    {
+      "epoch": 4.1475,
+      "grad_norm": 0.5889613100214705,
+      "learning_rate": 8.7582204778347e-08,
+      "loss": 0.0021,
+      "step": 18249
+    },
+    {
+      "epoch": 4.1477272727272725,
+      "grad_norm": 0.636101700938271,
+      "learning_rate": 8.75366468443594e-08,
+      "loss": 0.0017,
+      "step": 18250
+    },
+    {
+      "epoch": 4.147954545454546,
+      "grad_norm": 0.8988075963369333,
+      "learning_rate": 8.749109987018715e-08,
+      "loss": 0.0023,
+      "step": 18251
+    },
+    {
+      "epoch": 4.148181818181818,
+      "grad_norm": 1.0691211447383302,
+      "learning_rate": 8.74455638567593e-08,
+      "loss": 0.0079,
+      "step": 18252
+    },
+    {
+      "epoch": 4.148409090909091,
+      "grad_norm": 0.25515146765315466,
+      "learning_rate": 8.740003880500412e-08,
+      "loss": 0.0009,
+      "step": 18253
+    },
+    {
+      "epoch": 4.1486363636363635,
+      "grad_norm": 0.6212025257226483,
+      "learning_rate": 8.735452471584999e-08,
+      "loss": 0.0035,
+      "step": 18254
+    },
+    {
+      "epoch": 4.148863636363636,
+      "grad_norm": 0.9813385962431331,
+      "learning_rate": 8.730902159022503e-08,
+      "loss": 0.0046,
+      "step": 18255
+    },
+    {
+      "epoch": 4.149090909090909,
+      "grad_norm": 0.3692395126915699,
+      "learning_rate": 8.726352942905727e-08,
+      "loss": 0.0006,
+      "step": 18256
+    },
+    {
+      "epoch": 4.149318181818182,
+      "grad_norm": 0.3608091613439672,
+      "learning_rate": 8.72180482332742e-08,
+      "loss": 0.0024,
+      "step": 18257
+    },
+    {
+      "epoch": 4.149545454545454,
+      "grad_norm": 1.78018708624816,
+      "learning_rate": 8.717257800380338e-08,
+      "loss": 0.0086,
+      "step": 18258
+    },
+    {
+      "epoch": 4.149772727272727,
+      "grad_norm": 1.7131439462864895,
+      "learning_rate": 8.712711874157197e-08,
+      "loss": 0.0119,
+      "step": 18259
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 1.842722803668197,
+      "learning_rate": 8.708167044750693e-08,
+      "loss": 0.0132,
+      "step": 18260
+    },
+    {
+      "epoch": 4.150227272727273,
+      "grad_norm": 0.42713058413316646,
+      "learning_rate": 8.70362331225353e-08,
+      "loss": 0.0015,
+      "step": 18261
+    },
+    {
+      "epoch": 4.150454545454545,
+      "grad_norm": 0.7026161615467734,
+      "learning_rate": 8.699080676758321e-08,
+      "loss": 0.0033,
+      "step": 18262
+    },
+    {
+      "epoch": 4.150681818181818,
+      "grad_norm": 1.2862385658394349,
+      "learning_rate": 8.694539138357723e-08,
+      "loss": 0.0039,
+      "step": 18263
+    },
+    {
+      "epoch": 4.150909090909091,
+      "grad_norm": 1.6578254920031388,
+      "learning_rate": 8.68999869714434e-08,
+      "loss": 0.0056,
+      "step": 18264
+    },
+    {
+      "epoch": 4.151136363636364,
+      "grad_norm": 0.34141401041597463,
+      "learning_rate": 8.685459353210774e-08,
+      "loss": 0.0012,
+      "step": 18265
+    },
+    {
+      "epoch": 4.151363636363636,
+      "grad_norm": 0.14170488564107273,
+      "learning_rate": 8.680921106649558e-08,
+      "loss": 0.0008,
+      "step": 18266
+    },
+    {
+      "epoch": 4.151590909090909,
+      "grad_norm": 0.3061093214923374,
+      "learning_rate": 8.676383957553256e-08,
+      "loss": 0.0009,
+      "step": 18267
+    },
+    {
+      "epoch": 4.151818181818181,
+      "grad_norm": 0.5507530203018651,
+      "learning_rate": 8.671847906014385e-08,
+      "loss": 0.0016,
+      "step": 18268
+    },
+    {
+      "epoch": 4.152045454545455,
+      "grad_norm": 0.549681334757461,
+      "learning_rate": 8.667312952125446e-08,
+      "loss": 0.0033,
+      "step": 18269
+    },
+    {
+      "epoch": 4.152272727272727,
+      "grad_norm": 0.1185165553023937,
+      "learning_rate": 8.662779095978922e-08,
+      "loss": 0.0004,
+      "step": 18270
+    },
+    {
+      "epoch": 4.1525,
+      "grad_norm": 1.2429532507134373,
+      "learning_rate": 8.65824633766723e-08,
+      "loss": 0.0059,
+      "step": 18271
+    },
+    {
+      "epoch": 4.152727272727272,
+      "grad_norm": 0.9626394633700582,
+      "learning_rate": 8.653714677282842e-08,
+      "loss": 0.0058,
+      "step": 18272
+    },
+    {
+      "epoch": 4.152954545454546,
+      "grad_norm": 1.2113245839174756,
+      "learning_rate": 8.649184114918149e-08,
+      "loss": 0.0034,
+      "step": 18273
+    },
+    {
+      "epoch": 4.153181818181818,
+      "grad_norm": 1.01402391248653,
+      "learning_rate": 8.644654650665559e-08,
+      "loss": 0.0082,
+      "step": 18274
+    },
+    {
+      "epoch": 4.153409090909091,
+      "grad_norm": 0.47018842913605635,
+      "learning_rate": 8.640126284617398e-08,
+      "loss": 0.0013,
+      "step": 18275
+    },
+    {
+      "epoch": 4.153636363636363,
+      "grad_norm": 0.8658192242213478,
+      "learning_rate": 8.635599016866031e-08,
+      "loss": 0.0068,
+      "step": 18276
+    },
+    {
+      "epoch": 4.153863636363637,
+      "grad_norm": 1.3566333563249535,
+      "learning_rate": 8.631072847503765e-08,
+      "loss": 0.0098,
+      "step": 18277
+    },
+    {
+      "epoch": 4.154090909090909,
+      "grad_norm": 0.312478390665022,
+      "learning_rate": 8.626547776622924e-08,
+      "loss": 0.0008,
+      "step": 18278
+    },
+    {
+      "epoch": 4.154318181818182,
+      "grad_norm": 0.4594652544359086,
+      "learning_rate": 8.622023804315739e-08,
+      "loss": 0.0016,
+      "step": 18279
+    },
+    {
+      "epoch": 4.154545454545454,
+      "grad_norm": 2.157107431254845,
+      "learning_rate": 8.61750093067449e-08,
+      "loss": 0.0049,
+      "step": 18280
+    },
+    {
+      "epoch": 4.154772727272727,
+      "grad_norm": 0.5201932899689388,
+      "learning_rate": 8.612979155791401e-08,
+      "loss": 0.003,
+      "step": 18281
+    },
+    {
+      "epoch": 4.155,
+      "grad_norm": 0.6134871739034942,
+      "learning_rate": 8.608458479758677e-08,
+      "loss": 0.003,
+      "step": 18282
+    },
+    {
+      "epoch": 4.155227272727273,
+      "grad_norm": 0.6717258067195409,
+      "learning_rate": 8.603938902668515e-08,
+      "loss": 0.0013,
+      "step": 18283
+    },
+    {
+      "epoch": 4.155454545454545,
+      "grad_norm": 0.6803281487465842,
+      "learning_rate": 8.59942042461305e-08,
+      "loss": 0.0056,
+      "step": 18284
+    },
+    {
+      "epoch": 4.155681818181818,
+      "grad_norm": 1.041227769001917,
+      "learning_rate": 8.594903045684444e-08,
+      "loss": 0.0062,
+      "step": 18285
+    },
+    {
+      "epoch": 4.155909090909091,
+      "grad_norm": 1.6532417130768489,
+      "learning_rate": 8.590386765974804e-08,
+      "loss": 0.0093,
+      "step": 18286
+    },
+    {
+      "epoch": 4.156136363636364,
+      "grad_norm": 1.7622463835228295,
+      "learning_rate": 8.585871585576228e-08,
+      "loss": 0.0054,
+      "step": 18287
+    },
+    {
+      "epoch": 4.156363636363636,
+      "grad_norm": 0.8257224018179862,
+      "learning_rate": 8.581357504580786e-08,
+      "loss": 0.0025,
+      "step": 18288
+    },
+    {
+      "epoch": 4.156590909090909,
+      "grad_norm": 0.28498251775432276,
+      "learning_rate": 8.576844523080538e-08,
+      "loss": 0.0009,
+      "step": 18289
+    },
+    {
+      "epoch": 4.156818181818182,
+      "grad_norm": 0.832810544989828,
+      "learning_rate": 8.572332641167496e-08,
+      "loss": 0.0051,
+      "step": 18290
+    },
+    {
+      "epoch": 4.157045454545455,
+      "grad_norm": 2.5014271918154796,
+      "learning_rate": 8.567821858933678e-08,
+      "loss": 0.0064,
+      "step": 18291
+    },
+    {
+      "epoch": 4.157272727272727,
+      "grad_norm": 1.0329051845616783,
+      "learning_rate": 8.563312176471078e-08,
+      "loss": 0.0052,
+      "step": 18292
+    },
+    {
+      "epoch": 4.1575,
+      "grad_norm": 1.2644239384160505,
+      "learning_rate": 8.558803593871622e-08,
+      "loss": 0.0076,
+      "step": 18293
+    },
+    {
+      "epoch": 4.157727272727273,
+      "grad_norm": 0.42186709706067493,
+      "learning_rate": 8.554296111227267e-08,
+      "loss": 0.0035,
+      "step": 18294
+    },
+    {
+      "epoch": 4.157954545454546,
+      "grad_norm": 1.701108585389682,
+      "learning_rate": 8.54978972862993e-08,
+      "loss": 0.0038,
+      "step": 18295
+    },
+    {
+      "epoch": 4.158181818181818,
+      "grad_norm": 1.2781261682580873,
+      "learning_rate": 8.545284446171518e-08,
+      "loss": 0.0042,
+      "step": 18296
+    },
+    {
+      "epoch": 4.158409090909091,
+      "grad_norm": 1.987126527789785,
+      "learning_rate": 8.540780263943865e-08,
+      "loss": 0.0145,
+      "step": 18297
+    },
+    {
+      "epoch": 4.158636363636363,
+      "grad_norm": 0.9665774481924118,
+      "learning_rate": 8.536277182038845e-08,
+      "loss": 0.0047,
+      "step": 18298
+    },
+    {
+      "epoch": 4.158863636363637,
+      "grad_norm": 0.7838091869834937,
+      "learning_rate": 8.531775200548271e-08,
+      "loss": 0.0035,
+      "step": 18299
+    },
+    {
+      "epoch": 4.159090909090909,
+      "grad_norm": 0.3546919509946543,
+      "learning_rate": 8.52727431956396e-08,
+      "loss": 0.0054,
+      "step": 18300
+    },
+    {
+      "epoch": 4.159318181818182,
+      "grad_norm": 0.588641755848844,
+      "learning_rate": 8.522774539177687e-08,
+      "loss": 0.0053,
+      "step": 18301
+    },
+    {
+      "epoch": 4.159545454545454,
+      "grad_norm": 2.1019111903212857,
+      "learning_rate": 8.518275859481201e-08,
+      "loss": 0.0076,
+      "step": 18302
+    },
+    {
+      "epoch": 4.159772727272728,
+      "grad_norm": 0.9391266670129541,
+      "learning_rate": 8.513778280566258e-08,
+      "loss": 0.0075,
+      "step": 18303
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 1.1343534524846224,
+      "learning_rate": 8.509281802524548e-08,
+      "loss": 0.006,
+      "step": 18304
+    },
+    {
+      "epoch": 4.160227272727273,
+      "grad_norm": 0.6402112908365518,
+      "learning_rate": 8.504786425447796e-08,
+      "loss": 0.0031,
+      "step": 18305
+    },
+    {
+      "epoch": 4.160454545454545,
+      "grad_norm": 0.6857808376804004,
+      "learning_rate": 8.500292149427638e-08,
+      "loss": 0.0016,
+      "step": 18306
+    },
+    {
+      "epoch": 4.160681818181819,
+      "grad_norm": 0.8009285171785664,
+      "learning_rate": 8.495798974555721e-08,
+      "loss": 0.0025,
+      "step": 18307
+    },
+    {
+      "epoch": 4.160909090909091,
+      "grad_norm": 2.184258063865562,
+      "learning_rate": 8.49130690092368e-08,
+      "loss": 0.0076,
+      "step": 18308
+    },
+    {
+      "epoch": 4.161136363636364,
+      "grad_norm": 0.9419022086106181,
+      "learning_rate": 8.486815928623132e-08,
+      "loss": 0.0063,
+      "step": 18309
+    },
+    {
+      "epoch": 4.161363636363636,
+      "grad_norm": 1.703509499639608,
+      "learning_rate": 8.482326057745626e-08,
+      "loss": 0.0089,
+      "step": 18310
+    },
+    {
+      "epoch": 4.161590909090909,
+      "grad_norm": 0.17046011065484895,
+      "learning_rate": 8.47783728838273e-08,
+      "loss": 0.0004,
+      "step": 18311
+    },
+    {
+      "epoch": 4.161818181818182,
+      "grad_norm": 1.5318977574393378,
+      "learning_rate": 8.47334962062598e-08,
+      "loss": 0.0031,
+      "step": 18312
+    },
+    {
+      "epoch": 4.162045454545455,
+      "grad_norm": 0.7170114869429529,
+      "learning_rate": 8.46886305456688e-08,
+      "loss": 0.0061,
+      "step": 18313
+    },
+    {
+      "epoch": 4.162272727272727,
+      "grad_norm": 0.3280965640402128,
+      "learning_rate": 8.464377590296939e-08,
+      "loss": 0.0014,
+      "step": 18314
+    },
+    {
+      "epoch": 4.1625,
+      "grad_norm": 1.7093001955392015,
+      "learning_rate": 8.459893227907599e-08,
+      "loss": 0.0172,
+      "step": 18315
+    },
+    {
+      "epoch": 4.162727272727273,
+      "grad_norm": 0.7470017485446888,
+      "learning_rate": 8.455409967490315e-08,
+      "loss": 0.0021,
+      "step": 18316
+    },
+    {
+      "epoch": 4.162954545454546,
+      "grad_norm": 1.7563795135610956,
+      "learning_rate": 8.450927809136506e-08,
+      "loss": 0.0021,
+      "step": 18317
+    },
+    {
+      "epoch": 4.163181818181818,
+      "grad_norm": 0.5019790994571065,
+      "learning_rate": 8.446446752937571e-08,
+      "loss": 0.0024,
+      "step": 18318
+    },
+    {
+      "epoch": 4.163409090909091,
+      "grad_norm": 1.2116373800466251,
+      "learning_rate": 8.441966798984894e-08,
+      "loss": 0.0062,
+      "step": 18319
+    },
+    {
+      "epoch": 4.163636363636364,
+      "grad_norm": 0.5310267787197038,
+      "learning_rate": 8.437487947369816e-08,
+      "loss": 0.0017,
+      "step": 18320
+    },
+    {
+      "epoch": 4.1638636363636365,
+      "grad_norm": 0.572162126934574,
+      "learning_rate": 8.433010198183684e-08,
+      "loss": 0.003,
+      "step": 18321
+    },
+    {
+      "epoch": 4.164090909090909,
+      "grad_norm": 2.217401663082897,
+      "learning_rate": 8.428533551517807e-08,
+      "loss": 0.0045,
+      "step": 18322
+    },
+    {
+      "epoch": 4.164318181818182,
+      "grad_norm": 0.9058440069256827,
+      "learning_rate": 8.424058007463453e-08,
+      "loss": 0.0049,
+      "step": 18323
+    },
+    {
+      "epoch": 4.164545454545454,
+      "grad_norm": 1.7535577124165167,
+      "learning_rate": 8.419583566111895e-08,
+      "loss": 0.0093,
+      "step": 18324
+    },
+    {
+      "epoch": 4.1647727272727275,
+      "grad_norm": 1.3179324050129868,
+      "learning_rate": 8.415110227554376e-08,
+      "loss": 0.007,
+      "step": 18325
+    },
+    {
+      "epoch": 4.165,
+      "grad_norm": 0.9965532633892592,
+      "learning_rate": 8.410637991882117e-08,
+      "loss": 0.0079,
+      "step": 18326
+    },
+    {
+      "epoch": 4.165227272727273,
+      "grad_norm": 0.9279924936990306,
+      "learning_rate": 8.406166859186327e-08,
+      "loss": 0.0047,
+      "step": 18327
+    },
+    {
+      "epoch": 4.165454545454545,
+      "grad_norm": 1.5183692449290769,
+      "learning_rate": 8.401696829558152e-08,
+      "loss": 0.004,
+      "step": 18328
+    },
+    {
+      "epoch": 4.1656818181818185,
+      "grad_norm": 2.184857295133181,
+      "learning_rate": 8.397227903088757e-08,
+      "loss": 0.0079,
+      "step": 18329
+    },
+    {
+      "epoch": 4.165909090909091,
+      "grad_norm": 0.43393411386654834,
+      "learning_rate": 8.392760079869278e-08,
+      "loss": 0.0012,
+      "step": 18330
+    },
+    {
+      "epoch": 4.1661363636363635,
+      "grad_norm": 1.9316257848030225,
+      "learning_rate": 8.388293359990807e-08,
+      "loss": 0.0155,
+      "step": 18331
+    },
+    {
+      "epoch": 4.166363636363636,
+      "grad_norm": 0.3772923220520482,
+      "learning_rate": 8.383827743544446e-08,
+      "loss": 0.0011,
+      "step": 18332
+    },
+    {
+      "epoch": 4.1665909090909095,
+      "grad_norm": 3.506544429974403,
+      "learning_rate": 8.379363230621246e-08,
+      "loss": 0.0031,
+      "step": 18333
+    },
+    {
+      "epoch": 4.166818181818182,
+      "grad_norm": 0.3374542787643541,
+      "learning_rate": 8.374899821312246e-08,
+      "loss": 0.0009,
+      "step": 18334
+    },
+    {
+      "epoch": 4.1670454545454545,
+      "grad_norm": 1.4944483314669637,
+      "learning_rate": 8.370437515708465e-08,
+      "loss": 0.0142,
+      "step": 18335
+    },
+    {
+      "epoch": 4.167272727272727,
+      "grad_norm": 0.8674781647555466,
+      "learning_rate": 8.365976313900917e-08,
+      "loss": 0.0042,
+      "step": 18336
+    },
+    {
+      "epoch": 4.1675,
+      "grad_norm": 0.4407416952012621,
+      "learning_rate": 8.361516215980532e-08,
+      "loss": 0.0009,
+      "step": 18337
+    },
+    {
+      "epoch": 4.167727272727273,
+      "grad_norm": 1.0613269491399637,
+      "learning_rate": 8.357057222038286e-08,
+      "loss": 0.0085,
+      "step": 18338
+    },
+    {
+      "epoch": 4.1679545454545455,
+      "grad_norm": 2.0772028215777074,
+      "learning_rate": 8.352599332165097e-08,
+      "loss": 0.0185,
+      "step": 18339
+    },
+    {
+      "epoch": 4.168181818181818,
+      "grad_norm": 0.24877710717953405,
+      "learning_rate": 8.348142546451894e-08,
+      "loss": 0.0004,
+      "step": 18340
+    },
+    {
+      "epoch": 4.1684090909090905,
+      "grad_norm": 0.7260609630737784,
+      "learning_rate": 8.34368686498952e-08,
+      "loss": 0.004,
+      "step": 18341
+    },
+    {
+      "epoch": 4.168636363636364,
+      "grad_norm": 1.8581155685328425,
+      "learning_rate": 8.33923228786885e-08,
+      "loss": 0.0087,
+      "step": 18342
+    },
+    {
+      "epoch": 4.168863636363636,
+      "grad_norm": 0.4305197253177193,
+      "learning_rate": 8.334778815180723e-08,
+      "loss": 0.0022,
+      "step": 18343
+    },
+    {
+      "epoch": 4.169090909090909,
+      "grad_norm": 0.7190694383015246,
+      "learning_rate": 8.330326447015966e-08,
+      "loss": 0.0031,
+      "step": 18344
+    },
+    {
+      "epoch": 4.1693181818181815,
+      "grad_norm": 0.7534617541890928,
+      "learning_rate": 8.325875183465344e-08,
+      "loss": 0.0023,
+      "step": 18345
+    },
+    {
+      "epoch": 4.169545454545455,
+      "grad_norm": 0.47718093192338523,
+      "learning_rate": 8.321425024619641e-08,
+      "loss": 0.0025,
+      "step": 18346
+    },
+    {
+      "epoch": 4.169772727272727,
+      "grad_norm": 0.4873597273717066,
+      "learning_rate": 8.316975970569595e-08,
+      "loss": 0.0013,
+      "step": 18347
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 1.6868217292601557,
+      "learning_rate": 8.31252802140594e-08,
+      "loss": 0.0081,
+      "step": 18348
+    },
+    {
+      "epoch": 4.1702272727272724,
+      "grad_norm": 0.8551205760453998,
+      "learning_rate": 8.308081177219374e-08,
+      "loss": 0.0035,
+      "step": 18349
+    },
+    {
+      "epoch": 4.170454545454546,
+      "grad_norm": 1.0833773527121886,
+      "learning_rate": 8.303635438100575e-08,
+      "loss": 0.0044,
+      "step": 18350
+    },
+    {
+      "epoch": 4.170681818181818,
+      "grad_norm": 0.5411444221528516,
+      "learning_rate": 8.299190804140205e-08,
+      "loss": 0.0034,
+      "step": 18351
+    },
+    {
+      "epoch": 4.170909090909091,
+      "grad_norm": 0.4896019481662361,
+      "learning_rate": 8.294747275428887e-08,
+      "loss": 0.0027,
+      "step": 18352
+    },
+    {
+      "epoch": 4.171136363636363,
+      "grad_norm": 1.2445240458669835,
+      "learning_rate": 8.290304852057257e-08,
+      "loss": 0.0061,
+      "step": 18353
+    },
+    {
+      "epoch": 4.171363636363636,
+      "grad_norm": 3.693188554116214,
+      "learning_rate": 8.285863534115868e-08,
+      "loss": 0.0048,
+      "step": 18354
+    },
+    {
+      "epoch": 4.171590909090909,
+      "grad_norm": 1.0224399567109568,
+      "learning_rate": 8.281423321695299e-08,
+      "loss": 0.0085,
+      "step": 18355
+    },
+    {
+      "epoch": 4.171818181818182,
+      "grad_norm": 1.0349515333064194,
+      "learning_rate": 8.276984214886104e-08,
+      "loss": 0.0037,
+      "step": 18356
+    },
+    {
+      "epoch": 4.172045454545454,
+      "grad_norm": 2.032735834368802,
+      "learning_rate": 8.272546213778799e-08,
+      "loss": 0.0078,
+      "step": 18357
+    },
+    {
+      "epoch": 4.172272727272727,
+      "grad_norm": 0.3873721715987931,
+      "learning_rate": 8.26810931846389e-08,
+      "loss": 0.0008,
+      "step": 18358
+    },
+    {
+      "epoch": 4.1725,
+      "grad_norm": 1.4153866802172836,
+      "learning_rate": 8.263673529031839e-08,
+      "loss": 0.0111,
+      "step": 18359
+    },
+    {
+      "epoch": 4.172727272727273,
+      "grad_norm": 0.8639747729526623,
+      "learning_rate": 8.259238845573101e-08,
+      "loss": 0.0047,
+      "step": 18360
+    },
+    {
+      "epoch": 4.172954545454545,
+      "grad_norm": 1.3261472271851873,
+      "learning_rate": 8.254805268178109e-08,
+      "loss": 0.0056,
+      "step": 18361
+    },
+    {
+      "epoch": 4.173181818181818,
+      "grad_norm": 0.7372840421093849,
+      "learning_rate": 8.250372796937273e-08,
+      "loss": 0.0037,
+      "step": 18362
+    },
+    {
+      "epoch": 4.173409090909091,
+      "grad_norm": 0.6534821682627504,
+      "learning_rate": 8.245941431940983e-08,
+      "loss": 0.0015,
+      "step": 18363
+    },
+    {
+      "epoch": 4.173636363636364,
+      "grad_norm": 1.0335615512684273,
+      "learning_rate": 8.2415111732796e-08,
+      "loss": 0.0028,
+      "step": 18364
+    },
+    {
+      "epoch": 4.173863636363636,
+      "grad_norm": 1.45195303467763,
+      "learning_rate": 8.237082021043459e-08,
+      "loss": 0.0094,
+      "step": 18365
+    },
+    {
+      "epoch": 4.174090909090909,
+      "grad_norm": 1.5000809929336054,
+      "learning_rate": 8.232653975322899e-08,
+      "loss": 0.0062,
+      "step": 18366
+    },
+    {
+      "epoch": 4.174318181818181,
+      "grad_norm": 1.0296929885321553,
+      "learning_rate": 8.22822703620818e-08,
+      "loss": 0.0049,
+      "step": 18367
+    },
+    {
+      "epoch": 4.174545454545455,
+      "grad_norm": 0.8193476807746166,
+      "learning_rate": 8.223801203789599e-08,
+      "loss": 0.0021,
+      "step": 18368
+    },
+    {
+      "epoch": 4.174772727272727,
+      "grad_norm": 0.5023274404797866,
+      "learning_rate": 8.219376478157403e-08,
+      "loss": 0.0015,
+      "step": 18369
+    },
+    {
+      "epoch": 4.175,
+      "grad_norm": 1.0203161477843259,
+      "learning_rate": 8.214952859401812e-08,
+      "loss": 0.0079,
+      "step": 18370
+    },
+    {
+      "epoch": 4.175227272727272,
+      "grad_norm": 0.7568043057068787,
+      "learning_rate": 8.210530347613058e-08,
+      "loss": 0.0028,
+      "step": 18371
+    },
+    {
+      "epoch": 4.175454545454546,
+      "grad_norm": 0.5300181679132528,
+      "learning_rate": 8.206108942881284e-08,
+      "loss": 0.0029,
+      "step": 18372
+    },
+    {
+      "epoch": 4.175681818181818,
+      "grad_norm": 1.1957445233843276,
+      "learning_rate": 8.201688645296676e-08,
+      "loss": 0.0057,
+      "step": 18373
+    },
+    {
+      "epoch": 4.175909090909091,
+      "grad_norm": 1.5141436591591184,
+      "learning_rate": 8.197269454949357e-08,
+      "loss": 0.0107,
+      "step": 18374
+    },
+    {
+      "epoch": 4.176136363636363,
+      "grad_norm": 0.6947354356281039,
+      "learning_rate": 8.192851371929469e-08,
+      "loss": 0.0022,
+      "step": 18375
+    },
+    {
+      "epoch": 4.176363636363637,
+      "grad_norm": 0.4383121402324687,
+      "learning_rate": 8.18843439632707e-08,
+      "loss": 0.0023,
+      "step": 18376
+    },
+    {
+      "epoch": 4.176590909090909,
+      "grad_norm": 1.2178872181025542,
+      "learning_rate": 8.184018528232252e-08,
+      "loss": 0.003,
+      "step": 18377
+    },
+    {
+      "epoch": 4.176818181818182,
+      "grad_norm": 0.326865865933565,
+      "learning_rate": 8.17960376773505e-08,
+      "loss": 0.0005,
+      "step": 18378
+    },
+    {
+      "epoch": 4.177045454545454,
+      "grad_norm": 1.9680830970362413,
+      "learning_rate": 8.175190114925492e-08,
+      "loss": 0.0135,
+      "step": 18379
+    },
+    {
+      "epoch": 4.177272727272728,
+      "grad_norm": 0.9469663259033324,
+      "learning_rate": 8.170777569893581e-08,
+      "loss": 0.0059,
+      "step": 18380
+    },
+    {
+      "epoch": 4.1775,
+      "grad_norm": 0.6208803352293455,
+      "learning_rate": 8.166366132729301e-08,
+      "loss": 0.0021,
+      "step": 18381
+    },
+    {
+      "epoch": 4.177727272727273,
+      "grad_norm": 0.7543916257822901,
+      "learning_rate": 8.161955803522609e-08,
+      "loss": 0.0039,
+      "step": 18382
+    },
+    {
+      "epoch": 4.177954545454545,
+      "grad_norm": 1.1723312949974438,
+      "learning_rate": 8.157546582363437e-08,
+      "loss": 0.0043,
+      "step": 18383
+    },
+    {
+      "epoch": 4.178181818181818,
+      "grad_norm": 1.4104925969870377,
+      "learning_rate": 8.153138469341702e-08,
+      "loss": 0.0037,
+      "step": 18384
+    },
+    {
+      "epoch": 4.178409090909091,
+      "grad_norm": 0.5156214073381509,
+      "learning_rate": 8.148731464547282e-08,
+      "loss": 0.0025,
+      "step": 18385
+    },
+    {
+      "epoch": 4.178636363636364,
+      "grad_norm": 1.4626473185271793,
+      "learning_rate": 8.144325568070043e-08,
+      "loss": 0.0046,
+      "step": 18386
+    },
+    {
+      "epoch": 4.178863636363636,
+      "grad_norm": 0.6148199012983158,
+      "learning_rate": 8.139920779999832e-08,
+      "loss": 0.0012,
+      "step": 18387
+    },
+    {
+      "epoch": 4.179090909090909,
+      "grad_norm": 1.399078154557462,
+      "learning_rate": 8.135517100426489e-08,
+      "loss": 0.0065,
+      "step": 18388
+    },
+    {
+      "epoch": 4.179318181818182,
+      "grad_norm": 0.4945136921334482,
+      "learning_rate": 8.131114529439786e-08,
+      "loss": 0.0011,
+      "step": 18389
+    },
+    {
+      "epoch": 4.179545454545455,
+      "grad_norm": 1.0315864646258952,
+      "learning_rate": 8.126713067129504e-08,
+      "loss": 0.0042,
+      "step": 18390
+    },
+    {
+      "epoch": 4.179772727272727,
+      "grad_norm": 0.6309967536611649,
+      "learning_rate": 8.122312713585406e-08,
+      "loss": 0.0042,
+      "step": 18391
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 0.48185756874542096,
+      "learning_rate": 8.117913468897218e-08,
+      "loss": 0.0024,
+      "step": 18392
+    },
+    {
+      "epoch": 4.180227272727273,
+      "grad_norm": 1.2018740573907778,
+      "learning_rate": 8.113515333154645e-08,
+      "loss": 0.0023,
+      "step": 18393
+    },
+    {
+      "epoch": 4.180454545454546,
+      "grad_norm": 0.8638982909607567,
+      "learning_rate": 8.10911830644738e-08,
+      "loss": 0.0028,
+      "step": 18394
+    },
+    {
+      "epoch": 4.180681818181818,
+      "grad_norm": 0.6578935094992743,
+      "learning_rate": 8.104722388865088e-08,
+      "loss": 0.0013,
+      "step": 18395
+    },
+    {
+      "epoch": 4.180909090909091,
+      "grad_norm": 2.1636664052781294,
+      "learning_rate": 8.100327580497398e-08,
+      "loss": 0.0086,
+      "step": 18396
+    },
+    {
+      "epoch": 4.181136363636363,
+      "grad_norm": 1.7953669606330467,
+      "learning_rate": 8.095933881433947e-08,
+      "loss": 0.0132,
+      "step": 18397
+    },
+    {
+      "epoch": 4.181363636363637,
+      "grad_norm": 0.11013612413944031,
+      "learning_rate": 8.091541291764309e-08,
+      "loss": 0.0003,
+      "step": 18398
+    },
+    {
+      "epoch": 4.181590909090909,
+      "grad_norm": 0.621762599639719,
+      "learning_rate": 8.087149811578057e-08,
+      "loss": 0.0009,
+      "step": 18399
+    },
+    {
+      "epoch": 4.181818181818182,
+      "grad_norm": 1.3816640484348404,
+      "learning_rate": 8.082759440964754e-08,
+      "loss": 0.0038,
+      "step": 18400
+    },
+    {
+      "epoch": 4.182045454545454,
+      "grad_norm": 0.9421329764801715,
+      "learning_rate": 8.078370180013937e-08,
+      "loss": 0.0065,
+      "step": 18401
+    },
+    {
+      "epoch": 4.182272727272728,
+      "grad_norm": 1.9473518155605607,
+      "learning_rate": 8.073982028815083e-08,
+      "loss": 0.0065,
+      "step": 18402
+    },
+    {
+      "epoch": 4.1825,
+      "grad_norm": 1.3806025073131738,
+      "learning_rate": 8.069594987457679e-08,
+      "loss": 0.0053,
+      "step": 18403
+    },
+    {
+      "epoch": 4.182727272727273,
+      "grad_norm": 0.6715202056615132,
+      "learning_rate": 8.065209056031199e-08,
+      "loss": 0.005,
+      "step": 18404
+    },
+    {
+      "epoch": 4.182954545454545,
+      "grad_norm": 0.2971907436442667,
+      "learning_rate": 8.060824234625072e-08,
+      "loss": 0.001,
+      "step": 18405
+    },
+    {
+      "epoch": 4.183181818181819,
+      "grad_norm": 0.4601613984701938,
+      "learning_rate": 8.056440523328727e-08,
+      "loss": 0.0011,
+      "step": 18406
+    },
+    {
+      "epoch": 4.183409090909091,
+      "grad_norm": 0.7113541125615759,
+      "learning_rate": 8.052057922231529e-08,
+      "loss": 0.0041,
+      "step": 18407
+    },
+    {
+      "epoch": 4.183636363636364,
+      "grad_norm": 0.5282657265471666,
+      "learning_rate": 8.047676431422853e-08,
+      "loss": 0.003,
+      "step": 18408
+    },
+    {
+      "epoch": 4.183863636363636,
+      "grad_norm": 0.37627023251515057,
+      "learning_rate": 8.043296050992043e-08,
+      "loss": 0.0013,
+      "step": 18409
+    },
+    {
+      "epoch": 4.184090909090909,
+      "grad_norm": 0.3929455038199952,
+      "learning_rate": 8.038916781028458e-08,
+      "loss": 0.0007,
+      "step": 18410
+    },
+    {
+      "epoch": 4.184318181818182,
+      "grad_norm": 0.5431124134329117,
+      "learning_rate": 8.03453862162136e-08,
+      "loss": 0.0044,
+      "step": 18411
+    },
+    {
+      "epoch": 4.184545454545455,
+      "grad_norm": 2.6688825866493504,
+      "learning_rate": 8.030161572860038e-08,
+      "loss": 0.0077,
+      "step": 18412
+    },
+    {
+      "epoch": 4.184772727272727,
+      "grad_norm": 0.534531192352763,
+      "learning_rate": 8.025785634833756e-08,
+      "loss": 0.0026,
+      "step": 18413
+    },
+    {
+      "epoch": 4.185,
+      "grad_norm": 1.3740344981440349,
+      "learning_rate": 8.021410807631733e-08,
+      "loss": 0.0098,
+      "step": 18414
+    },
+    {
+      "epoch": 4.185227272727273,
+      "grad_norm": 0.8470230843962676,
+      "learning_rate": 8.017037091343202e-08,
+      "loss": 0.0034,
+      "step": 18415
+    },
+    {
+      "epoch": 4.1854545454545455,
+      "grad_norm": 0.8149847441097373,
+      "learning_rate": 8.012664486057322e-08,
+      "loss": 0.0033,
+      "step": 18416
+    },
+    {
+      "epoch": 4.185681818181818,
+      "grad_norm": 1.1784799640391468,
+      "learning_rate": 8.008292991863273e-08,
+      "loss": 0.0028,
+      "step": 18417
+    },
+    {
+      "epoch": 4.185909090909091,
+      "grad_norm": 0.40716381428729725,
+      "learning_rate": 8.003922608850197e-08,
+      "loss": 0.0017,
+      "step": 18418
+    },
+    {
+      "epoch": 4.186136363636364,
+      "grad_norm": 0.6430373972061861,
+      "learning_rate": 7.999553337107217e-08,
+      "loss": 0.0032,
+      "step": 18419
+    },
+    {
+      "epoch": 4.1863636363636365,
+      "grad_norm": 0.9130545067589699,
+      "learning_rate": 7.995185176723423e-08,
+      "loss": 0.0038,
+      "step": 18420
+    },
+    {
+      "epoch": 4.186590909090909,
+      "grad_norm": 18.391180009976992,
+      "learning_rate": 7.990818127787888e-08,
+      "loss": 0.0303,
+      "step": 18421
+    },
+    {
+      "epoch": 4.1868181818181816,
+      "grad_norm": 1.563774289821003,
+      "learning_rate": 7.986452190389668e-08,
+      "loss": 0.0061,
+      "step": 18422
+    },
+    {
+      "epoch": 4.187045454545455,
+      "grad_norm": 1.151666199367097,
+      "learning_rate": 7.982087364617794e-08,
+      "loss": 0.0035,
+      "step": 18423
+    },
+    {
+      "epoch": 4.1872727272727275,
+      "grad_norm": 0.7476546603502819,
+      "learning_rate": 7.977723650561264e-08,
+      "loss": 0.0017,
+      "step": 18424
+    },
+    {
+      "epoch": 4.1875,
+      "grad_norm": 1.3815892143819855,
+      "learning_rate": 7.973361048309072e-08,
+      "loss": 0.0056,
+      "step": 18425
+    },
+    {
+      "epoch": 4.1877272727272725,
+      "grad_norm": 1.356220880472182,
+      "learning_rate": 7.968999557950178e-08,
+      "loss": 0.0036,
+      "step": 18426
+    },
+    {
+      "epoch": 4.187954545454545,
+      "grad_norm": 0.29796608902875465,
+      "learning_rate": 7.964639179573512e-08,
+      "loss": 0.0007,
+      "step": 18427
+    },
+    {
+      "epoch": 4.1881818181818184,
+      "grad_norm": 1.0109055760057823,
+      "learning_rate": 7.960279913268011e-08,
+      "loss": 0.0037,
+      "step": 18428
+    },
+    {
+      "epoch": 4.188409090909091,
+      "grad_norm": 0.31570757767265845,
+      "learning_rate": 7.955921759122541e-08,
+      "loss": 0.0011,
+      "step": 18429
+    },
+    {
+      "epoch": 4.1886363636363635,
+      "grad_norm": 0.7408794277291135,
+      "learning_rate": 7.951564717225982e-08,
+      "loss": 0.0026,
+      "step": 18430
+    },
+    {
+      "epoch": 4.188863636363636,
+      "grad_norm": 0.4590220206836212,
+      "learning_rate": 7.947208787667186e-08,
+      "loss": 0.0014,
+      "step": 18431
+    },
+    {
+      "epoch": 4.189090909090909,
+      "grad_norm": 0.8060968935811551,
+      "learning_rate": 7.942853970534986e-08,
+      "loss": 0.0031,
+      "step": 18432
+    },
+    {
+      "epoch": 4.189318181818182,
+      "grad_norm": 0.6905436782285718,
+      "learning_rate": 7.938500265918165e-08,
+      "loss": 0.0032,
+      "step": 18433
+    },
+    {
+      "epoch": 4.1895454545454545,
+      "grad_norm": 0.3303989067609812,
+      "learning_rate": 7.934147673905504e-08,
+      "loss": 0.0008,
+      "step": 18434
+    },
+    {
+      "epoch": 4.189772727272727,
+      "grad_norm": 0.881681106203037,
+      "learning_rate": 7.929796194585767e-08,
+      "loss": 0.0056,
+      "step": 18435
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 1.6053153173780121,
+      "learning_rate": 7.925445828047697e-08,
+      "loss": 0.0075,
+      "step": 18436
+    },
+    {
+      "epoch": 4.190227272727273,
+      "grad_norm": 0.3734848426451548,
+      "learning_rate": 7.921096574380005e-08,
+      "loss": 0.0011,
+      "step": 18437
+    },
+    {
+      "epoch": 4.190454545454545,
+      "grad_norm": 0.8417553089961931,
+      "learning_rate": 7.916748433671344e-08,
+      "loss": 0.0022,
+      "step": 18438
+    },
+    {
+      "epoch": 4.190681818181818,
+      "grad_norm": 0.36219907459565215,
+      "learning_rate": 7.91240140601042e-08,
+      "loss": 0.0011,
+      "step": 18439
+    },
+    {
+      "epoch": 4.190909090909091,
+      "grad_norm": 1.1245816629949084,
+      "learning_rate": 7.90805549148587e-08,
+      "loss": 0.0078,
+      "step": 18440
+    },
+    {
+      "epoch": 4.191136363636364,
+      "grad_norm": 0.4312190780665295,
+      "learning_rate": 7.90371069018632e-08,
+      "loss": 0.0018,
+      "step": 18441
+    },
+    {
+      "epoch": 4.191363636363636,
+      "grad_norm": 0.6182906002297405,
+      "learning_rate": 7.899367002200349e-08,
+      "loss": 0.0016,
+      "step": 18442
+    },
+    {
+      "epoch": 4.191590909090909,
+      "grad_norm": 1.5598822558889236,
+      "learning_rate": 7.895024427616534e-08,
+      "loss": 0.0041,
+      "step": 18443
+    },
+    {
+      "epoch": 4.191818181818181,
+      "grad_norm": 0.7571456654181101,
+      "learning_rate": 7.890682966523434e-08,
+      "loss": 0.0026,
+      "step": 18444
+    },
+    {
+      "epoch": 4.192045454545455,
+      "grad_norm": 0.3335911253909567,
+      "learning_rate": 7.886342619009598e-08,
+      "loss": 0.002,
+      "step": 18445
+    },
+    {
+      "epoch": 4.192272727272727,
+      "grad_norm": 1.3679959224615892,
+      "learning_rate": 7.882003385163497e-08,
+      "loss": 0.0021,
+      "step": 18446
+    },
+    {
+      "epoch": 4.1925,
+      "grad_norm": 0.8812729709719217,
+      "learning_rate": 7.877665265073636e-08,
+      "loss": 0.004,
+      "step": 18447
+    },
+    {
+      "epoch": 4.192727272727272,
+      "grad_norm": 0.20734523179134276,
+      "learning_rate": 7.873328258828479e-08,
+      "loss": 0.0007,
+      "step": 18448
+    },
+    {
+      "epoch": 4.192954545454546,
+      "grad_norm": 0.6636409209952557,
+      "learning_rate": 7.868992366516455e-08,
+      "loss": 0.0021,
+      "step": 18449
+    },
+    {
+      "epoch": 4.193181818181818,
+      "grad_norm": 0.8734109019529945,
+      "learning_rate": 7.864657588225993e-08,
+      "loss": 0.0031,
+      "step": 18450
+    },
+    {
+      "epoch": 4.193409090909091,
+      "grad_norm": 1.5665337763238465,
+      "learning_rate": 7.860323924045475e-08,
+      "loss": 0.0112,
+      "step": 18451
+    },
+    {
+      "epoch": 4.193636363636363,
+      "grad_norm": 0.5571938514245576,
+      "learning_rate": 7.855991374063274e-08,
+      "loss": 0.0015,
+      "step": 18452
+    },
+    {
+      "epoch": 4.193863636363637,
+      "grad_norm": 0.984985269459338,
+      "learning_rate": 7.851659938367742e-08,
+      "loss": 0.007,
+      "step": 18453
+    },
+    {
+      "epoch": 4.194090909090909,
+      "grad_norm": 0.5849891525621264,
+      "learning_rate": 7.847329617047199e-08,
+      "loss": 0.0032,
+      "step": 18454
+    },
+    {
+      "epoch": 4.194318181818182,
+      "grad_norm": 0.3007165292429843,
+      "learning_rate": 7.843000410189954e-08,
+      "loss": 0.0013,
+      "step": 18455
+    },
+    {
+      "epoch": 4.194545454545454,
+      "grad_norm": 0.5139078902405407,
+      "learning_rate": 7.838672317884287e-08,
+      "loss": 0.0023,
+      "step": 18456
+    },
+    {
+      "epoch": 4.194772727272727,
+      "grad_norm": 0.5178584346883588,
+      "learning_rate": 7.834345340218451e-08,
+      "loss": 0.0011,
+      "step": 18457
+    },
+    {
+      "epoch": 4.195,
+      "grad_norm": 0.5062007563848491,
+      "learning_rate": 7.830019477280681e-08,
+      "loss": 0.0038,
+      "step": 18458
+    },
+    {
+      "epoch": 4.195227272727273,
+      "grad_norm": 1.6134104269162002,
+      "learning_rate": 7.825694729159207e-08,
+      "loss": 0.0052,
+      "step": 18459
+    },
+    {
+      "epoch": 4.195454545454545,
+      "grad_norm": 1.4588752717926559,
+      "learning_rate": 7.821371095942187e-08,
+      "loss": 0.0043,
+      "step": 18460
+    },
+    {
+      "epoch": 4.195681818181818,
+      "grad_norm": 0.9287594863440192,
+      "learning_rate": 7.817048577717805e-08,
+      "loss": 0.0046,
+      "step": 18461
+    },
+    {
+      "epoch": 4.195909090909091,
+      "grad_norm": 0.8963840102407017,
+      "learning_rate": 7.812727174574202e-08,
+      "loss": 0.003,
+      "step": 18462
+    },
+    {
+      "epoch": 4.196136363636364,
+      "grad_norm": 1.2602998032752177,
+      "learning_rate": 7.808406886599514e-08,
+      "loss": 0.0075,
+      "step": 18463
+    },
+    {
+      "epoch": 4.196363636363636,
+      "grad_norm": 0.6755236629212417,
+      "learning_rate": 7.804087713881803e-08,
+      "loss": 0.0015,
+      "step": 18464
+    },
+    {
+      "epoch": 4.196590909090909,
+      "grad_norm": 0.6923726853846482,
+      "learning_rate": 7.799769656509172e-08,
+      "loss": 0.0027,
+      "step": 18465
+    },
+    {
+      "epoch": 4.196818181818182,
+      "grad_norm": 0.15441005061231342,
+      "learning_rate": 7.79545271456967e-08,
+      "loss": 0.0003,
+      "step": 18466
+    },
+    {
+      "epoch": 4.197045454545455,
+      "grad_norm": 0.5856288842897714,
+      "learning_rate": 7.791136888151328e-08,
+      "loss": 0.0023,
+      "step": 18467
+    },
+    {
+      "epoch": 4.197272727272727,
+      "grad_norm": 1.7135580254063951,
+      "learning_rate": 7.786822177342134e-08,
+      "loss": 0.0026,
+      "step": 18468
+    },
+    {
+      "epoch": 4.1975,
+      "grad_norm": 0.5080970385276767,
+      "learning_rate": 7.782508582230099e-08,
+      "loss": 0.002,
+      "step": 18469
+    },
+    {
+      "epoch": 4.197727272727272,
+      "grad_norm": 1.1708573005995988,
+      "learning_rate": 7.778196102903166e-08,
+      "loss": 0.004,
+      "step": 18470
+    },
+    {
+      "epoch": 4.197954545454546,
+      "grad_norm": 0.20307675980344359,
+      "learning_rate": 7.773884739449288e-08,
+      "loss": 0.0005,
+      "step": 18471
+    },
+    {
+      "epoch": 4.198181818181818,
+      "grad_norm": 1.1844098905532716,
+      "learning_rate": 7.769574491956387e-08,
+      "loss": 0.0046,
+      "step": 18472
+    },
+    {
+      "epoch": 4.198409090909091,
+      "grad_norm": 0.32934177383076824,
+      "learning_rate": 7.765265360512337e-08,
+      "loss": 0.0015,
+      "step": 18473
+    },
+    {
+      "epoch": 4.198636363636363,
+      "grad_norm": 1.6573481614833723,
+      "learning_rate": 7.760957345205009e-08,
+      "loss": 0.0082,
+      "step": 18474
+    },
+    {
+      "epoch": 4.198863636363637,
+      "grad_norm": 2.333882048648306,
+      "learning_rate": 7.756650446122262e-08,
+      "loss": 0.013,
+      "step": 18475
+    },
+    {
+      "epoch": 4.199090909090909,
+      "grad_norm": 0.7829429716017142,
+      "learning_rate": 7.75234466335193e-08,
+      "loss": 0.0037,
+      "step": 18476
+    },
+    {
+      "epoch": 4.199318181818182,
+      "grad_norm": 1.058141467437493,
+      "learning_rate": 7.748039996981787e-08,
+      "loss": 0.007,
+      "step": 18477
+    },
+    {
+      "epoch": 4.199545454545454,
+      "grad_norm": 1.5624627929931143,
+      "learning_rate": 7.74373644709963e-08,
+      "loss": 0.0055,
+      "step": 18478
+    },
+    {
+      "epoch": 4.199772727272728,
+      "grad_norm": 0.613287611700521,
+      "learning_rate": 7.739434013793214e-08,
+      "loss": 0.0025,
+      "step": 18479
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 0.3945086632649101,
+      "learning_rate": 7.735132697150274e-08,
+      "loss": 0.0016,
+      "step": 18480
+    },
+    {
+      "epoch": 4.200227272727273,
+      "grad_norm": 0.20818277541517627,
+      "learning_rate": 7.73083249725854e-08,
+      "loss": 0.0007,
+      "step": 18481
+    },
+    {
+      "epoch": 4.200454545454545,
+      "grad_norm": 1.2668258205122014,
+      "learning_rate": 7.726533414205658e-08,
+      "loss": 0.011,
+      "step": 18482
+    },
+    {
+      "epoch": 4.200681818181819,
+      "grad_norm": 0.6149644573586586,
+      "learning_rate": 7.722235448079325e-08,
+      "loss": 0.0061,
+      "step": 18483
+    },
+    {
+      "epoch": 4.200909090909091,
+      "grad_norm": 1.659720424928277,
+      "learning_rate": 7.717938598967176e-08,
+      "loss": 0.0173,
+      "step": 18484
+    },
+    {
+      "epoch": 4.201136363636364,
+      "grad_norm": 3.4740892835473147,
+      "learning_rate": 7.713642866956826e-08,
+      "loss": 0.0064,
+      "step": 18485
+    },
+    {
+      "epoch": 4.201363636363636,
+      "grad_norm": 0.7243939730834722,
+      "learning_rate": 7.70934825213588e-08,
+      "loss": 0.0045,
+      "step": 18486
+    },
+    {
+      "epoch": 4.201590909090909,
+      "grad_norm": 0.5917332148323148,
+      "learning_rate": 7.705054754591913e-08,
+      "loss": 0.0041,
+      "step": 18487
+    },
+    {
+      "epoch": 4.201818181818182,
+      "grad_norm": 0.9065790844390362,
+      "learning_rate": 7.700762374412474e-08,
+      "loss": 0.0055,
+      "step": 18488
+    },
+    {
+      "epoch": 4.202045454545455,
+      "grad_norm": 3.1667648064248644,
+      "learning_rate": 7.696471111685103e-08,
+      "loss": 0.0146,
+      "step": 18489
+    },
+    {
+      "epoch": 4.202272727272727,
+      "grad_norm": 1.2747779203566802,
+      "learning_rate": 7.692180966497285e-08,
+      "loss": 0.0097,
+      "step": 18490
+    },
+    {
+      "epoch": 4.2025,
+      "grad_norm": 0.9407113215933448,
+      "learning_rate": 7.687891938936514e-08,
+      "loss": 0.0062,
+      "step": 18491
+    },
+    {
+      "epoch": 4.202727272727273,
+      "grad_norm": 1.2830161137809173,
+      "learning_rate": 7.683604029090248e-08,
+      "loss": 0.0035,
+      "step": 18492
+    },
+    {
+      "epoch": 4.202954545454546,
+      "grad_norm": 0.5175512523425415,
+      "learning_rate": 7.679317237045937e-08,
+      "loss": 0.0037,
+      "step": 18493
+    },
+    {
+      "epoch": 4.203181818181818,
+      "grad_norm": 1.4905896725500118,
+      "learning_rate": 7.675031562890997e-08,
+      "loss": 0.0101,
+      "step": 18494
+    },
+    {
+      "epoch": 4.203409090909091,
+      "grad_norm": 0.8782555940494857,
+      "learning_rate": 7.670747006712797e-08,
+      "loss": 0.004,
+      "step": 18495
+    },
+    {
+      "epoch": 4.203636363636364,
+      "grad_norm": 1.0825467505817818,
+      "learning_rate": 7.666463568598716e-08,
+      "loss": 0.004,
+      "step": 18496
+    },
+    {
+      "epoch": 4.203863636363637,
+      "grad_norm": 0.7318481187091429,
+      "learning_rate": 7.662181248636109e-08,
+      "loss": 0.0032,
+      "step": 18497
+    },
+    {
+      "epoch": 4.204090909090909,
+      "grad_norm": 0.42522390491644374,
+      "learning_rate": 7.657900046912296e-08,
+      "loss": 0.0019,
+      "step": 18498
+    },
+    {
+      "epoch": 4.204318181818182,
+      "grad_norm": 0.5870109283313263,
+      "learning_rate": 7.653619963514575e-08,
+      "loss": 0.003,
+      "step": 18499
+    },
+    {
+      "epoch": 4.204545454545454,
+      "grad_norm": 0.4541856401065745,
+      "learning_rate": 7.649340998530236e-08,
+      "loss": 0.0013,
+      "step": 18500
+    },
+    {
+      "epoch": 4.2047727272727276,
+      "grad_norm": 0.888483051805529,
+      "learning_rate": 7.645063152046523e-08,
+      "loss": 0.0034,
+      "step": 18501
+    },
+    {
+      "epoch": 4.205,
+      "grad_norm": 1.036387270818922,
+      "learning_rate": 7.640786424150678e-08,
+      "loss": 0.0013,
+      "step": 18502
+    },
+    {
+      "epoch": 4.205227272727273,
+      "grad_norm": 1.2373727232655856,
+      "learning_rate": 7.636510814929901e-08,
+      "loss": 0.0082,
+      "step": 18503
+    },
+    {
+      "epoch": 4.205454545454545,
+      "grad_norm": 2.675190308233045,
+      "learning_rate": 7.632236324471374e-08,
+      "loss": 0.0062,
+      "step": 18504
+    },
+    {
+      "epoch": 4.2056818181818185,
+      "grad_norm": 1.4546036556630961,
+      "learning_rate": 7.627962952862278e-08,
+      "loss": 0.0032,
+      "step": 18505
+    },
+    {
+      "epoch": 4.205909090909091,
+      "grad_norm": 0.9213951573104917,
+      "learning_rate": 7.623690700189749e-08,
+      "loss": 0.0018,
+      "step": 18506
+    },
+    {
+      "epoch": 4.206136363636364,
+      "grad_norm": 2.459551908790825,
+      "learning_rate": 7.619419566540908e-08,
+      "loss": 0.0126,
+      "step": 18507
+    },
+    {
+      "epoch": 4.206363636363636,
+      "grad_norm": 0.5617161489773999,
+      "learning_rate": 7.615149552002841e-08,
+      "loss": 0.0022,
+      "step": 18508
+    },
+    {
+      "epoch": 4.2065909090909095,
+      "grad_norm": 1.5248755652140924,
+      "learning_rate": 7.61088065666262e-08,
+      "loss": 0.0044,
+      "step": 18509
+    },
+    {
+      "epoch": 4.206818181818182,
+      "grad_norm": 0.5817720424054973,
+      "learning_rate": 7.606612880607308e-08,
+      "loss": 0.0008,
+      "step": 18510
+    },
+    {
+      "epoch": 4.2070454545454545,
+      "grad_norm": 1.173797369986704,
+      "learning_rate": 7.602346223923939e-08,
+      "loss": 0.004,
+      "step": 18511
+    },
+    {
+      "epoch": 4.207272727272727,
+      "grad_norm": 0.5395418482947939,
+      "learning_rate": 7.598080686699482e-08,
+      "loss": 0.0015,
+      "step": 18512
+    },
+    {
+      "epoch": 4.2075,
+      "grad_norm": 0.1977276107919317,
+      "learning_rate": 7.593816269020949e-08,
+      "loss": 0.0005,
+      "step": 18513
+    },
+    {
+      "epoch": 4.207727272727273,
+      "grad_norm": 1.0715579072344763,
+      "learning_rate": 7.589552970975293e-08,
+      "loss": 0.0032,
+      "step": 18514
+    },
+    {
+      "epoch": 4.2079545454545455,
+      "grad_norm": 0.9186957563471713,
+      "learning_rate": 7.585290792649448e-08,
+      "loss": 0.0046,
+      "step": 18515
+    },
+    {
+      "epoch": 4.208181818181818,
+      "grad_norm": 0.23595830839440154,
+      "learning_rate": 7.581029734130322e-08,
+      "loss": 0.001,
+      "step": 18516
+    },
+    {
+      "epoch": 4.2084090909090905,
+      "grad_norm": 0.3512499386233867,
+      "learning_rate": 7.57676979550482e-08,
+      "loss": 0.0018,
+      "step": 18517
+    },
+    {
+      "epoch": 4.208636363636364,
+      "grad_norm": 0.38184547288164966,
+      "learning_rate": 7.572510976859793e-08,
+      "loss": 0.0014,
+      "step": 18518
+    },
+    {
+      "epoch": 4.2088636363636365,
+      "grad_norm": 1.4488984092238326,
+      "learning_rate": 7.568253278282097e-08,
+      "loss": 0.0049,
+      "step": 18519
+    },
+    {
+      "epoch": 4.209090909090909,
+      "grad_norm": 0.5493789215531834,
+      "learning_rate": 7.563996699858563e-08,
+      "loss": 0.0019,
+      "step": 18520
+    },
+    {
+      "epoch": 4.2093181818181815,
+      "grad_norm": 1.7066747561692936,
+      "learning_rate": 7.55974124167596e-08,
+      "loss": 0.0182,
+      "step": 18521
+    },
+    {
+      "epoch": 4.209545454545455,
+      "grad_norm": 0.7444712950368018,
+      "learning_rate": 7.555486903821088e-08,
+      "loss": 0.0044,
+      "step": 18522
+    },
+    {
+      "epoch": 4.209772727272727,
+      "grad_norm": 0.3834426317764744,
+      "learning_rate": 7.551233686380694e-08,
+      "loss": 0.0017,
+      "step": 18523
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 1.5635513553578233,
+      "learning_rate": 7.546981589441518e-08,
+      "loss": 0.014,
+      "step": 18524
+    },
+    {
+      "epoch": 4.2102272727272725,
+      "grad_norm": 1.3070945898182071,
+      "learning_rate": 7.542730613090243e-08,
+      "loss": 0.0047,
+      "step": 18525
+    },
+    {
+      "epoch": 4.210454545454546,
+      "grad_norm": 1.6080466682553802,
+      "learning_rate": 7.53848075741357e-08,
+      "loss": 0.0099,
+      "step": 18526
+    },
+    {
+      "epoch": 4.210681818181818,
+      "grad_norm": 1.2652290313787378,
+      "learning_rate": 7.534232022498165e-08,
+      "loss": 0.0066,
+      "step": 18527
+    },
+    {
+      "epoch": 4.210909090909091,
+      "grad_norm": 0.7272446905205081,
+      "learning_rate": 7.529984408430661e-08,
+      "loss": 0.0035,
+      "step": 18528
+    },
+    {
+      "epoch": 4.2111363636363635,
+      "grad_norm": 1.0078823577876626,
+      "learning_rate": 7.525737915297669e-08,
+      "loss": 0.0052,
+      "step": 18529
+    },
+    {
+      "epoch": 4.211363636363636,
+      "grad_norm": 0.7491041458570048,
+      "learning_rate": 7.5214925431858e-08,
+      "loss": 0.0039,
+      "step": 18530
+    },
+    {
+      "epoch": 4.211590909090909,
+      "grad_norm": 1.4030386054109922,
+      "learning_rate": 7.517248292181602e-08,
+      "loss": 0.0026,
+      "step": 18531
+    },
+    {
+      "epoch": 4.211818181818182,
+      "grad_norm": 1.0059347303974013,
+      "learning_rate": 7.51300516237164e-08,
+      "loss": 0.0044,
+      "step": 18532
+    },
+    {
+      "epoch": 4.212045454545454,
+      "grad_norm": 1.0922239474880786,
+      "learning_rate": 7.508763153842448e-08,
+      "loss": 0.0064,
+      "step": 18533
+    },
+    {
+      "epoch": 4.212272727272727,
+      "grad_norm": 2.67664251227009,
+      "learning_rate": 7.504522266680491e-08,
+      "loss": 0.0124,
+      "step": 18534
+    },
+    {
+      "epoch": 4.2125,
+      "grad_norm": 2.597197771741436,
+      "learning_rate": 7.500282500972275e-08,
+      "loss": 0.0074,
+      "step": 18535
+    },
+    {
+      "epoch": 4.212727272727273,
+      "grad_norm": 1.222339500476602,
+      "learning_rate": 7.496043856804253e-08,
+      "loss": 0.0026,
+      "step": 18536
+    },
+    {
+      "epoch": 4.212954545454545,
+      "grad_norm": 1.1806964052934017,
+      "learning_rate": 7.491806334262857e-08,
+      "loss": 0.0055,
+      "step": 18537
+    },
+    {
+      "epoch": 4.213181818181818,
+      "grad_norm": 0.13558923384192995,
+      "learning_rate": 7.487569933434504e-08,
+      "loss": 0.0004,
+      "step": 18538
+    },
+    {
+      "epoch": 4.213409090909091,
+      "grad_norm": 0.8956636626539566,
+      "learning_rate": 7.483334654405566e-08,
+      "loss": 0.0028,
+      "step": 18539
+    },
+    {
+      "epoch": 4.213636363636364,
+      "grad_norm": 0.6575942752213243,
+      "learning_rate": 7.479100497262415e-08,
+      "loss": 0.0034,
+      "step": 18540
+    },
+    {
+      "epoch": 4.213863636363636,
+      "grad_norm": 0.2636353148606192,
+      "learning_rate": 7.474867462091397e-08,
+      "loss": 0.0012,
+      "step": 18541
+    },
+    {
+      "epoch": 4.214090909090909,
+      "grad_norm": 0.9524442956655418,
+      "learning_rate": 7.470635548978838e-08,
+      "loss": 0.0044,
+      "step": 18542
+    },
+    {
+      "epoch": 4.214318181818181,
+      "grad_norm": 0.535645364244061,
+      "learning_rate": 7.466404758011011e-08,
+      "loss": 0.002,
+      "step": 18543
+    },
+    {
+      "epoch": 4.214545454545455,
+      "grad_norm": 1.4228141134455747,
+      "learning_rate": 7.4621750892742e-08,
+      "loss": 0.0088,
+      "step": 18544
+    },
+    {
+      "epoch": 4.214772727272727,
+      "grad_norm": 0.6022915532445109,
+      "learning_rate": 7.457946542854665e-08,
+      "loss": 0.0035,
+      "step": 18545
+    },
+    {
+      "epoch": 4.215,
+      "grad_norm": 1.797135326735456,
+      "learning_rate": 7.453719118838618e-08,
+      "loss": 0.0104,
+      "step": 18546
+    },
+    {
+      "epoch": 4.215227272727272,
+      "grad_norm": 0.6487141981128898,
+      "learning_rate": 7.449492817312276e-08,
+      "loss": 0.0022,
+      "step": 18547
+    },
+    {
+      "epoch": 4.215454545454546,
+      "grad_norm": 0.8883763087858814,
+      "learning_rate": 7.445267638361814e-08,
+      "loss": 0.0028,
+      "step": 18548
+    },
+    {
+      "epoch": 4.215681818181818,
+      "grad_norm": 0.7479691057994936,
+      "learning_rate": 7.441043582073398e-08,
+      "loss": 0.0031,
+      "step": 18549
+    },
+    {
+      "epoch": 4.215909090909091,
+      "grad_norm": 0.2699813618248034,
+      "learning_rate": 7.436820648533155e-08,
+      "loss": 0.0014,
+      "step": 18550
+    },
+    {
+      "epoch": 4.216136363636363,
+      "grad_norm": 5.525813808673653,
+      "learning_rate": 7.432598837827217e-08,
+      "loss": 0.0098,
+      "step": 18551
+    },
+    {
+      "epoch": 4.216363636363637,
+      "grad_norm": 0.44674975434940645,
+      "learning_rate": 7.428378150041646e-08,
+      "loss": 0.0017,
+      "step": 18552
+    },
+    {
+      "epoch": 4.216590909090909,
+      "grad_norm": 0.8366880559626938,
+      "learning_rate": 7.424158585262519e-08,
+      "loss": 0.0019,
+      "step": 18553
+    },
+    {
+      "epoch": 4.216818181818182,
+      "grad_norm": 0.7571378371727074,
+      "learning_rate": 7.419940143575888e-08,
+      "loss": 0.0021,
+      "step": 18554
+    },
+    {
+      "epoch": 4.217045454545454,
+      "grad_norm": 0.6916367938219117,
+      "learning_rate": 7.415722825067779e-08,
+      "loss": 0.0035,
+      "step": 18555
+    },
+    {
+      "epoch": 4.217272727272727,
+      "grad_norm": 1.3701568776796542,
+      "learning_rate": 7.41150662982417e-08,
+      "loss": 0.0039,
+      "step": 18556
+    },
+    {
+      "epoch": 4.2175,
+      "grad_norm": 3.4713281738990007,
+      "learning_rate": 7.407291557931045e-08,
+      "loss": 0.0189,
+      "step": 18557
+    },
+    {
+      "epoch": 4.217727272727273,
+      "grad_norm": 0.3384381522726909,
+      "learning_rate": 7.403077609474368e-08,
+      "loss": 0.0011,
+      "step": 18558
+    },
+    {
+      "epoch": 4.217954545454545,
+      "grad_norm": 0.4978228316637448,
+      "learning_rate": 7.39886478454005e-08,
+      "loss": 0.0019,
+      "step": 18559
+    },
+    {
+      "epoch": 4.218181818181818,
+      "grad_norm": 2.024436858211643,
+      "learning_rate": 7.394653083214013e-08,
+      "loss": 0.0119,
+      "step": 18560
+    },
+    {
+      "epoch": 4.218409090909091,
+      "grad_norm": 1.9594028437349933,
+      "learning_rate": 7.390442505582137e-08,
+      "loss": 0.0125,
+      "step": 18561
+    },
+    {
+      "epoch": 4.218636363636364,
+      "grad_norm": 0.43669037533608585,
+      "learning_rate": 7.386233051730287e-08,
+      "loss": 0.0013,
+      "step": 18562
+    },
+    {
+      "epoch": 4.218863636363636,
+      "grad_norm": 0.490575140114271,
+      "learning_rate": 7.382024721744284e-08,
+      "loss": 0.0011,
+      "step": 18563
+    },
+    {
+      "epoch": 4.219090909090909,
+      "grad_norm": 0.25725366611542,
+      "learning_rate": 7.377817515709977e-08,
+      "loss": 0.001,
+      "step": 18564
+    },
+    {
+      "epoch": 4.219318181818182,
+      "grad_norm": 0.4491865758511594,
+      "learning_rate": 7.373611433713123e-08,
+      "loss": 0.0017,
+      "step": 18565
+    },
+    {
+      "epoch": 4.219545454545455,
+      "grad_norm": 1.668303245222958,
+      "learning_rate": 7.369406475839502e-08,
+      "loss": 0.0032,
+      "step": 18566
+    },
+    {
+      "epoch": 4.219772727272727,
+      "grad_norm": 1.4777213994812628,
+      "learning_rate": 7.365202642174871e-08,
+      "loss": 0.0056,
+      "step": 18567
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 2.333587465657223,
+      "learning_rate": 7.360999932804953e-08,
+      "loss": 0.0089,
+      "step": 18568
+    },
+    {
+      "epoch": 4.220227272727273,
+      "grad_norm": 1.1793840568114038,
+      "learning_rate": 7.356798347815427e-08,
+      "loss": 0.0035,
+      "step": 18569
+    },
+    {
+      "epoch": 4.220454545454546,
+      "grad_norm": 0.5812160614807824,
+      "learning_rate": 7.352597887291985e-08,
+      "loss": 0.0037,
+      "step": 18570
+    },
+    {
+      "epoch": 4.220681818181818,
+      "grad_norm": 0.22421046046452453,
+      "learning_rate": 7.348398551320287e-08,
+      "loss": 0.0011,
+      "step": 18571
+    },
+    {
+      "epoch": 4.220909090909091,
+      "grad_norm": 1.958113984940879,
+      "learning_rate": 7.344200339985959e-08,
+      "loss": 0.0062,
+      "step": 18572
+    },
+    {
+      "epoch": 4.221136363636363,
+      "grad_norm": 0.3816560577502344,
+      "learning_rate": 7.340003253374625e-08,
+      "loss": 0.0017,
+      "step": 18573
+    },
+    {
+      "epoch": 4.221363636363637,
+      "grad_norm": 1.7032661491299095,
+      "learning_rate": 7.335807291571837e-08,
+      "loss": 0.006,
+      "step": 18574
+    },
+    {
+      "epoch": 4.221590909090909,
+      "grad_norm": 0.7762456790291364,
+      "learning_rate": 7.331612454663184e-08,
+      "loss": 0.003,
+      "step": 18575
+    },
+    {
+      "epoch": 4.221818181818182,
+      "grad_norm": 0.5665441439232004,
+      "learning_rate": 7.327418742734197e-08,
+      "loss": 0.0017,
+      "step": 18576
+    },
+    {
+      "epoch": 4.222045454545454,
+      "grad_norm": 1.0490634617803118,
+      "learning_rate": 7.323226155870396e-08,
+      "loss": 0.0044,
+      "step": 18577
+    },
+    {
+      "epoch": 4.222272727272728,
+      "grad_norm": 0.9519073072042794,
+      "learning_rate": 7.319034694157276e-08,
+      "loss": 0.006,
+      "step": 18578
+    },
+    {
+      "epoch": 4.2225,
+      "grad_norm": 0.6174503196531008,
+      "learning_rate": 7.314844357680304e-08,
+      "loss": 0.0036,
+      "step": 18579
+    },
+    {
+      "epoch": 4.222727272727273,
+      "grad_norm": 1.2692056916020829,
+      "learning_rate": 7.31065514652493e-08,
+      "loss": 0.008,
+      "step": 18580
+    },
+    {
+      "epoch": 4.222954545454545,
+      "grad_norm": 2.1787418690198734,
+      "learning_rate": 7.306467060776583e-08,
+      "loss": 0.0079,
+      "step": 18581
+    },
+    {
+      "epoch": 4.223181818181819,
+      "grad_norm": 1.6379999825889164,
+      "learning_rate": 7.302280100520678e-08,
+      "loss": 0.0082,
+      "step": 18582
+    },
+    {
+      "epoch": 4.223409090909091,
+      "grad_norm": 0.23344352281650108,
+      "learning_rate": 7.298094265842561e-08,
+      "loss": 0.0003,
+      "step": 18583
+    },
+    {
+      "epoch": 4.223636363636364,
+      "grad_norm": 0.9750892937032357,
+      "learning_rate": 7.293909556827614e-08,
+      "loss": 0.0029,
+      "step": 18584
+    },
+    {
+      "epoch": 4.223863636363636,
+      "grad_norm": 0.8149333643679842,
+      "learning_rate": 7.28972597356116e-08,
+      "loss": 0.001,
+      "step": 18585
+    },
+    {
+      "epoch": 4.224090909090909,
+      "grad_norm": 1.5910793471237337,
+      "learning_rate": 7.285543516128528e-08,
+      "loss": 0.0087,
+      "step": 18586
+    },
+    {
+      "epoch": 4.224318181818182,
+      "grad_norm": 1.1994915632734295,
+      "learning_rate": 7.281362184614976e-08,
+      "loss": 0.0032,
+      "step": 18587
+    },
+    {
+      "epoch": 4.224545454545455,
+      "grad_norm": 1.1076224995446096,
+      "learning_rate": 7.277181979105784e-08,
+      "loss": 0.003,
+      "step": 18588
+    },
+    {
+      "epoch": 4.224772727272727,
+      "grad_norm": 2.0791939005225073,
+      "learning_rate": 7.273002899686188e-08,
+      "loss": 0.0049,
+      "step": 18589
+    },
+    {
+      "epoch": 4.225,
+      "grad_norm": 1.3905438310215827,
+      "learning_rate": 7.268824946441419e-08,
+      "loss": 0.0075,
+      "step": 18590
+    },
+    {
+      "epoch": 4.225227272727273,
+      "grad_norm": 0.5967351487907976,
+      "learning_rate": 7.264648119456659e-08,
+      "loss": 0.002,
+      "step": 18591
+    },
+    {
+      "epoch": 4.225454545454546,
+      "grad_norm": 1.513351345539386,
+      "learning_rate": 7.260472418817096e-08,
+      "loss": 0.0053,
+      "step": 18592
+    },
+    {
+      "epoch": 4.225681818181818,
+      "grad_norm": 1.073525567783954,
+      "learning_rate": 7.256297844607865e-08,
+      "loss": 0.0043,
+      "step": 18593
+    },
+    {
+      "epoch": 4.225909090909091,
+      "grad_norm": 1.457124294991254,
+      "learning_rate": 7.252124396914098e-08,
+      "loss": 0.0107,
+      "step": 18594
+    },
+    {
+      "epoch": 4.226136363636364,
+      "grad_norm": 1.1297171704990685,
+      "learning_rate": 7.247952075820915e-08,
+      "loss": 0.0051,
+      "step": 18595
+    },
+    {
+      "epoch": 4.2263636363636365,
+      "grad_norm": 0.525072098209144,
+      "learning_rate": 7.243780881413365e-08,
+      "loss": 0.0027,
+      "step": 18596
+    },
+    {
+      "epoch": 4.226590909090909,
+      "grad_norm": 1.3226434488860843,
+      "learning_rate": 7.239610813776527e-08,
+      "loss": 0.0037,
+      "step": 18597
+    },
+    {
+      "epoch": 4.226818181818182,
+      "grad_norm": 0.6331868690005417,
+      "learning_rate": 7.235441872995429e-08,
+      "loss": 0.0029,
+      "step": 18598
+    },
+    {
+      "epoch": 4.227045454545454,
+      "grad_norm": 1.0850764101098158,
+      "learning_rate": 7.2312740591551e-08,
+      "loss": 0.0062,
+      "step": 18599
+    },
+    {
+      "epoch": 4.2272727272727275,
+      "grad_norm": 0.3027581021932005,
+      "learning_rate": 7.227107372340503e-08,
+      "loss": 0.0023,
+      "step": 18600
+    },
+    {
+      "epoch": 4.2275,
+      "grad_norm": 0.7198438169599578,
+      "learning_rate": 7.222941812636608e-08,
+      "loss": 0.0042,
+      "step": 18601
+    },
+    {
+      "epoch": 4.227727272727273,
+      "grad_norm": 0.7448551253271823,
+      "learning_rate": 7.218777380128377e-08,
+      "loss": 0.0039,
+      "step": 18602
+    },
+    {
+      "epoch": 4.227954545454545,
+      "grad_norm": 1.8890104706872637,
+      "learning_rate": 7.214614074900713e-08,
+      "loss": 0.0039,
+      "step": 18603
+    },
+    {
+      "epoch": 4.2281818181818185,
+      "grad_norm": 1.9331019949602317,
+      "learning_rate": 7.210451897038526e-08,
+      "loss": 0.0054,
+      "step": 18604
+    },
+    {
+      "epoch": 4.228409090909091,
+      "grad_norm": 1.245848292150106,
+      "learning_rate": 7.206290846626677e-08,
+      "loss": 0.0055,
+      "step": 18605
+    },
+    {
+      "epoch": 4.2286363636363635,
+      "grad_norm": 0.9443633252055152,
+      "learning_rate": 7.202130923750009e-08,
+      "loss": 0.0037,
+      "step": 18606
+    },
+    {
+      "epoch": 4.228863636363636,
+      "grad_norm": 0.44488119616852806,
+      "learning_rate": 7.197972128493374e-08,
+      "loss": 0.002,
+      "step": 18607
+    },
+    {
+      "epoch": 4.2290909090909095,
+      "grad_norm": 0.3912574498748782,
+      "learning_rate": 7.193814460941584e-08,
+      "loss": 0.0018,
+      "step": 18608
+    },
+    {
+      "epoch": 4.229318181818182,
+      "grad_norm": 6.071909368366847,
+      "learning_rate": 7.189657921179386e-08,
+      "loss": 0.005,
+      "step": 18609
+    },
+    {
+      "epoch": 4.2295454545454545,
+      "grad_norm": 5.371173349675699,
+      "learning_rate": 7.185502509291557e-08,
+      "loss": 0.0106,
+      "step": 18610
+    },
+    {
+      "epoch": 4.229772727272727,
+      "grad_norm": 0.20747821517168094,
+      "learning_rate": 7.181348225362837e-08,
+      "loss": 0.0006,
+      "step": 18611
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 0.4072754755656923,
+      "learning_rate": 7.17719506947795e-08,
+      "loss": 0.001,
+      "step": 18612
+    },
+    {
+      "epoch": 4.230227272727273,
+      "grad_norm": 1.4422439930288313,
+      "learning_rate": 7.173043041721555e-08,
+      "loss": 0.0099,
+      "step": 18613
+    },
+    {
+      "epoch": 4.2304545454545455,
+      "grad_norm": 0.37792298756691844,
+      "learning_rate": 7.168892142178336e-08,
+      "loss": 0.0014,
+      "step": 18614
+    },
+    {
+      "epoch": 4.230681818181818,
+      "grad_norm": 1.0773504134730805,
+      "learning_rate": 7.164742370932939e-08,
+      "loss": 0.0027,
+      "step": 18615
+    },
+    {
+      "epoch": 4.2309090909090905,
+      "grad_norm": 3.043179368307199,
+      "learning_rate": 7.160593728069979e-08,
+      "loss": 0.0182,
+      "step": 18616
+    },
+    {
+      "epoch": 4.231136363636364,
+      "grad_norm": 1.0567004313692703,
+      "learning_rate": 7.156446213674076e-08,
+      "loss": 0.0052,
+      "step": 18617
+    },
+    {
+      "epoch": 4.231363636363636,
+      "grad_norm": 0.922687152998737,
+      "learning_rate": 7.152299827829771e-08,
+      "loss": 0.0046,
+      "step": 18618
+    },
+    {
+      "epoch": 4.231590909090909,
+      "grad_norm": 2.722292312291806,
+      "learning_rate": 7.148154570621631e-08,
+      "loss": 0.0221,
+      "step": 18619
+    },
+    {
+      "epoch": 4.2318181818181815,
+      "grad_norm": 2.0120056438321985,
+      "learning_rate": 7.144010442134184e-08,
+      "loss": 0.0092,
+      "step": 18620
+    },
+    {
+      "epoch": 4.232045454545455,
+      "grad_norm": 0.3406251968041876,
+      "learning_rate": 7.139867442451939e-08,
+      "loss": 0.0007,
+      "step": 18621
+    },
+    {
+      "epoch": 4.232272727272727,
+      "grad_norm": 0.5837643396415889,
+      "learning_rate": 7.135725571659382e-08,
+      "loss": 0.0034,
+      "step": 18622
+    },
+    {
+      "epoch": 4.2325,
+      "grad_norm": 0.9325005864598362,
+      "learning_rate": 7.131584829840963e-08,
+      "loss": 0.0035,
+      "step": 18623
+    },
+    {
+      "epoch": 4.2327272727272724,
+      "grad_norm": 0.33618036354098285,
+      "learning_rate": 7.127445217081128e-08,
+      "loss": 0.0011,
+      "step": 18624
+    },
+    {
+      "epoch": 4.232954545454546,
+      "grad_norm": 0.6212761877840213,
+      "learning_rate": 7.123306733464286e-08,
+      "loss": 0.0012,
+      "step": 18625
+    },
+    {
+      "epoch": 4.233181818181818,
+      "grad_norm": 0.3785430804548767,
+      "learning_rate": 7.11916937907485e-08,
+      "loss": 0.0012,
+      "step": 18626
+    },
+    {
+      "epoch": 4.233409090909091,
+      "grad_norm": 2.7751337892244514,
+      "learning_rate": 7.115033153997148e-08,
+      "loss": 0.0273,
+      "step": 18627
+    },
+    {
+      "epoch": 4.233636363636363,
+      "grad_norm": 1.2994941773576991,
+      "learning_rate": 7.110898058315545e-08,
+      "loss": 0.0046,
+      "step": 18628
+    },
+    {
+      "epoch": 4.233863636363636,
+      "grad_norm": 0.9511684757715709,
+      "learning_rate": 7.10676409211436e-08,
+      "loss": 0.0033,
+      "step": 18629
+    },
+    {
+      "epoch": 4.234090909090909,
+      "grad_norm": 0.6663628653050926,
+      "learning_rate": 7.102631255477909e-08,
+      "loss": 0.0024,
+      "step": 18630
+    },
+    {
+      "epoch": 4.234318181818182,
+      "grad_norm": 0.9545141237522997,
+      "learning_rate": 7.098499548490442e-08,
+      "loss": 0.0025,
+      "step": 18631
+    },
+    {
+      "epoch": 4.234545454545454,
+      "grad_norm": 0.752445569919558,
+      "learning_rate": 7.094368971236221e-08,
+      "loss": 0.0021,
+      "step": 18632
+    },
+    {
+      "epoch": 4.234772727272727,
+      "grad_norm": 0.9508322261606545,
+      "learning_rate": 7.090239523799475e-08,
+      "loss": 0.0068,
+      "step": 18633
+    },
+    {
+      "epoch": 4.235,
+      "grad_norm": 0.4766061091847654,
+      "learning_rate": 7.086111206264429e-08,
+      "loss": 0.0026,
+      "step": 18634
+    },
+    {
+      "epoch": 4.235227272727273,
+      "grad_norm": 1.1534922416425488,
+      "learning_rate": 7.081984018715216e-08,
+      "loss": 0.0042,
+      "step": 18635
+    },
+    {
+      "epoch": 4.235454545454545,
+      "grad_norm": 0.37101417697108197,
+      "learning_rate": 7.077857961236056e-08,
+      "loss": 0.001,
+      "step": 18636
+    },
+    {
+      "epoch": 4.235681818181818,
+      "grad_norm": 0.4134770657429357,
+      "learning_rate": 7.073733033911054e-08,
+      "loss": 0.0017,
+      "step": 18637
+    },
+    {
+      "epoch": 4.235909090909091,
+      "grad_norm": 1.826957125976424,
+      "learning_rate": 7.069609236824335e-08,
+      "loss": 0.0044,
+      "step": 18638
+    },
+    {
+      "epoch": 4.236136363636364,
+      "grad_norm": 0.6745692283992357,
+      "learning_rate": 7.065486570059996e-08,
+      "loss": 0.0041,
+      "step": 18639
+    },
+    {
+      "epoch": 4.236363636363636,
+      "grad_norm": 0.6910838713930482,
+      "learning_rate": 7.06136503370209e-08,
+      "loss": 0.0028,
+      "step": 18640
+    },
+    {
+      "epoch": 4.236590909090909,
+      "grad_norm": 1.1766127033294862,
+      "learning_rate": 7.057244627834664e-08,
+      "loss": 0.0055,
+      "step": 18641
+    },
+    {
+      "epoch": 4.236818181818181,
+      "grad_norm": 0.820275339196062,
+      "learning_rate": 7.053125352541746e-08,
+      "loss": 0.0029,
+      "step": 18642
+    },
+    {
+      "epoch": 4.237045454545455,
+      "grad_norm": 2.8491560482852094,
+      "learning_rate": 7.049007207907348e-08,
+      "loss": 0.0172,
+      "step": 18643
+    },
+    {
+      "epoch": 4.237272727272727,
+      "grad_norm": 0.6443026667130419,
+      "learning_rate": 7.044890194015421e-08,
+      "loss": 0.0023,
+      "step": 18644
+    },
+    {
+      "epoch": 4.2375,
+      "grad_norm": 0.9509532801461257,
+      "learning_rate": 7.040774310949926e-08,
+      "loss": 0.0028,
+      "step": 18645
+    },
+    {
+      "epoch": 4.237727272727272,
+      "grad_norm": 1.1615956638411091,
+      "learning_rate": 7.036659558794804e-08,
+      "loss": 0.002,
+      "step": 18646
+    },
+    {
+      "epoch": 4.237954545454546,
+      "grad_norm": 0.6242814431361299,
+      "learning_rate": 7.032545937633958e-08,
+      "loss": 0.0017,
+      "step": 18647
+    },
+    {
+      "epoch": 4.238181818181818,
+      "grad_norm": 1.5705511175793498,
+      "learning_rate": 7.028433447551261e-08,
+      "loss": 0.0035,
+      "step": 18648
+    },
+    {
+      "epoch": 4.238409090909091,
+      "grad_norm": 1.0345500130021599,
+      "learning_rate": 7.024322088630582e-08,
+      "loss": 0.0046,
+      "step": 18649
+    },
+    {
+      "epoch": 4.238636363636363,
+      "grad_norm": 0.6436616584326553,
+      "learning_rate": 7.020211860955757e-08,
+      "loss": 0.0019,
+      "step": 18650
+    },
+    {
+      "epoch": 4.238863636363637,
+      "grad_norm": 0.34003428285071235,
+      "learning_rate": 7.0161027646106e-08,
+      "loss": 0.0014,
+      "step": 18651
+    },
+    {
+      "epoch": 4.239090909090909,
+      "grad_norm": 1.4027390997177287,
+      "learning_rate": 7.011994799678908e-08,
+      "loss": 0.0026,
+      "step": 18652
+    },
+    {
+      "epoch": 4.239318181818182,
+      "grad_norm": 0.18865451594829066,
+      "learning_rate": 7.007887966244443e-08,
+      "loss": 0.0008,
+      "step": 18653
+    },
+    {
+      "epoch": 4.239545454545454,
+      "grad_norm": 2.4469542691154462,
+      "learning_rate": 7.003782264390961e-08,
+      "loss": 0.0175,
+      "step": 18654
+    },
+    {
+      "epoch": 4.239772727272728,
+      "grad_norm": 0.4928426026252541,
+      "learning_rate": 6.999677694202172e-08,
+      "loss": 0.0025,
+      "step": 18655
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 0.5062685686980289,
+      "learning_rate": 6.995574255761792e-08,
+      "loss": 0.0014,
+      "step": 18656
+    },
+    {
+      "epoch": 4.240227272727273,
+      "grad_norm": 0.36330280277472277,
+      "learning_rate": 6.99147194915347e-08,
+      "loss": 0.0015,
+      "step": 18657
+    },
+    {
+      "epoch": 4.240454545454545,
+      "grad_norm": 1.9929166891418237,
+      "learning_rate": 6.987370774460877e-08,
+      "loss": 0.0145,
+      "step": 18658
+    },
+    {
+      "epoch": 4.240681818181818,
+      "grad_norm": 1.370825380366018,
+      "learning_rate": 6.983270731767647e-08,
+      "loss": 0.0021,
+      "step": 18659
+    },
+    {
+      "epoch": 4.240909090909091,
+      "grad_norm": 0.8444586653655829,
+      "learning_rate": 6.979171821157379e-08,
+      "loss": 0.006,
+      "step": 18660
+    },
+    {
+      "epoch": 4.241136363636364,
+      "grad_norm": 3.376645705927254,
+      "learning_rate": 6.975074042713664e-08,
+      "loss": 0.0322,
+      "step": 18661
+    },
+    {
+      "epoch": 4.241363636363636,
+      "grad_norm": 0.991664791713044,
+      "learning_rate": 6.970977396520057e-08,
+      "loss": 0.0068,
+      "step": 18662
+    },
+    {
+      "epoch": 4.241590909090909,
+      "grad_norm": 0.4670606850552082,
+      "learning_rate": 6.966881882660086e-08,
+      "loss": 0.0012,
+      "step": 18663
+    },
+    {
+      "epoch": 4.241818181818182,
+      "grad_norm": 4.314396714562345,
+      "learning_rate": 6.962787501217282e-08,
+      "loss": 0.0095,
+      "step": 18664
+    },
+    {
+      "epoch": 4.242045454545455,
+      "grad_norm": 0.35418837285888294,
+      "learning_rate": 6.958694252275127e-08,
+      "loss": 0.0012,
+      "step": 18665
+    },
+    {
+      "epoch": 4.242272727272727,
+      "grad_norm": 3.311466775667751,
+      "learning_rate": 6.954602135917094e-08,
+      "loss": 0.0081,
+      "step": 18666
+    },
+    {
+      "epoch": 4.2425,
+      "grad_norm": 3.597890200249882,
+      "learning_rate": 6.950511152226635e-08,
+      "loss": 0.0303,
+      "step": 18667
+    },
+    {
+      "epoch": 4.242727272727273,
+      "grad_norm": 0.4533917129301661,
+      "learning_rate": 6.946421301287156e-08,
+      "loss": 0.0018,
+      "step": 18668
+    },
+    {
+      "epoch": 4.242954545454546,
+      "grad_norm": 0.6981268304744094,
+      "learning_rate": 6.942332583182082e-08,
+      "loss": 0.0024,
+      "step": 18669
+    },
+    {
+      "epoch": 4.243181818181818,
+      "grad_norm": 0.5838413548452491,
+      "learning_rate": 6.93824499799476e-08,
+      "loss": 0.0026,
+      "step": 18670
+    },
+    {
+      "epoch": 4.243409090909091,
+      "grad_norm": 0.817505791928124,
+      "learning_rate": 6.934158545808555e-08,
+      "loss": 0.0036,
+      "step": 18671
+    },
+    {
+      "epoch": 4.243636363636363,
+      "grad_norm": 0.22453405934930867,
+      "learning_rate": 6.930073226706795e-08,
+      "loss": 0.0007,
+      "step": 18672
+    },
+    {
+      "epoch": 4.243863636363637,
+      "grad_norm": 1.340804150138386,
+      "learning_rate": 6.925989040772788e-08,
+      "loss": 0.0048,
+      "step": 18673
+    },
+    {
+      "epoch": 4.244090909090909,
+      "grad_norm": 1.282213295660623,
+      "learning_rate": 6.921905988089831e-08,
+      "loss": 0.0067,
+      "step": 18674
+    },
+    {
+      "epoch": 4.244318181818182,
+      "grad_norm": 0.17606900929686517,
+      "learning_rate": 6.917824068741169e-08,
+      "loss": 0.0005,
+      "step": 18675
+    },
+    {
+      "epoch": 4.244545454545454,
+      "grad_norm": 0.9179590686599726,
+      "learning_rate": 6.913743282810033e-08,
+      "loss": 0.0033,
+      "step": 18676
+    },
+    {
+      "epoch": 4.244772727272728,
+      "grad_norm": 1.0320389632761795,
+      "learning_rate": 6.909663630379648e-08,
+      "loss": 0.0053,
+      "step": 18677
+    },
+    {
+      "epoch": 4.245,
+      "grad_norm": 0.5914081442991704,
+      "learning_rate": 6.90558511153322e-08,
+      "loss": 0.0024,
+      "step": 18678
+    },
+    {
+      "epoch": 4.245227272727273,
+      "grad_norm": 1.016988512917482,
+      "learning_rate": 6.901507726353888e-08,
+      "loss": 0.002,
+      "step": 18679
+    },
+    {
+      "epoch": 4.245454545454545,
+      "grad_norm": 0.7560708671929127,
+      "learning_rate": 6.897431474924811e-08,
+      "loss": 0.0018,
+      "step": 18680
+    },
+    {
+      "epoch": 4.245681818181819,
+      "grad_norm": 4.050629545600955,
+      "learning_rate": 6.893356357329108e-08,
+      "loss": 0.0225,
+      "step": 18681
+    },
+    {
+      "epoch": 4.245909090909091,
+      "grad_norm": 0.8204499821223732,
+      "learning_rate": 6.889282373649879e-08,
+      "loss": 0.0046,
+      "step": 18682
+    },
+    {
+      "epoch": 4.246136363636364,
+      "grad_norm": 1.5225389428923024,
+      "learning_rate": 6.885209523970204e-08,
+      "loss": 0.0101,
+      "step": 18683
+    },
+    {
+      "epoch": 4.246363636363636,
+      "grad_norm": 1.1082506676699189,
+      "learning_rate": 6.881137808373129e-08,
+      "loss": 0.0085,
+      "step": 18684
+    },
+    {
+      "epoch": 4.246590909090909,
+      "grad_norm": 0.3615955283896837,
+      "learning_rate": 6.877067226941688e-08,
+      "loss": 0.0013,
+      "step": 18685
+    },
+    {
+      "epoch": 4.246818181818182,
+      "grad_norm": 1.65395179727682,
+      "learning_rate": 6.872997779758888e-08,
+      "loss": 0.0043,
+      "step": 18686
+    },
+    {
+      "epoch": 4.247045454545455,
+      "grad_norm": 2.1955078333543967,
+      "learning_rate": 6.868929466907717e-08,
+      "loss": 0.0089,
+      "step": 18687
+    },
+    {
+      "epoch": 4.247272727272727,
+      "grad_norm": 2.9874256163847153,
+      "learning_rate": 6.864862288471122e-08,
+      "loss": 0.0201,
+      "step": 18688
+    },
+    {
+      "epoch": 4.2475,
+      "grad_norm": 1.0520285837154777,
+      "learning_rate": 6.860796244532043e-08,
+      "loss": 0.0065,
+      "step": 18689
+    },
+    {
+      "epoch": 4.247727272727273,
+      "grad_norm": 1.0696040335224966,
+      "learning_rate": 6.856731335173394e-08,
+      "loss": 0.0051,
+      "step": 18690
+    },
+    {
+      "epoch": 4.2479545454545455,
+      "grad_norm": 2.942011418753579,
+      "learning_rate": 6.852667560478081e-08,
+      "loss": 0.0194,
+      "step": 18691
+    },
+    {
+      "epoch": 4.248181818181818,
+      "grad_norm": 0.6646084616531788,
+      "learning_rate": 6.848604920528954e-08,
+      "loss": 0.0021,
+      "step": 18692
+    },
+    {
+      "epoch": 4.248409090909091,
+      "grad_norm": 1.1783378492288235,
+      "learning_rate": 6.844543415408856e-08,
+      "loss": 0.0044,
+      "step": 18693
+    },
+    {
+      "epoch": 4.248636363636364,
+      "grad_norm": 2.4497034181006683,
+      "learning_rate": 6.840483045200608e-08,
+      "loss": 0.0273,
+      "step": 18694
+    },
+    {
+      "epoch": 4.2488636363636365,
+      "grad_norm": 0.22244383724965425,
+      "learning_rate": 6.836423809987023e-08,
+      "loss": 0.0007,
+      "step": 18695
+    },
+    {
+      "epoch": 4.249090909090909,
+      "grad_norm": 4.103164924308661,
+      "learning_rate": 6.832365709850859e-08,
+      "loss": 0.0286,
+      "step": 18696
+    },
+    {
+      "epoch": 4.2493181818181816,
+      "grad_norm": 3.1023364989630924,
+      "learning_rate": 6.828308744874883e-08,
+      "loss": 0.0088,
+      "step": 18697
+    },
+    {
+      "epoch": 4.249545454545455,
+      "grad_norm": 0.48879078337859355,
+      "learning_rate": 6.824252915141813e-08,
+      "loss": 0.0016,
+      "step": 18698
+    },
+    {
+      "epoch": 4.2497727272727275,
+      "grad_norm": 1.2224962502345698,
+      "learning_rate": 6.820198220734361e-08,
+      "loss": 0.0074,
+      "step": 18699
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 1.6636111770401731,
+      "learning_rate": 6.816144661735212e-08,
+      "loss": 0.0052,
+      "step": 18700
+    },
+    {
+      "epoch": 4.2502272727272725,
+      "grad_norm": 0.9599618488301224,
+      "learning_rate": 6.812092238227014e-08,
+      "loss": 0.004,
+      "step": 18701
+    },
+    {
+      "epoch": 4.250454545454545,
+      "grad_norm": 0.5255595577649169,
+      "learning_rate": 6.808040950292403e-08,
+      "loss": 0.0029,
+      "step": 18702
+    },
+    {
+      "epoch": 4.2506818181818184,
+      "grad_norm": 1.4436424123934128,
+      "learning_rate": 6.803990798013998e-08,
+      "loss": 0.013,
+      "step": 18703
+    },
+    {
+      "epoch": 4.250909090909091,
+      "grad_norm": 1.12286864880849,
+      "learning_rate": 6.79994178147439e-08,
+      "loss": 0.0089,
+      "step": 18704
+    },
+    {
+      "epoch": 4.2511363636363635,
+      "grad_norm": 1.7542617266501392,
+      "learning_rate": 6.795893900756155e-08,
+      "loss": 0.0178,
+      "step": 18705
+    },
+    {
+      "epoch": 4.251363636363636,
+      "grad_norm": 0.5493843098735395,
+      "learning_rate": 6.791847155941805e-08,
+      "loss": 0.0013,
+      "step": 18706
+    },
+    {
+      "epoch": 4.251590909090909,
+      "grad_norm": 2.8874431424717133,
+      "learning_rate": 6.787801547113886e-08,
+      "loss": 0.014,
+      "step": 18707
+    },
+    {
+      "epoch": 4.251818181818182,
+      "grad_norm": 2.0893864333265273,
+      "learning_rate": 6.783757074354892e-08,
+      "loss": 0.0122,
+      "step": 18708
+    },
+    {
+      "epoch": 4.2520454545454545,
+      "grad_norm": 1.0497850315037032,
+      "learning_rate": 6.779713737747301e-08,
+      "loss": 0.0068,
+      "step": 18709
+    },
+    {
+      "epoch": 4.252272727272727,
+      "grad_norm": 1.0706819683919062,
+      "learning_rate": 6.775671537373546e-08,
+      "loss": 0.005,
+      "step": 18710
+    },
+    {
+      "epoch": 4.2525,
+      "grad_norm": 1.6091516167415016,
+      "learning_rate": 6.771630473316062e-08,
+      "loss": 0.0089,
+      "step": 18711
+    },
+    {
+      "epoch": 4.252727272727273,
+      "grad_norm": 0.8120673497865044,
+      "learning_rate": 6.76759054565726e-08,
+      "loss": 0.0025,
+      "step": 18712
+    },
+    {
+      "epoch": 4.252954545454545,
+      "grad_norm": 0.3062994985544057,
+      "learning_rate": 6.763551754479513e-08,
+      "loss": 0.0011,
+      "step": 18713
+    },
+    {
+      "epoch": 4.253181818181818,
+      "grad_norm": 0.31233871867140844,
+      "learning_rate": 6.759514099865187e-08,
+      "loss": 0.0019,
+      "step": 18714
+    },
+    {
+      "epoch": 4.253409090909091,
+      "grad_norm": 0.7005110732639785,
+      "learning_rate": 6.75547758189661e-08,
+      "loss": 0.0021,
+      "step": 18715
+    },
+    {
+      "epoch": 4.253636363636364,
+      "grad_norm": 1.361407135829946,
+      "learning_rate": 6.7514422006561e-08,
+      "loss": 0.0052,
+      "step": 18716
+    },
+    {
+      "epoch": 4.253863636363636,
+      "grad_norm": 0.7487232878290129,
+      "learning_rate": 6.747407956225943e-08,
+      "loss": 0.0021,
+      "step": 18717
+    },
+    {
+      "epoch": 4.254090909090909,
+      "grad_norm": 0.5284896864082517,
+      "learning_rate": 6.743374848688415e-08,
+      "loss": 0.0023,
+      "step": 18718
+    },
+    {
+      "epoch": 4.254318181818181,
+      "grad_norm": 2.0758336771773185,
+      "learning_rate": 6.739342878125728e-08,
+      "loss": 0.0095,
+      "step": 18719
+    },
+    {
+      "epoch": 4.254545454545455,
+      "grad_norm": 1.3684124660032198,
+      "learning_rate": 6.735312044620128e-08,
+      "loss": 0.0056,
+      "step": 18720
+    },
+    {
+      "epoch": 4.254772727272727,
+      "grad_norm": 0.5907162085646034,
+      "learning_rate": 6.731282348253799e-08,
+      "loss": 0.0019,
+      "step": 18721
+    },
+    {
+      "epoch": 4.255,
+      "grad_norm": 2.2818636656091664,
+      "learning_rate": 6.727253789108927e-08,
+      "loss": 0.0086,
+      "step": 18722
+    },
+    {
+      "epoch": 4.255227272727272,
+      "grad_norm": 0.7512972598428256,
+      "learning_rate": 6.72322636726764e-08,
+      "loss": 0.0046,
+      "step": 18723
+    },
+    {
+      "epoch": 4.255454545454546,
+      "grad_norm": 0.5487687910108013,
+      "learning_rate": 6.719200082812079e-08,
+      "loss": 0.0019,
+      "step": 18724
+    },
+    {
+      "epoch": 4.255681818181818,
+      "grad_norm": 1.2286064359748252,
+      "learning_rate": 6.71517493582434e-08,
+      "loss": 0.0048,
+      "step": 18725
+    },
+    {
+      "epoch": 4.255909090909091,
+      "grad_norm": 0.7308587150643328,
+      "learning_rate": 6.711150926386502e-08,
+      "loss": 0.0055,
+      "step": 18726
+    },
+    {
+      "epoch": 4.256136363636363,
+      "grad_norm": 0.2487237722739691,
+      "learning_rate": 6.707128054580631e-08,
+      "loss": 0.001,
+      "step": 18727
+    },
+    {
+      "epoch": 4.256363636363636,
+      "grad_norm": 2.173236752444706,
+      "learning_rate": 6.703106320488755e-08,
+      "loss": 0.0161,
+      "step": 18728
+    },
+    {
+      "epoch": 4.256590909090909,
+      "grad_norm": 1.8231923375693475,
+      "learning_rate": 6.69908572419288e-08,
+      "loss": 0.0079,
+      "step": 18729
+    },
+    {
+      "epoch": 4.256818181818182,
+      "grad_norm": 0.7029781856975544,
+      "learning_rate": 6.695066265775e-08,
+      "loss": 0.0032,
+      "step": 18730
+    },
+    {
+      "epoch": 4.257045454545454,
+      "grad_norm": 0.6716130293966355,
+      "learning_rate": 6.691047945317087e-08,
+      "loss": 0.0037,
+      "step": 18731
+    },
+    {
+      "epoch": 4.257272727272727,
+      "grad_norm": 0.803846098216572,
+      "learning_rate": 6.687030762901059e-08,
+      "loss": 0.0021,
+      "step": 18732
+    },
+    {
+      "epoch": 4.2575,
+      "grad_norm": 1.2611685036735916,
+      "learning_rate": 6.683014718608848e-08,
+      "loss": 0.0085,
+      "step": 18733
+    },
+    {
+      "epoch": 4.257727272727273,
+      "grad_norm": 1.788988789756406,
+      "learning_rate": 6.678999812522339e-08,
+      "loss": 0.0065,
+      "step": 18734
+    },
+    {
+      "epoch": 4.257954545454545,
+      "grad_norm": 2.7468754426383954,
+      "learning_rate": 6.674986044723428e-08,
+      "loss": 0.0148,
+      "step": 18735
+    },
+    {
+      "epoch": 4.258181818181818,
+      "grad_norm": 2.035943010595331,
+      "learning_rate": 6.670973415293928e-08,
+      "loss": 0.006,
+      "step": 18736
+    },
+    {
+      "epoch": 4.258409090909091,
+      "grad_norm": 1.108373466809747,
+      "learning_rate": 6.666961924315679e-08,
+      "loss": 0.0038,
+      "step": 18737
+    },
+    {
+      "epoch": 4.258636363636364,
+      "grad_norm": 1.0427578070326837,
+      "learning_rate": 6.662951571870482e-08,
+      "loss": 0.0087,
+      "step": 18738
+    },
+    {
+      "epoch": 4.258863636363636,
+      "grad_norm": 0.570813523195516,
+      "learning_rate": 6.658942358040112e-08,
+      "loss": 0.0009,
+      "step": 18739
+    },
+    {
+      "epoch": 4.259090909090909,
+      "grad_norm": 2.1590130494763744,
+      "learning_rate": 6.654934282906345e-08,
+      "loss": 0.0104,
+      "step": 18740
+    },
+    {
+      "epoch": 4.259318181818182,
+      "grad_norm": 0.26565108191871584,
+      "learning_rate": 6.65092734655088e-08,
+      "loss": 0.0008,
+      "step": 18741
+    },
+    {
+      "epoch": 4.259545454545455,
+      "grad_norm": 2.9512700338139277,
+      "learning_rate": 6.64692154905544e-08,
+      "loss": 0.0182,
+      "step": 18742
+    },
+    {
+      "epoch": 4.259772727272727,
+      "grad_norm": 1.8447091245049436,
+      "learning_rate": 6.642916890501714e-08,
+      "loss": 0.0067,
+      "step": 18743
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 0.7060377831249681,
+      "learning_rate": 6.638913370971361e-08,
+      "loss": 0.0046,
+      "step": 18744
+    },
+    {
+      "epoch": 4.260227272727272,
+      "grad_norm": 1.1439646414094136,
+      "learning_rate": 6.634910990546015e-08,
+      "loss": 0.0052,
+      "step": 18745
+    },
+    {
+      "epoch": 4.260454545454546,
+      "grad_norm": 2.446256043546572,
+      "learning_rate": 6.630909749307302e-08,
+      "loss": 0.0068,
+      "step": 18746
+    },
+    {
+      "epoch": 4.260681818181818,
+      "grad_norm": 2.0749969291974497,
+      "learning_rate": 6.626909647336803e-08,
+      "loss": 0.0166,
+      "step": 18747
+    },
+    {
+      "epoch": 4.260909090909091,
+      "grad_norm": 0.4730710630578731,
+      "learning_rate": 6.622910684716092e-08,
+      "loss": 0.0013,
+      "step": 18748
+    },
+    {
+      "epoch": 4.261136363636363,
+      "grad_norm": 0.8232313615895809,
+      "learning_rate": 6.61891286152673e-08,
+      "loss": 0.0063,
+      "step": 18749
+    },
+    {
+      "epoch": 4.261363636363637,
+      "grad_norm": 0.5408962565939549,
+      "learning_rate": 6.614916177850213e-08,
+      "loss": 0.0028,
+      "step": 18750
+    },
+    {
+      "epoch": 4.261590909090909,
+      "grad_norm": 1.0289278115903682,
+      "learning_rate": 6.610920633768045e-08,
+      "loss": 0.0057,
+      "step": 18751
+    },
+    {
+      "epoch": 4.261818181818182,
+      "grad_norm": 0.8524218872917242,
+      "learning_rate": 6.606926229361717e-08,
+      "loss": 0.0053,
+      "step": 18752
+    },
+    {
+      "epoch": 4.262045454545454,
+      "grad_norm": 0.6502793355153328,
+      "learning_rate": 6.602932964712684e-08,
+      "loss": 0.0025,
+      "step": 18753
+    },
+    {
+      "epoch": 4.262272727272728,
+      "grad_norm": 0.7995406217669744,
+      "learning_rate": 6.598940839902354e-08,
+      "loss": 0.0031,
+      "step": 18754
+    },
+    {
+      "epoch": 4.2625,
+      "grad_norm": 1.2866284506678891,
+      "learning_rate": 6.594949855012141e-08,
+      "loss": 0.0043,
+      "step": 18755
+    },
+    {
+      "epoch": 4.262727272727273,
+      "grad_norm": 1.7497643450923719,
+      "learning_rate": 6.590960010123438e-08,
+      "loss": 0.0133,
+      "step": 18756
+    },
+    {
+      "epoch": 4.262954545454545,
+      "grad_norm": 0.4304905271362198,
+      "learning_rate": 6.586971305317595e-08,
+      "loss": 0.0028,
+      "step": 18757
+    },
+    {
+      "epoch": 4.263181818181819,
+      "grad_norm": 1.3876095156585744,
+      "learning_rate": 6.582983740675952e-08,
+      "loss": 0.0067,
+      "step": 18758
+    },
+    {
+      "epoch": 4.263409090909091,
+      "grad_norm": 0.7112973091891258,
+      "learning_rate": 6.578997316279825e-08,
+      "loss": 0.0025,
+      "step": 18759
+    },
+    {
+      "epoch": 4.263636363636364,
+      "grad_norm": 0.7717569732054081,
+      "learning_rate": 6.575012032210503e-08,
+      "loss": 0.0026,
+      "step": 18760
+    },
+    {
+      "epoch": 4.263863636363636,
+      "grad_norm": 0.735694214249354,
+      "learning_rate": 6.571027888549256e-08,
+      "loss": 0.0031,
+      "step": 18761
+    },
+    {
+      "epoch": 4.264090909090909,
+      "grad_norm": 1.3786446166667328,
+      "learning_rate": 6.567044885377328e-08,
+      "loss": 0.0087,
+      "step": 18762
+    },
+    {
+      "epoch": 4.264318181818182,
+      "grad_norm": 0.8713961278434208,
+      "learning_rate": 6.563063022775928e-08,
+      "loss": 0.0048,
+      "step": 18763
+    },
+    {
+      "epoch": 4.264545454545455,
+      "grad_norm": 1.4318665367067658,
+      "learning_rate": 6.559082300826261e-08,
+      "loss": 0.0078,
+      "step": 18764
+    },
+    {
+      "epoch": 4.264772727272727,
+      "grad_norm": 1.8050135363421898,
+      "learning_rate": 6.555102719609498e-08,
+      "loss": 0.0071,
+      "step": 18765
+    },
+    {
+      "epoch": 4.265,
+      "grad_norm": 1.256592722899697,
+      "learning_rate": 6.551124279206803e-08,
+      "loss": 0.0041,
+      "step": 18766
+    },
+    {
+      "epoch": 4.265227272727273,
+      "grad_norm": 2.470590211267279,
+      "learning_rate": 6.547146979699277e-08,
+      "loss": 0.005,
+      "step": 18767
+    },
+    {
+      "epoch": 4.265454545454546,
+      "grad_norm": 0.5212207356579887,
+      "learning_rate": 6.543170821168044e-08,
+      "loss": 0.002,
+      "step": 18768
+    },
+    {
+      "epoch": 4.265681818181818,
+      "grad_norm": 1.900458678535844,
+      "learning_rate": 6.53919580369418e-08,
+      "loss": 0.0089,
+      "step": 18769
+    },
+    {
+      "epoch": 4.265909090909091,
+      "grad_norm": 0.6148912382826421,
+      "learning_rate": 6.535221927358754e-08,
+      "loss": 0.0022,
+      "step": 18770
+    },
+    {
+      "epoch": 4.266136363636363,
+      "grad_norm": 1.6961685732845124,
+      "learning_rate": 6.531249192242774e-08,
+      "loss": 0.0099,
+      "step": 18771
+    },
+    {
+      "epoch": 4.266363636363637,
+      "grad_norm": 2.120038052730565,
+      "learning_rate": 6.527277598427272e-08,
+      "loss": 0.0091,
+      "step": 18772
+    },
+    {
+      "epoch": 4.266590909090909,
+      "grad_norm": 0.9178555573419646,
+      "learning_rate": 6.523307145993224e-08,
+      "loss": 0.0035,
+      "step": 18773
+    },
+    {
+      "epoch": 4.266818181818182,
+      "grad_norm": 1.7335581246615275,
+      "learning_rate": 6.519337835021587e-08,
+      "loss": 0.0094,
+      "step": 18774
+    },
+    {
+      "epoch": 4.267045454545454,
+      "grad_norm": 1.2700293527806636,
+      "learning_rate": 6.515369665593344e-08,
+      "loss": 0.0064,
+      "step": 18775
+    },
+    {
+      "epoch": 4.2672727272727276,
+      "grad_norm": 0.9204592867420638,
+      "learning_rate": 6.51140263778937e-08,
+      "loss": 0.0027,
+      "step": 18776
+    },
+    {
+      "epoch": 4.2675,
+      "grad_norm": 1.3383687120916166,
+      "learning_rate": 6.507436751690577e-08,
+      "loss": 0.0035,
+      "step": 18777
+    },
+    {
+      "epoch": 4.267727272727273,
+      "grad_norm": 1.1589821502973627,
+      "learning_rate": 6.503472007377824e-08,
+      "loss": 0.006,
+      "step": 18778
+    },
+    {
+      "epoch": 4.267954545454545,
+      "grad_norm": 0.6486346774157495,
+      "learning_rate": 6.499508404931992e-08,
+      "loss": 0.0065,
+      "step": 18779
+    },
+    {
+      "epoch": 4.2681818181818185,
+      "grad_norm": 1.1115314841289308,
+      "learning_rate": 6.495545944433863e-08,
+      "loss": 0.0056,
+      "step": 18780
+    },
+    {
+      "epoch": 4.268409090909091,
+      "grad_norm": 3.5060307841576424,
+      "learning_rate": 6.491584625964255e-08,
+      "loss": 0.0105,
+      "step": 18781
+    },
+    {
+      "epoch": 4.268636363636364,
+      "grad_norm": 2.6686451116067818,
+      "learning_rate": 6.487624449603949e-08,
+      "loss": 0.0096,
+      "step": 18782
+    },
+    {
+      "epoch": 4.268863636363636,
+      "grad_norm": 3.4989214049124837,
+      "learning_rate": 6.483665415433702e-08,
+      "loss": 0.0148,
+      "step": 18783
+    },
+    {
+      "epoch": 4.2690909090909095,
+      "grad_norm": 0.6295742029484189,
+      "learning_rate": 6.479707523534259e-08,
+      "loss": 0.0044,
+      "step": 18784
+    },
+    {
+      "epoch": 4.269318181818182,
+      "grad_norm": 1.3025661179658645,
+      "learning_rate": 6.475750773986295e-08,
+      "loss": 0.0084,
+      "step": 18785
+    },
+    {
+      "epoch": 4.2695454545454545,
+      "grad_norm": 0.8524689416943719,
+      "learning_rate": 6.471795166870513e-08,
+      "loss": 0.0051,
+      "step": 18786
+    },
+    {
+      "epoch": 4.269772727272727,
+      "grad_norm": 1.5234279756365998,
+      "learning_rate": 6.467840702267573e-08,
+      "loss": 0.0039,
+      "step": 18787
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 0.20713194692538514,
+      "learning_rate": 6.463887380258119e-08,
+      "loss": 0.0006,
+      "step": 18788
+    },
+    {
+      "epoch": 4.270227272727273,
+      "grad_norm": 0.4852609509420961,
+      "learning_rate": 6.459935200922758e-08,
+      "loss": 0.0018,
+      "step": 18789
+    },
+    {
+      "epoch": 4.2704545454545455,
+      "grad_norm": 0.7257680328956748,
+      "learning_rate": 6.455984164342087e-08,
+      "loss": 0.0019,
+      "step": 18790
+    },
+    {
+      "epoch": 4.270681818181818,
+      "grad_norm": 0.7558115658073378,
+      "learning_rate": 6.452034270596672e-08,
+      "loss": 0.0036,
+      "step": 18791
+    },
+    {
+      "epoch": 4.2709090909090905,
+      "grad_norm": 1.0270888448200106,
+      "learning_rate": 6.448085519767077e-08,
+      "loss": 0.0033,
+      "step": 18792
+    },
+    {
+      "epoch": 4.271136363636364,
+      "grad_norm": 3.3555796490186385,
+      "learning_rate": 6.444137911933781e-08,
+      "loss": 0.0143,
+      "step": 18793
+    },
+    {
+      "epoch": 4.2713636363636365,
+      "grad_norm": 0.9985852986974327,
+      "learning_rate": 6.440191447177321e-08,
+      "loss": 0.003,
+      "step": 18794
+    },
+    {
+      "epoch": 4.271590909090909,
+      "grad_norm": 0.2700393379743801,
+      "learning_rate": 6.436246125578153e-08,
+      "loss": 0.0012,
+      "step": 18795
+    },
+    {
+      "epoch": 4.2718181818181815,
+      "grad_norm": 1.8249422701446256,
+      "learning_rate": 6.432301947216738e-08,
+      "loss": 0.0105,
+      "step": 18796
+    },
+    {
+      "epoch": 4.272045454545455,
+      "grad_norm": 1.575082984945921,
+      "learning_rate": 6.428358912173506e-08,
+      "loss": 0.0275,
+      "step": 18797
+    },
+    {
+      "epoch": 4.272272727272727,
+      "grad_norm": 1.7782335395519178,
+      "learning_rate": 6.424417020528858e-08,
+      "loss": 0.013,
+      "step": 18798
+    },
+    {
+      "epoch": 4.2725,
+      "grad_norm": 2.0241470489359963,
+      "learning_rate": 6.420476272363169e-08,
+      "loss": 0.0122,
+      "step": 18799
+    },
+    {
+      "epoch": 4.2727272727272725,
+      "grad_norm": 0.5085014009207661,
+      "learning_rate": 6.416536667756803e-08,
+      "loss": 0.0036,
+      "step": 18800
+    },
+    {
+      "epoch": 4.272954545454546,
+      "grad_norm": 1.863803661179137,
+      "learning_rate": 6.412598206790111e-08,
+      "loss": 0.019,
+      "step": 18801
+    },
+    {
+      "epoch": 4.273181818181818,
+      "grad_norm": 0.7195567675878495,
+      "learning_rate": 6.40866088954338e-08,
+      "loss": 0.0041,
+      "step": 18802
+    },
+    {
+      "epoch": 4.273409090909091,
+      "grad_norm": 0.3173855498295541,
+      "learning_rate": 6.404724716096906e-08,
+      "loss": 0.0017,
+      "step": 18803
+    },
+    {
+      "epoch": 4.2736363636363635,
+      "grad_norm": 1.2667931794666802,
+      "learning_rate": 6.400789686530962e-08,
+      "loss": 0.0126,
+      "step": 18804
+    },
+    {
+      "epoch": 4.273863636363636,
+      "grad_norm": 1.9182033352573435,
+      "learning_rate": 6.396855800925784e-08,
+      "loss": 0.0162,
+      "step": 18805
+    },
+    {
+      "epoch": 4.274090909090909,
+      "grad_norm": 0.6255797537406853,
+      "learning_rate": 6.392923059361609e-08,
+      "loss": 0.002,
+      "step": 18806
+    },
+    {
+      "epoch": 4.274318181818182,
+      "grad_norm": 0.534885605486534,
+      "learning_rate": 6.388991461918605e-08,
+      "loss": 0.0016,
+      "step": 18807
+    },
+    {
+      "epoch": 4.274545454545454,
+      "grad_norm": 1.7154286454802992,
+      "learning_rate": 6.38506100867696e-08,
+      "loss": 0.0103,
+      "step": 18808
+    },
+    {
+      "epoch": 4.274772727272727,
+      "grad_norm": 0.2767691349502281,
+      "learning_rate": 6.381131699716818e-08,
+      "loss": 0.0009,
+      "step": 18809
+    },
+    {
+      "epoch": 4.275,
+      "grad_norm": 4.606432474733516,
+      "learning_rate": 6.37720353511831e-08,
+      "loss": 0.0096,
+      "step": 18810
+    },
+    {
+      "epoch": 4.275227272727273,
+      "grad_norm": 2.0138396522664164,
+      "learning_rate": 6.37327651496153e-08,
+      "loss": 0.0167,
+      "step": 18811
+    },
+    {
+      "epoch": 4.275454545454545,
+      "grad_norm": 1.4742564062170478,
+      "learning_rate": 6.369350639326553e-08,
+      "loss": 0.0055,
+      "step": 18812
+    },
+    {
+      "epoch": 4.275681818181818,
+      "grad_norm": 0.6189122999023664,
+      "learning_rate": 6.365425908293449e-08,
+      "loss": 0.0027,
+      "step": 18813
+    },
+    {
+      "epoch": 4.27590909090909,
+      "grad_norm": 1.1915095018857156,
+      "learning_rate": 6.361502321942249e-08,
+      "loss": 0.0106,
+      "step": 18814
+    },
+    {
+      "epoch": 4.276136363636364,
+      "grad_norm": 0.9091151282782262,
+      "learning_rate": 6.357579880352946e-08,
+      "loss": 0.0035,
+      "step": 18815
+    },
+    {
+      "epoch": 4.276363636363636,
+      "grad_norm": 1.893887884478165,
+      "learning_rate": 6.353658583605537e-08,
+      "loss": 0.0093,
+      "step": 18816
+    },
+    {
+      "epoch": 4.276590909090909,
+      "grad_norm": 2.885646127643889,
+      "learning_rate": 6.34973843177998e-08,
+      "loss": 0.0185,
+      "step": 18817
+    },
+    {
+      "epoch": 4.276818181818181,
+      "grad_norm": 0.7293797617686762,
+      "learning_rate": 6.345819424956219e-08,
+      "loss": 0.0029,
+      "step": 18818
+    },
+    {
+      "epoch": 4.277045454545455,
+      "grad_norm": 4.31572821620937,
+      "learning_rate": 6.34190156321416e-08,
+      "loss": 0.0154,
+      "step": 18819
+    },
+    {
+      "epoch": 4.277272727272727,
+      "grad_norm": 1.2077501156254213,
+      "learning_rate": 6.337984846633707e-08,
+      "loss": 0.006,
+      "step": 18820
+    },
+    {
+      "epoch": 4.2775,
+      "grad_norm": 1.5262488412622768,
+      "learning_rate": 6.334069275294723e-08,
+      "loss": 0.0059,
+      "step": 18821
+    },
+    {
+      "epoch": 4.277727272727272,
+      "grad_norm": 0.42718369380278315,
+      "learning_rate": 6.330154849277049e-08,
+      "loss": 0.0015,
+      "step": 18822
+    },
+    {
+      "epoch": 4.277954545454546,
+      "grad_norm": 1.3548455860819006,
+      "learning_rate": 6.326241568660525e-08,
+      "loss": 0.0062,
+      "step": 18823
+    },
+    {
+      "epoch": 4.278181818181818,
+      "grad_norm": 2.0165456836662146,
+      "learning_rate": 6.322329433524927e-08,
+      "loss": 0.0124,
+      "step": 18824
+    },
+    {
+      "epoch": 4.278409090909091,
+      "grad_norm": 0.6631179817924295,
+      "learning_rate": 6.318418443950037e-08,
+      "loss": 0.0029,
+      "step": 18825
+    },
+    {
+      "epoch": 4.278636363636363,
+      "grad_norm": 0.3414235531904603,
+      "learning_rate": 6.314508600015603e-08,
+      "loss": 0.0008,
+      "step": 18826
+    },
+    {
+      "epoch": 4.278863636363637,
+      "grad_norm": 0.969047172663544,
+      "learning_rate": 6.31059990180137e-08,
+      "loss": 0.0051,
+      "step": 18827
+    },
+    {
+      "epoch": 4.279090909090909,
+      "grad_norm": 0.6384792918361845,
+      "learning_rate": 6.306692349387036e-08,
+      "loss": 0.0014,
+      "step": 18828
+    },
+    {
+      "epoch": 4.279318181818182,
+      "grad_norm": 1.0895174164384602,
+      "learning_rate": 6.302785942852275e-08,
+      "loss": 0.0038,
+      "step": 18829
+    },
+    {
+      "epoch": 4.279545454545454,
+      "grad_norm": 2.408797253175877,
+      "learning_rate": 6.298880682276742e-08,
+      "loss": 0.0282,
+      "step": 18830
+    },
+    {
+      "epoch": 4.279772727272727,
+      "grad_norm": 1.1149104309527291,
+      "learning_rate": 6.294976567740085e-08,
+      "loss": 0.0037,
+      "step": 18831
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 0.8881862517416119,
+      "learning_rate": 6.291073599321913e-08,
+      "loss": 0.0026,
+      "step": 18832
+    },
+    {
+      "epoch": 4.280227272727273,
+      "grad_norm": 0.6999933402324107,
+      "learning_rate": 6.287171777101808e-08,
+      "loss": 0.0024,
+      "step": 18833
+    },
+    {
+      "epoch": 4.280454545454545,
+      "grad_norm": 0.2977955189365957,
+      "learning_rate": 6.283271101159345e-08,
+      "loss": 0.0008,
+      "step": 18834
+    },
+    {
+      "epoch": 4.280681818181818,
+      "grad_norm": 1.3444224912292737,
+      "learning_rate": 6.279371571574051e-08,
+      "loss": 0.0076,
+      "step": 18835
+    },
+    {
+      "epoch": 4.280909090909091,
+      "grad_norm": 0.5022877508917529,
+      "learning_rate": 6.275473188425474e-08,
+      "loss": 0.0021,
+      "step": 18836
+    },
+    {
+      "epoch": 4.281136363636364,
+      "grad_norm": 0.8956098370786165,
+      "learning_rate": 6.27157595179307e-08,
+      "loss": 0.0082,
+      "step": 18837
+    },
+    {
+      "epoch": 4.281363636363636,
+      "grad_norm": 0.1041212647157879,
+      "learning_rate": 6.267679861756325e-08,
+      "loss": 0.0003,
+      "step": 18838
+    },
+    {
+      "epoch": 4.281590909090909,
+      "grad_norm": 0.889415553963287,
+      "learning_rate": 6.263784918394696e-08,
+      "loss": 0.0037,
+      "step": 18839
+    },
+    {
+      "epoch": 4.281818181818182,
+      "grad_norm": 0.6827239528590794,
+      "learning_rate": 6.259891121787605e-08,
+      "loss": 0.0029,
+      "step": 18840
+    },
+    {
+      "epoch": 4.282045454545455,
+      "grad_norm": 0.34681122958074395,
+      "learning_rate": 6.255998472014455e-08,
+      "loss": 0.001,
+      "step": 18841
+    },
+    {
+      "epoch": 4.282272727272727,
+      "grad_norm": 0.45712644366023086,
+      "learning_rate": 6.252106969154612e-08,
+      "loss": 0.0018,
+      "step": 18842
+    },
+    {
+      "epoch": 4.2825,
+      "grad_norm": 0.45942042560734436,
+      "learning_rate": 6.248216613287436e-08,
+      "loss": 0.0015,
+      "step": 18843
+    },
+    {
+      "epoch": 4.282727272727273,
+      "grad_norm": 1.3290416584932645,
+      "learning_rate": 6.24432740449226e-08,
+      "loss": 0.0107,
+      "step": 18844
+    },
+    {
+      "epoch": 4.282954545454546,
+      "grad_norm": 1.2286473875314077,
+      "learning_rate": 6.240439342848403e-08,
+      "loss": 0.0065,
+      "step": 18845
+    },
+    {
+      "epoch": 4.283181818181818,
+      "grad_norm": 0.3934978219061647,
+      "learning_rate": 6.236552428435127e-08,
+      "loss": 0.002,
+      "step": 18846
+    },
+    {
+      "epoch": 4.283409090909091,
+      "grad_norm": 3.3283325786346443,
+      "learning_rate": 6.232666661331703e-08,
+      "loss": 0.0148,
+      "step": 18847
+    },
+    {
+      "epoch": 4.283636363636363,
+      "grad_norm": 2.4307822906964276,
+      "learning_rate": 6.228782041617365e-08,
+      "loss": 0.0095,
+      "step": 18848
+    },
+    {
+      "epoch": 4.283863636363637,
+      "grad_norm": 0.5487792925018425,
+      "learning_rate": 6.224898569371342e-08,
+      "loss": 0.0013,
+      "step": 18849
+    },
+    {
+      "epoch": 4.284090909090909,
+      "grad_norm": 1.5708515158299734,
+      "learning_rate": 6.221016244672807e-08,
+      "loss": 0.0083,
+      "step": 18850
+    },
+    {
+      "epoch": 4.284318181818182,
+      "grad_norm": 0.9632344371839221,
+      "learning_rate": 6.217135067600938e-08,
+      "loss": 0.0074,
+      "step": 18851
+    },
+    {
+      "epoch": 4.284545454545454,
+      "grad_norm": 0.5365648310381945,
+      "learning_rate": 6.213255038234875e-08,
+      "loss": 0.0021,
+      "step": 18852
+    },
+    {
+      "epoch": 4.284772727272728,
+      "grad_norm": 1.430301422549913,
+      "learning_rate": 6.209376156653734e-08,
+      "loss": 0.007,
+      "step": 18853
+    },
+    {
+      "epoch": 4.285,
+      "grad_norm": 2.4303228613480874,
+      "learning_rate": 6.205498422936635e-08,
+      "loss": 0.0173,
+      "step": 18854
+    },
+    {
+      "epoch": 4.285227272727273,
+      "grad_norm": 0.7210062418353451,
+      "learning_rate": 6.201621837162622e-08,
+      "loss": 0.0054,
+      "step": 18855
+    },
+    {
+      "epoch": 4.285454545454545,
+      "grad_norm": 0.4881020646220097,
+      "learning_rate": 6.197746399410759e-08,
+      "loss": 0.0048,
+      "step": 18856
+    },
+    {
+      "epoch": 4.285681818181819,
+      "grad_norm": 0.8429104410422451,
+      "learning_rate": 6.193872109760072e-08,
+      "loss": 0.0045,
+      "step": 18857
+    },
+    {
+      "epoch": 4.285909090909091,
+      "grad_norm": 0.5361164539216042,
+      "learning_rate": 6.189998968289577e-08,
+      "loss": 0.0028,
+      "step": 18858
+    },
+    {
+      "epoch": 4.286136363636364,
+      "grad_norm": 1.6403745191471595,
+      "learning_rate": 6.186126975078221e-08,
+      "loss": 0.0075,
+      "step": 18859
+    },
+    {
+      "epoch": 4.286363636363636,
+      "grad_norm": 1.6341053132358672,
+      "learning_rate": 6.18225613020499e-08,
+      "loss": 0.0135,
+      "step": 18860
+    },
+    {
+      "epoch": 4.28659090909091,
+      "grad_norm": 3.3106842220616115,
+      "learning_rate": 6.178386433748802e-08,
+      "loss": 0.0298,
+      "step": 18861
+    },
+    {
+      "epoch": 4.286818181818182,
+      "grad_norm": 0.7975845273203294,
+      "learning_rate": 6.17451788578858e-08,
+      "loss": 0.002,
+      "step": 18862
+    },
+    {
+      "epoch": 4.287045454545455,
+      "grad_norm": 2.2958745685037134,
+      "learning_rate": 6.170650486403196e-08,
+      "loss": 0.0094,
+      "step": 18863
+    },
+    {
+      "epoch": 4.287272727272727,
+      "grad_norm": 1.2626539120003772,
+      "learning_rate": 6.166784235671526e-08,
+      "loss": 0.004,
+      "step": 18864
+    },
+    {
+      "epoch": 4.2875,
+      "grad_norm": 1.6159715333408267,
+      "learning_rate": 6.162919133672403e-08,
+      "loss": 0.0065,
+      "step": 18865
+    },
+    {
+      "epoch": 4.287727272727273,
+      "grad_norm": 0.4949787512391543,
+      "learning_rate": 6.159055180484638e-08,
+      "loss": 0.0015,
+      "step": 18866
+    },
+    {
+      "epoch": 4.287954545454546,
+      "grad_norm": 0.7302932835752618,
+      "learning_rate": 6.155192376187051e-08,
+      "loss": 0.003,
+      "step": 18867
+    },
+    {
+      "epoch": 4.288181818181818,
+      "grad_norm": 0.6801123836635675,
+      "learning_rate": 6.15133072085837e-08,
+      "loss": 0.0036,
+      "step": 18868
+    },
+    {
+      "epoch": 4.288409090909091,
+      "grad_norm": 0.6518584001889309,
+      "learning_rate": 6.147470214577366e-08,
+      "loss": 0.0035,
+      "step": 18869
+    },
+    {
+      "epoch": 4.288636363636364,
+      "grad_norm": 0.6323253223358676,
+      "learning_rate": 6.143610857422759e-08,
+      "loss": 0.0046,
+      "step": 18870
+    },
+    {
+      "epoch": 4.2888636363636365,
+      "grad_norm": 0.8232642453524613,
+      "learning_rate": 6.139752649473244e-08,
+      "loss": 0.003,
+      "step": 18871
+    },
+    {
+      "epoch": 4.289090909090909,
+      "grad_norm": 2.0376856490180018,
+      "learning_rate": 6.13589559080751e-08,
+      "loss": 0.0063,
+      "step": 18872
+    },
+    {
+      "epoch": 4.289318181818182,
+      "grad_norm": 0.7728045533058018,
+      "learning_rate": 6.132039681504185e-08,
+      "loss": 0.004,
+      "step": 18873
+    },
+    {
+      "epoch": 4.289545454545454,
+      "grad_norm": 1.156424577920418,
+      "learning_rate": 6.128184921641912e-08,
+      "loss": 0.0088,
+      "step": 18874
+    },
+    {
+      "epoch": 4.2897727272727275,
+      "grad_norm": 1.146559154991086,
+      "learning_rate": 6.12433131129929e-08,
+      "loss": 0.0165,
+      "step": 18875
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 0.9665005771161171,
+      "learning_rate": 6.120478850554918e-08,
+      "loss": 0.0083,
+      "step": 18876
+    },
+    {
+      "epoch": 4.290227272727273,
+      "grad_norm": 0.9768619508424918,
+      "learning_rate": 6.11662753948733e-08,
+      "loss": 0.0035,
+      "step": 18877
+    },
+    {
+      "epoch": 4.290454545454545,
+      "grad_norm": 0.5263966041345216,
+      "learning_rate": 6.112777378175076e-08,
+      "loss": 0.0017,
+      "step": 18878
+    },
+    {
+      "epoch": 4.2906818181818185,
+      "grad_norm": 0.6433999502034808,
+      "learning_rate": 6.108928366696667e-08,
+      "loss": 0.0013,
+      "step": 18879
+    },
+    {
+      "epoch": 4.290909090909091,
+      "grad_norm": 0.5988348532441592,
+      "learning_rate": 6.105080505130582e-08,
+      "loss": 0.0025,
+      "step": 18880
+    },
+    {
+      "epoch": 4.2911363636363635,
+      "grad_norm": 1.9874171220249577,
+      "learning_rate": 6.101233793555289e-08,
+      "loss": 0.0109,
+      "step": 18881
+    },
+    {
+      "epoch": 4.291363636363636,
+      "grad_norm": 0.6338830817147462,
+      "learning_rate": 6.097388232049237e-08,
+      "loss": 0.0029,
+      "step": 18882
+    },
+    {
+      "epoch": 4.2915909090909095,
+      "grad_norm": 1.8260362588367367,
+      "learning_rate": 6.093543820690841e-08,
+      "loss": 0.0207,
+      "step": 18883
+    },
+    {
+      "epoch": 4.291818181818182,
+      "grad_norm": 0.9826882146148944,
+      "learning_rate": 6.089700559558486e-08,
+      "loss": 0.0065,
+      "step": 18884
+    },
+    {
+      "epoch": 4.2920454545454545,
+      "grad_norm": 0.796406148903814,
+      "learning_rate": 6.085858448730567e-08,
+      "loss": 0.0048,
+      "step": 18885
+    },
+    {
+      "epoch": 4.292272727272727,
+      "grad_norm": 0.4290483480211586,
+      "learning_rate": 6.0820174882854e-08,
+      "loss": 0.0015,
+      "step": 18886
+    },
+    {
+      "epoch": 4.2925,
+      "grad_norm": 2.5153055399014552,
+      "learning_rate": 6.07817767830133e-08,
+      "loss": 0.0248,
+      "step": 18887
+    },
+    {
+      "epoch": 4.292727272727273,
+      "grad_norm": 1.1546838146725666,
+      "learning_rate": 6.074339018856644e-08,
+      "loss": 0.0065,
+      "step": 18888
+    },
+    {
+      "epoch": 4.2929545454545455,
+      "grad_norm": 0.2923639348760116,
+      "learning_rate": 6.070501510029637e-08,
+      "loss": 0.001,
+      "step": 18889
+    },
+    {
+      "epoch": 4.293181818181818,
+      "grad_norm": 0.32904894585811173,
+      "learning_rate": 6.066665151898541e-08,
+      "loss": 0.0008,
+      "step": 18890
+    },
+    {
+      "epoch": 4.2934090909090905,
+      "grad_norm": 2.6596217149898935,
+      "learning_rate": 6.06282994454159e-08,
+      "loss": 0.0138,
+      "step": 18891
+    },
+    {
+      "epoch": 4.293636363636364,
+      "grad_norm": 0.5911892190055629,
+      "learning_rate": 6.058995888037003e-08,
+      "loss": 0.0063,
+      "step": 18892
+    },
+    {
+      "epoch": 4.293863636363636,
+      "grad_norm": 2.0842721314629356,
+      "learning_rate": 6.055162982462955e-08,
+      "loss": 0.0088,
+      "step": 18893
+    },
+    {
+      "epoch": 4.294090909090909,
+      "grad_norm": 2.121064707794854,
+      "learning_rate": 6.051331227897605e-08,
+      "loss": 0.0079,
+      "step": 18894
+    },
+    {
+      "epoch": 4.2943181818181815,
+      "grad_norm": 1.7361619780549609,
+      "learning_rate": 6.047500624419098e-08,
+      "loss": 0.0173,
+      "step": 18895
+    },
+    {
+      "epoch": 4.294545454545455,
+      "grad_norm": 1.204150398269578,
+      "learning_rate": 6.043671172105532e-08,
+      "loss": 0.0045,
+      "step": 18896
+    },
+    {
+      "epoch": 4.294772727272727,
+      "grad_norm": 0.81775017384872,
+      "learning_rate": 6.039842871035005e-08,
+      "loss": 0.0032,
+      "step": 18897
+    },
+    {
+      "epoch": 4.295,
+      "grad_norm": 0.48189176124951355,
+      "learning_rate": 6.0360157212856e-08,
+      "loss": 0.0015,
+      "step": 18898
+    },
+    {
+      "epoch": 4.2952272727272724,
+      "grad_norm": 1.055942002764481,
+      "learning_rate": 6.032189722935324e-08,
+      "loss": 0.0033,
+      "step": 18899
+    },
+    {
+      "epoch": 4.295454545454546,
+      "grad_norm": 1.3317016787551645,
+      "learning_rate": 6.028364876062217e-08,
+      "loss": 0.0083,
+      "step": 18900
+    },
+    {
+      "epoch": 4.295681818181818,
+      "grad_norm": 0.9906796534722876,
+      "learning_rate": 6.024541180744268e-08,
+      "loss": 0.0049,
+      "step": 18901
+    },
+    {
+      "epoch": 4.295909090909091,
+      "grad_norm": 0.7477145737198015,
+      "learning_rate": 6.020718637059463e-08,
+      "loss": 0.0028,
+      "step": 18902
+    },
+    {
+      "epoch": 4.296136363636363,
+      "grad_norm": 0.40492026612486415,
+      "learning_rate": 6.01689724508573e-08,
+      "loss": 0.0012,
+      "step": 18903
+    },
+    {
+      "epoch": 4.296363636363637,
+      "grad_norm": 1.048947501233064,
+      "learning_rate": 6.013077004900996e-08,
+      "loss": 0.007,
+      "step": 18904
+    },
+    {
+      "epoch": 4.296590909090909,
+      "grad_norm": 0.8315215157346437,
+      "learning_rate": 6.009257916583176e-08,
+      "loss": 0.0021,
+      "step": 18905
+    },
+    {
+      "epoch": 4.296818181818182,
+      "grad_norm": 0.5555187701118668,
+      "learning_rate": 6.00543998021013e-08,
+      "loss": 0.003,
+      "step": 18906
+    },
+    {
+      "epoch": 4.297045454545454,
+      "grad_norm": 1.8538494111269714,
+      "learning_rate": 6.001623195859743e-08,
+      "loss": 0.0096,
+      "step": 18907
+    },
+    {
+      "epoch": 4.297272727272727,
+      "grad_norm": 1.36566152153383,
+      "learning_rate": 5.997807563609812e-08,
+      "loss": 0.0073,
+      "step": 18908
+    },
+    {
+      "epoch": 4.2975,
+      "grad_norm": 0.8876773508364451,
+      "learning_rate": 5.993993083538159e-08,
+      "loss": 0.0035,
+      "step": 18909
+    },
+    {
+      "epoch": 4.297727272727273,
+      "grad_norm": 2.5651966017188705,
+      "learning_rate": 5.990179755722566e-08,
+      "loss": 0.0092,
+      "step": 18910
+    },
+    {
+      "epoch": 4.297954545454545,
+      "grad_norm": 6.883163328605633,
+      "learning_rate": 5.9863675802408e-08,
+      "loss": 0.0197,
+      "step": 18911
+    },
+    {
+      "epoch": 4.298181818181818,
+      "grad_norm": 0.5886069371982388,
+      "learning_rate": 5.982556557170587e-08,
+      "loss": 0.0016,
+      "step": 18912
+    },
+    {
+      "epoch": 4.298409090909091,
+      "grad_norm": 1.2740480481215204,
+      "learning_rate": 5.978746686589646e-08,
+      "loss": 0.0094,
+      "step": 18913
+    },
+    {
+      "epoch": 4.298636363636364,
+      "grad_norm": 0.44922569118442673,
+      "learning_rate": 5.974937968575664e-08,
+      "loss": 0.0013,
+      "step": 18914
+    },
+    {
+      "epoch": 4.298863636363636,
+      "grad_norm": 0.6432532994662068,
+      "learning_rate": 5.971130403206328e-08,
+      "loss": 0.002,
+      "step": 18915
+    },
+    {
+      "epoch": 4.299090909090909,
+      "grad_norm": 0.40802821558202534,
+      "learning_rate": 5.967323990559249e-08,
+      "loss": 0.0017,
+      "step": 18916
+    },
+    {
+      "epoch": 4.299318181818181,
+      "grad_norm": 0.6643892943347631,
+      "learning_rate": 5.963518730712062e-08,
+      "loss": 0.004,
+      "step": 18917
+    },
+    {
+      "epoch": 4.299545454545455,
+      "grad_norm": 1.3917610741209419,
+      "learning_rate": 5.959714623742355e-08,
+      "loss": 0.004,
+      "step": 18918
+    },
+    {
+      "epoch": 4.299772727272727,
+      "grad_norm": 0.6291320384943272,
+      "learning_rate": 5.9559116697277127e-08,
+      "loss": 0.0025,
+      "step": 18919
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 0.45094328677597967,
+      "learning_rate": 5.952109868745691e-08,
+      "loss": 0.0017,
+      "step": 18920
+    },
+    {
+      "epoch": 4.300227272727272,
+      "grad_norm": 1.0304574772818613,
+      "learning_rate": 5.948309220873791e-08,
+      "loss": 0.0035,
+      "step": 18921
+    },
+    {
+      "epoch": 4.300454545454546,
+      "grad_norm": 0.4547308833698876,
+      "learning_rate": 5.944509726189519e-08,
+      "loss": 0.0015,
+      "step": 18922
+    },
+    {
+      "epoch": 4.300681818181818,
+      "grad_norm": 0.2471483018675285,
+      "learning_rate": 5.940711384770363e-08,
+      "loss": 0.0027,
+      "step": 18923
+    },
+    {
+      "epoch": 4.300909090909091,
+      "grad_norm": 0.6910224303435002,
+      "learning_rate": 5.936914196693774e-08,
+      "loss": 0.0025,
+      "step": 18924
+    },
+    {
+      "epoch": 4.301136363636363,
+      "grad_norm": 1.4398156425466888,
+      "learning_rate": 5.9331181620371905e-08,
+      "loss": 0.0043,
+      "step": 18925
+    },
+    {
+      "epoch": 4.301363636363637,
+      "grad_norm": 0.4892936052977675,
+      "learning_rate": 5.9293232808780094e-08,
+      "loss": 0.0019,
+      "step": 18926
+    },
+    {
+      "epoch": 4.301590909090909,
+      "grad_norm": 0.41412111308350824,
+      "learning_rate": 5.9255295532936185e-08,
+      "loss": 0.0014,
+      "step": 18927
+    },
+    {
+      "epoch": 4.301818181818182,
+      "grad_norm": 0.19543253455292048,
+      "learning_rate": 5.921736979361381e-08,
+      "loss": 0.0004,
+      "step": 18928
+    },
+    {
+      "epoch": 4.302045454545454,
+      "grad_norm": 1.0462830630060909,
+      "learning_rate": 5.9179455591586445e-08,
+      "loss": 0.004,
+      "step": 18929
+    },
+    {
+      "epoch": 4.302272727272728,
+      "grad_norm": 0.9699300611147663,
+      "learning_rate": 5.914155292762701e-08,
+      "loss": 0.0046,
+      "step": 18930
+    },
+    {
+      "epoch": 4.3025,
+      "grad_norm": 1.0898703321007108,
+      "learning_rate": 5.9103661802508494e-08,
+      "loss": 0.0028,
+      "step": 18931
+    },
+    {
+      "epoch": 4.302727272727273,
+      "grad_norm": 0.9270605872853536,
+      "learning_rate": 5.906578221700356e-08,
+      "loss": 0.0049,
+      "step": 18932
+    },
+    {
+      "epoch": 4.302954545454545,
+      "grad_norm": 0.25656169081517655,
+      "learning_rate": 5.902791417188476e-08,
+      "loss": 0.0007,
+      "step": 18933
+    },
+    {
+      "epoch": 4.303181818181818,
+      "grad_norm": 0.9415263952332613,
+      "learning_rate": 5.899005766792407e-08,
+      "loss": 0.0075,
+      "step": 18934
+    },
+    {
+      "epoch": 4.303409090909091,
+      "grad_norm": 1.4723825349537893,
+      "learning_rate": 5.895221270589363e-08,
+      "loss": 0.01,
+      "step": 18935
+    },
+    {
+      "epoch": 4.303636363636364,
+      "grad_norm": 1.012820464917433,
+      "learning_rate": 5.891437928656499e-08,
+      "loss": 0.0055,
+      "step": 18936
+    },
+    {
+      "epoch": 4.303863636363636,
+      "grad_norm": 0.5624946253734168,
+      "learning_rate": 5.887655741070995e-08,
+      "loss": 0.0034,
+      "step": 18937
+    },
+    {
+      "epoch": 4.304090909090909,
+      "grad_norm": 0.8260742711353544,
+      "learning_rate": 5.883874707909937e-08,
+      "loss": 0.0055,
+      "step": 18938
+    },
+    {
+      "epoch": 4.304318181818182,
+      "grad_norm": 0.6611049059535353,
+      "learning_rate": 5.8800948292504486e-08,
+      "loss": 0.0019,
+      "step": 18939
+    },
+    {
+      "epoch": 4.304545454545455,
+      "grad_norm": 4.482806917231233,
+      "learning_rate": 5.876316105169609e-08,
+      "loss": 0.0058,
+      "step": 18940
+    },
+    {
+      "epoch": 4.304772727272727,
+      "grad_norm": 0.44736100847873667,
+      "learning_rate": 5.87253853574446e-08,
+      "loss": 0.0017,
+      "step": 18941
+    },
+    {
+      "epoch": 4.305,
+      "grad_norm": 2.3272152540694324,
+      "learning_rate": 5.8687621210520574e-08,
+      "loss": 0.0132,
+      "step": 18942
+    },
+    {
+      "epoch": 4.305227272727273,
+      "grad_norm": 1.2733016638391095,
+      "learning_rate": 5.8649868611693886e-08,
+      "loss": 0.0045,
+      "step": 18943
+    },
+    {
+      "epoch": 4.305454545454546,
+      "grad_norm": 0.8642596613179551,
+      "learning_rate": 5.8612127561734416e-08,
+      "loss": 0.0051,
+      "step": 18944
+    },
+    {
+      "epoch": 4.305681818181818,
+      "grad_norm": 0.802836878672135,
+      "learning_rate": 5.857439806141174e-08,
+      "loss": 0.0048,
+      "step": 18945
+    },
+    {
+      "epoch": 4.305909090909091,
+      "grad_norm": 2.539153646096159,
+      "learning_rate": 5.85366801114954e-08,
+      "loss": 0.0065,
+      "step": 18946
+    },
+    {
+      "epoch": 4.306136363636364,
+      "grad_norm": 1.297873266495809,
+      "learning_rate": 5.8498973712754344e-08,
+      "loss": 0.007,
+      "step": 18947
+    },
+    {
+      "epoch": 4.306363636363637,
+      "grad_norm": 0.6810129538269851,
+      "learning_rate": 5.846127886595748e-08,
+      "loss": 0.0044,
+      "step": 18948
+    },
+    {
+      "epoch": 4.306590909090909,
+      "grad_norm": 1.1821139996547034,
+      "learning_rate": 5.842359557187351e-08,
+      "loss": 0.0042,
+      "step": 18949
+    },
+    {
+      "epoch": 4.306818181818182,
+      "grad_norm": 0.38251353953203043,
+      "learning_rate": 5.83859238312709e-08,
+      "loss": 0.0021,
+      "step": 18950
+    },
+    {
+      "epoch": 4.307045454545454,
+      "grad_norm": 0.5686797475277013,
+      "learning_rate": 5.8348263644917924e-08,
+      "loss": 0.0028,
+      "step": 18951
+    },
+    {
+      "epoch": 4.307272727272728,
+      "grad_norm": 2.064337188457424,
+      "learning_rate": 5.831061501358232e-08,
+      "loss": 0.0072,
+      "step": 18952
+    },
+    {
+      "epoch": 4.3075,
+      "grad_norm": 0.8971372035556956,
+      "learning_rate": 5.8272977938031925e-08,
+      "loss": 0.0038,
+      "step": 18953
+    },
+    {
+      "epoch": 4.307727272727273,
+      "grad_norm": 1.389801690853383,
+      "learning_rate": 5.823535241903419e-08,
+      "loss": 0.0073,
+      "step": 18954
+    },
+    {
+      "epoch": 4.307954545454545,
+      "grad_norm": 1.6774572348484003,
+      "learning_rate": 5.819773845735642e-08,
+      "loss": 0.0168,
+      "step": 18955
+    },
+    {
+      "epoch": 4.308181818181819,
+      "grad_norm": 0.688166131747783,
+      "learning_rate": 5.8160136053765565e-08,
+      "loss": 0.0025,
+      "step": 18956
+    },
+    {
+      "epoch": 4.308409090909091,
+      "grad_norm": 5.886484203536189,
+      "learning_rate": 5.812254520902845e-08,
+      "loss": 0.0056,
+      "step": 18957
+    },
+    {
+      "epoch": 4.308636363636364,
+      "grad_norm": 1.1845860707143554,
+      "learning_rate": 5.8084965923911675e-08,
+      "loss": 0.0051,
+      "step": 18958
+    },
+    {
+      "epoch": 4.308863636363636,
+      "grad_norm": 1.482644848730106,
+      "learning_rate": 5.8047398199181576e-08,
+      "loss": 0.01,
+      "step": 18959
+    },
+    {
+      "epoch": 4.309090909090909,
+      "grad_norm": 1.692028065674095,
+      "learning_rate": 5.8009842035603935e-08,
+      "loss": 0.0062,
+      "step": 18960
+    },
+    {
+      "epoch": 4.309318181818182,
+      "grad_norm": 1.8007509155186117,
+      "learning_rate": 5.797229743394486e-08,
+      "loss": 0.0076,
+      "step": 18961
+    },
+    {
+      "epoch": 4.309545454545455,
+      "grad_norm": 0.3400712352241173,
+      "learning_rate": 5.793476439496987e-08,
+      "loss": 0.001,
+      "step": 18962
+    },
+    {
+      "epoch": 4.309772727272727,
+      "grad_norm": 1.237978828404479,
+      "learning_rate": 5.78972429194443e-08,
+      "loss": 0.0035,
+      "step": 18963
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 1.2415609803295677,
+      "learning_rate": 5.785973300813347e-08,
+      "loss": 0.01,
+      "step": 18964
+    },
+    {
+      "epoch": 4.310227272727273,
+      "grad_norm": 0.09067262781736406,
+      "learning_rate": 5.782223466180195e-08,
+      "loss": 0.0002,
+      "step": 18965
+    },
+    {
+      "epoch": 4.3104545454545455,
+      "grad_norm": 1.149762281100995,
+      "learning_rate": 5.778474788121463e-08,
+      "loss": 0.0039,
+      "step": 18966
+    },
+    {
+      "epoch": 4.310681818181818,
+      "grad_norm": 2.930413362142535,
+      "learning_rate": 5.774727266713582e-08,
+      "loss": 0.0187,
+      "step": 18967
+    },
+    {
+      "epoch": 4.310909090909091,
+      "grad_norm": 3.3007579137922587,
+      "learning_rate": 5.770980902032986e-08,
+      "loss": 0.0081,
+      "step": 18968
+    },
+    {
+      "epoch": 4.311136363636364,
+      "grad_norm": 0.47507309860867586,
+      "learning_rate": 5.7672356941560484e-08,
+      "loss": 0.001,
+      "step": 18969
+    },
+    {
+      "epoch": 4.3113636363636365,
+      "grad_norm": 0.9419652081127731,
+      "learning_rate": 5.763491643159147e-08,
+      "loss": 0.003,
+      "step": 18970
+    },
+    {
+      "epoch": 4.311590909090909,
+      "grad_norm": 0.9287813135347287,
+      "learning_rate": 5.7597487491186236e-08,
+      "loss": 0.0039,
+      "step": 18971
+    },
+    {
+      "epoch": 4.3118181818181816,
+      "grad_norm": 0.13922684377202435,
+      "learning_rate": 5.756007012110827e-08,
+      "loss": 0.0005,
+      "step": 18972
+    },
+    {
+      "epoch": 4.312045454545455,
+      "grad_norm": 0.32387038742899094,
+      "learning_rate": 5.7522664322120506e-08,
+      "loss": 0.0008,
+      "step": 18973
+    },
+    {
+      "epoch": 4.3122727272727275,
+      "grad_norm": 1.6488794729807734,
+      "learning_rate": 5.7485270094985515e-08,
+      "loss": 0.0087,
+      "step": 18974
+    },
+    {
+      "epoch": 4.3125,
+      "grad_norm": 0.6465434299011433,
+      "learning_rate": 5.744788744046596e-08,
+      "loss": 0.0035,
+      "step": 18975
+    },
+    {
+      "epoch": 4.3127272727272725,
+      "grad_norm": 0.6664246735043554,
+      "learning_rate": 5.7410516359324156e-08,
+      "loss": 0.0021,
+      "step": 18976
+    },
+    {
+      "epoch": 4.312954545454545,
+      "grad_norm": 0.49731399686179684,
+      "learning_rate": 5.7373156852322255e-08,
+      "loss": 0.0019,
+      "step": 18977
+    },
+    {
+      "epoch": 4.3131818181818184,
+      "grad_norm": 1.5005438649871503,
+      "learning_rate": 5.733580892022181e-08,
+      "loss": 0.0116,
+      "step": 18978
+    },
+    {
+      "epoch": 4.313409090909091,
+      "grad_norm": 0.5653340129224498,
+      "learning_rate": 5.729847256378457e-08,
+      "loss": 0.0023,
+      "step": 18979
+    },
+    {
+      "epoch": 4.3136363636363635,
+      "grad_norm": 1.6287113105827444,
+      "learning_rate": 5.726114778377194e-08,
+      "loss": 0.0061,
+      "step": 18980
+    },
+    {
+      "epoch": 4.313863636363636,
+      "grad_norm": 2.4018202162051123,
+      "learning_rate": 5.722383458094505e-08,
+      "loss": 0.0152,
+      "step": 18981
+    },
+    {
+      "epoch": 4.314090909090909,
+      "grad_norm": 0.854674290081881,
+      "learning_rate": 5.718653295606461e-08,
+      "loss": 0.006,
+      "step": 18982
+    },
+    {
+      "epoch": 4.314318181818182,
+      "grad_norm": 2.14094772787142,
+      "learning_rate": 5.7149242909891324e-08,
+      "loss": 0.0125,
+      "step": 18983
+    },
+    {
+      "epoch": 4.3145454545454545,
+      "grad_norm": 0.6754076819821007,
+      "learning_rate": 5.71119644431857e-08,
+      "loss": 0.0015,
+      "step": 18984
+    },
+    {
+      "epoch": 4.314772727272727,
+      "grad_norm": 0.8982968555934376,
+      "learning_rate": 5.707469755670783e-08,
+      "loss": 0.0026,
+      "step": 18985
+    },
+    {
+      "epoch": 4.315,
+      "grad_norm": 0.9691745674028479,
+      "learning_rate": 5.7037442251217656e-08,
+      "loss": 0.0037,
+      "step": 18986
+    },
+    {
+      "epoch": 4.315227272727273,
+      "grad_norm": 3.0661334385151533,
+      "learning_rate": 5.7000198527474916e-08,
+      "loss": 0.0328,
+      "step": 18987
+    },
+    {
+      "epoch": 4.315454545454545,
+      "grad_norm": 0.25376196345679813,
+      "learning_rate": 5.696296638623908e-08,
+      "loss": 0.001,
+      "step": 18988
+    },
+    {
+      "epoch": 4.315681818181818,
+      "grad_norm": 1.107575246627382,
+      "learning_rate": 5.692574582826932e-08,
+      "loss": 0.0018,
+      "step": 18989
+    },
+    {
+      "epoch": 4.315909090909091,
+      "grad_norm": 0.7176922347708442,
+      "learning_rate": 5.688853685432483e-08,
+      "loss": 0.0026,
+      "step": 18990
+    },
+    {
+      "epoch": 4.316136363636364,
+      "grad_norm": 0.3370136951144057,
+      "learning_rate": 5.685133946516402e-08,
+      "loss": 0.0019,
+      "step": 18991
+    },
+    {
+      "epoch": 4.316363636363636,
+      "grad_norm": 2.118471468341355,
+      "learning_rate": 5.6814153661545676e-08,
+      "loss": 0.0212,
+      "step": 18992
+    },
+    {
+      "epoch": 4.316590909090909,
+      "grad_norm": 0.232550573701279,
+      "learning_rate": 5.677697944422793e-08,
+      "loss": 0.001,
+      "step": 18993
+    },
+    {
+      "epoch": 4.316818181818181,
+      "grad_norm": 0.3259359718008651,
+      "learning_rate": 5.673981681396892e-08,
+      "loss": 0.0009,
+      "step": 18994
+    },
+    {
+      "epoch": 4.317045454545455,
+      "grad_norm": 0.353163134186129,
+      "learning_rate": 5.670266577152653e-08,
+      "loss": 0.0012,
+      "step": 18995
+    },
+    {
+      "epoch": 4.317272727272727,
+      "grad_norm": 1.2280953059218858,
+      "learning_rate": 5.666552631765805e-08,
+      "loss": 0.0057,
+      "step": 18996
+    },
+    {
+      "epoch": 4.3175,
+      "grad_norm": 0.14685267659145348,
+      "learning_rate": 5.6628398453121086e-08,
+      "loss": 0.0003,
+      "step": 18997
+    },
+    {
+      "epoch": 4.317727272727272,
+      "grad_norm": 1.0527505186777961,
+      "learning_rate": 5.659128217867266e-08,
+      "loss": 0.0015,
+      "step": 18998
+    },
+    {
+      "epoch": 4.317954545454546,
+      "grad_norm": 0.9312103485268544,
+      "learning_rate": 5.655417749506974e-08,
+      "loss": 0.0019,
+      "step": 18999
+    },
+    {
+      "epoch": 4.318181818181818,
+      "grad_norm": 1.3873512297385666,
+      "learning_rate": 5.651708440306859e-08,
+      "loss": 0.0049,
+      "step": 19000
+    },
+    {
+      "epoch": 4.318409090909091,
+      "grad_norm": 0.5920204315390105,
+      "learning_rate": 5.648000290342598e-08,
+      "loss": 0.004,
+      "step": 19001
+    },
+    {
+      "epoch": 4.318636363636363,
+      "grad_norm": 0.833689441081542,
+      "learning_rate": 5.644293299689796e-08,
+      "loss": 0.0056,
+      "step": 19002
+    },
+    {
+      "epoch": 4.318863636363636,
+      "grad_norm": 0.8724136009114063,
+      "learning_rate": 5.6405874684240595e-08,
+      "loss": 0.0049,
+      "step": 19003
+    },
+    {
+      "epoch": 4.319090909090909,
+      "grad_norm": 0.6954417393794124,
+      "learning_rate": 5.636882796620925e-08,
+      "loss": 0.0036,
+      "step": 19004
+    },
+    {
+      "epoch": 4.319318181818182,
+      "grad_norm": 1.1563984433645536,
+      "learning_rate": 5.633179284355951e-08,
+      "loss": 0.0051,
+      "step": 19005
+    },
+    {
+      "epoch": 4.319545454545454,
+      "grad_norm": 1.5823207382603537,
+      "learning_rate": 5.629476931704667e-08,
+      "loss": 0.0076,
+      "step": 19006
+    },
+    {
+      "epoch": 4.319772727272727,
+      "grad_norm": 0.7856679930512557,
+      "learning_rate": 5.62577573874256e-08,
+      "loss": 0.0066,
+      "step": 19007
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 1.1576364447294385,
+      "learning_rate": 5.62207570554512e-08,
+      "loss": 0.0101,
+      "step": 19008
+    },
+    {
+      "epoch": 4.320227272727273,
+      "grad_norm": 1.2248525766134362,
+      "learning_rate": 5.6183768321877716e-08,
+      "loss": 0.0041,
+      "step": 19009
+    },
+    {
+      "epoch": 4.320454545454545,
+      "grad_norm": 1.0008415634895045,
+      "learning_rate": 5.6146791187459624e-08,
+      "loss": 0.0097,
+      "step": 19010
+    },
+    {
+      "epoch": 4.320681818181818,
+      "grad_norm": 1.6937911265524277,
+      "learning_rate": 5.6109825652950824e-08,
+      "loss": 0.0074,
+      "step": 19011
+    },
+    {
+      "epoch": 4.320909090909091,
+      "grad_norm": 0.5945167240638336,
+      "learning_rate": 5.607287171910531e-08,
+      "loss": 0.0019,
+      "step": 19012
+    },
+    {
+      "epoch": 4.321136363636364,
+      "grad_norm": 1.3740106541590569,
+      "learning_rate": 5.6035929386676344e-08,
+      "loss": 0.0115,
+      "step": 19013
+    },
+    {
+      "epoch": 4.321363636363636,
+      "grad_norm": 1.8631307966252328,
+      "learning_rate": 5.599899865641738e-08,
+      "loss": 0.013,
+      "step": 19014
+    },
+    {
+      "epoch": 4.321590909090909,
+      "grad_norm": 1.9126924448857991,
+      "learning_rate": 5.5962079529081545e-08,
+      "loss": 0.0092,
+      "step": 19015
+    },
+    {
+      "epoch": 4.321818181818182,
+      "grad_norm": 0.4351478519014374,
+      "learning_rate": 5.592517200542165e-08,
+      "loss": 0.001,
+      "step": 19016
+    },
+    {
+      "epoch": 4.322045454545455,
+      "grad_norm": 0.7345191760381374,
+      "learning_rate": 5.588827608619028e-08,
+      "loss": 0.0035,
+      "step": 19017
+    },
+    {
+      "epoch": 4.322272727272727,
+      "grad_norm": 2.179124609581903,
+      "learning_rate": 5.5851391772139827e-08,
+      "loss": 0.0104,
+      "step": 19018
+    },
+    {
+      "epoch": 4.3225,
+      "grad_norm": 1.968560319025573,
+      "learning_rate": 5.5814519064022465e-08,
+      "loss": 0.0115,
+      "step": 19019
+    },
+    {
+      "epoch": 4.322727272727272,
+      "grad_norm": 3.161288869986785,
+      "learning_rate": 5.577765796259002e-08,
+      "loss": 0.0037,
+      "step": 19020
+    },
+    {
+      "epoch": 4.322954545454546,
+      "grad_norm": 0.6772112237445393,
+      "learning_rate": 5.5740808468594324e-08,
+      "loss": 0.0029,
+      "step": 19021
+    },
+    {
+      "epoch": 4.323181818181818,
+      "grad_norm": 2.017122813872374,
+      "learning_rate": 5.570397058278658e-08,
+      "loss": 0.0036,
+      "step": 19022
+    },
+    {
+      "epoch": 4.323409090909091,
+      "grad_norm": 0.4141184490591745,
+      "learning_rate": 5.566714430591806e-08,
+      "loss": 0.0011,
+      "step": 19023
+    },
+    {
+      "epoch": 4.323636363636363,
+      "grad_norm": 0.6453146741240422,
+      "learning_rate": 5.5630329638739695e-08,
+      "loss": 0.0022,
+      "step": 19024
+    },
+    {
+      "epoch": 4.323863636363637,
+      "grad_norm": 0.5677377784868243,
+      "learning_rate": 5.559352658200234e-08,
+      "loss": 0.0019,
+      "step": 19025
+    },
+    {
+      "epoch": 4.324090909090909,
+      "grad_norm": 0.8883375056439182,
+      "learning_rate": 5.555673513645624e-08,
+      "loss": 0.0046,
+      "step": 19026
+    },
+    {
+      "epoch": 4.324318181818182,
+      "grad_norm": 1.2890500852346563,
+      "learning_rate": 5.551995530285181e-08,
+      "loss": 0.0063,
+      "step": 19027
+    },
+    {
+      "epoch": 4.324545454545454,
+      "grad_norm": 1.0244557008510682,
+      "learning_rate": 5.548318708193896e-08,
+      "loss": 0.0051,
+      "step": 19028
+    },
+    {
+      "epoch": 4.324772727272728,
+      "grad_norm": 0.12436755291954947,
+      "learning_rate": 5.54464304744675e-08,
+      "loss": 0.0004,
+      "step": 19029
+    },
+    {
+      "epoch": 4.325,
+      "grad_norm": 0.5744455131531109,
+      "learning_rate": 5.5409685481187036e-08,
+      "loss": 0.0021,
+      "step": 19030
+    },
+    {
+      "epoch": 4.325227272727273,
+      "grad_norm": 0.4144460650230973,
+      "learning_rate": 5.5372952102846694e-08,
+      "loss": 0.0015,
+      "step": 19031
+    },
+    {
+      "epoch": 4.325454545454545,
+      "grad_norm": 0.8136411240937212,
+      "learning_rate": 5.533623034019567e-08,
+      "loss": 0.0037,
+      "step": 19032
+    },
+    {
+      "epoch": 4.325681818181819,
+      "grad_norm": 1.577339036352665,
+      "learning_rate": 5.5299520193982726e-08,
+      "loss": 0.0086,
+      "step": 19033
+    },
+    {
+      "epoch": 4.325909090909091,
+      "grad_norm": 0.8425570661785317,
+      "learning_rate": 5.526282166495659e-08,
+      "loss": 0.002,
+      "step": 19034
+    },
+    {
+      "epoch": 4.326136363636364,
+      "grad_norm": 0.5657429642089921,
+      "learning_rate": 5.522613475386539e-08,
+      "loss": 0.0022,
+      "step": 19035
+    },
+    {
+      "epoch": 4.326363636363636,
+      "grad_norm": 3.304600231044813,
+      "learning_rate": 5.518945946145729e-08,
+      "loss": 0.0083,
+      "step": 19036
+    },
+    {
+      "epoch": 4.326590909090909,
+      "grad_norm": 0.7111833677054255,
+      "learning_rate": 5.5152795788480235e-08,
+      "loss": 0.0019,
+      "step": 19037
+    },
+    {
+      "epoch": 4.326818181818182,
+      "grad_norm": 0.7745799507827589,
+      "learning_rate": 5.511614373568189e-08,
+      "loss": 0.0013,
+      "step": 19038
+    },
+    {
+      "epoch": 4.327045454545455,
+      "grad_norm": 1.6821600184850432,
+      "learning_rate": 5.50795033038095e-08,
+      "loss": 0.0061,
+      "step": 19039
+    },
+    {
+      "epoch": 4.327272727272727,
+      "grad_norm": 0.44662677424357067,
+      "learning_rate": 5.504287449361039e-08,
+      "loss": 0.0015,
+      "step": 19040
+    },
+    {
+      "epoch": 4.3275,
+      "grad_norm": 0.9653247355826682,
+      "learning_rate": 5.500625730583131e-08,
+      "loss": 0.0041,
+      "step": 19041
+    },
+    {
+      "epoch": 4.327727272727273,
+      "grad_norm": 2.149058509212715,
+      "learning_rate": 5.496965174121911e-08,
+      "loss": 0.0117,
+      "step": 19042
+    },
+    {
+      "epoch": 4.327954545454546,
+      "grad_norm": 2.0833658741166428,
+      "learning_rate": 5.493305780052027e-08,
+      "loss": 0.0215,
+      "step": 19043
+    },
+    {
+      "epoch": 4.328181818181818,
+      "grad_norm": 0.5716454787450279,
+      "learning_rate": 5.4896475484480796e-08,
+      "loss": 0.0048,
+      "step": 19044
+    },
+    {
+      "epoch": 4.328409090909091,
+      "grad_norm": 0.3521654643023755,
+      "learning_rate": 5.4859904793846814e-08,
+      "loss": 0.0011,
+      "step": 19045
+    },
+    {
+      "epoch": 4.328636363636363,
+      "grad_norm": 1.3985261367572552,
+      "learning_rate": 5.482334572936406e-08,
+      "loss": 0.0071,
+      "step": 19046
+    },
+    {
+      "epoch": 4.328863636363637,
+      "grad_norm": 0.7548021384817978,
+      "learning_rate": 5.4786798291777963e-08,
+      "loss": 0.0028,
+      "step": 19047
+    },
+    {
+      "epoch": 4.329090909090909,
+      "grad_norm": 0.7503338610537821,
+      "learning_rate": 5.4750262481833846e-08,
+      "loss": 0.0033,
+      "step": 19048
+    },
+    {
+      "epoch": 4.329318181818182,
+      "grad_norm": 2.226771429686414,
+      "learning_rate": 5.471373830027674e-08,
+      "loss": 0.0055,
+      "step": 19049
+    },
+    {
+      "epoch": 4.329545454545454,
+      "grad_norm": 0.7546036988668613,
+      "learning_rate": 5.467722574785145e-08,
+      "loss": 0.0024,
+      "step": 19050
+    },
+    {
+      "epoch": 4.3297727272727276,
+      "grad_norm": 0.8966667747156807,
+      "learning_rate": 5.4640724825302526e-08,
+      "loss": 0.0022,
+      "step": 19051
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 1.4342884845971406,
+      "learning_rate": 5.460423553337431e-08,
+      "loss": 0.0122,
+      "step": 19052
+    },
+    {
+      "epoch": 4.330227272727273,
+      "grad_norm": 1.0066762444696733,
+      "learning_rate": 5.456775787281078e-08,
+      "loss": 0.0077,
+      "step": 19053
+    },
+    {
+      "epoch": 4.330454545454545,
+      "grad_norm": 0.9425474015611451,
+      "learning_rate": 5.4531291844355864e-08,
+      "loss": 0.0111,
+      "step": 19054
+    },
+    {
+      "epoch": 4.3306818181818185,
+      "grad_norm": 1.1399101160222604,
+      "learning_rate": 5.4494837448753056e-08,
+      "loss": 0.0036,
+      "step": 19055
+    },
+    {
+      "epoch": 4.330909090909091,
+      "grad_norm": 0.8246433978857775,
+      "learning_rate": 5.445839468674601e-08,
+      "loss": 0.0025,
+      "step": 19056
+    },
+    {
+      "epoch": 4.331136363636364,
+      "grad_norm": 1.7310874774182596,
+      "learning_rate": 5.442196355907751e-08,
+      "loss": 0.0063,
+      "step": 19057
+    },
+    {
+      "epoch": 4.331363636363636,
+      "grad_norm": 0.8603054455339617,
+      "learning_rate": 5.4385544066490606e-08,
+      "loss": 0.0066,
+      "step": 19058
+    },
+    {
+      "epoch": 4.3315909090909095,
+      "grad_norm": 0.8365710898400247,
+      "learning_rate": 5.434913620972794e-08,
+      "loss": 0.0058,
+      "step": 19059
+    },
+    {
+      "epoch": 4.331818181818182,
+      "grad_norm": 0.6775074487014018,
+      "learning_rate": 5.4312739989531984e-08,
+      "loss": 0.0035,
+      "step": 19060
+    },
+    {
+      "epoch": 4.3320454545454545,
+      "grad_norm": 0.5555297000627096,
+      "learning_rate": 5.427635540664486e-08,
+      "loss": 0.0013,
+      "step": 19061
+    },
+    {
+      "epoch": 4.332272727272727,
+      "grad_norm": 1.7759997727290564,
+      "learning_rate": 5.4239982461808466e-08,
+      "loss": 0.0118,
+      "step": 19062
+    },
+    {
+      "epoch": 4.3325,
+      "grad_norm": 0.3975491486544563,
+      "learning_rate": 5.4203621155764636e-08,
+      "loss": 0.0063,
+      "step": 19063
+    },
+    {
+      "epoch": 4.332727272727273,
+      "grad_norm": 0.4085640249518531,
+      "learning_rate": 5.4167271489254795e-08,
+      "loss": 0.0013,
+      "step": 19064
+    },
+    {
+      "epoch": 4.3329545454545455,
+      "grad_norm": 0.38136477790067236,
+      "learning_rate": 5.413093346302023e-08,
+      "loss": 0.0013,
+      "step": 19065
+    },
+    {
+      "epoch": 4.333181818181818,
+      "grad_norm": 1.2233001760931077,
+      "learning_rate": 5.40946070778018e-08,
+      "loss": 0.0051,
+      "step": 19066
+    },
+    {
+      "epoch": 4.3334090909090905,
+      "grad_norm": 0.5277108193706971,
+      "learning_rate": 5.40582923343403e-08,
+      "loss": 0.0033,
+      "step": 19067
+    },
+    {
+      "epoch": 4.333636363636364,
+      "grad_norm": 1.439004994376818,
+      "learning_rate": 5.402198923337633e-08,
+      "loss": 0.0092,
+      "step": 19068
+    },
+    {
+      "epoch": 4.3338636363636365,
+      "grad_norm": 1.7508991155275508,
+      "learning_rate": 5.398569777565019e-08,
+      "loss": 0.0055,
+      "step": 19069
+    },
+    {
+      "epoch": 4.334090909090909,
+      "grad_norm": 0.42448289901169234,
+      "learning_rate": 5.39494179619017e-08,
+      "loss": 0.0012,
+      "step": 19070
+    },
+    {
+      "epoch": 4.3343181818181815,
+      "grad_norm": 0.647066939371811,
+      "learning_rate": 5.391314979287091e-08,
+      "loss": 0.0027,
+      "step": 19071
+    },
+    {
+      "epoch": 4.334545454545455,
+      "grad_norm": 1.258269896456324,
+      "learning_rate": 5.387689326929728e-08,
+      "loss": 0.0101,
+      "step": 19072
+    },
+    {
+      "epoch": 4.334772727272727,
+      "grad_norm": 1.3001787645110139,
+      "learning_rate": 5.3840648391920225e-08,
+      "loss": 0.0077,
+      "step": 19073
+    },
+    {
+      "epoch": 4.335,
+      "grad_norm": 2.4217616228701626,
+      "learning_rate": 5.380441516147882e-08,
+      "loss": 0.0287,
+      "step": 19074
+    },
+    {
+      "epoch": 4.3352272727272725,
+      "grad_norm": 1.9369046270278707,
+      "learning_rate": 5.376819357871182e-08,
+      "loss": 0.0107,
+      "step": 19075
+    },
+    {
+      "epoch": 4.335454545454546,
+      "grad_norm": 8.23102092297842,
+      "learning_rate": 5.373198364435797e-08,
+      "loss": 0.0091,
+      "step": 19076
+    },
+    {
+      "epoch": 4.335681818181818,
+      "grad_norm": 2.3497397086660707,
+      "learning_rate": 5.3695785359155556e-08,
+      "loss": 0.011,
+      "step": 19077
+    },
+    {
+      "epoch": 4.335909090909091,
+      "grad_norm": 1.7166314798803057,
+      "learning_rate": 5.365959872384273e-08,
+      "loss": 0.0056,
+      "step": 19078
+    },
+    {
+      "epoch": 4.3361363636363635,
+      "grad_norm": 0.651118333481773,
+      "learning_rate": 5.362342373915752e-08,
+      "loss": 0.0032,
+      "step": 19079
+    },
+    {
+      "epoch": 4.336363636363636,
+      "grad_norm": 2.2180583420112456,
+      "learning_rate": 5.3587260405837466e-08,
+      "loss": 0.0044,
+      "step": 19080
+    },
+    {
+      "epoch": 4.336590909090909,
+      "grad_norm": 1.1508655877169922,
+      "learning_rate": 5.355110872462009e-08,
+      "loss": 0.0047,
+      "step": 19081
+    },
+    {
+      "epoch": 4.336818181818182,
+      "grad_norm": 1.706339061984474,
+      "learning_rate": 5.351496869624266e-08,
+      "loss": 0.0066,
+      "step": 19082
+    },
+    {
+      "epoch": 4.337045454545454,
+      "grad_norm": 0.8995632911923649,
+      "learning_rate": 5.347884032144195e-08,
+      "loss": 0.0052,
+      "step": 19083
+    },
+    {
+      "epoch": 4.337272727272727,
+      "grad_norm": 0.4973043296194708,
+      "learning_rate": 5.344272360095472e-08,
+      "loss": 0.0021,
+      "step": 19084
+    },
+    {
+      "epoch": 4.3375,
+      "grad_norm": 1.610186117719736,
+      "learning_rate": 5.340661853551747e-08,
+      "loss": 0.0159,
+      "step": 19085
+    },
+    {
+      "epoch": 4.337727272727273,
+      "grad_norm": 1.429333687957305,
+      "learning_rate": 5.337052512586649e-08,
+      "loss": 0.0033,
+      "step": 19086
+    },
+    {
+      "epoch": 4.337954545454545,
+      "grad_norm": 1.7825579334069297,
+      "learning_rate": 5.3334443372737926e-08,
+      "loss": 0.0101,
+      "step": 19087
+    },
+    {
+      "epoch": 4.338181818181818,
+      "grad_norm": 0.6252298039587593,
+      "learning_rate": 5.3298373276867226e-08,
+      "loss": 0.0037,
+      "step": 19088
+    },
+    {
+      "epoch": 4.33840909090909,
+      "grad_norm": 3.787998153007591,
+      "learning_rate": 5.3262314838990057e-08,
+      "loss": 0.007,
+      "step": 19089
+    },
+    {
+      "epoch": 4.338636363636364,
+      "grad_norm": 0.886293118263976,
+      "learning_rate": 5.3226268059841795e-08,
+      "loss": 0.0019,
+      "step": 19090
+    },
+    {
+      "epoch": 4.338863636363636,
+      "grad_norm": 0.6550847286313332,
+      "learning_rate": 5.319023294015742e-08,
+      "loss": 0.005,
+      "step": 19091
+    },
+    {
+      "epoch": 4.339090909090909,
+      "grad_norm": 1.3488984950551588,
+      "learning_rate": 5.3154209480671753e-08,
+      "loss": 0.012,
+      "step": 19092
+    },
+    {
+      "epoch": 4.339318181818181,
+      "grad_norm": 1.4366575616117925,
+      "learning_rate": 5.311819768211941e-08,
+      "loss": 0.0046,
+      "step": 19093
+    },
+    {
+      "epoch": 4.339545454545455,
+      "grad_norm": 0.7610428096365017,
+      "learning_rate": 5.308219754523475e-08,
+      "loss": 0.0034,
+      "step": 19094
+    },
+    {
+      "epoch": 4.339772727272727,
+      "grad_norm": 1.839775513206789,
+      "learning_rate": 5.304620907075182e-08,
+      "loss": 0.0093,
+      "step": 19095
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 1.3361441447642812,
+      "learning_rate": 5.301023225940463e-08,
+      "loss": 0.0078,
+      "step": 19096
+    },
+    {
+      "epoch": 4.340227272727272,
+      "grad_norm": 0.6424114497510734,
+      "learning_rate": 5.297426711192655e-08,
+      "loss": 0.0021,
+      "step": 19097
+    },
+    {
+      "epoch": 4.340454545454546,
+      "grad_norm": 0.28556200307205454,
+      "learning_rate": 5.293831362905115e-08,
+      "loss": 0.001,
+      "step": 19098
+    },
+    {
+      "epoch": 4.340681818181818,
+      "grad_norm": 1.1666048412531371,
+      "learning_rate": 5.290237181151153e-08,
+      "loss": 0.0051,
+      "step": 19099
+    },
+    {
+      "epoch": 4.340909090909091,
+      "grad_norm": 1.5026794674492236,
+      "learning_rate": 5.286644166004079e-08,
+      "loss": 0.0079,
+      "step": 19100
+    },
+    {
+      "epoch": 4.341136363636363,
+      "grad_norm": 1.3674285148507732,
+      "learning_rate": 5.283052317537125e-08,
+      "loss": 0.0094,
+      "step": 19101
+    },
+    {
+      "epoch": 4.341363636363637,
+      "grad_norm": 1.1478520555468752,
+      "learning_rate": 5.27946163582356e-08,
+      "loss": 0.0083,
+      "step": 19102
+    },
+    {
+      "epoch": 4.341590909090909,
+      "grad_norm": 0.2687154334958447,
+      "learning_rate": 5.275872120936602e-08,
+      "loss": 0.0017,
+      "step": 19103
+    },
+    {
+      "epoch": 4.341818181818182,
+      "grad_norm": 1.291057004282071,
+      "learning_rate": 5.2722837729494506e-08,
+      "loss": 0.003,
+      "step": 19104
+    },
+    {
+      "epoch": 4.342045454545454,
+      "grad_norm": 0.725337753426431,
+      "learning_rate": 5.268696591935261e-08,
+      "loss": 0.0022,
+      "step": 19105
+    },
+    {
+      "epoch": 4.342272727272727,
+      "grad_norm": 0.7791874817436968,
+      "learning_rate": 5.265110577967199e-08,
+      "loss": 0.0044,
+      "step": 19106
+    },
+    {
+      "epoch": 4.3425,
+      "grad_norm": 1.27565522590499,
+      "learning_rate": 5.2615257311183776e-08,
+      "loss": 0.0026,
+      "step": 19107
+    },
+    {
+      "epoch": 4.342727272727273,
+      "grad_norm": 1.5491356441238295,
+      "learning_rate": 5.257942051461906e-08,
+      "loss": 0.013,
+      "step": 19108
+    },
+    {
+      "epoch": 4.342954545454545,
+      "grad_norm": 0.6582654268984551,
+      "learning_rate": 5.254359539070865e-08,
+      "loss": 0.0038,
+      "step": 19109
+    },
+    {
+      "epoch": 4.343181818181818,
+      "grad_norm": 0.3657291819044734,
+      "learning_rate": 5.2507781940182996e-08,
+      "loss": 0.0013,
+      "step": 19110
+    },
+    {
+      "epoch": 4.343409090909091,
+      "grad_norm": 0.61666051207714,
+      "learning_rate": 5.247198016377242e-08,
+      "loss": 0.003,
+      "step": 19111
+    },
+    {
+      "epoch": 4.343636363636364,
+      "grad_norm": 1.4525412456104168,
+      "learning_rate": 5.243619006220703e-08,
+      "loss": 0.0053,
+      "step": 19112
+    },
+    {
+      "epoch": 4.343863636363636,
+      "grad_norm": 2.176939357537824,
+      "learning_rate": 5.240041163621673e-08,
+      "loss": 0.0063,
+      "step": 19113
+    },
+    {
+      "epoch": 4.344090909090909,
+      "grad_norm": 1.66755516105178,
+      "learning_rate": 5.2364644886530866e-08,
+      "loss": 0.0129,
+      "step": 19114
+    },
+    {
+      "epoch": 4.344318181818182,
+      "grad_norm": 0.845224026628158,
+      "learning_rate": 5.232888981387886e-08,
+      "loss": 0.0099,
+      "step": 19115
+    },
+    {
+      "epoch": 4.344545454545455,
+      "grad_norm": 0.36198059329844345,
+      "learning_rate": 5.2293146418989914e-08,
+      "loss": 0.0009,
+      "step": 19116
+    },
+    {
+      "epoch": 4.344772727272727,
+      "grad_norm": 1.948349330635055,
+      "learning_rate": 5.2257414702592894e-08,
+      "loss": 0.0025,
+      "step": 19117
+    },
+    {
+      "epoch": 4.345,
+      "grad_norm": 1.4479870391127818,
+      "learning_rate": 5.222169466541645e-08,
+      "loss": 0.0036,
+      "step": 19118
+    },
+    {
+      "epoch": 4.345227272727273,
+      "grad_norm": 0.6129165236119305,
+      "learning_rate": 5.218598630818882e-08,
+      "loss": 0.0016,
+      "step": 19119
+    },
+    {
+      "epoch": 4.345454545454546,
+      "grad_norm": 1.2650996727544281,
+      "learning_rate": 5.21502896316383e-08,
+      "loss": 0.0039,
+      "step": 19120
+    },
+    {
+      "epoch": 4.345681818181818,
+      "grad_norm": 1.9856109480031436,
+      "learning_rate": 5.211460463649273e-08,
+      "loss": 0.0103,
+      "step": 19121
+    },
+    {
+      "epoch": 4.345909090909091,
+      "grad_norm": 0.3862619015096834,
+      "learning_rate": 5.2078931323479785e-08,
+      "loss": 0.0009,
+      "step": 19122
+    },
+    {
+      "epoch": 4.346136363636363,
+      "grad_norm": 0.43945011360535347,
+      "learning_rate": 5.2043269693327e-08,
+      "loss": 0.0016,
+      "step": 19123
+    },
+    {
+      "epoch": 4.346363636363637,
+      "grad_norm": 1.7090422760110424,
+      "learning_rate": 5.200761974676151e-08,
+      "loss": 0.0115,
+      "step": 19124
+    },
+    {
+      "epoch": 4.346590909090909,
+      "grad_norm": 10.306691986847316,
+      "learning_rate": 5.19719814845103e-08,
+      "loss": 0.0022,
+      "step": 19125
+    },
+    {
+      "epoch": 4.346818181818182,
+      "grad_norm": 0.62051312135726,
+      "learning_rate": 5.193635490730022e-08,
+      "loss": 0.0022,
+      "step": 19126
+    },
+    {
+      "epoch": 4.347045454545454,
+      "grad_norm": 0.5699837028651509,
+      "learning_rate": 5.190074001585749e-08,
+      "loss": 0.0047,
+      "step": 19127
+    },
+    {
+      "epoch": 4.347272727272728,
+      "grad_norm": 1.7108891540799278,
+      "learning_rate": 5.186513681090847e-08,
+      "loss": 0.0052,
+      "step": 19128
+    },
+    {
+      "epoch": 4.3475,
+      "grad_norm": 0.8850478017130072,
+      "learning_rate": 5.1829545293179264e-08,
+      "loss": 0.0057,
+      "step": 19129
+    },
+    {
+      "epoch": 4.347727272727273,
+      "grad_norm": 1.7383239445213097,
+      "learning_rate": 5.179396546339553e-08,
+      "loss": 0.0067,
+      "step": 19130
+    },
+    {
+      "epoch": 4.347954545454545,
+      "grad_norm": 0.9604909754023971,
+      "learning_rate": 5.175839732228294e-08,
+      "loss": 0.0075,
+      "step": 19131
+    },
+    {
+      "epoch": 4.348181818181819,
+      "grad_norm": 0.5653185041641275,
+      "learning_rate": 5.1722840870566674e-08,
+      "loss": 0.0034,
+      "step": 19132
+    },
+    {
+      "epoch": 4.348409090909091,
+      "grad_norm": 0.9902833626834291,
+      "learning_rate": 5.1687296108971775e-08,
+      "loss": 0.005,
+      "step": 19133
+    },
+    {
+      "epoch": 4.348636363636364,
+      "grad_norm": 1.0917327016970677,
+      "learning_rate": 5.165176303822308e-08,
+      "loss": 0.0058,
+      "step": 19134
+    },
+    {
+      "epoch": 4.348863636363636,
+      "grad_norm": 0.6170323914924909,
+      "learning_rate": 5.161624165904537e-08,
+      "loss": 0.0043,
+      "step": 19135
+    },
+    {
+      "epoch": 4.34909090909091,
+      "grad_norm": 0.11536665475355948,
+      "learning_rate": 5.158073197216264e-08,
+      "loss": 0.0004,
+      "step": 19136
+    },
+    {
+      "epoch": 4.349318181818182,
+      "grad_norm": 0.6920542323701558,
+      "learning_rate": 5.154523397829923e-08,
+      "loss": 0.0023,
+      "step": 19137
+    },
+    {
+      "epoch": 4.349545454545455,
+      "grad_norm": 0.6109432520542253,
+      "learning_rate": 5.150974767817882e-08,
+      "loss": 0.0016,
+      "step": 19138
+    },
+    {
+      "epoch": 4.349772727272727,
+      "grad_norm": 1.5715126119280824,
+      "learning_rate": 5.147427307252532e-08,
+      "loss": 0.0109,
+      "step": 19139
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 1.3005171920351921,
+      "learning_rate": 5.143881016206193e-08,
+      "loss": 0.0102,
+      "step": 19140
+    },
+    {
+      "epoch": 4.350227272727273,
+      "grad_norm": 0.921411342832468,
+      "learning_rate": 5.140335894751179e-08,
+      "loss": 0.0061,
+      "step": 19141
+    },
+    {
+      "epoch": 4.350454545454546,
+      "grad_norm": 0.813760357892201,
+      "learning_rate": 5.136791942959789e-08,
+      "loss": 0.0048,
+      "step": 19142
+    },
+    {
+      "epoch": 4.350681818181818,
+      "grad_norm": 0.41322264158676214,
+      "learning_rate": 5.133249160904283e-08,
+      "loss": 0.0018,
+      "step": 19143
+    },
+    {
+      "epoch": 4.350909090909091,
+      "grad_norm": 0.6158093118261185,
+      "learning_rate": 5.129707548656923e-08,
+      "loss": 0.0024,
+      "step": 19144
+    },
+    {
+      "epoch": 4.351136363636364,
+      "grad_norm": 1.1334683182863317,
+      "learning_rate": 5.1261671062899016e-08,
+      "loss": 0.0184,
+      "step": 19145
+    },
+    {
+      "epoch": 4.3513636363636365,
+      "grad_norm": 1.3673137195793303,
+      "learning_rate": 5.122627833875425e-08,
+      "loss": 0.0088,
+      "step": 19146
+    },
+    {
+      "epoch": 4.351590909090909,
+      "grad_norm": 0.3035402912015344,
+      "learning_rate": 5.1190897314856713e-08,
+      "loss": 0.0014,
+      "step": 19147
+    },
+    {
+      "epoch": 4.351818181818182,
+      "grad_norm": 0.28261393768385307,
+      "learning_rate": 5.115552799192791e-08,
+      "loss": 0.001,
+      "step": 19148
+    },
+    {
+      "epoch": 4.352045454545454,
+      "grad_norm": 0.3639034177461748,
+      "learning_rate": 5.112017037068893e-08,
+      "loss": 0.001,
+      "step": 19149
+    },
+    {
+      "epoch": 4.3522727272727275,
+      "grad_norm": 1.1016772939143276,
+      "learning_rate": 5.108482445186087e-08,
+      "loss": 0.0067,
+      "step": 19150
+    },
+    {
+      "epoch": 4.3525,
+      "grad_norm": 0.7267758527995402,
+      "learning_rate": 5.104949023616447e-08,
+      "loss": 0.0039,
+      "step": 19151
+    },
+    {
+      "epoch": 4.352727272727273,
+      "grad_norm": 0.6576921434229963,
+      "learning_rate": 5.101416772432033e-08,
+      "loss": 0.0021,
+      "step": 19152
+    },
+    {
+      "epoch": 4.352954545454545,
+      "grad_norm": 1.208490275989519,
+      "learning_rate": 5.097885691704864e-08,
+      "loss": 0.003,
+      "step": 19153
+    },
+    {
+      "epoch": 4.3531818181818185,
+      "grad_norm": 1.478726174628654,
+      "learning_rate": 5.094355781506945e-08,
+      "loss": 0.0093,
+      "step": 19154
+    },
+    {
+      "epoch": 4.353409090909091,
+      "grad_norm": 0.7471412708164372,
+      "learning_rate": 5.0908270419102666e-08,
+      "loss": 0.0054,
+      "step": 19155
+    },
+    {
+      "epoch": 4.3536363636363635,
+      "grad_norm": 1.5256174855749094,
+      "learning_rate": 5.087299472986778e-08,
+      "loss": 0.0042,
+      "step": 19156
+    },
+    {
+      "epoch": 4.353863636363636,
+      "grad_norm": 1.7449905733298126,
+      "learning_rate": 5.083773074808431e-08,
+      "loss": 0.0058,
+      "step": 19157
+    },
+    {
+      "epoch": 4.3540909090909095,
+      "grad_norm": 1.0090929280682004,
+      "learning_rate": 5.080247847447109e-08,
+      "loss": 0.0047,
+      "step": 19158
+    },
+    {
+      "epoch": 4.354318181818182,
+      "grad_norm": 0.6590493611742388,
+      "learning_rate": 5.076723790974701e-08,
+      "loss": 0.0018,
+      "step": 19159
+    },
+    {
+      "epoch": 4.3545454545454545,
+      "grad_norm": 1.088919825302916,
+      "learning_rate": 5.07320090546308e-08,
+      "loss": 0.0043,
+      "step": 19160
+    },
+    {
+      "epoch": 4.354772727272727,
+      "grad_norm": 1.196015198405553,
+      "learning_rate": 5.0696791909840834e-08,
+      "loss": 0.004,
+      "step": 19161
+    },
+    {
+      "epoch": 4.355,
+      "grad_norm": 0.5081675142609845,
+      "learning_rate": 5.0661586476095165e-08,
+      "loss": 0.0025,
+      "step": 19162
+    },
+    {
+      "epoch": 4.355227272727273,
+      "grad_norm": 0.6647507082090017,
+      "learning_rate": 5.0626392754111744e-08,
+      "loss": 0.0024,
+      "step": 19163
+    },
+    {
+      "epoch": 4.3554545454545455,
+      "grad_norm": 0.9694141912200684,
+      "learning_rate": 5.0591210744608195e-08,
+      "loss": 0.0039,
+      "step": 19164
+    },
+    {
+      "epoch": 4.355681818181818,
+      "grad_norm": 1.5516962338122242,
+      "learning_rate": 5.055604044830194e-08,
+      "loss": 0.0037,
+      "step": 19165
+    },
+    {
+      "epoch": 4.3559090909090905,
+      "grad_norm": 1.0386840265472608,
+      "learning_rate": 5.052088186591031e-08,
+      "loss": 0.0031,
+      "step": 19166
+    },
+    {
+      "epoch": 4.356136363636364,
+      "grad_norm": 0.1521672723664277,
+      "learning_rate": 5.04857349981501e-08,
+      "loss": 0.0004,
+      "step": 19167
+    },
+    {
+      "epoch": 4.356363636363636,
+      "grad_norm": 0.7504273830224902,
+      "learning_rate": 5.045059984573789e-08,
+      "loss": 0.0022,
+      "step": 19168
+    },
+    {
+      "epoch": 4.356590909090909,
+      "grad_norm": 1.8177192802278574,
+      "learning_rate": 5.041547640939039e-08,
+      "loss": 0.0054,
+      "step": 19169
+    },
+    {
+      "epoch": 4.3568181818181815,
+      "grad_norm": 2.1952991234647232,
+      "learning_rate": 5.038036468982384e-08,
+      "loss": 0.0071,
+      "step": 19170
+    },
+    {
+      "epoch": 4.357045454545455,
+      "grad_norm": 1.2992445935574397,
+      "learning_rate": 5.0345264687754056e-08,
+      "loss": 0.0093,
+      "step": 19171
+    },
+    {
+      "epoch": 4.357272727272727,
+      "grad_norm": 0.9758884169250787,
+      "learning_rate": 5.0310176403896916e-08,
+      "loss": 0.0022,
+      "step": 19172
+    },
+    {
+      "epoch": 4.3575,
+      "grad_norm": 1.3650762373405814,
+      "learning_rate": 5.027509983896783e-08,
+      "loss": 0.0134,
+      "step": 19173
+    },
+    {
+      "epoch": 4.3577272727272724,
+      "grad_norm": 0.7520600198815186,
+      "learning_rate": 5.024003499368206e-08,
+      "loss": 0.0039,
+      "step": 19174
+    },
+    {
+      "epoch": 4.357954545454546,
+      "grad_norm": 0.5783781414741034,
+      "learning_rate": 5.020498186875493e-08,
+      "loss": 0.0014,
+      "step": 19175
+    },
+    {
+      "epoch": 4.358181818181818,
+      "grad_norm": 3.6370727330620887,
+      "learning_rate": 5.0169940464900795e-08,
+      "loss": 0.0258,
+      "step": 19176
+    },
+    {
+      "epoch": 4.358409090909091,
+      "grad_norm": 0.7857398405359505,
+      "learning_rate": 5.0134910782834444e-08,
+      "loss": 0.0025,
+      "step": 19177
+    },
+    {
+      "epoch": 4.358636363636363,
+      "grad_norm": 0.6116157166975449,
+      "learning_rate": 5.009989282327023e-08,
+      "loss": 0.0031,
+      "step": 19178
+    },
+    {
+      "epoch": 4.358863636363637,
+      "grad_norm": 2.5737939486889503,
+      "learning_rate": 5.006488658692223e-08,
+      "loss": 0.0243,
+      "step": 19179
+    },
+    {
+      "epoch": 4.359090909090909,
+      "grad_norm": 0.5511088113823099,
+      "learning_rate": 5.002989207450411e-08,
+      "loss": 0.0023,
+      "step": 19180
+    },
+    {
+      "epoch": 4.359318181818182,
+      "grad_norm": 4.253224302734607,
+      "learning_rate": 4.9994909286729606e-08,
+      "loss": 0.0044,
+      "step": 19181
+    },
+    {
+      "epoch": 4.359545454545454,
+      "grad_norm": 1.02600515656692,
+      "learning_rate": 4.995993822431204e-08,
+      "loss": 0.0055,
+      "step": 19182
+    },
+    {
+      "epoch": 4.359772727272727,
+      "grad_norm": 1.1537922330076607,
+      "learning_rate": 4.992497888796452e-08,
+      "loss": 0.0032,
+      "step": 19183
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 1.752287590639894,
+      "learning_rate": 4.989003127839995e-08,
+      "loss": 0.0079,
+      "step": 19184
+    },
+    {
+      "epoch": 4.360227272727273,
+      "grad_norm": 0.5231059716134333,
+      "learning_rate": 4.9855095396331017e-08,
+      "loss": 0.003,
+      "step": 19185
+    },
+    {
+      "epoch": 4.360454545454545,
+      "grad_norm": 0.36733130491694854,
+      "learning_rate": 4.9820171242470074e-08,
+      "loss": 0.0009,
+      "step": 19186
+    },
+    {
+      "epoch": 4.360681818181818,
+      "grad_norm": 0.7977050071766166,
+      "learning_rate": 4.978525881752932e-08,
+      "loss": 0.0063,
+      "step": 19187
+    },
+    {
+      "epoch": 4.360909090909091,
+      "grad_norm": 1.0967708493160426,
+      "learning_rate": 4.975035812222071e-08,
+      "loss": 0.0034,
+      "step": 19188
+    },
+    {
+      "epoch": 4.361136363636364,
+      "grad_norm": 2.2417866356614176,
+      "learning_rate": 4.9715469157255795e-08,
+      "loss": 0.0158,
+      "step": 19189
+    },
+    {
+      "epoch": 4.361363636363636,
+      "grad_norm": 1.0657010853815039,
+      "learning_rate": 4.968059192334611e-08,
+      "loss": 0.0054,
+      "step": 19190
+    },
+    {
+      "epoch": 4.361590909090909,
+      "grad_norm": 0.7619316879801722,
+      "learning_rate": 4.9645726421202815e-08,
+      "loss": 0.0015,
+      "step": 19191
+    },
+    {
+      "epoch": 4.361818181818181,
+      "grad_norm": 1.6565472104820367,
+      "learning_rate": 4.961087265153708e-08,
+      "loss": 0.0143,
+      "step": 19192
+    },
+    {
+      "epoch": 4.362045454545455,
+      "grad_norm": 4.59940419753315,
+      "learning_rate": 4.957603061505938e-08,
+      "loss": 0.0195,
+      "step": 19193
+    },
+    {
+      "epoch": 4.362272727272727,
+      "grad_norm": 0.48133702015300267,
+      "learning_rate": 4.9541200312480246e-08,
+      "loss": 0.0009,
+      "step": 19194
+    },
+    {
+      "epoch": 4.3625,
+      "grad_norm": 2.0114919194687677,
+      "learning_rate": 4.950638174450996e-08,
+      "loss": 0.005,
+      "step": 19195
+    },
+    {
+      "epoch": 4.362727272727272,
+      "grad_norm": 0.6632631889310524,
+      "learning_rate": 4.947157491185864e-08,
+      "loss": 0.0022,
+      "step": 19196
+    },
+    {
+      "epoch": 4.362954545454546,
+      "grad_norm": 0.31941997219307033,
+      "learning_rate": 4.9436779815235926e-08,
+      "loss": 0.0012,
+      "step": 19197
+    },
+    {
+      "epoch": 4.363181818181818,
+      "grad_norm": 1.1576983244985268,
+      "learning_rate": 4.9401996455351386e-08,
+      "loss": 0.0046,
+      "step": 19198
+    },
+    {
+      "epoch": 4.363409090909091,
+      "grad_norm": 2.849012900475155,
+      "learning_rate": 4.9367224832914324e-08,
+      "loss": 0.0058,
+      "step": 19199
+    },
+    {
+      "epoch": 4.363636363636363,
+      "grad_norm": 0.5398942613808689,
+      "learning_rate": 4.933246494863375e-08,
+      "loss": 0.0014,
+      "step": 19200
+    },
+    {
+      "epoch": 4.363863636363637,
+      "grad_norm": 0.33496015573647236,
+      "learning_rate": 4.929771680321875e-08,
+      "loss": 0.0012,
+      "step": 19201
+    },
+    {
+      "epoch": 4.364090909090909,
+      "grad_norm": 0.6787164765706889,
+      "learning_rate": 4.9262980397377437e-08,
+      "loss": 0.003,
+      "step": 19202
+    },
+    {
+      "epoch": 4.364318181818182,
+      "grad_norm": 1.9905221377862194,
+      "learning_rate": 4.922825573181848e-08,
+      "loss": 0.0047,
+      "step": 19203
+    },
+    {
+      "epoch": 4.364545454545454,
+      "grad_norm": 0.5816117350905479,
+      "learning_rate": 4.919354280724979e-08,
+      "loss": 0.0031,
+      "step": 19204
+    },
+    {
+      "epoch": 4.364772727272728,
+      "grad_norm": 0.778487834552075,
+      "learning_rate": 4.915884162437948e-08,
+      "loss": 0.0029,
+      "step": 19205
+    },
+    {
+      "epoch": 4.365,
+      "grad_norm": 1.3342610450571484,
+      "learning_rate": 4.912415218391489e-08,
+      "loss": 0.0051,
+      "step": 19206
+    },
+    {
+      "epoch": 4.365227272727273,
+      "grad_norm": 0.6880322749454951,
+      "learning_rate": 4.908947448656352e-08,
+      "loss": 0.0091,
+      "step": 19207
+    },
+    {
+      "epoch": 4.365454545454545,
+      "grad_norm": 1.0278421504681428,
+      "learning_rate": 4.9054808533032444e-08,
+      "loss": 0.005,
+      "step": 19208
+    },
+    {
+      "epoch": 4.365681818181818,
+      "grad_norm": 1.114729127836697,
+      "learning_rate": 4.902015432402866e-08,
+      "loss": 0.0081,
+      "step": 19209
+    },
+    {
+      "epoch": 4.365909090909091,
+      "grad_norm": 0.7180780896471857,
+      "learning_rate": 4.89855118602589e-08,
+      "loss": 0.0028,
+      "step": 19210
+    },
+    {
+      "epoch": 4.366136363636364,
+      "grad_norm": 0.8913666385863064,
+      "learning_rate": 4.895088114242927e-08,
+      "loss": 0.0058,
+      "step": 19211
+    },
+    {
+      "epoch": 4.366363636363636,
+      "grad_norm": 1.8117628989638004,
+      "learning_rate": 4.89162621712462e-08,
+      "loss": 0.0069,
+      "step": 19212
+    },
+    {
+      "epoch": 4.366590909090909,
+      "grad_norm": 0.4156706724051043,
+      "learning_rate": 4.8881654947415605e-08,
+      "loss": 0.0014,
+      "step": 19213
+    },
+    {
+      "epoch": 4.366818181818182,
+      "grad_norm": 3.269070008424606,
+      "learning_rate": 4.8847059471643155e-08,
+      "loss": 0.0066,
+      "step": 19214
+    },
+    {
+      "epoch": 4.367045454545455,
+      "grad_norm": 0.2510148118815246,
+      "learning_rate": 4.881247574463427e-08,
+      "loss": 0.0008,
+      "step": 19215
+    },
+    {
+      "epoch": 4.367272727272727,
+      "grad_norm": 0.6832016929577497,
+      "learning_rate": 4.877790376709428e-08,
+      "loss": 0.0027,
+      "step": 19216
+    },
+    {
+      "epoch": 4.3675,
+      "grad_norm": 0.32719865440504076,
+      "learning_rate": 4.874334353972804e-08,
+      "loss": 0.0023,
+      "step": 19217
+    },
+    {
+      "epoch": 4.367727272727273,
+      "grad_norm": 0.6578979648547895,
+      "learning_rate": 4.87087950632404e-08,
+      "loss": 0.0022,
+      "step": 19218
+    },
+    {
+      "epoch": 4.367954545454546,
+      "grad_norm": 0.9274543889301093,
+      "learning_rate": 4.867425833833595e-08,
+      "loss": 0.0057,
+      "step": 19219
+    },
+    {
+      "epoch": 4.368181818181818,
+      "grad_norm": 1.2365743368144713,
+      "learning_rate": 4.8639733365718686e-08,
+      "loss": 0.0029,
+      "step": 19220
+    },
+    {
+      "epoch": 4.368409090909091,
+      "grad_norm": 2.411520849956863,
+      "learning_rate": 4.860522014609278e-08,
+      "loss": 0.0093,
+      "step": 19221
+    },
+    {
+      "epoch": 4.368636363636364,
+      "grad_norm": 0.5663678507184038,
+      "learning_rate": 4.857071868016205e-08,
+      "loss": 0.0022,
+      "step": 19222
+    },
+    {
+      "epoch": 4.368863636363637,
+      "grad_norm": 1.7875385999181899,
+      "learning_rate": 4.853622896863003e-08,
+      "loss": 0.0079,
+      "step": 19223
+    },
+    {
+      "epoch": 4.369090909090909,
+      "grad_norm": 1.0753901604253502,
+      "learning_rate": 4.8501751012199964e-08,
+      "loss": 0.0035,
+      "step": 19224
+    },
+    {
+      "epoch": 4.369318181818182,
+      "grad_norm": 0.9819659281047429,
+      "learning_rate": 4.846728481157491e-08,
+      "loss": 0.003,
+      "step": 19225
+    },
+    {
+      "epoch": 4.369545454545454,
+      "grad_norm": 4.0547548805576366,
+      "learning_rate": 4.8432830367457775e-08,
+      "loss": 0.0273,
+      "step": 19226
+    },
+    {
+      "epoch": 4.369772727272728,
+      "grad_norm": 1.009201303225784,
+      "learning_rate": 4.839838768055105e-08,
+      "loss": 0.0027,
+      "step": 19227
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 0.2910975451196997,
+      "learning_rate": 4.836395675155717e-08,
+      "loss": 0.0007,
+      "step": 19228
+    },
+    {
+      "epoch": 4.370227272727273,
+      "grad_norm": 1.0799473314831107,
+      "learning_rate": 4.83295375811782e-08,
+      "loss": 0.0045,
+      "step": 19229
+    },
+    {
+      "epoch": 4.370454545454545,
+      "grad_norm": 2.5354524426402443,
+      "learning_rate": 4.829513017011601e-08,
+      "loss": 0.0091,
+      "step": 19230
+    },
+    {
+      "epoch": 4.370681818181819,
+      "grad_norm": 0.9084911338661196,
+      "learning_rate": 4.826073451907228e-08,
+      "loss": 0.005,
+      "step": 19231
+    },
+    {
+      "epoch": 4.370909090909091,
+      "grad_norm": 0.166189266487644,
+      "learning_rate": 4.822635062874843e-08,
+      "loss": 0.0004,
+      "step": 19232
+    },
+    {
+      "epoch": 4.371136363636364,
+      "grad_norm": 0.3982690013688005,
+      "learning_rate": 4.819197849984539e-08,
+      "loss": 0.0008,
+      "step": 19233
+    },
+    {
+      "epoch": 4.371363636363636,
+      "grad_norm": 0.5412511433475623,
+      "learning_rate": 4.815761813306424e-08,
+      "loss": 0.0014,
+      "step": 19234
+    },
+    {
+      "epoch": 4.371590909090909,
+      "grad_norm": 1.103249445254891,
+      "learning_rate": 4.812326952910555e-08,
+      "loss": 0.0052,
+      "step": 19235
+    },
+    {
+      "epoch": 4.371818181818182,
+      "grad_norm": 0.2224499387801343,
+      "learning_rate": 4.808893268867e-08,
+      "loss": 0.0003,
+      "step": 19236
+    },
+    {
+      "epoch": 4.372045454545455,
+      "grad_norm": 0.5095212709840045,
+      "learning_rate": 4.805460761245745e-08,
+      "loss": 0.001,
+      "step": 19237
+    },
+    {
+      "epoch": 4.372272727272727,
+      "grad_norm": 0.29108690178071256,
+      "learning_rate": 4.802029430116796e-08,
+      "loss": 0.0008,
+      "step": 19238
+    },
+    {
+      "epoch": 4.3725,
+      "grad_norm": 1.0498712277622742,
+      "learning_rate": 4.7985992755501324e-08,
+      "loss": 0.0036,
+      "step": 19239
+    },
+    {
+      "epoch": 4.372727272727273,
+      "grad_norm": 0.8664096842761777,
+      "learning_rate": 4.795170297615691e-08,
+      "loss": 0.0101,
+      "step": 19240
+    },
+    {
+      "epoch": 4.3729545454545455,
+      "grad_norm": 1.22881858172948,
+      "learning_rate": 4.791742496383409e-08,
+      "loss": 0.0091,
+      "step": 19241
+    },
+    {
+      "epoch": 4.373181818181818,
+      "grad_norm": 3.512080102216715,
+      "learning_rate": 4.788315871923167e-08,
+      "loss": 0.0232,
+      "step": 19242
+    },
+    {
+      "epoch": 4.373409090909091,
+      "grad_norm": 1.4211803978077462,
+      "learning_rate": 4.7848904243048484e-08,
+      "loss": 0.0133,
+      "step": 19243
+    },
+    {
+      "epoch": 4.373636363636364,
+      "grad_norm": 0.23520355310909108,
+      "learning_rate": 4.781466153598299e-08,
+      "loss": 0.0009,
+      "step": 19244
+    },
+    {
+      "epoch": 4.3738636363636365,
+      "grad_norm": 2.197730180829613,
+      "learning_rate": 4.7780430598733524e-08,
+      "loss": 0.0082,
+      "step": 19245
+    },
+    {
+      "epoch": 4.374090909090909,
+      "grad_norm": 0.8250177380382411,
+      "learning_rate": 4.774621143199813e-08,
+      "loss": 0.0057,
+      "step": 19246
+    },
+    {
+      "epoch": 4.3743181818181816,
+      "grad_norm": 2.010924026962996,
+      "learning_rate": 4.771200403647452e-08,
+      "loss": 0.0087,
+      "step": 19247
+    },
+    {
+      "epoch": 4.374545454545455,
+      "grad_norm": 1.8659992773995866,
+      "learning_rate": 4.7677808412860267e-08,
+      "loss": 0.0103,
+      "step": 19248
+    },
+    {
+      "epoch": 4.3747727272727275,
+      "grad_norm": 2.764387177753432,
+      "learning_rate": 4.7643624561852855e-08,
+      "loss": 0.0079,
+      "step": 19249
+    },
+    {
+      "epoch": 4.375,
+      "grad_norm": 2.682556620363947,
+      "learning_rate": 4.760945248414903e-08,
+      "loss": 0.0232,
+      "step": 19250
+    },
+    {
+      "epoch": 4.3752272727272725,
+      "grad_norm": 1.1501038725392532,
+      "learning_rate": 4.757529218044579e-08,
+      "loss": 0.0042,
+      "step": 19251
+    },
+    {
+      "epoch": 4.375454545454545,
+      "grad_norm": 0.496534756713114,
+      "learning_rate": 4.754114365143975e-08,
+      "loss": 0.0011,
+      "step": 19252
+    },
+    {
+      "epoch": 4.3756818181818184,
+      "grad_norm": 1.6721347598480083,
+      "learning_rate": 4.750700689782721e-08,
+      "loss": 0.0076,
+      "step": 19253
+    },
+    {
+      "epoch": 4.375909090909091,
+      "grad_norm": 0.24731753101281542,
+      "learning_rate": 4.747288192030436e-08,
+      "loss": 0.0012,
+      "step": 19254
+    },
+    {
+      "epoch": 4.3761363636363635,
+      "grad_norm": 2.0559973076930915,
+      "learning_rate": 4.743876871956696e-08,
+      "loss": 0.0232,
+      "step": 19255
+    },
+    {
+      "epoch": 4.376363636363636,
+      "grad_norm": 2.473928601867361,
+      "learning_rate": 4.740466729631063e-08,
+      "loss": 0.0218,
+      "step": 19256
+    },
+    {
+      "epoch": 4.376590909090909,
+      "grad_norm": 0.7748620908736213,
+      "learning_rate": 4.7370577651230795e-08,
+      "loss": 0.0016,
+      "step": 19257
+    },
+    {
+      "epoch": 4.376818181818182,
+      "grad_norm": 0.45449178778040134,
+      "learning_rate": 4.733649978502265e-08,
+      "loss": 0.0015,
+      "step": 19258
+    },
+    {
+      "epoch": 4.3770454545454545,
+      "grad_norm": 0.8284588984924871,
+      "learning_rate": 4.730243369838107e-08,
+      "loss": 0.007,
+      "step": 19259
+    },
+    {
+      "epoch": 4.377272727272727,
+      "grad_norm": 0.65211882618744,
+      "learning_rate": 4.72683793920007e-08,
+      "loss": 0.0025,
+      "step": 19260
+    },
+    {
+      "epoch": 4.3775,
+      "grad_norm": 0.9630795031647174,
+      "learning_rate": 4.723433686657598e-08,
+      "loss": 0.0067,
+      "step": 19261
+    },
+    {
+      "epoch": 4.377727272727273,
+      "grad_norm": 1.548357747332502,
+      "learning_rate": 4.720030612280116e-08,
+      "loss": 0.01,
+      "step": 19262
+    },
+    {
+      "epoch": 4.377954545454545,
+      "grad_norm": 2.4843110334733924,
+      "learning_rate": 4.716628716137005e-08,
+      "loss": 0.0013,
+      "step": 19263
+    },
+    {
+      "epoch": 4.378181818181818,
+      "grad_norm": 0.28018600123956094,
+      "learning_rate": 4.713227998297647e-08,
+      "loss": 0.0009,
+      "step": 19264
+    },
+    {
+      "epoch": 4.378409090909091,
+      "grad_norm": 0.9408710984393949,
+      "learning_rate": 4.709828458831382e-08,
+      "loss": 0.0029,
+      "step": 19265
+    },
+    {
+      "epoch": 4.378636363636364,
+      "grad_norm": 5.741772804542702,
+      "learning_rate": 4.706430097807531e-08,
+      "loss": 0.0289,
+      "step": 19266
+    },
+    {
+      "epoch": 4.378863636363636,
+      "grad_norm": 1.5571487584491686,
+      "learning_rate": 4.703032915295412e-08,
+      "loss": 0.0052,
+      "step": 19267
+    },
+    {
+      "epoch": 4.379090909090909,
+      "grad_norm": 0.7309880674179514,
+      "learning_rate": 4.699636911364276e-08,
+      "loss": 0.0017,
+      "step": 19268
+    },
+    {
+      "epoch": 4.379318181818181,
+      "grad_norm": 0.9867964575392205,
+      "learning_rate": 4.696242086083373e-08,
+      "loss": 0.0087,
+      "step": 19269
+    },
+    {
+      "epoch": 4.379545454545455,
+      "grad_norm": 0.7507266687320865,
+      "learning_rate": 4.692848439521946e-08,
+      "loss": 0.0146,
+      "step": 19270
+    },
+    {
+      "epoch": 4.379772727272727,
+      "grad_norm": 0.7278215162532655,
+      "learning_rate": 4.6894559717491976e-08,
+      "loss": 0.0014,
+      "step": 19271
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 0.8554662680810402,
+      "learning_rate": 4.686064682834286e-08,
+      "loss": 0.002,
+      "step": 19272
+    },
+    {
+      "epoch": 4.380227272727272,
+      "grad_norm": 4.497352804127197,
+      "learning_rate": 4.6826745728463794e-08,
+      "loss": 0.0161,
+      "step": 19273
+    },
+    {
+      "epoch": 4.380454545454546,
+      "grad_norm": 1.4702749895476486,
+      "learning_rate": 4.6792856418546026e-08,
+      "loss": 0.0035,
+      "step": 19274
+    },
+    {
+      "epoch": 4.380681818181818,
+      "grad_norm": 1.0288662493748626,
+      "learning_rate": 4.675897889928075e-08,
+      "loss": 0.0028,
+      "step": 19275
+    },
+    {
+      "epoch": 4.380909090909091,
+      "grad_norm": 0.6777210252435126,
+      "learning_rate": 4.6725113171358566e-08,
+      "loss": 0.0037,
+      "step": 19276
+    },
+    {
+      "epoch": 4.381136363636363,
+      "grad_norm": 1.4698012577718416,
+      "learning_rate": 4.669125923547027e-08,
+      "loss": 0.0055,
+      "step": 19277
+    },
+    {
+      "epoch": 4.381363636363636,
+      "grad_norm": 0.45372583485825546,
+      "learning_rate": 4.665741709230606e-08,
+      "loss": 0.004,
+      "step": 19278
+    },
+    {
+      "epoch": 4.381590909090909,
+      "grad_norm": 1.1647375098114414,
+      "learning_rate": 4.662358674255609e-08,
+      "loss": 0.004,
+      "step": 19279
+    },
+    {
+      "epoch": 4.381818181818182,
+      "grad_norm": 0.9529039272426597,
+      "learning_rate": 4.658976818691034e-08,
+      "loss": 0.0067,
+      "step": 19280
+    },
+    {
+      "epoch": 4.382045454545454,
+      "grad_norm": 0.4540677844746431,
+      "learning_rate": 4.6555961426058215e-08,
+      "loss": 0.0016,
+      "step": 19281
+    },
+    {
+      "epoch": 4.382272727272727,
+      "grad_norm": 1.033017590437786,
+      "learning_rate": 4.6522166460689214e-08,
+      "loss": 0.0093,
+      "step": 19282
+    },
+    {
+      "epoch": 4.3825,
+      "grad_norm": 2.287449059530188,
+      "learning_rate": 4.6488383291492385e-08,
+      "loss": 0.0209,
+      "step": 19283
+    },
+    {
+      "epoch": 4.382727272727273,
+      "grad_norm": 1.4882556156860214,
+      "learning_rate": 4.6454611919156817e-08,
+      "loss": 0.007,
+      "step": 19284
+    },
+    {
+      "epoch": 4.382954545454545,
+      "grad_norm": 2.6284586441422437,
+      "learning_rate": 4.6420852344370994e-08,
+      "loss": 0.015,
+      "step": 19285
+    },
+    {
+      "epoch": 4.383181818181818,
+      "grad_norm": 0.5411839847055943,
+      "learning_rate": 4.638710456782332e-08,
+      "loss": 0.0014,
+      "step": 19286
+    },
+    {
+      "epoch": 4.383409090909091,
+      "grad_norm": 0.5809155738262926,
+      "learning_rate": 4.635336859020199e-08,
+      "loss": 0.002,
+      "step": 19287
+    },
+    {
+      "epoch": 4.383636363636364,
+      "grad_norm": 2.2120450420607014,
+      "learning_rate": 4.631964441219501e-08,
+      "loss": 0.0091,
+      "step": 19288
+    },
+    {
+      "epoch": 4.383863636363636,
+      "grad_norm": 0.690747004734713,
+      "learning_rate": 4.628593203449008e-08,
+      "loss": 0.0065,
+      "step": 19289
+    },
+    {
+      "epoch": 4.384090909090909,
+      "grad_norm": 0.7537234358271537,
+      "learning_rate": 4.6252231457774576e-08,
+      "loss": 0.0043,
+      "step": 19290
+    },
+    {
+      "epoch": 4.384318181818182,
+      "grad_norm": 0.5889650458971784,
+      "learning_rate": 4.621854268273579e-08,
+      "loss": 0.0015,
+      "step": 19291
+    },
+    {
+      "epoch": 4.384545454545455,
+      "grad_norm": 0.6735408960844377,
+      "learning_rate": 4.618486571006067e-08,
+      "loss": 0.003,
+      "step": 19292
+    },
+    {
+      "epoch": 4.384772727272727,
+      "grad_norm": 0.5330843809826307,
+      "learning_rate": 4.615120054043597e-08,
+      "loss": 0.0019,
+      "step": 19293
+    },
+    {
+      "epoch": 4.385,
+      "grad_norm": 1.478021085860316,
+      "learning_rate": 4.611754717454814e-08,
+      "loss": 0.0115,
+      "step": 19294
+    },
+    {
+      "epoch": 4.385227272727272,
+      "grad_norm": 0.47852781402625844,
+      "learning_rate": 4.608390561308337e-08,
+      "loss": 0.0016,
+      "step": 19295
+    },
+    {
+      "epoch": 4.385454545454546,
+      "grad_norm": 0.3056503284992515,
+      "learning_rate": 4.605027585672772e-08,
+      "loss": 0.0014,
+      "step": 19296
+    },
+    {
+      "epoch": 4.385681818181818,
+      "grad_norm": 2.0896051008688192,
+      "learning_rate": 4.6016657906167025e-08,
+      "loss": 0.0129,
+      "step": 19297
+    },
+    {
+      "epoch": 4.385909090909091,
+      "grad_norm": 0.9027882269698935,
+      "learning_rate": 4.598305176208684e-08,
+      "loss": 0.0028,
+      "step": 19298
+    },
+    {
+      "epoch": 4.386136363636363,
+      "grad_norm": 1.5739582501180585,
+      "learning_rate": 4.594945742517232e-08,
+      "loss": 0.0039,
+      "step": 19299
+    },
+    {
+      "epoch": 4.386363636363637,
+      "grad_norm": 0.7595030266145955,
+      "learning_rate": 4.591587489610853e-08,
+      "loss": 0.0058,
+      "step": 19300
+    },
+    {
+      "epoch": 4.386590909090909,
+      "grad_norm": 0.21212389416664088,
+      "learning_rate": 4.588230417558036e-08,
+      "loss": 0.0006,
+      "step": 19301
+    },
+    {
+      "epoch": 4.386818181818182,
+      "grad_norm": 1.0330714433437251,
+      "learning_rate": 4.5848745264272384e-08,
+      "loss": 0.0084,
+      "step": 19302
+    },
+    {
+      "epoch": 4.387045454545454,
+      "grad_norm": 0.6166134576212348,
+      "learning_rate": 4.5815198162868846e-08,
+      "loss": 0.002,
+      "step": 19303
+    },
+    {
+      "epoch": 4.387272727272728,
+      "grad_norm": 0.626177543199823,
+      "learning_rate": 4.5781662872053795e-08,
+      "loss": 0.0049,
+      "step": 19304
+    },
+    {
+      "epoch": 4.3875,
+      "grad_norm": 0.6731881075573921,
+      "learning_rate": 4.5748139392511185e-08,
+      "loss": 0.0014,
+      "step": 19305
+    },
+    {
+      "epoch": 4.387727272727273,
+      "grad_norm": 1.2584894361411176,
+      "learning_rate": 4.5714627724924505e-08,
+      "loss": 0.0099,
+      "step": 19306
+    },
+    {
+      "epoch": 4.387954545454545,
+      "grad_norm": 1.5818180804502406,
+      "learning_rate": 4.568112786997724e-08,
+      "loss": 0.0059,
+      "step": 19307
+    },
+    {
+      "epoch": 4.388181818181819,
+      "grad_norm": 0.35813951462669796,
+      "learning_rate": 4.5647639828352437e-08,
+      "loss": 0.0017,
+      "step": 19308
+    },
+    {
+      "epoch": 4.388409090909091,
+      "grad_norm": 0.873377149164763,
+      "learning_rate": 4.5614163600733034e-08,
+      "loss": 0.0038,
+      "step": 19309
+    },
+    {
+      "epoch": 4.388636363636364,
+      "grad_norm": 1.5842694202772736,
+      "learning_rate": 4.5580699187801614e-08,
+      "loss": 0.0096,
+      "step": 19310
+    },
+    {
+      "epoch": 4.388863636363636,
+      "grad_norm": 3.10441492881688,
+      "learning_rate": 4.554724659024068e-08,
+      "loss": 0.0139,
+      "step": 19311
+    },
+    {
+      "epoch": 4.389090909090909,
+      "grad_norm": 1.7234326706717797,
+      "learning_rate": 4.551380580873219e-08,
+      "loss": 0.0096,
+      "step": 19312
+    },
+    {
+      "epoch": 4.389318181818182,
+      "grad_norm": 0.37877975961468147,
+      "learning_rate": 4.548037684395816e-08,
+      "loss": 0.0008,
+      "step": 19313
+    },
+    {
+      "epoch": 4.389545454545455,
+      "grad_norm": 1.6342990436351192,
+      "learning_rate": 4.544695969660027e-08,
+      "loss": 0.0094,
+      "step": 19314
+    },
+    {
+      "epoch": 4.389772727272727,
+      "grad_norm": 0.6254508147782661,
+      "learning_rate": 4.541355436734007e-08,
+      "loss": 0.0036,
+      "step": 19315
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 0.9450289438611179,
+      "learning_rate": 4.5380160856858534e-08,
+      "loss": 0.005,
+      "step": 19316
+    },
+    {
+      "epoch": 4.390227272727273,
+      "grad_norm": 1.1125638261272595,
+      "learning_rate": 4.534677916583671e-08,
+      "loss": 0.0038,
+      "step": 19317
+    },
+    {
+      "epoch": 4.390454545454546,
+      "grad_norm": 1.1227033721612993,
+      "learning_rate": 4.531340929495538e-08,
+      "loss": 0.0037,
+      "step": 19318
+    },
+    {
+      "epoch": 4.390681818181818,
+      "grad_norm": 0.3786200316748268,
+      "learning_rate": 4.52800512448949e-08,
+      "loss": 0.0014,
+      "step": 19319
+    },
+    {
+      "epoch": 4.390909090909091,
+      "grad_norm": 2.1578205084622026,
+      "learning_rate": 4.5246705016335557e-08,
+      "loss": 0.0117,
+      "step": 19320
+    },
+    {
+      "epoch": 4.391136363636363,
+      "grad_norm": 0.328390219164692,
+      "learning_rate": 4.521337060995737e-08,
+      "loss": 0.0016,
+      "step": 19321
+    },
+    {
+      "epoch": 4.391363636363637,
+      "grad_norm": 0.5403867571433054,
+      "learning_rate": 4.5180048026439995e-08,
+      "loss": 0.0012,
+      "step": 19322
+    },
+    {
+      "epoch": 4.391590909090909,
+      "grad_norm": 0.3871164522908397,
+      "learning_rate": 4.5146737266463025e-08,
+      "loss": 0.0015,
+      "step": 19323
+    },
+    {
+      "epoch": 4.391818181818182,
+      "grad_norm": 1.1122727166397446,
+      "learning_rate": 4.511343833070579e-08,
+      "loss": 0.0101,
+      "step": 19324
+    },
+    {
+      "epoch": 4.392045454545454,
+      "grad_norm": 0.9676672991373725,
+      "learning_rate": 4.508015121984711e-08,
+      "loss": 0.0064,
+      "step": 19325
+    },
+    {
+      "epoch": 4.3922727272727276,
+      "grad_norm": 0.34003740650032976,
+      "learning_rate": 4.504687593456583e-08,
+      "loss": 0.0015,
+      "step": 19326
+    },
+    {
+      "epoch": 4.3925,
+      "grad_norm": 0.871607705375347,
+      "learning_rate": 4.5013612475540566e-08,
+      "loss": 0.0009,
+      "step": 19327
+    },
+    {
+      "epoch": 4.392727272727273,
+      "grad_norm": 0.7014283888564188,
+      "learning_rate": 4.4980360843449725e-08,
+      "loss": 0.0017,
+      "step": 19328
+    },
+    {
+      "epoch": 4.392954545454545,
+      "grad_norm": 0.7383867523761036,
+      "learning_rate": 4.4947121038971113e-08,
+      "loss": 0.0017,
+      "step": 19329
+    },
+    {
+      "epoch": 4.3931818181818185,
+      "grad_norm": 0.394548819400589,
+      "learning_rate": 4.4913893062782654e-08,
+      "loss": 0.0012,
+      "step": 19330
+    },
+    {
+      "epoch": 4.393409090909091,
+      "grad_norm": 0.6052979356254392,
+      "learning_rate": 4.488067691556186e-08,
+      "loss": 0.0019,
+      "step": 19331
+    },
+    {
+      "epoch": 4.393636363636364,
+      "grad_norm": 1.2644687607596874,
+      "learning_rate": 4.4847472597986245e-08,
+      "loss": 0.0086,
+      "step": 19332
+    },
+    {
+      "epoch": 4.393863636363636,
+      "grad_norm": 0.8680416162638646,
+      "learning_rate": 4.481428011073284e-08,
+      "loss": 0.0038,
+      "step": 19333
+    },
+    {
+      "epoch": 4.3940909090909095,
+      "grad_norm": 0.39538693324874147,
+      "learning_rate": 4.4781099454478325e-08,
+      "loss": 0.002,
+      "step": 19334
+    },
+    {
+      "epoch": 4.394318181818182,
+      "grad_norm": 1.6149246023069654,
+      "learning_rate": 4.4747930629899375e-08,
+      "loss": 0.0179,
+      "step": 19335
+    },
+    {
+      "epoch": 4.3945454545454545,
+      "grad_norm": 0.6759975188959143,
+      "learning_rate": 4.4714773637672546e-08,
+      "loss": 0.0029,
+      "step": 19336
+    },
+    {
+      "epoch": 4.394772727272727,
+      "grad_norm": 0.45098038883810243,
+      "learning_rate": 4.4681628478473946e-08,
+      "loss": 0.0011,
+      "step": 19337
+    },
+    {
+      "epoch": 4.395,
+      "grad_norm": 4.060368395847448,
+      "learning_rate": 4.4648495152979225e-08,
+      "loss": 0.0476,
+      "step": 19338
+    },
+    {
+      "epoch": 4.395227272727273,
+      "grad_norm": 0.3084396186654877,
+      "learning_rate": 4.461537366186416e-08,
+      "loss": 0.0007,
+      "step": 19339
+    },
+    {
+      "epoch": 4.3954545454545455,
+      "grad_norm": 0.6375188769794907,
+      "learning_rate": 4.458226400580419e-08,
+      "loss": 0.0025,
+      "step": 19340
+    },
+    {
+      "epoch": 4.395681818181818,
+      "grad_norm": 0.9880327731095392,
+      "learning_rate": 4.454916618547446e-08,
+      "loss": 0.0072,
+      "step": 19341
+    },
+    {
+      "epoch": 4.3959090909090905,
+      "grad_norm": 1.0415722212480176,
+      "learning_rate": 4.4516080201549986e-08,
+      "loss": 0.0064,
+      "step": 19342
+    },
+    {
+      "epoch": 4.396136363636364,
+      "grad_norm": 0.9883194330368363,
+      "learning_rate": 4.4483006054705236e-08,
+      "loss": 0.0081,
+      "step": 19343
+    },
+    {
+      "epoch": 4.3963636363636365,
+      "grad_norm": 2.5081539482841175,
+      "learning_rate": 4.444994374561473e-08,
+      "loss": 0.0162,
+      "step": 19344
+    },
+    {
+      "epoch": 4.396590909090909,
+      "grad_norm": 0.7549270298921565,
+      "learning_rate": 4.441689327495273e-08,
+      "loss": 0.0027,
+      "step": 19345
+    },
+    {
+      "epoch": 4.3968181818181815,
+      "grad_norm": 1.31639453216117,
+      "learning_rate": 4.4383854643393206e-08,
+      "loss": 0.0088,
+      "step": 19346
+    },
+    {
+      "epoch": 4.397045454545455,
+      "grad_norm": 2.3022355027984336,
+      "learning_rate": 4.435082785160978e-08,
+      "loss": 0.0058,
+      "step": 19347
+    },
+    {
+      "epoch": 4.397272727272727,
+      "grad_norm": 0.5097650984141417,
+      "learning_rate": 4.4317812900275956e-08,
+      "loss": 0.0018,
+      "step": 19348
+    },
+    {
+      "epoch": 4.3975,
+      "grad_norm": 1.4216524173397878,
+      "learning_rate": 4.4284809790064996e-08,
+      "loss": 0.0056,
+      "step": 19349
+    },
+    {
+      "epoch": 4.3977272727272725,
+      "grad_norm": 2.1258355908819864,
+      "learning_rate": 4.425181852164992e-08,
+      "loss": 0.0103,
+      "step": 19350
+    },
+    {
+      "epoch": 4.397954545454546,
+      "grad_norm": 0.2735832501861137,
+      "learning_rate": 4.421883909570337e-08,
+      "loss": 0.0007,
+      "step": 19351
+    },
+    {
+      "epoch": 4.398181818181818,
+      "grad_norm": 0.4342675590201497,
+      "learning_rate": 4.418587151289788e-08,
+      "loss": 0.0008,
+      "step": 19352
+    },
+    {
+      "epoch": 4.398409090909091,
+      "grad_norm": 0.7452897919525028,
+      "learning_rate": 4.415291577390582e-08,
+      "loss": 0.0044,
+      "step": 19353
+    },
+    {
+      "epoch": 4.3986363636363635,
+      "grad_norm": 1.1070798385154952,
+      "learning_rate": 4.411997187939915e-08,
+      "loss": 0.0054,
+      "step": 19354
+    },
+    {
+      "epoch": 4.398863636363636,
+      "grad_norm": 0.778314287633086,
+      "learning_rate": 4.4087039830049763e-08,
+      "loss": 0.0024,
+      "step": 19355
+    },
+    {
+      "epoch": 4.399090909090909,
+      "grad_norm": 0.7198668413836852,
+      "learning_rate": 4.4054119626529e-08,
+      "loss": 0.0023,
+      "step": 19356
+    },
+    {
+      "epoch": 4.399318181818182,
+      "grad_norm": 1.2271455102433115,
+      "learning_rate": 4.40212112695082e-08,
+      "loss": 0.0089,
+      "step": 19357
+    },
+    {
+      "epoch": 4.399545454545454,
+      "grad_norm": 1.0576471843241333,
+      "learning_rate": 4.398831475965849e-08,
+      "loss": 0.0041,
+      "step": 19358
+    },
+    {
+      "epoch": 4.399772727272727,
+      "grad_norm": 0.5126407619347035,
+      "learning_rate": 4.39554300976508e-08,
+      "loss": 0.0019,
+      "step": 19359
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 0.4147749790028958,
+      "learning_rate": 4.392255728415548e-08,
+      "loss": 0.0016,
+      "step": 19360
+    },
+    {
+      "epoch": 4.400227272727273,
+      "grad_norm": 0.284608754468562,
+      "learning_rate": 4.388969631984291e-08,
+      "loss": 0.001,
+      "step": 19361
+    },
+    {
+      "epoch": 4.400454545454545,
+      "grad_norm": 1.86782185006196,
+      "learning_rate": 4.3856847205383316e-08,
+      "loss": 0.0053,
+      "step": 19362
+    },
+    {
+      "epoch": 4.400681818181818,
+      "grad_norm": 0.4430044474795293,
+      "learning_rate": 4.3824009941446365e-08,
+      "loss": 0.002,
+      "step": 19363
+    },
+    {
+      "epoch": 4.40090909090909,
+      "grad_norm": 1.1153362562727707,
+      "learning_rate": 4.379118452870194e-08,
+      "loss": 0.0042,
+      "step": 19364
+    },
+    {
+      "epoch": 4.401136363636364,
+      "grad_norm": 0.4302331285440092,
+      "learning_rate": 4.375837096781903e-08,
+      "loss": 0.0008,
+      "step": 19365
+    },
+    {
+      "epoch": 4.401363636363636,
+      "grad_norm": 1.095628467501074,
+      "learning_rate": 4.3725569259467085e-08,
+      "loss": 0.0087,
+      "step": 19366
+    },
+    {
+      "epoch": 4.401590909090909,
+      "grad_norm": 0.6373542253558473,
+      "learning_rate": 4.369277940431481e-08,
+      "loss": 0.0018,
+      "step": 19367
+    },
+    {
+      "epoch": 4.401818181818181,
+      "grad_norm": 0.5066559897327424,
+      "learning_rate": 4.366000140303105e-08,
+      "loss": 0.0018,
+      "step": 19368
+    },
+    {
+      "epoch": 4.402045454545455,
+      "grad_norm": 1.1938260352764658,
+      "learning_rate": 4.362723525628394e-08,
+      "loss": 0.0026,
+      "step": 19369
+    },
+    {
+      "epoch": 4.402272727272727,
+      "grad_norm": 0.3331030105070399,
+      "learning_rate": 4.359448096474178e-08,
+      "loss": 0.0016,
+      "step": 19370
+    },
+    {
+      "epoch": 4.4025,
+      "grad_norm": 1.305600776680663,
+      "learning_rate": 4.35617385290725e-08,
+      "loss": 0.0099,
+      "step": 19371
+    },
+    {
+      "epoch": 4.402727272727272,
+      "grad_norm": 0.9565976549040285,
+      "learning_rate": 4.352900794994383e-08,
+      "loss": 0.0037,
+      "step": 19372
+    },
+    {
+      "epoch": 4.402954545454546,
+      "grad_norm": 0.67862239993248,
+      "learning_rate": 4.3496289228023014e-08,
+      "loss": 0.0031,
+      "step": 19373
+    },
+    {
+      "epoch": 4.403181818181818,
+      "grad_norm": 0.7411250975476031,
+      "learning_rate": 4.3463582363977294e-08,
+      "loss": 0.0019,
+      "step": 19374
+    },
+    {
+      "epoch": 4.403409090909091,
+      "grad_norm": 1.4189984184626547,
+      "learning_rate": 4.343088735847378e-08,
+      "loss": 0.0049,
+      "step": 19375
+    },
+    {
+      "epoch": 4.403636363636363,
+      "grad_norm": 0.3520232913183985,
+      "learning_rate": 4.339820421217902e-08,
+      "loss": 0.0013,
+      "step": 19376
+    },
+    {
+      "epoch": 4.403863636363637,
+      "grad_norm": 0.39489990337428477,
+      "learning_rate": 4.336553292575964e-08,
+      "loss": 0.0013,
+      "step": 19377
+    },
+    {
+      "epoch": 4.404090909090909,
+      "grad_norm": 0.7147595116766311,
+      "learning_rate": 4.333287349988162e-08,
+      "loss": 0.0012,
+      "step": 19378
+    },
+    {
+      "epoch": 4.404318181818182,
+      "grad_norm": 1.3478249640616367,
+      "learning_rate": 4.3300225935211174e-08,
+      "loss": 0.0084,
+      "step": 19379
+    },
+    {
+      "epoch": 4.404545454545454,
+      "grad_norm": 0.5580822890884216,
+      "learning_rate": 4.326759023241388e-08,
+      "loss": 0.0034,
+      "step": 19380
+    },
+    {
+      "epoch": 4.404772727272727,
+      "grad_norm": 1.567993792838367,
+      "learning_rate": 4.323496639215531e-08,
+      "loss": 0.0115,
+      "step": 19381
+    },
+    {
+      "epoch": 4.405,
+      "grad_norm": 2.1197018715733296,
+      "learning_rate": 4.320235441510078e-08,
+      "loss": 0.0215,
+      "step": 19382
+    },
+    {
+      "epoch": 4.405227272727273,
+      "grad_norm": 2.5109278081468456,
+      "learning_rate": 4.3169754301915156e-08,
+      "loss": 0.0145,
+      "step": 19383
+    },
+    {
+      "epoch": 4.405454545454545,
+      "grad_norm": 0.6265390746444296,
+      "learning_rate": 4.3137166053263335e-08,
+      "loss": 0.0018,
+      "step": 19384
+    },
+    {
+      "epoch": 4.405681818181818,
+      "grad_norm": 1.6490419497291022,
+      "learning_rate": 4.310458966980999e-08,
+      "loss": 0.0101,
+      "step": 19385
+    },
+    {
+      "epoch": 4.405909090909091,
+      "grad_norm": 1.3917025505684346,
+      "learning_rate": 4.3072025152219035e-08,
+      "loss": 0.0132,
+      "step": 19386
+    },
+    {
+      "epoch": 4.406136363636364,
+      "grad_norm": 1.7095387766022967,
+      "learning_rate": 4.303947250115473e-08,
+      "loss": 0.0116,
+      "step": 19387
+    },
+    {
+      "epoch": 4.406363636363636,
+      "grad_norm": 0.23862783975525642,
+      "learning_rate": 4.300693171728092e-08,
+      "loss": 0.0013,
+      "step": 19388
+    },
+    {
+      "epoch": 4.406590909090909,
+      "grad_norm": 1.0144035035787489,
+      "learning_rate": 4.29744028012611e-08,
+      "loss": 0.005,
+      "step": 19389
+    },
+    {
+      "epoch": 4.406818181818182,
+      "grad_norm": 0.5231144649770078,
+      "learning_rate": 4.29418857537587e-08,
+      "loss": 0.0016,
+      "step": 19390
+    },
+    {
+      "epoch": 4.407045454545455,
+      "grad_norm": 2.5191931739848297,
+      "learning_rate": 4.290938057543659e-08,
+      "loss": 0.0015,
+      "step": 19391
+    },
+    {
+      "epoch": 4.407272727272727,
+      "grad_norm": 0.4148199840487481,
+      "learning_rate": 4.2876887266957714e-08,
+      "loss": 0.0016,
+      "step": 19392
+    },
+    {
+      "epoch": 4.4075,
+      "grad_norm": 1.0921523825803736,
+      "learning_rate": 4.2844405828984663e-08,
+      "loss": 0.0065,
+      "step": 19393
+    },
+    {
+      "epoch": 4.407727272727273,
+      "grad_norm": 0.7491881375879212,
+      "learning_rate": 4.281193626217983e-08,
+      "loss": 0.0025,
+      "step": 19394
+    },
+    {
+      "epoch": 4.407954545454546,
+      "grad_norm": 1.9032183526575996,
+      "learning_rate": 4.2779478567205325e-08,
+      "loss": 0.0065,
+      "step": 19395
+    },
+    {
+      "epoch": 4.408181818181818,
+      "grad_norm": 0.2935307992437522,
+      "learning_rate": 4.274703274472297e-08,
+      "loss": 0.0005,
+      "step": 19396
+    },
+    {
+      "epoch": 4.408409090909091,
+      "grad_norm": 1.2326872459110767,
+      "learning_rate": 4.27145987953944e-08,
+      "loss": 0.0024,
+      "step": 19397
+    },
+    {
+      "epoch": 4.408636363636363,
+      "grad_norm": 0.6351027714007547,
+      "learning_rate": 4.268217671988102e-08,
+      "loss": 0.0015,
+      "step": 19398
+    },
+    {
+      "epoch": 4.408863636363637,
+      "grad_norm": 1.2424221463447447,
+      "learning_rate": 4.264976651884405e-08,
+      "loss": 0.0055,
+      "step": 19399
+    },
+    {
+      "epoch": 4.409090909090909,
+      "grad_norm": 0.6733266601038879,
+      "learning_rate": 4.261736819294427e-08,
+      "loss": 0.0016,
+      "step": 19400
+    },
+    {
+      "epoch": 4.409318181818182,
+      "grad_norm": 0.9638413421004286,
+      "learning_rate": 4.258498174284233e-08,
+      "loss": 0.003,
+      "step": 19401
+    },
+    {
+      "epoch": 4.409545454545454,
+      "grad_norm": 0.583781709954035,
+      "learning_rate": 4.255260716919868e-08,
+      "loss": 0.0021,
+      "step": 19402
+    },
+    {
+      "epoch": 4.409772727272728,
+      "grad_norm": 1.1019842144529948,
+      "learning_rate": 4.2520244472673625e-08,
+      "loss": 0.0058,
+      "step": 19403
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 0.8876259196786963,
+      "learning_rate": 4.2487893653926844e-08,
+      "loss": 0.0025,
+      "step": 19404
+    },
+    {
+      "epoch": 4.410227272727273,
+      "grad_norm": 1.3652702954525664,
+      "learning_rate": 4.245555471361816e-08,
+      "loss": 0.0044,
+      "step": 19405
+    },
+    {
+      "epoch": 4.410454545454545,
+      "grad_norm": 1.6884788056118696,
+      "learning_rate": 4.242322765240704e-08,
+      "loss": 0.0081,
+      "step": 19406
+    },
+    {
+      "epoch": 4.410681818181819,
+      "grad_norm": 0.9124316188472403,
+      "learning_rate": 4.2390912470952824e-08,
+      "loss": 0.0043,
+      "step": 19407
+    },
+    {
+      "epoch": 4.410909090909091,
+      "grad_norm": 0.47973223555647715,
+      "learning_rate": 4.235860916991415e-08,
+      "loss": 0.0026,
+      "step": 19408
+    },
+    {
+      "epoch": 4.411136363636364,
+      "grad_norm": 1.8780067502332014,
+      "learning_rate": 4.2326317749949925e-08,
+      "loss": 0.0107,
+      "step": 19409
+    },
+    {
+      "epoch": 4.411363636363636,
+      "grad_norm": 0.8381508523416841,
+      "learning_rate": 4.22940382117186e-08,
+      "loss": 0.0022,
+      "step": 19410
+    },
+    {
+      "epoch": 4.41159090909091,
+      "grad_norm": 1.2083206453736228,
+      "learning_rate": 4.2261770555878445e-08,
+      "loss": 0.0035,
+      "step": 19411
+    },
+    {
+      "epoch": 4.411818181818182,
+      "grad_norm": 1.0267614660401199,
+      "learning_rate": 4.2229514783087425e-08,
+      "loss": 0.0039,
+      "step": 19412
+    },
+    {
+      "epoch": 4.412045454545455,
+      "grad_norm": 1.3043250183206556,
+      "learning_rate": 4.219727089400327e-08,
+      "loss": 0.0043,
+      "step": 19413
+    },
+    {
+      "epoch": 4.412272727272727,
+      "grad_norm": 0.36235046011732436,
+      "learning_rate": 4.2165038889283586e-08,
+      "loss": 0.0024,
+      "step": 19414
+    },
+    {
+      "epoch": 4.4125,
+      "grad_norm": 1.8875790242234363,
+      "learning_rate": 4.213281876958547e-08,
+      "loss": 0.0127,
+      "step": 19415
+    },
+    {
+      "epoch": 4.412727272727273,
+      "grad_norm": 0.5635733995597362,
+      "learning_rate": 4.2100610535566256e-08,
+      "loss": 0.0027,
+      "step": 19416
+    },
+    {
+      "epoch": 4.412954545454546,
+      "grad_norm": 0.7740983421150243,
+      "learning_rate": 4.2068414187882366e-08,
+      "loss": 0.0026,
+      "step": 19417
+    },
+    {
+      "epoch": 4.413181818181818,
+      "grad_norm": 0.5855480646402553,
+      "learning_rate": 4.203622972719049e-08,
+      "loss": 0.004,
+      "step": 19418
+    },
+    {
+      "epoch": 4.413409090909091,
+      "grad_norm": 0.5645593533438615,
+      "learning_rate": 4.2004057154146906e-08,
+      "loss": 0.0012,
+      "step": 19419
+    },
+    {
+      "epoch": 4.413636363636364,
+      "grad_norm": 0.727065244470546,
+      "learning_rate": 4.197189646940769e-08,
+      "loss": 0.0023,
+      "step": 19420
+    },
+    {
+      "epoch": 4.4138636363636365,
+      "grad_norm": 0.25037861714534027,
+      "learning_rate": 4.193974767362878e-08,
+      "loss": 0.0002,
+      "step": 19421
+    },
+    {
+      "epoch": 4.414090909090909,
+      "grad_norm": 0.5865394477223344,
+      "learning_rate": 4.190761076746548e-08,
+      "loss": 0.0033,
+      "step": 19422
+    },
+    {
+      "epoch": 4.414318181818182,
+      "grad_norm": 3.20053997932044,
+      "learning_rate": 4.187548575157331e-08,
+      "loss": 0.0165,
+      "step": 19423
+    },
+    {
+      "epoch": 4.414545454545454,
+      "grad_norm": 0.35747780919021854,
+      "learning_rate": 4.18433726266073e-08,
+      "loss": 0.0015,
+      "step": 19424
+    },
+    {
+      "epoch": 4.4147727272727275,
+      "grad_norm": 0.9003658189147266,
+      "learning_rate": 4.1811271393222274e-08,
+      "loss": 0.0028,
+      "step": 19425
+    },
+    {
+      "epoch": 4.415,
+      "grad_norm": 0.7773447702717295,
+      "learning_rate": 4.1779182052072844e-08,
+      "loss": 0.0037,
+      "step": 19426
+    },
+    {
+      "epoch": 4.415227272727273,
+      "grad_norm": 0.9313570815748241,
+      "learning_rate": 4.1747104603813356e-08,
+      "loss": 0.0024,
+      "step": 19427
+    },
+    {
+      "epoch": 4.415454545454545,
+      "grad_norm": 2.225162798530664,
+      "learning_rate": 4.1715039049098e-08,
+      "loss": 0.0027,
+      "step": 19428
+    },
+    {
+      "epoch": 4.4156818181818185,
+      "grad_norm": 2.5680786184979656,
+      "learning_rate": 4.1682985388580635e-08,
+      "loss": 0.0123,
+      "step": 19429
+    },
+    {
+      "epoch": 4.415909090909091,
+      "grad_norm": 1.6480582836280127,
+      "learning_rate": 4.165094362291483e-08,
+      "loss": 0.0052,
+      "step": 19430
+    },
+    {
+      "epoch": 4.4161363636363635,
+      "grad_norm": 1.9175518368462643,
+      "learning_rate": 4.1618913752753885e-08,
+      "loss": 0.0185,
+      "step": 19431
+    },
+    {
+      "epoch": 4.416363636363636,
+      "grad_norm": 1.2598011957072177,
+      "learning_rate": 4.158689577875116e-08,
+      "loss": 0.0052,
+      "step": 19432
+    },
+    {
+      "epoch": 4.4165909090909095,
+      "grad_norm": 0.41603834431993686,
+      "learning_rate": 4.155488970155941e-08,
+      "loss": 0.0017,
+      "step": 19433
+    },
+    {
+      "epoch": 4.416818181818182,
+      "grad_norm": 0.45980886544052535,
+      "learning_rate": 4.152289552183143e-08,
+      "loss": 0.0015,
+      "step": 19434
+    },
+    {
+      "epoch": 4.4170454545454545,
+      "grad_norm": 0.8882007715981922,
+      "learning_rate": 4.1490913240219423e-08,
+      "loss": 0.006,
+      "step": 19435
+    },
+    {
+      "epoch": 4.417272727272727,
+      "grad_norm": 0.3089557515708537,
+      "learning_rate": 4.14589428573757e-08,
+      "loss": 0.0014,
+      "step": 19436
+    },
+    {
+      "epoch": 4.4175,
+      "grad_norm": 0.3923667151121952,
+      "learning_rate": 4.142698437395218e-08,
+      "loss": 0.0009,
+      "step": 19437
+    },
+    {
+      "epoch": 4.417727272727273,
+      "grad_norm": 2.372932184442372,
+      "learning_rate": 4.1395037790600624e-08,
+      "loss": 0.0093,
+      "step": 19438
+    },
+    {
+      "epoch": 4.4179545454545455,
+      "grad_norm": 2.018587442707227,
+      "learning_rate": 4.136310310797233e-08,
+      "loss": 0.009,
+      "step": 19439
+    },
+    {
+      "epoch": 4.418181818181818,
+      "grad_norm": 0.9884619307610419,
+      "learning_rate": 4.1331180326718567e-08,
+      "loss": 0.004,
+      "step": 19440
+    },
+    {
+      "epoch": 4.4184090909090905,
+      "grad_norm": 0.256589923812497,
+      "learning_rate": 4.1299269447490287e-08,
+      "loss": 0.0009,
+      "step": 19441
+    },
+    {
+      "epoch": 4.418636363636364,
+      "grad_norm": 1.5471462222587018,
+      "learning_rate": 4.1267370470938275e-08,
+      "loss": 0.0078,
+      "step": 19442
+    },
+    {
+      "epoch": 4.418863636363636,
+      "grad_norm": 1.115982505651217,
+      "learning_rate": 4.123548339771292e-08,
+      "loss": 0.004,
+      "step": 19443
+    },
+    {
+      "epoch": 4.419090909090909,
+      "grad_norm": 1.210457460431509,
+      "learning_rate": 4.1203608228464476e-08,
+      "loss": 0.001,
+      "step": 19444
+    },
+    {
+      "epoch": 4.4193181818181815,
+      "grad_norm": 1.2722520823519978,
+      "learning_rate": 4.1171744963842964e-08,
+      "loss": 0.0048,
+      "step": 19445
+    },
+    {
+      "epoch": 4.419545454545455,
+      "grad_norm": 1.6586443384098533,
+      "learning_rate": 4.1139893604498144e-08,
+      "loss": 0.0065,
+      "step": 19446
+    },
+    {
+      "epoch": 4.419772727272727,
+      "grad_norm": 0.9533235227746673,
+      "learning_rate": 4.110805415107951e-08,
+      "loss": 0.0046,
+      "step": 19447
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 1.1405395574291666,
+      "learning_rate": 4.107622660423631e-08,
+      "loss": 0.0089,
+      "step": 19448
+    },
+    {
+      "epoch": 4.4202272727272724,
+      "grad_norm": 0.5456672213516084,
+      "learning_rate": 4.104441096461749e-08,
+      "loss": 0.0044,
+      "step": 19449
+    },
+    {
+      "epoch": 4.420454545454546,
+      "grad_norm": 1.0839864790934903,
+      "learning_rate": 4.10126072328719e-08,
+      "loss": 0.0066,
+      "step": 19450
+    },
+    {
+      "epoch": 4.420681818181818,
+      "grad_norm": 1.507850962128578,
+      "learning_rate": 4.098081540964818e-08,
+      "loss": 0.0075,
+      "step": 19451
+    },
+    {
+      "epoch": 4.420909090909091,
+      "grad_norm": 1.851977460801877,
+      "learning_rate": 4.094903549559444e-08,
+      "loss": 0.0142,
+      "step": 19452
+    },
+    {
+      "epoch": 4.421136363636363,
+      "grad_norm": 0.9408121464359107,
+      "learning_rate": 4.091726749135883e-08,
+      "loss": 0.0013,
+      "step": 19453
+    },
+    {
+      "epoch": 4.421363636363637,
+      "grad_norm": 0.9414353192278735,
+      "learning_rate": 4.088551139758902e-08,
+      "loss": 0.0036,
+      "step": 19454
+    },
+    {
+      "epoch": 4.421590909090909,
+      "grad_norm": 1.4553407031422638,
+      "learning_rate": 4.085376721493277e-08,
+      "loss": 0.0229,
+      "step": 19455
+    },
+    {
+      "epoch": 4.421818181818182,
+      "grad_norm": 1.6815749890448153,
+      "learning_rate": 4.082203494403726e-08,
+      "loss": 0.0047,
+      "step": 19456
+    },
+    {
+      "epoch": 4.422045454545454,
+      "grad_norm": 0.4015665136964545,
+      "learning_rate": 4.079031458554962e-08,
+      "loss": 0.0011,
+      "step": 19457
+    },
+    {
+      "epoch": 4.422272727272727,
+      "grad_norm": 0.5815236555318927,
+      "learning_rate": 4.075860614011669e-08,
+      "loss": 0.0013,
+      "step": 19458
+    },
+    {
+      "epoch": 4.4225,
+      "grad_norm": 0.6846001872465568,
+      "learning_rate": 4.072690960838503e-08,
+      "loss": 0.0017,
+      "step": 19459
+    },
+    {
+      "epoch": 4.422727272727273,
+      "grad_norm": 0.6845695011679188,
+      "learning_rate": 4.069522499100108e-08,
+      "loss": 0.0025,
+      "step": 19460
+    },
+    {
+      "epoch": 4.422954545454545,
+      "grad_norm": 0.8650213292822971,
+      "learning_rate": 4.066355228861077e-08,
+      "loss": 0.0054,
+      "step": 19461
+    },
+    {
+      "epoch": 4.423181818181818,
+      "grad_norm": 0.12866366552945616,
+      "learning_rate": 4.063189150186012e-08,
+      "loss": 0.0006,
+      "step": 19462
+    },
+    {
+      "epoch": 4.423409090909091,
+      "grad_norm": 0.5961617547011525,
+      "learning_rate": 4.060024263139464e-08,
+      "loss": 0.0016,
+      "step": 19463
+    },
+    {
+      "epoch": 4.423636363636364,
+      "grad_norm": 1.3412922912480043,
+      "learning_rate": 4.0568605677859736e-08,
+      "loss": 0.0055,
+      "step": 19464
+    },
+    {
+      "epoch": 4.423863636363636,
+      "grad_norm": 1.3787231643017808,
+      "learning_rate": 4.053698064190072e-08,
+      "loss": 0.003,
+      "step": 19465
+    },
+    {
+      "epoch": 4.424090909090909,
+      "grad_norm": 3.5228197560126766,
+      "learning_rate": 4.0505367524162205e-08,
+      "loss": 0.0147,
+      "step": 19466
+    },
+    {
+      "epoch": 4.424318181818181,
+      "grad_norm": 0.7457419623907624,
+      "learning_rate": 4.047376632528889e-08,
+      "loss": 0.002,
+      "step": 19467
+    },
+    {
+      "epoch": 4.424545454545455,
+      "grad_norm": 1.0590360067329925,
+      "learning_rate": 4.044217704592533e-08,
+      "loss": 0.0036,
+      "step": 19468
+    },
+    {
+      "epoch": 4.424772727272727,
+      "grad_norm": 1.0837775835955525,
+      "learning_rate": 4.041059968671567e-08,
+      "loss": 0.0027,
+      "step": 19469
+    },
+    {
+      "epoch": 4.425,
+      "grad_norm": 2.031577352667111,
+      "learning_rate": 4.037903424830361e-08,
+      "loss": 0.0052,
+      "step": 19470
+    },
+    {
+      "epoch": 4.425227272727272,
+      "grad_norm": 1.4231317675105568,
+      "learning_rate": 4.034748073133296e-08,
+      "loss": 0.0056,
+      "step": 19471
+    },
+    {
+      "epoch": 4.425454545454546,
+      "grad_norm": 0.18257168985589498,
+      "learning_rate": 4.0315939136447226e-08,
+      "loss": 0.0005,
+      "step": 19472
+    },
+    {
+      "epoch": 4.425681818181818,
+      "grad_norm": 0.28731501913365703,
+      "learning_rate": 4.02844094642895e-08,
+      "loss": 0.0008,
+      "step": 19473
+    },
+    {
+      "epoch": 4.425909090909091,
+      "grad_norm": 1.0882537809121067,
+      "learning_rate": 4.025289171550276e-08,
+      "loss": 0.0053,
+      "step": 19474
+    },
+    {
+      "epoch": 4.426136363636363,
+      "grad_norm": 0.45343052791918714,
+      "learning_rate": 4.022138589072973e-08,
+      "loss": 0.0019,
+      "step": 19475
+    },
+    {
+      "epoch": 4.426363636363637,
+      "grad_norm": 1.6212933693420897,
+      "learning_rate": 4.018989199061282e-08,
+      "loss": 0.0125,
+      "step": 19476
+    },
+    {
+      "epoch": 4.426590909090909,
+      "grad_norm": 0.6022477103563728,
+      "learning_rate": 4.015841001579423e-08,
+      "loss": 0.0019,
+      "step": 19477
+    },
+    {
+      "epoch": 4.426818181818182,
+      "grad_norm": 2.3907320835409407,
+      "learning_rate": 4.012693996691609e-08,
+      "loss": 0.0181,
+      "step": 19478
+    },
+    {
+      "epoch": 4.427045454545454,
+      "grad_norm": 1.0265745188098498,
+      "learning_rate": 4.009548184461996e-08,
+      "loss": 0.0031,
+      "step": 19479
+    },
+    {
+      "epoch": 4.427272727272728,
+      "grad_norm": 0.7122680705865765,
+      "learning_rate": 4.006403564954735e-08,
+      "loss": 0.0047,
+      "step": 19480
+    },
+    {
+      "epoch": 4.4275,
+      "grad_norm": 2.218649178204538,
+      "learning_rate": 4.003260138233954e-08,
+      "loss": 0.0123,
+      "step": 19481
+    },
+    {
+      "epoch": 4.427727272727273,
+      "grad_norm": 1.9226681150664544,
+      "learning_rate": 4.0001179043637625e-08,
+      "loss": 0.0073,
+      "step": 19482
+    },
+    {
+      "epoch": 4.427954545454545,
+      "grad_norm": 3.885552235516198,
+      "learning_rate": 3.9969768634082126e-08,
+      "loss": 0.0064,
+      "step": 19483
+    },
+    {
+      "epoch": 4.428181818181818,
+      "grad_norm": 1.7556114538922285,
+      "learning_rate": 3.993837015431365e-08,
+      "loss": 0.0083,
+      "step": 19484
+    },
+    {
+      "epoch": 4.428409090909091,
+      "grad_norm": 0.2914087786088134,
+      "learning_rate": 3.990698360497258e-08,
+      "loss": 0.0016,
+      "step": 19485
+    },
+    {
+      "epoch": 4.428636363636364,
+      "grad_norm": 0.9781727518615416,
+      "learning_rate": 3.987560898669883e-08,
+      "loss": 0.0073,
+      "step": 19486
+    },
+    {
+      "epoch": 4.428863636363636,
+      "grad_norm": 0.7227108350814433,
+      "learning_rate": 3.984424630013223e-08,
+      "loss": 0.0019,
+      "step": 19487
+    },
+    {
+      "epoch": 4.429090909090909,
+      "grad_norm": 2.2744921289036557,
+      "learning_rate": 3.9812895545912334e-08,
+      "loss": 0.0253,
+      "step": 19488
+    },
+    {
+      "epoch": 4.429318181818182,
+      "grad_norm": 1.5105436739842149,
+      "learning_rate": 3.978155672467836e-08,
+      "loss": 0.0129,
+      "step": 19489
+    },
+    {
+      "epoch": 4.429545454545455,
+      "grad_norm": 1.6897480863158678,
+      "learning_rate": 3.975022983706944e-08,
+      "loss": 0.0041,
+      "step": 19490
+    },
+    {
+      "epoch": 4.429772727272727,
+      "grad_norm": 3.822030542942741,
+      "learning_rate": 3.9718914883724455e-08,
+      "loss": 0.0333,
+      "step": 19491
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 2.6789853995357675,
+      "learning_rate": 3.9687611865281765e-08,
+      "loss": 0.0133,
+      "step": 19492
+    },
+    {
+      "epoch": 4.430227272727273,
+      "grad_norm": 0.41807788861596057,
+      "learning_rate": 3.9656320782379826e-08,
+      "loss": 0.0019,
+      "step": 19493
+    },
+    {
+      "epoch": 4.430454545454546,
+      "grad_norm": 0.46116734016625355,
+      "learning_rate": 3.962504163565667e-08,
+      "loss": 0.0012,
+      "step": 19494
+    },
+    {
+      "epoch": 4.430681818181818,
+      "grad_norm": 0.974282417015717,
+      "learning_rate": 3.9593774425750256e-08,
+      "loss": 0.0044,
+      "step": 19495
+    },
+    {
+      "epoch": 4.430909090909091,
+      "grad_norm": 2.305319306385842,
+      "learning_rate": 3.956251915329799e-08,
+      "loss": 0.007,
+      "step": 19496
+    },
+    {
+      "epoch": 4.431136363636364,
+      "grad_norm": 1.3168693421425077,
+      "learning_rate": 3.9531275818937285e-08,
+      "loss": 0.0026,
+      "step": 19497
+    },
+    {
+      "epoch": 4.431363636363637,
+      "grad_norm": 1.4555409071880003,
+      "learning_rate": 3.9500044423305335e-08,
+      "loss": 0.0042,
+      "step": 19498
+    },
+    {
+      "epoch": 4.431590909090909,
+      "grad_norm": 0.8154617209751884,
+      "learning_rate": 3.946882496703885e-08,
+      "loss": 0.0024,
+      "step": 19499
+    },
+    {
+      "epoch": 4.431818181818182,
+      "grad_norm": 2.261931549658542,
+      "learning_rate": 3.9437617450774695e-08,
+      "loss": 0.0125,
+      "step": 19500
+    },
+    {
+      "epoch": 4.432045454545454,
+      "grad_norm": 1.1763317487372529,
+      "learning_rate": 3.940642187514901e-08,
+      "loss": 0.0059,
+      "step": 19501
+    },
+    {
+      "epoch": 4.432272727272728,
+      "grad_norm": 0.4768231891269952,
+      "learning_rate": 3.9375238240797964e-08,
+      "loss": 0.0021,
+      "step": 19502
+    },
+    {
+      "epoch": 4.4325,
+      "grad_norm": 0.9538274112875951,
+      "learning_rate": 3.934406654835743e-08,
+      "loss": 0.0059,
+      "step": 19503
+    },
+    {
+      "epoch": 4.432727272727273,
+      "grad_norm": 2.0924601943122716,
+      "learning_rate": 3.9312906798463296e-08,
+      "loss": 0.009,
+      "step": 19504
+    },
+    {
+      "epoch": 4.432954545454545,
+      "grad_norm": 0.17978004581645715,
+      "learning_rate": 3.928175899175067e-08,
+      "loss": 0.0003,
+      "step": 19505
+    },
+    {
+      "epoch": 4.433181818181819,
+      "grad_norm": 2.446877586269523,
+      "learning_rate": 3.9250623128854815e-08,
+      "loss": 0.0105,
+      "step": 19506
+    },
+    {
+      "epoch": 4.433409090909091,
+      "grad_norm": 1.3801596438190797,
+      "learning_rate": 3.921949921041064e-08,
+      "loss": 0.004,
+      "step": 19507
+    },
+    {
+      "epoch": 4.433636363636364,
+      "grad_norm": 2.934891518736427,
+      "learning_rate": 3.918838723705291e-08,
+      "loss": 0.008,
+      "step": 19508
+    },
+    {
+      "epoch": 4.433863636363636,
+      "grad_norm": 1.1757030029987818,
+      "learning_rate": 3.915728720941592e-08,
+      "loss": 0.002,
+      "step": 19509
+    },
+    {
+      "epoch": 4.434090909090909,
+      "grad_norm": 1.184557917726705,
+      "learning_rate": 3.9126199128133867e-08,
+      "loss": 0.0061,
+      "step": 19510
+    },
+    {
+      "epoch": 4.434318181818182,
+      "grad_norm": 0.7304171139553373,
+      "learning_rate": 3.909512299384077e-08,
+      "loss": 0.0023,
+      "step": 19511
+    },
+    {
+      "epoch": 4.434545454545455,
+      "grad_norm": 1.163342421861292,
+      "learning_rate": 3.9064058807170224e-08,
+      "loss": 0.0054,
+      "step": 19512
+    },
+    {
+      "epoch": 4.434772727272727,
+      "grad_norm": 0.6658074042030467,
+      "learning_rate": 3.9033006568755874e-08,
+      "loss": 0.0048,
+      "step": 19513
+    },
+    {
+      "epoch": 4.435,
+      "grad_norm": 0.6085707890352605,
+      "learning_rate": 3.9001966279230694e-08,
+      "loss": 0.0013,
+      "step": 19514
+    },
+    {
+      "epoch": 4.435227272727273,
+      "grad_norm": 2.1114078906288225,
+      "learning_rate": 3.8970937939227794e-08,
+      "loss": 0.0193,
+      "step": 19515
+    },
+    {
+      "epoch": 4.4354545454545455,
+      "grad_norm": 0.8089481132299574,
+      "learning_rate": 3.893992154937978e-08,
+      "loss": 0.0034,
+      "step": 19516
+    },
+    {
+      "epoch": 4.435681818181818,
+      "grad_norm": 1.5393070794912256,
+      "learning_rate": 3.8908917110319217e-08,
+      "loss": 0.0062,
+      "step": 19517
+    },
+    {
+      "epoch": 4.435909090909091,
+      "grad_norm": 1.4057468347015663,
+      "learning_rate": 3.8877924622678366e-08,
+      "loss": 0.0034,
+      "step": 19518
+    },
+    {
+      "epoch": 4.436136363636364,
+      "grad_norm": 9.66151007400738,
+      "learning_rate": 3.8846944087089154e-08,
+      "loss": 0.0179,
+      "step": 19519
+    },
+    {
+      "epoch": 4.4363636363636365,
+      "grad_norm": 1.2320929838740942,
+      "learning_rate": 3.881597550418337e-08,
+      "loss": 0.0067,
+      "step": 19520
+    },
+    {
+      "epoch": 4.436590909090909,
+      "grad_norm": 1.2535121014777704,
+      "learning_rate": 3.8785018874592447e-08,
+      "loss": 0.006,
+      "step": 19521
+    },
+    {
+      "epoch": 4.4368181818181816,
+      "grad_norm": 0.46918025407294317,
+      "learning_rate": 3.875407419894783e-08,
+      "loss": 0.002,
+      "step": 19522
+    },
+    {
+      "epoch": 4.437045454545455,
+      "grad_norm": 1.4044835708693302,
+      "learning_rate": 3.8723141477880336e-08,
+      "loss": 0.0049,
+      "step": 19523
+    },
+    {
+      "epoch": 4.4372727272727275,
+      "grad_norm": 0.8164114761715068,
+      "learning_rate": 3.8692220712020776e-08,
+      "loss": 0.0026,
+      "step": 19524
+    },
+    {
+      "epoch": 4.4375,
+      "grad_norm": 2.789314918930477,
+      "learning_rate": 3.866131190199969e-08,
+      "loss": 0.0025,
+      "step": 19525
+    },
+    {
+      "epoch": 4.4377272727272725,
+      "grad_norm": 1.6349853021408969,
+      "learning_rate": 3.863041504844747e-08,
+      "loss": 0.013,
+      "step": 19526
+    },
+    {
+      "epoch": 4.437954545454545,
+      "grad_norm": 0.5210825477245615,
+      "learning_rate": 3.859953015199404e-08,
+      "loss": 0.0028,
+      "step": 19527
+    },
+    {
+      "epoch": 4.4381818181818184,
+      "grad_norm": 2.440057421639351,
+      "learning_rate": 3.8568657213269164e-08,
+      "loss": 0.016,
+      "step": 19528
+    },
+    {
+      "epoch": 4.438409090909091,
+      "grad_norm": 0.5009176749883211,
+      "learning_rate": 3.853779623290241e-08,
+      "loss": 0.0024,
+      "step": 19529
+    },
+    {
+      "epoch": 4.4386363636363635,
+      "grad_norm": 0.43865963856291545,
+      "learning_rate": 3.850694721152327e-08,
+      "loss": 0.0038,
+      "step": 19530
+    },
+    {
+      "epoch": 4.438863636363636,
+      "grad_norm": 1.2447048499331663,
+      "learning_rate": 3.847611014976056e-08,
+      "loss": 0.01,
+      "step": 19531
+    },
+    {
+      "epoch": 4.439090909090909,
+      "grad_norm": 0.4321904111980826,
+      "learning_rate": 3.8445285048243145e-08,
+      "loss": 0.0019,
+      "step": 19532
+    },
+    {
+      "epoch": 4.439318181818182,
+      "grad_norm": 1.018978229142318,
+      "learning_rate": 3.841447190759976e-08,
+      "loss": 0.0046,
+      "step": 19533
+    },
+    {
+      "epoch": 4.4395454545454545,
+      "grad_norm": 1.6117655268206181,
+      "learning_rate": 3.838367072845864e-08,
+      "loss": 0.0028,
+      "step": 19534
+    },
+    {
+      "epoch": 4.439772727272727,
+      "grad_norm": 0.3819913528469988,
+      "learning_rate": 3.8352881511447974e-08,
+      "loss": 0.0007,
+      "step": 19535
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 0.9473880064725155,
+      "learning_rate": 3.832210425719546e-08,
+      "loss": 0.0045,
+      "step": 19536
+    },
+    {
+      "epoch": 4.440227272727273,
+      "grad_norm": 1.62231618264992,
+      "learning_rate": 3.829133896632871e-08,
+      "loss": 0.0111,
+      "step": 19537
+    },
+    {
+      "epoch": 4.440454545454545,
+      "grad_norm": 0.8887492735905125,
+      "learning_rate": 3.8260585639475145e-08,
+      "loss": 0.0035,
+      "step": 19538
+    },
+    {
+      "epoch": 4.440681818181818,
+      "grad_norm": 0.263882567730443,
+      "learning_rate": 3.82298442772619e-08,
+      "loss": 0.0009,
+      "step": 19539
+    },
+    {
+      "epoch": 4.440909090909091,
+      "grad_norm": 0.5388957498659069,
+      "learning_rate": 3.819911488031576e-08,
+      "loss": 0.0023,
+      "step": 19540
+    },
+    {
+      "epoch": 4.441136363636364,
+      "grad_norm": 0.3944064734218467,
+      "learning_rate": 3.8168397449263455e-08,
+      "loss": 0.0013,
+      "step": 19541
+    },
+    {
+      "epoch": 4.441363636363636,
+      "grad_norm": 0.7049013845213546,
+      "learning_rate": 3.813769198473122e-08,
+      "loss": 0.0033,
+      "step": 19542
+    },
+    {
+      "epoch": 4.441590909090909,
+      "grad_norm": 1.3301725887950597,
+      "learning_rate": 3.8106998487345354e-08,
+      "loss": 0.0064,
+      "step": 19543
+    },
+    {
+      "epoch": 4.441818181818181,
+      "grad_norm": 3.0925181509205237,
+      "learning_rate": 3.8076316957731674e-08,
+      "loss": 0.0035,
+      "step": 19544
+    },
+    {
+      "epoch": 4.442045454545455,
+      "grad_norm": 0.33237566737974933,
+      "learning_rate": 3.804564739651587e-08,
+      "loss": 0.0012,
+      "step": 19545
+    },
+    {
+      "epoch": 4.442272727272727,
+      "grad_norm": 0.28233598164495644,
+      "learning_rate": 3.8014989804323197e-08,
+      "loss": 0.0011,
+      "step": 19546
+    },
+    {
+      "epoch": 4.4425,
+      "grad_norm": 1.3262844852423656,
+      "learning_rate": 3.798434418177899e-08,
+      "loss": 0.0097,
+      "step": 19547
+    },
+    {
+      "epoch": 4.442727272727272,
+      "grad_norm": 0.686033618483297,
+      "learning_rate": 3.795371052950809e-08,
+      "loss": 0.0085,
+      "step": 19548
+    },
+    {
+      "epoch": 4.442954545454546,
+      "grad_norm": 0.2923517579483069,
+      "learning_rate": 3.7923088848135225e-08,
+      "loss": 0.0006,
+      "step": 19549
+    },
+    {
+      "epoch": 4.443181818181818,
+      "grad_norm": 0.6457757137413441,
+      "learning_rate": 3.789247913828474e-08,
+      "loss": 0.002,
+      "step": 19550
+    },
+    {
+      "epoch": 4.443409090909091,
+      "grad_norm": 2.0389953310182576,
+      "learning_rate": 3.786188140058093e-08,
+      "loss": 0.013,
+      "step": 19551
+    },
+    {
+      "epoch": 4.443636363636363,
+      "grad_norm": 1.4210427973150526,
+      "learning_rate": 3.7831295635647744e-08,
+      "loss": 0.0073,
+      "step": 19552
+    },
+    {
+      "epoch": 4.443863636363636,
+      "grad_norm": 2.6434632245875833,
+      "learning_rate": 3.7800721844108714e-08,
+      "loss": 0.0103,
+      "step": 19553
+    },
+    {
+      "epoch": 4.444090909090909,
+      "grad_norm": 0.7453408809567538,
+      "learning_rate": 3.777016002658743e-08,
+      "loss": 0.0033,
+      "step": 19554
+    },
+    {
+      "epoch": 4.444318181818182,
+      "grad_norm": 3.3472582508190496,
+      "learning_rate": 3.7739610183707e-08,
+      "loss": 0.0041,
+      "step": 19555
+    },
+    {
+      "epoch": 4.444545454545454,
+      "grad_norm": 0.8994755387285205,
+      "learning_rate": 3.770907231609049e-08,
+      "loss": 0.0043,
+      "step": 19556
+    },
+    {
+      "epoch": 4.444772727272727,
+      "grad_norm": 0.5359244298119182,
+      "learning_rate": 3.767854642436064e-08,
+      "loss": 0.0028,
+      "step": 19557
+    },
+    {
+      "epoch": 4.445,
+      "grad_norm": 2.468019498312316,
+      "learning_rate": 3.7648032509139816e-08,
+      "loss": 0.0043,
+      "step": 19558
+    },
+    {
+      "epoch": 4.445227272727273,
+      "grad_norm": 0.39488618163064243,
+      "learning_rate": 3.761753057105022e-08,
+      "loss": 0.0008,
+      "step": 19559
+    },
+    {
+      "epoch": 4.445454545454545,
+      "grad_norm": 1.1159335561933321,
+      "learning_rate": 3.7587040610714e-08,
+      "loss": 0.0036,
+      "step": 19560
+    },
+    {
+      "epoch": 4.445681818181818,
+      "grad_norm": 2.104875665438524,
+      "learning_rate": 3.755656262875288e-08,
+      "loss": 0.0104,
+      "step": 19561
+    },
+    {
+      "epoch": 4.445909090909091,
+      "grad_norm": 0.4579267292893218,
+      "learning_rate": 3.752609662578816e-08,
+      "loss": 0.0014,
+      "step": 19562
+    },
+    {
+      "epoch": 4.446136363636364,
+      "grad_norm": 2.51037287198689,
+      "learning_rate": 3.749564260244129e-08,
+      "loss": 0.0164,
+      "step": 19563
+    },
+    {
+      "epoch": 4.446363636363636,
+      "grad_norm": 0.4742887331425271,
+      "learning_rate": 3.7465200559333173e-08,
+      "loss": 0.0025,
+      "step": 19564
+    },
+    {
+      "epoch": 4.446590909090909,
+      "grad_norm": 0.3968469896944992,
+      "learning_rate": 3.743477049708469e-08,
+      "loss": 0.0011,
+      "step": 19565
+    },
+    {
+      "epoch": 4.446818181818182,
+      "grad_norm": 0.8909129995775255,
+      "learning_rate": 3.740435241631639e-08,
+      "loss": 0.0037,
+      "step": 19566
+    },
+    {
+      "epoch": 4.447045454545455,
+      "grad_norm": 0.858717799067022,
+      "learning_rate": 3.7373946317648403e-08,
+      "loss": 0.0025,
+      "step": 19567
+    },
+    {
+      "epoch": 4.447272727272727,
+      "grad_norm": 0.25783934520746266,
+      "learning_rate": 3.734355220170079e-08,
+      "loss": 0.0011,
+      "step": 19568
+    },
+    {
+      "epoch": 4.4475,
+      "grad_norm": 0.4263727904480992,
+      "learning_rate": 3.7313170069093324e-08,
+      "loss": 0.0013,
+      "step": 19569
+    },
+    {
+      "epoch": 4.447727272727272,
+      "grad_norm": 0.16590441464765293,
+      "learning_rate": 3.7282799920445725e-08,
+      "loss": 0.0005,
+      "step": 19570
+    },
+    {
+      "epoch": 4.447954545454546,
+      "grad_norm": 0.9200143823079436,
+      "learning_rate": 3.725244175637709e-08,
+      "loss": 0.004,
+      "step": 19571
+    },
+    {
+      "epoch": 4.448181818181818,
+      "grad_norm": 1.5487988342733825,
+      "learning_rate": 3.722209557750649e-08,
+      "loss": 0.02,
+      "step": 19572
+    },
+    {
+      "epoch": 4.448409090909091,
+      "grad_norm": 1.401044482554512,
+      "learning_rate": 3.719176138445288e-08,
+      "loss": 0.007,
+      "step": 19573
+    },
+    {
+      "epoch": 4.448636363636363,
+      "grad_norm": 1.0998973545635897,
+      "learning_rate": 3.7161439177834794e-08,
+      "loss": 0.0053,
+      "step": 19574
+    },
+    {
+      "epoch": 4.448863636363637,
+      "grad_norm": 0.6760369474024549,
+      "learning_rate": 3.713112895827042e-08,
+      "loss": 0.0014,
+      "step": 19575
+    },
+    {
+      "epoch": 4.449090909090909,
+      "grad_norm": 1.9169441149717692,
+      "learning_rate": 3.7100830726377876e-08,
+      "loss": 0.0117,
+      "step": 19576
+    },
+    {
+      "epoch": 4.449318181818182,
+      "grad_norm": 2.60977944579798,
+      "learning_rate": 3.707054448277507e-08,
+      "loss": 0.015,
+      "step": 19577
+    },
+    {
+      "epoch": 4.449545454545454,
+      "grad_norm": 1.5172014434954502,
+      "learning_rate": 3.7040270228079555e-08,
+      "loss": 0.0034,
+      "step": 19578
+    },
+    {
+      "epoch": 4.449772727272728,
+      "grad_norm": 0.44414662761609425,
+      "learning_rate": 3.70100079629087e-08,
+      "loss": 0.002,
+      "step": 19579
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 1.6538735823914448,
+      "learning_rate": 3.697975768787956e-08,
+      "loss": 0.0068,
+      "step": 19580
+    },
+    {
+      "epoch": 4.450227272727273,
+      "grad_norm": 1.9351488979467888,
+      "learning_rate": 3.6949519403609094e-08,
+      "loss": 0.0093,
+      "step": 19581
+    },
+    {
+      "epoch": 4.450454545454545,
+      "grad_norm": 0.3579923107909528,
+      "learning_rate": 3.6919293110713736e-08,
+      "loss": 0.0015,
+      "step": 19582
+    },
+    {
+      "epoch": 4.450681818181819,
+      "grad_norm": 1.1671902704666726,
+      "learning_rate": 3.688907880981009e-08,
+      "loss": 0.0053,
+      "step": 19583
+    },
+    {
+      "epoch": 4.450909090909091,
+      "grad_norm": 1.199037336330572,
+      "learning_rate": 3.685887650151411e-08,
+      "loss": 0.0037,
+      "step": 19584
+    },
+    {
+      "epoch": 4.451136363636364,
+      "grad_norm": 2.599450072423746,
+      "learning_rate": 3.682868618644163e-08,
+      "loss": 0.0113,
+      "step": 19585
+    },
+    {
+      "epoch": 4.451363636363636,
+      "grad_norm": 0.5687161522498644,
+      "learning_rate": 3.679850786520833e-08,
+      "loss": 0.0048,
+      "step": 19586
+    },
+    {
+      "epoch": 4.451590909090909,
+      "grad_norm": 1.590257618619759,
+      "learning_rate": 3.67683415384297e-08,
+      "loss": 0.008,
+      "step": 19587
+    },
+    {
+      "epoch": 4.451818181818182,
+      "grad_norm": 2.133118671910525,
+      "learning_rate": 3.673818720672087e-08,
+      "loss": 0.0059,
+      "step": 19588
+    },
+    {
+      "epoch": 4.452045454545455,
+      "grad_norm": 0.9437514306354988,
+      "learning_rate": 3.670804487069663e-08,
+      "loss": 0.0065,
+      "step": 19589
+    },
+    {
+      "epoch": 4.452272727272727,
+      "grad_norm": 0.5401041127272436,
+      "learning_rate": 3.667791453097168e-08,
+      "loss": 0.0016,
+      "step": 19590
+    },
+    {
+      "epoch": 4.4525,
+      "grad_norm": 0.5226073375482051,
+      "learning_rate": 3.66477961881604e-08,
+      "loss": 0.0022,
+      "step": 19591
+    },
+    {
+      "epoch": 4.452727272727273,
+      "grad_norm": 1.0391950206501182,
+      "learning_rate": 3.661768984287703e-08,
+      "loss": 0.0028,
+      "step": 19592
+    },
+    {
+      "epoch": 4.452954545454546,
+      "grad_norm": 1.1101367879277166,
+      "learning_rate": 3.658759549573537e-08,
+      "loss": 0.004,
+      "step": 19593
+    },
+    {
+      "epoch": 4.453181818181818,
+      "grad_norm": 2.9282780057073614,
+      "learning_rate": 3.6557513147349245e-08,
+      "loss": 0.046,
+      "step": 19594
+    },
+    {
+      "epoch": 4.453409090909091,
+      "grad_norm": 0.6709972993164736,
+      "learning_rate": 3.652744279833205e-08,
+      "loss": 0.003,
+      "step": 19595
+    },
+    {
+      "epoch": 4.453636363636363,
+      "grad_norm": 0.2981847964317201,
+      "learning_rate": 3.649738444929697e-08,
+      "loss": 0.001,
+      "step": 19596
+    },
+    {
+      "epoch": 4.453863636363637,
+      "grad_norm": 1.690455438646056,
+      "learning_rate": 3.646733810085693e-08,
+      "loss": 0.0052,
+      "step": 19597
+    },
+    {
+      "epoch": 4.454090909090909,
+      "grad_norm": 0.3237987164553364,
+      "learning_rate": 3.6437303753624475e-08,
+      "loss": 0.0012,
+      "step": 19598
+    },
+    {
+      "epoch": 4.454318181818182,
+      "grad_norm": 1.3974377157379902,
+      "learning_rate": 3.6407281408212325e-08,
+      "loss": 0.0087,
+      "step": 19599
+    },
+    {
+      "epoch": 4.454545454545454,
+      "grad_norm": 0.682596167472254,
+      "learning_rate": 3.6377271065232495e-08,
+      "loss": 0.0039,
+      "step": 19600
+    },
+    {
+      "epoch": 4.4547727272727276,
+      "grad_norm": 0.5879248376286884,
+      "learning_rate": 3.6347272725297116e-08,
+      "loss": 0.001,
+      "step": 19601
+    },
+    {
+      "epoch": 4.455,
+      "grad_norm": 0.24962590614170144,
+      "learning_rate": 3.631728638901773e-08,
+      "loss": 0.0006,
+      "step": 19602
+    },
+    {
+      "epoch": 4.455227272727273,
+      "grad_norm": 0.49046638629151923,
+      "learning_rate": 3.628731205700592e-08,
+      "loss": 0.0023,
+      "step": 19603
+    },
+    {
+      "epoch": 4.455454545454545,
+      "grad_norm": 1.8629555905914428,
+      "learning_rate": 3.62573497298728e-08,
+      "loss": 0.0052,
+      "step": 19604
+    },
+    {
+      "epoch": 4.4556818181818185,
+      "grad_norm": 0.6966526387431534,
+      "learning_rate": 3.6227399408229616e-08,
+      "loss": 0.0043,
+      "step": 19605
+    },
+    {
+      "epoch": 4.455909090909091,
+      "grad_norm": 0.41585322770212524,
+      "learning_rate": 3.619746109268678e-08,
+      "loss": 0.0026,
+      "step": 19606
+    },
+    {
+      "epoch": 4.456136363636364,
+      "grad_norm": 0.5711810748339528,
+      "learning_rate": 3.616753478385492e-08,
+      "loss": 0.0038,
+      "step": 19607
+    },
+    {
+      "epoch": 4.456363636363636,
+      "grad_norm": 0.43941416056851446,
+      "learning_rate": 3.613762048234437e-08,
+      "loss": 0.0008,
+      "step": 19608
+    },
+    {
+      "epoch": 4.4565909090909095,
+      "grad_norm": 1.2268882676447987,
+      "learning_rate": 3.6107718188765e-08,
+      "loss": 0.0049,
+      "step": 19609
+    },
+    {
+      "epoch": 4.456818181818182,
+      "grad_norm": 1.8368769567100067,
+      "learning_rate": 3.607782790372667e-08,
+      "loss": 0.0188,
+      "step": 19610
+    },
+    {
+      "epoch": 4.4570454545454545,
+      "grad_norm": 0.8587438801990016,
+      "learning_rate": 3.604794962783882e-08,
+      "loss": 0.0028,
+      "step": 19611
+    },
+    {
+      "epoch": 4.457272727272727,
+      "grad_norm": 1.4877504034197955,
+      "learning_rate": 3.6018083361710824e-08,
+      "loss": 0.0108,
+      "step": 19612
+    },
+    {
+      "epoch": 4.4575,
+      "grad_norm": 0.4293375977934097,
+      "learning_rate": 3.598822910595157e-08,
+      "loss": 0.0009,
+      "step": 19613
+    },
+    {
+      "epoch": 4.457727272727273,
+      "grad_norm": 0.4775400811357879,
+      "learning_rate": 3.595838686117008e-08,
+      "loss": 0.002,
+      "step": 19614
+    },
+    {
+      "epoch": 4.4579545454545455,
+      "grad_norm": 0.2370978596758775,
+      "learning_rate": 3.592855662797455e-08,
+      "loss": 0.001,
+      "step": 19615
+    },
+    {
+      "epoch": 4.458181818181818,
+      "grad_norm": 0.641790191301729,
+      "learning_rate": 3.589873840697353e-08,
+      "loss": 0.0026,
+      "step": 19616
+    },
+    {
+      "epoch": 4.4584090909090905,
+      "grad_norm": 1.8368430940707363,
+      "learning_rate": 3.586893219877492e-08,
+      "loss": 0.0047,
+      "step": 19617
+    },
+    {
+      "epoch": 4.458636363636364,
+      "grad_norm": 1.3417912219094679,
+      "learning_rate": 3.583913800398665e-08,
+      "loss": 0.0098,
+      "step": 19618
+    },
+    {
+      "epoch": 4.4588636363636365,
+      "grad_norm": 3.2239959868764663,
+      "learning_rate": 3.580935582321615e-08,
+      "loss": 0.0127,
+      "step": 19619
+    },
+    {
+      "epoch": 4.459090909090909,
+      "grad_norm": 1.3294228720035262,
+      "learning_rate": 3.577958565707078e-08,
+      "loss": 0.0066,
+      "step": 19620
+    },
+    {
+      "epoch": 4.4593181818181815,
+      "grad_norm": 0.3642665649993246,
+      "learning_rate": 3.5749827506157614e-08,
+      "loss": 0.0014,
+      "step": 19621
+    },
+    {
+      "epoch": 4.459545454545455,
+      "grad_norm": 3.252244538787776,
+      "learning_rate": 3.5720081371083474e-08,
+      "loss": 0.0041,
+      "step": 19622
+    },
+    {
+      "epoch": 4.459772727272727,
+      "grad_norm": 1.591422054247374,
+      "learning_rate": 3.5690347252454946e-08,
+      "loss": 0.0134,
+      "step": 19623
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 1.0986592380304065,
+      "learning_rate": 3.566062515087829e-08,
+      "loss": 0.0041,
+      "step": 19624
+    },
+    {
+      "epoch": 4.4602272727272725,
+      "grad_norm": 1.2232807116005158,
+      "learning_rate": 3.5630915066959674e-08,
+      "loss": 0.008,
+      "step": 19625
+    },
+    {
+      "epoch": 4.460454545454546,
+      "grad_norm": 0.9866743055503978,
+      "learning_rate": 3.560121700130495e-08,
+      "loss": 0.0037,
+      "step": 19626
+    },
+    {
+      "epoch": 4.460681818181818,
+      "grad_norm": 1.097866633850882,
+      "learning_rate": 3.5571530954519666e-08,
+      "loss": 0.0052,
+      "step": 19627
+    },
+    {
+      "epoch": 4.460909090909091,
+      "grad_norm": 0.600585272216134,
+      "learning_rate": 3.554185692720917e-08,
+      "loss": 0.0024,
+      "step": 19628
+    },
+    {
+      "epoch": 4.4611363636363635,
+      "grad_norm": 0.2503754586653748,
+      "learning_rate": 3.551219491997854e-08,
+      "loss": 0.0009,
+      "step": 19629
+    },
+    {
+      "epoch": 4.461363636363636,
+      "grad_norm": 0.6093030101513474,
+      "learning_rate": 3.54825449334327e-08,
+      "loss": 0.0048,
+      "step": 19630
+    },
+    {
+      "epoch": 4.461590909090909,
+      "grad_norm": 1.7842466797510945,
+      "learning_rate": 3.545290696817631e-08,
+      "loss": 0.007,
+      "step": 19631
+    },
+    {
+      "epoch": 4.461818181818182,
+      "grad_norm": 0.9192022154330624,
+      "learning_rate": 3.542328102481361e-08,
+      "loss": 0.0049,
+      "step": 19632
+    },
+    {
+      "epoch": 4.462045454545454,
+      "grad_norm": 1.5515409751367835,
+      "learning_rate": 3.539366710394877e-08,
+      "loss": 0.0062,
+      "step": 19633
+    },
+    {
+      "epoch": 4.462272727272727,
+      "grad_norm": 2.4853009024724093,
+      "learning_rate": 3.536406520618568e-08,
+      "loss": 0.0124,
+      "step": 19634
+    },
+    {
+      "epoch": 4.4625,
+      "grad_norm": 0.7008283132695153,
+      "learning_rate": 3.533447533212796e-08,
+      "loss": 0.0016,
+      "step": 19635
+    },
+    {
+      "epoch": 4.462727272727273,
+      "grad_norm": 1.391219046943567,
+      "learning_rate": 3.530489748237907e-08,
+      "loss": 0.0063,
+      "step": 19636
+    },
+    {
+      "epoch": 4.462954545454545,
+      "grad_norm": 0.8287301034625406,
+      "learning_rate": 3.527533165754209e-08,
+      "loss": 0.0025,
+      "step": 19637
+    },
+    {
+      "epoch": 4.463181818181818,
+      "grad_norm": 2.2000680564491586,
+      "learning_rate": 3.524577785821992e-08,
+      "loss": 0.0083,
+      "step": 19638
+    },
+    {
+      "epoch": 4.46340909090909,
+      "grad_norm": 3.0015398107578717,
+      "learning_rate": 3.521623608501515e-08,
+      "loss": 0.026,
+      "step": 19639
+    },
+    {
+      "epoch": 4.463636363636364,
+      "grad_norm": 3.1875761366388997,
+      "learning_rate": 3.5186706338530344e-08,
+      "loss": 0.0168,
+      "step": 19640
+    },
+    {
+      "epoch": 4.463863636363636,
+      "grad_norm": 1.522069450922719,
+      "learning_rate": 3.5157188619367526e-08,
+      "loss": 0.0087,
+      "step": 19641
+    },
+    {
+      "epoch": 4.464090909090909,
+      "grad_norm": 2.3117096576655243,
+      "learning_rate": 3.512768292812871e-08,
+      "loss": 0.0242,
+      "step": 19642
+    },
+    {
+      "epoch": 4.464318181818181,
+      "grad_norm": 0.5842864818561232,
+      "learning_rate": 3.50981892654155e-08,
+      "loss": 0.0016,
+      "step": 19643
+    },
+    {
+      "epoch": 4.464545454545455,
+      "grad_norm": 1.2176054000167464,
+      "learning_rate": 3.506870763182929e-08,
+      "loss": 0.0058,
+      "step": 19644
+    },
+    {
+      "epoch": 4.464772727272727,
+      "grad_norm": 2.428607016418709,
+      "learning_rate": 3.5039238027971546e-08,
+      "loss": 0.0047,
+      "step": 19645
+    },
+    {
+      "epoch": 4.465,
+      "grad_norm": 0.3831324502546472,
+      "learning_rate": 3.500978045444282e-08,
+      "loss": 0.0016,
+      "step": 19646
+    },
+    {
+      "epoch": 4.465227272727272,
+      "grad_norm": 1.2771491439473999,
+      "learning_rate": 3.4980334911843965e-08,
+      "loss": 0.0034,
+      "step": 19647
+    },
+    {
+      "epoch": 4.465454545454546,
+      "grad_norm": 1.5079380563497335,
+      "learning_rate": 3.49509014007754e-08,
+      "loss": 0.0119,
+      "step": 19648
+    },
+    {
+      "epoch": 4.465681818181818,
+      "grad_norm": 1.7818362585824343,
+      "learning_rate": 3.492147992183746e-08,
+      "loss": 0.0077,
+      "step": 19649
+    },
+    {
+      "epoch": 4.465909090909091,
+      "grad_norm": 1.623565555884101,
+      "learning_rate": 3.4892070475629965e-08,
+      "loss": 0.0195,
+      "step": 19650
+    },
+    {
+      "epoch": 4.466136363636363,
+      "grad_norm": 0.3670671826711621,
+      "learning_rate": 3.4862673062752566e-08,
+      "loss": 0.001,
+      "step": 19651
+    },
+    {
+      "epoch": 4.466363636363637,
+      "grad_norm": 0.8595244242989762,
+      "learning_rate": 3.4833287683804855e-08,
+      "loss": 0.0055,
+      "step": 19652
+    },
+    {
+      "epoch": 4.466590909090909,
+      "grad_norm": 1.17000583842709,
+      "learning_rate": 3.4803914339386e-08,
+      "loss": 0.004,
+      "step": 19653
+    },
+    {
+      "epoch": 4.466818181818182,
+      "grad_norm": 2.4877227801131383,
+      "learning_rate": 3.477455303009497e-08,
+      "loss": 0.0126,
+      "step": 19654
+    },
+    {
+      "epoch": 4.467045454545454,
+      "grad_norm": 0.562137982197674,
+      "learning_rate": 3.474520375653059e-08,
+      "loss": 0.0042,
+      "step": 19655
+    },
+    {
+      "epoch": 4.467272727272727,
+      "grad_norm": 0.27305840803448994,
+      "learning_rate": 3.471586651929114e-08,
+      "loss": 0.0014,
+      "step": 19656
+    },
+    {
+      "epoch": 4.4675,
+      "grad_norm": 0.21810163882699982,
+      "learning_rate": 3.4686541318975086e-08,
+      "loss": 0.0006,
+      "step": 19657
+    },
+    {
+      "epoch": 4.467727272727273,
+      "grad_norm": 0.5154857585176086,
+      "learning_rate": 3.465722815618036e-08,
+      "loss": 0.0018,
+      "step": 19658
+    },
+    {
+      "epoch": 4.467954545454545,
+      "grad_norm": 0.5945293691102752,
+      "learning_rate": 3.462792703150454e-08,
+      "loss": 0.0043,
+      "step": 19659
+    },
+    {
+      "epoch": 4.468181818181818,
+      "grad_norm": 0.87648055000714,
+      "learning_rate": 3.459863794554527e-08,
+      "loss": 0.0034,
+      "step": 19660
+    },
+    {
+      "epoch": 4.468409090909091,
+      "grad_norm": 0.772727509866827,
+      "learning_rate": 3.4569360898899785e-08,
+      "loss": 0.0038,
+      "step": 19661
+    },
+    {
+      "epoch": 4.468636363636364,
+      "grad_norm": 0.8251910885422867,
+      "learning_rate": 3.454009589216524e-08,
+      "loss": 0.0049,
+      "step": 19662
+    },
+    {
+      "epoch": 4.468863636363636,
+      "grad_norm": 1.3977091541384612,
+      "learning_rate": 3.451084292593811e-08,
+      "loss": 0.0063,
+      "step": 19663
+    },
+    {
+      "epoch": 4.469090909090909,
+      "grad_norm": 0.8656648613800445,
+      "learning_rate": 3.448160200081506e-08,
+      "loss": 0.0051,
+      "step": 19664
+    },
+    {
+      "epoch": 4.469318181818182,
+      "grad_norm": 2.0354101599619923,
+      "learning_rate": 3.445237311739236e-08,
+      "loss": 0.0122,
+      "step": 19665
+    },
+    {
+      "epoch": 4.469545454545455,
+      "grad_norm": 0.8807078801218088,
+      "learning_rate": 3.442315627626605e-08,
+      "loss": 0.0033,
+      "step": 19666
+    },
+    {
+      "epoch": 4.469772727272727,
+      "grad_norm": 1.0561202527519329,
+      "learning_rate": 3.439395147803198e-08,
+      "loss": 0.0035,
+      "step": 19667
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 2.449286529172108,
+      "learning_rate": 3.4364758723285516e-08,
+      "loss": 0.0097,
+      "step": 19668
+    },
+    {
+      "epoch": 4.470227272727273,
+      "grad_norm": 1.5027024557053263,
+      "learning_rate": 3.433557801262206e-08,
+      "loss": 0.0044,
+      "step": 19669
+    },
+    {
+      "epoch": 4.470454545454546,
+      "grad_norm": 0.5001126143937817,
+      "learning_rate": 3.430640934663658e-08,
+      "loss": 0.003,
+      "step": 19670
+    },
+    {
+      "epoch": 4.470681818181818,
+      "grad_norm": 1.619100409378238,
+      "learning_rate": 3.4277252725923995e-08,
+      "loss": 0.0054,
+      "step": 19671
+    },
+    {
+      "epoch": 4.470909090909091,
+      "grad_norm": 0.5816336759245624,
+      "learning_rate": 3.424810815107878e-08,
+      "loss": 0.0015,
+      "step": 19672
+    },
+    {
+      "epoch": 4.471136363636363,
+      "grad_norm": 0.7389487718762707,
+      "learning_rate": 3.42189756226953e-08,
+      "loss": 0.0045,
+      "step": 19673
+    },
+    {
+      "epoch": 4.471363636363637,
+      "grad_norm": 0.9565688346764439,
+      "learning_rate": 3.4189855141367536e-08,
+      "loss": 0.0055,
+      "step": 19674
+    },
+    {
+      "epoch": 4.471590909090909,
+      "grad_norm": 1.6790503001104489,
+      "learning_rate": 3.416074670768946e-08,
+      "loss": 0.0089,
+      "step": 19675
+    },
+    {
+      "epoch": 4.471818181818182,
+      "grad_norm": 1.597790897509722,
+      "learning_rate": 3.4131650322254405e-08,
+      "loss": 0.0104,
+      "step": 19676
+    },
+    {
+      "epoch": 4.472045454545454,
+      "grad_norm": 2.3555173109948098,
+      "learning_rate": 3.4102565985655866e-08,
+      "loss": 0.0098,
+      "step": 19677
+    },
+    {
+      "epoch": 4.472272727272728,
+      "grad_norm": 3.4171067309051,
+      "learning_rate": 3.4073493698486905e-08,
+      "loss": 0.016,
+      "step": 19678
+    },
+    {
+      "epoch": 4.4725,
+      "grad_norm": 0.5407758556350787,
+      "learning_rate": 3.404443346134031e-08,
+      "loss": 0.0017,
+      "step": 19679
+    },
+    {
+      "epoch": 4.472727272727273,
+      "grad_norm": 0.5799834322202175,
+      "learning_rate": 3.4015385274808746e-08,
+      "loss": 0.0016,
+      "step": 19680
+    },
+    {
+      "epoch": 4.472954545454545,
+      "grad_norm": 1.078595459686797,
+      "learning_rate": 3.3986349139484495e-08,
+      "loss": 0.0033,
+      "step": 19681
+    },
+    {
+      "epoch": 4.473181818181819,
+      "grad_norm": 0.7044121269181329,
+      "learning_rate": 3.3957325055959594e-08,
+      "loss": 0.0037,
+      "step": 19682
+    },
+    {
+      "epoch": 4.473409090909091,
+      "grad_norm": 0.5343851493629921,
+      "learning_rate": 3.392831302482601e-08,
+      "loss": 0.0023,
+      "step": 19683
+    },
+    {
+      "epoch": 4.473636363636364,
+      "grad_norm": 0.26175025028119464,
+      "learning_rate": 3.389931304667526e-08,
+      "loss": 0.0009,
+      "step": 19684
+    },
+    {
+      "epoch": 4.473863636363636,
+      "grad_norm": 0.8241716866899924,
+      "learning_rate": 3.387032512209877e-08,
+      "loss": 0.003,
+      "step": 19685
+    },
+    {
+      "epoch": 4.47409090909091,
+      "grad_norm": 2.101394586709387,
+      "learning_rate": 3.3841349251687655e-08,
+      "loss": 0.0113,
+      "step": 19686
+    },
+    {
+      "epoch": 4.474318181818182,
+      "grad_norm": 0.2346594868027231,
+      "learning_rate": 3.3812385436032754e-08,
+      "loss": 0.0006,
+      "step": 19687
+    },
+    {
+      "epoch": 4.474545454545455,
+      "grad_norm": 0.7630933830126656,
+      "learning_rate": 3.3783433675724644e-08,
+      "loss": 0.0016,
+      "step": 19688
+    },
+    {
+      "epoch": 4.474772727272727,
+      "grad_norm": 1.0158025917871247,
+      "learning_rate": 3.37544939713539e-08,
+      "loss": 0.0096,
+      "step": 19689
+    },
+    {
+      "epoch": 4.475,
+      "grad_norm": 0.8851907926664049,
+      "learning_rate": 3.372556632351046e-08,
+      "loss": 0.0054,
+      "step": 19690
+    },
+    {
+      "epoch": 4.475227272727273,
+      "grad_norm": 2.3580217412597833,
+      "learning_rate": 3.369665073278419e-08,
+      "loss": 0.0076,
+      "step": 19691
+    },
+    {
+      "epoch": 4.475454545454546,
+      "grad_norm": 0.898567301528637,
+      "learning_rate": 3.3667747199764776e-08,
+      "loss": 0.0023,
+      "step": 19692
+    },
+    {
+      "epoch": 4.475681818181818,
+      "grad_norm": 0.6569684482468025,
+      "learning_rate": 3.3638855725041804e-08,
+      "loss": 0.0025,
+      "step": 19693
+    },
+    {
+      "epoch": 4.475909090909091,
+      "grad_norm": 4.344244413634947,
+      "learning_rate": 3.360997630920411e-08,
+      "loss": 0.0095,
+      "step": 19694
+    },
+    {
+      "epoch": 4.476136363636364,
+      "grad_norm": 0.5090649693720004,
+      "learning_rate": 3.358110895284074e-08,
+      "loss": 0.0022,
+      "step": 19695
+    },
+    {
+      "epoch": 4.4763636363636365,
+      "grad_norm": 1.0768384079177722,
+      "learning_rate": 3.355225365654031e-08,
+      "loss": 0.0043,
+      "step": 19696
+    },
+    {
+      "epoch": 4.476590909090909,
+      "grad_norm": 0.8132822050806889,
+      "learning_rate": 3.35234104208914e-08,
+      "loss": 0.0036,
+      "step": 19697
+    },
+    {
+      "epoch": 4.476818181818182,
+      "grad_norm": 1.0020581100109294,
+      "learning_rate": 3.3494579246481987e-08,
+      "loss": 0.0076,
+      "step": 19698
+    },
+    {
+      "epoch": 4.477045454545454,
+      "grad_norm": 0.8945326814760582,
+      "learning_rate": 3.346576013389995e-08,
+      "loss": 0.0023,
+      "step": 19699
+    },
+    {
+      "epoch": 4.4772727272727275,
+      "grad_norm": 0.3485735756502476,
+      "learning_rate": 3.3436953083733004e-08,
+      "loss": 0.0015,
+      "step": 19700
+    },
+    {
+      "epoch": 4.4775,
+      "grad_norm": 1.1745930776736444,
+      "learning_rate": 3.340815809656873e-08,
+      "loss": 0.0052,
+      "step": 19701
+    },
+    {
+      "epoch": 4.477727272727273,
+      "grad_norm": 0.6809320859641607,
+      "learning_rate": 3.337937517299425e-08,
+      "loss": 0.0025,
+      "step": 19702
+    },
+    {
+      "epoch": 4.477954545454545,
+      "grad_norm": 0.67391254668841,
+      "learning_rate": 3.335060431359644e-08,
+      "loss": 0.0041,
+      "step": 19703
+    },
+    {
+      "epoch": 4.4781818181818185,
+      "grad_norm": 1.0042120721989234,
+      "learning_rate": 3.3321845518961905e-08,
+      "loss": 0.0032,
+      "step": 19704
+    },
+    {
+      "epoch": 4.478409090909091,
+      "grad_norm": 1.5812293146726863,
+      "learning_rate": 3.329309878967721e-08,
+      "loss": 0.0113,
+      "step": 19705
+    },
+    {
+      "epoch": 4.4786363636363635,
+      "grad_norm": 1.607137021440815,
+      "learning_rate": 3.326436412632861e-08,
+      "loss": 0.0195,
+      "step": 19706
+    },
+    {
+      "epoch": 4.478863636363636,
+      "grad_norm": 1.72942668036827,
+      "learning_rate": 3.3235641529501883e-08,
+      "loss": 0.0111,
+      "step": 19707
+    },
+    {
+      "epoch": 4.4790909090909095,
+      "grad_norm": 0.5875049770812288,
+      "learning_rate": 3.3206930999782824e-08,
+      "loss": 0.0064,
+      "step": 19708
+    },
+    {
+      "epoch": 4.479318181818182,
+      "grad_norm": 1.1222956956316217,
+      "learning_rate": 3.3178232537756855e-08,
+      "loss": 0.0032,
+      "step": 19709
+    },
+    {
+      "epoch": 4.4795454545454545,
+      "grad_norm": 0.6887459177496518,
+      "learning_rate": 3.314954614400928e-08,
+      "loss": 0.0038,
+      "step": 19710
+    },
+    {
+      "epoch": 4.479772727272727,
+      "grad_norm": 0.4661709469093402,
+      "learning_rate": 3.312087181912504e-08,
+      "loss": 0.002,
+      "step": 19711
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 2.301144050729614,
+      "learning_rate": 3.309220956368868e-08,
+      "loss": 0.0109,
+      "step": 19712
+    },
+    {
+      "epoch": 4.480227272727273,
+      "grad_norm": 1.314205509065552,
+      "learning_rate": 3.3063559378284854e-08,
+      "loss": 0.0122,
+      "step": 19713
+    },
+    {
+      "epoch": 4.4804545454545455,
+      "grad_norm": 0.41547864042160026,
+      "learning_rate": 3.3034921263497764e-08,
+      "loss": 0.001,
+      "step": 19714
+    },
+    {
+      "epoch": 4.480681818181818,
+      "grad_norm": 0.843596902056836,
+      "learning_rate": 3.30062952199113e-08,
+      "loss": 0.0022,
+      "step": 19715
+    },
+    {
+      "epoch": 4.4809090909090905,
+      "grad_norm": 0.8427598909822113,
+      "learning_rate": 3.2977681248109324e-08,
+      "loss": 0.004,
+      "step": 19716
+    },
+    {
+      "epoch": 4.481136363636364,
+      "grad_norm": 0.6140663889331804,
+      "learning_rate": 3.294907934867524e-08,
+      "loss": 0.0016,
+      "step": 19717
+    },
+    {
+      "epoch": 4.481363636363636,
+      "grad_norm": 0.5907352263972239,
+      "learning_rate": 3.292048952219233e-08,
+      "loss": 0.0021,
+      "step": 19718
+    },
+    {
+      "epoch": 4.481590909090909,
+      "grad_norm": 0.8771578406856739,
+      "learning_rate": 3.289191176924368e-08,
+      "loss": 0.0024,
+      "step": 19719
+    },
+    {
+      "epoch": 4.4818181818181815,
+      "grad_norm": 1.171339765199351,
+      "learning_rate": 3.286334609041188e-08,
+      "loss": 0.0034,
+      "step": 19720
+    },
+    {
+      "epoch": 4.482045454545455,
+      "grad_norm": 2.7596927658243664,
+      "learning_rate": 3.283479248627944e-08,
+      "loss": 0.0144,
+      "step": 19721
+    },
+    {
+      "epoch": 4.482272727272727,
+      "grad_norm": 2.2668512224578516,
+      "learning_rate": 3.2806250957428673e-08,
+      "loss": 0.012,
+      "step": 19722
+    },
+    {
+      "epoch": 4.4825,
+      "grad_norm": 0.2904173755808036,
+      "learning_rate": 3.2777721504441556e-08,
+      "loss": 0.0011,
+      "step": 19723
+    },
+    {
+      "epoch": 4.4827272727272724,
+      "grad_norm": 0.52474003157408,
+      "learning_rate": 3.2749204127900046e-08,
+      "loss": 0.0027,
+      "step": 19724
+    },
+    {
+      "epoch": 4.482954545454546,
+      "grad_norm": 0.40231977574114247,
+      "learning_rate": 3.272069882838534e-08,
+      "loss": 0.0016,
+      "step": 19725
+    },
+    {
+      "epoch": 4.483181818181818,
+      "grad_norm": 0.7953709793340727,
+      "learning_rate": 3.269220560647893e-08,
+      "loss": 0.0027,
+      "step": 19726
+    },
+    {
+      "epoch": 4.483409090909091,
+      "grad_norm": 0.6726003615969915,
+      "learning_rate": 3.26637244627618e-08,
+      "loss": 0.0023,
+      "step": 19727
+    },
+    {
+      "epoch": 4.483636363636363,
+      "grad_norm": 0.2862484227306337,
+      "learning_rate": 3.263525539781473e-08,
+      "loss": 0.0006,
+      "step": 19728
+    },
+    {
+      "epoch": 4.483863636363637,
+      "grad_norm": 0.22109626395345255,
+      "learning_rate": 3.2606798412218104e-08,
+      "loss": 0.0006,
+      "step": 19729
+    },
+    {
+      "epoch": 4.484090909090909,
+      "grad_norm": 0.3009936041950953,
+      "learning_rate": 3.2578353506552416e-08,
+      "loss": 0.0016,
+      "step": 19730
+    },
+    {
+      "epoch": 4.484318181818182,
+      "grad_norm": 1.4637608170273961,
+      "learning_rate": 3.2549920681397704e-08,
+      "loss": 0.0086,
+      "step": 19731
+    },
+    {
+      "epoch": 4.484545454545454,
+      "grad_norm": 0.9017269679006605,
+      "learning_rate": 3.252149993733363e-08,
+      "loss": 0.0033,
+      "step": 19732
+    },
+    {
+      "epoch": 4.484772727272727,
+      "grad_norm": 2.115641226544033,
+      "learning_rate": 3.2493091274939865e-08,
+      "loss": 0.0069,
+      "step": 19733
+    },
+    {
+      "epoch": 4.485,
+      "grad_norm": 0.35653449245772384,
+      "learning_rate": 3.246469469479562e-08,
+      "loss": 0.0019,
+      "step": 19734
+    },
+    {
+      "epoch": 4.485227272727273,
+      "grad_norm": 0.6040538841824695,
+      "learning_rate": 3.243631019747999e-08,
+      "loss": 0.0022,
+      "step": 19735
+    },
+    {
+      "epoch": 4.485454545454545,
+      "grad_norm": 0.9898576352304169,
+      "learning_rate": 3.240793778357176e-08,
+      "loss": 0.0024,
+      "step": 19736
+    },
+    {
+      "epoch": 4.485681818181818,
+      "grad_norm": 1.0753426812624054,
+      "learning_rate": 3.237957745364964e-08,
+      "loss": 0.0068,
+      "step": 19737
+    },
+    {
+      "epoch": 4.485909090909091,
+      "grad_norm": 1.005572486607982,
+      "learning_rate": 3.23512292082917e-08,
+      "loss": 0.009,
+      "step": 19738
+    },
+    {
+      "epoch": 4.486136363636364,
+      "grad_norm": 0.6203746927081513,
+      "learning_rate": 3.2322893048076095e-08,
+      "loss": 0.0022,
+      "step": 19739
+    },
+    {
+      "epoch": 4.486363636363636,
+      "grad_norm": 0.8726093999909275,
+      "learning_rate": 3.229456897358069e-08,
+      "loss": 0.0039,
+      "step": 19740
+    },
+    {
+      "epoch": 4.486590909090909,
+      "grad_norm": 0.2938223763566416,
+      "learning_rate": 3.2266256985383214e-08,
+      "loss": 0.0008,
+      "step": 19741
+    },
+    {
+      "epoch": 4.486818181818181,
+      "grad_norm": 3.137316115694326,
+      "learning_rate": 3.2237957084060645e-08,
+      "loss": 0.017,
+      "step": 19742
+    },
+    {
+      "epoch": 4.487045454545455,
+      "grad_norm": 1.1847852691325218,
+      "learning_rate": 3.220966927019037e-08,
+      "loss": 0.0097,
+      "step": 19743
+    },
+    {
+      "epoch": 4.487272727272727,
+      "grad_norm": 0.5714835388742223,
+      "learning_rate": 3.218139354434907e-08,
+      "loss": 0.0017,
+      "step": 19744
+    },
+    {
+      "epoch": 4.4875,
+      "grad_norm": 1.294485108337363,
+      "learning_rate": 3.215312990711337e-08,
+      "loss": 0.0036,
+      "step": 19745
+    },
+    {
+      "epoch": 4.487727272727272,
+      "grad_norm": 0.7634524269162678,
+      "learning_rate": 3.212487835905963e-08,
+      "loss": 0.0014,
+      "step": 19746
+    },
+    {
+      "epoch": 4.487954545454546,
+      "grad_norm": 1.008536895695256,
+      "learning_rate": 3.209663890076396e-08,
+      "loss": 0.0132,
+      "step": 19747
+    },
+    {
+      "epoch": 4.488181818181818,
+      "grad_norm": 0.779603071251391,
+      "learning_rate": 3.206841153280217e-08,
+      "loss": 0.0027,
+      "step": 19748
+    },
+    {
+      "epoch": 4.488409090909091,
+      "grad_norm": 0.7759727500439129,
+      "learning_rate": 3.20401962557499e-08,
+      "loss": 0.005,
+      "step": 19749
+    },
+    {
+      "epoch": 4.488636363636363,
+      "grad_norm": 1.0184440840283862,
+      "learning_rate": 3.201199307018267e-08,
+      "loss": 0.0095,
+      "step": 19750
+    },
+    {
+      "epoch": 4.488863636363637,
+      "grad_norm": 0.7148510802347302,
+      "learning_rate": 3.1983801976675294e-08,
+      "loss": 0.0016,
+      "step": 19751
+    },
+    {
+      "epoch": 4.489090909090909,
+      "grad_norm": 0.6907200586677626,
+      "learning_rate": 3.1955622975802724e-08,
+      "loss": 0.006,
+      "step": 19752
+    },
+    {
+      "epoch": 4.489318181818182,
+      "grad_norm": 1.455765036651624,
+      "learning_rate": 3.192745606813964e-08,
+      "loss": 0.0022,
+      "step": 19753
+    },
+    {
+      "epoch": 4.489545454545454,
+      "grad_norm": 3.536261919671114,
+      "learning_rate": 3.1899301254260524e-08,
+      "loss": 0.0195,
+      "step": 19754
+    },
+    {
+      "epoch": 4.489772727272728,
+      "grad_norm": 0.8691157730521287,
+      "learning_rate": 3.187115853473928e-08,
+      "loss": 0.0042,
+      "step": 19755
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 0.4380246907542866,
+      "learning_rate": 3.184302791014983e-08,
+      "loss": 0.0013,
+      "step": 19756
+    },
+    {
+      "epoch": 4.490227272727273,
+      "grad_norm": 0.5630858430292732,
+      "learning_rate": 3.1814909381065947e-08,
+      "loss": 0.0039,
+      "step": 19757
+    },
+    {
+      "epoch": 4.490454545454545,
+      "grad_norm": 0.62062695394276,
+      "learning_rate": 3.178680294806086e-08,
+      "loss": 0.0026,
+      "step": 19758
+    },
+    {
+      "epoch": 4.490681818181818,
+      "grad_norm": 3.099947964186087,
+      "learning_rate": 3.175870861170778e-08,
+      "loss": 0.0128,
+      "step": 19759
+    },
+    {
+      "epoch": 4.490909090909091,
+      "grad_norm": 0.7868439902480834,
+      "learning_rate": 3.173062637257966e-08,
+      "loss": 0.0043,
+      "step": 19760
+    },
+    {
+      "epoch": 4.491136363636364,
+      "grad_norm": 0.8258760349306511,
+      "learning_rate": 3.170255623124903e-08,
+      "loss": 0.0031,
+      "step": 19761
+    },
+    {
+      "epoch": 4.491363636363636,
+      "grad_norm": 2.056346301644469,
+      "learning_rate": 3.167449818828834e-08,
+      "loss": 0.0151,
+      "step": 19762
+    },
+    {
+      "epoch": 4.491590909090909,
+      "grad_norm": 1.615782137868462,
+      "learning_rate": 3.1646452244269846e-08,
+      "loss": 0.0068,
+      "step": 19763
+    },
+    {
+      "epoch": 4.491818181818182,
+      "grad_norm": 7.645623517428731,
+      "learning_rate": 3.1618418399765246e-08,
+      "loss": 0.0128,
+      "step": 19764
+    },
+    {
+      "epoch": 4.492045454545455,
+      "grad_norm": 1.5755530822121395,
+      "learning_rate": 3.159039665534637e-08,
+      "loss": 0.0044,
+      "step": 19765
+    },
+    {
+      "epoch": 4.492272727272727,
+      "grad_norm": 0.5725261681060765,
+      "learning_rate": 3.1562387011584495e-08,
+      "loss": 0.0018,
+      "step": 19766
+    },
+    {
+      "epoch": 4.4925,
+      "grad_norm": 0.4872695085726364,
+      "learning_rate": 3.1534389469050905e-08,
+      "loss": 0.0008,
+      "step": 19767
+    },
+    {
+      "epoch": 4.492727272727273,
+      "grad_norm": 0.4649460147409852,
+      "learning_rate": 3.150640402831654e-08,
+      "loss": 0.0015,
+      "step": 19768
+    },
+    {
+      "epoch": 4.492954545454546,
+      "grad_norm": 3.9111845471331326,
+      "learning_rate": 3.147843068995189e-08,
+      "loss": 0.0129,
+      "step": 19769
+    },
+    {
+      "epoch": 4.493181818181818,
+      "grad_norm": 1.380168991112812,
+      "learning_rate": 3.145046945452757e-08,
+      "loss": 0.0071,
+      "step": 19770
+    },
+    {
+      "epoch": 4.493409090909091,
+      "grad_norm": 1.3009418804743333,
+      "learning_rate": 3.142252032261359e-08,
+      "loss": 0.0105,
+      "step": 19771
+    },
+    {
+      "epoch": 4.493636363636364,
+      "grad_norm": 0.5849500380421232,
+      "learning_rate": 3.1394583294780125e-08,
+      "loss": 0.0013,
+      "step": 19772
+    },
+    {
+      "epoch": 4.493863636363637,
+      "grad_norm": 2.4890691883829543,
+      "learning_rate": 3.136665837159658e-08,
+      "loss": 0.0101,
+      "step": 19773
+    },
+    {
+      "epoch": 4.494090909090909,
+      "grad_norm": 1.3035996544375794,
+      "learning_rate": 3.13387455536325e-08,
+      "loss": 0.003,
+      "step": 19774
+    },
+    {
+      "epoch": 4.494318181818182,
+      "grad_norm": 0.8475524534382132,
+      "learning_rate": 3.131084484145716e-08,
+      "loss": 0.0039,
+      "step": 19775
+    },
+    {
+      "epoch": 4.494545454545454,
+      "grad_norm": 0.1703987613557201,
+      "learning_rate": 3.12829562356394e-08,
+      "loss": 0.0005,
+      "step": 19776
+    },
+    {
+      "epoch": 4.494772727272728,
+      "grad_norm": 1.5903588596414906,
+      "learning_rate": 3.1255079736748006e-08,
+      "loss": 0.0223,
+      "step": 19777
+    },
+    {
+      "epoch": 4.495,
+      "grad_norm": 1.4266628066725147,
+      "learning_rate": 3.122721534535134e-08,
+      "loss": 0.0053,
+      "step": 19778
+    },
+    {
+      "epoch": 4.495227272727273,
+      "grad_norm": 0.23251732533790068,
+      "learning_rate": 3.119936306201762e-08,
+      "loss": 0.0007,
+      "step": 19779
+    },
+    {
+      "epoch": 4.495454545454545,
+      "grad_norm": 0.8447945598277933,
+      "learning_rate": 3.1171522887314885e-08,
+      "loss": 0.0062,
+      "step": 19780
+    },
+    {
+      "epoch": 4.495681818181819,
+      "grad_norm": 0.7106115467248537,
+      "learning_rate": 3.114369482181086e-08,
+      "loss": 0.0039,
+      "step": 19781
+    },
+    {
+      "epoch": 4.495909090909091,
+      "grad_norm": 2.5309226620869563,
+      "learning_rate": 3.111587886607287e-08,
+      "loss": 0.0121,
+      "step": 19782
+    },
+    {
+      "epoch": 4.496136363636364,
+      "grad_norm": 1.452450069385135,
+      "learning_rate": 3.1088075020668174e-08,
+      "loss": 0.0059,
+      "step": 19783
+    },
+    {
+      "epoch": 4.496363636363636,
+      "grad_norm": 0.23224392156090023,
+      "learning_rate": 3.1060283286163744e-08,
+      "loss": 0.0007,
+      "step": 19784
+    },
+    {
+      "epoch": 4.496590909090909,
+      "grad_norm": 1.922550525483337,
+      "learning_rate": 3.103250366312642e-08,
+      "loss": 0.0034,
+      "step": 19785
+    },
+    {
+      "epoch": 4.496818181818182,
+      "grad_norm": 2.728396072186387,
+      "learning_rate": 3.1004736152122486e-08,
+      "loss": 0.0101,
+      "step": 19786
+    },
+    {
+      "epoch": 4.497045454545455,
+      "grad_norm": 0.7533217630691078,
+      "learning_rate": 3.09769807537183e-08,
+      "loss": 0.004,
+      "step": 19787
+    },
+    {
+      "epoch": 4.497272727272727,
+      "grad_norm": 1.5700219592718738,
+      "learning_rate": 3.094923746847979e-08,
+      "loss": 0.0067,
+      "step": 19788
+    },
+    {
+      "epoch": 4.4975,
+      "grad_norm": 0.8073235352009462,
+      "learning_rate": 3.092150629697269e-08,
+      "loss": 0.0014,
+      "step": 19789
+    },
+    {
+      "epoch": 4.497727272727273,
+      "grad_norm": 1.911366605222308,
+      "learning_rate": 3.089378723976251e-08,
+      "loss": 0.0046,
+      "step": 19790
+    },
+    {
+      "epoch": 4.4979545454545455,
+      "grad_norm": 0.9423762757362053,
+      "learning_rate": 3.08660802974145e-08,
+      "loss": 0.0026,
+      "step": 19791
+    },
+    {
+      "epoch": 4.498181818181818,
+      "grad_norm": 1.289545263104169,
+      "learning_rate": 3.083838547049363e-08,
+      "loss": 0.0103,
+      "step": 19792
+    },
+    {
+      "epoch": 4.498409090909091,
+      "grad_norm": 0.7444386174564233,
+      "learning_rate": 3.081070275956465e-08,
+      "loss": 0.0045,
+      "step": 19793
+    },
+    {
+      "epoch": 4.498636363636364,
+      "grad_norm": 1.4308998470201555,
+      "learning_rate": 3.078303216519217e-08,
+      "loss": 0.0086,
+      "step": 19794
+    },
+    {
+      "epoch": 4.4988636363636365,
+      "grad_norm": 1.6439539464364874,
+      "learning_rate": 3.07553736879402e-08,
+      "loss": 0.0061,
+      "step": 19795
+    },
+    {
+      "epoch": 4.499090909090909,
+      "grad_norm": 1.9121695531023386,
+      "learning_rate": 3.0727727328372927e-08,
+      "loss": 0.0044,
+      "step": 19796
+    },
+    {
+      "epoch": 4.4993181818181816,
+      "grad_norm": 0.4885746476523435,
+      "learning_rate": 3.0700093087054e-08,
+      "loss": 0.0015,
+      "step": 19797
+    },
+    {
+      "epoch": 4.499545454545455,
+      "grad_norm": 1.098735612835077,
+      "learning_rate": 3.067247096454708e-08,
+      "loss": 0.0053,
+      "step": 19798
+    },
+    {
+      "epoch": 4.4997727272727275,
+      "grad_norm": 1.5875649083220562,
+      "learning_rate": 3.0644860961415306e-08,
+      "loss": 0.0061,
+      "step": 19799
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 1.42367554609349,
+      "learning_rate": 3.0617263078221716e-08,
+      "loss": 0.0093,
+      "step": 19800
+    },
+    {
+      "epoch": 4.5002272727272725,
+      "grad_norm": 0.5615540113349052,
+      "learning_rate": 3.058967731552905e-08,
+      "loss": 0.003,
+      "step": 19801
+    },
+    {
+      "epoch": 4.500454545454545,
+      "grad_norm": 0.22608428231747876,
+      "learning_rate": 3.0562103673899905e-08,
+      "loss": 0.0005,
+      "step": 19802
+    },
+    {
+      "epoch": 4.5006818181818184,
+      "grad_norm": 0.4158500634238999,
+      "learning_rate": 3.053454215389655e-08,
+      "loss": 0.002,
+      "step": 19803
+    },
+    {
+      "epoch": 4.500909090909091,
+      "grad_norm": 1.1164582330961226,
+      "learning_rate": 3.050699275608096e-08,
+      "loss": 0.0032,
+      "step": 19804
+    },
+    {
+      "epoch": 4.5011363636363635,
+      "grad_norm": 1.4830962636897367,
+      "learning_rate": 3.047945548101491e-08,
+      "loss": 0.007,
+      "step": 19805
+    },
+    {
+      "epoch": 4.501363636363636,
+      "grad_norm": 1.4101932713759704,
+      "learning_rate": 3.045193032925997e-08,
+      "loss": 0.0074,
+      "step": 19806
+    },
+    {
+      "epoch": 4.501590909090909,
+      "grad_norm": 2.537719165244629,
+      "learning_rate": 3.042441730137742e-08,
+      "loss": 0.0103,
+      "step": 19807
+    },
+    {
+      "epoch": 4.501818181818182,
+      "grad_norm": 1.2007003337478126,
+      "learning_rate": 3.0396916397928274e-08,
+      "loss": 0.0022,
+      "step": 19808
+    },
+    {
+      "epoch": 4.5020454545454545,
+      "grad_norm": 1.0043291970535217,
+      "learning_rate": 3.036942761947333e-08,
+      "loss": 0.0079,
+      "step": 19809
+    },
+    {
+      "epoch": 4.502272727272727,
+      "grad_norm": 0.9966944668212082,
+      "learning_rate": 3.034195096657318e-08,
+      "loss": 0.01,
+      "step": 19810
+    },
+    {
+      "epoch": 4.5024999999999995,
+      "grad_norm": 2.161685592673502,
+      "learning_rate": 3.031448643978807e-08,
+      "loss": 0.0059,
+      "step": 19811
+    },
+    {
+      "epoch": 4.502727272727273,
+      "grad_norm": 1.0532291738814632,
+      "learning_rate": 3.028703403967818e-08,
+      "loss": 0.0058,
+      "step": 19812
+    },
+    {
+      "epoch": 4.502954545454545,
+      "grad_norm": 0.7393209792526357,
+      "learning_rate": 3.025959376680314e-08,
+      "loss": 0.0052,
+      "step": 19813
+    },
+    {
+      "epoch": 4.503181818181818,
+      "grad_norm": 1.6023924599653028,
+      "learning_rate": 3.0232165621722486e-08,
+      "loss": 0.0048,
+      "step": 19814
+    },
+    {
+      "epoch": 4.503409090909091,
+      "grad_norm": 1.2682474942791777,
+      "learning_rate": 3.020474960499564e-08,
+      "loss": 0.0084,
+      "step": 19815
+    },
+    {
+      "epoch": 4.503636363636364,
+      "grad_norm": 1.564914250768411,
+      "learning_rate": 3.0177345717181745e-08,
+      "loss": 0.0035,
+      "step": 19816
+    },
+    {
+      "epoch": 4.503863636363636,
+      "grad_norm": 0.4631730010330217,
+      "learning_rate": 3.014995395883937e-08,
+      "loss": 0.0011,
+      "step": 19817
+    },
+    {
+      "epoch": 4.504090909090909,
+      "grad_norm": 1.6467281407021221,
+      "learning_rate": 3.0122574330527166e-08,
+      "loss": 0.0037,
+      "step": 19818
+    },
+    {
+      "epoch": 4.504318181818181,
+      "grad_norm": 1.1424701230064767,
+      "learning_rate": 3.009520683280358e-08,
+      "loss": 0.0052,
+      "step": 19819
+    },
+    {
+      "epoch": 4.504545454545455,
+      "grad_norm": 0.9996934186367843,
+      "learning_rate": 3.0067851466226565e-08,
+      "loss": 0.0029,
+      "step": 19820
+    },
+    {
+      "epoch": 4.504772727272727,
+      "grad_norm": 0.8423557540809252,
+      "learning_rate": 3.0040508231353946e-08,
+      "loss": 0.0033,
+      "step": 19821
+    },
+    {
+      "epoch": 4.505,
+      "grad_norm": 1.3782774919762062,
+      "learning_rate": 3.001317712874333e-08,
+      "loss": 0.0065,
+      "step": 19822
+    },
+    {
+      "epoch": 4.505227272727272,
+      "grad_norm": 0.8004219216452219,
+      "learning_rate": 2.998585815895211e-08,
+      "loss": 0.0022,
+      "step": 19823
+    },
+    {
+      "epoch": 4.505454545454546,
+      "grad_norm": 1.3537884095797117,
+      "learning_rate": 2.9958551322537206e-08,
+      "loss": 0.0079,
+      "step": 19824
+    },
+    {
+      "epoch": 4.505681818181818,
+      "grad_norm": 0.7072996756897524,
+      "learning_rate": 2.993125662005568e-08,
+      "loss": 0.0046,
+      "step": 19825
+    },
+    {
+      "epoch": 4.505909090909091,
+      "grad_norm": 1.100590777114988,
+      "learning_rate": 2.990397405206395e-08,
+      "loss": 0.0037,
+      "step": 19826
+    },
+    {
+      "epoch": 4.506136363636363,
+      "grad_norm": 0.8182122542261495,
+      "learning_rate": 2.9876703619118316e-08,
+      "loss": 0.0031,
+      "step": 19827
+    },
+    {
+      "epoch": 4.506363636363636,
+      "grad_norm": 0.7444383601036145,
+      "learning_rate": 2.984944532177493e-08,
+      "loss": 0.0037,
+      "step": 19828
+    },
+    {
+      "epoch": 4.506590909090909,
+      "grad_norm": 0.527612734474712,
+      "learning_rate": 2.98221991605898e-08,
+      "loss": 0.0022,
+      "step": 19829
+    },
+    {
+      "epoch": 4.506818181818182,
+      "grad_norm": 0.4065764333132421,
+      "learning_rate": 2.979496513611825e-08,
+      "loss": 0.0011,
+      "step": 19830
+    },
+    {
+      "epoch": 4.507045454545454,
+      "grad_norm": 2.3532389444109567,
+      "learning_rate": 2.9767743248915805e-08,
+      "loss": 0.0177,
+      "step": 19831
+    },
+    {
+      "epoch": 4.507272727272728,
+      "grad_norm": 0.5333064301219504,
+      "learning_rate": 2.974053349953744e-08,
+      "loss": 0.0018,
+      "step": 19832
+    },
+    {
+      "epoch": 4.5075,
+      "grad_norm": 0.6583374095861855,
+      "learning_rate": 2.9713335888538128e-08,
+      "loss": 0.0028,
+      "step": 19833
+    },
+    {
+      "epoch": 4.507727272727273,
+      "grad_norm": 0.9310864841848366,
+      "learning_rate": 2.968615041647256e-08,
+      "loss": 0.0034,
+      "step": 19834
+    },
+    {
+      "epoch": 4.507954545454545,
+      "grad_norm": 0.5542688540359091,
+      "learning_rate": 2.9658977083894812e-08,
+      "loss": 0.0014,
+      "step": 19835
+    },
+    {
+      "epoch": 4.508181818181818,
+      "grad_norm": 0.4872112644104225,
+      "learning_rate": 2.963181589135916e-08,
+      "loss": 0.004,
+      "step": 19836
+    },
+    {
+      "epoch": 4.508409090909091,
+      "grad_norm": 1.3174977860965154,
+      "learning_rate": 2.960466683941947e-08,
+      "loss": 0.0045,
+      "step": 19837
+    },
+    {
+      "epoch": 4.508636363636364,
+      "grad_norm": 1.0461417779194506,
+      "learning_rate": 2.957752992862932e-08,
+      "loss": 0.0033,
+      "step": 19838
+    },
+    {
+      "epoch": 4.508863636363636,
+      "grad_norm": 0.8680433590027327,
+      "learning_rate": 2.9550405159542163e-08,
+      "loss": 0.0035,
+      "step": 19839
+    },
+    {
+      "epoch": 4.509090909090909,
+      "grad_norm": 0.5305737038472209,
+      "learning_rate": 2.9523292532711024e-08,
+      "loss": 0.0027,
+      "step": 19840
+    },
+    {
+      "epoch": 4.509318181818182,
+      "grad_norm": 0.8375409062165048,
+      "learning_rate": 2.9496192048688867e-08,
+      "loss": 0.0043,
+      "step": 19841
+    },
+    {
+      "epoch": 4.509545454545455,
+      "grad_norm": 1.1518389802864364,
+      "learning_rate": 2.9469103708028303e-08,
+      "loss": 0.0031,
+      "step": 19842
+    },
+    {
+      "epoch": 4.509772727272727,
+      "grad_norm": 0.6038156058841687,
+      "learning_rate": 2.94420275112816e-08,
+      "loss": 0.0021,
+      "step": 19843
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 3.7289653151148654,
+      "learning_rate": 2.9414963459000954e-08,
+      "loss": 0.0039,
+      "step": 19844
+    },
+    {
+      "epoch": 4.510227272727272,
+      "grad_norm": 0.5924106190935818,
+      "learning_rate": 2.9387911551738286e-08,
+      "loss": 0.0023,
+      "step": 19845
+    },
+    {
+      "epoch": 4.510454545454546,
+      "grad_norm": 1.7193502545397852,
+      "learning_rate": 2.936087179004517e-08,
+      "loss": 0.0075,
+      "step": 19846
+    },
+    {
+      "epoch": 4.510681818181818,
+      "grad_norm": 0.4440394492885277,
+      "learning_rate": 2.9333844174473104e-08,
+      "loss": 0.0018,
+      "step": 19847
+    },
+    {
+      "epoch": 4.510909090909091,
+      "grad_norm": 1.2296130038175899,
+      "learning_rate": 2.930682870557311e-08,
+      "loss": 0.0054,
+      "step": 19848
+    },
+    {
+      "epoch": 4.511136363636363,
+      "grad_norm": 0.6319617369446479,
+      "learning_rate": 2.9279825383896066e-08,
+      "loss": 0.0026,
+      "step": 19849
+    },
+    {
+      "epoch": 4.511363636363637,
+      "grad_norm": 1.3884967229320224,
+      "learning_rate": 2.9252834209992713e-08,
+      "loss": 0.006,
+      "step": 19850
+    },
+    {
+      "epoch": 4.511590909090909,
+      "grad_norm": 0.1275670632501594,
+      "learning_rate": 2.922585518441344e-08,
+      "loss": 0.0004,
+      "step": 19851
+    },
+    {
+      "epoch": 4.511818181818182,
+      "grad_norm": 5.28234381263797,
+      "learning_rate": 2.91988883077083e-08,
+      "loss": 0.0224,
+      "step": 19852
+    },
+    {
+      "epoch": 4.512045454545454,
+      "grad_norm": 1.0152949077689937,
+      "learning_rate": 2.9171933580427332e-08,
+      "loss": 0.0026,
+      "step": 19853
+    },
+    {
+      "epoch": 4.512272727272728,
+      "grad_norm": 0.7733239509146067,
+      "learning_rate": 2.914499100312003e-08,
+      "loss": 0.0049,
+      "step": 19854
+    },
+    {
+      "epoch": 4.5125,
+      "grad_norm": 0.8649820290655499,
+      "learning_rate": 2.9118060576335953e-08,
+      "loss": 0.0027,
+      "step": 19855
+    },
+    {
+      "epoch": 4.512727272727273,
+      "grad_norm": 2.634885812646491,
+      "learning_rate": 2.909114230062425e-08,
+      "loss": 0.0164,
+      "step": 19856
+    },
+    {
+      "epoch": 4.512954545454545,
+      "grad_norm": 0.3238654363408137,
+      "learning_rate": 2.9064236176533714e-08,
+      "loss": 0.0013,
+      "step": 19857
+    },
+    {
+      "epoch": 4.513181818181819,
+      "grad_norm": 0.8090964623100879,
+      "learning_rate": 2.9037342204613074e-08,
+      "loss": 0.0021,
+      "step": 19858
+    },
+    {
+      "epoch": 4.513409090909091,
+      "grad_norm": 0.8092132901738497,
+      "learning_rate": 2.901046038541071e-08,
+      "loss": 0.0045,
+      "step": 19859
+    },
+    {
+      "epoch": 4.513636363636364,
+      "grad_norm": 1.5071165006134049,
+      "learning_rate": 2.898359071947493e-08,
+      "loss": 0.0047,
+      "step": 19860
+    },
+    {
+      "epoch": 4.513863636363636,
+      "grad_norm": 0.7427191642359573,
+      "learning_rate": 2.8956733207353427e-08,
+      "loss": 0.0029,
+      "step": 19861
+    },
+    {
+      "epoch": 4.514090909090909,
+      "grad_norm": 1.1489759370501382,
+      "learning_rate": 2.8929887849593953e-08,
+      "loss": 0.0029,
+      "step": 19862
+    },
+    {
+      "epoch": 4.514318181818182,
+      "grad_norm": 0.6529142389481837,
+      "learning_rate": 2.890305464674406e-08,
+      "loss": 0.0025,
+      "step": 19863
+    },
+    {
+      "epoch": 4.514545454545455,
+      "grad_norm": 1.1027826227862885,
+      "learning_rate": 2.8876233599350877e-08,
+      "loss": 0.0062,
+      "step": 19864
+    },
+    {
+      "epoch": 4.514772727272727,
+      "grad_norm": 0.25972167780442024,
+      "learning_rate": 2.884942470796119e-08,
+      "loss": 0.0009,
+      "step": 19865
+    },
+    {
+      "epoch": 4.515,
+      "grad_norm": 1.5937461049776231,
+      "learning_rate": 2.8822627973121785e-08,
+      "loss": 0.0048,
+      "step": 19866
+    },
+    {
+      "epoch": 4.515227272727273,
+      "grad_norm": 0.8159540129840782,
+      "learning_rate": 2.8795843395379096e-08,
+      "loss": 0.0037,
+      "step": 19867
+    },
+    {
+      "epoch": 4.515454545454546,
+      "grad_norm": 0.4203952369544212,
+      "learning_rate": 2.8769070975279215e-08,
+      "loss": 0.005,
+      "step": 19868
+    },
+    {
+      "epoch": 4.515681818181818,
+      "grad_norm": 0.751972916808237,
+      "learning_rate": 2.8742310713368306e-08,
+      "loss": 0.0039,
+      "step": 19869
+    },
+    {
+      "epoch": 4.515909090909091,
+      "grad_norm": 2.0396794919172856,
+      "learning_rate": 2.871556261019183e-08,
+      "loss": 0.0049,
+      "step": 19870
+    },
+    {
+      "epoch": 4.516136363636363,
+      "grad_norm": 0.4973665442635102,
+      "learning_rate": 2.8688826666295258e-08,
+      "loss": 0.0044,
+      "step": 19871
+    },
+    {
+      "epoch": 4.516363636363637,
+      "grad_norm": 0.4587935690864683,
+      "learning_rate": 2.8662102882223915e-08,
+      "loss": 0.002,
+      "step": 19872
+    },
+    {
+      "epoch": 4.516590909090909,
+      "grad_norm": 1.188980357998434,
+      "learning_rate": 2.8635391258522714e-08,
+      "loss": 0.0029,
+      "step": 19873
+    },
+    {
+      "epoch": 4.516818181818182,
+      "grad_norm": 1.571363978740501,
+      "learning_rate": 2.860869179573622e-08,
+      "loss": 0.0102,
+      "step": 19874
+    },
+    {
+      "epoch": 4.517045454545455,
+      "grad_norm": 0.6591213998287906,
+      "learning_rate": 2.8582004494408928e-08,
+      "loss": 0.0025,
+      "step": 19875
+    },
+    {
+      "epoch": 4.5172727272727276,
+      "grad_norm": 1.854574679261005,
+      "learning_rate": 2.855532935508512e-08,
+      "loss": 0.0202,
+      "step": 19876
+    },
+    {
+      "epoch": 4.5175,
+      "grad_norm": 0.5559983446513543,
+      "learning_rate": 2.8528666378308743e-08,
+      "loss": 0.0036,
+      "step": 19877
+    },
+    {
+      "epoch": 4.517727272727273,
+      "grad_norm": 0.6196274169389043,
+      "learning_rate": 2.8502015564623382e-08,
+      "loss": 0.0046,
+      "step": 19878
+    },
+    {
+      "epoch": 4.517954545454545,
+      "grad_norm": 1.3067962182112407,
+      "learning_rate": 2.8475376914572637e-08,
+      "loss": 0.0074,
+      "step": 19879
+    },
+    {
+      "epoch": 4.5181818181818185,
+      "grad_norm": 0.7987407715943255,
+      "learning_rate": 2.8448750428699616e-08,
+      "loss": 0.0057,
+      "step": 19880
+    },
+    {
+      "epoch": 4.518409090909091,
+      "grad_norm": 0.9953819073611256,
+      "learning_rate": 2.842213610754728e-08,
+      "loss": 0.0154,
+      "step": 19881
+    },
+    {
+      "epoch": 4.518636363636364,
+      "grad_norm": 0.5838412391270474,
+      "learning_rate": 2.8395533951658392e-08,
+      "loss": 0.0017,
+      "step": 19882
+    },
+    {
+      "epoch": 4.518863636363636,
+      "grad_norm": 1.1781327235472967,
+      "learning_rate": 2.8368943961575442e-08,
+      "loss": 0.005,
+      "step": 19883
+    },
+    {
+      "epoch": 4.5190909090909095,
+      "grad_norm": 1.3887564615643315,
+      "learning_rate": 2.8342366137840554e-08,
+      "loss": 0.006,
+      "step": 19884
+    },
+    {
+      "epoch": 4.519318181818182,
+      "grad_norm": 2.7854841130312775,
+      "learning_rate": 2.8315800480995735e-08,
+      "loss": 0.0115,
+      "step": 19885
+    },
+    {
+      "epoch": 4.5195454545454545,
+      "grad_norm": 1.4785187843384835,
+      "learning_rate": 2.8289246991582837e-08,
+      "loss": 0.008,
+      "step": 19886
+    },
+    {
+      "epoch": 4.519772727272727,
+      "grad_norm": 0.5375174489685585,
+      "learning_rate": 2.8262705670143095e-08,
+      "loss": 0.0023,
+      "step": 19887
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 1.0209722602797284,
+      "learning_rate": 2.8236176517217884e-08,
+      "loss": 0.0029,
+      "step": 19888
+    },
+    {
+      "epoch": 4.520227272727273,
+      "grad_norm": 0.6833209103548846,
+      "learning_rate": 2.820965953334816e-08,
+      "loss": 0.0023,
+      "step": 19889
+    },
+    {
+      "epoch": 4.5204545454545455,
+      "grad_norm": 1.3824268002798605,
+      "learning_rate": 2.8183154719074596e-08,
+      "loss": 0.0122,
+      "step": 19890
+    },
+    {
+      "epoch": 4.520681818181818,
+      "grad_norm": 0.2930015360256783,
+      "learning_rate": 2.8156662074937806e-08,
+      "loss": 0.0018,
+      "step": 19891
+    },
+    {
+      "epoch": 4.5209090909090905,
+      "grad_norm": 0.8805888636218304,
+      "learning_rate": 2.8130181601477847e-08,
+      "loss": 0.0023,
+      "step": 19892
+    },
+    {
+      "epoch": 4.521136363636364,
+      "grad_norm": 0.5719411223762777,
+      "learning_rate": 2.8103713299234766e-08,
+      "loss": 0.0025,
+      "step": 19893
+    },
+    {
+      "epoch": 4.5213636363636365,
+      "grad_norm": 0.9032495762588448,
+      "learning_rate": 2.8077257168748345e-08,
+      "loss": 0.0035,
+      "step": 19894
+    },
+    {
+      "epoch": 4.521590909090909,
+      "grad_norm": 1.492453189143574,
+      "learning_rate": 2.805081321055815e-08,
+      "loss": 0.0038,
+      "step": 19895
+    },
+    {
+      "epoch": 4.5218181818181815,
+      "grad_norm": 0.8063555152934554,
+      "learning_rate": 2.8024381425203127e-08,
+      "loss": 0.0022,
+      "step": 19896
+    },
+    {
+      "epoch": 4.522045454545455,
+      "grad_norm": 1.4318184054782346,
+      "learning_rate": 2.7997961813222424e-08,
+      "loss": 0.0076,
+      "step": 19897
+    },
+    {
+      "epoch": 4.522272727272727,
+      "grad_norm": 0.42433574307321276,
+      "learning_rate": 2.797155437515492e-08,
+      "loss": 0.0008,
+      "step": 19898
+    },
+    {
+      "epoch": 4.5225,
+      "grad_norm": 1.2891405214419518,
+      "learning_rate": 2.7945159111539e-08,
+      "loss": 0.0071,
+      "step": 19899
+    },
+    {
+      "epoch": 4.5227272727272725,
+      "grad_norm": 1.3490476443421573,
+      "learning_rate": 2.7918776022912918e-08,
+      "loss": 0.0035,
+      "step": 19900
+    },
+    {
+      "epoch": 4.522954545454546,
+      "grad_norm": 0.4874868344956792,
+      "learning_rate": 2.7892405109814572e-08,
+      "loss": 0.0014,
+      "step": 19901
+    },
+    {
+      "epoch": 4.523181818181818,
+      "grad_norm": 2.1046217591089276,
+      "learning_rate": 2.7866046372781868e-08,
+      "loss": 0.0084,
+      "step": 19902
+    },
+    {
+      "epoch": 4.523409090909091,
+      "grad_norm": 0.8050384399662746,
+      "learning_rate": 2.7839699812352156e-08,
+      "loss": 0.0056,
+      "step": 19903
+    },
+    {
+      "epoch": 4.5236363636363635,
+      "grad_norm": 0.4169179145802148,
+      "learning_rate": 2.7813365429062918e-08,
+      "loss": 0.0016,
+      "step": 19904
+    },
+    {
+      "epoch": 4.523863636363636,
+      "grad_norm": 0.865441522399283,
+      "learning_rate": 2.778704322345088e-08,
+      "loss": 0.0028,
+      "step": 19905
+    },
+    {
+      "epoch": 4.524090909090909,
+      "grad_norm": 1.980335942691865,
+      "learning_rate": 2.7760733196052976e-08,
+      "loss": 0.0108,
+      "step": 19906
+    },
+    {
+      "epoch": 4.524318181818182,
+      "grad_norm": 1.3622492846677958,
+      "learning_rate": 2.7734435347405648e-08,
+      "loss": 0.0056,
+      "step": 19907
+    },
+    {
+      "epoch": 4.524545454545454,
+      "grad_norm": 0.7687507085600713,
+      "learning_rate": 2.7708149678045206e-08,
+      "loss": 0.0012,
+      "step": 19908
+    },
+    {
+      "epoch": 4.524772727272727,
+      "grad_norm": 0.8218988975188288,
+      "learning_rate": 2.768187618850761e-08,
+      "loss": 0.0037,
+      "step": 19909
+    },
+    {
+      "epoch": 4.525,
+      "grad_norm": 1.0516371205678448,
+      "learning_rate": 2.765561487932861e-08,
+      "loss": 0.0063,
+      "step": 19910
+    },
+    {
+      "epoch": 4.525227272727273,
+      "grad_norm": 1.3111844042521417,
+      "learning_rate": 2.7629365751043676e-08,
+      "loss": 0.016,
+      "step": 19911
+    },
+    {
+      "epoch": 4.525454545454545,
+      "grad_norm": 1.4394629064484286,
+      "learning_rate": 2.7603128804188224e-08,
+      "loss": 0.005,
+      "step": 19912
+    },
+    {
+      "epoch": 4.525681818181818,
+      "grad_norm": 4.00150124925247,
+      "learning_rate": 2.7576904039297166e-08,
+      "loss": 0.0333,
+      "step": 19913
+    },
+    {
+      "epoch": 4.52590909090909,
+      "grad_norm": 0.3538524272056693,
+      "learning_rate": 2.755069145690535e-08,
+      "loss": 0.001,
+      "step": 19914
+    },
+    {
+      "epoch": 4.526136363636364,
+      "grad_norm": 1.0343344900602855,
+      "learning_rate": 2.7524491057547216e-08,
+      "loss": 0.0056,
+      "step": 19915
+    },
+    {
+      "epoch": 4.526363636363636,
+      "grad_norm": 0.3430573059179899,
+      "learning_rate": 2.7498302841756987e-08,
+      "loss": 0.0007,
+      "step": 19916
+    },
+    {
+      "epoch": 4.526590909090909,
+      "grad_norm": 0.5071964860654213,
+      "learning_rate": 2.747212681006896e-08,
+      "loss": 0.0027,
+      "step": 19917
+    },
+    {
+      "epoch": 4.526818181818182,
+      "grad_norm": 7.556375924760961,
+      "learning_rate": 2.744596296301652e-08,
+      "loss": 0.0067,
+      "step": 19918
+    },
+    {
+      "epoch": 4.527045454545455,
+      "grad_norm": 0.849306155693216,
+      "learning_rate": 2.7419811301133483e-08,
+      "loss": 0.0065,
+      "step": 19919
+    },
+    {
+      "epoch": 4.527272727272727,
+      "grad_norm": 0.7287115568940568,
+      "learning_rate": 2.7393671824952964e-08,
+      "loss": 0.0016,
+      "step": 19920
+    },
+    {
+      "epoch": 4.5275,
+      "grad_norm": 0.46612310288064673,
+      "learning_rate": 2.7367544535008146e-08,
+      "loss": 0.0018,
+      "step": 19921
+    },
+    {
+      "epoch": 4.527727272727272,
+      "grad_norm": 2.8266073198748356,
+      "learning_rate": 2.7341429431831658e-08,
+      "loss": 0.0106,
+      "step": 19922
+    },
+    {
+      "epoch": 4.527954545454546,
+      "grad_norm": 0.6724743481102438,
+      "learning_rate": 2.731532651595613e-08,
+      "loss": 0.0022,
+      "step": 19923
+    },
+    {
+      "epoch": 4.528181818181818,
+      "grad_norm": 3.4747207286030974,
+      "learning_rate": 2.7289235787913772e-08,
+      "loss": 0.0768,
+      "step": 19924
+    },
+    {
+      "epoch": 4.528409090909091,
+      "grad_norm": 0.8066872130037379,
+      "learning_rate": 2.7263157248236665e-08,
+      "loss": 0.0055,
+      "step": 19925
+    },
+    {
+      "epoch": 4.528636363636363,
+      "grad_norm": 0.1725120877611145,
+      "learning_rate": 2.7237090897456534e-08,
+      "loss": 0.0006,
+      "step": 19926
+    },
+    {
+      "epoch": 4.528863636363637,
+      "grad_norm": 2.5448465067612966,
+      "learning_rate": 2.7211036736105036e-08,
+      "loss": 0.0424,
+      "step": 19927
+    },
+    {
+      "epoch": 4.529090909090909,
+      "grad_norm": 2.028250858036064,
+      "learning_rate": 2.718499476471341e-08,
+      "loss": 0.0089,
+      "step": 19928
+    },
+    {
+      "epoch": 4.529318181818182,
+      "grad_norm": 2.8575591538088427,
+      "learning_rate": 2.7158964983812695e-08,
+      "loss": 0.012,
+      "step": 19929
+    },
+    {
+      "epoch": 4.529545454545454,
+      "grad_norm": 0.8324778204817777,
+      "learning_rate": 2.713294739393371e-08,
+      "loss": 0.0021,
+      "step": 19930
+    },
+    {
+      "epoch": 4.529772727272727,
+      "grad_norm": 0.5321958942005631,
+      "learning_rate": 2.7106941995606874e-08,
+      "loss": 0.0016,
+      "step": 19931
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 0.6081203636510831,
+      "learning_rate": 2.7080948789362656e-08,
+      "loss": 0.0028,
+      "step": 19932
+    },
+    {
+      "epoch": 4.530227272727273,
+      "grad_norm": 0.6484004451863987,
+      "learning_rate": 2.705496777573091e-08,
+      "loss": 0.0057,
+      "step": 19933
+    },
+    {
+      "epoch": 4.530454545454545,
+      "grad_norm": 2.6999700765667094,
+      "learning_rate": 2.7028998955241637e-08,
+      "loss": 0.0134,
+      "step": 19934
+    },
+    {
+      "epoch": 4.530681818181818,
+      "grad_norm": 1.1452149660985784,
+      "learning_rate": 2.7003042328424337e-08,
+      "loss": 0.0034,
+      "step": 19935
+    },
+    {
+      "epoch": 4.530909090909091,
+      "grad_norm": 0.4092030127253345,
+      "learning_rate": 2.697709789580817e-08,
+      "loss": 0.0015,
+      "step": 19936
+    },
+    {
+      "epoch": 4.531136363636364,
+      "grad_norm": 0.5699009041392175,
+      "learning_rate": 2.69511656579223e-08,
+      "loss": 0.001,
+      "step": 19937
+    },
+    {
+      "epoch": 4.531363636363636,
+      "grad_norm": 1.3632858970689048,
+      "learning_rate": 2.692524561529554e-08,
+      "loss": 0.0104,
+      "step": 19938
+    },
+    {
+      "epoch": 4.531590909090909,
+      "grad_norm": 0.42948014602585965,
+      "learning_rate": 2.689933776845642e-08,
+      "loss": 0.0043,
+      "step": 19939
+    },
+    {
+      "epoch": 4.531818181818182,
+      "grad_norm": 2.0647727351713066,
+      "learning_rate": 2.6873442117933274e-08,
+      "loss": 0.0078,
+      "step": 19940
+    },
+    {
+      "epoch": 4.532045454545455,
+      "grad_norm": 0.8916561912814966,
+      "learning_rate": 2.6847558664254078e-08,
+      "loss": 0.006,
+      "step": 19941
+    },
+    {
+      "epoch": 4.532272727272727,
+      "grad_norm": 1.7767744581816158,
+      "learning_rate": 2.6821687407946674e-08,
+      "loss": 0.0061,
+      "step": 19942
+    },
+    {
+      "epoch": 4.5325,
+      "grad_norm": 0.8847802743185063,
+      "learning_rate": 2.67958283495387e-08,
+      "loss": 0.0036,
+      "step": 19943
+    },
+    {
+      "epoch": 4.532727272727273,
+      "grad_norm": 0.2448207215969205,
+      "learning_rate": 2.6769981489557367e-08,
+      "loss": 0.0007,
+      "step": 19944
+    },
+    {
+      "epoch": 4.532954545454546,
+      "grad_norm": 0.9691219742391703,
+      "learning_rate": 2.674414682852976e-08,
+      "loss": 0.0064,
+      "step": 19945
+    },
+    {
+      "epoch": 4.533181818181818,
+      "grad_norm": 0.519658152154364,
+      "learning_rate": 2.671832436698274e-08,
+      "loss": 0.0025,
+      "step": 19946
+    },
+    {
+      "epoch": 4.533409090909091,
+      "grad_norm": 1.1313171027389362,
+      "learning_rate": 2.6692514105442837e-08,
+      "loss": 0.005,
+      "step": 19947
+    },
+    {
+      "epoch": 4.533636363636363,
+      "grad_norm": 2.0552650257765075,
+      "learning_rate": 2.6666716044436505e-08,
+      "loss": 0.0084,
+      "step": 19948
+    },
+    {
+      "epoch": 4.533863636363637,
+      "grad_norm": 0.884237959301465,
+      "learning_rate": 2.664093018448957e-08,
+      "loss": 0.0058,
+      "step": 19949
+    },
+    {
+      "epoch": 4.534090909090909,
+      "grad_norm": 2.6969292277921086,
+      "learning_rate": 2.6615156526127935e-08,
+      "loss": 0.0132,
+      "step": 19950
+    },
+    {
+      "epoch": 4.534318181818182,
+      "grad_norm": 2.2867542123891718,
+      "learning_rate": 2.6589395069877218e-08,
+      "loss": 0.0069,
+      "step": 19951
+    },
+    {
+      "epoch": 4.534545454545454,
+      "grad_norm": 0.6673493288916528,
+      "learning_rate": 2.6563645816262832e-08,
+      "loss": 0.0011,
+      "step": 19952
+    },
+    {
+      "epoch": 4.534772727272728,
+      "grad_norm": 0.45855330764591123,
+      "learning_rate": 2.653790876580964e-08,
+      "loss": 0.0018,
+      "step": 19953
+    },
+    {
+      "epoch": 4.535,
+      "grad_norm": 1.0002813306317537,
+      "learning_rate": 2.651218391904256e-08,
+      "loss": 0.0057,
+      "step": 19954
+    },
+    {
+      "epoch": 4.535227272727273,
+      "grad_norm": 0.5903068830554141,
+      "learning_rate": 2.648647127648611e-08,
+      "loss": 0.0013,
+      "step": 19955
+    },
+    {
+      "epoch": 4.535454545454545,
+      "grad_norm": 0.37173624451482884,
+      "learning_rate": 2.6460770838664732e-08,
+      "loss": 0.0011,
+      "step": 19956
+    },
+    {
+      "epoch": 4.535681818181818,
+      "grad_norm": 0.5837297827355671,
+      "learning_rate": 2.6435082606102447e-08,
+      "loss": 0.003,
+      "step": 19957
+    },
+    {
+      "epoch": 4.535909090909091,
+      "grad_norm": 1.4598667032633132,
+      "learning_rate": 2.6409406579323074e-08,
+      "loss": 0.0093,
+      "step": 19958
+    },
+    {
+      "epoch": 4.536136363636364,
+      "grad_norm": 2.5874828468759783,
+      "learning_rate": 2.6383742758850224e-08,
+      "loss": 0.0111,
+      "step": 19959
+    },
+    {
+      "epoch": 4.536363636363636,
+      "grad_norm": 1.3175831515456606,
+      "learning_rate": 2.6358091145207155e-08,
+      "loss": 0.0043,
+      "step": 19960
+    },
+    {
+      "epoch": 4.53659090909091,
+      "grad_norm": 1.7376325852584322,
+      "learning_rate": 2.633245173891713e-08,
+      "loss": 0.0047,
+      "step": 19961
+    },
+    {
+      "epoch": 4.536818181818182,
+      "grad_norm": 1.1620073586678343,
+      "learning_rate": 2.630682454050272e-08,
+      "loss": 0.0027,
+      "step": 19962
+    },
+    {
+      "epoch": 4.537045454545455,
+      "grad_norm": 1.1485542556865487,
+      "learning_rate": 2.628120955048663e-08,
+      "loss": 0.0027,
+      "step": 19963
+    },
+    {
+      "epoch": 4.537272727272727,
+      "grad_norm": 1.4292837894474903,
+      "learning_rate": 2.625560676939122e-08,
+      "loss": 0.0043,
+      "step": 19964
+    },
+    {
+      "epoch": 4.5375,
+      "grad_norm": 0.6845873435273083,
+      "learning_rate": 2.6230016197738643e-08,
+      "loss": 0.0036,
+      "step": 19965
+    },
+    {
+      "epoch": 4.537727272727273,
+      "grad_norm": 0.7437549862403995,
+      "learning_rate": 2.620443783605056e-08,
+      "loss": 0.0027,
+      "step": 19966
+    },
+    {
+      "epoch": 4.537954545454546,
+      "grad_norm": 1.6240837304544349,
+      "learning_rate": 2.6178871684848645e-08,
+      "loss": 0.0048,
+      "step": 19967
+    },
+    {
+      "epoch": 4.538181818181818,
+      "grad_norm": 0.8052883874393287,
+      "learning_rate": 2.615331774465421e-08,
+      "loss": 0.003,
+      "step": 19968
+    },
+    {
+      "epoch": 4.538409090909091,
+      "grad_norm": 4.500246934539869,
+      "learning_rate": 2.612777601598844e-08,
+      "loss": 0.0062,
+      "step": 19969
+    },
+    {
+      "epoch": 4.538636363636364,
+      "grad_norm": 1.2439805449326578,
+      "learning_rate": 2.6102246499372165e-08,
+      "loss": 0.0045,
+      "step": 19970
+    },
+    {
+      "epoch": 4.5388636363636365,
+      "grad_norm": 2.5824351942763553,
+      "learning_rate": 2.6076729195325806e-08,
+      "loss": 0.0119,
+      "step": 19971
+    },
+    {
+      "epoch": 4.539090909090909,
+      "grad_norm": 0.5645070439636554,
+      "learning_rate": 2.6051224104369846e-08,
+      "loss": 0.0017,
+      "step": 19972
+    },
+    {
+      "epoch": 4.539318181818182,
+      "grad_norm": 0.5387422758791711,
+      "learning_rate": 2.6025731227024286e-08,
+      "loss": 0.0011,
+      "step": 19973
+    },
+    {
+      "epoch": 4.539545454545454,
+      "grad_norm": 2.0423367851289482,
+      "learning_rate": 2.600025056380913e-08,
+      "loss": 0.0063,
+      "step": 19974
+    },
+    {
+      "epoch": 4.5397727272727275,
+      "grad_norm": 0.6225965739146015,
+      "learning_rate": 2.5974782115243818e-08,
+      "loss": 0.0039,
+      "step": 19975
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 1.9157165144967778,
+      "learning_rate": 2.5949325881847733e-08,
+      "loss": 0.0081,
+      "step": 19976
+    },
+    {
+      "epoch": 4.540227272727273,
+      "grad_norm": 0.8158030444981935,
+      "learning_rate": 2.5923881864140034e-08,
+      "loss": 0.0026,
+      "step": 19977
+    },
+    {
+      "epoch": 4.540454545454545,
+      "grad_norm": 1.6673156928778914,
+      "learning_rate": 2.589845006263955e-08,
+      "loss": 0.006,
+      "step": 19978
+    },
+    {
+      "epoch": 4.5406818181818185,
+      "grad_norm": 0.5270838436273989,
+      "learning_rate": 2.5873030477864893e-08,
+      "loss": 0.0041,
+      "step": 19979
+    },
+    {
+      "epoch": 4.540909090909091,
+      "grad_norm": 0.6749600707364176,
+      "learning_rate": 2.5847623110334326e-08,
+      "loss": 0.0061,
+      "step": 19980
+    },
+    {
+      "epoch": 4.5411363636363635,
+      "grad_norm": 1.9590978313118363,
+      "learning_rate": 2.582222796056598e-08,
+      "loss": 0.0042,
+      "step": 19981
+    },
+    {
+      "epoch": 4.541363636363636,
+      "grad_norm": 1.8701791349859047,
+      "learning_rate": 2.579684502907777e-08,
+      "loss": 0.0072,
+      "step": 19982
+    },
+    {
+      "epoch": 4.5415909090909095,
+      "grad_norm": 1.6392372275531055,
+      "learning_rate": 2.5771474316387338e-08,
+      "loss": 0.0039,
+      "step": 19983
+    },
+    {
+      "epoch": 4.541818181818182,
+      "grad_norm": 0.33801768603334825,
+      "learning_rate": 2.5746115823011842e-08,
+      "loss": 0.0011,
+      "step": 19984
+    },
+    {
+      "epoch": 4.5420454545454545,
+      "grad_norm": 2.3765115983122107,
+      "learning_rate": 2.5720769549468504e-08,
+      "loss": 0.01,
+      "step": 19985
+    },
+    {
+      "epoch": 4.542272727272727,
+      "grad_norm": 0.5419218068316622,
+      "learning_rate": 2.5695435496274138e-08,
+      "loss": 0.003,
+      "step": 19986
+    },
+    {
+      "epoch": 4.5425,
+      "grad_norm": 1.565186330182604,
+      "learning_rate": 2.567011366394548e-08,
+      "loss": 0.0091,
+      "step": 19987
+    },
+    {
+      "epoch": 4.542727272727273,
+      "grad_norm": 0.4133506807005697,
+      "learning_rate": 2.5644804052998712e-08,
+      "loss": 0.0011,
+      "step": 19988
+    },
+    {
+      "epoch": 4.5429545454545455,
+      "grad_norm": 2.716631021281454,
+      "learning_rate": 2.5619506663950093e-08,
+      "loss": 0.005,
+      "step": 19989
+    },
+    {
+      "epoch": 4.543181818181818,
+      "grad_norm": 0.38104408219059444,
+      "learning_rate": 2.5594221497315316e-08,
+      "loss": 0.0012,
+      "step": 19990
+    },
+    {
+      "epoch": 4.5434090909090905,
+      "grad_norm": 0.8221441337988642,
+      "learning_rate": 2.5568948553610155e-08,
+      "loss": 0.0044,
+      "step": 19991
+    },
+    {
+      "epoch": 4.543636363636364,
+      "grad_norm": 0.18590515271942665,
+      "learning_rate": 2.5543687833349957e-08,
+      "loss": 0.0005,
+      "step": 19992
+    },
+    {
+      "epoch": 4.543863636363636,
+      "grad_norm": 1.0892037326732438,
+      "learning_rate": 2.5518439337049658e-08,
+      "loss": 0.0106,
+      "step": 19993
+    },
+    {
+      "epoch": 4.544090909090909,
+      "grad_norm": 0.6459760774520635,
+      "learning_rate": 2.5493203065224333e-08,
+      "loss": 0.0019,
+      "step": 19994
+    },
+    {
+      "epoch": 4.5443181818181815,
+      "grad_norm": 1.31413303808974,
+      "learning_rate": 2.5467979018388434e-08,
+      "loss": 0.0094,
+      "step": 19995
+    },
+    {
+      "epoch": 4.544545454545455,
+      "grad_norm": 0.7677370678587537,
+      "learning_rate": 2.5442767197056405e-08,
+      "loss": 0.0059,
+      "step": 19996
+    },
+    {
+      "epoch": 4.544772727272727,
+      "grad_norm": 1.7267791685203413,
+      "learning_rate": 2.5417567601742354e-08,
+      "loss": 0.0109,
+      "step": 19997
+    },
+    {
+      "epoch": 4.545,
+      "grad_norm": 0.7794343573216899,
+      "learning_rate": 2.5392380232960102e-08,
+      "loss": 0.0018,
+      "step": 19998
+    },
+    {
+      "epoch": 4.5452272727272724,
+      "grad_norm": 1.1938016320973563,
+      "learning_rate": 2.5367205091223272e-08,
+      "loss": 0.0087,
+      "step": 19999
+    },
+    {
+      "epoch": 4.545454545454545,
+      "grad_norm": 2.055202516364959,
+      "learning_rate": 2.5342042177045195e-08,
+      "loss": 0.0068,
+      "step": 20000
+    },
+    {
+      "epoch": 4.545681818181818,
+      "grad_norm": 0.4353551112646081,
+      "learning_rate": 2.5316891490939217e-08,
+      "loss": 0.0009,
+      "step": 20001
+    },
+    {
+      "epoch": 4.545909090909091,
+      "grad_norm": 0.26635827713762966,
+      "learning_rate": 2.5291753033417908e-08,
+      "loss": 0.0004,
+      "step": 20002
+    },
+    {
+      "epoch": 4.546136363636363,
+      "grad_norm": 1.2554726442535622,
+      "learning_rate": 2.526662680499399e-08,
+      "loss": 0.0068,
+      "step": 20003
+    },
+    {
+      "epoch": 4.546363636363637,
+      "grad_norm": 0.8882067865502363,
+      "learning_rate": 2.5241512806179826e-08,
+      "loss": 0.003,
+      "step": 20004
+    },
+    {
+      "epoch": 4.546590909090909,
+      "grad_norm": 0.4088862139997798,
+      "learning_rate": 2.5216411037487578e-08,
+      "loss": 0.0029,
+      "step": 20005
+    },
+    {
+      "epoch": 4.546818181818182,
+      "grad_norm": 0.40881744445957025,
+      "learning_rate": 2.5191321499429128e-08,
+      "loss": 0.0028,
+      "step": 20006
+    },
+    {
+      "epoch": 4.547045454545454,
+      "grad_norm": 2.2296696627155925,
+      "learning_rate": 2.516624419251601e-08,
+      "loss": 0.0089,
+      "step": 20007
+    },
+    {
+      "epoch": 4.547272727272727,
+      "grad_norm": 0.9856105941410991,
+      "learning_rate": 2.514117911725962e-08,
+      "loss": 0.006,
+      "step": 20008
+    },
+    {
+      "epoch": 4.5475,
+      "grad_norm": 1.5223536062747498,
+      "learning_rate": 2.511612627417122e-08,
+      "loss": 0.0124,
+      "step": 20009
+    },
+    {
+      "epoch": 4.547727272727273,
+      "grad_norm": 0.5558931897758878,
+      "learning_rate": 2.5091085663761434e-08,
+      "loss": 0.0019,
+      "step": 20010
+    },
+    {
+      "epoch": 4.547954545454545,
+      "grad_norm": 0.6464389034204825,
+      "learning_rate": 2.5066057286541044e-08,
+      "loss": 0.0025,
+      "step": 20011
+    },
+    {
+      "epoch": 4.548181818181818,
+      "grad_norm": 1.0260193133978768,
+      "learning_rate": 2.5041041143020404e-08,
+      "loss": 0.0054,
+      "step": 20012
+    },
+    {
+      "epoch": 4.548409090909091,
+      "grad_norm": 1.5915146915882143,
+      "learning_rate": 2.5016037233709657e-08,
+      "loss": 0.0126,
+      "step": 20013
+    },
+    {
+      "epoch": 4.548636363636364,
+      "grad_norm": 0.879119432711272,
+      "learning_rate": 2.4991045559118603e-08,
+      "loss": 0.0037,
+      "step": 20014
+    },
+    {
+      "epoch": 4.548863636363636,
+      "grad_norm": 0.4476079128972119,
+      "learning_rate": 2.496606611975691e-08,
+      "loss": 0.0024,
+      "step": 20015
+    },
+    {
+      "epoch": 4.549090909090909,
+      "grad_norm": 1.1550747229712217,
+      "learning_rate": 2.4941098916133958e-08,
+      "loss": 0.0047,
+      "step": 20016
+    },
+    {
+      "epoch": 4.549318181818181,
+      "grad_norm": 0.8352057979919217,
+      "learning_rate": 2.4916143948758784e-08,
+      "loss": 0.0028,
+      "step": 20017
+    },
+    {
+      "epoch": 4.549545454545455,
+      "grad_norm": 0.09368421618721212,
+      "learning_rate": 2.4891201218140358e-08,
+      "loss": 0.0002,
+      "step": 20018
+    },
+    {
+      "epoch": 4.549772727272727,
+      "grad_norm": 2.170971018851526,
+      "learning_rate": 2.4866270724787368e-08,
+      "loss": 0.0037,
+      "step": 20019
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 1.2270570570518433,
+      "learning_rate": 2.484135246920802e-08,
+      "loss": 0.0028,
+      "step": 20020
+    },
+    {
+      "epoch": 4.550227272727272,
+      "grad_norm": 2.684102151677205,
+      "learning_rate": 2.4816446451910585e-08,
+      "loss": 0.0182,
+      "step": 20021
+    },
+    {
+      "epoch": 4.550454545454546,
+      "grad_norm": 0.8825144809974783,
+      "learning_rate": 2.4791552673402927e-08,
+      "loss": 0.0037,
+      "step": 20022
+    },
+    {
+      "epoch": 4.550681818181818,
+      "grad_norm": 1.9706754483520255,
+      "learning_rate": 2.476667113419255e-08,
+      "loss": 0.0083,
+      "step": 20023
+    },
+    {
+      "epoch": 4.550909090909091,
+      "grad_norm": 1.176054265801073,
+      "learning_rate": 2.4741801834786902e-08,
+      "loss": 0.0039,
+      "step": 20024
+    },
+    {
+      "epoch": 4.551136363636363,
+      "grad_norm": 1.8327256048191023,
+      "learning_rate": 2.471694477569314e-08,
+      "loss": 0.0125,
+      "step": 20025
+    },
+    {
+      "epoch": 4.551363636363637,
+      "grad_norm": 1.5187028194269945,
+      "learning_rate": 2.4692099957418157e-08,
+      "loss": 0.0151,
+      "step": 20026
+    },
+    {
+      "epoch": 4.551590909090909,
+      "grad_norm": 1.6326851029489553,
+      "learning_rate": 2.4667267380468553e-08,
+      "loss": 0.0053,
+      "step": 20027
+    },
+    {
+      "epoch": 4.551818181818182,
+      "grad_norm": 0.1042991989956587,
+      "learning_rate": 2.4642447045350736e-08,
+      "loss": 0.0002,
+      "step": 20028
+    },
+    {
+      "epoch": 4.552045454545454,
+      "grad_norm": 2.3369762295047387,
+      "learning_rate": 2.4617638952570688e-08,
+      "loss": 0.0067,
+      "step": 20029
+    },
+    {
+      "epoch": 4.552272727272728,
+      "grad_norm": 0.6585959408688014,
+      "learning_rate": 2.4592843102634463e-08,
+      "loss": 0.0043,
+      "step": 20030
+    },
+    {
+      "epoch": 4.5525,
+      "grad_norm": 0.1377892155699207,
+      "learning_rate": 2.4568059496047764e-08,
+      "loss": 0.0003,
+      "step": 20031
+    },
+    {
+      "epoch": 4.552727272727273,
+      "grad_norm": 1.9009765009166986,
+      "learning_rate": 2.4543288133315676e-08,
+      "loss": 0.0087,
+      "step": 20032
+    },
+    {
+      "epoch": 4.552954545454545,
+      "grad_norm": 0.9733222214409696,
+      "learning_rate": 2.451852901494356e-08,
+      "loss": 0.0038,
+      "step": 20033
+    },
+    {
+      "epoch": 4.553181818181818,
+      "grad_norm": 1.1917594119546637,
+      "learning_rate": 2.449378214143622e-08,
+      "loss": 0.009,
+      "step": 20034
+    },
+    {
+      "epoch": 4.553409090909091,
+      "grad_norm": 4.240203077500053,
+      "learning_rate": 2.4469047513298178e-08,
+      "loss": 0.0163,
+      "step": 20035
+    },
+    {
+      "epoch": 4.553636363636364,
+      "grad_norm": 1.8741950610243674,
+      "learning_rate": 2.4444325131034175e-08,
+      "loss": 0.0072,
+      "step": 20036
+    },
+    {
+      "epoch": 4.553863636363636,
+      "grad_norm": 0.43486525000782283,
+      "learning_rate": 2.4419614995147973e-08,
+      "loss": 0.0006,
+      "step": 20037
+    },
+    {
+      "epoch": 4.554090909090909,
+      "grad_norm": 0.3721078597350748,
+      "learning_rate": 2.4394917106143612e-08,
+      "loss": 0.0013,
+      "step": 20038
+    },
+    {
+      "epoch": 4.554318181818182,
+      "grad_norm": 0.6005647806587291,
+      "learning_rate": 2.437023146452465e-08,
+      "loss": 0.0015,
+      "step": 20039
+    },
+    {
+      "epoch": 4.554545454545455,
+      "grad_norm": 0.38471586597174606,
+      "learning_rate": 2.434555807079471e-08,
+      "loss": 0.001,
+      "step": 20040
+    },
+    {
+      "epoch": 4.554772727272727,
+      "grad_norm": 0.7838336329227569,
+      "learning_rate": 2.4320896925456584e-08,
+      "loss": 0.0051,
+      "step": 20041
+    },
+    {
+      "epoch": 4.555,
+      "grad_norm": 1.1361725773518527,
+      "learning_rate": 2.429624802901334e-08,
+      "loss": 0.0028,
+      "step": 20042
+    },
+    {
+      "epoch": 4.555227272727272,
+      "grad_norm": 0.33648322016958604,
+      "learning_rate": 2.4271611381967568e-08,
+      "loss": 0.0011,
+      "step": 20043
+    },
+    {
+      "epoch": 4.555454545454546,
+      "grad_norm": 1.6926869195410676,
+      "learning_rate": 2.424698698482171e-08,
+      "loss": 0.007,
+      "step": 20044
+    },
+    {
+      "epoch": 4.555681818181818,
+      "grad_norm": 1.9016915013912545,
+      "learning_rate": 2.4222374838077864e-08,
+      "loss": 0.0042,
+      "step": 20045
+    },
+    {
+      "epoch": 4.555909090909091,
+      "grad_norm": 1.508848605751838,
+      "learning_rate": 2.419777494223785e-08,
+      "loss": 0.0205,
+      "step": 20046
+    },
+    {
+      "epoch": 4.556136363636364,
+      "grad_norm": 1.5985473023951209,
+      "learning_rate": 2.417318729780342e-08,
+      "loss": 0.0048,
+      "step": 20047
+    },
+    {
+      "epoch": 4.556363636363637,
+      "grad_norm": 1.1720725890789214,
+      "learning_rate": 2.4148611905275844e-08,
+      "loss": 0.0042,
+      "step": 20048
+    },
+    {
+      "epoch": 4.556590909090909,
+      "grad_norm": 0.4262160041584269,
+      "learning_rate": 2.412404876515631e-08,
+      "loss": 0.0012,
+      "step": 20049
+    },
+    {
+      "epoch": 4.556818181818182,
+      "grad_norm": 0.7235339499843138,
+      "learning_rate": 2.4099497877945745e-08,
+      "loss": 0.0025,
+      "step": 20050
+    },
+    {
+      "epoch": 4.557045454545454,
+      "grad_norm": 0.3799341610555203,
+      "learning_rate": 2.4074959244144713e-08,
+      "loss": 0.0011,
+      "step": 20051
+    },
+    {
+      "epoch": 4.557272727272728,
+      "grad_norm": 0.46023557775226887,
+      "learning_rate": 2.4050432864253652e-08,
+      "loss": 0.0014,
+      "step": 20052
+    },
+    {
+      "epoch": 4.5575,
+      "grad_norm": 0.5140918346202471,
+      "learning_rate": 2.402591873877279e-08,
+      "loss": 0.0026,
+      "step": 20053
+    },
+    {
+      "epoch": 4.557727272727273,
+      "grad_norm": 0.7065833035575134,
+      "learning_rate": 2.4001416868201787e-08,
+      "loss": 0.0012,
+      "step": 20054
+    },
+    {
+      "epoch": 4.557954545454545,
+      "grad_norm": 0.6349016593108694,
+      "learning_rate": 2.3976927253040393e-08,
+      "loss": 0.0015,
+      "step": 20055
+    },
+    {
+      "epoch": 4.558181818181819,
+      "grad_norm": 0.8438120221108635,
+      "learning_rate": 2.395244989378799e-08,
+      "loss": 0.0043,
+      "step": 20056
+    },
+    {
+      "epoch": 4.558409090909091,
+      "grad_norm": 1.3327205199560348,
+      "learning_rate": 2.3927984790943702e-08,
+      "loss": 0.0115,
+      "step": 20057
+    },
+    {
+      "epoch": 4.558636363636364,
+      "grad_norm": 0.5178766049172272,
+      "learning_rate": 2.3903531945006568e-08,
+      "loss": 0.0026,
+      "step": 20058
+    },
+    {
+      "epoch": 4.558863636363636,
+      "grad_norm": 0.2937625253520414,
+      "learning_rate": 2.3879091356474938e-08,
+      "loss": 0.0015,
+      "step": 20059
+    },
+    {
+      "epoch": 4.559090909090909,
+      "grad_norm": 0.3691400902912058,
+      "learning_rate": 2.3854663025847448e-08,
+      "loss": 0.0038,
+      "step": 20060
+    },
+    {
+      "epoch": 4.559318181818182,
+      "grad_norm": 0.4667786470354639,
+      "learning_rate": 2.383024695362203e-08,
+      "loss": 0.0023,
+      "step": 20061
+    },
+    {
+      "epoch": 4.559545454545455,
+      "grad_norm": 2.250948053417554,
+      "learning_rate": 2.3805843140296824e-08,
+      "loss": 0.0145,
+      "step": 20062
+    },
+    {
+      "epoch": 4.559772727272727,
+      "grad_norm": 0.9758428552938321,
+      "learning_rate": 2.3781451586369213e-08,
+      "loss": 0.0062,
+      "step": 20063
+    },
+    {
+      "epoch": 4.5600000000000005,
+      "grad_norm": 0.3050881980004224,
+      "learning_rate": 2.375707229233665e-08,
+      "loss": 0.0007,
+      "step": 20064
+    },
+    {
+      "epoch": 4.560227272727273,
+      "grad_norm": 2.8996468169610745,
+      "learning_rate": 2.3732705258696296e-08,
+      "loss": 0.0137,
+      "step": 20065
+    },
+    {
+      "epoch": 4.5604545454545455,
+      "grad_norm": 0.6739105961089594,
+      "learning_rate": 2.3708350485945193e-08,
+      "loss": 0.003,
+      "step": 20066
+    },
+    {
+      "epoch": 4.560681818181818,
+      "grad_norm": 0.615749638377527,
+      "learning_rate": 2.3684007974579745e-08,
+      "loss": 0.0039,
+      "step": 20067
+    },
+    {
+      "epoch": 4.560909090909091,
+      "grad_norm": 0.8165796281940674,
+      "learning_rate": 2.3659677725096365e-08,
+      "loss": 0.0036,
+      "step": 20068
+    },
+    {
+      "epoch": 4.561136363636364,
+      "grad_norm": 1.304695809021939,
+      "learning_rate": 2.3635359737991317e-08,
+      "loss": 0.0053,
+      "step": 20069
+    },
+    {
+      "epoch": 4.5613636363636365,
+      "grad_norm": 1.9033768339825086,
+      "learning_rate": 2.361105401376046e-08,
+      "loss": 0.0082,
+      "step": 20070
+    },
+    {
+      "epoch": 4.561590909090909,
+      "grad_norm": 1.1004203894941273,
+      "learning_rate": 2.3586760552899364e-08,
+      "loss": 0.0037,
+      "step": 20071
+    },
+    {
+      "epoch": 4.5618181818181816,
+      "grad_norm": 0.3387164096017668,
+      "learning_rate": 2.356247935590347e-08,
+      "loss": 0.0012,
+      "step": 20072
+    },
+    {
+      "epoch": 4.562045454545455,
+      "grad_norm": 0.7715300368740432,
+      "learning_rate": 2.3538210423267803e-08,
+      "loss": 0.0032,
+      "step": 20073
+    },
+    {
+      "epoch": 4.5622727272727275,
+      "grad_norm": 1.5345076498216998,
+      "learning_rate": 2.3513953755487373e-08,
+      "loss": 0.0062,
+      "step": 20074
+    },
+    {
+      "epoch": 4.5625,
+      "grad_norm": 0.7985572530799855,
+      "learning_rate": 2.3489709353056862e-08,
+      "loss": 0.0034,
+      "step": 20075
+    },
+    {
+      "epoch": 4.5627272727272725,
+      "grad_norm": 0.3274133701330426,
+      "learning_rate": 2.346547721647052e-08,
+      "loss": 0.0013,
+      "step": 20076
+    },
+    {
+      "epoch": 4.562954545454545,
+      "grad_norm": 0.8345524801260756,
+      "learning_rate": 2.3441257346222478e-08,
+      "loss": 0.006,
+      "step": 20077
+    },
+    {
+      "epoch": 4.5631818181818184,
+      "grad_norm": 1.4791924569044872,
+      "learning_rate": 2.34170497428067e-08,
+      "loss": 0.0034,
+      "step": 20078
+    },
+    {
+      "epoch": 4.563409090909091,
+      "grad_norm": 3.1786776688428815,
+      "learning_rate": 2.3392854406716762e-08,
+      "loss": 0.0282,
+      "step": 20079
+    },
+    {
+      "epoch": 4.5636363636363635,
+      "grad_norm": 0.22608917818947405,
+      "learning_rate": 2.3368671338446156e-08,
+      "loss": 0.0007,
+      "step": 20080
+    },
+    {
+      "epoch": 4.563863636363636,
+      "grad_norm": 1.0111836869144837,
+      "learning_rate": 2.3344500538487886e-08,
+      "loss": 0.0061,
+      "step": 20081
+    },
+    {
+      "epoch": 4.564090909090909,
+      "grad_norm": 1.532774238613387,
+      "learning_rate": 2.332034200733489e-08,
+      "loss": 0.0177,
+      "step": 20082
+    },
+    {
+      "epoch": 4.564318181818182,
+      "grad_norm": 1.6939558234169645,
+      "learning_rate": 2.329619574547984e-08,
+      "loss": 0.0066,
+      "step": 20083
+    },
+    {
+      "epoch": 4.5645454545454545,
+      "grad_norm": 1.7312884156301038,
+      "learning_rate": 2.3272061753415175e-08,
+      "loss": 0.0095,
+      "step": 20084
+    },
+    {
+      "epoch": 4.564772727272727,
+      "grad_norm": 0.7047954533171881,
+      "learning_rate": 2.3247940031632804e-08,
+      "loss": 0.0026,
+      "step": 20085
+    },
+    {
+      "epoch": 4.5649999999999995,
+      "grad_norm": 2.003489661018332,
+      "learning_rate": 2.322383058062483e-08,
+      "loss": 0.0195,
+      "step": 20086
+    },
+    {
+      "epoch": 4.565227272727273,
+      "grad_norm": 1.0766871465435306,
+      "learning_rate": 2.319973340088273e-08,
+      "loss": 0.0076,
+      "step": 20087
+    },
+    {
+      "epoch": 4.565454545454545,
+      "grad_norm": 1.5670117099738032,
+      "learning_rate": 2.3175648492898136e-08,
+      "loss": 0.0025,
+      "step": 20088
+    },
+    {
+      "epoch": 4.565681818181818,
+      "grad_norm": 1.2423980568131032,
+      "learning_rate": 2.3151575857161825e-08,
+      "loss": 0.0058,
+      "step": 20089
+    },
+    {
+      "epoch": 4.565909090909091,
+      "grad_norm": 2.267692982147501,
+      "learning_rate": 2.3127515494164935e-08,
+      "loss": 0.0185,
+      "step": 20090
+    },
+    {
+      "epoch": 4.566136363636364,
+      "grad_norm": 0.5058637291573156,
+      "learning_rate": 2.310346740439798e-08,
+      "loss": 0.0017,
+      "step": 20091
+    },
+    {
+      "epoch": 4.566363636363636,
+      "grad_norm": 0.449512526201359,
+      "learning_rate": 2.3079431588351464e-08,
+      "loss": 0.0055,
+      "step": 20092
+    },
+    {
+      "epoch": 4.566590909090909,
+      "grad_norm": 0.8471996540335579,
+      "learning_rate": 2.3055408046515487e-08,
+      "loss": 0.0037,
+      "step": 20093
+    },
+    {
+      "epoch": 4.566818181818181,
+      "grad_norm": 1.4769802144261126,
+      "learning_rate": 2.3031396779379724e-08,
+      "loss": 0.0042,
+      "step": 20094
+    },
+    {
+      "epoch": 4.567045454545455,
+      "grad_norm": 1.0351499601495668,
+      "learning_rate": 2.3007397787434064e-08,
+      "loss": 0.0077,
+      "step": 20095
+    },
+    {
+      "epoch": 4.567272727272727,
+      "grad_norm": 0.6253474501171675,
+      "learning_rate": 2.2983411071167765e-08,
+      "loss": 0.0039,
+      "step": 20096
+    },
+    {
+      "epoch": 4.5675,
+      "grad_norm": 1.0221178846518828,
+      "learning_rate": 2.2959436631070086e-08,
+      "loss": 0.004,
+      "step": 20097
+    },
+    {
+      "epoch": 4.567727272727272,
+      "grad_norm": 1.2419548824196265,
+      "learning_rate": 2.293547446762974e-08,
+      "loss": 0.0037,
+      "step": 20098
+    },
+    {
+      "epoch": 4.567954545454546,
+      "grad_norm": 0.606651248055708,
+      "learning_rate": 2.2911524581335428e-08,
+      "loss": 0.001,
+      "step": 20099
+    },
+    {
+      "epoch": 4.568181818181818,
+      "grad_norm": 0.4872137520081244,
+      "learning_rate": 2.288758697267551e-08,
+      "loss": 0.0014,
+      "step": 20100
+    },
+    {
+      "epoch": 4.568409090909091,
+      "grad_norm": 1.680791462892659,
+      "learning_rate": 2.2863661642138142e-08,
+      "loss": 0.0073,
+      "step": 20101
+    },
+    {
+      "epoch": 4.568636363636363,
+      "grad_norm": 0.2551359528311373,
+      "learning_rate": 2.2839748590211264e-08,
+      "loss": 0.0007,
+      "step": 20102
+    },
+    {
+      "epoch": 4.568863636363636,
+      "grad_norm": 0.5678343409015254,
+      "learning_rate": 2.28158478173824e-08,
+      "loss": 0.0017,
+      "step": 20103
+    },
+    {
+      "epoch": 4.569090909090909,
+      "grad_norm": 0.6039297663155445,
+      "learning_rate": 2.2791959324138942e-08,
+      "loss": 0.0034,
+      "step": 20104
+    },
+    {
+      "epoch": 4.569318181818182,
+      "grad_norm": 2.518957937652886,
+      "learning_rate": 2.2768083110968065e-08,
+      "loss": 0.0096,
+      "step": 20105
+    },
+    {
+      "epoch": 4.569545454545454,
+      "grad_norm": 0.7525871737398552,
+      "learning_rate": 2.2744219178356676e-08,
+      "loss": 0.0037,
+      "step": 20106
+    },
+    {
+      "epoch": 4.569772727272728,
+      "grad_norm": 2.1340165838257383,
+      "learning_rate": 2.2720367526791325e-08,
+      "loss": 0.0212,
+      "step": 20107
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 0.7870064116176505,
+      "learning_rate": 2.2696528156758366e-08,
+      "loss": 0.0029,
+      "step": 20108
+    },
+    {
+      "epoch": 4.570227272727273,
+      "grad_norm": 0.7771671603587207,
+      "learning_rate": 2.267270106874393e-08,
+      "loss": 0.0055,
+      "step": 20109
+    },
+    {
+      "epoch": 4.570454545454545,
+      "grad_norm": 1.2041464852659103,
+      "learning_rate": 2.2648886263234027e-08,
+      "loss": 0.0093,
+      "step": 20110
+    },
+    {
+      "epoch": 4.570681818181818,
+      "grad_norm": 0.36941420331612945,
+      "learning_rate": 2.262508374071416e-08,
+      "loss": 0.0008,
+      "step": 20111
+    },
+    {
+      "epoch": 4.570909090909091,
+      "grad_norm": 0.5032352925706307,
+      "learning_rate": 2.2601293501669715e-08,
+      "loss": 0.001,
+      "step": 20112
+    },
+    {
+      "epoch": 4.571136363636364,
+      "grad_norm": 0.3540416626415914,
+      "learning_rate": 2.2577515546585854e-08,
+      "loss": 0.0016,
+      "step": 20113
+    },
+    {
+      "epoch": 4.571363636363636,
+      "grad_norm": 0.8179312510029644,
+      "learning_rate": 2.2553749875947466e-08,
+      "loss": 0.0051,
+      "step": 20114
+    },
+    {
+      "epoch": 4.571590909090909,
+      "grad_norm": 1.2169532474234246,
+      "learning_rate": 2.2529996490239167e-08,
+      "loss": 0.0082,
+      "step": 20115
+    },
+    {
+      "epoch": 4.571818181818182,
+      "grad_norm": 1.9221296989139687,
+      "learning_rate": 2.250625538994522e-08,
+      "loss": 0.0164,
+      "step": 20116
+    },
+    {
+      "epoch": 4.572045454545455,
+      "grad_norm": 0.23249160244008743,
+      "learning_rate": 2.248252657554989e-08,
+      "loss": 0.0008,
+      "step": 20117
+    },
+    {
+      "epoch": 4.572272727272727,
+      "grad_norm": 2.1742429988298952,
+      "learning_rate": 2.2458810047536957e-08,
+      "loss": 0.0089,
+      "step": 20118
+    },
+    {
+      "epoch": 4.5725,
+      "grad_norm": 1.2867232729135685,
+      "learning_rate": 2.2435105806390205e-08,
+      "loss": 0.0067,
+      "step": 20119
+    },
+    {
+      "epoch": 4.572727272727272,
+      "grad_norm": 0.7302232410885194,
+      "learning_rate": 2.241141385259271e-08,
+      "loss": 0.004,
+      "step": 20120
+    },
+    {
+      "epoch": 4.572954545454546,
+      "grad_norm": 1.8907256070222915,
+      "learning_rate": 2.2387734186627848e-08,
+      "loss": 0.0072,
+      "step": 20121
+    },
+    {
+      "epoch": 4.573181818181818,
+      "grad_norm": 1.5031664857854474,
+      "learning_rate": 2.2364066808978346e-08,
+      "loss": 0.0045,
+      "step": 20122
+    },
+    {
+      "epoch": 4.573409090909091,
+      "grad_norm": 2.2162559959787185,
+      "learning_rate": 2.234041172012688e-08,
+      "loss": 0.0171,
+      "step": 20123
+    },
+    {
+      "epoch": 4.573636363636363,
+      "grad_norm": 0.7648822442748268,
+      "learning_rate": 2.2316768920555843e-08,
+      "loss": 0.0022,
+      "step": 20124
+    },
+    {
+      "epoch": 4.573863636363637,
+      "grad_norm": 1.0323264482485641,
+      "learning_rate": 2.2293138410747344e-08,
+      "loss": 0.0033,
+      "step": 20125
+    },
+    {
+      "epoch": 4.574090909090909,
+      "grad_norm": 0.8040388826173421,
+      "learning_rate": 2.2269520191183154e-08,
+      "loss": 0.004,
+      "step": 20126
+    },
+    {
+      "epoch": 4.574318181818182,
+      "grad_norm": 0.3777595552579785,
+      "learning_rate": 2.224591426234504e-08,
+      "loss": 0.0033,
+      "step": 20127
+    },
+    {
+      "epoch": 4.574545454545454,
+      "grad_norm": 0.3175929761339825,
+      "learning_rate": 2.2222320624714355e-08,
+      "loss": 0.0002,
+      "step": 20128
+    },
+    {
+      "epoch": 4.574772727272728,
+      "grad_norm": 0.40932581533101947,
+      "learning_rate": 2.21987392787721e-08,
+      "loss": 0.0015,
+      "step": 20129
+    },
+    {
+      "epoch": 4.575,
+      "grad_norm": 0.5701189683920854,
+      "learning_rate": 2.217517022499914e-08,
+      "loss": 0.0011,
+      "step": 20130
+    },
+    {
+      "epoch": 4.575227272727273,
+      "grad_norm": 1.7848405359854274,
+      "learning_rate": 2.2151613463876205e-08,
+      "loss": 0.0139,
+      "step": 20131
+    },
+    {
+      "epoch": 4.575454545454545,
+      "grad_norm": 0.7959825307188035,
+      "learning_rate": 2.21280689958836e-08,
+      "loss": 0.0024,
+      "step": 20132
+    },
+    {
+      "epoch": 4.575681818181819,
+      "grad_norm": 0.32443529135198584,
+      "learning_rate": 2.210453682150143e-08,
+      "loss": 0.0031,
+      "step": 20133
+    },
+    {
+      "epoch": 4.575909090909091,
+      "grad_norm": 2.0602529127048723,
+      "learning_rate": 2.2081016941209524e-08,
+      "loss": 0.0099,
+      "step": 20134
+    },
+    {
+      "epoch": 4.576136363636364,
+      "grad_norm": 2.119132101569688,
+      "learning_rate": 2.2057509355487562e-08,
+      "loss": 0.0054,
+      "step": 20135
+    },
+    {
+      "epoch": 4.576363636363636,
+      "grad_norm": 0.1238006599661712,
+      "learning_rate": 2.2034014064814884e-08,
+      "loss": 0.0003,
+      "step": 20136
+    },
+    {
+      "epoch": 4.576590909090909,
+      "grad_norm": 1.7102950571275124,
+      "learning_rate": 2.2010531069670618e-08,
+      "loss": 0.0075,
+      "step": 20137
+    },
+    {
+      "epoch": 4.576818181818182,
+      "grad_norm": 2.0566008760885808,
+      "learning_rate": 2.1987060370533553e-08,
+      "loss": 0.0058,
+      "step": 20138
+    },
+    {
+      "epoch": 4.577045454545455,
+      "grad_norm": 0.21112278507858157,
+      "learning_rate": 2.196360196788233e-08,
+      "loss": 0.0006,
+      "step": 20139
+    },
+    {
+      "epoch": 4.577272727272727,
+      "grad_norm": 0.4897376485767561,
+      "learning_rate": 2.194015586219532e-08,
+      "loss": 0.0032,
+      "step": 20140
+    },
+    {
+      "epoch": 4.5775,
+      "grad_norm": 3.1419281764249907,
+      "learning_rate": 2.191672205395061e-08,
+      "loss": 0.0187,
+      "step": 20141
+    },
+    {
+      "epoch": 4.577727272727273,
+      "grad_norm": 2.430665128780345,
+      "learning_rate": 2.1893300543626082e-08,
+      "loss": 0.0117,
+      "step": 20142
+    },
+    {
+      "epoch": 4.577954545454546,
+      "grad_norm": 0.36804375106880993,
+      "learning_rate": 2.1869891331699338e-08,
+      "loss": 0.0012,
+      "step": 20143
+    },
+    {
+      "epoch": 4.578181818181818,
+      "grad_norm": 0.6932272176978798,
+      "learning_rate": 2.184649441864778e-08,
+      "loss": 0.0031,
+      "step": 20144
+    },
+    {
+      "epoch": 4.578409090909091,
+      "grad_norm": 0.4410829755844138,
+      "learning_rate": 2.182310980494845e-08,
+      "loss": 0.001,
+      "step": 20145
+    },
+    {
+      "epoch": 4.578636363636363,
+      "grad_norm": 1.2402951873080854,
+      "learning_rate": 2.1799737491078193e-08,
+      "loss": 0.008,
+      "step": 20146
+    },
+    {
+      "epoch": 4.578863636363637,
+      "grad_norm": 1.0335217480868486,
+      "learning_rate": 2.177637747751357e-08,
+      "loss": 0.0087,
+      "step": 20147
+    },
+    {
+      "epoch": 4.579090909090909,
+      "grad_norm": 1.3075632683173573,
+      "learning_rate": 2.175302976473108e-08,
+      "loss": 0.0058,
+      "step": 20148
+    },
+    {
+      "epoch": 4.579318181818182,
+      "grad_norm": 1.3254579490257852,
+      "learning_rate": 2.1729694353206653e-08,
+      "loss": 0.0054,
+      "step": 20149
+    },
+    {
+      "epoch": 4.579545454545455,
+      "grad_norm": 0.43175647721893445,
+      "learning_rate": 2.1706371243416303e-08,
+      "loss": 0.0012,
+      "step": 20150
+    },
+    {
+      "epoch": 4.5797727272727276,
+      "grad_norm": 0.2178609611276076,
+      "learning_rate": 2.1683060435835554e-08,
+      "loss": 0.0005,
+      "step": 20151
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 1.758707441210586,
+      "learning_rate": 2.1659761930939647e-08,
+      "loss": 0.0048,
+      "step": 20152
+    },
+    {
+      "epoch": 4.580227272727273,
+      "grad_norm": 0.36563691261518605,
+      "learning_rate": 2.1636475729203827e-08,
+      "loss": 0.0029,
+      "step": 20153
+    },
+    {
+      "epoch": 4.580454545454545,
+      "grad_norm": 0.8727382949533181,
+      "learning_rate": 2.1613201831102922e-08,
+      "loss": 0.0042,
+      "step": 20154
+    },
+    {
+      "epoch": 4.5806818181818185,
+      "grad_norm": 1.3567981053023226,
+      "learning_rate": 2.1589940237111416e-08,
+      "loss": 0.0056,
+      "step": 20155
+    },
+    {
+      "epoch": 4.580909090909091,
+      "grad_norm": 0.3496551333286933,
+      "learning_rate": 2.156669094770379e-08,
+      "loss": 0.0015,
+      "step": 20156
+    },
+    {
+      "epoch": 4.581136363636364,
+      "grad_norm": 0.863142096347651,
+      "learning_rate": 2.154345396335411e-08,
+      "loss": 0.0034,
+      "step": 20157
+    },
+    {
+      "epoch": 4.581363636363636,
+      "grad_norm": 0.7991635545183807,
+      "learning_rate": 2.152022928453616e-08,
+      "loss": 0.0039,
+      "step": 20158
+    },
+    {
+      "epoch": 4.5815909090909095,
+      "grad_norm": 0.43371052715581226,
+      "learning_rate": 2.1497016911723595e-08,
+      "loss": 0.0016,
+      "step": 20159
+    },
+    {
+      "epoch": 4.581818181818182,
+      "grad_norm": 2.1964075399588,
+      "learning_rate": 2.1473816845389712e-08,
+      "loss": 0.0115,
+      "step": 20160
+    },
+    {
+      "epoch": 4.5820454545454545,
+      "grad_norm": 1.168725459664931,
+      "learning_rate": 2.1450629086007605e-08,
+      "loss": 0.0125,
+      "step": 20161
+    },
+    {
+      "epoch": 4.582272727272727,
+      "grad_norm": 0.39850279596463983,
+      "learning_rate": 2.1427453634050092e-08,
+      "loss": 0.0015,
+      "step": 20162
+    },
+    {
+      "epoch": 4.5825,
+      "grad_norm": 1.1241475034813562,
+      "learning_rate": 2.140429048998985e-08,
+      "loss": 0.0063,
+      "step": 20163
+    },
+    {
+      "epoch": 4.582727272727273,
+      "grad_norm": 0.2552689629684154,
+      "learning_rate": 2.138113965429914e-08,
+      "loss": 0.0009,
+      "step": 20164
+    },
+    {
+      "epoch": 4.5829545454545455,
+      "grad_norm": 2.148779844455158,
+      "learning_rate": 2.1358001127450012e-08,
+      "loss": 0.0219,
+      "step": 20165
+    },
+    {
+      "epoch": 4.583181818181818,
+      "grad_norm": 1.0473896786154167,
+      "learning_rate": 2.1334874909914385e-08,
+      "loss": 0.002,
+      "step": 20166
+    },
+    {
+      "epoch": 4.5834090909090905,
+      "grad_norm": 0.9298145727727417,
+      "learning_rate": 2.1311761002163894e-08,
+      "loss": 0.0032,
+      "step": 20167
+    },
+    {
+      "epoch": 4.583636363636364,
+      "grad_norm": 1.9887682468780559,
+      "learning_rate": 2.128865940466969e-08,
+      "loss": 0.025,
+      "step": 20168
+    },
+    {
+      "epoch": 4.5838636363636365,
+      "grad_norm": 1.4859809536525805,
+      "learning_rate": 2.1265570117902922e-08,
+      "loss": 0.0058,
+      "step": 20169
+    },
+    {
+      "epoch": 4.584090909090909,
+      "grad_norm": 0.729848915783496,
+      "learning_rate": 2.1242493142334468e-08,
+      "loss": 0.0021,
+      "step": 20170
+    },
+    {
+      "epoch": 4.5843181818181815,
+      "grad_norm": 1.728115978862249,
+      "learning_rate": 2.1219428478434917e-08,
+      "loss": 0.0053,
+      "step": 20171
+    },
+    {
+      "epoch": 4.584545454545455,
+      "grad_norm": 1.3356383402182193,
+      "learning_rate": 2.1196376126674592e-08,
+      "loss": 0.0038,
+      "step": 20172
+    },
+    {
+      "epoch": 4.584772727272727,
+      "grad_norm": 1.0657393972389675,
+      "learning_rate": 2.1173336087523532e-08,
+      "loss": 0.0055,
+      "step": 20173
+    },
+    {
+      "epoch": 4.585,
+      "grad_norm": 0.7358046063733428,
+      "learning_rate": 2.115030836145157e-08,
+      "loss": 0.0007,
+      "step": 20174
+    },
+    {
+      "epoch": 4.5852272727272725,
+      "grad_norm": 2.0814338248121587,
+      "learning_rate": 2.1127292948928263e-08,
+      "loss": 0.0078,
+      "step": 20175
+    },
+    {
+      "epoch": 4.585454545454546,
+      "grad_norm": 1.1408149257323532,
+      "learning_rate": 2.1104289850423092e-08,
+      "loss": 0.0098,
+      "step": 20176
+    },
+    {
+      "epoch": 4.585681818181818,
+      "grad_norm": 1.3645808962305073,
+      "learning_rate": 2.1081299066404923e-08,
+      "loss": 0.0068,
+      "step": 20177
+    },
+    {
+      "epoch": 4.585909090909091,
+      "grad_norm": 1.3106797740604337,
+      "learning_rate": 2.105832059734261e-08,
+      "loss": 0.0101,
+      "step": 20178
+    },
+    {
+      "epoch": 4.5861363636363635,
+      "grad_norm": 2.8692119358936803,
+      "learning_rate": 2.1035354443704815e-08,
+      "loss": 0.015,
+      "step": 20179
+    },
+    {
+      "epoch": 4.586363636363636,
+      "grad_norm": 2.166708839341907,
+      "learning_rate": 2.101240060595984e-08,
+      "loss": 0.0034,
+      "step": 20180
+    },
+    {
+      "epoch": 4.586590909090909,
+      "grad_norm": 0.8075526156118578,
+      "learning_rate": 2.098945908457578e-08,
+      "loss": 0.0041,
+      "step": 20181
+    },
+    {
+      "epoch": 4.586818181818182,
+      "grad_norm": 2.2147434276115705,
+      "learning_rate": 2.0966529880020318e-08,
+      "loss": 0.0079,
+      "step": 20182
+    },
+    {
+      "epoch": 4.587045454545454,
+      "grad_norm": 1.4767779982746736,
+      "learning_rate": 2.094361299276114e-08,
+      "loss": 0.0071,
+      "step": 20183
+    },
+    {
+      "epoch": 4.587272727272727,
+      "grad_norm": 0.9729378620385266,
+      "learning_rate": 2.0920708423265502e-08,
+      "loss": 0.007,
+      "step": 20184
+    },
+    {
+      "epoch": 4.5875,
+      "grad_norm": 0.7472254074053024,
+      "learning_rate": 2.089781617200047e-08,
+      "loss": 0.0035,
+      "step": 20185
+    },
+    {
+      "epoch": 4.587727272727273,
+      "grad_norm": 1.330840590030063,
+      "learning_rate": 2.0874936239432957e-08,
+      "loss": 0.0036,
+      "step": 20186
+    },
+    {
+      "epoch": 4.587954545454545,
+      "grad_norm": 0.46217557478273674,
+      "learning_rate": 2.0852068626029398e-08,
+      "loss": 0.0025,
+      "step": 20187
+    },
+    {
+      "epoch": 4.588181818181818,
+      "grad_norm": 1.1411962192947238,
+      "learning_rate": 2.0829213332256154e-08,
+      "loss": 0.0044,
+      "step": 20188
+    },
+    {
+      "epoch": 4.58840909090909,
+      "grad_norm": 0.5411707692912947,
+      "learning_rate": 2.080637035857931e-08,
+      "loss": 0.005,
+      "step": 20189
+    },
+    {
+      "epoch": 4.588636363636364,
+      "grad_norm": 2.2434545769771477,
+      "learning_rate": 2.078353970546461e-08,
+      "loss": 0.0107,
+      "step": 20190
+    },
+    {
+      "epoch": 4.588863636363636,
+      "grad_norm": 0.8737136480392106,
+      "learning_rate": 2.0760721373377715e-08,
+      "loss": 0.0042,
+      "step": 20191
+    },
+    {
+      "epoch": 4.589090909090909,
+      "grad_norm": 1.317502447351545,
+      "learning_rate": 2.0737915362783814e-08,
+      "loss": 0.0059,
+      "step": 20192
+    },
+    {
+      "epoch": 4.589318181818182,
+      "grad_norm": 0.5243512776006692,
+      "learning_rate": 2.071512167414795e-08,
+      "loss": 0.0076,
+      "step": 20193
+    },
+    {
+      "epoch": 4.589545454545455,
+      "grad_norm": 0.834768117431828,
+      "learning_rate": 2.069234030793517e-08,
+      "loss": 0.003,
+      "step": 20194
+    },
+    {
+      "epoch": 4.589772727272727,
+      "grad_norm": 1.0039528333435668,
+      "learning_rate": 2.066957126460968e-08,
+      "loss": 0.0053,
+      "step": 20195
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 1.5496314587578321,
+      "learning_rate": 2.0646814544635978e-08,
+      "loss": 0.0101,
+      "step": 20196
+    },
+    {
+      "epoch": 4.590227272727272,
+      "grad_norm": 0.9999553009627375,
+      "learning_rate": 2.0624070148478128e-08,
+      "loss": 0.0056,
+      "step": 20197
+    },
+    {
+      "epoch": 4.590454545454546,
+      "grad_norm": 1.0413882648872532,
+      "learning_rate": 2.0601338076599865e-08,
+      "loss": 0.0042,
+      "step": 20198
+    },
+    {
+      "epoch": 4.590681818181818,
+      "grad_norm": 2.1492109809757816,
+      "learning_rate": 2.057861832946477e-08,
+      "loss": 0.0051,
+      "step": 20199
+    },
+    {
+      "epoch": 4.590909090909091,
+      "grad_norm": 0.46283281758437,
+      "learning_rate": 2.0555910907536018e-08,
+      "loss": 0.0012,
+      "step": 20200
+    },
+    {
+      "epoch": 4.591136363636363,
+      "grad_norm": 2.0028620841310976,
+      "learning_rate": 2.0533215811276846e-08,
+      "loss": 0.0188,
+      "step": 20201
+    },
+    {
+      "epoch": 4.591363636363637,
+      "grad_norm": 1.2832487675756352,
+      "learning_rate": 2.0510533041149874e-08,
+      "loss": 0.0047,
+      "step": 20202
+    },
+    {
+      "epoch": 4.591590909090909,
+      "grad_norm": 0.6218838795968117,
+      "learning_rate": 2.0487862597617786e-08,
+      "loss": 0.0036,
+      "step": 20203
+    },
+    {
+      "epoch": 4.591818181818182,
+      "grad_norm": 1.7086301000963595,
+      "learning_rate": 2.0465204481142782e-08,
+      "loss": 0.0114,
+      "step": 20204
+    },
+    {
+      "epoch": 4.592045454545454,
+      "grad_norm": 1.651541214836777,
+      "learning_rate": 2.044255869218692e-08,
+      "loss": 0.0051,
+      "step": 20205
+    },
+    {
+      "epoch": 4.592272727272727,
+      "grad_norm": 0.97027937775502,
+      "learning_rate": 2.0419925231211986e-08,
+      "loss": 0.0036,
+      "step": 20206
+    },
+    {
+      "epoch": 4.5925,
+      "grad_norm": 0.9031306354468518,
+      "learning_rate": 2.0397304098679628e-08,
+      "loss": 0.0056,
+      "step": 20207
+    },
+    {
+      "epoch": 4.592727272727273,
+      "grad_norm": 0.41762613860137926,
+      "learning_rate": 2.0374695295050928e-08,
+      "loss": 0.0018,
+      "step": 20208
+    },
+    {
+      "epoch": 4.592954545454545,
+      "grad_norm": 0.8123889180421361,
+      "learning_rate": 2.035209882078705e-08,
+      "loss": 0.0022,
+      "step": 20209
+    },
+    {
+      "epoch": 4.593181818181818,
+      "grad_norm": 0.6366472354681536,
+      "learning_rate": 2.0329514676348736e-08,
+      "loss": 0.0009,
+      "step": 20210
+    },
+    {
+      "epoch": 4.593409090909091,
+      "grad_norm": 1.7897651403473978,
+      "learning_rate": 2.0306942862196586e-08,
+      "loss": 0.0072,
+      "step": 20211
+    },
+    {
+      "epoch": 4.593636363636364,
+      "grad_norm": 4.287752735152976,
+      "learning_rate": 2.0284383378790726e-08,
+      "loss": 0.0137,
+      "step": 20212
+    },
+    {
+      "epoch": 4.593863636363636,
+      "grad_norm": 1.9014659840033807,
+      "learning_rate": 2.0261836226591266e-08,
+      "loss": 0.028,
+      "step": 20213
+    },
+    {
+      "epoch": 4.594090909090909,
+      "grad_norm": 1.4180752449867098,
+      "learning_rate": 2.0239301406058053e-08,
+      "loss": 0.0031,
+      "step": 20214
+    },
+    {
+      "epoch": 4.594318181818182,
+      "grad_norm": 0.503363850333403,
+      "learning_rate": 2.0216778917650505e-08,
+      "loss": 0.0037,
+      "step": 20215
+    },
+    {
+      "epoch": 4.594545454545455,
+      "grad_norm": 1.0208091471058234,
+      "learning_rate": 2.019426876182791e-08,
+      "loss": 0.0047,
+      "step": 20216
+    },
+    {
+      "epoch": 4.594772727272727,
+      "grad_norm": 1.7124649787325266,
+      "learning_rate": 2.0171770939049345e-08,
+      "loss": 0.0127,
+      "step": 20217
+    },
+    {
+      "epoch": 4.595,
+      "grad_norm": 1.9574948039709132,
+      "learning_rate": 2.014928544977354e-08,
+      "loss": 0.0119,
+      "step": 20218
+    },
+    {
+      "epoch": 4.595227272727273,
+      "grad_norm": 1.3216029631817632,
+      "learning_rate": 2.012681229445909e-08,
+      "loss": 0.0023,
+      "step": 20219
+    },
+    {
+      "epoch": 4.595454545454546,
+      "grad_norm": 0.5853144676159775,
+      "learning_rate": 2.0104351473564168e-08,
+      "loss": 0.0015,
+      "step": 20220
+    },
+    {
+      "epoch": 4.595681818181818,
+      "grad_norm": 0.40694147671592157,
+      "learning_rate": 2.008190298754681e-08,
+      "loss": 0.0012,
+      "step": 20221
+    },
+    {
+      "epoch": 4.595909090909091,
+      "grad_norm": 0.4901384048427826,
+      "learning_rate": 2.005946683686477e-08,
+      "loss": 0.0015,
+      "step": 20222
+    },
+    {
+      "epoch": 4.596136363636363,
+      "grad_norm": 0.5793376558331029,
+      "learning_rate": 2.003704302197554e-08,
+      "loss": 0.0021,
+      "step": 20223
+    },
+    {
+      "epoch": 4.596363636363637,
+      "grad_norm": 1.151093452273118,
+      "learning_rate": 2.0014631543336523e-08,
+      "loss": 0.007,
+      "step": 20224
+    },
+    {
+      "epoch": 4.596590909090909,
+      "grad_norm": 0.8440653127314828,
+      "learning_rate": 1.9992232401404575e-08,
+      "loss": 0.0023,
+      "step": 20225
+    },
+    {
+      "epoch": 4.596818181818182,
+      "grad_norm": 0.4049479135676388,
+      "learning_rate": 1.996984559663649e-08,
+      "loss": 0.0019,
+      "step": 20226
+    },
+    {
+      "epoch": 4.597045454545454,
+      "grad_norm": 0.6425065861593034,
+      "learning_rate": 1.9947471129488777e-08,
+      "loss": 0.0039,
+      "step": 20227
+    },
+    {
+      "epoch": 4.597272727272728,
+      "grad_norm": 1.9814780010764816,
+      "learning_rate": 1.9925109000417733e-08,
+      "loss": 0.0075,
+      "step": 20228
+    },
+    {
+      "epoch": 4.5975,
+      "grad_norm": 0.44996599540322635,
+      "learning_rate": 1.9902759209879315e-08,
+      "loss": 0.0022,
+      "step": 20229
+    },
+    {
+      "epoch": 4.597727272727273,
+      "grad_norm": 0.06904805425466769,
+      "learning_rate": 1.988042175832927e-08,
+      "loss": 0.0003,
+      "step": 20230
+    },
+    {
+      "epoch": 4.597954545454545,
+      "grad_norm": 0.5874754400915659,
+      "learning_rate": 1.9858096646223136e-08,
+      "loss": 0.0028,
+      "step": 20231
+    },
+    {
+      "epoch": 4.598181818181818,
+      "grad_norm": 2.0568237004059156,
+      "learning_rate": 1.983578387401604e-08,
+      "loss": 0.0048,
+      "step": 20232
+    },
+    {
+      "epoch": 4.598409090909091,
+      "grad_norm": 1.4680201085003484,
+      "learning_rate": 1.981348344216323e-08,
+      "loss": 0.0032,
+      "step": 20233
+    },
+    {
+      "epoch": 4.598636363636364,
+      "grad_norm": 1.1511403328397927,
+      "learning_rate": 1.9791195351119215e-08,
+      "loss": 0.0088,
+      "step": 20234
+    },
+    {
+      "epoch": 4.598863636363636,
+      "grad_norm": 0.2772690294386796,
+      "learning_rate": 1.976891960133863e-08,
+      "loss": 0.001,
+      "step": 20235
+    },
+    {
+      "epoch": 4.59909090909091,
+      "grad_norm": 2.160174316513933,
+      "learning_rate": 1.9746656193275552e-08,
+      "loss": 0.014,
+      "step": 20236
+    },
+    {
+      "epoch": 4.599318181818182,
+      "grad_norm": 2.0464917590726364,
+      "learning_rate": 1.9724405127384132e-08,
+      "loss": 0.0107,
+      "step": 20237
+    },
+    {
+      "epoch": 4.599545454545455,
+      "grad_norm": 1.3958444797031,
+      "learning_rate": 1.9702166404118178e-08,
+      "loss": 0.0045,
+      "step": 20238
+    },
+    {
+      "epoch": 4.599772727272727,
+      "grad_norm": 0.782665704028619,
+      "learning_rate": 1.9679940023930867e-08,
+      "loss": 0.0043,
+      "step": 20239
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 0.7738888516654694,
+      "learning_rate": 1.965772598727572e-08,
+      "loss": 0.0067,
+      "step": 20240
+    },
+    {
+      "epoch": 4.600227272727273,
+      "grad_norm": 0.15995341001409336,
+      "learning_rate": 1.9635524294605578e-08,
+      "loss": 0.0006,
+      "step": 20241
+    },
+    {
+      "epoch": 4.600454545454546,
+      "grad_norm": 0.9008867611270605,
+      "learning_rate": 1.9613334946373268e-08,
+      "loss": 0.0095,
+      "step": 20242
+    },
+    {
+      "epoch": 4.600681818181818,
+      "grad_norm": 0.5059982690929605,
+      "learning_rate": 1.959115794303114e-08,
+      "loss": 0.0032,
+      "step": 20243
+    },
+    {
+      "epoch": 4.600909090909091,
+      "grad_norm": 1.8139363838786193,
+      "learning_rate": 1.9568993285031536e-08,
+      "loss": 0.0067,
+      "step": 20244
+    },
+    {
+      "epoch": 4.601136363636364,
+      "grad_norm": 1.3922664457659124,
+      "learning_rate": 1.9546840972826318e-08,
+      "loss": 0.0144,
+      "step": 20245
+    },
+    {
+      "epoch": 4.6013636363636365,
+      "grad_norm": 2.101112004448128,
+      "learning_rate": 1.9524701006867348e-08,
+      "loss": 0.0047,
+      "step": 20246
+    },
+    {
+      "epoch": 4.601590909090909,
+      "grad_norm": 2.2039374702414145,
+      "learning_rate": 1.9502573387606e-08,
+      "loss": 0.0124,
+      "step": 20247
+    },
+    {
+      "epoch": 4.601818181818182,
+      "grad_norm": 0.35223490530701,
+      "learning_rate": 1.948045811549358e-08,
+      "loss": 0.0013,
+      "step": 20248
+    },
+    {
+      "epoch": 4.602045454545454,
+      "grad_norm": 0.265767323748949,
+      "learning_rate": 1.945835519098098e-08,
+      "loss": 0.0009,
+      "step": 20249
+    },
+    {
+      "epoch": 4.6022727272727275,
+      "grad_norm": 0.9404671933505102,
+      "learning_rate": 1.9436264614518877e-08,
+      "loss": 0.0032,
+      "step": 20250
+    },
+    {
+      "epoch": 4.6025,
+      "grad_norm": 0.2765237804972594,
+      "learning_rate": 1.9414186386557955e-08,
+      "loss": 0.0008,
+      "step": 20251
+    },
+    {
+      "epoch": 4.602727272727273,
+      "grad_norm": 1.7109329799650546,
+      "learning_rate": 1.9392120507548198e-08,
+      "loss": 0.006,
+      "step": 20252
+    },
+    {
+      "epoch": 4.602954545454545,
+      "grad_norm": 1.5635628854574182,
+      "learning_rate": 1.9370066977939598e-08,
+      "loss": 0.0103,
+      "step": 20253
+    },
+    {
+      "epoch": 4.6031818181818185,
+      "grad_norm": 1.4493029019666426,
+      "learning_rate": 1.934802579818193e-08,
+      "loss": 0.0033,
+      "step": 20254
+    },
+    {
+      "epoch": 4.603409090909091,
+      "grad_norm": 1.6328379390225956,
+      "learning_rate": 1.9325996968724698e-08,
+      "loss": 0.0082,
+      "step": 20255
+    },
+    {
+      "epoch": 4.6036363636363635,
+      "grad_norm": 0.6319977133970687,
+      "learning_rate": 1.9303980490016986e-08,
+      "loss": 0.0026,
+      "step": 20256
+    },
+    {
+      "epoch": 4.603863636363636,
+      "grad_norm": 1.732186433781712,
+      "learning_rate": 1.928197636250781e-08,
+      "loss": 0.007,
+      "step": 20257
+    },
+    {
+      "epoch": 4.6040909090909095,
+      "grad_norm": 0.8124725206433352,
+      "learning_rate": 1.9259984586645907e-08,
+      "loss": 0.0043,
+      "step": 20258
+    },
+    {
+      "epoch": 4.604318181818182,
+      "grad_norm": 1.036471104799119,
+      "learning_rate": 1.9238005162879673e-08,
+      "loss": 0.005,
+      "step": 20259
+    },
+    {
+      "epoch": 4.6045454545454545,
+      "grad_norm": 1.2039610450230807,
+      "learning_rate": 1.9216038091657356e-08,
+      "loss": 0.003,
+      "step": 20260
+    },
+    {
+      "epoch": 4.604772727272727,
+      "grad_norm": 1.3627263874184612,
+      "learning_rate": 1.9194083373426793e-08,
+      "loss": 0.0028,
+      "step": 20261
+    },
+    {
+      "epoch": 4.605,
+      "grad_norm": 0.551758554707767,
+      "learning_rate": 1.917214100863568e-08,
+      "loss": 0.0026,
+      "step": 20262
+    },
+    {
+      "epoch": 4.605227272727273,
+      "grad_norm": 0.3330417913348629,
+      "learning_rate": 1.915021099773165e-08,
+      "loss": 0.0033,
+      "step": 20263
+    },
+    {
+      "epoch": 4.6054545454545455,
+      "grad_norm": 0.6409320381649694,
+      "learning_rate": 1.912829334116184e-08,
+      "loss": 0.002,
+      "step": 20264
+    },
+    {
+      "epoch": 4.605681818181818,
+      "grad_norm": 0.7369006175663797,
+      "learning_rate": 1.910638803937305e-08,
+      "loss": 0.0064,
+      "step": 20265
+    },
+    {
+      "epoch": 4.6059090909090905,
+      "grad_norm": 1.4357766177987323,
+      "learning_rate": 1.9084495092812073e-08,
+      "loss": 0.013,
+      "step": 20266
+    },
+    {
+      "epoch": 4.606136363636364,
+      "grad_norm": 1.1374830689178834,
+      "learning_rate": 1.9062614501925287e-08,
+      "loss": 0.0064,
+      "step": 20267
+    },
+    {
+      "epoch": 4.606363636363636,
+      "grad_norm": 0.6471561301205899,
+      "learning_rate": 1.9040746267159002e-08,
+      "loss": 0.0016,
+      "step": 20268
+    },
+    {
+      "epoch": 4.606590909090909,
+      "grad_norm": 0.7756024339584915,
+      "learning_rate": 1.9018890388958973e-08,
+      "loss": 0.0018,
+      "step": 20269
+    },
+    {
+      "epoch": 4.6068181818181815,
+      "grad_norm": 2.9718102180667043,
+      "learning_rate": 1.8997046867770957e-08,
+      "loss": 0.0095,
+      "step": 20270
+    },
+    {
+      "epoch": 4.607045454545455,
+      "grad_norm": 0.5987634716293896,
+      "learning_rate": 1.8975215704040358e-08,
+      "loss": 0.002,
+      "step": 20271
+    },
+    {
+      "epoch": 4.607272727272727,
+      "grad_norm": 1.9268987240277544,
+      "learning_rate": 1.8953396898212446e-08,
+      "loss": 0.0166,
+      "step": 20272
+    },
+    {
+      "epoch": 4.6075,
+      "grad_norm": 0.3558224288279283,
+      "learning_rate": 1.8931590450732078e-08,
+      "loss": 0.0007,
+      "step": 20273
+    },
+    {
+      "epoch": 4.6077272727272724,
+      "grad_norm": 0.6656777681712286,
+      "learning_rate": 1.8909796362043888e-08,
+      "loss": 0.0037,
+      "step": 20274
+    },
+    {
+      "epoch": 4.607954545454545,
+      "grad_norm": 1.0064945111620858,
+      "learning_rate": 1.888801463259232e-08,
+      "loss": 0.0029,
+      "step": 20275
+    },
+    {
+      "epoch": 4.608181818181818,
+      "grad_norm": 3.366068435140205,
+      "learning_rate": 1.8866245262821602e-08,
+      "loss": 0.012,
+      "step": 20276
+    },
+    {
+      "epoch": 4.608409090909091,
+      "grad_norm": 2.4933387742214674,
+      "learning_rate": 1.8844488253175544e-08,
+      "loss": 0.0151,
+      "step": 20277
+    },
+    {
+      "epoch": 4.608636363636363,
+      "grad_norm": 0.7659746666354575,
+      "learning_rate": 1.8822743604097886e-08,
+      "loss": 0.0066,
+      "step": 20278
+    },
+    {
+      "epoch": 4.608863636363637,
+      "grad_norm": 0.5430109880433233,
+      "learning_rate": 1.8801011316032027e-08,
+      "loss": 0.0023,
+      "step": 20279
+    },
+    {
+      "epoch": 4.609090909090909,
+      "grad_norm": 0.7016056654653751,
+      "learning_rate": 1.8779291389421153e-08,
+      "loss": 0.0043,
+      "step": 20280
+    },
+    {
+      "epoch": 4.609318181818182,
+      "grad_norm": 1.855285834884007,
+      "learning_rate": 1.8757583824708105e-08,
+      "loss": 0.0099,
+      "step": 20281
+    },
+    {
+      "epoch": 4.609545454545454,
+      "grad_norm": 0.404802140972389,
+      "learning_rate": 1.873588862233558e-08,
+      "loss": 0.0024,
+      "step": 20282
+    },
+    {
+      "epoch": 4.609772727272727,
+      "grad_norm": 0.2989126310212012,
+      "learning_rate": 1.8714205782746006e-08,
+      "loss": 0.0015,
+      "step": 20283
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 0.24458525566164593,
+      "learning_rate": 1.869253530638146e-08,
+      "loss": 0.0006,
+      "step": 20284
+    },
+    {
+      "epoch": 4.610227272727273,
+      "grad_norm": 0.8805330825125967,
+      "learning_rate": 1.867087719368388e-08,
+      "loss": 0.0032,
+      "step": 20285
+    },
+    {
+      "epoch": 4.610454545454545,
+      "grad_norm": 0.8385580827674878,
+      "learning_rate": 1.8649231445094924e-08,
+      "loss": 0.0047,
+      "step": 20286
+    },
+    {
+      "epoch": 4.610681818181818,
+      "grad_norm": 0.3351780288765914,
+      "learning_rate": 1.8627598061055975e-08,
+      "loss": 0.0008,
+      "step": 20287
+    },
+    {
+      "epoch": 4.610909090909091,
+      "grad_norm": 0.8485531733208366,
+      "learning_rate": 1.8605977042008143e-08,
+      "loss": 0.0072,
+      "step": 20288
+    },
+    {
+      "epoch": 4.611136363636364,
+      "grad_norm": 1.6336057004706195,
+      "learning_rate": 1.8584368388392394e-08,
+      "loss": 0.0064,
+      "step": 20289
+    },
+    {
+      "epoch": 4.611363636363636,
+      "grad_norm": 1.7615891144108193,
+      "learning_rate": 1.8562772100649344e-08,
+      "loss": 0.0123,
+      "step": 20290
+    },
+    {
+      "epoch": 4.611590909090909,
+      "grad_norm": 0.7053306953992126,
+      "learning_rate": 1.8541188179219265e-08,
+      "loss": 0.0036,
+      "step": 20291
+    },
+    {
+      "epoch": 4.611818181818181,
+      "grad_norm": 0.886450644210592,
+      "learning_rate": 1.8519616624542435e-08,
+      "loss": 0.0051,
+      "step": 20292
+    },
+    {
+      "epoch": 4.612045454545455,
+      "grad_norm": 1.183893440352902,
+      "learning_rate": 1.849805743705864e-08,
+      "loss": 0.0025,
+      "step": 20293
+    },
+    {
+      "epoch": 4.612272727272727,
+      "grad_norm": 2.999160141801076,
+      "learning_rate": 1.8476510617207594e-08,
+      "loss": 0.0146,
+      "step": 20294
+    },
+    {
+      "epoch": 4.6125,
+      "grad_norm": 0.8878863046451919,
+      "learning_rate": 1.845497616542874e-08,
+      "loss": 0.0046,
+      "step": 20295
+    },
+    {
+      "epoch": 4.612727272727272,
+      "grad_norm": 1.0357531099483874,
+      "learning_rate": 1.843345408216096e-08,
+      "loss": 0.0042,
+      "step": 20296
+    },
+    {
+      "epoch": 4.612954545454546,
+      "grad_norm": 0.7590673420274321,
+      "learning_rate": 1.8411944367843286e-08,
+      "loss": 0.0062,
+      "step": 20297
+    },
+    {
+      "epoch": 4.613181818181818,
+      "grad_norm": 3.074049406598694,
+      "learning_rate": 1.8390447022914383e-08,
+      "loss": 0.0178,
+      "step": 20298
+    },
+    {
+      "epoch": 4.613409090909091,
+      "grad_norm": 1.0110007830922416,
+      "learning_rate": 1.8368962047812523e-08,
+      "loss": 0.008,
+      "step": 20299
+    },
+    {
+      "epoch": 4.613636363636363,
+      "grad_norm": 1.0382032357066495,
+      "learning_rate": 1.8347489442975817e-08,
+      "loss": 0.0017,
+      "step": 20300
+    },
+    {
+      "epoch": 4.613863636363637,
+      "grad_norm": 1.9495009311811606,
+      "learning_rate": 1.8326029208842188e-08,
+      "loss": 0.0086,
+      "step": 20301
+    },
+    {
+      "epoch": 4.614090909090909,
+      "grad_norm": 3.285217627572585,
+      "learning_rate": 1.83045813458492e-08,
+      "loss": 0.0084,
+      "step": 20302
+    },
+    {
+      "epoch": 4.614318181818182,
+      "grad_norm": 0.6727462855523674,
+      "learning_rate": 1.8283145854434277e-08,
+      "loss": 0.0038,
+      "step": 20303
+    },
+    {
+      "epoch": 4.614545454545454,
+      "grad_norm": 0.5422685614074438,
+      "learning_rate": 1.82617227350345e-08,
+      "loss": 0.0015,
+      "step": 20304
+    },
+    {
+      "epoch": 4.614772727272728,
+      "grad_norm": 0.9637599819958073,
+      "learning_rate": 1.824031198808668e-08,
+      "loss": 0.0021,
+      "step": 20305
+    },
+    {
+      "epoch": 4.615,
+      "grad_norm": 0.6681351902777157,
+      "learning_rate": 1.8218913614027473e-08,
+      "loss": 0.0035,
+      "step": 20306
+    },
+    {
+      "epoch": 4.615227272727273,
+      "grad_norm": 0.9273259934454944,
+      "learning_rate": 1.819752761329313e-08,
+      "loss": 0.0048,
+      "step": 20307
+    },
+    {
+      "epoch": 4.615454545454545,
+      "grad_norm": 0.6626990735388141,
+      "learning_rate": 1.81761539863199e-08,
+      "loss": 0.0027,
+      "step": 20308
+    },
+    {
+      "epoch": 4.615681818181818,
+      "grad_norm": 1.5869635713732861,
+      "learning_rate": 1.8154792733543482e-08,
+      "loss": 0.0066,
+      "step": 20309
+    },
+    {
+      "epoch": 4.615909090909091,
+      "grad_norm": 1.477611690235562,
+      "learning_rate": 1.8133443855399628e-08,
+      "loss": 0.0047,
+      "step": 20310
+    },
+    {
+      "epoch": 4.616136363636364,
+      "grad_norm": 0.5268787422808833,
+      "learning_rate": 1.8112107352323478e-08,
+      "loss": 0.002,
+      "step": 20311
+    },
+    {
+      "epoch": 4.616363636363636,
+      "grad_norm": 0.5622455073583327,
+      "learning_rate": 1.8090783224750385e-08,
+      "loss": 0.0023,
+      "step": 20312
+    },
+    {
+      "epoch": 4.616590909090909,
+      "grad_norm": 0.6797392154514701,
+      "learning_rate": 1.8069471473114927e-08,
+      "loss": 0.0022,
+      "step": 20313
+    },
+    {
+      "epoch": 4.616818181818182,
+      "grad_norm": 0.8622966393142705,
+      "learning_rate": 1.8048172097851826e-08,
+      "loss": 0.0038,
+      "step": 20314
+    },
+    {
+      "epoch": 4.617045454545455,
+      "grad_norm": 1.4175929998395989,
+      "learning_rate": 1.802688509939532e-08,
+      "loss": 0.0069,
+      "step": 20315
+    },
+    {
+      "epoch": 4.617272727272727,
+      "grad_norm": 3.4763266352387987,
+      "learning_rate": 1.8005610478179643e-08,
+      "loss": 0.0194,
+      "step": 20316
+    },
+    {
+      "epoch": 4.6175,
+      "grad_norm": 1.843709449621031,
+      "learning_rate": 1.7984348234638545e-08,
+      "loss": 0.0118,
+      "step": 20317
+    },
+    {
+      "epoch": 4.617727272727272,
+      "grad_norm": 2.8056893512258405,
+      "learning_rate": 1.7963098369205497e-08,
+      "loss": 0.0182,
+      "step": 20318
+    },
+    {
+      "epoch": 4.617954545454546,
+      "grad_norm": 0.9942236537192816,
+      "learning_rate": 1.794186088231391e-08,
+      "loss": 0.0044,
+      "step": 20319
+    },
+    {
+      "epoch": 4.618181818181818,
+      "grad_norm": 0.3162029511568735,
+      "learning_rate": 1.79206357743969e-08,
+      "loss": 0.0013,
+      "step": 20320
+    },
+    {
+      "epoch": 4.618409090909091,
+      "grad_norm": 1.3106898086146668,
+      "learning_rate": 1.7899423045887182e-08,
+      "loss": 0.0045,
+      "step": 20321
+    },
+    {
+      "epoch": 4.618636363636364,
+      "grad_norm": 0.42715244419804027,
+      "learning_rate": 1.7878222697217397e-08,
+      "loss": 0.0023,
+      "step": 20322
+    },
+    {
+      "epoch": 4.618863636363637,
+      "grad_norm": 1.753704684015323,
+      "learning_rate": 1.7857034728819835e-08,
+      "loss": 0.0074,
+      "step": 20323
+    },
+    {
+      "epoch": 4.619090909090909,
+      "grad_norm": 0.3440400098200101,
+      "learning_rate": 1.783585914112651e-08,
+      "loss": 0.0015,
+      "step": 20324
+    },
+    {
+      "epoch": 4.619318181818182,
+      "grad_norm": 0.6120267262539397,
+      "learning_rate": 1.7814695934569302e-08,
+      "loss": 0.0044,
+      "step": 20325
+    },
+    {
+      "epoch": 4.619545454545454,
+      "grad_norm": 1.1631306030731146,
+      "learning_rate": 1.7793545109579813e-08,
+      "loss": 0.0038,
+      "step": 20326
+    },
+    {
+      "epoch": 4.619772727272728,
+      "grad_norm": 0.45407300997248,
+      "learning_rate": 1.777240666658922e-08,
+      "loss": 0.0012,
+      "step": 20327
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 3.0635853504357446,
+      "learning_rate": 1.775128060602864e-08,
+      "loss": 0.0117,
+      "step": 20328
+    },
+    {
+      "epoch": 4.620227272727273,
+      "grad_norm": 0.3382068308588993,
+      "learning_rate": 1.7730166928328768e-08,
+      "loss": 0.001,
+      "step": 20329
+    },
+    {
+      "epoch": 4.620454545454545,
+      "grad_norm": 0.24276624327268254,
+      "learning_rate": 1.7709065633920372e-08,
+      "loss": 0.0009,
+      "step": 20330
+    },
+    {
+      "epoch": 4.620681818181819,
+      "grad_norm": 0.5119168412779894,
+      "learning_rate": 1.7687976723233525e-08,
+      "loss": 0.0025,
+      "step": 20331
+    },
+    {
+      "epoch": 4.620909090909091,
+      "grad_norm": 1.6701621996057396,
+      "learning_rate": 1.76669001966983e-08,
+      "loss": 0.0035,
+      "step": 20332
+    },
+    {
+      "epoch": 4.621136363636364,
+      "grad_norm": 1.2686359061307109,
+      "learning_rate": 1.7645836054744555e-08,
+      "loss": 0.0055,
+      "step": 20333
+    },
+    {
+      "epoch": 4.621363636363636,
+      "grad_norm": 1.438886422086256,
+      "learning_rate": 1.7624784297801886e-08,
+      "loss": 0.0055,
+      "step": 20334
+    },
+    {
+      "epoch": 4.621590909090909,
+      "grad_norm": 0.380463151358392,
+      "learning_rate": 1.760374492629939e-08,
+      "loss": 0.0013,
+      "step": 20335
+    },
+    {
+      "epoch": 4.621818181818182,
+      "grad_norm": 2.5129126011517675,
+      "learning_rate": 1.7582717940666238e-08,
+      "loss": 0.0087,
+      "step": 20336
+    },
+    {
+      "epoch": 4.622045454545455,
+      "grad_norm": 2.7959440625116034,
+      "learning_rate": 1.7561703341331186e-08,
+      "loss": 0.0094,
+      "step": 20337
+    },
+    {
+      "epoch": 4.622272727272727,
+      "grad_norm": 2.2775336598561666,
+      "learning_rate": 1.7540701128722636e-08,
+      "loss": 0.0163,
+      "step": 20338
+    },
+    {
+      "epoch": 4.6225000000000005,
+      "grad_norm": 3.0941735856016535,
+      "learning_rate": 1.751971130326907e-08,
+      "loss": 0.0168,
+      "step": 20339
+    },
+    {
+      "epoch": 4.622727272727273,
+      "grad_norm": 1.5056867002171712,
+      "learning_rate": 1.749873386539834e-08,
+      "loss": 0.0244,
+      "step": 20340
+    },
+    {
+      "epoch": 4.6229545454545455,
+      "grad_norm": 0.3699812348122988,
+      "learning_rate": 1.7477768815538295e-08,
+      "loss": 0.0006,
+      "step": 20341
+    },
+    {
+      "epoch": 4.623181818181818,
+      "grad_norm": 1.252663747149098,
+      "learning_rate": 1.7456816154116374e-08,
+      "loss": 0.0064,
+      "step": 20342
+    },
+    {
+      "epoch": 4.623409090909091,
+      "grad_norm": 1.3029538528238989,
+      "learning_rate": 1.743587588156001e-08,
+      "loss": 0.0045,
+      "step": 20343
+    },
+    {
+      "epoch": 4.623636363636364,
+      "grad_norm": 1.127259610630913,
+      "learning_rate": 1.7414947998296015e-08,
+      "loss": 0.0083,
+      "step": 20344
+    },
+    {
+      "epoch": 4.6238636363636365,
+      "grad_norm": 0.9396516944841272,
+      "learning_rate": 1.7394032504751205e-08,
+      "loss": 0.0044,
+      "step": 20345
+    },
+    {
+      "epoch": 4.624090909090909,
+      "grad_norm": 0.27477853996662727,
+      "learning_rate": 1.7373129401352107e-08,
+      "loss": 0.0006,
+      "step": 20346
+    },
+    {
+      "epoch": 4.6243181818181816,
+      "grad_norm": 2.813497815544435,
+      "learning_rate": 1.735223868852491e-08,
+      "loss": 0.0261,
+      "step": 20347
+    },
+    {
+      "epoch": 4.624545454545455,
+      "grad_norm": 2.7993642429571297,
+      "learning_rate": 1.7331360366695803e-08,
+      "loss": 0.0142,
+      "step": 20348
+    },
+    {
+      "epoch": 4.6247727272727275,
+      "grad_norm": 0.7311923437679234,
+      "learning_rate": 1.731049443629035e-08,
+      "loss": 0.003,
+      "step": 20349
+    },
+    {
+      "epoch": 4.625,
+      "grad_norm": 0.3212091680831667,
+      "learning_rate": 1.7289640897734036e-08,
+      "loss": 0.0017,
+      "step": 20350
+    },
+    {
+      "epoch": 4.6252272727272725,
+      "grad_norm": 0.22495073620878286,
+      "learning_rate": 1.7268799751452154e-08,
+      "loss": 0.0005,
+      "step": 20351
+    },
+    {
+      "epoch": 4.625454545454545,
+      "grad_norm": 1.6150866848682608,
+      "learning_rate": 1.7247970997869706e-08,
+      "loss": 0.014,
+      "step": 20352
+    },
+    {
+      "epoch": 4.6256818181818184,
+      "grad_norm": 0.9838780356263643,
+      "learning_rate": 1.7227154637411356e-08,
+      "loss": 0.0034,
+      "step": 20353
+    },
+    {
+      "epoch": 4.625909090909091,
+      "grad_norm": 0.7342315883849917,
+      "learning_rate": 1.7206350670501693e-08,
+      "loss": 0.0027,
+      "step": 20354
+    },
+    {
+      "epoch": 4.6261363636363635,
+      "grad_norm": 0.6731272166022619,
+      "learning_rate": 1.7185559097564892e-08,
+      "loss": 0.0019,
+      "step": 20355
+    },
+    {
+      "epoch": 4.626363636363636,
+      "grad_norm": 2.01354958646277,
+      "learning_rate": 1.7164779919024988e-08,
+      "loss": 0.0061,
+      "step": 20356
+    },
+    {
+      "epoch": 4.626590909090909,
+      "grad_norm": 0.7659962638270925,
+      "learning_rate": 1.71440131353056e-08,
+      "loss": 0.0035,
+      "step": 20357
+    },
+    {
+      "epoch": 4.626818181818182,
+      "grad_norm": 0.9351733189636556,
+      "learning_rate": 1.7123258746830283e-08,
+      "loss": 0.0058,
+      "step": 20358
+    },
+    {
+      "epoch": 4.6270454545454545,
+      "grad_norm": 2.3718692062742575,
+      "learning_rate": 1.7102516754022164e-08,
+      "loss": 0.0139,
+      "step": 20359
+    },
+    {
+      "epoch": 4.627272727272727,
+      "grad_norm": 0.6849200036305416,
+      "learning_rate": 1.708178715730424e-08,
+      "loss": 0.0035,
+      "step": 20360
+    },
+    {
+      "epoch": 4.6274999999999995,
+      "grad_norm": 0.7742736925999659,
+      "learning_rate": 1.706106995709937e-08,
+      "loss": 0.0039,
+      "step": 20361
+    },
+    {
+      "epoch": 4.627727272727273,
+      "grad_norm": 0.7893487337880416,
+      "learning_rate": 1.7040365153829786e-08,
+      "loss": 0.0014,
+      "step": 20362
+    },
+    {
+      "epoch": 4.627954545454545,
+      "grad_norm": 0.7435046601578098,
+      "learning_rate": 1.7019672747917854e-08,
+      "loss": 0.0041,
+      "step": 20363
+    },
+    {
+      "epoch": 4.628181818181818,
+      "grad_norm": 0.3581476400694386,
+      "learning_rate": 1.699899273978539e-08,
+      "loss": 0.0011,
+      "step": 20364
+    },
+    {
+      "epoch": 4.628409090909091,
+      "grad_norm": 6.7936794435584975,
+      "learning_rate": 1.6978325129854283e-08,
+      "loss": 0.0116,
+      "step": 20365
+    },
+    {
+      "epoch": 4.628636363636364,
+      "grad_norm": 0.29154283204401915,
+      "learning_rate": 1.695766991854586e-08,
+      "loss": 0.001,
+      "step": 20366
+    },
+    {
+      "epoch": 4.628863636363636,
+      "grad_norm": 0.3848004078844347,
+      "learning_rate": 1.693702710628131e-08,
+      "loss": 0.0012,
+      "step": 20367
+    },
+    {
+      "epoch": 4.629090909090909,
+      "grad_norm": 0.5636667741736369,
+      "learning_rate": 1.6916396693481552e-08,
+      "loss": 0.0053,
+      "step": 20368
+    },
+    {
+      "epoch": 4.629318181818181,
+      "grad_norm": 0.5724557588843338,
+      "learning_rate": 1.6895778680567355e-08,
+      "loss": 0.0015,
+      "step": 20369
+    },
+    {
+      "epoch": 4.629545454545455,
+      "grad_norm": 0.22678247437450383,
+      "learning_rate": 1.6875173067959147e-08,
+      "loss": 0.0008,
+      "step": 20370
+    },
+    {
+      "epoch": 4.629772727272727,
+      "grad_norm": 1.1025366430177086,
+      "learning_rate": 1.685457985607708e-08,
+      "loss": 0.004,
+      "step": 20371
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 0.5275706587623399,
+      "learning_rate": 1.6833999045341095e-08,
+      "loss": 0.0012,
+      "step": 20372
+    },
+    {
+      "epoch": 4.630227272727272,
+      "grad_norm": 0.8284641510793103,
+      "learning_rate": 1.681343063617086e-08,
+      "loss": 0.0025,
+      "step": 20373
+    },
+    {
+      "epoch": 4.630454545454546,
+      "grad_norm": 1.2208678446323011,
+      "learning_rate": 1.679287462898589e-08,
+      "loss": 0.0119,
+      "step": 20374
+    },
+    {
+      "epoch": 4.630681818181818,
+      "grad_norm": 1.7009774871093621,
+      "learning_rate": 1.6772331024205236e-08,
+      "loss": 0.0084,
+      "step": 20375
+    },
+    {
+      "epoch": 4.630909090909091,
+      "grad_norm": 0.610347356815404,
+      "learning_rate": 1.6751799822247865e-08,
+      "loss": 0.0034,
+      "step": 20376
+    },
+    {
+      "epoch": 4.631136363636363,
+      "grad_norm": 0.5704663101353973,
+      "learning_rate": 1.67312810235324e-08,
+      "loss": 0.0023,
+      "step": 20377
+    },
+    {
+      "epoch": 4.631363636363636,
+      "grad_norm": 0.21844292039254143,
+      "learning_rate": 1.6710774628477394e-08,
+      "loss": 0.0005,
+      "step": 20378
+    },
+    {
+      "epoch": 4.631590909090909,
+      "grad_norm": 0.9030603091068,
+      "learning_rate": 1.6690280637500848e-08,
+      "loss": 0.0032,
+      "step": 20379
+    },
+    {
+      "epoch": 4.631818181818182,
+      "grad_norm": 0.8173104187891981,
+      "learning_rate": 1.6669799051020757e-08,
+      "loss": 0.0041,
+      "step": 20380
+    },
+    {
+      "epoch": 4.632045454545454,
+      "grad_norm": 1.0455742686888723,
+      "learning_rate": 1.6649329869454708e-08,
+      "loss": 0.003,
+      "step": 20381
+    },
+    {
+      "epoch": 4.632272727272728,
+      "grad_norm": 0.4952089525534768,
+      "learning_rate": 1.6628873093220144e-08,
+      "loss": 0.0027,
+      "step": 20382
+    },
+    {
+      "epoch": 4.6325,
+      "grad_norm": 1.130583216893397,
+      "learning_rate": 1.660842872273423e-08,
+      "loss": 0.0048,
+      "step": 20383
+    },
+    {
+      "epoch": 4.632727272727273,
+      "grad_norm": 0.8906528551721525,
+      "learning_rate": 1.6587996758413923e-08,
+      "loss": 0.0106,
+      "step": 20384
+    },
+    {
+      "epoch": 4.632954545454545,
+      "grad_norm": 0.838607310464892,
+      "learning_rate": 1.6567577200675698e-08,
+      "loss": 0.0036,
+      "step": 20385
+    },
+    {
+      "epoch": 4.633181818181818,
+      "grad_norm": 0.7557548596423544,
+      "learning_rate": 1.6547170049936097e-08,
+      "loss": 0.0048,
+      "step": 20386
+    },
+    {
+      "epoch": 4.633409090909091,
+      "grad_norm": 0.8419330225881219,
+      "learning_rate": 1.6526775306611244e-08,
+      "loss": 0.0059,
+      "step": 20387
+    },
+    {
+      "epoch": 4.633636363636364,
+      "grad_norm": 0.444231719942697,
+      "learning_rate": 1.650639297111692e-08,
+      "loss": 0.0011,
+      "step": 20388
+    },
+    {
+      "epoch": 4.633863636363636,
+      "grad_norm": 0.8437815132896279,
+      "learning_rate": 1.648602304386883e-08,
+      "loss": 0.0043,
+      "step": 20389
+    },
+    {
+      "epoch": 4.634090909090909,
+      "grad_norm": 0.3514352308116961,
+      "learning_rate": 1.6465665525282336e-08,
+      "loss": 0.001,
+      "step": 20390
+    },
+    {
+      "epoch": 4.634318181818182,
+      "grad_norm": 3.0318465492326125,
+      "learning_rate": 1.6445320415772596e-08,
+      "loss": 0.0167,
+      "step": 20391
+    },
+    {
+      "epoch": 4.634545454545455,
+      "grad_norm": 1.6787906356016418,
+      "learning_rate": 1.6424987715754417e-08,
+      "loss": 0.0058,
+      "step": 20392
+    },
+    {
+      "epoch": 4.634772727272727,
+      "grad_norm": 2.3146445304856806,
+      "learning_rate": 1.6404667425642464e-08,
+      "loss": 0.0186,
+      "step": 20393
+    },
+    {
+      "epoch": 4.635,
+      "grad_norm": 0.38244061936058626,
+      "learning_rate": 1.638435954585113e-08,
+      "loss": 0.0034,
+      "step": 20394
+    },
+    {
+      "epoch": 4.635227272727272,
+      "grad_norm": 0.4159514327243404,
+      "learning_rate": 1.636406407679446e-08,
+      "loss": 0.0023,
+      "step": 20395
+    },
+    {
+      "epoch": 4.635454545454546,
+      "grad_norm": 1.4897129972446554,
+      "learning_rate": 1.6343781018886423e-08,
+      "loss": 0.0105,
+      "step": 20396
+    },
+    {
+      "epoch": 4.635681818181818,
+      "grad_norm": 1.273952991057156,
+      "learning_rate": 1.6323510372540443e-08,
+      "loss": 0.0038,
+      "step": 20397
+    },
+    {
+      "epoch": 4.635909090909091,
+      "grad_norm": 0.24542156479524407,
+      "learning_rate": 1.6303252138170008e-08,
+      "loss": 0.0009,
+      "step": 20398
+    },
+    {
+      "epoch": 4.636136363636363,
+      "grad_norm": 1.6222542352795302,
+      "learning_rate": 1.628300631618819e-08,
+      "loss": 0.0048,
+      "step": 20399
+    },
+    {
+      "epoch": 4.636363636363637,
+      "grad_norm": 3.142921738095657,
+      "learning_rate": 1.6262772907007854e-08,
+      "loss": 0.0141,
+      "step": 20400
+    },
+    {
+      "epoch": 4.636590909090909,
+      "grad_norm": 0.5042929965160852,
+      "learning_rate": 1.6242551911041584e-08,
+      "loss": 0.0023,
+      "step": 20401
+    },
+    {
+      "epoch": 4.636818181818182,
+      "grad_norm": 0.5022793665636126,
+      "learning_rate": 1.622234332870169e-08,
+      "loss": 0.0068,
+      "step": 20402
+    },
+    {
+      "epoch": 4.637045454545454,
+      "grad_norm": 1.056385405830029,
+      "learning_rate": 1.620214716040028e-08,
+      "loss": 0.0022,
+      "step": 20403
+    },
+    {
+      "epoch": 4.637272727272728,
+      "grad_norm": 0.8716421084562664,
+      "learning_rate": 1.6181963406549233e-08,
+      "loss": 0.0053,
+      "step": 20404
+    },
+    {
+      "epoch": 4.6375,
+      "grad_norm": 0.12483231370531392,
+      "learning_rate": 1.6161792067560105e-08,
+      "loss": 0.0004,
+      "step": 20405
+    },
+    {
+      "epoch": 4.637727272727273,
+      "grad_norm": 0.7313521504638333,
+      "learning_rate": 1.6141633143844163e-08,
+      "loss": 0.0023,
+      "step": 20406
+    },
+    {
+      "epoch": 4.637954545454545,
+      "grad_norm": 1.7074792316562601,
+      "learning_rate": 1.61214866358126e-08,
+      "loss": 0.008,
+      "step": 20407
+    },
+    {
+      "epoch": 4.638181818181819,
+      "grad_norm": 1.080545772375185,
+      "learning_rate": 1.610135254387607e-08,
+      "loss": 0.0049,
+      "step": 20408
+    },
+    {
+      "epoch": 4.638409090909091,
+      "grad_norm": 0.6930771383501843,
+      "learning_rate": 1.6081230868445415e-08,
+      "loss": 0.0028,
+      "step": 20409
+    },
+    {
+      "epoch": 4.638636363636364,
+      "grad_norm": 0.2992685341106531,
+      "learning_rate": 1.606112160993066e-08,
+      "loss": 0.0013,
+      "step": 20410
+    },
+    {
+      "epoch": 4.638863636363636,
+      "grad_norm": 1.1348907344058528,
+      "learning_rate": 1.6041024768741962e-08,
+      "loss": 0.0029,
+      "step": 20411
+    },
+    {
+      "epoch": 4.639090909090909,
+      "grad_norm": 0.6896906167249666,
+      "learning_rate": 1.6020940345289198e-08,
+      "loss": 0.0033,
+      "step": 20412
+    },
+    {
+      "epoch": 4.639318181818182,
+      "grad_norm": 0.6968609525745602,
+      "learning_rate": 1.6000868339981836e-08,
+      "loss": 0.0019,
+      "step": 20413
+    },
+    {
+      "epoch": 4.639545454545455,
+      "grad_norm": 0.5688383159949,
+      "learning_rate": 1.598080875322927e-08,
+      "loss": 0.0032,
+      "step": 20414
+    },
+    {
+      "epoch": 4.639772727272727,
+      "grad_norm": 1.0557360959422726,
+      "learning_rate": 1.596076158544048e-08,
+      "loss": 0.0044,
+      "step": 20415
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 0.9008257510530399,
+      "learning_rate": 1.5940726837024302e-08,
+      "loss": 0.0045,
+      "step": 20416
+    },
+    {
+      "epoch": 4.640227272727273,
+      "grad_norm": 0.6220453661962648,
+      "learning_rate": 1.592070450838923e-08,
+      "loss": 0.0034,
+      "step": 20417
+    },
+    {
+      "epoch": 4.640454545454546,
+      "grad_norm": 1.8726327424210678,
+      "learning_rate": 1.5900694599943685e-08,
+      "loss": 0.014,
+      "step": 20418
+    },
+    {
+      "epoch": 4.640681818181818,
+      "grad_norm": 0.5093715093198635,
+      "learning_rate": 1.588069711209554e-08,
+      "loss": 0.002,
+      "step": 20419
+    },
+    {
+      "epoch": 4.640909090909091,
+      "grad_norm": 1.029523595952919,
+      "learning_rate": 1.5860712045252662e-08,
+      "loss": 0.0071,
+      "step": 20420
+    },
+    {
+      "epoch": 4.641136363636363,
+      "grad_norm": 1.2882107419709887,
+      "learning_rate": 1.5840739399822572e-08,
+      "loss": 0.0061,
+      "step": 20421
+    },
+    {
+      "epoch": 4.641363636363637,
+      "grad_norm": 0.8755906730462707,
+      "learning_rate": 1.582077917621258e-08,
+      "loss": 0.0072,
+      "step": 20422
+    },
+    {
+      "epoch": 4.641590909090909,
+      "grad_norm": 0.5138842663241249,
+      "learning_rate": 1.5800831374829655e-08,
+      "loss": 0.0026,
+      "step": 20423
+    },
+    {
+      "epoch": 4.641818181818182,
+      "grad_norm": 2.142691807026711,
+      "learning_rate": 1.5780895996080554e-08,
+      "loss": 0.0213,
+      "step": 20424
+    },
+    {
+      "epoch": 4.642045454545455,
+      "grad_norm": 0.435253222591622,
+      "learning_rate": 1.576097304037183e-08,
+      "loss": 0.0012,
+      "step": 20425
+    },
+    {
+      "epoch": 4.6422727272727276,
+      "grad_norm": 0.8957666925409736,
+      "learning_rate": 1.5741062508109752e-08,
+      "loss": 0.0032,
+      "step": 20426
+    },
+    {
+      "epoch": 4.6425,
+      "grad_norm": 1.1185155989816737,
+      "learning_rate": 1.572116439970031e-08,
+      "loss": 0.0061,
+      "step": 20427
+    },
+    {
+      "epoch": 4.642727272727273,
+      "grad_norm": 0.3898121154261233,
+      "learning_rate": 1.5701278715549302e-08,
+      "loss": 0.0016,
+      "step": 20428
+    },
+    {
+      "epoch": 4.642954545454545,
+      "grad_norm": 1.0214566566913337,
+      "learning_rate": 1.5681405456062087e-08,
+      "loss": 0.0023,
+      "step": 20429
+    },
+    {
+      "epoch": 4.6431818181818185,
+      "grad_norm": 0.5664046957646219,
+      "learning_rate": 1.5661544621644112e-08,
+      "loss": 0.0052,
+      "step": 20430
+    },
+    {
+      "epoch": 4.643409090909091,
+      "grad_norm": 0.6483313807742641,
+      "learning_rate": 1.5641696212700327e-08,
+      "loss": 0.0024,
+      "step": 20431
+    },
+    {
+      "epoch": 4.643636363636364,
+      "grad_norm": 0.8462247129970645,
+      "learning_rate": 1.562186022963541e-08,
+      "loss": 0.0047,
+      "step": 20432
+    },
+    {
+      "epoch": 4.643863636363636,
+      "grad_norm": 1.9738361098020738,
+      "learning_rate": 1.5602036672853897e-08,
+      "loss": 0.0091,
+      "step": 20433
+    },
+    {
+      "epoch": 4.6440909090909095,
+      "grad_norm": 1.381346544373438,
+      "learning_rate": 1.558222554275998e-08,
+      "loss": 0.0061,
+      "step": 20434
+    },
+    {
+      "epoch": 4.644318181818182,
+      "grad_norm": 1.9902153851472035,
+      "learning_rate": 1.5562426839757713e-08,
+      "loss": 0.0165,
+      "step": 20435
+    },
+    {
+      "epoch": 4.6445454545454545,
+      "grad_norm": 1.2994847586492158,
+      "learning_rate": 1.5542640564250727e-08,
+      "loss": 0.0022,
+      "step": 20436
+    },
+    {
+      "epoch": 4.644772727272727,
+      "grad_norm": 0.35517921487543413,
+      "learning_rate": 1.552286671664259e-08,
+      "loss": 0.0011,
+      "step": 20437
+    },
+    {
+      "epoch": 4.645,
+      "grad_norm": 14.248627633763824,
+      "learning_rate": 1.550310529733645e-08,
+      "loss": 0.0099,
+      "step": 20438
+    },
+    {
+      "epoch": 4.645227272727273,
+      "grad_norm": 0.4600227830494147,
+      "learning_rate": 1.548335630673539e-08,
+      "loss": 0.0022,
+      "step": 20439
+    },
+    {
+      "epoch": 4.6454545454545455,
+      "grad_norm": 0.6433294213101037,
+      "learning_rate": 1.5463619745242076e-08,
+      "loss": 0.0023,
+      "step": 20440
+    },
+    {
+      "epoch": 4.645681818181818,
+      "grad_norm": 2.137020461150948,
+      "learning_rate": 1.5443895613258887e-08,
+      "loss": 0.0227,
+      "step": 20441
+    },
+    {
+      "epoch": 4.6459090909090905,
+      "grad_norm": 1.7600092511903516,
+      "learning_rate": 1.5424183911188078e-08,
+      "loss": 0.0066,
+      "step": 20442
+    },
+    {
+      "epoch": 4.646136363636364,
+      "grad_norm": 0.8076234906452949,
+      "learning_rate": 1.5404484639431616e-08,
+      "loss": 0.0043,
+      "step": 20443
+    },
+    {
+      "epoch": 4.6463636363636365,
+      "grad_norm": 0.6354712182985257,
+      "learning_rate": 1.5384797798391263e-08,
+      "loss": 0.0019,
+      "step": 20444
+    },
+    {
+      "epoch": 4.646590909090909,
+      "grad_norm": 1.922779002588427,
+      "learning_rate": 1.5365123388468363e-08,
+      "loss": 0.0068,
+      "step": 20445
+    },
+    {
+      "epoch": 4.6468181818181815,
+      "grad_norm": 2.378064275906947,
+      "learning_rate": 1.5345461410064198e-08,
+      "loss": 0.0091,
+      "step": 20446
+    },
+    {
+      "epoch": 4.647045454545455,
+      "grad_norm": 1.2928900470412212,
+      "learning_rate": 1.532581186357962e-08,
+      "loss": 0.0037,
+      "step": 20447
+    },
+    {
+      "epoch": 4.647272727272727,
+      "grad_norm": 0.37877469759463184,
+      "learning_rate": 1.5306174749415425e-08,
+      "loss": 0.002,
+      "step": 20448
+    },
+    {
+      "epoch": 4.6475,
+      "grad_norm": 0.5536930473313743,
+      "learning_rate": 1.5286550067972057e-08,
+      "loss": 0.0029,
+      "step": 20449
+    },
+    {
+      "epoch": 4.6477272727272725,
+      "grad_norm": 1.871702609596857,
+      "learning_rate": 1.5266937819649537e-08,
+      "loss": 0.0101,
+      "step": 20450
+    },
+    {
+      "epoch": 4.647954545454546,
+      "grad_norm": 0.8792942929345599,
+      "learning_rate": 1.524733800484797e-08,
+      "loss": 0.0022,
+      "step": 20451
+    },
+    {
+      "epoch": 4.648181818181818,
+      "grad_norm": 0.45371933157922806,
+      "learning_rate": 1.5227750623966892e-08,
+      "loss": 0.0025,
+      "step": 20452
+    },
+    {
+      "epoch": 4.648409090909091,
+      "grad_norm": 1.428231782126673,
+      "learning_rate": 1.5208175677405844e-08,
+      "loss": 0.0055,
+      "step": 20453
+    },
+    {
+      "epoch": 4.6486363636363635,
+      "grad_norm": 0.8762484453725152,
+      "learning_rate": 1.5188613165563886e-08,
+      "loss": 0.0067,
+      "step": 20454
+    },
+    {
+      "epoch": 4.648863636363636,
+      "grad_norm": 2.6666043615035493,
+      "learning_rate": 1.5169063088840004e-08,
+      "loss": 0.0097,
+      "step": 20455
+    },
+    {
+      "epoch": 4.649090909090909,
+      "grad_norm": 1.8209934670673718,
+      "learning_rate": 1.5149525447632763e-08,
+      "loss": 0.0088,
+      "step": 20456
+    },
+    {
+      "epoch": 4.649318181818182,
+      "grad_norm": 1.1721493399326812,
+      "learning_rate": 1.5130000242340736e-08,
+      "loss": 0.0048,
+      "step": 20457
+    },
+    {
+      "epoch": 4.649545454545454,
+      "grad_norm": 0.423883289011149,
+      "learning_rate": 1.511048747336194e-08,
+      "loss": 0.0017,
+      "step": 20458
+    },
+    {
+      "epoch": 4.649772727272727,
+      "grad_norm": 0.8482135583092716,
+      "learning_rate": 1.509098714109425e-08,
+      "loss": 0.0021,
+      "step": 20459
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 0.9759315318927599,
+      "learning_rate": 1.507149924593547e-08,
+      "loss": 0.0025,
+      "step": 20460
+    },
+    {
+      "epoch": 4.650227272727273,
+      "grad_norm": 0.8552102481533482,
+      "learning_rate": 1.5052023788282853e-08,
+      "loss": 0.0039,
+      "step": 20461
+    },
+    {
+      "epoch": 4.650454545454545,
+      "grad_norm": 2.0486737537483797,
+      "learning_rate": 1.5032560768533654e-08,
+      "loss": 0.0095,
+      "step": 20462
+    },
+    {
+      "epoch": 4.650681818181818,
+      "grad_norm": 0.3649517888264847,
+      "learning_rate": 1.5013110187084636e-08,
+      "loss": 0.0011,
+      "step": 20463
+    },
+    {
+      "epoch": 4.65090909090909,
+      "grad_norm": 0.5675572843707917,
+      "learning_rate": 1.4993672044332427e-08,
+      "loss": 0.0038,
+      "step": 20464
+    },
+    {
+      "epoch": 4.651136363636364,
+      "grad_norm": 1.8667194339693471,
+      "learning_rate": 1.4974246340673516e-08,
+      "loss": 0.0112,
+      "step": 20465
+    },
+    {
+      "epoch": 4.651363636363636,
+      "grad_norm": 0.5553626307372128,
+      "learning_rate": 1.4954833076504043e-08,
+      "loss": 0.0017,
+      "step": 20466
+    },
+    {
+      "epoch": 4.651590909090909,
+      "grad_norm": 0.25333961587124415,
+      "learning_rate": 1.493543225221973e-08,
+      "loss": 0.0005,
+      "step": 20467
+    },
+    {
+      "epoch": 4.651818181818182,
+      "grad_norm": 0.46400898134971785,
+      "learning_rate": 1.491604386821624e-08,
+      "loss": 0.0028,
+      "step": 20468
+    },
+    {
+      "epoch": 4.652045454545455,
+      "grad_norm": 1.8298432751145908,
+      "learning_rate": 1.4896667924889017e-08,
+      "loss": 0.0226,
+      "step": 20469
+    },
+    {
+      "epoch": 4.652272727272727,
+      "grad_norm": 2.8277376674153993,
+      "learning_rate": 1.487730442263309e-08,
+      "loss": 0.0055,
+      "step": 20470
+    },
+    {
+      "epoch": 4.6525,
+      "grad_norm": 1.7292233205360643,
+      "learning_rate": 1.4857953361843356e-08,
+      "loss": 0.0044,
+      "step": 20471
+    },
+    {
+      "epoch": 4.652727272727272,
+      "grad_norm": 1.763843317292033,
+      "learning_rate": 1.4838614742914426e-08,
+      "loss": 0.0036,
+      "step": 20472
+    },
+    {
+      "epoch": 4.652954545454546,
+      "grad_norm": 0.9970504372367873,
+      "learning_rate": 1.481928856624057e-08,
+      "loss": 0.0014,
+      "step": 20473
+    },
+    {
+      "epoch": 4.653181818181818,
+      "grad_norm": 1.961221605896905,
+      "learning_rate": 1.4799974832215985e-08,
+      "loss": 0.0139,
+      "step": 20474
+    },
+    {
+      "epoch": 4.653409090909091,
+      "grad_norm": 0.6810126133850388,
+      "learning_rate": 1.4780673541234456e-08,
+      "loss": 0.0032,
+      "step": 20475
+    },
+    {
+      "epoch": 4.653636363636363,
+      "grad_norm": 0.40655469925493637,
+      "learning_rate": 1.4761384693689554e-08,
+      "loss": 0.0015,
+      "step": 20476
+    },
+    {
+      "epoch": 4.653863636363637,
+      "grad_norm": 1.075660796740642,
+      "learning_rate": 1.4742108289974646e-08,
+      "loss": 0.0104,
+      "step": 20477
+    },
+    {
+      "epoch": 4.654090909090909,
+      "grad_norm": 0.19359607640345525,
+      "learning_rate": 1.4722844330482753e-08,
+      "loss": 0.0007,
+      "step": 20478
+    },
+    {
+      "epoch": 4.654318181818182,
+      "grad_norm": 1.1269213868898282,
+      "learning_rate": 1.470359281560689e-08,
+      "loss": 0.0047,
+      "step": 20479
+    },
+    {
+      "epoch": 4.654545454545454,
+      "grad_norm": 1.271909606797696,
+      "learning_rate": 1.4684353745739385e-08,
+      "loss": 0.0052,
+      "step": 20480
+    },
+    {
+      "epoch": 4.654772727272727,
+      "grad_norm": 0.2603207849318073,
+      "learning_rate": 1.4665127121272702e-08,
+      "loss": 0.001,
+      "step": 20481
+    },
+    {
+      "epoch": 4.655,
+      "grad_norm": 0.6218711475225029,
+      "learning_rate": 1.4645912942598817e-08,
+      "loss": 0.0039,
+      "step": 20482
+    },
+    {
+      "epoch": 4.655227272727273,
+      "grad_norm": 0.33955042504428545,
+      "learning_rate": 1.4626711210109639e-08,
+      "loss": 0.0014,
+      "step": 20483
+    },
+    {
+      "epoch": 4.655454545454545,
+      "grad_norm": 1.5280504834304287,
+      "learning_rate": 1.4607521924196733e-08,
+      "loss": 0.0058,
+      "step": 20484
+    },
+    {
+      "epoch": 4.655681818181818,
+      "grad_norm": 1.0913189835015369,
+      "learning_rate": 1.4588345085251241e-08,
+      "loss": 0.0082,
+      "step": 20485
+    },
+    {
+      "epoch": 4.655909090909091,
+      "grad_norm": 1.5401207970165667,
+      "learning_rate": 1.4569180693664309e-08,
+      "loss": 0.0078,
+      "step": 20486
+    },
+    {
+      "epoch": 4.656136363636364,
+      "grad_norm": 0.26091229387574955,
+      "learning_rate": 1.4550028749826737e-08,
+      "loss": 0.0017,
+      "step": 20487
+    },
+    {
+      "epoch": 4.656363636363636,
+      "grad_norm": 0.867686076376672,
+      "learning_rate": 1.4530889254129113e-08,
+      "loss": 0.0029,
+      "step": 20488
+    },
+    {
+      "epoch": 4.656590909090909,
+      "grad_norm": 1.139774380223876,
+      "learning_rate": 1.4511762206961683e-08,
+      "loss": 0.0094,
+      "step": 20489
+    },
+    {
+      "epoch": 4.656818181818182,
+      "grad_norm": 1.8936187177155788,
+      "learning_rate": 1.449264760871441e-08,
+      "loss": 0.0239,
+      "step": 20490
+    },
+    {
+      "epoch": 4.657045454545455,
+      "grad_norm": 0.8403444238471377,
+      "learning_rate": 1.4473545459777195e-08,
+      "loss": 0.0025,
+      "step": 20491
+    },
+    {
+      "epoch": 4.657272727272727,
+      "grad_norm": 1.1488475087123202,
+      "learning_rate": 1.4454455760539512e-08,
+      "loss": 0.0038,
+      "step": 20492
+    },
+    {
+      "epoch": 4.6575,
+      "grad_norm": 0.14362072255586641,
+      "learning_rate": 1.4435378511390637e-08,
+      "loss": 0.0003,
+      "step": 20493
+    },
+    {
+      "epoch": 4.657727272727273,
+      "grad_norm": 0.7599282580191256,
+      "learning_rate": 1.4416313712719565e-08,
+      "loss": 0.0023,
+      "step": 20494
+    },
+    {
+      "epoch": 4.657954545454546,
+      "grad_norm": 1.3715628083580111,
+      "learning_rate": 1.439726136491508e-08,
+      "loss": 0.0115,
+      "step": 20495
+    },
+    {
+      "epoch": 4.658181818181818,
+      "grad_norm": 0.8865584576090725,
+      "learning_rate": 1.4378221468365691e-08,
+      "loss": 0.0048,
+      "step": 20496
+    },
+    {
+      "epoch": 4.658409090909091,
+      "grad_norm": 3.60061274234025,
+      "learning_rate": 1.43591940234597e-08,
+      "loss": 0.0115,
+      "step": 20497
+    },
+    {
+      "epoch": 4.658636363636363,
+      "grad_norm": 1.6913523041078726,
+      "learning_rate": 1.4340179030584989e-08,
+      "loss": 0.0033,
+      "step": 20498
+    },
+    {
+      "epoch": 4.658863636363637,
+      "grad_norm": 0.5450414263290356,
+      "learning_rate": 1.4321176490129444e-08,
+      "loss": 0.0015,
+      "step": 20499
+    },
+    {
+      "epoch": 4.659090909090909,
+      "grad_norm": 0.9411725455797338,
+      "learning_rate": 1.4302186402480461e-08,
+      "loss": 0.0013,
+      "step": 20500
+    },
+    {
+      "epoch": 4.659318181818182,
+      "grad_norm": 1.0298123734084152,
+      "learning_rate": 1.4283208768025371e-08,
+      "loss": 0.0036,
+      "step": 20501
+    },
+    {
+      "epoch": 4.659545454545454,
+      "grad_norm": 0.8213084887263702,
+      "learning_rate": 1.426424358715102e-08,
+      "loss": 0.0031,
+      "step": 20502
+    },
+    {
+      "epoch": 4.659772727272728,
+      "grad_norm": 0.2843703336911199,
+      "learning_rate": 1.4245290860244315e-08,
+      "loss": 0.0009,
+      "step": 20503
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 2.2139093997413744,
+      "learning_rate": 1.4226350587691548e-08,
+      "loss": 0.0133,
+      "step": 20504
+    },
+    {
+      "epoch": 4.660227272727273,
+      "grad_norm": 0.3428342351332154,
+      "learning_rate": 1.4207422769879075e-08,
+      "loss": 0.0019,
+      "step": 20505
+    },
+    {
+      "epoch": 4.660454545454545,
+      "grad_norm": 1.7569671329087453,
+      "learning_rate": 1.4188507407192909e-08,
+      "loss": 0.0051,
+      "step": 20506
+    },
+    {
+      "epoch": 4.660681818181818,
+      "grad_norm": 0.5475394889784192,
+      "learning_rate": 1.4169604500018641e-08,
+      "loss": 0.0018,
+      "step": 20507
+    },
+    {
+      "epoch": 4.660909090909091,
+      "grad_norm": 1.077107285013809,
+      "learning_rate": 1.4150714048741798e-08,
+      "loss": 0.0045,
+      "step": 20508
+    },
+    {
+      "epoch": 4.661136363636364,
+      "grad_norm": 0.6530104906995529,
+      "learning_rate": 1.4131836053747559e-08,
+      "loss": 0.0015,
+      "step": 20509
+    },
+    {
+      "epoch": 4.661363636363636,
+      "grad_norm": 1.024545981259957,
+      "learning_rate": 1.411297051542096e-08,
+      "loss": 0.0033,
+      "step": 20510
+    },
+    {
+      "epoch": 4.66159090909091,
+      "grad_norm": 0.6228722517894767,
+      "learning_rate": 1.4094117434146626e-08,
+      "loss": 0.0027,
+      "step": 20511
+    },
+    {
+      "epoch": 4.661818181818182,
+      "grad_norm": 1.4985654893976137,
+      "learning_rate": 1.4075276810308971e-08,
+      "loss": 0.0141,
+      "step": 20512
+    },
+    {
+      "epoch": 4.662045454545455,
+      "grad_norm": 1.9257578834271847,
+      "learning_rate": 1.4056448644292272e-08,
+      "loss": 0.0197,
+      "step": 20513
+    },
+    {
+      "epoch": 4.662272727272727,
+      "grad_norm": 0.2208243839575528,
+      "learning_rate": 1.4037632936480458e-08,
+      "loss": 0.0004,
+      "step": 20514
+    },
+    {
+      "epoch": 4.6625,
+      "grad_norm": 0.7435452408533143,
+      "learning_rate": 1.401882968725718e-08,
+      "loss": 0.0014,
+      "step": 20515
+    },
+    {
+      "epoch": 4.662727272727273,
+      "grad_norm": 0.912446156945775,
+      "learning_rate": 1.400003889700588e-08,
+      "loss": 0.0025,
+      "step": 20516
+    },
+    {
+      "epoch": 4.662954545454546,
+      "grad_norm": 1.0416382523551164,
+      "learning_rate": 1.3981260566109656e-08,
+      "loss": 0.0016,
+      "step": 20517
+    },
+    {
+      "epoch": 4.663181818181818,
+      "grad_norm": 1.0663103969471925,
+      "learning_rate": 1.3962494694951603e-08,
+      "loss": 0.004,
+      "step": 20518
+    },
+    {
+      "epoch": 4.663409090909091,
+      "grad_norm": 0.30227055956268567,
+      "learning_rate": 1.3943741283914265e-08,
+      "loss": 0.0009,
+      "step": 20519
+    },
+    {
+      "epoch": 4.663636363636364,
+      "grad_norm": 0.498924437963042,
+      "learning_rate": 1.392500033338004e-08,
+      "loss": 0.0029,
+      "step": 20520
+    },
+    {
+      "epoch": 4.6638636363636365,
+      "grad_norm": 1.9236050931636408,
+      "learning_rate": 1.3906271843731196e-08,
+      "loss": 0.0119,
+      "step": 20521
+    },
+    {
+      "epoch": 4.664090909090909,
+      "grad_norm": 1.069415258521584,
+      "learning_rate": 1.3887555815349576e-08,
+      "loss": 0.011,
+      "step": 20522
+    },
+    {
+      "epoch": 4.664318181818182,
+      "grad_norm": 1.3867794215295592,
+      "learning_rate": 1.3868852248616893e-08,
+      "loss": 0.0038,
+      "step": 20523
+    },
+    {
+      "epoch": 4.664545454545454,
+      "grad_norm": 3.1262462306467063,
+      "learning_rate": 1.3850161143914436e-08,
+      "loss": 0.0062,
+      "step": 20524
+    },
+    {
+      "epoch": 4.6647727272727275,
+      "grad_norm": 1.279649405045721,
+      "learning_rate": 1.3831482501623362e-08,
+      "loss": 0.0037,
+      "step": 20525
+    },
+    {
+      "epoch": 4.665,
+      "grad_norm": 2.115198606309296,
+      "learning_rate": 1.3812816322124612e-08,
+      "loss": 0.0067,
+      "step": 20526
+    },
+    {
+      "epoch": 4.665227272727273,
+      "grad_norm": 0.841493683076323,
+      "learning_rate": 1.3794162605798855e-08,
+      "loss": 0.0032,
+      "step": 20527
+    },
+    {
+      "epoch": 4.665454545454545,
+      "grad_norm": 1.1457859013531524,
+      "learning_rate": 1.3775521353026483e-08,
+      "loss": 0.0054,
+      "step": 20528
+    },
+    {
+      "epoch": 4.6656818181818185,
+      "grad_norm": 1.3688431128367682,
+      "learning_rate": 1.3756892564187468e-08,
+      "loss": 0.0088,
+      "step": 20529
+    },
+    {
+      "epoch": 4.665909090909091,
+      "grad_norm": 0.9252528247887694,
+      "learning_rate": 1.3738276239661786e-08,
+      "loss": 0.0062,
+      "step": 20530
+    },
+    {
+      "epoch": 4.6661363636363635,
+      "grad_norm": 0.7659812310993981,
+      "learning_rate": 1.371967237982906e-08,
+      "loss": 0.0024,
+      "step": 20531
+    },
+    {
+      "epoch": 4.666363636363636,
+      "grad_norm": 1.511981299933046,
+      "learning_rate": 1.3701080985068782e-08,
+      "loss": 0.0074,
+      "step": 20532
+    },
+    {
+      "epoch": 4.6665909090909095,
+      "grad_norm": 0.6003554187667984,
+      "learning_rate": 1.3682502055759814e-08,
+      "loss": 0.0031,
+      "step": 20533
+    },
+    {
+      "epoch": 4.666818181818182,
+      "grad_norm": 0.20119924234037895,
+      "learning_rate": 1.3663935592281088e-08,
+      "loss": 0.0005,
+      "step": 20534
+    },
+    {
+      "epoch": 4.6670454545454545,
+      "grad_norm": 0.8573085329610445,
+      "learning_rate": 1.3645381595011332e-08,
+      "loss": 0.0019,
+      "step": 20535
+    },
+    {
+      "epoch": 4.667272727272727,
+      "grad_norm": 0.9290662463835615,
+      "learning_rate": 1.3626840064328781e-08,
+      "loss": 0.0037,
+      "step": 20536
+    },
+    {
+      "epoch": 4.6675,
+      "grad_norm": 1.0676154964909077,
+      "learning_rate": 1.3608311000611538e-08,
+      "loss": 0.0123,
+      "step": 20537
+    },
+    {
+      "epoch": 4.667727272727273,
+      "grad_norm": 1.9556066225040727,
+      "learning_rate": 1.3589794404237494e-08,
+      "loss": 0.0173,
+      "step": 20538
+    },
+    {
+      "epoch": 4.6679545454545455,
+      "grad_norm": 2.0497227091702532,
+      "learning_rate": 1.3571290275584195e-08,
+      "loss": 0.0044,
+      "step": 20539
+    },
+    {
+      "epoch": 4.668181818181818,
+      "grad_norm": 2.4373415126802147,
+      "learning_rate": 1.3552798615028978e-08,
+      "loss": 0.0068,
+      "step": 20540
+    },
+    {
+      "epoch": 4.6684090909090905,
+      "grad_norm": 0.7452233276207422,
+      "learning_rate": 1.3534319422948971e-08,
+      "loss": 0.0012,
+      "step": 20541
+    },
+    {
+      "epoch": 4.668636363636364,
+      "grad_norm": 1.2586768135832316,
+      "learning_rate": 1.3515852699720958e-08,
+      "loss": 0.0039,
+      "step": 20542
+    },
+    {
+      "epoch": 4.668863636363636,
+      "grad_norm": 2.020889898092911,
+      "learning_rate": 1.349739844572144e-08,
+      "loss": 0.0069,
+      "step": 20543
+    },
+    {
+      "epoch": 4.669090909090909,
+      "grad_norm": 1.55106970105805,
+      "learning_rate": 1.3478956661326786e-08,
+      "loss": 0.0098,
+      "step": 20544
+    },
+    {
+      "epoch": 4.6693181818181815,
+      "grad_norm": 0.36463270298831596,
+      "learning_rate": 1.3460527346913152e-08,
+      "loss": 0.0013,
+      "step": 20545
+    },
+    {
+      "epoch": 4.669545454545455,
+      "grad_norm": 1.877443112655435,
+      "learning_rate": 1.344211050285621e-08,
+      "loss": 0.011,
+      "step": 20546
+    },
+    {
+      "epoch": 4.669772727272727,
+      "grad_norm": 0.6201931129368303,
+      "learning_rate": 1.3423706129531492e-08,
+      "loss": 0.0027,
+      "step": 20547
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 1.8820157743094983,
+      "learning_rate": 1.3405314227314391e-08,
+      "loss": 0.0193,
+      "step": 20548
+    },
+    {
+      "epoch": 4.6702272727272724,
+      "grad_norm": 1.789549219272369,
+      "learning_rate": 1.3386934796579959e-08,
+      "loss": 0.0057,
+      "step": 20549
+    },
+    {
+      "epoch": 4.670454545454545,
+      "grad_norm": 0.5390852943640815,
+      "learning_rate": 1.3368567837702892e-08,
+      "loss": 0.0019,
+      "step": 20550
+    },
+    {
+      "epoch": 4.670681818181818,
+      "grad_norm": 0.8862686660111858,
+      "learning_rate": 1.3350213351057823e-08,
+      "loss": 0.0103,
+      "step": 20551
+    },
+    {
+      "epoch": 4.670909090909091,
+      "grad_norm": 0.6380396403983427,
+      "learning_rate": 1.3331871337018965e-08,
+      "loss": 0.0013,
+      "step": 20552
+    },
+    {
+      "epoch": 4.671136363636363,
+      "grad_norm": 0.9485752364651432,
+      "learning_rate": 1.3313541795960325e-08,
+      "loss": 0.009,
+      "step": 20553
+    },
+    {
+      "epoch": 4.671363636363637,
+      "grad_norm": 0.31463299452777055,
+      "learning_rate": 1.329522472825584e-08,
+      "loss": 0.0011,
+      "step": 20554
+    },
+    {
+      "epoch": 4.671590909090909,
+      "grad_norm": 1.5184364348301305,
+      "learning_rate": 1.3276920134278753e-08,
+      "loss": 0.0053,
+      "step": 20555
+    },
+    {
+      "epoch": 4.671818181818182,
+      "grad_norm": 2.6938694045623994,
+      "learning_rate": 1.3258628014402586e-08,
+      "loss": 0.0064,
+      "step": 20556
+    },
+    {
+      "epoch": 4.672045454545454,
+      "grad_norm": 1.8393402361737987,
+      "learning_rate": 1.3240348369000164e-08,
+      "loss": 0.0048,
+      "step": 20557
+    },
+    {
+      "epoch": 4.672272727272727,
+      "grad_norm": 0.4444398001671905,
+      "learning_rate": 1.3222081198444383e-08,
+      "loss": 0.002,
+      "step": 20558
+    },
+    {
+      "epoch": 4.6725,
+      "grad_norm": 0.4898046691611408,
+      "learning_rate": 1.3203826503107586e-08,
+      "loss": 0.0024,
+      "step": 20559
+    },
+    {
+      "epoch": 4.672727272727273,
+      "grad_norm": 1.708207690147831,
+      "learning_rate": 1.3185584283362114e-08,
+      "loss": 0.0048,
+      "step": 20560
+    },
+    {
+      "epoch": 4.672954545454545,
+      "grad_norm": 1.9252456399985307,
+      "learning_rate": 1.3167354539579957e-08,
+      "loss": 0.0113,
+      "step": 20561
+    },
+    {
+      "epoch": 4.673181818181818,
+      "grad_norm": 0.6568828219276688,
+      "learning_rate": 1.3149137272132905e-08,
+      "loss": 0.0026,
+      "step": 20562
+    },
+    {
+      "epoch": 4.673409090909091,
+      "grad_norm": 0.7098694623768653,
+      "learning_rate": 1.3130932481392328e-08,
+      "loss": 0.0032,
+      "step": 20563
+    },
+    {
+      "epoch": 4.673636363636364,
+      "grad_norm": 0.9288860868296707,
+      "learning_rate": 1.3112740167729523e-08,
+      "loss": 0.0029,
+      "step": 20564
+    },
+    {
+      "epoch": 4.673863636363636,
+      "grad_norm": 1.1396219306250213,
+      "learning_rate": 1.3094560331515377e-08,
+      "loss": 0.0036,
+      "step": 20565
+    },
+    {
+      "epoch": 4.674090909090909,
+      "grad_norm": 0.7980644898012831,
+      "learning_rate": 1.3076392973120702e-08,
+      "loss": 0.0073,
+      "step": 20566
+    },
+    {
+      "epoch": 4.674318181818181,
+      "grad_norm": 0.9308919961211688,
+      "learning_rate": 1.3058238092915967e-08,
+      "loss": 0.0038,
+      "step": 20567
+    },
+    {
+      "epoch": 4.674545454545455,
+      "grad_norm": 0.6429343129323831,
+      "learning_rate": 1.3040095691271293e-08,
+      "loss": 0.0028,
+      "step": 20568
+    },
+    {
+      "epoch": 4.674772727272727,
+      "grad_norm": 0.38737393893770056,
+      "learning_rate": 1.3021965768556729e-08,
+      "loss": 0.0022,
+      "step": 20569
+    },
+    {
+      "epoch": 4.675,
+      "grad_norm": 0.350881575998914,
+      "learning_rate": 1.3003848325141982e-08,
+      "loss": 0.0017,
+      "step": 20570
+    },
+    {
+      "epoch": 4.675227272727272,
+      "grad_norm": 0.7374645610391216,
+      "learning_rate": 1.2985743361396338e-08,
+      "loss": 0.0024,
+      "step": 20571
+    },
+    {
+      "epoch": 4.675454545454546,
+      "grad_norm": 0.3968721742348304,
+      "learning_rate": 1.2967650877689225e-08,
+      "loss": 0.0008,
+      "step": 20572
+    },
+    {
+      "epoch": 4.675681818181818,
+      "grad_norm": 0.03847946202754264,
+      "learning_rate": 1.2949570874389445e-08,
+      "loss": 0.0001,
+      "step": 20573
+    },
+    {
+      "epoch": 4.675909090909091,
+      "grad_norm": 0.40842607468178366,
+      "learning_rate": 1.2931503351865659e-08,
+      "loss": 0.0019,
+      "step": 20574
+    },
+    {
+      "epoch": 4.676136363636363,
+      "grad_norm": 1.339828732248278,
+      "learning_rate": 1.2913448310486326e-08,
+      "loss": 0.0054,
+      "step": 20575
+    },
+    {
+      "epoch": 4.676363636363637,
+      "grad_norm": 0.8332883757043111,
+      "learning_rate": 1.2895405750619621e-08,
+      "loss": 0.0023,
+      "step": 20576
+    },
+    {
+      "epoch": 4.676590909090909,
+      "grad_norm": 1.6559228698247275,
+      "learning_rate": 1.2877375672633516e-08,
+      "loss": 0.0039,
+      "step": 20577
+    },
+    {
+      "epoch": 4.676818181818182,
+      "grad_norm": 1.9038681738306344,
+      "learning_rate": 1.2859358076895561e-08,
+      "loss": 0.0075,
+      "step": 20578
+    },
+    {
+      "epoch": 4.677045454545454,
+      "grad_norm": 1.3677571011122538,
+      "learning_rate": 1.2841352963773244e-08,
+      "loss": 0.0037,
+      "step": 20579
+    },
+    {
+      "epoch": 4.677272727272728,
+      "grad_norm": 0.8492581270961828,
+      "learning_rate": 1.2823360333633697e-08,
+      "loss": 0.0085,
+      "step": 20580
+    },
+    {
+      "epoch": 4.6775,
+      "grad_norm": 1.57214001305,
+      "learning_rate": 1.2805380186843853e-08,
+      "loss": 0.0061,
+      "step": 20581
+    },
+    {
+      "epoch": 4.677727272727273,
+      "grad_norm": 1.6683504480761944,
+      "learning_rate": 1.278741252377036e-08,
+      "loss": 0.017,
+      "step": 20582
+    },
+    {
+      "epoch": 4.677954545454545,
+      "grad_norm": 1.081190308508576,
+      "learning_rate": 1.2769457344779593e-08,
+      "loss": 0.0057,
+      "step": 20583
+    },
+    {
+      "epoch": 4.678181818181818,
+      "grad_norm": 0.8846251827388035,
+      "learning_rate": 1.2751514650237648e-08,
+      "loss": 0.0043,
+      "step": 20584
+    },
+    {
+      "epoch": 4.678409090909091,
+      "grad_norm": 1.06653080421137,
+      "learning_rate": 1.2733584440510552e-08,
+      "loss": 0.0049,
+      "step": 20585
+    },
+    {
+      "epoch": 4.678636363636364,
+      "grad_norm": 1.5555954806864423,
+      "learning_rate": 1.2715666715963707e-08,
+      "loss": 0.0175,
+      "step": 20586
+    },
+    {
+      "epoch": 4.678863636363636,
+      "grad_norm": 0.781130621244288,
+      "learning_rate": 1.2697761476962653e-08,
+      "loss": 0.0011,
+      "step": 20587
+    },
+    {
+      "epoch": 4.679090909090909,
+      "grad_norm": 2.820254028042363,
+      "learning_rate": 1.2679868723872445e-08,
+      "loss": 0.0083,
+      "step": 20588
+    },
+    {
+      "epoch": 4.679318181818182,
+      "grad_norm": 1.1017072955448608,
+      "learning_rate": 1.2661988457058003e-08,
+      "loss": 0.0038,
+      "step": 20589
+    },
+    {
+      "epoch": 4.679545454545455,
+      "grad_norm": 0.9311986624702482,
+      "learning_rate": 1.2644120676883823e-08,
+      "loss": 0.0054,
+      "step": 20590
+    },
+    {
+      "epoch": 4.679772727272727,
+      "grad_norm": 1.2542813980610275,
+      "learning_rate": 1.2626265383714405e-08,
+      "loss": 0.0038,
+      "step": 20591
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 0.7280104678314955,
+      "learning_rate": 1.2608422577913768e-08,
+      "loss": 0.0027,
+      "step": 20592
+    },
+    {
+      "epoch": 4.680227272727272,
+      "grad_norm": 1.2148693181951122,
+      "learning_rate": 1.2590592259845712e-08,
+      "loss": 0.0038,
+      "step": 20593
+    },
+    {
+      "epoch": 4.680454545454546,
+      "grad_norm": 1.409402390970752,
+      "learning_rate": 1.2572774429873979e-08,
+      "loss": 0.0056,
+      "step": 20594
+    },
+    {
+      "epoch": 4.680681818181818,
+      "grad_norm": 0.39520882975754346,
+      "learning_rate": 1.2554969088361748e-08,
+      "loss": 0.0013,
+      "step": 20595
+    },
+    {
+      "epoch": 4.680909090909091,
+      "grad_norm": 0.3824641485331314,
+      "learning_rate": 1.2537176235672202e-08,
+      "loss": 0.0011,
+      "step": 20596
+    },
+    {
+      "epoch": 4.681136363636364,
+      "grad_norm": 1.6325106682310453,
+      "learning_rate": 1.2519395872168036e-08,
+      "loss": 0.0132,
+      "step": 20597
+    },
+    {
+      "epoch": 4.681363636363637,
+      "grad_norm": 0.5734791744882902,
+      "learning_rate": 1.2501627998212015e-08,
+      "loss": 0.0019,
+      "step": 20598
+    },
+    {
+      "epoch": 4.681590909090909,
+      "grad_norm": 0.6770075829783645,
+      "learning_rate": 1.248387261416635e-08,
+      "loss": 0.002,
+      "step": 20599
+    },
+    {
+      "epoch": 4.681818181818182,
+      "grad_norm": 1.1360207823863682,
+      "learning_rate": 1.2466129720393113e-08,
+      "loss": 0.0037,
+      "step": 20600
+    },
+    {
+      "epoch": 4.682045454545454,
+      "grad_norm": 0.8053048869403825,
+      "learning_rate": 1.2448399317254097e-08,
+      "loss": 0.0021,
+      "step": 20601
+    },
+    {
+      "epoch": 4.682272727272728,
+      "grad_norm": 0.793287456469555,
+      "learning_rate": 1.2430681405110959e-08,
+      "loss": 0.0047,
+      "step": 20602
+    },
+    {
+      "epoch": 4.6825,
+      "grad_norm": 1.8948281043720099,
+      "learning_rate": 1.2412975984324796e-08,
+      "loss": 0.0056,
+      "step": 20603
+    },
+    {
+      "epoch": 4.682727272727273,
+      "grad_norm": 2.312286273122176,
+      "learning_rate": 1.2395283055256848e-08,
+      "loss": 0.0309,
+      "step": 20604
+    },
+    {
+      "epoch": 4.682954545454545,
+      "grad_norm": 1.1875010011338172,
+      "learning_rate": 1.23776026182678e-08,
+      "loss": 0.0042,
+      "step": 20605
+    },
+    {
+      "epoch": 4.683181818181819,
+      "grad_norm": 1.2120127403957004,
+      "learning_rate": 1.2359934673718196e-08,
+      "loss": 0.0053,
+      "step": 20606
+    },
+    {
+      "epoch": 4.683409090909091,
+      "grad_norm": 0.544640581416949,
+      "learning_rate": 1.2342279221968372e-08,
+      "loss": 0.0015,
+      "step": 20607
+    },
+    {
+      "epoch": 4.683636363636364,
+      "grad_norm": 1.784582748721483,
+      "learning_rate": 1.232463626337825e-08,
+      "loss": 0.0041,
+      "step": 20608
+    },
+    {
+      "epoch": 4.683863636363636,
+      "grad_norm": 1.0243504624194704,
+      "learning_rate": 1.230700579830768e-08,
+      "loss": 0.0013,
+      "step": 20609
+    },
+    {
+      "epoch": 4.684090909090909,
+      "grad_norm": 1.05762421834271,
+      "learning_rate": 1.2289387827116236e-08,
+      "loss": 0.0032,
+      "step": 20610
+    },
+    {
+      "epoch": 4.684318181818182,
+      "grad_norm": 0.6009176910403986,
+      "learning_rate": 1.2271782350163006e-08,
+      "loss": 0.0032,
+      "step": 20611
+    },
+    {
+      "epoch": 4.684545454545455,
+      "grad_norm": 0.40755011497009724,
+      "learning_rate": 1.2254189367807146e-08,
+      "loss": 0.0015,
+      "step": 20612
+    },
+    {
+      "epoch": 4.684772727272727,
+      "grad_norm": 1.1069608905690247,
+      "learning_rate": 1.2236608880407327e-08,
+      "loss": 0.0039,
+      "step": 20613
+    },
+    {
+      "epoch": 4.6850000000000005,
+      "grad_norm": 0.5918184418276531,
+      "learning_rate": 1.2219040888322154e-08,
+      "loss": 0.0039,
+      "step": 20614
+    },
+    {
+      "epoch": 4.685227272727273,
+      "grad_norm": 1.3184728670137442,
+      "learning_rate": 1.220148539190974e-08,
+      "loss": 0.0049,
+      "step": 20615
+    },
+    {
+      "epoch": 4.6854545454545455,
+      "grad_norm": 0.891306147103832,
+      "learning_rate": 1.2183942391528134e-08,
+      "loss": 0.0025,
+      "step": 20616
+    },
+    {
+      "epoch": 4.685681818181818,
+      "grad_norm": 0.12995080265904338,
+      "learning_rate": 1.2166411887535105e-08,
+      "loss": 0.0004,
+      "step": 20617
+    },
+    {
+      "epoch": 4.685909090909091,
+      "grad_norm": 3.2799541733695983,
+      "learning_rate": 1.2148893880288004e-08,
+      "loss": 0.0172,
+      "step": 20618
+    },
+    {
+      "epoch": 4.686136363636364,
+      "grad_norm": 3.9611847667645774,
+      "learning_rate": 1.2131388370144187e-08,
+      "loss": 0.0163,
+      "step": 20619
+    },
+    {
+      "epoch": 4.6863636363636365,
+      "grad_norm": 1.504986746214008,
+      "learning_rate": 1.2113895357460589e-08,
+      "loss": 0.0062,
+      "step": 20620
+    },
+    {
+      "epoch": 4.686590909090909,
+      "grad_norm": 1.2308950088810653,
+      "learning_rate": 1.20964148425938e-08,
+      "loss": 0.0035,
+      "step": 20621
+    },
+    {
+      "epoch": 4.6868181818181816,
+      "grad_norm": 0.5400861949473172,
+      "learning_rate": 1.207894682590048e-08,
+      "loss": 0.0016,
+      "step": 20622
+    },
+    {
+      "epoch": 4.687045454545455,
+      "grad_norm": 1.4022481239448488,
+      "learning_rate": 1.2061491307736664e-08,
+      "loss": 0.002,
+      "step": 20623
+    },
+    {
+      "epoch": 4.6872727272727275,
+      "grad_norm": 1.1392545762486477,
+      "learning_rate": 1.2044048288458454e-08,
+      "loss": 0.0048,
+      "step": 20624
+    },
+    {
+      "epoch": 4.6875,
+      "grad_norm": 2.7119491511832874,
+      "learning_rate": 1.20266177684214e-08,
+      "loss": 0.0092,
+      "step": 20625
+    },
+    {
+      "epoch": 4.6877272727272725,
+      "grad_norm": 1.577031289093678,
+      "learning_rate": 1.2009199747980981e-08,
+      "loss": 0.0085,
+      "step": 20626
+    },
+    {
+      "epoch": 4.687954545454545,
+      "grad_norm": 0.9000236895412632,
+      "learning_rate": 1.1991794227492401e-08,
+      "loss": 0.0046,
+      "step": 20627
+    },
+    {
+      "epoch": 4.6881818181818184,
+      "grad_norm": 4.244437456452733,
+      "learning_rate": 1.1974401207310582e-08,
+      "loss": 0.0119,
+      "step": 20628
+    },
+    {
+      "epoch": 4.688409090909091,
+      "grad_norm": 1.096492181986499,
+      "learning_rate": 1.195702068779031e-08,
+      "loss": 0.0026,
+      "step": 20629
+    },
+    {
+      "epoch": 4.6886363636363635,
+      "grad_norm": 3.015898999011623,
+      "learning_rate": 1.1939652669285818e-08,
+      "loss": 0.0046,
+      "step": 20630
+    },
+    {
+      "epoch": 4.688863636363636,
+      "grad_norm": 0.874765235912184,
+      "learning_rate": 1.1922297152151332e-08,
+      "loss": 0.0036,
+      "step": 20631
+    },
+    {
+      "epoch": 4.689090909090909,
+      "grad_norm": 2.0318563993886003,
+      "learning_rate": 1.190495413674081e-08,
+      "loss": 0.0161,
+      "step": 20632
+    },
+    {
+      "epoch": 4.689318181818182,
+      "grad_norm": 0.2815542359213111,
+      "learning_rate": 1.1887623623407924e-08,
+      "loss": 0.0009,
+      "step": 20633
+    },
+    {
+      "epoch": 4.6895454545454545,
+      "grad_norm": 1.1973424588955794,
+      "learning_rate": 1.1870305612506003e-08,
+      "loss": 0.0089,
+      "step": 20634
+    },
+    {
+      "epoch": 4.689772727272727,
+      "grad_norm": 0.5888525310954346,
+      "learning_rate": 1.1853000104388169e-08,
+      "loss": 0.0029,
+      "step": 20635
+    },
+    {
+      "epoch": 4.6899999999999995,
+      "grad_norm": 1.5690327845162253,
+      "learning_rate": 1.1835707099407332e-08,
+      "loss": 0.0038,
+      "step": 20636
+    },
+    {
+      "epoch": 4.690227272727273,
+      "grad_norm": 0.7288134091671394,
+      "learning_rate": 1.1818426597916267e-08,
+      "loss": 0.0027,
+      "step": 20637
+    },
+    {
+      "epoch": 4.690454545454545,
+      "grad_norm": 2.701918831454231,
+      "learning_rate": 1.180115860026712e-08,
+      "loss": 0.0298,
+      "step": 20638
+    },
+    {
+      "epoch": 4.690681818181818,
+      "grad_norm": 2.7128512636918933,
+      "learning_rate": 1.1783903106812183e-08,
+      "loss": 0.0154,
+      "step": 20639
+    },
+    {
+      "epoch": 4.690909090909091,
+      "grad_norm": 1.4843445128009942,
+      "learning_rate": 1.1766660117903255e-08,
+      "loss": 0.004,
+      "step": 20640
+    },
+    {
+      "epoch": 4.691136363636364,
+      "grad_norm": 0.340992400391068,
+      "learning_rate": 1.1749429633892001e-08,
+      "loss": 0.0011,
+      "step": 20641
+    },
+    {
+      "epoch": 4.691363636363636,
+      "grad_norm": 1.5799135510627125,
+      "learning_rate": 1.1732211655129735e-08,
+      "loss": 0.0085,
+      "step": 20642
+    },
+    {
+      "epoch": 4.691590909090909,
+      "grad_norm": 1.6167646298851022,
+      "learning_rate": 1.1715006181967567e-08,
+      "loss": 0.0038,
+      "step": 20643
+    },
+    {
+      "epoch": 4.691818181818181,
+      "grad_norm": 1.2313321891050415,
+      "learning_rate": 1.1697813214756326e-08,
+      "loss": 0.0088,
+      "step": 20644
+    },
+    {
+      "epoch": 4.692045454545455,
+      "grad_norm": 1.2766035855988953,
+      "learning_rate": 1.1680632753846704e-08,
+      "loss": 0.0089,
+      "step": 20645
+    },
+    {
+      "epoch": 4.692272727272727,
+      "grad_norm": 1.1636611655423972,
+      "learning_rate": 1.1663464799588978e-08,
+      "loss": 0.0034,
+      "step": 20646
+    },
+    {
+      "epoch": 4.6925,
+      "grad_norm": 0.9132789961688176,
+      "learning_rate": 1.1646309352333213e-08,
+      "loss": 0.0037,
+      "step": 20647
+    },
+    {
+      "epoch": 4.692727272727272,
+      "grad_norm": 1.4142247604040945,
+      "learning_rate": 1.1629166412429199e-08,
+      "loss": 0.007,
+      "step": 20648
+    },
+    {
+      "epoch": 4.692954545454546,
+      "grad_norm": 1.4453920476269455,
+      "learning_rate": 1.1612035980226658e-08,
+      "loss": 0.0083,
+      "step": 20649
+    },
+    {
+      "epoch": 4.693181818181818,
+      "grad_norm": 0.7486874381521975,
+      "learning_rate": 1.1594918056074753e-08,
+      "loss": 0.0062,
+      "step": 20650
+    },
+    {
+      "epoch": 4.693409090909091,
+      "grad_norm": 1.5328806789853122,
+      "learning_rate": 1.157781264032265e-08,
+      "loss": 0.0054,
+      "step": 20651
+    },
+    {
+      "epoch": 4.693636363636363,
+      "grad_norm": 0.33282204513753644,
+      "learning_rate": 1.1560719733319098e-08,
+      "loss": 0.0006,
+      "step": 20652
+    },
+    {
+      "epoch": 4.693863636363636,
+      "grad_norm": 0.2427815516904312,
+      "learning_rate": 1.1543639335412637e-08,
+      "loss": 0.0008,
+      "step": 20653
+    },
+    {
+      "epoch": 4.694090909090909,
+      "grad_norm": 0.5410196145422252,
+      "learning_rate": 1.152657144695167e-08,
+      "loss": 0.0016,
+      "step": 20654
+    },
+    {
+      "epoch": 4.694318181818182,
+      "grad_norm": 2.2723604222176284,
+      "learning_rate": 1.1509516068284183e-08,
+      "loss": 0.0099,
+      "step": 20655
+    },
+    {
+      "epoch": 4.694545454545454,
+      "grad_norm": 0.21923170824333676,
+      "learning_rate": 1.1492473199757881e-08,
+      "loss": 0.0007,
+      "step": 20656
+    },
+    {
+      "epoch": 4.694772727272728,
+      "grad_norm": 1.8836352840733837,
+      "learning_rate": 1.1475442841720405e-08,
+      "loss": 0.0067,
+      "step": 20657
+    },
+    {
+      "epoch": 4.695,
+      "grad_norm": 0.7381293846074484,
+      "learning_rate": 1.1458424994518976e-08,
+      "loss": 0.0059,
+      "step": 20658
+    },
+    {
+      "epoch": 4.695227272727273,
+      "grad_norm": 0.2044330640833906,
+      "learning_rate": 1.1441419658500748e-08,
+      "loss": 0.0005,
+      "step": 20659
+    },
+    {
+      "epoch": 4.695454545454545,
+      "grad_norm": 1.441150716391157,
+      "learning_rate": 1.1424426834012319e-08,
+      "loss": 0.0055,
+      "step": 20660
+    },
+    {
+      "epoch": 4.695681818181818,
+      "grad_norm": 2.0128901124997696,
+      "learning_rate": 1.1407446521400217e-08,
+      "loss": 0.0237,
+      "step": 20661
+    },
+    {
+      "epoch": 4.695909090909091,
+      "grad_norm": 3.5927164307714152,
+      "learning_rate": 1.1390478721010831e-08,
+      "loss": 0.0059,
+      "step": 20662
+    },
+    {
+      "epoch": 4.696136363636364,
+      "grad_norm": 0.6311543942932788,
+      "learning_rate": 1.1373523433189995e-08,
+      "loss": 0.0026,
+      "step": 20663
+    },
+    {
+      "epoch": 4.696363636363636,
+      "grad_norm": 1.085798967991149,
+      "learning_rate": 1.1356580658283684e-08,
+      "loss": 0.0082,
+      "step": 20664
+    },
+    {
+      "epoch": 4.696590909090909,
+      "grad_norm": 0.6364499217191188,
+      "learning_rate": 1.1339650396637175e-08,
+      "loss": 0.0045,
+      "step": 20665
+    },
+    {
+      "epoch": 4.696818181818182,
+      "grad_norm": 1.5656343219439761,
+      "learning_rate": 1.132273264859575e-08,
+      "loss": 0.0061,
+      "step": 20666
+    },
+    {
+      "epoch": 4.697045454545455,
+      "grad_norm": 0.6163686936705723,
+      "learning_rate": 1.1305827414504409e-08,
+      "loss": 0.0026,
+      "step": 20667
+    },
+    {
+      "epoch": 4.697272727272727,
+      "grad_norm": 1.5692265154819638,
+      "learning_rate": 1.1288934694708014e-08,
+      "loss": 0.0077,
+      "step": 20668
+    },
+    {
+      "epoch": 4.6975,
+      "grad_norm": 0.5292218005182903,
+      "learning_rate": 1.1272054489550807e-08,
+      "loss": 0.0012,
+      "step": 20669
+    },
+    {
+      "epoch": 4.697727272727272,
+      "grad_norm": 1.1016657678039352,
+      "learning_rate": 1.1255186799377163e-08,
+      "loss": 0.0098,
+      "step": 20670
+    },
+    {
+      "epoch": 4.697954545454546,
+      "grad_norm": 0.5360394317328436,
+      "learning_rate": 1.1238331624530973e-08,
+      "loss": 0.006,
+      "step": 20671
+    },
+    {
+      "epoch": 4.698181818181818,
+      "grad_norm": 0.6445888991665621,
+      "learning_rate": 1.1221488965355992e-08,
+      "loss": 0.0023,
+      "step": 20672
+    },
+    {
+      "epoch": 4.698409090909091,
+      "grad_norm": 1.1988292227373485,
+      "learning_rate": 1.1204658822195624e-08,
+      "loss": 0.0039,
+      "step": 20673
+    },
+    {
+      "epoch": 4.698636363636363,
+      "grad_norm": 0.8031239833450657,
+      "learning_rate": 1.1187841195393067e-08,
+      "loss": 0.0042,
+      "step": 20674
+    },
+    {
+      "epoch": 4.698863636363637,
+      "grad_norm": 1.368810623385112,
+      "learning_rate": 1.1171036085291312e-08,
+      "loss": 0.0054,
+      "step": 20675
+    },
+    {
+      "epoch": 4.699090909090909,
+      "grad_norm": 0.4807148119803926,
+      "learning_rate": 1.1154243492233069e-08,
+      "loss": 0.0013,
+      "step": 20676
+    },
+    {
+      "epoch": 4.699318181818182,
+      "grad_norm": 0.12435774567829644,
+      "learning_rate": 1.1137463416560703e-08,
+      "loss": 0.0004,
+      "step": 20677
+    },
+    {
+      "epoch": 4.699545454545454,
+      "grad_norm": 0.44303284865396053,
+      "learning_rate": 1.1120695858616304e-08,
+      "loss": 0.0011,
+      "step": 20678
+    },
+    {
+      "epoch": 4.699772727272728,
+      "grad_norm": 1.1892015869491839,
+      "learning_rate": 1.1103940818741956e-08,
+      "loss": 0.0046,
+      "step": 20679
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 1.243479285987481,
+      "learning_rate": 1.1087198297279265e-08,
+      "loss": 0.0065,
+      "step": 20680
+    },
+    {
+      "epoch": 4.700227272727273,
+      "grad_norm": 1.426558129781912,
+      "learning_rate": 1.107046829456962e-08,
+      "loss": 0.0101,
+      "step": 20681
+    },
+    {
+      "epoch": 4.700454545454545,
+      "grad_norm": 0.8058496359434362,
+      "learning_rate": 1.1053750810954139e-08,
+      "loss": 0.0051,
+      "step": 20682
+    },
+    {
+      "epoch": 4.700681818181819,
+      "grad_norm": 0.3943818365051008,
+      "learning_rate": 1.10370458467738e-08,
+      "loss": 0.0017,
+      "step": 20683
+    },
+    {
+      "epoch": 4.700909090909091,
+      "grad_norm": 1.0324432029937378,
+      "learning_rate": 1.1020353402369233e-08,
+      "loss": 0.0071,
+      "step": 20684
+    },
+    {
+      "epoch": 4.701136363636364,
+      "grad_norm": 0.4642195282977715,
+      "learning_rate": 1.100367347808079e-08,
+      "loss": 0.002,
+      "step": 20685
+    },
+    {
+      "epoch": 4.701363636363636,
+      "grad_norm": 1.0524395857830415,
+      "learning_rate": 1.0987006074248615e-08,
+      "loss": 0.0061,
+      "step": 20686
+    },
+    {
+      "epoch": 4.701590909090909,
+      "grad_norm": 1.2492381144263465,
+      "learning_rate": 1.0970351191212577e-08,
+      "loss": 0.0036,
+      "step": 20687
+    },
+    {
+      "epoch": 4.701818181818182,
+      "grad_norm": 0.805304997816066,
+      "learning_rate": 1.0953708829312334e-08,
+      "loss": 0.0036,
+      "step": 20688
+    },
+    {
+      "epoch": 4.702045454545455,
+      "grad_norm": 0.9194565177264126,
+      "learning_rate": 1.0937078988887198e-08,
+      "loss": 0.0035,
+      "step": 20689
+    },
+    {
+      "epoch": 4.702272727272727,
+      "grad_norm": 1.3373210122231125,
+      "learning_rate": 1.0920461670276341e-08,
+      "loss": 0.0044,
+      "step": 20690
+    },
+    {
+      "epoch": 4.7025,
+      "grad_norm": 1.374040838258778,
+      "learning_rate": 1.090385687381859e-08,
+      "loss": 0.0056,
+      "step": 20691
+    },
+    {
+      "epoch": 4.702727272727273,
+      "grad_norm": 1.2477001106779617,
+      "learning_rate": 1.0887264599852565e-08,
+      "loss": 0.0033,
+      "step": 20692
+    },
+    {
+      "epoch": 4.702954545454546,
+      "grad_norm": 0.16329922872758526,
+      "learning_rate": 1.0870684848716533e-08,
+      "loss": 0.0004,
+      "step": 20693
+    },
+    {
+      "epoch": 4.703181818181818,
+      "grad_norm": 0.38866785659178194,
+      "learning_rate": 1.08541176207487e-08,
+      "loss": 0.0007,
+      "step": 20694
+    },
+    {
+      "epoch": 4.703409090909091,
+      "grad_norm": 2.3649932927241144,
+      "learning_rate": 1.0837562916286848e-08,
+      "loss": 0.015,
+      "step": 20695
+    },
+    {
+      "epoch": 4.703636363636363,
+      "grad_norm": 0.3622448452019646,
+      "learning_rate": 1.0821020735668555e-08,
+      "loss": 0.0007,
+      "step": 20696
+    },
+    {
+      "epoch": 4.703863636363637,
+      "grad_norm": 1.030804269688041,
+      "learning_rate": 1.0804491079231122e-08,
+      "loss": 0.0064,
+      "step": 20697
+    },
+    {
+      "epoch": 4.704090909090909,
+      "grad_norm": 1.3580148700412407,
+      "learning_rate": 1.078797394731164e-08,
+      "loss": 0.0091,
+      "step": 20698
+    },
+    {
+      "epoch": 4.704318181818182,
+      "grad_norm": 0.7545084871229778,
+      "learning_rate": 1.077146934024699e-08,
+      "loss": 0.0029,
+      "step": 20699
+    },
+    {
+      "epoch": 4.704545454545455,
+      "grad_norm": 0.4811496707011908,
+      "learning_rate": 1.0754977258373572e-08,
+      "loss": 0.0022,
+      "step": 20700
+    },
+    {
+      "epoch": 4.7047727272727276,
+      "grad_norm": 0.07988361979841331,
+      "learning_rate": 1.0738497702027783e-08,
+      "loss": 0.0002,
+      "step": 20701
+    },
+    {
+      "epoch": 4.705,
+      "grad_norm": 0.4854181478017579,
+      "learning_rate": 1.0722030671545675e-08,
+      "loss": 0.0014,
+      "step": 20702
+    },
+    {
+      "epoch": 4.705227272727273,
+      "grad_norm": 2.2336774949980476,
+      "learning_rate": 1.0705576167263018e-08,
+      "loss": 0.0182,
+      "step": 20703
+    },
+    {
+      "epoch": 4.705454545454545,
+      "grad_norm": 1.344334304589032,
+      "learning_rate": 1.0689134189515379e-08,
+      "loss": 0.0084,
+      "step": 20704
+    },
+    {
+      "epoch": 4.7056818181818185,
+      "grad_norm": 1.7416732809862356,
+      "learning_rate": 1.0672704738637976e-08,
+      "loss": 0.0109,
+      "step": 20705
+    },
+    {
+      "epoch": 4.705909090909091,
+      "grad_norm": 0.6158581888882708,
+      "learning_rate": 1.0656287814965888e-08,
+      "loss": 0.0028,
+      "step": 20706
+    },
+    {
+      "epoch": 4.706136363636364,
+      "grad_norm": 0.7967458348573633,
+      "learning_rate": 1.0639883418833915e-08,
+      "loss": 0.0024,
+      "step": 20707
+    },
+    {
+      "epoch": 4.706363636363636,
+      "grad_norm": 1.145434665487577,
+      "learning_rate": 1.0623491550576514e-08,
+      "loss": 0.0039,
+      "step": 20708
+    },
+    {
+      "epoch": 4.7065909090909095,
+      "grad_norm": 1.173997980891536,
+      "learning_rate": 1.0607112210528e-08,
+      "loss": 0.0035,
+      "step": 20709
+    },
+    {
+      "epoch": 4.706818181818182,
+      "grad_norm": 3.8563444453095816,
+      "learning_rate": 1.0590745399022273e-08,
+      "loss": 0.0127,
+      "step": 20710
+    },
+    {
+      "epoch": 4.7070454545454545,
+      "grad_norm": 0.45225134386534777,
+      "learning_rate": 1.0574391116393163e-08,
+      "loss": 0.0035,
+      "step": 20711
+    },
+    {
+      "epoch": 4.707272727272727,
+      "grad_norm": 1.008805082728042,
+      "learning_rate": 1.0558049362974152e-08,
+      "loss": 0.0067,
+      "step": 20712
+    },
+    {
+      "epoch": 4.7075,
+      "grad_norm": 1.1187980874233958,
+      "learning_rate": 1.0541720139098446e-08,
+      "loss": 0.0098,
+      "step": 20713
+    },
+    {
+      "epoch": 4.707727272727273,
+      "grad_norm": 1.6577299164902528,
+      "learning_rate": 1.0525403445099045e-08,
+      "loss": 0.0105,
+      "step": 20714
+    },
+    {
+      "epoch": 4.7079545454545455,
+      "grad_norm": 0.730406038841891,
+      "learning_rate": 1.0509099281308666e-08,
+      "loss": 0.0031,
+      "step": 20715
+    },
+    {
+      "epoch": 4.708181818181818,
+      "grad_norm": 1.331147313430137,
+      "learning_rate": 1.0492807648059822e-08,
+      "loss": 0.004,
+      "step": 20716
+    },
+    {
+      "epoch": 4.7084090909090905,
+      "grad_norm": 0.8175831359534805,
+      "learning_rate": 1.0476528545684678e-08,
+      "loss": 0.0046,
+      "step": 20717
+    },
+    {
+      "epoch": 4.708636363636364,
+      "grad_norm": 1.1126416092817148,
+      "learning_rate": 1.046026197451526e-08,
+      "loss": 0.0016,
+      "step": 20718
+    },
+    {
+      "epoch": 4.7088636363636365,
+      "grad_norm": 0.318708155941551,
+      "learning_rate": 1.0444007934883177e-08,
+      "loss": 0.0006,
+      "step": 20719
+    },
+    {
+      "epoch": 4.709090909090909,
+      "grad_norm": 1.0653058746552468,
+      "learning_rate": 1.0427766427119901e-08,
+      "loss": 0.0045,
+      "step": 20720
+    },
+    {
+      "epoch": 4.7093181818181815,
+      "grad_norm": 1.1625410452937655,
+      "learning_rate": 1.0411537451556694e-08,
+      "loss": 0.0038,
+      "step": 20721
+    },
+    {
+      "epoch": 4.709545454545455,
+      "grad_norm": 2.6649117512832574,
+      "learning_rate": 1.0395321008524473e-08,
+      "loss": 0.0397,
+      "step": 20722
+    },
+    {
+      "epoch": 4.709772727272727,
+      "grad_norm": 1.602041047546573,
+      "learning_rate": 1.0379117098353805e-08,
+      "loss": 0.0168,
+      "step": 20723
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 2.0356094251690577,
+      "learning_rate": 1.0362925721375261e-08,
+      "loss": 0.011,
+      "step": 20724
+    },
+    {
+      "epoch": 4.7102272727272725,
+      "grad_norm": 2.610917990680262,
+      "learning_rate": 1.0346746877918993e-08,
+      "loss": 0.0076,
+      "step": 20725
+    },
+    {
+      "epoch": 4.710454545454546,
+      "grad_norm": 1.7424370902485602,
+      "learning_rate": 1.0330580568314804e-08,
+      "loss": 0.0071,
+      "step": 20726
+    },
+    {
+      "epoch": 4.710681818181818,
+      "grad_norm": 0.7097549782729607,
+      "learning_rate": 1.0314426792892435e-08,
+      "loss": 0.0022,
+      "step": 20727
+    },
+    {
+      "epoch": 4.710909090909091,
+      "grad_norm": 2.0341357753567,
+      "learning_rate": 1.029828555198134e-08,
+      "loss": 0.0101,
+      "step": 20728
+    },
+    {
+      "epoch": 4.7111363636363635,
+      "grad_norm": 2.413942675396229,
+      "learning_rate": 1.0282156845910565e-08,
+      "loss": 0.0101,
+      "step": 20729
+    },
+    {
+      "epoch": 4.711363636363636,
+      "grad_norm": 1.069728518357844,
+      "learning_rate": 1.0266040675009081e-08,
+      "loss": 0.0025,
+      "step": 20730
+    },
+    {
+      "epoch": 4.711590909090909,
+      "grad_norm": 1.1215269452986216,
+      "learning_rate": 1.0249937039605445e-08,
+      "loss": 0.0052,
+      "step": 20731
+    },
+    {
+      "epoch": 4.711818181818182,
+      "grad_norm": 0.7092318019510793,
+      "learning_rate": 1.0233845940028075e-08,
+      "loss": 0.0024,
+      "step": 20732
+    },
+    {
+      "epoch": 4.712045454545454,
+      "grad_norm": 0.4075468122077592,
+      "learning_rate": 1.0217767376605112e-08,
+      "loss": 0.0015,
+      "step": 20733
+    },
+    {
+      "epoch": 4.712272727272727,
+      "grad_norm": 0.6436006869486732,
+      "learning_rate": 1.0201701349664417e-08,
+      "loss": 0.0058,
+      "step": 20734
+    },
+    {
+      "epoch": 4.7125,
+      "grad_norm": 1.5990638560042338,
+      "learning_rate": 1.0185647859533646e-08,
+      "loss": 0.0038,
+      "step": 20735
+    },
+    {
+      "epoch": 4.712727272727273,
+      "grad_norm": 0.8210105165256325,
+      "learning_rate": 1.0169606906540036e-08,
+      "loss": 0.0079,
+      "step": 20736
+    },
+    {
+      "epoch": 4.712954545454545,
+      "grad_norm": 0.499929117388147,
+      "learning_rate": 1.0153578491010827e-08,
+      "loss": 0.0012,
+      "step": 20737
+    },
+    {
+      "epoch": 4.713181818181818,
+      "grad_norm": 0.4786709253038024,
+      "learning_rate": 1.0137562613272838e-08,
+      "loss": 0.0025,
+      "step": 20738
+    },
+    {
+      "epoch": 4.71340909090909,
+      "grad_norm": 3.4569711599970026,
+      "learning_rate": 1.0121559273652618e-08,
+      "loss": 0.0158,
+      "step": 20739
+    },
+    {
+      "epoch": 4.713636363636364,
+      "grad_norm": 0.9767042018966672,
+      "learning_rate": 1.01055684724765e-08,
+      "loss": 0.0084,
+      "step": 20740
+    },
+    {
+      "epoch": 4.713863636363636,
+      "grad_norm": 1.50940984420847,
+      "learning_rate": 1.008959021007061e-08,
+      "loss": 0.0039,
+      "step": 20741
+    },
+    {
+      "epoch": 4.714090909090909,
+      "grad_norm": 1.263760186816864,
+      "learning_rate": 1.0073624486760734e-08,
+      "loss": 0.0045,
+      "step": 20742
+    },
+    {
+      "epoch": 4.714318181818182,
+      "grad_norm": 0.9758608403346037,
+      "learning_rate": 1.005767130287251e-08,
+      "loss": 0.0056,
+      "step": 20743
+    },
+    {
+      "epoch": 4.714545454545455,
+      "grad_norm": 0.6342224547061484,
+      "learning_rate": 1.0041730658731097e-08,
+      "loss": 0.0022,
+      "step": 20744
+    },
+    {
+      "epoch": 4.714772727272727,
+      "grad_norm": 3.386364635810552,
+      "learning_rate": 1.0025802554661651e-08,
+      "loss": 0.0108,
+      "step": 20745
+    },
+    {
+      "epoch": 4.715,
+      "grad_norm": 0.49459837042042193,
+      "learning_rate": 1.0009886990988981e-08,
+      "loss": 0.001,
+      "step": 20746
+    },
+    {
+      "epoch": 4.715227272727272,
+      "grad_norm": 2.6947388425631327,
+      "learning_rate": 9.99398396803769e-09,
+      "loss": 0.0176,
+      "step": 20747
+    },
+    {
+      "epoch": 4.715454545454546,
+      "grad_norm": 1.1129637571731332,
+      "learning_rate": 9.978093486131961e-09,
+      "loss": 0.0086,
+      "step": 20748
+    },
+    {
+      "epoch": 4.715681818181818,
+      "grad_norm": 0.3649740166390375,
+      "learning_rate": 9.962215545595844e-09,
+      "loss": 0.0007,
+      "step": 20749
+    },
+    {
+      "epoch": 4.715909090909091,
+      "grad_norm": 0.6824125354356656,
+      "learning_rate": 9.946350146753172e-09,
+      "loss": 0.0046,
+      "step": 20750
+    },
+    {
+      "epoch": 4.716136363636363,
+      "grad_norm": 0.9190630583397678,
+      "learning_rate": 9.93049728992744e-09,
+      "loss": 0.0055,
+      "step": 20751
+    },
+    {
+      "epoch": 4.716363636363637,
+      "grad_norm": 1.602647205360712,
+      "learning_rate": 9.91465697544193e-09,
+      "loss": 0.0055,
+      "step": 20752
+    },
+    {
+      "epoch": 4.716590909090909,
+      "grad_norm": 0.8953578941869361,
+      "learning_rate": 9.898829203619648e-09,
+      "loss": 0.0018,
+      "step": 20753
+    },
+    {
+      "epoch": 4.716818181818182,
+      "grad_norm": 1.7711989375203192,
+      "learning_rate": 9.883013974783251e-09,
+      "loss": 0.0156,
+      "step": 20754
+    },
+    {
+      "epoch": 4.717045454545454,
+      "grad_norm": 0.829314209958302,
+      "learning_rate": 9.8672112892554e-09,
+      "loss": 0.0063,
+      "step": 20755
+    },
+    {
+      "epoch": 4.717272727272727,
+      "grad_norm": 2.379956274996303,
+      "learning_rate": 9.851421147358334e-09,
+      "loss": 0.0149,
+      "step": 20756
+    },
+    {
+      "epoch": 4.7175,
+      "grad_norm": 0.31208235750695384,
+      "learning_rate": 9.83564354941388e-09,
+      "loss": 0.0015,
+      "step": 20757
+    },
+    {
+      "epoch": 4.717727272727273,
+      "grad_norm": 2.8947502051972034,
+      "learning_rate": 9.819878495743867e-09,
+      "loss": 0.0061,
+      "step": 20758
+    },
+    {
+      "epoch": 4.717954545454545,
+      "grad_norm": 0.7216156688016349,
+      "learning_rate": 9.80412598666984e-09,
+      "loss": 0.002,
+      "step": 20759
+    },
+    {
+      "epoch": 4.718181818181818,
+      "grad_norm": 0.7786088806949987,
+      "learning_rate": 9.788386022512933e-09,
+      "loss": 0.0031,
+      "step": 20760
+    },
+    {
+      "epoch": 4.718409090909091,
+      "grad_norm": 1.1461507221890153,
+      "learning_rate": 9.772658603594068e-09,
+      "loss": 0.0052,
+      "step": 20761
+    },
+    {
+      "epoch": 4.718636363636364,
+      "grad_norm": 0.7115202715502875,
+      "learning_rate": 9.7569437302341e-09,
+      "loss": 0.0029,
+      "step": 20762
+    },
+    {
+      "epoch": 4.718863636363636,
+      "grad_norm": 2.124753715465928,
+      "learning_rate": 9.741241402753329e-09,
+      "loss": 0.0078,
+      "step": 20763
+    },
+    {
+      "epoch": 4.719090909090909,
+      "grad_norm": 0.2590346761026701,
+      "learning_rate": 9.725551621472053e-09,
+      "loss": 0.0006,
+      "step": 20764
+    },
+    {
+      "epoch": 4.719318181818182,
+      "grad_norm": 1.3153389956074917,
+      "learning_rate": 9.709874386710158e-09,
+      "loss": 0.0137,
+      "step": 20765
+    },
+    {
+      "epoch": 4.719545454545455,
+      "grad_norm": 2.0501938121700585,
+      "learning_rate": 9.694209698787386e-09,
+      "loss": 0.0122,
+      "step": 20766
+    },
+    {
+      "epoch": 4.719772727272727,
+      "grad_norm": 0.500026169695284,
+      "learning_rate": 9.678557558023133e-09,
+      "loss": 0.003,
+      "step": 20767
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 1.539348250672017,
+      "learning_rate": 9.66291796473659e-09,
+      "loss": 0.0038,
+      "step": 20768
+    },
+    {
+      "epoch": 4.720227272727273,
+      "grad_norm": 1.8396009077259727,
+      "learning_rate": 9.647290919246668e-09,
+      "loss": 0.0036,
+      "step": 20769
+    },
+    {
+      "epoch": 4.720454545454546,
+      "grad_norm": 0.4219886087647318,
+      "learning_rate": 9.631676421872003e-09,
+      "loss": 0.001,
+      "step": 20770
+    },
+    {
+      "epoch": 4.720681818181818,
+      "grad_norm": 2.0436580881596043,
+      "learning_rate": 9.616074472931016e-09,
+      "loss": 0.0131,
+      "step": 20771
+    },
+    {
+      "epoch": 4.720909090909091,
+      "grad_norm": 0.4988106401412272,
+      "learning_rate": 9.60048507274186e-09,
+      "loss": 0.0022,
+      "step": 20772
+    },
+    {
+      "epoch": 4.721136363636363,
+      "grad_norm": 3.8364501416845807,
+      "learning_rate": 9.584908221622475e-09,
+      "loss": 0.0062,
+      "step": 20773
+    },
+    {
+      "epoch": 4.721363636363637,
+      "grad_norm": 1.033850745218083,
+      "learning_rate": 9.569343919890519e-09,
+      "loss": 0.0065,
+      "step": 20774
+    },
+    {
+      "epoch": 4.721590909090909,
+      "grad_norm": 1.1867438194529472,
+      "learning_rate": 9.553792167863243e-09,
+      "loss": 0.0034,
+      "step": 20775
+    },
+    {
+      "epoch": 4.721818181818182,
+      "grad_norm": 0.40846737159660174,
+      "learning_rate": 9.538252965857823e-09,
+      "loss": 0.0032,
+      "step": 20776
+    },
+    {
+      "epoch": 4.722045454545454,
+      "grad_norm": 1.3560444209217892,
+      "learning_rate": 9.522726314191156e-09,
+      "loss": 0.0071,
+      "step": 20777
+    },
+    {
+      "epoch": 4.722272727272728,
+      "grad_norm": 0.3201233507806434,
+      "learning_rate": 9.507212213179934e-09,
+      "loss": 0.0018,
+      "step": 20778
+    },
+    {
+      "epoch": 4.7225,
+      "grad_norm": 2.1954438365861426,
+      "learning_rate": 9.491710663140364e-09,
+      "loss": 0.005,
+      "step": 20779
+    },
+    {
+      "epoch": 4.722727272727273,
+      "grad_norm": 2.390662422521881,
+      "learning_rate": 9.476221664388648e-09,
+      "loss": 0.0203,
+      "step": 20780
+    },
+    {
+      "epoch": 4.722954545454545,
+      "grad_norm": 1.439981045991122,
+      "learning_rate": 9.46074521724065e-09,
+      "loss": 0.0074,
+      "step": 20781
+    },
+    {
+      "epoch": 4.723181818181818,
+      "grad_norm": 1.1248003583029123,
+      "learning_rate": 9.445281322011877e-09,
+      "loss": 0.0045,
+      "step": 20782
+    },
+    {
+      "epoch": 4.723409090909091,
+      "grad_norm": 1.7283304775879995,
+      "learning_rate": 9.42982997901777e-09,
+      "loss": 0.009,
+      "step": 20783
+    },
+    {
+      "epoch": 4.723636363636364,
+      "grad_norm": 0.7073118066830586,
+      "learning_rate": 9.41439118857329e-09,
+      "loss": 0.002,
+      "step": 20784
+    },
+    {
+      "epoch": 4.723863636363636,
+      "grad_norm": 0.7856620811906824,
+      "learning_rate": 9.398964950993323e-09,
+      "loss": 0.0064,
+      "step": 20785
+    },
+    {
+      "epoch": 4.72409090909091,
+      "grad_norm": 1.322286056172406,
+      "learning_rate": 9.383551266592475e-09,
+      "loss": 0.0031,
+      "step": 20786
+    },
+    {
+      "epoch": 4.724318181818182,
+      "grad_norm": 1.7023608874716871,
+      "learning_rate": 9.368150135685011e-09,
+      "loss": 0.0137,
+      "step": 20787
+    },
+    {
+      "epoch": 4.724545454545455,
+      "grad_norm": 1.6602851822836957,
+      "learning_rate": 9.352761558584985e-09,
+      "loss": 0.0059,
+      "step": 20788
+    },
+    {
+      "epoch": 4.724772727272727,
+      "grad_norm": 0.6896873332848041,
+      "learning_rate": 9.337385535606244e-09,
+      "loss": 0.0023,
+      "step": 20789
+    },
+    {
+      "epoch": 4.725,
+      "grad_norm": 1.8560233686157706,
+      "learning_rate": 9.322022067062217e-09,
+      "loss": 0.0077,
+      "step": 20790
+    },
+    {
+      "epoch": 4.725227272727273,
+      "grad_norm": 0.5440008007451316,
+      "learning_rate": 9.306671153266333e-09,
+      "loss": 0.0015,
+      "step": 20791
+    },
+    {
+      "epoch": 4.725454545454546,
+      "grad_norm": 0.7891078388181847,
+      "learning_rate": 9.291332794531538e-09,
+      "loss": 0.0032,
+      "step": 20792
+    },
+    {
+      "epoch": 4.725681818181818,
+      "grad_norm": 1.3735191065929746,
+      "learning_rate": 9.276006991170636e-09,
+      "loss": 0.0104,
+      "step": 20793
+    },
+    {
+      "epoch": 4.725909090909091,
+      "grad_norm": 0.9417996322370654,
+      "learning_rate": 9.260693743496086e-09,
+      "loss": 0.0056,
+      "step": 20794
+    },
+    {
+      "epoch": 4.726136363636364,
+      "grad_norm": 1.2663658944170557,
+      "learning_rate": 9.245393051820206e-09,
+      "loss": 0.0067,
+      "step": 20795
+    },
+    {
+      "epoch": 4.7263636363636365,
+      "grad_norm": 1.4271697261504415,
+      "learning_rate": 9.23010491645511e-09,
+      "loss": 0.0099,
+      "step": 20796
+    },
+    {
+      "epoch": 4.726590909090909,
+      "grad_norm": 1.361214564657646,
+      "learning_rate": 9.214829337712422e-09,
+      "loss": 0.0075,
+      "step": 20797
+    },
+    {
+      "epoch": 4.726818181818182,
+      "grad_norm": 3.2276211608543175,
+      "learning_rate": 9.19956631590356e-09,
+      "loss": 0.0143,
+      "step": 20798
+    },
+    {
+      "epoch": 4.727045454545454,
+      "grad_norm": 1.0256036576533334,
+      "learning_rate": 9.184315851339942e-09,
+      "loss": 0.0053,
+      "step": 20799
+    },
+    {
+      "epoch": 4.7272727272727275,
+      "grad_norm": 0.9466319660967275,
+      "learning_rate": 9.1690779443325e-09,
+      "loss": 0.0042,
+      "step": 20800
+    },
+    {
+      "epoch": 4.7275,
+      "grad_norm": 1.5203696457991935,
+      "learning_rate": 9.153852595191887e-09,
+      "loss": 0.0091,
+      "step": 20801
+    },
+    {
+      "epoch": 4.727727272727273,
+      "grad_norm": 1.1082041940260123,
+      "learning_rate": 9.13863980422862e-09,
+      "loss": 0.0077,
+      "step": 20802
+    },
+    {
+      "epoch": 4.727954545454545,
+      "grad_norm": 0.8535181458076451,
+      "learning_rate": 9.123439571752935e-09,
+      "loss": 0.0049,
+      "step": 20803
+    },
+    {
+      "epoch": 4.7281818181818185,
+      "grad_norm": 0.8446644197290586,
+      "learning_rate": 9.108251898074794e-09,
+      "loss": 0.0039,
+      "step": 20804
+    },
+    {
+      "epoch": 4.728409090909091,
+      "grad_norm": 0.8143540511754087,
+      "learning_rate": 9.09307678350381e-09,
+      "loss": 0.0037,
+      "step": 20805
+    },
+    {
+      "epoch": 4.7286363636363635,
+      "grad_norm": 0.8292054238672023,
+      "learning_rate": 9.077914228349527e-09,
+      "loss": 0.0045,
+      "step": 20806
+    },
+    {
+      "epoch": 4.728863636363636,
+      "grad_norm": 0.6020056054133207,
+      "learning_rate": 9.062764232921142e-09,
+      "loss": 0.0021,
+      "step": 20807
+    },
+    {
+      "epoch": 4.7290909090909095,
+      "grad_norm": 0.4060961688861924,
+      "learning_rate": 9.047626797527505e-09,
+      "loss": 0.0013,
+      "step": 20808
+    },
+    {
+      "epoch": 4.729318181818182,
+      "grad_norm": 0.8686784478533496,
+      "learning_rate": 9.0325019224774e-09,
+      "loss": 0.0046,
+      "step": 20809
+    },
+    {
+      "epoch": 4.7295454545454545,
+      "grad_norm": 1.416821371209233,
+      "learning_rate": 9.017389608079117e-09,
+      "loss": 0.0073,
+      "step": 20810
+    },
+    {
+      "epoch": 4.729772727272727,
+      "grad_norm": 0.3664735001244591,
+      "learning_rate": 9.002289854640952e-09,
+      "loss": 0.003,
+      "step": 20811
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 0.7314805921314824,
+      "learning_rate": 8.987202662470787e-09,
+      "loss": 0.0029,
+      "step": 20812
+    },
+    {
+      "epoch": 4.730227272727273,
+      "grad_norm": 0.4146044274856041,
+      "learning_rate": 8.97212803187622e-09,
+      "loss": 0.0014,
+      "step": 20813
+    },
+    {
+      "epoch": 4.7304545454545455,
+      "grad_norm": 1.448745130681519,
+      "learning_rate": 8.957065963164716e-09,
+      "loss": 0.0048,
+      "step": 20814
+    },
+    {
+      "epoch": 4.730681818181818,
+      "grad_norm": 1.983767763437437,
+      "learning_rate": 8.942016456643318e-09,
+      "loss": 0.0197,
+      "step": 20815
+    },
+    {
+      "epoch": 4.7309090909090905,
+      "grad_norm": 0.4069128483202107,
+      "learning_rate": 8.926979512619075e-09,
+      "loss": 0.0037,
+      "step": 20816
+    },
+    {
+      "epoch": 4.731136363636364,
+      "grad_norm": 1.61969424727575,
+      "learning_rate": 8.911955131398406e-09,
+      "loss": 0.008,
+      "step": 20817
+    },
+    {
+      "epoch": 4.731363636363636,
+      "grad_norm": 0.2071292900551337,
+      "learning_rate": 8.896943313287873e-09,
+      "loss": 0.0011,
+      "step": 20818
+    },
+    {
+      "epoch": 4.731590909090909,
+      "grad_norm": 0.5332428364633827,
+      "learning_rate": 8.88194405859355e-09,
+      "loss": 0.0009,
+      "step": 20819
+    },
+    {
+      "epoch": 4.7318181818181815,
+      "grad_norm": 1.1705807490655284,
+      "learning_rate": 8.866957367621165e-09,
+      "loss": 0.007,
+      "step": 20820
+    },
+    {
+      "epoch": 4.732045454545455,
+      "grad_norm": 2.3705404131271273,
+      "learning_rate": 8.851983240676446e-09,
+      "loss": 0.0074,
+      "step": 20821
+    },
+    {
+      "epoch": 4.732272727272727,
+      "grad_norm": 0.5900860912394806,
+      "learning_rate": 8.837021678064772e-09,
+      "loss": 0.0025,
+      "step": 20822
+    },
+    {
+      "epoch": 4.7325,
+      "grad_norm": 1.6053361119657303,
+      "learning_rate": 8.822072680091178e-09,
+      "loss": 0.0078,
+      "step": 20823
+    },
+    {
+      "epoch": 4.7327272727272724,
+      "grad_norm": 0.9554018517069377,
+      "learning_rate": 8.807136247060422e-09,
+      "loss": 0.0038,
+      "step": 20824
+    },
+    {
+      "epoch": 4.732954545454545,
+      "grad_norm": 0.1990277819388118,
+      "learning_rate": 8.792212379277187e-09,
+      "loss": 0.0006,
+      "step": 20825
+    },
+    {
+      "epoch": 4.733181818181818,
+      "grad_norm": 1.9410078756217357,
+      "learning_rate": 8.777301077045816e-09,
+      "loss": 0.0126,
+      "step": 20826
+    },
+    {
+      "epoch": 4.733409090909091,
+      "grad_norm": 1.432762342622609,
+      "learning_rate": 8.76240234067037e-09,
+      "loss": 0.0075,
+      "step": 20827
+    },
+    {
+      "epoch": 4.733636363636363,
+      "grad_norm": 1.4489808361000198,
+      "learning_rate": 8.747516170454564e-09,
+      "loss": 0.0068,
+      "step": 20828
+    },
+    {
+      "epoch": 4.733863636363637,
+      "grad_norm": 1.334136227379228,
+      "learning_rate": 8.732642566701977e-09,
+      "loss": 0.0062,
+      "step": 20829
+    },
+    {
+      "epoch": 4.734090909090909,
+      "grad_norm": 1.2516229100221838,
+      "learning_rate": 8.717781529715974e-09,
+      "loss": 0.0072,
+      "step": 20830
+    },
+    {
+      "epoch": 4.734318181818182,
+      "grad_norm": 0.7550160173100816,
+      "learning_rate": 8.702933059799648e-09,
+      "loss": 0.0017,
+      "step": 20831
+    },
+    {
+      "epoch": 4.734545454545454,
+      "grad_norm": 6.540005705536809,
+      "learning_rate": 8.688097157255673e-09,
+      "loss": 0.0132,
+      "step": 20832
+    },
+    {
+      "epoch": 4.734772727272727,
+      "grad_norm": 1.0317882378128371,
+      "learning_rate": 8.673273822386585e-09,
+      "loss": 0.0072,
+      "step": 20833
+    },
+    {
+      "epoch": 4.735,
+      "grad_norm": 0.8639851266708087,
+      "learning_rate": 8.658463055494643e-09,
+      "loss": 0.0047,
+      "step": 20834
+    },
+    {
+      "epoch": 4.735227272727273,
+      "grad_norm": 0.4154734081270038,
+      "learning_rate": 8.643664856882033e-09,
+      "loss": 0.0013,
+      "step": 20835
+    },
+    {
+      "epoch": 4.735454545454545,
+      "grad_norm": 0.34429029106132925,
+      "learning_rate": 8.628879226850253e-09,
+      "loss": 0.0012,
+      "step": 20836
+    },
+    {
+      "epoch": 4.735681818181818,
+      "grad_norm": 1.1175824908343486,
+      "learning_rate": 8.614106165701005e-09,
+      "loss": 0.012,
+      "step": 20837
+    },
+    {
+      "epoch": 4.735909090909091,
+      "grad_norm": 0.915521807138756,
+      "learning_rate": 8.599345673735508e-09,
+      "loss": 0.0056,
+      "step": 20838
+    },
+    {
+      "epoch": 4.736136363636364,
+      "grad_norm": 1.486596553940759,
+      "learning_rate": 8.584597751254628e-09,
+      "loss": 0.0034,
+      "step": 20839
+    },
+    {
+      "epoch": 4.736363636363636,
+      "grad_norm": 0.9669150975072517,
+      "learning_rate": 8.569862398559239e-09,
+      "loss": 0.0075,
+      "step": 20840
+    },
+    {
+      "epoch": 4.736590909090909,
+      "grad_norm": 1.2665555017044066,
+      "learning_rate": 8.555139615949795e-09,
+      "loss": 0.0057,
+      "step": 20841
+    },
+    {
+      "epoch": 4.736818181818181,
+      "grad_norm": 0.21531657698838771,
+      "learning_rate": 8.540429403726542e-09,
+      "loss": 0.0008,
+      "step": 20842
+    },
+    {
+      "epoch": 4.737045454545455,
+      "grad_norm": 1.5466722734935712,
+      "learning_rate": 8.525731762189306e-09,
+      "loss": 0.0026,
+      "step": 20843
+    },
+    {
+      "epoch": 4.737272727272727,
+      "grad_norm": 0.4703605950277923,
+      "learning_rate": 8.511046691637992e-09,
+      "loss": 0.0015,
+      "step": 20844
+    },
+    {
+      "epoch": 4.7375,
+      "grad_norm": 0.758519291979777,
+      "learning_rate": 8.49637419237194e-09,
+      "loss": 0.0028,
+      "step": 20845
+    },
+    {
+      "epoch": 4.737727272727272,
+      "grad_norm": 1.7468440197353452,
+      "learning_rate": 8.481714264690356e-09,
+      "loss": 0.0084,
+      "step": 20846
+    },
+    {
+      "epoch": 4.737954545454546,
+      "grad_norm": 1.2680396159523015,
+      "learning_rate": 8.467066908892166e-09,
+      "loss": 0.0028,
+      "step": 20847
+    },
+    {
+      "epoch": 4.738181818181818,
+      "grad_norm": 1.0991573416876004,
+      "learning_rate": 8.45243212527616e-09,
+      "loss": 0.0017,
+      "step": 20848
+    },
+    {
+      "epoch": 4.738409090909091,
+      "grad_norm": 0.25944334044219713,
+      "learning_rate": 8.437809914140575e-09,
+      "loss": 0.0009,
+      "step": 20849
+    },
+    {
+      "epoch": 4.738636363636363,
+      "grad_norm": 0.2738571299142838,
+      "learning_rate": 8.423200275783778e-09,
+      "loss": 0.0005,
+      "step": 20850
+    },
+    {
+      "epoch": 4.738863636363637,
+      "grad_norm": 1.5189547572116797,
+      "learning_rate": 8.408603210503588e-09,
+      "loss": 0.0113,
+      "step": 20851
+    },
+    {
+      "epoch": 4.739090909090909,
+      "grad_norm": 1.2122747307571666,
+      "learning_rate": 8.394018718597616e-09,
+      "loss": 0.003,
+      "step": 20852
+    },
+    {
+      "epoch": 4.739318181818182,
+      "grad_norm": 3.245600222025475,
+      "learning_rate": 8.3794468003634e-09,
+      "loss": 0.007,
+      "step": 20853
+    },
+    {
+      "epoch": 4.739545454545454,
+      "grad_norm": 1.853254576538358,
+      "learning_rate": 8.364887456097994e-09,
+      "loss": 0.0067,
+      "step": 20854
+    },
+    {
+      "epoch": 4.739772727272728,
+      "grad_norm": 0.8968011951693697,
+      "learning_rate": 8.350340686098315e-09,
+      "loss": 0.0046,
+      "step": 20855
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 2.7900439008548767,
+      "learning_rate": 8.335806490660999e-09,
+      "loss": 0.0213,
+      "step": 20856
+    },
+    {
+      "epoch": 4.740227272727273,
+      "grad_norm": 1.2524076716274382,
+      "learning_rate": 8.321284870082477e-09,
+      "loss": 0.0071,
+      "step": 20857
+    },
+    {
+      "epoch": 4.740454545454545,
+      "grad_norm": 0.5287893040256882,
+      "learning_rate": 8.306775824658761e-09,
+      "loss": 0.0022,
+      "step": 20858
+    },
+    {
+      "epoch": 4.740681818181818,
+      "grad_norm": 0.4242225642947342,
+      "learning_rate": 8.292279354685726e-09,
+      "loss": 0.0015,
+      "step": 20859
+    },
+    {
+      "epoch": 4.740909090909091,
+      "grad_norm": 1.515560027949728,
+      "learning_rate": 8.277795460459037e-09,
+      "loss": 0.0171,
+      "step": 20860
+    },
+    {
+      "epoch": 4.741136363636364,
+      "grad_norm": 0.43141207337634474,
+      "learning_rate": 8.263324142274084e-09,
+      "loss": 0.0015,
+      "step": 20861
+    },
+    {
+      "epoch": 4.741363636363636,
+      "grad_norm": 0.6261838962893296,
+      "learning_rate": 8.248865400425909e-09,
+      "loss": 0.002,
+      "step": 20862
+    },
+    {
+      "epoch": 4.741590909090909,
+      "grad_norm": 1.3920713789274324,
+      "learning_rate": 8.234419235209276e-09,
+      "loss": 0.0046,
+      "step": 20863
+    },
+    {
+      "epoch": 4.741818181818182,
+      "grad_norm": 0.352225448191239,
+      "learning_rate": 8.21998564691888e-09,
+      "loss": 0.0009,
+      "step": 20864
+    },
+    {
+      "epoch": 4.742045454545455,
+      "grad_norm": 1.9606009707803573,
+      "learning_rate": 8.205564635848998e-09,
+      "loss": 0.0133,
+      "step": 20865
+    },
+    {
+      "epoch": 4.742272727272727,
+      "grad_norm": 1.298875137894175,
+      "learning_rate": 8.19115620229377e-09,
+      "loss": 0.0047,
+      "step": 20866
+    },
+    {
+      "epoch": 4.7425,
+      "grad_norm": 1.5007019187711859,
+      "learning_rate": 8.176760346546921e-09,
+      "loss": 0.0104,
+      "step": 20867
+    },
+    {
+      "epoch": 4.742727272727272,
+      "grad_norm": 1.5736545488957352,
+      "learning_rate": 8.162377068901966e-09,
+      "loss": 0.0099,
+      "step": 20868
+    },
+    {
+      "epoch": 4.742954545454546,
+      "grad_norm": 1.221787040689112,
+      "learning_rate": 8.14800636965235e-09,
+      "loss": 0.007,
+      "step": 20869
+    },
+    {
+      "epoch": 4.743181818181818,
+      "grad_norm": 0.6849617279213503,
+      "learning_rate": 8.133648249090965e-09,
+      "loss": 0.0031,
+      "step": 20870
+    },
+    {
+      "epoch": 4.743409090909091,
+      "grad_norm": 1.7564274972978298,
+      "learning_rate": 8.119302707510698e-09,
+      "loss": 0.0054,
+      "step": 20871
+    },
+    {
+      "epoch": 4.743636363636364,
+      "grad_norm": 1.445650078922327,
+      "learning_rate": 8.104969745204098e-09,
+      "loss": 0.0055,
+      "step": 20872
+    },
+    {
+      "epoch": 4.743863636363637,
+      "grad_norm": 0.4372828672092051,
+      "learning_rate": 8.090649362463356e-09,
+      "loss": 0.0012,
+      "step": 20873
+    },
+    {
+      "epoch": 4.744090909090909,
+      "grad_norm": 0.45876640595563434,
+      "learning_rate": 8.076341559580606e-09,
+      "loss": 0.003,
+      "step": 20874
+    },
+    {
+      "epoch": 4.744318181818182,
+      "grad_norm": 1.6828080186489003,
+      "learning_rate": 8.062046336847485e-09,
+      "loss": 0.0051,
+      "step": 20875
+    },
+    {
+      "epoch": 4.744545454545454,
+      "grad_norm": 1.5389809166502988,
+      "learning_rate": 8.047763694555566e-09,
+      "loss": 0.0048,
+      "step": 20876
+    },
+    {
+      "epoch": 4.744772727272728,
+      "grad_norm": 1.1911523291399668,
+      "learning_rate": 8.033493632996006e-09,
+      "loss": 0.0047,
+      "step": 20877
+    },
+    {
+      "epoch": 4.745,
+      "grad_norm": 0.6225003068142642,
+      "learning_rate": 8.019236152459961e-09,
+      "loss": 0.0024,
+      "step": 20878
+    },
+    {
+      "epoch": 4.745227272727273,
+      "grad_norm": 1.4908289853678842,
+      "learning_rate": 8.004991253238031e-09,
+      "loss": 0.0022,
+      "step": 20879
+    },
+    {
+      "epoch": 4.745454545454545,
+      "grad_norm": 0.3233411641250432,
+      "learning_rate": 7.990758935620749e-09,
+      "loss": 0.0008,
+      "step": 20880
+    },
+    {
+      "epoch": 4.745681818181819,
+      "grad_norm": 0.6756061488280105,
+      "learning_rate": 7.976539199898298e-09,
+      "loss": 0.0021,
+      "step": 20881
+    },
+    {
+      "epoch": 4.745909090909091,
+      "grad_norm": 1.3770742471120165,
+      "learning_rate": 7.962332046360655e-09,
+      "loss": 0.0033,
+      "step": 20882
+    },
+    {
+      "epoch": 4.746136363636364,
+      "grad_norm": 1.5876673832159331,
+      "learning_rate": 7.948137475297588e-09,
+      "loss": 0.0042,
+      "step": 20883
+    },
+    {
+      "epoch": 4.746363636363636,
+      "grad_norm": 0.8114655868567411,
+      "learning_rate": 7.933955486998449e-09,
+      "loss": 0.0067,
+      "step": 20884
+    },
+    {
+      "epoch": 4.746590909090909,
+      "grad_norm": 1.1900748041980451,
+      "learning_rate": 7.91978608175252e-09,
+      "loss": 0.0101,
+      "step": 20885
+    },
+    {
+      "epoch": 4.746818181818182,
+      "grad_norm": 0.4952628912286291,
+      "learning_rate": 7.905629259848738e-09,
+      "loss": 0.002,
+      "step": 20886
+    },
+    {
+      "epoch": 4.747045454545455,
+      "grad_norm": 1.8382071001542357,
+      "learning_rate": 7.89148502157569e-09,
+      "loss": 0.0104,
+      "step": 20887
+    },
+    {
+      "epoch": 4.747272727272727,
+      "grad_norm": 0.2400627209734323,
+      "learning_rate": 7.877353367221966e-09,
+      "loss": 0.0009,
+      "step": 20888
+    },
+    {
+      "epoch": 4.7475000000000005,
+      "grad_norm": 0.22385979499021844,
+      "learning_rate": 7.863234297075598e-09,
+      "loss": 0.0006,
+      "step": 20889
+    },
+    {
+      "epoch": 4.747727272727273,
+      "grad_norm": 0.3725990282229779,
+      "learning_rate": 7.849127811424482e-09,
+      "loss": 0.0011,
+      "step": 20890
+    },
+    {
+      "epoch": 4.7479545454545455,
+      "grad_norm": 1.0935770992383602,
+      "learning_rate": 7.835033910556373e-09,
+      "loss": 0.0058,
+      "step": 20891
+    },
+    {
+      "epoch": 4.748181818181818,
+      "grad_norm": 1.138633255244419,
+      "learning_rate": 7.82095259475861e-09,
+      "loss": 0.006,
+      "step": 20892
+    },
+    {
+      "epoch": 4.748409090909091,
+      "grad_norm": 1.3931983827227055,
+      "learning_rate": 7.806883864318396e-09,
+      "loss": 0.011,
+      "step": 20893
+    },
+    {
+      "epoch": 4.748636363636364,
+      "grad_norm": 2.124969307819026,
+      "learning_rate": 7.792827719522513e-09,
+      "loss": 0.0056,
+      "step": 20894
+    },
+    {
+      "epoch": 4.7488636363636365,
+      "grad_norm": 0.557276048971298,
+      "learning_rate": 7.778784160657679e-09,
+      "loss": 0.0041,
+      "step": 20895
+    },
+    {
+      "epoch": 4.749090909090909,
+      "grad_norm": 1.0935937893593286,
+      "learning_rate": 7.76475318801019e-09,
+      "loss": 0.0061,
+      "step": 20896
+    },
+    {
+      "epoch": 4.7493181818181816,
+      "grad_norm": 0.3903559075372121,
+      "learning_rate": 7.750734801866278e-09,
+      "loss": 0.0021,
+      "step": 20897
+    },
+    {
+      "epoch": 4.749545454545455,
+      "grad_norm": 0.5505342679953515,
+      "learning_rate": 7.736729002511753e-09,
+      "loss": 0.0029,
+      "step": 20898
+    },
+    {
+      "epoch": 4.7497727272727275,
+      "grad_norm": 0.2772543965864388,
+      "learning_rate": 7.722735790232083e-09,
+      "loss": 0.001,
+      "step": 20899
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 0.6857780980686359,
+      "learning_rate": 7.708755165312804e-09,
+      "loss": 0.0039,
+      "step": 20900
+    },
+    {
+      "epoch": 4.7502272727272725,
+      "grad_norm": 0.48403117463831274,
+      "learning_rate": 7.694787128038963e-09,
+      "loss": 0.0011,
+      "step": 20901
+    },
+    {
+      "epoch": 4.750454545454545,
+      "grad_norm": 1.5372988436942627,
+      "learning_rate": 7.680831678695266e-09,
+      "loss": 0.0145,
+      "step": 20902
+    },
+    {
+      "epoch": 4.7506818181818184,
+      "grad_norm": 0.8562074018620122,
+      "learning_rate": 7.666888817566483e-09,
+      "loss": 0.0067,
+      "step": 20903
+    },
+    {
+      "epoch": 4.750909090909091,
+      "grad_norm": 1.7096220675240867,
+      "learning_rate": 7.652958544936765e-09,
+      "loss": 0.0065,
+      "step": 20904
+    },
+    {
+      "epoch": 4.7511363636363635,
+      "grad_norm": 0.2675647585976415,
+      "learning_rate": 7.639040861090325e-09,
+      "loss": 0.0007,
+      "step": 20905
+    },
+    {
+      "epoch": 4.751363636363636,
+      "grad_norm": 0.41993538873477737,
+      "learning_rate": 7.625135766310757e-09,
+      "loss": 0.0027,
+      "step": 20906
+    },
+    {
+      "epoch": 4.751590909090909,
+      "grad_norm": 1.1465147229782109,
+      "learning_rate": 7.611243260881862e-09,
+      "loss": 0.0063,
+      "step": 20907
+    },
+    {
+      "epoch": 4.751818181818182,
+      "grad_norm": 1.6127465612149792,
+      "learning_rate": 7.597363345086747e-09,
+      "loss": 0.0063,
+      "step": 20908
+    },
+    {
+      "epoch": 4.7520454545454545,
+      "grad_norm": 2.1283654761714135,
+      "learning_rate": 7.583496019208519e-09,
+      "loss": 0.0099,
+      "step": 20909
+    },
+    {
+      "epoch": 4.752272727272727,
+      "grad_norm": 1.4226998652930862,
+      "learning_rate": 7.569641283530008e-09,
+      "loss": 0.0038,
+      "step": 20910
+    },
+    {
+      "epoch": 4.7524999999999995,
+      "grad_norm": 0.10163282771694986,
+      "learning_rate": 7.555799138333555e-09,
+      "loss": 0.0003,
+      "step": 20911
+    },
+    {
+      "epoch": 4.752727272727273,
+      "grad_norm": 2.207971194621103,
+      "learning_rate": 7.541969583901645e-09,
+      "loss": 0.0176,
+      "step": 20912
+    },
+    {
+      "epoch": 4.752954545454545,
+      "grad_norm": 1.2247534536314553,
+      "learning_rate": 7.528152620516136e-09,
+      "loss": 0.0104,
+      "step": 20913
+    },
+    {
+      "epoch": 4.753181818181818,
+      "grad_norm": 1.581136990544135,
+      "learning_rate": 7.514348248458883e-09,
+      "loss": 0.0094,
+      "step": 20914
+    },
+    {
+      "epoch": 4.753409090909091,
+      "grad_norm": 2.2130796235893335,
+      "learning_rate": 7.50055646801133e-09,
+      "loss": 0.0049,
+      "step": 20915
+    },
+    {
+      "epoch": 4.753636363636364,
+      "grad_norm": 0.7872572249180493,
+      "learning_rate": 7.486777279454641e-09,
+      "loss": 0.0011,
+      "step": 20916
+    },
+    {
+      "epoch": 4.753863636363636,
+      "grad_norm": 0.3712861193548923,
+      "learning_rate": 7.473010683069979e-09,
+      "loss": 0.0018,
+      "step": 20917
+    },
+    {
+      "epoch": 4.754090909090909,
+      "grad_norm": 0.7016399094624921,
+      "learning_rate": 7.459256679137883e-09,
+      "loss": 0.0021,
+      "step": 20918
+    },
+    {
+      "epoch": 4.754318181818181,
+      "grad_norm": 1.491119110160955,
+      "learning_rate": 7.445515267938963e-09,
+      "loss": 0.011,
+      "step": 20919
+    },
+    {
+      "epoch": 4.754545454545455,
+      "grad_norm": 1.1669279235439474,
+      "learning_rate": 7.431786449753342e-09,
+      "loss": 0.0032,
+      "step": 20920
+    },
+    {
+      "epoch": 4.754772727272727,
+      "grad_norm": 2.8034625493189185,
+      "learning_rate": 7.418070224861004e-09,
+      "loss": 0.0163,
+      "step": 20921
+    },
+    {
+      "epoch": 4.755,
+      "grad_norm": 1.8345877458511393,
+      "learning_rate": 7.4043665935416566e-09,
+      "loss": 0.0064,
+      "step": 20922
+    },
+    {
+      "epoch": 4.755227272727272,
+      "grad_norm": 0.284655508598094,
+      "learning_rate": 7.390675556074728e-09,
+      "loss": 0.0027,
+      "step": 20923
+    },
+    {
+      "epoch": 4.755454545454546,
+      "grad_norm": 0.6006960048929482,
+      "learning_rate": 7.37699711273937e-09,
+      "loss": 0.002,
+      "step": 20924
+    },
+    {
+      "epoch": 4.755681818181818,
+      "grad_norm": 0.43078356536742624,
+      "learning_rate": 7.363331263814596e-09,
+      "loss": 0.0013,
+      "step": 20925
+    },
+    {
+      "epoch": 4.755909090909091,
+      "grad_norm": 0.7147765872236193,
+      "learning_rate": 7.349678009578934e-09,
+      "loss": 0.0034,
+      "step": 20926
+    },
+    {
+      "epoch": 4.756136363636363,
+      "grad_norm": 2.2638541337863898,
+      "learning_rate": 7.336037350310979e-09,
+      "loss": 0.0164,
+      "step": 20927
+    },
+    {
+      "epoch": 4.756363636363636,
+      "grad_norm": 0.5575727405312721,
+      "learning_rate": 7.322409286288773e-09,
+      "loss": 0.0017,
+      "step": 20928
+    },
+    {
+      "epoch": 4.756590909090909,
+      "grad_norm": 0.3122448693108503,
+      "learning_rate": 7.308793817790219e-09,
+      "loss": 0.0012,
+      "step": 20929
+    },
+    {
+      "epoch": 4.756818181818182,
+      "grad_norm": 0.19386885188978756,
+      "learning_rate": 7.295190945093011e-09,
+      "loss": 0.0005,
+      "step": 20930
+    },
+    {
+      "epoch": 4.757045454545454,
+      "grad_norm": 1.2705634245551574,
+      "learning_rate": 7.281600668474428e-09,
+      "loss": 0.0027,
+      "step": 20931
+    },
+    {
+      "epoch": 4.757272727272728,
+      "grad_norm": 2.5683979850270826,
+      "learning_rate": 7.268022988211748e-09,
+      "loss": 0.0069,
+      "step": 20932
+    },
+    {
+      "epoch": 4.7575,
+      "grad_norm": 0.2471762374318464,
+      "learning_rate": 7.254457904581763e-09,
+      "loss": 0.001,
+      "step": 20933
+    },
+    {
+      "epoch": 4.757727272727273,
+      "grad_norm": 0.9049866886258068,
+      "learning_rate": 7.240905417861058e-09,
+      "loss": 0.0029,
+      "step": 20934
+    },
+    {
+      "epoch": 4.757954545454545,
+      "grad_norm": 0.4988888391971366,
+      "learning_rate": 7.227365528326078e-09,
+      "loss": 0.0014,
+      "step": 20935
+    },
+    {
+      "epoch": 4.758181818181818,
+      "grad_norm": 0.5748220390450273,
+      "learning_rate": 7.213838236252921e-09,
+      "loss": 0.0024,
+      "step": 20936
+    },
+    {
+      "epoch": 4.758409090909091,
+      "grad_norm": 0.5827824990321995,
+      "learning_rate": 7.200323541917339e-09,
+      "loss": 0.0021,
+      "step": 20937
+    },
+    {
+      "epoch": 4.758636363636364,
+      "grad_norm": 0.14602167079489559,
+      "learning_rate": 7.186821445594947e-09,
+      "loss": 0.0003,
+      "step": 20938
+    },
+    {
+      "epoch": 4.758863636363636,
+      "grad_norm": 0.6017191885307104,
+      "learning_rate": 7.173331947561077e-09,
+      "loss": 0.0024,
+      "step": 20939
+    },
+    {
+      "epoch": 4.759090909090909,
+      "grad_norm": 0.33794990554994336,
+      "learning_rate": 7.159855048090858e-09,
+      "loss": 0.0017,
+      "step": 20940
+    },
+    {
+      "epoch": 4.759318181818182,
+      "grad_norm": 0.36119761205298384,
+      "learning_rate": 7.146390747459139e-09,
+      "loss": 0.0011,
+      "step": 20941
+    },
+    {
+      "epoch": 4.759545454545455,
+      "grad_norm": 1.1888095215957764,
+      "learning_rate": 7.132939045940354e-09,
+      "loss": 0.0102,
+      "step": 20942
+    },
+    {
+      "epoch": 4.759772727272727,
+      "grad_norm": 0.2528880709728548,
+      "learning_rate": 7.119499943808866e-09,
+      "loss": 0.001,
+      "step": 20943
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 0.45934181455243484,
+      "learning_rate": 7.1060734413387625e-09,
+      "loss": 0.0012,
+      "step": 20944
+    },
+    {
+      "epoch": 4.760227272727272,
+      "grad_norm": 0.4406633963290122,
+      "learning_rate": 7.092659538803781e-09,
+      "loss": 0.0006,
+      "step": 20945
+    },
+    {
+      "epoch": 4.760454545454546,
+      "grad_norm": 2.4171049887169156,
+      "learning_rate": 7.0792582364774545e-09,
+      "loss": 0.0075,
+      "step": 20946
+    },
+    {
+      "epoch": 4.760681818181818,
+      "grad_norm": 2.885844453618077,
+      "learning_rate": 7.065869534633036e-09,
+      "loss": 0.0071,
+      "step": 20947
+    },
+    {
+      "epoch": 4.760909090909091,
+      "grad_norm": 0.20838441823237178,
+      "learning_rate": 7.052493433543639e-09,
+      "loss": 0.0007,
+      "step": 20948
+    },
+    {
+      "epoch": 4.761136363636363,
+      "grad_norm": 1.3057977014195141,
+      "learning_rate": 7.0391299334819644e-09,
+      "loss": 0.0048,
+      "step": 20949
+    },
+    {
+      "epoch": 4.761363636363637,
+      "grad_norm": 0.365286954260398,
+      "learning_rate": 7.02577903472057e-09,
+      "loss": 0.0017,
+      "step": 20950
+    },
+    {
+      "epoch": 4.761590909090909,
+      "grad_norm": 0.2335956701066095,
+      "learning_rate": 7.012440737531531e-09,
+      "loss": 0.003,
+      "step": 20951
+    },
+    {
+      "epoch": 4.761818181818182,
+      "grad_norm": 0.5591057265109656,
+      "learning_rate": 6.9991150421870605e-09,
+      "loss": 0.0021,
+      "step": 20952
+    },
+    {
+      "epoch": 4.762045454545454,
+      "grad_norm": 2.1832218730690904,
+      "learning_rate": 6.985801948958745e-09,
+      "loss": 0.0119,
+      "step": 20953
+    },
+    {
+      "epoch": 4.762272727272728,
+      "grad_norm": 0.9141307889185911,
+      "learning_rate": 6.972501458118175e-09,
+      "loss": 0.0023,
+      "step": 20954
+    },
+    {
+      "epoch": 4.7625,
+      "grad_norm": 1.1714584550976563,
+      "learning_rate": 6.959213569936521e-09,
+      "loss": 0.0033,
+      "step": 20955
+    },
+    {
+      "epoch": 4.762727272727273,
+      "grad_norm": 1.714258268687425,
+      "learning_rate": 6.945938284684678e-09,
+      "loss": 0.0122,
+      "step": 20956
+    },
+    {
+      "epoch": 4.762954545454545,
+      "grad_norm": 0.8726859269625069,
+      "learning_rate": 6.932675602633401e-09,
+      "loss": 0.0047,
+      "step": 20957
+    },
+    {
+      "epoch": 4.763181818181819,
+      "grad_norm": 1.3411229459848002,
+      "learning_rate": 6.9194255240531696e-09,
+      "loss": 0.0044,
+      "step": 20958
+    },
+    {
+      "epoch": 4.763409090909091,
+      "grad_norm": 1.6399436879881681,
+      "learning_rate": 6.906188049214113e-09,
+      "loss": 0.0145,
+      "step": 20959
+    },
+    {
+      "epoch": 4.763636363636364,
+      "grad_norm": 1.5423157216447612,
+      "learning_rate": 6.892963178386294e-09,
+      "loss": 0.0104,
+      "step": 20960
+    },
+    {
+      "epoch": 4.763863636363636,
+      "grad_norm": 0.5795400444529918,
+      "learning_rate": 6.87975091183922e-09,
+      "loss": 0.0028,
+      "step": 20961
+    },
+    {
+      "epoch": 4.764090909090909,
+      "grad_norm": 2.597532072978824,
+      "learning_rate": 6.8665512498423266e-09,
+      "loss": 0.008,
+      "step": 20962
+    },
+    {
+      "epoch": 4.764318181818182,
+      "grad_norm": 0.598807854281596,
+      "learning_rate": 6.853364192664982e-09,
+      "loss": 0.0024,
+      "step": 20963
+    },
+    {
+      "epoch": 4.764545454545455,
+      "grad_norm": 0.35351485414193773,
+      "learning_rate": 6.840189740575931e-09,
+      "loss": 0.0014,
+      "step": 20964
+    },
+    {
+      "epoch": 4.764772727272727,
+      "grad_norm": 0.8256767850862057,
+      "learning_rate": 6.827027893843846e-09,
+      "loss": 0.0036,
+      "step": 20965
+    },
+    {
+      "epoch": 4.765,
+      "grad_norm": 2.2414037115194994,
+      "learning_rate": 6.813878652737124e-09,
+      "loss": 0.0057,
+      "step": 20966
+    },
+    {
+      "epoch": 4.765227272727273,
+      "grad_norm": 0.8906728768634651,
+      "learning_rate": 6.800742017523884e-09,
+      "loss": 0.0052,
+      "step": 20967
+    },
+    {
+      "epoch": 4.765454545454546,
+      "grad_norm": 2.6550344757468447,
+      "learning_rate": 6.787617988472037e-09,
+      "loss": 0.0043,
+      "step": 20968
+    },
+    {
+      "epoch": 4.765681818181818,
+      "grad_norm": 1.1289232664380626,
+      "learning_rate": 6.7745065658491455e-09,
+      "loss": 0.0048,
+      "step": 20969
+    },
+    {
+      "epoch": 4.765909090909091,
+      "grad_norm": 1.0422343093913289,
+      "learning_rate": 6.761407749922636e-09,
+      "loss": 0.0063,
+      "step": 20970
+    },
+    {
+      "epoch": 4.766136363636363,
+      "grad_norm": 0.26651914195793475,
+      "learning_rate": 6.748321540959654e-09,
+      "loss": 0.0009,
+      "step": 20971
+    },
+    {
+      "epoch": 4.766363636363637,
+      "grad_norm": 0.6005293720070114,
+      "learning_rate": 6.735247939227002e-09,
+      "loss": 0.0024,
+      "step": 20972
+    },
+    {
+      "epoch": 4.766590909090909,
+      "grad_norm": 1.2310774631136965,
+      "learning_rate": 6.722186944991202e-09,
+      "loss": 0.0037,
+      "step": 20973
+    },
+    {
+      "epoch": 4.766818181818182,
+      "grad_norm": 0.3368455095180143,
+      "learning_rate": 6.709138558518638e-09,
+      "loss": 0.0009,
+      "step": 20974
+    },
+    {
+      "epoch": 4.767045454545455,
+      "grad_norm": 1.7139242467175542,
+      "learning_rate": 6.696102780075486e-09,
+      "loss": 0.0097,
+      "step": 20975
+    },
+    {
+      "epoch": 4.7672727272727276,
+      "grad_norm": 1.0650748420205427,
+      "learning_rate": 6.683079609927437e-09,
+      "loss": 0.0063,
+      "step": 20976
+    },
+    {
+      "epoch": 4.7675,
+      "grad_norm": 1.5645510510974303,
+      "learning_rate": 6.670069048340111e-09,
+      "loss": 0.0065,
+      "step": 20977
+    },
+    {
+      "epoch": 4.767727272727273,
+      "grad_norm": 0.7001363504410303,
+      "learning_rate": 6.657071095578782e-09,
+      "loss": 0.0027,
+      "step": 20978
+    },
+    {
+      "epoch": 4.767954545454545,
+      "grad_norm": 2.089259433990285,
+      "learning_rate": 6.644085751908585e-09,
+      "loss": 0.0061,
+      "step": 20979
+    },
+    {
+      "epoch": 4.7681818181818185,
+      "grad_norm": 0.4817060933926435,
+      "learning_rate": 6.63111301759424e-09,
+      "loss": 0.0013,
+      "step": 20980
+    },
+    {
+      "epoch": 4.768409090909091,
+      "grad_norm": 0.5228086662922866,
+      "learning_rate": 6.618152892900326e-09,
+      "loss": 0.0027,
+      "step": 20981
+    },
+    {
+      "epoch": 4.768636363636364,
+      "grad_norm": 0.7428327727673828,
+      "learning_rate": 6.6052053780910755e-09,
+      "loss": 0.0049,
+      "step": 20982
+    },
+    {
+      "epoch": 4.768863636363636,
+      "grad_norm": 0.6178072756034176,
+      "learning_rate": 6.592270473430515e-09,
+      "loss": 0.003,
+      "step": 20983
+    },
+    {
+      "epoch": 4.7690909090909095,
+      "grad_norm": 1.2745821128026897,
+      "learning_rate": 6.5793481791824595e-09,
+      "loss": 0.0064,
+      "step": 20984
+    },
+    {
+      "epoch": 4.769318181818182,
+      "grad_norm": 0.8216173252096284,
+      "learning_rate": 6.56643849561045e-09,
+      "loss": 0.0051,
+      "step": 20985
+    },
+    {
+      "epoch": 4.7695454545454545,
+      "grad_norm": 0.6651867689341914,
+      "learning_rate": 6.553541422977608e-09,
+      "loss": 0.0022,
+      "step": 20986
+    },
+    {
+      "epoch": 4.769772727272727,
+      "grad_norm": 0.4308396843247123,
+      "learning_rate": 6.5406569615469875e-09,
+      "loss": 0.0018,
+      "step": 20987
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 1.4430109684499262,
+      "learning_rate": 6.527785111581364e-09,
+      "loss": 0.0093,
+      "step": 20988
+    },
+    {
+      "epoch": 4.770227272727273,
+      "grad_norm": 1.0097292786211014,
+      "learning_rate": 6.514925873343167e-09,
+      "loss": 0.0055,
+      "step": 20989
+    },
+    {
+      "epoch": 4.7704545454545455,
+      "grad_norm": 0.678241645384635,
+      "learning_rate": 6.502079247094617e-09,
+      "loss": 0.0012,
+      "step": 20990
+    },
+    {
+      "epoch": 4.770681818181818,
+      "grad_norm": 1.3522971432894815,
+      "learning_rate": 6.489245233097727e-09,
+      "loss": 0.004,
+      "step": 20991
+    },
+    {
+      "epoch": 4.7709090909090905,
+      "grad_norm": 0.9095294518649839,
+      "learning_rate": 6.476423831614162e-09,
+      "loss": 0.0034,
+      "step": 20992
+    },
+    {
+      "epoch": 4.771136363636364,
+      "grad_norm": 0.6159616229308169,
+      "learning_rate": 6.4636150429054495e-09,
+      "loss": 0.0046,
+      "step": 20993
+    },
+    {
+      "epoch": 4.7713636363636365,
+      "grad_norm": 1.5172353028117123,
+      "learning_rate": 6.450818867232631e-09,
+      "loss": 0.0062,
+      "step": 20994
+    },
+    {
+      "epoch": 4.771590909090909,
+      "grad_norm": 1.5437877956897326,
+      "learning_rate": 6.438035304856816e-09,
+      "loss": 0.0102,
+      "step": 20995
+    },
+    {
+      "epoch": 4.7718181818181815,
+      "grad_norm": 2.6008658108111304,
+      "learning_rate": 6.42526435603856e-09,
+      "loss": 0.0349,
+      "step": 20996
+    },
+    {
+      "epoch": 4.772045454545455,
+      "grad_norm": 0.2631552160591311,
+      "learning_rate": 6.4125060210382815e-09,
+      "loss": 0.0009,
+      "step": 20997
+    },
+    {
+      "epoch": 4.772272727272727,
+      "grad_norm": 0.39074944988044963,
+      "learning_rate": 6.399760300116256e-09,
+      "loss": 0.0031,
+      "step": 20998
+    },
+    {
+      "epoch": 4.7725,
+      "grad_norm": 0.2935849341744365,
+      "learning_rate": 6.387027193532347e-09,
+      "loss": 0.001,
+      "step": 20999
+    },
+    {
+      "epoch": 4.7727272727272725,
+      "grad_norm": 1.1750948956623246,
+      "learning_rate": 6.374306701546138e-09,
+      "loss": 0.0048,
+      "step": 21000
+    },
+    {
+      "epoch": 4.772954545454546,
+      "grad_norm": 1.4517910301862003,
+      "learning_rate": 6.361598824417073e-09,
+      "loss": 0.0032,
+      "step": 21001
+    },
+    {
+      "epoch": 4.773181818181818,
+      "grad_norm": 1.943599609696048,
+      "learning_rate": 6.348903562404321e-09,
+      "loss": 0.0177,
+      "step": 21002
+    },
+    {
+      "epoch": 4.773409090909091,
+      "grad_norm": 0.4911757826035164,
+      "learning_rate": 6.336220915766633e-09,
+      "loss": 0.0023,
+      "step": 21003
+    },
+    {
+      "epoch": 4.7736363636363635,
+      "grad_norm": 0.8267782090129974,
+      "learning_rate": 6.323550884762761e-09,
+      "loss": 0.005,
+      "step": 21004
+    },
+    {
+      "epoch": 4.773863636363636,
+      "grad_norm": 0.7584619801957369,
+      "learning_rate": 6.310893469651039e-09,
+      "loss": 0.0058,
+      "step": 21005
+    },
+    {
+      "epoch": 4.774090909090909,
+      "grad_norm": 0.8939405653545773,
+      "learning_rate": 6.2982486706895244e-09,
+      "loss": 0.0054,
+      "step": 21006
+    },
+    {
+      "epoch": 4.774318181818182,
+      "grad_norm": 0.6337019781322554,
+      "learning_rate": 6.285616488136137e-09,
+      "loss": 0.0018,
+      "step": 21007
+    },
+    {
+      "epoch": 4.774545454545454,
+      "grad_norm": 0.43267807649799406,
+      "learning_rate": 6.272996922248448e-09,
+      "loss": 0.002,
+      "step": 21008
+    },
+    {
+      "epoch": 4.774772727272727,
+      "grad_norm": 0.3853382630077475,
+      "learning_rate": 6.260389973283751e-09,
+      "loss": 0.0011,
+      "step": 21009
+    },
+    {
+      "epoch": 4.775,
+      "grad_norm": 1.1928201628319726,
+      "learning_rate": 6.247795641499202e-09,
+      "loss": 0.0127,
+      "step": 21010
+    },
+    {
+      "epoch": 4.775227272727273,
+      "grad_norm": 0.837975213190348,
+      "learning_rate": 6.23521392715154e-09,
+      "loss": 0.003,
+      "step": 21011
+    },
+    {
+      "epoch": 4.775454545454545,
+      "grad_norm": 1.3549702124879257,
+      "learning_rate": 6.222644830497365e-09,
+      "loss": 0.0073,
+      "step": 21012
+    },
+    {
+      "epoch": 4.775681818181818,
+      "grad_norm": 0.26002257442154275,
+      "learning_rate": 6.210088351792931e-09,
+      "loss": 0.0008,
+      "step": 21013
+    },
+    {
+      "epoch": 4.77590909090909,
+      "grad_norm": 1.189609481944409,
+      "learning_rate": 6.197544491294422e-09,
+      "loss": 0.0074,
+      "step": 21014
+    },
+    {
+      "epoch": 4.776136363636364,
+      "grad_norm": 0.6423829398718289,
+      "learning_rate": 6.185013249257465e-09,
+      "loss": 0.0047,
+      "step": 21015
+    },
+    {
+      "epoch": 4.776363636363636,
+      "grad_norm": 0.9030441001202814,
+      "learning_rate": 6.172494625937692e-09,
+      "loss": 0.0118,
+      "step": 21016
+    },
+    {
+      "epoch": 4.776590909090909,
+      "grad_norm": 0.6200597119879763,
+      "learning_rate": 6.159988621590382e-09,
+      "loss": 0.0013,
+      "step": 21017
+    },
+    {
+      "epoch": 4.776818181818182,
+      "grad_norm": 1.5085560236320223,
+      "learning_rate": 6.147495236470541e-09,
+      "loss": 0.0054,
+      "step": 21018
+    },
+    {
+      "epoch": 4.777045454545455,
+      "grad_norm": 0.3396731932317007,
+      "learning_rate": 6.1350144708328254e-09,
+      "loss": 0.0012,
+      "step": 21019
+    },
+    {
+      "epoch": 4.777272727272727,
+      "grad_norm": 0.9318705966969756,
+      "learning_rate": 6.122546324931963e-09,
+      "loss": 0.0033,
+      "step": 21020
+    },
+    {
+      "epoch": 4.7775,
+      "grad_norm": 1.414894600523297,
+      "learning_rate": 6.110090799021987e-09,
+      "loss": 0.0046,
+      "step": 21021
+    },
+    {
+      "epoch": 4.777727272727272,
+      "grad_norm": 0.5063727377373554,
+      "learning_rate": 6.097647893356997e-09,
+      "loss": 0.0019,
+      "step": 21022
+    },
+    {
+      "epoch": 4.777954545454546,
+      "grad_norm": 0.8638530758803918,
+      "learning_rate": 6.085217608190683e-09,
+      "loss": 0.0028,
+      "step": 21023
+    },
+    {
+      "epoch": 4.778181818181818,
+      "grad_norm": 2.591951432419879,
+      "learning_rate": 6.072799943776589e-09,
+      "loss": 0.0041,
+      "step": 21024
+    },
+    {
+      "epoch": 4.778409090909091,
+      "grad_norm": 0.5784882025910163,
+      "learning_rate": 6.060394900367777e-09,
+      "loss": 0.0034,
+      "step": 21025
+    },
+    {
+      "epoch": 4.778636363636363,
+      "grad_norm": 1.705412877121961,
+      "learning_rate": 6.0480024782173795e-09,
+      "loss": 0.0106,
+      "step": 21026
+    },
+    {
+      "epoch": 4.778863636363637,
+      "grad_norm": 4.670492549493878,
+      "learning_rate": 6.03562267757804e-09,
+      "loss": 0.0092,
+      "step": 21027
+    },
+    {
+      "epoch": 4.779090909090909,
+      "grad_norm": 0.9477512175691069,
+      "learning_rate": 6.023255498702266e-09,
+      "loss": 0.0024,
+      "step": 21028
+    },
+    {
+      "epoch": 4.779318181818182,
+      "grad_norm": 0.49334610792960915,
+      "learning_rate": 6.010900941842077e-09,
+      "loss": 0.0011,
+      "step": 21029
+    },
+    {
+      "epoch": 4.779545454545454,
+      "grad_norm": 1.8794887837466772,
+      "learning_rate": 5.998559007249493e-09,
+      "loss": 0.0058,
+      "step": 21030
+    },
+    {
+      "epoch": 4.779772727272727,
+      "grad_norm": 0.76577165298433,
+      "learning_rate": 5.98622969517626e-09,
+      "loss": 0.0025,
+      "step": 21031
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 0.6038806969964796,
+      "learning_rate": 5.973913005873702e-09,
+      "loss": 0.0019,
+      "step": 21032
+    },
+    {
+      "epoch": 4.780227272727273,
+      "grad_norm": 1.1864012184358765,
+      "learning_rate": 5.961608939593008e-09,
+      "loss": 0.0068,
+      "step": 21033
+    },
+    {
+      "epoch": 4.780454545454545,
+      "grad_norm": 1.150594748047977,
+      "learning_rate": 5.9493174965850886e-09,
+      "loss": 0.0045,
+      "step": 21034
+    },
+    {
+      "epoch": 4.780681818181818,
+      "grad_norm": 1.8632630548354545,
+      "learning_rate": 5.937038677100576e-09,
+      "loss": 0.0083,
+      "step": 21035
+    },
+    {
+      "epoch": 4.780909090909091,
+      "grad_norm": 0.7507851049935841,
+      "learning_rate": 5.924772481389826e-09,
+      "loss": 0.0027,
+      "step": 21036
+    },
+    {
+      "epoch": 4.781136363636364,
+      "grad_norm": 1.379805034412775,
+      "learning_rate": 5.912518909703055e-09,
+      "loss": 0.0097,
+      "step": 21037
+    },
+    {
+      "epoch": 4.781363636363636,
+      "grad_norm": 0.5579551207433088,
+      "learning_rate": 5.900277962290063e-09,
+      "loss": 0.0025,
+      "step": 21038
+    },
+    {
+      "epoch": 4.781590909090909,
+      "grad_norm": 0.5287515281987316,
+      "learning_rate": 5.888049639400442e-09,
+      "loss": 0.0056,
+      "step": 21039
+    },
+    {
+      "epoch": 4.781818181818182,
+      "grad_norm": 0.6110410564034986,
+      "learning_rate": 5.8758339412836455e-09,
+      "loss": 0.0031,
+      "step": 21040
+    },
+    {
+      "epoch": 4.782045454545455,
+      "grad_norm": 0.7287161412747513,
+      "learning_rate": 5.8636308681886414e-09,
+      "loss": 0.0028,
+      "step": 21041
+    },
+    {
+      "epoch": 4.782272727272727,
+      "grad_norm": 1.2407131797728241,
+      "learning_rate": 5.851440420364466e-09,
+      "loss": 0.0042,
+      "step": 21042
+    },
+    {
+      "epoch": 4.7825,
+      "grad_norm": 0.15969863966497772,
+      "learning_rate": 5.839262598059461e-09,
+      "loss": 0.0005,
+      "step": 21043
+    },
+    {
+      "epoch": 4.782727272727273,
+      "grad_norm": 0.7343238737682197,
+      "learning_rate": 5.8270974015221805e-09,
+      "loss": 0.006,
+      "step": 21044
+    },
+    {
+      "epoch": 4.782954545454546,
+      "grad_norm": 2.7553764302976145,
+      "learning_rate": 5.8149448310004795e-09,
+      "loss": 0.0084,
+      "step": 21045
+    },
+    {
+      "epoch": 4.783181818181818,
+      "grad_norm": 0.4520386920422124,
+      "learning_rate": 5.8028048867424245e-09,
+      "loss": 0.0011,
+      "step": 21046
+    },
+    {
+      "epoch": 4.783409090909091,
+      "grad_norm": 2.6007400853871574,
+      "learning_rate": 5.790677568995318e-09,
+      "loss": 0.0156,
+      "step": 21047
+    },
+    {
+      "epoch": 4.783636363636363,
+      "grad_norm": 0.6231988434701182,
+      "learning_rate": 5.7785628780066e-09,
+      "loss": 0.0028,
+      "step": 21048
+    },
+    {
+      "epoch": 4.783863636363637,
+      "grad_norm": 1.628583159757475,
+      "learning_rate": 5.766460814023225e-09,
+      "loss": 0.0137,
+      "step": 21049
+    },
+    {
+      "epoch": 4.784090909090909,
+      "grad_norm": 0.3725159282188987,
+      "learning_rate": 5.754371377292081e-09,
+      "loss": 0.0011,
+      "step": 21050
+    },
+    {
+      "epoch": 4.784318181818182,
+      "grad_norm": 0.3174584223559383,
+      "learning_rate": 5.7422945680596364e-09,
+      "loss": 0.0012,
+      "step": 21051
+    },
+    {
+      "epoch": 4.784545454545454,
+      "grad_norm": 1.1259590489478768,
+      "learning_rate": 5.730230386572153e-09,
+      "loss": 0.0062,
+      "step": 21052
+    },
+    {
+      "epoch": 4.784772727272728,
+      "grad_norm": 2.3273543021323952,
+      "learning_rate": 5.718178833075614e-09,
+      "loss": 0.0152,
+      "step": 21053
+    },
+    {
+      "epoch": 4.785,
+      "grad_norm": 0.6045882331131912,
+      "learning_rate": 5.706139907815863e-09,
+      "loss": 0.0035,
+      "step": 21054
+    },
+    {
+      "epoch": 4.785227272727273,
+      "grad_norm": 0.447096923012277,
+      "learning_rate": 5.694113611038332e-09,
+      "loss": 0.0014,
+      "step": 21055
+    },
+    {
+      "epoch": 4.785454545454545,
+      "grad_norm": 1.8932717502571335,
+      "learning_rate": 5.682099942988239e-09,
+      "loss": 0.0061,
+      "step": 21056
+    },
+    {
+      "epoch": 4.785681818181818,
+      "grad_norm": 0.5923928220832203,
+      "learning_rate": 5.670098903910598e-09,
+      "loss": 0.0034,
+      "step": 21057
+    },
+    {
+      "epoch": 4.785909090909091,
+      "grad_norm": 0.7941427440409715,
+      "learning_rate": 5.658110494050073e-09,
+      "loss": 0.0032,
+      "step": 21058
+    },
+    {
+      "epoch": 4.786136363636364,
+      "grad_norm": 3.1128733159588555,
+      "learning_rate": 5.646134713651261e-09,
+      "loss": 0.0066,
+      "step": 21059
+    },
+    {
+      "epoch": 4.786363636363636,
+      "grad_norm": 0.5754539283435537,
+      "learning_rate": 5.6341715629582015e-09,
+      "loss": 0.0017,
+      "step": 21060
+    },
+    {
+      "epoch": 4.78659090909091,
+      "grad_norm": 1.0217537375290666,
+      "learning_rate": 5.622221042215006e-09,
+      "loss": 0.0072,
+      "step": 21061
+    },
+    {
+      "epoch": 4.786818181818182,
+      "grad_norm": 0.42174363427910533,
+      "learning_rate": 5.61028315166523e-09,
+      "loss": 0.0009,
+      "step": 21062
+    },
+    {
+      "epoch": 4.787045454545455,
+      "grad_norm": 1.1690970715649933,
+      "learning_rate": 5.598357891552358e-09,
+      "loss": 0.0093,
+      "step": 21063
+    },
+    {
+      "epoch": 4.787272727272727,
+      "grad_norm": 0.5590111100539003,
+      "learning_rate": 5.5864452621196e-09,
+      "loss": 0.0016,
+      "step": 21064
+    },
+    {
+      "epoch": 4.7875,
+      "grad_norm": 0.6955994367223968,
+      "learning_rate": 5.574545263609885e-09,
+      "loss": 0.0024,
+      "step": 21065
+    },
+    {
+      "epoch": 4.787727272727273,
+      "grad_norm": 0.46243284575814264,
+      "learning_rate": 5.5626578962657966e-09,
+      "loss": 0.0014,
+      "step": 21066
+    },
+    {
+      "epoch": 4.787954545454546,
+      "grad_norm": 0.9639820652387412,
+      "learning_rate": 5.550783160329781e-09,
+      "loss": 0.0079,
+      "step": 21067
+    },
+    {
+      "epoch": 4.788181818181818,
+      "grad_norm": 1.149133018206508,
+      "learning_rate": 5.538921056043936e-09,
+      "loss": 0.0078,
+      "step": 21068
+    },
+    {
+      "epoch": 4.788409090909091,
+      "grad_norm": 1.2166893506883196,
+      "learning_rate": 5.52707158365029e-09,
+      "loss": 0.0043,
+      "step": 21069
+    },
+    {
+      "epoch": 4.788636363636364,
+      "grad_norm": 1.2610691590268732,
+      "learning_rate": 5.515234743390316e-09,
+      "loss": 0.002,
+      "step": 21070
+    },
+    {
+      "epoch": 4.7888636363636365,
+      "grad_norm": 0.8897448679825083,
+      "learning_rate": 5.503410535505488e-09,
+      "loss": 0.0051,
+      "step": 21071
+    },
+    {
+      "epoch": 4.789090909090909,
+      "grad_norm": 0.8062162187173973,
+      "learning_rate": 5.4915989602369325e-09,
+      "loss": 0.0034,
+      "step": 21072
+    },
+    {
+      "epoch": 4.789318181818182,
+      "grad_norm": 2.7045846505703564,
+      "learning_rate": 5.479800017825429e-09,
+      "loss": 0.0062,
+      "step": 21073
+    },
+    {
+      "epoch": 4.789545454545454,
+      "grad_norm": 1.7707143701449233,
+      "learning_rate": 5.4680137085115495e-09,
+      "loss": 0.0115,
+      "step": 21074
+    },
+    {
+      "epoch": 4.7897727272727275,
+      "grad_norm": 1.0052578754622405,
+      "learning_rate": 5.456240032535795e-09,
+      "loss": 0.0056,
+      "step": 21075
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 0.7677813547320759,
+      "learning_rate": 5.444478990138113e-09,
+      "loss": 0.0045,
+      "step": 21076
+    },
+    {
+      "epoch": 4.790227272727273,
+      "grad_norm": 1.5765832374090811,
+      "learning_rate": 5.4327305815584515e-09,
+      "loss": 0.0112,
+      "step": 21077
+    },
+    {
+      "epoch": 4.790454545454545,
+      "grad_norm": 0.2692005986308452,
+      "learning_rate": 5.4209948070362e-09,
+      "loss": 0.0005,
+      "step": 21078
+    },
+    {
+      "epoch": 4.7906818181818185,
+      "grad_norm": 3.2190242812197374,
+      "learning_rate": 5.409271666810823e-09,
+      "loss": 0.0086,
+      "step": 21079
+    },
+    {
+      "epoch": 4.790909090909091,
+      "grad_norm": 0.5062888695863084,
+      "learning_rate": 5.397561161121362e-09,
+      "loss": 0.0037,
+      "step": 21080
+    },
+    {
+      "epoch": 4.7911363636363635,
+      "grad_norm": 0.877205970419486,
+      "learning_rate": 5.385863290206517e-09,
+      "loss": 0.0038,
+      "step": 21081
+    },
+    {
+      "epoch": 4.791363636363636,
+      "grad_norm": 0.9445554503766618,
+      "learning_rate": 5.374178054304985e-09,
+      "loss": 0.0044,
+      "step": 21082
+    },
+    {
+      "epoch": 4.7915909090909095,
+      "grad_norm": 2.1645872057020368,
+      "learning_rate": 5.362505453654909e-09,
+      "loss": 0.024,
+      "step": 21083
+    },
+    {
+      "epoch": 4.791818181818182,
+      "grad_norm": 1.153151623919523,
+      "learning_rate": 5.350845488494363e-09,
+      "loss": 0.0077,
+      "step": 21084
+    },
+    {
+      "epoch": 4.7920454545454545,
+      "grad_norm": 1.4113913409938412,
+      "learning_rate": 5.339198159061074e-09,
+      "loss": 0.0147,
+      "step": 21085
+    },
+    {
+      "epoch": 4.792272727272727,
+      "grad_norm": 0.9285977309854292,
+      "learning_rate": 5.3275634655926974e-09,
+      "loss": 0.0071,
+      "step": 21086
+    },
+    {
+      "epoch": 4.7925,
+      "grad_norm": 1.0068516085579742,
+      "learning_rate": 5.315941408326267e-09,
+      "loss": 0.0027,
+      "step": 21087
+    },
+    {
+      "epoch": 4.792727272727273,
+      "grad_norm": 2.32296219873131,
+      "learning_rate": 5.304331987498953e-09,
+      "loss": 0.0167,
+      "step": 21088
+    },
+    {
+      "epoch": 4.7929545454545455,
+      "grad_norm": 3.2910456638725405,
+      "learning_rate": 5.292735203347443e-09,
+      "loss": 0.0115,
+      "step": 21089
+    },
+    {
+      "epoch": 4.793181818181818,
+      "grad_norm": 0.8851354887489066,
+      "learning_rate": 5.281151056108214e-09,
+      "loss": 0.0033,
+      "step": 21090
+    },
+    {
+      "epoch": 4.7934090909090905,
+      "grad_norm": 1.739747354610571,
+      "learning_rate": 5.269579546017395e-09,
+      "loss": 0.0052,
+      "step": 21091
+    },
+    {
+      "epoch": 4.793636363636364,
+      "grad_norm": 0.753377392668986,
+      "learning_rate": 5.258020673311049e-09,
+      "loss": 0.0032,
+      "step": 21092
+    },
+    {
+      "epoch": 4.793863636363636,
+      "grad_norm": 2.489745195925557,
+      "learning_rate": 5.246474438224888e-09,
+      "loss": 0.0124,
+      "step": 21093
+    },
+    {
+      "epoch": 4.794090909090909,
+      "grad_norm": 0.25612050235856954,
+      "learning_rate": 5.234940840994351e-09,
+      "loss": 0.0008,
+      "step": 21094
+    },
+    {
+      "epoch": 4.7943181818181815,
+      "grad_norm": 1.7065076693457428,
+      "learning_rate": 5.223419881854597e-09,
+      "loss": 0.0047,
+      "step": 21095
+    },
+    {
+      "epoch": 4.794545454545455,
+      "grad_norm": 0.3352534540233844,
+      "learning_rate": 5.211911561040506e-09,
+      "loss": 0.0014,
+      "step": 21096
+    },
+    {
+      "epoch": 4.794772727272727,
+      "grad_norm": 0.3047781876694,
+      "learning_rate": 5.200415878786891e-09,
+      "loss": 0.0007,
+      "step": 21097
+    },
+    {
+      "epoch": 4.795,
+      "grad_norm": 0.921212642363645,
+      "learning_rate": 5.1889328353280785e-09,
+      "loss": 0.0035,
+      "step": 21098
+    },
+    {
+      "epoch": 4.7952272727272724,
+      "grad_norm": 3.073229222181897,
+      "learning_rate": 5.1774624308982556e-09,
+      "loss": 0.0328,
+      "step": 21099
+    },
+    {
+      "epoch": 4.795454545454545,
+      "grad_norm": 2.5925816064689373,
+      "learning_rate": 5.166004665731264e-09,
+      "loss": 0.0106,
+      "step": 21100
+    },
+    {
+      "epoch": 4.795681818181818,
+      "grad_norm": 1.6213968869857691,
+      "learning_rate": 5.154559540060805e-09,
+      "loss": 0.0073,
+      "step": 21101
+    },
+    {
+      "epoch": 4.795909090909091,
+      "grad_norm": 2.076031422717543,
+      "learning_rate": 5.143127054120303e-09,
+      "loss": 0.0098,
+      "step": 21102
+    },
+    {
+      "epoch": 4.796136363636363,
+      "grad_norm": 1.094108793841688,
+      "learning_rate": 5.131707208142836e-09,
+      "loss": 0.0051,
+      "step": 21103
+    },
+    {
+      "epoch": 4.796363636363637,
+      "grad_norm": 0.37803872026307633,
+      "learning_rate": 5.1203000023612725e-09,
+      "loss": 0.0012,
+      "step": 21104
+    },
+    {
+      "epoch": 4.796590909090909,
+      "grad_norm": 0.40502669431443783,
+      "learning_rate": 5.108905437008205e-09,
+      "loss": 0.0013,
+      "step": 21105
+    },
+    {
+      "epoch": 4.796818181818182,
+      "grad_norm": 0.9522931430755236,
+      "learning_rate": 5.097523512316016e-09,
+      "loss": 0.0056,
+      "step": 21106
+    },
+    {
+      "epoch": 4.797045454545454,
+      "grad_norm": 0.6737943558762879,
+      "learning_rate": 5.086154228516812e-09,
+      "loss": 0.0012,
+      "step": 21107
+    },
+    {
+      "epoch": 4.797272727272727,
+      "grad_norm": 0.13835201779726738,
+      "learning_rate": 5.074797585842492e-09,
+      "loss": 0.0003,
+      "step": 21108
+    },
+    {
+      "epoch": 4.7975,
+      "grad_norm": 1.062289395232293,
+      "learning_rate": 5.063453584524467e-09,
+      "loss": 0.0028,
+      "step": 21109
+    },
+    {
+      "epoch": 4.797727272727273,
+      "grad_norm": 0.3360146765915773,
+      "learning_rate": 5.052122224794217e-09,
+      "loss": 0.0008,
+      "step": 21110
+    },
+    {
+      "epoch": 4.797954545454545,
+      "grad_norm": 1.4486909455499428,
+      "learning_rate": 5.040803506882741e-09,
+      "loss": 0.0041,
+      "step": 21111
+    },
+    {
+      "epoch": 4.798181818181818,
+      "grad_norm": 1.518895992285405,
+      "learning_rate": 5.029497431020824e-09,
+      "loss": 0.0153,
+      "step": 21112
+    },
+    {
+      "epoch": 4.798409090909091,
+      "grad_norm": 0.5179635079792152,
+      "learning_rate": 5.0182039974390475e-09,
+      "loss": 0.0028,
+      "step": 21113
+    },
+    {
+      "epoch": 4.798636363636364,
+      "grad_norm": 6.955604274130558,
+      "learning_rate": 5.006923206367781e-09,
+      "loss": 0.0097,
+      "step": 21114
+    },
+    {
+      "epoch": 4.798863636363636,
+      "grad_norm": 2.6510484650654336,
+      "learning_rate": 4.9956550580369106e-09,
+      "loss": 0.0236,
+      "step": 21115
+    },
+    {
+      "epoch": 4.799090909090909,
+      "grad_norm": 0.7012130051268548,
+      "learning_rate": 4.984399552676392e-09,
+      "loss": 0.0017,
+      "step": 21116
+    },
+    {
+      "epoch": 4.799318181818181,
+      "grad_norm": 4.411580559243821,
+      "learning_rate": 4.973156690515554e-09,
+      "loss": 0.0189,
+      "step": 21117
+    },
+    {
+      "epoch": 4.799545454545455,
+      "grad_norm": 0.7750076571187259,
+      "learning_rate": 4.961926471783799e-09,
+      "loss": 0.0022,
+      "step": 21118
+    },
+    {
+      "epoch": 4.799772727272727,
+      "grad_norm": 0.8966478715337294,
+      "learning_rate": 4.950708896710038e-09,
+      "loss": 0.008,
+      "step": 21119
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.7434923713592979,
+      "learning_rate": 4.939503965523049e-09,
+      "loss": 0.0018,
+      "step": 21120
+    },
+    {
+      "epoch": 4.800227272727272,
+      "grad_norm": 1.778979643930183,
+      "learning_rate": 4.928311678451397e-09,
+      "loss": 0.0044,
+      "step": 21121
+    },
+    {
+      "epoch": 4.800454545454546,
+      "grad_norm": 0.3290844548114461,
+      "learning_rate": 4.9171320357231656e-09,
+      "loss": 0.0011,
+      "step": 21122
+    },
+    {
+      "epoch": 4.800681818181818,
+      "grad_norm": 0.9426075668522762,
+      "learning_rate": 4.905965037566435e-09,
+      "loss": 0.0055,
+      "step": 21123
+    },
+    {
+      "epoch": 4.800909090909091,
+      "grad_norm": 3.5085255264681035,
+      "learning_rate": 4.89481068420887e-09,
+      "loss": 0.0067,
+      "step": 21124
+    },
+    {
+      "epoch": 4.801136363636363,
+      "grad_norm": 0.5527293499556173,
+      "learning_rate": 4.883668975878067e-09,
+      "loss": 0.0021,
+      "step": 21125
+    },
+    {
+      "epoch": 4.801363636363637,
+      "grad_norm": 0.5500715455327922,
+      "learning_rate": 4.872539912800997e-09,
+      "loss": 0.0017,
+      "step": 21126
+    },
+    {
+      "epoch": 4.801590909090909,
+      "grad_norm": 0.6443822347189283,
+      "learning_rate": 4.861423495204701e-09,
+      "loss": 0.0028,
+      "step": 21127
+    },
+    {
+      "epoch": 4.801818181818182,
+      "grad_norm": 1.3200362262356258,
+      "learning_rate": 4.850319723315872e-09,
+      "loss": 0.0055,
+      "step": 21128
+    },
+    {
+      "epoch": 4.802045454545454,
+      "grad_norm": 0.6480272469833869,
+      "learning_rate": 4.839228597360996e-09,
+      "loss": 0.0023,
+      "step": 21129
+    },
+    {
+      "epoch": 4.802272727272728,
+      "grad_norm": 0.8765599323860201,
+      "learning_rate": 4.828150117566141e-09,
+      "loss": 0.0022,
+      "step": 21130
+    },
+    {
+      "epoch": 4.8025,
+      "grad_norm": 0.17723580605322647,
+      "learning_rate": 4.817084284157239e-09,
+      "loss": 0.0006,
+      "step": 21131
+    },
+    {
+      "epoch": 4.802727272727273,
+      "grad_norm": 1.4698615229345569,
+      "learning_rate": 4.806031097359942e-09,
+      "loss": 0.0089,
+      "step": 21132
+    },
+    {
+      "epoch": 4.802954545454545,
+      "grad_norm": 0.741738640418225,
+      "learning_rate": 4.794990557399695e-09,
+      "loss": 0.0035,
+      "step": 21133
+    },
+    {
+      "epoch": 4.803181818181818,
+      "grad_norm": 0.8817503500911517,
+      "learning_rate": 4.783962664501596e-09,
+      "loss": 0.008,
+      "step": 21134
+    },
+    {
+      "epoch": 4.803409090909091,
+      "grad_norm": 0.7640450556319167,
+      "learning_rate": 4.7729474188905335e-09,
+      "loss": 0.0028,
+      "step": 21135
+    },
+    {
+      "epoch": 4.803636363636364,
+      "grad_norm": 1.0886787011581638,
+      "learning_rate": 4.76194482079112e-09,
+      "loss": 0.0094,
+      "step": 21136
+    },
+    {
+      "epoch": 4.803863636363636,
+      "grad_norm": 2.0129066976954126,
+      "learning_rate": 4.7509548704276904e-09,
+      "loss": 0.0131,
+      "step": 21137
+    },
+    {
+      "epoch": 4.804090909090909,
+      "grad_norm": 0.9697906442271989,
+      "learning_rate": 4.73997756802444e-09,
+      "loss": 0.0017,
+      "step": 21138
+    },
+    {
+      "epoch": 4.804318181818182,
+      "grad_norm": 0.9395283037519686,
+      "learning_rate": 4.729012913805078e-09,
+      "loss": 0.0055,
+      "step": 21139
+    },
+    {
+      "epoch": 4.804545454545455,
+      "grad_norm": 0.5675955104929612,
+      "learning_rate": 4.7180609079932465e-09,
+      "loss": 0.0024,
+      "step": 21140
+    },
+    {
+      "epoch": 4.804772727272727,
+      "grad_norm": 0.5414108552306298,
+      "learning_rate": 4.707121550812307e-09,
+      "loss": 0.0017,
+      "step": 21141
+    },
+    {
+      "epoch": 4.805,
+      "grad_norm": 2.243804714488599,
+      "learning_rate": 4.696194842485347e-09,
+      "loss": 0.0072,
+      "step": 21142
+    },
+    {
+      "epoch": 4.805227272727272,
+      "grad_norm": 0.46679140402580954,
+      "learning_rate": 4.685280783235171e-09,
+      "loss": 0.0011,
+      "step": 21143
+    },
+    {
+      "epoch": 4.805454545454546,
+      "grad_norm": 0.40177442573798755,
+      "learning_rate": 4.674379373284313e-09,
+      "loss": 0.0008,
+      "step": 21144
+    },
+    {
+      "epoch": 4.805681818181818,
+      "grad_norm": 1.168813021109882,
+      "learning_rate": 4.663490612855023e-09,
+      "loss": 0.004,
+      "step": 21145
+    },
+    {
+      "epoch": 4.805909090909091,
+      "grad_norm": 0.36483916028483543,
+      "learning_rate": 4.652614502169417e-09,
+      "loss": 0.0018,
+      "step": 21146
+    },
+    {
+      "epoch": 4.806136363636364,
+      "grad_norm": 1.0369512288361622,
+      "learning_rate": 4.64175104144933e-09,
+      "loss": 0.0021,
+      "step": 21147
+    },
+    {
+      "epoch": 4.806363636363637,
+      "grad_norm": 2.4943215815788107,
+      "learning_rate": 4.6309002309161135e-09,
+      "loss": 0.0113,
+      "step": 21148
+    },
+    {
+      "epoch": 4.806590909090909,
+      "grad_norm": 0.19601565460314996,
+      "learning_rate": 4.620062070791187e-09,
+      "loss": 0.001,
+      "step": 21149
+    },
+    {
+      "epoch": 4.806818181818182,
+      "grad_norm": 0.6858147927005149,
+      "learning_rate": 4.609236561295486e-09,
+      "loss": 0.0021,
+      "step": 21150
+    },
+    {
+      "epoch": 4.807045454545454,
+      "grad_norm": 2.322640928691125,
+      "learning_rate": 4.598423702649873e-09,
+      "loss": 0.008,
+      "step": 21151
+    },
+    {
+      "epoch": 4.807272727272728,
+      "grad_norm": 2.5501459787194656,
+      "learning_rate": 4.587623495074662e-09,
+      "loss": 0.0274,
+      "step": 21152
+    },
+    {
+      "epoch": 4.8075,
+      "grad_norm": 0.42016094590115416,
+      "learning_rate": 4.576835938790158e-09,
+      "loss": 0.0012,
+      "step": 21153
+    },
+    {
+      "epoch": 4.807727272727273,
+      "grad_norm": 2.101426486175178,
+      "learning_rate": 4.566061034016397e-09,
+      "loss": 0.0152,
+      "step": 21154
+    },
+    {
+      "epoch": 4.807954545454545,
+      "grad_norm": 0.6276904398936745,
+      "learning_rate": 4.555298780972994e-09,
+      "loss": 0.0022,
+      "step": 21155
+    },
+    {
+      "epoch": 4.808181818181819,
+      "grad_norm": 0.6593603361959515,
+      "learning_rate": 4.544549179879565e-09,
+      "loss": 0.0023,
+      "step": 21156
+    },
+    {
+      "epoch": 4.808409090909091,
+      "grad_norm": 0.14491135712201833,
+      "learning_rate": 4.5338122309551706e-09,
+      "loss": 0.0004,
+      "step": 21157
+    },
+    {
+      "epoch": 4.808636363636364,
+      "grad_norm": 0.535565684143062,
+      "learning_rate": 4.523087934418871e-09,
+      "loss": 0.0019,
+      "step": 21158
+    },
+    {
+      "epoch": 4.808863636363636,
+      "grad_norm": 0.8504228999729521,
+      "learning_rate": 4.5123762904891734e-09,
+      "loss": 0.003,
+      "step": 21159
+    },
+    {
+      "epoch": 4.809090909090909,
+      "grad_norm": 0.32681864177376546,
+      "learning_rate": 4.501677299384722e-09,
+      "loss": 0.0018,
+      "step": 21160
+    },
+    {
+      "epoch": 4.809318181818182,
+      "grad_norm": 1.7254136733265402,
+      "learning_rate": 4.490990961323535e-09,
+      "loss": 0.0063,
+      "step": 21161
+    },
+    {
+      "epoch": 4.809545454545455,
+      "grad_norm": 0.43567937539124757,
+      "learning_rate": 4.480317276523635e-09,
+      "loss": 0.0016,
+      "step": 21162
+    },
+    {
+      "epoch": 4.809772727272727,
+      "grad_norm": 1.8890434325712238,
+      "learning_rate": 4.469656245202555e-09,
+      "loss": 0.0054,
+      "step": 21163
+    },
+    {
+      "epoch": 4.8100000000000005,
+      "grad_norm": 0.8160275904901516,
+      "learning_rate": 4.459007867577761e-09,
+      "loss": 0.0027,
+      "step": 21164
+    },
+    {
+      "epoch": 4.810227272727273,
+      "grad_norm": 1.1565563621471329,
+      "learning_rate": 4.4483721438665086e-09,
+      "loss": 0.0027,
+      "step": 21165
+    },
+    {
+      "epoch": 4.8104545454545455,
+      "grad_norm": 1.3893166609792913,
+      "learning_rate": 4.43774907428543e-09,
+      "loss": 0.0081,
+      "step": 21166
+    },
+    {
+      "epoch": 4.810681818181818,
+      "grad_norm": 0.6341771998282602,
+      "learning_rate": 4.427138659051297e-09,
+      "loss": 0.0015,
+      "step": 21167
+    },
+    {
+      "epoch": 4.810909090909091,
+      "grad_norm": 0.8968683270001077,
+      "learning_rate": 4.416540898380464e-09,
+      "loss": 0.0051,
+      "step": 21168
+    },
+    {
+      "epoch": 4.811136363636364,
+      "grad_norm": 0.9037054346357981,
+      "learning_rate": 4.4059557924890775e-09,
+      "loss": 0.0033,
+      "step": 21169
+    },
+    {
+      "epoch": 4.8113636363636365,
+      "grad_norm": 1.0959420289858153,
+      "learning_rate": 4.395383341592938e-09,
+      "loss": 0.0054,
+      "step": 21170
+    },
+    {
+      "epoch": 4.811590909090909,
+      "grad_norm": 2.0472143237991136,
+      "learning_rate": 4.384823545907565e-09,
+      "loss": 0.022,
+      "step": 21171
+    },
+    {
+      "epoch": 4.8118181818181816,
+      "grad_norm": 0.295262697487013,
+      "learning_rate": 4.374276405648414e-09,
+      "loss": 0.0008,
+      "step": 21172
+    },
+    {
+      "epoch": 4.812045454545455,
+      "grad_norm": 0.5473811482064034,
+      "learning_rate": 4.36374192103052e-09,
+      "loss": 0.0031,
+      "step": 21173
+    },
+    {
+      "epoch": 4.8122727272727275,
+      "grad_norm": 0.5529496370010679,
+      "learning_rate": 4.353220092268642e-09,
+      "loss": 0.0009,
+      "step": 21174
+    },
+    {
+      "epoch": 4.8125,
+      "grad_norm": 0.9067430156038524,
+      "learning_rate": 4.342710919577401e-09,
+      "loss": 0.0072,
+      "step": 21175
+    },
+    {
+      "epoch": 4.8127272727272725,
+      "grad_norm": 1.0536745885616279,
+      "learning_rate": 4.332214403171067e-09,
+      "loss": 0.0063,
+      "step": 21176
+    },
+    {
+      "epoch": 4.812954545454545,
+      "grad_norm": 0.8149824224391962,
+      "learning_rate": 4.321730543263708e-09,
+      "loss": 0.0034,
+      "step": 21177
+    },
+    {
+      "epoch": 4.8131818181818184,
+      "grad_norm": 3.4315593921104144,
+      "learning_rate": 4.31125934006911e-09,
+      "loss": 0.018,
+      "step": 21178
+    },
+    {
+      "epoch": 4.813409090909091,
+      "grad_norm": 0.38255717405916095,
+      "learning_rate": 4.300800793800783e-09,
+      "loss": 0.001,
+      "step": 21179
+    },
+    {
+      "epoch": 4.8136363636363635,
+      "grad_norm": 0.7839861940106196,
+      "learning_rate": 4.290354904671959e-09,
+      "loss": 0.0067,
+      "step": 21180
+    },
+    {
+      "epoch": 4.813863636363636,
+      "grad_norm": 0.4806380356224576,
+      "learning_rate": 4.2799216728957315e-09,
+      "loss": 0.0016,
+      "step": 21181
+    },
+    {
+      "epoch": 4.814090909090909,
+      "grad_norm": 1.6324853097917957,
+      "learning_rate": 4.2695010986848475e-09,
+      "loss": 0.0141,
+      "step": 21182
+    },
+    {
+      "epoch": 4.814318181818182,
+      "grad_norm": 0.7402480275929266,
+      "learning_rate": 4.259093182251637e-09,
+      "loss": 0.0033,
+      "step": 21183
+    },
+    {
+      "epoch": 4.8145454545454545,
+      "grad_norm": 0.2651280510658424,
+      "learning_rate": 4.248697923808568e-09,
+      "loss": 0.001,
+      "step": 21184
+    },
+    {
+      "epoch": 4.814772727272727,
+      "grad_norm": 0.2763856081016614,
+      "learning_rate": 4.2383153235674155e-09,
+      "loss": 0.0011,
+      "step": 21185
+    },
+    {
+      "epoch": 4.8149999999999995,
+      "grad_norm": 1.7455330481215436,
+      "learning_rate": 4.227945381740026e-09,
+      "loss": 0.0043,
+      "step": 21186
+    },
+    {
+      "epoch": 4.815227272727273,
+      "grad_norm": 0.5323807251650612,
+      "learning_rate": 4.217588098537895e-09,
+      "loss": 0.0022,
+      "step": 21187
+    },
+    {
+      "epoch": 4.815454545454545,
+      "grad_norm": 1.9670131090058365,
+      "learning_rate": 4.207243474172104e-09,
+      "loss": 0.0102,
+      "step": 21188
+    },
+    {
+      "epoch": 4.815681818181818,
+      "grad_norm": 0.9920645666321647,
+      "learning_rate": 4.196911508853596e-09,
+      "loss": 0.0035,
+      "step": 21189
+    },
+    {
+      "epoch": 4.815909090909091,
+      "grad_norm": 0.5277718335184781,
+      "learning_rate": 4.186592202793174e-09,
+      "loss": 0.0018,
+      "step": 21190
+    },
+    {
+      "epoch": 4.816136363636364,
+      "grad_norm": 3.3400478029624554,
+      "learning_rate": 4.176285556201226e-09,
+      "loss": 0.008,
+      "step": 21191
+    },
+    {
+      "epoch": 4.816363636363636,
+      "grad_norm": 0.5370490844095163,
+      "learning_rate": 4.165991569287861e-09,
+      "loss": 0.0049,
+      "step": 21192
+    },
+    {
+      "epoch": 4.816590909090909,
+      "grad_norm": 1.8381348831046591,
+      "learning_rate": 4.15571024226305e-09,
+      "loss": 0.0058,
+      "step": 21193
+    },
+    {
+      "epoch": 4.816818181818181,
+      "grad_norm": 1.211236965572865,
+      "learning_rate": 4.1454415753364174e-09,
+      "loss": 0.0055,
+      "step": 21194
+    },
+    {
+      "epoch": 4.817045454545455,
+      "grad_norm": 0.3597843338433516,
+      "learning_rate": 4.135185568717378e-09,
+      "loss": 0.0027,
+      "step": 21195
+    },
+    {
+      "epoch": 4.817272727272727,
+      "grad_norm": 2.843270045568999,
+      "learning_rate": 4.124942222615072e-09,
+      "loss": 0.0074,
+      "step": 21196
+    },
+    {
+      "epoch": 4.8175,
+      "grad_norm": 0.9042747313845517,
+      "learning_rate": 4.114711537238358e-09,
+      "loss": 0.0069,
+      "step": 21197
+    },
+    {
+      "epoch": 4.817727272727272,
+      "grad_norm": 1.9072949340133756,
+      "learning_rate": 4.10449351279589e-09,
+      "loss": 0.0044,
+      "step": 21198
+    },
+    {
+      "epoch": 4.817954545454546,
+      "grad_norm": 1.3639078904583426,
+      "learning_rate": 4.094288149495973e-09,
+      "loss": 0.0045,
+      "step": 21199
+    },
+    {
+      "epoch": 4.818181818181818,
+      "grad_norm": 0.574413212350717,
+      "learning_rate": 4.084095447546843e-09,
+      "loss": 0.0017,
+      "step": 21200
+    },
+    {
+      "epoch": 4.818409090909091,
+      "grad_norm": 0.587192146328614,
+      "learning_rate": 4.0739154071561125e-09,
+      "loss": 0.0026,
+      "step": 21201
+    },
+    {
+      "epoch": 4.818636363636363,
+      "grad_norm": 0.19371903295198356,
+      "learning_rate": 4.0637480285316e-09,
+      "loss": 0.0003,
+      "step": 21202
+    },
+    {
+      "epoch": 4.818863636363636,
+      "grad_norm": 0.37216263036931935,
+      "learning_rate": 4.053593311880502e-09,
+      "loss": 0.0007,
+      "step": 21203
+    },
+    {
+      "epoch": 4.819090909090909,
+      "grad_norm": 0.40051555385714627,
+      "learning_rate": 4.043451257410014e-09,
+      "loss": 0.0016,
+      "step": 21204
+    },
+    {
+      "epoch": 4.819318181818182,
+      "grad_norm": 0.7881316315932371,
+      "learning_rate": 4.0333218653267754e-09,
+      "loss": 0.0031,
+      "step": 21205
+    },
+    {
+      "epoch": 4.819545454545454,
+      "grad_norm": 0.44519400690685795,
+      "learning_rate": 4.0232051358374965e-09,
+      "loss": 0.002,
+      "step": 21206
+    },
+    {
+      "epoch": 4.819772727272728,
+      "grad_norm": 1.679291366266106,
+      "learning_rate": 4.013101069148401e-09,
+      "loss": 0.0091,
+      "step": 21207
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 0.8020908106883081,
+      "learning_rate": 4.0030096654655735e-09,
+      "loss": 0.0018,
+      "step": 21208
+    },
+    {
+      "epoch": 4.820227272727273,
+      "grad_norm": 1.3786427710159639,
+      "learning_rate": 3.992930924994754e-09,
+      "loss": 0.0066,
+      "step": 21209
+    },
+    {
+      "epoch": 4.820454545454545,
+      "grad_norm": 0.41270115398494195,
+      "learning_rate": 3.98286484794147e-09,
+      "loss": 0.002,
+      "step": 21210
+    },
+    {
+      "epoch": 4.820681818181818,
+      "grad_norm": 1.4856053034292336,
+      "learning_rate": 3.972811434510976e-09,
+      "loss": 0.0069,
+      "step": 21211
+    },
+    {
+      "epoch": 4.820909090909091,
+      "grad_norm": 0.8334379032116424,
+      "learning_rate": 3.962770684908315e-09,
+      "loss": 0.0027,
+      "step": 21212
+    },
+    {
+      "epoch": 4.821136363636364,
+      "grad_norm": 2.084021391214472,
+      "learning_rate": 3.952742599338255e-09,
+      "loss": 0.015,
+      "step": 21213
+    },
+    {
+      "epoch": 4.821363636363636,
+      "grad_norm": 1.2921901179646826,
+      "learning_rate": 3.9427271780052154e-09,
+      "loss": 0.0076,
+      "step": 21214
+    },
+    {
+      "epoch": 4.821590909090909,
+      "grad_norm": 0.7138429805085318,
+      "learning_rate": 3.932724421113476e-09,
+      "loss": 0.0034,
+      "step": 21215
+    },
+    {
+      "epoch": 4.821818181818182,
+      "grad_norm": 2.635978329465957,
+      "learning_rate": 3.9227343288669725e-09,
+      "loss": 0.0107,
+      "step": 21216
+    },
+    {
+      "epoch": 4.822045454545455,
+      "grad_norm": 0.6037725284776897,
+      "learning_rate": 3.9127569014695e-09,
+      "loss": 0.0026,
+      "step": 21217
+    },
+    {
+      "epoch": 4.822272727272727,
+      "grad_norm": 1.0350680403628287,
+      "learning_rate": 3.902792139124368e-09,
+      "loss": 0.0044,
+      "step": 21218
+    },
+    {
+      "epoch": 4.8225,
+      "grad_norm": 1.9779270600512266,
+      "learning_rate": 3.892840042034954e-09,
+      "loss": 0.0113,
+      "step": 21219
+    },
+    {
+      "epoch": 4.822727272727272,
+      "grad_norm": 1.1434646935961974,
+      "learning_rate": 3.882900610404083e-09,
+      "loss": 0.011,
+      "step": 21220
+    },
+    {
+      "epoch": 4.822954545454546,
+      "grad_norm": 1.7104573945283121,
+      "learning_rate": 3.872973844434441e-09,
+      "loss": 0.0091,
+      "step": 21221
+    },
+    {
+      "epoch": 4.823181818181818,
+      "grad_norm": 0.38037245346784004,
+      "learning_rate": 3.8630597443285036e-09,
+      "loss": 0.0015,
+      "step": 21222
+    },
+    {
+      "epoch": 4.823409090909091,
+      "grad_norm": 1.1261671026549758,
+      "learning_rate": 3.853158310288402e-09,
+      "loss": 0.0031,
+      "step": 21223
+    },
+    {
+      "epoch": 4.823636363636363,
+      "grad_norm": 1.5868805117566382,
+      "learning_rate": 3.843269542516057e-09,
+      "loss": 0.0066,
+      "step": 21224
+    },
+    {
+      "epoch": 4.823863636363637,
+      "grad_norm": 0.6475123827018073,
+      "learning_rate": 3.833393441213113e-09,
+      "loss": 0.0039,
+      "step": 21225
+    },
+    {
+      "epoch": 4.824090909090909,
+      "grad_norm": 2.5574949996799616,
+      "learning_rate": 3.823530006581006e-09,
+      "loss": 0.003,
+      "step": 21226
+    },
+    {
+      "epoch": 4.824318181818182,
+      "grad_norm": 0.26522255893946683,
+      "learning_rate": 3.8136792388207555e-09,
+      "loss": 0.0031,
+      "step": 21227
+    },
+    {
+      "epoch": 4.824545454545454,
+      "grad_norm": 0.21287058976847212,
+      "learning_rate": 3.803841138133313e-09,
+      "loss": 0.001,
+      "step": 21228
+    },
+    {
+      "epoch": 4.824772727272728,
+      "grad_norm": 0.8541557444717018,
+      "learning_rate": 3.794015704719281e-09,
+      "loss": 0.003,
+      "step": 21229
+    },
+    {
+      "epoch": 4.825,
+      "grad_norm": 0.9028478562188702,
+      "learning_rate": 3.7842029387790555e-09,
+      "loss": 0.0041,
+      "step": 21230
+    },
+    {
+      "epoch": 4.825227272727273,
+      "grad_norm": 1.8624482705118657,
+      "learning_rate": 3.774402840512753e-09,
+      "loss": 0.0043,
+      "step": 21231
+    },
+    {
+      "epoch": 4.825454545454545,
+      "grad_norm": 0.9466898926829608,
+      "learning_rate": 3.764615410120076e-09,
+      "loss": 0.0029,
+      "step": 21232
+    },
+    {
+      "epoch": 4.825681818181819,
+      "grad_norm": 0.6606220912147475,
+      "learning_rate": 3.754840647800725e-09,
+      "loss": 0.0021,
+      "step": 21233
+    },
+    {
+      "epoch": 4.825909090909091,
+      "grad_norm": 2.378679481292716,
+      "learning_rate": 3.745078553753986e-09,
+      "loss": 0.0247,
+      "step": 21234
+    },
+    {
+      "epoch": 4.826136363636364,
+      "grad_norm": 1.8208081254164104,
+      "learning_rate": 3.735329128178936e-09,
+      "loss": 0.013,
+      "step": 21235
+    },
+    {
+      "epoch": 4.826363636363636,
+      "grad_norm": 1.0445752349278445,
+      "learning_rate": 3.725592371274373e-09,
+      "loss": 0.0039,
+      "step": 21236
+    },
+    {
+      "epoch": 4.826590909090909,
+      "grad_norm": 1.675649821266311,
+      "learning_rate": 3.7158682832388188e-09,
+      "loss": 0.0045,
+      "step": 21237
+    },
+    {
+      "epoch": 4.826818181818182,
+      "grad_norm": 1.2428707944983723,
+      "learning_rate": 3.7061568642706574e-09,
+      "loss": 0.0051,
+      "step": 21238
+    },
+    {
+      "epoch": 4.827045454545455,
+      "grad_norm": 0.6943988326277768,
+      "learning_rate": 3.696458114567855e-09,
+      "loss": 0.0029,
+      "step": 21239
+    },
+    {
+      "epoch": 4.827272727272727,
+      "grad_norm": 2.2838656047579624,
+      "learning_rate": 3.6867720343282397e-09,
+      "loss": 0.0107,
+      "step": 21240
+    },
+    {
+      "epoch": 4.8275,
+      "grad_norm": 3.748664518850677,
+      "learning_rate": 3.6770986237492227e-09,
+      "loss": 0.0176,
+      "step": 21241
+    },
+    {
+      "epoch": 4.827727272727273,
+      "grad_norm": 1.6294308768674712,
+      "learning_rate": 3.667437883028077e-09,
+      "loss": 0.0058,
+      "step": 21242
+    },
+    {
+      "epoch": 4.827954545454546,
+      "grad_norm": 0.5397980931988774,
+      "learning_rate": 3.6577898123619367e-09,
+      "loss": 0.0022,
+      "step": 21243
+    },
+    {
+      "epoch": 4.828181818181818,
+      "grad_norm": 0.810640635946598,
+      "learning_rate": 3.648154411947449e-09,
+      "loss": 0.0051,
+      "step": 21244
+    },
+    {
+      "epoch": 4.828409090909091,
+      "grad_norm": 1.7216531801591708,
+      "learning_rate": 3.6385316819810557e-09,
+      "loss": 0.0116,
+      "step": 21245
+    },
+    {
+      "epoch": 4.828636363636363,
+      "grad_norm": 2.4930423612884383,
+      "learning_rate": 3.628921622659057e-09,
+      "loss": 0.0092,
+      "step": 21246
+    },
+    {
+      "epoch": 4.828863636363637,
+      "grad_norm": 0.7264636290581223,
+      "learning_rate": 3.6193242341773377e-09,
+      "loss": 0.0019,
+      "step": 21247
+    },
+    {
+      "epoch": 4.829090909090909,
+      "grad_norm": 0.24383600298467442,
+      "learning_rate": 3.6097395167317138e-09,
+      "loss": 0.0011,
+      "step": 21248
+    },
+    {
+      "epoch": 4.829318181818182,
+      "grad_norm": 0.4158083994179854,
+      "learning_rate": 3.6001674705175153e-09,
+      "loss": 0.0011,
+      "step": 21249
+    },
+    {
+      "epoch": 4.829545454545455,
+      "grad_norm": 1.2342137103842534,
+      "learning_rate": 3.5906080957300723e-09,
+      "loss": 0.0066,
+      "step": 21250
+    },
+    {
+      "epoch": 4.8297727272727276,
+      "grad_norm": 0.5952464387037373,
+      "learning_rate": 3.5810613925641584e-09,
+      "loss": 0.0009,
+      "step": 21251
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 1.3360722958957751,
+      "learning_rate": 3.571527361214619e-09,
+      "loss": 0.0132,
+      "step": 21252
+    },
+    {
+      "epoch": 4.830227272727273,
+      "grad_norm": 0.62640713301523,
+      "learning_rate": 3.562006001875673e-09,
+      "loss": 0.0037,
+      "step": 21253
+    },
+    {
+      "epoch": 4.830454545454545,
+      "grad_norm": 0.45519570873252607,
+      "learning_rate": 3.552497314741679e-09,
+      "loss": 0.0016,
+      "step": 21254
+    },
+    {
+      "epoch": 4.8306818181818185,
+      "grad_norm": 1.400817934962429,
+      "learning_rate": 3.5430013000063712e-09,
+      "loss": 0.0095,
+      "step": 21255
+    },
+    {
+      "epoch": 4.830909090909091,
+      "grad_norm": 0.47950063734335957,
+      "learning_rate": 3.533517957863483e-09,
+      "loss": 0.0028,
+      "step": 21256
+    },
+    {
+      "epoch": 4.831136363636364,
+      "grad_norm": 1.145762233447895,
+      "learning_rate": 3.524047288506402e-09,
+      "loss": 0.0098,
+      "step": 21257
+    },
+    {
+      "epoch": 4.831363636363636,
+      "grad_norm": 0.8666271233068182,
+      "learning_rate": 3.5145892921281677e-09,
+      "loss": 0.0061,
+      "step": 21258
+    },
+    {
+      "epoch": 4.8315909090909095,
+      "grad_norm": 1.177742591289263,
+      "learning_rate": 3.505143968921751e-09,
+      "loss": 0.0039,
+      "step": 21259
+    },
+    {
+      "epoch": 4.831818181818182,
+      "grad_norm": 0.6044521808501061,
+      "learning_rate": 3.4957113190796365e-09,
+      "loss": 0.0012,
+      "step": 21260
+    },
+    {
+      "epoch": 4.8320454545454545,
+      "grad_norm": 0.7899438708139179,
+      "learning_rate": 3.4862913427943097e-09,
+      "loss": 0.005,
+      "step": 21261
+    },
+    {
+      "epoch": 4.832272727272727,
+      "grad_norm": 0.4120973243462116,
+      "learning_rate": 3.4768840402577692e-09,
+      "loss": 0.0022,
+      "step": 21262
+    },
+    {
+      "epoch": 4.8325,
+      "grad_norm": 1.6255944501011683,
+      "learning_rate": 3.467489411661876e-09,
+      "loss": 0.006,
+      "step": 21263
+    },
+    {
+      "epoch": 4.832727272727273,
+      "grad_norm": 0.8878790395083684,
+      "learning_rate": 3.458107457198143e-09,
+      "loss": 0.0027,
+      "step": 21264
+    },
+    {
+      "epoch": 4.8329545454545455,
+      "grad_norm": 0.2435798295814479,
+      "learning_rate": 3.4487381770580143e-09,
+      "loss": 0.0009,
+      "step": 21265
+    },
+    {
+      "epoch": 4.833181818181818,
+      "grad_norm": 3.4355488680813058,
+      "learning_rate": 3.439381571432379e-09,
+      "loss": 0.0146,
+      "step": 21266
+    },
+    {
+      "epoch": 4.8334090909090905,
+      "grad_norm": 2.282247022311551,
+      "learning_rate": 3.430037640512196e-09,
+      "loss": 0.01,
+      "step": 21267
+    },
+    {
+      "epoch": 4.833636363636364,
+      "grad_norm": 1.13487804247915,
+      "learning_rate": 3.4207063844879374e-09,
+      "loss": 0.0104,
+      "step": 21268
+    },
+    {
+      "epoch": 4.8338636363636365,
+      "grad_norm": 1.0459901450945384,
+      "learning_rate": 3.411387803549868e-09,
+      "loss": 0.0019,
+      "step": 21269
+    },
+    {
+      "epoch": 4.834090909090909,
+      "grad_norm": 2.0435621785013782,
+      "learning_rate": 3.4020818978880434e-09,
+      "loss": 0.0228,
+      "step": 21270
+    },
+    {
+      "epoch": 4.8343181818181815,
+      "grad_norm": 2.661376903257343,
+      "learning_rate": 3.3927886676921734e-09,
+      "loss": 0.0065,
+      "step": 21271
+    },
+    {
+      "epoch": 4.834545454545455,
+      "grad_norm": 1.1786012344886434,
+      "learning_rate": 3.3835081131517592e-09,
+      "loss": 0.0027,
+      "step": 21272
+    },
+    {
+      "epoch": 4.834772727272727,
+      "grad_norm": 1.7197405743247007,
+      "learning_rate": 3.3742402344561635e-09,
+      "loss": 0.0141,
+      "step": 21273
+    },
+    {
+      "epoch": 4.835,
+      "grad_norm": 0.837147093328227,
+      "learning_rate": 3.3649850317942627e-09,
+      "loss": 0.0052,
+      "step": 21274
+    },
+    {
+      "epoch": 4.8352272727272725,
+      "grad_norm": 7.610284584487426,
+      "learning_rate": 3.355742505354795e-09,
+      "loss": 0.0101,
+      "step": 21275
+    },
+    {
+      "epoch": 4.835454545454546,
+      "grad_norm": 1.9341887681278522,
+      "learning_rate": 3.3465126553262895e-09,
+      "loss": 0.0037,
+      "step": 21276
+    },
+    {
+      "epoch": 4.835681818181818,
+      "grad_norm": 0.8134182063613922,
+      "learning_rate": 3.33729548189686e-09,
+      "loss": 0.0024,
+      "step": 21277
+    },
+    {
+      "epoch": 4.835909090909091,
+      "grad_norm": 1.895500648672958,
+      "learning_rate": 3.3280909852546202e-09,
+      "loss": 0.0195,
+      "step": 21278
+    },
+    {
+      "epoch": 4.8361363636363635,
+      "grad_norm": 0.9666826986586958,
+      "learning_rate": 3.318899165587128e-09,
+      "loss": 0.0031,
+      "step": 21279
+    },
+    {
+      "epoch": 4.836363636363636,
+      "grad_norm": 0.8011804938028884,
+      "learning_rate": 3.309720023081872e-09,
+      "loss": 0.0058,
+      "step": 21280
+    },
+    {
+      "epoch": 4.836590909090909,
+      "grad_norm": 0.8120566311178182,
+      "learning_rate": 3.300553557925995e-09,
+      "loss": 0.0027,
+      "step": 21281
+    },
+    {
+      "epoch": 4.836818181818182,
+      "grad_norm": 0.6740695952883143,
+      "learning_rate": 3.2913997703064303e-09,
+      "loss": 0.0029,
+      "step": 21282
+    },
+    {
+      "epoch": 4.837045454545454,
+      "grad_norm": 0.6955387208831608,
+      "learning_rate": 3.282258660409904e-09,
+      "loss": 0.0017,
+      "step": 21283
+    },
+    {
+      "epoch": 4.837272727272727,
+      "grad_norm": 1.3136070221070006,
+      "learning_rate": 3.273130228422794e-09,
+      "loss": 0.0056,
+      "step": 21284
+    },
+    {
+      "epoch": 4.8375,
+      "grad_norm": 0.9658582780468092,
+      "learning_rate": 3.2640144745311325e-09,
+      "loss": 0.0031,
+      "step": 21285
+    },
+    {
+      "epoch": 4.837727272727273,
+      "grad_norm": 1.0297644381978786,
+      "learning_rate": 3.2549113989209515e-09,
+      "loss": 0.0048,
+      "step": 21286
+    },
+    {
+      "epoch": 4.837954545454545,
+      "grad_norm": 3.2066349594332904,
+      "learning_rate": 3.245821001777796e-09,
+      "loss": 0.0205,
+      "step": 21287
+    },
+    {
+      "epoch": 4.838181818181818,
+      "grad_norm": 0.9213139336871925,
+      "learning_rate": 3.236743283287144e-09,
+      "loss": 0.0044,
+      "step": 21288
+    },
+    {
+      "epoch": 4.83840909090909,
+      "grad_norm": 0.7506448248192695,
+      "learning_rate": 3.2276782436339167e-09,
+      "loss": 0.0025,
+      "step": 21289
+    },
+    {
+      "epoch": 4.838636363636364,
+      "grad_norm": 1.2424997213466742,
+      "learning_rate": 3.2186258830031044e-09,
+      "loss": 0.0117,
+      "step": 21290
+    },
+    {
+      "epoch": 4.838863636363636,
+      "grad_norm": 0.8818732553415282,
+      "learning_rate": 3.2095862015792817e-09,
+      "loss": 0.0022,
+      "step": 21291
+    },
+    {
+      "epoch": 4.839090909090909,
+      "grad_norm": 3.1423131785306055,
+      "learning_rate": 3.200559199546746e-09,
+      "loss": 0.0125,
+      "step": 21292
+    },
+    {
+      "epoch": 4.839318181818182,
+      "grad_norm": 1.156843536291129,
+      "learning_rate": 3.1915448770896557e-09,
+      "loss": 0.0046,
+      "step": 21293
+    },
+    {
+      "epoch": 4.839545454545455,
+      "grad_norm": 0.9126903780130887,
+      "learning_rate": 3.1825432343916828e-09,
+      "loss": 0.0053,
+      "step": 21294
+    },
+    {
+      "epoch": 4.839772727272727,
+      "grad_norm": 1.4954982995099593,
+      "learning_rate": 3.1735542716365e-09,
+      "loss": 0.0081,
+      "step": 21295
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 1.6452065555526305,
+      "learning_rate": 3.1645779890074335e-09,
+      "loss": 0.0061,
+      "step": 21296
+    },
+    {
+      "epoch": 4.840227272727272,
+      "grad_norm": 0.9718925508810184,
+      "learning_rate": 3.155614386687392e-09,
+      "loss": 0.0036,
+      "step": 21297
+    },
+    {
+      "epoch": 4.840454545454546,
+      "grad_norm": 0.19704276147917296,
+      "learning_rate": 3.146663464859284e-09,
+      "loss": 0.0004,
+      "step": 21298
+    },
+    {
+      "epoch": 4.840681818181818,
+      "grad_norm": 1.4431484948422952,
+      "learning_rate": 3.137725223705604e-09,
+      "loss": 0.0144,
+      "step": 21299
+    },
+    {
+      "epoch": 4.840909090909091,
+      "grad_norm": 0.8409533880531139,
+      "learning_rate": 3.128799663408566e-09,
+      "loss": 0.0034,
+      "step": 21300
+    },
+    {
+      "epoch": 4.841136363636363,
+      "grad_norm": 4.057748647847217,
+      "learning_rate": 3.1198867841502474e-09,
+      "loss": 0.0201,
+      "step": 21301
+    },
+    {
+      "epoch": 4.841363636363637,
+      "grad_norm": 1.0373309077782824,
+      "learning_rate": 3.1109865861123774e-09,
+      "loss": 0.0034,
+      "step": 21302
+    },
+    {
+      "epoch": 4.841590909090909,
+      "grad_norm": 0.6914483367588196,
+      "learning_rate": 3.102099069476408e-09,
+      "loss": 0.0059,
+      "step": 21303
+    },
+    {
+      "epoch": 4.841818181818182,
+      "grad_norm": 0.8784912056135904,
+      "learning_rate": 3.093224234423653e-09,
+      "loss": 0.0035,
+      "step": 21304
+    },
+    {
+      "epoch": 4.842045454545454,
+      "grad_norm": 1.4516651664690532,
+      "learning_rate": 3.084362081135009e-09,
+      "loss": 0.0045,
+      "step": 21305
+    },
+    {
+      "epoch": 4.842272727272727,
+      "grad_norm": 0.7365337512310405,
+      "learning_rate": 3.0755126097911647e-09,
+      "loss": 0.0014,
+      "step": 21306
+    },
+    {
+      "epoch": 4.8425,
+      "grad_norm": 1.2574544254875806,
+      "learning_rate": 3.0666758205726704e-09,
+      "loss": 0.0087,
+      "step": 21307
+    },
+    {
+      "epoch": 4.842727272727273,
+      "grad_norm": 2.1211217261978277,
+      "learning_rate": 3.0578517136596597e-09,
+      "loss": 0.012,
+      "step": 21308
+    },
+    {
+      "epoch": 4.842954545454545,
+      "grad_norm": 0.4224493634286749,
+      "learning_rate": 3.0490402892320582e-09,
+      "loss": 0.001,
+      "step": 21309
+    },
+    {
+      "epoch": 4.843181818181818,
+      "grad_norm": 0.9897673265035436,
+      "learning_rate": 3.040241547469583e-09,
+      "loss": 0.0059,
+      "step": 21310
+    },
+    {
+      "epoch": 4.843409090909091,
+      "grad_norm": 2.069617467780976,
+      "learning_rate": 3.031455488551674e-09,
+      "loss": 0.013,
+      "step": 21311
+    },
+    {
+      "epoch": 4.843636363636364,
+      "grad_norm": 0.7654129547956069,
+      "learning_rate": 3.0226821126574933e-09,
+      "loss": 0.0021,
+      "step": 21312
+    },
+    {
+      "epoch": 4.843863636363636,
+      "grad_norm": 0.5681090539709875,
+      "learning_rate": 3.013921419965926e-09,
+      "loss": 0.0023,
+      "step": 21313
+    },
+    {
+      "epoch": 4.844090909090909,
+      "grad_norm": 1.4457805077114294,
+      "learning_rate": 3.005173410655579e-09,
+      "loss": 0.0128,
+      "step": 21314
+    },
+    {
+      "epoch": 4.844318181818182,
+      "grad_norm": 0.30062215650065877,
+      "learning_rate": 2.9964380849049206e-09,
+      "loss": 0.0006,
+      "step": 21315
+    },
+    {
+      "epoch": 4.844545454545455,
+      "grad_norm": 0.974455418391349,
+      "learning_rate": 2.987715442892003e-09,
+      "loss": 0.002,
+      "step": 21316
+    },
+    {
+      "epoch": 4.844772727272727,
+      "grad_norm": 1.3134216083349306,
+      "learning_rate": 2.9790054847947394e-09,
+      "loss": 0.0034,
+      "step": 21317
+    },
+    {
+      "epoch": 4.845,
+      "grad_norm": 0.632826176165839,
+      "learning_rate": 2.9703082107906957e-09,
+      "loss": 0.0021,
+      "step": 21318
+    },
+    {
+      "epoch": 4.845227272727273,
+      "grad_norm": 0.9506055647790341,
+      "learning_rate": 2.9616236210572997e-09,
+      "loss": 0.0043,
+      "step": 21319
+    },
+    {
+      "epoch": 4.845454545454546,
+      "grad_norm": 3.4139908666550047,
+      "learning_rate": 2.9529517157715628e-09,
+      "loss": 0.0082,
+      "step": 21320
+    },
+    {
+      "epoch": 4.845681818181818,
+      "grad_norm": 1.2635823410593217,
+      "learning_rate": 2.944292495110426e-09,
+      "loss": 0.0072,
+      "step": 21321
+    },
+    {
+      "epoch": 4.845909090909091,
+      "grad_norm": 1.6301310268598286,
+      "learning_rate": 2.9356459592503466e-09,
+      "loss": 0.0132,
+      "step": 21322
+    },
+    {
+      "epoch": 4.846136363636363,
+      "grad_norm": 0.5235989059145895,
+      "learning_rate": 2.9270121083677107e-09,
+      "loss": 0.0019,
+      "step": 21323
+    },
+    {
+      "epoch": 4.846363636363637,
+      "grad_norm": 1.4423982642154476,
+      "learning_rate": 2.9183909426385576e-09,
+      "loss": 0.004,
+      "step": 21324
+    },
+    {
+      "epoch": 4.846590909090909,
+      "grad_norm": 0.5657566514945609,
+      "learning_rate": 2.9097824622387193e-09,
+      "loss": 0.0031,
+      "step": 21325
+    },
+    {
+      "epoch": 4.846818181818182,
+      "grad_norm": 0.7739846118409374,
+      "learning_rate": 2.9011866673436805e-09,
+      "loss": 0.0035,
+      "step": 21326
+    },
+    {
+      "epoch": 4.847045454545454,
+      "grad_norm": 0.5576098717664729,
+      "learning_rate": 2.892603558128787e-09,
+      "loss": 0.0014,
+      "step": 21327
+    },
+    {
+      "epoch": 4.847272727272728,
+      "grad_norm": 1.5158277590004603,
+      "learning_rate": 2.884033134768968e-09,
+      "loss": 0.0077,
+      "step": 21328
+    },
+    {
+      "epoch": 4.8475,
+      "grad_norm": 1.6766704514904494,
+      "learning_rate": 2.8754753974391537e-09,
+      "loss": 0.0074,
+      "step": 21329
+    },
+    {
+      "epoch": 4.847727272727273,
+      "grad_norm": 1.714696865348992,
+      "learning_rate": 2.8669303463136484e-09,
+      "loss": 0.0094,
+      "step": 21330
+    },
+    {
+      "epoch": 4.847954545454545,
+      "grad_norm": 1.786561123149584,
+      "learning_rate": 2.858397981566896e-09,
+      "loss": 0.0183,
+      "step": 21331
+    },
+    {
+      "epoch": 4.848181818181818,
+      "grad_norm": 0.3884788922644131,
+      "learning_rate": 2.849878303372716e-09,
+      "loss": 0.0014,
+      "step": 21332
+    },
+    {
+      "epoch": 4.848409090909091,
+      "grad_norm": 1.6375821682566645,
+      "learning_rate": 2.8413713119049973e-09,
+      "loss": 0.0103,
+      "step": 21333
+    },
+    {
+      "epoch": 4.848636363636364,
+      "grad_norm": 2.8873789662216227,
+      "learning_rate": 2.832877007337073e-09,
+      "loss": 0.0169,
+      "step": 21334
+    },
+    {
+      "epoch": 4.848863636363636,
+      "grad_norm": 0.3184158093751422,
+      "learning_rate": 2.8243953898422077e-09,
+      "loss": 0.0009,
+      "step": 21335
+    },
+    {
+      "epoch": 4.84909090909091,
+      "grad_norm": 0.3181026935359169,
+      "learning_rate": 2.815926459593457e-09,
+      "loss": 0.0017,
+      "step": 21336
+    },
+    {
+      "epoch": 4.849318181818182,
+      "grad_norm": 1.4528579492607787,
+      "learning_rate": 2.8074702167633223e-09,
+      "loss": 0.0133,
+      "step": 21337
+    },
+    {
+      "epoch": 4.849545454545455,
+      "grad_norm": 1.2519744622482338,
+      "learning_rate": 2.799026661524373e-09,
+      "loss": 0.0046,
+      "step": 21338
+    },
+    {
+      "epoch": 4.849772727272727,
+      "grad_norm": 3.451939540830859,
+      "learning_rate": 2.7905957940487645e-09,
+      "loss": 0.0169,
+      "step": 21339
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 0.8524880321819377,
+      "learning_rate": 2.782177614508441e-09,
+      "loss": 0.0019,
+      "step": 21340
+    },
+    {
+      "epoch": 4.850227272727273,
+      "grad_norm": 0.4665033671440423,
+      "learning_rate": 2.7737721230750025e-09,
+      "loss": 0.001,
+      "step": 21341
+    },
+    {
+      "epoch": 4.850454545454546,
+      "grad_norm": 0.22203124721627196,
+      "learning_rate": 2.7653793199199087e-09,
+      "loss": 0.0005,
+      "step": 21342
+    },
+    {
+      "epoch": 4.850681818181818,
+      "grad_norm": 0.3137328186090868,
+      "learning_rate": 2.756999205214203e-09,
+      "loss": 0.0015,
+      "step": 21343
+    },
+    {
+      "epoch": 4.850909090909091,
+      "grad_norm": 0.3271442811852377,
+      "learning_rate": 2.7486317791289295e-09,
+      "loss": 0.001,
+      "step": 21344
+    },
+    {
+      "epoch": 4.851136363636364,
+      "grad_norm": 0.21063261683146903,
+      "learning_rate": 2.7402770418346465e-09,
+      "loss": 0.0007,
+      "step": 21345
+    },
+    {
+      "epoch": 4.8513636363636365,
+      "grad_norm": 0.6600846329593244,
+      "learning_rate": 2.7319349935016336e-09,
+      "loss": 0.0034,
+      "step": 21346
+    },
+    {
+      "epoch": 4.851590909090909,
+      "grad_norm": 1.574843118160513,
+      "learning_rate": 2.7236056343001026e-09,
+      "loss": 0.0059,
+      "step": 21347
+    },
+    {
+      "epoch": 4.851818181818182,
+      "grad_norm": 0.5826079565953904,
+      "learning_rate": 2.715288964399848e-09,
+      "loss": 0.0022,
+      "step": 21348
+    },
+    {
+      "epoch": 4.852045454545454,
+      "grad_norm": 0.45770794811092047,
+      "learning_rate": 2.706984983970526e-09,
+      "loss": 0.001,
+      "step": 21349
+    },
+    {
+      "epoch": 4.8522727272727275,
+      "grad_norm": 0.5172548819265963,
+      "learning_rate": 2.698693693181445e-09,
+      "loss": 0.0014,
+      "step": 21350
+    },
+    {
+      "epoch": 4.8525,
+      "grad_norm": 0.9952233619087092,
+      "learning_rate": 2.690415092201637e-09,
+      "loss": 0.008,
+      "step": 21351
+    },
+    {
+      "epoch": 4.852727272727273,
+      "grad_norm": 2.3367055210584047,
+      "learning_rate": 2.682149181199925e-09,
+      "loss": 0.006,
+      "step": 21352
+    },
+    {
+      "epoch": 4.852954545454545,
+      "grad_norm": 0.5927524709579929,
+      "learning_rate": 2.6738959603449242e-09,
+      "loss": 0.0026,
+      "step": 21353
+    },
+    {
+      "epoch": 4.8531818181818185,
+      "grad_norm": 2.1323153297192996,
+      "learning_rate": 2.6656554298048335e-09,
+      "loss": 0.0057,
+      "step": 21354
+    },
+    {
+      "epoch": 4.853409090909091,
+      "grad_norm": 2.2928423867167433,
+      "learning_rate": 2.6574275897477825e-09,
+      "loss": 0.0094,
+      "step": 21355
+    },
+    {
+      "epoch": 4.8536363636363635,
+      "grad_norm": 1.0267146198029176,
+      "learning_rate": 2.649212440341553e-09,
+      "loss": 0.0027,
+      "step": 21356
+    },
+    {
+      "epoch": 4.853863636363636,
+      "grad_norm": 1.651950535309702,
+      "learning_rate": 2.641009981753581e-09,
+      "loss": 0.014,
+      "step": 21357
+    },
+    {
+      "epoch": 4.8540909090909095,
+      "grad_norm": 0.3979702490135378,
+      "learning_rate": 2.6328202141512327e-09,
+      "loss": 0.0016,
+      "step": 21358
+    },
+    {
+      "epoch": 4.854318181818182,
+      "grad_norm": 0.8146185608725767,
+      "learning_rate": 2.624643137701388e-09,
+      "loss": 0.0045,
+      "step": 21359
+    },
+    {
+      "epoch": 4.8545454545454545,
+      "grad_norm": 0.9866106388343562,
+      "learning_rate": 2.616478752570928e-09,
+      "loss": 0.0052,
+      "step": 21360
+    },
+    {
+      "epoch": 4.854772727272727,
+      "grad_norm": 1.2217632233887983,
+      "learning_rate": 2.6083270589262467e-09,
+      "loss": 0.004,
+      "step": 21361
+    },
+    {
+      "epoch": 4.855,
+      "grad_norm": 0.24232499277682953,
+      "learning_rate": 2.6001880569336e-09,
+      "loss": 0.0008,
+      "step": 21362
+    },
+    {
+      "epoch": 4.855227272727273,
+      "grad_norm": 1.4591079666114508,
+      "learning_rate": 2.592061746758967e-09,
+      "loss": 0.0048,
+      "step": 21363
+    },
+    {
+      "epoch": 4.8554545454545455,
+      "grad_norm": 1.7110111333309372,
+      "learning_rate": 2.583948128568048e-09,
+      "loss": 0.0046,
+      "step": 21364
+    },
+    {
+      "epoch": 4.855681818181818,
+      "grad_norm": 1.6400467655073088,
+      "learning_rate": 2.575847202526266e-09,
+      "loss": 0.0116,
+      "step": 21365
+    },
+    {
+      "epoch": 4.8559090909090905,
+      "grad_norm": 1.0277865674974143,
+      "learning_rate": 2.5677589687988366e-09,
+      "loss": 0.0043,
+      "step": 21366
+    },
+    {
+      "epoch": 4.856136363636364,
+      "grad_norm": 0.6662681546642583,
+      "learning_rate": 2.5596834275506965e-09,
+      "loss": 0.0027,
+      "step": 21367
+    },
+    {
+      "epoch": 4.856363636363636,
+      "grad_norm": 1.2806179310275643,
+      "learning_rate": 2.551620578946576e-09,
+      "loss": 0.0125,
+      "step": 21368
+    },
+    {
+      "epoch": 4.856590909090909,
+      "grad_norm": 1.3613942824505365,
+      "learning_rate": 2.543570423150718e-09,
+      "loss": 0.0041,
+      "step": 21369
+    },
+    {
+      "epoch": 4.8568181818181815,
+      "grad_norm": 0.5239129659393059,
+      "learning_rate": 2.5355329603275054e-09,
+      "loss": 0.0014,
+      "step": 21370
+    },
+    {
+      "epoch": 4.857045454545455,
+      "grad_norm": 0.8221628523596297,
+      "learning_rate": 2.5275081906406263e-09,
+      "loss": 0.0045,
+      "step": 21371
+    },
+    {
+      "epoch": 4.857272727272727,
+      "grad_norm": 1.8357292187646188,
+      "learning_rate": 2.5194961142539083e-09,
+      "loss": 0.0247,
+      "step": 21372
+    },
+    {
+      "epoch": 4.8575,
+      "grad_norm": 0.7030201316878544,
+      "learning_rate": 2.5114967313306234e-09,
+      "loss": 0.0068,
+      "step": 21373
+    },
+    {
+      "epoch": 4.8577272727272724,
+      "grad_norm": 0.6743233204088351,
+      "learning_rate": 2.5035100420338352e-09,
+      "loss": 0.0025,
+      "step": 21374
+    },
+    {
+      "epoch": 4.857954545454545,
+      "grad_norm": 1.6145389262052208,
+      "learning_rate": 2.4955360465265394e-09,
+      "loss": 0.0052,
+      "step": 21375
+    },
+    {
+      "epoch": 4.858181818181818,
+      "grad_norm": 0.9537630616930668,
+      "learning_rate": 2.4875747449713132e-09,
+      "loss": 0.0032,
+      "step": 21376
+    },
+    {
+      "epoch": 4.858409090909091,
+      "grad_norm": 1.2475930774109714,
+      "learning_rate": 2.4796261375304583e-09,
+      "loss": 0.0065,
+      "step": 21377
+    },
+    {
+      "epoch": 4.858636363636363,
+      "grad_norm": 1.4292832710454335,
+      "learning_rate": 2.4716902243660664e-09,
+      "loss": 0.0069,
+      "step": 21378
+    },
+    {
+      "epoch": 4.858863636363637,
+      "grad_norm": 1.6919082703440909,
+      "learning_rate": 2.4637670056399527e-09,
+      "loss": 0.0078,
+      "step": 21379
+    },
+    {
+      "epoch": 4.859090909090909,
+      "grad_norm": 1.8626458832923962,
+      "learning_rate": 2.4558564815137243e-09,
+      "loss": 0.0064,
+      "step": 21380
+    },
+    {
+      "epoch": 4.859318181818182,
+      "grad_norm": 0.5974414255511772,
+      "learning_rate": 2.447958652148641e-09,
+      "loss": 0.0034,
+      "step": 21381
+    },
+    {
+      "epoch": 4.859545454545454,
+      "grad_norm": 0.8776877427965629,
+      "learning_rate": 2.4400735177057544e-09,
+      "loss": 0.0039,
+      "step": 21382
+    },
+    {
+      "epoch": 4.859772727272727,
+      "grad_norm": 0.6014418985683285,
+      "learning_rate": 2.4322010783459777e-09,
+      "loss": 0.0034,
+      "step": 21383
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 1.764141670320305,
+      "learning_rate": 2.4243413342296684e-09,
+      "loss": 0.0056,
+      "step": 21384
+    },
+    {
+      "epoch": 4.860227272727273,
+      "grad_norm": 0.4701531385947711,
+      "learning_rate": 2.4164942855171845e-09,
+      "loss": 0.0019,
+      "step": 21385
+    },
+    {
+      "epoch": 4.860454545454545,
+      "grad_norm": 0.7448973476013979,
+      "learning_rate": 2.4086599323686066e-09,
+      "loss": 0.003,
+      "step": 21386
+    },
+    {
+      "epoch": 4.860681818181818,
+      "grad_norm": 0.36709460725202575,
+      "learning_rate": 2.400838274943529e-09,
+      "loss": 0.0012,
+      "step": 21387
+    },
+    {
+      "epoch": 4.860909090909091,
+      "grad_norm": 1.4004053134971228,
+      "learning_rate": 2.3930293134016152e-09,
+      "loss": 0.0067,
+      "step": 21388
+    },
+    {
+      "epoch": 4.861136363636364,
+      "grad_norm": 0.36177830987273574,
+      "learning_rate": 2.3852330479020443e-09,
+      "loss": 0.001,
+      "step": 21389
+    },
+    {
+      "epoch": 4.861363636363636,
+      "grad_norm": 0.3451319497942192,
+      "learning_rate": 2.3774494786037163e-09,
+      "loss": 0.001,
+      "step": 21390
+    },
+    {
+      "epoch": 4.861590909090909,
+      "grad_norm": 0.4803630486904989,
+      "learning_rate": 2.3696786056654626e-09,
+      "loss": 0.0019,
+      "step": 21391
+    },
+    {
+      "epoch": 4.861818181818181,
+      "grad_norm": 1.5736614107197648,
+      "learning_rate": 2.3619204292456982e-09,
+      "loss": 0.0064,
+      "step": 21392
+    },
+    {
+      "epoch": 4.862045454545455,
+      "grad_norm": 0.9013831505121126,
+      "learning_rate": 2.3541749495026302e-09,
+      "loss": 0.004,
+      "step": 21393
+    },
+    {
+      "epoch": 4.862272727272727,
+      "grad_norm": 0.646954066189672,
+      "learning_rate": 2.346442166594187e-09,
+      "loss": 0.0041,
+      "step": 21394
+    },
+    {
+      "epoch": 4.8625,
+      "grad_norm": 1.6248982848866493,
+      "learning_rate": 2.338722080678091e-09,
+      "loss": 0.0119,
+      "step": 21395
+    },
+    {
+      "epoch": 4.862727272727272,
+      "grad_norm": 0.7521526530079525,
+      "learning_rate": 2.331014691911715e-09,
+      "loss": 0.0024,
+      "step": 21396
+    },
+    {
+      "epoch": 4.862954545454546,
+      "grad_norm": 0.62748490873031,
+      "learning_rate": 2.3233200004522947e-09,
+      "loss": 0.0011,
+      "step": 21397
+    },
+    {
+      "epoch": 4.863181818181818,
+      "grad_norm": 0.5774808369861496,
+      "learning_rate": 2.3156380064566495e-09,
+      "loss": 0.0028,
+      "step": 21398
+    },
+    {
+      "epoch": 4.863409090909091,
+      "grad_norm": 0.655501411067765,
+      "learning_rate": 2.3079687100814596e-09,
+      "loss": 0.0041,
+      "step": 21399
+    },
+    {
+      "epoch": 4.863636363636363,
+      "grad_norm": 0.6735329621578324,
+      "learning_rate": 2.300312111483197e-09,
+      "loss": 0.0019,
+      "step": 21400
+    },
+    {
+      "epoch": 4.863863636363637,
+      "grad_norm": 0.1922105246885915,
+      "learning_rate": 2.2926682108179164e-09,
+      "loss": 0.0007,
+      "step": 21401
+    },
+    {
+      "epoch": 4.864090909090909,
+      "grad_norm": 1.4010246712293317,
+      "learning_rate": 2.285037008241467e-09,
+      "loss": 0.0082,
+      "step": 21402
+    },
+    {
+      "epoch": 4.864318181818182,
+      "grad_norm": 0.6209382337728498,
+      "learning_rate": 2.277418503909487e-09,
+      "loss": 0.0024,
+      "step": 21403
+    },
+    {
+      "epoch": 4.864545454545454,
+      "grad_norm": 3.568685696364345,
+      "learning_rate": 2.2698126979774082e-09,
+      "loss": 0.0073,
+      "step": 21404
+    },
+    {
+      "epoch": 4.864772727272728,
+      "grad_norm": 1.2882053484699136,
+      "learning_rate": 2.2622195906002454e-09,
+      "loss": 0.011,
+      "step": 21405
+    },
+    {
+      "epoch": 4.865,
+      "grad_norm": 0.6307903613228046,
+      "learning_rate": 2.254639181932805e-09,
+      "loss": 0.0042,
+      "step": 21406
+    },
+    {
+      "epoch": 4.865227272727273,
+      "grad_norm": 1.3509557020960863,
+      "learning_rate": 2.2470714721297555e-09,
+      "loss": 0.0029,
+      "step": 21407
+    },
+    {
+      "epoch": 4.865454545454545,
+      "grad_norm": 1.5758756193826091,
+      "learning_rate": 2.239516461345348e-09,
+      "loss": 0.0057,
+      "step": 21408
+    },
+    {
+      "epoch": 4.865681818181818,
+      "grad_norm": 0.9974557258809192,
+      "learning_rate": 2.2319741497336956e-09,
+      "loss": 0.0032,
+      "step": 21409
+    },
+    {
+      "epoch": 4.865909090909091,
+      "grad_norm": 1.813357333126763,
+      "learning_rate": 2.2244445374484946e-09,
+      "loss": 0.008,
+      "step": 21410
+    },
+    {
+      "epoch": 4.866136363636364,
+      "grad_norm": 1.0868325418144396,
+      "learning_rate": 2.216927624643442e-09,
+      "loss": 0.0064,
+      "step": 21411
+    },
+    {
+      "epoch": 4.866363636363636,
+      "grad_norm": 0.8544802253429548,
+      "learning_rate": 2.209423411471748e-09,
+      "loss": 0.0026,
+      "step": 21412
+    },
+    {
+      "epoch": 4.866590909090909,
+      "grad_norm": 0.5764436998869062,
+      "learning_rate": 2.2019318980863456e-09,
+      "loss": 0.0022,
+      "step": 21413
+    },
+    {
+      "epoch": 4.866818181818182,
+      "grad_norm": 0.588282561308941,
+      "learning_rate": 2.194453084640169e-09,
+      "loss": 0.0031,
+      "step": 21414
+    },
+    {
+      "epoch": 4.867045454545455,
+      "grad_norm": 0.6954916610468963,
+      "learning_rate": 2.1869869712855264e-09,
+      "loss": 0.0017,
+      "step": 21415
+    },
+    {
+      "epoch": 4.867272727272727,
+      "grad_norm": 0.7717505450698416,
+      "learning_rate": 2.179533558174865e-09,
+      "loss": 0.0039,
+      "step": 21416
+    },
+    {
+      "epoch": 4.8675,
+      "grad_norm": 1.7110579614343673,
+      "learning_rate": 2.1720928454600777e-09,
+      "loss": 0.0067,
+      "step": 21417
+    },
+    {
+      "epoch": 4.867727272727272,
+      "grad_norm": 1.3462224723952299,
+      "learning_rate": 2.164664833292848e-09,
+      "loss": 0.0072,
+      "step": 21418
+    },
+    {
+      "epoch": 4.867954545454546,
+      "grad_norm": 0.5035214321384535,
+      "learning_rate": 2.157249521824792e-09,
+      "loss": 0.0022,
+      "step": 21419
+    },
+    {
+      "epoch": 4.868181818181818,
+      "grad_norm": 0.33391073631477897,
+      "learning_rate": 2.1498469112069682e-09,
+      "loss": 0.0011,
+      "step": 21420
+    },
+    {
+      "epoch": 4.868409090909091,
+      "grad_norm": 0.7160449556160929,
+      "learning_rate": 2.142457001590367e-09,
+      "loss": 0.0069,
+      "step": 21421
+    },
+    {
+      "epoch": 4.868636363636364,
+      "grad_norm": 0.827923935640624,
+      "learning_rate": 2.1350797931257717e-09,
+      "loss": 0.0043,
+      "step": 21422
+    },
+    {
+      "epoch": 4.868863636363637,
+      "grad_norm": 0.7704786971275727,
+      "learning_rate": 2.1277152859634774e-09,
+      "loss": 0.0019,
+      "step": 21423
+    },
+    {
+      "epoch": 4.869090909090909,
+      "grad_norm": 1.2196874463979654,
+      "learning_rate": 2.1203634802537816e-09,
+      "loss": 0.0043,
+      "step": 21424
+    },
+    {
+      "epoch": 4.869318181818182,
+      "grad_norm": 2.418276923843272,
+      "learning_rate": 2.113024376146494e-09,
+      "loss": 0.0073,
+      "step": 21425
+    },
+    {
+      "epoch": 4.869545454545454,
+      "grad_norm": 0.8950454249802737,
+      "learning_rate": 2.105697973791357e-09,
+      "loss": 0.0038,
+      "step": 21426
+    },
+    {
+      "epoch": 4.869772727272728,
+      "grad_norm": 0.8798113689022109,
+      "learning_rate": 2.0983842733376946e-09,
+      "loss": 0.0057,
+      "step": 21427
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 1.4394392540578336,
+      "learning_rate": 2.091083274934763e-09,
+      "loss": 0.0035,
+      "step": 21428
+    },
+    {
+      "epoch": 4.870227272727273,
+      "grad_norm": 2.3502391631187667,
+      "learning_rate": 2.083794978731332e-09,
+      "loss": 0.0053,
+      "step": 21429
+    },
+    {
+      "epoch": 4.870454545454545,
+      "grad_norm": 0.24921289499061167,
+      "learning_rate": 2.0765193848760327e-09,
+      "loss": 0.0008,
+      "step": 21430
+    },
+    {
+      "epoch": 4.870681818181819,
+      "grad_norm": 1.0542612937749698,
+      "learning_rate": 2.069256493517288e-09,
+      "loss": 0.0038,
+      "step": 21431
+    },
+    {
+      "epoch": 4.870909090909091,
+      "grad_norm": 0.4385710293784371,
+      "learning_rate": 2.062006304803174e-09,
+      "loss": 0.0016,
+      "step": 21432
+    },
+    {
+      "epoch": 4.871136363636364,
+      "grad_norm": 0.29650181200304343,
+      "learning_rate": 2.054768818881489e-09,
+      "loss": 0.0011,
+      "step": 21433
+    },
+    {
+      "epoch": 4.871363636363636,
+      "grad_norm": 0.4618579465077135,
+      "learning_rate": 2.0475440358998927e-09,
+      "loss": 0.0015,
+      "step": 21434
+    },
+    {
+      "epoch": 4.871590909090909,
+      "grad_norm": 1.521352790202688,
+      "learning_rate": 2.0403319560056285e-09,
+      "loss": 0.0053,
+      "step": 21435
+    },
+    {
+      "epoch": 4.871818181818182,
+      "grad_norm": 0.5921182384606707,
+      "learning_rate": 2.0331325793458005e-09,
+      "loss": 0.0028,
+      "step": 21436
+    },
+    {
+      "epoch": 4.872045454545455,
+      "grad_norm": 0.4100171733660847,
+      "learning_rate": 2.025945906067306e-09,
+      "loss": 0.0012,
+      "step": 21437
+    },
+    {
+      "epoch": 4.872272727272727,
+      "grad_norm": 1.3460313228547818,
+      "learning_rate": 2.018771936316555e-09,
+      "loss": 0.0058,
+      "step": 21438
+    },
+    {
+      "epoch": 4.8725000000000005,
+      "grad_norm": 0.628795449779835,
+      "learning_rate": 2.0116106702398887e-09,
+      "loss": 0.0018,
+      "step": 21439
+    },
+    {
+      "epoch": 4.872727272727273,
+      "grad_norm": 0.6699776311308691,
+      "learning_rate": 2.004462107983371e-09,
+      "loss": 0.0029,
+      "step": 21440
+    },
+    {
+      "epoch": 4.8729545454545455,
+      "grad_norm": 2.4330511753714736,
+      "learning_rate": 1.9973262496927882e-09,
+      "loss": 0.0087,
+      "step": 21441
+    },
+    {
+      "epoch": 4.873181818181818,
+      "grad_norm": 1.5412156834414088,
+      "learning_rate": 1.99020309551351e-09,
+      "loss": 0.0144,
+      "step": 21442
+    },
+    {
+      "epoch": 4.873409090909091,
+      "grad_norm": 0.369173147919116,
+      "learning_rate": 1.9830926455909756e-09,
+      "loss": 0.0008,
+      "step": 21443
+    },
+    {
+      "epoch": 4.873636363636364,
+      "grad_norm": 1.2968211957022744,
+      "learning_rate": 1.9759949000699996e-09,
+      "loss": 0.0084,
+      "step": 21444
+    },
+    {
+      "epoch": 4.8738636363636365,
+      "grad_norm": 0.6798561975393406,
+      "learning_rate": 1.968909859095466e-09,
+      "loss": 0.0011,
+      "step": 21445
+    },
+    {
+      "epoch": 4.874090909090909,
+      "grad_norm": 1.2561440131132027,
+      "learning_rate": 1.9618375228118438e-09,
+      "loss": 0.0069,
+      "step": 21446
+    },
+    {
+      "epoch": 4.8743181818181816,
+      "grad_norm": 1.447268362936246,
+      "learning_rate": 1.954777891363252e-09,
+      "loss": 0.008,
+      "step": 21447
+    },
+    {
+      "epoch": 4.874545454545455,
+      "grad_norm": 1.3948544785092347,
+      "learning_rate": 1.9477309648936737e-09,
+      "loss": 0.0087,
+      "step": 21448
+    },
+    {
+      "epoch": 4.8747727272727275,
+      "grad_norm": 0.8674690528726561,
+      "learning_rate": 1.9406967435468827e-09,
+      "loss": 0.0023,
+      "step": 21449
+    },
+    {
+      "epoch": 4.875,
+      "grad_norm": 0.6011907277030133,
+      "learning_rate": 1.9336752274663056e-09,
+      "loss": 0.0042,
+      "step": 21450
+    },
+    {
+      "epoch": 4.8752272727272725,
+      "grad_norm": 0.952531121566364,
+      "learning_rate": 1.9266664167950226e-09,
+      "loss": 0.0023,
+      "step": 21451
+    },
+    {
+      "epoch": 4.875454545454545,
+      "grad_norm": 0.8695705451473722,
+      "learning_rate": 1.9196703116760447e-09,
+      "loss": 0.0051,
+      "step": 21452
+    },
+    {
+      "epoch": 4.8756818181818184,
+      "grad_norm": 2.1533158082759427,
+      "learning_rate": 1.912686912252035e-09,
+      "loss": 0.0075,
+      "step": 21453
+    },
+    {
+      "epoch": 4.875909090909091,
+      "grad_norm": 0.7951209306927998,
+      "learning_rate": 1.9057162186653098e-09,
+      "loss": 0.0063,
+      "step": 21454
+    },
+    {
+      "epoch": 4.8761363636363635,
+      "grad_norm": 2.7590903693323474,
+      "learning_rate": 1.898758231058118e-09,
+      "loss": 0.0108,
+      "step": 21455
+    },
+    {
+      "epoch": 4.876363636363636,
+      "grad_norm": 1.1254727574291803,
+      "learning_rate": 1.891812949572358e-09,
+      "loss": 0.004,
+      "step": 21456
+    },
+    {
+      "epoch": 4.876590909090909,
+      "grad_norm": 0.3129103393874435,
+      "learning_rate": 1.8848803743495154e-09,
+      "loss": 0.0025,
+      "step": 21457
+    },
+    {
+      "epoch": 4.876818181818182,
+      "grad_norm": 1.2008199208656296,
+      "learning_rate": 1.8779605055310736e-09,
+      "loss": 0.0022,
+      "step": 21458
+    },
+    {
+      "epoch": 4.8770454545454545,
+      "grad_norm": 1.1876588274636957,
+      "learning_rate": 1.8710533432581694e-09,
+      "loss": 0.0057,
+      "step": 21459
+    },
+    {
+      "epoch": 4.877272727272727,
+      "grad_norm": 0.4424210874011084,
+      "learning_rate": 1.8641588876715243e-09,
+      "loss": 0.0022,
+      "step": 21460
+    },
+    {
+      "epoch": 4.8774999999999995,
+      "grad_norm": 0.40127179690150266,
+      "learning_rate": 1.8572771389118587e-09,
+      "loss": 0.0005,
+      "step": 21461
+    },
+    {
+      "epoch": 4.877727272727273,
+      "grad_norm": 3.1877932889119776,
+      "learning_rate": 1.8504080971194077e-09,
+      "loss": 0.0113,
+      "step": 21462
+    },
+    {
+      "epoch": 4.877954545454545,
+      "grad_norm": 5.162171002809758,
+      "learning_rate": 1.8435517624342675e-09,
+      "loss": 0.0112,
+      "step": 21463
+    },
+    {
+      "epoch": 4.878181818181818,
+      "grad_norm": 0.5791272668335747,
+      "learning_rate": 1.8367081349962568e-09,
+      "loss": 0.005,
+      "step": 21464
+    },
+    {
+      "epoch": 4.878409090909091,
+      "grad_norm": 1.7690397970522937,
+      "learning_rate": 1.829877214944986e-09,
+      "loss": 0.0046,
+      "step": 21465
+    },
+    {
+      "epoch": 4.878636363636364,
+      "grad_norm": 1.218778153423266,
+      "learning_rate": 1.82305900241965e-09,
+      "loss": 0.004,
+      "step": 21466
+    },
+    {
+      "epoch": 4.878863636363636,
+      "grad_norm": 0.4840673450329766,
+      "learning_rate": 1.8162534975593727e-09,
+      "loss": 0.0017,
+      "step": 21467
+    },
+    {
+      "epoch": 4.879090909090909,
+      "grad_norm": 0.7357397163027962,
+      "learning_rate": 1.809460700502863e-09,
+      "loss": 0.002,
+      "step": 21468
+    },
+    {
+      "epoch": 4.879318181818181,
+      "grad_norm": 0.864158679272739,
+      "learning_rate": 1.8026806113886907e-09,
+      "loss": 0.0037,
+      "step": 21469
+    },
+    {
+      "epoch": 4.879545454545455,
+      "grad_norm": 0.7260614532561702,
+      "learning_rate": 1.7959132303550786e-09,
+      "loss": 0.0016,
+      "step": 21470
+    },
+    {
+      "epoch": 4.879772727272727,
+      "grad_norm": 2.281643667412178,
+      "learning_rate": 1.7891585575399717e-09,
+      "loss": 0.0098,
+      "step": 21471
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 0.7364226439103143,
+      "learning_rate": 1.782416593081246e-09,
+      "loss": 0.0027,
+      "step": 21472
+    },
+    {
+      "epoch": 4.880227272727272,
+      "grad_norm": 1.3008710370170051,
+      "learning_rate": 1.7756873371162915e-09,
+      "loss": 0.0029,
+      "step": 21473
+    },
+    {
+      "epoch": 4.880454545454546,
+      "grad_norm": 0.8591715907767645,
+      "learning_rate": 1.7689707897823594e-09,
+      "loss": 0.0031,
+      "step": 21474
+    },
+    {
+      "epoch": 4.880681818181818,
+      "grad_norm": 0.8994871135003054,
+      "learning_rate": 1.7622669512163542e-09,
+      "loss": 0.0055,
+      "step": 21475
+    },
+    {
+      "epoch": 4.880909090909091,
+      "grad_norm": 1.3444674382300619,
+      "learning_rate": 1.7555758215551112e-09,
+      "loss": 0.0089,
+      "step": 21476
+    },
+    {
+      "epoch": 4.881136363636363,
+      "grad_norm": 0.9339089590607687,
+      "learning_rate": 1.7488974009349097e-09,
+      "loss": 0.0024,
+      "step": 21477
+    },
+    {
+      "epoch": 4.881363636363636,
+      "grad_norm": 1.034223795902771,
+      "learning_rate": 1.74223168949203e-09,
+      "loss": 0.0035,
+      "step": 21478
+    },
+    {
+      "epoch": 4.881590909090909,
+      "grad_norm": 2.026301584109011,
+      "learning_rate": 1.7355786873623353e-09,
+      "loss": 0.0122,
+      "step": 21479
+    },
+    {
+      "epoch": 4.881818181818182,
+      "grad_norm": 0.3350930940926397,
+      "learning_rate": 1.7289383946816203e-09,
+      "loss": 0.0008,
+      "step": 21480
+    },
+    {
+      "epoch": 4.882045454545454,
+      "grad_norm": 0.4484612726444554,
+      "learning_rate": 1.7223108115851927e-09,
+      "loss": 0.0013,
+      "step": 21481
+    },
+    {
+      "epoch": 4.882272727272728,
+      "grad_norm": 2.130855223159859,
+      "learning_rate": 1.7156959382081534e-09,
+      "loss": 0.0081,
+      "step": 21482
+    },
+    {
+      "epoch": 4.8825,
+      "grad_norm": 0.7152657243868823,
+      "learning_rate": 1.7090937746854636e-09,
+      "loss": 0.0013,
+      "step": 21483
+    },
+    {
+      "epoch": 4.882727272727273,
+      "grad_norm": 0.6964505915928368,
+      "learning_rate": 1.702504321151807e-09,
+      "loss": 0.004,
+      "step": 21484
+    },
+    {
+      "epoch": 4.882954545454545,
+      "grad_norm": 0.727050816873663,
+      "learning_rate": 1.695927577741452e-09,
+      "loss": 0.0039,
+      "step": 21485
+    },
+    {
+      "epoch": 4.883181818181818,
+      "grad_norm": 0.19882060122532846,
+      "learning_rate": 1.6893635445885265e-09,
+      "loss": 0.0006,
+      "step": 21486
+    },
+    {
+      "epoch": 4.883409090909091,
+      "grad_norm": 2.388578057098591,
+      "learning_rate": 1.6828122218269515e-09,
+      "loss": 0.03,
+      "step": 21487
+    },
+    {
+      "epoch": 4.883636363636364,
+      "grad_norm": 0.057088192608467606,
+      "learning_rate": 1.6762736095902315e-09,
+      "loss": 0.0001,
+      "step": 21488
+    },
+    {
+      "epoch": 4.883863636363636,
+      "grad_norm": 0.15854303802464648,
+      "learning_rate": 1.6697477080117319e-09,
+      "loss": 0.0005,
+      "step": 21489
+    },
+    {
+      "epoch": 4.884090909090909,
+      "grad_norm": 1.1659095305327551,
+      "learning_rate": 1.66323451722461e-09,
+      "loss": 0.0028,
+      "step": 21490
+    },
+    {
+      "epoch": 4.884318181818182,
+      "grad_norm": 0.4447655653188191,
+      "learning_rate": 1.6567340373615376e-09,
+      "loss": 0.001,
+      "step": 21491
+    },
+    {
+      "epoch": 4.884545454545455,
+      "grad_norm": 1.0991035550892476,
+      "learning_rate": 1.6502462685551863e-09,
+      "loss": 0.006,
+      "step": 21492
+    },
+    {
+      "epoch": 4.884772727272727,
+      "grad_norm": 0.25428217188043206,
+      "learning_rate": 1.6437712109377418e-09,
+      "loss": 0.0008,
+      "step": 21493
+    },
+    {
+      "epoch": 4.885,
+      "grad_norm": 0.4285613328093244,
+      "learning_rate": 1.637308864641321e-09,
+      "loss": 0.0008,
+      "step": 21494
+    },
+    {
+      "epoch": 4.885227272727272,
+      "grad_norm": 6.326578846395397,
+      "learning_rate": 1.6308592297976933e-09,
+      "loss": 0.0104,
+      "step": 21495
+    },
+    {
+      "epoch": 4.885454545454546,
+      "grad_norm": 0.7356672205949645,
+      "learning_rate": 1.6244223065384202e-09,
+      "loss": 0.0034,
+      "step": 21496
+    },
+    {
+      "epoch": 4.885681818181818,
+      "grad_norm": 0.5918134486121417,
+      "learning_rate": 1.6179980949946466e-09,
+      "loss": 0.0021,
+      "step": 21497
+    },
+    {
+      "epoch": 4.885909090909091,
+      "grad_norm": 0.586712156604727,
+      "learning_rate": 1.6115865952975179e-09,
+      "loss": 0.0023,
+      "step": 21498
+    },
+    {
+      "epoch": 4.886136363636363,
+      "grad_norm": 0.7317763235385547,
+      "learning_rate": 1.6051878075776933e-09,
+      "loss": 0.0053,
+      "step": 21499
+    },
+    {
+      "epoch": 4.886363636363637,
+      "grad_norm": 1.135485680005004,
+      "learning_rate": 1.598801731965624e-09,
+      "loss": 0.0093,
+      "step": 21500
+    },
+    {
+      "epoch": 4.886590909090909,
+      "grad_norm": 1.6717899981444175,
+      "learning_rate": 1.5924283685915532e-09,
+      "loss": 0.0135,
+      "step": 21501
+    },
+    {
+      "epoch": 4.886818181818182,
+      "grad_norm": 0.740006063216855,
+      "learning_rate": 1.586067717585446e-09,
+      "loss": 0.0042,
+      "step": 21502
+    },
+    {
+      "epoch": 4.887045454545454,
+      "grad_norm": 0.7229837742204044,
+      "learning_rate": 1.5797197790771295e-09,
+      "loss": 0.0036,
+      "step": 21503
+    },
+    {
+      "epoch": 4.887272727272728,
+      "grad_norm": 1.1062075907591378,
+      "learning_rate": 1.5733845531958059e-09,
+      "loss": 0.0054,
+      "step": 21504
+    },
+    {
+      "epoch": 4.8875,
+      "grad_norm": 0.6744116572809152,
+      "learning_rate": 1.5670620400708852e-09,
+      "loss": 0.0058,
+      "step": 21505
+    },
+    {
+      "epoch": 4.887727272727273,
+      "grad_norm": 0.3931009154034634,
+      "learning_rate": 1.560752239831223e-09,
+      "loss": 0.0011,
+      "step": 21506
+    },
+    {
+      "epoch": 4.887954545454545,
+      "grad_norm": 1.6486125102930844,
+      "learning_rate": 1.554455152605397e-09,
+      "loss": 0.012,
+      "step": 21507
+    },
+    {
+      "epoch": 4.888181818181819,
+      "grad_norm": 1.2465437658327683,
+      "learning_rate": 1.5481707785219153e-09,
+      "loss": 0.0091,
+      "step": 21508
+    },
+    {
+      "epoch": 4.888409090909091,
+      "grad_norm": 2.075991497607531,
+      "learning_rate": 1.5418991177089393e-09,
+      "loss": 0.0099,
+      "step": 21509
+    },
+    {
+      "epoch": 4.888636363636364,
+      "grad_norm": 1.1107329905179928,
+      "learning_rate": 1.5356401702942836e-09,
+      "loss": 0.0035,
+      "step": 21510
+    },
+    {
+      "epoch": 4.888863636363636,
+      "grad_norm": 0.38674548643734563,
+      "learning_rate": 1.5293939364056238e-09,
+      "loss": 0.0021,
+      "step": 21511
+    },
+    {
+      "epoch": 4.889090909090909,
+      "grad_norm": 0.9271854451418498,
+      "learning_rate": 1.5231604161703578e-09,
+      "loss": 0.0047,
+      "step": 21512
+    },
+    {
+      "epoch": 4.889318181818182,
+      "grad_norm": 0.6486712321584946,
+      "learning_rate": 1.5169396097156065e-09,
+      "loss": 0.0034,
+      "step": 21513
+    },
+    {
+      "epoch": 4.889545454545455,
+      "grad_norm": 1.4356898934425923,
+      "learning_rate": 1.5107315171680737e-09,
+      "loss": 0.0065,
+      "step": 21514
+    },
+    {
+      "epoch": 4.889772727272727,
+      "grad_norm": 1.4209427341535812,
+      "learning_rate": 1.5045361386545333e-09,
+      "loss": 0.0085,
+      "step": 21515
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 2.968090778043487,
+      "learning_rate": 1.4983534743012036e-09,
+      "loss": 0.0131,
+      "step": 21516
+    },
+    {
+      "epoch": 4.890227272727273,
+      "grad_norm": 0.4979266181496989,
+      "learning_rate": 1.4921835242342336e-09,
+      "loss": 0.0028,
+      "step": 21517
+    },
+    {
+      "epoch": 4.890454545454546,
+      "grad_norm": 0.9720844930283951,
+      "learning_rate": 1.4860262885794258e-09,
+      "loss": 0.0039,
+      "step": 21518
+    },
+    {
+      "epoch": 4.890681818181818,
+      "grad_norm": 0.3446476435068884,
+      "learning_rate": 1.4798817674623045e-09,
+      "loss": 0.0018,
+      "step": 21519
+    },
+    {
+      "epoch": 4.890909090909091,
+      "grad_norm": 0.7666549543400665,
+      "learning_rate": 1.4737499610081862e-09,
+      "loss": 0.0025,
+      "step": 21520
+    },
+    {
+      "epoch": 4.891136363636363,
+      "grad_norm": 1.9667870220587444,
+      "learning_rate": 1.46763086934211e-09,
+      "loss": 0.0094,
+      "step": 21521
+    },
+    {
+      "epoch": 4.891363636363637,
+      "grad_norm": 1.0641641265832658,
+      "learning_rate": 1.4615244925888368e-09,
+      "loss": 0.0098,
+      "step": 21522
+    },
+    {
+      "epoch": 4.891590909090909,
+      "grad_norm": 0.4619323295277406,
+      "learning_rate": 1.4554308308729198e-09,
+      "loss": 0.0017,
+      "step": 21523
+    },
+    {
+      "epoch": 4.891818181818182,
+      "grad_norm": 0.28896154455495693,
+      "learning_rate": 1.4493498843185655e-09,
+      "loss": 0.0008,
+      "step": 21524
+    },
+    {
+      "epoch": 4.892045454545455,
+      "grad_norm": 0.6690334970259477,
+      "learning_rate": 1.443281653049841e-09,
+      "loss": 0.0021,
+      "step": 21525
+    },
+    {
+      "epoch": 4.8922727272727276,
+      "grad_norm": 1.2786130786419534,
+      "learning_rate": 1.4372261371904672e-09,
+      "loss": 0.0061,
+      "step": 21526
+    },
+    {
+      "epoch": 4.8925,
+      "grad_norm": 2.6462011952743425,
+      "learning_rate": 1.4311833368638866e-09,
+      "loss": 0.0129,
+      "step": 21527
+    },
+    {
+      "epoch": 4.892727272727273,
+      "grad_norm": 1.2812804479182383,
+      "learning_rate": 1.4251532521933343e-09,
+      "loss": 0.005,
+      "step": 21528
+    },
+    {
+      "epoch": 4.892954545454545,
+      "grad_norm": 0.8072137157946815,
+      "learning_rate": 1.4191358833019058e-09,
+      "loss": 0.0038,
+      "step": 21529
+    },
+    {
+      "epoch": 4.8931818181818185,
+      "grad_norm": 0.3264131657049567,
+      "learning_rate": 1.413131230312073e-09,
+      "loss": 0.0017,
+      "step": 21530
+    },
+    {
+      "epoch": 4.893409090909091,
+      "grad_norm": 0.7751557732961439,
+      "learning_rate": 1.4071392933464461e-09,
+      "loss": 0.0039,
+      "step": 21531
+    },
+    {
+      "epoch": 4.893636363636364,
+      "grad_norm": 1.7717212999636156,
+      "learning_rate": 1.4011600725271495e-09,
+      "loss": 0.0107,
+      "step": 21532
+    },
+    {
+      "epoch": 4.893863636363636,
+      "grad_norm": 0.7456700699322765,
+      "learning_rate": 1.395193567976169e-09,
+      "loss": 0.002,
+      "step": 21533
+    },
+    {
+      "epoch": 4.8940909090909095,
+      "grad_norm": 0.6003096911803082,
+      "learning_rate": 1.3892397798150742e-09,
+      "loss": 0.002,
+      "step": 21534
+    },
+    {
+      "epoch": 4.894318181818182,
+      "grad_norm": 3.2813255815545013,
+      "learning_rate": 1.383298708165365e-09,
+      "loss": 0.0065,
+      "step": 21535
+    },
+    {
+      "epoch": 4.8945454545454545,
+      "grad_norm": 0.9596889967063733,
+      "learning_rate": 1.377370353148194e-09,
+      "loss": 0.0056,
+      "step": 21536
+    },
+    {
+      "epoch": 4.894772727272727,
+      "grad_norm": 0.8783561168681518,
+      "learning_rate": 1.3714547148843683e-09,
+      "loss": 0.0044,
+      "step": 21537
+    },
+    {
+      "epoch": 4.895,
+      "grad_norm": 1.6664545815041194,
+      "learning_rate": 1.3655517934945544e-09,
+      "loss": 0.0051,
+      "step": 21538
+    },
+    {
+      "epoch": 4.895227272727273,
+      "grad_norm": 0.7597240105646822,
+      "learning_rate": 1.3596615890990732e-09,
+      "loss": 0.0046,
+      "step": 21539
+    },
+    {
+      "epoch": 4.8954545454545455,
+      "grad_norm": 1.8032023394918553,
+      "learning_rate": 1.3537841018181758e-09,
+      "loss": 0.0055,
+      "step": 21540
+    },
+    {
+      "epoch": 4.895681818181818,
+      "grad_norm": 1.4696083322051534,
+      "learning_rate": 1.347919331771627e-09,
+      "loss": 0.0049,
+      "step": 21541
+    },
+    {
+      "epoch": 4.8959090909090905,
+      "grad_norm": 0.9841691371525089,
+      "learning_rate": 1.3420672790789846e-09,
+      "loss": 0.0022,
+      "step": 21542
+    },
+    {
+      "epoch": 4.896136363636364,
+      "grad_norm": 0.9794938491856505,
+      "learning_rate": 1.3362279438596665e-09,
+      "loss": 0.0024,
+      "step": 21543
+    },
+    {
+      "epoch": 4.8963636363636365,
+      "grad_norm": 0.42956552649492646,
+      "learning_rate": 1.3304013262326748e-09,
+      "loss": 0.0018,
+      "step": 21544
+    },
+    {
+      "epoch": 4.896590909090909,
+      "grad_norm": 1.9219098788726607,
+      "learning_rate": 1.3245874263168036e-09,
+      "loss": 0.0151,
+      "step": 21545
+    },
+    {
+      "epoch": 4.8968181818181815,
+      "grad_norm": 2.7278816571424818,
+      "learning_rate": 1.3187862442307076e-09,
+      "loss": 0.0293,
+      "step": 21546
+    },
+    {
+      "epoch": 4.897045454545455,
+      "grad_norm": 0.3828748906932365,
+      "learning_rate": 1.3129977800926258e-09,
+      "loss": 0.0007,
+      "step": 21547
+    },
+    {
+      "epoch": 4.897272727272727,
+      "grad_norm": 0.6415332062901632,
+      "learning_rate": 1.307222034020658e-09,
+      "loss": 0.0031,
+      "step": 21548
+    },
+    {
+      "epoch": 4.8975,
+      "grad_norm": 6.590637485620236,
+      "learning_rate": 1.301459006132419e-09,
+      "loss": 0.0024,
+      "step": 21549
+    },
+    {
+      "epoch": 4.8977272727272725,
+      "grad_norm": 1.6031324323882894,
+      "learning_rate": 1.2957086965455917e-09,
+      "loss": 0.0046,
+      "step": 21550
+    },
+    {
+      "epoch": 4.897954545454546,
+      "grad_norm": 0.4046663591204857,
+      "learning_rate": 1.2899711053773744e-09,
+      "loss": 0.002,
+      "step": 21551
+    },
+    {
+      "epoch": 4.898181818181818,
+      "grad_norm": 1.5575833426209864,
+      "learning_rate": 1.2842462327447569e-09,
+      "loss": 0.0065,
+      "step": 21552
+    },
+    {
+      "epoch": 4.898409090909091,
+      "grad_norm": 1.769765540717639,
+      "learning_rate": 1.2785340787645207e-09,
+      "loss": 0.0111,
+      "step": 21553
+    },
+    {
+      "epoch": 4.8986363636363635,
+      "grad_norm": 1.4214074104166399,
+      "learning_rate": 1.2728346435530309e-09,
+      "loss": 0.0082,
+      "step": 21554
+    },
+    {
+      "epoch": 4.898863636363636,
+      "grad_norm": 0.8915157706605128,
+      "learning_rate": 1.2671479272266528e-09,
+      "loss": 0.0015,
+      "step": 21555
+    },
+    {
+      "epoch": 4.899090909090909,
+      "grad_norm": 0.22400023101957722,
+      "learning_rate": 1.2614739299013357e-09,
+      "loss": 0.0006,
+      "step": 21556
+    },
+    {
+      "epoch": 4.899318181818182,
+      "grad_norm": 0.9812942541241185,
+      "learning_rate": 1.255812651692681e-09,
+      "loss": 0.0055,
+      "step": 21557
+    },
+    {
+      "epoch": 4.899545454545454,
+      "grad_norm": 1.0902526604724851,
+      "learning_rate": 1.2501640927162218e-09,
+      "loss": 0.0049,
+      "step": 21558
+    },
+    {
+      "epoch": 4.899772727272727,
+      "grad_norm": 1.0169845417645074,
+      "learning_rate": 1.244528253087074e-09,
+      "loss": 0.0062,
+      "step": 21559
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 2.240207462143034,
+      "learning_rate": 1.238905132920215e-09,
+      "loss": 0.0068,
+      "step": 21560
+    },
+    {
+      "epoch": 4.900227272727273,
+      "grad_norm": 0.693059481272778,
+      "learning_rate": 1.2332947323302757e-09,
+      "loss": 0.0022,
+      "step": 21561
+    },
+    {
+      "epoch": 4.900454545454545,
+      "grad_norm": 0.6944401299200142,
+      "learning_rate": 1.2276970514316783e-09,
+      "loss": 0.0066,
+      "step": 21562
+    },
+    {
+      "epoch": 4.900681818181818,
+      "grad_norm": 0.5556186934041265,
+      "learning_rate": 1.2221120903385675e-09,
+      "loss": 0.0016,
+      "step": 21563
+    },
+    {
+      "epoch": 4.90090909090909,
+      "grad_norm": 1.2166268955439736,
+      "learning_rate": 1.21653984916488e-09,
+      "loss": 0.0028,
+      "step": 21564
+    },
+    {
+      "epoch": 4.901136363636364,
+      "grad_norm": 2.1271404155405564,
+      "learning_rate": 1.2109803280241361e-09,
+      "loss": 0.0069,
+      "step": 21565
+    },
+    {
+      "epoch": 4.901363636363636,
+      "grad_norm": 1.2312290421805303,
+      "learning_rate": 1.2054335270297868e-09,
+      "loss": 0.0053,
+      "step": 21566
+    },
+    {
+      "epoch": 4.901590909090909,
+      "grad_norm": 1.1079610883569686,
+      "learning_rate": 1.199899446294936e-09,
+      "loss": 0.0051,
+      "step": 21567
+    },
+    {
+      "epoch": 4.901818181818182,
+      "grad_norm": 0.8026744907851796,
+      "learning_rate": 1.1943780859323405e-09,
+      "loss": 0.0021,
+      "step": 21568
+    },
+    {
+      "epoch": 4.902045454545455,
+      "grad_norm": 2.3920539016465496,
+      "learning_rate": 1.1888694460547578e-09,
+      "loss": 0.0143,
+      "step": 21569
+    },
+    {
+      "epoch": 4.902272727272727,
+      "grad_norm": 1.3942038396704506,
+      "learning_rate": 1.1833735267743896e-09,
+      "loss": 0.0066,
+      "step": 21570
+    },
+    {
+      "epoch": 4.9025,
+      "grad_norm": 1.0897798705236386,
+      "learning_rate": 1.1778903282032987e-09,
+      "loss": 0.0012,
+      "step": 21571
+    },
+    {
+      "epoch": 4.902727272727272,
+      "grad_norm": 1.4131434629208433,
+      "learning_rate": 1.17241985045341e-09,
+      "loss": 0.0066,
+      "step": 21572
+    },
+    {
+      "epoch": 4.902954545454546,
+      "grad_norm": 0.3644246929323946,
+      "learning_rate": 1.166962093636162e-09,
+      "loss": 0.0018,
+      "step": 21573
+    },
+    {
+      "epoch": 4.903181818181818,
+      "grad_norm": 2.967078309314931,
+      "learning_rate": 1.1615170578629237e-09,
+      "loss": 0.0214,
+      "step": 21574
+    },
+    {
+      "epoch": 4.903409090909091,
+      "grad_norm": 0.5281085722724143,
+      "learning_rate": 1.1560847432446482e-09,
+      "loss": 0.002,
+      "step": 21575
+    },
+    {
+      "epoch": 4.903636363636363,
+      "grad_norm": 1.5106482310245422,
+      "learning_rate": 1.1506651498922188e-09,
+      "loss": 0.0159,
+      "step": 21576
+    },
+    {
+      "epoch": 4.903863636363637,
+      "grad_norm": 1.714799699988457,
+      "learning_rate": 1.1452582779160337e-09,
+      "loss": 0.0082,
+      "step": 21577
+    },
+    {
+      "epoch": 4.904090909090909,
+      "grad_norm": 1.9246233154682337,
+      "learning_rate": 1.1398641274264904e-09,
+      "loss": 0.0053,
+      "step": 21578
+    },
+    {
+      "epoch": 4.904318181818182,
+      "grad_norm": 1.3032207413331482,
+      "learning_rate": 1.1344826985334318e-09,
+      "loss": 0.0039,
+      "step": 21579
+    },
+    {
+      "epoch": 4.904545454545454,
+      "grad_norm": 0.8520202886388344,
+      "learning_rate": 1.1291139913467003e-09,
+      "loss": 0.0046,
+      "step": 21580
+    },
+    {
+      "epoch": 4.904772727272727,
+      "grad_norm": 0.743529112761142,
+      "learning_rate": 1.123758005975653e-09,
+      "loss": 0.0042,
+      "step": 21581
+    },
+    {
+      "epoch": 4.905,
+      "grad_norm": 1.382350516800048,
+      "learning_rate": 1.1184147425297165e-09,
+      "loss": 0.0037,
+      "step": 21582
+    },
+    {
+      "epoch": 4.905227272727273,
+      "grad_norm": 0.5087505393382871,
+      "learning_rate": 1.113084201117623e-09,
+      "loss": 0.0016,
+      "step": 21583
+    },
+    {
+      "epoch": 4.905454545454545,
+      "grad_norm": 0.39115888486417566,
+      "learning_rate": 1.1077663818482437e-09,
+      "loss": 0.0025,
+      "step": 21584
+    },
+    {
+      "epoch": 4.905681818181818,
+      "grad_norm": 1.2299566684621877,
+      "learning_rate": 1.1024612848298948e-09,
+      "loss": 0.0099,
+      "step": 21585
+    },
+    {
+      "epoch": 4.905909090909091,
+      "grad_norm": 1.3996324672011384,
+      "learning_rate": 1.0971689101708233e-09,
+      "loss": 0.0053,
+      "step": 21586
+    },
+    {
+      "epoch": 4.906136363636364,
+      "grad_norm": 1.1686034480909067,
+      "learning_rate": 1.091889257978998e-09,
+      "loss": 0.0075,
+      "step": 21587
+    },
+    {
+      "epoch": 4.906363636363636,
+      "grad_norm": 0.7056583123367088,
+      "learning_rate": 1.086622328361972e-09,
+      "loss": 0.0047,
+      "step": 21588
+    },
+    {
+      "epoch": 4.906590909090909,
+      "grad_norm": 0.36283948085135426,
+      "learning_rate": 1.0813681214271594e-09,
+      "loss": 0.0016,
+      "step": 21589
+    },
+    {
+      "epoch": 4.906818181818182,
+      "grad_norm": 0.7982006512826816,
+      "learning_rate": 1.0761266372817658e-09,
+      "loss": 0.0027,
+      "step": 21590
+    },
+    {
+      "epoch": 4.907045454545455,
+      "grad_norm": 0.3307278459980054,
+      "learning_rate": 1.0708978760326505e-09,
+      "loss": 0.0013,
+      "step": 21591
+    },
+    {
+      "epoch": 4.907272727272727,
+      "grad_norm": 1.0740714152490787,
+      "learning_rate": 1.0656818377863947e-09,
+      "loss": 0.0038,
+      "step": 21592
+    },
+    {
+      "epoch": 4.9075,
+      "grad_norm": 0.799602016141559,
+      "learning_rate": 1.0604785226494414e-09,
+      "loss": 0.0039,
+      "step": 21593
+    },
+    {
+      "epoch": 4.907727272727273,
+      "grad_norm": 0.6174154001946199,
+      "learning_rate": 1.0552879307278857e-09,
+      "loss": 0.002,
+      "step": 21594
+    },
+    {
+      "epoch": 4.907954545454546,
+      "grad_norm": 0.46154692685578264,
+      "learning_rate": 1.0501100621274768e-09,
+      "loss": 0.0016,
+      "step": 21595
+    },
+    {
+      "epoch": 4.908181818181818,
+      "grad_norm": 1.6116989502526067,
+      "learning_rate": 1.0449449169538939e-09,
+      "loss": 0.0092,
+      "step": 21596
+    },
+    {
+      "epoch": 4.908409090909091,
+      "grad_norm": 2.4852908582057465,
+      "learning_rate": 1.0397924953124e-09,
+      "loss": 0.0063,
+      "step": 21597
+    },
+    {
+      "epoch": 4.908636363636363,
+      "grad_norm": 0.3418652673128655,
+      "learning_rate": 1.0346527973081194e-09,
+      "loss": 0.0011,
+      "step": 21598
+    },
+    {
+      "epoch": 4.908863636363637,
+      "grad_norm": 1.9944727511722533,
+      "learning_rate": 1.0295258230458294e-09,
+      "loss": 0.008,
+      "step": 21599
+    },
+    {
+      "epoch": 4.909090909090909,
+      "grad_norm": 2.065147504393907,
+      "learning_rate": 1.024411572630099e-09,
+      "loss": 0.0038,
+      "step": 21600
+    },
+    {
+      "epoch": 4.909318181818182,
+      "grad_norm": 1.039395494757524,
+      "learning_rate": 1.01931004616522e-09,
+      "loss": 0.0057,
+      "step": 21601
+    },
+    {
+      "epoch": 4.909545454545454,
+      "grad_norm": 2.859865197706339,
+      "learning_rate": 1.014221243755137e-09,
+      "loss": 0.0254,
+      "step": 21602
+    },
+    {
+      "epoch": 4.909772727272728,
+      "grad_norm": 0.4838483869225091,
+      "learning_rate": 1.009145165503725e-09,
+      "loss": 0.0022,
+      "step": 21603
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 0.15524274297160814,
+      "learning_rate": 1.0040818115145124e-09,
+      "loss": 0.0005,
+      "step": 21604
+    },
+    {
+      "epoch": 4.910227272727273,
+      "grad_norm": 0.8998324247659335,
+      "learning_rate": 9.990311818906111e-10,
+      "loss": 0.0062,
+      "step": 21605
+    },
+    {
+      "epoch": 4.910454545454545,
+      "grad_norm": 1.8246279483570815,
+      "learning_rate": 9.939932767351336e-10,
+      "loss": 0.0087,
+      "step": 21606
+    },
+    {
+      "epoch": 4.910681818181818,
+      "grad_norm": 0.8595595574124267,
+      "learning_rate": 9.889680961507053e-10,
+      "loss": 0.0018,
+      "step": 21607
+    },
+    {
+      "epoch": 4.910909090909091,
+      "grad_norm": 0.6644605200782808,
+      "learning_rate": 9.83955640239953e-10,
+      "loss": 0.0038,
+      "step": 21608
+    },
+    {
+      "epoch": 4.911136363636364,
+      "grad_norm": 1.634783860296358,
+      "learning_rate": 9.789559091049479e-10,
+      "loss": 0.0107,
+      "step": 21609
+    },
+    {
+      "epoch": 4.911363636363636,
+      "grad_norm": 1.0919287652141612,
+      "learning_rate": 9.739689028476912e-10,
+      "loss": 0.0045,
+      "step": 21610
+    },
+    {
+      "epoch": 4.91159090909091,
+      "grad_norm": 1.2640930793710181,
+      "learning_rate": 9.689946215699076e-10,
+      "loss": 0.004,
+      "step": 21611
+    },
+    {
+      "epoch": 4.911818181818182,
+      "grad_norm": 1.1304770878066528,
+      "learning_rate": 9.640330653730435e-10,
+      "loss": 0.004,
+      "step": 21612
+    },
+    {
+      "epoch": 4.912045454545455,
+      "grad_norm": 1.6564802857270682,
+      "learning_rate": 9.59084234358129e-10,
+      "loss": 0.003,
+      "step": 21613
+    },
+    {
+      "epoch": 4.912272727272727,
+      "grad_norm": 0.4648928750434537,
+      "learning_rate": 9.541481286262643e-10,
+      "loss": 0.0015,
+      "step": 21614
+    },
+    {
+      "epoch": 4.9125,
+      "grad_norm": 3.2190607353009333,
+      "learning_rate": 9.492247482779936e-10,
+      "loss": 0.0064,
+      "step": 21615
+    },
+    {
+      "epoch": 4.912727272727273,
+      "grad_norm": 1.839477039680068,
+      "learning_rate": 9.443140934137924e-10,
+      "loss": 0.0106,
+      "step": 21616
+    },
+    {
+      "epoch": 4.912954545454546,
+      "grad_norm": 1.793492442241465,
+      "learning_rate": 9.39416164133719e-10,
+      "loss": 0.0088,
+      "step": 21617
+    },
+    {
+      "epoch": 4.913181818181818,
+      "grad_norm": 0.6813453802854537,
+      "learning_rate": 9.34530960537694e-10,
+      "loss": 0.0162,
+      "step": 21618
+    },
+    {
+      "epoch": 4.913409090909091,
+      "grad_norm": 0.7391977315232678,
+      "learning_rate": 9.296584827252903e-10,
+      "loss": 0.0013,
+      "step": 21619
+    },
+    {
+      "epoch": 4.913636363636364,
+      "grad_norm": 1.3493844809327986,
+      "learning_rate": 9.247987307959422e-10,
+      "loss": 0.0063,
+      "step": 21620
+    },
+    {
+      "epoch": 4.9138636363636365,
+      "grad_norm": 0.5609598259713651,
+      "learning_rate": 9.199517048486678e-10,
+      "loss": 0.0024,
+      "step": 21621
+    },
+    {
+      "epoch": 4.914090909090909,
+      "grad_norm": 0.45849839370867507,
+      "learning_rate": 9.151174049824158e-10,
+      "loss": 0.0021,
+      "step": 21622
+    },
+    {
+      "epoch": 4.914318181818182,
+      "grad_norm": 1.1308892494706757,
+      "learning_rate": 9.10295831295649e-10,
+      "loss": 0.0035,
+      "step": 21623
+    },
+    {
+      "epoch": 4.914545454545454,
+      "grad_norm": 0.9152501959122217,
+      "learning_rate": 9.054869838867608e-10,
+      "loss": 0.0052,
+      "step": 21624
+    },
+    {
+      "epoch": 4.9147727272727275,
+      "grad_norm": 0.5631935177797535,
+      "learning_rate": 9.006908628537287e-10,
+      "loss": 0.0023,
+      "step": 21625
+    },
+    {
+      "epoch": 4.915,
+      "grad_norm": 0.5896960692940932,
+      "learning_rate": 8.959074682944601e-10,
+      "loss": 0.0023,
+      "step": 21626
+    },
+    {
+      "epoch": 4.915227272727273,
+      "grad_norm": 0.7056587194873003,
+      "learning_rate": 8.911368003064468e-10,
+      "loss": 0.002,
+      "step": 21627
+    },
+    {
+      "epoch": 4.915454545454545,
+      "grad_norm": 1.940254515703756,
+      "learning_rate": 8.863788589869027e-10,
+      "loss": 0.0033,
+      "step": 21628
+    },
+    {
+      "epoch": 4.9156818181818185,
+      "grad_norm": 0.7510040933894904,
+      "learning_rate": 8.81633644432972e-10,
+      "loss": 0.0062,
+      "step": 21629
+    },
+    {
+      "epoch": 4.915909090909091,
+      "grad_norm": 0.18873144832553543,
+      "learning_rate": 8.769011567413832e-10,
+      "loss": 0.002,
+      "step": 21630
+    },
+    {
+      "epoch": 4.9161363636363635,
+      "grad_norm": 0.8069321339768415,
+      "learning_rate": 8.721813960085868e-10,
+      "loss": 0.0049,
+      "step": 21631
+    },
+    {
+      "epoch": 4.916363636363636,
+      "grad_norm": 2.8185974616442175,
+      "learning_rate": 8.674743623308946e-10,
+      "loss": 0.0128,
+      "step": 21632
+    },
+    {
+      "epoch": 4.9165909090909095,
+      "grad_norm": 2.7644443977598963,
+      "learning_rate": 8.627800558042715e-10,
+      "loss": 0.0329,
+      "step": 21633
+    },
+    {
+      "epoch": 4.916818181818182,
+      "grad_norm": 1.6318067174220083,
+      "learning_rate": 8.580984765244049e-10,
+      "loss": 0.0069,
+      "step": 21634
+    },
+    {
+      "epoch": 4.9170454545454545,
+      "grad_norm": 0.6708791153166452,
+      "learning_rate": 8.534296245867741e-10,
+      "loss": 0.0015,
+      "step": 21635
+    },
+    {
+      "epoch": 4.917272727272727,
+      "grad_norm": 1.2413158832363942,
+      "learning_rate": 8.487735000865804e-10,
+      "loss": 0.0061,
+      "step": 21636
+    },
+    {
+      "epoch": 4.9175,
+      "grad_norm": 0.5264936244039287,
+      "learning_rate": 8.441301031188176e-10,
+      "loss": 0.0007,
+      "step": 21637
+    },
+    {
+      "epoch": 4.917727272727273,
+      "grad_norm": 0.1903557920029646,
+      "learning_rate": 8.39499433778132e-10,
+      "loss": 0.0006,
+      "step": 21638
+    },
+    {
+      "epoch": 4.9179545454545455,
+      "grad_norm": 0.9116646174275432,
+      "learning_rate": 8.348814921589621e-10,
+      "loss": 0.0033,
+      "step": 21639
+    },
+    {
+      "epoch": 4.918181818181818,
+      "grad_norm": 0.37965248591191375,
+      "learning_rate": 8.302762783555379e-10,
+      "loss": 0.0013,
+      "step": 21640
+    },
+    {
+      "epoch": 4.9184090909090905,
+      "grad_norm": 1.8747341526777088,
+      "learning_rate": 8.25683792461604e-10,
+      "loss": 0.006,
+      "step": 21641
+    },
+    {
+      "epoch": 4.918636363636364,
+      "grad_norm": 0.6754090029853829,
+      "learning_rate": 8.211040345709742e-10,
+      "loss": 0.0054,
+      "step": 21642
+    },
+    {
+      "epoch": 4.918863636363636,
+      "grad_norm": 4.904296294032903,
+      "learning_rate": 8.165370047769073e-10,
+      "loss": 0.0116,
+      "step": 21643
+    },
+    {
+      "epoch": 4.919090909090909,
+      "grad_norm": 0.10454947409169107,
+      "learning_rate": 8.119827031726618e-10,
+      "loss": 0.0003,
+      "step": 21644
+    },
+    {
+      "epoch": 4.9193181818181815,
+      "grad_norm": 1.7427297606682148,
+      "learning_rate": 8.074411298510109e-10,
+      "loss": 0.0031,
+      "step": 21645
+    },
+    {
+      "epoch": 4.919545454545455,
+      "grad_norm": 1.9339998877206745,
+      "learning_rate": 8.02912284904589e-10,
+      "loss": 0.0132,
+      "step": 21646
+    },
+    {
+      "epoch": 4.919772727272727,
+      "grad_norm": 1.0926261365061263,
+      "learning_rate": 7.983961684258218e-10,
+      "loss": 0.0102,
+      "step": 21647
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 0.9332780755127442,
+      "learning_rate": 7.938927805066499e-10,
+      "loss": 0.0014,
+      "step": 21648
+    },
+    {
+      "epoch": 4.9202272727272724,
+      "grad_norm": 1.2051403425448206,
+      "learning_rate": 7.894021212390829e-10,
+      "loss": 0.0057,
+      "step": 21649
+    },
+    {
+      "epoch": 4.920454545454545,
+      "grad_norm": 0.5227529292534705,
+      "learning_rate": 7.849241907145061e-10,
+      "loss": 0.0033,
+      "step": 21650
+    },
+    {
+      "epoch": 4.920681818181818,
+      "grad_norm": 0.524168378685508,
+      "learning_rate": 7.804589890243741e-10,
+      "loss": 0.0011,
+      "step": 21651
+    },
+    {
+      "epoch": 4.920909090909091,
+      "grad_norm": 0.8954190494429527,
+      "learning_rate": 7.760065162597252e-10,
+      "loss": 0.0023,
+      "step": 21652
+    },
+    {
+      "epoch": 4.921136363636363,
+      "grad_norm": 0.8754883642373098,
+      "learning_rate": 7.715667725113202e-10,
+      "loss": 0.0055,
+      "step": 21653
+    },
+    {
+      "epoch": 4.921363636363637,
+      "grad_norm": 2.3047349673543476,
+      "learning_rate": 7.671397578696421e-10,
+      "loss": 0.0228,
+      "step": 21654
+    },
+    {
+      "epoch": 4.921590909090909,
+      "grad_norm": 0.4310362742791826,
+      "learning_rate": 7.627254724250355e-10,
+      "loss": 0.0018,
+      "step": 21655
+    },
+    {
+      "epoch": 4.921818181818182,
+      "grad_norm": 1.278081088969159,
+      "learning_rate": 7.583239162674977e-10,
+      "loss": 0.0045,
+      "step": 21656
+    },
+    {
+      "epoch": 4.922045454545454,
+      "grad_norm": 1.6056455040163053,
+      "learning_rate": 7.53935089486818e-10,
+      "loss": 0.0082,
+      "step": 21657
+    },
+    {
+      "epoch": 4.922272727272727,
+      "grad_norm": 0.48142553185437126,
+      "learning_rate": 7.495589921724389e-10,
+      "loss": 0.0018,
+      "step": 21658
+    },
+    {
+      "epoch": 4.9225,
+      "grad_norm": 1.6324934546963856,
+      "learning_rate": 7.451956244135944e-10,
+      "loss": 0.0072,
+      "step": 21659
+    },
+    {
+      "epoch": 4.922727272727273,
+      "grad_norm": 0.7865720248471949,
+      "learning_rate": 7.408449862993106e-10,
+      "loss": 0.0025,
+      "step": 21660
+    },
+    {
+      "epoch": 4.922954545454545,
+      "grad_norm": 0.4248837101327759,
+      "learning_rate": 7.365070779182665e-10,
+      "loss": 0.0014,
+      "step": 21661
+    },
+    {
+      "epoch": 4.923181818181818,
+      "grad_norm": 0.23687558300269582,
+      "learning_rate": 7.32181899358933e-10,
+      "loss": 0.0005,
+      "step": 21662
+    },
+    {
+      "epoch": 4.923409090909091,
+      "grad_norm": 0.8579270071459284,
+      "learning_rate": 7.278694507095035e-10,
+      "loss": 0.0019,
+      "step": 21663
+    },
+    {
+      "epoch": 4.923636363636364,
+      "grad_norm": 1.9644057442468774,
+      "learning_rate": 7.235697320578938e-10,
+      "loss": 0.0169,
+      "step": 21664
+    },
+    {
+      "epoch": 4.923863636363636,
+      "grad_norm": 1.7691624564063368,
+      "learning_rate": 7.192827434918808e-10,
+      "loss": 0.0061,
+      "step": 21665
+    },
+    {
+      "epoch": 4.924090909090909,
+      "grad_norm": 0.7554220667833854,
+      "learning_rate": 7.15008485098756e-10,
+      "loss": 0.0022,
+      "step": 21666
+    },
+    {
+      "epoch": 4.924318181818181,
+      "grad_norm": 0.6553477272215413,
+      "learning_rate": 7.10746956965741e-10,
+      "loss": 0.002,
+      "step": 21667
+    },
+    {
+      "epoch": 4.924545454545455,
+      "grad_norm": 0.3302674401884509,
+      "learning_rate": 7.06498159179711e-10,
+      "loss": 0.0012,
+      "step": 21668
+    },
+    {
+      "epoch": 4.924772727272727,
+      "grad_norm": 0.4747537399010183,
+      "learning_rate": 7.022620918273327e-10,
+      "loss": 0.0019,
+      "step": 21669
+    },
+    {
+      "epoch": 4.925,
+      "grad_norm": 0.4948255941155982,
+      "learning_rate": 6.980387549949952e-10,
+      "loss": 0.0023,
+      "step": 21670
+    },
+    {
+      "epoch": 4.925227272727272,
+      "grad_norm": 1.0197763332188314,
+      "learning_rate": 6.938281487688103e-10,
+      "loss": 0.0099,
+      "step": 21671
+    },
+    {
+      "epoch": 4.925454545454546,
+      "grad_norm": 0.8325955155438898,
+      "learning_rate": 6.896302732345427e-10,
+      "loss": 0.0036,
+      "step": 21672
+    },
+    {
+      "epoch": 4.925681818181818,
+      "grad_norm": 0.8197130194727335,
+      "learning_rate": 6.854451284779573e-10,
+      "loss": 0.0036,
+      "step": 21673
+    },
+    {
+      "epoch": 4.925909090909091,
+      "grad_norm": 0.9380921980137082,
+      "learning_rate": 6.812727145843328e-10,
+      "loss": 0.0023,
+      "step": 21674
+    },
+    {
+      "epoch": 4.926136363636363,
+      "grad_norm": 0.5300325760800639,
+      "learning_rate": 6.771130316387403e-10,
+      "loss": 0.0018,
+      "step": 21675
+    },
+    {
+      "epoch": 4.926363636363637,
+      "grad_norm": 2.878312943957312,
+      "learning_rate": 6.72966079725973e-10,
+      "loss": 0.0126,
+      "step": 21676
+    },
+    {
+      "epoch": 4.926590909090909,
+      "grad_norm": 0.6816814813179293,
+      "learning_rate": 6.688318589306159e-10,
+      "loss": 0.0016,
+      "step": 21677
+    },
+    {
+      "epoch": 4.926818181818182,
+      "grad_norm": 0.11464630610636162,
+      "learning_rate": 6.64710369337046e-10,
+      "loss": 0.0003,
+      "step": 21678
+    },
+    {
+      "epoch": 4.927045454545454,
+      "grad_norm": 2.290145672920192,
+      "learning_rate": 6.60601611029224e-10,
+      "loss": 0.0057,
+      "step": 21679
+    },
+    {
+      "epoch": 4.927272727272728,
+      "grad_norm": 0.7756345401646391,
+      "learning_rate": 6.565055840909024e-10,
+      "loss": 0.0031,
+      "step": 21680
+    },
+    {
+      "epoch": 4.9275,
+      "grad_norm": 2.1719517358132387,
+      "learning_rate": 6.524222886056947e-10,
+      "loss": 0.011,
+      "step": 21681
+    },
+    {
+      "epoch": 4.927727272727273,
+      "grad_norm": 1.249915172712491,
+      "learning_rate": 6.483517246567983e-10,
+      "loss": 0.0059,
+      "step": 21682
+    },
+    {
+      "epoch": 4.927954545454545,
+      "grad_norm": 1.3240384048712444,
+      "learning_rate": 6.442938923273412e-10,
+      "loss": 0.0012,
+      "step": 21683
+    },
+    {
+      "epoch": 4.928181818181818,
+      "grad_norm": 0.2309170887562606,
+      "learning_rate": 6.402487916998962e-10,
+      "loss": 0.0009,
+      "step": 21684
+    },
+    {
+      "epoch": 4.928409090909091,
+      "grad_norm": 1.2938530186154569,
+      "learning_rate": 6.362164228570361e-10,
+      "loss": 0.0046,
+      "step": 21685
+    },
+    {
+      "epoch": 4.928636363636364,
+      "grad_norm": 0.5260498199515805,
+      "learning_rate": 6.321967858810562e-10,
+      "loss": 0.0025,
+      "step": 21686
+    },
+    {
+      "epoch": 4.928863636363636,
+      "grad_norm": 1.6212031116960317,
+      "learning_rate": 6.281898808537662e-10,
+      "loss": 0.0056,
+      "step": 21687
+    },
+    {
+      "epoch": 4.929090909090909,
+      "grad_norm": 2.6917341850650036,
+      "learning_rate": 6.241957078570449e-10,
+      "loss": 0.0337,
+      "step": 21688
+    },
+    {
+      "epoch": 4.929318181818182,
+      "grad_norm": 1.6490524102501352,
+      "learning_rate": 6.202142669722161e-10,
+      "loss": 0.008,
+      "step": 21689
+    },
+    {
+      "epoch": 4.929545454545455,
+      "grad_norm": 1.7911788322473563,
+      "learning_rate": 6.16245558280465e-10,
+      "loss": 0.0068,
+      "step": 21690
+    },
+    {
+      "epoch": 4.929772727272727,
+      "grad_norm": 1.3846388455460128,
+      "learning_rate": 6.122895818627683e-10,
+      "loss": 0.0065,
+      "step": 21691
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 1.5845687632580743,
+      "learning_rate": 6.083463377998949e-10,
+      "loss": 0.005,
+      "step": 21692
+    },
+    {
+      "epoch": 4.930227272727272,
+      "grad_norm": 3.2272570779114993,
+      "learning_rate": 6.044158261720585e-10,
+      "loss": 0.0153,
+      "step": 21693
+    },
+    {
+      "epoch": 4.930454545454546,
+      "grad_norm": 0.5343938138988811,
+      "learning_rate": 6.004980470595417e-10,
+      "loss": 0.0013,
+      "step": 21694
+    },
+    {
+      "epoch": 4.930681818181818,
+      "grad_norm": 1.3878783849754526,
+      "learning_rate": 5.965930005421422e-10,
+      "loss": 0.0043,
+      "step": 21695
+    },
+    {
+      "epoch": 4.930909090909091,
+      "grad_norm": 0.7759785508215206,
+      "learning_rate": 5.927006866995877e-10,
+      "loss": 0.0024,
+      "step": 21696
+    },
+    {
+      "epoch": 4.931136363636364,
+      "grad_norm": 1.38339106849964,
+      "learning_rate": 5.888211056111898e-10,
+      "loss": 0.008,
+      "step": 21697
+    },
+    {
+      "epoch": 4.931363636363637,
+      "grad_norm": 0.807944384640507,
+      "learning_rate": 5.849542573560518e-10,
+      "loss": 0.0035,
+      "step": 21698
+    },
+    {
+      "epoch": 4.931590909090909,
+      "grad_norm": 0.6784865421117199,
+      "learning_rate": 5.811001420130691e-10,
+      "loss": 0.0023,
+      "step": 21699
+    },
+    {
+      "epoch": 4.931818181818182,
+      "grad_norm": 2.54669776415481,
+      "learning_rate": 5.772587596608592e-10,
+      "loss": 0.0048,
+      "step": 21700
+    },
+    {
+      "epoch": 4.932045454545454,
+      "grad_norm": 0.37371312376216476,
+      "learning_rate": 5.734301103776929e-10,
+      "loss": 0.0016,
+      "step": 21701
+    },
+    {
+      "epoch": 4.932272727272728,
+      "grad_norm": 0.6793810804471192,
+      "learning_rate": 5.696141942416328e-10,
+      "loss": 0.005,
+      "step": 21702
+    },
+    {
+      "epoch": 4.9325,
+      "grad_norm": 1.8205225805080543,
+      "learning_rate": 5.658110113304638e-10,
+      "loss": 0.0254,
+      "step": 21703
+    },
+    {
+      "epoch": 4.932727272727273,
+      "grad_norm": 0.9708256860506994,
+      "learning_rate": 5.620205617219016e-10,
+      "loss": 0.0016,
+      "step": 21704
+    },
+    {
+      "epoch": 4.932954545454545,
+      "grad_norm": 0.44936874282307354,
+      "learning_rate": 5.582428454930372e-10,
+      "loss": 0.0011,
+      "step": 21705
+    },
+    {
+      "epoch": 4.933181818181819,
+      "grad_norm": 0.9599896881451281,
+      "learning_rate": 5.544778627210312e-10,
+      "loss": 0.0026,
+      "step": 21706
+    },
+    {
+      "epoch": 4.933409090909091,
+      "grad_norm": 2.0297955172614954,
+      "learning_rate": 5.507256134826278e-10,
+      "loss": 0.0097,
+      "step": 21707
+    },
+    {
+      "epoch": 4.933636363636364,
+      "grad_norm": 0.807969376690933,
+      "learning_rate": 5.469860978543629e-10,
+      "loss": 0.002,
+      "step": 21708
+    },
+    {
+      "epoch": 4.933863636363636,
+      "grad_norm": 2.5609463156102974,
+      "learning_rate": 5.432593159124949e-10,
+      "loss": 0.0231,
+      "step": 21709
+    },
+    {
+      "epoch": 4.934090909090909,
+      "grad_norm": 0.29927105569922285,
+      "learning_rate": 5.395452677329354e-10,
+      "loss": 0.0008,
+      "step": 21710
+    },
+    {
+      "epoch": 4.934318181818182,
+      "grad_norm": 0.462756133667458,
+      "learning_rate": 5.358439533915266e-10,
+      "loss": 0.0017,
+      "step": 21711
+    },
+    {
+      "epoch": 4.934545454545455,
+      "grad_norm": 2.207968953846225,
+      "learning_rate": 5.321553729636248e-10,
+      "loss": 0.0118,
+      "step": 21712
+    },
+    {
+      "epoch": 4.934772727272727,
+      "grad_norm": 0.17654359420785906,
+      "learning_rate": 5.284795265245863e-10,
+      "loss": 0.0006,
+      "step": 21713
+    },
+    {
+      "epoch": 4.9350000000000005,
+      "grad_norm": 3.622937559492159,
+      "learning_rate": 5.248164141492818e-10,
+      "loss": 0.0033,
+      "step": 21714
+    },
+    {
+      "epoch": 4.935227272727273,
+      "grad_norm": 1.0779410425212803,
+      "learning_rate": 5.211660359124432e-10,
+      "loss": 0.0101,
+      "step": 21715
+    },
+    {
+      "epoch": 4.9354545454545455,
+      "grad_norm": 0.3791404299221079,
+      "learning_rate": 5.175283918884555e-10,
+      "loss": 0.001,
+      "step": 21716
+    },
+    {
+      "epoch": 4.935681818181818,
+      "grad_norm": 0.5594967153181855,
+      "learning_rate": 5.139034821515649e-10,
+      "loss": 0.002,
+      "step": 21717
+    },
+    {
+      "epoch": 4.935909090909091,
+      "grad_norm": 1.5408077710822736,
+      "learning_rate": 5.102913067756704e-10,
+      "loss": 0.0096,
+      "step": 21718
+    },
+    {
+      "epoch": 4.936136363636364,
+      "grad_norm": 0.27600527789450485,
+      "learning_rate": 5.066918658343245e-10,
+      "loss": 0.0009,
+      "step": 21719
+    },
+    {
+      "epoch": 4.9363636363636365,
+      "grad_norm": 0.7265973930328169,
+      "learning_rate": 5.031051594011488e-10,
+      "loss": 0.0025,
+      "step": 21720
+    },
+    {
+      "epoch": 4.936590909090909,
+      "grad_norm": 0.7244956464047348,
+      "learning_rate": 4.99531187549071e-10,
+      "loss": 0.0037,
+      "step": 21721
+    },
+    {
+      "epoch": 4.9368181818181816,
+      "grad_norm": 0.7127966678110425,
+      "learning_rate": 4.95969950351019e-10,
+      "loss": 0.003,
+      "step": 21722
+    },
+    {
+      "epoch": 4.937045454545455,
+      "grad_norm": 1.898524727728228,
+      "learning_rate": 4.924214478797123e-10,
+      "loss": 0.0158,
+      "step": 21723
+    },
+    {
+      "epoch": 4.9372727272727275,
+      "grad_norm": 1.512459902640325,
+      "learning_rate": 4.888856802074544e-10,
+      "loss": 0.0023,
+      "step": 21724
+    },
+    {
+      "epoch": 4.9375,
+      "grad_norm": 0.9930051418542704,
+      "learning_rate": 4.8536264740634e-10,
+      "loss": 0.0045,
+      "step": 21725
+    },
+    {
+      "epoch": 4.9377272727272725,
+      "grad_norm": 0.6533065961202341,
+      "learning_rate": 4.818523495481869e-10,
+      "loss": 0.0015,
+      "step": 21726
+    },
+    {
+      "epoch": 4.937954545454545,
+      "grad_norm": 0.5025508935714601,
+      "learning_rate": 4.783547867046045e-10,
+      "loss": 0.0018,
+      "step": 21727
+    },
+    {
+      "epoch": 4.9381818181818184,
+      "grad_norm": 2.590258672728226,
+      "learning_rate": 4.748699589468553e-10,
+      "loss": 0.0103,
+      "step": 21728
+    },
+    {
+      "epoch": 4.938409090909091,
+      "grad_norm": 0.4887599096736404,
+      "learning_rate": 4.713978663461322e-10,
+      "loss": 0.0025,
+      "step": 21729
+    },
+    {
+      "epoch": 4.9386363636363635,
+      "grad_norm": 2.5110299327594268,
+      "learning_rate": 4.679385089730732e-10,
+      "loss": 0.0144,
+      "step": 21730
+    },
+    {
+      "epoch": 4.938863636363636,
+      "grad_norm": 2.1941239088846767,
+      "learning_rate": 4.6449188689838554e-10,
+      "loss": 0.0095,
+      "step": 21731
+    },
+    {
+      "epoch": 4.939090909090909,
+      "grad_norm": 1.3758723748789103,
+      "learning_rate": 4.6105800019222157e-10,
+      "loss": 0.0116,
+      "step": 21732
+    },
+    {
+      "epoch": 4.939318181818182,
+      "grad_norm": 0.3237493256578243,
+      "learning_rate": 4.576368489246641e-10,
+      "loss": 0.0013,
+      "step": 21733
+    },
+    {
+      "epoch": 4.9395454545454545,
+      "grad_norm": 1.4588053091301025,
+      "learning_rate": 4.5422843316544896e-10,
+      "loss": 0.0069,
+      "step": 21734
+    },
+    {
+      "epoch": 4.939772727272727,
+      "grad_norm": 0.7779643673237319,
+      "learning_rate": 4.508327529841039e-10,
+      "loss": 0.0051,
+      "step": 21735
+    },
+    {
+      "epoch": 4.9399999999999995,
+      "grad_norm": 1.036086766511288,
+      "learning_rate": 4.4744980844987906e-10,
+      "loss": 0.0026,
+      "step": 21736
+    },
+    {
+      "epoch": 4.940227272727273,
+      "grad_norm": 0.8572363817764304,
+      "learning_rate": 4.4407959963174706e-10,
+      "loss": 0.0033,
+      "step": 21737
+    },
+    {
+      "epoch": 4.940454545454545,
+      "grad_norm": 0.14996335106471947,
+      "learning_rate": 4.40722126598403e-10,
+      "loss": 0.0003,
+      "step": 21738
+    },
+    {
+      "epoch": 4.940681818181818,
+      "grad_norm": 0.437178847620183,
+      "learning_rate": 4.373773894183336e-10,
+      "loss": 0.0011,
+      "step": 21739
+    },
+    {
+      "epoch": 4.940909090909091,
+      "grad_norm": 1.343323831172476,
+      "learning_rate": 4.340453881597484e-10,
+      "loss": 0.0028,
+      "step": 21740
+    },
+    {
+      "epoch": 4.941136363636364,
+      "grad_norm": 0.834138217607777,
+      "learning_rate": 4.307261228906484e-10,
+      "loss": 0.0061,
+      "step": 21741
+    },
+    {
+      "epoch": 4.941363636363636,
+      "grad_norm": 0.46892024073031563,
+      "learning_rate": 4.2741959367861854e-10,
+      "loss": 0.0028,
+      "step": 21742
+    },
+    {
+      "epoch": 4.941590909090909,
+      "grad_norm": 0.7026823691962334,
+      "learning_rate": 4.2412580059117416e-10,
+      "loss": 0.0045,
+      "step": 21743
+    },
+    {
+      "epoch": 4.941818181818181,
+      "grad_norm": 1.2756068692236993,
+      "learning_rate": 4.20844743695345e-10,
+      "loss": 0.0046,
+      "step": 21744
+    },
+    {
+      "epoch": 4.942045454545455,
+      "grad_norm": 0.5124614083769946,
+      "learning_rate": 4.175764230582302e-10,
+      "loss": 0.0031,
+      "step": 21745
+    },
+    {
+      "epoch": 4.942272727272727,
+      "grad_norm": 0.8634154085676605,
+      "learning_rate": 4.143208387463044e-10,
+      "loss": 0.0076,
+      "step": 21746
+    },
+    {
+      "epoch": 4.9425,
+      "grad_norm": 0.24709798208636874,
+      "learning_rate": 4.1107799082604203e-10,
+      "loss": 0.0005,
+      "step": 21747
+    },
+    {
+      "epoch": 4.942727272727272,
+      "grad_norm": 0.8085576808575883,
+      "learning_rate": 4.0784787936357097e-10,
+      "loss": 0.0021,
+      "step": 21748
+    },
+    {
+      "epoch": 4.942954545454546,
+      "grad_norm": 0.736848180984571,
+      "learning_rate": 4.046305044247412e-10,
+      "loss": 0.0032,
+      "step": 21749
+    },
+    {
+      "epoch": 4.943181818181818,
+      "grad_norm": 0.6516876638527895,
+      "learning_rate": 4.014258660751946e-10,
+      "loss": 0.0033,
+      "step": 21750
+    },
+    {
+      "epoch": 4.943409090909091,
+      "grad_norm": 2.2162115456659635,
+      "learning_rate": 3.982339643801569e-10,
+      "loss": 0.0079,
+      "step": 21751
+    },
+    {
+      "epoch": 4.943636363636363,
+      "grad_norm": 2.4286276662461703,
+      "learning_rate": 3.950547994049231e-10,
+      "loss": 0.014,
+      "step": 21752
+    },
+    {
+      "epoch": 4.943863636363636,
+      "grad_norm": 1.189267543022769,
+      "learning_rate": 3.918883712140942e-10,
+      "loss": 0.0064,
+      "step": 21753
+    },
+    {
+      "epoch": 4.944090909090909,
+      "grad_norm": 0.646845772155418,
+      "learning_rate": 3.887346798724101e-10,
+      "loss": 0.0031,
+      "step": 21754
+    },
+    {
+      "epoch": 4.944318181818182,
+      "grad_norm": 0.3589853844914332,
+      "learning_rate": 3.85593725444125e-10,
+      "loss": 0.0013,
+      "step": 21755
+    },
+    {
+      "epoch": 4.944545454545454,
+      "grad_norm": 0.8641449032250371,
+      "learning_rate": 3.8246550799328483e-10,
+      "loss": 0.0022,
+      "step": 21756
+    },
+    {
+      "epoch": 4.944772727272728,
+      "grad_norm": 0.7019620988508698,
+      "learning_rate": 3.793500275836581e-10,
+      "loss": 0.0048,
+      "step": 21757
+    },
+    {
+      "epoch": 4.945,
+      "grad_norm": 1.5518594123589435,
+      "learning_rate": 3.762472842788051e-10,
+      "loss": 0.0084,
+      "step": 21758
+    },
+    {
+      "epoch": 4.945227272727273,
+      "grad_norm": 0.6047957831129473,
+      "learning_rate": 3.7315727814193905e-10,
+      "loss": 0.0044,
+      "step": 21759
+    },
+    {
+      "epoch": 4.945454545454545,
+      "grad_norm": 2.5199486521167973,
+      "learning_rate": 3.700800092361345e-10,
+      "loss": 0.0166,
+      "step": 21760
+    },
+    {
+      "epoch": 4.945681818181818,
+      "grad_norm": 0.8564049663939604,
+      "learning_rate": 3.6701547762418856e-10,
+      "loss": 0.0107,
+      "step": 21761
+    },
+    {
+      "epoch": 4.945909090909091,
+      "grad_norm": 1.97093500131234,
+      "learning_rate": 3.6396368336848176e-10,
+      "loss": 0.0065,
+      "step": 21762
+    },
+    {
+      "epoch": 4.946136363636364,
+      "grad_norm": 1.31735223921282,
+      "learning_rate": 3.609246265313254e-10,
+      "loss": 0.0022,
+      "step": 21763
+    },
+    {
+      "epoch": 4.946363636363636,
+      "grad_norm": 0.7581467780512511,
+      "learning_rate": 3.578983071746145e-10,
+      "loss": 0.001,
+      "step": 21764
+    },
+    {
+      "epoch": 4.946590909090909,
+      "grad_norm": 1.000323164100525,
+      "learning_rate": 3.548847253601051e-10,
+      "loss": 0.0065,
+      "step": 21765
+    },
+    {
+      "epoch": 4.946818181818182,
+      "grad_norm": 0.6547681357159573,
+      "learning_rate": 3.5188388114927585e-10,
+      "loss": 0.0027,
+      "step": 21766
+    },
+    {
+      "epoch": 4.947045454545455,
+      "grad_norm": 0.13808415412347988,
+      "learning_rate": 3.488957746032584e-10,
+      "loss": 0.0003,
+      "step": 21767
+    },
+    {
+      "epoch": 4.947272727272727,
+      "grad_norm": 0.5040661589133586,
+      "learning_rate": 3.4592040578304564e-10,
+      "loss": 0.0017,
+      "step": 21768
+    },
+    {
+      "epoch": 4.9475,
+      "grad_norm": 0.7351228193889002,
+      "learning_rate": 3.429577747492835e-10,
+      "loss": 0.0038,
+      "step": 21769
+    },
+    {
+      "epoch": 4.947727272727272,
+      "grad_norm": 0.6750825163761375,
+      "learning_rate": 3.400078815624097e-10,
+      "loss": 0.0032,
+      "step": 21770
+    },
+    {
+      "epoch": 4.947954545454546,
+      "grad_norm": 0.7238622972692462,
+      "learning_rate": 3.3707072628251514e-10,
+      "loss": 0.0074,
+      "step": 21771
+    },
+    {
+      "epoch": 4.948181818181818,
+      "grad_norm": 0.8051576662468826,
+      "learning_rate": 3.341463089695518e-10,
+      "loss": 0.0076,
+      "step": 21772
+    },
+    {
+      "epoch": 4.948409090909091,
+      "grad_norm": 2.1127874010407726,
+      "learning_rate": 3.312346296831248e-10,
+      "loss": 0.0101,
+      "step": 21773
+    },
+    {
+      "epoch": 4.948636363636363,
+      "grad_norm": 0.35814207444310053,
+      "learning_rate": 3.283356884826311e-10,
+      "loss": 0.0008,
+      "step": 21774
+    },
+    {
+      "epoch": 4.948863636363637,
+      "grad_norm": 0.8131928565467004,
+      "learning_rate": 3.2544948542719e-10,
+      "loss": 0.0041,
+      "step": 21775
+    },
+    {
+      "epoch": 4.949090909090909,
+      "grad_norm": 0.8838658081261603,
+      "learning_rate": 3.22576020575574e-10,
+      "loss": 0.0024,
+      "step": 21776
+    },
+    {
+      "epoch": 4.949318181818182,
+      "grad_norm": 0.15522670577621953,
+      "learning_rate": 3.1971529398648614e-10,
+      "loss": 0.0005,
+      "step": 21777
+    },
+    {
+      "epoch": 4.949545454545454,
+      "grad_norm": 0.7385208047748344,
+      "learning_rate": 3.168673057182131e-10,
+      "loss": 0.0019,
+      "step": 21778
+    },
+    {
+      "epoch": 4.949772727272728,
+      "grad_norm": 0.5586558580761917,
+      "learning_rate": 3.1403205582883343e-10,
+      "loss": 0.0022,
+      "step": 21779
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 0.630346002210814,
+      "learning_rate": 3.1120954437614813e-10,
+      "loss": 0.0024,
+      "step": 21780
+    },
+    {
+      "epoch": 4.950227272727273,
+      "grad_norm": 0.4389058944036576,
+      "learning_rate": 3.0839977141775004e-10,
+      "loss": 0.0009,
+      "step": 21781
+    },
+    {
+      "epoch": 4.950454545454545,
+      "grad_norm": 3.2587359685218393,
+      "learning_rate": 3.0560273701095433e-10,
+      "loss": 0.0172,
+      "step": 21782
+    },
+    {
+      "epoch": 4.950681818181819,
+      "grad_norm": 2.2582815059043657,
+      "learning_rate": 3.028184412127294e-10,
+      "loss": 0.0085,
+      "step": 21783
+    },
+    {
+      "epoch": 4.950909090909091,
+      "grad_norm": 1.159836540133609,
+      "learning_rate": 3.000468840798354e-10,
+      "loss": 0.0038,
+      "step": 21784
+    },
+    {
+      "epoch": 4.951136363636364,
+      "grad_norm": 0.9375822148318017,
+      "learning_rate": 2.972880656688937e-10,
+      "loss": 0.0052,
+      "step": 21785
+    },
+    {
+      "epoch": 4.951363636363636,
+      "grad_norm": 0.6176900712253548,
+      "learning_rate": 2.945419860360399e-10,
+      "loss": 0.0011,
+      "step": 21786
+    },
+    {
+      "epoch": 4.951590909090909,
+      "grad_norm": 0.592383957886272,
+      "learning_rate": 2.918086452374097e-10,
+      "loss": 0.0019,
+      "step": 21787
+    },
+    {
+      "epoch": 4.951818181818182,
+      "grad_norm": 0.37493016717985456,
+      "learning_rate": 2.8908804332858366e-10,
+      "loss": 0.0013,
+      "step": 21788
+    },
+    {
+      "epoch": 4.952045454545455,
+      "grad_norm": 0.8326741032440841,
+      "learning_rate": 2.8638018036521164e-10,
+      "loss": 0.0042,
+      "step": 21789
+    },
+    {
+      "epoch": 4.952272727272727,
+      "grad_norm": 0.7996220571144561,
+      "learning_rate": 2.836850564023191e-10,
+      "loss": 0.002,
+      "step": 21790
+    },
+    {
+      "epoch": 4.9525,
+      "grad_norm": 2.3078058644268484,
+      "learning_rate": 2.810026714950009e-10,
+      "loss": 0.0214,
+      "step": 21791
+    },
+    {
+      "epoch": 4.952727272727273,
+      "grad_norm": 0.8411406603064271,
+      "learning_rate": 2.7833302569793555e-10,
+      "loss": 0.0039,
+      "step": 21792
+    },
+    {
+      "epoch": 4.952954545454546,
+      "grad_norm": 0.9308644437422426,
+      "learning_rate": 2.756761190655932e-10,
+      "loss": 0.005,
+      "step": 21793
+    },
+    {
+      "epoch": 4.953181818181818,
+      "grad_norm": 0.9734435882202142,
+      "learning_rate": 2.73031951652028e-10,
+      "loss": 0.0079,
+      "step": 21794
+    },
+    {
+      "epoch": 4.953409090909091,
+      "grad_norm": 0.3208662567331407,
+      "learning_rate": 2.7040052351129387e-10,
+      "loss": 0.0009,
+      "step": 21795
+    },
+    {
+      "epoch": 4.953636363636363,
+      "grad_norm": 0.706204427615534,
+      "learning_rate": 2.6778183469695906e-10,
+      "loss": 0.0015,
+      "step": 21796
+    },
+    {
+      "epoch": 4.953863636363637,
+      "grad_norm": 0.6103645779322839,
+      "learning_rate": 2.6517588526245307e-10,
+      "loss": 0.0014,
+      "step": 21797
+    },
+    {
+      "epoch": 4.954090909090909,
+      "grad_norm": 0.8476787096822158,
+      "learning_rate": 2.6258267526092783e-10,
+      "loss": 0.0056,
+      "step": 21798
+    },
+    {
+      "epoch": 4.954318181818182,
+      "grad_norm": 1.0223815816734811,
+      "learning_rate": 2.600022047452577e-10,
+      "loss": 0.0037,
+      "step": 21799
+    },
+    {
+      "epoch": 4.954545454545455,
+      "grad_norm": 1.1028901582353352,
+      "learning_rate": 2.574344737681089e-10,
+      "loss": 0.0099,
+      "step": 21800
+    },
+    {
+      "epoch": 4.9547727272727276,
+      "grad_norm": 2.1525537482380583,
+      "learning_rate": 2.5487948238173134e-10,
+      "loss": 0.0083,
+      "step": 21801
+    },
+    {
+      "epoch": 4.955,
+      "grad_norm": 0.7831879755571658,
+      "learning_rate": 2.523372306383054e-10,
+      "loss": 0.0028,
+      "step": 21802
+    },
+    {
+      "epoch": 4.955227272727273,
+      "grad_norm": 0.480865365528704,
+      "learning_rate": 2.4980771858966467e-10,
+      "loss": 0.0014,
+      "step": 21803
+    },
+    {
+      "epoch": 4.955454545454545,
+      "grad_norm": 0.8960888346033904,
+      "learning_rate": 2.472909462873652e-10,
+      "loss": 0.0044,
+      "step": 21804
+    },
+    {
+      "epoch": 4.9556818181818185,
+      "grad_norm": 0.6455125432657944,
+      "learning_rate": 2.4478691378275466e-10,
+      "loss": 0.0029,
+      "step": 21805
+    },
+    {
+      "epoch": 4.955909090909091,
+      "grad_norm": 0.9581007035790685,
+      "learning_rate": 2.422956211269034e-10,
+      "loss": 0.0026,
+      "step": 21806
+    },
+    {
+      "epoch": 4.956136363636364,
+      "grad_norm": 0.6061097879320956,
+      "learning_rate": 2.3981706837053477e-10,
+      "loss": 0.0017,
+      "step": 21807
+    },
+    {
+      "epoch": 4.956363636363636,
+      "grad_norm": 1.282509951034276,
+      "learning_rate": 2.373512555643026e-10,
+      "loss": 0.0024,
+      "step": 21808
+    },
+    {
+      "epoch": 4.9565909090909095,
+      "grad_norm": 2.25996118323506,
+      "learning_rate": 2.348981827584446e-10,
+      "loss": 0.0069,
+      "step": 21809
+    },
+    {
+      "epoch": 4.956818181818182,
+      "grad_norm": 0.5853358756646698,
+      "learning_rate": 2.3245785000292065e-10,
+      "loss": 0.003,
+      "step": 21810
+    },
+    {
+      "epoch": 4.9570454545454545,
+      "grad_norm": 1.459907529701215,
+      "learning_rate": 2.3003025734762153e-10,
+      "loss": 0.0049,
+      "step": 21811
+    },
+    {
+      "epoch": 4.957272727272727,
+      "grad_norm": 0.6427584836950417,
+      "learning_rate": 2.2761540484195211e-10,
+      "loss": 0.0038,
+      "step": 21812
+    },
+    {
+      "epoch": 4.9575,
+      "grad_norm": 0.180602321284944,
+      "learning_rate": 2.2521329253510914e-10,
+      "loss": 0.0003,
+      "step": 21813
+    },
+    {
+      "epoch": 4.957727272727273,
+      "grad_norm": 0.42725123744903987,
+      "learning_rate": 2.2282392047622003e-10,
+      "loss": 0.0009,
+      "step": 21814
+    },
+    {
+      "epoch": 4.9579545454545455,
+      "grad_norm": 0.25043639840114035,
+      "learning_rate": 2.204472887139264e-10,
+      "loss": 0.001,
+      "step": 21815
+    },
+    {
+      "epoch": 4.958181818181818,
+      "grad_norm": 2.26869777562958,
+      "learning_rate": 2.1808339729666174e-10,
+      "loss": 0.0095,
+      "step": 21816
+    },
+    {
+      "epoch": 4.9584090909090905,
+      "grad_norm": 1.2303763746935559,
+      "learning_rate": 2.157322462727207e-10,
+      "loss": 0.0029,
+      "step": 21817
+    },
+    {
+      "epoch": 4.958636363636364,
+      "grad_norm": 6.568955615858876,
+      "learning_rate": 2.1339383568998174e-10,
+      "loss": 0.0067,
+      "step": 21818
+    },
+    {
+      "epoch": 4.9588636363636365,
+      "grad_norm": 3.0183770408587773,
+      "learning_rate": 2.11068165596115e-10,
+      "loss": 0.0119,
+      "step": 21819
+    },
+    {
+      "epoch": 4.959090909090909,
+      "grad_norm": 0.4870801133493534,
+      "learning_rate": 2.087552360385825e-10,
+      "loss": 0.0024,
+      "step": 21820
+    },
+    {
+      "epoch": 4.9593181818181815,
+      "grad_norm": 0.5091163865805056,
+      "learning_rate": 2.0645504706456876e-10,
+      "loss": 0.0018,
+      "step": 21821
+    },
+    {
+      "epoch": 4.959545454545455,
+      "grad_norm": 1.6526828128819147,
+      "learning_rate": 2.041675987209113e-10,
+      "loss": 0.0072,
+      "step": 21822
+    },
+    {
+      "epoch": 4.959772727272727,
+      "grad_norm": 1.0235247982060478,
+      "learning_rate": 2.0189289105430886e-10,
+      "loss": 0.0028,
+      "step": 21823
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 1.2722056066149228,
+      "learning_rate": 1.996309241111133e-10,
+      "loss": 0.0056,
+      "step": 21824
+    },
+    {
+      "epoch": 4.9602272727272725,
+      "grad_norm": 0.7014052658258695,
+      "learning_rate": 1.9738169793753758e-10,
+      "loss": 0.0034,
+      "step": 21825
+    },
+    {
+      "epoch": 4.960454545454546,
+      "grad_norm": 1.8432489322842784,
+      "learning_rate": 1.9514521257930906e-10,
+      "loss": 0.0058,
+      "step": 21826
+    },
+    {
+      "epoch": 4.960681818181818,
+      "grad_norm": 10.672427294484125,
+      "learning_rate": 1.9292146808208567e-10,
+      "loss": 0.01,
+      "step": 21827
+    },
+    {
+      "epoch": 4.960909090909091,
+      "grad_norm": 0.7229248526076923,
+      "learning_rate": 1.9071046449124774e-10,
+      "loss": 0.0047,
+      "step": 21828
+    },
+    {
+      "epoch": 4.9611363636363635,
+      "grad_norm": 0.955490719401935,
+      "learning_rate": 1.8851220185189811e-10,
+      "loss": 0.0036,
+      "step": 21829
+    },
+    {
+      "epoch": 4.961363636363636,
+      "grad_norm": 1.7686808958649909,
+      "learning_rate": 1.8632668020879263e-10,
+      "loss": 0.0084,
+      "step": 21830
+    },
+    {
+      "epoch": 4.961590909090909,
+      "grad_norm": 1.2612785885931355,
+      "learning_rate": 1.841538996065484e-10,
+      "loss": 0.0062,
+      "step": 21831
+    },
+    {
+      "epoch": 4.961818181818182,
+      "grad_norm": 0.49015226383281396,
+      "learning_rate": 1.8199386008943553e-10,
+      "loss": 0.003,
+      "step": 21832
+    },
+    {
+      "epoch": 4.962045454545454,
+      "grad_norm": 0.17236535528789082,
+      "learning_rate": 1.7984656170158544e-10,
+      "loss": 0.0004,
+      "step": 21833
+    },
+    {
+      "epoch": 4.962272727272727,
+      "grad_norm": 0.46479305420001826,
+      "learning_rate": 1.777120044866437e-10,
+      "loss": 0.0017,
+      "step": 21834
+    },
+    {
+      "epoch": 4.9625,
+      "grad_norm": 0.8405632429575065,
+      "learning_rate": 1.7559018848825604e-10,
+      "loss": 0.0025,
+      "step": 21835
+    },
+    {
+      "epoch": 4.962727272727273,
+      "grad_norm": 1.3091081661975752,
+      "learning_rate": 1.734811137496517e-10,
+      "loss": 0.0032,
+      "step": 21836
+    },
+    {
+      "epoch": 4.962954545454545,
+      "grad_norm": 0.5165603739804019,
+      "learning_rate": 1.7138478031385187e-10,
+      "loss": 0.0012,
+      "step": 21837
+    },
+    {
+      "epoch": 4.963181818181818,
+      "grad_norm": 2.144612946421606,
+      "learning_rate": 1.693011882235307e-10,
+      "loss": 0.0067,
+      "step": 21838
+    },
+    {
+      "epoch": 4.96340909090909,
+      "grad_norm": 1.3552246728129846,
+      "learning_rate": 1.6723033752129303e-10,
+      "loss": 0.0186,
+      "step": 21839
+    },
+    {
+      "epoch": 4.963636363636364,
+      "grad_norm": 0.7900286392264282,
+      "learning_rate": 1.6517222824925793e-10,
+      "loss": 0.0027,
+      "step": 21840
+    },
+    {
+      "epoch": 4.963863636363636,
+      "grad_norm": 1.2836433302605457,
+      "learning_rate": 1.6312686044947511e-10,
+      "loss": 0.01,
+      "step": 21841
+    },
+    {
+      "epoch": 4.964090909090909,
+      "grad_norm": 0.9145540524949495,
+      "learning_rate": 1.6109423416364733e-10,
+      "loss": 0.005,
+      "step": 21842
+    },
+    {
+      "epoch": 4.964318181818182,
+      "grad_norm": 0.5872011062807613,
+      "learning_rate": 1.5907434943319978e-10,
+      "loss": 0.0016,
+      "step": 21843
+    },
+    {
+      "epoch": 4.964545454545455,
+      "grad_norm": 1.7936300028084777,
+      "learning_rate": 1.5706720629928008e-10,
+      "loss": 0.0083,
+      "step": 21844
+    },
+    {
+      "epoch": 4.964772727272727,
+      "grad_norm": 1.5170509719892802,
+      "learning_rate": 1.550728048028971e-10,
+      "loss": 0.0084,
+      "step": 21845
+    },
+    {
+      "epoch": 4.965,
+      "grad_norm": 0.48485942704300694,
+      "learning_rate": 1.5309114498464339e-10,
+      "loss": 0.0016,
+      "step": 21846
+    },
+    {
+      "epoch": 4.965227272727272,
+      "grad_norm": 0.6980376599895802,
+      "learning_rate": 1.511222268849727e-10,
+      "loss": 0.0012,
+      "step": 21847
+    },
+    {
+      "epoch": 4.965454545454546,
+      "grad_norm": 0.7450315093552944,
+      "learning_rate": 1.491660505440612e-10,
+      "loss": 0.0032,
+      "step": 21848
+    },
+    {
+      "epoch": 4.965681818181818,
+      "grad_norm": 0.4259434385890639,
+      "learning_rate": 1.4722261600173816e-10,
+      "loss": 0.0009,
+      "step": 21849
+    },
+    {
+      "epoch": 4.965909090909091,
+      "grad_norm": 1.4530935618073066,
+      "learning_rate": 1.4529192329769403e-10,
+      "loss": 0.0045,
+      "step": 21850
+    },
+    {
+      "epoch": 4.966136363636363,
+      "grad_norm": 0.2402480267503221,
+      "learning_rate": 1.4337397247120298e-10,
+      "loss": 0.0009,
+      "step": 21851
+    },
+    {
+      "epoch": 4.966363636363637,
+      "grad_norm": 0.4873925145964774,
+      "learning_rate": 1.4146876356146977e-10,
+      "loss": 0.0018,
+      "step": 21852
+    },
+    {
+      "epoch": 4.966590909090909,
+      "grad_norm": 1.8026506776397067,
+      "learning_rate": 1.3957629660735217e-10,
+      "loss": 0.0138,
+      "step": 21853
+    },
+    {
+      "epoch": 4.966818181818182,
+      "grad_norm": 1.040419023590425,
+      "learning_rate": 1.3769657164736105e-10,
+      "loss": 0.0042,
+      "step": 21854
+    },
+    {
+      "epoch": 4.967045454545454,
+      "grad_norm": 1.2695282414488622,
+      "learning_rate": 1.358295887198685e-10,
+      "loss": 0.0054,
+      "step": 21855
+    },
+    {
+      "epoch": 4.967272727272727,
+      "grad_norm": 1.8787418717505804,
+      "learning_rate": 1.3397534786296907e-10,
+      "loss": 0.0062,
+      "step": 21856
+    },
+    {
+      "epoch": 4.9675,
+      "grad_norm": 1.1900087525900978,
+      "learning_rate": 1.321338491144103e-10,
+      "loss": 0.0072,
+      "step": 21857
+    },
+    {
+      "epoch": 4.967727272727273,
+      "grad_norm": 0.352296558071497,
+      "learning_rate": 1.3030509251180102e-10,
+      "loss": 0.0021,
+      "step": 21858
+    },
+    {
+      "epoch": 4.967954545454545,
+      "grad_norm": 0.42259285538829383,
+      "learning_rate": 1.2848907809247247e-10,
+      "loss": 0.0019,
+      "step": 21859
+    },
+    {
+      "epoch": 4.968181818181818,
+      "grad_norm": 2.6857004519842773,
+      "learning_rate": 1.2668580589333956e-10,
+      "loss": 0.0089,
+      "step": 21860
+    },
+    {
+      "epoch": 4.968409090909091,
+      "grad_norm": 0.7713704619444671,
+      "learning_rate": 1.2489527595124785e-10,
+      "loss": 0.002,
+      "step": 21861
+    },
+    {
+      "epoch": 4.968636363636364,
+      "grad_norm": 5.104648016145386,
+      "learning_rate": 1.2311748830269587e-10,
+      "loss": 0.0094,
+      "step": 21862
+    },
+    {
+      "epoch": 4.968863636363636,
+      "grad_norm": 0.7399944573773225,
+      "learning_rate": 1.213524429839047e-10,
+      "loss": 0.0054,
+      "step": 21863
+    },
+    {
+      "epoch": 4.969090909090909,
+      "grad_norm": 1.5301214423306753,
+      "learning_rate": 1.1960014003095655e-10,
+      "loss": 0.004,
+      "step": 21864
+    },
+    {
+      "epoch": 4.969318181818182,
+      "grad_norm": 1.1926676480066751,
+      "learning_rate": 1.1786057947944794e-10,
+      "loss": 0.0144,
+      "step": 21865
+    },
+    {
+      "epoch": 4.969545454545455,
+      "grad_norm": 2.0717939817104334,
+      "learning_rate": 1.1613376136504484e-10,
+      "loss": 0.0092,
+      "step": 21866
+    },
+    {
+      "epoch": 4.969772727272727,
+      "grad_norm": 0.4248351727834252,
+      "learning_rate": 1.1441968572271923e-10,
+      "loss": 0.0016,
+      "step": 21867
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 0.44206250806565756,
+      "learning_rate": 1.1271835258765129e-10,
+      "loss": 0.0014,
+      "step": 21868
+    },
+    {
+      "epoch": 4.970227272727273,
+      "grad_norm": 1.04474721175951,
+      "learning_rate": 1.1102976199439675e-10,
+      "loss": 0.0025,
+      "step": 21869
+    },
+    {
+      "epoch": 4.970454545454546,
+      "grad_norm": 2.4562300082169317,
+      "learning_rate": 1.093539139773725e-10,
+      "loss": 0.0179,
+      "step": 21870
+    },
+    {
+      "epoch": 4.970681818181818,
+      "grad_norm": 1.014742552619562,
+      "learning_rate": 1.0769080857085667e-10,
+      "loss": 0.0028,
+      "step": 21871
+    },
+    {
+      "epoch": 4.970909090909091,
+      "grad_norm": 0.8323900249694393,
+      "learning_rate": 1.0604044580871109e-10,
+      "loss": 0.0053,
+      "step": 21872
+    },
+    {
+      "epoch": 4.971136363636363,
+      "grad_norm": 0.697175766277631,
+      "learning_rate": 1.0440282572458937e-10,
+      "loss": 0.0019,
+      "step": 21873
+    },
+    {
+      "epoch": 4.971363636363637,
+      "grad_norm": 0.6947183345912733,
+      "learning_rate": 1.0277794835186761e-10,
+      "loss": 0.0016,
+      "step": 21874
+    },
+    {
+      "epoch": 4.971590909090909,
+      "grad_norm": 1.6662753629485376,
+      "learning_rate": 1.011658137237137e-10,
+      "loss": 0.0062,
+      "step": 21875
+    },
+    {
+      "epoch": 4.971818181818182,
+      "grad_norm": 0.44299654440456576,
+      "learning_rate": 9.956642187294863e-11,
+      "loss": 0.0015,
+      "step": 21876
+    },
+    {
+      "epoch": 4.972045454545454,
+      "grad_norm": 1.5744092086633839,
+      "learning_rate": 9.797977283225457e-11,
+      "loss": 0.0028,
+      "step": 21877
+    },
+    {
+      "epoch": 4.972272727272728,
+      "grad_norm": 0.3818815258335783,
+      "learning_rate": 9.640586663389739e-11,
+      "loss": 0.0009,
+      "step": 21878
+    },
+    {
+      "epoch": 4.9725,
+      "grad_norm": 1.250792880627434,
+      "learning_rate": 9.484470331007356e-11,
+      "loss": 0.0081,
+      "step": 21879
+    },
+    {
+      "epoch": 4.972727272727273,
+      "grad_norm": 0.3530876992000406,
+      "learning_rate": 9.32962828926326e-11,
+      "loss": 0.0015,
+      "step": 21880
+    },
+    {
+      "epoch": 4.972954545454545,
+      "grad_norm": 0.5587997274521209,
+      "learning_rate": 9.176060541300768e-11,
+      "loss": 0.0017,
+      "step": 21881
+    },
+    {
+      "epoch": 4.973181818181818,
+      "grad_norm": 1.2223230500641058,
+      "learning_rate": 9.023767090263202e-11,
+      "loss": 0.0077,
+      "step": 21882
+    },
+    {
+      "epoch": 4.973409090909091,
+      "grad_norm": 1.7929179956038352,
+      "learning_rate": 8.872747939252247e-11,
+      "loss": 0.0176,
+      "step": 21883
+    },
+    {
+      "epoch": 4.973636363636364,
+      "grad_norm": 1.6670009573772062,
+      "learning_rate": 8.723003091341831e-11,
+      "loss": 0.0086,
+      "step": 21884
+    },
+    {
+      "epoch": 4.973863636363636,
+      "grad_norm": 2.2079291986125646,
+      "learning_rate": 8.574532549598946e-11,
+      "loss": 0.0047,
+      "step": 21885
+    },
+    {
+      "epoch": 4.97409090909091,
+      "grad_norm": 0.36393820959107515,
+      "learning_rate": 8.427336317042012e-11,
+      "loss": 0.0008,
+      "step": 21886
+    },
+    {
+      "epoch": 4.974318181818182,
+      "grad_norm": 0.2979522197565185,
+      "learning_rate": 8.281414396675569e-11,
+      "loss": 0.0008,
+      "step": 21887
+    },
+    {
+      "epoch": 4.974545454545455,
+      "grad_norm": 2.104570560556785,
+      "learning_rate": 8.136766791469463e-11,
+      "loss": 0.0149,
+      "step": 21888
+    },
+    {
+      "epoch": 4.974772727272727,
+      "grad_norm": 1.5107090848321854,
+      "learning_rate": 7.993393504379665e-11,
+      "loss": 0.0065,
+      "step": 21889
+    },
+    {
+      "epoch": 4.975,
+      "grad_norm": 0.6030524407418578,
+      "learning_rate": 7.851294538327447e-11,
+      "loss": 0.0018,
+      "step": 21890
+    },
+    {
+      "epoch": 4.975227272727273,
+      "grad_norm": 0.31458767747202315,
+      "learning_rate": 7.710469896213268e-11,
+      "loss": 0.001,
+      "step": 21891
+    },
+    {
+      "epoch": 4.975454545454546,
+      "grad_norm": 0.6104635352950436,
+      "learning_rate": 7.57091958090289e-11,
+      "loss": 0.0016,
+      "step": 21892
+    },
+    {
+      "epoch": 4.975681818181818,
+      "grad_norm": 0.8766964941543017,
+      "learning_rate": 7.432643595248202e-11,
+      "loss": 0.003,
+      "step": 21893
+    },
+    {
+      "epoch": 4.975909090909091,
+      "grad_norm": 0.7236307247968701,
+      "learning_rate": 7.29564194206639e-11,
+      "loss": 0.0034,
+      "step": 21894
+    },
+    {
+      "epoch": 4.976136363636364,
+      "grad_norm": 0.6610071099196569,
+      "learning_rate": 7.159914624146891e-11,
+      "loss": 0.0085,
+      "step": 21895
+    },
+    {
+      "epoch": 4.9763636363636365,
+      "grad_norm": 1.2885274039405725,
+      "learning_rate": 7.025461644265263e-11,
+      "loss": 0.0035,
+      "step": 21896
+    },
+    {
+      "epoch": 4.976590909090909,
+      "grad_norm": 1.512406192761858,
+      "learning_rate": 6.892283005162371e-11,
+      "loss": 0.0083,
+      "step": 21897
+    },
+    {
+      "epoch": 4.976818181818182,
+      "grad_norm": 1.4831102039515105,
+      "learning_rate": 6.76037870954438e-11,
+      "loss": 0.0045,
+      "step": 21898
+    },
+    {
+      "epoch": 4.977045454545454,
+      "grad_norm": 1.8093050019148231,
+      "learning_rate": 6.629748760110522e-11,
+      "loss": 0.0028,
+      "step": 21899
+    },
+    {
+      "epoch": 4.9772727272727275,
+      "grad_norm": 0.825982382697242,
+      "learning_rate": 6.500393159525332e-11,
+      "loss": 0.0039,
+      "step": 21900
+    },
+    {
+      "epoch": 4.9775,
+      "grad_norm": 0.36975978229213435,
+      "learning_rate": 6.37231191041865e-11,
+      "loss": 0.0008,
+      "step": 21901
+    },
+    {
+      "epoch": 4.977727272727273,
+      "grad_norm": 1.229898489680273,
+      "learning_rate": 6.245505015406439e-11,
+      "loss": 0.0042,
+      "step": 21902
+    },
+    {
+      "epoch": 4.977954545454545,
+      "grad_norm": 1.2027424120422305,
+      "learning_rate": 6.119972477076908e-11,
+      "loss": 0.0047,
+      "step": 21903
+    },
+    {
+      "epoch": 4.9781818181818185,
+      "grad_norm": 0.2453223759044651,
+      "learning_rate": 5.995714297983568e-11,
+      "loss": 0.0013,
+      "step": 21904
+    },
+    {
+      "epoch": 4.978409090909091,
+      "grad_norm": 1.4309065297790975,
+      "learning_rate": 5.872730480666056e-11,
+      "loss": 0.0086,
+      "step": 21905
+    },
+    {
+      "epoch": 4.9786363636363635,
+      "grad_norm": 0.8666474698182552,
+      "learning_rate": 5.75102102763625e-11,
+      "loss": 0.0024,
+      "step": 21906
+    },
+    {
+      "epoch": 4.978863636363636,
+      "grad_norm": 0.809357073384839,
+      "learning_rate": 5.630585941364397e-11,
+      "loss": 0.0029,
+      "step": 21907
+    },
+    {
+      "epoch": 4.9790909090909095,
+      "grad_norm": 2.4059435542133674,
+      "learning_rate": 5.511425224313805e-11,
+      "loss": 0.0279,
+      "step": 21908
+    },
+    {
+      "epoch": 4.979318181818182,
+      "grad_norm": 0.8263140455341361,
+      "learning_rate": 5.393538878913085e-11,
+      "loss": 0.0027,
+      "step": 21909
+    },
+    {
+      "epoch": 4.9795454545454545,
+      "grad_norm": 1.6303901770724487,
+      "learning_rate": 5.276926907563096e-11,
+      "loss": 0.0127,
+      "step": 21910
+    },
+    {
+      "epoch": 4.979772727272727,
+      "grad_norm": 0.5950168490290192,
+      "learning_rate": 5.161589312650816e-11,
+      "loss": 0.0018,
+      "step": 21911
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 1.709655663656172,
+      "learning_rate": 5.0475260965146545e-11,
+      "loss": 0.0034,
+      "step": 21912
+    },
+    {
+      "epoch": 4.980227272727273,
+      "grad_norm": 0.7004571554966754,
+      "learning_rate": 4.934737261493017e-11,
+      "loss": 0.0029,
+      "step": 21913
+    },
+    {
+      "epoch": 4.9804545454545455,
+      "grad_norm": 2.253864318717863,
+      "learning_rate": 4.823222809875738e-11,
+      "loss": 0.0219,
+      "step": 21914
+    },
+    {
+      "epoch": 4.980681818181818,
+      "grad_norm": 0.6767311437583081,
+      "learning_rate": 4.7129827439457155e-11,
+      "loss": 0.005,
+      "step": 21915
+    },
+    {
+      "epoch": 4.9809090909090905,
+      "grad_norm": 1.3139773479917267,
+      "learning_rate": 4.6040170659511495e-11,
+      "loss": 0.0065,
+      "step": 21916
+    },
+    {
+      "epoch": 4.981136363636364,
+      "grad_norm": 0.4713533475190403,
+      "learning_rate": 4.496325778105548e-11,
+      "loss": 0.0014,
+      "step": 21917
+    },
+    {
+      "epoch": 4.981363636363636,
+      "grad_norm": 1.5579051094906071,
+      "learning_rate": 4.38990888260854e-11,
+      "loss": 0.0058,
+      "step": 21918
+    },
+    {
+      "epoch": 4.981590909090909,
+      "grad_norm": 1.5310679886549998,
+      "learning_rate": 4.284766381632e-11,
+      "loss": 0.0086,
+      "step": 21919
+    },
+    {
+      "epoch": 4.9818181818181815,
+      "grad_norm": 0.9066177509669014,
+      "learning_rate": 4.180898277313106e-11,
+      "loss": 0.005,
+      "step": 21920
+    },
+    {
+      "epoch": 4.982045454545455,
+      "grad_norm": 0.4790671865499382,
+      "learning_rate": 4.078304571782099e-11,
+      "loss": 0.0013,
+      "step": 21921
+    },
+    {
+      "epoch": 4.982272727272727,
+      "grad_norm": 0.9705629505651928,
+      "learning_rate": 3.9769852671206474e-11,
+      "loss": 0.0073,
+      "step": 21922
+    },
+    {
+      "epoch": 4.9825,
+      "grad_norm": 1.005521293190892,
+      "learning_rate": 3.876940365403481e-11,
+      "loss": 0.0031,
+      "step": 21923
+    },
+    {
+      "epoch": 4.9827272727272724,
+      "grad_norm": 0.765873733235613,
+      "learning_rate": 3.778169868663695e-11,
+      "loss": 0.0045,
+      "step": 21924
+    },
+    {
+      "epoch": 4.982954545454545,
+      "grad_norm": 1.20550358195799,
+      "learning_rate": 3.680673778920507e-11,
+      "loss": 0.0051,
+      "step": 21925
+    },
+    {
+      "epoch": 4.983181818181818,
+      "grad_norm": 2.6188981865688703,
+      "learning_rate": 3.584452098151503e-11,
+      "loss": 0.0105,
+      "step": 21926
+    },
+    {
+      "epoch": 4.983409090909091,
+      "grad_norm": 0.5124793009742068,
+      "learning_rate": 3.489504828334267e-11,
+      "loss": 0.0011,
+      "step": 21927
+    },
+    {
+      "epoch": 4.983636363636363,
+      "grad_norm": 0.4113595754937458,
+      "learning_rate": 3.395831971397812e-11,
+      "loss": 0.0012,
+      "step": 21928
+    },
+    {
+      "epoch": 4.983863636363637,
+      "grad_norm": 1.6744747366022363,
+      "learning_rate": 3.303433529243394e-11,
+      "loss": 0.0157,
+      "step": 21929
+    },
+    {
+      "epoch": 4.984090909090909,
+      "grad_norm": 0.6126880661080939,
+      "learning_rate": 3.2123095037722704e-11,
+      "loss": 0.0028,
+      "step": 21930
+    },
+    {
+      "epoch": 4.984318181818182,
+      "grad_norm": 0.6012386923036791,
+      "learning_rate": 3.122459896830188e-11,
+      "loss": 0.0021,
+      "step": 21931
+    },
+    {
+      "epoch": 4.984545454545454,
+      "grad_norm": 6.44289109001708,
+      "learning_rate": 3.033884710249013e-11,
+      "loss": 0.0205,
+      "step": 21932
+    },
+    {
+      "epoch": 4.984772727272727,
+      "grad_norm": 0.8037176901834192,
+      "learning_rate": 2.946583945839798e-11,
+      "loss": 0.0026,
+      "step": 21933
+    },
+    {
+      "epoch": 4.985,
+      "grad_norm": 0.3342013824976586,
+      "learning_rate": 2.8605576053858386e-11,
+      "loss": 0.0012,
+      "step": 21934
+    },
+    {
+      "epoch": 4.985227272727273,
+      "grad_norm": 2.717413919178886,
+      "learning_rate": 2.7758056906357357e-11,
+      "loss": 0.0273,
+      "step": 21935
+    },
+    {
+      "epoch": 4.985454545454545,
+      "grad_norm": 0.37684000966037784,
+      "learning_rate": 2.6923282033172738e-11,
+      "loss": 0.0011,
+      "step": 21936
+    },
+    {
+      "epoch": 4.985681818181818,
+      "grad_norm": 1.0692147188980654,
+      "learning_rate": 2.610125145137421e-11,
+      "loss": 0.0086,
+      "step": 21937
+    },
+    {
+      "epoch": 4.985909090909091,
+      "grad_norm": 0.17557540627139462,
+      "learning_rate": 2.5291965177753896e-11,
+      "loss": 0.0004,
+      "step": 21938
+    },
+    {
+      "epoch": 4.986136363636364,
+      "grad_norm": 0.43943464977177454,
+      "learning_rate": 2.4495423228687587e-11,
+      "loss": 0.0014,
+      "step": 21939
+    },
+    {
+      "epoch": 4.986363636363636,
+      "grad_norm": 1.069963382114333,
+      "learning_rate": 2.3711625620481683e-11,
+      "loss": 0.0051,
+      "step": 21940
+    },
+    {
+      "epoch": 4.986590909090909,
+      "grad_norm": 0.7531099718236848,
+      "learning_rate": 2.2940572369165025e-11,
+      "loss": 0.0025,
+      "step": 21941
+    },
+    {
+      "epoch": 4.986818181818181,
+      "grad_norm": 0.505967736624618,
+      "learning_rate": 2.218226349035013e-11,
+      "loss": 0.0031,
+      "step": 21942
+    },
+    {
+      "epoch": 4.987045454545455,
+      "grad_norm": 1.2992496687864492,
+      "learning_rate": 2.1436698999649508e-11,
+      "loss": 0.0047,
+      "step": 21943
+    },
+    {
+      "epoch": 4.987272727272727,
+      "grad_norm": 0.536719495712336,
+      "learning_rate": 2.0703878912189944e-11,
+      "loss": 0.0017,
+      "step": 21944
+    },
+    {
+      "epoch": 4.9875,
+      "grad_norm": 1.0285499757828935,
+      "learning_rate": 1.9983803242890064e-11,
+      "loss": 0.0085,
+      "step": 21945
+    },
+    {
+      "epoch": 4.987727272727272,
+      "grad_norm": 1.4014045831259792,
+      "learning_rate": 1.9276472006460322e-11,
+      "loss": 0.0076,
+      "step": 21946
+    },
+    {
+      "epoch": 4.987954545454546,
+      "grad_norm": 0.5667439916648468,
+      "learning_rate": 1.8581885217264227e-11,
+      "loss": 0.0014,
+      "step": 21947
+    },
+    {
+      "epoch": 4.988181818181818,
+      "grad_norm": 1.5822089134242907,
+      "learning_rate": 1.7900042889595903e-11,
+      "loss": 0.0113,
+      "step": 21948
+    },
+    {
+      "epoch": 4.988409090909091,
+      "grad_norm": 0.7645074550977614,
+      "learning_rate": 1.7230945037263748e-11,
+      "loss": 0.0019,
+      "step": 21949
+    },
+    {
+      "epoch": 4.988636363636363,
+      "grad_norm": 1.12982789376798,
+      "learning_rate": 1.657459167393738e-11,
+      "loss": 0.0089,
+      "step": 21950
+    },
+    {
+      "epoch": 4.988863636363637,
+      "grad_norm": 0.6290855668185851,
+      "learning_rate": 1.5930982813008865e-11,
+      "loss": 0.002,
+      "step": 21951
+    },
+    {
+      "epoch": 4.989090909090909,
+      "grad_norm": 2.4632450677324638,
+      "learning_rate": 1.5300118467592717e-11,
+      "loss": 0.0056,
+      "step": 21952
+    },
+    {
+      "epoch": 4.989318181818182,
+      "grad_norm": 0.8974663918673784,
+      "learning_rate": 1.4681998650525886e-11,
+      "loss": 0.0026,
+      "step": 21953
+    },
+    {
+      "epoch": 4.989545454545454,
+      "grad_norm": 0.4294231263187667,
+      "learning_rate": 1.407662337443716e-11,
+      "loss": 0.0014,
+      "step": 21954
+    },
+    {
+      "epoch": 4.989772727272728,
+      "grad_norm": 4.623920895604957,
+      "learning_rate": 1.3483992651747158e-11,
+      "loss": 0.0069,
+      "step": 21955
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 1.0630046715845076,
+      "learning_rate": 1.2904106494390778e-11,
+      "loss": 0.0032,
+      "step": 21956
+    },
+    {
+      "epoch": 4.990227272727273,
+      "grad_norm": 0.3925044708701705,
+      "learning_rate": 1.2336964914302918e-11,
+      "loss": 0.0022,
+      "step": 21957
+    },
+    {
+      "epoch": 4.990454545454545,
+      "grad_norm": 2.860246834683172,
+      "learning_rate": 1.1782567923002141e-11,
+      "loss": 0.0099,
+      "step": 21958
+    },
+    {
+      "epoch": 4.990681818181818,
+      "grad_norm": 1.8113630126503562,
+      "learning_rate": 1.1240915531798845e-11,
+      "loss": 0.0065,
+      "step": 21959
+    },
+    {
+      "epoch": 4.990909090909091,
+      "grad_norm": 0.833729393780592,
+      "learning_rate": 1.0712007751795261e-11,
+      "loss": 0.0036,
+      "step": 21960
+    },
+    {
+      "epoch": 4.991136363636364,
+      "grad_norm": 0.970628260519193,
+      "learning_rate": 1.0195844593677285e-11,
+      "loss": 0.0041,
+      "step": 21961
+    },
+    {
+      "epoch": 4.991363636363636,
+      "grad_norm": 1.7708148249301285,
+      "learning_rate": 9.692426067992034e-12,
+      "loss": 0.0041,
+      "step": 21962
+    },
+    {
+      "epoch": 4.991590909090909,
+      "grad_norm": 0.7961868632391601,
+      "learning_rate": 9.201752185078461e-12,
+      "loss": 0.0029,
+      "step": 21963
+    },
+    {
+      "epoch": 4.991818181818182,
+      "grad_norm": 0.8050196873560773,
+      "learning_rate": 8.723822954928573e-12,
+      "loss": 0.0043,
+      "step": 21964
+    },
+    {
+      "epoch": 4.992045454545455,
+      "grad_norm": 1.5150134187264017,
+      "learning_rate": 8.258638387187434e-12,
+      "loss": 0.0084,
+      "step": 21965
+    },
+    {
+      "epoch": 4.992272727272727,
+      "grad_norm": 1.2504636326084382,
+      "learning_rate": 7.806198491430717e-12,
+      "loss": 0.0044,
+      "step": 21966
+    },
+    {
+      "epoch": 4.9925,
+      "grad_norm": 0.699653717632133,
+      "learning_rate": 7.36650327688715e-12,
+      "loss": 0.0028,
+      "step": 21967
+    },
+    {
+      "epoch": 4.992727272727272,
+      "grad_norm": 1.9520530767919761,
+      "learning_rate": 6.939552752507905e-12,
+      "loss": 0.0082,
+      "step": 21968
+    },
+    {
+      "epoch": 4.992954545454546,
+      "grad_norm": 1.4330087302614938,
+      "learning_rate": 6.525346926966602e-12,
+      "loss": 0.0078,
+      "step": 21969
+    },
+    {
+      "epoch": 4.993181818181818,
+      "grad_norm": 0.39829239552825274,
+      "learning_rate": 6.1238858087286914e-12,
+      "loss": 0.0014,
+      "step": 21970
+    },
+    {
+      "epoch": 4.993409090909091,
+      "grad_norm": 0.5978818171631448,
+      "learning_rate": 5.735169406051455e-12,
+      "loss": 0.0024,
+      "step": 21971
+    },
+    {
+      "epoch": 4.993636363636364,
+      "grad_norm": 1.6322281028992824,
+      "learning_rate": 5.359197726706455e-12,
+      "loss": 0.0097,
+      "step": 21972
+    },
+    {
+      "epoch": 4.993863636363637,
+      "grad_norm": 0.41138087902544285,
+      "learning_rate": 4.995970778465254e-12,
+      "loss": 0.0017,
+      "step": 21973
+    },
+    {
+      "epoch": 4.994090909090909,
+      "grad_norm": 1.5744429031378668,
+      "learning_rate": 4.645488568752465e-12,
+      "loss": 0.0073,
+      "step": 21974
+    },
+    {
+      "epoch": 4.994318181818182,
+      "grad_norm": 0.12889098220755696,
+      "learning_rate": 4.307751104715152e-12,
+      "loss": 0.0004,
+      "step": 21975
+    },
+    {
+      "epoch": 4.994545454545454,
+      "grad_norm": 1.6446149810391273,
+      "learning_rate": 3.98275839315343e-12,
+      "loss": 0.0037,
+      "step": 21976
+    },
+    {
+      "epoch": 4.994772727272728,
+      "grad_norm": 0.559439765483483,
+      "learning_rate": 3.670510440728636e-12,
+      "loss": 0.0019,
+      "step": 21977
+    },
+    {
+      "epoch": 4.995,
+      "grad_norm": 0.4128778909515117,
+      "learning_rate": 3.3710072538939433e-12,
+      "loss": 0.0011,
+      "step": 21978
+    },
+    {
+      "epoch": 4.995227272727273,
+      "grad_norm": 2.9253681199439785,
+      "learning_rate": 3.084248838616799e-12,
+      "loss": 0.0101,
+      "step": 21979
+    },
+    {
+      "epoch": 4.995454545454545,
+      "grad_norm": 1.3331559051567672,
+      "learning_rate": 2.8102352008646525e-12,
+      "loss": 0.0045,
+      "step": 21980
+    },
+    {
+      "epoch": 4.995681818181819,
+      "grad_norm": 0.6269864815153362,
+      "learning_rate": 2.5489663461886192e-12,
+      "loss": 0.0027,
+      "step": 21981
+    },
+    {
+      "epoch": 4.995909090909091,
+      "grad_norm": 0.6206460314898734,
+      "learning_rate": 2.300442279931647e-12,
+      "loss": 0.0013,
+      "step": 21982
+    },
+    {
+      "epoch": 4.996136363636364,
+      "grad_norm": 2.171266426965877,
+      "learning_rate": 2.06466300708974e-12,
+      "loss": 0.0069,
+      "step": 21983
+    },
+    {
+      "epoch": 4.996363636363636,
+      "grad_norm": 0.28062128265875863,
+      "learning_rate": 1.841628532520123e-12,
+      "loss": 0.0008,
+      "step": 21984
+    },
+    {
+      "epoch": 4.996590909090909,
+      "grad_norm": 0.6859116553992395,
+      "learning_rate": 1.6313388607330781e-12,
+      "loss": 0.0031,
+      "step": 21985
+    },
+    {
+      "epoch": 4.996818181818182,
+      "grad_norm": 1.0265783635863404,
+      "learning_rate": 1.433793996100108e-12,
+      "loss": 0.0024,
+      "step": 21986
+    },
+    {
+      "epoch": 4.997045454545455,
+      "grad_norm": 0.8620364208586103,
+      "learning_rate": 1.2489939426457708e-12,
+      "loss": 0.0027,
+      "step": 21987
+    },
+    {
+      "epoch": 4.997272727272727,
+      "grad_norm": 1.4741336125285454,
+      "learning_rate": 1.076938703978292e-12,
+      "loss": 0.0034,
+      "step": 21988
+    },
+    {
+      "epoch": 4.9975000000000005,
+      "grad_norm": 1.137370959964085,
+      "learning_rate": 9.176282838446737e-13,
+      "loss": 0.0042,
+      "step": 21989
+    },
+    {
+      "epoch": 4.997727272727273,
+      "grad_norm": 1.6498754954414496,
+      "learning_rate": 7.710626852980296e-13,
+      "loss": 0.0025,
+      "step": 21990
+    },
+    {
+      "epoch": 4.9979545454545455,
+      "grad_norm": 0.567194981033282,
+      "learning_rate": 6.37241911391473e-13,
+      "loss": 0.0019,
+      "step": 21991
+    },
+    {
+      "epoch": 4.998181818181818,
+      "grad_norm": 2.0360173984659125,
+      "learning_rate": 5.161659649005612e-13,
+      "loss": 0.0141,
+      "step": 21992
+    },
+    {
+      "epoch": 4.998409090909091,
+      "grad_norm": 1.3019095903066222,
+      "learning_rate": 4.078348482539074e-13,
+      "loss": 0.0061,
+      "step": 21993
+    },
+    {
+      "epoch": 4.998636363636364,
+      "grad_norm": 1.4578062289086773,
+      "learning_rate": 3.1224856367195744e-13,
+      "loss": 0.01,
+      "step": 21994
+    },
+    {
+      "epoch": 4.9988636363636365,
+      "grad_norm": 0.12373193188924148,
+      "learning_rate": 2.294071130282127e-13,
+      "loss": 0.0004,
+      "step": 21995
+    },
+    {
+      "epoch": 4.999090909090909,
+      "grad_norm": 0.37091000367456994,
+      "learning_rate": 1.5931049812678568e-13,
+      "loss": 0.0009,
+      "step": 21996
+    },
+    {
+      "epoch": 4.9993181818181816,
+      "grad_norm": 0.6555394233060091,
+      "learning_rate": 1.0195872035545506e-13,
+      "loss": 0.0086,
+      "step": 21997
+    },
+    {
+      "epoch": 4.999545454545455,
+      "grad_norm": 1.017816094581103,
+      "learning_rate": 5.735178089383287e-14,
+      "loss": 0.0048,
+      "step": 21998
+    },
+    {
+      "epoch": 4.9997727272727275,
+      "grad_norm": 1.8123685740455593,
+      "learning_rate": 2.548968064397528e-14,
+      "loss": 0.0059,
+      "step": 21999
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.3724647017072994,
+      "learning_rate": 6.37242016099382e-15,
+      "loss": 0.0018,
+      "step": 22000
+    },
+    {
+      "epoch": 5.0,
+      "step": 22000,
+      "total_flos": 61272169512960.0,
+      "train_loss": 0.03695063419647406,
+      "train_runtime": 26285.5206,
+      "train_samples_per_second": 1.674,
+      "train_steps_per_second": 0.837
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 22000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 1100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 61272169512960.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}