diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,21742 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.993958920660491,
+  "eval_steps": 500,
+  "global_step": 3100,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0016109544905356424,
+      "grad_norm": 6.032404008426598,
+      "learning_rate": 1.2903225806451614e-07,
+      "loss": 0.8102,
+      "step": 1
+    },
+    {
+      "epoch": 0.003221908981071285,
+      "grad_norm": 6.091224660092794,
+      "learning_rate": 2.580645161290323e-07,
+      "loss": 0.8201,
+      "step": 2
+    },
+    {
+      "epoch": 0.004832863471606927,
+      "grad_norm": 6.242173191014313,
+      "learning_rate": 3.870967741935484e-07,
+      "loss": 0.8364,
+      "step": 3
+    },
+    {
+      "epoch": 0.00644381796214257,
+      "grad_norm": 5.9113732582345815,
+      "learning_rate": 5.161290322580646e-07,
+      "loss": 0.8033,
+      "step": 4
+    },
+    {
+      "epoch": 0.008054772452678213,
+      "grad_norm": 5.836423086831506,
+      "learning_rate": 6.451612903225807e-07,
+      "loss": 0.7877,
+      "step": 5
+    },
+    {
+      "epoch": 0.009665726943213854,
+      "grad_norm": 5.755126086223614,
+      "learning_rate": 7.741935483870968e-07,
+      "loss": 0.7791,
+      "step": 6
+    },
+    {
+      "epoch": 0.011276681433749497,
+      "grad_norm": 5.649783806242555,
+      "learning_rate": 9.032258064516129e-07,
+      "loss": 0.7931,
+      "step": 7
+    },
+    {
+      "epoch": 0.01288763592428514,
+      "grad_norm": 5.479452292994705,
+      "learning_rate": 1.0322580645161291e-06,
+      "loss": 0.7854,
+      "step": 8
+    },
+    {
+      "epoch": 0.01449859041482078,
+      "grad_norm": 5.366601689873154,
+      "learning_rate": 1.1612903225806454e-06,
+      "loss": 0.7727,
+      "step": 9
+    },
+    {
+      "epoch": 0.016109544905356425,
+      "grad_norm": 4.434943674510098,
+      "learning_rate": 1.2903225806451614e-06,
+      "loss": 0.7384,
+      "step": 10
+    },
+    {
+      "epoch": 0.017720499395892066,
+      "grad_norm": 4.293198637826971,
+      "learning_rate": 1.4193548387096776e-06,
+      "loss": 0.7394,
+      "step": 11
+    },
+    {
+      "epoch": 0.019331453886427707,
+      "grad_norm": 4.083095267873627,
+      "learning_rate": 1.5483870967741937e-06,
+      "loss": 0.74,
+      "step": 12
+    },
+    {
+      "epoch": 0.020942408376963352,
+      "grad_norm": 2.4257793361125777,
+      "learning_rate": 1.67741935483871e-06,
+      "loss": 0.6943,
+      "step": 13
+    },
+    {
+      "epoch": 0.022553362867498993,
+      "grad_norm": 2.404485912753832,
+      "learning_rate": 1.8064516129032258e-06,
+      "loss": 0.7221,
+      "step": 14
+    },
+    {
+      "epoch": 0.024164317358034634,
+      "grad_norm": 2.1157005039902415,
+      "learning_rate": 1.935483870967742e-06,
+      "loss": 0.7026,
+      "step": 15
+    },
+    {
+      "epoch": 0.02577527184857028,
+      "grad_norm": 1.9329284071198143,
+      "learning_rate": 2.0645161290322582e-06,
+      "loss": 0.6697,
+      "step": 16
+    },
+    {
+      "epoch": 0.02738622633910592,
+      "grad_norm": 1.856103169273838,
+      "learning_rate": 2.1935483870967745e-06,
+      "loss": 0.672,
+      "step": 17
+    },
+    {
+      "epoch": 0.02899718082964156,
+      "grad_norm": 3.1117388382933022,
+      "learning_rate": 2.3225806451612907e-06,
+      "loss": 0.6815,
+      "step": 18
+    },
+    {
+      "epoch": 0.030608135320177206,
+      "grad_norm": 3.2415743455441484,
+      "learning_rate": 2.4516129032258066e-06,
+      "loss": 0.6679,
+      "step": 19
+    },
+    {
+      "epoch": 0.03221908981071285,
+      "grad_norm": 3.20677352745239,
+      "learning_rate": 2.580645161290323e-06,
+      "loss": 0.6551,
+      "step": 20
+    },
+    {
+      "epoch": 0.03383004430124849,
+      "grad_norm": 2.823762323190898,
+      "learning_rate": 2.709677419354839e-06,
+      "loss": 0.6319,
+      "step": 21
+    },
+    {
+      "epoch": 0.03544099879178413,
+      "grad_norm": 2.6726511366702796,
+      "learning_rate": 2.8387096774193553e-06,
+      "loss": 0.6458,
+      "step": 22
+    },
+    {
+      "epoch": 0.03705195328231978,
+      "grad_norm": 2.2046688512156916,
+      "learning_rate": 2.967741935483871e-06,
+      "loss": 0.6299,
+      "step": 23
+    },
+    {
+      "epoch": 0.038662907772855415,
+      "grad_norm": 1.6050271766621365,
+      "learning_rate": 3.0967741935483874e-06,
+      "loss": 0.6186,
+      "step": 24
+    },
+    {
+      "epoch": 0.04027386226339106,
+      "grad_norm": 1.3573669680309588,
+      "learning_rate": 3.225806451612903e-06,
+      "loss": 0.5967,
+      "step": 25
+    },
+    {
+      "epoch": 0.041884816753926704,
+      "grad_norm": 1.1627320193357804,
+      "learning_rate": 3.35483870967742e-06,
+      "loss": 0.6048,
+      "step": 26
+    },
+    {
+      "epoch": 0.04349577124446234,
+      "grad_norm": 0.9897690073046419,
+      "learning_rate": 3.4838709677419357e-06,
+      "loss": 0.5834,
+      "step": 27
+    },
+    {
+      "epoch": 0.045106725734997986,
+      "grad_norm": 0.9903767157665542,
+      "learning_rate": 3.6129032258064515e-06,
+      "loss": 0.5815,
+      "step": 28
+    },
+    {
+      "epoch": 0.04671768022553363,
+      "grad_norm": 1.1368680688613164,
+      "learning_rate": 3.741935483870968e-06,
+      "loss": 0.5707,
+      "step": 29
+    },
+    {
+      "epoch": 0.04832863471606927,
+      "grad_norm": 1.0426684828392754,
+      "learning_rate": 3.870967741935484e-06,
+      "loss": 0.5461,
+      "step": 30
+    },
+    {
+      "epoch": 0.04993958920660491,
+      "grad_norm": 0.9337404457709985,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.5549,
+      "step": 31
+    },
+    {
+      "epoch": 0.05155054369714056,
+      "grad_norm": 0.7242627578457423,
+      "learning_rate": 4.1290322580645165e-06,
+      "loss": 0.5507,
+      "step": 32
+    },
+    {
+      "epoch": 0.053161498187676196,
+      "grad_norm": 0.6685301803699689,
+      "learning_rate": 4.258064516129032e-06,
+      "loss": 0.5403,
+      "step": 33
+    },
+    {
+      "epoch": 0.05477245267821184,
+      "grad_norm": 0.6968371169058664,
+      "learning_rate": 4.387096774193549e-06,
+      "loss": 0.5268,
+      "step": 34
+    },
+    {
+      "epoch": 0.056383407168747485,
+      "grad_norm": 0.8728817789992291,
+      "learning_rate": 4.516129032258065e-06,
+      "loss": 0.5347,
+      "step": 35
+    },
+    {
+      "epoch": 0.05799436165928312,
+      "grad_norm": 0.8602530490558115,
+      "learning_rate": 4.6451612903225815e-06,
+      "loss": 0.5462,
+      "step": 36
+    },
+    {
+      "epoch": 0.05960531614981877,
+      "grad_norm": 0.6378392806884997,
+      "learning_rate": 4.774193548387097e-06,
+      "loss": 0.5199,
+      "step": 37
+    },
+    {
+      "epoch": 0.06121627064035441,
+      "grad_norm": 0.5998483205329147,
+      "learning_rate": 4.903225806451613e-06,
+      "loss": 0.5378,
+      "step": 38
+    },
+    {
+      "epoch": 0.06282722513089005,
+      "grad_norm": 0.6059446636403915,
+      "learning_rate": 5.032258064516129e-06,
+      "loss": 0.5161,
+      "step": 39
+    },
+    {
+      "epoch": 0.0644381796214257,
+      "grad_norm": 0.6817404761352738,
+      "learning_rate": 5.161290322580646e-06,
+      "loss": 0.5194,
+      "step": 40
+    },
+    {
+      "epoch": 0.06604913411196134,
+      "grad_norm": 0.6331844245949004,
+      "learning_rate": 5.290322580645162e-06,
+      "loss": 0.5384,
+      "step": 41
+    },
+    {
+      "epoch": 0.06766008860249698,
+      "grad_norm": 0.4948231352853153,
+      "learning_rate": 5.419354838709678e-06,
+      "loss": 0.5169,
+      "step": 42
+    },
+    {
+      "epoch": 0.06927104309303263,
+      "grad_norm": 0.5035080428321753,
+      "learning_rate": 5.548387096774194e-06,
+      "loss": 0.5135,
+      "step": 43
+    },
+    {
+      "epoch": 0.07088199758356827,
+      "grad_norm": 0.536385019533745,
+      "learning_rate": 5.677419354838711e-06,
+      "loss": 0.4819,
+      "step": 44
+    },
+    {
+      "epoch": 0.0724929520741039,
+      "grad_norm": 0.5872680433979439,
+      "learning_rate": 5.806451612903226e-06,
+      "loss": 0.5404,
+      "step": 45
+    },
+    {
+      "epoch": 0.07410390656463955,
+      "grad_norm": 0.553950328048959,
+      "learning_rate": 5.935483870967742e-06,
+      "loss": 0.5012,
+      "step": 46
+    },
+    {
+      "epoch": 0.07571486105517519,
+      "grad_norm": 0.46649858895580004,
+      "learning_rate": 6.064516129032259e-06,
+      "loss": 0.5084,
+      "step": 47
+    },
+    {
+      "epoch": 0.07732581554571083,
+      "grad_norm": 0.49980550410748753,
+      "learning_rate": 6.193548387096775e-06,
+      "loss": 0.4997,
+      "step": 48
+    },
+    {
+      "epoch": 0.07893677003624648,
+      "grad_norm": 0.500125771445445,
+      "learning_rate": 6.3225806451612906e-06,
+      "loss": 0.5211,
+      "step": 49
+    },
+    {
+      "epoch": 0.08054772452678212,
+      "grad_norm": 0.43794258077350834,
+      "learning_rate": 6.451612903225806e-06,
+      "loss": 0.5019,
+      "step": 50
+    },
+    {
+      "epoch": 0.08215867901731776,
+      "grad_norm": 0.41542158356072195,
+      "learning_rate": 6.580645161290323e-06,
+      "loss": 0.5,
+      "step": 51
+    },
+    {
+      "epoch": 0.08376963350785341,
+      "grad_norm": 0.4398106247413648,
+      "learning_rate": 6.70967741935484e-06,
+      "loss": 0.482,
+      "step": 52
+    },
+    {
+      "epoch": 0.08538058799838905,
+      "grad_norm": 0.4658104909549922,
+      "learning_rate": 6.838709677419355e-06,
+      "loss": 0.51,
+      "step": 53
+    },
+    {
+      "epoch": 0.08699154248892468,
+      "grad_norm": 0.44586970110991897,
+      "learning_rate": 6.967741935483871e-06,
+      "loss": 0.5066,
+      "step": 54
+    },
+    {
+      "epoch": 0.08860249697946034,
+      "grad_norm": 0.36952288687065493,
+      "learning_rate": 7.096774193548388e-06,
+      "loss": 0.4868,
+      "step": 55
+    },
+    {
+      "epoch": 0.09021345146999597,
+      "grad_norm": 0.37449257402238395,
+      "learning_rate": 7.225806451612903e-06,
+      "loss": 0.4874,
+      "step": 56
+    },
+    {
+      "epoch": 0.09182440596053161,
+      "grad_norm": 0.43146034558081636,
+      "learning_rate": 7.35483870967742e-06,
+      "loss": 0.4961,
+      "step": 57
+    },
+    {
+      "epoch": 0.09343536045106726,
+      "grad_norm": 0.33924473498486496,
+      "learning_rate": 7.483870967741936e-06,
+      "loss": 0.4659,
+      "step": 58
+    },
+    {
+      "epoch": 0.0950463149416029,
+      "grad_norm": 0.3650384114275885,
+      "learning_rate": 7.612903225806451e-06,
+      "loss": 0.4922,
+      "step": 59
+    },
+    {
+      "epoch": 0.09665726943213854,
+      "grad_norm": 0.42392175050797987,
+      "learning_rate": 7.741935483870968e-06,
+      "loss": 0.4825,
+      "step": 60
+    },
+    {
+      "epoch": 0.09826822392267419,
+      "grad_norm": 0.41566274005192855,
+      "learning_rate": 7.870967741935484e-06,
+      "loss": 0.496,
+      "step": 61
+    },
+    {
+      "epoch": 0.09987917841320983,
+      "grad_norm": 0.3532541757044527,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.4836,
+      "step": 62
+    },
+    {
+      "epoch": 0.10149013290374546,
+      "grad_norm": 0.38676034561653744,
+      "learning_rate": 8.129032258064517e-06,
+      "loss": 0.4617,
+      "step": 63
+    },
+    {
+      "epoch": 0.10310108739428112,
+      "grad_norm": 0.39235511829919395,
+      "learning_rate": 8.258064516129033e-06,
+      "loss": 0.4699,
+      "step": 64
+    },
+    {
+      "epoch": 0.10471204188481675,
+      "grad_norm": 0.42853549136039776,
+      "learning_rate": 8.387096774193549e-06,
+      "loss": 0.4967,
+      "step": 65
+    },
+    {
+      "epoch": 0.10632299637535239,
+      "grad_norm": 0.4006077949721882,
+      "learning_rate": 8.516129032258065e-06,
+      "loss": 0.4745,
+      "step": 66
+    },
+    {
+      "epoch": 0.10793395086588804,
+      "grad_norm": 0.365858425655604,
+      "learning_rate": 8.64516129032258e-06,
+      "loss": 0.486,
+      "step": 67
+    },
+    {
+      "epoch": 0.10954490535642368,
+      "grad_norm": 0.4029132008489552,
+      "learning_rate": 8.774193548387098e-06,
+      "loss": 0.4578,
+      "step": 68
+    },
+    {
+      "epoch": 0.11115585984695932,
+      "grad_norm": 0.4416121827528027,
+      "learning_rate": 8.903225806451614e-06,
+      "loss": 0.4838,
+      "step": 69
+    },
+    {
+      "epoch": 0.11276681433749497,
+      "grad_norm": 0.37290691713598234,
+      "learning_rate": 9.03225806451613e-06,
+      "loss": 0.4626,
+      "step": 70
+    },
+    {
+      "epoch": 0.11437776882803061,
+      "grad_norm": 0.39437106054528326,
+      "learning_rate": 9.161290322580645e-06,
+      "loss": 0.4606,
+      "step": 71
+    },
+    {
+      "epoch": 0.11598872331856624,
+      "grad_norm": 0.46230012352472577,
+      "learning_rate": 9.290322580645163e-06,
+      "loss": 0.4843,
+      "step": 72
+    },
+    {
+      "epoch": 0.1175996778091019,
+      "grad_norm": 0.36168746091549053,
+      "learning_rate": 9.419354838709677e-06,
+      "loss": 0.4776,
+      "step": 73
+    },
+    {
+      "epoch": 0.11921063229963753,
+      "grad_norm": 0.3887609362159801,
+      "learning_rate": 9.548387096774195e-06,
+      "loss": 0.4913,
+      "step": 74
+    },
+    {
+      "epoch": 0.12082158679017317,
+      "grad_norm": 0.4212414686324945,
+      "learning_rate": 9.67741935483871e-06,
+      "loss": 0.472,
+      "step": 75
+    },
+    {
+      "epoch": 0.12243254128070882,
+      "grad_norm": 0.3736240767300341,
+      "learning_rate": 9.806451612903226e-06,
+      "loss": 0.463,
+      "step": 76
+    },
+    {
+      "epoch": 0.12404349577124446,
+      "grad_norm": 0.3897026748050938,
+      "learning_rate": 9.935483870967742e-06,
+      "loss": 0.4869,
+      "step": 77
+    },
+    {
+      "epoch": 0.1256544502617801,
+      "grad_norm": 0.35819030543571656,
+      "learning_rate": 1.0064516129032258e-05,
+      "loss": 0.4821,
+      "step": 78
+    },
+    {
+      "epoch": 0.12726540475231574,
+      "grad_norm": 0.419675735054713,
+      "learning_rate": 1.0193548387096774e-05,
+      "loss": 0.4572,
+      "step": 79
+    },
+    {
+      "epoch": 0.1288763592428514,
+      "grad_norm": 0.45017621890591875,
+      "learning_rate": 1.0322580645161291e-05,
+      "loss": 0.4519,
+      "step": 80
+    },
+    {
+      "epoch": 0.13048731373338704,
+      "grad_norm": 0.38873544789688036,
+      "learning_rate": 1.0451612903225807e-05,
+      "loss": 0.4704,
+      "step": 81
+    },
+    {
+      "epoch": 0.13209826822392268,
+      "grad_norm": 0.4590008765481535,
+      "learning_rate": 1.0580645161290325e-05,
+      "loss": 0.4624,
+      "step": 82
+    },
+    {
+      "epoch": 0.13370922271445831,
+      "grad_norm": 0.4017641572848416,
+      "learning_rate": 1.070967741935484e-05,
+      "loss": 0.4552,
+      "step": 83
+    },
+    {
+      "epoch": 0.13532017720499395,
+      "grad_norm": 0.4919488749370493,
+      "learning_rate": 1.0838709677419356e-05,
+      "loss": 0.4824,
+      "step": 84
+    },
+    {
+      "epoch": 0.1369311316955296,
+      "grad_norm": 0.46571573865087124,
+      "learning_rate": 1.096774193548387e-05,
+      "loss": 0.4593,
+      "step": 85
+    },
+    {
+      "epoch": 0.13854208618606526,
+      "grad_norm": 0.483375616479342,
+      "learning_rate": 1.1096774193548388e-05,
+      "loss": 0.4762,
+      "step": 86
+    },
+    {
+      "epoch": 0.1401530406766009,
+      "grad_norm": 0.390666645368303,
+      "learning_rate": 1.1225806451612904e-05,
+      "loss": 0.4543,
+      "step": 87
+    },
+    {
+      "epoch": 0.14176399516713653,
+      "grad_norm": 0.5003728860520759,
+      "learning_rate": 1.1354838709677421e-05,
+      "loss": 0.4466,
+      "step": 88
+    },
+    {
+      "epoch": 0.14337494965767217,
+      "grad_norm": 0.4761849607513898,
+      "learning_rate": 1.1483870967741937e-05,
+      "loss": 0.4744,
+      "step": 89
+    },
+    {
+      "epoch": 0.1449859041482078,
+      "grad_norm": 0.4431212670764865,
+      "learning_rate": 1.1612903225806453e-05,
+      "loss": 0.4785,
+      "step": 90
+    },
+    {
+      "epoch": 0.14659685863874344,
+      "grad_norm": 0.4645484282151763,
+      "learning_rate": 1.1741935483870967e-05,
+      "loss": 0.4747,
+      "step": 91
+    },
+    {
+      "epoch": 0.1482078131292791,
+      "grad_norm": 0.4591410377370998,
+      "learning_rate": 1.1870967741935484e-05,
+      "loss": 0.4501,
+      "step": 92
+    },
+    {
+      "epoch": 0.14981876761981475,
+      "grad_norm": 0.43211703633867105,
+      "learning_rate": 1.2e-05,
+      "loss": 0.4318,
+      "step": 93
+    },
+    {
+      "epoch": 0.15142972211035038,
+      "grad_norm": 0.4134987376570425,
+      "learning_rate": 1.2129032258064518e-05,
+      "loss": 0.4554,
+      "step": 94
+    },
+    {
+      "epoch": 0.15304067660088602,
+      "grad_norm": 0.40687971876636914,
+      "learning_rate": 1.2258064516129034e-05,
+      "loss": 0.4463,
+      "step": 95
+    },
+    {
+      "epoch": 0.15465163109142166,
+      "grad_norm": 0.4821367578390539,
+      "learning_rate": 1.238709677419355e-05,
+      "loss": 0.4684,
+      "step": 96
+    },
+    {
+      "epoch": 0.1562625855819573,
+      "grad_norm": 0.4008872991192848,
+      "learning_rate": 1.2516129032258067e-05,
+      "loss": 0.4587,
+      "step": 97
+    },
+    {
+      "epoch": 0.15787354007249296,
+      "grad_norm": 0.4709306605581316,
+      "learning_rate": 1.2645161290322581e-05,
+      "loss": 0.4669,
+      "step": 98
+    },
+    {
+      "epoch": 0.1594844945630286,
+      "grad_norm": 0.4270863547298053,
+      "learning_rate": 1.2774193548387097e-05,
+      "loss": 0.453,
+      "step": 99
+    },
+    {
+      "epoch": 0.16109544905356424,
+      "grad_norm": 0.47904696715631884,
+      "learning_rate": 1.2903225806451613e-05,
+      "loss": 0.4331,
+      "step": 100
+    },
+    {
+      "epoch": 0.16270640354409988,
+      "grad_norm": 0.5646915331058875,
+      "learning_rate": 1.303225806451613e-05,
+      "loss": 0.4647,
+      "step": 101
+    },
+    {
+      "epoch": 0.1643173580346355,
+      "grad_norm": 0.421375992229102,
+      "learning_rate": 1.3161290322580646e-05,
+      "loss": 0.4628,
+      "step": 102
+    },
+    {
+      "epoch": 0.16592831252517115,
+      "grad_norm": 0.4577925255478774,
+      "learning_rate": 1.3290322580645164e-05,
+      "loss": 0.4432,
+      "step": 103
+    },
+    {
+      "epoch": 0.16753926701570682,
+      "grad_norm": 0.5080123906152704,
+      "learning_rate": 1.341935483870968e-05,
+      "loss": 0.4581,
+      "step": 104
+    },
+    {
+      "epoch": 0.16915022150624245,
+      "grad_norm": 0.4302757240951249,
+      "learning_rate": 1.3548387096774194e-05,
+      "loss": 0.4252,
+      "step": 105
+    },
+    {
+      "epoch": 0.1707611759967781,
+      "grad_norm": 0.543321933222442,
+      "learning_rate": 1.367741935483871e-05,
+      "loss": 0.4556,
+      "step": 106
+    },
+    {
+      "epoch": 0.17237213048731373,
+      "grad_norm": 0.4342499733326806,
+      "learning_rate": 1.3806451612903227e-05,
+      "loss": 0.4474,
+      "step": 107
+    },
+    {
+      "epoch": 0.17398308497784937,
+      "grad_norm": 0.625382634227691,
+      "learning_rate": 1.3935483870967743e-05,
+      "loss": 0.4262,
+      "step": 108
+    },
+    {
+      "epoch": 0.175594039468385,
+      "grad_norm": 0.4889154545842034,
+      "learning_rate": 1.406451612903226e-05,
+      "loss": 0.4633,
+      "step": 109
+    },
+    {
+      "epoch": 0.17720499395892067,
+      "grad_norm": 0.468374811877519,
+      "learning_rate": 1.4193548387096776e-05,
+      "loss": 0.4685,
+      "step": 110
+    },
+    {
+      "epoch": 0.1788159484494563,
+      "grad_norm": 0.5750906393209809,
+      "learning_rate": 1.4322580645161292e-05,
+      "loss": 0.4521,
+      "step": 111
+    },
+    {
+      "epoch": 0.18042690293999195,
+      "grad_norm": 0.43929194658608134,
+      "learning_rate": 1.4451612903225806e-05,
+      "loss": 0.458,
+      "step": 112
+    },
+    {
+      "epoch": 0.18203785743052758,
+      "grad_norm": 0.5255857904785238,
+      "learning_rate": 1.4580645161290324e-05,
+      "loss": 0.4354,
+      "step": 113
+    },
+    {
+      "epoch": 0.18364881192106322,
+      "grad_norm": 0.5697953246124879,
+      "learning_rate": 1.470967741935484e-05,
+      "loss": 0.4597,
+      "step": 114
+    },
+    {
+      "epoch": 0.18525976641159886,
+      "grad_norm": 0.684779610893239,
+      "learning_rate": 1.4838709677419357e-05,
+      "loss": 0.4513,
+      "step": 115
+    },
+    {
+      "epoch": 0.18687072090213452,
+      "grad_norm": 0.7663595177962778,
+      "learning_rate": 1.4967741935483873e-05,
+      "loss": 0.4501,
+      "step": 116
+    },
+    {
+      "epoch": 0.18848167539267016,
+      "grad_norm": 0.6231547313160818,
+      "learning_rate": 1.5096774193548389e-05,
+      "loss": 0.4654,
+      "step": 117
+    },
+    {
+      "epoch": 0.1900926298832058,
+      "grad_norm": 0.762366532637053,
+      "learning_rate": 1.5225806451612903e-05,
+      "loss": 0.4379,
+      "step": 118
+    },
+    {
+      "epoch": 0.19170358437374144,
+      "grad_norm": 0.6014548144293846,
+      "learning_rate": 1.535483870967742e-05,
+      "loss": 0.4433,
+      "step": 119
+    },
+    {
+      "epoch": 0.19331453886427707,
+      "grad_norm": 0.5976102150815602,
+      "learning_rate": 1.5483870967741936e-05,
+      "loss": 0.4412,
+      "step": 120
+    },
+    {
+      "epoch": 0.19492549335481274,
+      "grad_norm": 0.7203469874012064,
+      "learning_rate": 1.5612903225806454e-05,
+      "loss": 0.4746,
+      "step": 121
+    },
+    {
+      "epoch": 0.19653644784534838,
+      "grad_norm": 0.7853276912774007,
+      "learning_rate": 1.5741935483870968e-05,
+      "loss": 0.457,
+      "step": 122
+    },
+    {
+      "epoch": 0.19814740233588402,
+      "grad_norm": 0.5932744430290895,
+      "learning_rate": 1.5870967741935485e-05,
+      "loss": 0.4576,
+      "step": 123
+    },
+    {
+      "epoch": 0.19975835682641965,
+      "grad_norm": 0.7267109233788118,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.4467,
+      "step": 124
+    },
+    {
+      "epoch": 0.2013693113169553,
+      "grad_norm": 0.5406828652336474,
+      "learning_rate": 1.6129032258064517e-05,
+      "loss": 0.4462,
+      "step": 125
+    },
+    {
+      "epoch": 0.20298026580749093,
+      "grad_norm": 0.5858021844743821,
+      "learning_rate": 1.6258064516129034e-05,
+      "loss": 0.4626,
+      "step": 126
+    },
+    {
+      "epoch": 0.2045912202980266,
+      "grad_norm": 0.5877766270343217,
+      "learning_rate": 1.638709677419355e-05,
+      "loss": 0.455,
+      "step": 127
+    },
+    {
+      "epoch": 0.20620217478856223,
+      "grad_norm": 0.47863537726359784,
+      "learning_rate": 1.6516129032258066e-05,
+      "loss": 0.4539,
+      "step": 128
+    },
+    {
+      "epoch": 0.20781312927909787,
+      "grad_norm": 0.439070948766706,
+      "learning_rate": 1.6645161290322583e-05,
+      "loss": 0.4443,
+      "step": 129
+    },
+    {
+      "epoch": 0.2094240837696335,
+      "grad_norm": 0.5181283081353542,
+      "learning_rate": 1.6774193548387098e-05,
+      "loss": 0.4522,
+      "step": 130
+    },
+    {
+      "epoch": 0.21103503826016914,
+      "grad_norm": 0.4577673620554301,
+      "learning_rate": 1.6903225806451615e-05,
+      "loss": 0.479,
+      "step": 131
+    },
+    {
+      "epoch": 0.21264599275070478,
+      "grad_norm": 0.4523908387978445,
+      "learning_rate": 1.703225806451613e-05,
+      "loss": 0.4456,
+      "step": 132
+    },
+    {
+      "epoch": 0.21425694724124045,
+      "grad_norm": 0.5290727008449524,
+      "learning_rate": 1.7161290322580647e-05,
+      "loss": 0.4221,
+      "step": 133
+    },
+    {
+      "epoch": 0.21586790173177609,
+      "grad_norm": 0.4980372169033664,
+      "learning_rate": 1.729032258064516e-05,
+      "loss": 0.4563,
+      "step": 134
+    },
+    {
+      "epoch": 0.21747885622231172,
+      "grad_norm": 0.48606495543694916,
+      "learning_rate": 1.741935483870968e-05,
+      "loss": 0.4324,
+      "step": 135
+    },
+    {
+      "epoch": 0.21908981071284736,
+      "grad_norm": 0.6957991032575429,
+      "learning_rate": 1.7548387096774196e-05,
+      "loss": 0.4722,
+      "step": 136
+    },
+    {
+      "epoch": 0.220700765203383,
+      "grad_norm": 0.4643809416764509,
+      "learning_rate": 1.7677419354838713e-05,
+      "loss": 0.4521,
+      "step": 137
+    },
+    {
+      "epoch": 0.22231171969391864,
+      "grad_norm": 0.6664029651503106,
+      "learning_rate": 1.7806451612903228e-05,
+      "loss": 0.4481,
+      "step": 138
+    },
+    {
+      "epoch": 0.2239226741844543,
+      "grad_norm": 0.5260601890755087,
+      "learning_rate": 1.7935483870967742e-05,
+      "loss": 0.4525,
+      "step": 139
+    },
+    {
+      "epoch": 0.22553362867498994,
+      "grad_norm": 0.5161740937718994,
+      "learning_rate": 1.806451612903226e-05,
+      "loss": 0.4455,
+      "step": 140
+    },
+    {
+      "epoch": 0.22714458316552558,
+      "grad_norm": 0.5760986430160835,
+      "learning_rate": 1.8193548387096777e-05,
+      "loss": 0.4582,
+      "step": 141
+    },
+    {
+      "epoch": 0.22875553765606121,
+      "grad_norm": 0.44639939244600985,
+      "learning_rate": 1.832258064516129e-05,
+      "loss": 0.4296,
+      "step": 142
+    },
+    {
+      "epoch": 0.23036649214659685,
+      "grad_norm": 0.4882062915992461,
+      "learning_rate": 1.845161290322581e-05,
+      "loss": 0.446,
+      "step": 143
+    },
+    {
+      "epoch": 0.2319774466371325,
+      "grad_norm": 0.4896222422614565,
+      "learning_rate": 1.8580645161290326e-05,
+      "loss": 0.4093,
+      "step": 144
+    },
+    {
+      "epoch": 0.23358840112766815,
+      "grad_norm": 0.5601501144229972,
+      "learning_rate": 1.870967741935484e-05,
+      "loss": 0.4505,
+      "step": 145
+    },
+    {
+      "epoch": 0.2351993556182038,
+      "grad_norm": 0.5437487520198745,
+      "learning_rate": 1.8838709677419354e-05,
+      "loss": 0.4322,
+      "step": 146
+    },
+    {
+      "epoch": 0.23681031010873943,
+      "grad_norm": 0.526186951905453,
+      "learning_rate": 1.896774193548387e-05,
+      "loss": 0.4521,
+      "step": 147
+    },
+    {
+      "epoch": 0.23842126459927507,
+      "grad_norm": 0.5667856097490956,
+      "learning_rate": 1.909677419354839e-05,
+      "loss": 0.4271,
+      "step": 148
+    },
+    {
+      "epoch": 0.2400322190898107,
+      "grad_norm": 0.49095301330200275,
+      "learning_rate": 1.9225806451612907e-05,
+      "loss": 0.4253,
+      "step": 149
+    },
+    {
+      "epoch": 0.24164317358034634,
+      "grad_norm": 0.5980300594255382,
+      "learning_rate": 1.935483870967742e-05,
+      "loss": 0.4244,
+      "step": 150
+    },
+    {
+      "epoch": 0.243254128070882,
+      "grad_norm": 0.47376445836894726,
+      "learning_rate": 1.948387096774194e-05,
+      "loss": 0.4372,
+      "step": 151
+    },
+    {
+      "epoch": 0.24486508256141765,
+      "grad_norm": 0.5782178587487257,
+      "learning_rate": 1.9612903225806452e-05,
+      "loss": 0.4616,
+      "step": 152
+    },
+    {
+      "epoch": 0.24647603705195328,
+      "grad_norm": 0.6641367840166066,
+      "learning_rate": 1.974193548387097e-05,
+      "loss": 0.4373,
+      "step": 153
+    },
+    {
+      "epoch": 0.24808699154248892,
+      "grad_norm": 0.4915152452969149,
+      "learning_rate": 1.9870967741935484e-05,
+      "loss": 0.4375,
+      "step": 154
+    },
+    {
+      "epoch": 0.24969794603302456,
+      "grad_norm": 0.4715056217318344,
+      "learning_rate": 2e-05,
+      "loss": 0.4602,
+      "step": 155
+    },
+    {
+      "epoch": 0.2513089005235602,
+      "grad_norm": 0.4588234482596573,
+      "learning_rate": 2.0129032258064516e-05,
+      "loss": 0.4364,
+      "step": 156
+    },
+    {
+      "epoch": 0.25291985501409586,
+      "grad_norm": 0.49507022488662084,
+      "learning_rate": 2.0258064516129033e-05,
+      "loss": 0.4328,
+      "step": 157
+    },
+    {
+      "epoch": 0.2545308095046315,
+      "grad_norm": 0.475753423255671,
+      "learning_rate": 2.0387096774193547e-05,
+      "loss": 0.428,
+      "step": 158
+    },
+    {
+      "epoch": 0.25614176399516714,
+      "grad_norm": 0.5183928270573065,
+      "learning_rate": 2.051612903225807e-05,
+      "loss": 0.4392,
+      "step": 159
+    },
+    {
+      "epoch": 0.2577527184857028,
+      "grad_norm": 0.4774005808982356,
+      "learning_rate": 2.0645161290322582e-05,
+      "loss": 0.4428,
+      "step": 160
+    },
+    {
+      "epoch": 0.2593636729762384,
+      "grad_norm": 0.4572472041520854,
+      "learning_rate": 2.07741935483871e-05,
+      "loss": 0.4634,
+      "step": 161
+    },
+    {
+      "epoch": 0.2609746274667741,
+      "grad_norm": 0.6081326727542029,
+      "learning_rate": 2.0903225806451614e-05,
+      "loss": 0.428,
+      "step": 162
+    },
+    {
+      "epoch": 0.2625855819573097,
+      "grad_norm": 0.5023244790140038,
+      "learning_rate": 2.1032258064516128e-05,
+      "loss": 0.462,
+      "step": 163
+    },
+    {
+      "epoch": 0.26419653644784535,
+      "grad_norm": 0.5479494420185165,
+      "learning_rate": 2.116129032258065e-05,
+      "loss": 0.4469,
+      "step": 164
+    },
+    {
+      "epoch": 0.26580749093838096,
+      "grad_norm": 0.4584032900157975,
+      "learning_rate": 2.1290322580645163e-05,
+      "loss": 0.4477,
+      "step": 165
+    },
+    {
+      "epoch": 0.26741844542891663,
+      "grad_norm": 0.6145948964239977,
+      "learning_rate": 2.141935483870968e-05,
+      "loss": 0.4371,
+      "step": 166
+    },
+    {
+      "epoch": 0.2690293999194523,
+      "grad_norm": 0.5594470471380588,
+      "learning_rate": 2.1548387096774195e-05,
+      "loss": 0.456,
+      "step": 167
+    },
+    {
+      "epoch": 0.2706403544099879,
+      "grad_norm": 0.45972051481741444,
+      "learning_rate": 2.1677419354838712e-05,
+      "loss": 0.4391,
+      "step": 168
+    },
+    {
+      "epoch": 0.27225130890052357,
+      "grad_norm": 0.5461702197752573,
+      "learning_rate": 2.1806451612903227e-05,
+      "loss": 0.4409,
+      "step": 169
+    },
+    {
+      "epoch": 0.2738622633910592,
+      "grad_norm": 0.5906536061989351,
+      "learning_rate": 2.193548387096774e-05,
+      "loss": 0.4586,
+      "step": 170
+    },
+    {
+      "epoch": 0.27547321788159485,
+      "grad_norm": 0.599985383963263,
+      "learning_rate": 2.206451612903226e-05,
+      "loss": 0.4503,
+      "step": 171
+    },
+    {
+      "epoch": 0.2770841723721305,
+      "grad_norm": 0.5604973755851727,
+      "learning_rate": 2.2193548387096776e-05,
+      "loss": 0.442,
+      "step": 172
+    },
+    {
+      "epoch": 0.2786951268626661,
+      "grad_norm": 0.4798615338782793,
+      "learning_rate": 2.2322580645161293e-05,
+      "loss": 0.4436,
+      "step": 173
+    },
+    {
+      "epoch": 0.2803060813532018,
+      "grad_norm": 0.6325144823189365,
+      "learning_rate": 2.2451612903225807e-05,
+      "loss": 0.4439,
+      "step": 174
+    },
+    {
+      "epoch": 0.2819170358437374,
+      "grad_norm": 0.5641277931011702,
+      "learning_rate": 2.2580645161290328e-05,
+      "loss": 0.4288,
+      "step": 175
+    },
+    {
+      "epoch": 0.28352799033427306,
+      "grad_norm": 0.4953611842384428,
+      "learning_rate": 2.2709677419354842e-05,
+      "loss": 0.4449,
+      "step": 176
+    },
+    {
+      "epoch": 0.2851389448248087,
+      "grad_norm": 0.5224583314537559,
+      "learning_rate": 2.2838709677419357e-05,
+      "loss": 0.4162,
+      "step": 177
+    },
+    {
+      "epoch": 0.28674989931534434,
+      "grad_norm": 0.5001528193199506,
+      "learning_rate": 2.2967741935483874e-05,
+      "loss": 0.4393,
+      "step": 178
+    },
+    {
+      "epoch": 0.28836085380588,
+      "grad_norm": 0.48777820697444463,
+      "learning_rate": 2.3096774193548388e-05,
+      "loss": 0.4261,
+      "step": 179
+    },
+    {
+      "epoch": 0.2899718082964156,
+      "grad_norm": 0.64062502637235,
+      "learning_rate": 2.3225806451612906e-05,
+      "loss": 0.4442,
+      "step": 180
+    },
+    {
+      "epoch": 0.2915827627869513,
+      "grad_norm": 0.5683627649259488,
+      "learning_rate": 2.335483870967742e-05,
+      "loss": 0.447,
+      "step": 181
+    },
+    {
+      "epoch": 0.2931937172774869,
+      "grad_norm": 0.7696086123267061,
+      "learning_rate": 2.3483870967741934e-05,
+      "loss": 0.4219,
+      "step": 182
+    },
+    {
+      "epoch": 0.29480467176802255,
+      "grad_norm": 0.7701985708034406,
+      "learning_rate": 2.3612903225806455e-05,
+      "loss": 0.4178,
+      "step": 183
+    },
+    {
+      "epoch": 0.2964156262585582,
+      "grad_norm": 0.655838324079945,
+      "learning_rate": 2.374193548387097e-05,
+      "loss": 0.4229,
+      "step": 184
+    },
+    {
+      "epoch": 0.29802658074909383,
+      "grad_norm": 0.5294585112609449,
+      "learning_rate": 2.3870967741935486e-05,
+      "loss": 0.4374,
+      "step": 185
+    },
+    {
+      "epoch": 0.2996375352396295,
+      "grad_norm": 0.6213012416475787,
+      "learning_rate": 2.4e-05,
+      "loss": 0.4307,
+      "step": 186
+    },
+    {
+      "epoch": 0.3012484897301651,
+      "grad_norm": 0.7049316455544041,
+      "learning_rate": 2.4129032258064518e-05,
+      "loss": 0.4251,
+      "step": 187
+    },
+    {
+      "epoch": 0.30285944422070077,
+      "grad_norm": 0.5847111370562961,
+      "learning_rate": 2.4258064516129036e-05,
+      "loss": 0.4322,
+      "step": 188
+    },
+    {
+      "epoch": 0.30447039871123643,
+      "grad_norm": 0.5607890554854233,
+      "learning_rate": 2.438709677419355e-05,
+      "loss": 0.4283,
+      "step": 189
+    },
+    {
+      "epoch": 0.30608135320177204,
+      "grad_norm": 0.4807120710699517,
+      "learning_rate": 2.4516129032258067e-05,
+      "loss": 0.4321,
+      "step": 190
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 0.6290000689322379,
+      "learning_rate": 2.464516129032258e-05,
+      "loss": 0.4398,
+      "step": 191
+    },
+    {
+      "epoch": 0.3093032621828433,
+      "grad_norm": 0.5860760329948072,
+      "learning_rate": 2.47741935483871e-05,
+      "loss": 0.4239,
+      "step": 192
+    },
+    {
+      "epoch": 0.310914216673379,
+      "grad_norm": 0.5713603124420665,
+      "learning_rate": 2.4903225806451613e-05,
+      "loss": 0.4501,
+      "step": 193
+    },
+    {
+      "epoch": 0.3125251711639146,
+      "grad_norm": 0.7292654172275151,
+      "learning_rate": 2.5032258064516134e-05,
+      "loss": 0.44,
+      "step": 194
+    },
+    {
+      "epoch": 0.31413612565445026,
+      "grad_norm": 0.557304847589959,
+      "learning_rate": 2.5161290322580648e-05,
+      "loss": 0.4391,
+      "step": 195
+    },
+    {
+      "epoch": 0.3157470801449859,
+      "grad_norm": 0.8185353596554609,
+      "learning_rate": 2.5290322580645162e-05,
+      "loss": 0.4495,
+      "step": 196
+    },
+    {
+      "epoch": 0.31735803463552154,
+      "grad_norm": 0.6696922756603706,
+      "learning_rate": 2.541935483870968e-05,
+      "loss": 0.4229,
+      "step": 197
+    },
+    {
+      "epoch": 0.3189689891260572,
+      "grad_norm": 0.5867477527038533,
+      "learning_rate": 2.5548387096774194e-05,
+      "loss": 0.4281,
+      "step": 198
+    },
+    {
+      "epoch": 0.3205799436165928,
+      "grad_norm": 0.7383292533663941,
+      "learning_rate": 2.567741935483871e-05,
+      "loss": 0.4335,
+      "step": 199
+    },
+    {
+      "epoch": 0.3221908981071285,
+      "grad_norm": 0.5494545052520469,
+      "learning_rate": 2.5806451612903226e-05,
+      "loss": 0.4123,
+      "step": 200
+    },
+    {
+      "epoch": 0.32380185259766414,
+      "grad_norm": 0.787263534027303,
+      "learning_rate": 2.5935483870967746e-05,
+      "loss": 0.4334,
+      "step": 201
+    },
+    {
+      "epoch": 0.32541280708819975,
+      "grad_norm": 0.671237794078286,
+      "learning_rate": 2.606451612903226e-05,
+      "loss": 0.4469,
+      "step": 202
+    },
+    {
+      "epoch": 0.3270237615787354,
+      "grad_norm": 0.6087290607998392,
+      "learning_rate": 2.6193548387096775e-05,
+      "loss": 0.4468,
+      "step": 203
+    },
+    {
+      "epoch": 0.328634716069271,
+      "grad_norm": 0.8456692623524803,
+      "learning_rate": 2.6322580645161292e-05,
+      "loss": 0.4349,
+      "step": 204
+    },
+    {
+      "epoch": 0.3302456705598067,
+      "grad_norm": 0.5789430896858069,
+      "learning_rate": 2.6451612903225806e-05,
+      "loss": 0.4386,
+      "step": 205
+    },
+    {
+      "epoch": 0.3318566250503423,
+      "grad_norm": 0.5227774356619872,
+      "learning_rate": 2.6580645161290327e-05,
+      "loss": 0.4237,
+      "step": 206
+    },
+    {
+      "epoch": 0.33346757954087797,
+      "grad_norm": 0.5151727075586936,
+      "learning_rate": 2.670967741935484e-05,
+      "loss": 0.4397,
+      "step": 207
+    },
+    {
+      "epoch": 0.33507853403141363,
+      "grad_norm": 0.48417771982658486,
+      "learning_rate": 2.683870967741936e-05,
+      "loss": 0.4479,
+      "step": 208
+    },
+    {
+      "epoch": 0.33668948852194924,
+      "grad_norm": 0.5481117200758377,
+      "learning_rate": 2.6967741935483873e-05,
+      "loss": 0.4186,
+      "step": 209
+    },
+    {
+      "epoch": 0.3383004430124849,
+      "grad_norm": 0.4448179461564387,
+      "learning_rate": 2.7096774193548387e-05,
+      "loss": 0.4084,
+      "step": 210
+    },
+    {
+      "epoch": 0.3399113975030205,
+      "grad_norm": 0.49449277894946425,
+      "learning_rate": 2.7225806451612905e-05,
+      "loss": 0.442,
+      "step": 211
+    },
+    {
+      "epoch": 0.3415223519935562,
+      "grad_norm": 0.5207685357019004,
+      "learning_rate": 2.735483870967742e-05,
+      "loss": 0.4353,
+      "step": 212
+    },
+    {
+      "epoch": 0.34313330648409185,
+      "grad_norm": 0.7050029152494538,
+      "learning_rate": 2.748387096774194e-05,
+      "loss": 0.4226,
+      "step": 213
+    },
+    {
+      "epoch": 0.34474426097462746,
+      "grad_norm": 0.6948829474376914,
+      "learning_rate": 2.7612903225806454e-05,
+      "loss": 0.4246,
+      "step": 214
+    },
+    {
+      "epoch": 0.3463552154651631,
+      "grad_norm": 0.5285793312066525,
+      "learning_rate": 2.774193548387097e-05,
+      "loss": 0.4387,
+      "step": 215
+    },
+    {
+      "epoch": 0.34796616995569873,
+      "grad_norm": 0.5892451924441414,
+      "learning_rate": 2.7870967741935485e-05,
+      "loss": 0.4455,
+      "step": 216
+    },
+    {
+      "epoch": 0.3495771244462344,
+      "grad_norm": 0.8021238678624306,
+      "learning_rate": 2.8e-05,
+      "loss": 0.425,
+      "step": 217
+    },
+    {
+      "epoch": 0.35118807893677,
+      "grad_norm": 0.5945002361956657,
+      "learning_rate": 2.812903225806452e-05,
+      "loss": 0.4286,
+      "step": 218
+    },
+    {
+      "epoch": 0.3527990334273057,
+      "grad_norm": 0.428799155792937,
+      "learning_rate": 2.8258064516129035e-05,
+      "loss": 0.408,
+      "step": 219
+    },
+    {
+      "epoch": 0.35440998791784134,
+      "grad_norm": 0.5963355738995051,
+      "learning_rate": 2.8387096774193552e-05,
+      "loss": 0.4067,
+      "step": 220
+    },
+    {
+      "epoch": 0.35602094240837695,
+      "grad_norm": 0.6365882727245175,
+      "learning_rate": 2.8516129032258066e-05,
+      "loss": 0.4408,
+      "step": 221
+    },
+    {
+      "epoch": 0.3576318968989126,
+      "grad_norm": 0.46796304829903396,
+      "learning_rate": 2.8645161290322584e-05,
+      "loss": 0.4338,
+      "step": 222
+    },
+    {
+      "epoch": 0.3592428513894482,
+      "grad_norm": 0.47402969771177833,
+      "learning_rate": 2.8774193548387098e-05,
+      "loss": 0.4128,
+      "step": 223
+    },
+    {
+      "epoch": 0.3608538058799839,
+      "grad_norm": 0.48803059781651736,
+      "learning_rate": 2.8903225806451612e-05,
+      "loss": 0.4407,
+      "step": 224
+    },
+    {
+      "epoch": 0.36246476037051956,
+      "grad_norm": 0.49604586682206475,
+      "learning_rate": 2.9032258064516133e-05,
+      "loss": 0.4232,
+      "step": 225
+    },
+    {
+      "epoch": 0.36407571486105517,
+      "grad_norm": 0.48189900699189114,
+      "learning_rate": 2.9161290322580647e-05,
+      "loss": 0.4377,
+      "step": 226
+    },
+    {
+      "epoch": 0.36568666935159083,
+      "grad_norm": 0.48989884967543235,
+      "learning_rate": 2.9290322580645165e-05,
+      "loss": 0.4182,
+      "step": 227
+    },
+    {
+      "epoch": 0.36729762384212644,
+      "grad_norm": 0.5085990736148946,
+      "learning_rate": 2.941935483870968e-05,
+      "loss": 0.4141,
+      "step": 228
+    },
+    {
+      "epoch": 0.3689085783326621,
+      "grad_norm": 0.5542764435566585,
+      "learning_rate": 2.95483870967742e-05,
+      "loss": 0.4239,
+      "step": 229
+    },
+    {
+      "epoch": 0.3705195328231977,
+      "grad_norm": 0.5509845441337561,
+      "learning_rate": 2.9677419354838714e-05,
+      "loss": 0.4327,
+      "step": 230
+    },
+    {
+      "epoch": 0.3721304873137334,
+      "grad_norm": 0.5374789273502775,
+      "learning_rate": 2.9806451612903228e-05,
+      "loss": 0.4046,
+      "step": 231
+    },
+    {
+      "epoch": 0.37374144180426905,
+      "grad_norm": 0.512328932947813,
+      "learning_rate": 2.9935483870967745e-05,
+      "loss": 0.4416,
+      "step": 232
+    },
+    {
+      "epoch": 0.37535239629480466,
+      "grad_norm": 0.5980263212180401,
+      "learning_rate": 3.006451612903226e-05,
+      "loss": 0.437,
+      "step": 233
+    },
+    {
+      "epoch": 0.3769633507853403,
+      "grad_norm": 0.4528657439497093,
+      "learning_rate": 3.0193548387096777e-05,
+      "loss": 0.4151,
+      "step": 234
+    },
+    {
+      "epoch": 0.37857430527587593,
+      "grad_norm": 0.5472677539799308,
+      "learning_rate": 3.032258064516129e-05,
+      "loss": 0.4435,
+      "step": 235
+    },
+    {
+      "epoch": 0.3801852597664116,
+      "grad_norm": 0.5284722744720981,
+      "learning_rate": 3.0451612903225805e-05,
+      "loss": 0.421,
+      "step": 236
+    },
+    {
+      "epoch": 0.38179621425694726,
+      "grad_norm": 0.5145172039518136,
+      "learning_rate": 3.0580645161290326e-05,
+      "loss": 0.4255,
+      "step": 237
+    },
+    {
+      "epoch": 0.3834071687474829,
+      "grad_norm": 0.48595473394452576,
+      "learning_rate": 3.070967741935484e-05,
+      "loss": 0.4237,
+      "step": 238
+    },
+    {
+      "epoch": 0.38501812323801854,
+      "grad_norm": 0.6578784457151443,
+      "learning_rate": 3.083870967741936e-05,
+      "loss": 0.4372,
+      "step": 239
+    },
+    {
+      "epoch": 0.38662907772855415,
+      "grad_norm": 0.6381289705882743,
+      "learning_rate": 3.096774193548387e-05,
+      "loss": 0.4405,
+      "step": 240
+    },
+    {
+      "epoch": 0.3882400322190898,
+      "grad_norm": 0.5736653750108794,
+      "learning_rate": 3.109677419354839e-05,
+      "loss": 0.4304,
+      "step": 241
+    },
+    {
+      "epoch": 0.3898509867096255,
+      "grad_norm": 0.61679678041202,
+      "learning_rate": 3.122580645161291e-05,
+      "loss": 0.4299,
+      "step": 242
+    },
+    {
+      "epoch": 0.3914619412001611,
+      "grad_norm": 0.6665472313912354,
+      "learning_rate": 3.135483870967742e-05,
+      "loss": 0.4312,
+      "step": 243
+    },
+    {
+      "epoch": 0.39307289569069676,
+      "grad_norm": 0.6298402893744898,
+      "learning_rate": 3.1483870967741935e-05,
+      "loss": 0.4161,
+      "step": 244
+    },
+    {
+      "epoch": 0.39468385018123237,
+      "grad_norm": 0.5300708226898071,
+      "learning_rate": 3.161290322580645e-05,
+      "loss": 0.4277,
+      "step": 245
+    },
+    {
+      "epoch": 0.39629480467176803,
+      "grad_norm": 0.5893160355497045,
+      "learning_rate": 3.174193548387097e-05,
+      "loss": 0.4285,
+      "step": 246
+    },
+    {
+      "epoch": 0.39790575916230364,
+      "grad_norm": 0.6980967474518777,
+      "learning_rate": 3.187096774193549e-05,
+      "loss": 0.4181,
+      "step": 247
+    },
+    {
+      "epoch": 0.3995167136528393,
+      "grad_norm": 0.5413223642212789,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 0.4269,
+      "step": 248
+    },
+    {
+      "epoch": 0.40112766814337497,
+      "grad_norm": 0.40290216945992247,
+      "learning_rate": 3.2129032258064516e-05,
+      "loss": 0.4347,
+      "step": 249
+    },
+    {
+      "epoch": 0.4027386226339106,
+      "grad_norm": 0.5063526982810828,
+      "learning_rate": 3.2258064516129034e-05,
+      "loss": 0.4387,
+      "step": 250
+    },
+    {
+      "epoch": 0.40434957712444625,
+      "grad_norm": 0.5220460662141977,
+      "learning_rate": 3.238709677419355e-05,
+      "loss": 0.4307,
+      "step": 251
+    },
+    {
+      "epoch": 0.40596053161498186,
+      "grad_norm": 0.4914288485416426,
+      "learning_rate": 3.251612903225807e-05,
+      "loss": 0.4212,
+      "step": 252
+    },
+    {
+      "epoch": 0.4075714861055175,
+      "grad_norm": 0.5074074899301333,
+      "learning_rate": 3.2645161290322586e-05,
+      "loss": 0.4247,
+      "step": 253
+    },
+    {
+      "epoch": 0.4091824405960532,
+      "grad_norm": 0.5117479204044484,
+      "learning_rate": 3.27741935483871e-05,
+      "loss": 0.4199,
+      "step": 254
+    },
+    {
+      "epoch": 0.4107933950865888,
+      "grad_norm": 0.5057623752314662,
+      "learning_rate": 3.2903225806451614e-05,
+      "loss": 0.4377,
+      "step": 255
+    },
+    {
+      "epoch": 0.41240434957712446,
+      "grad_norm": 0.45973399006678667,
+      "learning_rate": 3.303225806451613e-05,
+      "loss": 0.4294,
+      "step": 256
+    },
+    {
+      "epoch": 0.4140153040676601,
+      "grad_norm": 0.4690587970020925,
+      "learning_rate": 3.316129032258064e-05,
+      "loss": 0.43,
+      "step": 257
+    },
+    {
+      "epoch": 0.41562625855819574,
+      "grad_norm": 0.5079472609720749,
+      "learning_rate": 3.329032258064517e-05,
+      "loss": 0.4218,
+      "step": 258
+    },
+    {
+      "epoch": 0.41723721304873135,
+      "grad_norm": 0.5254404696098615,
+      "learning_rate": 3.341935483870968e-05,
+      "loss": 0.4236,
+      "step": 259
+    },
+    {
+      "epoch": 0.418848167539267,
+      "grad_norm": 0.4097556199416461,
+      "learning_rate": 3.3548387096774195e-05,
+      "loss": 0.4297,
+      "step": 260
+    },
+    {
+      "epoch": 0.4204591220298027,
+      "grad_norm": 0.5308773641962137,
+      "learning_rate": 3.367741935483871e-05,
+      "loss": 0.4144,
+      "step": 261
+    },
+    {
+      "epoch": 0.4220700765203383,
+      "grad_norm": 0.5114920055423455,
+      "learning_rate": 3.380645161290323e-05,
+      "loss": 0.4063,
+      "step": 262
+    },
+    {
+      "epoch": 0.42368103101087395,
+      "grad_norm": 0.5684133487028641,
+      "learning_rate": 3.393548387096775e-05,
+      "loss": 0.3987,
+      "step": 263
+    },
+    {
+      "epoch": 0.42529198550140956,
+      "grad_norm": 0.4879815808284312,
+      "learning_rate": 3.406451612903226e-05,
+      "loss": 0.4301,
+      "step": 264
+    },
+    {
+      "epoch": 0.42690293999194523,
+      "grad_norm": 0.5519017376318576,
+      "learning_rate": 3.4193548387096776e-05,
+      "loss": 0.4282,
+      "step": 265
+    },
+    {
+      "epoch": 0.4285138944824809,
+      "grad_norm": 0.5172565641895582,
+      "learning_rate": 3.4322580645161294e-05,
+      "loss": 0.4293,
+      "step": 266
+    },
+    {
+      "epoch": 0.4301248489730165,
+      "grad_norm": 0.5978796630371895,
+      "learning_rate": 3.445161290322581e-05,
+      "loss": 0.4256,
+      "step": 267
+    },
+    {
+      "epoch": 0.43173580346355217,
+      "grad_norm": 0.46900318232253063,
+      "learning_rate": 3.458064516129032e-05,
+      "loss": 0.4472,
+      "step": 268
+    },
+    {
+      "epoch": 0.4333467579540878,
+      "grad_norm": 0.5359562617290479,
+      "learning_rate": 3.4709677419354846e-05,
+      "loss": 0.434,
+      "step": 269
+    },
+    {
+      "epoch": 0.43495771244462345,
+      "grad_norm": 0.5467288956413452,
+      "learning_rate": 3.483870967741936e-05,
+      "loss": 0.4273,
+      "step": 270
+    },
+    {
+      "epoch": 0.43656866693515906,
+      "grad_norm": 0.5657718857551184,
+      "learning_rate": 3.4967741935483874e-05,
+      "loss": 0.4353,
+      "step": 271
+    },
+    {
+      "epoch": 0.4381796214256947,
+      "grad_norm": 0.5222662238820099,
+      "learning_rate": 3.509677419354839e-05,
+      "loss": 0.4054,
+      "step": 272
+    },
+    {
+      "epoch": 0.4397905759162304,
+      "grad_norm": 0.5104771085486108,
+      "learning_rate": 3.52258064516129e-05,
+      "loss": 0.4296,
+      "step": 273
+    },
+    {
+      "epoch": 0.441401530406766,
+      "grad_norm": 0.573976020499011,
+      "learning_rate": 3.535483870967743e-05,
+      "loss": 0.4336,
+      "step": 274
+    },
+    {
+      "epoch": 0.44301248489730166,
+      "grad_norm": 0.5651641640641581,
+      "learning_rate": 3.548387096774194e-05,
+      "loss": 0.4332,
+      "step": 275
+    },
+    {
+      "epoch": 0.44462343938783727,
+      "grad_norm": 0.4432761779379142,
+      "learning_rate": 3.5612903225806455e-05,
+      "loss": 0.4372,
+      "step": 276
+    },
+    {
+      "epoch": 0.44623439387837294,
+      "grad_norm": 0.5288954385305018,
+      "learning_rate": 3.574193548387097e-05,
+      "loss": 0.4319,
+      "step": 277
+    },
+    {
+      "epoch": 0.4478453483689086,
+      "grad_norm": 0.4467858794531326,
+      "learning_rate": 3.5870967741935483e-05,
+      "loss": 0.398,
+      "step": 278
+    },
+    {
+      "epoch": 0.4494563028594442,
+      "grad_norm": 0.4854143357523973,
+      "learning_rate": 3.6e-05,
+      "loss": 0.405,
+      "step": 279
+    },
+    {
+      "epoch": 0.4510672573499799,
+      "grad_norm": 0.47026604377994136,
+      "learning_rate": 3.612903225806452e-05,
+      "loss": 0.4406,
+      "step": 280
+    },
+    {
+      "epoch": 0.4526782118405155,
+      "grad_norm": 0.49527144296582787,
+      "learning_rate": 3.6258064516129036e-05,
+      "loss": 0.4397,
+      "step": 281
+    },
+    {
+      "epoch": 0.45428916633105115,
+      "grad_norm": 0.5999996141858859,
+      "learning_rate": 3.6387096774193553e-05,
+      "loss": 0.4234,
+      "step": 282
+    },
+    {
+      "epoch": 0.45590012082158676,
+      "grad_norm": 0.5220722098689459,
+      "learning_rate": 3.651612903225807e-05,
+      "loss": 0.4346,
+      "step": 283
+    },
+    {
+      "epoch": 0.45751107531212243,
+      "grad_norm": 0.43711076153629586,
+      "learning_rate": 3.664516129032258e-05,
+      "loss": 0.4355,
+      "step": 284
+    },
+    {
+      "epoch": 0.4591220298026581,
+      "grad_norm": 0.6172985122046,
+      "learning_rate": 3.67741935483871e-05,
+      "loss": 0.4184,
+      "step": 285
+    },
+    {
+      "epoch": 0.4607329842931937,
+      "grad_norm": 0.5867425345199105,
+      "learning_rate": 3.690322580645162e-05,
+      "loss": 0.4438,
+      "step": 286
+    },
+    {
+      "epoch": 0.46234393878372937,
+      "grad_norm": 0.4457544342019048,
+      "learning_rate": 3.7032258064516134e-05,
+      "loss": 0.4194,
+      "step": 287
+    },
+    {
+      "epoch": 0.463954893274265,
+      "grad_norm": 0.666524347966105,
+      "learning_rate": 3.716129032258065e-05,
+      "loss": 0.422,
+      "step": 288
+    },
+    {
+      "epoch": 0.46556584776480064,
+      "grad_norm": 0.48549716093426315,
+      "learning_rate": 3.729032258064516e-05,
+      "loss": 0.4163,
+      "step": 289
+    },
+    {
+      "epoch": 0.4671768022553363,
+      "grad_norm": 0.528960986822585,
+      "learning_rate": 3.741935483870968e-05,
+      "loss": 0.4244,
+      "step": 290
+    },
+    {
+      "epoch": 0.4687877567458719,
+      "grad_norm": 0.5886855824754854,
+      "learning_rate": 3.75483870967742e-05,
+      "loss": 0.4131,
+      "step": 291
+    },
+    {
+      "epoch": 0.4703987112364076,
+      "grad_norm": 0.5041181448802059,
+      "learning_rate": 3.767741935483871e-05,
+      "loss": 0.4293,
+      "step": 292
+    },
+    {
+      "epoch": 0.4720096657269432,
+      "grad_norm": 0.5833712014204475,
+      "learning_rate": 3.780645161290323e-05,
+      "loss": 0.4189,
+      "step": 293
+    },
+    {
+      "epoch": 0.47362062021747886,
+      "grad_norm": 0.6344962543797149,
+      "learning_rate": 3.793548387096774e-05,
+      "loss": 0.4148,
+      "step": 294
+    },
+    {
+      "epoch": 0.4752315747080145,
+      "grad_norm": 0.4209164542717526,
+      "learning_rate": 3.806451612903226e-05,
+      "loss": 0.3986,
+      "step": 295
+    },
+    {
+      "epoch": 0.47684252919855014,
+      "grad_norm": 0.5561556783839265,
+      "learning_rate": 3.819354838709678e-05,
+      "loss": 0.4354,
+      "step": 296
+    },
+    {
+      "epoch": 0.4784534836890858,
+      "grad_norm": 0.6257224120221472,
+      "learning_rate": 3.832258064516129e-05,
+      "loss": 0.4151,
+      "step": 297
+    },
+    {
+      "epoch": 0.4800644381796214,
+      "grad_norm": 0.6035957973374384,
+      "learning_rate": 3.8451612903225813e-05,
+      "loss": 0.4116,
+      "step": 298
+    },
+    {
+      "epoch": 0.4816753926701571,
+      "grad_norm": 0.5050188538136214,
+      "learning_rate": 3.8580645161290324e-05,
+      "loss": 0.4205,
+      "step": 299
+    },
+    {
+      "epoch": 0.4832863471606927,
+      "grad_norm": 0.4716292658982768,
+      "learning_rate": 3.870967741935484e-05,
+      "loss": 0.4277,
+      "step": 300
+    },
+    {
+      "epoch": 0.48489730165122835,
+      "grad_norm": 0.4741578168809942,
+      "learning_rate": 3.883870967741936e-05,
+      "loss": 0.4212,
+      "step": 301
+    },
+    {
+      "epoch": 0.486508256141764,
+      "grad_norm": 0.4987417243199203,
+      "learning_rate": 3.896774193548388e-05,
+      "loss": 0.433,
+      "step": 302
+    },
+    {
+      "epoch": 0.4881192106322996,
+      "grad_norm": 0.47028866203783526,
+      "learning_rate": 3.909677419354839e-05,
+      "loss": 0.4409,
+      "step": 303
+    },
+    {
+      "epoch": 0.4897301651228353,
+      "grad_norm": 0.5419362505448754,
+      "learning_rate": 3.9225806451612905e-05,
+      "loss": 0.4072,
+      "step": 304
+    },
+    {
+      "epoch": 0.4913411196133709,
+      "grad_norm": 0.6217806303253193,
+      "learning_rate": 3.935483870967742e-05,
+      "loss": 0.4285,
+      "step": 305
+    },
+    {
+      "epoch": 0.49295207410390657,
+      "grad_norm": 0.5420446824564362,
+      "learning_rate": 3.948387096774194e-05,
+      "loss": 0.4405,
+      "step": 306
+    },
+    {
+      "epoch": 0.49456302859444223,
+      "grad_norm": 0.6297593472421277,
+      "learning_rate": 3.961290322580646e-05,
+      "loss": 0.4336,
+      "step": 307
+    },
+    {
+      "epoch": 0.49617398308497784,
+      "grad_norm": 0.6587570423137625,
+      "learning_rate": 3.974193548387097e-05,
+      "loss": 0.4241,
+      "step": 308
+    },
+    {
+      "epoch": 0.4977849375755135,
+      "grad_norm": 0.670548883274989,
+      "learning_rate": 3.9870967741935486e-05,
+      "loss": 0.4347,
+      "step": 309
+    },
+    {
+      "epoch": 0.4993958920660491,
+      "grad_norm": 0.6932491705418964,
+      "learning_rate": 4e-05,
+      "loss": 0.3976,
+      "step": 310
+    },
+    {
+      "epoch": 0.5010068465565848,
+      "grad_norm": 0.4742576328363187,
+      "learning_rate": 3.999998732081634e-05,
+      "loss": 0.4291,
+      "step": 311
+    },
+    {
+      "epoch": 0.5026178010471204,
+      "grad_norm": 0.529513963401129,
+      "learning_rate": 3.999994928328141e-05,
+      "loss": 0.4305,
+      "step": 312
+    },
+    {
+      "epoch": 0.5042287555376561,
+      "grad_norm": 0.6091004305460854,
+      "learning_rate": 3.9999885887443455e-05,
+      "loss": 0.4216,
+      "step": 313
+    },
+    {
+      "epoch": 0.5058397100281917,
+      "grad_norm": 0.5381494437696771,
+      "learning_rate": 3.9999797133382855e-05,
+      "loss": 0.4165,
+      "step": 314
+    },
+    {
+      "epoch": 0.5074506645187273,
+      "grad_norm": 0.4317011506932105,
+      "learning_rate": 3.9999683021212134e-05,
+      "loss": 0.443,
+      "step": 315
+    },
+    {
+      "epoch": 0.509061619009263,
+      "grad_norm": 0.47285920628489064,
+      "learning_rate": 3.999954355107598e-05,
+      "loss": 0.4282,
+      "step": 316
+    },
+    {
+      "epoch": 0.5106725734997987,
+      "grad_norm": 0.5242267633007971,
+      "learning_rate": 3.999937872315124e-05,
+      "loss": 0.4179,
+      "step": 317
+    },
+    {
+      "epoch": 0.5122835279903343,
+      "grad_norm": 0.4123011973763891,
+      "learning_rate": 3.9999188537646894e-05,
+      "loss": 0.4438,
+      "step": 318
+    },
+    {
+      "epoch": 0.5138944824808699,
+      "grad_norm": 0.4057247056617174,
+      "learning_rate": 3.999897299480408e-05,
+      "loss": 0.4036,
+      "step": 319
+    },
+    {
+      "epoch": 0.5155054369714056,
+      "grad_norm": 0.4188342427708109,
+      "learning_rate": 3.9998732094896084e-05,
+      "loss": 0.4299,
+      "step": 320
+    },
+    {
+      "epoch": 0.5171163914619412,
+      "grad_norm": 0.43672985709660445,
+      "learning_rate": 3.999846583822836e-05,
+      "loss": 0.4237,
+      "step": 321
+    },
+    {
+      "epoch": 0.5187273459524768,
+      "grad_norm": 0.48310073400224307,
+      "learning_rate": 3.999817422513849e-05,
+      "loss": 0.4187,
+      "step": 322
+    },
+    {
+      "epoch": 0.5203383004430124,
+      "grad_norm": 0.5530788167112765,
+      "learning_rate": 3.999785725599623e-05,
+      "loss": 0.4311,
+      "step": 323
+    },
+    {
+      "epoch": 0.5219492549335482,
+      "grad_norm": 0.5381555047474669,
+      "learning_rate": 3.999751493120345e-05,
+      "loss": 0.4291,
+      "step": 324
+    },
+    {
+      "epoch": 0.5235602094240838,
+      "grad_norm": 0.5827345030273974,
+      "learning_rate": 3.99971472511942e-05,
+      "loss": 0.4305,
+      "step": 325
+    },
+    {
+      "epoch": 0.5251711639146194,
+      "grad_norm": 0.38155681688778703,
+      "learning_rate": 3.999675421643467e-05,
+      "loss": 0.4174,
+      "step": 326
+    },
+    {
+      "epoch": 0.5267821184051551,
+      "grad_norm": 0.5492754517767143,
+      "learning_rate": 3.99963358274232e-05,
+      "loss": 0.4212,
+      "step": 327
+    },
+    {
+      "epoch": 0.5283930728956907,
+      "grad_norm": 0.48789623482177874,
+      "learning_rate": 3.9995892084690256e-05,
+      "loss": 0.4215,
+      "step": 328
+    },
+    {
+      "epoch": 0.5300040273862263,
+      "grad_norm": 0.4422563535366751,
+      "learning_rate": 3.9995422988798494e-05,
+      "loss": 0.4118,
+      "step": 329
+    },
+    {
+      "epoch": 0.5316149818767619,
+      "grad_norm": 0.6826787557912818,
+      "learning_rate": 3.999492854034266e-05,
+      "loss": 0.4322,
+      "step": 330
+    },
+    {
+      "epoch": 0.5332259363672976,
+      "grad_norm": 0.7161197167511507,
+      "learning_rate": 3.99944087399497e-05,
+      "loss": 0.4297,
+      "step": 331
+    },
+    {
+      "epoch": 0.5348368908578333,
+      "grad_norm": 0.45307187386139897,
+      "learning_rate": 3.999386358827866e-05,
+      "loss": 0.4017,
+      "step": 332
+    },
+    {
+      "epoch": 0.5364478453483689,
+      "grad_norm": 0.5080846848977575,
+      "learning_rate": 3.999329308602076e-05,
+      "loss": 0.4176,
+      "step": 333
+    },
+    {
+      "epoch": 0.5380587998389046,
+      "grad_norm": 0.44961265160037006,
+      "learning_rate": 3.9992697233899345e-05,
+      "loss": 0.4146,
+      "step": 334
+    },
+    {
+      "epoch": 0.5396697543294402,
+      "grad_norm": 0.44569468180373156,
+      "learning_rate": 3.9992076032669905e-05,
+      "loss": 0.4276,
+      "step": 335
+    },
+    {
+      "epoch": 0.5412807088199758,
+      "grad_norm": 0.4589301924407286,
+      "learning_rate": 3.999142948312007e-05,
+      "loss": 0.4068,
+      "step": 336
+    },
+    {
+      "epoch": 0.5428916633105115,
+      "grad_norm": 0.46283215335566513,
+      "learning_rate": 3.999075758606963e-05,
+      "loss": 0.4102,
+      "step": 337
+    },
+    {
+      "epoch": 0.5445026178010471,
+      "grad_norm": 0.4428187054855388,
+      "learning_rate": 3.999006034237047e-05,
+      "loss": 0.4076,
+      "step": 338
+    },
+    {
+      "epoch": 0.5461135722915827,
+      "grad_norm": 0.3941621296766751,
+      "learning_rate": 3.9989337752906656e-05,
+      "loss": 0.4248,
+      "step": 339
+    },
+    {
+      "epoch": 0.5477245267821184,
+      "grad_norm": 0.4911947798210392,
+      "learning_rate": 3.998858981859436e-05,
+      "loss": 0.4234,
+      "step": 340
+    },
+    {
+      "epoch": 0.5493354812726541,
+      "grad_norm": 0.45142619025228986,
+      "learning_rate": 3.998781654038192e-05,
+      "loss": 0.4098,
+      "step": 341
+    },
+    {
+      "epoch": 0.5509464357631897,
+      "grad_norm": 0.45716417313795044,
+      "learning_rate": 3.998701791924977e-05,
+      "loss": 0.4232,
+      "step": 342
+    },
+    {
+      "epoch": 0.5525573902537253,
+      "grad_norm": 0.4811829968487867,
+      "learning_rate": 3.998619395621051e-05,
+      "loss": 0.4301,
+      "step": 343
+    },
+    {
+      "epoch": 0.554168344744261,
+      "grad_norm": 0.3793740271087329,
+      "learning_rate": 3.9985344652308846e-05,
+      "loss": 0.4215,
+      "step": 344
+    },
+    {
+      "epoch": 0.5557792992347966,
+      "grad_norm": 0.42003419142799425,
+      "learning_rate": 3.998447000862164e-05,
+      "loss": 0.4509,
+      "step": 345
+    },
+    {
+      "epoch": 0.5573902537253322,
+      "grad_norm": 0.3906557697989691,
+      "learning_rate": 3.9983570026257844e-05,
+      "loss": 0.4231,
+      "step": 346
+    },
+    {
+      "epoch": 0.5590012082158679,
+      "grad_norm": 0.39196036899925196,
+      "learning_rate": 3.9982644706358596e-05,
+      "loss": 0.4089,
+      "step": 347
+    },
+    {
+      "epoch": 0.5606121627064036,
+      "grad_norm": 0.4913202549527037,
+      "learning_rate": 3.998169405009711e-05,
+      "loss": 0.422,
+      "step": 348
+    },
+    {
+      "epoch": 0.5622231171969392,
+      "grad_norm": 0.4279358660405233,
+      "learning_rate": 3.9980718058678733e-05,
+      "loss": 0.4248,
+      "step": 349
+    },
+    {
+      "epoch": 0.5638340716874748,
+      "grad_norm": 0.49969446549842206,
+      "learning_rate": 3.997971673334095e-05,
+      "loss": 0.4202,
+      "step": 350
+    },
+    {
+      "epoch": 0.5654450261780105,
+      "grad_norm": 0.41732878429281683,
+      "learning_rate": 3.997869007535336e-05,
+      "loss": 0.422,
+      "step": 351
+    },
+    {
+      "epoch": 0.5670559806685461,
+      "grad_norm": 0.4294137359439838,
+      "learning_rate": 3.997763808601768e-05,
+      "loss": 0.4252,
+      "step": 352
+    },
+    {
+      "epoch": 0.5686669351590817,
+      "grad_norm": 0.3707661272418532,
+      "learning_rate": 3.997656076666776e-05,
+      "loss": 0.4027,
+      "step": 353
+    },
+    {
+      "epoch": 0.5702778896496175,
+      "grad_norm": 0.3766908978268678,
+      "learning_rate": 3.997545811866952e-05,
+      "loss": 0.4312,
+      "step": 354
+    },
+    {
+      "epoch": 0.5718888441401531,
+      "grad_norm": 0.45306234795270156,
+      "learning_rate": 3.997433014342106e-05,
+      "loss": 0.4223,
+      "step": 355
+    },
+    {
+      "epoch": 0.5734997986306887,
+      "grad_norm": 0.4793083302776694,
+      "learning_rate": 3.997317684235254e-05,
+      "loss": 0.4229,
+      "step": 356
+    },
+    {
+      "epoch": 0.5751107531212243,
+      "grad_norm": 0.4675021055165329,
+      "learning_rate": 3.9971998216926274e-05,
+      "loss": 0.4434,
+      "step": 357
+    },
+    {
+      "epoch": 0.57672170761176,
+      "grad_norm": 0.4460710699457162,
+      "learning_rate": 3.997079426863664e-05,
+      "loss": 0.4013,
+      "step": 358
+    },
+    {
+      "epoch": 0.5783326621022956,
+      "grad_norm": 0.45964782363764733,
+      "learning_rate": 3.996956499901015e-05,
+      "loss": 0.4245,
+      "step": 359
+    },
+    {
+      "epoch": 0.5799436165928312,
+      "grad_norm": 0.5182341589054947,
+      "learning_rate": 3.996831040960543e-05,
+      "loss": 0.4038,
+      "step": 360
+    },
+    {
+      "epoch": 0.581554571083367,
+      "grad_norm": 0.4117269217562021,
+      "learning_rate": 3.996703050201319e-05,
+      "loss": 0.4153,
+      "step": 361
+    },
+    {
+      "epoch": 0.5831655255739026,
+      "grad_norm": 0.46882737702461946,
+      "learning_rate": 3.996572527785625e-05,
+      "loss": 0.3976,
+      "step": 362
+    },
+    {
+      "epoch": 0.5847764800644382,
+      "grad_norm": 0.3796364352520464,
+      "learning_rate": 3.996439473878952e-05,
+      "loss": 0.4208,
+      "step": 363
+    },
+    {
+      "epoch": 0.5863874345549738,
+      "grad_norm": 0.4992194206692767,
+      "learning_rate": 3.996303888650002e-05,
+      "loss": 0.4266,
+      "step": 364
+    },
+    {
+      "epoch": 0.5879983890455095,
+      "grad_norm": 0.4787895226087798,
+      "learning_rate": 3.9961657722706864e-05,
+      "loss": 0.4301,
+      "step": 365
+    },
+    {
+      "epoch": 0.5896093435360451,
+      "grad_norm": 0.5407494009323632,
+      "learning_rate": 3.996025124916125e-05,
+      "loss": 0.4157,
+      "step": 366
+    },
+    {
+      "epoch": 0.5912202980265807,
+      "grad_norm": 0.4261943946875255,
+      "learning_rate": 3.995881946764647e-05,
+      "loss": 0.4089,
+      "step": 367
+    },
+    {
+      "epoch": 0.5928312525171164,
+      "grad_norm": 0.3850462450614437,
+      "learning_rate": 3.995736237997792e-05,
+      "loss": 0.4272,
+      "step": 368
+    },
+    {
+      "epoch": 0.594442207007652,
+      "grad_norm": 0.45463889935104884,
+      "learning_rate": 3.9955879988003046e-05,
+      "loss": 0.4133,
+      "step": 369
+    },
+    {
+      "epoch": 0.5960531614981877,
+      "grad_norm": 0.48825127888326364,
+      "learning_rate": 3.9954372293601415e-05,
+      "loss": 0.4336,
+      "step": 370
+    },
+    {
+      "epoch": 0.5976641159887233,
+      "grad_norm": 0.4068129703538119,
+      "learning_rate": 3.9952839298684656e-05,
+      "loss": 0.4001,
+      "step": 371
+    },
+    {
+      "epoch": 0.599275070479259,
+      "grad_norm": 0.49552613300447507,
+      "learning_rate": 3.9951281005196486e-05,
+      "loss": 0.4023,
+      "step": 372
+    },
+    {
+      "epoch": 0.6008860249697946,
+      "grad_norm": 0.46945317770031153,
+      "learning_rate": 3.994969741511269e-05,
+      "loss": 0.3857,
+      "step": 373
+    },
+    {
+      "epoch": 0.6024969794603302,
+      "grad_norm": 0.4741587306281566,
+      "learning_rate": 3.994808853044113e-05,
+      "loss": 0.4461,
+      "step": 374
+    },
+    {
+      "epoch": 0.6041079339508659,
+      "grad_norm": 0.4685179794707909,
+      "learning_rate": 3.994645435322174e-05,
+      "loss": 0.4094,
+      "step": 375
+    },
+    {
+      "epoch": 0.6057188884414015,
+      "grad_norm": 0.45483642283060166,
+      "learning_rate": 3.994479488552652e-05,
+      "loss": 0.4228,
+      "step": 376
+    },
+    {
+      "epoch": 0.6073298429319371,
+      "grad_norm": 0.4244946382664049,
+      "learning_rate": 3.9943110129459555e-05,
+      "loss": 0.4279,
+      "step": 377
+    },
+    {
+      "epoch": 0.6089407974224729,
+      "grad_norm": 0.4294198854653389,
+      "learning_rate": 3.994140008715697e-05,
+      "loss": 0.4032,
+      "step": 378
+    },
+    {
+      "epoch": 0.6105517519130085,
+      "grad_norm": 0.3989456663260737,
+      "learning_rate": 3.993966476078694e-05,
+      "loss": 0.4147,
+      "step": 379
+    },
+    {
+      "epoch": 0.6121627064035441,
+      "grad_norm": 0.3713937522581636,
+      "learning_rate": 3.9937904152549746e-05,
+      "loss": 0.4192,
+      "step": 380
+    },
+    {
+      "epoch": 0.6137736608940797,
+      "grad_norm": 0.4384496746610661,
+      "learning_rate": 3.993611826467768e-05,
+      "loss": 0.4356,
+      "step": 381
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.42889466156762074,
+      "learning_rate": 3.99343070994351e-05,
+      "loss": 0.4299,
+      "step": 382
+    },
+    {
+      "epoch": 0.616995569875151,
+      "grad_norm": 0.3555646074061858,
+      "learning_rate": 3.993247065911844e-05,
+      "loss": 0.4309,
+      "step": 383
+    },
+    {
+      "epoch": 0.6186065243656866,
+      "grad_norm": 0.43396045989866394,
+      "learning_rate": 3.993060894605612e-05,
+      "loss": 0.3994,
+      "step": 384
+    },
+    {
+      "epoch": 0.6202174788562224,
+      "grad_norm": 0.43776606757803926,
+      "learning_rate": 3.992872196260866e-05,
+      "loss": 0.4097,
+      "step": 385
+    },
+    {
+      "epoch": 0.621828433346758,
+      "grad_norm": 0.420636085604741,
+      "learning_rate": 3.99268097111686e-05,
+      "loss": 0.417,
+      "step": 386
+    },
+    {
+      "epoch": 0.6234393878372936,
+      "grad_norm": 0.5064346878522746,
+      "learning_rate": 3.992487219416052e-05,
+      "loss": 0.4183,
+      "step": 387
+    },
+    {
+      "epoch": 0.6250503423278292,
+      "grad_norm": 0.5057851487040337,
+      "learning_rate": 3.992290941404103e-05,
+      "loss": 0.3974,
+      "step": 388
+    },
+    {
+      "epoch": 0.6266612968183649,
+      "grad_norm": 0.5521508424756063,
+      "learning_rate": 3.992092137329878e-05,
+      "loss": 0.4108,
+      "step": 389
+    },
+    {
+      "epoch": 0.6282722513089005,
+      "grad_norm": 0.5405974200010049,
+      "learning_rate": 3.991890807445443e-05,
+      "loss": 0.4477,
+      "step": 390
+    },
+    {
+      "epoch": 0.6298832057994361,
+      "grad_norm": 0.39506341263585665,
+      "learning_rate": 3.991686952006069e-05,
+      "loss": 0.4326,
+      "step": 391
+    },
+    {
+      "epoch": 0.6314941602899719,
+      "grad_norm": 0.511124831235241,
+      "learning_rate": 3.991480571270228e-05,
+      "loss": 0.4226,
+      "step": 392
+    },
+    {
+      "epoch": 0.6331051147805075,
+      "grad_norm": 0.42776265768914273,
+      "learning_rate": 3.991271665499594e-05,
+      "loss": 0.4046,
+      "step": 393
+    },
+    {
+      "epoch": 0.6347160692710431,
+      "grad_norm": 0.4959908838554333,
+      "learning_rate": 3.991060234959042e-05,
+      "loss": 0.416,
+      "step": 394
+    },
+    {
+      "epoch": 0.6363270237615787,
+      "grad_norm": 0.4670521894890168,
+      "learning_rate": 3.990846279916649e-05,
+      "loss": 0.4224,
+      "step": 395
+    },
+    {
+      "epoch": 0.6379379782521144,
+      "grad_norm": 0.4006152685344792,
+      "learning_rate": 3.9906298006436924e-05,
+      "loss": 0.4164,
+      "step": 396
+    },
+    {
+      "epoch": 0.63954893274265,
+      "grad_norm": 0.4463899466226375,
+      "learning_rate": 3.99041079741465e-05,
+      "loss": 0.4298,
+      "step": 397
+    },
+    {
+      "epoch": 0.6411598872331856,
+      "grad_norm": 0.39436113561085406,
+      "learning_rate": 3.9901892705072004e-05,
+      "loss": 0.4139,
+      "step": 398
+    },
+    {
+      "epoch": 0.6427708417237213,
+      "grad_norm": 0.48684252012399337,
+      "learning_rate": 3.989965220202221e-05,
+      "loss": 0.4157,
+      "step": 399
+    },
+    {
+      "epoch": 0.644381796214257,
+      "grad_norm": 0.36781064362322613,
+      "learning_rate": 3.98973864678379e-05,
+      "loss": 0.4094,
+      "step": 400
+    },
+    {
+      "epoch": 0.6459927507047926,
+      "grad_norm": 0.4545736363221886,
+      "learning_rate": 3.989509550539185e-05,
+      "loss": 0.416,
+      "step": 401
+    },
+    {
+      "epoch": 0.6476037051953283,
+      "grad_norm": 0.4021213653694868,
+      "learning_rate": 3.989277931758879e-05,
+      "loss": 0.3986,
+      "step": 402
+    },
+    {
+      "epoch": 0.6492146596858639,
+      "grad_norm": 0.3938202729838156,
+      "learning_rate": 3.989043790736547e-05,
+      "loss": 0.3977,
+      "step": 403
+    },
+    {
+      "epoch": 0.6508256141763995,
+      "grad_norm": 0.3714283205343456,
+      "learning_rate": 3.9888071277690606e-05,
+      "loss": 0.4008,
+      "step": 404
+    },
+    {
+      "epoch": 0.6524365686669351,
+      "grad_norm": 0.49605923508034744,
+      "learning_rate": 3.988567943156489e-05,
+      "loss": 0.4044,
+      "step": 405
+    },
+    {
+      "epoch": 0.6540475231574708,
+      "grad_norm": 0.5342861440227792,
+      "learning_rate": 3.988326237202099e-05,
+      "loss": 0.4053,
+      "step": 406
+    },
+    {
+      "epoch": 0.6556584776480064,
+      "grad_norm": 0.3428099270926289,
+      "learning_rate": 3.988082010212354e-05,
+      "loss": 0.3928,
+      "step": 407
+    },
+    {
+      "epoch": 0.657269432138542,
+      "grad_norm": 0.4004574130249575,
+      "learning_rate": 3.987835262496913e-05,
+      "loss": 0.4171,
+      "step": 408
+    },
+    {
+      "epoch": 0.6588803866290778,
+      "grad_norm": 0.35970332120506987,
+      "learning_rate": 3.9875859943686335e-05,
+      "loss": 0.4024,
+      "step": 409
+    },
+    {
+      "epoch": 0.6604913411196134,
+      "grad_norm": 0.4224502636249679,
+      "learning_rate": 3.9873342061435664e-05,
+      "loss": 0.4187,
+      "step": 410
+    },
+    {
+      "epoch": 0.662102295610149,
+      "grad_norm": 0.3764938020888867,
+      "learning_rate": 3.987079898140958e-05,
+      "loss": 0.4072,
+      "step": 411
+    },
+    {
+      "epoch": 0.6637132501006846,
+      "grad_norm": 0.3694724316905374,
+      "learning_rate": 3.98682307068325e-05,
+      "loss": 0.415,
+      "step": 412
+    },
+    {
+      "epoch": 0.6653242045912203,
+      "grad_norm": 0.36693082697539087,
+      "learning_rate": 3.98656372409608e-05,
+      "loss": 0.4046,
+      "step": 413
+    },
+    {
+      "epoch": 0.6669351590817559,
+      "grad_norm": 0.44730430916292946,
+      "learning_rate": 3.986301858708278e-05,
+      "loss": 0.4133,
+      "step": 414
+    },
+    {
+      "epoch": 0.6685461135722915,
+      "grad_norm": 0.3809823415227652,
+      "learning_rate": 3.9860374748518676e-05,
+      "loss": 0.4211,
+      "step": 415
+    },
+    {
+      "epoch": 0.6701570680628273,
+      "grad_norm": 0.3563474324193386,
+      "learning_rate": 3.985770572862065e-05,
+      "loss": 0.4088,
+      "step": 416
+    },
+    {
+      "epoch": 0.6717680225533629,
+      "grad_norm": 0.3805173842463904,
+      "learning_rate": 3.985501153077282e-05,
+      "loss": 0.4124,
+      "step": 417
+    },
+    {
+      "epoch": 0.6733789770438985,
+      "grad_norm": 0.43552193509232306,
+      "learning_rate": 3.985229215839119e-05,
+      "loss": 0.4261,
+      "step": 418
+    },
+    {
+      "epoch": 0.6749899315344342,
+      "grad_norm": 0.42439761528072495,
+      "learning_rate": 3.984954761492372e-05,
+      "loss": 0.4017,
+      "step": 419
+    },
+    {
+      "epoch": 0.6766008860249698,
+      "grad_norm": 0.3105332394794023,
+      "learning_rate": 3.984677790385025e-05,
+      "loss": 0.3985,
+      "step": 420
+    },
+    {
+      "epoch": 0.6782118405155054,
+      "grad_norm": 0.4027207084117632,
+      "learning_rate": 3.9843983028682555e-05,
+      "loss": 0.4026,
+      "step": 421
+    },
+    {
+      "epoch": 0.679822795006041,
+      "grad_norm": 0.3508235171579207,
+      "learning_rate": 3.984116299296432e-05,
+      "loss": 0.3896,
+      "step": 422
+    },
+    {
+      "epoch": 0.6814337494965768,
+      "grad_norm": 0.3573507910316362,
+      "learning_rate": 3.9838317800271105e-05,
+      "loss": 0.4283,
+      "step": 423
+    },
+    {
+      "epoch": 0.6830447039871124,
+      "grad_norm": 0.37632509861430463,
+      "learning_rate": 3.983544745421038e-05,
+      "loss": 0.3997,
+      "step": 424
+    },
+    {
+      "epoch": 0.684655658477648,
+      "grad_norm": 0.3388804904806458,
+      "learning_rate": 3.983255195842152e-05,
+      "loss": 0.4124,
+      "step": 425
+    },
+    {
+      "epoch": 0.6862666129681837,
+      "grad_norm": 0.4833298766343206,
+      "learning_rate": 3.982963131657577e-05,
+      "loss": 0.4177,
+      "step": 426
+    },
+    {
+      "epoch": 0.6878775674587193,
+      "grad_norm": 0.42070364198235066,
+      "learning_rate": 3.982668553237628e-05,
+      "loss": 0.3904,
+      "step": 427
+    },
+    {
+      "epoch": 0.6894885219492549,
+      "grad_norm": 0.4482515006894441,
+      "learning_rate": 3.982371460955804e-05,
+      "loss": 0.4311,
+      "step": 428
+    },
+    {
+      "epoch": 0.6910994764397905,
+      "grad_norm": 0.40436392123161874,
+      "learning_rate": 3.982071855188796e-05,
+      "loss": 0.4045,
+      "step": 429
+    },
+    {
+      "epoch": 0.6927104309303262,
+      "grad_norm": 0.3383709764320333,
+      "learning_rate": 3.981769736316478e-05,
+      "loss": 0.4006,
+      "step": 430
+    },
+    {
+      "epoch": 0.6943213854208619,
+      "grad_norm": 0.42754996232658854,
+      "learning_rate": 3.9814651047219135e-05,
+      "loss": 0.4289,
+      "step": 431
+    },
+    {
+      "epoch": 0.6959323399113975,
+      "grad_norm": 0.3876602252178334,
+      "learning_rate": 3.9811579607913486e-05,
+      "loss": 0.4246,
+      "step": 432
+    },
+    {
+      "epoch": 0.6975432944019332,
+      "grad_norm": 0.4278854255881262,
+      "learning_rate": 3.9808483049142185e-05,
+      "loss": 0.4448,
+      "step": 433
+    },
+    {
+      "epoch": 0.6991542488924688,
+      "grad_norm": 0.35865693620507294,
+      "learning_rate": 3.980536137483141e-05,
+      "loss": 0.4142,
+      "step": 434
+    },
+    {
+      "epoch": 0.7007652033830044,
+      "grad_norm": 0.4625593533325866,
+      "learning_rate": 3.980221458893919e-05,
+      "loss": 0.4171,
+      "step": 435
+    },
+    {
+      "epoch": 0.70237615787354,
+      "grad_norm": 0.32682701461120106,
+      "learning_rate": 3.979904269545538e-05,
+      "loss": 0.4266,
+      "step": 436
+    },
+    {
+      "epoch": 0.7039871123640757,
+      "grad_norm": 0.415104731241559,
+      "learning_rate": 3.979584569840171e-05,
+      "loss": 0.4204,
+      "step": 437
+    },
+    {
+      "epoch": 0.7055980668546113,
+      "grad_norm": 0.3592034123608851,
+      "learning_rate": 3.979262360183169e-05,
+      "loss": 0.4067,
+      "step": 438
+    },
+    {
+      "epoch": 0.707209021345147,
+      "grad_norm": 0.43562091204913506,
+      "learning_rate": 3.9789376409830674e-05,
+      "loss": 0.4081,
+      "step": 439
+    },
+    {
+      "epoch": 0.7088199758356827,
+      "grad_norm": 0.43196190361541487,
+      "learning_rate": 3.978610412651584e-05,
+      "loss": 0.401,
+      "step": 440
+    },
+    {
+      "epoch": 0.7104309303262183,
+      "grad_norm": 0.4071385089261009,
+      "learning_rate": 3.978280675603618e-05,
+      "loss": 0.4163,
+      "step": 441
+    },
+    {
+      "epoch": 0.7120418848167539,
+      "grad_norm": 0.41016039223243267,
+      "learning_rate": 3.97794843025725e-05,
+      "loss": 0.3942,
+      "step": 442
+    },
+    {
+      "epoch": 0.7136528393072896,
+      "grad_norm": 0.3841321139543235,
+      "learning_rate": 3.977613677033738e-05,
+      "loss": 0.4118,
+      "step": 443
+    },
+    {
+      "epoch": 0.7152637937978252,
+      "grad_norm": 0.35310267037819615,
+      "learning_rate": 3.977276416357523e-05,
+      "loss": 0.4218,
+      "step": 444
+    },
+    {
+      "epoch": 0.7168747482883608,
+      "grad_norm": 0.4197080640230907,
+      "learning_rate": 3.976936648656223e-05,
+      "loss": 0.4261,
+      "step": 445
+    },
+    {
+      "epoch": 0.7184857027788965,
+      "grad_norm": 0.3546111322565111,
+      "learning_rate": 3.976594374360637e-05,
+      "loss": 0.4133,
+      "step": 446
+    },
+    {
+      "epoch": 0.7200966572694322,
+      "grad_norm": 0.4394668646642138,
+      "learning_rate": 3.97624959390474e-05,
+      "loss": 0.4168,
+      "step": 447
+    },
+    {
+      "epoch": 0.7217076117599678,
+      "grad_norm": 0.3221466833352809,
+      "learning_rate": 3.975902307725686e-05,
+      "loss": 0.4175,
+      "step": 448
+    },
+    {
+      "epoch": 0.7233185662505034,
+      "grad_norm": 0.46121007238081874,
+      "learning_rate": 3.975552516263804e-05,
+      "loss": 0.3995,
+      "step": 449
+    },
+    {
+      "epoch": 0.7249295207410391,
+      "grad_norm": 0.40160398472276715,
+      "learning_rate": 3.9752002199626035e-05,
+      "loss": 0.4152,
+      "step": 450
+    },
+    {
+      "epoch": 0.7265404752315747,
+      "grad_norm": 0.37117019417124525,
+      "learning_rate": 3.974845419268766e-05,
+      "loss": 0.3937,
+      "step": 451
+    },
+    {
+      "epoch": 0.7281514297221103,
+      "grad_norm": 0.31852553564843433,
+      "learning_rate": 3.97448811463215e-05,
+      "loss": 0.396,
+      "step": 452
+    },
+    {
+      "epoch": 0.7297623842126459,
+      "grad_norm": 0.42093190036658035,
+      "learning_rate": 3.974128306505788e-05,
+      "loss": 0.4105,
+      "step": 453
+    },
+    {
+      "epoch": 0.7313733387031817,
+      "grad_norm": 0.3427182108489802,
+      "learning_rate": 3.973765995345889e-05,
+      "loss": 0.3974,
+      "step": 454
+    },
+    {
+      "epoch": 0.7329842931937173,
+      "grad_norm": 0.38214069037483106,
+      "learning_rate": 3.973401181611832e-05,
+      "loss": 0.4309,
+      "step": 455
+    },
+    {
+      "epoch": 0.7345952476842529,
+      "grad_norm": 0.41629666543067606,
+      "learning_rate": 3.973033865766172e-05,
+      "loss": 0.4094,
+      "step": 456
+    },
+    {
+      "epoch": 0.7362062021747886,
+      "grad_norm": 0.34200464296520006,
+      "learning_rate": 3.972664048274636e-05,
+      "loss": 0.4161,
+      "step": 457
+    },
+    {
+      "epoch": 0.7378171566653242,
+      "grad_norm": 0.3471638897635492,
+      "learning_rate": 3.9722917296061216e-05,
+      "loss": 0.4297,
+      "step": 458
+    },
+    {
+      "epoch": 0.7394281111558598,
+      "grad_norm": 0.45161457244236064,
+      "learning_rate": 3.971916910232699e-05,
+      "loss": 0.4295,
+      "step": 459
+    },
+    {
+      "epoch": 0.7410390656463954,
+      "grad_norm": 0.42974197063134223,
+      "learning_rate": 3.971539590629608e-05,
+      "loss": 0.399,
+      "step": 460
+    },
+    {
+      "epoch": 0.7426500201369312,
+      "grad_norm": 0.43361932874840936,
+      "learning_rate": 3.971159771275259e-05,
+      "loss": 0.4096,
+      "step": 461
+    },
+    {
+      "epoch": 0.7442609746274668,
+      "grad_norm": 0.36249109522453676,
+      "learning_rate": 3.9707774526512334e-05,
+      "loss": 0.3998,
+      "step": 462
+    },
+    {
+      "epoch": 0.7458719291180024,
+      "grad_norm": 0.42345130014954085,
+      "learning_rate": 3.970392635242278e-05,
+      "loss": 0.4123,
+      "step": 463
+    },
+    {
+      "epoch": 0.7474828836085381,
+      "grad_norm": 0.4028131720862331,
+      "learning_rate": 3.970005319536311e-05,
+      "loss": 0.4048,
+      "step": 464
+    },
+    {
+      "epoch": 0.7490938380990737,
+      "grad_norm": 0.3922591817364784,
+      "learning_rate": 3.9696155060244166e-05,
+      "loss": 0.416,
+      "step": 465
+    },
+    {
+      "epoch": 0.7507047925896093,
+      "grad_norm": 0.394910086124614,
+      "learning_rate": 3.969223195200847e-05,
+      "loss": 0.409,
+      "step": 466
+    },
+    {
+      "epoch": 0.752315747080145,
+      "grad_norm": 0.4010895179099205,
+      "learning_rate": 3.9688283875630193e-05,
+      "loss": 0.4221,
+      "step": 467
+    },
+    {
+      "epoch": 0.7539267015706806,
+      "grad_norm": 0.3344274716823261,
+      "learning_rate": 3.96843108361152e-05,
+      "loss": 0.3966,
+      "step": 468
+    },
+    {
+      "epoch": 0.7555376560612163,
+      "grad_norm": 0.36508649487932493,
+      "learning_rate": 3.968031283850094e-05,
+      "loss": 0.3896,
+      "step": 469
+    },
+    {
+      "epoch": 0.7571486105517519,
+      "grad_norm": 0.3194248632705373,
+      "learning_rate": 3.967628988785658e-05,
+      "loss": 0.3913,
+      "step": 470
+    },
+    {
+      "epoch": 0.7587595650422876,
+      "grad_norm": 0.3348385327401639,
+      "learning_rate": 3.967224198928289e-05,
+      "loss": 0.4321,
+      "step": 471
+    },
+    {
+      "epoch": 0.7603705195328232,
+      "grad_norm": 0.3331982859049919,
+      "learning_rate": 3.966816914791226e-05,
+      "loss": 0.4088,
+      "step": 472
+    },
+    {
+      "epoch": 0.7619814740233588,
+      "grad_norm": 0.3252671342709306,
+      "learning_rate": 3.9664071368908726e-05,
+      "loss": 0.4324,
+      "step": 473
+    },
+    {
+      "epoch": 0.7635924285138945,
+      "grad_norm": 0.3778011510229453,
+      "learning_rate": 3.965994865746794e-05,
+      "loss": 0.4083,
+      "step": 474
+    },
+    {
+      "epoch": 0.7652033830044301,
+      "grad_norm": 0.3697341056634576,
+      "learning_rate": 3.9655801018817166e-05,
+      "loss": 0.417,
+      "step": 475
+    },
+    {
+      "epoch": 0.7668143374949657,
+      "grad_norm": 0.3686614640695823,
+      "learning_rate": 3.965162845821526e-05,
+      "loss": 0.404,
+      "step": 476
+    },
+    {
+      "epoch": 0.7684252919855014,
+      "grad_norm": 0.3198502413551383,
+      "learning_rate": 3.96474309809527e-05,
+      "loss": 0.3918,
+      "step": 477
+    },
+    {
+      "epoch": 0.7700362464760371,
+      "grad_norm": 0.404426337364835,
+      "learning_rate": 3.964320859235155e-05,
+      "loss": 0.3978,
+      "step": 478
+    },
+    {
+      "epoch": 0.7716472009665727,
+      "grad_norm": 0.37411618257407514,
+      "learning_rate": 3.963896129776544e-05,
+      "loss": 0.4447,
+      "step": 479
+    },
+    {
+      "epoch": 0.7732581554571083,
+      "grad_norm": 0.3647067289069705,
+      "learning_rate": 3.963468910257959e-05,
+      "loss": 0.4123,
+      "step": 480
+    },
+    {
+      "epoch": 0.774869109947644,
+      "grad_norm": 0.3646906272324245,
+      "learning_rate": 3.9630392012210804e-05,
+      "loss": 0.4106,
+      "step": 481
+    },
+    {
+      "epoch": 0.7764800644381796,
+      "grad_norm": 0.34012637826464404,
+      "learning_rate": 3.962607003210745e-05,
+      "loss": 0.4228,
+      "step": 482
+    },
+    {
+      "epoch": 0.7780910189287152,
+      "grad_norm": 0.3104491216102974,
+      "learning_rate": 3.9621723167749424e-05,
+      "loss": 0.4041,
+      "step": 483
+    },
+    {
+      "epoch": 0.779701973419251,
+      "grad_norm": 0.35559733908050584,
+      "learning_rate": 3.9617351424648215e-05,
+      "loss": 0.4224,
+      "step": 484
+    },
+    {
+      "epoch": 0.7813129279097866,
+      "grad_norm": 0.3461370454635196,
+      "learning_rate": 3.961295480834683e-05,
+      "loss": 0.4032,
+      "step": 485
+    },
+    {
+      "epoch": 0.7829238824003222,
+      "grad_norm": 0.3625961813341553,
+      "learning_rate": 3.960853332441981e-05,
+      "loss": 0.4031,
+      "step": 486
+    },
+    {
+      "epoch": 0.7845348368908578,
+      "grad_norm": 0.329895268062928,
+      "learning_rate": 3.960408697847324e-05,
+      "loss": 0.4115,
+      "step": 487
+    },
+    {
+      "epoch": 0.7861457913813935,
+      "grad_norm": 0.38660120590435787,
+      "learning_rate": 3.959961577614474e-05,
+      "loss": 0.4166,
+      "step": 488
+    },
+    {
+      "epoch": 0.7877567458719291,
+      "grad_norm": 0.33405951123771366,
+      "learning_rate": 3.9595119723103416e-05,
+      "loss": 0.4125,
+      "step": 489
+    },
+    {
+      "epoch": 0.7893677003624647,
+      "grad_norm": 0.37519732027389735,
+      "learning_rate": 3.9590598825049896e-05,
+      "loss": 0.4171,
+      "step": 490
+    },
+    {
+      "epoch": 0.7909786548530005,
+      "grad_norm": 0.3309560769864468,
+      "learning_rate": 3.95860530877163e-05,
+      "loss": 0.3816,
+      "step": 491
+    },
+    {
+      "epoch": 0.7925896093435361,
+      "grad_norm": 0.5272231298290175,
+      "learning_rate": 3.958148251686628e-05,
+      "loss": 0.4174,
+      "step": 492
+    },
+    {
+      "epoch": 0.7942005638340717,
+      "grad_norm": 0.3928743178231115,
+      "learning_rate": 3.9576887118294915e-05,
+      "loss": 0.4315,
+      "step": 493
+    },
+    {
+      "epoch": 0.7958115183246073,
+      "grad_norm": 0.4279333846270503,
+      "learning_rate": 3.957226689782882e-05,
+      "loss": 0.3968,
+      "step": 494
+    },
+    {
+      "epoch": 0.797422472815143,
+      "grad_norm": 0.5131288201799059,
+      "learning_rate": 3.956762186132604e-05,
+      "loss": 0.4145,
+      "step": 495
+    },
+    {
+      "epoch": 0.7990334273056786,
+      "grad_norm": 0.4048881402151487,
+      "learning_rate": 3.9562952014676116e-05,
+      "loss": 0.4184,
+      "step": 496
+    },
+    {
+      "epoch": 0.8006443817962142,
+      "grad_norm": 0.5006403294292532,
+      "learning_rate": 3.955825736380002e-05,
+      "loss": 0.4098,
+      "step": 497
+    },
+    {
+      "epoch": 0.8022553362867499,
+      "grad_norm": 0.45269029887323053,
+      "learning_rate": 3.95535379146502e-05,
+      "loss": 0.3942,
+      "step": 498
+    },
+    {
+      "epoch": 0.8038662907772856,
+      "grad_norm": 0.4401085637177939,
+      "learning_rate": 3.9548793673210515e-05,
+      "loss": 0.4112,
+      "step": 499
+    },
+    {
+      "epoch": 0.8054772452678212,
+      "grad_norm": 0.4536636237917159,
+      "learning_rate": 3.954402464549628e-05,
+      "loss": 0.4069,
+      "step": 500
+    },
+    {
+      "epoch": 0.8070881997583568,
+      "grad_norm": 0.4549086907445366,
+      "learning_rate": 3.9539230837554253e-05,
+      "loss": 0.3907,
+      "step": 501
+    },
+    {
+      "epoch": 0.8086991542488925,
+      "grad_norm": 0.4852155503335648,
+      "learning_rate": 3.953441225546257e-05,
+      "loss": 0.3988,
+      "step": 502
+    },
+    {
+      "epoch": 0.8103101087394281,
+      "grad_norm": 0.44412599397084784,
+      "learning_rate": 3.95295689053308e-05,
+      "loss": 0.4119,
+      "step": 503
+    },
+    {
+      "epoch": 0.8119210632299637,
+      "grad_norm": 0.44492278549668857,
+      "learning_rate": 3.9524700793299926e-05,
+      "loss": 0.4121,
+      "step": 504
+    },
+    {
+      "epoch": 0.8135320177204994,
+      "grad_norm": 0.34508494742909007,
+      "learning_rate": 3.951980792554231e-05,
+      "loss": 0.4033,
+      "step": 505
+    },
+    {
+      "epoch": 0.815142972211035,
+      "grad_norm": 0.4855681935693533,
+      "learning_rate": 3.9514890308261706e-05,
+      "loss": 0.4152,
+      "step": 506
+    },
+    {
+      "epoch": 0.8167539267015707,
+      "grad_norm": 0.3132705383046582,
+      "learning_rate": 3.9509947947693266e-05,
+      "loss": 0.3828,
+      "step": 507
+    },
+    {
+      "epoch": 0.8183648811921064,
+      "grad_norm": 0.4283053623030758,
+      "learning_rate": 3.950498085010348e-05,
+      "loss": 0.4005,
+      "step": 508
+    },
+    {
+      "epoch": 0.819975835682642,
+      "grad_norm": 0.3609521715579133,
+      "learning_rate": 3.949998902179024e-05,
+      "loss": 0.4155,
+      "step": 509
+    },
+    {
+      "epoch": 0.8215867901731776,
+      "grad_norm": 0.36832356239597386,
+      "learning_rate": 3.9494972469082764e-05,
+      "loss": 0.4141,
+      "step": 510
+    },
+    {
+      "epoch": 0.8231977446637132,
+      "grad_norm": 0.4503424316996966,
+      "learning_rate": 3.948993119834164e-05,
+      "loss": 0.4112,
+      "step": 511
+    },
+    {
+      "epoch": 0.8248086991542489,
+      "grad_norm": 0.41571779430634015,
+      "learning_rate": 3.948486521595878e-05,
+      "loss": 0.427,
+      "step": 512
+    },
+    {
+      "epoch": 0.8264196536447845,
+      "grad_norm": 0.3883909259491303,
+      "learning_rate": 3.9479774528357445e-05,
+      "loss": 0.3969,
+      "step": 513
+    },
+    {
+      "epoch": 0.8280306081353201,
+      "grad_norm": 0.3932810351114709,
+      "learning_rate": 3.9474659141992197e-05,
+      "loss": 0.4121,
+      "step": 514
+    },
+    {
+      "epoch": 0.8296415626258559,
+      "grad_norm": 0.4308748780875734,
+      "learning_rate": 3.946951906334895e-05,
+      "loss": 0.4076,
+      "step": 515
+    },
+    {
+      "epoch": 0.8312525171163915,
+      "grad_norm": 0.35777478573661275,
+      "learning_rate": 3.946435429894488e-05,
+      "loss": 0.4132,
+      "step": 516
+    },
+    {
+      "epoch": 0.8328634716069271,
+      "grad_norm": 0.48261083312441627,
+      "learning_rate": 3.94591648553285e-05,
+      "loss": 0.4006,
+      "step": 517
+    },
+    {
+      "epoch": 0.8344744260974627,
+      "grad_norm": 0.40863315410267975,
+      "learning_rate": 3.94539507390796e-05,
+      "loss": 0.4063,
+      "step": 518
+    },
+    {
+      "epoch": 0.8360853805879984,
+      "grad_norm": 0.43340425011865114,
+      "learning_rate": 3.944871195680926e-05,
+      "loss": 0.4138,
+      "step": 519
+    },
+    {
+      "epoch": 0.837696335078534,
+      "grad_norm": 0.39224215584003436,
+      "learning_rate": 3.9443448515159815e-05,
+      "loss": 0.3847,
+      "step": 520
+    },
+    {
+      "epoch": 0.8393072895690696,
+      "grad_norm": 0.4188929922999027,
+      "learning_rate": 3.9438160420804886e-05,
+      "loss": 0.4059,
+      "step": 521
+    },
+    {
+      "epoch": 0.8409182440596054,
+      "grad_norm": 0.41744558940813187,
+      "learning_rate": 3.943284768044935e-05,
+      "loss": 0.4052,
+      "step": 522
+    },
+    {
+      "epoch": 0.842529198550141,
+      "grad_norm": 0.46993836475096185,
+      "learning_rate": 3.942751030082932e-05,
+      "loss": 0.4243,
+      "step": 523
+    },
+    {
+      "epoch": 0.8441401530406766,
+      "grad_norm": 0.36251580903388136,
+      "learning_rate": 3.942214828871216e-05,
+      "loss": 0.4198,
+      "step": 524
+    },
+    {
+      "epoch": 0.8457511075312123,
+      "grad_norm": 0.3940717167893435,
+      "learning_rate": 3.9416761650896456e-05,
+      "loss": 0.4082,
+      "step": 525
+    },
+    {
+      "epoch": 0.8473620620217479,
+      "grad_norm": 0.4274934534112222,
+      "learning_rate": 3.941135039421204e-05,
+      "loss": 0.3946,
+      "step": 526
+    },
+    {
+      "epoch": 0.8489730165122835,
+      "grad_norm": 0.36645772020289197,
+      "learning_rate": 3.940591452551993e-05,
+      "loss": 0.4068,
+      "step": 527
+    },
+    {
+      "epoch": 0.8505839710028191,
+      "grad_norm": 0.4316640291709618,
+      "learning_rate": 3.9400454051712375e-05,
+      "loss": 0.4009,
+      "step": 528
+    },
+    {
+      "epoch": 0.8521949254933548,
+      "grad_norm": 0.34398127722906846,
+      "learning_rate": 3.939496897971281e-05,
+      "loss": 0.3907,
+      "step": 529
+    },
+    {
+      "epoch": 0.8538058799838905,
+      "grad_norm": 0.40687953024824,
+      "learning_rate": 3.938945931647585e-05,
+      "loss": 0.3901,
+      "step": 530
+    },
+    {
+      "epoch": 0.8554168344744261,
+      "grad_norm": 0.35490989918542215,
+      "learning_rate": 3.9383925068987306e-05,
+      "loss": 0.3928,
+      "step": 531
+    },
+    {
+      "epoch": 0.8570277889649618,
+      "grad_norm": 0.3548604235172149,
+      "learning_rate": 3.937836624426414e-05,
+      "loss": 0.4037,
+      "step": 532
+    },
+    {
+      "epoch": 0.8586387434554974,
+      "grad_norm": 0.3846710598884646,
+      "learning_rate": 3.9372782849354496e-05,
+      "loss": 0.4046,
+      "step": 533
+    },
+    {
+      "epoch": 0.860249697946033,
+      "grad_norm": 0.46337690960360517,
+      "learning_rate": 3.936717489133768e-05,
+      "loss": 0.4049,
+      "step": 534
+    },
+    {
+      "epoch": 0.8618606524365686,
+      "grad_norm": 0.3347915864544789,
+      "learning_rate": 3.936154237732409e-05,
+      "loss": 0.4199,
+      "step": 535
+    },
+    {
+      "epoch": 0.8634716069271043,
+      "grad_norm": 0.4109719486832367,
+      "learning_rate": 3.9355885314455316e-05,
+      "loss": 0.4145,
+      "step": 536
+    },
+    {
+      "epoch": 0.86508256141764,
+      "grad_norm": 0.4039943914910843,
+      "learning_rate": 3.935020370990405e-05,
+      "loss": 0.4094,
+      "step": 537
+    },
+    {
+      "epoch": 0.8666935159081756,
+      "grad_norm": 0.33381054763996676,
+      "learning_rate": 3.9344497570874105e-05,
+      "loss": 0.3997,
+      "step": 538
+    },
+    {
+      "epoch": 0.8683044703987113,
+      "grad_norm": 0.37323165887081494,
+      "learning_rate": 3.933876690460039e-05,
+      "loss": 0.4003,
+      "step": 539
+    },
+    {
+      "epoch": 0.8699154248892469,
+      "grad_norm": 0.4031148093361396,
+      "learning_rate": 3.9333011718348925e-05,
+      "loss": 0.4066,
+      "step": 540
+    },
+    {
+      "epoch": 0.8715263793797825,
+      "grad_norm": 0.35075255673198685,
+      "learning_rate": 3.932723201941683e-05,
+      "loss": 0.3941,
+      "step": 541
+    },
+    {
+      "epoch": 0.8731373338703181,
+      "grad_norm": 0.40922218226065216,
+      "learning_rate": 3.932142781513227e-05,
+      "loss": 0.4147,
+      "step": 542
+    },
+    {
+      "epoch": 0.8747482883608538,
+      "grad_norm": 0.3354099083614772,
+      "learning_rate": 3.9315599112854513e-05,
+      "loss": 0.4039,
+      "step": 543
+    },
+    {
+      "epoch": 0.8763592428513894,
+      "grad_norm": 0.3748405697078558,
+      "learning_rate": 3.930974591997387e-05,
+      "loss": 0.4069,
+      "step": 544
+    },
+    {
+      "epoch": 0.877970197341925,
+      "grad_norm": 0.3578120297382228,
+      "learning_rate": 3.930386824391173e-05,
+      "loss": 0.4237,
+      "step": 545
+    },
+    {
+      "epoch": 0.8795811518324608,
+      "grad_norm": 0.3498596617280635,
+      "learning_rate": 3.9297966092120494e-05,
+      "loss": 0.4201,
+      "step": 546
+    },
+    {
+      "epoch": 0.8811921063229964,
+      "grad_norm": 0.32782617117586416,
+      "learning_rate": 3.9292039472083604e-05,
+      "loss": 0.4052,
+      "step": 547
+    },
+    {
+      "epoch": 0.882803060813532,
+      "grad_norm": 0.32420421899810425,
+      "learning_rate": 3.928608839131554e-05,
+      "loss": 0.3967,
+      "step": 548
+    },
+    {
+      "epoch": 0.8844140153040677,
+      "grad_norm": 0.3610196304954956,
+      "learning_rate": 3.9280112857361785e-05,
+      "loss": 0.4073,
+      "step": 549
+    },
+    {
+      "epoch": 0.8860249697946033,
+      "grad_norm": 0.3462247564175289,
+      "learning_rate": 3.927411287779882e-05,
+      "loss": 0.419,
+      "step": 550
+    },
+    {
+      "epoch": 0.8876359242851389,
+      "grad_norm": 0.38210585668638986,
+      "learning_rate": 3.926808846023414e-05,
+      "loss": 0.398,
+      "step": 551
+    },
+    {
+      "epoch": 0.8892468787756745,
+      "grad_norm": 0.38393152658210905,
+      "learning_rate": 3.926203961230621e-05,
+      "loss": 0.425,
+      "step": 552
+    },
+    {
+      "epoch": 0.8908578332662103,
+      "grad_norm": 0.38660450532200585,
+      "learning_rate": 3.925596634168447e-05,
+      "loss": 0.4139,
+      "step": 553
+    },
+    {
+      "epoch": 0.8924687877567459,
+      "grad_norm": 0.4455772806196931,
+      "learning_rate": 3.9249868656069346e-05,
+      "loss": 0.4076,
+      "step": 554
+    },
+    {
+      "epoch": 0.8940797422472815,
+      "grad_norm": 0.30617516032050024,
+      "learning_rate": 3.9243746563192184e-05,
+      "loss": 0.4047,
+      "step": 555
+    },
+    {
+      "epoch": 0.8956906967378172,
+      "grad_norm": 0.4660640316019242,
+      "learning_rate": 3.923760007081532e-05,
+      "loss": 0.4023,
+      "step": 556
+    },
+    {
+      "epoch": 0.8973016512283528,
+      "grad_norm": 0.3429596427602513,
+      "learning_rate": 3.9231429186731996e-05,
+      "loss": 0.38,
+      "step": 557
+    },
+    {
+      "epoch": 0.8989126057188884,
+      "grad_norm": 0.41289741497886173,
+      "learning_rate": 3.922523391876638e-05,
+      "loss": 0.3973,
+      "step": 558
+    },
+    {
+      "epoch": 0.900523560209424,
+      "grad_norm": 0.4494827289948899,
+      "learning_rate": 3.921901427477358e-05,
+      "loss": 0.3997,
+      "step": 559
+    },
+    {
+      "epoch": 0.9021345146999598,
+      "grad_norm": 0.33632315973719623,
+      "learning_rate": 3.921277026263959e-05,
+      "loss": 0.409,
+      "step": 560
+    },
+    {
+      "epoch": 0.9037454691904954,
+      "grad_norm": 0.5070254674512086,
+      "learning_rate": 3.9206501890281305e-05,
+      "loss": 0.3968,
+      "step": 561
+    },
+    {
+      "epoch": 0.905356423681031,
+      "grad_norm": 0.4187055671468531,
+      "learning_rate": 3.920020916564652e-05,
+      "loss": 0.4142,
+      "step": 562
+    },
+    {
+      "epoch": 0.9069673781715667,
+      "grad_norm": 0.4460483809169745,
+      "learning_rate": 3.9193892096713886e-05,
+      "loss": 0.3944,
+      "step": 563
+    },
+    {
+      "epoch": 0.9085783326621023,
+      "grad_norm": 0.39905515724281176,
+      "learning_rate": 3.918755069149293e-05,
+      "loss": 0.4127,
+      "step": 564
+    },
+    {
+      "epoch": 0.9101892871526379,
+      "grad_norm": 0.3437786125257829,
+      "learning_rate": 3.9181184958024045e-05,
+      "loss": 0.4223,
+      "step": 565
+    },
+    {
+      "epoch": 0.9118002416431735,
+      "grad_norm": 0.3388066571758449,
+      "learning_rate": 3.917479490437845e-05,
+      "loss": 0.402,
+      "step": 566
+    },
+    {
+      "epoch": 0.9134111961337092,
+      "grad_norm": 0.38012535779270673,
+      "learning_rate": 3.9168380538658224e-05,
+      "loss": 0.4161,
+      "step": 567
+    },
+    {
+      "epoch": 0.9150221506242449,
+      "grad_norm": 0.3391977799991008,
+      "learning_rate": 3.916194186899626e-05,
+      "loss": 0.4038,
+      "step": 568
+    },
+    {
+      "epoch": 0.9166331051147805,
+      "grad_norm": 0.3646302581201986,
+      "learning_rate": 3.915547890355625e-05,
+      "loss": 0.4148,
+      "step": 569
+    },
+    {
+      "epoch": 0.9182440596053162,
+      "grad_norm": 0.3135049185381983,
+      "learning_rate": 3.914899165053272e-05,
+      "loss": 0.4111,
+      "step": 570
+    },
+    {
+      "epoch": 0.9198550140958518,
+      "grad_norm": 0.3535605107932701,
+      "learning_rate": 3.9142480118150964e-05,
+      "loss": 0.3874,
+      "step": 571
+    },
+    {
+      "epoch": 0.9214659685863874,
+      "grad_norm": 0.3418138874943149,
+      "learning_rate": 3.913594431466709e-05,
+      "loss": 0.395,
+      "step": 572
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 0.40111797732706755,
+      "learning_rate": 3.912938424836795e-05,
+      "loss": 0.412,
+      "step": 573
+    },
+    {
+      "epoch": 0.9246878775674587,
+      "grad_norm": 0.3565226075692776,
+      "learning_rate": 3.912279992757117e-05,
+      "loss": 0.4047,
+      "step": 574
+    },
+    {
+      "epoch": 0.9262988320579943,
+      "grad_norm": 0.43319435674290785,
+      "learning_rate": 3.911619136062515e-05,
+      "loss": 0.3937,
+      "step": 575
+    },
+    {
+      "epoch": 0.92790978654853,
+      "grad_norm": 0.40127033071910445,
+      "learning_rate": 3.9109558555909005e-05,
+      "loss": 0.4009,
+      "step": 576
+    },
+    {
+      "epoch": 0.9295207410390657,
+      "grad_norm": 0.40228778188193176,
+      "learning_rate": 3.910290152183258e-05,
+      "loss": 0.4035,
+      "step": 577
+    },
+    {
+      "epoch": 0.9311316955296013,
+      "grad_norm": 0.41880827476759785,
+      "learning_rate": 3.909622026683647e-05,
+      "loss": 0.3902,
+      "step": 578
+    },
+    {
+      "epoch": 0.9327426500201369,
+      "grad_norm": 0.4339612846528539,
+      "learning_rate": 3.9089514799391936e-05,
+      "loss": 0.427,
+      "step": 579
+    },
+    {
+      "epoch": 0.9343536045106726,
+      "grad_norm": 0.40920471224475635,
+      "learning_rate": 3.908278512800098e-05,
+      "loss": 0.4053,
+      "step": 580
+    },
+    {
+      "epoch": 0.9359645590012082,
+      "grad_norm": 0.3903565760977215,
+      "learning_rate": 3.907603126119627e-05,
+      "loss": 0.3851,
+      "step": 581
+    },
+    {
+      "epoch": 0.9375755134917438,
+      "grad_norm": 0.36452212891712393,
+      "learning_rate": 3.9069253207541165e-05,
+      "loss": 0.4102,
+      "step": 582
+    },
+    {
+      "epoch": 0.9391864679822794,
+      "grad_norm": 0.40963179376690034,
+      "learning_rate": 3.906245097562968e-05,
+      "loss": 0.4042,
+      "step": 583
+    },
+    {
+      "epoch": 0.9407974224728152,
+      "grad_norm": 0.47835605254275715,
+      "learning_rate": 3.9055624574086485e-05,
+      "loss": 0.4083,
+      "step": 584
+    },
+    {
+      "epoch": 0.9424083769633508,
+      "grad_norm": 0.35302712606627323,
+      "learning_rate": 3.9048774011566906e-05,
+      "loss": 0.399,
+      "step": 585
+    },
+    {
+      "epoch": 0.9440193314538864,
+      "grad_norm": 0.45147544517664695,
+      "learning_rate": 3.904189929675689e-05,
+      "loss": 0.3835,
+      "step": 586
+    },
+    {
+      "epoch": 0.9456302859444221,
+      "grad_norm": 0.3576207044412354,
+      "learning_rate": 3.903500043837302e-05,
+      "loss": 0.3983,
+      "step": 587
+    },
+    {
+      "epoch": 0.9472412404349577,
+      "grad_norm": 0.47971033814742575,
+      "learning_rate": 3.9028077445162486e-05,
+      "loss": 0.3876,
+      "step": 588
+    },
+    {
+      "epoch": 0.9488521949254933,
+      "grad_norm": 0.46648498519940407,
+      "learning_rate": 3.9021130325903076e-05,
+      "loss": 0.4104,
+      "step": 589
+    },
+    {
+      "epoch": 0.950463149416029,
+      "grad_norm": 0.36648880409549883,
+      "learning_rate": 3.9014159089403167e-05,
+      "loss": 0.4145,
+      "step": 590
+    },
+    {
+      "epoch": 0.9520741039065647,
+      "grad_norm": 0.4553737406525612,
+      "learning_rate": 3.9007163744501723e-05,
+      "loss": 0.4032,
+      "step": 591
+    },
+    {
+      "epoch": 0.9536850583971003,
+      "grad_norm": 0.4171288408426286,
+      "learning_rate": 3.900014430006827e-05,
+      "loss": 0.4131,
+      "step": 592
+    },
+    {
+      "epoch": 0.9552960128876359,
+      "grad_norm": 0.37296598077856796,
+      "learning_rate": 3.8993100765002886e-05,
+      "loss": 0.4098,
+      "step": 593
+    },
+    {
+      "epoch": 0.9569069673781716,
+      "grad_norm": 0.3668599752905086,
+      "learning_rate": 3.8986033148236206e-05,
+      "loss": 0.4006,
+      "step": 594
+    },
+    {
+      "epoch": 0.9585179218687072,
+      "grad_norm": 0.32605216315392055,
+      "learning_rate": 3.897894145872939e-05,
+      "loss": 0.389,
+      "step": 595
+    },
+    {
+      "epoch": 0.9601288763592428,
+      "grad_norm": 0.3408714924804675,
+      "learning_rate": 3.8971825705474104e-05,
+      "loss": 0.3854,
+      "step": 596
+    },
+    {
+      "epoch": 0.9617398308497785,
+      "grad_norm": 0.37330680663486493,
+      "learning_rate": 3.8964685897492566e-05,
+      "loss": 0.4025,
+      "step": 597
+    },
+    {
+      "epoch": 0.9633507853403142,
+      "grad_norm": 0.3549667131762123,
+      "learning_rate": 3.895752204383746e-05,
+      "loss": 0.4013,
+      "step": 598
+    },
+    {
+      "epoch": 0.9649617398308498,
+      "grad_norm": 0.3660249510077813,
+      "learning_rate": 3.895033415359196e-05,
+      "loss": 0.4057,
+      "step": 599
+    },
+    {
+      "epoch": 0.9665726943213854,
+      "grad_norm": 0.39751018282200024,
+      "learning_rate": 3.894312223586974e-05,
+      "loss": 0.4048,
+      "step": 600
+    },
+    {
+      "epoch": 0.9681836488119211,
+      "grad_norm": 0.31674666928386697,
+      "learning_rate": 3.8935886299814904e-05,
+      "loss": 0.3884,
+      "step": 601
+    },
+    {
+      "epoch": 0.9697946033024567,
+      "grad_norm": 0.42282251139146454,
+      "learning_rate": 3.8928626354602046e-05,
+      "loss": 0.373,
+      "step": 602
+    },
+    {
+      "epoch": 0.9714055577929923,
+      "grad_norm": 0.31609372557688337,
+      "learning_rate": 3.8921342409436175e-05,
+      "loss": 0.3946,
+      "step": 603
+    },
+    {
+      "epoch": 0.973016512283528,
+      "grad_norm": 0.44022277125892706,
+      "learning_rate": 3.891403447355274e-05,
+      "loss": 0.3961,
+      "step": 604
+    },
+    {
+      "epoch": 0.9746274667740636,
+      "grad_norm": 0.39405832611427605,
+      "learning_rate": 3.890670255621761e-05,
+      "loss": 0.3799,
+      "step": 605
+    },
+    {
+      "epoch": 0.9762384212645993,
+      "grad_norm": 0.3119144672899951,
+      "learning_rate": 3.889934666672706e-05,
+      "loss": 0.3837,
+      "step": 606
+    },
+    {
+      "epoch": 0.9778493757551349,
+      "grad_norm": 0.3896876780989498,
+      "learning_rate": 3.8891966814407745e-05,
+      "loss": 0.3998,
+      "step": 607
+    },
+    {
+      "epoch": 0.9794603302456706,
+      "grad_norm": 0.360270898809426,
+      "learning_rate": 3.8884563008616726e-05,
+      "loss": 0.4064,
+      "step": 608
+    },
+    {
+      "epoch": 0.9810712847362062,
+      "grad_norm": 0.42631951285751807,
+      "learning_rate": 3.8877135258741416e-05,
+      "loss": 0.392,
+      "step": 609
+    },
+    {
+      "epoch": 0.9826822392267418,
+      "grad_norm": 0.4415412873812909,
+      "learning_rate": 3.886968357419961e-05,
+      "loss": 0.4088,
+      "step": 610
+    },
+    {
+      "epoch": 0.9842931937172775,
+      "grad_norm": 0.36666132641993976,
+      "learning_rate": 3.886220796443942e-05,
+      "loss": 0.3939,
+      "step": 611
+    },
+    {
+      "epoch": 0.9859041482078131,
+      "grad_norm": 0.4202033590716815,
+      "learning_rate": 3.885470843893932e-05,
+      "loss": 0.3816,
+      "step": 612
+    },
+    {
+      "epoch": 0.9875151026983487,
+      "grad_norm": 0.3516656089440837,
+      "learning_rate": 3.884718500720808e-05,
+      "loss": 0.3865,
+      "step": 613
+    },
+    {
+      "epoch": 0.9891260571888845,
+      "grad_norm": 0.42108876415853347,
+      "learning_rate": 3.8839637678784815e-05,
+      "loss": 0.4058,
+      "step": 614
+    },
+    {
+      "epoch": 0.9907370116794201,
+      "grad_norm": 0.38505023099974517,
+      "learning_rate": 3.883206646323892e-05,
+      "loss": 0.4097,
+      "step": 615
+    },
+    {
+      "epoch": 0.9923479661699557,
+      "grad_norm": 0.40822421339312126,
+      "learning_rate": 3.882447137017007e-05,
+      "loss": 0.3982,
+      "step": 616
+    },
+    {
+      "epoch": 0.9939589206604913,
+      "grad_norm": 0.40702788912386473,
+      "learning_rate": 3.881685240920822e-05,
+      "loss": 0.4071,
+      "step": 617
+    },
+    {
+      "epoch": 0.995569875151027,
+      "grad_norm": 0.3723642576686159,
+      "learning_rate": 3.8809209590013606e-05,
+      "loss": 0.3869,
+      "step": 618
+    },
+    {
+      "epoch": 0.9971808296415626,
+      "grad_norm": 0.346526709604658,
+      "learning_rate": 3.8801542922276686e-05,
+      "loss": 0.3939,
+      "step": 619
+    },
+    {
+      "epoch": 0.9987917841320982,
+      "grad_norm": 0.4059535452467616,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 0.3963,
+      "step": 620
+    },
+    {
+      "epoch": 1.000402738622634,
+      "grad_norm": 0.41606971570933265,
+      "learning_rate": 3.8786138080089e-05,
+      "loss": 0.4643,
+      "step": 621
+    },
+    {
+      "epoch": 1.0020136931131696,
+      "grad_norm": 0.4642191778743235,
+      "learning_rate": 3.877839992517031e-05,
+      "loss": 0.3676,
+      "step": 622
+    },
+    {
+      "epoch": 1.0036246476037052,
+      "grad_norm": 0.4205733804052529,
+      "learning_rate": 3.8770637960773465e-05,
+      "loss": 0.358,
+      "step": 623
+    },
+    {
+      "epoch": 1.0052356020942408,
+      "grad_norm": 0.40959547235243765,
+      "learning_rate": 3.8762852196739994e-05,
+      "loss": 0.3635,
+      "step": 624
+    },
+    {
+      "epoch": 1.0068465565847764,
+      "grad_norm": 0.3791990728099719,
+      "learning_rate": 3.875504264294161e-05,
+      "loss": 0.3383,
+      "step": 625
+    },
+    {
+      "epoch": 1.0084575110753122,
+      "grad_norm": 0.5130910962490318,
+      "learning_rate": 3.8747209309280195e-05,
+      "loss": 0.3667,
+      "step": 626
+    },
+    {
+      "epoch": 1.0100684655658478,
+      "grad_norm": 0.39767900122923594,
+      "learning_rate": 3.873935220568776e-05,
+      "loss": 0.3597,
+      "step": 627
+    },
+    {
+      "epoch": 1.0116794200563834,
+      "grad_norm": 0.44532028330504037,
+      "learning_rate": 3.8731471342126495e-05,
+      "loss": 0.356,
+      "step": 628
+    },
+    {
+      "epoch": 1.013290374546919,
+      "grad_norm": 0.4136333579628707,
+      "learning_rate": 3.872356672858868e-05,
+      "loss": 0.3356,
+      "step": 629
+    },
+    {
+      "epoch": 1.0149013290374547,
+      "grad_norm": 0.40179390937304404,
+      "learning_rate": 3.871563837509672e-05,
+      "loss": 0.3317,
+      "step": 630
+    },
+    {
+      "epoch": 1.0165122835279903,
+      "grad_norm": 0.4824835553730714,
+      "learning_rate": 3.870768629170311e-05,
+      "loss": 0.3426,
+      "step": 631
+    },
+    {
+      "epoch": 1.018123238018526,
+      "grad_norm": 0.367967163870901,
+      "learning_rate": 3.869971048849046e-05,
+      "loss": 0.3409,
+      "step": 632
+    },
+    {
+      "epoch": 1.0197341925090617,
+      "grad_norm": 0.49092699271229584,
+      "learning_rate": 3.8691710975571425e-05,
+      "loss": 0.3658,
+      "step": 633
+    },
+    {
+      "epoch": 1.0213451469995973,
+      "grad_norm": 0.344714835725969,
+      "learning_rate": 3.8683687763088745e-05,
+      "loss": 0.3418,
+      "step": 634
+    },
+    {
+      "epoch": 1.022956101490133,
+      "grad_norm": 0.43359371603108976,
+      "learning_rate": 3.867564086121519e-05,
+      "loss": 0.3684,
+      "step": 635
+    },
+    {
+      "epoch": 1.0245670559806686,
+      "grad_norm": 0.3727355148405864,
+      "learning_rate": 3.866757028015357e-05,
+      "loss": 0.3721,
+      "step": 636
+    },
+    {
+      "epoch": 1.0261780104712042,
+      "grad_norm": 0.39174926745367894,
+      "learning_rate": 3.865947603013674e-05,
+      "loss": 0.3309,
+      "step": 637
+    },
+    {
+      "epoch": 1.0277889649617398,
+      "grad_norm": 0.3255706942445588,
+      "learning_rate": 3.865135812142753e-05,
+      "loss": 0.3429,
+      "step": 638
+    },
+    {
+      "epoch": 1.0293999194522754,
+      "grad_norm": 0.40845531163767246,
+      "learning_rate": 3.86432165643188e-05,
+      "loss": 0.3456,
+      "step": 639
+    },
+    {
+      "epoch": 1.0310108739428112,
+      "grad_norm": 0.32752549293401095,
+      "learning_rate": 3.863505136913337e-05,
+      "loss": 0.3561,
+      "step": 640
+    },
+    {
+      "epoch": 1.0326218284333468,
+      "grad_norm": 0.4306213011485377,
+      "learning_rate": 3.862686254622405e-05,
+      "loss": 0.3675,
+      "step": 641
+    },
+    {
+      "epoch": 1.0342327829238824,
+      "grad_norm": 0.3768521574814054,
+      "learning_rate": 3.8618650105973586e-05,
+      "loss": 0.3551,
+      "step": 642
+    },
+    {
+      "epoch": 1.035843737414418,
+      "grad_norm": 0.3398666804464596,
+      "learning_rate": 3.8610414058794695e-05,
+      "loss": 0.3195,
+      "step": 643
+    },
+    {
+      "epoch": 1.0374546919049537,
+      "grad_norm": 0.355463485377364,
+      "learning_rate": 3.860215441513001e-05,
+      "loss": 0.3427,
+      "step": 644
+    },
+    {
+      "epoch": 1.0390656463954893,
+      "grad_norm": 0.29625164040082436,
+      "learning_rate": 3.8593871185452074e-05,
+      "loss": 0.3446,
+      "step": 645
+    },
+    {
+      "epoch": 1.0406766008860249,
+      "grad_norm": 0.3640138737256208,
+      "learning_rate": 3.858556438026335e-05,
+      "loss": 0.3728,
+      "step": 646
+    },
+    {
+      "epoch": 1.0422875553765607,
+      "grad_norm": 0.32386946809285116,
+      "learning_rate": 3.8577234010096206e-05,
+      "loss": 0.3468,
+      "step": 647
+    },
+    {
+      "epoch": 1.0438985098670963,
+      "grad_norm": 0.36609144049870035,
+      "learning_rate": 3.856888008551285e-05,
+      "loss": 0.3617,
+      "step": 648
+    },
+    {
+      "epoch": 1.045509464357632,
+      "grad_norm": 0.4000523742064122,
+      "learning_rate": 3.856050261710539e-05,
+      "loss": 0.3591,
+      "step": 649
+    },
+    {
+      "epoch": 1.0471204188481675,
+      "grad_norm": 0.3847983140548682,
+      "learning_rate": 3.8552101615495755e-05,
+      "loss": 0.3566,
+      "step": 650
+    },
+    {
+      "epoch": 1.0487313733387031,
+      "grad_norm": 0.38930820839888375,
+      "learning_rate": 3.854367709133575e-05,
+      "loss": 0.3632,
+      "step": 651
+    },
+    {
+      "epoch": 1.0503423278292388,
+      "grad_norm": 0.3309535311423929,
+      "learning_rate": 3.853522905530698e-05,
+      "loss": 0.3477,
+      "step": 652
+    },
+    {
+      "epoch": 1.0519532823197744,
+      "grad_norm": 0.36099065328392765,
+      "learning_rate": 3.8526757518120846e-05,
+      "loss": 0.3498,
+      "step": 653
+    },
+    {
+      "epoch": 1.0535642368103102,
+      "grad_norm": 0.37075788058964326,
+      "learning_rate": 3.8518262490518585e-05,
+      "loss": 0.3658,
+      "step": 654
+    },
+    {
+      "epoch": 1.0551751913008458,
+      "grad_norm": 0.34165627904419094,
+      "learning_rate": 3.8509743983271196e-05,
+      "loss": 0.3506,
+      "step": 655
+    },
+    {
+      "epoch": 1.0567861457913814,
+      "grad_norm": 0.30950248679253206,
+      "learning_rate": 3.8501202007179447e-05,
+      "loss": 0.3307,
+      "step": 656
+    },
+    {
+      "epoch": 1.058397100281917,
+      "grad_norm": 0.3237159191997492,
+      "learning_rate": 3.8492636573073866e-05,
+      "loss": 0.3319,
+      "step": 657
+    },
+    {
+      "epoch": 1.0600080547724526,
+      "grad_norm": 0.3517039667110292,
+      "learning_rate": 3.8484047691814724e-05,
+      "loss": 0.3707,
+      "step": 658
+    },
+    {
+      "epoch": 1.0616190092629882,
+      "grad_norm": 0.339096140285127,
+      "learning_rate": 3.847543537429202e-05,
+      "loss": 0.3516,
+      "step": 659
+    },
+    {
+      "epoch": 1.0632299637535239,
+      "grad_norm": 0.38263230843949914,
+      "learning_rate": 3.8466799631425474e-05,
+      "loss": 0.3811,
+      "step": 660
+    },
+    {
+      "epoch": 1.0648409182440597,
+      "grad_norm": 0.36914183180517873,
+      "learning_rate": 3.8458140474164503e-05,
+      "loss": 0.3729,
+      "step": 661
+    },
+    {
+      "epoch": 1.0664518727345953,
+      "grad_norm": 0.3636472527013866,
+      "learning_rate": 3.8449457913488205e-05,
+      "loss": 0.344,
+      "step": 662
+    },
+    {
+      "epoch": 1.068062827225131,
+      "grad_norm": 0.3219280442977279,
+      "learning_rate": 3.8440751960405365e-05,
+      "loss": 0.3574,
+      "step": 663
+    },
+    {
+      "epoch": 1.0696737817156665,
+      "grad_norm": 0.3405685714287699,
+      "learning_rate": 3.843202262595442e-05,
+      "loss": 0.3402,
+      "step": 664
+    },
+    {
+      "epoch": 1.0712847362062021,
+      "grad_norm": 0.3498606308270493,
+      "learning_rate": 3.842326992120345e-05,
+      "loss": 0.3639,
+      "step": 665
+    },
+    {
+      "epoch": 1.0728956906967377,
+      "grad_norm": 0.32207988596751275,
+      "learning_rate": 3.841449385725018e-05,
+      "loss": 0.3409,
+      "step": 666
+    },
+    {
+      "epoch": 1.0745066451872733,
+      "grad_norm": 0.3491136718479199,
+      "learning_rate": 3.8405694445221924e-05,
+      "loss": 0.3462,
+      "step": 667
+    },
+    {
+      "epoch": 1.0761175996778092,
+      "grad_norm": 0.36346049492233184,
+      "learning_rate": 3.839687169627564e-05,
+      "loss": 0.3675,
+      "step": 668
+    },
+    {
+      "epoch": 1.0777285541683448,
+      "grad_norm": 0.3270189746678473,
+      "learning_rate": 3.838802562159783e-05,
+      "loss": 0.3428,
+      "step": 669
+    },
+    {
+      "epoch": 1.0793395086588804,
+      "grad_norm": 0.3572958227451133,
+      "learning_rate": 3.837915623240462e-05,
+      "loss": 0.3538,
+      "step": 670
+    },
+    {
+      "epoch": 1.080950463149416,
+      "grad_norm": 0.3646744219922745,
+      "learning_rate": 3.8370263539941647e-05,
+      "loss": 0.3521,
+      "step": 671
+    },
+    {
+      "epoch": 1.0825614176399516,
+      "grad_norm": 0.3403146272839342,
+      "learning_rate": 3.8361347555484136e-05,
+      "loss": 0.3447,
+      "step": 672
+    },
+    {
+      "epoch": 1.0841723721304872,
+      "grad_norm": 0.32652840028581526,
+      "learning_rate": 3.835240829033682e-05,
+      "loss": 0.3304,
+      "step": 673
+    },
+    {
+      "epoch": 1.085783326621023,
+      "grad_norm": 0.3400766178269568,
+      "learning_rate": 3.834344575583396e-05,
+      "loss": 0.3599,
+      "step": 674
+    },
+    {
+      "epoch": 1.0873942811115587,
+      "grad_norm": 0.38815523922016976,
+      "learning_rate": 3.833445996333932e-05,
+      "loss": 0.3359,
+      "step": 675
+    },
+    {
+      "epoch": 1.0890052356020943,
+      "grad_norm": 0.41704665471764435,
+      "learning_rate": 3.832545092424615e-05,
+      "loss": 0.3596,
+      "step": 676
+    },
+    {
+      "epoch": 1.09061619009263,
+      "grad_norm": 0.32531971390606135,
+      "learning_rate": 3.831641864997717e-05,
+      "loss": 0.3263,
+      "step": 677
+    },
+    {
+      "epoch": 1.0922271445831655,
+      "grad_norm": 0.3585802875329914,
+      "learning_rate": 3.830736315198457e-05,
+      "loss": 0.3641,
+      "step": 678
+    },
+    {
+      "epoch": 1.093838099073701,
+      "grad_norm": 0.31768810788040425,
+      "learning_rate": 3.8298284441749985e-05,
+      "loss": 0.3707,
+      "step": 679
+    },
+    {
+      "epoch": 1.0954490535642367,
+      "grad_norm": 0.32746796457773686,
+      "learning_rate": 3.828918253078448e-05,
+      "loss": 0.3433,
+      "step": 680
+    },
+    {
+      "epoch": 1.0970600080547726,
+      "grad_norm": 0.3224365136634425,
+      "learning_rate": 3.828005743062853e-05,
+      "loss": 0.3825,
+      "step": 681
+    },
+    {
+      "epoch": 1.0986709625453082,
+      "grad_norm": 0.3486055061501249,
+      "learning_rate": 3.827090915285202e-05,
+      "loss": 0.3626,
+      "step": 682
+    },
+    {
+      "epoch": 1.1002819170358438,
+      "grad_norm": 0.33829450630818075,
+      "learning_rate": 3.826173770905422e-05,
+      "loss": 0.35,
+      "step": 683
+    },
+    {
+      "epoch": 1.1018928715263794,
+      "grad_norm": 0.31882058047495504,
+      "learning_rate": 3.825254311086377e-05,
+      "loss": 0.3405,
+      "step": 684
+    },
+    {
+      "epoch": 1.103503826016915,
+      "grad_norm": 0.37226089185500716,
+      "learning_rate": 3.8243325369938674e-05,
+      "loss": 0.3469,
+      "step": 685
+    },
+    {
+      "epoch": 1.1051147805074506,
+      "grad_norm": 0.4101836669177939,
+      "learning_rate": 3.823408449796627e-05,
+      "loss": 0.345,
+      "step": 686
+    },
+    {
+      "epoch": 1.1067257349979862,
+      "grad_norm": 0.3137847522951066,
+      "learning_rate": 3.822482050666322e-05,
+      "loss": 0.3221,
+      "step": 687
+    },
+    {
+      "epoch": 1.108336689488522,
+      "grad_norm": 0.36142432810250685,
+      "learning_rate": 3.821553340777553e-05,
+      "loss": 0.3635,
+      "step": 688
+    },
+    {
+      "epoch": 1.1099476439790577,
+      "grad_norm": 0.33266260220336297,
+      "learning_rate": 3.820622321307847e-05,
+      "loss": 0.3545,
+      "step": 689
+    },
+    {
+      "epoch": 1.1115585984695933,
+      "grad_norm": 0.31351708791895283,
+      "learning_rate": 3.8196889934376617e-05,
+      "loss": 0.3681,
+      "step": 690
+    },
+    {
+      "epoch": 1.1131695529601289,
+      "grad_norm": 0.28063065842857104,
+      "learning_rate": 3.818753358350379e-05,
+      "loss": 0.3443,
+      "step": 691
+    },
+    {
+      "epoch": 1.1147805074506645,
+      "grad_norm": 0.36353977199002524,
+      "learning_rate": 3.8178154172323094e-05,
+      "loss": 0.3753,
+      "step": 692
+    },
+    {
+      "epoch": 1.1163914619412,
+      "grad_norm": 0.2768651552529601,
+      "learning_rate": 3.8168751712726846e-05,
+      "loss": 0.3446,
+      "step": 693
+    },
+    {
+      "epoch": 1.1180024164317357,
+      "grad_norm": 0.3342687611239813,
+      "learning_rate": 3.815932621663661e-05,
+      "loss": 0.3724,
+      "step": 694
+    },
+    {
+      "epoch": 1.1196133709222715,
+      "grad_norm": 0.33647821887748514,
+      "learning_rate": 3.814987769600312e-05,
+      "loss": 0.3455,
+      "step": 695
+    },
+    {
+      "epoch": 1.1212243254128071,
+      "grad_norm": 0.31273923882238536,
+      "learning_rate": 3.814040616280636e-05,
+      "loss": 0.3462,
+      "step": 696
+    },
+    {
+      "epoch": 1.1228352799033428,
+      "grad_norm": 0.34039417308107917,
+      "learning_rate": 3.8130911629055443e-05,
+      "loss": 0.3398,
+      "step": 697
+    },
+    {
+      "epoch": 1.1244462343938784,
+      "grad_norm": 0.33579456080236614,
+      "learning_rate": 3.812139410678866e-05,
+      "loss": 0.3596,
+      "step": 698
+    },
+    {
+      "epoch": 1.126057188884414,
+      "grad_norm": 0.3230165075161485,
+      "learning_rate": 3.811185360807347e-05,
+      "loss": 0.3497,
+      "step": 699
+    },
+    {
+      "epoch": 1.1276681433749496,
+      "grad_norm": 0.3241972598923344,
+      "learning_rate": 3.810229014500643e-05,
+      "loss": 0.3324,
+      "step": 700
+    },
+    {
+      "epoch": 1.1292790978654854,
+      "grad_norm": 0.3296772176461001,
+      "learning_rate": 3.809270372971323e-05,
+      "loss": 0.3506,
+      "step": 701
+    },
+    {
+      "epoch": 1.130890052356021,
+      "grad_norm": 0.37285713064822357,
+      "learning_rate": 3.8083094374348676e-05,
+      "loss": 0.3732,
+      "step": 702
+    },
+    {
+      "epoch": 1.1325010068465566,
+      "grad_norm": 0.2874339346820618,
+      "learning_rate": 3.807346209109663e-05,
+      "loss": 0.3483,
+      "step": 703
+    },
+    {
+      "epoch": 1.1341119613370922,
+      "grad_norm": 0.37713299713195747,
+      "learning_rate": 3.8063806892170055e-05,
+      "loss": 0.3707,
+      "step": 704
+    },
+    {
+      "epoch": 1.1357229158276279,
+      "grad_norm": 0.36683564876617375,
+      "learning_rate": 3.805412878981095e-05,
+      "loss": 0.3412,
+      "step": 705
+    },
+    {
+      "epoch": 1.1373338703181635,
+      "grad_norm": 0.37846828032041174,
+      "learning_rate": 3.804442779629035e-05,
+      "loss": 0.3823,
+      "step": 706
+    },
+    {
+      "epoch": 1.138944824808699,
+      "grad_norm": 0.39303680121813295,
+      "learning_rate": 3.803470392390834e-05,
+      "loss": 0.335,
+      "step": 707
+    },
+    {
+      "epoch": 1.140555779299235,
+      "grad_norm": 0.32030974316744143,
+      "learning_rate": 3.8024957184993986e-05,
+      "loss": 0.3496,
+      "step": 708
+    },
+    {
+      "epoch": 1.1421667337897705,
+      "grad_norm": 0.40038802750352454,
+      "learning_rate": 3.8015187591905356e-05,
+      "loss": 0.3631,
+      "step": 709
+    },
+    {
+      "epoch": 1.1437776882803061,
+      "grad_norm": 0.3889019319477048,
+      "learning_rate": 3.800539515702949e-05,
+      "loss": 0.3617,
+      "step": 710
+    },
+    {
+      "epoch": 1.1453886427708417,
+      "grad_norm": 0.3372232290880548,
+      "learning_rate": 3.799557989278241e-05,
+      "loss": 0.3763,
+      "step": 711
+    },
+    {
+      "epoch": 1.1469995972613773,
+      "grad_norm": 0.37103844691689164,
+      "learning_rate": 3.798574181160907e-05,
+      "loss": 0.3707,
+      "step": 712
+    },
+    {
+      "epoch": 1.148610551751913,
+      "grad_norm": 0.3193426723569192,
+      "learning_rate": 3.7975880925983345e-05,
+      "loss": 0.3419,
+      "step": 713
+    },
+    {
+      "epoch": 1.1502215062424486,
+      "grad_norm": 0.3735797639088716,
+      "learning_rate": 3.796599724840803e-05,
+      "loss": 0.3612,
+      "step": 714
+    },
+    {
+      "epoch": 1.1518324607329844,
+      "grad_norm": 0.3144326051598097,
+      "learning_rate": 3.795609079141484e-05,
+      "loss": 0.344,
+      "step": 715
+    },
+    {
+      "epoch": 1.15344341522352,
+      "grad_norm": 0.3226419021515401,
+      "learning_rate": 3.794616156756433e-05,
+      "loss": 0.3352,
+      "step": 716
+    },
+    {
+      "epoch": 1.1550543697140556,
+      "grad_norm": 0.2952244005389135,
+      "learning_rate": 3.793620958944596e-05,
+      "loss": 0.369,
+      "step": 717
+    },
+    {
+      "epoch": 1.1566653242045912,
+      "grad_norm": 0.3017482997087285,
+      "learning_rate": 3.792623486967802e-05,
+      "loss": 0.3554,
+      "step": 718
+    },
+    {
+      "epoch": 1.1582762786951268,
+      "grad_norm": 0.3332567088466206,
+      "learning_rate": 3.791623742090765e-05,
+      "loss": 0.3549,
+      "step": 719
+    },
+    {
+      "epoch": 1.1598872331856624,
+      "grad_norm": 0.3496541869504861,
+      "learning_rate": 3.790621725581079e-05,
+      "loss": 0.3477,
+      "step": 720
+    },
+    {
+      "epoch": 1.161498187676198,
+      "grad_norm": 0.29711261872162453,
+      "learning_rate": 3.7896174387092194e-05,
+      "loss": 0.338,
+      "step": 721
+    },
+    {
+      "epoch": 1.163109142166734,
+      "grad_norm": 0.35281267419179585,
+      "learning_rate": 3.788610882748539e-05,
+      "loss": 0.329,
+      "step": 722
+    },
+    {
+      "epoch": 1.1647200966572695,
+      "grad_norm": 0.40685923102090615,
+      "learning_rate": 3.78760205897527e-05,
+      "loss": 0.4075,
+      "step": 723
+    },
+    {
+      "epoch": 1.166331051147805,
+      "grad_norm": 0.32685402009954984,
+      "learning_rate": 3.786590968668518e-05,
+      "loss": 0.3441,
+      "step": 724
+    },
+    {
+      "epoch": 1.1679420056383407,
+      "grad_norm": 0.30475399585683605,
+      "learning_rate": 3.785577613110264e-05,
+      "loss": 0.3271,
+      "step": 725
+    },
+    {
+      "epoch": 1.1695529601288763,
+      "grad_norm": 0.39855984307309933,
+      "learning_rate": 3.784561993585358e-05,
+      "loss": 0.3795,
+      "step": 726
+    },
+    {
+      "epoch": 1.171163914619412,
+      "grad_norm": 0.320784768450466,
+      "learning_rate": 3.783544111381524e-05,
+      "loss": 0.3476,
+      "step": 727
+    },
+    {
+      "epoch": 1.1727748691099475,
+      "grad_norm": 0.398225649315675,
+      "learning_rate": 3.782523967789354e-05,
+      "loss": 0.3998,
+      "step": 728
+    },
+    {
+      "epoch": 1.1743858236004834,
+      "grad_norm": 0.33274129125722207,
+      "learning_rate": 3.781501564102305e-05,
+      "loss": 0.3594,
+      "step": 729
+    },
+    {
+      "epoch": 1.175996778091019,
+      "grad_norm": 0.29733991185377096,
+      "learning_rate": 3.7804769016167036e-05,
+      "loss": 0.329,
+      "step": 730
+    },
+    {
+      "epoch": 1.1776077325815546,
+      "grad_norm": 0.34843064282370634,
+      "learning_rate": 3.779449981631737e-05,
+      "loss": 0.3725,
+      "step": 731
+    },
+    {
+      "epoch": 1.1792186870720902,
+      "grad_norm": 0.29488591686422405,
+      "learning_rate": 3.7784208054494554e-05,
+      "loss": 0.3618,
+      "step": 732
+    },
+    {
+      "epoch": 1.1808296415626258,
+      "grad_norm": 0.32309697260812176,
+      "learning_rate": 3.777389374374772e-05,
+      "loss": 0.3567,
+      "step": 733
+    },
+    {
+      "epoch": 1.1824405960531614,
+      "grad_norm": 0.2949421891798728,
+      "learning_rate": 3.776355689715455e-05,
+      "loss": 0.3425,
+      "step": 734
+    },
+    {
+      "epoch": 1.184051550543697,
+      "grad_norm": 0.3234000905104195,
+      "learning_rate": 3.775319752782133e-05,
+      "loss": 0.3556,
+      "step": 735
+    },
+    {
+      "epoch": 1.1856625050342329,
+      "grad_norm": 0.31297392729871315,
+      "learning_rate": 3.7742815648882906e-05,
+      "loss": 0.3596,
+      "step": 736
+    },
+    {
+      "epoch": 1.1872734595247685,
+      "grad_norm": 0.3651886549670445,
+      "learning_rate": 3.773241127350264e-05,
+      "loss": 0.3675,
+      "step": 737
+    },
+    {
+      "epoch": 1.188884414015304,
+      "grad_norm": 0.30965174965919956,
+      "learning_rate": 3.772198441487243e-05,
+      "loss": 0.343,
+      "step": 738
+    },
+    {
+      "epoch": 1.1904953685058397,
+      "grad_norm": 0.3346542890794813,
+      "learning_rate": 3.771153508621269e-05,
+      "loss": 0.3554,
+      "step": 739
+    },
+    {
+      "epoch": 1.1921063229963753,
+      "grad_norm": 0.3132631499207701,
+      "learning_rate": 3.770106330077231e-05,
+      "loss": 0.3428,
+      "step": 740
+    },
+    {
+      "epoch": 1.193717277486911,
+      "grad_norm": 0.36626598122283643,
+      "learning_rate": 3.769056907182866e-05,
+      "loss": 0.3674,
+      "step": 741
+    },
+    {
+      "epoch": 1.1953282319774465,
+      "grad_norm": 0.3296955790235176,
+      "learning_rate": 3.768005241268757e-05,
+      "loss": 0.3422,
+      "step": 742
+    },
+    {
+      "epoch": 1.1969391864679824,
+      "grad_norm": 0.3214817268876617,
+      "learning_rate": 3.76695133366833e-05,
+      "loss": 0.3456,
+      "step": 743
+    },
+    {
+      "epoch": 1.198550140958518,
+      "grad_norm": 0.2844760445477903,
+      "learning_rate": 3.7658951857178544e-05,
+      "loss": 0.3428,
+      "step": 744
+    },
+    {
+      "epoch": 1.2001610954490536,
+      "grad_norm": 0.3371872324687368,
+      "learning_rate": 3.764836798756439e-05,
+      "loss": 0.3618,
+      "step": 745
+    },
+    {
+      "epoch": 1.2017720499395892,
+      "grad_norm": 0.2908980367000638,
+      "learning_rate": 3.763776174126031e-05,
+      "loss": 0.3335,
+      "step": 746
+    },
+    {
+      "epoch": 1.2033830044301248,
+      "grad_norm": 0.3261395185565908,
+      "learning_rate": 3.762713313171419e-05,
+      "loss": 0.3677,
+      "step": 747
+    },
+    {
+      "epoch": 1.2049939589206604,
+      "grad_norm": 0.31868194245423315,
+      "learning_rate": 3.761648217240221e-05,
+      "loss": 0.3708,
+      "step": 748
+    },
+    {
+      "epoch": 1.206604913411196,
+      "grad_norm": 0.29164880164956636,
+      "learning_rate": 3.760580887682892e-05,
+      "loss": 0.3267,
+      "step": 749
+    },
+    {
+      "epoch": 1.2082158679017319,
+      "grad_norm": 0.28510156507177037,
+      "learning_rate": 3.7595113258527206e-05,
+      "loss": 0.3437,
+      "step": 750
+    },
+    {
+      "epoch": 1.2098268223922675,
+      "grad_norm": 0.2957794927406389,
+      "learning_rate": 3.758439533105822e-05,
+      "loss": 0.3238,
+      "step": 751
+    },
+    {
+      "epoch": 1.211437776882803,
+      "grad_norm": 0.348201286744665,
+      "learning_rate": 3.757365510801143e-05,
+      "loss": 0.3804,
+      "step": 752
+    },
+    {
+      "epoch": 1.2130487313733387,
+      "grad_norm": 0.31209813548990145,
+      "learning_rate": 3.756289260300456e-05,
+      "loss": 0.3738,
+      "step": 753
+    },
+    {
+      "epoch": 1.2146596858638743,
+      "grad_norm": 0.3194574953898529,
+      "learning_rate": 3.755210782968358e-05,
+      "loss": 0.3411,
+      "step": 754
+    },
+    {
+      "epoch": 1.21627064035441,
+      "grad_norm": 0.315228642494671,
+      "learning_rate": 3.7541300801722715e-05,
+      "loss": 0.3442,
+      "step": 755
+    },
+    {
+      "epoch": 1.2178815948449455,
+      "grad_norm": 0.2897162651173343,
+      "learning_rate": 3.7530471532824385e-05,
+      "loss": 0.3496,
+      "step": 756
+    },
+    {
+      "epoch": 1.2194925493354813,
+      "grad_norm": 0.34000075648700634,
+      "learning_rate": 3.751962003671922e-05,
+      "loss": 0.3824,
+      "step": 757
+    },
+    {
+      "epoch": 1.221103503826017,
+      "grad_norm": 0.2976452200617564,
+      "learning_rate": 3.750874632716604e-05,
+      "loss": 0.3424,
+      "step": 758
+    },
+    {
+      "epoch": 1.2227144583165526,
+      "grad_norm": 0.29249500001948164,
+      "learning_rate": 3.74978504179518e-05,
+      "loss": 0.3723,
+      "step": 759
+    },
+    {
+      "epoch": 1.2243254128070882,
+      "grad_norm": 0.3355440693648072,
+      "learning_rate": 3.7486932322891646e-05,
+      "loss": 0.354,
+      "step": 760
+    },
+    {
+      "epoch": 1.2259363672976238,
+      "grad_norm": 0.2774173083048779,
+      "learning_rate": 3.747599205582882e-05,
+      "loss": 0.3388,
+      "step": 761
+    },
+    {
+      "epoch": 1.2275473217881594,
+      "grad_norm": 0.307261387050007,
+      "learning_rate": 3.746502963063469e-05,
+      "loss": 0.3536,
+      "step": 762
+    },
+    {
+      "epoch": 1.229158276278695,
+      "grad_norm": 0.2779362555047775,
+      "learning_rate": 3.745404506120872e-05,
+      "loss": 0.3418,
+      "step": 763
+    },
+    {
+      "epoch": 1.2307692307692308,
+      "grad_norm": 0.2970571621131108,
+      "learning_rate": 3.744303836147844e-05,
+      "loss": 0.3885,
+      "step": 764
+    },
+    {
+      "epoch": 1.2323801852597664,
+      "grad_norm": 0.299475415748665,
+      "learning_rate": 3.743200954539945e-05,
+      "loss": 0.3357,
+      "step": 765
+    },
+    {
+      "epoch": 1.233991139750302,
+      "grad_norm": 0.28066511695868956,
+      "learning_rate": 3.7420958626955395e-05,
+      "loss": 0.3186,
+      "step": 766
+    },
+    {
+      "epoch": 1.2356020942408377,
+      "grad_norm": 0.32650220462849366,
+      "learning_rate": 3.7409885620157925e-05,
+      "loss": 0.374,
+      "step": 767
+    },
+    {
+      "epoch": 1.2372130487313733,
+      "grad_norm": 0.33812078920671956,
+      "learning_rate": 3.739879053904672e-05,
+      "loss": 0.3508,
+      "step": 768
+    },
+    {
+      "epoch": 1.2388240032219089,
+      "grad_norm": 0.34383445727592993,
+      "learning_rate": 3.738767339768942e-05,
+      "loss": 0.3779,
+      "step": 769
+    },
+    {
+      "epoch": 1.2404349577124447,
+      "grad_norm": 0.3202923747231639,
+      "learning_rate": 3.737653421018168e-05,
+      "loss": 0.3426,
+      "step": 770
+    },
+    {
+      "epoch": 1.2420459122029803,
+      "grad_norm": 0.35409199577635775,
+      "learning_rate": 3.736537299064706e-05,
+      "loss": 0.3572,
+      "step": 771
+    },
+    {
+      "epoch": 1.243656866693516,
+      "grad_norm": 0.3436578084892543,
+      "learning_rate": 3.735418975323708e-05,
+      "loss": 0.3388,
+      "step": 772
+    },
+    {
+      "epoch": 1.2452678211840515,
+      "grad_norm": 0.29421403073429603,
+      "learning_rate": 3.734298451213117e-05,
+      "loss": 0.3496,
+      "step": 773
+    },
+    {
+      "epoch": 1.2468787756745872,
+      "grad_norm": 0.3674454300701155,
+      "learning_rate": 3.7331757281536665e-05,
+      "loss": 0.3457,
+      "step": 774
+    },
+    {
+      "epoch": 1.2484897301651228,
+      "grad_norm": 0.2658262781818514,
+      "learning_rate": 3.732050807568878e-05,
+      "loss": 0.3353,
+      "step": 775
+    },
+    {
+      "epoch": 1.2501006846556586,
+      "grad_norm": 0.3558402384689967,
+      "learning_rate": 3.7309236908850574e-05,
+      "loss": 0.3531,
+      "step": 776
+    },
+    {
+      "epoch": 1.251711639146194,
+      "grad_norm": 0.2968753461483703,
+      "learning_rate": 3.729794379531299e-05,
+      "loss": 0.3463,
+      "step": 777
+    },
+    {
+      "epoch": 1.2533225936367298,
+      "grad_norm": 0.3068748390315143,
+      "learning_rate": 3.7286628749394754e-05,
+      "loss": 0.3495,
+      "step": 778
+    },
+    {
+      "epoch": 1.2549335481272654,
+      "grad_norm": 0.2853593430828942,
+      "learning_rate": 3.727529178544243e-05,
+      "loss": 0.349,
+      "step": 779
+    },
+    {
+      "epoch": 1.256544502617801,
+      "grad_norm": 0.36814006315805403,
+      "learning_rate": 3.726393291783036e-05,
+      "loss": 0.3683,
+      "step": 780
+    },
+    {
+      "epoch": 1.2581554571083367,
+      "grad_norm": 0.34204662393847385,
+      "learning_rate": 3.7252552160960666e-05,
+      "loss": 0.3395,
+      "step": 781
+    },
+    {
+      "epoch": 1.2597664115988723,
+      "grad_norm": 0.25952729501669486,
+      "learning_rate": 3.724114952926322e-05,
+      "loss": 0.3426,
+      "step": 782
+    },
+    {
+      "epoch": 1.261377366089408,
+      "grad_norm": 0.3434914251843314,
+      "learning_rate": 3.722972503719561e-05,
+      "loss": 0.3277,
+      "step": 783
+    },
+    {
+      "epoch": 1.2629883205799435,
+      "grad_norm": 0.36202104323985773,
+      "learning_rate": 3.7218278699243176e-05,
+      "loss": 0.3722,
+      "step": 784
+    },
+    {
+      "epoch": 1.2645992750704793,
+      "grad_norm": 0.35145897106947177,
+      "learning_rate": 3.7206810529918935e-05,
+      "loss": 0.3711,
+      "step": 785
+    },
+    {
+      "epoch": 1.266210229561015,
+      "grad_norm": 0.3119487302914577,
+      "learning_rate": 3.7195320543763596e-05,
+      "loss": 0.3352,
+      "step": 786
+    },
+    {
+      "epoch": 1.2678211840515505,
+      "grad_norm": 0.35035093132752193,
+      "learning_rate": 3.718380875534552e-05,
+      "loss": 0.3743,
+      "step": 787
+    },
+    {
+      "epoch": 1.2694321385420861,
+      "grad_norm": 0.3418946608898889,
+      "learning_rate": 3.71722751792607e-05,
+      "loss": 0.3429,
+      "step": 788
+    },
+    {
+      "epoch": 1.2710430930326218,
+      "grad_norm": 0.31270337962677086,
+      "learning_rate": 3.7160719830132794e-05,
+      "loss": 0.344,
+      "step": 789
+    },
+    {
+      "epoch": 1.2726540475231576,
+      "grad_norm": 0.3274369342826082,
+      "learning_rate": 3.714914272261302e-05,
+      "loss": 0.3499,
+      "step": 790
+    },
+    {
+      "epoch": 1.2742650020136932,
+      "grad_norm": 0.34598565436756357,
+      "learning_rate": 3.7137543871380224e-05,
+      "loss": 0.3829,
+      "step": 791
+    },
+    {
+      "epoch": 1.2758759565042288,
+      "grad_norm": 0.28889266818776443,
+      "learning_rate": 3.712592329114079e-05,
+      "loss": 0.3381,
+      "step": 792
+    },
+    {
+      "epoch": 1.2774869109947644,
+      "grad_norm": 0.3717773811020485,
+      "learning_rate": 3.7114280996628666e-05,
+      "loss": 0.3721,
+      "step": 793
+    },
+    {
+      "epoch": 1.2790978654853,
+      "grad_norm": 0.37135926966232047,
+      "learning_rate": 3.710261700260534e-05,
+      "loss": 0.3718,
+      "step": 794
+    },
+    {
+      "epoch": 1.2807088199758356,
+      "grad_norm": 0.33049624744013434,
+      "learning_rate": 3.7090931323859794e-05,
+      "loss": 0.3586,
+      "step": 795
+    },
+    {
+      "epoch": 1.2823197744663712,
+      "grad_norm": 0.3101839548871945,
+      "learning_rate": 3.707922397520852e-05,
+      "loss": 0.3397,
+      "step": 796
+    },
+    {
+      "epoch": 1.283930728956907,
+      "grad_norm": 0.30737107820444354,
+      "learning_rate": 3.706749497149547e-05,
+      "loss": 0.336,
+      "step": 797
+    },
+    {
+      "epoch": 1.2855416834474427,
+      "grad_norm": 0.3019490959641579,
+      "learning_rate": 3.705574432759208e-05,
+      "loss": 0.3512,
+      "step": 798
+    },
+    {
+      "epoch": 1.2871526379379783,
+      "grad_norm": 0.3503088570735374,
+      "learning_rate": 3.70439720583972e-05,
+      "loss": 0.3822,
+      "step": 799
+    },
+    {
+      "epoch": 1.288763592428514,
+      "grad_norm": 0.311137345604332,
+      "learning_rate": 3.70321781788371e-05,
+      "loss": 0.3641,
+      "step": 800
+    },
+    {
+      "epoch": 1.2903745469190495,
+      "grad_norm": 0.31960209562676384,
+      "learning_rate": 3.702036270386547e-05,
+      "loss": 0.3813,
+      "step": 801
+    },
+    {
+      "epoch": 1.2919855014095851,
+      "grad_norm": 0.3425160205698806,
+      "learning_rate": 3.700852564846335e-05,
+      "loss": 0.354,
+      "step": 802
+    },
+    {
+      "epoch": 1.2935964559001207,
+      "grad_norm": 0.30986308980528005,
+      "learning_rate": 3.6996667027639174e-05,
+      "loss": 0.3579,
+      "step": 803
+    },
+    {
+      "epoch": 1.2952074103906566,
+      "grad_norm": 0.3908250025109488,
+      "learning_rate": 3.6984786856428705e-05,
+      "loss": 0.3531,
+      "step": 804
+    },
+    {
+      "epoch": 1.2968183648811922,
+      "grad_norm": 0.3475929563748349,
+      "learning_rate": 3.697288514989502e-05,
+      "loss": 0.374,
+      "step": 805
+    },
+    {
+      "epoch": 1.2984293193717278,
+      "grad_norm": 0.36294126212027755,
+      "learning_rate": 3.696096192312852e-05,
+      "loss": 0.3576,
+      "step": 806
+    },
+    {
+      "epoch": 1.3000402738622634,
+      "grad_norm": 0.32338264256863225,
+      "learning_rate": 3.694901719124688e-05,
+      "loss": 0.3312,
+      "step": 807
+    },
+    {
+      "epoch": 1.301651228352799,
+      "grad_norm": 0.35527955178483733,
+      "learning_rate": 3.6937050969395055e-05,
+      "loss": 0.3836,
+      "step": 808
+    },
+    {
+      "epoch": 1.3032621828433346,
+      "grad_norm": 0.3628976303009859,
+      "learning_rate": 3.6925063272745214e-05,
+      "loss": 0.3629,
+      "step": 809
+    },
+    {
+      "epoch": 1.3048731373338702,
+      "grad_norm": 0.3022352084491082,
+      "learning_rate": 3.6913054116496797e-05,
+      "loss": 0.3605,
+      "step": 810
+    },
+    {
+      "epoch": 1.306484091824406,
+      "grad_norm": 0.3393294470418059,
+      "learning_rate": 3.690102351587643e-05,
+      "loss": 0.3512,
+      "step": 811
+    },
+    {
+      "epoch": 1.3080950463149417,
+      "grad_norm": 0.28363397227191833,
+      "learning_rate": 3.688897148613794e-05,
+      "loss": 0.3429,
+      "step": 812
+    },
+    {
+      "epoch": 1.3097060008054773,
+      "grad_norm": 0.33963097783454826,
+      "learning_rate": 3.687689804256229e-05,
+      "loss": 0.3332,
+      "step": 813
+    },
+    {
+      "epoch": 1.3113169552960129,
+      "grad_norm": 0.3600086874002223,
+      "learning_rate": 3.6864803200457646e-05,
+      "loss": 0.3625,
+      "step": 814
+    },
+    {
+      "epoch": 1.3129279097865485,
+      "grad_norm": 0.2816481879831538,
+      "learning_rate": 3.685268697515928e-05,
+      "loss": 0.3684,
+      "step": 815
+    },
+    {
+      "epoch": 1.314538864277084,
+      "grad_norm": 0.3139691237821825,
+      "learning_rate": 3.684054938202956e-05,
+      "loss": 0.3393,
+      "step": 816
+    },
+    {
+      "epoch": 1.3161498187676197,
+      "grad_norm": 0.29165261191102926,
+      "learning_rate": 3.682839043645798e-05,
+      "loss": 0.3228,
+      "step": 817
+    },
+    {
+      "epoch": 1.3177607732581555,
+      "grad_norm": 0.2718062507025103,
+      "learning_rate": 3.681621015386108e-05,
+      "loss": 0.3769,
+      "step": 818
+    },
+    {
+      "epoch": 1.3193717277486912,
+      "grad_norm": 0.3635524032153051,
+      "learning_rate": 3.680400854968246e-05,
+      "loss": 0.36,
+      "step": 819
+    },
+    {
+      "epoch": 1.3209826822392268,
+      "grad_norm": 0.28780321276593324,
+      "learning_rate": 3.679178563939278e-05,
+      "loss": 0.3479,
+      "step": 820
+    },
+    {
+      "epoch": 1.3225936367297624,
+      "grad_norm": 0.3238210640271435,
+      "learning_rate": 3.677954143848967e-05,
+      "loss": 0.3767,
+      "step": 821
+    },
+    {
+      "epoch": 1.324204591220298,
+      "grad_norm": 0.3080441012695841,
+      "learning_rate": 3.676727596249779e-05,
+      "loss": 0.3561,
+      "step": 822
+    },
+    {
+      "epoch": 1.3258155457108336,
+      "grad_norm": 0.32081488942296216,
+      "learning_rate": 3.675498922696876e-05,
+      "loss": 0.3692,
+      "step": 823
+    },
+    {
+      "epoch": 1.3274265002013692,
+      "grad_norm": 0.3278673509495895,
+      "learning_rate": 3.6742681247481144e-05,
+      "loss": 0.3231,
+      "step": 824
+    },
+    {
+      "epoch": 1.329037454691905,
+      "grad_norm": 0.3096169164731479,
+      "learning_rate": 3.6730352039640476e-05,
+      "loss": 0.3821,
+      "step": 825
+    },
+    {
+      "epoch": 1.3306484091824407,
+      "grad_norm": 0.33349256230646473,
+      "learning_rate": 3.671800161907917e-05,
+      "loss": 0.3268,
+      "step": 826
+    },
+    {
+      "epoch": 1.3322593636729763,
+      "grad_norm": 0.3295767828308085,
+      "learning_rate": 3.6705630001456556e-05,
+      "loss": 0.335,
+      "step": 827
+    },
+    {
+      "epoch": 1.3338703181635119,
+      "grad_norm": 0.2912807748587227,
+      "learning_rate": 3.669323720245884e-05,
+      "loss": 0.3409,
+      "step": 828
+    },
+    {
+      "epoch": 1.3354812726540475,
+      "grad_norm": 0.30649973311303386,
+      "learning_rate": 3.668082323779907e-05,
+      "loss": 0.3512,
+      "step": 829
+    },
+    {
+      "epoch": 1.337092227144583,
+      "grad_norm": 0.2959906133761005,
+      "learning_rate": 3.6668388123217154e-05,
+      "loss": 0.3612,
+      "step": 830
+    },
+    {
+      "epoch": 1.3387031816351187,
+      "grad_norm": 0.29861778533728894,
+      "learning_rate": 3.6655931874479783e-05,
+      "loss": 0.3356,
+      "step": 831
+    },
+    {
+      "epoch": 1.3403141361256545,
+      "grad_norm": 0.36174475564403585,
+      "learning_rate": 3.664345450738048e-05,
+      "loss": 0.3857,
+      "step": 832
+    },
+    {
+      "epoch": 1.3419250906161901,
+      "grad_norm": 0.2964990585195445,
+      "learning_rate": 3.663095603773952e-05,
+      "loss": 0.3703,
+      "step": 833
+    },
+    {
+      "epoch": 1.3435360451067258,
+      "grad_norm": 0.3444117342564229,
+      "learning_rate": 3.6618436481403945e-05,
+      "loss": 0.3477,
+      "step": 834
+    },
+    {
+      "epoch": 1.3451469995972614,
+      "grad_norm": 0.2921618958671085,
+      "learning_rate": 3.6605895854247534e-05,
+      "loss": 0.3663,
+      "step": 835
+    },
+    {
+      "epoch": 1.346757954087797,
+      "grad_norm": 0.3135599574497408,
+      "learning_rate": 3.659333417217076e-05,
+      "loss": 0.3556,
+      "step": 836
+    },
+    {
+      "epoch": 1.3483689085783326,
+      "grad_norm": 0.28548160767634045,
+      "learning_rate": 3.658075145110083e-05,
+      "loss": 0.3401,
+      "step": 837
+    },
+    {
+      "epoch": 1.3499798630688682,
+      "grad_norm": 0.33185921583222566,
+      "learning_rate": 3.6568147706991616e-05,
+      "loss": 0.3671,
+      "step": 838
+    },
+    {
+      "epoch": 1.351590817559404,
+      "grad_norm": 0.29106614505484624,
+      "learning_rate": 3.655552295582361e-05,
+      "loss": 0.3751,
+      "step": 839
+    },
+    {
+      "epoch": 1.3532017720499396,
+      "grad_norm": 0.32102312365154567,
+      "learning_rate": 3.654287721360398e-05,
+      "loss": 0.3854,
+      "step": 840
+    },
+    {
+      "epoch": 1.3548127265404752,
+      "grad_norm": 0.28054378659585727,
+      "learning_rate": 3.653021049636648e-05,
+      "loss": 0.3331,
+      "step": 841
+    },
+    {
+      "epoch": 1.3564236810310109,
+      "grad_norm": 0.3182024975537731,
+      "learning_rate": 3.65175228201715e-05,
+      "loss": 0.3423,
+      "step": 842
+    },
+    {
+      "epoch": 1.3580346355215465,
+      "grad_norm": 0.29631447266060057,
+      "learning_rate": 3.650481420110596e-05,
+      "loss": 0.3454,
+      "step": 843
+    },
+    {
+      "epoch": 1.3596455900120823,
+      "grad_norm": 0.39903373865432623,
+      "learning_rate": 3.6492084655283355e-05,
+      "loss": 0.369,
+      "step": 844
+    },
+    {
+      "epoch": 1.3612565445026177,
+      "grad_norm": 0.29989201553650346,
+      "learning_rate": 3.647933419884371e-05,
+      "loss": 0.3304,
+      "step": 845
+    },
+    {
+      "epoch": 1.3628674989931535,
+      "grad_norm": 0.35056143882565566,
+      "learning_rate": 3.646656284795357e-05,
+      "loss": 0.3723,
+      "step": 846
+    },
+    {
+      "epoch": 1.3644784534836891,
+      "grad_norm": 0.3441551125299749,
+      "learning_rate": 3.645377061880595e-05,
+      "loss": 0.3557,
+      "step": 847
+    },
+    {
+      "epoch": 1.3660894079742247,
+      "grad_norm": 0.3274845208948303,
+      "learning_rate": 3.644095752762036e-05,
+      "loss": 0.3399,
+      "step": 848
+    },
+    {
+      "epoch": 1.3677003624647603,
+      "grad_norm": 0.3024421060736529,
+      "learning_rate": 3.642812359064276e-05,
+      "loss": 0.3596,
+      "step": 849
+    },
+    {
+      "epoch": 1.369311316955296,
+      "grad_norm": 0.31600146060711926,
+      "learning_rate": 3.641526882414553e-05,
+      "loss": 0.3423,
+      "step": 850
+    },
+    {
+      "epoch": 1.3709222714458318,
+      "grad_norm": 0.30071046711567967,
+      "learning_rate": 3.640239324442746e-05,
+      "loss": 0.3424,
+      "step": 851
+    },
+    {
+      "epoch": 1.3725332259363672,
+      "grad_norm": 0.3203775229603514,
+      "learning_rate": 3.638949686781374e-05,
+      "loss": 0.3563,
+      "step": 852
+    },
+    {
+      "epoch": 1.374144180426903,
+      "grad_norm": 0.2791617091294289,
+      "learning_rate": 3.6376579710655915e-05,
+      "loss": 0.3473,
+      "step": 853
+    },
+    {
+      "epoch": 1.3757551349174386,
+      "grad_norm": 0.3262506737731935,
+      "learning_rate": 3.63636417893319e-05,
+      "loss": 0.3535,
+      "step": 854
+    },
+    {
+      "epoch": 1.3773660894079742,
+      "grad_norm": 0.37963304107161205,
+      "learning_rate": 3.6350683120245906e-05,
+      "loss": 0.3877,
+      "step": 855
+    },
+    {
+      "epoch": 1.3789770438985098,
+      "grad_norm": 0.26345344067107274,
+      "learning_rate": 3.633770371982848e-05,
+      "loss": 0.3344,
+      "step": 856
+    },
+    {
+      "epoch": 1.3805879983890454,
+      "grad_norm": 0.3589958673951733,
+      "learning_rate": 3.632470360453643e-05,
+      "loss": 0.3624,
+      "step": 857
+    },
+    {
+      "epoch": 1.3821989528795813,
+      "grad_norm": 0.2870645821680028,
+      "learning_rate": 3.631168279085286e-05,
+      "loss": 0.3442,
+      "step": 858
+    },
+    {
+      "epoch": 1.3838099073701167,
+      "grad_norm": 0.3154900838253569,
+      "learning_rate": 3.629864129528709e-05,
+      "loss": 0.3544,
+      "step": 859
+    },
+    {
+      "epoch": 1.3854208618606525,
+      "grad_norm": 0.32587561605795656,
+      "learning_rate": 3.6285579134374655e-05,
+      "loss": 0.3365,
+      "step": 860
+    },
+    {
+      "epoch": 1.387031816351188,
+      "grad_norm": 0.28524552874452264,
+      "learning_rate": 3.627249632467733e-05,
+      "loss": 0.3513,
+      "step": 861
+    },
+    {
+      "epoch": 1.3886427708417237,
+      "grad_norm": 0.3043822768243937,
+      "learning_rate": 3.625939288278304e-05,
+      "loss": 0.3505,
+      "step": 862
+    },
+    {
+      "epoch": 1.3902537253322593,
+      "grad_norm": 0.3283867868538535,
+      "learning_rate": 3.6246268825305886e-05,
+      "loss": 0.328,
+      "step": 863
+    },
+    {
+      "epoch": 1.391864679822795,
+      "grad_norm": 0.27028256421648467,
+      "learning_rate": 3.6233124168886094e-05,
+      "loss": 0.337,
+      "step": 864
+    },
+    {
+      "epoch": 1.3934756343133308,
+      "grad_norm": 0.31823659410494604,
+      "learning_rate": 3.621995893019003e-05,
+      "loss": 0.3483,
+      "step": 865
+    },
+    {
+      "epoch": 1.3950865888038662,
+      "grad_norm": 0.2853031194725599,
+      "learning_rate": 3.620677312591012e-05,
+      "loss": 0.3672,
+      "step": 866
+    },
+    {
+      "epoch": 1.396697543294402,
+      "grad_norm": 0.3703793870257088,
+      "learning_rate": 3.61935667727649e-05,
+      "loss": 0.3466,
+      "step": 867
+    },
+    {
+      "epoch": 1.3983084977849376,
+      "grad_norm": 0.3194336151665436,
+      "learning_rate": 3.6180339887498953e-05,
+      "loss": 0.3268,
+      "step": 868
+    },
+    {
+      "epoch": 1.3999194522754732,
+      "grad_norm": 0.33386957554709146,
+      "learning_rate": 3.616709248688288e-05,
+      "loss": 0.3419,
+      "step": 869
+    },
+    {
+      "epoch": 1.4015304067660088,
+      "grad_norm": 0.30916892946803903,
+      "learning_rate": 3.61538245877133e-05,
+      "loss": 0.3446,
+      "step": 870
+    },
+    {
+      "epoch": 1.4031413612565444,
+      "grad_norm": 0.29073836232064454,
+      "learning_rate": 3.614053620681284e-05,
+      "loss": 0.3247,
+      "step": 871
+    },
+    {
+      "epoch": 1.4047523157470803,
+      "grad_norm": 0.3341684803586826,
+      "learning_rate": 3.6127227361030076e-05,
+      "loss": 0.3811,
+      "step": 872
+    },
+    {
+      "epoch": 1.4063632702376159,
+      "grad_norm": 0.31287760315769936,
+      "learning_rate": 3.611389806723953e-05,
+      "loss": 0.3606,
+      "step": 873
+    },
+    {
+      "epoch": 1.4079742247281515,
+      "grad_norm": 0.3063668454004494,
+      "learning_rate": 3.610054834234167e-05,
+      "loss": 0.3625,
+      "step": 874
+    },
+    {
+      "epoch": 1.409585179218687,
+      "grad_norm": 0.3168490265948009,
+      "learning_rate": 3.608717820326285e-05,
+      "loss": 0.373,
+      "step": 875
+    },
+    {
+      "epoch": 1.4111961337092227,
+      "grad_norm": 0.3722746286249624,
+      "learning_rate": 3.6073787666955326e-05,
+      "loss": 0.3765,
+      "step": 876
+    },
+    {
+      "epoch": 1.4128070881997583,
+      "grad_norm": 0.2752473149335231,
+      "learning_rate": 3.6060376750397187e-05,
+      "loss": 0.3144,
+      "step": 877
+    },
+    {
+      "epoch": 1.414418042690294,
+      "grad_norm": 0.35239569804844095,
+      "learning_rate": 3.6046945470592395e-05,
+      "loss": 0.3759,
+      "step": 878
+    },
+    {
+      "epoch": 1.4160289971808298,
+      "grad_norm": 0.3259555420279317,
+      "learning_rate": 3.6033493844570704e-05,
+      "loss": 0.3643,
+      "step": 879
+    },
+    {
+      "epoch": 1.4176399516713654,
+      "grad_norm": 0.28683471938292443,
+      "learning_rate": 3.602002188938769e-05,
+      "loss": 0.3337,
+      "step": 880
+    },
+    {
+      "epoch": 1.419250906161901,
+      "grad_norm": 0.32883729844020054,
+      "learning_rate": 3.6006529622124694e-05,
+      "loss": 0.345,
+      "step": 881
+    },
+    {
+      "epoch": 1.4208618606524366,
+      "grad_norm": 0.2947797991693968,
+      "learning_rate": 3.59930170598888e-05,
+      "loss": 0.3356,
+      "step": 882
+    },
+    {
+      "epoch": 1.4224728151429722,
+      "grad_norm": 0.2826135754232561,
+      "learning_rate": 3.597948421981283e-05,
+      "loss": 0.3107,
+      "step": 883
+    },
+    {
+      "epoch": 1.4240837696335078,
+      "grad_norm": 0.3096811885777464,
+      "learning_rate": 3.596593111905533e-05,
+      "loss": 0.3754,
+      "step": 884
+    },
+    {
+      "epoch": 1.4256947241240434,
+      "grad_norm": 0.2813838416003549,
+      "learning_rate": 3.5952357774800526e-05,
+      "loss": 0.3484,
+      "step": 885
+    },
+    {
+      "epoch": 1.4273056786145792,
+      "grad_norm": 0.3147799342493275,
+      "learning_rate": 3.5938764204258306e-05,
+      "loss": 0.3464,
+      "step": 886
+    },
+    {
+      "epoch": 1.4289166331051149,
+      "grad_norm": 0.27371781021865504,
+      "learning_rate": 3.5925150424664206e-05,
+      "loss": 0.3348,
+      "step": 887
+    },
+    {
+      "epoch": 1.4305275875956505,
+      "grad_norm": 0.2929800361539922,
+      "learning_rate": 3.591151645327939e-05,
+      "loss": 0.3638,
+      "step": 888
+    },
+    {
+      "epoch": 1.432138542086186,
+      "grad_norm": 0.3091933237540173,
+      "learning_rate": 3.589786230739062e-05,
+      "loss": 0.3751,
+      "step": 889
+    },
+    {
+      "epoch": 1.4337494965767217,
+      "grad_norm": 0.30038308262391095,
+      "learning_rate": 3.5884188004310244e-05,
+      "loss": 0.3685,
+      "step": 890
+    },
+    {
+      "epoch": 1.4353604510672573,
+      "grad_norm": 0.2702977590995165,
+      "learning_rate": 3.587049356137615e-05,
+      "loss": 0.3421,
+      "step": 891
+    },
+    {
+      "epoch": 1.436971405557793,
+      "grad_norm": 0.31963763306891574,
+      "learning_rate": 3.5856778995951794e-05,
+      "loss": 0.3574,
+      "step": 892
+    },
+    {
+      "epoch": 1.4385823600483287,
+      "grad_norm": 0.2789441221702962,
+      "learning_rate": 3.58430443254261e-05,
+      "loss": 0.3707,
+      "step": 893
+    },
+    {
+      "epoch": 1.4401933145388643,
+      "grad_norm": 0.32066299952132377,
+      "learning_rate": 3.582928956721352e-05,
+      "loss": 0.3041,
+      "step": 894
+    },
+    {
+      "epoch": 1.4418042690294,
+      "grad_norm": 0.2805130759068286,
+      "learning_rate": 3.581551473875397e-05,
+      "loss": 0.3738,
+      "step": 895
+    },
+    {
+      "epoch": 1.4434152235199356,
+      "grad_norm": 0.32812904359127354,
+      "learning_rate": 3.580171985751281e-05,
+      "loss": 0.346,
+      "step": 896
+    },
+    {
+      "epoch": 1.4450261780104712,
+      "grad_norm": 0.27111375470537674,
+      "learning_rate": 3.578790494098081e-05,
+      "loss": 0.3306,
+      "step": 897
+    },
+    {
+      "epoch": 1.4466371325010068,
+      "grad_norm": 0.38149897363461827,
+      "learning_rate": 3.5774070006674164e-05,
+      "loss": 0.3516,
+      "step": 898
+    },
+    {
+      "epoch": 1.4482480869915424,
+      "grad_norm": 0.2892182943854451,
+      "learning_rate": 3.576021507213444e-05,
+      "loss": 0.3594,
+      "step": 899
+    },
+    {
+      "epoch": 1.4498590414820782,
+      "grad_norm": 0.3372883515472436,
+      "learning_rate": 3.574634015492857e-05,
+      "loss": 0.3532,
+      "step": 900
+    },
+    {
+      "epoch": 1.4514699959726138,
+      "grad_norm": 0.26358417061513745,
+      "learning_rate": 3.57324452726488e-05,
+      "loss": 0.3288,
+      "step": 901
+    },
+    {
+      "epoch": 1.4530809504631494,
+      "grad_norm": 0.31566876672673594,
+      "learning_rate": 3.571853044291271e-05,
+      "loss": 0.3362,
+      "step": 902
+    },
+    {
+      "epoch": 1.454691904953685,
+      "grad_norm": 0.32363249428251567,
+      "learning_rate": 3.5704595683363187e-05,
+      "loss": 0.3781,
+      "step": 903
+    },
+    {
+      "epoch": 1.4563028594442207,
+      "grad_norm": 0.3230781499413541,
+      "learning_rate": 3.569064101166835e-05,
+      "loss": 0.3541,
+      "step": 904
+    },
+    {
+      "epoch": 1.4579138139347563,
+      "grad_norm": 0.3156721284052765,
+      "learning_rate": 3.567666644552159e-05,
+      "loss": 0.3609,
+      "step": 905
+    },
+    {
+      "epoch": 1.4595247684252919,
+      "grad_norm": 0.24785776600781328,
+      "learning_rate": 3.566267200264151e-05,
+      "loss": 0.3287,
+      "step": 906
+    },
+    {
+      "epoch": 1.4611357229158277,
+      "grad_norm": 0.3480224482497804,
+      "learning_rate": 3.564865770077193e-05,
+      "loss": 0.3661,
+      "step": 907
+    },
+    {
+      "epoch": 1.4627466774063633,
+      "grad_norm": 0.2646519126771718,
+      "learning_rate": 3.563462355768184e-05,
+      "loss": 0.3346,
+      "step": 908
+    },
+    {
+      "epoch": 1.464357631896899,
+      "grad_norm": 0.3172875453636419,
+      "learning_rate": 3.562056959116538e-05,
+      "loss": 0.358,
+      "step": 909
+    },
+    {
+      "epoch": 1.4659685863874345,
+      "grad_norm": 0.3022971119624446,
+      "learning_rate": 3.560649581904184e-05,
+      "loss": 0.3738,
+      "step": 910
+    },
+    {
+      "epoch": 1.4675795408779702,
+      "grad_norm": 0.3594652213181189,
+      "learning_rate": 3.559240225915561e-05,
+      "loss": 0.364,
+      "step": 911
+    },
+    {
+      "epoch": 1.4691904953685058,
+      "grad_norm": 0.29252293105230515,
+      "learning_rate": 3.557828892937617e-05,
+      "loss": 0.3324,
+      "step": 912
+    },
+    {
+      "epoch": 1.4708014498590414,
+      "grad_norm": 0.33755638959110873,
+      "learning_rate": 3.5564155847598085e-05,
+      "loss": 0.3463,
+      "step": 913
+    },
+    {
+      "epoch": 1.4724124043495772,
+      "grad_norm": 0.28967085216937327,
+      "learning_rate": 3.555000303174093e-05,
+      "loss": 0.3529,
+      "step": 914
+    },
+    {
+      "epoch": 1.4740233588401128,
+      "grad_norm": 0.3396436534360357,
+      "learning_rate": 3.553583049974933e-05,
+      "loss": 0.3821,
+      "step": 915
+    },
+    {
+      "epoch": 1.4756343133306484,
+      "grad_norm": 0.27806521100206133,
+      "learning_rate": 3.55216382695929e-05,
+      "loss": 0.3594,
+      "step": 916
+    },
+    {
+      "epoch": 1.477245267821184,
+      "grad_norm": 0.30304501601924577,
+      "learning_rate": 3.550742635926622e-05,
+      "loss": 0.332,
+      "step": 917
+    },
+    {
+      "epoch": 1.4788562223117196,
+      "grad_norm": 0.28397653690883456,
+      "learning_rate": 3.549319478678885e-05,
+      "loss": 0.3478,
+      "step": 918
+    },
+    {
+      "epoch": 1.4804671768022553,
+      "grad_norm": 0.3534165168824709,
+      "learning_rate": 3.547894357020525e-05,
+      "loss": 0.3502,
+      "step": 919
+    },
+    {
+      "epoch": 1.4820781312927909,
+      "grad_norm": 0.30300554924196443,
+      "learning_rate": 3.546467272758479e-05,
+      "loss": 0.3427,
+      "step": 920
+    },
+    {
+      "epoch": 1.4836890857833267,
+      "grad_norm": 0.2821833761521695,
+      "learning_rate": 3.5450382277021745e-05,
+      "loss": 0.336,
+      "step": 921
+    },
+    {
+      "epoch": 1.4853000402738623,
+      "grad_norm": 0.3067490433896909,
+      "learning_rate": 3.543607223663524e-05,
+      "loss": 0.3478,
+      "step": 922
+    },
+    {
+      "epoch": 1.486910994764398,
+      "grad_norm": 0.35973200105393427,
+      "learning_rate": 3.542174262456924e-05,
+      "loss": 0.3744,
+      "step": 923
+    },
+    {
+      "epoch": 1.4885219492549335,
+      "grad_norm": 0.3624137649224464,
+      "learning_rate": 3.540739345899252e-05,
+      "loss": 0.3801,
+      "step": 924
+    },
+    {
+      "epoch": 1.4901329037454691,
+      "grad_norm": 0.30165267187605543,
+      "learning_rate": 3.5393024758098645e-05,
+      "loss": 0.3286,
+      "step": 925
+    },
+    {
+      "epoch": 1.491743858236005,
+      "grad_norm": 0.28896166859244726,
+      "learning_rate": 3.537863654010597e-05,
+      "loss": 0.3387,
+      "step": 926
+    },
+    {
+      "epoch": 1.4933548127265404,
+      "grad_norm": 0.3064859067705259,
+      "learning_rate": 3.5364228823257565e-05,
+      "loss": 0.3568,
+      "step": 927
+    },
+    {
+      "epoch": 1.4949657672170762,
+      "grad_norm": 0.32160688390289044,
+      "learning_rate": 3.534980162582124e-05,
+      "loss": 0.3688,
+      "step": 928
+    },
+    {
+      "epoch": 1.4965767217076118,
+      "grad_norm": 0.3013538609491535,
+      "learning_rate": 3.5335354966089514e-05,
+      "loss": 0.3376,
+      "step": 929
+    },
+    {
+      "epoch": 1.4981876761981474,
+      "grad_norm": 0.3057453257092756,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 0.3426,
+      "step": 930
+    },
+    {
+      "epoch": 1.499798630688683,
+      "grad_norm": 0.3070652933287628,
+      "learning_rate": 3.530640333303323e-05,
+      "loss": 0.3572,
+      "step": 931
+    },
+    {
+      "epoch": 1.5014095851792186,
+      "grad_norm": 0.303242364147413,
+      "learning_rate": 3.5291898396416984e-05,
+      "loss": 0.3766,
+      "step": 932
+    },
+    {
+      "epoch": 1.5030205396697545,
+      "grad_norm": 0.30911330099386514,
+      "learning_rate": 3.5277374070921904e-05,
+      "loss": 0.3606,
+      "step": 933
+    },
+    {
+      "epoch": 1.5046314941602899,
+      "grad_norm": 0.26387501860850937,
+      "learning_rate": 3.5262830374963636e-05,
+      "loss": 0.3132,
+      "step": 934
+    },
+    {
+      "epoch": 1.5062424486508257,
+      "grad_norm": 0.31366992473226274,
+      "learning_rate": 3.524826732698241e-05,
+      "loss": 0.345,
+      "step": 935
+    },
+    {
+      "epoch": 1.5078534031413613,
+      "grad_norm": 0.36305603623046623,
+      "learning_rate": 3.523368494544298e-05,
+      "loss": 0.3481,
+      "step": 936
+    },
+    {
+      "epoch": 1.509464357631897,
+      "grad_norm": 0.3028100237793981,
+      "learning_rate": 3.521908324883462e-05,
+      "loss": 0.3494,
+      "step": 937
+    },
+    {
+      "epoch": 1.5110753121224325,
+      "grad_norm": 0.34440302527582833,
+      "learning_rate": 3.520446225567108e-05,
+      "loss": 0.3635,
+      "step": 938
+    },
+    {
+      "epoch": 1.5126862666129681,
+      "grad_norm": 0.2883451420805599,
+      "learning_rate": 3.518982198449059e-05,
+      "loss": 0.356,
+      "step": 939
+    },
+    {
+      "epoch": 1.514297221103504,
+      "grad_norm": 0.3013627859596229,
+      "learning_rate": 3.517516245385582e-05,
+      "loss": 0.3289,
+      "step": 940
+    },
+    {
+      "epoch": 1.5159081755940393,
+      "grad_norm": 0.3018221027079222,
+      "learning_rate": 3.516048368235386e-05,
+      "loss": 0.3604,
+      "step": 941
+    },
+    {
+      "epoch": 1.5175191300845752,
+      "grad_norm": 0.30511935925361944,
+      "learning_rate": 3.5145785688596184e-05,
+      "loss": 0.3234,
+      "step": 942
+    },
+    {
+      "epoch": 1.5191300845751108,
+      "grad_norm": 0.3007192258635043,
+      "learning_rate": 3.513106849121866e-05,
+      "loss": 0.3544,
+      "step": 943
+    },
+    {
+      "epoch": 1.5207410390656464,
+      "grad_norm": 0.2641898297732925,
+      "learning_rate": 3.5116332108881486e-05,
+      "loss": 0.3408,
+      "step": 944
+    },
+    {
+      "epoch": 1.522351993556182,
+      "grad_norm": 0.3410443834630316,
+      "learning_rate": 3.5101576560269195e-05,
+      "loss": 0.3533,
+      "step": 945
+    },
+    {
+      "epoch": 1.5239629480467176,
+      "grad_norm": 0.28303295707105,
+      "learning_rate": 3.508680186409062e-05,
+      "loss": 0.3367,
+      "step": 946
+    },
+    {
+      "epoch": 1.5255739025372534,
+      "grad_norm": 0.3345908288880972,
+      "learning_rate": 3.507200803907886e-05,
+      "loss": 0.3861,
+      "step": 947
+    },
+    {
+      "epoch": 1.5271848570277888,
+      "grad_norm": 2.8670378142195654,
+      "learning_rate": 3.505719510399129e-05,
+      "loss": 0.4205,
+      "step": 948
+    },
+    {
+      "epoch": 1.5287958115183247,
+      "grad_norm": 0.4210670854339028,
+      "learning_rate": 3.504236307760949e-05,
+      "loss": 0.3497,
+      "step": 949
+    },
+    {
+      "epoch": 1.5304067660088603,
+      "grad_norm": 0.2976798975045218,
+      "learning_rate": 3.502751197873927e-05,
+      "loss": 0.3504,
+      "step": 950
+    },
+    {
+      "epoch": 1.5320177204993959,
+      "grad_norm": 0.3591627202188393,
+      "learning_rate": 3.501264182621061e-05,
+      "loss": 0.3746,
+      "step": 951
+    },
+    {
+      "epoch": 1.5336286749899315,
+      "grad_norm": 0.3210770679629104,
+      "learning_rate": 3.499775263887764e-05,
+      "loss": 0.3468,
+      "step": 952
+    },
+    {
+      "epoch": 1.535239629480467,
+      "grad_norm": 0.33550226114199827,
+      "learning_rate": 3.4982844435618643e-05,
+      "loss": 0.3508,
+      "step": 953
+    },
+    {
+      "epoch": 1.536850583971003,
+      "grad_norm": 0.32509668801494995,
+      "learning_rate": 3.4967917235336e-05,
+      "loss": 0.3508,
+      "step": 954
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 0.338038211291058,
+      "learning_rate": 3.4952971056956186e-05,
+      "loss": 0.3496,
+      "step": 955
+    },
+    {
+      "epoch": 1.5400724929520742,
+      "grad_norm": 0.3141904048502338,
+      "learning_rate": 3.4938005919429724e-05,
+      "loss": 0.3418,
+      "step": 956
+    },
+    {
+      "epoch": 1.5416834474426098,
+      "grad_norm": 0.2815520350465673,
+      "learning_rate": 3.49230218417312e-05,
+      "loss": 0.3312,
+      "step": 957
+    },
+    {
+      "epoch": 1.5432944019331454,
+      "grad_norm": 0.36948297955531856,
+      "learning_rate": 3.49080188428592e-05,
+      "loss": 0.3603,
+      "step": 958
+    },
+    {
+      "epoch": 1.544905356423681,
+      "grad_norm": 0.3084260379112921,
+      "learning_rate": 3.489299694183629e-05,
+      "loss": 0.357,
+      "step": 959
+    },
+    {
+      "epoch": 1.5465163109142166,
+      "grad_norm": 0.3549939328103352,
+      "learning_rate": 3.4877956157709024e-05,
+      "loss": 0.3565,
+      "step": 960
+    },
+    {
+      "epoch": 1.5481272654047524,
+      "grad_norm": 0.28779131717746126,
+      "learning_rate": 3.4862896509547886e-05,
+      "loss": 0.3468,
+      "step": 961
+    },
+    {
+      "epoch": 1.5497382198952878,
+      "grad_norm": 0.3232280682124179,
+      "learning_rate": 3.4847818016447284e-05,
+      "loss": 0.362,
+      "step": 962
+    },
+    {
+      "epoch": 1.5513491743858236,
+      "grad_norm": 0.30311024972844897,
+      "learning_rate": 3.483272069752551e-05,
+      "loss": 0.3175,
+      "step": 963
+    },
+    {
+      "epoch": 1.5529601288763593,
+      "grad_norm": 0.325020064316006,
+      "learning_rate": 3.481760457192474e-05,
+      "loss": 0.3729,
+      "step": 964
+    },
+    {
+      "epoch": 1.5545710833668949,
+      "grad_norm": 0.308178367218141,
+      "learning_rate": 3.4802469658810984e-05,
+      "loss": 0.3433,
+      "step": 965
+    },
+    {
+      "epoch": 1.5561820378574305,
+      "grad_norm": 0.3139480265008325,
+      "learning_rate": 3.478731597737407e-05,
+      "loss": 0.3455,
+      "step": 966
+    },
+    {
+      "epoch": 1.557792992347966,
+      "grad_norm": 0.32224249506902913,
+      "learning_rate": 3.4772143546827635e-05,
+      "loss": 0.3526,
+      "step": 967
+    },
+    {
+      "epoch": 1.559403946838502,
+      "grad_norm": 0.38577637148488075,
+      "learning_rate": 3.475695238640908e-05,
+      "loss": 0.3668,
+      "step": 968
+    },
+    {
+      "epoch": 1.5610149013290373,
+      "grad_norm": 0.3160738969499725,
+      "learning_rate": 3.474174251537956e-05,
+      "loss": 0.3389,
+      "step": 969
+    },
+    {
+      "epoch": 1.5626258558195731,
+      "grad_norm": 0.348308437450324,
+      "learning_rate": 3.4726513953023944e-05,
+      "loss": 0.3841,
+      "step": 970
+    },
+    {
+      "epoch": 1.5642368103101087,
+      "grad_norm": 0.3192517617664544,
+      "learning_rate": 3.471126671865082e-05,
+      "loss": 0.3518,
+      "step": 971
+    },
+    {
+      "epoch": 1.5658477648006444,
+      "grad_norm": 0.3315975814114015,
+      "learning_rate": 3.469600083159241e-05,
+      "loss": 0.3501,
+      "step": 972
+    },
+    {
+      "epoch": 1.56745871929118,
+      "grad_norm": 0.3429653180075547,
+      "learning_rate": 3.468071631120464e-05,
+      "loss": 0.3673,
+      "step": 973
+    },
+    {
+      "epoch": 1.5690696737817156,
+      "grad_norm": 0.2936081834216568,
+      "learning_rate": 3.466541317686702e-05,
+      "loss": 0.3726,
+      "step": 974
+    },
+    {
+      "epoch": 1.5706806282722514,
+      "grad_norm": 0.2870880348305391,
+      "learning_rate": 3.465009144798268e-05,
+      "loss": 0.3473,
+      "step": 975
+    },
+    {
+      "epoch": 1.5722915827627868,
+      "grad_norm": 0.31292709888328285,
+      "learning_rate": 3.4634751143978317e-05,
+      "loss": 0.3494,
+      "step": 976
+    },
+    {
+      "epoch": 1.5739025372533226,
+      "grad_norm": 0.3344663746754926,
+      "learning_rate": 3.461939228430419e-05,
+      "loss": 0.3594,
+      "step": 977
+    },
+    {
+      "epoch": 1.5755134917438582,
+      "grad_norm": 0.3243550278021458,
+      "learning_rate": 3.460401488843407e-05,
+      "loss": 0.3842,
+      "step": 978
+    },
+    {
+      "epoch": 1.5771244462343939,
+      "grad_norm": 0.28798557728982177,
+      "learning_rate": 3.458861897586524e-05,
+      "loss": 0.3267,
+      "step": 979
+    },
+    {
+      "epoch": 1.5787354007249297,
+      "grad_norm": 0.3366500322864135,
+      "learning_rate": 3.4573204566118476e-05,
+      "loss": 0.3929,
+      "step": 980
+    },
+    {
+      "epoch": 1.580346355215465,
+      "grad_norm": 0.32529408271268495,
+      "learning_rate": 3.455777167873798e-05,
+      "loss": 0.3484,
+      "step": 981
+    },
+    {
+      "epoch": 1.581957309706001,
+      "grad_norm": 0.3119605526527732,
+      "learning_rate": 3.454232033329139e-05,
+      "loss": 0.3498,
+      "step": 982
+    },
+    {
+      "epoch": 1.5835682641965363,
+      "grad_norm": 0.3155331712421188,
+      "learning_rate": 3.452685054936976e-05,
+      "loss": 0.3472,
+      "step": 983
+    },
+    {
+      "epoch": 1.5851792186870721,
+      "grad_norm": 0.2915430968431837,
+      "learning_rate": 3.4511362346587504e-05,
+      "loss": 0.3407,
+      "step": 984
+    },
+    {
+      "epoch": 1.5867901731776077,
+      "grad_norm": 0.3137253719745165,
+      "learning_rate": 3.44958557445824e-05,
+      "loss": 0.355,
+      "step": 985
+    },
+    {
+      "epoch": 1.5884011276681433,
+      "grad_norm": 0.3188776562958085,
+      "learning_rate": 3.4480330763015564e-05,
+      "loss": 0.3558,
+      "step": 986
+    },
+    {
+      "epoch": 1.5900120821586792,
+      "grad_norm": 0.3129181773241444,
+      "learning_rate": 3.4464787421571395e-05,
+      "loss": 0.3584,
+      "step": 987
+    },
+    {
+      "epoch": 1.5916230366492146,
+      "grad_norm": 0.3197808791803997,
+      "learning_rate": 3.444922573995758e-05,
+      "loss": 0.3538,
+      "step": 988
+    },
+    {
+      "epoch": 1.5932339911397504,
+      "grad_norm": 0.29636880126371085,
+      "learning_rate": 3.443364573790507e-05,
+      "loss": 0.3371,
+      "step": 989
+    },
+    {
+      "epoch": 1.5948449456302858,
+      "grad_norm": 0.30432129789976997,
+      "learning_rate": 3.4418047435168025e-05,
+      "loss": 0.3474,
+      "step": 990
+    },
+    {
+      "epoch": 1.5964559001208216,
+      "grad_norm": 0.26763048833686587,
+      "learning_rate": 3.4402430851523824e-05,
+      "loss": 0.3407,
+      "step": 991
+    },
+    {
+      "epoch": 1.5980668546113572,
+      "grad_norm": 0.27316984480658096,
+      "learning_rate": 3.438679600677303e-05,
+      "loss": 0.3422,
+      "step": 992
+    },
+    {
+      "epoch": 1.5996778091018928,
+      "grad_norm": 0.28509710496947416,
+      "learning_rate": 3.437114292073933e-05,
+      "loss": 0.3438,
+      "step": 993
+    },
+    {
+      "epoch": 1.6012887635924287,
+      "grad_norm": 0.2820397711644288,
+      "learning_rate": 3.435547161326958e-05,
+      "loss": 0.3758,
+      "step": 994
+    },
+    {
+      "epoch": 1.602899718082964,
+      "grad_norm": 0.27746741501473066,
+      "learning_rate": 3.43397821042337e-05,
+      "loss": 0.357,
+      "step": 995
+    },
+    {
+      "epoch": 1.6045106725734999,
+      "grad_norm": 0.329470817735326,
+      "learning_rate": 3.4324074413524725e-05,
+      "loss": 0.3732,
+      "step": 996
+    },
+    {
+      "epoch": 1.6061216270640355,
+      "grad_norm": 0.2883189529050093,
+      "learning_rate": 3.430834856105871e-05,
+      "loss": 0.3113,
+      "step": 997
+    },
+    {
+      "epoch": 1.607732581554571,
+      "grad_norm": 0.30812120490912964,
+      "learning_rate": 3.429260456677475e-05,
+      "loss": 0.3413,
+      "step": 998
+    },
+    {
+      "epoch": 1.6093435360451067,
+      "grad_norm": 0.3016161794889211,
+      "learning_rate": 3.4276842450634964e-05,
+      "loss": 0.3732,
+      "step": 999
+    },
+    {
+      "epoch": 1.6109544905356423,
+      "grad_norm": 0.30608350201014334,
+      "learning_rate": 3.4261062232624405e-05,
+      "loss": 0.345,
+      "step": 1000
+    },
+    {
+      "epoch": 1.6125654450261782,
+      "grad_norm": 0.287021015994268,
+      "learning_rate": 3.4245263932751124e-05,
+      "loss": 0.3675,
+      "step": 1001
+    },
+    {
+      "epoch": 1.6141763995167135,
+      "grad_norm": 0.29513629823587945,
+      "learning_rate": 3.4229447571046055e-05,
+      "loss": 0.3726,
+      "step": 1002
+    },
+    {
+      "epoch": 1.6157873540072494,
+      "grad_norm": 0.28497512595967667,
+      "learning_rate": 3.421361316756307e-05,
+      "loss": 0.3417,
+      "step": 1003
+    },
+    {
+      "epoch": 1.617398308497785,
+      "grad_norm": 0.27472976631378027,
+      "learning_rate": 3.4197760742378886e-05,
+      "loss": 0.3826,
+      "step": 1004
+    },
+    {
+      "epoch": 1.6190092629883206,
+      "grad_norm": 0.32583486377402415,
+      "learning_rate": 3.4181890315593104e-05,
+      "loss": 0.3429,
+      "step": 1005
+    },
+    {
+      "epoch": 1.6206202174788562,
+      "grad_norm": 0.3158909644175841,
+      "learning_rate": 3.41660019073281e-05,
+      "loss": 0.339,
+      "step": 1006
+    },
+    {
+      "epoch": 1.6222311719693918,
+      "grad_norm": 0.2874717190588641,
+      "learning_rate": 3.41500955377291e-05,
+      "loss": 0.3448,
+      "step": 1007
+    },
+    {
+      "epoch": 1.6238421264599276,
+      "grad_norm": 0.32333997510660806,
+      "learning_rate": 3.413417122696408e-05,
+      "loss": 0.3574,
+      "step": 1008
+    },
+    {
+      "epoch": 1.625453080950463,
+      "grad_norm": 0.32153546396431953,
+      "learning_rate": 3.411822899522376e-05,
+      "loss": 0.343,
+      "step": 1009
+    },
+    {
+      "epoch": 1.6270640354409989,
+      "grad_norm": 0.33818702288315083,
+      "learning_rate": 3.410226886272159e-05,
+      "loss": 0.3592,
+      "step": 1010
+    },
+    {
+      "epoch": 1.6286749899315345,
+      "grad_norm": 0.2779482392923402,
+      "learning_rate": 3.408629084969372e-05,
+      "loss": 0.3575,
+      "step": 1011
+    },
+    {
+      "epoch": 1.63028594442207,
+      "grad_norm": 0.2958503892623944,
+      "learning_rate": 3.407029497639896e-05,
+      "loss": 0.3431,
+      "step": 1012
+    },
+    {
+      "epoch": 1.6318968989126057,
+      "grad_norm": 0.27651659921242794,
+      "learning_rate": 3.405428126311878e-05,
+      "loss": 0.3476,
+      "step": 1013
+    },
+    {
+      "epoch": 1.6335078534031413,
+      "grad_norm": 0.2709900192224873,
+      "learning_rate": 3.403824973015725e-05,
+      "loss": 0.376,
+      "step": 1014
+    },
+    {
+      "epoch": 1.6351188078936771,
+      "grad_norm": 0.33820245973489416,
+      "learning_rate": 3.4022200397841056e-05,
+      "loss": 0.3518,
+      "step": 1015
+    },
+    {
+      "epoch": 1.6367297623842125,
+      "grad_norm": 0.2802490380827594,
+      "learning_rate": 3.4006133286519435e-05,
+      "loss": 0.3777,
+      "step": 1016
+    },
+    {
+      "epoch": 1.6383407168747484,
+      "grad_norm": 0.3219486083753848,
+      "learning_rate": 3.399004841656417e-05,
+      "loss": 0.334,
+      "step": 1017
+    },
+    {
+      "epoch": 1.639951671365284,
+      "grad_norm": 0.2804170578284511,
+      "learning_rate": 3.3973945808369566e-05,
+      "loss": 0.3353,
+      "step": 1018
+    },
+    {
+      "epoch": 1.6415626258558196,
+      "grad_norm": 0.3397988556608258,
+      "learning_rate": 3.395782548235242e-05,
+      "loss": 0.3929,
+      "step": 1019
+    },
+    {
+      "epoch": 1.6431735803463552,
+      "grad_norm": 0.25874400518260726,
+      "learning_rate": 3.394168745895199e-05,
+      "loss": 0.3425,
+      "step": 1020
+    },
+    {
+      "epoch": 1.6447845348368908,
+      "grad_norm": 0.29769463217777464,
+      "learning_rate": 3.392553175862996e-05,
+      "loss": 0.3454,
+      "step": 1021
+    },
+    {
+      "epoch": 1.6463954893274266,
+      "grad_norm": 0.27496871301352516,
+      "learning_rate": 3.390935840187045e-05,
+      "loss": 0.3574,
+      "step": 1022
+    },
+    {
+      "epoch": 1.648006443817962,
+      "grad_norm": 0.3130236579001918,
+      "learning_rate": 3.3893167409179945e-05,
+      "loss": 0.353,
+      "step": 1023
+    },
+    {
+      "epoch": 1.6496173983084979,
+      "grad_norm": 0.30244268078948694,
+      "learning_rate": 3.387695880108732e-05,
+      "loss": 0.3554,
+      "step": 1024
+    },
+    {
+      "epoch": 1.6512283527990335,
+      "grad_norm": 0.2637267118602355,
+      "learning_rate": 3.3860732598143754e-05,
+      "loss": 0.3254,
+      "step": 1025
+    },
+    {
+      "epoch": 1.652839307289569,
+      "grad_norm": 0.27265533782673024,
+      "learning_rate": 3.3844488820922755e-05,
+      "loss": 0.3352,
+      "step": 1026
+    },
+    {
+      "epoch": 1.6544502617801047,
+      "grad_norm": 0.2847443661542053,
+      "learning_rate": 3.3828227490020096e-05,
+      "loss": 0.345,
+      "step": 1027
+    },
+    {
+      "epoch": 1.6560612162706403,
+      "grad_norm": 0.2910046942769011,
+      "learning_rate": 3.381194862605383e-05,
+      "loss": 0.3406,
+      "step": 1028
+    },
+    {
+      "epoch": 1.6576721707611761,
+      "grad_norm": 0.30278444886358213,
+      "learning_rate": 3.3795652249664216e-05,
+      "loss": 0.3744,
+      "step": 1029
+    },
+    {
+      "epoch": 1.6592831252517115,
+      "grad_norm": 0.3181445159389258,
+      "learning_rate": 3.377933838151374e-05,
+      "loss": 0.3571,
+      "step": 1030
+    },
+    {
+      "epoch": 1.6608940797422473,
+      "grad_norm": 0.2555750516659823,
+      "learning_rate": 3.376300704228704e-05,
+      "loss": 0.3293,
+      "step": 1031
+    },
+    {
+      "epoch": 1.662505034232783,
+      "grad_norm": 0.2994395275689444,
+      "learning_rate": 3.374665825269093e-05,
+      "loss": 0.3625,
+      "step": 1032
+    },
+    {
+      "epoch": 1.6641159887233186,
+      "grad_norm": 0.2941866705604041,
+      "learning_rate": 3.373029203345435e-05,
+      "loss": 0.3466,
+      "step": 1033
+    },
+    {
+      "epoch": 1.6657269432138542,
+      "grad_norm": 0.2520003523058466,
+      "learning_rate": 3.3713908405328316e-05,
+      "loss": 0.3492,
+      "step": 1034
+    },
+    {
+      "epoch": 1.6673378977043898,
+      "grad_norm": 0.27879316282952243,
+      "learning_rate": 3.369750738908593e-05,
+      "loss": 0.3658,
+      "step": 1035
+    },
+    {
+      "epoch": 1.6689488521949256,
+      "grad_norm": 0.26750661336616033,
+      "learning_rate": 3.368108900552236e-05,
+      "loss": 0.3479,
+      "step": 1036
+    },
+    {
+      "epoch": 1.670559806685461,
+      "grad_norm": 0.2703473771389898,
+      "learning_rate": 3.366465327545475e-05,
+      "loss": 0.3246,
+      "step": 1037
+    },
+    {
+      "epoch": 1.6721707611759968,
+      "grad_norm": 0.2492707097376767,
+      "learning_rate": 3.3648200219722285e-05,
+      "loss": 0.3379,
+      "step": 1038
+    },
+    {
+      "epoch": 1.6737817156665324,
+      "grad_norm": 0.28088000970492083,
+      "learning_rate": 3.3631729859186086e-05,
+      "loss": 0.3772,
+      "step": 1039
+    },
+    {
+      "epoch": 1.675392670157068,
+      "grad_norm": 0.6148118511176305,
+      "learning_rate": 3.3615242214729226e-05,
+      "loss": 0.3457,
+      "step": 1040
+    },
+    {
+      "epoch": 1.6770036246476037,
+      "grad_norm": 0.2741941991125401,
+      "learning_rate": 3.35987373072567e-05,
+      "loss": 0.3679,
+      "step": 1041
+    },
+    {
+      "epoch": 1.6786145791381393,
+      "grad_norm": 0.30929820639855726,
+      "learning_rate": 3.3582215157695376e-05,
+      "loss": 0.3595,
+      "step": 1042
+    },
+    {
+      "epoch": 1.680225533628675,
+      "grad_norm": 0.2855629710854195,
+      "learning_rate": 3.3565675786994e-05,
+      "loss": 0.3485,
+      "step": 1043
+    },
+    {
+      "epoch": 1.6818364881192105,
+      "grad_norm": 0.3160188622883023,
+      "learning_rate": 3.3549119216123125e-05,
+      "loss": 0.3469,
+      "step": 1044
+    },
+    {
+      "epoch": 1.6834474426097463,
+      "grad_norm": 0.3178650447881005,
+      "learning_rate": 3.353254546607515e-05,
+      "loss": 0.3631,
+      "step": 1045
+    },
+    {
+      "epoch": 1.685058397100282,
+      "grad_norm": 0.26522002369737246,
+      "learning_rate": 3.351595455786423e-05,
+      "loss": 0.3164,
+      "step": 1046
+    },
+    {
+      "epoch": 1.6866693515908175,
+      "grad_norm": 0.3103051673772948,
+      "learning_rate": 3.3499346512526286e-05,
+      "loss": 0.3537,
+      "step": 1047
+    },
+    {
+      "epoch": 1.6882803060813532,
+      "grad_norm": 0.26705305622759873,
+      "learning_rate": 3.348272135111895e-05,
+      "loss": 0.3207,
+      "step": 1048
+    },
+    {
+      "epoch": 1.6898912605718888,
+      "grad_norm": 0.2965365817221479,
+      "learning_rate": 3.346607909472159e-05,
+      "loss": 0.3301,
+      "step": 1049
+    },
+    {
+      "epoch": 1.6915022150624246,
+      "grad_norm": 0.3039530865497769,
+      "learning_rate": 3.344941976443521e-05,
+      "loss": 0.3541,
+      "step": 1050
+    },
+    {
+      "epoch": 1.69311316955296,
+      "grad_norm": 0.3019357707645832,
+      "learning_rate": 3.3432743381382494e-05,
+      "loss": 0.3476,
+      "step": 1051
+    },
+    {
+      "epoch": 1.6947241240434958,
+      "grad_norm": 0.30486830582214286,
+      "learning_rate": 3.341604996670773e-05,
+      "loss": 0.3692,
+      "step": 1052
+    },
+    {
+      "epoch": 1.6963350785340314,
+      "grad_norm": 0.2846970874116654,
+      "learning_rate": 3.33993395415768e-05,
+      "loss": 0.3489,
+      "step": 1053
+    },
+    {
+      "epoch": 1.697946033024567,
+      "grad_norm": 0.2844019284784835,
+      "learning_rate": 3.3382612127177166e-05,
+      "loss": 0.3484,
+      "step": 1054
+    },
+    {
+      "epoch": 1.6995569875151026,
+      "grad_norm": 0.31814090002272866,
+      "learning_rate": 3.3365867744717827e-05,
+      "loss": 0.3313,
+      "step": 1055
+    },
+    {
+      "epoch": 1.7011679420056383,
+      "grad_norm": 0.3018174083258602,
+      "learning_rate": 3.334910641542928e-05,
+      "loss": 0.3928,
+      "step": 1056
+    },
+    {
+      "epoch": 1.702778896496174,
+      "grad_norm": 0.2805949905286555,
+      "learning_rate": 3.3332328160563534e-05,
+      "loss": 0.3138,
+      "step": 1057
+    },
+    {
+      "epoch": 1.7043898509867095,
+      "grad_norm": 0.294154845413384,
+      "learning_rate": 3.331553300139404e-05,
+      "loss": 0.3715,
+      "step": 1058
+    },
+    {
+      "epoch": 1.7060008054772453,
+      "grad_norm": 0.28060148879956204,
+      "learning_rate": 3.3298720959215686e-05,
+      "loss": 0.3404,
+      "step": 1059
+    },
+    {
+      "epoch": 1.707611759967781,
+      "grad_norm": 0.2898378970822988,
+      "learning_rate": 3.328189205534479e-05,
+      "loss": 0.3707,
+      "step": 1060
+    },
+    {
+      "epoch": 1.7092227144583165,
+      "grad_norm": 0.29125221651243705,
+      "learning_rate": 3.3265046311118996e-05,
+      "loss": 0.3513,
+      "step": 1061
+    },
+    {
+      "epoch": 1.7108336689488524,
+      "grad_norm": 0.2944155426474975,
+      "learning_rate": 3.3248183747897354e-05,
+      "loss": 0.3294,
+      "step": 1062
+    },
+    {
+      "epoch": 1.7124446234393877,
+      "grad_norm": 0.2934301800909067,
+      "learning_rate": 3.3231304387060215e-05,
+      "loss": 0.3756,
+      "step": 1063
+    },
+    {
+      "epoch": 1.7140555779299236,
+      "grad_norm": 0.3258895549591974,
+      "learning_rate": 3.321440825000923e-05,
+      "loss": 0.3307,
+      "step": 1064
+    },
+    {
+      "epoch": 1.715666532420459,
+      "grad_norm": 0.25585412371136995,
+      "learning_rate": 3.3197495358167314e-05,
+      "loss": 0.3319,
+      "step": 1065
+    },
+    {
+      "epoch": 1.7172774869109948,
+      "grad_norm": 0.33852886154261486,
+      "learning_rate": 3.318056573297864e-05,
+      "loss": 0.345,
+      "step": 1066
+    },
+    {
+      "epoch": 1.7188884414015304,
+      "grad_norm": 0.29308836939109273,
+      "learning_rate": 3.3163619395908594e-05,
+      "loss": 0.3521,
+      "step": 1067
+    },
+    {
+      "epoch": 1.720499395892066,
+      "grad_norm": 0.3260413931230068,
+      "learning_rate": 3.314665636844374e-05,
+      "loss": 0.3579,
+      "step": 1068
+    },
+    {
+      "epoch": 1.7221103503826019,
+      "grad_norm": 0.315853677609962,
+      "learning_rate": 3.3129676672091814e-05,
+      "loss": 0.3693,
+      "step": 1069
+    },
+    {
+      "epoch": 1.7237213048731372,
+      "grad_norm": 0.25252028334754817,
+      "learning_rate": 3.311268032838169e-05,
+      "loss": 0.311,
+      "step": 1070
+    },
+    {
+      "epoch": 1.725332259363673,
+      "grad_norm": 0.30596679620107337,
+      "learning_rate": 3.309566735886334e-05,
+      "loss": 0.361,
+      "step": 1071
+    },
+    {
+      "epoch": 1.7269432138542085,
+      "grad_norm": 0.2605593059669489,
+      "learning_rate": 3.307863778510782e-05,
+      "loss": 0.3559,
+      "step": 1072
+    },
+    {
+      "epoch": 1.7285541683447443,
+      "grad_norm": 0.30530466414998403,
+      "learning_rate": 3.306159162870724e-05,
+      "loss": 0.3668,
+      "step": 1073
+    },
+    {
+      "epoch": 1.73016512283528,
+      "grad_norm": 0.28036764595429337,
+      "learning_rate": 3.304452891127474e-05,
+      "loss": 0.3556,
+      "step": 1074
+    },
+    {
+      "epoch": 1.7317760773258155,
+      "grad_norm": 0.3020190132811284,
+      "learning_rate": 3.302744965444445e-05,
+      "loss": 0.3526,
+      "step": 1075
+    },
+    {
+      "epoch": 1.7333870318163513,
+      "grad_norm": 0.25272390853900173,
+      "learning_rate": 3.301035387987146e-05,
+      "loss": 0.3276,
+      "step": 1076
+    },
+    {
+      "epoch": 1.7349979863068867,
+      "grad_norm": 0.342414793506701,
+      "learning_rate": 3.299324160923184e-05,
+      "loss": 0.3355,
+      "step": 1077
+    },
+    {
+      "epoch": 1.7366089407974226,
+      "grad_norm": 0.3002251269987107,
+      "learning_rate": 3.297611286422254e-05,
+      "loss": 0.3466,
+      "step": 1078
+    },
+    {
+      "epoch": 1.738219895287958,
+      "grad_norm": 0.32281240413837947,
+      "learning_rate": 3.295896766656141e-05,
+      "loss": 0.3627,
+      "step": 1079
+    },
+    {
+      "epoch": 1.7398308497784938,
+      "grad_norm": 0.27577305817853215,
+      "learning_rate": 3.294180603798716e-05,
+      "loss": 0.3264,
+      "step": 1080
+    },
+    {
+      "epoch": 1.7414418042690294,
+      "grad_norm": 0.3366688008037847,
+      "learning_rate": 3.292462800025933e-05,
+      "loss": 0.3553,
+      "step": 1081
+    },
+    {
+      "epoch": 1.743052758759565,
+      "grad_norm": 0.2864923341321958,
+      "learning_rate": 3.290743357515829e-05,
+      "loss": 0.3715,
+      "step": 1082
+    },
+    {
+      "epoch": 1.7446637132501008,
+      "grad_norm": 0.2885887017629758,
+      "learning_rate": 3.289022278448513e-05,
+      "loss": 0.3452,
+      "step": 1083
+    },
+    {
+      "epoch": 1.7462746677406362,
+      "grad_norm": 0.28829785150227605,
+      "learning_rate": 3.287299565006177e-05,
+      "loss": 0.3275,
+      "step": 1084
+    },
+    {
+      "epoch": 1.747885622231172,
+      "grad_norm": 0.3095311480709755,
+      "learning_rate": 3.285575219373079e-05,
+      "loss": 0.3884,
+      "step": 1085
+    },
+    {
+      "epoch": 1.7494965767217077,
+      "grad_norm": 0.3841547619344348,
+      "learning_rate": 3.2838492437355487e-05,
+      "loss": 0.3251,
+      "step": 1086
+    },
+    {
+      "epoch": 1.7511075312122433,
+      "grad_norm": 0.2583438353752805,
+      "learning_rate": 3.2821216402819814e-05,
+      "loss": 0.3298,
+      "step": 1087
+    },
+    {
+      "epoch": 1.7527184857027789,
+      "grad_norm": 0.3520798827537718,
+      "learning_rate": 3.280392411202838e-05,
+      "loss": 0.3611,
+      "step": 1088
+    },
+    {
+      "epoch": 1.7543294401933145,
+      "grad_norm": 0.24181815291715242,
+      "learning_rate": 3.27866155869064e-05,
+      "loss": 0.3353,
+      "step": 1089
+    },
+    {
+      "epoch": 1.7559403946838503,
+      "grad_norm": 0.26861982821219954,
+      "learning_rate": 3.276929084939967e-05,
+      "loss": 0.3268,
+      "step": 1090
+    },
+    {
+      "epoch": 1.7575513491743857,
+      "grad_norm": 0.30793537607947774,
+      "learning_rate": 3.275194992147455e-05,
+      "loss": 0.3758,
+      "step": 1091
+    },
+    {
+      "epoch": 1.7591623036649215,
+      "grad_norm": 0.2735491056478904,
+      "learning_rate": 3.27345928251179e-05,
+      "loss": 0.3522,
+      "step": 1092
+    },
+    {
+      "epoch": 1.7607732581554572,
+      "grad_norm": 0.31416721865401886,
+      "learning_rate": 3.271721958233713e-05,
+      "loss": 0.3887,
+      "step": 1093
+    },
+    {
+      "epoch": 1.7623842126459928,
+      "grad_norm": 0.23777442546977362,
+      "learning_rate": 3.269983021516006e-05,
+      "loss": 0.341,
+      "step": 1094
+    },
+    {
+      "epoch": 1.7639951671365284,
+      "grad_norm": 0.2966716961575548,
+      "learning_rate": 3.268242474563502e-05,
+      "loss": 0.3582,
+      "step": 1095
+    },
+    {
+      "epoch": 1.765606121627064,
+      "grad_norm": 0.2695248187187801,
+      "learning_rate": 3.2665003195830705e-05,
+      "loss": 0.3663,
+      "step": 1096
+    },
+    {
+      "epoch": 1.7672170761175998,
+      "grad_norm": 0.27648681082467197,
+      "learning_rate": 3.2647565587836224e-05,
+      "loss": 0.3367,
+      "step": 1097
+    },
+    {
+      "epoch": 1.7688280306081352,
+      "grad_norm": 0.24769125048939025,
+      "learning_rate": 3.2630111943761035e-05,
+      "loss": 0.3161,
+      "step": 1098
+    },
+    {
+      "epoch": 1.770438985098671,
+      "grad_norm": 0.32602085521437335,
+      "learning_rate": 3.261264228573495e-05,
+      "loss": 0.377,
+      "step": 1099
+    },
+    {
+      "epoch": 1.7720499395892066,
+      "grad_norm": 0.26152980147545596,
+      "learning_rate": 3.259515663590805e-05,
+      "loss": 0.3594,
+      "step": 1100
+    },
+    {
+      "epoch": 1.7736608940797423,
+      "grad_norm": 0.2382508327122932,
+      "learning_rate": 3.257765501645072e-05,
+      "loss": 0.3169,
+      "step": 1101
+    },
+    {
+      "epoch": 1.7752718485702779,
+      "grad_norm": 0.3243836829426945,
+      "learning_rate": 3.256013744955359e-05,
+      "loss": 0.3719,
+      "step": 1102
+    },
+    {
+      "epoch": 1.7768828030608135,
+      "grad_norm": 0.25232903385465316,
+      "learning_rate": 3.25426039574275e-05,
+      "loss": 0.3608,
+      "step": 1103
+    },
+    {
+      "epoch": 1.7784937575513493,
+      "grad_norm": 0.2871596146859017,
+      "learning_rate": 3.2525054562303485e-05,
+      "loss": 0.36,
+      "step": 1104
+    },
+    {
+      "epoch": 1.7801047120418847,
+      "grad_norm": 0.27660826788802817,
+      "learning_rate": 3.250748928643274e-05,
+      "loss": 0.3321,
+      "step": 1105
+    },
+    {
+      "epoch": 1.7817156665324205,
+      "grad_norm": 0.26281335753440443,
+      "learning_rate": 3.248990815208661e-05,
+      "loss": 0.3365,
+      "step": 1106
+    },
+    {
+      "epoch": 1.7833266210229561,
+      "grad_norm": 0.3107781337125716,
+      "learning_rate": 3.247231118155654e-05,
+      "loss": 0.3673,
+      "step": 1107
+    },
+    {
+      "epoch": 1.7849375755134917,
+      "grad_norm": 0.2783443100448748,
+      "learning_rate": 3.245469839715404e-05,
+      "loss": 0.3394,
+      "step": 1108
+    },
+    {
+      "epoch": 1.7865485300040274,
+      "grad_norm": 0.3194930726854231,
+      "learning_rate": 3.24370698212107e-05,
+      "loss": 0.3464,
+      "step": 1109
+    },
+    {
+      "epoch": 1.788159484494563,
+      "grad_norm": 0.29259588976237577,
+      "learning_rate": 3.24194254760781e-05,
+      "loss": 0.351,
+      "step": 1110
+    },
+    {
+      "epoch": 1.7897704389850988,
+      "grad_norm": 0.2833032383542838,
+      "learning_rate": 3.240176538412783e-05,
+      "loss": 0.3112,
+      "step": 1111
+    },
+    {
+      "epoch": 1.7913813934756342,
+      "grad_norm": 0.3062775036699522,
+      "learning_rate": 3.2384089567751464e-05,
+      "loss": 0.3538,
+      "step": 1112
+    },
+    {
+      "epoch": 1.79299234796617,
+      "grad_norm": 0.31520368622639755,
+      "learning_rate": 3.236639804936047e-05,
+      "loss": 0.365,
+      "step": 1113
+    },
+    {
+      "epoch": 1.7946033024567056,
+      "grad_norm": 0.2706287135324734,
+      "learning_rate": 3.234869085138626e-05,
+      "loss": 0.3057,
+      "step": 1114
+    },
+    {
+      "epoch": 1.7962142569472412,
+      "grad_norm": 0.2776347182974663,
+      "learning_rate": 3.233096799628012e-05,
+      "loss": 0.3591,
+      "step": 1115
+    },
+    {
+      "epoch": 1.7978252114377768,
+      "grad_norm": 0.23921615357001141,
+      "learning_rate": 3.2313229506513167e-05,
+      "loss": 0.3287,
+      "step": 1116
+    },
+    {
+      "epoch": 1.7994361659283125,
+      "grad_norm": 0.27114468404862,
+      "learning_rate": 3.229547540457638e-05,
+      "loss": 0.361,
+      "step": 1117
+    },
+    {
+      "epoch": 1.8010471204188483,
+      "grad_norm": 0.24463881898739914,
+      "learning_rate": 3.2277705712980495e-05,
+      "loss": 0.3404,
+      "step": 1118
+    },
+    {
+      "epoch": 1.8026580749093837,
+      "grad_norm": 0.3166270096090458,
+      "learning_rate": 3.225992045425604e-05,
+      "loss": 0.36,
+      "step": 1119
+    },
+    {
+      "epoch": 1.8042690293999195,
+      "grad_norm": 0.2723260812159827,
+      "learning_rate": 3.224211965095326e-05,
+      "loss": 0.3402,
+      "step": 1120
+    },
+    {
+      "epoch": 1.8058799838904551,
+      "grad_norm": 0.28717184216313857,
+      "learning_rate": 3.222430332564213e-05,
+      "loss": 0.3424,
+      "step": 1121
+    },
+    {
+      "epoch": 1.8074909383809907,
+      "grad_norm": 0.30590726225785186,
+      "learning_rate": 3.220647150091229e-05,
+      "loss": 0.384,
+      "step": 1122
+    },
+    {
+      "epoch": 1.8091018928715263,
+      "grad_norm": 0.3295960133235172,
+      "learning_rate": 3.2188624199373054e-05,
+      "loss": 0.3477,
+      "step": 1123
+    },
+    {
+      "epoch": 1.810712847362062,
+      "grad_norm": 0.24641937154261023,
+      "learning_rate": 3.217076144365332e-05,
+      "loss": 0.3291,
+      "step": 1124
+    },
+    {
+      "epoch": 1.8123238018525978,
+      "grad_norm": 0.36214805461319893,
+      "learning_rate": 3.215288325640161e-05,
+      "loss": 0.37,
+      "step": 1125
+    },
+    {
+      "epoch": 1.8139347563431332,
+      "grad_norm": 0.27639672336136023,
+      "learning_rate": 3.213498966028603e-05,
+      "loss": 0.3474,
+      "step": 1126
+    },
+    {
+      "epoch": 1.815545710833669,
+      "grad_norm": 0.2905471387956144,
+      "learning_rate": 3.2117080677994156e-05,
+      "loss": 0.3489,
+      "step": 1127
+    },
+    {
+      "epoch": 1.8171566653242046,
+      "grad_norm": 0.2742465600770904,
+      "learning_rate": 3.2099156332233155e-05,
+      "loss": 0.3212,
+      "step": 1128
+    },
+    {
+      "epoch": 1.8187676198147402,
+      "grad_norm": 0.32581522778160565,
+      "learning_rate": 3.2081216645729615e-05,
+      "loss": 0.3789,
+      "step": 1129
+    },
+    {
+      "epoch": 1.8203785743052758,
+      "grad_norm": 0.25895277713528797,
+      "learning_rate": 3.20632616412296e-05,
+      "loss": 0.3548,
+      "step": 1130
+    },
+    {
+      "epoch": 1.8219895287958114,
+      "grad_norm": 0.2567453896587822,
+      "learning_rate": 3.204529134149858e-05,
+      "loss": 0.337,
+      "step": 1131
+    },
+    {
+      "epoch": 1.8236004832863473,
+      "grad_norm": 0.2545418099367893,
+      "learning_rate": 3.2027305769321446e-05,
+      "loss": 0.3757,
+      "step": 1132
+    },
+    {
+      "epoch": 1.8252114377768827,
+      "grad_norm": 0.26861543778444336,
+      "learning_rate": 3.2009304947502415e-05,
+      "loss": 0.3309,
+      "step": 1133
+    },
+    {
+      "epoch": 1.8268223922674185,
+      "grad_norm": 0.25906425044357323,
+      "learning_rate": 3.1991288898865076e-05,
+      "loss": 0.3425,
+      "step": 1134
+    },
+    {
+      "epoch": 1.828433346757954,
+      "grad_norm": 0.2791495649866975,
+      "learning_rate": 3.19732576462523e-05,
+      "loss": 0.3626,
+      "step": 1135
+    },
+    {
+      "epoch": 1.8300443012484897,
+      "grad_norm": 0.28566136887901167,
+      "learning_rate": 3.195521121252625e-05,
+      "loss": 0.3427,
+      "step": 1136
+    },
+    {
+      "epoch": 1.8316552557390253,
+      "grad_norm": 0.3006511911637088,
+      "learning_rate": 3.193714962056832e-05,
+      "loss": 0.3689,
+      "step": 1137
+    },
+    {
+      "epoch": 1.833266210229561,
+      "grad_norm": 0.2575714390405587,
+      "learning_rate": 3.1919072893279144e-05,
+      "loss": 0.3254,
+      "step": 1138
+    },
+    {
+      "epoch": 1.8348771647200968,
+      "grad_norm": 0.29714340630201763,
+      "learning_rate": 3.190098105357853e-05,
+      "loss": 0.3114,
+      "step": 1139
+    },
+    {
+      "epoch": 1.8364881192106322,
+      "grad_norm": 0.2739715615694121,
+      "learning_rate": 3.188287412440546e-05,
+      "loss": 0.3628,
+      "step": 1140
+    },
+    {
+      "epoch": 1.838099073701168,
+      "grad_norm": 0.31490538424728864,
+      "learning_rate": 3.186475212871803e-05,
+      "loss": 0.3681,
+      "step": 1141
+    },
+    {
+      "epoch": 1.8397100281917036,
+      "grad_norm": 0.25725247708626503,
+      "learning_rate": 3.1846615089493465e-05,
+      "loss": 0.3373,
+      "step": 1142
+    },
+    {
+      "epoch": 1.8413209826822392,
+      "grad_norm": 0.29237114605953757,
+      "learning_rate": 3.182846302972804e-05,
+      "loss": 0.3928,
+      "step": 1143
+    },
+    {
+      "epoch": 1.8429319371727748,
+      "grad_norm": 0.24162054032640948,
+      "learning_rate": 3.181029597243709e-05,
+      "loss": 0.3066,
+      "step": 1144
+    },
+    {
+      "epoch": 1.8445428916633104,
+      "grad_norm": 0.2746441036778749,
+      "learning_rate": 3.1792113940654976e-05,
+      "loss": 0.3429,
+      "step": 1145
+    },
+    {
+      "epoch": 1.8461538461538463,
+      "grad_norm": 0.27290548211325333,
+      "learning_rate": 3.1773916957435e-05,
+      "loss": 0.3567,
+      "step": 1146
+    },
+    {
+      "epoch": 1.8477648006443816,
+      "grad_norm": 0.3015414783683767,
+      "learning_rate": 3.1755705045849465e-05,
+      "loss": 0.3386,
+      "step": 1147
+    },
+    {
+      "epoch": 1.8493757551349175,
+      "grad_norm": 0.2540351001266061,
+      "learning_rate": 3.173747822898959e-05,
+      "loss": 0.3452,
+      "step": 1148
+    },
+    {
+      "epoch": 1.850986709625453,
+      "grad_norm": 0.31221816074924413,
+      "learning_rate": 3.1719236529965494e-05,
+      "loss": 0.3592,
+      "step": 1149
+    },
+    {
+      "epoch": 1.8525976641159887,
+      "grad_norm": 0.2423803158509095,
+      "learning_rate": 3.170097997190615e-05,
+      "loss": 0.3195,
+      "step": 1150
+    },
+    {
+      "epoch": 1.8542086186065245,
+      "grad_norm": 0.2966064384297838,
+      "learning_rate": 3.16827085779594e-05,
+      "loss": 0.3355,
+      "step": 1151
+    },
+    {
+      "epoch": 1.85581957309706,
+      "grad_norm": 0.2810752575128961,
+      "learning_rate": 3.1664422371291866e-05,
+      "loss": 0.3336,
+      "step": 1152
+    },
+    {
+      "epoch": 1.8574305275875957,
+      "grad_norm": 0.2784162936236359,
+      "learning_rate": 3.164612137508898e-05,
+      "loss": 0.3542,
+      "step": 1153
+    },
+    {
+      "epoch": 1.8590414820781311,
+      "grad_norm": 0.32474379464442327,
+      "learning_rate": 3.162780561255489e-05,
+      "loss": 0.368,
+      "step": 1154
+    },
+    {
+      "epoch": 1.860652436568667,
+      "grad_norm": 0.25214198273713495,
+      "learning_rate": 3.16094751069125e-05,
+      "loss": 0.3208,
+      "step": 1155
+    },
+    {
+      "epoch": 1.8622633910592026,
+      "grad_norm": 0.2804333800265253,
+      "learning_rate": 3.15911298814034e-05,
+      "loss": 0.3633,
+      "step": 1156
+    },
+    {
+      "epoch": 1.8638743455497382,
+      "grad_norm": 0.3177102741885549,
+      "learning_rate": 3.157276995928783e-05,
+      "loss": 0.3548,
+      "step": 1157
+    },
+    {
+      "epoch": 1.865485300040274,
+      "grad_norm": 0.2779089270761033,
+      "learning_rate": 3.155439536384467e-05,
+      "loss": 0.3397,
+      "step": 1158
+    },
+    {
+      "epoch": 1.8670962545308094,
+      "grad_norm": 0.30157569369109105,
+      "learning_rate": 3.153600611837142e-05,
+      "loss": 0.3538,
+      "step": 1159
+    },
+    {
+      "epoch": 1.8687072090213452,
+      "grad_norm": 0.3222782938570124,
+      "learning_rate": 3.151760224618413e-05,
+      "loss": 0.3584,
+      "step": 1160
+    },
+    {
+      "epoch": 1.8703181635118806,
+      "grad_norm": 0.26202863886970124,
+      "learning_rate": 3.1499183770617414e-05,
+      "loss": 0.3294,
+      "step": 1161
+    },
+    {
+      "epoch": 1.8719291180024165,
+      "grad_norm": 0.3313404569532206,
+      "learning_rate": 3.1480750715024396e-05,
+      "loss": 0.3607,
+      "step": 1162
+    },
+    {
+      "epoch": 1.873540072492952,
+      "grad_norm": 0.2804570203558461,
+      "learning_rate": 3.146230310277668e-05,
+      "loss": 0.3684,
+      "step": 1163
+    },
+    {
+      "epoch": 1.8751510269834877,
+      "grad_norm": 0.328059825473741,
+      "learning_rate": 3.144384095726433e-05,
+      "loss": 0.3345,
+      "step": 1164
+    },
+    {
+      "epoch": 1.8767619814740235,
+      "grad_norm": 0.3310382940626733,
+      "learning_rate": 3.142536430189585e-05,
+      "loss": 0.3592,
+      "step": 1165
+    },
+    {
+      "epoch": 1.878372935964559,
+      "grad_norm": 0.2571589985927003,
+      "learning_rate": 3.140687316009812e-05,
+      "loss": 0.334,
+      "step": 1166
+    },
+    {
+      "epoch": 1.8799838904550947,
+      "grad_norm": 0.3084210247719844,
+      "learning_rate": 3.138836755531641e-05,
+      "loss": 0.3549,
+      "step": 1167
+    },
+    {
+      "epoch": 1.8815948449456303,
+      "grad_norm": 0.3136180657588817,
+      "learning_rate": 3.13698475110143e-05,
+      "loss": 0.3385,
+      "step": 1168
+    },
+    {
+      "epoch": 1.883205799436166,
+      "grad_norm": 0.28698454787368805,
+      "learning_rate": 3.135131305067372e-05,
+      "loss": 0.3459,
+      "step": 1169
+    },
+    {
+      "epoch": 1.8848167539267016,
+      "grad_norm": 0.3203660869415405,
+      "learning_rate": 3.1332764197794825e-05,
+      "loss": 0.3665,
+      "step": 1170
+    },
+    {
+      "epoch": 1.8864277084172372,
+      "grad_norm": 0.2760307046037802,
+      "learning_rate": 3.131420097589606e-05,
+      "loss": 0.3535,
+      "step": 1171
+    },
+    {
+      "epoch": 1.888038662907773,
+      "grad_norm": 0.304173676179774,
+      "learning_rate": 3.129562340851408e-05,
+      "loss": 0.3598,
+      "step": 1172
+    },
+    {
+      "epoch": 1.8896496173983084,
+      "grad_norm": 0.24789364158706606,
+      "learning_rate": 3.127703151920371e-05,
+      "loss": 0.3423,
+      "step": 1173
+    },
+    {
+      "epoch": 1.8912605718888442,
+      "grad_norm": 0.27869905089039404,
+      "learning_rate": 3.125842533153796e-05,
+      "loss": 0.3801,
+      "step": 1174
+    },
+    {
+      "epoch": 1.8928715263793798,
+      "grad_norm": 0.2659917053384485,
+      "learning_rate": 3.1239804869107943e-05,
+      "loss": 0.3162,
+      "step": 1175
+    },
+    {
+      "epoch": 1.8944824808699154,
+      "grad_norm": 0.2740676239526807,
+      "learning_rate": 3.1221170155522896e-05,
+      "loss": 0.3763,
+      "step": 1176
+    },
+    {
+      "epoch": 1.896093435360451,
+      "grad_norm": 0.3036149904699523,
+      "learning_rate": 3.1202521214410116e-05,
+      "loss": 0.3324,
+      "step": 1177
+    },
+    {
+      "epoch": 1.8977043898509867,
+      "grad_norm": 0.2704344928402855,
+      "learning_rate": 3.1183858069414936e-05,
+      "loss": 0.3508,
+      "step": 1178
+    },
+    {
+      "epoch": 1.8993153443415225,
+      "grad_norm": 0.2745284876347354,
+      "learning_rate": 3.1165180744200704e-05,
+      "loss": 0.3194,
+      "step": 1179
+    },
+    {
+      "epoch": 1.9009262988320579,
+      "grad_norm": 0.2584540162273995,
+      "learning_rate": 3.114648926244873e-05,
+      "loss": 0.3385,
+      "step": 1180
+    },
+    {
+      "epoch": 1.9025372533225937,
+      "grad_norm": 0.3116807537951856,
+      "learning_rate": 3.11277836478583e-05,
+      "loss": 0.3586,
+      "step": 1181
+    },
+    {
+      "epoch": 1.9041482078131293,
+      "grad_norm": 0.2623123292830474,
+      "learning_rate": 3.11090639241466e-05,
+      "loss": 0.3502,
+      "step": 1182
+    },
+    {
+      "epoch": 1.905759162303665,
+      "grad_norm": 0.2806030867053086,
+      "learning_rate": 3.1090330115048716e-05,
+      "loss": 0.3602,
+      "step": 1183
+    },
+    {
+      "epoch": 1.9073701167942005,
+      "grad_norm": 0.30229750912862674,
+      "learning_rate": 3.107158224431759e-05,
+      "loss": 0.3303,
+      "step": 1184
+    },
+    {
+      "epoch": 1.9089810712847362,
+      "grad_norm": 0.3148634173453989,
+      "learning_rate": 3.105282033572398e-05,
+      "loss": 0.3854,
+      "step": 1185
+    },
+    {
+      "epoch": 1.910592025775272,
+      "grad_norm": 0.28827408407660937,
+      "learning_rate": 3.1034044413056465e-05,
+      "loss": 0.3304,
+      "step": 1186
+    },
+    {
+      "epoch": 1.9122029802658074,
+      "grad_norm": 0.3063959496143443,
+      "learning_rate": 3.1015254500121376e-05,
+      "loss": 0.3747,
+      "step": 1187
+    },
+    {
+      "epoch": 1.9138139347563432,
+      "grad_norm": 0.2714542673376483,
+      "learning_rate": 3.09964506207428e-05,
+      "loss": 0.3437,
+      "step": 1188
+    },
+    {
+      "epoch": 1.9154248892468788,
+      "grad_norm": 0.2550631224442582,
+      "learning_rate": 3.097763279876251e-05,
+      "loss": 0.3379,
+      "step": 1189
+    },
+    {
+      "epoch": 1.9170358437374144,
+      "grad_norm": 0.27675565441191935,
+      "learning_rate": 3.095880105803997e-05,
+      "loss": 0.3567,
+      "step": 1190
+    },
+    {
+      "epoch": 1.91864679822795,
+      "grad_norm": 0.25174769021841753,
+      "learning_rate": 3.09399554224523e-05,
+      "loss": 0.3066,
+      "step": 1191
+    },
+    {
+      "epoch": 1.9202577527184856,
+      "grad_norm": 0.3167995968880189,
+      "learning_rate": 3.092109591589421e-05,
+      "loss": 0.3712,
+      "step": 1192
+    },
+    {
+      "epoch": 1.9218687072090215,
+      "grad_norm": 0.25594727349036206,
+      "learning_rate": 3.0902222562278025e-05,
+      "loss": 0.3463,
+      "step": 1193
+    },
+    {
+      "epoch": 1.9234796616995569,
+      "grad_norm": 0.2623623493033658,
+      "learning_rate": 3.088333538553361e-05,
+      "loss": 0.3485,
+      "step": 1194
+    },
+    {
+      "epoch": 1.9250906161900927,
+      "grad_norm": 0.24387996552506205,
+      "learning_rate": 3.086443440960838e-05,
+      "loss": 0.3247,
+      "step": 1195
+    },
+    {
+      "epoch": 1.9267015706806283,
+      "grad_norm": 0.27837207857209045,
+      "learning_rate": 3.084551965846721e-05,
+      "loss": 0.3378,
+      "step": 1196
+    },
+    {
+      "epoch": 1.928312525171164,
+      "grad_norm": 0.30825795217226376,
+      "learning_rate": 3.0826591156092474e-05,
+      "loss": 0.4079,
+      "step": 1197
+    },
+    {
+      "epoch": 1.9299234796616995,
+      "grad_norm": 0.2893650060453642,
+      "learning_rate": 3.080764892648396e-05,
+      "loss": 0.3492,
+      "step": 1198
+    },
+    {
+      "epoch": 1.9315344341522351,
+      "grad_norm": 0.2649652705405745,
+      "learning_rate": 3.0788692993658874e-05,
+      "loss": 0.3357,
+      "step": 1199
+    },
+    {
+      "epoch": 1.933145388642771,
+      "grad_norm": 0.29712091496353926,
+      "learning_rate": 3.076972338165178e-05,
+      "loss": 0.3542,
+      "step": 1200
+    },
+    {
+      "epoch": 1.9347563431333064,
+      "grad_norm": 0.27057553138069074,
+      "learning_rate": 3.075074011451461e-05,
+      "loss": 0.3755,
+      "step": 1201
+    },
+    {
+      "epoch": 1.9363672976238422,
+      "grad_norm": 0.2491519925272819,
+      "learning_rate": 3.07317432163166e-05,
+      "loss": 0.3191,
+      "step": 1202
+    },
+    {
+      "epoch": 1.9379782521143778,
+      "grad_norm": 0.26410068635338846,
+      "learning_rate": 3.0712732711144254e-05,
+      "loss": 0.3462,
+      "step": 1203
+    },
+    {
+      "epoch": 1.9395892066049134,
+      "grad_norm": 0.3053998571246562,
+      "learning_rate": 3.0693708623101345e-05,
+      "loss": 0.3587,
+      "step": 1204
+    },
+    {
+      "epoch": 1.941200161095449,
+      "grad_norm": 0.2766139786219459,
+      "learning_rate": 3.067467097630886e-05,
+      "loss": 0.3506,
+      "step": 1205
+    },
+    {
+      "epoch": 1.9428111155859846,
+      "grad_norm": 0.30093579529593656,
+      "learning_rate": 3.0655619794905e-05,
+      "loss": 0.3676,
+      "step": 1206
+    },
+    {
+      "epoch": 1.9444220700765205,
+      "grad_norm": 0.26128628640625445,
+      "learning_rate": 3.063655510304508e-05,
+      "loss": 0.3358,
+      "step": 1207
+    },
+    {
+      "epoch": 1.9460330245670558,
+      "grad_norm": 0.29154302004266824,
+      "learning_rate": 3.061747692490159e-05,
+      "loss": 0.3621,
+      "step": 1208
+    },
+    {
+      "epoch": 1.9476439790575917,
+      "grad_norm": 0.2518019757597743,
+      "learning_rate": 3.05983852846641e-05,
+      "loss": 0.3323,
+      "step": 1209
+    },
+    {
+      "epoch": 1.9492549335481273,
+      "grad_norm": 0.2691404649318548,
+      "learning_rate": 3.057928020653925e-05,
+      "loss": 0.3441,
+      "step": 1210
+    },
+    {
+      "epoch": 1.950865888038663,
+      "grad_norm": 0.2548234265739506,
+      "learning_rate": 3.056016171475072e-05,
+      "loss": 0.3453,
+      "step": 1211
+    },
+    {
+      "epoch": 1.9524768425291985,
+      "grad_norm": 0.25140408398192154,
+      "learning_rate": 3.0541029833539195e-05,
+      "loss": 0.3427,
+      "step": 1212
+    },
+    {
+      "epoch": 1.9540877970197341,
+      "grad_norm": 0.27769435781256435,
+      "learning_rate": 3.0521884587162344e-05,
+      "loss": 0.3448,
+      "step": 1213
+    },
+    {
+      "epoch": 1.95569875151027,
+      "grad_norm": 0.2572998449613547,
+      "learning_rate": 3.050272599989477e-05,
+      "loss": 0.3495,
+      "step": 1214
+    },
+    {
+      "epoch": 1.9573097060008053,
+      "grad_norm": 0.2557887816223843,
+      "learning_rate": 3.0483554096027998e-05,
+      "loss": 0.328,
+      "step": 1215
+    },
+    {
+      "epoch": 1.9589206604913412,
+      "grad_norm": 0.2614863654535399,
+      "learning_rate": 3.046436889987044e-05,
+      "loss": 0.3426,
+      "step": 1216
+    },
+    {
+      "epoch": 1.9605316149818768,
+      "grad_norm": 0.2375877870665729,
+      "learning_rate": 3.0445170435747364e-05,
+      "loss": 0.3031,
+      "step": 1217
+    },
+    {
+      "epoch": 1.9621425694724124,
+      "grad_norm": 0.2894524906451252,
+      "learning_rate": 3.0425958728000845e-05,
+      "loss": 0.3805,
+      "step": 1218
+    },
+    {
+      "epoch": 1.963753523962948,
+      "grad_norm": 0.2593732546970939,
+      "learning_rate": 3.0406733800989766e-05,
+      "loss": 0.3456,
+      "step": 1219
+    },
+    {
+      "epoch": 1.9653644784534836,
+      "grad_norm": 0.2533421503146536,
+      "learning_rate": 3.0387495679089753e-05,
+      "loss": 0.3617,
+      "step": 1220
+    },
+    {
+      "epoch": 1.9669754329440194,
+      "grad_norm": 0.2737014852345156,
+      "learning_rate": 3.0368244386693196e-05,
+      "loss": 0.3451,
+      "step": 1221
+    },
+    {
+      "epoch": 1.9685863874345548,
+      "grad_norm": 0.2385100617519932,
+      "learning_rate": 3.034897994820915e-05,
+      "loss": 0.3462,
+      "step": 1222
+    },
+    {
+      "epoch": 1.9701973419250907,
+      "grad_norm": 0.2679673464061188,
+      "learning_rate": 3.0329702388063348e-05,
+      "loss": 0.3474,
+      "step": 1223
+    },
+    {
+      "epoch": 1.9718082964156263,
+      "grad_norm": 0.25455746591228023,
+      "learning_rate": 3.0310411730698166e-05,
+      "loss": 0.3403,
+      "step": 1224
+    },
+    {
+      "epoch": 1.9734192509061619,
+      "grad_norm": 0.2559764269490091,
+      "learning_rate": 3.029110800057258e-05,
+      "loss": 0.3268,
+      "step": 1225
+    },
+    {
+      "epoch": 1.9750302053966975,
+      "grad_norm": 0.26102725959877315,
+      "learning_rate": 3.0271791222162145e-05,
+      "loss": 0.3513,
+      "step": 1226
+    },
+    {
+      "epoch": 1.976641159887233,
+      "grad_norm": 0.2837350041203044,
+      "learning_rate": 3.0252461419958968e-05,
+      "loss": 0.355,
+      "step": 1227
+    },
+    {
+      "epoch": 1.978252114377769,
+      "grad_norm": 0.276477357654969,
+      "learning_rate": 3.023311861847165e-05,
+      "loss": 0.3156,
+      "step": 1228
+    },
+    {
+      "epoch": 1.9798630688683043,
+      "grad_norm": 0.2630117478712932,
+      "learning_rate": 3.0213762842225284e-05,
+      "loss": 0.3311,
+      "step": 1229
+    },
+    {
+      "epoch": 1.9814740233588402,
+      "grad_norm": 0.3217425165655156,
+      "learning_rate": 3.0194394115761415e-05,
+      "loss": 0.3413,
+      "step": 1230
+    },
+    {
+      "epoch": 1.9830849778493758,
+      "grad_norm": 0.3176180655742061,
+      "learning_rate": 3.0175012463638016e-05,
+      "loss": 0.3541,
+      "step": 1231
+    },
+    {
+      "epoch": 1.9846959323399114,
+      "grad_norm": 0.32893088484252825,
+      "learning_rate": 3.0155617910429426e-05,
+      "loss": 0.3467,
+      "step": 1232
+    },
+    {
+      "epoch": 1.9863068868304472,
+      "grad_norm": 0.3081265438017872,
+      "learning_rate": 3.0136210480726365e-05,
+      "loss": 0.3385,
+      "step": 1233
+    },
+    {
+      "epoch": 1.9879178413209826,
+      "grad_norm": 0.31493139090080025,
+      "learning_rate": 3.011679019913587e-05,
+      "loss": 0.3539,
+      "step": 1234
+    },
+    {
+      "epoch": 1.9895287958115184,
+      "grad_norm": 0.3293171737255103,
+      "learning_rate": 3.0097357090281267e-05,
+      "loss": 0.3587,
+      "step": 1235
+    },
+    {
+      "epoch": 1.9911397503020538,
+      "grad_norm": 0.2763097409073324,
+      "learning_rate": 3.0077911178802152e-05,
+      "loss": 0.3572,
+      "step": 1236
+    },
+    {
+      "epoch": 1.9927507047925896,
+      "grad_norm": 0.29206192980238854,
+      "learning_rate": 3.0058452489354358e-05,
+      "loss": 0.334,
+      "step": 1237
+    },
+    {
+      "epoch": 1.9943616592831253,
+      "grad_norm": 0.31050171676421406,
+      "learning_rate": 3.0038981046609915e-05,
+      "loss": 0.3542,
+      "step": 1238
+    },
+    {
+      "epoch": 1.9959726137736609,
+      "grad_norm": 0.2887676714430983,
+      "learning_rate": 3.0019496875257012e-05,
+      "loss": 0.3474,
+      "step": 1239
+    },
+    {
+      "epoch": 1.9975835682641967,
+      "grad_norm": 0.2960252968104312,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.3486,
+      "step": 1240
+    },
+    {
+      "epoch": 1.999194522754732,
+      "grad_norm": 0.28911173594595874,
+      "learning_rate": 2.9980490445559325e-05,
+      "loss": 0.3589,
+      "step": 1241
+    },
+    {
+      "epoch": 2.000805477245268,
+      "grad_norm": 0.39770939551092377,
+      "learning_rate": 2.9960968236671504e-05,
+      "loss": 0.3328,
+      "step": 1242
+    },
+    {
+      "epoch": 2.0024164317358033,
+      "grad_norm": 0.2857834417846077,
+      "learning_rate": 2.9941433398089098e-05,
+      "loss": 0.3006,
+      "step": 1243
+    },
+    {
+      "epoch": 2.004027386226339,
+      "grad_norm": 0.6949713663880401,
+      "learning_rate": 2.99218859545807e-05,
+      "loss": 0.2604,
+      "step": 1244
+    },
+    {
+      "epoch": 2.005638340716875,
+      "grad_norm": 0.5111752099726601,
+      "learning_rate": 2.9902325930930868e-05,
+      "loss": 0.2976,
+      "step": 1245
+    },
+    {
+      "epoch": 2.0072492952074104,
+      "grad_norm": 0.39163668448086514,
+      "learning_rate": 2.9882753351940115e-05,
+      "loss": 0.2594,
+      "step": 1246
+    },
+    {
+      "epoch": 2.008860249697946,
+      "grad_norm": 0.33796868345118525,
+      "learning_rate": 2.9863168242424873e-05,
+      "loss": 0.2955,
+      "step": 1247
+    },
+    {
+      "epoch": 2.0104712041884816,
+      "grad_norm": 0.3941724192343171,
+      "learning_rate": 2.9843570627217463e-05,
+      "loss": 0.2846,
+      "step": 1248
+    },
+    {
+      "epoch": 2.0120821586790174,
+      "grad_norm": 0.31642626423981934,
+      "learning_rate": 2.9823960531166065e-05,
+      "loss": 0.2781,
+      "step": 1249
+    },
+    {
+      "epoch": 2.013693113169553,
+      "grad_norm": 0.37669153988473264,
+      "learning_rate": 2.980433797913467e-05,
+      "loss": 0.2853,
+      "step": 1250
+    },
+    {
+      "epoch": 2.0153040676600886,
+      "grad_norm": 0.30754959333932375,
+      "learning_rate": 2.978470299600308e-05,
+      "loss": 0.2796,
+      "step": 1251
+    },
+    {
+      "epoch": 2.0169150221506245,
+      "grad_norm": 0.3422825055523034,
+      "learning_rate": 2.9765055606666844e-05,
+      "loss": 0.2664,
+      "step": 1252
+    },
+    {
+      "epoch": 2.01852597664116,
+      "grad_norm": 0.31316330889075605,
+      "learning_rate": 2.9745395836037265e-05,
+      "loss": 0.2814,
+      "step": 1253
+    },
+    {
+      "epoch": 2.0201369311316957,
+      "grad_norm": 0.39624358182523745,
+      "learning_rate": 2.9725723709041304e-05,
+      "loss": 0.2882,
+      "step": 1254
+    },
+    {
+      "epoch": 2.021747885622231,
+      "grad_norm": 0.29976044657757916,
+      "learning_rate": 2.9706039250621626e-05,
+      "loss": 0.2763,
+      "step": 1255
+    },
+    {
+      "epoch": 2.023358840112767,
+      "grad_norm": 0.3659709335531895,
+      "learning_rate": 2.968634248573651e-05,
+      "loss": 0.2939,
+      "step": 1256
+    },
+    {
+      "epoch": 2.0249697946033023,
+      "grad_norm": 0.30126786571812936,
+      "learning_rate": 2.9666633439359857e-05,
+      "loss": 0.2716,
+      "step": 1257
+    },
+    {
+      "epoch": 2.026580749093838,
+      "grad_norm": 0.3122695764846602,
+      "learning_rate": 2.9646912136481116e-05,
+      "loss": 0.2658,
+      "step": 1258
+    },
+    {
+      "epoch": 2.028191703584374,
+      "grad_norm": 0.2867439501044589,
+      "learning_rate": 2.9627178602105296e-05,
+      "loss": 0.2676,
+      "step": 1259
+    },
+    {
+      "epoch": 2.0298026580749093,
+      "grad_norm": 0.30992230207590665,
+      "learning_rate": 2.960743286125291e-05,
+      "loss": 0.2955,
+      "step": 1260
+    },
+    {
+      "epoch": 2.031413612565445,
+      "grad_norm": 0.3058682460420191,
+      "learning_rate": 2.958767493895994e-05,
+      "loss": 0.2896,
+      "step": 1261
+    },
+    {
+      "epoch": 2.0330245670559806,
+      "grad_norm": 0.26862316860853597,
+      "learning_rate": 2.9567904860277825e-05,
+      "loss": 0.2566,
+      "step": 1262
+    },
+    {
+      "epoch": 2.0346355215465164,
+      "grad_norm": 0.25194544144940745,
+      "learning_rate": 2.95481226502734e-05,
+      "loss": 0.2478,
+      "step": 1263
+    },
+    {
+      "epoch": 2.036246476037052,
+      "grad_norm": 0.30167552158426153,
+      "learning_rate": 2.9528328334028903e-05,
+      "loss": 0.2744,
+      "step": 1264
+    },
+    {
+      "epoch": 2.0378574305275876,
+      "grad_norm": 0.28046902794293055,
+      "learning_rate": 2.9508521936641906e-05,
+      "loss": 0.2833,
+      "step": 1265
+    },
+    {
+      "epoch": 2.0394683850181234,
+      "grad_norm": 0.310202694342102,
+      "learning_rate": 2.948870348322531e-05,
+      "loss": 0.3095,
+      "step": 1266
+    },
+    {
+      "epoch": 2.041079339508659,
+      "grad_norm": 0.2675039895958826,
+      "learning_rate": 2.9468872998907285e-05,
+      "loss": 0.2643,
+      "step": 1267
+    },
+    {
+      "epoch": 2.0426902939991947,
+      "grad_norm": 0.2594449033037284,
+      "learning_rate": 2.9449030508831273e-05,
+      "loss": 0.2682,
+      "step": 1268
+    },
+    {
+      "epoch": 2.04430124848973,
+      "grad_norm": 0.28891086210870076,
+      "learning_rate": 2.9429176038155938e-05,
+      "loss": 0.2621,
+      "step": 1269
+    },
+    {
+      "epoch": 2.045912202980266,
+      "grad_norm": 0.3343156311987029,
+      "learning_rate": 2.9409309612055116e-05,
+      "loss": 0.3005,
+      "step": 1270
+    },
+    {
+      "epoch": 2.0475231574708013,
+      "grad_norm": 0.31461832017319397,
+      "learning_rate": 2.938943125571782e-05,
+      "loss": 0.281,
+      "step": 1271
+    },
+    {
+      "epoch": 2.049134111961337,
+      "grad_norm": 0.30370641290272277,
+      "learning_rate": 2.9369540994348175e-05,
+      "loss": 0.2864,
+      "step": 1272
+    },
+    {
+      "epoch": 2.050745066451873,
+      "grad_norm": 0.27610662631816574,
+      "learning_rate": 2.9349638853165427e-05,
+      "loss": 0.2672,
+      "step": 1273
+    },
+    {
+      "epoch": 2.0523560209424083,
+      "grad_norm": 0.3276272216535393,
+      "learning_rate": 2.932972485740385e-05,
+      "loss": 0.282,
+      "step": 1274
+    },
+    {
+      "epoch": 2.053966975432944,
+      "grad_norm": 0.3009539346491537,
+      "learning_rate": 2.9309799032312775e-05,
+      "loss": 0.253,
+      "step": 1275
+    },
+    {
+      "epoch": 2.0555779299234795,
+      "grad_norm": 0.28633699999733225,
+      "learning_rate": 2.9289861403156504e-05,
+      "loss": 0.2791,
+      "step": 1276
+    },
+    {
+      "epoch": 2.0571888844140154,
+      "grad_norm": 0.35069115032081216,
+      "learning_rate": 2.9269911995214354e-05,
+      "loss": 0.2621,
+      "step": 1277
+    },
+    {
+      "epoch": 2.0587998389045508,
+      "grad_norm": 0.2799484554842468,
+      "learning_rate": 2.924995083378051e-05,
+      "loss": 0.2723,
+      "step": 1278
+    },
+    {
+      "epoch": 2.0604107933950866,
+      "grad_norm": 0.36643461872273314,
+      "learning_rate": 2.922997794416412e-05,
+      "loss": 0.2745,
+      "step": 1279
+    },
+    {
+      "epoch": 2.0620217478856224,
+      "grad_norm": 0.2855006354725177,
+      "learning_rate": 2.920999335168917e-05,
+      "loss": 0.2604,
+      "step": 1280
+    },
+    {
+      "epoch": 2.063632702376158,
+      "grad_norm": 0.37126926892403106,
+      "learning_rate": 2.9189997081694493e-05,
+      "loss": 0.2772,
+      "step": 1281
+    },
+    {
+      "epoch": 2.0652436568666936,
+      "grad_norm": 0.3006772235704741,
+      "learning_rate": 2.916998915953373e-05,
+      "loss": 0.302,
+      "step": 1282
+    },
+    {
+      "epoch": 2.066854611357229,
+      "grad_norm": 0.3753958972390217,
+      "learning_rate": 2.914996961057528e-05,
+      "loss": 0.2811,
+      "step": 1283
+    },
+    {
+      "epoch": 2.068465565847765,
+      "grad_norm": 0.2994073547515185,
+      "learning_rate": 2.9129938460202312e-05,
+      "loss": 0.2859,
+      "step": 1284
+    },
+    {
+      "epoch": 2.0700765203383003,
+      "grad_norm": 0.31202683354041316,
+      "learning_rate": 2.910989573381268e-05,
+      "loss": 0.2733,
+      "step": 1285
+    },
+    {
+      "epoch": 2.071687474828836,
+      "grad_norm": 0.2767590884579919,
+      "learning_rate": 2.9089841456818935e-05,
+      "loss": 0.2515,
+      "step": 1286
+    },
+    {
+      "epoch": 2.073298429319372,
+      "grad_norm": 0.2735836950606523,
+      "learning_rate": 2.906977565464825e-05,
+      "loss": 0.2903,
+      "step": 1287
+    },
+    {
+      "epoch": 2.0749093838099073,
+      "grad_norm": 0.2921931649866422,
+      "learning_rate": 2.9049698352742438e-05,
+      "loss": 0.2981,
+      "step": 1288
+    },
+    {
+      "epoch": 2.076520338300443,
+      "grad_norm": 0.267867103961345,
+      "learning_rate": 2.9029609576557863e-05,
+      "loss": 0.2652,
+      "step": 1289
+    },
+    {
+      "epoch": 2.0781312927909785,
+      "grad_norm": 0.29224521078187127,
+      "learning_rate": 2.9009509351565462e-05,
+      "loss": 0.2647,
+      "step": 1290
+    },
+    {
+      "epoch": 2.0797422472815144,
+      "grad_norm": 0.2729335202374728,
+      "learning_rate": 2.8989397703250682e-05,
+      "loss": 0.2798,
+      "step": 1291
+    },
+    {
+      "epoch": 2.0813532017720497,
+      "grad_norm": 0.26817050857073244,
+      "learning_rate": 2.8969274657113452e-05,
+      "loss": 0.2839,
+      "step": 1292
+    },
+    {
+      "epoch": 2.0829641562625856,
+      "grad_norm": 0.2954611997992363,
+      "learning_rate": 2.8949140238668145e-05,
+      "loss": 0.3008,
+      "step": 1293
+    },
+    {
+      "epoch": 2.0845751107531214,
+      "grad_norm": 0.2584560509593604,
+      "learning_rate": 2.8928994473443557e-05,
+      "loss": 0.2863,
+      "step": 1294
+    },
+    {
+      "epoch": 2.086186065243657,
+      "grad_norm": 0.2697921301073831,
+      "learning_rate": 2.890883738698289e-05,
+      "loss": 0.2786,
+      "step": 1295
+    },
+    {
+      "epoch": 2.0877970197341926,
+      "grad_norm": 0.2803128875813109,
+      "learning_rate": 2.8888669004843665e-05,
+      "loss": 0.2734,
+      "step": 1296
+    },
+    {
+      "epoch": 2.089407974224728,
+      "grad_norm": 0.2381547290712765,
+      "learning_rate": 2.8868489352597762e-05,
+      "loss": 0.251,
+      "step": 1297
+    },
+    {
+      "epoch": 2.091018928715264,
+      "grad_norm": 0.3323977346030221,
+      "learning_rate": 2.8848298455831317e-05,
+      "loss": 0.308,
+      "step": 1298
+    },
+    {
+      "epoch": 2.0926298832057992,
+      "grad_norm": 0.27153635242123286,
+      "learning_rate": 2.882809634014475e-05,
+      "loss": 0.2589,
+      "step": 1299
+    },
+    {
+      "epoch": 2.094240837696335,
+      "grad_norm": 0.285739554905134,
+      "learning_rate": 2.880788303115269e-05,
+      "loss": 0.273,
+      "step": 1300
+    },
+    {
+      "epoch": 2.095851792186871,
+      "grad_norm": 0.3010326450465779,
+      "learning_rate": 2.878765855448396e-05,
+      "loss": 0.3033,
+      "step": 1301
+    },
+    {
+      "epoch": 2.0974627466774063,
+      "grad_norm": 0.24696037255401815,
+      "learning_rate": 2.876742293578155e-05,
+      "loss": 0.2609,
+      "step": 1302
+    },
+    {
+      "epoch": 2.099073701167942,
+      "grad_norm": 0.27166822832389825,
+      "learning_rate": 2.8747176200702572e-05,
+      "loss": 0.2551,
+      "step": 1303
+    },
+    {
+      "epoch": 2.1006846556584775,
+      "grad_norm": 0.2735038670067162,
+      "learning_rate": 2.8726918374918233e-05,
+      "loss": 0.2758,
+      "step": 1304
+    },
+    {
+      "epoch": 2.1022956101490133,
+      "grad_norm": 0.3028520148216462,
+      "learning_rate": 2.87066494841138e-05,
+      "loss": 0.2911,
+      "step": 1305
+    },
+    {
+      "epoch": 2.1039065646395487,
+      "grad_norm": 0.2872325900858017,
+      "learning_rate": 2.8686369553988576e-05,
+      "loss": 0.2843,
+      "step": 1306
+    },
+    {
+      "epoch": 2.1055175191300846,
+      "grad_norm": 0.27879094380561814,
+      "learning_rate": 2.8666078610255854e-05,
+      "loss": 0.2792,
+      "step": 1307
+    },
+    {
+      "epoch": 2.1071284736206204,
+      "grad_norm": 0.30483411415338657,
+      "learning_rate": 2.8645776678642893e-05,
+      "loss": 0.2495,
+      "step": 1308
+    },
+    {
+      "epoch": 2.108739428111156,
+      "grad_norm": 0.2683464007417797,
+      "learning_rate": 2.8625463784890884e-05,
+      "loss": 0.2867,
+      "step": 1309
+    },
+    {
+      "epoch": 2.1103503826016916,
+      "grad_norm": 0.29749249023584406,
+      "learning_rate": 2.8605139954754923e-05,
+      "loss": 0.278,
+      "step": 1310
+    },
+    {
+      "epoch": 2.111961337092227,
+      "grad_norm": 0.2687970108131372,
+      "learning_rate": 2.8584805214003967e-05,
+      "loss": 0.2504,
+      "step": 1311
+    },
+    {
+      "epoch": 2.113572291582763,
+      "grad_norm": 0.32466154003508846,
+      "learning_rate": 2.8564459588420807e-05,
+      "loss": 0.2991,
+      "step": 1312
+    },
+    {
+      "epoch": 2.115183246073298,
+      "grad_norm": 0.26356374059791937,
+      "learning_rate": 2.854410310380203e-05,
+      "loss": 0.2656,
+      "step": 1313
+    },
+    {
+      "epoch": 2.116794200563834,
+      "grad_norm": 0.2949078910210175,
+      "learning_rate": 2.8523735785958e-05,
+      "loss": 0.2684,
+      "step": 1314
+    },
+    {
+      "epoch": 2.11840515505437,
+      "grad_norm": 0.31103645726920576,
+      "learning_rate": 2.8503357660712815e-05,
+      "loss": 0.3014,
+      "step": 1315
+    },
+    {
+      "epoch": 2.1200161095449053,
+      "grad_norm": 0.3295336108394426,
+      "learning_rate": 2.8482968753904277e-05,
+      "loss": 0.2945,
+      "step": 1316
+    },
+    {
+      "epoch": 2.121627064035441,
+      "grad_norm": 0.3043674560354506,
+      "learning_rate": 2.8462569091383853e-05,
+      "loss": 0.2761,
+      "step": 1317
+    },
+    {
+      "epoch": 2.1232380185259765,
+      "grad_norm": 0.2945740096658187,
+      "learning_rate": 2.844215869901664e-05,
+      "loss": 0.2933,
+      "step": 1318
+    },
+    {
+      "epoch": 2.1248489730165123,
+      "grad_norm": 0.2751385832645397,
+      "learning_rate": 2.8421737602681364e-05,
+      "loss": 0.2657,
+      "step": 1319
+    },
+    {
+      "epoch": 2.1264599275070477,
+      "grad_norm": 0.2966644566850293,
+      "learning_rate": 2.8401305828270302e-05,
+      "loss": 0.2744,
+      "step": 1320
+    },
+    {
+      "epoch": 2.1280708819975835,
+      "grad_norm": 0.27307884228282997,
+      "learning_rate": 2.838086340168927e-05,
+      "loss": 0.2798,
+      "step": 1321
+    },
+    {
+      "epoch": 2.1296818364881194,
+      "grad_norm": 0.2877534471726608,
+      "learning_rate": 2.836041034885761e-05,
+      "loss": 0.2493,
+      "step": 1322
+    },
+    {
+      "epoch": 2.1312927909786548,
+      "grad_norm": 0.2614334519008766,
+      "learning_rate": 2.833994669570811e-05,
+      "loss": 0.2874,
+      "step": 1323
+    },
+    {
+      "epoch": 2.1329037454691906,
+      "grad_norm": 0.29160322965169916,
+      "learning_rate": 2.831947246818702e-05,
+      "loss": 0.2856,
+      "step": 1324
+    },
+    {
+      "epoch": 2.134514699959726,
+      "grad_norm": 0.2660119080149266,
+      "learning_rate": 2.829898769225399e-05,
+      "loss": 0.3064,
+      "step": 1325
+    },
+    {
+      "epoch": 2.136125654450262,
+      "grad_norm": 0.2912904811381936,
+      "learning_rate": 2.8278492393882032e-05,
+      "loss": 0.2657,
+      "step": 1326
+    },
+    {
+      "epoch": 2.1377366089407976,
+      "grad_norm": 0.2448795402010965,
+      "learning_rate": 2.8257986599057524e-05,
+      "loss": 0.2437,
+      "step": 1327
+    },
+    {
+      "epoch": 2.139347563431333,
+      "grad_norm": 0.27855483837999373,
+      "learning_rate": 2.8237470333780136e-05,
+      "loss": 0.281,
+      "step": 1328
+    },
+    {
+      "epoch": 2.140958517921869,
+      "grad_norm": 0.28864912755883265,
+      "learning_rate": 2.8216943624062815e-05,
+      "loss": 0.2634,
+      "step": 1329
+    },
+    {
+      "epoch": 2.1425694724124043,
+      "grad_norm": 0.2935166591639771,
+      "learning_rate": 2.8196406495931753e-05,
+      "loss": 0.2896,
+      "step": 1330
+    },
+    {
+      "epoch": 2.14418042690294,
+      "grad_norm": 0.27542513543126673,
+      "learning_rate": 2.8175858975426358e-05,
+      "loss": 0.2674,
+      "step": 1331
+    },
+    {
+      "epoch": 2.1457913813934755,
+      "grad_norm": 0.28218015707916827,
+      "learning_rate": 2.81553010885992e-05,
+      "loss": 0.2784,
+      "step": 1332
+    },
+    {
+      "epoch": 2.1474023358840113,
+      "grad_norm": 0.26365559112393655,
+      "learning_rate": 2.813473286151601e-05,
+      "loss": 0.2661,
+      "step": 1333
+    },
+    {
+      "epoch": 2.1490132903745467,
+      "grad_norm": 0.2649219889260671,
+      "learning_rate": 2.8114154320255612e-05,
+      "loss": 0.2729,
+      "step": 1334
+    },
+    {
+      "epoch": 2.1506242448650825,
+      "grad_norm": 0.2526894475726239,
+      "learning_rate": 2.809356549090992e-05,
+      "loss": 0.2877,
+      "step": 1335
+    },
+    {
+      "epoch": 2.1522351993556184,
+      "grad_norm": 0.2846315930814484,
+      "learning_rate": 2.8072966399583897e-05,
+      "loss": 0.2967,
+      "step": 1336
+    },
+    {
+      "epoch": 2.1538461538461537,
+      "grad_norm": 0.24679278451222952,
+      "learning_rate": 2.8052357072395494e-05,
+      "loss": 0.2792,
+      "step": 1337
+    },
+    {
+      "epoch": 2.1554571083366896,
+      "grad_norm": 0.25433364102855244,
+      "learning_rate": 2.8031737535475668e-05,
+      "loss": 0.273,
+      "step": 1338
+    },
+    {
+      "epoch": 2.157068062827225,
+      "grad_norm": 0.32637103264618766,
+      "learning_rate": 2.801110781496831e-05,
+      "loss": 0.29,
+      "step": 1339
+    },
+    {
+      "epoch": 2.158679017317761,
+      "grad_norm": 0.2554277138169469,
+      "learning_rate": 2.799046793703021e-05,
+      "loss": 0.268,
+      "step": 1340
+    },
+    {
+      "epoch": 2.1602899718082966,
+      "grad_norm": 0.26969310352840337,
+      "learning_rate": 2.796981792783105e-05,
+      "loss": 0.2853,
+      "step": 1341
+    },
+    {
+      "epoch": 2.161900926298832,
+      "grad_norm": 0.2648111637517603,
+      "learning_rate": 2.7949157813553366e-05,
+      "loss": 0.2626,
+      "step": 1342
+    },
+    {
+      "epoch": 2.163511880789368,
+      "grad_norm": 0.2559746677208276,
+      "learning_rate": 2.7928487620392487e-05,
+      "loss": 0.3044,
+      "step": 1343
+    },
+    {
+      "epoch": 2.1651228352799032,
+      "grad_norm": 0.2714805738931639,
+      "learning_rate": 2.790780737455654e-05,
+      "loss": 0.2633,
+      "step": 1344
+    },
+    {
+      "epoch": 2.166733789770439,
+      "grad_norm": 0.24864674470379333,
+      "learning_rate": 2.7887117102266373e-05,
+      "loss": 0.2552,
+      "step": 1345
+    },
+    {
+      "epoch": 2.1683447442609745,
+      "grad_norm": 0.2547234928945848,
+      "learning_rate": 2.786641682975558e-05,
+      "loss": 0.2819,
+      "step": 1346
+    },
+    {
+      "epoch": 2.1699556987515103,
+      "grad_norm": 0.26524020894968475,
+      "learning_rate": 2.78457065832704e-05,
+      "loss": 0.3123,
+      "step": 1347
+    },
+    {
+      "epoch": 2.171566653242046,
+      "grad_norm": 0.2449565344886634,
+      "learning_rate": 2.782498638906975e-05,
+      "loss": 0.2643,
+      "step": 1348
+    },
+    {
+      "epoch": 2.1731776077325815,
+      "grad_norm": 0.2893223634942532,
+      "learning_rate": 2.780425627342514e-05,
+      "loss": 0.2816,
+      "step": 1349
+    },
+    {
+      "epoch": 2.1747885622231173,
+      "grad_norm": 0.2641617143950078,
+      "learning_rate": 2.7783516262620657e-05,
+      "loss": 0.2727,
+      "step": 1350
+    },
+    {
+      "epoch": 2.1763995167136527,
+      "grad_norm": 0.24627345310846985,
+      "learning_rate": 2.7762766382952948e-05,
+      "loss": 0.2814,
+      "step": 1351
+    },
+    {
+      "epoch": 2.1780104712041886,
+      "grad_norm": 0.28246010031678387,
+      "learning_rate": 2.7742006660731164e-05,
+      "loss": 0.2872,
+      "step": 1352
+    },
+    {
+      "epoch": 2.179621425694724,
+      "grad_norm": 0.27546870573554,
+      "learning_rate": 2.7721237122276944e-05,
+      "loss": 0.2822,
+      "step": 1353
+    },
+    {
+      "epoch": 2.18123238018526,
+      "grad_norm": 0.2520502824176606,
+      "learning_rate": 2.7700457793924357e-05,
+      "loss": 0.2802,
+      "step": 1354
+    },
+    {
+      "epoch": 2.1828433346757956,
+      "grad_norm": 0.26219847355655707,
+      "learning_rate": 2.767966870201991e-05,
+      "loss": 0.3008,
+      "step": 1355
+    },
+    {
+      "epoch": 2.184454289166331,
+      "grad_norm": 0.24699012142387217,
+      "learning_rate": 2.765886987292246e-05,
+      "loss": 0.2777,
+      "step": 1356
+    },
+    {
+      "epoch": 2.186065243656867,
+      "grad_norm": 0.23388920141854022,
+      "learning_rate": 2.7638061333003236e-05,
+      "loss": 0.2587,
+      "step": 1357
+    },
+    {
+      "epoch": 2.187676198147402,
+      "grad_norm": 0.26392578878534023,
+      "learning_rate": 2.7617243108645753e-05,
+      "loss": 0.3032,
+      "step": 1358
+    },
+    {
+      "epoch": 2.189287152637938,
+      "grad_norm": 0.23761776973362883,
+      "learning_rate": 2.759641522624583e-05,
+      "loss": 0.2498,
+      "step": 1359
+    },
+    {
+      "epoch": 2.1908981071284734,
+      "grad_norm": 0.27175194130700814,
+      "learning_rate": 2.7575577712211524e-05,
+      "loss": 0.2634,
+      "step": 1360
+    },
+    {
+      "epoch": 2.1925090616190093,
+      "grad_norm": 0.2676460221850602,
+      "learning_rate": 2.755473059296309e-05,
+      "loss": 0.2786,
+      "step": 1361
+    },
+    {
+      "epoch": 2.194120016109545,
+      "grad_norm": 0.276401287454354,
+      "learning_rate": 2.7533873894932996e-05,
+      "loss": 0.2798,
+      "step": 1362
+    },
+    {
+      "epoch": 2.1957309706000805,
+      "grad_norm": 0.24034592871532384,
+      "learning_rate": 2.7513007644565806e-05,
+      "loss": 0.262,
+      "step": 1363
+    },
+    {
+      "epoch": 2.1973419250906163,
+      "grad_norm": 0.25935727042899787,
+      "learning_rate": 2.7492131868318247e-05,
+      "loss": 0.2785,
+      "step": 1364
+    },
+    {
+      "epoch": 2.1989528795811517,
+      "grad_norm": 0.26803580380287895,
+      "learning_rate": 2.7471246592659075e-05,
+      "loss": 0.286,
+      "step": 1365
+    },
+    {
+      "epoch": 2.2005638340716875,
+      "grad_norm": 0.2801935730796939,
+      "learning_rate": 2.745035184406913e-05,
+      "loss": 0.2848,
+      "step": 1366
+    },
+    {
+      "epoch": 2.202174788562223,
+      "grad_norm": 0.25675623918826773,
+      "learning_rate": 2.7429447649041243e-05,
+      "loss": 0.2602,
+      "step": 1367
+    },
+    {
+      "epoch": 2.2037857430527588,
+      "grad_norm": 0.25791438655113136,
+      "learning_rate": 2.7408534034080228e-05,
+      "loss": 0.262,
+      "step": 1368
+    },
+    {
+      "epoch": 2.2053966975432946,
+      "grad_norm": 0.2753646395273099,
+      "learning_rate": 2.7387611025702837e-05,
+      "loss": 0.2958,
+      "step": 1369
+    },
+    {
+      "epoch": 2.20700765203383,
+      "grad_norm": 0.2441198130082228,
+      "learning_rate": 2.736667865043775e-05,
+      "loss": 0.2689,
+      "step": 1370
+    },
+    {
+      "epoch": 2.208618606524366,
+      "grad_norm": 0.25462467078769574,
+      "learning_rate": 2.734573693482549e-05,
+      "loss": 0.2802,
+      "step": 1371
+    },
+    {
+      "epoch": 2.210229561014901,
+      "grad_norm": 0.25852451115604996,
+      "learning_rate": 2.732478590541846e-05,
+      "loss": 0.2712,
+      "step": 1372
+    },
+    {
+      "epoch": 2.211840515505437,
+      "grad_norm": 0.25961347780273764,
+      "learning_rate": 2.7303825588780844e-05,
+      "loss": 0.2633,
+      "step": 1373
+    },
+    {
+      "epoch": 2.2134514699959724,
+      "grad_norm": 0.3037253510918743,
+      "learning_rate": 2.7282856011488615e-05,
+      "loss": 0.2878,
+      "step": 1374
+    },
+    {
+      "epoch": 2.2150624244865083,
+      "grad_norm": 0.22592118401629108,
+      "learning_rate": 2.7261877200129495e-05,
+      "loss": 0.2416,
+      "step": 1375
+    },
+    {
+      "epoch": 2.216673378977044,
+      "grad_norm": 0.3573981841929647,
+      "learning_rate": 2.724088918130289e-05,
+      "loss": 0.3017,
+      "step": 1376
+    },
+    {
+      "epoch": 2.2182843334675795,
+      "grad_norm": 0.2512932467094342,
+      "learning_rate": 2.721989198161991e-05,
+      "loss": 0.2794,
+      "step": 1377
+    },
+    {
+      "epoch": 2.2198952879581153,
+      "grad_norm": 0.3159257756922127,
+      "learning_rate": 2.7198885627703266e-05,
+      "loss": 0.2618,
+      "step": 1378
+    },
+    {
+      "epoch": 2.2215062424486507,
+      "grad_norm": 0.2599794920778706,
+      "learning_rate": 2.7177870146187323e-05,
+      "loss": 0.2728,
+      "step": 1379
+    },
+    {
+      "epoch": 2.2231171969391865,
+      "grad_norm": 0.2978036146497738,
+      "learning_rate": 2.7156845563717987e-05,
+      "loss": 0.2991,
+      "step": 1380
+    },
+    {
+      "epoch": 2.224728151429722,
+      "grad_norm": 0.24686915917646665,
+      "learning_rate": 2.7135811906952714e-05,
+      "loss": 0.2664,
+      "step": 1381
+    },
+    {
+      "epoch": 2.2263391059202577,
+      "grad_norm": 0.28407633766585894,
+      "learning_rate": 2.711476920256046e-05,
+      "loss": 0.3044,
+      "step": 1382
+    },
+    {
+      "epoch": 2.2279500604107936,
+      "grad_norm": 0.260934439271912,
+      "learning_rate": 2.709371747722166e-05,
+      "loss": 0.2783,
+      "step": 1383
+    },
+    {
+      "epoch": 2.229561014901329,
+      "grad_norm": 0.24471982062639003,
+      "learning_rate": 2.7072656757628188e-05,
+      "loss": 0.2686,
+      "step": 1384
+    },
+    {
+      "epoch": 2.231171969391865,
+      "grad_norm": 0.2633341856655095,
+      "learning_rate": 2.7051587070483307e-05,
+      "loss": 0.28,
+      "step": 1385
+    },
+    {
+      "epoch": 2.2327829238824,
+      "grad_norm": 0.26507841081942923,
+      "learning_rate": 2.7030508442501667e-05,
+      "loss": 0.2783,
+      "step": 1386
+    },
+    {
+      "epoch": 2.234393878372936,
+      "grad_norm": 0.25464764785100497,
+      "learning_rate": 2.7009420900409237e-05,
+      "loss": 0.2821,
+      "step": 1387
+    },
+    {
+      "epoch": 2.2360048328634714,
+      "grad_norm": 0.2868050675323047,
+      "learning_rate": 2.6988324470943315e-05,
+      "loss": 0.283,
+      "step": 1388
+    },
+    {
+      "epoch": 2.2376157873540072,
+      "grad_norm": 0.24611229723847267,
+      "learning_rate": 2.6967219180852448e-05,
+      "loss": 0.2822,
+      "step": 1389
+    },
+    {
+      "epoch": 2.239226741844543,
+      "grad_norm": 0.27187189363506403,
+      "learning_rate": 2.6946105056896406e-05,
+      "loss": 0.2994,
+      "step": 1390
+    },
+    {
+      "epoch": 2.2408376963350785,
+      "grad_norm": 0.21615468885402625,
+      "learning_rate": 2.692498212584619e-05,
+      "loss": 0.2415,
+      "step": 1391
+    },
+    {
+      "epoch": 2.2424486508256143,
+      "grad_norm": 0.26600283632882193,
+      "learning_rate": 2.6903850414483953e-05,
+      "loss": 0.2836,
+      "step": 1392
+    },
+    {
+      "epoch": 2.2440596053161497,
+      "grad_norm": 0.23728976141998126,
+      "learning_rate": 2.6882709949602966e-05,
+      "loss": 0.2841,
+      "step": 1393
+    },
+    {
+      "epoch": 2.2456705598066855,
+      "grad_norm": 0.25798754292795073,
+      "learning_rate": 2.6861560758007627e-05,
+      "loss": 0.2658,
+      "step": 1394
+    },
+    {
+      "epoch": 2.247281514297221,
+      "grad_norm": 0.2734561974922546,
+      "learning_rate": 2.684040286651338e-05,
+      "loss": 0.306,
+      "step": 1395
+    },
+    {
+      "epoch": 2.2488924687877567,
+      "grad_norm": 0.2858311588641443,
+      "learning_rate": 2.6819236301946697e-05,
+      "loss": 0.2793,
+      "step": 1396
+    },
+    {
+      "epoch": 2.2505034232782926,
+      "grad_norm": 0.2584315966942385,
+      "learning_rate": 2.6798061091145062e-05,
+      "loss": 0.2597,
+      "step": 1397
+    },
+    {
+      "epoch": 2.252114377768828,
+      "grad_norm": 0.27927968714043205,
+      "learning_rate": 2.677687726095691e-05,
+      "loss": 0.2817,
+      "step": 1398
+    },
+    {
+      "epoch": 2.253725332259364,
+      "grad_norm": 0.33168687660726137,
+      "learning_rate": 2.6755684838241614e-05,
+      "loss": 0.3129,
+      "step": 1399
+    },
+    {
+      "epoch": 2.255336286749899,
+      "grad_norm": 0.2554635016264873,
+      "learning_rate": 2.673448384986943e-05,
+      "loss": 0.2854,
+      "step": 1400
+    },
+    {
+      "epoch": 2.256947241240435,
+      "grad_norm": 0.2580831506179042,
+      "learning_rate": 2.6713274322721484e-05,
+      "loss": 0.2674,
+      "step": 1401
+    },
+    {
+      "epoch": 2.258558195730971,
+      "grad_norm": 0.2716906409548147,
+      "learning_rate": 2.669205628368972e-05,
+      "loss": 0.2931,
+      "step": 1402
+    },
+    {
+      "epoch": 2.260169150221506,
+      "grad_norm": 0.22672601567358644,
+      "learning_rate": 2.6670829759676882e-05,
+      "loss": 0.2621,
+      "step": 1403
+    },
+    {
+      "epoch": 2.261780104712042,
+      "grad_norm": 0.27308559599455845,
+      "learning_rate": 2.6649594777596476e-05,
+      "loss": 0.2802,
+      "step": 1404
+    },
+    {
+      "epoch": 2.2633910592025774,
+      "grad_norm": 0.25346462509049156,
+      "learning_rate": 2.6628351364372717e-05,
+      "loss": 0.2546,
+      "step": 1405
+    },
+    {
+      "epoch": 2.2650020136931133,
+      "grad_norm": 0.24996293140763204,
+      "learning_rate": 2.6607099546940526e-05,
+      "loss": 0.2944,
+      "step": 1406
+    },
+    {
+      "epoch": 2.2666129681836487,
+      "grad_norm": 0.24233262846738599,
+      "learning_rate": 2.6585839352245467e-05,
+      "loss": 0.2774,
+      "step": 1407
+    },
+    {
+      "epoch": 2.2682239226741845,
+      "grad_norm": 0.2583809561916111,
+      "learning_rate": 2.6564570807243728e-05,
+      "loss": 0.2653,
+      "step": 1408
+    },
+    {
+      "epoch": 2.26983487716472,
+      "grad_norm": 0.23405502541905018,
+      "learning_rate": 2.65432939389021e-05,
+      "loss": 0.2733,
+      "step": 1409
+    },
+    {
+      "epoch": 2.2714458316552557,
+      "grad_norm": 0.27446386749155977,
+      "learning_rate": 2.6522008774197902e-05,
+      "loss": 0.3221,
+      "step": 1410
+    },
+    {
+      "epoch": 2.2730567861457915,
+      "grad_norm": 0.2384738855739092,
+      "learning_rate": 2.6500715340118995e-05,
+      "loss": 0.2705,
+      "step": 1411
+    },
+    {
+      "epoch": 2.274667740636327,
+      "grad_norm": 0.24307003681995026,
+      "learning_rate": 2.6479413663663706e-05,
+      "loss": 0.2785,
+      "step": 1412
+    },
+    {
+      "epoch": 2.2762786951268628,
+      "grad_norm": 0.2526921045428496,
+      "learning_rate": 2.6458103771840835e-05,
+      "loss": 0.2871,
+      "step": 1413
+    },
+    {
+      "epoch": 2.277889649617398,
+      "grad_norm": 0.24931242626377267,
+      "learning_rate": 2.6436785691669578e-05,
+      "loss": 0.2808,
+      "step": 1414
+    },
+    {
+      "epoch": 2.279500604107934,
+      "grad_norm": 0.2887386630317653,
+      "learning_rate": 2.6415459450179515e-05,
+      "loss": 0.3049,
+      "step": 1415
+    },
+    {
+      "epoch": 2.28111155859847,
+      "grad_norm": 0.25033808397227236,
+      "learning_rate": 2.6394125074410584e-05,
+      "loss": 0.2904,
+      "step": 1416
+    },
+    {
+      "epoch": 2.282722513089005,
+      "grad_norm": 0.2699612396823536,
+      "learning_rate": 2.6372782591413034e-05,
+      "loss": 0.2729,
+      "step": 1417
+    },
+    {
+      "epoch": 2.284333467579541,
+      "grad_norm": 0.23924615054624318,
+      "learning_rate": 2.635143202824739e-05,
+      "loss": 0.2686,
+      "step": 1418
+    },
+    {
+      "epoch": 2.2859444220700764,
+      "grad_norm": 0.279887893837138,
+      "learning_rate": 2.6330073411984418e-05,
+      "loss": 0.2697,
+      "step": 1419
+    },
+    {
+      "epoch": 2.2875553765606123,
+      "grad_norm": 0.2545245668178236,
+      "learning_rate": 2.6308706769705118e-05,
+      "loss": 0.2698,
+      "step": 1420
+    },
+    {
+      "epoch": 2.2891663310511476,
+      "grad_norm": 0.2544237823113252,
+      "learning_rate": 2.6287332128500616e-05,
+      "loss": 0.2784,
+      "step": 1421
+    },
+    {
+      "epoch": 2.2907772855416835,
+      "grad_norm": 0.2588334468257935,
+      "learning_rate": 2.6265949515472247e-05,
+      "loss": 0.2856,
+      "step": 1422
+    },
+    {
+      "epoch": 2.292388240032219,
+      "grad_norm": 0.2640309275817744,
+      "learning_rate": 2.6244558957731386e-05,
+      "loss": 0.2921,
+      "step": 1423
+    },
+    {
+      "epoch": 2.2939991945227547,
+      "grad_norm": 0.27035992387226954,
+      "learning_rate": 2.622316048239954e-05,
+      "loss": 0.2719,
+      "step": 1424
+    },
+    {
+      "epoch": 2.2956101490132905,
+      "grad_norm": 0.2763448005510581,
+      "learning_rate": 2.6201754116608222e-05,
+      "loss": 0.2977,
+      "step": 1425
+    },
+    {
+      "epoch": 2.297221103503826,
+      "grad_norm": 0.2454033455212278,
+      "learning_rate": 2.618033988749895e-05,
+      "loss": 0.261,
+      "step": 1426
+    },
+    {
+      "epoch": 2.2988320579943617,
+      "grad_norm": 0.28206219976356545,
+      "learning_rate": 2.615891782222322e-05,
+      "loss": 0.3032,
+      "step": 1427
+    },
+    {
+      "epoch": 2.300443012484897,
+      "grad_norm": 0.25919572585437717,
+      "learning_rate": 2.6137487947942472e-05,
+      "loss": 0.2932,
+      "step": 1428
+    },
+    {
+      "epoch": 2.302053966975433,
+      "grad_norm": 0.23936651714279733,
+      "learning_rate": 2.6116050291828026e-05,
+      "loss": 0.2885,
+      "step": 1429
+    },
+    {
+      "epoch": 2.303664921465969,
+      "grad_norm": 0.26079396517622616,
+      "learning_rate": 2.6094604881061076e-05,
+      "loss": 0.2644,
+      "step": 1430
+    },
+    {
+      "epoch": 2.305275875956504,
+      "grad_norm": 0.27314314971219317,
+      "learning_rate": 2.607315174283267e-05,
+      "loss": 0.3197,
+      "step": 1431
+    },
+    {
+      "epoch": 2.30688683044704,
+      "grad_norm": 0.23566335984437972,
+      "learning_rate": 2.6051690904343616e-05,
+      "loss": 0.2795,
+      "step": 1432
+    },
+    {
+      "epoch": 2.3084977849375754,
+      "grad_norm": 0.2598558733558277,
+      "learning_rate": 2.6030222392804526e-05,
+      "loss": 0.2833,
+      "step": 1433
+    },
+    {
+      "epoch": 2.3101087394281112,
+      "grad_norm": 0.2226975838547479,
+      "learning_rate": 2.60087462354357e-05,
+      "loss": 0.2551,
+      "step": 1434
+    },
+    {
+      "epoch": 2.3117196939186466,
+      "grad_norm": 0.2798036466100915,
+      "learning_rate": 2.5987262459467168e-05,
+      "loss": 0.3226,
+      "step": 1435
+    },
+    {
+      "epoch": 2.3133306484091825,
+      "grad_norm": 0.2241641247015054,
+      "learning_rate": 2.5965771092138586e-05,
+      "loss": 0.2618,
+      "step": 1436
+    },
+    {
+      "epoch": 2.314941602899718,
+      "grad_norm": 0.28580401578871045,
+      "learning_rate": 2.5944272160699272e-05,
+      "loss": 0.2888,
+      "step": 1437
+    },
+    {
+      "epoch": 2.3165525573902537,
+      "grad_norm": 0.256570118424594,
+      "learning_rate": 2.5922765692408112e-05,
+      "loss": 0.2771,
+      "step": 1438
+    },
+    {
+      "epoch": 2.3181635118807895,
+      "grad_norm": 0.2578027303979569,
+      "learning_rate": 2.5901251714533543e-05,
+      "loss": 0.2574,
+      "step": 1439
+    },
+    {
+      "epoch": 2.319774466371325,
+      "grad_norm": 0.2798470005022122,
+      "learning_rate": 2.5879730254353543e-05,
+      "loss": 0.2981,
+      "step": 1440
+    },
+    {
+      "epoch": 2.3213854208618607,
+      "grad_norm": 0.27607698811706227,
+      "learning_rate": 2.585820133915555e-05,
+      "loss": 0.2865,
+      "step": 1441
+    },
+    {
+      "epoch": 2.322996375352396,
+      "grad_norm": 0.26912286490683257,
+      "learning_rate": 2.58366649962365e-05,
+      "loss": 0.2695,
+      "step": 1442
+    },
+    {
+      "epoch": 2.324607329842932,
+      "grad_norm": 0.26457014248103583,
+      "learning_rate": 2.581512125290269e-05,
+      "loss": 0.2812,
+      "step": 1443
+    },
+    {
+      "epoch": 2.326218284333468,
+      "grad_norm": 0.2598505097593347,
+      "learning_rate": 2.579357013646985e-05,
+      "loss": 0.2606,
+      "step": 1444
+    },
+    {
+      "epoch": 2.327829238824003,
+      "grad_norm": 0.31694547112574056,
+      "learning_rate": 2.5772011674263017e-05,
+      "loss": 0.2661,
+      "step": 1445
+    },
+    {
+      "epoch": 2.329440193314539,
+      "grad_norm": 0.24909334654252163,
+      "learning_rate": 2.575044589361657e-05,
+      "loss": 0.2705,
+      "step": 1446
+    },
+    {
+      "epoch": 2.3310511478050744,
+      "grad_norm": 0.2774620303831127,
+      "learning_rate": 2.5728872821874155e-05,
+      "loss": 0.2874,
+      "step": 1447
+    },
+    {
+      "epoch": 2.33266210229561,
+      "grad_norm": 0.2735139105570509,
+      "learning_rate": 2.5707292486388675e-05,
+      "loss": 0.3037,
+      "step": 1448
+    },
+    {
+      "epoch": 2.3342730567861456,
+      "grad_norm": 0.2740635594254122,
+      "learning_rate": 2.5685704914522223e-05,
+      "loss": 0.2897,
+      "step": 1449
+    },
+    {
+      "epoch": 2.3358840112766814,
+      "grad_norm": 0.24084315748689009,
+      "learning_rate": 2.566411013364608e-05,
+      "loss": 0.2486,
+      "step": 1450
+    },
+    {
+      "epoch": 2.3374949657672173,
+      "grad_norm": 0.287578601994385,
+      "learning_rate": 2.5642508171140675e-05,
+      "loss": 0.3072,
+      "step": 1451
+    },
+    {
+      "epoch": 2.3391059202577527,
+      "grad_norm": 0.2550922001271773,
+      "learning_rate": 2.562089905439552e-05,
+      "loss": 0.2716,
+      "step": 1452
+    },
+    {
+      "epoch": 2.3407168747482885,
+      "grad_norm": 0.28349657757580804,
+      "learning_rate": 2.5599282810809222e-05,
+      "loss": 0.2776,
+      "step": 1453
+    },
+    {
+      "epoch": 2.342327829238824,
+      "grad_norm": 0.2539299708169048,
+      "learning_rate": 2.5577659467789397e-05,
+      "loss": 0.2882,
+      "step": 1454
+    },
+    {
+      "epoch": 2.3439387837293597,
+      "grad_norm": 0.2638598040941746,
+      "learning_rate": 2.5556029052752704e-05,
+      "loss": 0.2753,
+      "step": 1455
+    },
+    {
+      "epoch": 2.345549738219895,
+      "grad_norm": 0.26546655271952696,
+      "learning_rate": 2.5534391593124717e-05,
+      "loss": 0.2727,
+      "step": 1456
+    },
+    {
+      "epoch": 2.347160692710431,
+      "grad_norm": 0.274114516324995,
+      "learning_rate": 2.5512747116339985e-05,
+      "loss": 0.2679,
+      "step": 1457
+    },
+    {
+      "epoch": 2.3487716472009668,
+      "grad_norm": 0.27615786351600663,
+      "learning_rate": 2.5491095649841936e-05,
+      "loss": 0.29,
+      "step": 1458
+    },
+    {
+      "epoch": 2.350382601691502,
+      "grad_norm": 0.2523360738477017,
+      "learning_rate": 2.5469437221082855e-05,
+      "loss": 0.2876,
+      "step": 1459
+    },
+    {
+      "epoch": 2.351993556182038,
+      "grad_norm": 0.26083088114961767,
+      "learning_rate": 2.5447771857523868e-05,
+      "loss": 0.2615,
+      "step": 1460
+    },
+    {
+      "epoch": 2.3536045106725734,
+      "grad_norm": 0.26930621483118783,
+      "learning_rate": 2.5426099586634885e-05,
+      "loss": 0.2798,
+      "step": 1461
+    },
+    {
+      "epoch": 2.355215465163109,
+      "grad_norm": 0.25754096649737024,
+      "learning_rate": 2.5404420435894578e-05,
+      "loss": 0.2668,
+      "step": 1462
+    },
+    {
+      "epoch": 2.356826419653645,
+      "grad_norm": 0.2699366881095369,
+      "learning_rate": 2.538273443279033e-05,
+      "loss": 0.3052,
+      "step": 1463
+    },
+    {
+      "epoch": 2.3584373741441804,
+      "grad_norm": 0.2635570721152839,
+      "learning_rate": 2.5361041604818244e-05,
+      "loss": 0.2682,
+      "step": 1464
+    },
+    {
+      "epoch": 2.3600483286347163,
+      "grad_norm": 0.2407160146785441,
+      "learning_rate": 2.5339341979483037e-05,
+      "loss": 0.2746,
+      "step": 1465
+    },
+    {
+      "epoch": 2.3616592831252516,
+      "grad_norm": 0.29061883733407995,
+      "learning_rate": 2.531763558429807e-05,
+      "loss": 0.2968,
+      "step": 1466
+    },
+    {
+      "epoch": 2.3632702376157875,
+      "grad_norm": 0.24834565149196813,
+      "learning_rate": 2.5295922446785275e-05,
+      "loss": 0.2667,
+      "step": 1467
+    },
+    {
+      "epoch": 2.364881192106323,
+      "grad_norm": 0.2583516235073483,
+      "learning_rate": 2.527420259447514e-05,
+      "loss": 0.272,
+      "step": 1468
+    },
+    {
+      "epoch": 2.3664921465968587,
+      "grad_norm": 0.23935397713698203,
+      "learning_rate": 2.5252476054906668e-05,
+      "loss": 0.2852,
+      "step": 1469
+    },
+    {
+      "epoch": 2.368103101087394,
+      "grad_norm": 0.2530197056590334,
+      "learning_rate": 2.523074285562734e-05,
+      "loss": 0.2705,
+      "step": 1470
+    },
+    {
+      "epoch": 2.36971405557793,
+      "grad_norm": 0.25163841004710474,
+      "learning_rate": 2.5209003024193067e-05,
+      "loss": 0.2905,
+      "step": 1471
+    },
+    {
+      "epoch": 2.3713250100684657,
+      "grad_norm": 0.23237404626614086,
+      "learning_rate": 2.518725658816819e-05,
+      "loss": 0.2526,
+      "step": 1472
+    },
+    {
+      "epoch": 2.372935964559001,
+      "grad_norm": 0.26024820542438537,
+      "learning_rate": 2.5165503575125413e-05,
+      "loss": 0.2846,
+      "step": 1473
+    },
+    {
+      "epoch": 2.374546919049537,
+      "grad_norm": 0.2585972661841443,
+      "learning_rate": 2.514374401264578e-05,
+      "loss": 0.2749,
+      "step": 1474
+    },
+    {
+      "epoch": 2.3761578735400724,
+      "grad_norm": 0.2845138208353183,
+      "learning_rate": 2.5121977928318638e-05,
+      "loss": 0.2748,
+      "step": 1475
+    },
+    {
+      "epoch": 2.377768828030608,
+      "grad_norm": 0.2495500233521992,
+      "learning_rate": 2.5100205349741602e-05,
+      "loss": 0.2787,
+      "step": 1476
+    },
+    {
+      "epoch": 2.379379782521144,
+      "grad_norm": 0.25151165206564124,
+      "learning_rate": 2.507842630452054e-05,
+      "loss": 0.3044,
+      "step": 1477
+    },
+    {
+      "epoch": 2.3809907370116794,
+      "grad_norm": 0.26429785064547257,
+      "learning_rate": 2.5056640820269484e-05,
+      "loss": 0.2835,
+      "step": 1478
+    },
+    {
+      "epoch": 2.3826016915022152,
+      "grad_norm": 0.2461207371041576,
+      "learning_rate": 2.503484892461066e-05,
+      "loss": 0.2774,
+      "step": 1479
+    },
+    {
+      "epoch": 2.3842126459927506,
+      "grad_norm": 0.27735647570825706,
+      "learning_rate": 2.5013050645174414e-05,
+      "loss": 0.2757,
+      "step": 1480
+    },
+    {
+      "epoch": 2.3858236004832865,
+      "grad_norm": 0.2755004610517415,
+      "learning_rate": 2.499124600959918e-05,
+      "loss": 0.2728,
+      "step": 1481
+    },
+    {
+      "epoch": 2.387434554973822,
+      "grad_norm": 0.2752976426963208,
+      "learning_rate": 2.4969435045531457e-05,
+      "loss": 0.2604,
+      "step": 1482
+    },
+    {
+      "epoch": 2.3890455094643577,
+      "grad_norm": 0.2716284118402911,
+      "learning_rate": 2.494761778062577e-05,
+      "loss": 0.282,
+      "step": 1483
+    },
+    {
+      "epoch": 2.390656463954893,
+      "grad_norm": 0.25175456114152706,
+      "learning_rate": 2.4925794242544626e-05,
+      "loss": 0.2972,
+      "step": 1484
+    },
+    {
+      "epoch": 2.392267418445429,
+      "grad_norm": 0.2690581205659193,
+      "learning_rate": 2.490396445895849e-05,
+      "loss": 0.2779,
+      "step": 1485
+    },
+    {
+      "epoch": 2.3938783729359647,
+      "grad_norm": 0.25560689396934266,
+      "learning_rate": 2.4882128457545748e-05,
+      "loss": 0.2728,
+      "step": 1486
+    },
+    {
+      "epoch": 2.3954893274265,
+      "grad_norm": 0.25135288220419316,
+      "learning_rate": 2.4860286265992667e-05,
+      "loss": 0.2888,
+      "step": 1487
+    },
+    {
+      "epoch": 2.397100281917036,
+      "grad_norm": 0.2492488977293215,
+      "learning_rate": 2.4838437911993355e-05,
+      "loss": 0.2714,
+      "step": 1488
+    },
+    {
+      "epoch": 2.3987112364075713,
+      "grad_norm": 0.2374341024869927,
+      "learning_rate": 2.4816583423249756e-05,
+      "loss": 0.2416,
+      "step": 1489
+    },
+    {
+      "epoch": 2.400322190898107,
+      "grad_norm": 0.25762720542042816,
+      "learning_rate": 2.479472282747157e-05,
+      "loss": 0.3118,
+      "step": 1490
+    },
+    {
+      "epoch": 2.401933145388643,
+      "grad_norm": 0.24506926236957213,
+      "learning_rate": 2.4772856152376244e-05,
+      "loss": 0.2944,
+      "step": 1491
+    },
+    {
+      "epoch": 2.4035440998791784,
+      "grad_norm": 0.2640437205274289,
+      "learning_rate": 2.4750983425688945e-05,
+      "loss": 0.2895,
+      "step": 1492
+    },
+    {
+      "epoch": 2.405155054369714,
+      "grad_norm": 0.23704849670354022,
+      "learning_rate": 2.4729104675142496e-05,
+      "loss": 0.2877,
+      "step": 1493
+    },
+    {
+      "epoch": 2.4067660088602496,
+      "grad_norm": 0.24521000147844632,
+      "learning_rate": 2.4707219928477372e-05,
+      "loss": 0.2936,
+      "step": 1494
+    },
+    {
+      "epoch": 2.4083769633507854,
+      "grad_norm": 0.23091038574228756,
+      "learning_rate": 2.4685329213441645e-05,
+      "loss": 0.2655,
+      "step": 1495
+    },
+    {
+      "epoch": 2.409987917841321,
+      "grad_norm": 0.2466614443346807,
+      "learning_rate": 2.4663432557790955e-05,
+      "loss": 0.2762,
+      "step": 1496
+    },
+    {
+      "epoch": 2.4115988723318567,
+      "grad_norm": 0.24469535305837437,
+      "learning_rate": 2.464152998928848e-05,
+      "loss": 0.2855,
+      "step": 1497
+    },
+    {
+      "epoch": 2.413209826822392,
+      "grad_norm": 0.2505015124713008,
+      "learning_rate": 2.461962153570487e-05,
+      "loss": 0.3012,
+      "step": 1498
+    },
+    {
+      "epoch": 2.414820781312928,
+      "grad_norm": 0.23025616103553312,
+      "learning_rate": 2.459770722481827e-05,
+      "loss": 0.2608,
+      "step": 1499
+    },
+    {
+      "epoch": 2.4164317358034637,
+      "grad_norm": 0.23710616228759754,
+      "learning_rate": 2.4575787084414244e-05,
+      "loss": 0.2949,
+      "step": 1500
+    },
+    {
+      "epoch": 2.418042690293999,
+      "grad_norm": 0.26950446786419663,
+      "learning_rate": 2.4553861142285718e-05,
+      "loss": 0.2945,
+      "step": 1501
+    },
+    {
+      "epoch": 2.419653644784535,
+      "grad_norm": 0.22872827384203387,
+      "learning_rate": 2.4531929426233017e-05,
+      "loss": 0.2743,
+      "step": 1502
+    },
+    {
+      "epoch": 2.4212645992750703,
+      "grad_norm": 0.24607877318799967,
+      "learning_rate": 2.4509991964063762e-05,
+      "loss": 0.2788,
+      "step": 1503
+    },
+    {
+      "epoch": 2.422875553765606,
+      "grad_norm": 0.2392918287186858,
+      "learning_rate": 2.4488048783592864e-05,
+      "loss": 0.2588,
+      "step": 1504
+    },
+    {
+      "epoch": 2.424486508256142,
+      "grad_norm": 0.2546711994645602,
+      "learning_rate": 2.446609991264248e-05,
+      "loss": 0.2949,
+      "step": 1505
+    },
+    {
+      "epoch": 2.4260974627466774,
+      "grad_norm": 0.24785163146676142,
+      "learning_rate": 2.4444145379041987e-05,
+      "loss": 0.2906,
+      "step": 1506
+    },
+    {
+      "epoch": 2.427708417237213,
+      "grad_norm": 0.23900765965369322,
+      "learning_rate": 2.4422185210627943e-05,
+      "loss": 0.2707,
+      "step": 1507
+    },
+    {
+      "epoch": 2.4293193717277486,
+      "grad_norm": 0.2849435663798419,
+      "learning_rate": 2.4400219435244047e-05,
+      "loss": 0.3101,
+      "step": 1508
+    },
+    {
+      "epoch": 2.4309303262182844,
+      "grad_norm": 0.2335728091178472,
+      "learning_rate": 2.4378248080741123e-05,
+      "loss": 0.2767,
+      "step": 1509
+    },
+    {
+      "epoch": 2.43254128070882,
+      "grad_norm": 0.24500309892560315,
+      "learning_rate": 2.435627117497703e-05,
+      "loss": 0.2848,
+      "step": 1510
+    },
+    {
+      "epoch": 2.4341522351993556,
+      "grad_norm": 0.26504756906620003,
+      "learning_rate": 2.4334288745816714e-05,
+      "loss": 0.3254,
+      "step": 1511
+    },
+    {
+      "epoch": 2.435763189689891,
+      "grad_norm": 0.22953986456533704,
+      "learning_rate": 2.4312300821132087e-05,
+      "loss": 0.2533,
+      "step": 1512
+    },
+    {
+      "epoch": 2.437374144180427,
+      "grad_norm": 0.2685582896959977,
+      "learning_rate": 2.4290307428802047e-05,
+      "loss": 0.2936,
+      "step": 1513
+    },
+    {
+      "epoch": 2.4389850986709627,
+      "grad_norm": 0.257894081779025,
+      "learning_rate": 2.426830859671242e-05,
+      "loss": 0.2694,
+      "step": 1514
+    },
+    {
+      "epoch": 2.440596053161498,
+      "grad_norm": 0.27245685735959463,
+      "learning_rate": 2.4246304352755924e-05,
+      "loss": 0.2743,
+      "step": 1515
+    },
+    {
+      "epoch": 2.442207007652034,
+      "grad_norm": 0.2744009692046602,
+      "learning_rate": 2.4224294724832152e-05,
+      "loss": 0.2851,
+      "step": 1516
+    },
+    {
+      "epoch": 2.4438179621425693,
+      "grad_norm": 0.2923107453438156,
+      "learning_rate": 2.420227974084751e-05,
+      "loss": 0.2836,
+      "step": 1517
+    },
+    {
+      "epoch": 2.445428916633105,
+      "grad_norm": 0.259635977803087,
+      "learning_rate": 2.4180259428715203e-05,
+      "loss": 0.2757,
+      "step": 1518
+    },
+    {
+      "epoch": 2.447039871123641,
+      "grad_norm": 0.2792257618006009,
+      "learning_rate": 2.4158233816355185e-05,
+      "loss": 0.2601,
+      "step": 1519
+    },
+    {
+      "epoch": 2.4486508256141764,
+      "grad_norm": 0.2922956877345361,
+      "learning_rate": 2.413620293169415e-05,
+      "loss": 0.2859,
+      "step": 1520
+    },
+    {
+      "epoch": 2.450261780104712,
+      "grad_norm": 0.27963346219976065,
+      "learning_rate": 2.4114166802665437e-05,
+      "loss": 0.2903,
+      "step": 1521
+    },
+    {
+      "epoch": 2.4518727345952476,
+      "grad_norm": 0.26021125040145265,
+      "learning_rate": 2.409212545720908e-05,
+      "loss": 0.2755,
+      "step": 1522
+    },
+    {
+      "epoch": 2.4534836890857834,
+      "grad_norm": 0.32453764777695815,
+      "learning_rate": 2.4070078923271688e-05,
+      "loss": 0.3209,
+      "step": 1523
+    },
+    {
+      "epoch": 2.455094643576319,
+      "grad_norm": 0.25241900447923254,
+      "learning_rate": 2.404802722880649e-05,
+      "loss": 0.2571,
+      "step": 1524
+    },
+    {
+      "epoch": 2.4567055980668546,
+      "grad_norm": 0.28154790258027407,
+      "learning_rate": 2.4025970401773204e-05,
+      "loss": 0.2609,
+      "step": 1525
+    },
+    {
+      "epoch": 2.45831655255739,
+      "grad_norm": 0.2591231087696456,
+      "learning_rate": 2.4003908470138106e-05,
+      "loss": 0.2897,
+      "step": 1526
+    },
+    {
+      "epoch": 2.459927507047926,
+      "grad_norm": 0.285177493044488,
+      "learning_rate": 2.3981841461873927e-05,
+      "loss": 0.2781,
+      "step": 1527
+    },
+    {
+      "epoch": 2.4615384615384617,
+      "grad_norm": 0.28975036947156085,
+      "learning_rate": 2.3959769404959817e-05,
+      "loss": 0.28,
+      "step": 1528
+    },
+    {
+      "epoch": 2.463149416028997,
+      "grad_norm": 0.26691677035364914,
+      "learning_rate": 2.3937692327381356e-05,
+      "loss": 0.2775,
+      "step": 1529
+    },
+    {
+      "epoch": 2.464760370519533,
+      "grad_norm": 0.2681414848359182,
+      "learning_rate": 2.3915610257130464e-05,
+      "loss": 0.2622,
+      "step": 1530
+    },
+    {
+      "epoch": 2.4663713250100683,
+      "grad_norm": 0.24939280874639216,
+      "learning_rate": 2.3893523222205416e-05,
+      "loss": 0.2722,
+      "step": 1531
+    },
+    {
+      "epoch": 2.467982279500604,
+      "grad_norm": 0.2669734237482074,
+      "learning_rate": 2.3871431250610765e-05,
+      "loss": 0.2888,
+      "step": 1532
+    },
+    {
+      "epoch": 2.46959323399114,
+      "grad_norm": 0.24728738996882668,
+      "learning_rate": 2.3849334370357325e-05,
+      "loss": 0.2927,
+      "step": 1533
+    },
+    {
+      "epoch": 2.4712041884816753,
+      "grad_norm": 0.22346148184383519,
+      "learning_rate": 2.382723260946213e-05,
+      "loss": 0.2542,
+      "step": 1534
+    },
+    {
+      "epoch": 2.472815142972211,
+      "grad_norm": 0.2713126696973073,
+      "learning_rate": 2.3805125995948422e-05,
+      "loss": 0.2982,
+      "step": 1535
+    },
+    {
+      "epoch": 2.4744260974627466,
+      "grad_norm": 0.24113638235375398,
+      "learning_rate": 2.3783014557845573e-05,
+      "loss": 0.2818,
+      "step": 1536
+    },
+    {
+      "epoch": 2.4760370519532824,
+      "grad_norm": 0.2340220966184417,
+      "learning_rate": 2.376089832318909e-05,
+      "loss": 0.2654,
+      "step": 1537
+    },
+    {
+      "epoch": 2.4776480064438178,
+      "grad_norm": 0.2515922263908132,
+      "learning_rate": 2.3738777320020544e-05,
+      "loss": 0.2906,
+      "step": 1538
+    },
+    {
+      "epoch": 2.4792589609343536,
+      "grad_norm": 0.23936876164172788,
+      "learning_rate": 2.3716651576387562e-05,
+      "loss": 0.2705,
+      "step": 1539
+    },
+    {
+      "epoch": 2.4808699154248894,
+      "grad_norm": 0.23480838263802561,
+      "learning_rate": 2.369452112034379e-05,
+      "loss": 0.2845,
+      "step": 1540
+    },
+    {
+      "epoch": 2.482480869915425,
+      "grad_norm": 0.27839276582383377,
+      "learning_rate": 2.3672385979948825e-05,
+      "loss": 0.303,
+      "step": 1541
+    },
+    {
+      "epoch": 2.4840918244059607,
+      "grad_norm": 0.23665728826604437,
+      "learning_rate": 2.3650246183268238e-05,
+      "loss": 0.2787,
+      "step": 1542
+    },
+    {
+      "epoch": 2.485702778896496,
+      "grad_norm": 0.25851421490627297,
+      "learning_rate": 2.3628101758373464e-05,
+      "loss": 0.2929,
+      "step": 1543
+    },
+    {
+      "epoch": 2.487313733387032,
+      "grad_norm": 0.23517510701071315,
+      "learning_rate": 2.360595273334184e-05,
+      "loss": 0.2683,
+      "step": 1544
+    },
+    {
+      "epoch": 2.4889246878775673,
+      "grad_norm": 0.24605001495252662,
+      "learning_rate": 2.3583799136256505e-05,
+      "loss": 0.2844,
+      "step": 1545
+    },
+    {
+      "epoch": 2.490535642368103,
+      "grad_norm": 0.25239596472454423,
+      "learning_rate": 2.356164099520643e-05,
+      "loss": 0.2629,
+      "step": 1546
+    },
+    {
+      "epoch": 2.492146596858639,
+      "grad_norm": 0.2693185313862507,
+      "learning_rate": 2.3539478338286325e-05,
+      "loss": 0.2827,
+      "step": 1547
+    },
+    {
+      "epoch": 2.4937575513491743,
+      "grad_norm": 0.3241129729749029,
+      "learning_rate": 2.351731119359662e-05,
+      "loss": 0.2942,
+      "step": 1548
+    },
+    {
+      "epoch": 2.49536850583971,
+      "grad_norm": 0.2568486623179716,
+      "learning_rate": 2.3495139589243455e-05,
+      "loss": 0.2826,
+      "step": 1549
+    },
+    {
+      "epoch": 2.4969794603302455,
+      "grad_norm": 0.30686580882418024,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 0.2742,
+      "step": 1550
+    },
+    {
+      "epoch": 2.4985904148207814,
+      "grad_norm": 0.2576518847992097,
+      "learning_rate": 2.3450783113999487e-05,
+      "loss": 0.28,
+      "step": 1551
+    },
+    {
+      "epoch": 2.500201369311317,
+      "grad_norm": 0.3350355156604317,
+      "learning_rate": 2.3428598299349076e-05,
+      "loss": 0.2764,
+      "step": 1552
+    },
+    {
+      "epoch": 2.5018123238018526,
+      "grad_norm": 0.2557894175372604,
+      "learning_rate": 2.3406409137515912e-05,
+      "loss": 0.2601,
+      "step": 1553
+    },
+    {
+      "epoch": 2.503423278292388,
+      "grad_norm": 0.26635573164209964,
+      "learning_rate": 2.338421565663403e-05,
+      "loss": 0.287,
+      "step": 1554
+    },
+    {
+      "epoch": 2.505034232782924,
+      "grad_norm": 0.28770740938014006,
+      "learning_rate": 2.3362017884842967e-05,
+      "loss": 0.2796,
+      "step": 1555
+    },
+    {
+      "epoch": 2.5066451872734596,
+      "grad_norm": 0.25946436577333,
+      "learning_rate": 2.3339815850287676e-05,
+      "loss": 0.2753,
+      "step": 1556
+    },
+    {
+      "epoch": 2.508256141763995,
+      "grad_norm": 0.2588519714140495,
+      "learning_rate": 2.3317609581118527e-05,
+      "loss": 0.2652,
+      "step": 1557
+    },
+    {
+      "epoch": 2.509867096254531,
+      "grad_norm": 0.2690897104911625,
+      "learning_rate": 2.3295399105491256e-05,
+      "loss": 0.2791,
+      "step": 1558
+    },
+    {
+      "epoch": 2.5114780507450662,
+      "grad_norm": 0.2520218508663738,
+      "learning_rate": 2.3273184451566934e-05,
+      "loss": 0.2593,
+      "step": 1559
+    },
+    {
+      "epoch": 2.513089005235602,
+      "grad_norm": 0.2621023752863465,
+      "learning_rate": 2.325096564751193e-05,
+      "loss": 0.2736,
+      "step": 1560
+    },
+    {
+      "epoch": 2.514699959726138,
+      "grad_norm": 0.22160638366693367,
+      "learning_rate": 2.322874272149787e-05,
+      "loss": 0.2715,
+      "step": 1561
+    },
+    {
+      "epoch": 2.5163109142166733,
+      "grad_norm": 0.28383611341277853,
+      "learning_rate": 2.3206515701701612e-05,
+      "loss": 0.3236,
+      "step": 1562
+    },
+    {
+      "epoch": 2.517921868707209,
+      "grad_norm": 0.23283625465888177,
+      "learning_rate": 2.3184284616305205e-05,
+      "loss": 0.2733,
+      "step": 1563
+    },
+    {
+      "epoch": 2.5195328231977445,
+      "grad_norm": 0.24394026083888837,
+      "learning_rate": 2.316204949349585e-05,
+      "loss": 0.2741,
+      "step": 1564
+    },
+    {
+      "epoch": 2.5211437776882804,
+      "grad_norm": 0.230477810923903,
+      "learning_rate": 2.3139810361465854e-05,
+      "loss": 0.2541,
+      "step": 1565
+    },
+    {
+      "epoch": 2.522754732178816,
+      "grad_norm": 0.2515917527239852,
+      "learning_rate": 2.311756724841265e-05,
+      "loss": 0.2867,
+      "step": 1566
+    },
+    {
+      "epoch": 2.5243656866693516,
+      "grad_norm": 0.24050394233763317,
+      "learning_rate": 2.3095320182538657e-05,
+      "loss": 0.2944,
+      "step": 1567
+    },
+    {
+      "epoch": 2.525976641159887,
+      "grad_norm": 0.24985185288588335,
+      "learning_rate": 2.3073069192051364e-05,
+      "loss": 0.2617,
+      "step": 1568
+    },
+    {
+      "epoch": 2.527587595650423,
+      "grad_norm": 0.24018241301765708,
+      "learning_rate": 2.305081430516319e-05,
+      "loss": 0.2595,
+      "step": 1569
+    },
+    {
+      "epoch": 2.5291985501409586,
+      "grad_norm": 0.25522923876634196,
+      "learning_rate": 2.3028555550091536e-05,
+      "loss": 0.2948,
+      "step": 1570
+    },
+    {
+      "epoch": 2.530809504631494,
+      "grad_norm": 0.2517439540443441,
+      "learning_rate": 2.300629295505867e-05,
+      "loss": 0.2434,
+      "step": 1571
+    },
+    {
+      "epoch": 2.53242045912203,
+      "grad_norm": 0.23925236470373543,
+      "learning_rate": 2.2984026548291752e-05,
+      "loss": 0.2815,
+      "step": 1572
+    },
+    {
+      "epoch": 2.5340314136125652,
+      "grad_norm": 0.2870555456129622,
+      "learning_rate": 2.2961756358022765e-05,
+      "loss": 0.313,
+      "step": 1573
+    },
+    {
+      "epoch": 2.535642368103101,
+      "grad_norm": 0.24767206562519722,
+      "learning_rate": 2.2939482412488498e-05,
+      "loss": 0.2575,
+      "step": 1574
+    },
+    {
+      "epoch": 2.537253322593637,
+      "grad_norm": 0.2722004878155017,
+      "learning_rate": 2.291720473993049e-05,
+      "loss": 0.2865,
+      "step": 1575
+    },
+    {
+      "epoch": 2.5388642770841723,
+      "grad_norm": 0.27997751844902186,
+      "learning_rate": 2.289492336859501e-05,
+      "loss": 0.2742,
+      "step": 1576
+    },
+    {
+      "epoch": 2.540475231574708,
+      "grad_norm": 0.3147302217916788,
+      "learning_rate": 2.2872638326733018e-05,
+      "loss": 0.2786,
+      "step": 1577
+    },
+    {
+      "epoch": 2.5420861860652435,
+      "grad_norm": 0.24157360086493543,
+      "learning_rate": 2.2850349642600137e-05,
+      "loss": 0.2671,
+      "step": 1578
+    },
+    {
+      "epoch": 2.5436971405557793,
+      "grad_norm": 0.2931844117704698,
+      "learning_rate": 2.282805734445659e-05,
+      "loss": 0.2828,
+      "step": 1579
+    },
+    {
+      "epoch": 2.545308095046315,
+      "grad_norm": 0.27831191484940104,
+      "learning_rate": 2.2805761460567197e-05,
+      "loss": 0.2876,
+      "step": 1580
+    },
+    {
+      "epoch": 2.5469190495368506,
+      "grad_norm": 0.27135097897303195,
+      "learning_rate": 2.278346201920131e-05,
+      "loss": 0.3,
+      "step": 1581
+    },
+    {
+      "epoch": 2.5485300040273864,
+      "grad_norm": 0.24836372557953207,
+      "learning_rate": 2.2761159048632813e-05,
+      "loss": 0.2653,
+      "step": 1582
+    },
+    {
+      "epoch": 2.5501409585179218,
+      "grad_norm": 0.28745989489603807,
+      "learning_rate": 2.273885257714004e-05,
+      "loss": 0.263,
+      "step": 1583
+    },
+    {
+      "epoch": 2.5517519130084576,
+      "grad_norm": 0.2616915357039271,
+      "learning_rate": 2.2716542633005777e-05,
+      "loss": 0.2864,
+      "step": 1584
+    },
+    {
+      "epoch": 2.553362867498993,
+      "grad_norm": 0.2804002104585373,
+      "learning_rate": 2.2694229244517226e-05,
+      "loss": 0.271,
+      "step": 1585
+    },
+    {
+      "epoch": 2.554973821989529,
+      "grad_norm": 0.30579400972491694,
+      "learning_rate": 2.2671912439965923e-05,
+      "loss": 0.2869,
+      "step": 1586
+    },
+    {
+      "epoch": 2.556584776480064,
+      "grad_norm": 0.2577374199945713,
+      "learning_rate": 2.264959224764777e-05,
+      "loss": 0.2946,
+      "step": 1587
+    },
+    {
+      "epoch": 2.5581957309706,
+      "grad_norm": 0.2970356900027897,
+      "learning_rate": 2.262726869586293e-05,
+      "loss": 0.282,
+      "step": 1588
+    },
+    {
+      "epoch": 2.559806685461136,
+      "grad_norm": 0.2555540175655143,
+      "learning_rate": 2.260494181291587e-05,
+      "loss": 0.2948,
+      "step": 1589
+    },
+    {
+      "epoch": 2.5614176399516713,
+      "grad_norm": 0.2410510585798002,
+      "learning_rate": 2.258261162711523e-05,
+      "loss": 0.2694,
+      "step": 1590
+    },
+    {
+      "epoch": 2.563028594442207,
+      "grad_norm": 0.29264739126391853,
+      "learning_rate": 2.256027816677388e-05,
+      "loss": 0.2778,
+      "step": 1591
+    },
+    {
+      "epoch": 2.5646395489327425,
+      "grad_norm": 0.2495793628483905,
+      "learning_rate": 2.2537941460208818e-05,
+      "loss": 0.2792,
+      "step": 1592
+    },
+    {
+      "epoch": 2.5662505034232783,
+      "grad_norm": 0.25120155950587036,
+      "learning_rate": 2.2515601535741168e-05,
+      "loss": 0.2727,
+      "step": 1593
+    },
+    {
+      "epoch": 2.567861457913814,
+      "grad_norm": 0.2895173143422783,
+      "learning_rate": 2.2493258421696124e-05,
+      "loss": 0.268,
+      "step": 1594
+    },
+    {
+      "epoch": 2.5694724124043495,
+      "grad_norm": 0.235477824563421,
+      "learning_rate": 2.2470912146402935e-05,
+      "loss": 0.2559,
+      "step": 1595
+    },
+    {
+      "epoch": 2.5710833668948854,
+      "grad_norm": 0.23324303749320113,
+      "learning_rate": 2.244856273819485e-05,
+      "loss": 0.276,
+      "step": 1596
+    },
+    {
+      "epoch": 2.5726943213854208,
+      "grad_norm": 0.2874685117002799,
+      "learning_rate": 2.24262102254091e-05,
+      "loss": 0.3234,
+      "step": 1597
+    },
+    {
+      "epoch": 2.5743052758759566,
+      "grad_norm": 0.22530125349976493,
+      "learning_rate": 2.2403854636386843e-05,
+      "loss": 0.262,
+      "step": 1598
+    },
+    {
+      "epoch": 2.5759162303664924,
+      "grad_norm": 0.25364847023690723,
+      "learning_rate": 2.238149599947314e-05,
+      "loss": 0.2724,
+      "step": 1599
+    },
+    {
+      "epoch": 2.577527184857028,
+      "grad_norm": 0.25154728625547484,
+      "learning_rate": 2.2359134343016926e-05,
+      "loss": 0.2899,
+      "step": 1600
+    },
+    {
+      "epoch": 2.579138139347563,
+      "grad_norm": 0.2631559112148431,
+      "learning_rate": 2.233676969537094e-05,
+      "loss": 0.2735,
+      "step": 1601
+    },
+    {
+      "epoch": 2.580749093838099,
+      "grad_norm": 0.26024689995575573,
+      "learning_rate": 2.2314402084891746e-05,
+      "loss": 0.2723,
+      "step": 1602
+    },
+    {
+      "epoch": 2.582360048328635,
+      "grad_norm": 0.2432406812780365,
+      "learning_rate": 2.2292031539939635e-05,
+      "loss": 0.2651,
+      "step": 1603
+    },
+    {
+      "epoch": 2.5839710028191702,
+      "grad_norm": 0.26945019384770447,
+      "learning_rate": 2.2269658088878638e-05,
+      "loss": 0.2906,
+      "step": 1604
+    },
+    {
+      "epoch": 2.585581957309706,
+      "grad_norm": 0.25788545777611716,
+      "learning_rate": 2.2247281760076468e-05,
+      "loss": 0.2883,
+      "step": 1605
+    },
+    {
+      "epoch": 2.5871929118002415,
+      "grad_norm": 0.2732432882264997,
+      "learning_rate": 2.2224902581904476e-05,
+      "loss": 0.268,
+      "step": 1606
+    },
+    {
+      "epoch": 2.5888038662907773,
+      "grad_norm": 0.25613794655168626,
+      "learning_rate": 2.2202520582737635e-05,
+      "loss": 0.2596,
+      "step": 1607
+    },
+    {
+      "epoch": 2.590414820781313,
+      "grad_norm": 0.3055836045703691,
+      "learning_rate": 2.2180135790954494e-05,
+      "loss": 0.2931,
+      "step": 1608
+    },
+    {
+      "epoch": 2.5920257752718485,
+      "grad_norm": 0.23324193325162323,
+      "learning_rate": 2.215774823493715e-05,
+      "loss": 0.2667,
+      "step": 1609
+    },
+    {
+      "epoch": 2.5936367297623844,
+      "grad_norm": 0.24811631478589213,
+      "learning_rate": 2.213535794307118e-05,
+      "loss": 0.2757,
+      "step": 1610
+    },
+    {
+      "epoch": 2.5952476842529197,
+      "grad_norm": 0.2758092631908418,
+      "learning_rate": 2.211296494374566e-05,
+      "loss": 0.2924,
+      "step": 1611
+    },
+    {
+      "epoch": 2.5968586387434556,
+      "grad_norm": 0.2643433716278074,
+      "learning_rate": 2.209056926535307e-05,
+      "loss": 0.3036,
+      "step": 1612
+    },
+    {
+      "epoch": 2.5984695932339914,
+      "grad_norm": 0.2685767270547775,
+      "learning_rate": 2.2068170936289323e-05,
+      "loss": 0.2643,
+      "step": 1613
+    },
+    {
+      "epoch": 2.600080547724527,
+      "grad_norm": 0.24441427558762532,
+      "learning_rate": 2.2045769984953652e-05,
+      "loss": 0.2685,
+      "step": 1614
+    },
+    {
+      "epoch": 2.601691502215062,
+      "grad_norm": 0.29930220047366574,
+      "learning_rate": 2.2023366439748647e-05,
+      "loss": 0.2746,
+      "step": 1615
+    },
+    {
+      "epoch": 2.603302456705598,
+      "grad_norm": 0.22953726566232438,
+      "learning_rate": 2.2000960329080166e-05,
+      "loss": 0.2864,
+      "step": 1616
+    },
+    {
+      "epoch": 2.604913411196134,
+      "grad_norm": 0.25919477109412487,
+      "learning_rate": 2.197855168135734e-05,
+      "loss": 0.2724,
+      "step": 1617
+    },
+    {
+      "epoch": 2.6065243656866692,
+      "grad_norm": 0.24583444723040715,
+      "learning_rate": 2.1956140524992495e-05,
+      "loss": 0.2721,
+      "step": 1618
+    },
+    {
+      "epoch": 2.608135320177205,
+      "grad_norm": 0.2488426900524242,
+      "learning_rate": 2.1933726888401146e-05,
+      "loss": 0.3036,
+      "step": 1619
+    },
+    {
+      "epoch": 2.6097462746677405,
+      "grad_norm": 0.2709750897529024,
+      "learning_rate": 2.1911310800001967e-05,
+      "loss": 0.2761,
+      "step": 1620
+    },
+    {
+      "epoch": 2.6113572291582763,
+      "grad_norm": 0.27314838210070563,
+      "learning_rate": 2.188889228821671e-05,
+      "loss": 0.2854,
+      "step": 1621
+    },
+    {
+      "epoch": 2.612968183648812,
+      "grad_norm": 0.24503356817611527,
+      "learning_rate": 2.186647138147024e-05,
+      "loss": 0.2713,
+      "step": 1622
+    },
+    {
+      "epoch": 2.6145791381393475,
+      "grad_norm": 0.23765691589614468,
+      "learning_rate": 2.184404810819041e-05,
+      "loss": 0.2682,
+      "step": 1623
+    },
+    {
+      "epoch": 2.6161900926298833,
+      "grad_norm": 0.25031315550528577,
+      "learning_rate": 2.182162249680813e-05,
+      "loss": 0.2755,
+      "step": 1624
+    },
+    {
+      "epoch": 2.6178010471204187,
+      "grad_norm": 0.24667393832777065,
+      "learning_rate": 2.179919457575722e-05,
+      "loss": 0.2879,
+      "step": 1625
+    },
+    {
+      "epoch": 2.6194120016109546,
+      "grad_norm": 0.22980591927089686,
+      "learning_rate": 2.1776764373474465e-05,
+      "loss": 0.267,
+      "step": 1626
+    },
+    {
+      "epoch": 2.6210229561014904,
+      "grad_norm": 0.24144423356769742,
+      "learning_rate": 2.1754331918399526e-05,
+      "loss": 0.2771,
+      "step": 1627
+    },
+    {
+      "epoch": 2.6226339105920258,
+      "grad_norm": 0.2665613251605162,
+      "learning_rate": 2.1731897238974926e-05,
+      "loss": 0.3079,
+      "step": 1628
+    },
+    {
+      "epoch": 2.624244865082561,
+      "grad_norm": 0.22659947309530976,
+      "learning_rate": 2.170946036364601e-05,
+      "loss": 0.2617,
+      "step": 1629
+    },
+    {
+      "epoch": 2.625855819573097,
+      "grad_norm": 0.2773783867044038,
+      "learning_rate": 2.1687021320860893e-05,
+      "loss": 0.2679,
+      "step": 1630
+    },
+    {
+      "epoch": 2.627466774063633,
+      "grad_norm": 0.25665405302808597,
+      "learning_rate": 2.166458013907047e-05,
+      "loss": 0.295,
+      "step": 1631
+    },
+    {
+      "epoch": 2.629077728554168,
+      "grad_norm": 0.2625122154554604,
+      "learning_rate": 2.1642136846728313e-05,
+      "loss": 0.2548,
+      "step": 1632
+    },
+    {
+      "epoch": 2.630688683044704,
+      "grad_norm": 0.24524445937268322,
+      "learning_rate": 2.1619691472290692e-05,
+      "loss": 0.278,
+      "step": 1633
+    },
+    {
+      "epoch": 2.6322996375352394,
+      "grad_norm": 0.24739471796623,
+      "learning_rate": 2.159724404421649e-05,
+      "loss": 0.2447,
+      "step": 1634
+    },
+    {
+      "epoch": 2.6339105920257753,
+      "grad_norm": 0.24322288209895856,
+      "learning_rate": 2.157479459096724e-05,
+      "loss": 0.2823,
+      "step": 1635
+    },
+    {
+      "epoch": 2.635521546516311,
+      "grad_norm": 0.22906031471382138,
+      "learning_rate": 2.1552343141007e-05,
+      "loss": 0.2674,
+      "step": 1636
+    },
+    {
+      "epoch": 2.6371325010068465,
+      "grad_norm": 0.26408427275553314,
+      "learning_rate": 2.1529889722802384e-05,
+      "loss": 0.3259,
+      "step": 1637
+    },
+    {
+      "epoch": 2.6387434554973823,
+      "grad_norm": 0.20577040124306828,
+      "learning_rate": 2.1507434364822487e-05,
+      "loss": 0.2442,
+      "step": 1638
+    },
+    {
+      "epoch": 2.6403544099879177,
+      "grad_norm": 0.25336351837263493,
+      "learning_rate": 2.148497709553887e-05,
+      "loss": 0.2966,
+      "step": 1639
+    },
+    {
+      "epoch": 2.6419653644784535,
+      "grad_norm": 0.2179088401032362,
+      "learning_rate": 2.1462517943425523e-05,
+      "loss": 0.2399,
+      "step": 1640
+    },
+    {
+      "epoch": 2.6435763189689894,
+      "grad_norm": 0.2609474199496771,
+      "learning_rate": 2.1440056936958815e-05,
+      "loss": 0.2897,
+      "step": 1641
+    },
+    {
+      "epoch": 2.6451872734595248,
+      "grad_norm": 0.250702552124162,
+      "learning_rate": 2.141759410461746e-05,
+      "loss": 0.2797,
+      "step": 1642
+    },
+    {
+      "epoch": 2.64679822795006,
+      "grad_norm": 0.25202435733045636,
+      "learning_rate": 2.1395129474882507e-05,
+      "loss": 0.258,
+      "step": 1643
+    },
+    {
+      "epoch": 2.648409182440596,
+      "grad_norm": 0.23559656524296946,
+      "learning_rate": 2.1372663076237273e-05,
+      "loss": 0.3022,
+      "step": 1644
+    },
+    {
+      "epoch": 2.650020136931132,
+      "grad_norm": 0.2600623544743553,
+      "learning_rate": 2.1350194937167307e-05,
+      "loss": 0.284,
+      "step": 1645
+    },
+    {
+      "epoch": 2.651631091421667,
+      "grad_norm": 0.23042665562467768,
+      "learning_rate": 2.1327725086160385e-05,
+      "loss": 0.2659,
+      "step": 1646
+    },
+    {
+      "epoch": 2.653242045912203,
+      "grad_norm": 0.2319917641049967,
+      "learning_rate": 2.1305253551706442e-05,
+      "loss": 0.2755,
+      "step": 1647
+    },
+    {
+      "epoch": 2.6548530004027384,
+      "grad_norm": 0.24708094663257485,
+      "learning_rate": 2.1282780362297544e-05,
+      "loss": 0.2895,
+      "step": 1648
+    },
+    {
+      "epoch": 2.6564639548932742,
+      "grad_norm": 0.2463749021745463,
+      "learning_rate": 2.1260305546427867e-05,
+      "loss": 0.288,
+      "step": 1649
+    },
+    {
+      "epoch": 2.65807490938381,
+      "grad_norm": 0.2540138947859815,
+      "learning_rate": 2.123782913259364e-05,
+      "loss": 0.2826,
+      "step": 1650
+    },
+    {
+      "epoch": 2.6596858638743455,
+      "grad_norm": 0.23487737142224063,
+      "learning_rate": 2.121535114929312e-05,
+      "loss": 0.272,
+      "step": 1651
+    },
+    {
+      "epoch": 2.6612968183648813,
+      "grad_norm": 0.2282373426932882,
+      "learning_rate": 2.1192871625026553e-05,
+      "loss": 0.2343,
+      "step": 1652
+    },
+    {
+      "epoch": 2.6629077728554167,
+      "grad_norm": 0.2591476141147764,
+      "learning_rate": 2.1170390588296148e-05,
+      "loss": 0.3235,
+      "step": 1653
+    },
+    {
+      "epoch": 2.6645187273459525,
+      "grad_norm": 0.2636302573560719,
+      "learning_rate": 2.1147908067606012e-05,
+      "loss": 0.2785,
+      "step": 1654
+    },
+    {
+      "epoch": 2.6661296818364884,
+      "grad_norm": 0.23618006055822752,
+      "learning_rate": 2.112542409146217e-05,
+      "loss": 0.2655,
+      "step": 1655
+    },
+    {
+      "epoch": 2.6677406363270237,
+      "grad_norm": 0.25207523326650344,
+      "learning_rate": 2.1102938688372436e-05,
+      "loss": 0.2836,
+      "step": 1656
+    },
+    {
+      "epoch": 2.669351590817559,
+      "grad_norm": 0.25173559248973704,
+      "learning_rate": 2.1080451886846486e-05,
+      "loss": 0.261,
+      "step": 1657
+    },
+    {
+      "epoch": 2.670962545308095,
+      "grad_norm": 0.2431590866242753,
+      "learning_rate": 2.1057963715395746e-05,
+      "loss": 0.3086,
+      "step": 1658
+    },
+    {
+      "epoch": 2.672573499798631,
+      "grad_norm": 0.25202253459066404,
+      "learning_rate": 2.1035474202533385e-05,
+      "loss": 0.2567,
+      "step": 1659
+    },
+    {
+      "epoch": 2.674184454289166,
+      "grad_norm": 0.2555417920210164,
+      "learning_rate": 2.1012983376774255e-05,
+      "loss": 0.2739,
+      "step": 1660
+    },
+    {
+      "epoch": 2.675795408779702,
+      "grad_norm": 0.2420625615147683,
+      "learning_rate": 2.0990491266634903e-05,
+      "loss": 0.2653,
+      "step": 1661
+    },
+    {
+      "epoch": 2.6774063632702374,
+      "grad_norm": 0.2396489356305853,
+      "learning_rate": 2.0967997900633482e-05,
+      "loss": 0.3129,
+      "step": 1662
+    },
+    {
+      "epoch": 2.6790173177607732,
+      "grad_norm": 0.2131386533167181,
+      "learning_rate": 2.094550330728974e-05,
+      "loss": 0.2642,
+      "step": 1663
+    },
+    {
+      "epoch": 2.680628272251309,
+      "grad_norm": 0.21745561076267786,
+      "learning_rate": 2.092300751512499e-05,
+      "loss": 0.2642,
+      "step": 1664
+    },
+    {
+      "epoch": 2.6822392267418445,
+      "grad_norm": 0.2395086233145289,
+      "learning_rate": 2.0900510552662057e-05,
+      "loss": 0.2924,
+      "step": 1665
+    },
+    {
+      "epoch": 2.6838501812323803,
+      "grad_norm": 0.2353999626748506,
+      "learning_rate": 2.0878012448425258e-05,
+      "loss": 0.2591,
+      "step": 1666
+    },
+    {
+      "epoch": 2.6854611357229157,
+      "grad_norm": 0.24095219112832106,
+      "learning_rate": 2.085551323094035e-05,
+      "loss": 0.2962,
+      "step": 1667
+    },
+    {
+      "epoch": 2.6870720902134515,
+      "grad_norm": 0.2553288916093136,
+      "learning_rate": 2.08330129287345e-05,
+      "loss": 0.3125,
+      "step": 1668
+    },
+    {
+      "epoch": 2.6886830447039873,
+      "grad_norm": 0.24345181609728936,
+      "learning_rate": 2.0810511570336262e-05,
+      "loss": 0.2681,
+      "step": 1669
+    },
+    {
+      "epoch": 2.6902939991945227,
+      "grad_norm": 0.256453338841592,
+      "learning_rate": 2.0788009184275514e-05,
+      "loss": 0.2694,
+      "step": 1670
+    },
+    {
+      "epoch": 2.6919049536850586,
+      "grad_norm": 0.24077026988656322,
+      "learning_rate": 2.0765505799083452e-05,
+      "loss": 0.2741,
+      "step": 1671
+    },
+    {
+      "epoch": 2.693515908175594,
+      "grad_norm": 0.2534029520291385,
+      "learning_rate": 2.074300144329252e-05,
+      "loss": 0.3124,
+      "step": 1672
+    },
+    {
+      "epoch": 2.6951268626661298,
+      "grad_norm": 0.24039030268215603,
+      "learning_rate": 2.0720496145436423e-05,
+      "loss": 0.2761,
+      "step": 1673
+    },
+    {
+      "epoch": 2.696737817156665,
+      "grad_norm": 0.2441294858776203,
+      "learning_rate": 2.0697989934050025e-05,
+      "loss": 0.2702,
+      "step": 1674
+    },
+    {
+      "epoch": 2.698348771647201,
+      "grad_norm": 0.23858821941312688,
+      "learning_rate": 2.0675482837669367e-05,
+      "loss": 0.2585,
+      "step": 1675
+    },
+    {
+      "epoch": 2.6999597261377364,
+      "grad_norm": 0.22912907435783178,
+      "learning_rate": 2.0652974884831612e-05,
+      "loss": 0.2735,
+      "step": 1676
+    },
+    {
+      "epoch": 2.701570680628272,
+      "grad_norm": 0.25427589376548887,
+      "learning_rate": 2.063046610407501e-05,
+      "loss": 0.3006,
+      "step": 1677
+    },
+    {
+      "epoch": 2.703181635118808,
+      "grad_norm": 0.25451643357504333,
+      "learning_rate": 2.060795652393886e-05,
+      "loss": 0.2771,
+      "step": 1678
+    },
+    {
+      "epoch": 2.7047925896093434,
+      "grad_norm": 0.2470762873381859,
+      "learning_rate": 2.0585446172963457e-05,
+      "loss": 0.2967,
+      "step": 1679
+    },
+    {
+      "epoch": 2.7064035440998793,
+      "grad_norm": 0.2405780999475265,
+      "learning_rate": 2.05629350796901e-05,
+      "loss": 0.2728,
+      "step": 1680
+    },
+    {
+      "epoch": 2.7080144985904147,
+      "grad_norm": 0.2680658731906045,
+      "learning_rate": 2.0540423272661024e-05,
+      "loss": 0.2846,
+      "step": 1681
+    },
+    {
+      "epoch": 2.7096254530809505,
+      "grad_norm": 0.22930174904587397,
+      "learning_rate": 2.0517910780419355e-05,
+      "loss": 0.279,
+      "step": 1682
+    },
+    {
+      "epoch": 2.7112364075714863,
+      "grad_norm": 0.2513745451496423,
+      "learning_rate": 2.0495397631509092e-05,
+      "loss": 0.2498,
+      "step": 1683
+    },
+    {
+      "epoch": 2.7128473620620217,
+      "grad_norm": 0.2607357243459458,
+      "learning_rate": 2.047288385447507e-05,
+      "loss": 0.29,
+      "step": 1684
+    },
+    {
+      "epoch": 2.7144583165525575,
+      "grad_norm": 0.2680985429045293,
+      "learning_rate": 2.0450369477862922e-05,
+      "loss": 0.2788,
+      "step": 1685
+    },
+    {
+      "epoch": 2.716069271043093,
+      "grad_norm": 0.28257542550577647,
+      "learning_rate": 2.042785453021905e-05,
+      "loss": 0.2705,
+      "step": 1686
+    },
+    {
+      "epoch": 2.7176802255336288,
+      "grad_norm": 0.24906888756996964,
+      "learning_rate": 2.0405339040090557e-05,
+      "loss": 0.2845,
+      "step": 1687
+    },
+    {
+      "epoch": 2.7192911800241646,
+      "grad_norm": 0.23828748868464236,
+      "learning_rate": 2.0382823036025243e-05,
+      "loss": 0.2871,
+      "step": 1688
+    },
+    {
+      "epoch": 2.7209021345147,
+      "grad_norm": 0.2522899452359346,
+      "learning_rate": 2.0360306546571582e-05,
+      "loss": 0.2743,
+      "step": 1689
+    },
+    {
+      "epoch": 2.7225130890052354,
+      "grad_norm": 0.24592334075728844,
+      "learning_rate": 2.0337789600278623e-05,
+      "loss": 0.27,
+      "step": 1690
+    },
+    {
+      "epoch": 2.724124043495771,
+      "grad_norm": 0.2554007578885009,
+      "learning_rate": 2.0315272225696034e-05,
+      "loss": 0.3087,
+      "step": 1691
+    },
+    {
+      "epoch": 2.725734997986307,
+      "grad_norm": 0.23843192140396555,
+      "learning_rate": 2.0292754451373992e-05,
+      "loss": 0.2685,
+      "step": 1692
+    },
+    {
+      "epoch": 2.7273459524768424,
+      "grad_norm": 0.22520952757862236,
+      "learning_rate": 2.027023630586321e-05,
+      "loss": 0.261,
+      "step": 1693
+    },
+    {
+      "epoch": 2.7289569069673782,
+      "grad_norm": 0.26311834190007133,
+      "learning_rate": 2.024771781771485e-05,
+      "loss": 0.2881,
+      "step": 1694
+    },
+    {
+      "epoch": 2.7305678614579136,
+      "grad_norm": 0.20959220644219456,
+      "learning_rate": 2.0225199015480518e-05,
+      "loss": 0.2437,
+      "step": 1695
+    },
+    {
+      "epoch": 2.7321788159484495,
+      "grad_norm": 0.26113815468998086,
+      "learning_rate": 2.0202679927712224e-05,
+      "loss": 0.2894,
+      "step": 1696
+    },
+    {
+      "epoch": 2.7337897704389853,
+      "grad_norm": 0.2469561663058487,
+      "learning_rate": 2.018016058296232e-05,
+      "loss": 0.2702,
+      "step": 1697
+    },
+    {
+      "epoch": 2.7354007249295207,
+      "grad_norm": 0.2641432503130729,
+      "learning_rate": 2.0157641009783512e-05,
+      "loss": 0.297,
+      "step": 1698
+    },
+    {
+      "epoch": 2.7370116794200565,
+      "grad_norm": 0.23840297914176908,
+      "learning_rate": 2.0135121236728762e-05,
+      "loss": 0.2674,
+      "step": 1699
+    },
+    {
+      "epoch": 2.738622633910592,
+      "grad_norm": 0.2764405943973875,
+      "learning_rate": 2.0112601292351322e-05,
+      "loss": 0.2947,
+      "step": 1700
+    },
+    {
+      "epoch": 2.7402335884011277,
+      "grad_norm": 0.23575772294949454,
+      "learning_rate": 2.009008120520463e-05,
+      "loss": 0.2672,
+      "step": 1701
+    },
+    {
+      "epoch": 2.7418445428916636,
+      "grad_norm": 0.2521615855835443,
+      "learning_rate": 2.006756100384233e-05,
+      "loss": 0.2814,
+      "step": 1702
+    },
+    {
+      "epoch": 2.743455497382199,
+      "grad_norm": 0.2538001655861593,
+      "learning_rate": 2.0045040716818184e-05,
+      "loss": 0.2802,
+      "step": 1703
+    },
+    {
+      "epoch": 2.7450664518727343,
+      "grad_norm": 0.2281414090679116,
+      "learning_rate": 2.0022520372686092e-05,
+      "loss": 0.2723,
+      "step": 1704
+    },
+    {
+      "epoch": 2.74667740636327,
+      "grad_norm": 0.27249255745775103,
+      "learning_rate": 2e-05,
+      "loss": 0.2783,
+      "step": 1705
+    },
+    {
+      "epoch": 2.748288360853806,
+      "grad_norm": 0.21913760224877482,
+      "learning_rate": 1.9977479627313918e-05,
+      "loss": 0.2666,
+      "step": 1706
+    },
+    {
+      "epoch": 2.7498993153443414,
+      "grad_norm": 0.271276451382289,
+      "learning_rate": 1.995495928318182e-05,
+      "loss": 0.2592,
+      "step": 1707
+    },
+    {
+      "epoch": 2.7515102698348772,
+      "grad_norm": 0.25181265659436014,
+      "learning_rate": 1.9932438996157678e-05,
+      "loss": 0.2672,
+      "step": 1708
+    },
+    {
+      "epoch": 2.7531212243254126,
+      "grad_norm": 0.23000287341824674,
+      "learning_rate": 1.9909918794795378e-05,
+      "loss": 0.2883,
+      "step": 1709
+    },
+    {
+      "epoch": 2.7547321788159485,
+      "grad_norm": 0.2509709521664799,
+      "learning_rate": 1.988739870764869e-05,
+      "loss": 0.2955,
+      "step": 1710
+    },
+    {
+      "epoch": 2.7563431333064843,
+      "grad_norm": 0.255309204845929,
+      "learning_rate": 1.986487876327124e-05,
+      "loss": 0.2939,
+      "step": 1711
+    },
+    {
+      "epoch": 2.7579540877970197,
+      "grad_norm": 0.2389408162743876,
+      "learning_rate": 1.9842358990216498e-05,
+      "loss": 0.276,
+      "step": 1712
+    },
+    {
+      "epoch": 2.7595650422875555,
+      "grad_norm": 0.22548655871923712,
+      "learning_rate": 1.9819839417037688e-05,
+      "loss": 0.2551,
+      "step": 1713
+    },
+    {
+      "epoch": 2.761175996778091,
+      "grad_norm": 0.2514072456845176,
+      "learning_rate": 1.9797320072287786e-05,
+      "loss": 0.2797,
+      "step": 1714
+    },
+    {
+      "epoch": 2.7627869512686267,
+      "grad_norm": 0.2389348677377125,
+      "learning_rate": 1.9774800984519485e-05,
+      "loss": 0.2538,
+      "step": 1715
+    },
+    {
+      "epoch": 2.7643979057591626,
+      "grad_norm": 0.24399474291305062,
+      "learning_rate": 1.9752282182285158e-05,
+      "loss": 0.2698,
+      "step": 1716
+    },
+    {
+      "epoch": 2.766008860249698,
+      "grad_norm": 0.24492996641517525,
+      "learning_rate": 1.9729763694136796e-05,
+      "loss": 0.2571,
+      "step": 1717
+    },
+    {
+      "epoch": 2.7676198147402333,
+      "grad_norm": 0.25512566620189414,
+      "learning_rate": 1.9707245548626008e-05,
+      "loss": 0.2931,
+      "step": 1718
+    },
+    {
+      "epoch": 2.769230769230769,
+      "grad_norm": 0.24336440793633282,
+      "learning_rate": 1.968472777430397e-05,
+      "loss": 0.2718,
+      "step": 1719
+    },
+    {
+      "epoch": 2.770841723721305,
+      "grad_norm": 0.2365556332430933,
+      "learning_rate": 1.966221039972138e-05,
+      "loss": 0.2833,
+      "step": 1720
+    },
+    {
+      "epoch": 2.7724526782118404,
+      "grad_norm": 0.2489637128328502,
+      "learning_rate": 1.9639693453428428e-05,
+      "loss": 0.2614,
+      "step": 1721
+    },
+    {
+      "epoch": 2.774063632702376,
+      "grad_norm": 0.23427312408705342,
+      "learning_rate": 1.9617176963974757e-05,
+      "loss": 0.237,
+      "step": 1722
+    },
+    {
+      "epoch": 2.7756745871929116,
+      "grad_norm": 0.30355928331474424,
+      "learning_rate": 1.959466095990945e-05,
+      "loss": 0.3156,
+      "step": 1723
+    },
+    {
+      "epoch": 2.7772855416834474,
+      "grad_norm": 0.23089885134745378,
+      "learning_rate": 1.9572145469780957e-05,
+      "loss": 0.262,
+      "step": 1724
+    },
+    {
+      "epoch": 2.7788964961739833,
+      "grad_norm": 0.2650910801407474,
+      "learning_rate": 1.9549630522137084e-05,
+      "loss": 0.2691,
+      "step": 1725
+    },
+    {
+      "epoch": 2.7805074506645187,
+      "grad_norm": 0.27538303821293014,
+      "learning_rate": 1.9527116145524934e-05,
+      "loss": 0.2905,
+      "step": 1726
+    },
+    {
+      "epoch": 2.7821184051550545,
+      "grad_norm": 0.2529807875761806,
+      "learning_rate": 1.9504602368490918e-05,
+      "loss": 0.2971,
+      "step": 1727
+    },
+    {
+      "epoch": 2.78372935964559,
+      "grad_norm": 0.2799965953217938,
+      "learning_rate": 1.9482089219580655e-05,
+      "loss": 0.2752,
+      "step": 1728
+    },
+    {
+      "epoch": 2.7853403141361257,
+      "grad_norm": 0.25007926005011655,
+      "learning_rate": 1.9459576727338986e-05,
+      "loss": 0.2604,
+      "step": 1729
+    },
+    {
+      "epoch": 2.7869512686266615,
+      "grad_norm": 0.254990987065823,
+      "learning_rate": 1.9437064920309895e-05,
+      "loss": 0.2971,
+      "step": 1730
+    },
+    {
+      "epoch": 2.788562223117197,
+      "grad_norm": 0.26362113970903406,
+      "learning_rate": 1.941455382703655e-05,
+      "loss": 0.2538,
+      "step": 1731
+    },
+    {
+      "epoch": 2.7901731776077323,
+      "grad_norm": 0.246885971572867,
+      "learning_rate": 1.939204347606115e-05,
+      "loss": 0.2721,
+      "step": 1732
+    },
+    {
+      "epoch": 2.791784132098268,
+      "grad_norm": 0.266283943370594,
+      "learning_rate": 1.9369533895924992e-05,
+      "loss": 0.2866,
+      "step": 1733
+    },
+    {
+      "epoch": 2.793395086588804,
+      "grad_norm": 0.2680098900263936,
+      "learning_rate": 1.934702511516839e-05,
+      "loss": 0.2967,
+      "step": 1734
+    },
+    {
+      "epoch": 2.7950060410793394,
+      "grad_norm": 0.22740978433404274,
+      "learning_rate": 1.932451716233064e-05,
+      "loss": 0.2694,
+      "step": 1735
+    },
+    {
+      "epoch": 2.796616995569875,
+      "grad_norm": 0.2561037233597848,
+      "learning_rate": 1.930201006594999e-05,
+      "loss": 0.2679,
+      "step": 1736
+    },
+    {
+      "epoch": 2.7982279500604106,
+      "grad_norm": 0.23434576263054702,
+      "learning_rate": 1.9279503854563584e-05,
+      "loss": 0.2799,
+      "step": 1737
+    },
+    {
+      "epoch": 2.7998389045509464,
+      "grad_norm": 0.22674515230572093,
+      "learning_rate": 1.925699855670748e-05,
+      "loss": 0.2728,
+      "step": 1738
+    },
+    {
+      "epoch": 2.8014498590414822,
+      "grad_norm": 0.2428990371238118,
+      "learning_rate": 1.9234494200916554e-05,
+      "loss": 0.3036,
+      "step": 1739
+    },
+    {
+      "epoch": 2.8030608135320176,
+      "grad_norm": 0.21954338652190178,
+      "learning_rate": 1.9211990815724496e-05,
+      "loss": 0.2474,
+      "step": 1740
+    },
+    {
+      "epoch": 2.8046717680225535,
+      "grad_norm": 0.25787658937930996,
+      "learning_rate": 1.918948842966374e-05,
+      "loss": 0.2964,
+      "step": 1741
+    },
+    {
+      "epoch": 2.806282722513089,
+      "grad_norm": 0.225083476296253,
+      "learning_rate": 1.9166987071265506e-05,
+      "loss": 0.2625,
+      "step": 1742
+    },
+    {
+      "epoch": 2.8078936770036247,
+      "grad_norm": 0.26059530040772755,
+      "learning_rate": 1.914448676905966e-05,
+      "loss": 0.2689,
+      "step": 1743
+    },
+    {
+      "epoch": 2.8095046314941605,
+      "grad_norm": 0.2149852949576493,
+      "learning_rate": 1.9121987551574745e-05,
+      "loss": 0.2373,
+      "step": 1744
+    },
+    {
+      "epoch": 2.811115585984696,
+      "grad_norm": 0.26351487458727435,
+      "learning_rate": 1.9099489447337946e-05,
+      "loss": 0.2786,
+      "step": 1745
+    },
+    {
+      "epoch": 2.8127265404752317,
+      "grad_norm": 0.26753566423015324,
+      "learning_rate": 1.9076992484875014e-05,
+      "loss": 0.3151,
+      "step": 1746
+    },
+    {
+      "epoch": 2.814337494965767,
+      "grad_norm": 0.23423084235876232,
+      "learning_rate": 1.905449669271027e-05,
+      "loss": 0.276,
+      "step": 1747
+    },
+    {
+      "epoch": 2.815948449456303,
+      "grad_norm": 0.239399463127817,
+      "learning_rate": 1.9032002099366528e-05,
+      "loss": 0.2747,
+      "step": 1748
+    },
+    {
+      "epoch": 2.8175594039468383,
+      "grad_norm": 0.24223640625161968,
+      "learning_rate": 1.9009508733365103e-05,
+      "loss": 0.2576,
+      "step": 1749
+    },
+    {
+      "epoch": 2.819170358437374,
+      "grad_norm": 0.2741888988118937,
+      "learning_rate": 1.8987016623225748e-05,
+      "loss": 0.2846,
+      "step": 1750
+    },
+    {
+      "epoch": 2.8207813129279096,
+      "grad_norm": 0.26900885193390967,
+      "learning_rate": 1.896452579746663e-05,
+      "loss": 0.2628,
+      "step": 1751
+    },
+    {
+      "epoch": 2.8223922674184454,
+      "grad_norm": 0.25791432339739123,
+      "learning_rate": 1.8942036284604254e-05,
+      "loss": 0.2684,
+      "step": 1752
+    },
+    {
+      "epoch": 2.8240032219089812,
+      "grad_norm": 0.24175592345180597,
+      "learning_rate": 1.8919548113153517e-05,
+      "loss": 0.299,
+      "step": 1753
+    },
+    {
+      "epoch": 2.8256141763995166,
+      "grad_norm": 0.24350609167590054,
+      "learning_rate": 1.889706131162757e-05,
+      "loss": 0.2491,
+      "step": 1754
+    },
+    {
+      "epoch": 2.8272251308900525,
+      "grad_norm": 0.24783023887132075,
+      "learning_rate": 1.887457590853784e-05,
+      "loss": 0.2805,
+      "step": 1755
+    },
+    {
+      "epoch": 2.828836085380588,
+      "grad_norm": 0.23044942043966332,
+      "learning_rate": 1.8852091932393984e-05,
+      "loss": 0.2919,
+      "step": 1756
+    },
+    {
+      "epoch": 2.8304470398711237,
+      "grad_norm": 0.23547804631271316,
+      "learning_rate": 1.8829609411703855e-05,
+      "loss": 0.2729,
+      "step": 1757
+    },
+    {
+      "epoch": 2.8320579943616595,
+      "grad_norm": 0.2241584722151506,
+      "learning_rate": 1.8807128374973454e-05,
+      "loss": 0.274,
+      "step": 1758
+    },
+    {
+      "epoch": 2.833668948852195,
+      "grad_norm": 0.24199888275581322,
+      "learning_rate": 1.8784648850706883e-05,
+      "loss": 0.2853,
+      "step": 1759
+    },
+    {
+      "epoch": 2.8352799033427307,
+      "grad_norm": 0.2298411732619803,
+      "learning_rate": 1.8762170867406366e-05,
+      "loss": 0.2626,
+      "step": 1760
+    },
+    {
+      "epoch": 2.836890857833266,
+      "grad_norm": 0.21489914910143904,
+      "learning_rate": 1.873969445357214e-05,
+      "loss": 0.2683,
+      "step": 1761
+    },
+    {
+      "epoch": 2.838501812323802,
+      "grad_norm": 0.2618260680636698,
+      "learning_rate": 1.871721963770246e-05,
+      "loss": 0.2862,
+      "step": 1762
+    },
+    {
+      "epoch": 2.8401127668143373,
+      "grad_norm": 0.2845350731450781,
+      "learning_rate": 1.869474644829356e-05,
+      "loss": 0.2909,
+      "step": 1763
+    },
+    {
+      "epoch": 2.841723721304873,
+      "grad_norm": 0.21927422061321766,
+      "learning_rate": 1.867227491383962e-05,
+      "loss": 0.265,
+      "step": 1764
+    },
+    {
+      "epoch": 2.8433346757954086,
+      "grad_norm": 0.2603099243315801,
+      "learning_rate": 1.8649805062832697e-05,
+      "loss": 0.3039,
+      "step": 1765
+    },
+    {
+      "epoch": 2.8449456302859444,
+      "grad_norm": 0.23315130279336155,
+      "learning_rate": 1.8627336923762737e-05,
+      "loss": 0.2669,
+      "step": 1766
+    },
+    {
+      "epoch": 2.84655658477648,
+      "grad_norm": 0.2386078527910061,
+      "learning_rate": 1.8604870525117496e-05,
+      "loss": 0.2713,
+      "step": 1767
+    },
+    {
+      "epoch": 2.8481675392670156,
+      "grad_norm": 0.25248905060755344,
+      "learning_rate": 1.8582405895382544e-05,
+      "loss": 0.2866,
+      "step": 1768
+    },
+    {
+      "epoch": 2.8497784937575514,
+      "grad_norm": 0.2272473915570634,
+      "learning_rate": 1.8559943063041195e-05,
+      "loss": 0.2566,
+      "step": 1769
+    },
+    {
+      "epoch": 2.851389448248087,
+      "grad_norm": 0.2377600612126955,
+      "learning_rate": 1.853748205657448e-05,
+      "loss": 0.2657,
+      "step": 1770
+    },
+    {
+      "epoch": 2.8530004027386227,
+      "grad_norm": 0.25639490523892583,
+      "learning_rate": 1.8515022904461136e-05,
+      "loss": 0.2707,
+      "step": 1771
+    },
+    {
+      "epoch": 2.8546113572291585,
+      "grad_norm": 0.22335477679276985,
+      "learning_rate": 1.849256563517752e-05,
+      "loss": 0.2578,
+      "step": 1772
+    },
+    {
+      "epoch": 2.856222311719694,
+      "grad_norm": 0.26678857888253893,
+      "learning_rate": 1.8470110277197622e-05,
+      "loss": 0.2823,
+      "step": 1773
+    },
+    {
+      "epoch": 2.8578332662102297,
+      "grad_norm": 0.2658334462945636,
+      "learning_rate": 1.8447656858993e-05,
+      "loss": 0.2657,
+      "step": 1774
+    },
+    {
+      "epoch": 2.859444220700765,
+      "grad_norm": 0.23914934535135607,
+      "learning_rate": 1.8425205409032767e-05,
+      "loss": 0.2807,
+      "step": 1775
+    },
+    {
+      "epoch": 2.861055175191301,
+      "grad_norm": 0.23755248829706438,
+      "learning_rate": 1.8402755955783514e-05,
+      "loss": 0.2806,
+      "step": 1776
+    },
+    {
+      "epoch": 2.8626661296818368,
+      "grad_norm": 0.23622676175290067,
+      "learning_rate": 1.838030852770932e-05,
+      "loss": 0.2822,
+      "step": 1777
+    },
+    {
+      "epoch": 2.864277084172372,
+      "grad_norm": 0.2774308646019204,
+      "learning_rate": 1.835786315327169e-05,
+      "loss": 0.2953,
+      "step": 1778
+    },
+    {
+      "epoch": 2.8658880386629075,
+      "grad_norm": 0.23104176491388728,
+      "learning_rate": 1.8335419860929532e-05,
+      "loss": 0.252,
+      "step": 1779
+    },
+    {
+      "epoch": 2.8674989931534434,
+      "grad_norm": 0.2640215420459852,
+      "learning_rate": 1.831297867913911e-05,
+      "loss": 0.3052,
+      "step": 1780
+    },
+    {
+      "epoch": 2.869109947643979,
+      "grad_norm": 0.2508476637940116,
+      "learning_rate": 1.8290539636354e-05,
+      "loss": 0.2686,
+      "step": 1781
+    },
+    {
+      "epoch": 2.8707209021345146,
+      "grad_norm": 0.259821408101572,
+      "learning_rate": 1.8268102761025077e-05,
+      "loss": 0.2747,
+      "step": 1782
+    },
+    {
+      "epoch": 2.8723318566250504,
+      "grad_norm": 0.24412116210280163,
+      "learning_rate": 1.8245668081600477e-05,
+      "loss": 0.2755,
+      "step": 1783
+    },
+    {
+      "epoch": 2.873942811115586,
+      "grad_norm": 0.24896780062311544,
+      "learning_rate": 1.8223235626525542e-05,
+      "loss": 0.2714,
+      "step": 1784
+    },
+    {
+      "epoch": 2.8755537656061216,
+      "grad_norm": 0.24529433746790313,
+      "learning_rate": 1.820080542424278e-05,
+      "loss": 0.2637,
+      "step": 1785
+    },
+    {
+      "epoch": 2.8771647200966575,
+      "grad_norm": 0.272289653576584,
+      "learning_rate": 1.8178377503191875e-05,
+      "loss": 0.2916,
+      "step": 1786
+    },
+    {
+      "epoch": 2.878775674587193,
+      "grad_norm": 0.24673919563032362,
+      "learning_rate": 1.8155951891809592e-05,
+      "loss": 0.2817,
+      "step": 1787
+    },
+    {
+      "epoch": 2.8803866290777287,
+      "grad_norm": 0.24031746045696809,
+      "learning_rate": 1.813352861852977e-05,
+      "loss": 0.2788,
+      "step": 1788
+    },
+    {
+      "epoch": 2.881997583568264,
+      "grad_norm": 0.25670613044192553,
+      "learning_rate": 1.8111107711783293e-05,
+      "loss": 0.2658,
+      "step": 1789
+    },
+    {
+      "epoch": 2.8836085380588,
+      "grad_norm": 0.24396556057950303,
+      "learning_rate": 1.808868919999804e-05,
+      "loss": 0.256,
+      "step": 1790
+    },
+    {
+      "epoch": 2.8852194925493357,
+      "grad_norm": 0.24389592967270443,
+      "learning_rate": 1.806627311159886e-05,
+      "loss": 0.278,
+      "step": 1791
+    },
+    {
+      "epoch": 2.886830447039871,
+      "grad_norm": 0.25159917649723634,
+      "learning_rate": 1.8043859475007515e-05,
+      "loss": 0.2755,
+      "step": 1792
+    },
+    {
+      "epoch": 2.8884414015304065,
+      "grad_norm": 0.21554818631946265,
+      "learning_rate": 1.8021448318642666e-05,
+      "loss": 0.2546,
+      "step": 1793
+    },
+    {
+      "epoch": 2.8900523560209423,
+      "grad_norm": 0.22780147011061957,
+      "learning_rate": 1.7999039670919837e-05,
+      "loss": 0.2675,
+      "step": 1794
+    },
+    {
+      "epoch": 2.891663310511478,
+      "grad_norm": 0.2556179680309161,
+      "learning_rate": 1.797663356025136e-05,
+      "loss": 0.268,
+      "step": 1795
+    },
+    {
+      "epoch": 2.8932742650020136,
+      "grad_norm": 0.22179477121268426,
+      "learning_rate": 1.795423001504635e-05,
+      "loss": 0.2625,
+      "step": 1796
+    },
+    {
+      "epoch": 2.8948852194925494,
+      "grad_norm": 0.2566657026812676,
+      "learning_rate": 1.793182906371068e-05,
+      "loss": 0.3041,
+      "step": 1797
+    },
+    {
+      "epoch": 2.896496173983085,
+      "grad_norm": 0.22649500722803623,
+      "learning_rate": 1.7909430734646936e-05,
+      "loss": 0.2626,
+      "step": 1798
+    },
+    {
+      "epoch": 2.8981071284736206,
+      "grad_norm": 0.2382002490806491,
+      "learning_rate": 1.788703505625435e-05,
+      "loss": 0.2697,
+      "step": 1799
+    },
+    {
+      "epoch": 2.8997180829641565,
+      "grad_norm": 0.23333967874773118,
+      "learning_rate": 1.7864642056928823e-05,
+      "loss": 0.282,
+      "step": 1800
+    },
+    {
+      "epoch": 2.901329037454692,
+      "grad_norm": 0.22810377571609813,
+      "learning_rate": 1.7842251765062858e-05,
+      "loss": 0.2703,
+      "step": 1801
+    },
+    {
+      "epoch": 2.9029399919452277,
+      "grad_norm": 0.2475532042645999,
+      "learning_rate": 1.7819864209045512e-05,
+      "loss": 0.2893,
+      "step": 1802
+    },
+    {
+      "epoch": 2.904550946435763,
+      "grad_norm": 0.23819963629896387,
+      "learning_rate": 1.7797479417262375e-05,
+      "loss": 0.2693,
+      "step": 1803
+    },
+    {
+      "epoch": 2.906161900926299,
+      "grad_norm": 0.2498421038014751,
+      "learning_rate": 1.777509741809553e-05,
+      "loss": 0.2685,
+      "step": 1804
+    },
+    {
+      "epoch": 2.9077728554168347,
+      "grad_norm": 0.22386243561742944,
+      "learning_rate": 1.775271823992354e-05,
+      "loss": 0.2425,
+      "step": 1805
+    },
+    {
+      "epoch": 2.90938380990737,
+      "grad_norm": 0.2524226759643067,
+      "learning_rate": 1.773034191112137e-05,
+      "loss": 0.3005,
+      "step": 1806
+    },
+    {
+      "epoch": 2.9109947643979055,
+      "grad_norm": 0.23459453267664337,
+      "learning_rate": 1.7707968460060375e-05,
+      "loss": 0.2858,
+      "step": 1807
+    },
+    {
+      "epoch": 2.9126057188884413,
+      "grad_norm": 0.24955656470007692,
+      "learning_rate": 1.7685597915108257e-05,
+      "loss": 0.2793,
+      "step": 1808
+    },
+    {
+      "epoch": 2.914216673378977,
+      "grad_norm": 0.22015213316381949,
+      "learning_rate": 1.7663230304629066e-05,
+      "loss": 0.2702,
+      "step": 1809
+    },
+    {
+      "epoch": 2.9158276278695126,
+      "grad_norm": 0.2655548953653868,
+      "learning_rate": 1.7640865656983084e-05,
+      "loss": 0.3032,
+      "step": 1810
+    },
+    {
+      "epoch": 2.9174385823600484,
+      "grad_norm": 0.2323494023577661,
+      "learning_rate": 1.7618504000526863e-05,
+      "loss": 0.2761,
+      "step": 1811
+    },
+    {
+      "epoch": 2.9190495368505838,
+      "grad_norm": 0.2109588847916537,
+      "learning_rate": 1.759614536361316e-05,
+      "loss": 0.2614,
+      "step": 1812
+    },
+    {
+      "epoch": 2.9206604913411196,
+      "grad_norm": 0.25601356381819473,
+      "learning_rate": 1.7573789774590903e-05,
+      "loss": 0.3105,
+      "step": 1813
+    },
+    {
+      "epoch": 2.9222714458316554,
+      "grad_norm": 0.22547870004031134,
+      "learning_rate": 1.755143726180516e-05,
+      "loss": 0.2899,
+      "step": 1814
+    },
+    {
+      "epoch": 2.923882400322191,
+      "grad_norm": 0.2509635812735881,
+      "learning_rate": 1.7529087853597072e-05,
+      "loss": 0.2821,
+      "step": 1815
+    },
+    {
+      "epoch": 2.9254933548127267,
+      "grad_norm": 0.23583864792349135,
+      "learning_rate": 1.7506741578303883e-05,
+      "loss": 0.2623,
+      "step": 1816
+    },
+    {
+      "epoch": 2.927104309303262,
+      "grad_norm": 0.24318522703606663,
+      "learning_rate": 1.748439846425884e-05,
+      "loss": 0.2868,
+      "step": 1817
+    },
+    {
+      "epoch": 2.928715263793798,
+      "grad_norm": 0.22768641363630263,
+      "learning_rate": 1.7462058539791192e-05,
+      "loss": 0.2942,
+      "step": 1818
+    },
+    {
+      "epoch": 2.9303262182843337,
+      "grad_norm": 0.22037526237850094,
+      "learning_rate": 1.743972183322612e-05,
+      "loss": 0.2651,
+      "step": 1819
+    },
+    {
+      "epoch": 2.931937172774869,
+      "grad_norm": 0.2261074846645067,
+      "learning_rate": 1.7417388372884775e-05,
+      "loss": 0.2863,
+      "step": 1820
+    },
+    {
+      "epoch": 2.9335481272654045,
+      "grad_norm": 0.22274024245003757,
+      "learning_rate": 1.739505818708414e-05,
+      "loss": 0.2594,
+      "step": 1821
+    },
+    {
+      "epoch": 2.9351590817559403,
+      "grad_norm": 0.24040076004716376,
+      "learning_rate": 1.7372731304137072e-05,
+      "loss": 0.2691,
+      "step": 1822
+    },
+    {
+      "epoch": 2.936770036246476,
+      "grad_norm": 0.2506765713731466,
+      "learning_rate": 1.735040775235224e-05,
+      "loss": 0.2998,
+      "step": 1823
+    },
+    {
+      "epoch": 2.9383809907370115,
+      "grad_norm": 0.22705774588973596,
+      "learning_rate": 1.732808756003408e-05,
+      "loss": 0.2528,
+      "step": 1824
+    },
+    {
+      "epoch": 2.9399919452275474,
+      "grad_norm": 0.22020216475260657,
+      "learning_rate": 1.7305770755482788e-05,
+      "loss": 0.2594,
+      "step": 1825
+    },
+    {
+      "epoch": 2.9416028997180828,
+      "grad_norm": 0.2525904487369489,
+      "learning_rate": 1.7283457366994226e-05,
+      "loss": 0.2634,
+      "step": 1826
+    },
+    {
+      "epoch": 2.9432138542086186,
+      "grad_norm": 0.24948049883181214,
+      "learning_rate": 1.7261147422859967e-05,
+      "loss": 0.2824,
+      "step": 1827
+    },
+    {
+      "epoch": 2.9448248086991544,
+      "grad_norm": 0.22594459892161928,
+      "learning_rate": 1.7238840951367194e-05,
+      "loss": 0.2685,
+      "step": 1828
+    },
+    {
+      "epoch": 2.94643576318969,
+      "grad_norm": 0.23969345388513782,
+      "learning_rate": 1.72165379807987e-05,
+      "loss": 0.2776,
+      "step": 1829
+    },
+    {
+      "epoch": 2.9480467176802256,
+      "grad_norm": 0.23702835747098586,
+      "learning_rate": 1.7194238539432807e-05,
+      "loss": 0.286,
+      "step": 1830
+    },
+    {
+      "epoch": 2.949657672170761,
+      "grad_norm": 0.2308017426275451,
+      "learning_rate": 1.7171942655543415e-05,
+      "loss": 0.2553,
+      "step": 1831
+    },
+    {
+      "epoch": 2.951268626661297,
+      "grad_norm": 0.25865741662991343,
+      "learning_rate": 1.714965035739987e-05,
+      "loss": 0.295,
+      "step": 1832
+    },
+    {
+      "epoch": 2.9528795811518327,
+      "grad_norm": 0.22524224832126727,
+      "learning_rate": 1.7127361673266982e-05,
+      "loss": 0.2379,
+      "step": 1833
+    },
+    {
+      "epoch": 2.954490535642368,
+      "grad_norm": 0.2874573108392332,
+      "learning_rate": 1.7105076631404994e-05,
+      "loss": 0.3053,
+      "step": 1834
+    },
+    {
+      "epoch": 2.956101490132904,
+      "grad_norm": 0.2334494538238959,
+      "learning_rate": 1.7082795260069515e-05,
+      "loss": 0.2591,
+      "step": 1835
+    },
+    {
+      "epoch": 2.9577124446234393,
+      "grad_norm": 0.25784150732290884,
+      "learning_rate": 1.7060517587511512e-05,
+      "loss": 0.2823,
+      "step": 1836
+    },
+    {
+      "epoch": 2.959323399113975,
+      "grad_norm": 0.23058702287993874,
+      "learning_rate": 1.7038243641977238e-05,
+      "loss": 0.2464,
+      "step": 1837
+    },
+    {
+      "epoch": 2.9609343536045105,
+      "grad_norm": 0.22596215888674936,
+      "learning_rate": 1.701597345170825e-05,
+      "loss": 0.286,
+      "step": 1838
+    },
+    {
+      "epoch": 2.9625453080950463,
+      "grad_norm": 0.23090028001401205,
+      "learning_rate": 1.6993707044941334e-05,
+      "loss": 0.2828,
+      "step": 1839
+    },
+    {
+      "epoch": 2.9641562625855817,
+      "grad_norm": 0.22216529160650983,
+      "learning_rate": 1.6971444449908474e-05,
+      "loss": 0.2655,
+      "step": 1840
+    },
+    {
+      "epoch": 2.9657672170761176,
+      "grad_norm": 0.23782152766999168,
+      "learning_rate": 1.6949185694836806e-05,
+      "loss": 0.3075,
+      "step": 1841
+    },
+    {
+      "epoch": 2.9673781715666534,
+      "grad_norm": 0.23394803632471256,
+      "learning_rate": 1.6926930807948646e-05,
+      "loss": 0.2607,
+      "step": 1842
+    },
+    {
+      "epoch": 2.968989126057189,
+      "grad_norm": 0.22966128510957112,
+      "learning_rate": 1.6904679817461347e-05,
+      "loss": 0.2653,
+      "step": 1843
+    },
+    {
+      "epoch": 2.9706000805477246,
+      "grad_norm": 0.23928795508961923,
+      "learning_rate": 1.688243275158736e-05,
+      "loss": 0.2515,
+      "step": 1844
+    },
+    {
+      "epoch": 2.97221103503826,
+      "grad_norm": 0.2426394963369243,
+      "learning_rate": 1.6860189638534142e-05,
+      "loss": 0.269,
+      "step": 1845
+    },
+    {
+      "epoch": 2.973821989528796,
+      "grad_norm": 0.2321987864421601,
+      "learning_rate": 1.6837950506504158e-05,
+      "loss": 0.2831,
+      "step": 1846
+    },
+    {
+      "epoch": 2.9754329440193317,
+      "grad_norm": 0.25421214491796396,
+      "learning_rate": 1.6815715383694805e-05,
+      "loss": 0.284,
+      "step": 1847
+    },
+    {
+      "epoch": 2.977043898509867,
+      "grad_norm": 0.2443429609504397,
+      "learning_rate": 1.6793484298298387e-05,
+      "loss": 0.284,
+      "step": 1848
+    },
+    {
+      "epoch": 2.978654853000403,
+      "grad_norm": 0.21013676562255554,
+      "learning_rate": 1.6771257278502135e-05,
+      "loss": 0.2368,
+      "step": 1849
+    },
+    {
+      "epoch": 2.9802658074909383,
+      "grad_norm": 0.25821210868342026,
+      "learning_rate": 1.6749034352488077e-05,
+      "loss": 0.2834,
+      "step": 1850
+    },
+    {
+      "epoch": 2.981876761981474,
+      "grad_norm": 0.23933753025125024,
+      "learning_rate": 1.6726815548433072e-05,
+      "loss": 0.2712,
+      "step": 1851
+    },
+    {
+      "epoch": 2.98348771647201,
+      "grad_norm": 0.23023492804389561,
+      "learning_rate": 1.6704600894508743e-05,
+      "loss": 0.2598,
+      "step": 1852
+    },
+    {
+      "epoch": 2.9850986709625453,
+      "grad_norm": 0.270960478809936,
+      "learning_rate": 1.668239041888148e-05,
+      "loss": 0.3171,
+      "step": 1853
+    },
+    {
+      "epoch": 2.9867096254530807,
+      "grad_norm": 0.2165989069224987,
+      "learning_rate": 1.666018414971233e-05,
+      "loss": 0.2687,
+      "step": 1854
+    },
+    {
+      "epoch": 2.9883205799436166,
+      "grad_norm": 0.23657876878647432,
+      "learning_rate": 1.663798211515704e-05,
+      "loss": 0.2732,
+      "step": 1855
+    },
+    {
+      "epoch": 2.9899315344341524,
+      "grad_norm": 0.2514218073257401,
+      "learning_rate": 1.661578434336597e-05,
+      "loss": 0.2824,
+      "step": 1856
+    },
+    {
+      "epoch": 2.9915424889246878,
+      "grad_norm": 0.22874574938184158,
+      "learning_rate": 1.6593590862484095e-05,
+      "loss": 0.2753,
+      "step": 1857
+    },
+    {
+      "epoch": 2.9931534434152236,
+      "grad_norm": 0.2270308899508822,
+      "learning_rate": 1.6571401700650934e-05,
+      "loss": 0.2533,
+      "step": 1858
+    },
+    {
+      "epoch": 2.994764397905759,
+      "grad_norm": 0.23310437321259292,
+      "learning_rate": 1.6549216886000513e-05,
+      "loss": 0.2709,
+      "step": 1859
+    },
+    {
+      "epoch": 2.996375352396295,
+      "grad_norm": 0.26450041707096045,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 0.2978,
+      "step": 1860
+    },
+    {
+      "epoch": 2.9979863068868307,
+      "grad_norm": 0.2340175999215853,
+      "learning_rate": 1.6504860410756548e-05,
+      "loss": 0.2694,
+      "step": 1861
+    },
+    {
+      "epoch": 2.999597261377366,
+      "grad_norm": 0.27305237704293234,
+      "learning_rate": 1.6482688806403383e-05,
+      "loss": 0.3064,
+      "step": 1862
+    },
+    {
+      "epoch": 3.001208215867902,
+      "grad_norm": 0.3397063087612353,
+      "learning_rate": 1.646052166171368e-05,
+      "loss": 0.2409,
+      "step": 1863
+    },
+    {
+      "epoch": 3.0028191703584373,
+      "grad_norm": 0.30805634582005703,
+      "learning_rate": 1.6438359004793572e-05,
+      "loss": 0.2068,
+      "step": 1864
+    },
+    {
+      "epoch": 3.004430124848973,
+      "grad_norm": 0.45744869192317866,
+      "learning_rate": 1.64162008637435e-05,
+      "loss": 0.2269,
+      "step": 1865
+    },
+    {
+      "epoch": 3.0060410793395085,
+      "grad_norm": 0.2781379556860123,
+      "learning_rate": 1.639404726665817e-05,
+      "loss": 0.1878,
+      "step": 1866
+    },
+    {
+      "epoch": 3.0076520338300443,
+      "grad_norm": 0.45200098984478193,
+      "learning_rate": 1.637189824162654e-05,
+      "loss": 0.2083,
+      "step": 1867
+    },
+    {
+      "epoch": 3.00926298832058,
+      "grad_norm": 0.2842219362339001,
+      "learning_rate": 1.634975381673177e-05,
+      "loss": 0.1975,
+      "step": 1868
+    },
+    {
+      "epoch": 3.0108739428111155,
+      "grad_norm": 0.3164436406112575,
+      "learning_rate": 1.632761402005118e-05,
+      "loss": 0.1975,
+      "step": 1869
+    },
+    {
+      "epoch": 3.0124848973016514,
+      "grad_norm": 0.28217933369594117,
+      "learning_rate": 1.630547887965622e-05,
+      "loss": 0.2004,
+      "step": 1870
+    },
+    {
+      "epoch": 3.0140958517921868,
+      "grad_norm": 0.2569624519531185,
+      "learning_rate": 1.628334842361244e-05,
+      "loss": 0.1879,
+      "step": 1871
+    },
+    {
+      "epoch": 3.0157068062827226,
+      "grad_norm": 0.27431269684331583,
+      "learning_rate": 1.6261222679979462e-05,
+      "loss": 0.2135,
+      "step": 1872
+    },
+    {
+      "epoch": 3.017317760773258,
+      "grad_norm": 0.24794482330292733,
+      "learning_rate": 1.6239101676810917e-05,
+      "loss": 0.2132,
+      "step": 1873
+    },
+    {
+      "epoch": 3.018928715263794,
+      "grad_norm": 0.28290856893562144,
+      "learning_rate": 1.6216985442154427e-05,
+      "loss": 0.2237,
+      "step": 1874
+    },
+    {
+      "epoch": 3.0205396697543296,
+      "grad_norm": 0.24267628598445265,
+      "learning_rate": 1.619487400405158e-05,
+      "loss": 0.1798,
+      "step": 1875
+    },
+    {
+      "epoch": 3.022150624244865,
+      "grad_norm": 0.27946258366684773,
+      "learning_rate": 1.6172767390537874e-05,
+      "loss": 0.2003,
+      "step": 1876
+    },
+    {
+      "epoch": 3.023761578735401,
+      "grad_norm": 0.2635177433404853,
+      "learning_rate": 1.6150665629642685e-05,
+      "loss": 0.1947,
+      "step": 1877
+    },
+    {
+      "epoch": 3.0253725332259362,
+      "grad_norm": 0.2592624875109387,
+      "learning_rate": 1.6128568749389238e-05,
+      "loss": 0.1899,
+      "step": 1878
+    },
+    {
+      "epoch": 3.026983487716472,
+      "grad_norm": 0.25701500127549515,
+      "learning_rate": 1.6106476777794587e-05,
+      "loss": 0.2059,
+      "step": 1879
+    },
+    {
+      "epoch": 3.0285944422070075,
+      "grad_norm": 0.2531036287332326,
+      "learning_rate": 1.6084389742869543e-05,
+      "loss": 0.2029,
+      "step": 1880
+    },
+    {
+      "epoch": 3.0302053966975433,
+      "grad_norm": 0.26247322863177375,
+      "learning_rate": 1.6062307672618654e-05,
+      "loss": 0.2158,
+      "step": 1881
+    },
+    {
+      "epoch": 3.031816351188079,
+      "grad_norm": 0.2514848759579883,
+      "learning_rate": 1.6040230595040186e-05,
+      "loss": 0.2065,
+      "step": 1882
+    },
+    {
+      "epoch": 3.0334273056786145,
+      "grad_norm": 0.26086089307040033,
+      "learning_rate": 1.601815853812608e-05,
+      "loss": 0.1832,
+      "step": 1883
+    },
+    {
+      "epoch": 3.0350382601691503,
+      "grad_norm": 0.2650496609324423,
+      "learning_rate": 1.5996091529861897e-05,
+      "loss": 0.2189,
+      "step": 1884
+    },
+    {
+      "epoch": 3.0366492146596857,
+      "grad_norm": 0.24975861001178923,
+      "learning_rate": 1.5974029598226796e-05,
+      "loss": 0.1956,
+      "step": 1885
+    },
+    {
+      "epoch": 3.0382601691502216,
+      "grad_norm": 0.26398918888430833,
+      "learning_rate": 1.595197277119352e-05,
+      "loss": 0.2089,
+      "step": 1886
+    },
+    {
+      "epoch": 3.039871123640757,
+      "grad_norm": 0.23695742331210073,
+      "learning_rate": 1.5929921076728316e-05,
+      "loss": 0.1951,
+      "step": 1887
+    },
+    {
+      "epoch": 3.041482078131293,
+      "grad_norm": 0.23211753996183163,
+      "learning_rate": 1.590787454279093e-05,
+      "loss": 0.1923,
+      "step": 1888
+    },
+    {
+      "epoch": 3.0430930326218286,
+      "grad_norm": 0.24973175927852212,
+      "learning_rate": 1.5885833197334563e-05,
+      "loss": 0.186,
+      "step": 1889
+    },
+    {
+      "epoch": 3.044703987112364,
+      "grad_norm": 0.2529151185889554,
+      "learning_rate": 1.586379706830586e-05,
+      "loss": 0.2087,
+      "step": 1890
+    },
+    {
+      "epoch": 3.0463149416029,
+      "grad_norm": 0.2482208969171952,
+      "learning_rate": 1.584176618364482e-05,
+      "loss": 0.2011,
+      "step": 1891
+    },
+    {
+      "epoch": 3.0479258960934352,
+      "grad_norm": 0.2550637551299936,
+      "learning_rate": 1.5819740571284807e-05,
+      "loss": 0.2177,
+      "step": 1892
+    },
+    {
+      "epoch": 3.049536850583971,
+      "grad_norm": 0.23750126804753882,
+      "learning_rate": 1.5797720259152496e-05,
+      "loss": 0.174,
+      "step": 1893
+    },
+    {
+      "epoch": 3.0511478050745064,
+      "grad_norm": 0.24172300528999488,
+      "learning_rate": 1.5775705275167854e-05,
+      "loss": 0.2076,
+      "step": 1894
+    },
+    {
+      "epoch": 3.0527587595650423,
+      "grad_norm": 0.26506057370117664,
+      "learning_rate": 1.5753695647244083e-05,
+      "loss": 0.2068,
+      "step": 1895
+    },
+    {
+      "epoch": 3.054369714055578,
+      "grad_norm": 0.22576990598377747,
+      "learning_rate": 1.5731691403287595e-05,
+      "loss": 0.1925,
+      "step": 1896
+    },
+    {
+      "epoch": 3.0559806685461135,
+      "grad_norm": 0.2545620829654115,
+      "learning_rate": 1.5709692571197957e-05,
+      "loss": 0.2,
+      "step": 1897
+    },
+    {
+      "epoch": 3.0575916230366493,
+      "grad_norm": 0.26218591276871567,
+      "learning_rate": 1.568769917886792e-05,
+      "loss": 0.1976,
+      "step": 1898
+    },
+    {
+      "epoch": 3.0592025775271847,
+      "grad_norm": 0.23957115050109906,
+      "learning_rate": 1.5665711254183293e-05,
+      "loss": 0.1963,
+      "step": 1899
+    },
+    {
+      "epoch": 3.0608135320177206,
+      "grad_norm": 0.2595037940259688,
+      "learning_rate": 1.564372882502297e-05,
+      "loss": 0.2051,
+      "step": 1900
+    },
+    {
+      "epoch": 3.062424486508256,
+      "grad_norm": 0.2224941487256705,
+      "learning_rate": 1.5621751919258884e-05,
+      "loss": 0.1869,
+      "step": 1901
+    },
+    {
+      "epoch": 3.0640354409987918,
+      "grad_norm": 0.25468643872804336,
+      "learning_rate": 1.5599780564755956e-05,
+      "loss": 0.2183,
+      "step": 1902
+    },
+    {
+      "epoch": 3.0656463954893276,
+      "grad_norm": 0.25613678517781996,
+      "learning_rate": 1.5577814789372064e-05,
+      "loss": 0.1931,
+      "step": 1903
+    },
+    {
+      "epoch": 3.067257349979863,
+      "grad_norm": 0.24029809222854032,
+      "learning_rate": 1.555585462095802e-05,
+      "loss": 0.1899,
+      "step": 1904
+    },
+    {
+      "epoch": 3.068868304470399,
+      "grad_norm": 0.25815577075329543,
+      "learning_rate": 1.5533900087357527e-05,
+      "loss": 0.2053,
+      "step": 1905
+    },
+    {
+      "epoch": 3.070479258960934,
+      "grad_norm": 0.25550205510849705,
+      "learning_rate": 1.5511951216407142e-05,
+      "loss": 0.202,
+      "step": 1906
+    },
+    {
+      "epoch": 3.07209021345147,
+      "grad_norm": 0.24282903913932946,
+      "learning_rate": 1.5490008035936245e-05,
+      "loss": 0.1987,
+      "step": 1907
+    },
+    {
+      "epoch": 3.0737011679420054,
+      "grad_norm": 0.24787825404381872,
+      "learning_rate": 1.5468070573766982e-05,
+      "loss": 0.2126,
+      "step": 1908
+    },
+    {
+      "epoch": 3.0753121224325413,
+      "grad_norm": 0.23805631453760473,
+      "learning_rate": 1.5446138857714285e-05,
+      "loss": 0.2152,
+      "step": 1909
+    },
+    {
+      "epoch": 3.076923076923077,
+      "grad_norm": 0.22261953448924351,
+      "learning_rate": 1.5424212915585766e-05,
+      "loss": 0.2049,
+      "step": 1910
+    },
+    {
+      "epoch": 3.0785340314136125,
+      "grad_norm": 0.23389483167172603,
+      "learning_rate": 1.5402292775181732e-05,
+      "loss": 0.1841,
+      "step": 1911
+    },
+    {
+      "epoch": 3.0801449859041483,
+      "grad_norm": 0.24520165353665813,
+      "learning_rate": 1.5380378464295133e-05,
+      "loss": 0.2121,
+      "step": 1912
+    },
+    {
+      "epoch": 3.0817559403946837,
+      "grad_norm": 0.2355879009329018,
+      "learning_rate": 1.535847001071153e-05,
+      "loss": 0.1831,
+      "step": 1913
+    },
+    {
+      "epoch": 3.0833668948852195,
+      "grad_norm": 0.25353870145003116,
+      "learning_rate": 1.5336567442209052e-05,
+      "loss": 0.2093,
+      "step": 1914
+    },
+    {
+      "epoch": 3.084977849375755,
+      "grad_norm": 0.23181775476490288,
+      "learning_rate": 1.5314670786558358e-05,
+      "loss": 0.1993,
+      "step": 1915
+    },
+    {
+      "epoch": 3.0865888038662908,
+      "grad_norm": 0.24826824746417567,
+      "learning_rate": 1.5292780071522634e-05,
+      "loss": 0.2159,
+      "step": 1916
+    },
+    {
+      "epoch": 3.0881997583568266,
+      "grad_norm": 0.23097701612004543,
+      "learning_rate": 1.527089532485751e-05,
+      "loss": 0.1875,
+      "step": 1917
+    },
+    {
+      "epoch": 3.089810712847362,
+      "grad_norm": 0.23252448832192432,
+      "learning_rate": 1.5249016574311069e-05,
+      "loss": 0.1925,
+      "step": 1918
+    },
+    {
+      "epoch": 3.091421667337898,
+      "grad_norm": 0.2335044424719987,
+      "learning_rate": 1.5227143847623759e-05,
+      "loss": 0.1917,
+      "step": 1919
+    },
+    {
+      "epoch": 3.093032621828433,
+      "grad_norm": 0.2467731183493127,
+      "learning_rate": 1.5205277172528438e-05,
+      "loss": 0.2028,
+      "step": 1920
+    },
+    {
+      "epoch": 3.094643576318969,
+      "grad_norm": 0.2478688042585764,
+      "learning_rate": 1.5183416576750251e-05,
+      "loss": 0.2155,
+      "step": 1921
+    },
+    {
+      "epoch": 3.096254530809505,
+      "grad_norm": 0.23559604410809817,
+      "learning_rate": 1.5161562088006649e-05,
+      "loss": 0.1927,
+      "step": 1922
+    },
+    {
+      "epoch": 3.0978654853000402,
+      "grad_norm": 0.25132210120716414,
+      "learning_rate": 1.513971373400734e-05,
+      "loss": 0.2165,
+      "step": 1923
+    },
+    {
+      "epoch": 3.099476439790576,
+      "grad_norm": 0.21838775169451463,
+      "learning_rate": 1.5117871542454259e-05,
+      "loss": 0.1959,
+      "step": 1924
+    },
+    {
+      "epoch": 3.1010873942811115,
+      "grad_norm": 0.24648218446618922,
+      "learning_rate": 1.509603554104152e-05,
+      "loss": 0.215,
+      "step": 1925
+    },
+    {
+      "epoch": 3.1026983487716473,
+      "grad_norm": 0.23608736727692337,
+      "learning_rate": 1.5074205757455382e-05,
+      "loss": 0.1913,
+      "step": 1926
+    },
+    {
+      "epoch": 3.1043093032621827,
+      "grad_norm": 0.24688604035464398,
+      "learning_rate": 1.5052382219374238e-05,
+      "loss": 0.2028,
+      "step": 1927
+    },
+    {
+      "epoch": 3.1059202577527185,
+      "grad_norm": 0.21415210914040747,
+      "learning_rate": 1.5030564954468548e-05,
+      "loss": 0.1712,
+      "step": 1928
+    },
+    {
+      "epoch": 3.107531212243254,
+      "grad_norm": 0.28181734443533013,
+      "learning_rate": 1.500875399040083e-05,
+      "loss": 0.2383,
+      "step": 1929
+    },
+    {
+      "epoch": 3.1091421667337897,
+      "grad_norm": 0.23835213355733328,
+      "learning_rate": 1.498694935482559e-05,
+      "loss": 0.1919,
+      "step": 1930
+    },
+    {
+      "epoch": 3.1107531212243256,
+      "grad_norm": 0.239890464679267,
+      "learning_rate": 1.4965151075389344e-05,
+      "loss": 0.1857,
+      "step": 1931
+    },
+    {
+      "epoch": 3.112364075714861,
+      "grad_norm": 0.24414783199380563,
+      "learning_rate": 1.4943359179730523e-05,
+      "loss": 0.2172,
+      "step": 1932
+    },
+    {
+      "epoch": 3.113975030205397,
+      "grad_norm": 0.23874299205907634,
+      "learning_rate": 1.492157369547947e-05,
+      "loss": 0.1947,
+      "step": 1933
+    },
+    {
+      "epoch": 3.115585984695932,
+      "grad_norm": 0.26898773638453133,
+      "learning_rate": 1.4899794650258397e-05,
+      "loss": 0.224,
+      "step": 1934
+    },
+    {
+      "epoch": 3.117196939186468,
+      "grad_norm": 0.2577407154196406,
+      "learning_rate": 1.4878022071681368e-05,
+      "loss": 0.1898,
+      "step": 1935
+    },
+    {
+      "epoch": 3.118807893677004,
+      "grad_norm": 0.29459886996326795,
+      "learning_rate": 1.4856255987354232e-05,
+      "loss": 0.2289,
+      "step": 1936
+    },
+    {
+      "epoch": 3.1204188481675392,
+      "grad_norm": 0.2512628187386988,
+      "learning_rate": 1.4834496424874587e-05,
+      "loss": 0.2068,
+      "step": 1937
+    },
+    {
+      "epoch": 3.122029802658075,
+      "grad_norm": 0.267387282459161,
+      "learning_rate": 1.4812743411831814e-05,
+      "loss": 0.211,
+      "step": 1938
+    },
+    {
+      "epoch": 3.1236407571486104,
+      "grad_norm": 0.25843048547184483,
+      "learning_rate": 1.479099697580694e-05,
+      "loss": 0.2177,
+      "step": 1939
+    },
+    {
+      "epoch": 3.1252517116391463,
+      "grad_norm": 0.26113540198237006,
+      "learning_rate": 1.4769257144372668e-05,
+      "loss": 0.1874,
+      "step": 1940
+    },
+    {
+      "epoch": 3.1268626661296817,
+      "grad_norm": 0.2508889947235532,
+      "learning_rate": 1.4747523945093332e-05,
+      "loss": 0.2058,
+      "step": 1941
+    },
+    {
+      "epoch": 3.1284736206202175,
+      "grad_norm": 0.24792171081568534,
+      "learning_rate": 1.4725797405524866e-05,
+      "loss": 0.1922,
+      "step": 1942
+    },
+    {
+      "epoch": 3.130084575110753,
+      "grad_norm": 0.2536281872683442,
+      "learning_rate": 1.4704077553214734e-05,
+      "loss": 0.187,
+      "step": 1943
+    },
+    {
+      "epoch": 3.1316955296012887,
+      "grad_norm": 0.24634192517432316,
+      "learning_rate": 1.468236441570194e-05,
+      "loss": 0.2169,
+      "step": 1944
+    },
+    {
+      "epoch": 3.1333064840918246,
+      "grad_norm": 0.26202184945627943,
+      "learning_rate": 1.4660658020516966e-05,
+      "loss": 0.2107,
+      "step": 1945
+    },
+    {
+      "epoch": 3.13491743858236,
+      "grad_norm": 0.24692155351876013,
+      "learning_rate": 1.463895839518176e-05,
+      "loss": 0.2009,
+      "step": 1946
+    },
+    {
+      "epoch": 3.1365283930728958,
+      "grad_norm": 0.2651977196274272,
+      "learning_rate": 1.4617265567209673e-05,
+      "loss": 0.2161,
+      "step": 1947
+    },
+    {
+      "epoch": 3.138139347563431,
+      "grad_norm": 0.2254192338184251,
+      "learning_rate": 1.4595579564105432e-05,
+      "loss": 0.198,
+      "step": 1948
+    },
+    {
+      "epoch": 3.139750302053967,
+      "grad_norm": 0.2456828260242202,
+      "learning_rate": 1.457390041336512e-05,
+      "loss": 0.2169,
+      "step": 1949
+    },
+    {
+      "epoch": 3.141361256544503,
+      "grad_norm": 0.2383954425741448,
+      "learning_rate": 1.4552228142476138e-05,
+      "loss": 0.1913,
+      "step": 1950
+    },
+    {
+      "epoch": 3.142972211035038,
+      "grad_norm": 0.2388875846403241,
+      "learning_rate": 1.453056277891715e-05,
+      "loss": 0.2074,
+      "step": 1951
+    },
+    {
+      "epoch": 3.144583165525574,
+      "grad_norm": 0.24254252857096423,
+      "learning_rate": 1.4508904350158069e-05,
+      "loss": 0.184,
+      "step": 1952
+    },
+    {
+      "epoch": 3.1461941200161094,
+      "grad_norm": 0.2677691092809072,
+      "learning_rate": 1.4487252883660019e-05,
+      "loss": 0.1969,
+      "step": 1953
+    },
+    {
+      "epoch": 3.1478050745066453,
+      "grad_norm": 0.25451569692994286,
+      "learning_rate": 1.446560840687529e-05,
+      "loss": 0.2038,
+      "step": 1954
+    },
+    {
+      "epoch": 3.1494160289971806,
+      "grad_norm": 0.26165425081978877,
+      "learning_rate": 1.4443970947247308e-05,
+      "loss": 0.2257,
+      "step": 1955
+    },
+    {
+      "epoch": 3.1510269834877165,
+      "grad_norm": 0.2586993585923267,
+      "learning_rate": 1.4422340532210601e-05,
+      "loss": 0.2184,
+      "step": 1956
+    },
+    {
+      "epoch": 3.1526379379782523,
+      "grad_norm": 0.22749947793278547,
+      "learning_rate": 1.4400717189190783e-05,
+      "loss": 0.1886,
+      "step": 1957
+    },
+    {
+      "epoch": 3.1542488924687877,
+      "grad_norm": 0.25165536537988603,
+      "learning_rate": 1.4379100945604486e-05,
+      "loss": 0.1957,
+      "step": 1958
+    },
+    {
+      "epoch": 3.1558598469593235,
+      "grad_norm": 0.26090767826355016,
+      "learning_rate": 1.4357491828859333e-05,
+      "loss": 0.2062,
+      "step": 1959
+    },
+    {
+      "epoch": 3.157470801449859,
+      "grad_norm": 0.23823889920583088,
+      "learning_rate": 1.433588986635392e-05,
+      "loss": 0.1916,
+      "step": 1960
+    },
+    {
+      "epoch": 3.1590817559403948,
+      "grad_norm": 0.25301272984117323,
+      "learning_rate": 1.4314295085477784e-05,
+      "loss": 0.2129,
+      "step": 1961
+    },
+    {
+      "epoch": 3.16069271043093,
+      "grad_norm": 0.24454168144111912,
+      "learning_rate": 1.4292707513611332e-05,
+      "loss": 0.191,
+      "step": 1962
+    },
+    {
+      "epoch": 3.162303664921466,
+      "grad_norm": 0.2519013664831757,
+      "learning_rate": 1.4271127178125843e-05,
+      "loss": 0.213,
+      "step": 1963
+    },
+    {
+      "epoch": 3.163914619412002,
+      "grad_norm": 0.2322029065504692,
+      "learning_rate": 1.4249554106383432e-05,
+      "loss": 0.1915,
+      "step": 1964
+    },
+    {
+      "epoch": 3.165525573902537,
+      "grad_norm": 0.268939820675357,
+      "learning_rate": 1.4227988325736991e-05,
+      "loss": 0.2017,
+      "step": 1965
+    },
+    {
+      "epoch": 3.167136528393073,
+      "grad_norm": 0.24459588031927953,
+      "learning_rate": 1.420642986353016e-05,
+      "loss": 0.2095,
+      "step": 1966
+    },
+    {
+      "epoch": 3.1687474828836084,
+      "grad_norm": 0.23268508577761784,
+      "learning_rate": 1.4184878747097308e-05,
+      "loss": 0.2032,
+      "step": 1967
+    },
+    {
+      "epoch": 3.1703584373741442,
+      "grad_norm": 0.24314702305936226,
+      "learning_rate": 1.4163335003763506e-05,
+      "loss": 0.2094,
+      "step": 1968
+    },
+    {
+      "epoch": 3.1719693918646796,
+      "grad_norm": 0.2466331707682189,
+      "learning_rate": 1.414179866084445e-05,
+      "loss": 0.2039,
+      "step": 1969
+    },
+    {
+      "epoch": 3.1735803463552155,
+      "grad_norm": 0.24010384240031207,
+      "learning_rate": 1.4120269745646469e-05,
+      "loss": 0.2131,
+      "step": 1970
+    },
+    {
+      "epoch": 3.1751913008457513,
+      "grad_norm": 0.2315431702165574,
+      "learning_rate": 1.4098748285466464e-05,
+      "loss": 0.1836,
+      "step": 1971
+    },
+    {
+      "epoch": 3.1768022553362867,
+      "grad_norm": 0.24912715996998305,
+      "learning_rate": 1.4077234307591896e-05,
+      "loss": 0.2044,
+      "step": 1972
+    },
+    {
+      "epoch": 3.1784132098268225,
+      "grad_norm": 0.23840799641981053,
+      "learning_rate": 1.4055727839300733e-05,
+      "loss": 0.2022,
+      "step": 1973
+    },
+    {
+      "epoch": 3.180024164317358,
+      "grad_norm": 0.23128654085099198,
+      "learning_rate": 1.4034228907861414e-05,
+      "loss": 0.2028,
+      "step": 1974
+    },
+    {
+      "epoch": 3.1816351188078937,
+      "grad_norm": 0.26483087418421175,
+      "learning_rate": 1.4012737540532842e-05,
+      "loss": 0.2106,
+      "step": 1975
+    },
+    {
+      "epoch": 3.183246073298429,
+      "grad_norm": 0.23542978958227243,
+      "learning_rate": 1.3991253764564308e-05,
+      "loss": 0.2013,
+      "step": 1976
+    },
+    {
+      "epoch": 3.184857027788965,
+      "grad_norm": 0.23883779350162376,
+      "learning_rate": 1.3969777607195485e-05,
+      "loss": 0.1905,
+      "step": 1977
+    },
+    {
+      "epoch": 3.186467982279501,
+      "grad_norm": 0.22791901305740134,
+      "learning_rate": 1.3948309095656382e-05,
+      "loss": 0.2003,
+      "step": 1978
+    },
+    {
+      "epoch": 3.188078936770036,
+      "grad_norm": 0.2585822611597578,
+      "learning_rate": 1.3926848257167336e-05,
+      "loss": 0.2197,
+      "step": 1979
+    },
+    {
+      "epoch": 3.189689891260572,
+      "grad_norm": 0.2573536529025018,
+      "learning_rate": 1.3905395118938929e-05,
+      "loss": 0.206,
+      "step": 1980
+    },
+    {
+      "epoch": 3.1913008457511074,
+      "grad_norm": 0.24021103836798227,
+      "learning_rate": 1.3883949708171987e-05,
+      "loss": 0.208,
+      "step": 1981
+    },
+    {
+      "epoch": 3.1929118002416432,
+      "grad_norm": 0.24221200646286703,
+      "learning_rate": 1.3862512052057535e-05,
+      "loss": 0.2255,
+      "step": 1982
+    },
+    {
+      "epoch": 3.1945227547321786,
+      "grad_norm": 0.23090236397527916,
+      "learning_rate": 1.384108217777678e-05,
+      "loss": 0.1995,
+      "step": 1983
+    },
+    {
+      "epoch": 3.1961337092227144,
+      "grad_norm": 0.21761509754767042,
+      "learning_rate": 1.3819660112501054e-05,
+      "loss": 0.1835,
+      "step": 1984
+    },
+    {
+      "epoch": 3.1977446637132503,
+      "grad_norm": 0.2703261644476621,
+      "learning_rate": 1.3798245883391788e-05,
+      "loss": 0.2289,
+      "step": 1985
+    },
+    {
+      "epoch": 3.1993556182037857,
+      "grad_norm": 0.2230354526114237,
+      "learning_rate": 1.3776839517600458e-05,
+      "loss": 0.1979,
+      "step": 1986
+    },
+    {
+      "epoch": 3.2009665726943215,
+      "grad_norm": 0.23711275306860455,
+      "learning_rate": 1.3755441042268615e-05,
+      "loss": 0.1923,
+      "step": 1987
+    },
+    {
+      "epoch": 3.202577527184857,
+      "grad_norm": 0.24705039155672698,
+      "learning_rate": 1.3734050484527765e-05,
+      "loss": 0.2165,
+      "step": 1988
+    },
+    {
+      "epoch": 3.2041884816753927,
+      "grad_norm": 0.2377534616627376,
+      "learning_rate": 1.3712667871499385e-05,
+      "loss": 0.1929,
+      "step": 1989
+    },
+    {
+      "epoch": 3.205799436165928,
+      "grad_norm": 0.2449633515338068,
+      "learning_rate": 1.369129323029489e-05,
+      "loss": 0.2004,
+      "step": 1990
+    },
+    {
+      "epoch": 3.207410390656464,
+      "grad_norm": 0.2593360096012776,
+      "learning_rate": 1.3669926588015585e-05,
+      "loss": 0.2185,
+      "step": 1991
+    },
+    {
+      "epoch": 3.2090213451469998,
+      "grad_norm": 0.23324589375655463,
+      "learning_rate": 1.364856797175262e-05,
+      "loss": 0.1875,
+      "step": 1992
+    },
+    {
+      "epoch": 3.210632299637535,
+      "grad_norm": 0.2413326158872254,
+      "learning_rate": 1.362721740858697e-05,
+      "loss": 0.1826,
+      "step": 1993
+    },
+    {
+      "epoch": 3.212243254128071,
+      "grad_norm": 0.23918589432715148,
+      "learning_rate": 1.3605874925589419e-05,
+      "loss": 0.2068,
+      "step": 1994
+    },
+    {
+      "epoch": 3.2138542086186064,
+      "grad_norm": 0.23689262549972226,
+      "learning_rate": 1.3584540549820493e-05,
+      "loss": 0.2253,
+      "step": 1995
+    },
+    {
+      "epoch": 3.215465163109142,
+      "grad_norm": 0.23319941479929307,
+      "learning_rate": 1.3563214308330434e-05,
+      "loss": 0.1865,
+      "step": 1996
+    },
+    {
+      "epoch": 3.217076117599678,
+      "grad_norm": 0.23879194349455857,
+      "learning_rate": 1.3541896228159165e-05,
+      "loss": 0.213,
+      "step": 1997
+    },
+    {
+      "epoch": 3.2186870720902134,
+      "grad_norm": 0.25094170517305836,
+      "learning_rate": 1.3520586336336296e-05,
+      "loss": 0.2182,
+      "step": 1998
+    },
+    {
+      "epoch": 3.2202980265807493,
+      "grad_norm": 0.25223079196864345,
+      "learning_rate": 1.3499284659881013e-05,
+      "loss": 0.2001,
+      "step": 1999
+    },
+    {
+      "epoch": 3.2219089810712846,
+      "grad_norm": 0.24328823956234302,
+      "learning_rate": 1.3477991225802103e-05,
+      "loss": 0.2045,
+      "step": 2000
+    },
+    {
+      "epoch": 3.2235199355618205,
+      "grad_norm": 0.23787972021657403,
+      "learning_rate": 1.3456706061097905e-05,
+      "loss": 0.1807,
+      "step": 2001
+    },
+    {
+      "epoch": 3.225130890052356,
+      "grad_norm": 0.28090320878999836,
+      "learning_rate": 1.3435429192756275e-05,
+      "loss": 0.2231,
+      "step": 2002
+    },
+    {
+      "epoch": 3.2267418445428917,
+      "grad_norm": 0.2591510357821683,
+      "learning_rate": 1.3414160647754547e-05,
+      "loss": 0.2125,
+      "step": 2003
+    },
+    {
+      "epoch": 3.228352799033427,
+      "grad_norm": 0.22470518404451317,
+      "learning_rate": 1.339290045305948e-05,
+      "loss": 0.1817,
+      "step": 2004
+    },
+    {
+      "epoch": 3.229963753523963,
+      "grad_norm": 0.23711240462541178,
+      "learning_rate": 1.3371648635627285e-05,
+      "loss": 0.202,
+      "step": 2005
+    },
+    {
+      "epoch": 3.2315747080144988,
+      "grad_norm": 0.26273477899693026,
+      "learning_rate": 1.3350405222403529e-05,
+      "loss": 0.2079,
+      "step": 2006
+    },
+    {
+      "epoch": 3.233185662505034,
+      "grad_norm": 0.2325874086405899,
+      "learning_rate": 1.3329170240323124e-05,
+      "loss": 0.2106,
+      "step": 2007
+    },
+    {
+      "epoch": 3.23479661699557,
+      "grad_norm": 0.23774767585628861,
+      "learning_rate": 1.330794371631028e-05,
+      "loss": 0.1882,
+      "step": 2008
+    },
+    {
+      "epoch": 3.2364075714861054,
+      "grad_norm": 0.22890478281530627,
+      "learning_rate": 1.3286725677278525e-05,
+      "loss": 0.1993,
+      "step": 2009
+    },
+    {
+      "epoch": 3.238018525976641,
+      "grad_norm": 0.25495173404834265,
+      "learning_rate": 1.3265516150130577e-05,
+      "loss": 0.2195,
+      "step": 2010
+    },
+    {
+      "epoch": 3.239629480467177,
+      "grad_norm": 0.2433403725605887,
+      "learning_rate": 1.3244315161758392e-05,
+      "loss": 0.2033,
+      "step": 2011
+    },
+    {
+      "epoch": 3.2412404349577124,
+      "grad_norm": 0.24981637506327542,
+      "learning_rate": 1.3223122739043091e-05,
+      "loss": 0.2122,
+      "step": 2012
+    },
+    {
+      "epoch": 3.2428513894482482,
+      "grad_norm": 0.23690220270645826,
+      "learning_rate": 1.3201938908854942e-05,
+      "loss": 0.2007,
+      "step": 2013
+    },
+    {
+      "epoch": 3.2444623439387836,
+      "grad_norm": 0.23365777497487655,
+      "learning_rate": 1.3180763698053311e-05,
+      "loss": 0.2043,
+      "step": 2014
+    },
+    {
+      "epoch": 3.2460732984293195,
+      "grad_norm": 0.24922388772794682,
+      "learning_rate": 1.3159597133486628e-05,
+      "loss": 0.2049,
+      "step": 2015
+    },
+    {
+      "epoch": 3.247684252919855,
+      "grad_norm": 0.2615394743002068,
+      "learning_rate": 1.3138439241992376e-05,
+      "loss": 0.1982,
+      "step": 2016
+    },
+    {
+      "epoch": 3.2492952074103907,
+      "grad_norm": 0.23085725558413792,
+      "learning_rate": 1.3117290050397036e-05,
+      "loss": 0.2063,
+      "step": 2017
+    },
+    {
+      "epoch": 3.250906161900926,
+      "grad_norm": 0.2524638411576523,
+      "learning_rate": 1.3096149585516059e-05,
+      "loss": 0.201,
+      "step": 2018
+    },
+    {
+      "epoch": 3.252517116391462,
+      "grad_norm": 0.22321898517204689,
+      "learning_rate": 1.3075017874153808e-05,
+      "loss": 0.1873,
+      "step": 2019
+    },
+    {
+      "epoch": 3.2541280708819977,
+      "grad_norm": 0.24208007420834118,
+      "learning_rate": 1.3053894943103598e-05,
+      "loss": 0.2028,
+      "step": 2020
+    },
+    {
+      "epoch": 3.255739025372533,
+      "grad_norm": 0.23464995395813093,
+      "learning_rate": 1.303278081914756e-05,
+      "loss": 0.1874,
+      "step": 2021
+    },
+    {
+      "epoch": 3.257349979863069,
+      "grad_norm": 0.23529260639203314,
+      "learning_rate": 1.3011675529056688e-05,
+      "loss": 0.1984,
+      "step": 2022
+    },
+    {
+      "epoch": 3.2589609343536043,
+      "grad_norm": 0.23749352729237172,
+      "learning_rate": 1.2990579099590763e-05,
+      "loss": 0.201,
+      "step": 2023
+    },
+    {
+      "epoch": 3.26057188884414,
+      "grad_norm": 0.2427641865797294,
+      "learning_rate": 1.2969491557498342e-05,
+      "loss": 0.2031,
+      "step": 2024
+    },
+    {
+      "epoch": 3.262182843334676,
+      "grad_norm": 0.2375077667321114,
+      "learning_rate": 1.2948412929516703e-05,
+      "loss": 0.1918,
+      "step": 2025
+    },
+    {
+      "epoch": 3.2637937978252114,
+      "grad_norm": 0.24542218137105598,
+      "learning_rate": 1.2927343242371815e-05,
+      "loss": 0.207,
+      "step": 2026
+    },
+    {
+      "epoch": 3.2654047523157472,
+      "grad_norm": 0.2390333630741558,
+      "learning_rate": 1.2906282522778341e-05,
+      "loss": 0.1998,
+      "step": 2027
+    },
+    {
+      "epoch": 3.2670157068062826,
+      "grad_norm": 0.24837545936493144,
+      "learning_rate": 1.2885230797439543e-05,
+      "loss": 0.2157,
+      "step": 2028
+    },
+    {
+      "epoch": 3.2686266612968184,
+      "grad_norm": 0.2519965003627847,
+      "learning_rate": 1.2864188093047291e-05,
+      "loss": 0.1995,
+      "step": 2029
+    },
+    {
+      "epoch": 3.270237615787354,
+      "grad_norm": 0.23774917985531094,
+      "learning_rate": 1.2843154436282014e-05,
+      "loss": 0.2081,
+      "step": 2030
+    },
+    {
+      "epoch": 3.2718485702778897,
+      "grad_norm": 0.24180882546372334,
+      "learning_rate": 1.2822129853812682e-05,
+      "loss": 0.1794,
+      "step": 2031
+    },
+    {
+      "epoch": 3.273459524768425,
+      "grad_norm": 0.22947773377827083,
+      "learning_rate": 1.2801114372296742e-05,
+      "loss": 0.204,
+      "step": 2032
+    },
+    {
+      "epoch": 3.275070479258961,
+      "grad_norm": 0.2455803270540699,
+      "learning_rate": 1.2780108018380103e-05,
+      "loss": 0.2038,
+      "step": 2033
+    },
+    {
+      "epoch": 3.2766814337494967,
+      "grad_norm": 0.24059114783030533,
+      "learning_rate": 1.2759110818697114e-05,
+      "loss": 0.203,
+      "step": 2034
+    },
+    {
+      "epoch": 3.278292388240032,
+      "grad_norm": 0.22748506298956164,
+      "learning_rate": 1.273812279987051e-05,
+      "loss": 0.2093,
+      "step": 2035
+    },
+    {
+      "epoch": 3.279903342730568,
+      "grad_norm": 0.2469088454838881,
+      "learning_rate": 1.2717143988511392e-05,
+      "loss": 0.2227,
+      "step": 2036
+    },
+    {
+      "epoch": 3.2815142972211033,
+      "grad_norm": 0.23674243065437753,
+      "learning_rate": 1.2696174411219164e-05,
+      "loss": 0.1981,
+      "step": 2037
+    },
+    {
+      "epoch": 3.283125251711639,
+      "grad_norm": 0.22507856798719222,
+      "learning_rate": 1.2675214094581547e-05,
+      "loss": 0.1886,
+      "step": 2038
+    },
+    {
+      "epoch": 3.284736206202175,
+      "grad_norm": 0.24558240776697116,
+      "learning_rate": 1.2654263065174515e-05,
+      "loss": 0.2104,
+      "step": 2039
+    },
+    {
+      "epoch": 3.2863471606927104,
+      "grad_norm": 0.2247916689809979,
+      "learning_rate": 1.263332134956226e-05,
+      "loss": 0.1996,
+      "step": 2040
+    },
+    {
+      "epoch": 3.287958115183246,
+      "grad_norm": 0.24098803655370316,
+      "learning_rate": 1.2612388974297161e-05,
+      "loss": 0.2301,
+      "step": 2041
+    },
+    {
+      "epoch": 3.2895690696737816,
+      "grad_norm": 0.22790610018869215,
+      "learning_rate": 1.259146596591978e-05,
+      "loss": 0.2044,
+      "step": 2042
+    },
+    {
+      "epoch": 3.2911800241643174,
+      "grad_norm": 0.29439964547743164,
+      "learning_rate": 1.2570552350958764e-05,
+      "loss": 0.1872,
+      "step": 2043
+    },
+    {
+      "epoch": 3.292790978654853,
+      "grad_norm": 0.23266526895741785,
+      "learning_rate": 1.2549648155930875e-05,
+      "loss": 0.221,
+      "step": 2044
+    },
+    {
+      "epoch": 3.2944019331453887,
+      "grad_norm": 0.23179183784366264,
+      "learning_rate": 1.2528753407340929e-05,
+      "loss": 0.194,
+      "step": 2045
+    },
+    {
+      "epoch": 3.296012887635924,
+      "grad_norm": 0.21779866661814773,
+      "learning_rate": 1.250786813168176e-05,
+      "loss": 0.1891,
+      "step": 2046
+    },
+    {
+      "epoch": 3.29762384212646,
+      "grad_norm": 0.2532975434083456,
+      "learning_rate": 1.2486992355434197e-05,
+      "loss": 0.2207,
+      "step": 2047
+    },
+    {
+      "epoch": 3.2992347966169957,
+      "grad_norm": 0.22096278667714603,
+      "learning_rate": 1.2466126105067014e-05,
+      "loss": 0.1821,
+      "step": 2048
+    },
+    {
+      "epoch": 3.300845751107531,
+      "grad_norm": 0.2530562863101158,
+      "learning_rate": 1.2445269407036908e-05,
+      "loss": 0.2133,
+      "step": 2049
+    },
+    {
+      "epoch": 3.302456705598067,
+      "grad_norm": 0.24462052555468472,
+      "learning_rate": 1.242442228778848e-05,
+      "loss": 0.1947,
+      "step": 2050
+    },
+    {
+      "epoch": 3.3040676600886023,
+      "grad_norm": 0.22993112034177768,
+      "learning_rate": 1.2403584773754176e-05,
+      "loss": 0.1884,
+      "step": 2051
+    },
+    {
+      "epoch": 3.305678614579138,
+      "grad_norm": 0.2375252762228901,
+      "learning_rate": 1.238275689135425e-05,
+      "loss": 0.2068,
+      "step": 2052
+    },
+    {
+      "epoch": 3.307289569069674,
+      "grad_norm": 0.23974997189809513,
+      "learning_rate": 1.2361938666996772e-05,
+      "loss": 0.2094,
+      "step": 2053
+    },
+    {
+      "epoch": 3.3089005235602094,
+      "grad_norm": 0.23634430658941374,
+      "learning_rate": 1.2341130127077548e-05,
+      "loss": 0.1967,
+      "step": 2054
+    },
+    {
+      "epoch": 3.310511478050745,
+      "grad_norm": 0.23446190903752556,
+      "learning_rate": 1.2320331297980097e-05,
+      "loss": 0.183,
+      "step": 2055
+    },
+    {
+      "epoch": 3.3121224325412806,
+      "grad_norm": 0.23976096133391758,
+      "learning_rate": 1.2299542206075641e-05,
+      "loss": 0.2365,
+      "step": 2056
+    },
+    {
+      "epoch": 3.3137333870318164,
+      "grad_norm": 0.23801927478749976,
+      "learning_rate": 1.2278762877723058e-05,
+      "loss": 0.2101,
+      "step": 2057
+    },
+    {
+      "epoch": 3.315344341522352,
+      "grad_norm": 0.24082087940178765,
+      "learning_rate": 1.2257993339268843e-05,
+      "loss": 0.1928,
+      "step": 2058
+    },
+    {
+      "epoch": 3.3169552960128876,
+      "grad_norm": 0.24713092173760648,
+      "learning_rate": 1.223723361704706e-05,
+      "loss": 0.1997,
+      "step": 2059
+    },
+    {
+      "epoch": 3.3185662505034235,
+      "grad_norm": 0.24274357069914468,
+      "learning_rate": 1.221648373737935e-05,
+      "loss": 0.2094,
+      "step": 2060
+    },
+    {
+      "epoch": 3.320177204993959,
+      "grad_norm": 0.23991670470326582,
+      "learning_rate": 1.2195743726574869e-05,
+      "loss": 0.2056,
+      "step": 2061
+    },
+    {
+      "epoch": 3.3217881594844947,
+      "grad_norm": 0.23272961485583524,
+      "learning_rate": 1.2175013610930253e-05,
+      "loss": 0.2045,
+      "step": 2062
+    },
+    {
+      "epoch": 3.32339911397503,
+      "grad_norm": 0.2549934176914694,
+      "learning_rate": 1.2154293416729606e-05,
+      "loss": 0.2088,
+      "step": 2063
+    },
+    {
+      "epoch": 3.325010068465566,
+      "grad_norm": 0.2627470044380523,
+      "learning_rate": 1.2133583170244422e-05,
+      "loss": 0.2154,
+      "step": 2064
+    },
+    {
+      "epoch": 3.3266210229561013,
+      "grad_norm": 0.2418643009081854,
+      "learning_rate": 1.2112882897733634e-05,
+      "loss": 0.1826,
+      "step": 2065
+    },
+    {
+      "epoch": 3.328231977446637,
+      "grad_norm": 0.23964309123146355,
+      "learning_rate": 1.2092192625443469e-05,
+      "loss": 0.1956,
+      "step": 2066
+    },
+    {
+      "epoch": 3.329842931937173,
+      "grad_norm": 0.26877059633483413,
+      "learning_rate": 1.2071512379607515e-05,
+      "loss": 0.2125,
+      "step": 2067
+    },
+    {
+      "epoch": 3.3314538864277083,
+      "grad_norm": 0.2718834091071389,
+      "learning_rate": 1.2050842186446636e-05,
+      "loss": 0.1953,
+      "step": 2068
+    },
+    {
+      "epoch": 3.333064840918244,
+      "grad_norm": 0.24287090855325613,
+      "learning_rate": 1.2030182072168957e-05,
+      "loss": 0.1866,
+      "step": 2069
+    },
+    {
+      "epoch": 3.3346757954087796,
+      "grad_norm": 0.25823157441435457,
+      "learning_rate": 1.2009532062969801e-05,
+      "loss": 0.2084,
+      "step": 2070
+    },
+    {
+      "epoch": 3.3362867498993154,
+      "grad_norm": 0.239590496772935,
+      "learning_rate": 1.1988892185031697e-05,
+      "loss": 0.1934,
+      "step": 2071
+    },
+    {
+      "epoch": 3.3378977043898512,
+      "grad_norm": 0.25170794325761425,
+      "learning_rate": 1.1968262464524333e-05,
+      "loss": 0.2099,
+      "step": 2072
+    },
+    {
+      "epoch": 3.3395086588803866,
+      "grad_norm": 0.2774141741661976,
+      "learning_rate": 1.1947642927604507e-05,
+      "loss": 0.1951,
+      "step": 2073
+    },
+    {
+      "epoch": 3.3411196133709224,
+      "grad_norm": 0.25992979182223386,
+      "learning_rate": 1.1927033600416113e-05,
+      "loss": 0.1895,
+      "step": 2074
+    },
+    {
+      "epoch": 3.342730567861458,
+      "grad_norm": 0.2676459769274524,
+      "learning_rate": 1.190643450909008e-05,
+      "loss": 0.2062,
+      "step": 2075
+    },
+    {
+      "epoch": 3.3443415223519937,
+      "grad_norm": 0.24491077877038162,
+      "learning_rate": 1.1885845679744396e-05,
+      "loss": 0.1842,
+      "step": 2076
+    },
+    {
+      "epoch": 3.345952476842529,
+      "grad_norm": 0.2573204990604907,
+      "learning_rate": 1.1865267138484e-05,
+      "loss": 0.2252,
+      "step": 2077
+    },
+    {
+      "epoch": 3.347563431333065,
+      "grad_norm": 0.2488993418599943,
+      "learning_rate": 1.1844698911400805e-05,
+      "loss": 0.2058,
+      "step": 2078
+    },
+    {
+      "epoch": 3.3491743858236003,
+      "grad_norm": 0.2221579767832753,
+      "learning_rate": 1.1824141024573647e-05,
+      "loss": 0.2012,
+      "step": 2079
+    },
+    {
+      "epoch": 3.350785340314136,
+      "grad_norm": 0.25173011551503893,
+      "learning_rate": 1.1803593504068256e-05,
+      "loss": 0.2202,
+      "step": 2080
+    },
+    {
+      "epoch": 3.352396294804672,
+      "grad_norm": 0.2403210358197002,
+      "learning_rate": 1.1783056375937193e-05,
+      "loss": 0.2103,
+      "step": 2081
+    },
+    {
+      "epoch": 3.3540072492952073,
+      "grad_norm": 0.23783146032565713,
+      "learning_rate": 1.1762529666219869e-05,
+      "loss": 0.2067,
+      "step": 2082
+    },
+    {
+      "epoch": 3.355618203785743,
+      "grad_norm": 0.2423185833751914,
+      "learning_rate": 1.174201340094248e-05,
+      "loss": 0.2043,
+      "step": 2083
+    },
+    {
+      "epoch": 3.3572291582762785,
+      "grad_norm": 0.24591036904977595,
+      "learning_rate": 1.172150760611797e-05,
+      "loss": 0.1968,
+      "step": 2084
+    },
+    {
+      "epoch": 3.3588401127668144,
+      "grad_norm": 0.23724633382039845,
+      "learning_rate": 1.1701012307746021e-05,
+      "loss": 0.2087,
+      "step": 2085
+    },
+    {
+      "epoch": 3.36045106725735,
+      "grad_norm": 0.24067309607867643,
+      "learning_rate": 1.168052753181298e-05,
+      "loss": 0.2029,
+      "step": 2086
+    },
+    {
+      "epoch": 3.3620620217478856,
+      "grad_norm": 0.24166724371741746,
+      "learning_rate": 1.1660053304291894e-05,
+      "loss": 0.1864,
+      "step": 2087
+    },
+    {
+      "epoch": 3.3636729762384214,
+      "grad_norm": 0.24147797064255336,
+      "learning_rate": 1.16395896511424e-05,
+      "loss": 0.2112,
+      "step": 2088
+    },
+    {
+      "epoch": 3.365283930728957,
+      "grad_norm": 0.24098212734239127,
+      "learning_rate": 1.1619136598310725e-05,
+      "loss": 0.2045,
+      "step": 2089
+    },
+    {
+      "epoch": 3.3668948852194927,
+      "grad_norm": 0.25019441770575873,
+      "learning_rate": 1.1598694171729703e-05,
+      "loss": 0.211,
+      "step": 2090
+    },
+    {
+      "epoch": 3.368505839710028,
+      "grad_norm": 0.2365678451068785,
+      "learning_rate": 1.1578262397318642e-05,
+      "loss": 0.1937,
+      "step": 2091
+    },
+    {
+      "epoch": 3.370116794200564,
+      "grad_norm": 0.23164431406226096,
+      "learning_rate": 1.1557841300983363e-05,
+      "loss": 0.1858,
+      "step": 2092
+    },
+    {
+      "epoch": 3.3717277486910993,
+      "grad_norm": 0.24563176728263572,
+      "learning_rate": 1.1537430908616152e-05,
+      "loss": 0.2159,
+      "step": 2093
+    },
+    {
+      "epoch": 3.373338703181635,
+      "grad_norm": 0.23552543012415264,
+      "learning_rate": 1.151703124609573e-05,
+      "loss": 0.2163,
+      "step": 2094
+    },
+    {
+      "epoch": 3.374949657672171,
+      "grad_norm": 0.23518558644454512,
+      "learning_rate": 1.1496642339287191e-05,
+      "loss": 0.2063,
+      "step": 2095
+    },
+    {
+      "epoch": 3.3765606121627063,
+      "grad_norm": 0.2441080853367346,
+      "learning_rate": 1.1476264214042004e-05,
+      "loss": 0.2042,
+      "step": 2096
+    },
+    {
+      "epoch": 3.378171566653242,
+      "grad_norm": 0.451665532949196,
+      "learning_rate": 1.1455896896197974e-05,
+      "loss": 0.2149,
+      "step": 2097
+    },
+    {
+      "epoch": 3.3797825211437775,
+      "grad_norm": 0.25128065966750873,
+      "learning_rate": 1.1435540411579203e-05,
+      "loss": 0.2113,
+      "step": 2098
+    },
+    {
+      "epoch": 3.3813934756343134,
+      "grad_norm": 0.2225475197951388,
+      "learning_rate": 1.1415194785996045e-05,
+      "loss": 0.1594,
+      "step": 2099
+    },
+    {
+      "epoch": 3.383004430124849,
+      "grad_norm": 0.2617602545629556,
+      "learning_rate": 1.1394860045245084e-05,
+      "loss": 0.229,
+      "step": 2100
+    },
+    {
+      "epoch": 3.3846153846153846,
+      "grad_norm": 0.24500084267347086,
+      "learning_rate": 1.137453621510912e-05,
+      "loss": 0.2098,
+      "step": 2101
+    },
+    {
+      "epoch": 3.3862263391059204,
+      "grad_norm": 0.23416531343487373,
+      "learning_rate": 1.1354223321357119e-05,
+      "loss": 0.2084,
+      "step": 2102
+    },
+    {
+      "epoch": 3.387837293596456,
+      "grad_norm": 0.22091357857993194,
+      "learning_rate": 1.1333921389744153e-05,
+      "loss": 0.1967,
+      "step": 2103
+    },
+    {
+      "epoch": 3.3894482480869916,
+      "grad_norm": 0.2511364869911767,
+      "learning_rate": 1.1313630446011427e-05,
+      "loss": 0.216,
+      "step": 2104
+    },
+    {
+      "epoch": 3.391059202577527,
+      "grad_norm": 0.23267127034952573,
+      "learning_rate": 1.1293350515886203e-05,
+      "loss": 0.2022,
+      "step": 2105
+    },
+    {
+      "epoch": 3.392670157068063,
+      "grad_norm": 0.23555886598588704,
+      "learning_rate": 1.1273081625081777e-05,
+      "loss": 0.1995,
+      "step": 2106
+    },
+    {
+      "epoch": 3.3942811115585982,
+      "grad_norm": 0.23868806139569818,
+      "learning_rate": 1.1252823799297433e-05,
+      "loss": 0.1773,
+      "step": 2107
+    },
+    {
+      "epoch": 3.395892066049134,
+      "grad_norm": 0.27452473674180444,
+      "learning_rate": 1.123257706421845e-05,
+      "loss": 0.2199,
+      "step": 2108
+    },
+    {
+      "epoch": 3.39750302053967,
+      "grad_norm": 0.2394943226611453,
+      "learning_rate": 1.1212341445516043e-05,
+      "loss": 0.1981,
+      "step": 2109
+    },
+    {
+      "epoch": 3.3991139750302053,
+      "grad_norm": 0.24612894971879087,
+      "learning_rate": 1.1192116968847313e-05,
+      "loss": 0.1968,
+      "step": 2110
+    },
+    {
+      "epoch": 3.400724929520741,
+      "grad_norm": 0.25883844195430433,
+      "learning_rate": 1.1171903659855256e-05,
+      "loss": 0.1949,
+      "step": 2111
+    },
+    {
+      "epoch": 3.4023358840112765,
+      "grad_norm": 0.23984551387483893,
+      "learning_rate": 1.1151701544168685e-05,
+      "loss": 0.1993,
+      "step": 2112
+    },
+    {
+      "epoch": 3.4039468385018123,
+      "grad_norm": 0.23016178489654152,
+      "learning_rate": 1.1131510647402246e-05,
+      "loss": 0.1948,
+      "step": 2113
+    },
+    {
+      "epoch": 3.405557792992348,
+      "grad_norm": 0.2631642324887254,
+      "learning_rate": 1.1111330995156335e-05,
+      "loss": 0.1918,
+      "step": 2114
+    },
+    {
+      "epoch": 3.4071687474828836,
+      "grad_norm": 0.2506220508839031,
+      "learning_rate": 1.1091162613017113e-05,
+      "loss": 0.2062,
+      "step": 2115
+    },
+    {
+      "epoch": 3.4087797019734194,
+      "grad_norm": 0.25264234683049214,
+      "learning_rate": 1.1071005526556444e-05,
+      "loss": 0.2141,
+      "step": 2116
+    },
+    {
+      "epoch": 3.410390656463955,
+      "grad_norm": 0.25278773105827357,
+      "learning_rate": 1.1050859761331867e-05,
+      "loss": 0.1915,
+      "step": 2117
+    },
+    {
+      "epoch": 3.4120016109544906,
+      "grad_norm": 0.23224228492195387,
+      "learning_rate": 1.1030725342886556e-05,
+      "loss": 0.1938,
+      "step": 2118
+    },
+    {
+      "epoch": 3.413612565445026,
+      "grad_norm": 0.2142415135543403,
+      "learning_rate": 1.101060229674932e-05,
+      "loss": 0.1936,
+      "step": 2119
+    },
+    {
+      "epoch": 3.415223519935562,
+      "grad_norm": 0.23554798219735495,
+      "learning_rate": 1.0990490648434541e-05,
+      "loss": 0.1969,
+      "step": 2120
+    },
+    {
+      "epoch": 3.4168344744260972,
+      "grad_norm": 0.26354880573854594,
+      "learning_rate": 1.0970390423442142e-05,
+      "loss": 0.2096,
+      "step": 2121
+    },
+    {
+      "epoch": 3.418445428916633,
+      "grad_norm": 0.24440684389786652,
+      "learning_rate": 1.0950301647257572e-05,
+      "loss": 0.2051,
+      "step": 2122
+    },
+    {
+      "epoch": 3.420056383407169,
+      "grad_norm": 0.2552755780564851,
+      "learning_rate": 1.0930224345351751e-05,
+      "loss": 0.2117,
+      "step": 2123
+    },
+    {
+      "epoch": 3.4216673378977043,
+      "grad_norm": 0.29304190170396727,
+      "learning_rate": 1.0910158543181073e-05,
+      "loss": 0.1947,
+      "step": 2124
+    },
+    {
+      "epoch": 3.42327829238824,
+      "grad_norm": 0.2458155097174551,
+      "learning_rate": 1.089010426618732e-05,
+      "loss": 0.213,
+      "step": 2125
+    },
+    {
+      "epoch": 3.4248892468787755,
+      "grad_norm": 0.2462846208675487,
+      "learning_rate": 1.0870061539797696e-05,
+      "loss": 0.1995,
+      "step": 2126
+    },
+    {
+      "epoch": 3.4265002013693113,
+      "grad_norm": 0.2367304617090815,
+      "learning_rate": 1.0850030389424724e-05,
+      "loss": 0.2052,
+      "step": 2127
+    },
+    {
+      "epoch": 3.428111155859847,
+      "grad_norm": 0.24464716596958344,
+      "learning_rate": 1.0830010840466282e-05,
+      "loss": 0.2172,
+      "step": 2128
+    },
+    {
+      "epoch": 3.4297221103503825,
+      "grad_norm": 0.24560443043574962,
+      "learning_rate": 1.0810002918305508e-05,
+      "loss": 0.2054,
+      "step": 2129
+    },
+    {
+      "epoch": 3.4313330648409184,
+      "grad_norm": 0.24409525550351965,
+      "learning_rate": 1.0790006648310828e-05,
+      "loss": 0.2041,
+      "step": 2130
+    },
+    {
+      "epoch": 3.4329440193314538,
+      "grad_norm": 0.23782893656758072,
+      "learning_rate": 1.0770022055835882e-05,
+      "loss": 0.2036,
+      "step": 2131
+    },
+    {
+      "epoch": 3.4345549738219896,
+      "grad_norm": 0.23124086461460552,
+      "learning_rate": 1.075004916621949e-05,
+      "loss": 0.199,
+      "step": 2132
+    },
+    {
+      "epoch": 3.436165928312525,
+      "grad_norm": 0.2438499354960223,
+      "learning_rate": 1.073008800478566e-05,
+      "loss": 0.1827,
+      "step": 2133
+    },
+    {
+      "epoch": 3.437776882803061,
+      "grad_norm": 0.23923418585538098,
+      "learning_rate": 1.0710138596843494e-05,
+      "loss": 0.2235,
+      "step": 2134
+    },
+    {
+      "epoch": 3.439387837293596,
+      "grad_norm": 0.23302741192753726,
+      "learning_rate": 1.0690200967687234e-05,
+      "loss": 0.1979,
+      "step": 2135
+    },
+    {
+      "epoch": 3.440998791784132,
+      "grad_norm": 0.23649073229436382,
+      "learning_rate": 1.0670275142596154e-05,
+      "loss": 0.2074,
+      "step": 2136
+    },
+    {
+      "epoch": 3.442609746274668,
+      "grad_norm": 0.23258425943151656,
+      "learning_rate": 1.065036114683458e-05,
+      "loss": 0.2043,
+      "step": 2137
+    },
+    {
+      "epoch": 3.4442207007652033,
+      "grad_norm": 0.23498177622326072,
+      "learning_rate": 1.0630459005651823e-05,
+      "loss": 0.2005,
+      "step": 2138
+    },
+    {
+      "epoch": 3.445831655255739,
+      "grad_norm": 0.26706198990505337,
+      "learning_rate": 1.061056874428219e-05,
+      "loss": 0.2026,
+      "step": 2139
+    },
+    {
+      "epoch": 3.4474426097462745,
+      "grad_norm": 0.23061635890286963,
+      "learning_rate": 1.059069038794489e-05,
+      "loss": 0.214,
+      "step": 2140
+    },
+    {
+      "epoch": 3.4490535642368103,
+      "grad_norm": 0.22470610502390026,
+      "learning_rate": 1.0570823961844065e-05,
+      "loss": 0.1919,
+      "step": 2141
+    },
+    {
+      "epoch": 3.450664518727346,
+      "grad_norm": 0.24074853759574122,
+      "learning_rate": 1.055096949116873e-05,
+      "loss": 0.2134,
+      "step": 2142
+    },
+    {
+      "epoch": 3.4522754732178815,
+      "grad_norm": 0.238304166701547,
+      "learning_rate": 1.0531127001092715e-05,
+      "loss": 0.2032,
+      "step": 2143
+    },
+    {
+      "epoch": 3.4538864277084174,
+      "grad_norm": 0.22940631731584865,
+      "learning_rate": 1.0511296516774699e-05,
+      "loss": 0.1904,
+      "step": 2144
+    },
+    {
+      "epoch": 3.4554973821989527,
+      "grad_norm": 0.24654104415440423,
+      "learning_rate": 1.0491478063358096e-05,
+      "loss": 0.2021,
+      "step": 2145
+    },
+    {
+      "epoch": 3.4571083366894886,
+      "grad_norm": 0.26612320358388847,
+      "learning_rate": 1.0471671665971104e-05,
+      "loss": 0.2139,
+      "step": 2146
+    },
+    {
+      "epoch": 3.4587192911800244,
+      "grad_norm": 0.40048278104943835,
+      "learning_rate": 1.0451877349726605e-05,
+      "loss": 0.2229,
+      "step": 2147
+    },
+    {
+      "epoch": 3.46033024567056,
+      "grad_norm": 0.23711292098323433,
+      "learning_rate": 1.0432095139722187e-05,
+      "loss": 0.1839,
+      "step": 2148
+    },
+    {
+      "epoch": 3.4619412001610956,
+      "grad_norm": 0.26048839006315294,
+      "learning_rate": 1.0412325061040063e-05,
+      "loss": 0.1968,
+      "step": 2149
+    },
+    {
+      "epoch": 3.463552154651631,
+      "grad_norm": 0.2326769954633612,
+      "learning_rate": 1.0392567138747101e-05,
+      "loss": 0.2178,
+      "step": 2150
+    },
+    {
+      "epoch": 3.465163109142167,
+      "grad_norm": 0.25045178314044103,
+      "learning_rate": 1.0372821397894709e-05,
+      "loss": 0.1879,
+      "step": 2151
+    },
+    {
+      "epoch": 3.4667740636327022,
+      "grad_norm": 0.23568688727213294,
+      "learning_rate": 1.0353087863518894e-05,
+      "loss": 0.1919,
+      "step": 2152
+    },
+    {
+      "epoch": 3.468385018123238,
+      "grad_norm": 0.2412245199550461,
+      "learning_rate": 1.0333366560640151e-05,
+      "loss": 0.1995,
+      "step": 2153
+    },
+    {
+      "epoch": 3.4699959726137735,
+      "grad_norm": 0.2634416881896074,
+      "learning_rate": 1.0313657514263492e-05,
+      "loss": 0.2353,
+      "step": 2154
+    },
+    {
+      "epoch": 3.4716069271043093,
+      "grad_norm": 0.22072632069118556,
+      "learning_rate": 1.0293960749378384e-05,
+      "loss": 0.1939,
+      "step": 2155
+    },
+    {
+      "epoch": 3.473217881594845,
+      "grad_norm": 0.23393038614986122,
+      "learning_rate": 1.0274276290958701e-05,
+      "loss": 0.2051,
+      "step": 2156
+    },
+    {
+      "epoch": 3.4748288360853805,
+      "grad_norm": 0.23009879755108045,
+      "learning_rate": 1.0254604163962747e-05,
+      "loss": 0.1869,
+      "step": 2157
+    },
+    {
+      "epoch": 3.4764397905759163,
+      "grad_norm": 0.24631286359057225,
+      "learning_rate": 1.0234944393333155e-05,
+      "loss": 0.2072,
+      "step": 2158
+    },
+    {
+      "epoch": 3.4780507450664517,
+      "grad_norm": 0.2396424841750826,
+      "learning_rate": 1.0215297003996927e-05,
+      "loss": 0.1985,
+      "step": 2159
+    },
+    {
+      "epoch": 3.4796616995569876,
+      "grad_norm": 0.2229192233390955,
+      "learning_rate": 1.0195662020865333e-05,
+      "loss": 0.1877,
+      "step": 2160
+    },
+    {
+      "epoch": 3.4812726540475234,
+      "grad_norm": 0.27640686696076017,
+      "learning_rate": 1.0176039468833938e-05,
+      "loss": 0.2072,
+      "step": 2161
+    },
+    {
+      "epoch": 3.482883608538059,
+      "grad_norm": 0.2501269581468127,
+      "learning_rate": 1.015642937278254e-05,
+      "loss": 0.2093,
+      "step": 2162
+    },
+    {
+      "epoch": 3.4844945630285946,
+      "grad_norm": 0.23553356609239753,
+      "learning_rate": 1.0136831757575134e-05,
+      "loss": 0.2097,
+      "step": 2163
+    },
+    {
+      "epoch": 3.48610551751913,
+      "grad_norm": 0.2513360823413565,
+      "learning_rate": 1.0117246648059888e-05,
+      "loss": 0.2027,
+      "step": 2164
+    },
+    {
+      "epoch": 3.487716472009666,
+      "grad_norm": 0.25559246873602354,
+      "learning_rate": 1.0097674069069132e-05,
+      "loss": 0.2042,
+      "step": 2165
+    },
+    {
+      "epoch": 3.4893274265002012,
+      "grad_norm": 0.24381340477783897,
+      "learning_rate": 1.0078114045419305e-05,
+      "loss": 0.2193,
+      "step": 2166
+    },
+    {
+      "epoch": 3.490938380990737,
+      "grad_norm": 0.23737314305360308,
+      "learning_rate": 1.0058566601910903e-05,
+      "loss": 0.2057,
+      "step": 2167
+    },
+    {
+      "epoch": 3.4925493354812724,
+      "grad_norm": 0.24161527730895105,
+      "learning_rate": 1.0039031763328506e-05,
+      "loss": 0.1997,
+      "step": 2168
+    },
+    {
+      "epoch": 3.4941602899718083,
+      "grad_norm": 0.24433075314270097,
+      "learning_rate": 1.0019509554440678e-05,
+      "loss": 0.2156,
+      "step": 2169
+    },
+    {
+      "epoch": 3.495771244462344,
+      "grad_norm": 0.2280058138376306,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.1945,
+      "step": 2170
+    },
+    {
+      "epoch": 3.4973821989528795,
+      "grad_norm": 0.24779688243659267,
+      "learning_rate": 9.980503124742988e-06,
+      "loss": 0.2093,
+      "step": 2171
+    },
+    {
+      "epoch": 3.4989931534434153,
+      "grad_norm": 0.2428801961690144,
+      "learning_rate": 9.961018953390086e-06,
+      "loss": 0.2065,
+      "step": 2172
+    },
+    {
+      "epoch": 3.5006041079339507,
+      "grad_norm": 0.2219328596608684,
+      "learning_rate": 9.941547510645647e-06,
+      "loss": 0.1885,
+      "step": 2173
+    },
+    {
+      "epoch": 3.5022150624244865,
+      "grad_norm": 0.2267440935335129,
+      "learning_rate": 9.922088821197854e-06,
+      "loss": 0.2089,
+      "step": 2174
+    },
+    {
+      "epoch": 3.5038260169150224,
+      "grad_norm": 0.24159131404047698,
+      "learning_rate": 9.902642909718737e-06,
+      "loss": 0.2101,
+      "step": 2175
+    },
+    {
+      "epoch": 3.5054369714055578,
+      "grad_norm": 0.23555256415137021,
+      "learning_rate": 9.88320980086413e-06,
+      "loss": 0.2021,
+      "step": 2176
+    },
+    {
+      "epoch": 3.5070479258960936,
+      "grad_norm": 0.2274219679214926,
+      "learning_rate": 9.863789519273638e-06,
+      "loss": 0.1952,
+      "step": 2177
+    },
+    {
+      "epoch": 3.508658880386629,
+      "grad_norm": 0.23187334643843013,
+      "learning_rate": 9.844382089570584e-06,
+      "loss": 0.2084,
+      "step": 2178
+    },
+    {
+      "epoch": 3.510269834877165,
+      "grad_norm": 0.22771801310950746,
+      "learning_rate": 9.824987536361992e-06,
+      "loss": 0.1949,
+      "step": 2179
+    },
+    {
+      "epoch": 3.5118807893677,
+      "grad_norm": 0.22370527590647052,
+      "learning_rate": 9.805605884238587e-06,
+      "loss": 0.205,
+      "step": 2180
+    },
+    {
+      "epoch": 3.513491743858236,
+      "grad_norm": 0.23845020170094708,
+      "learning_rate": 9.786237157774726e-06,
+      "loss": 0.2032,
+      "step": 2181
+    },
+    {
+      "epoch": 3.5151026983487714,
+      "grad_norm": 0.2283710304261633,
+      "learning_rate": 9.766881381528357e-06,
+      "loss": 0.2032,
+      "step": 2182
+    },
+    {
+      "epoch": 3.5167136528393073,
+      "grad_norm": 0.2378550286458637,
+      "learning_rate": 9.747538580041034e-06,
+      "loss": 0.2106,
+      "step": 2183
+    },
+    {
+      "epoch": 3.518324607329843,
+      "grad_norm": 0.22430006122712037,
+      "learning_rate": 9.728208777837858e-06,
+      "loss": 0.2024,
+      "step": 2184
+    },
+    {
+      "epoch": 3.5199355618203785,
+      "grad_norm": 0.22469949464520347,
+      "learning_rate": 9.70889199942743e-06,
+      "loss": 0.2076,
+      "step": 2185
+    },
+    {
+      "epoch": 3.5215465163109143,
+      "grad_norm": 0.2416906697350346,
+      "learning_rate": 9.689588269301842e-06,
+      "loss": 0.2171,
+      "step": 2186
+    },
+    {
+      "epoch": 3.5231574708014497,
+      "grad_norm": 0.22087168616579303,
+      "learning_rate": 9.670297611936657e-06,
+      "loss": 0.2009,
+      "step": 2187
+    },
+    {
+      "epoch": 3.5247684252919855,
+      "grad_norm": 0.23173269867776017,
+      "learning_rate": 9.651020051790858e-06,
+      "loss": 0.2073,
+      "step": 2188
+    },
+    {
+      "epoch": 3.5263793797825214,
+      "grad_norm": 0.2277755728185813,
+      "learning_rate": 9.631755613306814e-06,
+      "loss": 0.2008,
+      "step": 2189
+    },
+    {
+      "epoch": 3.5279903342730567,
+      "grad_norm": 0.22892461332859246,
+      "learning_rate": 9.612504320910249e-06,
+      "loss": 0.22,
+      "step": 2190
+    },
+    {
+      "epoch": 3.5296012887635926,
+      "grad_norm": 0.22441902310723352,
+      "learning_rate": 9.59326619901024e-06,
+      "loss": 0.2079,
+      "step": 2191
+    },
+    {
+      "epoch": 3.531212243254128,
+      "grad_norm": 0.2331194408439935,
+      "learning_rate": 9.574041271999163e-06,
+      "loss": 0.2019,
+      "step": 2192
+    },
+    {
+      "epoch": 3.532823197744664,
+      "grad_norm": 0.2317551080792225,
+      "learning_rate": 9.55482956425264e-06,
+      "loss": 0.2088,
+      "step": 2193
+    },
+    {
+      "epoch": 3.534434152235199,
+      "grad_norm": 0.24198116715159212,
+      "learning_rate": 9.535631100129556e-06,
+      "loss": 0.2047,
+      "step": 2194
+    },
+    {
+      "epoch": 3.536045106725735,
+      "grad_norm": 0.2404786868991303,
+      "learning_rate": 9.516445903972005e-06,
+      "loss": 0.203,
+      "step": 2195
+    },
+    {
+      "epoch": 3.5376560612162704,
+      "grad_norm": 0.21582358266853396,
+      "learning_rate": 9.497274000105239e-06,
+      "loss": 0.1868,
+      "step": 2196
+    },
+    {
+      "epoch": 3.5392670157068062,
+      "grad_norm": 0.24330479425323304,
+      "learning_rate": 9.478115412837661e-06,
+      "loss": 0.2046,
+      "step": 2197
+    },
+    {
+      "epoch": 3.540877970197342,
+      "grad_norm": 0.2290973058402262,
+      "learning_rate": 9.458970166460804e-06,
+      "loss": 0.1997,
+      "step": 2198
+    },
+    {
+      "epoch": 3.5424889246878775,
+      "grad_norm": 0.24487357902072016,
+      "learning_rate": 9.439838285249285e-06,
+      "loss": 0.2109,
+      "step": 2199
+    },
+    {
+      "epoch": 3.5440998791784133,
+      "grad_norm": 0.23275375857106345,
+      "learning_rate": 9.420719793460758e-06,
+      "loss": 0.2045,
+      "step": 2200
+    },
+    {
+      "epoch": 3.5457108336689487,
+      "grad_norm": 0.23599591116541208,
+      "learning_rate": 9.401614715335905e-06,
+      "loss": 0.1868,
+      "step": 2201
+    },
+    {
+      "epoch": 3.5473217881594845,
+      "grad_norm": 0.23362100236229594,
+      "learning_rate": 9.38252307509841e-06,
+      "loss": 0.193,
+      "step": 2202
+    },
+    {
+      "epoch": 3.5489327426500203,
+      "grad_norm": 0.24650562505944937,
+      "learning_rate": 9.363444896954927e-06,
+      "loss": 0.2109,
+      "step": 2203
+    },
+    {
+      "epoch": 3.5505436971405557,
+      "grad_norm": 0.23856000850206951,
+      "learning_rate": 9.344380205095017e-06,
+      "loss": 0.1888,
+      "step": 2204
+    },
+    {
+      "epoch": 3.5521546516310916,
+      "grad_norm": 0.23064058859091333,
+      "learning_rate": 9.325329023691137e-06,
+      "loss": 0.2069,
+      "step": 2205
+    },
+    {
+      "epoch": 3.553765606121627,
+      "grad_norm": 0.2318135856582349,
+      "learning_rate": 9.30629137689866e-06,
+      "loss": 0.2023,
+      "step": 2206
+    },
+    {
+      "epoch": 3.555376560612163,
+      "grad_norm": 0.24342035330502437,
+      "learning_rate": 9.287267288855756e-06,
+      "loss": 0.2067,
+      "step": 2207
+    },
+    {
+      "epoch": 3.5569875151026986,
+      "grad_norm": 0.25166127958235246,
+      "learning_rate": 9.268256783683408e-06,
+      "loss": 0.1903,
+      "step": 2208
+    },
+    {
+      "epoch": 3.558598469593234,
+      "grad_norm": 0.22256962504013172,
+      "learning_rate": 9.24925988548539e-06,
+      "loss": 0.1935,
+      "step": 2209
+    },
+    {
+      "epoch": 3.5602094240837694,
+      "grad_norm": 0.28219584567539696,
+      "learning_rate": 9.230276618348224e-06,
+      "loss": 0.2108,
+      "step": 2210
+    },
+    {
+      "epoch": 3.5618203785743052,
+      "grad_norm": 0.24218404180288422,
+      "learning_rate": 9.21130700634114e-06,
+      "loss": 0.2002,
+      "step": 2211
+    },
+    {
+      "epoch": 3.563431333064841,
+      "grad_norm": 0.24262274443668091,
+      "learning_rate": 9.192351073516047e-06,
+      "loss": 0.1875,
+      "step": 2212
+    },
+    {
+      "epoch": 3.5650422875553764,
+      "grad_norm": 0.2364784795121163,
+      "learning_rate": 9.173408843907529e-06,
+      "loss": 0.1873,
+      "step": 2213
+    },
+    {
+      "epoch": 3.5666532420459123,
+      "grad_norm": 0.2317621350392338,
+      "learning_rate": 9.154480341532797e-06,
+      "loss": 0.2008,
+      "step": 2214
+    },
+    {
+      "epoch": 3.5682641965364477,
+      "grad_norm": 0.24422946032063308,
+      "learning_rate": 9.135565590391633e-06,
+      "loss": 0.2241,
+      "step": 2215
+    },
+    {
+      "epoch": 3.5698751510269835,
+      "grad_norm": 0.23009545062501383,
+      "learning_rate": 9.116664614466386e-06,
+      "loss": 0.2057,
+      "step": 2216
+    },
+    {
+      "epoch": 3.5714861055175193,
+      "grad_norm": 0.23204535283854022,
+      "learning_rate": 9.097777437721982e-06,
+      "loss": 0.1953,
+      "step": 2217
+    },
+    {
+      "epoch": 3.5730970600080547,
+      "grad_norm": 0.24756670044336804,
+      "learning_rate": 9.078904084105802e-06,
+      "loss": 0.2071,
+      "step": 2218
+    },
+    {
+      "epoch": 3.5747080144985905,
+      "grad_norm": 0.22696668780158338,
+      "learning_rate": 9.060044577547711e-06,
+      "loss": 0.197,
+      "step": 2219
+    },
+    {
+      "epoch": 3.576318968989126,
+      "grad_norm": 0.21813119741031062,
+      "learning_rate": 9.04119894196003e-06,
+      "loss": 0.1943,
+      "step": 2220
+    },
+    {
+      "epoch": 3.5779299234796618,
+      "grad_norm": 0.24547436858026264,
+      "learning_rate": 9.022367201237495e-06,
+      "loss": 0.1908,
+      "step": 2221
+    },
+    {
+      "epoch": 3.5795408779701976,
+      "grad_norm": 0.24906258689712765,
+      "learning_rate": 9.00354937925721e-06,
+      "loss": 0.2125,
+      "step": 2222
+    },
+    {
+      "epoch": 3.581151832460733,
+      "grad_norm": 0.2358356210611667,
+      "learning_rate": 8.98474549987862e-06,
+      "loss": 0.2037,
+      "step": 2223
+    },
+    {
+      "epoch": 3.5827627869512684,
+      "grad_norm": 0.23154346550342686,
+      "learning_rate": 8.965955586943538e-06,
+      "loss": 0.1915,
+      "step": 2224
+    },
+    {
+      "epoch": 3.584373741441804,
+      "grad_norm": 0.23932511820741448,
+      "learning_rate": 8.947179664276028e-06,
+      "loss": 0.2049,
+      "step": 2225
+    },
+    {
+      "epoch": 3.58598469593234,
+      "grad_norm": 0.216916084634077,
+      "learning_rate": 8.928417755682416e-06,
+      "loss": 0.1896,
+      "step": 2226
+    },
+    {
+      "epoch": 3.5875956504228754,
+      "grad_norm": 0.2200896036801776,
+      "learning_rate": 8.909669884951284e-06,
+      "loss": 0.2,
+      "step": 2227
+    },
+    {
+      "epoch": 3.5892066049134113,
+      "grad_norm": 0.2485175178385473,
+      "learning_rate": 8.890936075853403e-06,
+      "loss": 0.2034,
+      "step": 2228
+    },
+    {
+      "epoch": 3.5908175594039466,
+      "grad_norm": 0.2178344656723329,
+      "learning_rate": 8.87221635214171e-06,
+      "loss": 0.1798,
+      "step": 2229
+    },
+    {
+      "epoch": 3.5924285138944825,
+      "grad_norm": 0.2302593584653443,
+      "learning_rate": 8.853510737551274e-06,
+      "loss": 0.2048,
+      "step": 2230
+    },
+    {
+      "epoch": 3.5940394683850183,
+      "grad_norm": 0.2195145608319814,
+      "learning_rate": 8.8348192557993e-06,
+      "loss": 0.189,
+      "step": 2231
+    },
+    {
+      "epoch": 3.5956504228755537,
+      "grad_norm": 0.2384324191704394,
+      "learning_rate": 8.816141930585067e-06,
+      "loss": 0.191,
+      "step": 2232
+    },
+    {
+      "epoch": 3.5972613773660895,
+      "grad_norm": 0.27318029242884734,
+      "learning_rate": 8.79747878558989e-06,
+      "loss": 0.2054,
+      "step": 2233
+    },
+    {
+      "epoch": 3.598872331856625,
+      "grad_norm": 0.2443948830306952,
+      "learning_rate": 8.778829844477099e-06,
+      "loss": 0.1963,
+      "step": 2234
+    },
+    {
+      "epoch": 3.6004832863471607,
+      "grad_norm": 0.24789244954904005,
+      "learning_rate": 8.76019513089206e-06,
+      "loss": 0.2092,
+      "step": 2235
+    },
+    {
+      "epoch": 3.6020942408376966,
+      "grad_norm": 0.24650887122937143,
+      "learning_rate": 8.741574668462053e-06,
+      "loss": 0.2046,
+      "step": 2236
+    },
+    {
+      "epoch": 3.603705195328232,
+      "grad_norm": 0.23165112489745265,
+      "learning_rate": 8.722968480796294e-06,
+      "loss": 0.1855,
+      "step": 2237
+    },
+    {
+      "epoch": 3.6053161498187674,
+      "grad_norm": 0.2576959078546253,
+      "learning_rate": 8.704376591485923e-06,
+      "loss": 0.2059,
+      "step": 2238
+    },
+    {
+      "epoch": 3.606927104309303,
+      "grad_norm": 0.23768867314811368,
+      "learning_rate": 8.685799024103942e-06,
+      "loss": 0.1906,
+      "step": 2239
+    },
+    {
+      "epoch": 3.608538058799839,
+      "grad_norm": 0.2395303763373646,
+      "learning_rate": 8.667235802205183e-06,
+      "loss": 0.1901,
+      "step": 2240
+    },
+    {
+      "epoch": 3.6101490132903744,
+      "grad_norm": 0.2413793337525353,
+      "learning_rate": 8.648686949326286e-06,
+      "loss": 0.1961,
+      "step": 2241
+    },
+    {
+      "epoch": 3.6117599677809102,
+      "grad_norm": 0.23591510967044874,
+      "learning_rate": 8.630152488985694e-06,
+      "loss": 0.2082,
+      "step": 2242
+    },
+    {
+      "epoch": 3.6133709222714456,
+      "grad_norm": 0.2368705104725338,
+      "learning_rate": 8.611632444683595e-06,
+      "loss": 0.2016,
+      "step": 2243
+    },
+    {
+      "epoch": 3.6149818767619815,
+      "grad_norm": 0.22430359624959098,
+      "learning_rate": 8.593126839901886e-06,
+      "loss": 0.2003,
+      "step": 2244
+    },
+    {
+      "epoch": 3.6165928312525173,
+      "grad_norm": 0.2328209779003292,
+      "learning_rate": 8.57463569810415e-06,
+      "loss": 0.2193,
+      "step": 2245
+    },
+    {
+      "epoch": 3.6182037857430527,
+      "grad_norm": 0.21765260530670116,
+      "learning_rate": 8.556159042735672e-06,
+      "loss": 0.1884,
+      "step": 2246
+    },
+    {
+      "epoch": 3.6198147402335885,
+      "grad_norm": 0.2373213750878368,
+      "learning_rate": 8.537696897223331e-06,
+      "loss": 0.1916,
+      "step": 2247
+    },
+    {
+      "epoch": 3.621425694724124,
+      "grad_norm": 0.23641497550916923,
+      "learning_rate": 8.519249284975611e-06,
+      "loss": 0.2159,
+      "step": 2248
+    },
+    {
+      "epoch": 3.6230366492146597,
+      "grad_norm": 0.23628799850862556,
+      "learning_rate": 8.500816229382584e-06,
+      "loss": 0.2091,
+      "step": 2249
+    },
+    {
+      "epoch": 3.6246476037051956,
+      "grad_norm": 0.2234944894492603,
+      "learning_rate": 8.482397753815872e-06,
+      "loss": 0.2012,
+      "step": 2250
+    },
+    {
+      "epoch": 3.626258558195731,
+      "grad_norm": 0.21626870143948573,
+      "learning_rate": 8.463993881628586e-06,
+      "loss": 0.1922,
+      "step": 2251
+    },
+    {
+      "epoch": 3.6278695126862663,
+      "grad_norm": 0.23755306979760044,
+      "learning_rate": 8.445604636155328e-06,
+      "loss": 0.2078,
+      "step": 2252
+    },
+    {
+      "epoch": 3.629480467176802,
+      "grad_norm": 0.24117810761942085,
+      "learning_rate": 8.427230040712171e-06,
+      "loss": 0.2085,
+      "step": 2253
+    },
+    {
+      "epoch": 3.631091421667338,
+      "grad_norm": 0.22854665517211442,
+      "learning_rate": 8.408870118596606e-06,
+      "loss": 0.1999,
+      "step": 2254
+    },
+    {
+      "epoch": 3.6327023761578734,
+      "grad_norm": 0.22293728621187936,
+      "learning_rate": 8.390524893087505e-06,
+      "loss": 0.1836,
+      "step": 2255
+    },
+    {
+      "epoch": 3.6343133306484092,
+      "grad_norm": 0.22707952398983391,
+      "learning_rate": 8.37219438744511e-06,
+      "loss": 0.1859,
+      "step": 2256
+    },
+    {
+      "epoch": 3.6359242851389446,
+      "grad_norm": 0.23511966018985556,
+      "learning_rate": 8.353878624911026e-06,
+      "loss": 0.2185,
+      "step": 2257
+    },
+    {
+      "epoch": 3.6375352396294804,
+      "grad_norm": 0.22703099897283688,
+      "learning_rate": 8.335577628708137e-06,
+      "loss": 0.1927,
+      "step": 2258
+    },
+    {
+      "epoch": 3.6391461941200163,
+      "grad_norm": 0.25017536611561936,
+      "learning_rate": 8.317291422040601e-06,
+      "loss": 0.2087,
+      "step": 2259
+    },
+    {
+      "epoch": 3.6407571486105517,
+      "grad_norm": 0.23049249283018605,
+      "learning_rate": 8.299020028093844e-06,
+      "loss": 0.202,
+      "step": 2260
+    },
+    {
+      "epoch": 3.6423681031010875,
+      "grad_norm": 0.2271488606251097,
+      "learning_rate": 8.28076347003451e-06,
+      "loss": 0.1997,
+      "step": 2261
+    },
+    {
+      "epoch": 3.643979057591623,
+      "grad_norm": 0.23883238541566312,
+      "learning_rate": 8.262521771010417e-06,
+      "loss": 0.2128,
+      "step": 2262
+    },
+    {
+      "epoch": 3.6455900120821587,
+      "grad_norm": 0.22315904417254107,
+      "learning_rate": 8.24429495415054e-06,
+      "loss": 0.1997,
+      "step": 2263
+    },
+    {
+      "epoch": 3.6472009665726945,
+      "grad_norm": 0.2234185102011281,
+      "learning_rate": 8.226083042565006e-06,
+      "loss": 0.2003,
+      "step": 2264
+    },
+    {
+      "epoch": 3.64881192106323,
+      "grad_norm": 0.23225689520155557,
+      "learning_rate": 8.207886059345034e-06,
+      "loss": 0.1995,
+      "step": 2265
+    },
+    {
+      "epoch": 3.6504228755537658,
+      "grad_norm": 0.23187295084042797,
+      "learning_rate": 8.189704027562913e-06,
+      "loss": 0.1775,
+      "step": 2266
+    },
+    {
+      "epoch": 3.652033830044301,
+      "grad_norm": 0.23529056309699636,
+      "learning_rate": 8.171536970271963e-06,
+      "loss": 0.1999,
+      "step": 2267
+    },
+    {
+      "epoch": 3.653644784534837,
+      "grad_norm": 0.23236799884409995,
+      "learning_rate": 8.153384910506539e-06,
+      "loss": 0.1871,
+      "step": 2268
+    },
+    {
+      "epoch": 3.6552557390253724,
+      "grad_norm": 0.23485228349145965,
+      "learning_rate": 8.135247871281977e-06,
+      "loss": 0.2062,
+      "step": 2269
+    },
+    {
+      "epoch": 3.656866693515908,
+      "grad_norm": 0.22646757698120465,
+      "learning_rate": 8.11712587559455e-06,
+      "loss": 0.1832,
+      "step": 2270
+    },
+    {
+      "epoch": 3.6584776480064436,
+      "grad_norm": 0.2514790956885834,
+      "learning_rate": 8.099018946421473e-06,
+      "loss": 0.2196,
+      "step": 2271
+    },
+    {
+      "epoch": 3.6600886024969794,
+      "grad_norm": 0.2340164808100686,
+      "learning_rate": 8.080927106720862e-06,
+      "loss": 0.2239,
+      "step": 2272
+    },
+    {
+      "epoch": 3.6616995569875153,
+      "grad_norm": 0.2238317436694524,
+      "learning_rate": 8.062850379431689e-06,
+      "loss": 0.1888,
+      "step": 2273
+    },
+    {
+      "epoch": 3.6633105114780506,
+      "grad_norm": 0.23788481343752213,
+      "learning_rate": 8.044788787473756e-06,
+      "loss": 0.2094,
+      "step": 2274
+    },
+    {
+      "epoch": 3.6649214659685865,
+      "grad_norm": 0.23426880337639244,
+      "learning_rate": 8.026742353747698e-06,
+      "loss": 0.1888,
+      "step": 2275
+    },
+    {
+      "epoch": 3.666532420459122,
+      "grad_norm": 0.23338079652138916,
+      "learning_rate": 8.008711101134928e-06,
+      "loss": 0.1917,
+      "step": 2276
+    },
+    {
+      "epoch": 3.6681433749496577,
+      "grad_norm": 0.22694613559354795,
+      "learning_rate": 7.99069505249759e-06,
+      "loss": 0.2077,
+      "step": 2277
+    },
+    {
+      "epoch": 3.6697543294401935,
+      "grad_norm": 0.23368251076797544,
+      "learning_rate": 7.972694230678562e-06,
+      "loss": 0.2019,
+      "step": 2278
+    },
+    {
+      "epoch": 3.671365283930729,
+      "grad_norm": 0.2274192071332042,
+      "learning_rate": 7.95470865850142e-06,
+      "loss": 0.2081,
+      "step": 2279
+    },
+    {
+      "epoch": 3.6729762384212647,
+      "grad_norm": 0.23272909394736677,
+      "learning_rate": 7.936738358770409e-06,
+      "loss": 0.2012,
+      "step": 2280
+    },
+    {
+      "epoch": 3.6745871929118,
+      "grad_norm": 0.2283629788794831,
+      "learning_rate": 7.918783354270388e-06,
+      "loss": 0.1951,
+      "step": 2281
+    },
+    {
+      "epoch": 3.676198147402336,
+      "grad_norm": 0.23382239466066568,
+      "learning_rate": 7.900843667766845e-06,
+      "loss": 0.2083,
+      "step": 2282
+    },
+    {
+      "epoch": 3.677809101892872,
+      "grad_norm": 0.22454604048146312,
+      "learning_rate": 7.882919322005844e-06,
+      "loss": 0.1948,
+      "step": 2283
+    },
+    {
+      "epoch": 3.679420056383407,
+      "grad_norm": 0.22453891515871524,
+      "learning_rate": 7.865010339713986e-06,
+      "loss": 0.1785,
+      "step": 2284
+    },
+    {
+      "epoch": 3.6810310108739426,
+      "grad_norm": 0.23166411493780942,
+      "learning_rate": 7.847116743598388e-06,
+      "loss": 0.2198,
+      "step": 2285
+    },
+    {
+      "epoch": 3.6826419653644784,
+      "grad_norm": 0.2347847692976649,
+      "learning_rate": 7.829238556346683e-06,
+      "loss": 0.1903,
+      "step": 2286
+    },
+    {
+      "epoch": 3.6842529198550142,
+      "grad_norm": 0.2388300143720348,
+      "learning_rate": 7.811375800626954e-06,
+      "loss": 0.2082,
+      "step": 2287
+    },
+    {
+      "epoch": 3.6858638743455496,
+      "grad_norm": 0.22396488198134365,
+      "learning_rate": 7.793528499087709e-06,
+      "loss": 0.2033,
+      "step": 2288
+    },
+    {
+      "epoch": 3.6874748288360855,
+      "grad_norm": 0.2336053446605039,
+      "learning_rate": 7.775696674357876e-06,
+      "loss": 0.2105,
+      "step": 2289
+    },
+    {
+      "epoch": 3.689085783326621,
+      "grad_norm": 0.23591046953244488,
+      "learning_rate": 7.757880349046742e-06,
+      "loss": 0.1914,
+      "step": 2290
+    },
+    {
+      "epoch": 3.6906967378171567,
+      "grad_norm": 0.23748095616700868,
+      "learning_rate": 7.74007954574397e-06,
+      "loss": 0.2083,
+      "step": 2291
+    },
+    {
+      "epoch": 3.6923076923076925,
+      "grad_norm": 0.2481715938640332,
+      "learning_rate": 7.722294287019509e-06,
+      "loss": 0.1945,
+      "step": 2292
+    },
+    {
+      "epoch": 3.693918646798228,
+      "grad_norm": 0.24850684859956976,
+      "learning_rate": 7.704524595423631e-06,
+      "loss": 0.2224,
+      "step": 2293
+    },
+    {
+      "epoch": 3.6955296012887637,
+      "grad_norm": 0.22449191890739228,
+      "learning_rate": 7.686770493486835e-06,
+      "loss": 0.1917,
+      "step": 2294
+    },
+    {
+      "epoch": 3.697140555779299,
+      "grad_norm": 0.22240414266437813,
+      "learning_rate": 7.669032003719894e-06,
+      "loss": 0.2043,
+      "step": 2295
+    },
+    {
+      "epoch": 3.698751510269835,
+      "grad_norm": 0.23811833779559768,
+      "learning_rate": 7.651309148613745e-06,
+      "loss": 0.2228,
+      "step": 2296
+    },
+    {
+      "epoch": 3.700362464760371,
+      "grad_norm": 0.2346602801883088,
+      "learning_rate": 7.633601950639532e-06,
+      "loss": 0.2044,
+      "step": 2297
+    },
+    {
+      "epoch": 3.701973419250906,
+      "grad_norm": 0.24881467892424516,
+      "learning_rate": 7.615910432248546e-06,
+      "loss": 0.1971,
+      "step": 2298
+    },
+    {
+      "epoch": 3.7035843737414416,
+      "grad_norm": 0.21959606461798073,
+      "learning_rate": 7.598234615872169e-06,
+      "loss": 0.1723,
+      "step": 2299
+    },
+    {
+      "epoch": 3.7051953282319774,
+      "grad_norm": 0.2275339064144191,
+      "learning_rate": 7.580574523921906e-06,
+      "loss": 0.1943,
+      "step": 2300
+    },
+    {
+      "epoch": 3.7068062827225132,
+      "grad_norm": 0.2638835977196525,
+      "learning_rate": 7.562930178789305e-06,
+      "loss": 0.225,
+      "step": 2301
+    },
+    {
+      "epoch": 3.7084172372130486,
+      "grad_norm": 0.22139223269245323,
+      "learning_rate": 7.545301602845965e-06,
+      "loss": 0.1974,
+      "step": 2302
+    },
+    {
+      "epoch": 3.7100281917035844,
+      "grad_norm": 0.23262025318923577,
+      "learning_rate": 7.527688818443466e-06,
+      "loss": 0.1918,
+      "step": 2303
+    },
+    {
+      "epoch": 3.71163914619412,
+      "grad_norm": 0.2281510665207774,
+      "learning_rate": 7.510091847913396e-06,
+      "loss": 0.1827,
+      "step": 2304
+    },
+    {
+      "epoch": 3.7132501006846557,
+      "grad_norm": 0.2315480501775712,
+      "learning_rate": 7.492510713567265e-06,
+      "loss": 0.2254,
+      "step": 2305
+    },
+    {
+      "epoch": 3.7148610551751915,
+      "grad_norm": 0.22423715853143875,
+      "learning_rate": 7.474945437696528e-06,
+      "loss": 0.18,
+      "step": 2306
+    },
+    {
+      "epoch": 3.716472009665727,
+      "grad_norm": 0.23128163413988403,
+      "learning_rate": 7.457396042572507e-06,
+      "loss": 0.2008,
+      "step": 2307
+    },
+    {
+      "epoch": 3.7180829641562627,
+      "grad_norm": 0.23426184961659158,
+      "learning_rate": 7.4398625504464105e-06,
+      "loss": 0.195,
+      "step": 2308
+    },
+    {
+      "epoch": 3.719693918646798,
+      "grad_norm": 0.2242770154556143,
+      "learning_rate": 7.422344983549281e-06,
+      "loss": 0.1905,
+      "step": 2309
+    },
+    {
+      "epoch": 3.721304873137334,
+      "grad_norm": 0.22078020401371204,
+      "learning_rate": 7.404843364091951e-06,
+      "loss": 0.1849,
+      "step": 2310
+    },
+    {
+      "epoch": 3.7229158276278698,
+      "grad_norm": 0.2319635828085481,
+      "learning_rate": 7.387357714265057e-06,
+      "loss": 0.1882,
+      "step": 2311
+    },
+    {
+      "epoch": 3.724526782118405,
+      "grad_norm": 0.245118103968261,
+      "learning_rate": 7.369888056238963e-06,
+      "loss": 0.2013,
+      "step": 2312
+    },
+    {
+      "epoch": 3.7261377366089405,
+      "grad_norm": 0.22519823792644417,
+      "learning_rate": 7.352434412163785e-06,
+      "loss": 0.2004,
+      "step": 2313
+    },
+    {
+      "epoch": 3.7277486910994764,
+      "grad_norm": 0.21704653831383208,
+      "learning_rate": 7.334996804169301e-06,
+      "loss": 0.1881,
+      "step": 2314
+    },
+    {
+      "epoch": 3.729359645590012,
+      "grad_norm": 0.22862567502117337,
+      "learning_rate": 7.31757525436499e-06,
+      "loss": 0.2199,
+      "step": 2315
+    },
+    {
+      "epoch": 3.7309706000805476,
+      "grad_norm": 0.21180370521549158,
+      "learning_rate": 7.300169784839941e-06,
+      "loss": 0.1748,
+      "step": 2316
+    },
+    {
+      "epoch": 3.7325815545710834,
+      "grad_norm": 0.24645022888252976,
+      "learning_rate": 7.282780417662885e-06,
+      "loss": 0.2207,
+      "step": 2317
+    },
+    {
+      "epoch": 3.734192509061619,
+      "grad_norm": 0.21457737876895103,
+      "learning_rate": 7.265407174882102e-06,
+      "loss": 0.1932,
+      "step": 2318
+    },
+    {
+      "epoch": 3.7358034635521546,
+      "grad_norm": 0.2338070679618665,
+      "learning_rate": 7.248050078525463e-06,
+      "loss": 0.1986,
+      "step": 2319
+    },
+    {
+      "epoch": 3.7374144180426905,
+      "grad_norm": 0.23031422564032217,
+      "learning_rate": 7.2307091506003325e-06,
+      "loss": 0.1966,
+      "step": 2320
+    },
+    {
+      "epoch": 3.739025372533226,
+      "grad_norm": 0.23174071582081254,
+      "learning_rate": 7.2133844130936e-06,
+      "loss": 0.2106,
+      "step": 2321
+    },
+    {
+      "epoch": 3.7406363270237617,
+      "grad_norm": 0.22866751182170306,
+      "learning_rate": 7.1960758879716255e-06,
+      "loss": 0.1961,
+      "step": 2322
+    },
+    {
+      "epoch": 3.742247281514297,
+      "grad_norm": 0.24817377490035214,
+      "learning_rate": 7.178783597180192e-06,
+      "loss": 0.2104,
+      "step": 2323
+    },
+    {
+      "epoch": 3.743858236004833,
+      "grad_norm": 0.21967760847457896,
+      "learning_rate": 7.161507562644525e-06,
+      "loss": 0.1868,
+      "step": 2324
+    },
+    {
+      "epoch": 3.7454691904953687,
+      "grad_norm": 0.2256324233021988,
+      "learning_rate": 7.1442478062692135e-06,
+      "loss": 0.217,
+      "step": 2325
+    },
+    {
+      "epoch": 3.747080144985904,
+      "grad_norm": 0.22260244060453915,
+      "learning_rate": 7.127004349938234e-06,
+      "loss": 0.2126,
+      "step": 2326
+    },
+    {
+      "epoch": 3.7486910994764395,
+      "grad_norm": 0.23048377400215084,
+      "learning_rate": 7.109777215514866e-06,
+      "loss": 0.2021,
+      "step": 2327
+    },
+    {
+      "epoch": 3.7503020539669754,
+      "grad_norm": 0.22938279327652358,
+      "learning_rate": 7.092566424841724e-06,
+      "loss": 0.2079,
+      "step": 2328
+    },
+    {
+      "epoch": 3.751913008457511,
+      "grad_norm": 0.22136550368979707,
+      "learning_rate": 7.0753719997406725e-06,
+      "loss": 0.1841,
+      "step": 2329
+    },
+    {
+      "epoch": 3.7535239629480466,
+      "grad_norm": 0.24306246271158838,
+      "learning_rate": 7.0581939620128515e-06,
+      "loss": 0.1993,
+      "step": 2330
+    },
+    {
+      "epoch": 3.7551349174385824,
+      "grad_norm": 0.2292465000453504,
+      "learning_rate": 7.0410323334386e-06,
+      "loss": 0.1881,
+      "step": 2331
+    },
+    {
+      "epoch": 3.756745871929118,
+      "grad_norm": 0.24266774620522708,
+      "learning_rate": 7.0238871357774655e-06,
+      "loss": 0.2136,
+      "step": 2332
+    },
+    {
+      "epoch": 3.7583568264196536,
+      "grad_norm": 0.22008050143973384,
+      "learning_rate": 7.0067583907681645e-06,
+      "loss": 0.1973,
+      "step": 2333
+    },
+    {
+      "epoch": 3.7599677809101895,
+      "grad_norm": 0.21463510410453734,
+      "learning_rate": 6.989646120128537e-06,
+      "loss": 0.1983,
+      "step": 2334
+    },
+    {
+      "epoch": 3.761578735400725,
+      "grad_norm": 0.23707104664701026,
+      "learning_rate": 6.97255034555556e-06,
+      "loss": 0.2136,
+      "step": 2335
+    },
+    {
+      "epoch": 3.7631896898912607,
+      "grad_norm": 0.21494086360905443,
+      "learning_rate": 6.955471088725263e-06,
+      "loss": 0.1896,
+      "step": 2336
+    },
+    {
+      "epoch": 3.764800644381796,
+      "grad_norm": 0.22994725268735505,
+      "learning_rate": 6.938408371292764e-06,
+      "loss": 0.1971,
+      "step": 2337
+    },
+    {
+      "epoch": 3.766411598872332,
+      "grad_norm": 0.23686841356483834,
+      "learning_rate": 6.921362214892182e-06,
+      "loss": 0.2062,
+      "step": 2338
+    },
+    {
+      "epoch": 3.7680225533628677,
+      "grad_norm": 0.21759254488600177,
+      "learning_rate": 6.904332641136668e-06,
+      "loss": 0.1889,
+      "step": 2339
+    },
+    {
+      "epoch": 3.769633507853403,
+      "grad_norm": 0.23896380350462712,
+      "learning_rate": 6.887319671618315e-06,
+      "loss": 0.2313,
+      "step": 2340
+    },
+    {
+      "epoch": 3.7712444623439385,
+      "grad_norm": 0.22446891923766324,
+      "learning_rate": 6.870323327908193e-06,
+      "loss": 0.2126,
+      "step": 2341
+    },
+    {
+      "epoch": 3.7728554168344743,
+      "grad_norm": 0.2192309543722886,
+      "learning_rate": 6.8533436315562665e-06,
+      "loss": 0.2035,
+      "step": 2342
+    },
+    {
+      "epoch": 3.77446637132501,
+      "grad_norm": 0.22492224295878233,
+      "learning_rate": 6.836380604091411e-06,
+      "loss": 0.1933,
+      "step": 2343
+    },
+    {
+      "epoch": 3.7760773258155456,
+      "grad_norm": 0.23367721504266944,
+      "learning_rate": 6.819434267021366e-06,
+      "loss": 0.207,
+      "step": 2344
+    },
+    {
+      "epoch": 3.7776882803060814,
+      "grad_norm": 0.21684454903299977,
+      "learning_rate": 6.80250464183269e-06,
+      "loss": 0.1925,
+      "step": 2345
+    },
+    {
+      "epoch": 3.779299234796617,
+      "grad_norm": 0.2220089960826439,
+      "learning_rate": 6.785591749990779e-06,
+      "loss": 0.1939,
+      "step": 2346
+    },
+    {
+      "epoch": 3.7809101892871526,
+      "grad_norm": 0.2361697872498856,
+      "learning_rate": 6.768695612939789e-06,
+      "loss": 0.2059,
+      "step": 2347
+    },
+    {
+      "epoch": 3.7825211437776884,
+      "grad_norm": 0.21291861216027663,
+      "learning_rate": 6.751816252102652e-06,
+      "loss": 0.1933,
+      "step": 2348
+    },
+    {
+      "epoch": 3.784132098268224,
+      "grad_norm": 0.22876029211036183,
+      "learning_rate": 6.734953688881007e-06,
+      "loss": 0.2097,
+      "step": 2349
+    },
+    {
+      "epoch": 3.7857430527587597,
+      "grad_norm": 0.23320502678881608,
+      "learning_rate": 6.7181079446552165e-06,
+      "loss": 0.1858,
+      "step": 2350
+    },
+    {
+      "epoch": 3.787354007249295,
+      "grad_norm": 0.23397548643850805,
+      "learning_rate": 6.701279040784314e-06,
+      "loss": 0.2063,
+      "step": 2351
+    },
+    {
+      "epoch": 3.788964961739831,
+      "grad_norm": 0.221465934796803,
+      "learning_rate": 6.68446699860597e-06,
+      "loss": 0.2066,
+      "step": 2352
+    },
+    {
+      "epoch": 3.7905759162303667,
+      "grad_norm": 0.24482731312630718,
+      "learning_rate": 6.667671839436473e-06,
+      "loss": 0.2112,
+      "step": 2353
+    },
+    {
+      "epoch": 3.792186870720902,
+      "grad_norm": 0.23761431330849628,
+      "learning_rate": 6.650893584570724e-06,
+      "loss": 0.2135,
+      "step": 2354
+    },
+    {
+      "epoch": 3.793797825211438,
+      "grad_norm": 0.23123489562147165,
+      "learning_rate": 6.634132255282182e-06,
+      "loss": 0.1959,
+      "step": 2355
+    },
+    {
+      "epoch": 3.7954087797019733,
+      "grad_norm": 0.22293241276120573,
+      "learning_rate": 6.617387872822842e-06,
+      "loss": 0.1976,
+      "step": 2356
+    },
+    {
+      "epoch": 3.797019734192509,
+      "grad_norm": 0.23135387901728863,
+      "learning_rate": 6.6006604584232044e-06,
+      "loss": 0.1956,
+      "step": 2357
+    },
+    {
+      "epoch": 3.7986306886830445,
+      "grad_norm": 0.2325139006331464,
+      "learning_rate": 6.583950033292277e-06,
+      "loss": 0.208,
+      "step": 2358
+    },
+    {
+      "epoch": 3.8002416431735804,
+      "grad_norm": 0.233386229971809,
+      "learning_rate": 6.567256618617515e-06,
+      "loss": 0.2013,
+      "step": 2359
+    },
+    {
+      "epoch": 3.8018525976641158,
+      "grad_norm": 0.2191968818040746,
+      "learning_rate": 6.550580235564794e-06,
+      "loss": 0.1866,
+      "step": 2360
+    },
+    {
+      "epoch": 3.8034635521546516,
+      "grad_norm": 0.23952247025452705,
+      "learning_rate": 6.533920905278415e-06,
+      "loss": 0.1995,
+      "step": 2361
+    },
+    {
+      "epoch": 3.8050745066451874,
+      "grad_norm": 0.21794716771682257,
+      "learning_rate": 6.517278648881054e-06,
+      "loss": 0.1924,
+      "step": 2362
+    },
+    {
+      "epoch": 3.806685461135723,
+      "grad_norm": 0.24189326859388585,
+      "learning_rate": 6.500653487473727e-06,
+      "loss": 0.197,
+      "step": 2363
+    },
+    {
+      "epoch": 3.8082964156262586,
+      "grad_norm": 0.22118662032379693,
+      "learning_rate": 6.4840454421357755e-06,
+      "loss": 0.1838,
+      "step": 2364
+    },
+    {
+      "epoch": 3.809907370116794,
+      "grad_norm": 0.23506496434883922,
+      "learning_rate": 6.46745453392485e-06,
+      "loss": 0.2199,
+      "step": 2365
+    },
+    {
+      "epoch": 3.81151832460733,
+      "grad_norm": 0.2226823424824965,
+      "learning_rate": 6.450880783876878e-06,
+      "loss": 0.2025,
+      "step": 2366
+    },
+    {
+      "epoch": 3.8131292790978657,
+      "grad_norm": 0.22964938395731235,
+      "learning_rate": 6.434324213006013e-06,
+      "loss": 0.1831,
+      "step": 2367
+    },
+    {
+      "epoch": 3.814740233588401,
+      "grad_norm": 0.24890163240709035,
+      "learning_rate": 6.417784842304628e-06,
+      "loss": 0.2056,
+      "step": 2368
+    },
+    {
+      "epoch": 3.816351188078937,
+      "grad_norm": 0.23386480273146806,
+      "learning_rate": 6.401262692743302e-06,
+      "loss": 0.2001,
+      "step": 2369
+    },
+    {
+      "epoch": 3.8179621425694723,
+      "grad_norm": 0.23767647586351798,
+      "learning_rate": 6.384757785270777e-06,
+      "loss": 0.1918,
+      "step": 2370
+    },
+    {
+      "epoch": 3.819573097060008,
+      "grad_norm": 0.2378296385486382,
+      "learning_rate": 6.368270140813917e-06,
+      "loss": 0.2073,
+      "step": 2371
+    },
+    {
+      "epoch": 3.821184051550544,
+      "grad_norm": 0.24265218985960949,
+      "learning_rate": 6.351799780277716e-06,
+      "loss": 0.2087,
+      "step": 2372
+    },
+    {
+      "epoch": 3.8227950060410794,
+      "grad_norm": 0.21839384113148216,
+      "learning_rate": 6.335346724545255e-06,
+      "loss": 0.203,
+      "step": 2373
+    },
+    {
+      "epoch": 3.8244059605316147,
+      "grad_norm": 0.21506408724015144,
+      "learning_rate": 6.318910994477654e-06,
+      "loss": 0.1757,
+      "step": 2374
+    },
+    {
+      "epoch": 3.8260169150221506,
+      "grad_norm": 0.2272335909081642,
+      "learning_rate": 6.3024926109140725e-06,
+      "loss": 0.2218,
+      "step": 2375
+    },
+    {
+      "epoch": 3.8276278695126864,
+      "grad_norm": 0.22244961279714315,
+      "learning_rate": 6.286091594671688e-06,
+      "loss": 0.1857,
+      "step": 2376
+    },
+    {
+      "epoch": 3.829238824003222,
+      "grad_norm": 0.2588268718136773,
+      "learning_rate": 6.2697079665456575e-06,
+      "loss": 0.2055,
+      "step": 2377
+    },
+    {
+      "epoch": 3.8308497784937576,
+      "grad_norm": 0.229250689941906,
+      "learning_rate": 6.253341747309076e-06,
+      "loss": 0.1985,
+      "step": 2378
+    },
+    {
+      "epoch": 3.832460732984293,
+      "grad_norm": 0.22729115522993168,
+      "learning_rate": 6.236992957712968e-06,
+      "loss": 0.1987,
+      "step": 2379
+    },
+    {
+      "epoch": 3.834071687474829,
+      "grad_norm": 0.2203706251189883,
+      "learning_rate": 6.220661618486268e-06,
+      "loss": 0.2028,
+      "step": 2380
+    },
+    {
+      "epoch": 3.8356826419653647,
+      "grad_norm": 0.23167585742706767,
+      "learning_rate": 6.204347750335791e-06,
+      "loss": 0.2056,
+      "step": 2381
+    },
+    {
+      "epoch": 3.8372935964559,
+      "grad_norm": 0.22942562541407732,
+      "learning_rate": 6.188051373946182e-06,
+      "loss": 0.1891,
+      "step": 2382
+    },
+    {
+      "epoch": 3.838904550946436,
+      "grad_norm": 0.22056932768319534,
+      "learning_rate": 6.171772509979903e-06,
+      "loss": 0.2052,
+      "step": 2383
+    },
+    {
+      "epoch": 3.8405155054369713,
+      "grad_norm": 0.22982020120028385,
+      "learning_rate": 6.155511179077252e-06,
+      "loss": 0.1968,
+      "step": 2384
+    },
+    {
+      "epoch": 3.842126459927507,
+      "grad_norm": 0.24493108816991999,
+      "learning_rate": 6.1392674018562525e-06,
+      "loss": 0.2133,
+      "step": 2385
+    },
+    {
+      "epoch": 3.843737414418043,
+      "grad_norm": 0.24068488930223084,
+      "learning_rate": 6.1230411989126825e-06,
+      "loss": 0.2114,
+      "step": 2386
+    },
+    {
+      "epoch": 3.8453483689085783,
+      "grad_norm": 0.20980965674211466,
+      "learning_rate": 6.106832590820053e-06,
+      "loss": 0.1814,
+      "step": 2387
+    },
+    {
+      "epoch": 3.8469593233991137,
+      "grad_norm": 0.22052853718987608,
+      "learning_rate": 6.090641598129559e-06,
+      "loss": 0.1941,
+      "step": 2388
+    },
+    {
+      "epoch": 3.8485702778896496,
+      "grad_norm": 0.24442699536324203,
+      "learning_rate": 6.074468241370053e-06,
+      "loss": 0.2184,
+      "step": 2389
+    },
+    {
+      "epoch": 3.8501812323801854,
+      "grad_norm": 0.22931064215122915,
+      "learning_rate": 6.058312541048021e-06,
+      "loss": 0.1814,
+      "step": 2390
+    },
+    {
+      "epoch": 3.851792186870721,
+      "grad_norm": 0.21834781868614564,
+      "learning_rate": 6.042174517647583e-06,
+      "loss": 0.2028,
+      "step": 2391
+    },
+    {
+      "epoch": 3.8534031413612566,
+      "grad_norm": 0.23946089513989083,
+      "learning_rate": 6.026054191630439e-06,
+      "loss": 0.1887,
+      "step": 2392
+    },
+    {
+      "epoch": 3.855014095851792,
+      "grad_norm": 0.22970941758659105,
+      "learning_rate": 6.00995158343584e-06,
+      "loss": 0.2213,
+      "step": 2393
+    },
+    {
+      "epoch": 3.856625050342328,
+      "grad_norm": 0.2314117292257891,
+      "learning_rate": 5.993866713480567e-06,
+      "loss": 0.2038,
+      "step": 2394
+    },
+    {
+      "epoch": 3.8582360048328637,
+      "grad_norm": 0.2439904102574889,
+      "learning_rate": 5.977799602158949e-06,
+      "loss": 0.2155,
+      "step": 2395
+    },
+    {
+      "epoch": 3.859846959323399,
+      "grad_norm": 0.20257437837005451,
+      "learning_rate": 5.961750269842754e-06,
+      "loss": 0.1751,
+      "step": 2396
+    },
+    {
+      "epoch": 3.861457913813935,
+      "grad_norm": 0.22955556617831127,
+      "learning_rate": 5.945718736881225e-06,
+      "loss": 0.2154,
+      "step": 2397
+    },
+    {
+      "epoch": 3.8630688683044703,
+      "grad_norm": 0.20870764713948106,
+      "learning_rate": 5.929705023601038e-06,
+      "loss": 0.1858,
+      "step": 2398
+    },
+    {
+      "epoch": 3.864679822795006,
+      "grad_norm": 0.22806265886533345,
+      "learning_rate": 5.913709150306284e-06,
+      "loss": 0.1966,
+      "step": 2399
+    },
+    {
+      "epoch": 3.866290777285542,
+      "grad_norm": 0.21759470538337408,
+      "learning_rate": 5.897731137278417e-06,
+      "loss": 0.206,
+      "step": 2400
+    },
+    {
+      "epoch": 3.8679017317760773,
+      "grad_norm": 0.22726788817470528,
+      "learning_rate": 5.881771004776243e-06,
+      "loss": 0.218,
+      "step": 2401
+    },
+    {
+      "epoch": 3.8695126862666127,
+      "grad_norm": 0.223006978113695,
+      "learning_rate": 5.865828773035927e-06,
+      "loss": 0.2127,
+      "step": 2402
+    },
+    {
+      "epoch": 3.8711236407571485,
+      "grad_norm": 0.2113739938673185,
+      "learning_rate": 5.849904462270908e-06,
+      "loss": 0.1758,
+      "step": 2403
+    },
+    {
+      "epoch": 3.8727345952476844,
+      "grad_norm": 0.2209374461562865,
+      "learning_rate": 5.833998092671911e-06,
+      "loss": 0.1949,
+      "step": 2404
+    },
+    {
+      "epoch": 3.8743455497382198,
+      "grad_norm": 0.21472108401460543,
+      "learning_rate": 5.8181096844069055e-06,
+      "loss": 0.21,
+      "step": 2405
+    },
+    {
+      "epoch": 3.8759565042287556,
+      "grad_norm": 0.2263705045461306,
+      "learning_rate": 5.802239257621116e-06,
+      "loss": 0.2014,
+      "step": 2406
+    },
+    {
+      "epoch": 3.877567458719291,
+      "grad_norm": 0.2206794394457454,
+      "learning_rate": 5.786386832436938e-06,
+      "loss": 0.1958,
+      "step": 2407
+    },
+    {
+      "epoch": 3.879178413209827,
+      "grad_norm": 0.21542639161588065,
+      "learning_rate": 5.770552428953946e-06,
+      "loss": 0.205,
+      "step": 2408
+    },
+    {
+      "epoch": 3.8807893677003626,
+      "grad_norm": 0.21783160109169053,
+      "learning_rate": 5.754736067248878e-06,
+      "loss": 0.2038,
+      "step": 2409
+    },
+    {
+      "epoch": 3.882400322190898,
+      "grad_norm": 0.22638816278271062,
+      "learning_rate": 5.738937767375596e-06,
+      "loss": 0.2086,
+      "step": 2410
+    },
+    {
+      "epoch": 3.884011276681434,
+      "grad_norm": 0.23034123131475734,
+      "learning_rate": 5.723157549365046e-06,
+      "loss": 0.1996,
+      "step": 2411
+    },
+    {
+      "epoch": 3.8856222311719693,
+      "grad_norm": 0.2403842941222503,
+      "learning_rate": 5.707395433225247e-06,
+      "loss": 0.1928,
+      "step": 2412
+    },
+    {
+      "epoch": 3.887233185662505,
+      "grad_norm": 0.22829443560895596,
+      "learning_rate": 5.691651438941297e-06,
+      "loss": 0.1937,
+      "step": 2413
+    },
+    {
+      "epoch": 3.888844140153041,
+      "grad_norm": 0.22379900687151044,
+      "learning_rate": 5.675925586475286e-06,
+      "loss": 0.2068,
+      "step": 2414
+    },
+    {
+      "epoch": 3.8904550946435763,
+      "grad_norm": 0.21890279750969147,
+      "learning_rate": 5.660217895766302e-06,
+      "loss": 0.2006,
+      "step": 2415
+    },
+    {
+      "epoch": 3.8920660491341117,
+      "grad_norm": 0.24292430296852172,
+      "learning_rate": 5.644528386730424e-06,
+      "loss": 0.2124,
+      "step": 2416
+    },
+    {
+      "epoch": 3.8936770036246475,
+      "grad_norm": 0.23023882143912577,
+      "learning_rate": 5.628857079260672e-06,
+      "loss": 0.1948,
+      "step": 2417
+    },
+    {
+      "epoch": 3.8952879581151834,
+      "grad_norm": 0.2182554614510497,
+      "learning_rate": 5.613203993226981e-06,
+      "loss": 0.1901,
+      "step": 2418
+    },
+    {
+      "epoch": 3.8968989126057187,
+      "grad_norm": 0.2219288373965724,
+      "learning_rate": 5.597569148476178e-06,
+      "loss": 0.2212,
+      "step": 2419
+    },
+    {
+      "epoch": 3.8985098670962546,
+      "grad_norm": 0.21537709000847655,
+      "learning_rate": 5.581952564831978e-06,
+      "loss": 0.2052,
+      "step": 2420
+    },
+    {
+      "epoch": 3.90012082158679,
+      "grad_norm": 0.22629553986166404,
+      "learning_rate": 5.56635426209494e-06,
+      "loss": 0.1981,
+      "step": 2421
+    },
+    {
+      "epoch": 3.901731776077326,
+      "grad_norm": 0.22348100651315547,
+      "learning_rate": 5.550774260042428e-06,
+      "loss": 0.2071,
+      "step": 2422
+    },
+    {
+      "epoch": 3.9033427305678616,
+      "grad_norm": 0.22867321367174948,
+      "learning_rate": 5.5352125784286085e-06,
+      "loss": 0.2048,
+      "step": 2423
+    },
+    {
+      "epoch": 3.904953685058397,
+      "grad_norm": 0.23348086911174862,
+      "learning_rate": 5.519669236984442e-06,
+      "loss": 0.2142,
+      "step": 2424
+    },
+    {
+      "epoch": 3.906564639548933,
+      "grad_norm": 0.21081715913992716,
+      "learning_rate": 5.504144255417605e-06,
+      "loss": 0.1903,
+      "step": 2425
+    },
+    {
+      "epoch": 3.9081755940394682,
+      "grad_norm": 0.22628394919950884,
+      "learning_rate": 5.488637653412501e-06,
+      "loss": 0.1932,
+      "step": 2426
+    },
+    {
+      "epoch": 3.909786548530004,
+      "grad_norm": 0.24684525974817506,
+      "learning_rate": 5.473149450630242e-06,
+      "loss": 0.2193,
+      "step": 2427
+    },
+    {
+      "epoch": 3.91139750302054,
+      "grad_norm": 0.22045186763448354,
+      "learning_rate": 5.4576796667086125e-06,
+      "loss": 0.1781,
+      "step": 2428
+    },
+    {
+      "epoch": 3.9130084575110753,
+      "grad_norm": 0.2120017802530591,
+      "learning_rate": 5.442228321262029e-06,
+      "loss": 0.2007,
+      "step": 2429
+    },
+    {
+      "epoch": 3.914619412001611,
+      "grad_norm": 0.23287109451180657,
+      "learning_rate": 5.426795433881527e-06,
+      "loss": 0.206,
+      "step": 2430
+    },
+    {
+      "epoch": 3.9162303664921465,
+      "grad_norm": 0.22899308427743442,
+      "learning_rate": 5.411381024134756e-06,
+      "loss": 0.202,
+      "step": 2431
+    },
+    {
+      "epoch": 3.9178413209826823,
+      "grad_norm": 0.23426089115523857,
+      "learning_rate": 5.395985111565938e-06,
+      "loss": 0.2043,
+      "step": 2432
+    },
+    {
+      "epoch": 3.9194522754732177,
+      "grad_norm": 0.2205496319307363,
+      "learning_rate": 5.380607715695822e-06,
+      "loss": 0.1944,
+      "step": 2433
+    },
+    {
+      "epoch": 3.9210632299637536,
+      "grad_norm": 0.23047985224867817,
+      "learning_rate": 5.3652488560216875e-06,
+      "loss": 0.2084,
+      "step": 2434
+    },
+    {
+      "epoch": 3.922674184454289,
+      "grad_norm": 0.225561512407316,
+      "learning_rate": 5.349908552017323e-06,
+      "loss": 0.2004,
+      "step": 2435
+    },
+    {
+      "epoch": 3.924285138944825,
+      "grad_norm": 0.23594038731291309,
+      "learning_rate": 5.334586823132983e-06,
+      "loss": 0.2138,
+      "step": 2436
+    },
+    {
+      "epoch": 3.9258960934353606,
+      "grad_norm": 0.22443378504100486,
+      "learning_rate": 5.31928368879536e-06,
+      "loss": 0.1862,
+      "step": 2437
+    },
+    {
+      "epoch": 3.927507047925896,
+      "grad_norm": 0.23429297324806259,
+      "learning_rate": 5.303999168407585e-06,
+      "loss": 0.2156,
+      "step": 2438
+    },
+    {
+      "epoch": 3.929118002416432,
+      "grad_norm": 0.22875055640793499,
+      "learning_rate": 5.288733281349186e-06,
+      "loss": 0.1828,
+      "step": 2439
+    },
+    {
+      "epoch": 3.930728956906967,
+      "grad_norm": 0.2247145591887911,
+      "learning_rate": 5.273486046976057e-06,
+      "loss": 0.1846,
+      "step": 2440
+    },
+    {
+      "epoch": 3.932339911397503,
+      "grad_norm": 0.23458620492491555,
+      "learning_rate": 5.258257484620441e-06,
+      "loss": 0.2149,
+      "step": 2441
+    },
+    {
+      "epoch": 3.933950865888039,
+      "grad_norm": 0.22576936655233082,
+      "learning_rate": 5.243047613590919e-06,
+      "loss": 0.1738,
+      "step": 2442
+    },
+    {
+      "epoch": 3.9355618203785743,
+      "grad_norm": 0.2407340144249399,
+      "learning_rate": 5.227856453172371e-06,
+      "loss": 0.1849,
+      "step": 2443
+    },
+    {
+      "epoch": 3.93717277486911,
+      "grad_norm": 0.21933224499037723,
+      "learning_rate": 5.212684022625938e-06,
+      "loss": 0.2001,
+      "step": 2444
+    },
+    {
+      "epoch": 3.9387837293596455,
+      "grad_norm": 0.22669845460715846,
+      "learning_rate": 5.1975303411890235e-06,
+      "loss": 0.1974,
+      "step": 2445
+    },
+    {
+      "epoch": 3.9403946838501813,
+      "grad_norm": 0.21697797198281793,
+      "learning_rate": 5.182395428075262e-06,
+      "loss": 0.183,
+      "step": 2446
+    },
+    {
+      "epoch": 3.9420056383407167,
+      "grad_norm": 0.23111635845943326,
+      "learning_rate": 5.167279302474493e-06,
+      "loss": 0.2058,
+      "step": 2447
+    },
+    {
+      "epoch": 3.9436165928312525,
+      "grad_norm": 0.21966365881724723,
+      "learning_rate": 5.152181983552718e-06,
+      "loss": 0.1975,
+      "step": 2448
+    },
+    {
+      "epoch": 3.945227547321788,
+      "grad_norm": 0.23026550236047766,
+      "learning_rate": 5.137103490452113e-06,
+      "loss": 0.1977,
+      "step": 2449
+    },
+    {
+      "epoch": 3.9468385018123238,
+      "grad_norm": 0.22759257630671995,
+      "learning_rate": 5.12204384229098e-06,
+      "loss": 0.204,
+      "step": 2450
+    },
+    {
+      "epoch": 3.9484494563028596,
+      "grad_norm": 0.21139382664115194,
+      "learning_rate": 5.10700305816372e-06,
+      "loss": 0.1867,
+      "step": 2451
+    },
+    {
+      "epoch": 3.950060410793395,
+      "grad_norm": 0.21338019048344403,
+      "learning_rate": 5.091981157140808e-06,
+      "loss": 0.1819,
+      "step": 2452
+    },
+    {
+      "epoch": 3.951671365283931,
+      "grad_norm": 0.23306725549135782,
+      "learning_rate": 5.076978158268801e-06,
+      "loss": 0.1956,
+      "step": 2453
+    },
+    {
+      "epoch": 3.953282319774466,
+      "grad_norm": 0.2281881871246552,
+      "learning_rate": 5.061994080570281e-06,
+      "loss": 0.1836,
+      "step": 2454
+    },
+    {
+      "epoch": 3.954893274265002,
+      "grad_norm": 0.2225882034388656,
+      "learning_rate": 5.047028943043826e-06,
+      "loss": 0.181,
+      "step": 2455
+    },
+    {
+      "epoch": 3.956504228755538,
+      "grad_norm": 0.22366930281729092,
+      "learning_rate": 5.0320827646640054e-06,
+      "loss": 0.2067,
+      "step": 2456
+    },
+    {
+      "epoch": 3.9581151832460733,
+      "grad_norm": 0.2220234197721443,
+      "learning_rate": 5.01715556438136e-06,
+      "loss": 0.2225,
+      "step": 2457
+    },
+    {
+      "epoch": 3.959726137736609,
+      "grad_norm": 0.20193439968881782,
+      "learning_rate": 5.0022473611223635e-06,
+      "loss": 0.1899,
+      "step": 2458
+    },
+    {
+      "epoch": 3.9613370922271445,
+      "grad_norm": 0.22650822798732484,
+      "learning_rate": 4.987358173789394e-06,
+      "loss": 0.2058,
+      "step": 2459
+    },
+    {
+      "epoch": 3.9629480467176803,
+      "grad_norm": 0.22033364513660997,
+      "learning_rate": 4.972488021260733e-06,
+      "loss": 0.2028,
+      "step": 2460
+    },
+    {
+      "epoch": 3.964559001208216,
+      "grad_norm": 0.21837427939672938,
+      "learning_rate": 4.95763692239051e-06,
+      "loss": 0.2115,
+      "step": 2461
+    },
+    {
+      "epoch": 3.9661699556987515,
+      "grad_norm": 0.2279553953983244,
+      "learning_rate": 4.942804896008717e-06,
+      "loss": 0.2063,
+      "step": 2462
+    },
+    {
+      "epoch": 3.967780910189287,
+      "grad_norm": 0.2138127959490929,
+      "learning_rate": 4.927991960921141e-06,
+      "loss": 0.1793,
+      "step": 2463
+    },
+    {
+      "epoch": 3.9693918646798227,
+      "grad_norm": 0.22416253513747308,
+      "learning_rate": 4.9131981359093826e-06,
+      "loss": 0.2017,
+      "step": 2464
+    },
+    {
+      "epoch": 3.9710028191703586,
+      "grad_norm": 0.228356776210779,
+      "learning_rate": 4.8984234397308086e-06,
+      "loss": 0.1888,
+      "step": 2465
+    },
+    {
+      "epoch": 3.972613773660894,
+      "grad_norm": 0.24338949114710415,
+      "learning_rate": 4.883667891118515e-06,
+      "loss": 0.2384,
+      "step": 2466
+    },
+    {
+      "epoch": 3.97422472815143,
+      "grad_norm": 0.2099842060282153,
+      "learning_rate": 4.868931508781345e-06,
+      "loss": 0.1863,
+      "step": 2467
+    },
+    {
+      "epoch": 3.975835682641965,
+      "grad_norm": 0.217213579023591,
+      "learning_rate": 4.854214311403818e-06,
+      "loss": 0.207,
+      "step": 2468
+    },
+    {
+      "epoch": 3.977446637132501,
+      "grad_norm": 0.2198373072030007,
+      "learning_rate": 4.839516317646149e-06,
+      "loss": 0.198,
+      "step": 2469
+    },
+    {
+      "epoch": 3.979057591623037,
+      "grad_norm": 0.2262853783161159,
+      "learning_rate": 4.824837546144183e-06,
+      "loss": 0.2114,
+      "step": 2470
+    },
+    {
+      "epoch": 3.9806685461135722,
+      "grad_norm": 0.21939682500082897,
+      "learning_rate": 4.810178015509415e-06,
+      "loss": 0.2064,
+      "step": 2471
+    },
+    {
+      "epoch": 3.982279500604108,
+      "grad_norm": 0.22515465224528589,
+      "learning_rate": 4.795537744328924e-06,
+      "loss": 0.1987,
+      "step": 2472
+    },
+    {
+      "epoch": 3.9838904550946435,
+      "grad_norm": 0.21740935835993958,
+      "learning_rate": 4.780916751165388e-06,
+      "loss": 0.2032,
+      "step": 2473
+    },
+    {
+      "epoch": 3.9855014095851793,
+      "grad_norm": 0.22140692196370554,
+      "learning_rate": 4.76631505455702e-06,
+      "loss": 0.1975,
+      "step": 2474
+    },
+    {
+      "epoch": 3.987112364075715,
+      "grad_norm": 0.2284279380672877,
+      "learning_rate": 4.751732673017589e-06,
+      "loss": 0.2144,
+      "step": 2475
+    },
+    {
+      "epoch": 3.9887233185662505,
+      "grad_norm": 0.22252220249176013,
+      "learning_rate": 4.737169625036369e-06,
+      "loss": 0.1961,
+      "step": 2476
+    },
+    {
+      "epoch": 3.990334273056786,
+      "grad_norm": 0.2213040546419634,
+      "learning_rate": 4.722625929078102e-06,
+      "loss": 0.2141,
+      "step": 2477
+    },
+    {
+      "epoch": 3.9919452275473217,
+      "grad_norm": 0.22175328254802687,
+      "learning_rate": 4.708101603583019e-06,
+      "loss": 0.184,
+      "step": 2478
+    },
+    {
+      "epoch": 3.9935561820378576,
+      "grad_norm": 0.22787646419219595,
+      "learning_rate": 4.693596666966771e-06,
+      "loss": 0.2179,
+      "step": 2479
+    },
+    {
+      "epoch": 3.995167136528393,
+      "grad_norm": 0.21931238676949583,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.2055,
+      "step": 2480
+    },
+    {
+      "epoch": 3.996778091018929,
+      "grad_norm": 0.22831017006597473,
+      "learning_rate": 4.664645033910491e-06,
+      "loss": 0.2113,
+      "step": 2481
+    },
+    {
+      "epoch": 3.998389045509464,
+      "grad_norm": 0.2179416073465815,
+      "learning_rate": 4.650198374178767e-06,
+      "loss": 0.1947,
+      "step": 2482
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.26286690041897504,
+      "learning_rate": 4.635771176742443e-06,
+      "loss": 0.2311,
+      "step": 2483
+    },
+    {
+      "epoch": 4.001610954490536,
+      "grad_norm": 0.394604874099715,
+      "learning_rate": 4.621363459894039e-06,
+      "loss": 0.1444,
+      "step": 2484
+    },
+    {
+      "epoch": 4.003221908981072,
+      "grad_norm": 0.30419988334377224,
+      "learning_rate": 4.606975241901354e-06,
+      "loss": 0.1577,
+      "step": 2485
+    },
+    {
+      "epoch": 4.004832863471607,
+      "grad_norm": 0.24739619219289316,
+      "learning_rate": 4.592606541007481e-06,
+      "loss": 0.152,
+      "step": 2486
+    },
+    {
+      "epoch": 4.006443817962142,
+      "grad_norm": 0.37269612926471524,
+      "learning_rate": 4.578257375430764e-06,
+      "loss": 0.1528,
+      "step": 2487
+    },
+    {
+      "epoch": 4.008054772452678,
+      "grad_norm": 0.41792249971665374,
+      "learning_rate": 4.563927763364759e-06,
+      "loss": 0.1466,
+      "step": 2488
+    },
+    {
+      "epoch": 4.009665726943214,
+      "grad_norm": 0.3578527160556587,
+      "learning_rate": 4.549617722978259e-06,
+      "loss": 0.1601,
+      "step": 2489
+    },
+    {
+      "epoch": 4.01127668143375,
+      "grad_norm": 0.2889923380752588,
+      "learning_rate": 4.535327272415215e-06,
+      "loss": 0.1456,
+      "step": 2490
+    },
+    {
+      "epoch": 4.012887635924285,
+      "grad_norm": 0.3818532366043304,
+      "learning_rate": 4.521056429794763e-06,
+      "loss": 0.158,
+      "step": 2491
+    },
+    {
+      "epoch": 4.014498590414821,
+      "grad_norm": 0.36182018754446954,
+      "learning_rate": 4.506805213211154e-06,
+      "loss": 0.1496,
+      "step": 2492
+    },
+    {
+      "epoch": 4.0161095449053565,
+      "grad_norm": 0.32110946672977075,
+      "learning_rate": 4.492573640733781e-06,
+      "loss": 0.1518,
+      "step": 2493
+    },
+    {
+      "epoch": 4.017720499395892,
+      "grad_norm": 0.23828594766969258,
+      "learning_rate": 4.478361730407104e-06,
+      "loss": 0.1427,
+      "step": 2494
+    },
+    {
+      "epoch": 4.019331453886427,
+      "grad_norm": 0.28940264741575655,
+      "learning_rate": 4.464169500250677e-06,
+      "loss": 0.137,
+      "step": 2495
+    },
+    {
+      "epoch": 4.020942408376963,
+      "grad_norm": 0.32475999960770263,
+      "learning_rate": 4.449996968259074e-06,
+      "loss": 0.1552,
+      "step": 2496
+    },
+    {
+      "epoch": 4.022553362867499,
+      "grad_norm": 0.3106520553812506,
+      "learning_rate": 4.435844152401925e-06,
+      "loss": 0.1604,
+      "step": 2497
+    },
+    {
+      "epoch": 4.024164317358035,
+      "grad_norm": 0.25339318934044625,
+      "learning_rate": 4.4217110706238305e-06,
+      "loss": 0.1398,
+      "step": 2498
+    },
+    {
+      "epoch": 4.025775271848571,
+      "grad_norm": 0.23074024688964512,
+      "learning_rate": 4.407597740844393e-06,
+      "loss": 0.1536,
+      "step": 2499
+    },
+    {
+      "epoch": 4.027386226339106,
+      "grad_norm": 0.2298221033798615,
+      "learning_rate": 4.393504180958166e-06,
+      "loss": 0.1467,
+      "step": 2500
+    },
+    {
+      "epoch": 4.028997180829641,
+      "grad_norm": 0.24652202184560998,
+      "learning_rate": 4.3794304088346215e-06,
+      "loss": 0.1499,
+      "step": 2501
+    },
+    {
+      "epoch": 4.030608135320177,
+      "grad_norm": 0.25842548848300634,
+      "learning_rate": 4.365376442318168e-06,
+      "loss": 0.1491,
+      "step": 2502
+    },
+    {
+      "epoch": 4.032219089810713,
+      "grad_norm": 0.2241710771974074,
+      "learning_rate": 4.351342299228072e-06,
+      "loss": 0.1581,
+      "step": 2503
+    },
+    {
+      "epoch": 4.033830044301249,
+      "grad_norm": 0.226375077124321,
+      "learning_rate": 4.337327997358494e-06,
+      "loss": 0.1454,
+      "step": 2504
+    },
+    {
+      "epoch": 4.035440998791784,
+      "grad_norm": 0.2386861417217738,
+      "learning_rate": 4.323333554478415e-06,
+      "loss": 0.1489,
+      "step": 2505
+    },
+    {
+      "epoch": 4.03705195328232,
+      "grad_norm": 0.24161342029014413,
+      "learning_rate": 4.309358988331658e-06,
+      "loss": 0.1391,
+      "step": 2506
+    },
+    {
+      "epoch": 4.0386629077728555,
+      "grad_norm": 0.24826467464092444,
+      "learning_rate": 4.2954043166368176e-06,
+      "loss": 0.1428,
+      "step": 2507
+    },
+    {
+      "epoch": 4.040273862263391,
+      "grad_norm": 0.22543955458616258,
+      "learning_rate": 4.281469557087292e-06,
+      "loss": 0.1512,
+      "step": 2508
+    },
+    {
+      "epoch": 4.041884816753926,
+      "grad_norm": 0.22191099450465107,
+      "learning_rate": 4.267554727351209e-06,
+      "loss": 0.1518,
+      "step": 2509
+    },
+    {
+      "epoch": 4.043495771244462,
+      "grad_norm": 0.22413069894128723,
+      "learning_rate": 4.253659845071436e-06,
+      "loss": 0.1486,
+      "step": 2510
+    },
+    {
+      "epoch": 4.045106725734998,
+      "grad_norm": 0.23604198643903934,
+      "learning_rate": 4.239784927865562e-06,
+      "loss": 0.1563,
+      "step": 2511
+    },
+    {
+      "epoch": 4.046717680225534,
+      "grad_norm": 0.23765897381752274,
+      "learning_rate": 4.225929993325837e-06,
+      "loss": 0.1357,
+      "step": 2512
+    },
+    {
+      "epoch": 4.04832863471607,
+      "grad_norm": 0.2325992132869097,
+      "learning_rate": 4.2120950590191945e-06,
+      "loss": 0.1412,
+      "step": 2513
+    },
+    {
+      "epoch": 4.049939589206605,
+      "grad_norm": 0.219156691846048,
+      "learning_rate": 4.198280142487194e-06,
+      "loss": 0.1386,
+      "step": 2514
+    },
+    {
+      "epoch": 4.05155054369714,
+      "grad_norm": 0.22634529978022389,
+      "learning_rate": 4.184485261246032e-06,
+      "loss": 0.1401,
+      "step": 2515
+    },
+    {
+      "epoch": 4.053161498187676,
+      "grad_norm": 0.22623994036256284,
+      "learning_rate": 4.17071043278648e-06,
+      "loss": 0.1512,
+      "step": 2516
+    },
+    {
+      "epoch": 4.054772452678212,
+      "grad_norm": 0.2348031400592028,
+      "learning_rate": 4.156955674573908e-06,
+      "loss": 0.1514,
+      "step": 2517
+    },
+    {
+      "epoch": 4.056383407168748,
+      "grad_norm": 0.2422395543425095,
+      "learning_rate": 4.143221004048215e-06,
+      "loss": 0.1545,
+      "step": 2518
+    },
+    {
+      "epoch": 4.057994361659283,
+      "grad_norm": 0.22710278247393073,
+      "learning_rate": 4.129506438623854e-06,
+      "loss": 0.154,
+      "step": 2519
+    },
+    {
+      "epoch": 4.059605316149819,
+      "grad_norm": 0.2119993093135068,
+      "learning_rate": 4.11581199568976e-06,
+      "loss": 0.1508,
+      "step": 2520
+    },
+    {
+      "epoch": 4.0612162706403545,
+      "grad_norm": 0.2266486757464885,
+      "learning_rate": 4.10213769260938e-06,
+      "loss": 0.1475,
+      "step": 2521
+    },
+    {
+      "epoch": 4.06282722513089,
+      "grad_norm": 0.22567150872123293,
+      "learning_rate": 4.088483546720614e-06,
+      "loss": 0.1552,
+      "step": 2522
+    },
+    {
+      "epoch": 4.064438179621425,
+      "grad_norm": 0.23043810924965868,
+      "learning_rate": 4.074849575335804e-06,
+      "loss": 0.1552,
+      "step": 2523
+    },
+    {
+      "epoch": 4.066049134111961,
+      "grad_norm": 0.23119706591599065,
+      "learning_rate": 4.061235795741702e-06,
+      "loss": 0.1391,
+      "step": 2524
+    },
+    {
+      "epoch": 4.067660088602497,
+      "grad_norm": 0.2167544238933369,
+      "learning_rate": 4.04764222519948e-06,
+      "loss": 0.1439,
+      "step": 2525
+    },
+    {
+      "epoch": 4.069271043093033,
+      "grad_norm": 0.22872180237610718,
+      "learning_rate": 4.0340688809446745e-06,
+      "loss": 0.1652,
+      "step": 2526
+    },
+    {
+      "epoch": 4.070881997583569,
+      "grad_norm": 0.22662520444169074,
+      "learning_rate": 4.020515780187173e-06,
+      "loss": 0.1401,
+      "step": 2527
+    },
+    {
+      "epoch": 4.072492952074104,
+      "grad_norm": 0.22758725109112826,
+      "learning_rate": 4.006982940111204e-06,
+      "loss": 0.1474,
+      "step": 2528
+    },
+    {
+      "epoch": 4.074103906564639,
+      "grad_norm": 0.21647537990769658,
+      "learning_rate": 3.993470377875311e-06,
+      "loss": 0.153,
+      "step": 2529
+    },
+    {
+      "epoch": 4.075714861055175,
+      "grad_norm": 0.21820740866956023,
+      "learning_rate": 3.979978110612313e-06,
+      "loss": 0.165,
+      "step": 2530
+    },
+    {
+      "epoch": 4.077325815545711,
+      "grad_norm": 0.2133491338774458,
+      "learning_rate": 3.9665061554292946e-06,
+      "loss": 0.1397,
+      "step": 2531
+    },
+    {
+      "epoch": 4.078936770036247,
+      "grad_norm": 0.22608875122217786,
+      "learning_rate": 3.9530545294076075e-06,
+      "loss": 0.1541,
+      "step": 2532
+    },
+    {
+      "epoch": 4.080547724526782,
+      "grad_norm": 0.21979115321693685,
+      "learning_rate": 3.9396232496028176e-06,
+      "loss": 0.1507,
+      "step": 2533
+    },
+    {
+      "epoch": 4.082158679017318,
+      "grad_norm": 0.21782116069499285,
+      "learning_rate": 3.926212333044683e-06,
+      "loss": 0.1501,
+      "step": 2534
+    },
+    {
+      "epoch": 4.0837696335078535,
+      "grad_norm": 0.22037323116617433,
+      "learning_rate": 3.9128217967371515e-06,
+      "loss": 0.1522,
+      "step": 2535
+    },
+    {
+      "epoch": 4.085380587998389,
+      "grad_norm": 0.2144296673882382,
+      "learning_rate": 3.899451657658331e-06,
+      "loss": 0.1486,
+      "step": 2536
+    },
+    {
+      "epoch": 4.086991542488924,
+      "grad_norm": 0.21177670307177632,
+      "learning_rate": 3.8861019327604736e-06,
+      "loss": 0.143,
+      "step": 2537
+    },
+    {
+      "epoch": 4.08860249697946,
+      "grad_norm": 0.22542118104376185,
+      "learning_rate": 3.872772638969929e-06,
+      "loss": 0.1463,
+      "step": 2538
+    },
+    {
+      "epoch": 4.090213451469996,
+      "grad_norm": 0.21745441278678204,
+      "learning_rate": 3.859463793187159e-06,
+      "loss": 0.1361,
+      "step": 2539
+    },
+    {
+      "epoch": 4.091824405960532,
+      "grad_norm": 0.21992591585331572,
+      "learning_rate": 3.846175412286701e-06,
+      "loss": 0.1572,
+      "step": 2540
+    },
+    {
+      "epoch": 4.093435360451068,
+      "grad_norm": 0.22629768234535663,
+      "learning_rate": 3.83290751311713e-06,
+      "loss": 0.1419,
+      "step": 2541
+    },
+    {
+      "epoch": 4.0950463149416025,
+      "grad_norm": 0.2298622699155759,
+      "learning_rate": 3.819660112501053e-06,
+      "loss": 0.1355,
+      "step": 2542
+    },
+    {
+      "epoch": 4.096657269432138,
+      "grad_norm": 0.21678800682951568,
+      "learning_rate": 3.8064332272351e-06,
+      "loss": 0.1496,
+      "step": 2543
+    },
+    {
+      "epoch": 4.098268223922674,
+      "grad_norm": 0.22949965448343948,
+      "learning_rate": 3.7932268740898836e-06,
+      "loss": 0.1556,
+      "step": 2544
+    },
+    {
+      "epoch": 4.09987917841321,
+      "grad_norm": 0.22071553901670435,
+      "learning_rate": 3.7800410698099808e-06,
+      "loss": 0.1574,
+      "step": 2545
+    },
+    {
+      "epoch": 4.101490132903746,
+      "grad_norm": 0.2298993165539254,
+      "learning_rate": 3.7668758311139077e-06,
+      "loss": 0.151,
+      "step": 2546
+    },
+    {
+      "epoch": 4.103101087394281,
+      "grad_norm": 0.2225517731771703,
+      "learning_rate": 3.753731174694117e-06,
+      "loss": 0.1467,
+      "step": 2547
+    },
+    {
+      "epoch": 4.104712041884817,
+      "grad_norm": 0.225093485838971,
+      "learning_rate": 3.7406071172169634e-06,
+      "loss": 0.1352,
+      "step": 2548
+    },
+    {
+      "epoch": 4.1063229963753525,
+      "grad_norm": 0.23400738843222163,
+      "learning_rate": 3.727503675322681e-06,
+      "loss": 0.1421,
+      "step": 2549
+    },
+    {
+      "epoch": 4.107933950865888,
+      "grad_norm": 0.2366245815845293,
+      "learning_rate": 3.7144208656253476e-06,
+      "loss": 0.1485,
+      "step": 2550
+    },
+    {
+      "epoch": 4.109544905356423,
+      "grad_norm": 0.2154306414839747,
+      "learning_rate": 3.701358704712923e-06,
+      "loss": 0.1463,
+      "step": 2551
+    },
+    {
+      "epoch": 4.111155859846959,
+      "grad_norm": 0.2207385286375425,
+      "learning_rate": 3.6883172091471474e-06,
+      "loss": 0.1588,
+      "step": 2552
+    },
+    {
+      "epoch": 4.112766814337495,
+      "grad_norm": 0.22269838904544906,
+      "learning_rate": 3.67529639546357e-06,
+      "loss": 0.1452,
+      "step": 2553
+    },
+    {
+      "epoch": 4.114377768828031,
+      "grad_norm": 0.2313604111125716,
+      "learning_rate": 3.6622962801715243e-06,
+      "loss": 0.1342,
+      "step": 2554
+    },
+    {
+      "epoch": 4.115988723318567,
+      "grad_norm": 0.2284422968399893,
+      "learning_rate": 3.649316879754099e-06,
+      "loss": 0.1417,
+      "step": 2555
+    },
+    {
+      "epoch": 4.1175996778091015,
+      "grad_norm": 0.2465718086622103,
+      "learning_rate": 3.6363582106681115e-06,
+      "loss": 0.1408,
+      "step": 2556
+    },
+    {
+      "epoch": 4.119210632299637,
+      "grad_norm": 0.23546119698074278,
+      "learning_rate": 3.623420289344088e-06,
+      "loss": 0.1367,
+      "step": 2557
+    },
+    {
+      "epoch": 4.120821586790173,
+      "grad_norm": 0.23013205586732102,
+      "learning_rate": 3.610503132186265e-06,
+      "loss": 0.1566,
+      "step": 2558
+    },
+    {
+      "epoch": 4.122432541280709,
+      "grad_norm": 0.22709715305840766,
+      "learning_rate": 3.597606755572545e-06,
+      "loss": 0.1421,
+      "step": 2559
+    },
+    {
+      "epoch": 4.124043495771245,
+      "grad_norm": 0.2270302334606441,
+      "learning_rate": 3.584731175854479e-06,
+      "loss": 0.1417,
+      "step": 2560
+    },
+    {
+      "epoch": 4.12565445026178,
+      "grad_norm": 0.22673844401199308,
+      "learning_rate": 3.57187640935724e-06,
+      "loss": 0.1433,
+      "step": 2561
+    },
+    {
+      "epoch": 4.127265404752316,
+      "grad_norm": 0.22556426103212388,
+      "learning_rate": 3.559042472379639e-06,
+      "loss": 0.1545,
+      "step": 2562
+    },
+    {
+      "epoch": 4.1288763592428515,
+      "grad_norm": 0.22805435104245703,
+      "learning_rate": 3.546229381194057e-06,
+      "loss": 0.1462,
+      "step": 2563
+    },
+    {
+      "epoch": 4.130487313733387,
+      "grad_norm": 0.21900564349736037,
+      "learning_rate": 3.5334371520464373e-06,
+      "loss": 0.1522,
+      "step": 2564
+    },
+    {
+      "epoch": 4.132098268223922,
+      "grad_norm": 0.2139940599235903,
+      "learning_rate": 3.520665801156289e-06,
+      "loss": 0.1372,
+      "step": 2565
+    },
+    {
+      "epoch": 4.133709222714458,
+      "grad_norm": 0.22062229403040695,
+      "learning_rate": 3.507915344716648e-06,
+      "loss": 0.1491,
+      "step": 2566
+    },
+    {
+      "epoch": 4.135320177204994,
+      "grad_norm": 0.23722546949444184,
+      "learning_rate": 3.4951857988940475e-06,
+      "loss": 0.1515,
+      "step": 2567
+    },
+    {
+      "epoch": 4.13693113169553,
+      "grad_norm": 0.22131787664167898,
+      "learning_rate": 3.482477179828505e-06,
+      "loss": 0.1412,
+      "step": 2568
+    },
+    {
+      "epoch": 4.138542086186066,
+      "grad_norm": 0.21918030255310963,
+      "learning_rate": 3.4697895036335204e-06,
+      "loss": 0.1545,
+      "step": 2569
+    },
+    {
+      "epoch": 4.1401530406766005,
+      "grad_norm": 0.22054026589218842,
+      "learning_rate": 3.457122786396032e-06,
+      "loss": 0.14,
+      "step": 2570
+    },
+    {
+      "epoch": 4.141763995167136,
+      "grad_norm": 0.22504736458991076,
+      "learning_rate": 3.4444770441764043e-06,
+      "loss": 0.1484,
+      "step": 2571
+    },
+    {
+      "epoch": 4.143374949657672,
+      "grad_norm": 0.21902661316640898,
+      "learning_rate": 3.431852293008391e-06,
+      "loss": 0.1498,
+      "step": 2572
+    },
+    {
+      "epoch": 4.144985904148208,
+      "grad_norm": 0.2210425406056193,
+      "learning_rate": 3.419248548899168e-06,
+      "loss": 0.1465,
+      "step": 2573
+    },
+    {
+      "epoch": 4.146596858638744,
+      "grad_norm": 0.22089200704994782,
+      "learning_rate": 3.406665827829243e-06,
+      "loss": 0.1562,
+      "step": 2574
+    },
+    {
+      "epoch": 4.148207813129279,
+      "grad_norm": 0.2252789843720791,
+      "learning_rate": 3.3941041457524748e-06,
+      "loss": 0.1465,
+      "step": 2575
+    },
+    {
+      "epoch": 4.149818767619815,
+      "grad_norm": 0.23272666378438298,
+      "learning_rate": 3.381563518596056e-06,
+      "loss": 0.1466,
+      "step": 2576
+    },
+    {
+      "epoch": 4.15142972211035,
+      "grad_norm": 0.22958528962750838,
+      "learning_rate": 3.3690439622604832e-06,
+      "loss": 0.1347,
+      "step": 2577
+    },
+    {
+      "epoch": 4.153040676600886,
+      "grad_norm": 0.2255442065793069,
+      "learning_rate": 3.3565454926195252e-06,
+      "loss": 0.1469,
+      "step": 2578
+    },
+    {
+      "epoch": 4.154651631091422,
+      "grad_norm": 0.21542881986949647,
+      "learning_rate": 3.344068125520219e-06,
+      "loss": 0.1515,
+      "step": 2579
+    },
+    {
+      "epoch": 4.156262585581957,
+      "grad_norm": 0.22530311793594246,
+      "learning_rate": 3.3316118767828498e-06,
+      "loss": 0.1542,
+      "step": 2580
+    },
+    {
+      "epoch": 4.157873540072493,
+      "grad_norm": 0.2156195896321625,
+      "learning_rate": 3.3191767622009307e-06,
+      "loss": 0.1528,
+      "step": 2581
+    },
+    {
+      "epoch": 4.159484494563029,
+      "grad_norm": 0.21841843601381836,
+      "learning_rate": 3.3067627975411675e-06,
+      "loss": 0.1496,
+      "step": 2582
+    },
+    {
+      "epoch": 4.1610954490535645,
+      "grad_norm": 0.23200002731345346,
+      "learning_rate": 3.294369998543443e-06,
+      "loss": 0.1515,
+      "step": 2583
+    },
+    {
+      "epoch": 4.1627064035440995,
+      "grad_norm": 0.23379438316230453,
+      "learning_rate": 3.2819983809208346e-06,
+      "loss": 0.1475,
+      "step": 2584
+    },
+    {
+      "epoch": 4.164317358034635,
+      "grad_norm": 0.21329943795679254,
+      "learning_rate": 3.269647960359532e-06,
+      "loss": 0.1473,
+      "step": 2585
+    },
+    {
+      "epoch": 4.165928312525171,
+      "grad_norm": 0.23135542202637302,
+      "learning_rate": 3.257318752518859e-06,
+      "loss": 0.1548,
+      "step": 2586
+    },
+    {
+      "epoch": 4.167539267015707,
+      "grad_norm": 0.21343772694404733,
+      "learning_rate": 3.2450107730312473e-06,
+      "loss": 0.1397,
+      "step": 2587
+    },
+    {
+      "epoch": 4.169150221506243,
+      "grad_norm": 0.23309642536401526,
+      "learning_rate": 3.232724037502215e-06,
+      "loss": 0.1599,
+      "step": 2588
+    },
+    {
+      "epoch": 4.170761175996778,
+      "grad_norm": 0.23103447802019342,
+      "learning_rate": 3.2204585615103355e-06,
+      "loss": 0.1534,
+      "step": 2589
+    },
+    {
+      "epoch": 4.172372130487314,
+      "grad_norm": 0.2270645147859799,
+      "learning_rate": 3.20821436060722e-06,
+      "loss": 0.1529,
+      "step": 2590
+    },
+    {
+      "epoch": 4.173983084977849,
+      "grad_norm": 0.23814634653183514,
+      "learning_rate": 3.195991450317537e-06,
+      "loss": 0.1466,
+      "step": 2591
+    },
+    {
+      "epoch": 4.175594039468385,
+      "grad_norm": 0.22830584321137043,
+      "learning_rate": 3.183789846138927e-06,
+      "loss": 0.1445,
+      "step": 2592
+    },
+    {
+      "epoch": 4.177204993958921,
+      "grad_norm": 0.21871275817927927,
+      "learning_rate": 3.1716095635420265e-06,
+      "loss": 0.1581,
+      "step": 2593
+    },
+    {
+      "epoch": 4.178815948449456,
+      "grad_norm": 0.22189336798505455,
+      "learning_rate": 3.159450617970441e-06,
+      "loss": 0.1572,
+      "step": 2594
+    },
+    {
+      "epoch": 4.180426902939992,
+      "grad_norm": 0.21870086409019102,
+      "learning_rate": 3.1473130248407278e-06,
+      "loss": 0.1366,
+      "step": 2595
+    },
+    {
+      "epoch": 4.182037857430528,
+      "grad_norm": 0.22775597716347107,
+      "learning_rate": 3.1351967995423594e-06,
+      "loss": 0.1527,
+      "step": 2596
+    },
+    {
+      "epoch": 4.1836488119210635,
+      "grad_norm": 0.22276143260948325,
+      "learning_rate": 3.1231019574377153e-06,
+      "loss": 0.1452,
+      "step": 2597
+    },
+    {
+      "epoch": 4.1852597664115985,
+      "grad_norm": 0.2178191763823298,
+      "learning_rate": 3.111028513862071e-06,
+      "loss": 0.149,
+      "step": 2598
+    },
+    {
+      "epoch": 4.186870720902134,
+      "grad_norm": 0.22160329576960924,
+      "learning_rate": 3.0989764841235746e-06,
+      "loss": 0.1484,
+      "step": 2599
+    },
+    {
+      "epoch": 4.18848167539267,
+      "grad_norm": 0.22102821797818814,
+      "learning_rate": 3.0869458835032097e-06,
+      "loss": 0.1469,
+      "step": 2600
+    },
+    {
+      "epoch": 4.190092629883206,
+      "grad_norm": 0.22697752307908553,
+      "learning_rate": 3.074936727254785e-06,
+      "loss": 0.1544,
+      "step": 2601
+    },
+    {
+      "epoch": 4.191703584373742,
+      "grad_norm": 0.22392044668301703,
+      "learning_rate": 3.0629490306049536e-06,
+      "loss": 0.1477,
+      "step": 2602
+    },
+    {
+      "epoch": 4.193314538864277,
+      "grad_norm": 0.22404327159580825,
+      "learning_rate": 3.0509828087531224e-06,
+      "loss": 0.1547,
+      "step": 2603
+    },
+    {
+      "epoch": 4.194925493354813,
+      "grad_norm": 0.23084898184198646,
+      "learning_rate": 3.039038076871481e-06,
+      "loss": 0.1544,
+      "step": 2604
+    },
+    {
+      "epoch": 4.196536447845348,
+      "grad_norm": 0.2232139079228995,
+      "learning_rate": 3.0271148501049796e-06,
+      "loss": 0.1402,
+      "step": 2605
+    },
+    {
+      "epoch": 4.198147402335884,
+      "grad_norm": 0.22551484154069532,
+      "learning_rate": 3.0152131435713007e-06,
+      "loss": 0.1501,
+      "step": 2606
+    },
+    {
+      "epoch": 4.19975835682642,
+      "grad_norm": 0.2282570119452304,
+      "learning_rate": 3.003332972360831e-06,
+      "loss": 0.154,
+      "step": 2607
+    },
+    {
+      "epoch": 4.201369311316955,
+      "grad_norm": 0.2162893984726015,
+      "learning_rate": 2.9914743515366516e-06,
+      "loss": 0.1439,
+      "step": 2608
+    },
+    {
+      "epoch": 4.202980265807491,
+      "grad_norm": 0.2221742722251089,
+      "learning_rate": 2.9796372961345364e-06,
+      "loss": 0.1533,
+      "step": 2609
+    },
+    {
+      "epoch": 4.204591220298027,
+      "grad_norm": 0.2333853238695215,
+      "learning_rate": 2.967821821162904e-06,
+      "loss": 0.1454,
+      "step": 2610
+    },
+    {
+      "epoch": 4.2062021747885625,
+      "grad_norm": 0.2120253738307176,
+      "learning_rate": 2.9560279416028102e-06,
+      "loss": 0.1506,
+      "step": 2611
+    },
+    {
+      "epoch": 4.2078131292790975,
+      "grad_norm": 0.2140333611346206,
+      "learning_rate": 2.944255672407925e-06,
+      "loss": 0.1478,
+      "step": 2612
+    },
+    {
+      "epoch": 4.209424083769633,
+      "grad_norm": 0.23068436963355107,
+      "learning_rate": 2.932505028504531e-06,
+      "loss": 0.1441,
+      "step": 2613
+    },
+    {
+      "epoch": 4.211035038260169,
+      "grad_norm": 0.21776274650570995,
+      "learning_rate": 2.9207760247914895e-06,
+      "loss": 0.1479,
+      "step": 2614
+    },
+    {
+      "epoch": 4.212645992750705,
+      "grad_norm": 0.23364491125107575,
+      "learning_rate": 2.909068676140212e-06,
+      "loss": 0.1507,
+      "step": 2615
+    },
+    {
+      "epoch": 4.214256947241241,
+      "grad_norm": 0.2212861004899993,
+      "learning_rate": 2.8973829973946645e-06,
+      "loss": 0.1443,
+      "step": 2616
+    },
+    {
+      "epoch": 4.215867901731776,
+      "grad_norm": 0.23588997528901365,
+      "learning_rate": 2.8857190033713343e-06,
+      "loss": 0.1414,
+      "step": 2617
+    },
+    {
+      "epoch": 4.217478856222312,
+      "grad_norm": 0.22836799572319763,
+      "learning_rate": 2.874076708859217e-06,
+      "loss": 0.1497,
+      "step": 2618
+    },
+    {
+      "epoch": 4.219089810712847,
+      "grad_norm": 0.21435982154109828,
+      "learning_rate": 2.8624561286197793e-06,
+      "loss": 0.1328,
+      "step": 2619
+    },
+    {
+      "epoch": 4.220700765203383,
+      "grad_norm": 0.2221623295059318,
+      "learning_rate": 2.850857277386978e-06,
+      "loss": 0.1594,
+      "step": 2620
+    },
+    {
+      "epoch": 4.222311719693919,
+      "grad_norm": 0.2224970582246465,
+      "learning_rate": 2.8392801698672135e-06,
+      "loss": 0.162,
+      "step": 2621
+    },
+    {
+      "epoch": 4.223922674184454,
+      "grad_norm": 0.2155490975094429,
+      "learning_rate": 2.827724820739306e-06,
+      "loss": 0.1299,
+      "step": 2622
+    },
+    {
+      "epoch": 4.22553362867499,
+      "grad_norm": 0.22520553516276554,
+      "learning_rate": 2.8161912446544916e-06,
+      "loss": 0.1516,
+      "step": 2623
+    },
+    {
+      "epoch": 4.227144583165526,
+      "grad_norm": 0.2258052727971873,
+      "learning_rate": 2.80467945623641e-06,
+      "loss": 0.1415,
+      "step": 2624
+    },
+    {
+      "epoch": 4.2287555376560615,
+      "grad_norm": 0.21802617466661803,
+      "learning_rate": 2.7931894700810703e-06,
+      "loss": 0.141,
+      "step": 2625
+    },
+    {
+      "epoch": 4.230366492146596,
+      "grad_norm": 0.22525103510627617,
+      "learning_rate": 2.781721300756828e-06,
+      "loss": 0.1526,
+      "step": 2626
+    },
+    {
+      "epoch": 4.231977446637132,
+      "grad_norm": 0.2225143843203524,
+      "learning_rate": 2.7702749628043933e-06,
+      "loss": 0.1539,
+      "step": 2627
+    },
+    {
+      "epoch": 4.233588401127668,
+      "grad_norm": 0.22398176140606088,
+      "learning_rate": 2.7588504707367913e-06,
+      "loss": 0.155,
+      "step": 2628
+    },
+    {
+      "epoch": 4.235199355618204,
+      "grad_norm": 0.21834762012587516,
+      "learning_rate": 2.747447839039339e-06,
+      "loss": 0.1572,
+      "step": 2629
+    },
+    {
+      "epoch": 4.23681031010874,
+      "grad_norm": 0.21892692454842144,
+      "learning_rate": 2.7360670821696422e-06,
+      "loss": 0.1458,
+      "step": 2630
+    },
+    {
+      "epoch": 4.238421264599275,
+      "grad_norm": 0.22088363148564538,
+      "learning_rate": 2.724708214557572e-06,
+      "loss": 0.1455,
+      "step": 2631
+    },
+    {
+      "epoch": 4.2400322190898105,
+      "grad_norm": 0.23026390864295404,
+      "learning_rate": 2.71337125060525e-06,
+      "loss": 0.1396,
+      "step": 2632
+    },
+    {
+      "epoch": 4.241643173580346,
+      "grad_norm": 0.2290115443147785,
+      "learning_rate": 2.702056204687018e-06,
+      "loss": 0.1526,
+      "step": 2633
+    },
+    {
+      "epoch": 4.243254128070882,
+      "grad_norm": 0.2329308138537628,
+      "learning_rate": 2.6907630911494287e-06,
+      "loss": 0.156,
+      "step": 2634
+    },
+    {
+      "epoch": 4.244865082561418,
+      "grad_norm": 0.2129377543079953,
+      "learning_rate": 2.679491924311226e-06,
+      "loss": 0.1425,
+      "step": 2635
+    },
+    {
+      "epoch": 4.246476037051953,
+      "grad_norm": 0.2277121827224005,
+      "learning_rate": 2.668242718463341e-06,
+      "loss": 0.1403,
+      "step": 2636
+    },
+    {
+      "epoch": 4.248086991542489,
+      "grad_norm": 0.22813667865109089,
+      "learning_rate": 2.657015487868835e-06,
+      "loss": 0.1505,
+      "step": 2637
+    },
+    {
+      "epoch": 4.249697946033025,
+      "grad_norm": 0.22983529188220994,
+      "learning_rate": 2.6458102467629275e-06,
+      "loss": 0.1402,
+      "step": 2638
+    },
+    {
+      "epoch": 4.2513089005235605,
+      "grad_norm": 0.21977565245784658,
+      "learning_rate": 2.6346270093529457e-06,
+      "loss": 0.1324,
+      "step": 2639
+    },
+    {
+      "epoch": 4.252919855014095,
+      "grad_norm": 0.2170974293585535,
+      "learning_rate": 2.623465789818327e-06,
+      "loss": 0.141,
+      "step": 2640
+    },
+    {
+      "epoch": 4.254530809504631,
+      "grad_norm": 0.2310399402068678,
+      "learning_rate": 2.6123266023105774e-06,
+      "loss": 0.1571,
+      "step": 2641
+    },
+    {
+      "epoch": 4.256141763995167,
+      "grad_norm": 0.22177073988062584,
+      "learning_rate": 2.6012094609532845e-06,
+      "loss": 0.1489,
+      "step": 2642
+    },
+    {
+      "epoch": 4.257752718485703,
+      "grad_norm": 0.2151846397977056,
+      "learning_rate": 2.5901143798420792e-06,
+      "loss": 0.1506,
+      "step": 2643
+    },
+    {
+      "epoch": 4.259363672976239,
+      "grad_norm": 0.21855697442042452,
+      "learning_rate": 2.579041373044613e-06,
+      "loss": 0.1439,
+      "step": 2644
+    },
+    {
+      "epoch": 4.260974627466774,
+      "grad_norm": 0.2208457852679748,
+      "learning_rate": 2.5679904546005507e-06,
+      "loss": 0.1447,
+      "step": 2645
+    },
+    {
+      "epoch": 4.2625855819573095,
+      "grad_norm": 0.22198003753575765,
+      "learning_rate": 2.5569616385215625e-06,
+      "loss": 0.1485,
+      "step": 2646
+    },
+    {
+      "epoch": 4.264196536447845,
+      "grad_norm": 0.21034373583782578,
+      "learning_rate": 2.5459549387912843e-06,
+      "loss": 0.1474,
+      "step": 2647
+    },
+    {
+      "epoch": 4.265807490938381,
+      "grad_norm": 0.22413997527867038,
+      "learning_rate": 2.5349703693653103e-06,
+      "loss": 0.1454,
+      "step": 2648
+    },
+    {
+      "epoch": 4.267418445428917,
+      "grad_norm": 0.22479134118096378,
+      "learning_rate": 2.5240079441711853e-06,
+      "loss": 0.1425,
+      "step": 2649
+    },
+    {
+      "epoch": 4.269029399919452,
+      "grad_norm": 0.20947722912794117,
+      "learning_rate": 2.5130676771083585e-06,
+      "loss": 0.1489,
+      "step": 2650
+    },
+    {
+      "epoch": 4.270640354409988,
+      "grad_norm": 0.22886596887508937,
+      "learning_rate": 2.5021495820482057e-06,
+      "loss": 0.1481,
+      "step": 2651
+    },
+    {
+      "epoch": 4.272251308900524,
+      "grad_norm": 0.23491024852118703,
+      "learning_rate": 2.4912536728339707e-06,
+      "loss": 0.1446,
+      "step": 2652
+    },
+    {
+      "epoch": 4.2738622633910595,
+      "grad_norm": 0.2414302632894677,
+      "learning_rate": 2.4803799632807836e-06,
+      "loss": 0.1521,
+      "step": 2653
+    },
+    {
+      "epoch": 4.275473217881595,
+      "grad_norm": 0.21026196842028957,
+      "learning_rate": 2.4695284671756215e-06,
+      "loss": 0.1489,
+      "step": 2654
+    },
+    {
+      "epoch": 4.27708417237213,
+      "grad_norm": 0.228563137088654,
+      "learning_rate": 2.45869919827729e-06,
+      "loss": 0.1428,
+      "step": 2655
+    },
+    {
+      "epoch": 4.278695126862666,
+      "grad_norm": 0.21717230893266137,
+      "learning_rate": 2.4478921703164236e-06,
+      "loss": 0.1423,
+      "step": 2656
+    },
+    {
+      "epoch": 4.280306081353202,
+      "grad_norm": 0.22308055219912287,
+      "learning_rate": 2.4371073969954463e-06,
+      "loss": 0.1366,
+      "step": 2657
+    },
+    {
+      "epoch": 4.281917035843738,
+      "grad_norm": 0.22580670108095363,
+      "learning_rate": 2.4263448919885745e-06,
+      "loss": 0.1457,
+      "step": 2658
+    },
+    {
+      "epoch": 4.283527990334273,
+      "grad_norm": 0.21867003917030386,
+      "learning_rate": 2.4156046689417823e-06,
+      "loss": 0.149,
+      "step": 2659
+    },
+    {
+      "epoch": 4.2851389448248085,
+      "grad_norm": 0.21703864427650707,
+      "learning_rate": 2.4048867414728004e-06,
+      "loss": 0.1553,
+      "step": 2660
+    },
+    {
+      "epoch": 4.286749899315344,
+      "grad_norm": 0.22892483066011338,
+      "learning_rate": 2.394191123171081e-06,
+      "loss": 0.1364,
+      "step": 2661
+    },
+    {
+      "epoch": 4.28836085380588,
+      "grad_norm": 0.22083545583834652,
+      "learning_rate": 2.3835178275978012e-06,
+      "loss": 0.1294,
+      "step": 2662
+    },
+    {
+      "epoch": 4.289971808296416,
+      "grad_norm": 0.22446324510789786,
+      "learning_rate": 2.3728668682858193e-06,
+      "loss": 0.1419,
+      "step": 2663
+    },
+    {
+      "epoch": 4.291582762786951,
+      "grad_norm": 0.23116257281704497,
+      "learning_rate": 2.3622382587396907e-06,
+      "loss": 0.1445,
+      "step": 2664
+    },
+    {
+      "epoch": 4.293193717277487,
+      "grad_norm": 0.21998960949875104,
+      "learning_rate": 2.3516320124356186e-06,
+      "loss": 0.151,
+      "step": 2665
+    },
+    {
+      "epoch": 4.294804671768023,
+      "grad_norm": 0.2147293966021974,
+      "learning_rate": 2.3410481428214602e-06,
+      "loss": 0.1543,
+      "step": 2666
+    },
+    {
+      "epoch": 4.296415626258558,
+      "grad_norm": 0.21872334682296563,
+      "learning_rate": 2.330486663316702e-06,
+      "loss": 0.1503,
+      "step": 2667
+    },
+    {
+      "epoch": 4.298026580749093,
+      "grad_norm": 0.22709801396680795,
+      "learning_rate": 2.31994758731243e-06,
+      "loss": 0.1394,
+      "step": 2668
+    },
+    {
+      "epoch": 4.299637535239629,
+      "grad_norm": 0.22239888881280748,
+      "learning_rate": 2.309430928171341e-06,
+      "loss": 0.1456,
+      "step": 2669
+    },
+    {
+      "epoch": 4.301248489730165,
+      "grad_norm": 0.22267661198438218,
+      "learning_rate": 2.2989366992276917e-06,
+      "loss": 0.14,
+      "step": 2670
+    },
+    {
+      "epoch": 4.302859444220701,
+      "grad_norm": 0.22089398672152466,
+      "learning_rate": 2.288464913787316e-06,
+      "loss": 0.1505,
+      "step": 2671
+    },
+    {
+      "epoch": 4.304470398711237,
+      "grad_norm": 0.22234097375462933,
+      "learning_rate": 2.278015585127573e-06,
+      "loss": 0.1365,
+      "step": 2672
+    },
+    {
+      "epoch": 4.306081353201772,
+      "grad_norm": 0.2268142943570914,
+      "learning_rate": 2.2675887264973694e-06,
+      "loss": 0.1521,
+      "step": 2673
+    },
+    {
+      "epoch": 4.3076923076923075,
+      "grad_norm": 0.20658475916341448,
+      "learning_rate": 2.257184351117101e-06,
+      "loss": 0.1387,
+      "step": 2674
+    },
+    {
+      "epoch": 4.309303262182843,
+      "grad_norm": 0.2302415242479479,
+      "learning_rate": 2.246802472178675e-06,
+      "loss": 0.1446,
+      "step": 2675
+    },
+    {
+      "epoch": 4.310914216673379,
+      "grad_norm": 0.22560731900901113,
+      "learning_rate": 2.236443102845458e-06,
+      "loss": 0.1525,
+      "step": 2676
+    },
+    {
+      "epoch": 4.312525171163915,
+      "grad_norm": 0.2234748667770128,
+      "learning_rate": 2.2261062562522896e-06,
+      "loss": 0.1549,
+      "step": 2677
+    },
+    {
+      "epoch": 4.31413612565445,
+      "grad_norm": 0.22870884705411906,
+      "learning_rate": 2.21579194550545e-06,
+      "loss": 0.1386,
+      "step": 2678
+    },
+    {
+      "epoch": 4.315747080144986,
+      "grad_norm": 0.2224606918323668,
+      "learning_rate": 2.2055001836826364e-06,
+      "loss": 0.1455,
+      "step": 2679
+    },
+    {
+      "epoch": 4.317358034635522,
+      "grad_norm": 0.22036372420401462,
+      "learning_rate": 2.19523098383297e-06,
+      "loss": 0.1508,
+      "step": 2680
+    },
+    {
+      "epoch": 4.318968989126057,
+      "grad_norm": 0.21510144959446054,
+      "learning_rate": 2.1849843589769513e-06,
+      "loss": 0.143,
+      "step": 2681
+    },
+    {
+      "epoch": 4.320579943616593,
+      "grad_norm": 0.2359493205380365,
+      "learning_rate": 2.1747603221064684e-06,
+      "loss": 0.1585,
+      "step": 2682
+    },
+    {
+      "epoch": 4.322190898107128,
+      "grad_norm": 0.21593001357688618,
+      "learning_rate": 2.164558886184762e-06,
+      "loss": 0.1323,
+      "step": 2683
+    },
+    {
+      "epoch": 4.323801852597664,
+      "grad_norm": 0.2258254732968412,
+      "learning_rate": 2.1543800641464263e-06,
+      "loss": 0.141,
+      "step": 2684
+    },
+    {
+      "epoch": 4.3254128070882,
+      "grad_norm": 0.22439808272135298,
+      "learning_rate": 2.1442238688973682e-06,
+      "loss": 0.1396,
+      "step": 2685
+    },
+    {
+      "epoch": 4.327023761578736,
+      "grad_norm": 0.21143922501895449,
+      "learning_rate": 2.1340903133148205e-06,
+      "loss": 0.1527,
+      "step": 2686
+    },
+    {
+      "epoch": 4.328634716069271,
+      "grad_norm": 0.22994834674226478,
+      "learning_rate": 2.123979410247301e-06,
+      "loss": 0.1544,
+      "step": 2687
+    },
+    {
+      "epoch": 4.3302456705598065,
+      "grad_norm": 0.22793741030782735,
+      "learning_rate": 2.1138911725146106e-06,
+      "loss": 0.156,
+      "step": 2688
+    },
+    {
+      "epoch": 4.331856625050342,
+      "grad_norm": 0.22320356349655823,
+      "learning_rate": 2.103825612907815e-06,
+      "loss": 0.1474,
+      "step": 2689
+    },
+    {
+      "epoch": 4.333467579540878,
+      "grad_norm": 0.23523038616188513,
+      "learning_rate": 2.093782744189217e-06,
+      "loss": 0.1417,
+      "step": 2690
+    },
+    {
+      "epoch": 4.335078534031414,
+      "grad_norm": 0.23099424056260986,
+      "learning_rate": 2.0837625790923543e-06,
+      "loss": 0.1432,
+      "step": 2691
+    },
+    {
+      "epoch": 4.336689488521949,
+      "grad_norm": 0.21683386647539887,
+      "learning_rate": 2.0737651303219787e-06,
+      "loss": 0.1584,
+      "step": 2692
+    },
+    {
+      "epoch": 4.338300443012485,
+      "grad_norm": 0.22280820135116566,
+      "learning_rate": 2.0637904105540428e-06,
+      "loss": 0.1394,
+      "step": 2693
+    },
+    {
+      "epoch": 4.339911397503021,
+      "grad_norm": 0.2281065955008403,
+      "learning_rate": 2.053838432435673e-06,
+      "loss": 0.1492,
+      "step": 2694
+    },
+    {
+      "epoch": 4.341522351993556,
+      "grad_norm": 0.20990894608507565,
+      "learning_rate": 2.0439092085851685e-06,
+      "loss": 0.1541,
+      "step": 2695
+    },
+    {
+      "epoch": 4.343133306484092,
+      "grad_norm": 0.2410571666539787,
+      "learning_rate": 2.0340027515919704e-06,
+      "loss": 0.1504,
+      "step": 2696
+    },
+    {
+      "epoch": 4.344744260974627,
+      "grad_norm": 0.21521786871981052,
+      "learning_rate": 2.024119074016664e-06,
+      "loss": 0.1458,
+      "step": 2697
+    },
+    {
+      "epoch": 4.346355215465163,
+      "grad_norm": 0.23352779441098917,
+      "learning_rate": 2.014258188390936e-06,
+      "loss": 0.1319,
+      "step": 2698
+    },
+    {
+      "epoch": 4.347966169955699,
+      "grad_norm": 0.21861563397638975,
+      "learning_rate": 2.0044201072175884e-06,
+      "loss": 0.1493,
+      "step": 2699
+    },
+    {
+      "epoch": 4.349577124446235,
+      "grad_norm": 0.23187235646614845,
+      "learning_rate": 1.9946048429705133e-06,
+      "loss": 0.1424,
+      "step": 2700
+    },
+    {
+      "epoch": 4.35118807893677,
+      "grad_norm": 0.23314062934543647,
+      "learning_rate": 1.984812408094656e-06,
+      "loss": 0.1443,
+      "step": 2701
+    },
+    {
+      "epoch": 4.3527990334273055,
+      "grad_norm": 0.22716318100135857,
+      "learning_rate": 1.975042815006023e-06,
+      "loss": 0.1506,
+      "step": 2702
+    },
+    {
+      "epoch": 4.354409987917841,
+      "grad_norm": 0.22789140251356194,
+      "learning_rate": 1.9652960760916627e-06,
+      "loss": 0.1509,
+      "step": 2703
+    },
+    {
+      "epoch": 4.356020942408377,
+      "grad_norm": 0.2298570051084699,
+      "learning_rate": 1.95557220370965e-06,
+      "loss": 0.1388,
+      "step": 2704
+    },
+    {
+      "epoch": 4.357631896898913,
+      "grad_norm": 0.22855340474841207,
+      "learning_rate": 1.945871210189054e-06,
+      "loss": 0.1609,
+      "step": 2705
+    },
+    {
+      "epoch": 4.359242851389448,
+      "grad_norm": 0.22885905350588037,
+      "learning_rate": 1.9361931078299443e-06,
+      "loss": 0.166,
+      "step": 2706
+    },
+    {
+      "epoch": 4.360853805879984,
+      "grad_norm": 0.21861981366054756,
+      "learning_rate": 1.92653790890337e-06,
+      "loss": 0.138,
+      "step": 2707
+    },
+    {
+      "epoch": 4.36246476037052,
+      "grad_norm": 0.23241034579490025,
+      "learning_rate": 1.916905625651331e-06,
+      "loss": 0.148,
+      "step": 2708
+    },
+    {
+      "epoch": 4.364075714861055,
+      "grad_norm": 0.22772695647417238,
+      "learning_rate": 1.9072962702867714e-06,
+      "loss": 0.1388,
+      "step": 2709
+    },
+    {
+      "epoch": 4.365686669351591,
+      "grad_norm": 0.21875331128783337,
+      "learning_rate": 1.8977098549935745e-06,
+      "loss": 0.1417,
+      "step": 2710
+    },
+    {
+      "epoch": 4.367297623842126,
+      "grad_norm": 0.22606749157296668,
+      "learning_rate": 1.8881463919265374e-06,
+      "loss": 0.1512,
+      "step": 2711
+    },
+    {
+      "epoch": 4.368908578332662,
+      "grad_norm": 0.232376385563654,
+      "learning_rate": 1.8786058932113428e-06,
+      "loss": 0.1519,
+      "step": 2712
+    },
+    {
+      "epoch": 4.370519532823198,
+      "grad_norm": 0.22299262190690977,
+      "learning_rate": 1.8690883709445652e-06,
+      "loss": 0.137,
+      "step": 2713
+    },
+    {
+      "epoch": 4.372130487313734,
+      "grad_norm": 0.22002068793259277,
+      "learning_rate": 1.859593837193645e-06,
+      "loss": 0.1624,
+      "step": 2714
+    },
+    {
+      "epoch": 4.373741441804269,
+      "grad_norm": 0.21192028521319847,
+      "learning_rate": 1.850122303996882e-06,
+      "loss": 0.1363,
+      "step": 2715
+    },
+    {
+      "epoch": 4.375352396294804,
+      "grad_norm": 0.21021284454907238,
+      "learning_rate": 1.8406737833634024e-06,
+      "loss": 0.142,
+      "step": 2716
+    },
+    {
+      "epoch": 4.37696335078534,
+      "grad_norm": 0.21658493528620826,
+      "learning_rate": 1.8312482872731553e-06,
+      "loss": 0.1456,
+      "step": 2717
+    },
+    {
+      "epoch": 4.378574305275876,
+      "grad_norm": 0.226219598466499,
+      "learning_rate": 1.8218458276769091e-06,
+      "loss": 0.1518,
+      "step": 2718
+    },
+    {
+      "epoch": 4.380185259766412,
+      "grad_norm": 0.21769790494716323,
+      "learning_rate": 1.8124664164962124e-06,
+      "loss": 0.1481,
+      "step": 2719
+    },
+    {
+      "epoch": 4.381796214256947,
+      "grad_norm": 0.22650672628751944,
+      "learning_rate": 1.803110065623388e-06,
+      "loss": 0.1566,
+      "step": 2720
+    },
+    {
+      "epoch": 4.383407168747483,
+      "grad_norm": 0.2125040635428062,
+      "learning_rate": 1.7937767869215284e-06,
+      "loss": 0.1363,
+      "step": 2721
+    },
+    {
+      "epoch": 4.3850181232380185,
+      "grad_norm": 0.22322895400847495,
+      "learning_rate": 1.784466592224472e-06,
+      "loss": 0.1463,
+      "step": 2722
+    },
+    {
+      "epoch": 4.386629077728554,
+      "grad_norm": 0.21766007587217553,
+      "learning_rate": 1.7751794933367828e-06,
+      "loss": 0.1509,
+      "step": 2723
+    },
+    {
+      "epoch": 4.38824003221909,
+      "grad_norm": 0.22239483211233982,
+      "learning_rate": 1.7659155020337392e-06,
+      "loss": 0.1552,
+      "step": 2724
+    },
+    {
+      "epoch": 4.389850986709625,
+      "grad_norm": 0.22028648628947187,
+      "learning_rate": 1.7566746300613325e-06,
+      "loss": 0.1551,
+      "step": 2725
+    },
+    {
+      "epoch": 4.391461941200161,
+      "grad_norm": 0.2212460302725368,
+      "learning_rate": 1.7474568891362342e-06,
+      "loss": 0.1388,
+      "step": 2726
+    },
+    {
+      "epoch": 4.393072895690697,
+      "grad_norm": 0.2215072355482207,
+      "learning_rate": 1.738262290945787e-06,
+      "loss": 0.1539,
+      "step": 2727
+    },
+    {
+      "epoch": 4.394683850181233,
+      "grad_norm": 0.21970291081765841,
+      "learning_rate": 1.7290908471479805e-06,
+      "loss": 0.1612,
+      "step": 2728
+    },
+    {
+      "epoch": 4.3962948046717685,
+      "grad_norm": 0.23823342781119003,
+      "learning_rate": 1.7199425693714733e-06,
+      "loss": 0.1489,
+      "step": 2729
+    },
+    {
+      "epoch": 4.397905759162303,
+      "grad_norm": 0.21723293567882335,
+      "learning_rate": 1.7108174692155266e-06,
+      "loss": 0.1489,
+      "step": 2730
+    },
+    {
+      "epoch": 4.399516713652839,
+      "grad_norm": 0.22473337832629095,
+      "learning_rate": 1.701715558250019e-06,
+      "loss": 0.1339,
+      "step": 2731
+    },
+    {
+      "epoch": 4.401127668143375,
+      "grad_norm": 0.23218370282919404,
+      "learning_rate": 1.6926368480154344e-06,
+      "loss": 0.1534,
+      "step": 2732
+    },
+    {
+      "epoch": 4.402738622633911,
+      "grad_norm": 0.22527228406674732,
+      "learning_rate": 1.683581350022838e-06,
+      "loss": 0.1492,
+      "step": 2733
+    },
+    {
+      "epoch": 4.404349577124446,
+      "grad_norm": 0.22299618930067308,
+      "learning_rate": 1.674549075753862e-06,
+      "loss": 0.1559,
+      "step": 2734
+    },
+    {
+      "epoch": 4.405960531614982,
+      "grad_norm": 0.2173548439071365,
+      "learning_rate": 1.6655400366606867e-06,
+      "loss": 0.1378,
+      "step": 2735
+    },
+    {
+      "epoch": 4.4075714861055175,
+      "grad_norm": 0.21748819640917166,
+      "learning_rate": 1.656554244166042e-06,
+      "loss": 0.1418,
+      "step": 2736
+    },
+    {
+      "epoch": 4.409182440596053,
+      "grad_norm": 0.22286227976513667,
+      "learning_rate": 1.6475917096631855e-06,
+      "loss": 0.1665,
+      "step": 2737
+    },
+    {
+      "epoch": 4.410793395086589,
+      "grad_norm": 0.21834466918449652,
+      "learning_rate": 1.6386524445158714e-06,
+      "loss": 0.1451,
+      "step": 2738
+    },
+    {
+      "epoch": 4.412404349577124,
+      "grad_norm": 0.2155920749793083,
+      "learning_rate": 1.6297364600583554e-06,
+      "loss": 0.1488,
+      "step": 2739
+    },
+    {
+      "epoch": 4.41401530406766,
+      "grad_norm": 0.21694950266493748,
+      "learning_rate": 1.620843767595388e-06,
+      "loss": 0.1281,
+      "step": 2740
+    },
+    {
+      "epoch": 4.415626258558196,
+      "grad_norm": 0.22513033325891293,
+      "learning_rate": 1.6119743784021725e-06,
+      "loss": 0.1465,
+      "step": 2741
+    },
+    {
+      "epoch": 4.417237213048732,
+      "grad_norm": 0.2110624349498471,
+      "learning_rate": 1.6031283037243684e-06,
+      "loss": 0.1494,
+      "step": 2742
+    },
+    {
+      "epoch": 4.418848167539267,
+      "grad_norm": 0.22603749621498606,
+      "learning_rate": 1.594305554778075e-06,
+      "loss": 0.1497,
+      "step": 2743
+    },
+    {
+      "epoch": 4.420459122029802,
+      "grad_norm": 0.23203113436121725,
+      "learning_rate": 1.5855061427498263e-06,
+      "loss": 0.1441,
+      "step": 2744
+    },
+    {
+      "epoch": 4.422070076520338,
+      "grad_norm": 0.23890073771122725,
+      "learning_rate": 1.5767300787965512e-06,
+      "loss": 0.1569,
+      "step": 2745
+    },
+    {
+      "epoch": 4.423681031010874,
+      "grad_norm": 0.21934839826909436,
+      "learning_rate": 1.5679773740455817e-06,
+      "loss": 0.156,
+      "step": 2746
+    },
+    {
+      "epoch": 4.42529198550141,
+      "grad_norm": 0.21682203320832888,
+      "learning_rate": 1.5592480395946342e-06,
+      "loss": 0.1399,
+      "step": 2747
+    },
+    {
+      "epoch": 4.426902939991945,
+      "grad_norm": 0.21588619784464438,
+      "learning_rate": 1.5505420865117993e-06,
+      "loss": 0.1427,
+      "step": 2748
+    },
+    {
+      "epoch": 4.428513894482481,
+      "grad_norm": 0.2196132757827625,
+      "learning_rate": 1.541859525835505e-06,
+      "loss": 0.1416,
+      "step": 2749
+    },
+    {
+      "epoch": 4.4301248489730165,
+      "grad_norm": 0.22650353348090568,
+      "learning_rate": 1.5332003685745279e-06,
+      "loss": 0.1428,
+      "step": 2750
+    },
+    {
+      "epoch": 4.431735803463552,
+      "grad_norm": 0.2163231272238479,
+      "learning_rate": 1.524564625707985e-06,
+      "loss": 0.1592,
+      "step": 2751
+    },
+    {
+      "epoch": 4.433346757954088,
+      "grad_norm": 0.21345314169858762,
+      "learning_rate": 1.5159523081852867e-06,
+      "loss": 0.1491,
+      "step": 2752
+    },
+    {
+      "epoch": 4.434957712444623,
+      "grad_norm": 0.21724471459393288,
+      "learning_rate": 1.5073634269261427e-06,
+      "loss": 0.1437,
+      "step": 2753
+    },
+    {
+      "epoch": 4.436568666935159,
+      "grad_norm": 0.21107110747167726,
+      "learning_rate": 1.4987979928205599e-06,
+      "loss": 0.1498,
+      "step": 2754
+    },
+    {
+      "epoch": 4.438179621425695,
+      "grad_norm": 0.23243728370562228,
+      "learning_rate": 1.4902560167288105e-06,
+      "loss": 0.1445,
+      "step": 2755
+    },
+    {
+      "epoch": 4.439790575916231,
+      "grad_norm": 0.20420200944872982,
+      "learning_rate": 1.4817375094814202e-06,
+      "loss": 0.1411,
+      "step": 2756
+    },
+    {
+      "epoch": 4.441401530406766,
+      "grad_norm": 0.23431831960543478,
+      "learning_rate": 1.473242481879158e-06,
+      "loss": 0.1596,
+      "step": 2757
+    },
+    {
+      "epoch": 4.443012484897301,
+      "grad_norm": 0.22563203510054342,
+      "learning_rate": 1.464770944693028e-06,
+      "loss": 0.1306,
+      "step": 2758
+    },
+    {
+      "epoch": 4.444623439387837,
+      "grad_norm": 0.21898437843550095,
+      "learning_rate": 1.4563229086642538e-06,
+      "loss": 0.1639,
+      "step": 2759
+    },
+    {
+      "epoch": 4.446234393878373,
+      "grad_norm": 0.21799291473907836,
+      "learning_rate": 1.4478983845042493e-06,
+      "loss": 0.1409,
+      "step": 2760
+    },
+    {
+      "epoch": 4.447845348368909,
+      "grad_norm": 0.21834671784511098,
+      "learning_rate": 1.439497382894617e-06,
+      "loss": 0.1451,
+      "step": 2761
+    },
+    {
+      "epoch": 4.449456302859444,
+      "grad_norm": 0.21784261633030536,
+      "learning_rate": 1.4311199144871534e-06,
+      "loss": 0.1498,
+      "step": 2762
+    },
+    {
+      "epoch": 4.45106725734998,
+      "grad_norm": 0.21551061217502962,
+      "learning_rate": 1.4227659899038004e-06,
+      "loss": 0.1483,
+      "step": 2763
+    },
+    {
+      "epoch": 4.4526782118405155,
+      "grad_norm": 0.22482929437855542,
+      "learning_rate": 1.4144356197366494e-06,
+      "loss": 0.1447,
+      "step": 2764
+    },
+    {
+      "epoch": 4.454289166331051,
+      "grad_norm": 0.21817533638318648,
+      "learning_rate": 1.406128814547929e-06,
+      "loss": 0.1583,
+      "step": 2765
+    },
+    {
+      "epoch": 4.455900120821587,
+      "grad_norm": 0.22826403127027628,
+      "learning_rate": 1.397845584869999e-06,
+      "loss": 0.1391,
+      "step": 2766
+    },
+    {
+      "epoch": 4.457511075312122,
+      "grad_norm": 0.22368572318707833,
+      "learning_rate": 1.3895859412053093e-06,
+      "loss": 0.1378,
+      "step": 2767
+    },
+    {
+      "epoch": 4.459122029802658,
+      "grad_norm": 0.2249853087961331,
+      "learning_rate": 1.3813498940264136e-06,
+      "loss": 0.1463,
+      "step": 2768
+    },
+    {
+      "epoch": 4.460732984293194,
+      "grad_norm": 0.2185064787669113,
+      "learning_rate": 1.3731374537759544e-06,
+      "loss": 0.1396,
+      "step": 2769
+    },
+    {
+      "epoch": 4.46234393878373,
+      "grad_norm": 0.21864596897479208,
+      "learning_rate": 1.3649486308666314e-06,
+      "loss": 0.1565,
+      "step": 2770
+    },
+    {
+      "epoch": 4.4639548932742645,
+      "grad_norm": 0.22727139547878636,
+      "learning_rate": 1.3567834356812015e-06,
+      "loss": 0.1437,
+      "step": 2771
+    },
+    {
+      "epoch": 4.4655658477648,
+      "grad_norm": 0.2233584478305029,
+      "learning_rate": 1.3486418785724697e-06,
+      "loss": 0.144,
+      "step": 2772
+    },
+    {
+      "epoch": 4.467176802255336,
+      "grad_norm": 0.351620925278512,
+      "learning_rate": 1.3405239698632654e-06,
+      "loss": 0.1532,
+      "step": 2773
+    },
+    {
+      "epoch": 4.468787756745872,
+      "grad_norm": 0.2215381503484612,
+      "learning_rate": 1.332429719846433e-06,
+      "loss": 0.1398,
+      "step": 2774
+    },
+    {
+      "epoch": 4.470398711236408,
+      "grad_norm": 0.2232986049913044,
+      "learning_rate": 1.3243591387848164e-06,
+      "loss": 0.1485,
+      "step": 2775
+    },
+    {
+      "epoch": 4.472009665726943,
+      "grad_norm": 0.2202202269211146,
+      "learning_rate": 1.3163122369112591e-06,
+      "loss": 0.1574,
+      "step": 2776
+    },
+    {
+      "epoch": 4.473620620217479,
+      "grad_norm": 0.22918240497697517,
+      "learning_rate": 1.3082890244285773e-06,
+      "loss": 0.1447,
+      "step": 2777
+    },
+    {
+      "epoch": 4.4752315747080145,
+      "grad_norm": 0.22566546540240492,
+      "learning_rate": 1.300289511509547e-06,
+      "loss": 0.1463,
+      "step": 2778
+    },
+    {
+      "epoch": 4.47684252919855,
+      "grad_norm": 0.21415981839053075,
+      "learning_rate": 1.292313708296893e-06,
+      "loss": 0.1542,
+      "step": 2779
+    },
+    {
+      "epoch": 4.478453483689086,
+      "grad_norm": 0.221053938368025,
+      "learning_rate": 1.2843616249032874e-06,
+      "loss": 0.1522,
+      "step": 2780
+    },
+    {
+      "epoch": 4.480064438179621,
+      "grad_norm": 0.21618116765332307,
+      "learning_rate": 1.2764332714113258e-06,
+      "loss": 0.1546,
+      "step": 2781
+    },
+    {
+      "epoch": 4.481675392670157,
+      "grad_norm": 0.21974552955078566,
+      "learning_rate": 1.2685286578735045e-06,
+      "loss": 0.143,
+      "step": 2782
+    },
+    {
+      "epoch": 4.483286347160693,
+      "grad_norm": 0.22078570986976645,
+      "learning_rate": 1.2606477943122352e-06,
+      "loss": 0.1449,
+      "step": 2783
+    },
+    {
+      "epoch": 4.484897301651229,
+      "grad_norm": 0.22916711988692676,
+      "learning_rate": 1.2527906907198094e-06,
+      "loss": 0.1422,
+      "step": 2784
+    },
+    {
+      "epoch": 4.486508256141764,
+      "grad_norm": 0.22001716488332673,
+      "learning_rate": 1.244957357058394e-06,
+      "loss": 0.1442,
+      "step": 2785
+    },
+    {
+      "epoch": 4.488119210632299,
+      "grad_norm": 0.22386383475134952,
+      "learning_rate": 1.2371478032600083e-06,
+      "loss": 0.1601,
+      "step": 2786
+    },
+    {
+      "epoch": 4.489730165122835,
+      "grad_norm": 0.21664981131161265,
+      "learning_rate": 1.2293620392265338e-06,
+      "loss": 0.1415,
+      "step": 2787
+    },
+    {
+      "epoch": 4.491341119613371,
+      "grad_norm": 0.22629217846588368,
+      "learning_rate": 1.2216000748296897e-06,
+      "loss": 0.1433,
+      "step": 2788
+    },
+    {
+      "epoch": 4.492952074103907,
+      "grad_norm": 0.235019336289307,
+      "learning_rate": 1.213861919911008e-06,
+      "loss": 0.1494,
+      "step": 2789
+    },
+    {
+      "epoch": 4.494563028594442,
+      "grad_norm": 0.21964906508530987,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.1425,
+      "step": 2790
+    },
+    {
+      "epoch": 4.496173983084978,
+      "grad_norm": 0.21452579914651837,
+      "learning_rate": 1.1984570777233184e-06,
+      "loss": 0.1548,
+      "step": 2791
+    },
+    {
+      "epoch": 4.4977849375755135,
+      "grad_norm": 0.22707174901989768,
+      "learning_rate": 1.1907904099863999e-06,
+      "loss": 0.1631,
+      "step": 2792
+    },
+    {
+      "epoch": 4.499395892066049,
+      "grad_norm": 0.2155179519101744,
+      "learning_rate": 1.18314759079178e-06,
+      "loss": 0.1443,
+      "step": 2793
+    },
+    {
+      "epoch": 4.501006846556585,
+      "grad_norm": 0.2226635289792672,
+      "learning_rate": 1.1755286298299339e-06,
+      "loss": 0.161,
+      "step": 2794
+    },
+    {
+      "epoch": 4.50261780104712,
+      "grad_norm": 0.22080540831878404,
+      "learning_rate": 1.1679335367610855e-06,
+      "loss": 0.1413,
+      "step": 2795
+    },
+    {
+      "epoch": 4.504228755537656,
+      "grad_norm": 0.22308278536499784,
+      "learning_rate": 1.1603623212151872e-06,
+      "loss": 0.1365,
+      "step": 2796
+    },
+    {
+      "epoch": 4.505839710028192,
+      "grad_norm": 0.23108606533673448,
+      "learning_rate": 1.152814992791922e-06,
+      "loss": 0.1393,
+      "step": 2797
+    },
+    {
+      "epoch": 4.507450664518728,
+      "grad_norm": 0.21861727723987914,
+      "learning_rate": 1.1452915610606885e-06,
+      "loss": 0.1497,
+      "step": 2798
+    },
+    {
+      "epoch": 4.5090616190092625,
+      "grad_norm": 0.2218833710624545,
+      "learning_rate": 1.1377920355605854e-06,
+      "loss": 0.1468,
+      "step": 2799
+    },
+    {
+      "epoch": 4.510672573499798,
+      "grad_norm": 0.21908566335116011,
+      "learning_rate": 1.1303164258003974e-06,
+      "loss": 0.1473,
+      "step": 2800
+    },
+    {
+      "epoch": 4.512283527990334,
+      "grad_norm": 0.21884018169901615,
+      "learning_rate": 1.1228647412585847e-06,
+      "loss": 0.1318,
+      "step": 2801
+    },
+    {
+      "epoch": 4.51389448248087,
+      "grad_norm": 0.22464962629724394,
+      "learning_rate": 1.1154369913832762e-06,
+      "loss": 0.1461,
+      "step": 2802
+    },
+    {
+      "epoch": 4.515505436971406,
+      "grad_norm": 0.2233609804572511,
+      "learning_rate": 1.1080331855922588e-06,
+      "loss": 0.1448,
+      "step": 2803
+    },
+    {
+      "epoch": 4.517116391461942,
+      "grad_norm": 0.21225483736906917,
+      "learning_rate": 1.100653333272943e-06,
+      "loss": 0.1452,
+      "step": 2804
+    },
+    {
+      "epoch": 4.518727345952477,
+      "grad_norm": 0.2269518895274179,
+      "learning_rate": 1.0932974437823884e-06,
+      "loss": 0.1451,
+      "step": 2805
+    },
+    {
+      "epoch": 4.520338300443012,
+      "grad_norm": 0.2261411870254825,
+      "learning_rate": 1.0859655264472568e-06,
+      "loss": 0.1538,
+      "step": 2806
+    },
+    {
+      "epoch": 4.521949254933548,
+      "grad_norm": 0.22345780419863742,
+      "learning_rate": 1.078657590563823e-06,
+      "loss": 0.1295,
+      "step": 2807
+    },
+    {
+      "epoch": 4.523560209424084,
+      "grad_norm": 0.21931553583779848,
+      "learning_rate": 1.0713736453979528e-06,
+      "loss": 0.1419,
+      "step": 2808
+    },
+    {
+      "epoch": 4.525171163914619,
+      "grad_norm": 0.22080634847733432,
+      "learning_rate": 1.064113700185092e-06,
+      "loss": 0.1439,
+      "step": 2809
+    },
+    {
+      "epoch": 4.526782118405155,
+      "grad_norm": 0.2261480900316319,
+      "learning_rate": 1.0568777641302663e-06,
+      "loss": 0.153,
+      "step": 2810
+    },
+    {
+      "epoch": 4.528393072895691,
+      "grad_norm": 0.22439464056795858,
+      "learning_rate": 1.0496658464080434e-06,
+      "loss": 0.1537,
+      "step": 2811
+    },
+    {
+      "epoch": 4.5300040273862265,
+      "grad_norm": 0.2248930635120482,
+      "learning_rate": 1.0424779561625465e-06,
+      "loss": 0.1488,
+      "step": 2812
+    },
+    {
+      "epoch": 4.531614981876762,
+      "grad_norm": 0.2069570637163575,
+      "learning_rate": 1.0353141025074364e-06,
+      "loss": 0.1647,
+      "step": 2813
+    },
+    {
+      "epoch": 4.533225936367297,
+      "grad_norm": 0.23132704696687886,
+      "learning_rate": 1.0281742945258987e-06,
+      "loss": 0.1462,
+      "step": 2814
+    },
+    {
+      "epoch": 4.534836890857833,
+      "grad_norm": 0.21337858571941506,
+      "learning_rate": 1.0210585412706187e-06,
+      "loss": 0.1487,
+      "step": 2815
+    },
+    {
+      "epoch": 4.536447845348369,
+      "grad_norm": 0.22309163223275602,
+      "learning_rate": 1.0139668517637991e-06,
+      "loss": 0.144,
+      "step": 2816
+    },
+    {
+      "epoch": 4.538058799838905,
+      "grad_norm": 0.21907142466944238,
+      "learning_rate": 1.006899234997114e-06,
+      "loss": 0.1446,
+      "step": 2817
+    },
+    {
+      "epoch": 4.53966975432944,
+      "grad_norm": 0.217901044651781,
+      "learning_rate": 9.998556999317334e-07,
+      "loss": 0.1367,
+      "step": 2818
+    },
+    {
+      "epoch": 4.541280708819976,
+      "grad_norm": 0.21574027266533383,
+      "learning_rate": 9.928362554982796e-07,
+      "loss": 0.1456,
+      "step": 2819
+    },
+    {
+      "epoch": 4.542891663310511,
+      "grad_norm": 0.22778052410416838,
+      "learning_rate": 9.858409105968337e-07,
+      "loss": 0.1468,
+      "step": 2820
+    },
+    {
+      "epoch": 4.544502617801047,
+      "grad_norm": 0.23571660112324896,
+      "learning_rate": 9.788696740969295e-07,
+      "loss": 0.1455,
+      "step": 2821
+    },
+    {
+      "epoch": 4.546113572291583,
+      "grad_norm": 0.2201132602209927,
+      "learning_rate": 9.71922554837521e-07,
+      "loss": 0.1446,
+      "step": 2822
+    },
+    {
+      "epoch": 4.547724526782118,
+      "grad_norm": 0.2170027257100728,
+      "learning_rate": 9.649995616269847e-07,
+      "loss": 0.1421,
+      "step": 2823
+    },
+    {
+      "epoch": 4.549335481272654,
+      "grad_norm": 0.21147627065327013,
+      "learning_rate": 9.581007032431144e-07,
+      "loss": 0.1486,
+      "step": 2824
+    },
+    {
+      "epoch": 4.55094643576319,
+      "grad_norm": 0.21922962408896876,
+      "learning_rate": 9.512259884331021e-07,
+      "loss": 0.138,
+      "step": 2825
+    },
+    {
+      "epoch": 4.5525573902537255,
+      "grad_norm": 0.21981668661472575,
+      "learning_rate": 9.443754259135197e-07,
+      "loss": 0.1415,
+      "step": 2826
+    },
+    {
+      "epoch": 4.554168344744261,
+      "grad_norm": 0.21062790215118363,
+      "learning_rate": 9.375490243703255e-07,
+      "loss": 0.1388,
+      "step": 2827
+    },
+    {
+      "epoch": 4.555779299234796,
+      "grad_norm": 0.20844877721721428,
+      "learning_rate": 9.307467924588364e-07,
+      "loss": 0.1425,
+      "step": 2828
+    },
+    {
+      "epoch": 4.557390253725332,
+      "grad_norm": 0.2169075920951353,
+      "learning_rate": 9.239687388037311e-07,
+      "loss": 0.1451,
+      "step": 2829
+    },
+    {
+      "epoch": 4.559001208215868,
+      "grad_norm": 0.2209905968566893,
+      "learning_rate": 9.172148719990237e-07,
+      "loss": 0.1419,
+      "step": 2830
+    },
+    {
+      "epoch": 4.560612162706404,
+      "grad_norm": 0.22168194810993105,
+      "learning_rate": 9.104852006080689e-07,
+      "loss": 0.1425,
+      "step": 2831
+    },
+    {
+      "epoch": 4.56222311719694,
+      "grad_norm": 0.23429944659967786,
+      "learning_rate": 9.03779733163539e-07,
+      "loss": 0.1352,
+      "step": 2832
+    },
+    {
+      "epoch": 4.563834071687475,
+      "grad_norm": 0.22251765674022653,
+      "learning_rate": 8.970984781674197e-07,
+      "loss": 0.158,
+      "step": 2833
+    },
+    {
+      "epoch": 4.56544502617801,
+      "grad_norm": 0.21439636612358917,
+      "learning_rate": 8.904414440909992e-07,
+      "loss": 0.1658,
+      "step": 2834
+    },
+    {
+      "epoch": 4.567055980668546,
+      "grad_norm": 0.21055760967769271,
+      "learning_rate": 8.83808639374848e-07,
+      "loss": 0.1406,
+      "step": 2835
+    },
+    {
+      "epoch": 4.568666935159082,
+      "grad_norm": 0.22812053637083266,
+      "learning_rate": 8.772000724288277e-07,
+      "loss": 0.1551,
+      "step": 2836
+    },
+    {
+      "epoch": 4.570277889649617,
+      "grad_norm": 0.23203123173864854,
+      "learning_rate": 8.706157516320557e-07,
+      "loss": 0.1457,
+      "step": 2837
+    },
+    {
+      "epoch": 4.571888844140153,
+      "grad_norm": 0.21647238240445146,
+      "learning_rate": 8.640556853329185e-07,
+      "loss": 0.1366,
+      "step": 2838
+    },
+    {
+      "epoch": 4.573499798630689,
+      "grad_norm": 0.2262132480851578,
+      "learning_rate": 8.575198818490405e-07,
+      "loss": 0.1532,
+      "step": 2839
+    },
+    {
+      "epoch": 4.5751107531212245,
+      "grad_norm": 0.22036901946478252,
+      "learning_rate": 8.510083494672905e-07,
+      "loss": 0.15,
+      "step": 2840
+    },
+    {
+      "epoch": 4.57672170761176,
+      "grad_norm": 0.20881526264861075,
+      "learning_rate": 8.445210964437556e-07,
+      "loss": 0.1572,
+      "step": 2841
+    },
+    {
+      "epoch": 4.578332662102295,
+      "grad_norm": 0.21997273950205165,
+      "learning_rate": 8.380581310037472e-07,
+      "loss": 0.1483,
+      "step": 2842
+    },
+    {
+      "epoch": 4.579943616592831,
+      "grad_norm": 0.2115529692321746,
+      "learning_rate": 8.316194613417749e-07,
+      "loss": 0.15,
+      "step": 2843
+    },
+    {
+      "epoch": 4.581554571083367,
+      "grad_norm": 0.2101065492853683,
+      "learning_rate": 8.252050956215462e-07,
+      "loss": 0.1327,
+      "step": 2844
+    },
+    {
+      "epoch": 4.583165525573903,
+      "grad_norm": 0.22367025563934423,
+      "learning_rate": 8.188150419759577e-07,
+      "loss": 0.1539,
+      "step": 2845
+    },
+    {
+      "epoch": 4.584776480064438,
+      "grad_norm": 0.21257442264917925,
+      "learning_rate": 8.124493085070706e-07,
+      "loss": 0.1459,
+      "step": 2846
+    },
+    {
+      "epoch": 4.5863874345549736,
+      "grad_norm": 0.2175995649893487,
+      "learning_rate": 8.061079032861197e-07,
+      "loss": 0.1362,
+      "step": 2847
+    },
+    {
+      "epoch": 4.587998389045509,
+      "grad_norm": 0.22731145011308046,
+      "learning_rate": 7.997908343534844e-07,
+      "loss": 0.1416,
+      "step": 2848
+    },
+    {
+      "epoch": 4.589609343536045,
+      "grad_norm": 0.230612422282459,
+      "learning_rate": 7.934981097186977e-07,
+      "loss": 0.1616,
+      "step": 2849
+    },
+    {
+      "epoch": 4.591220298026581,
+      "grad_norm": 0.22856353525173428,
+      "learning_rate": 7.872297373604154e-07,
+      "loss": 0.1481,
+      "step": 2850
+    },
+    {
+      "epoch": 4.592831252517117,
+      "grad_norm": 0.21655429075000812,
+      "learning_rate": 7.809857252264263e-07,
+      "loss": 0.1478,
+      "step": 2851
+    },
+    {
+      "epoch": 4.594442207007652,
+      "grad_norm": 0.2267420054274175,
+      "learning_rate": 7.747660812336221e-07,
+      "loss": 0.1386,
+      "step": 2852
+    },
+    {
+      "epoch": 4.596053161498188,
+      "grad_norm": 0.21517894418513714,
+      "learning_rate": 7.685708132680125e-07,
+      "loss": 0.1575,
+      "step": 2853
+    },
+    {
+      "epoch": 4.5976641159887235,
+      "grad_norm": 0.22771611741406794,
+      "learning_rate": 7.623999291846829e-07,
+      "loss": 0.1523,
+      "step": 2854
+    },
+    {
+      "epoch": 4.599275070479259,
+      "grad_norm": 0.22020469314547333,
+      "learning_rate": 7.562534368078167e-07,
+      "loss": 0.1312,
+      "step": 2855
+    },
+    {
+      "epoch": 4.600886024969794,
+      "grad_norm": 0.21911457185337718,
+      "learning_rate": 7.501313439306623e-07,
+      "loss": 0.1471,
+      "step": 2856
+    },
+    {
+      "epoch": 4.60249697946033,
+      "grad_norm": 0.220369515835811,
+      "learning_rate": 7.440336583155306e-07,
+      "loss": 0.1467,
+      "step": 2857
+    },
+    {
+      "epoch": 4.604107933950866,
+      "grad_norm": 0.2429223255819812,
+      "learning_rate": 7.379603876937969e-07,
+      "loss": 0.1393,
+      "step": 2858
+    },
+    {
+      "epoch": 4.605718888441402,
+      "grad_norm": 0.22295148519094363,
+      "learning_rate": 7.319115397658639e-07,
+      "loss": 0.1477,
+      "step": 2859
+    },
+    {
+      "epoch": 4.607329842931938,
+      "grad_norm": 0.2155898298714699,
+      "learning_rate": 7.258871222011832e-07,
+      "loss": 0.1365,
+      "step": 2860
+    },
+    {
+      "epoch": 4.6089407974224725,
+      "grad_norm": 0.21243596261299383,
+      "learning_rate": 7.198871426382203e-07,
+      "loss": 0.1558,
+      "step": 2861
+    },
+    {
+      "epoch": 4.610551751913008,
+      "grad_norm": 0.22122501445417958,
+      "learning_rate": 7.139116086844655e-07,
+      "loss": 0.161,
+      "step": 2862
+    },
+    {
+      "epoch": 4.612162706403544,
+      "grad_norm": 0.21598365968425512,
+      "learning_rate": 7.079605279163982e-07,
+      "loss": 0.1493,
+      "step": 2863
+    },
+    {
+      "epoch": 4.61377366089408,
+      "grad_norm": 0.2597213786138549,
+      "learning_rate": 7.020339078795136e-07,
+      "loss": 0.1559,
+      "step": 2864
+    },
+    {
+      "epoch": 4.615384615384615,
+      "grad_norm": 0.22332642167151248,
+      "learning_rate": 6.961317560882741e-07,
+      "loss": 0.1501,
+      "step": 2865
+    },
+    {
+      "epoch": 4.616995569875151,
+      "grad_norm": 0.21904255318165475,
+      "learning_rate": 6.902540800261292e-07,
+      "loss": 0.1627,
+      "step": 2866
+    },
+    {
+      "epoch": 4.618606524365687,
+      "grad_norm": 0.2257051902642088,
+      "learning_rate": 6.844008871454955e-07,
+      "loss": 0.1383,
+      "step": 2867
+    },
+    {
+      "epoch": 4.6202174788562225,
+      "grad_norm": 0.2028287385389341,
+      "learning_rate": 6.785721848677406e-07,
+      "loss": 0.1522,
+      "step": 2868
+    },
+    {
+      "epoch": 4.621828433346758,
+      "grad_norm": 0.21967388469905091,
+      "learning_rate": 6.727679805831821e-07,
+      "loss": 0.136,
+      "step": 2869
+    },
+    {
+      "epoch": 4.623439387837293,
+      "grad_norm": 0.2254164663996422,
+      "learning_rate": 6.669882816510776e-07,
+      "loss": 0.139,
+      "step": 2870
+    },
+    {
+      "epoch": 4.625050342327829,
+      "grad_norm": 0.21946694917764376,
+      "learning_rate": 6.61233095399616e-07,
+      "loss": 0.1387,
+      "step": 2871
+    },
+    {
+      "epoch": 4.626661296818365,
+      "grad_norm": 0.22027332476566938,
+      "learning_rate": 6.555024291259005e-07,
+      "loss": 0.1447,
+      "step": 2872
+    },
+    {
+      "epoch": 4.628272251308901,
+      "grad_norm": 0.21396999455794583,
+      "learning_rate": 6.497962900959542e-07,
+      "loss": 0.1556,
+      "step": 2873
+    },
+    {
+      "epoch": 4.629883205799436,
+      "grad_norm": 0.23178083820005146,
+      "learning_rate": 6.441146855446856e-07,
+      "loss": 0.135,
+      "step": 2874
+    },
+    {
+      "epoch": 4.6314941602899715,
+      "grad_norm": 0.21644352493998795,
+      "learning_rate": 6.384576226759165e-07,
+      "loss": 0.1559,
+      "step": 2875
+    },
+    {
+      "epoch": 4.633105114780507,
+      "grad_norm": 0.22303322809496479,
+      "learning_rate": 6.328251086623294e-07,
+      "loss": 0.1427,
+      "step": 2876
+    },
+    {
+      "epoch": 4.634716069271043,
+      "grad_norm": 0.21071259795990352,
+      "learning_rate": 6.272171506455005e-07,
+      "loss": 0.1459,
+      "step": 2877
+    },
+    {
+      "epoch": 4.636327023761579,
+      "grad_norm": 0.20980391695821105,
+      "learning_rate": 6.21633755735862e-07,
+      "loss": 0.138,
+      "step": 2878
+    },
+    {
+      "epoch": 4.637937978252115,
+      "grad_norm": 0.21263059217171254,
+      "learning_rate": 6.160749310127023e-07,
+      "loss": 0.1484,
+      "step": 2879
+    },
+    {
+      "epoch": 4.63954893274265,
+      "grad_norm": 0.204952813257361,
+      "learning_rate": 6.105406835241545e-07,
+      "loss": 0.1406,
+      "step": 2880
+    },
+    {
+      "epoch": 4.641159887233186,
+      "grad_norm": 0.21197948810614625,
+      "learning_rate": 6.050310202871922e-07,
+      "loss": 0.1281,
+      "step": 2881
+    },
+    {
+      "epoch": 4.6427708417237215,
+      "grad_norm": 0.20655087363236846,
+      "learning_rate": 5.995459482876253e-07,
+      "loss": 0.1522,
+      "step": 2882
+    },
+    {
+      "epoch": 4.644381796214257,
+      "grad_norm": 0.22413109041423177,
+      "learning_rate": 5.940854744800706e-07,
+      "loss": 0.1642,
+      "step": 2883
+    },
+    {
+      "epoch": 4.645992750704792,
+      "grad_norm": 0.21624222816321237,
+      "learning_rate": 5.886496057879676e-07,
+      "loss": 0.1359,
+      "step": 2884
+    },
+    {
+      "epoch": 4.647603705195328,
+      "grad_norm": 0.23678058476570954,
+      "learning_rate": 5.832383491035499e-07,
+      "loss": 0.1397,
+      "step": 2885
+    },
+    {
+      "epoch": 4.649214659685864,
+      "grad_norm": 0.21030051110070672,
+      "learning_rate": 5.778517112878512e-07,
+      "loss": 0.1485,
+      "step": 2886
+    },
+    {
+      "epoch": 4.6508256141764,
+      "grad_norm": 0.21914807690463745,
+      "learning_rate": 5.724896991706885e-07,
+      "loss": 0.1456,
+      "step": 2887
+    },
+    {
+      "epoch": 4.652436568666936,
+      "grad_norm": 0.21409204336065774,
+      "learning_rate": 5.671523195506567e-07,
+      "loss": 0.1543,
+      "step": 2888
+    },
+    {
+      "epoch": 4.6540475231574705,
+      "grad_norm": 0.21075887789724632,
+      "learning_rate": 5.618395791951159e-07,
+      "loss": 0.15,
+      "step": 2889
+    },
+    {
+      "epoch": 4.655658477648006,
+      "grad_norm": 0.22077872594297945,
+      "learning_rate": 5.565514848401887e-07,
+      "loss": 0.1354,
+      "step": 2890
+    },
+    {
+      "epoch": 4.657269432138542,
+      "grad_norm": 0.21354385616306004,
+      "learning_rate": 5.512880431907452e-07,
+      "loss": 0.1408,
+      "step": 2891
+    },
+    {
+      "epoch": 4.658880386629078,
+      "grad_norm": 0.212847283457817,
+      "learning_rate": 5.460492609203982e-07,
+      "loss": 0.1433,
+      "step": 2892
+    },
+    {
+      "epoch": 4.660491341119613,
+      "grad_norm": 0.2187454345219124,
+      "learning_rate": 5.40835144671501e-07,
+      "loss": 0.1465,
+      "step": 2893
+    },
+    {
+      "epoch": 4.662102295610149,
+      "grad_norm": 0.21640789825365564,
+      "learning_rate": 5.356457010551253e-07,
+      "loss": 0.1455,
+      "step": 2894
+    },
+    {
+      "epoch": 4.663713250100685,
+      "grad_norm": 0.21575352229791922,
+      "learning_rate": 5.304809366510566e-07,
+      "loss": 0.1556,
+      "step": 2895
+    },
+    {
+      "epoch": 4.66532420459122,
+      "grad_norm": 0.21723870718473914,
+      "learning_rate": 5.253408580078035e-07,
+      "loss": 0.1584,
+      "step": 2896
+    },
+    {
+      "epoch": 4.666935159081756,
+      "grad_norm": 0.22077657920712968,
+      "learning_rate": 5.202254716425636e-07,
+      "loss": 0.1502,
+      "step": 2897
+    },
+    {
+      "epoch": 4.668546113572291,
+      "grad_norm": 0.22750870337063364,
+      "learning_rate": 5.151347840412224e-07,
+      "loss": 0.1404,
+      "step": 2898
+    },
+    {
+      "epoch": 4.670157068062827,
+      "grad_norm": 0.21780471723045816,
+      "learning_rate": 5.100688016583632e-07,
+      "loss": 0.1317,
+      "step": 2899
+    },
+    {
+      "epoch": 4.671768022553363,
+      "grad_norm": 0.22372509521913855,
+      "learning_rate": 5.05027530917237e-07,
+      "loss": 0.1507,
+      "step": 2900
+    },
+    {
+      "epoch": 4.673378977043899,
+      "grad_norm": 0.21577961205407864,
+      "learning_rate": 5.00010978209764e-07,
+      "loss": 0.1498,
+      "step": 2901
+    },
+    {
+      "epoch": 4.6749899315344345,
+      "grad_norm": 0.22095270932077324,
+      "learning_rate": 4.950191498965207e-07,
+      "loss": 0.1567,
+      "step": 2902
+    },
+    {
+      "epoch": 4.6766008860249695,
+      "grad_norm": 0.22385680458863616,
+      "learning_rate": 4.900520523067376e-07,
+      "loss": 0.1387,
+      "step": 2903
+    },
+    {
+      "epoch": 4.678211840515505,
+      "grad_norm": 0.21480467606962006,
+      "learning_rate": 4.851096917382925e-07,
+      "loss": 0.1429,
+      "step": 2904
+    },
+    {
+      "epoch": 4.679822795006041,
+      "grad_norm": 0.21972524220304457,
+      "learning_rate": 4.801920744576949e-07,
+      "loss": 0.1466,
+      "step": 2905
+    },
+    {
+      "epoch": 4.681433749496577,
+      "grad_norm": 0.2143359141084018,
+      "learning_rate": 4.7529920670007724e-07,
+      "loss": 0.1559,
+      "step": 2906
+    },
+    {
+      "epoch": 4.683044703987113,
+      "grad_norm": 0.2195510458192481,
+      "learning_rate": 4.704310946692037e-07,
+      "loss": 0.1436,
+      "step": 2907
+    },
+    {
+      "epoch": 4.684655658477648,
+      "grad_norm": 0.2244650853064686,
+      "learning_rate": 4.6558774453743684e-07,
+      "loss": 0.1539,
+      "step": 2908
+    },
+    {
+      "epoch": 4.686266612968184,
+      "grad_norm": 0.2215405869664864,
+      "learning_rate": 4.607691624457511e-07,
+      "loss": 0.1387,
+      "step": 2909
+    },
+    {
+      "epoch": 4.687877567458719,
+      "grad_norm": 0.21906085385910223,
+      "learning_rate": 4.559753545037171e-07,
+      "loss": 0.1373,
+      "step": 2910
+    },
+    {
+      "epoch": 4.689488521949255,
+      "grad_norm": 0.21588538081675898,
+      "learning_rate": 4.512063267894906e-07,
+      "loss": 0.1598,
+      "step": 2911
+    },
+    {
+      "epoch": 4.69109947643979,
+      "grad_norm": 0.21543206275284182,
+      "learning_rate": 4.4646208534980807e-07,
+      "loss": 0.135,
+      "step": 2912
+    },
+    {
+      "epoch": 4.692710430930326,
+      "grad_norm": 0.20168686340533482,
+      "learning_rate": 4.4174263619998217e-07,
+      "loss": 0.138,
+      "step": 2913
+    },
+    {
+      "epoch": 4.694321385420862,
+      "grad_norm": 0.22114311160392375,
+      "learning_rate": 4.3704798532388624e-07,
+      "loss": 0.1521,
+      "step": 2914
+    },
+    {
+      "epoch": 4.695932339911398,
+      "grad_norm": 0.2165961430758482,
+      "learning_rate": 4.3237813867396117e-07,
+      "loss": 0.1398,
+      "step": 2915
+    },
+    {
+      "epoch": 4.6975432944019335,
+      "grad_norm": 0.21089525598082778,
+      "learning_rate": 4.2773310217118394e-07,
+      "loss": 0.1494,
+      "step": 2916
+    },
+    {
+      "epoch": 4.6991542488924685,
+      "grad_norm": 0.20868547364457915,
+      "learning_rate": 4.2311288170508336e-07,
+      "loss": 0.1356,
+      "step": 2917
+    },
+    {
+      "epoch": 4.700765203383004,
+      "grad_norm": 0.21119647807605038,
+      "learning_rate": 4.1851748313372463e-07,
+      "loss": 0.1551,
+      "step": 2918
+    },
+    {
+      "epoch": 4.70237615787354,
+      "grad_norm": 0.21654917077953273,
+      "learning_rate": 4.139469122836981e-07,
+      "loss": 0.1387,
+      "step": 2919
+    },
+    {
+      "epoch": 4.703987112364076,
+      "grad_norm": 0.21662833427984002,
+      "learning_rate": 4.094011749501103e-07,
+      "loss": 0.1502,
+      "step": 2920
+    },
+    {
+      "epoch": 4.705598066854611,
+      "grad_norm": 0.20946333432367129,
+      "learning_rate": 4.048802768965887e-07,
+      "loss": 0.1413,
+      "step": 2921
+    },
+    {
+      "epoch": 4.707209021345147,
+      "grad_norm": 0.22564245087432352,
+      "learning_rate": 4.003842238552613e-07,
+      "loss": 0.1545,
+      "step": 2922
+    },
+    {
+      "epoch": 4.708819975835683,
+      "grad_norm": 0.22076168759631085,
+      "learning_rate": 3.9591302152675703e-07,
+      "loss": 0.1591,
+      "step": 2923
+    },
+    {
+      "epoch": 4.710430930326218,
+      "grad_norm": 0.2097867230868185,
+      "learning_rate": 3.9146667558019433e-07,
+      "loss": 0.1571,
+      "step": 2924
+    },
+    {
+      "epoch": 4.712041884816754,
+      "grad_norm": 0.22088949709015748,
+      "learning_rate": 3.8704519165317923e-07,
+      "loss": 0.1565,
+      "step": 2925
+    },
+    {
+      "epoch": 4.71365283930729,
+      "grad_norm": 0.22150240058770918,
+      "learning_rate": 3.8264857535178943e-07,
+      "loss": 0.1324,
+      "step": 2926
+    },
+    {
+      "epoch": 4.715263793797825,
+      "grad_norm": 0.20698205346690662,
+      "learning_rate": 3.782768322505792e-07,
+      "loss": 0.1315,
+      "step": 2927
+    },
+    {
+      "epoch": 4.716874748288361,
+      "grad_norm": 0.20845720756492594,
+      "learning_rate": 3.7392996789255673e-07,
+      "loss": 0.1442,
+      "step": 2928
+    },
+    {
+      "epoch": 4.718485702778897,
+      "grad_norm": 0.2156287236788521,
+      "learning_rate": 3.6960798778919784e-07,
+      "loss": 0.1499,
+      "step": 2929
+    },
+    {
+      "epoch": 4.7200966572694325,
+      "grad_norm": 0.2196744399794609,
+      "learning_rate": 3.653108974204145e-07,
+      "loss": 0.1551,
+      "step": 2930
+    },
+    {
+      "epoch": 4.7217076117599674,
+      "grad_norm": 0.22139174195462538,
+      "learning_rate": 3.610387022345685e-07,
+      "loss": 0.1534,
+      "step": 2931
+    },
+    {
+      "epoch": 4.723318566250503,
+      "grad_norm": 0.2277220395224189,
+      "learning_rate": 3.567914076484558e-07,
+      "loss": 0.1498,
+      "step": 2932
+    },
+    {
+      "epoch": 4.724929520741039,
+      "grad_norm": 0.22109687230688657,
+      "learning_rate": 3.5256901904729967e-07,
+      "loss": 0.1562,
+      "step": 2933
+    },
+    {
+      "epoch": 4.726540475231575,
+      "grad_norm": 0.21580426849176798,
+      "learning_rate": 3.483715417847422e-07,
+      "loss": 0.1382,
+      "step": 2934
+    },
+    {
+      "epoch": 4.728151429722111,
+      "grad_norm": 0.2137311430685286,
+      "learning_rate": 3.441989811828417e-07,
+      "loss": 0.1354,
+      "step": 2935
+    },
+    {
+      "epoch": 4.729762384212646,
+      "grad_norm": 0.21046823458686373,
+      "learning_rate": 3.4005134253206393e-07,
+      "loss": 0.1464,
+      "step": 2936
+    },
+    {
+      "epoch": 4.7313733387031816,
+      "grad_norm": 0.20552298607096997,
+      "learning_rate": 3.3592863109128016e-07,
+      "loss": 0.1471,
+      "step": 2937
+    },
+    {
+      "epoch": 4.732984293193717,
+      "grad_norm": 0.2121097718281581,
+      "learning_rate": 3.318308520877489e-07,
+      "loss": 0.15,
+      "step": 2938
+    },
+    {
+      "epoch": 4.734595247684253,
+      "grad_norm": 0.22848848728985083,
+      "learning_rate": 3.277580107171163e-07,
+      "loss": 0.1485,
+      "step": 2939
+    },
+    {
+      "epoch": 4.736206202174788,
+      "grad_norm": 0.21565593846592726,
+      "learning_rate": 3.2371011214342053e-07,
+      "loss": 0.1582,
+      "step": 2940
+    },
+    {
+      "epoch": 4.737817156665324,
+      "grad_norm": 0.22164738957939945,
+      "learning_rate": 3.1968716149906043e-07,
+      "loss": 0.1453,
+      "step": 2941
+    },
+    {
+      "epoch": 4.73942811115586,
+      "grad_norm": 0.2168031711486235,
+      "learning_rate": 3.156891638848092e-07,
+      "loss": 0.1442,
+      "step": 2942
+    },
+    {
+      "epoch": 4.741039065646396,
+      "grad_norm": 0.2167016023126979,
+      "learning_rate": 3.117161243698052e-07,
+      "loss": 0.1426,
+      "step": 2943
+    },
+    {
+      "epoch": 4.7426500201369315,
+      "grad_norm": 0.21297127531599902,
+      "learning_rate": 3.077680479915368e-07,
+      "loss": 0.1462,
+      "step": 2944
+    },
+    {
+      "epoch": 4.744260974627466,
+      "grad_norm": 0.21958491536222166,
+      "learning_rate": 3.038449397558396e-07,
+      "loss": 0.1338,
+      "step": 2945
+    },
+    {
+      "epoch": 4.745871929118002,
+      "grad_norm": 0.21879043332580358,
+      "learning_rate": 2.9994680463689697e-07,
+      "loss": 0.1502,
+      "step": 2946
+    },
+    {
+      "epoch": 4.747482883608538,
+      "grad_norm": 0.21967915410535746,
+      "learning_rate": 2.9607364757722635e-07,
+      "loss": 0.15,
+      "step": 2947
+    },
+    {
+      "epoch": 4.749093838099074,
+      "grad_norm": 0.2113428697983794,
+      "learning_rate": 2.9222547348767504e-07,
+      "loss": 0.1267,
+      "step": 2948
+    },
+    {
+      "epoch": 4.750704792589609,
+      "grad_norm": 0.21023635797177284,
+      "learning_rate": 2.884022872474113e-07,
+      "loss": 0.1336,
+      "step": 2949
+    },
+    {
+      "epoch": 4.752315747080145,
+      "grad_norm": 0.2076654674576389,
+      "learning_rate": 2.8460409370392405e-07,
+      "loss": 0.1338,
+      "step": 2950
+    },
+    {
+      "epoch": 4.7539267015706805,
+      "grad_norm": 0.2371363457394752,
+      "learning_rate": 2.808308976730145e-07,
+      "loss": 0.1463,
+      "step": 2951
+    },
+    {
+      "epoch": 4.755537656061216,
+      "grad_norm": 0.21635183048072082,
+      "learning_rate": 2.770827039387869e-07,
+      "loss": 0.1327,
+      "step": 2952
+    },
+    {
+      "epoch": 4.757148610551752,
+      "grad_norm": 0.2234269293362455,
+      "learning_rate": 2.7335951725364185e-07,
+      "loss": 0.1455,
+      "step": 2953
+    },
+    {
+      "epoch": 4.758759565042288,
+      "grad_norm": 0.21747410916134383,
+      "learning_rate": 2.696613423382788e-07,
+      "loss": 0.1566,
+      "step": 2954
+    },
+    {
+      "epoch": 4.760370519532823,
+      "grad_norm": 0.22281641279316053,
+      "learning_rate": 2.6598818388168246e-07,
+      "loss": 0.1429,
+      "step": 2955
+    },
+    {
+      "epoch": 4.761981474023359,
+      "grad_norm": 0.2238241839767047,
+      "learning_rate": 2.6234004654111854e-07,
+      "loss": 0.1544,
+      "step": 2956
+    },
+    {
+      "epoch": 4.763592428513895,
+      "grad_norm": 0.21188445273823583,
+      "learning_rate": 2.5871693494212036e-07,
+      "loss": 0.1399,
+      "step": 2957
+    },
+    {
+      "epoch": 4.7652033830044305,
+      "grad_norm": 0.210545430187273,
+      "learning_rate": 2.551188536785043e-07,
+      "loss": 0.156,
+      "step": 2958
+    },
+    {
+      "epoch": 4.766814337494965,
+      "grad_norm": 0.22470921530004012,
+      "learning_rate": 2.5154580731234333e-07,
+      "loss": 0.1457,
+      "step": 2959
+    },
+    {
+      "epoch": 4.768425291985501,
+      "grad_norm": 0.21323401360872982,
+      "learning_rate": 2.479978003739669e-07,
+      "loss": 0.1601,
+      "step": 2960
+    },
+    {
+      "epoch": 4.770036246476037,
+      "grad_norm": 0.20568485058813663,
+      "learning_rate": 2.4447483736195877e-07,
+      "loss": 0.1345,
+      "step": 2961
+    },
+    {
+      "epoch": 4.771647200966573,
+      "grad_norm": 0.20709133915444555,
+      "learning_rate": 2.40976922743148e-07,
+      "loss": 0.1491,
+      "step": 2962
+    },
+    {
+      "epoch": 4.773258155457109,
+      "grad_norm": 0.22004597986350832,
+      "learning_rate": 2.3750406095260469e-07,
+      "loss": 0.145,
+      "step": 2963
+    },
+    {
+      "epoch": 4.774869109947644,
+      "grad_norm": 0.21770693112333073,
+      "learning_rate": 2.340562563936355e-07,
+      "loss": 0.1523,
+      "step": 2964
+    },
+    {
+      "epoch": 4.7764800644381795,
+      "grad_norm": 0.20635080684518206,
+      "learning_rate": 2.3063351343777241e-07,
+      "loss": 0.1487,
+      "step": 2965
+    },
+    {
+      "epoch": 4.778091018928715,
+      "grad_norm": 0.2116532190342573,
+      "learning_rate": 2.272358364247773e-07,
+      "loss": 0.1403,
+      "step": 2966
+    },
+    {
+      "epoch": 4.779701973419251,
+      "grad_norm": 0.22159722188901532,
+      "learning_rate": 2.238632296626242e-07,
+      "loss": 0.1546,
+      "step": 2967
+    },
+    {
+      "epoch": 4.781312927909786,
+      "grad_norm": 0.21591067485783777,
+      "learning_rate": 2.205156974275058e-07,
+      "loss": 0.1511,
+      "step": 2968
+    },
+    {
+      "epoch": 4.782923882400322,
+      "grad_norm": 0.2191530577323583,
+      "learning_rate": 2.1719324396381802e-07,
+      "loss": 0.151,
+      "step": 2969
+    },
+    {
+      "epoch": 4.784534836890858,
+      "grad_norm": 0.21158599130781444,
+      "learning_rate": 2.138958734841623e-07,
+      "loss": 0.1434,
+      "step": 2970
+    },
+    {
+      "epoch": 4.786145791381394,
+      "grad_norm": 0.22489697455911142,
+      "learning_rate": 2.106235901693321e-07,
+      "loss": 0.1453,
+      "step": 2971
+    },
+    {
+      "epoch": 4.7877567458719295,
+      "grad_norm": 0.20816987525634345,
+      "learning_rate": 2.0737639816831967e-07,
+      "loss": 0.1417,
+      "step": 2972
+    },
+    {
+      "epoch": 4.789367700362464,
+      "grad_norm": 0.2208234962177496,
+      "learning_rate": 2.0415430159829608e-07,
+      "loss": 0.1449,
+      "step": 2973
+    },
+    {
+      "epoch": 4.790978654853,
+      "grad_norm": 0.21872410598917383,
+      "learning_rate": 2.0095730454461781e-07,
+      "loss": 0.1539,
+      "step": 2974
+    },
+    {
+      "epoch": 4.792589609343536,
+      "grad_norm": 0.2368138202783192,
+      "learning_rate": 1.9778541106081572e-07,
+      "loss": 0.144,
+      "step": 2975
+    },
+    {
+      "epoch": 4.794200563834072,
+      "grad_norm": 0.2517246270809971,
+      "learning_rate": 1.9463862516859277e-07,
+      "loss": 0.1639,
+      "step": 2976
+    },
+    {
+      "epoch": 4.795811518324607,
+      "grad_norm": 0.20392676903290943,
+      "learning_rate": 1.915169508578174e-07,
+      "loss": 0.1362,
+      "step": 2977
+    },
+    {
+      "epoch": 4.797422472815143,
+      "grad_norm": 0.2251301182098052,
+      "learning_rate": 1.8842039208651685e-07,
+      "loss": 0.1366,
+      "step": 2978
+    },
+    {
+      "epoch": 4.7990334273056785,
+      "grad_norm": 0.21438203052030705,
+      "learning_rate": 1.8534895278087272e-07,
+      "loss": 0.1443,
+      "step": 2979
+    },
+    {
+      "epoch": 4.800644381796214,
+      "grad_norm": 0.21536055584983677,
+      "learning_rate": 1.823026368352232e-07,
+      "loss": 0.1456,
+      "step": 2980
+    },
+    {
+      "epoch": 4.80225533628675,
+      "grad_norm": 0.21399263776091337,
+      "learning_rate": 1.792814481120453e-07,
+      "loss": 0.1454,
+      "step": 2981
+    },
+    {
+      "epoch": 4.803866290777286,
+      "grad_norm": 0.2215447311821123,
+      "learning_rate": 1.7628539044195924e-07,
+      "loss": 0.1538,
+      "step": 2982
+    },
+    {
+      "epoch": 4.805477245267821,
+      "grad_norm": 0.22845562785090617,
+      "learning_rate": 1.7331446762372638e-07,
+      "loss": 0.1379,
+      "step": 2983
+    },
+    {
+      "epoch": 4.807088199758357,
+      "grad_norm": 0.21877133621326636,
+      "learning_rate": 1.7036868342422687e-07,
+      "loss": 0.1466,
+      "step": 2984
+    },
+    {
+      "epoch": 4.808699154248893,
+      "grad_norm": 0.22197790124470415,
+      "learning_rate": 1.6744804157848183e-07,
+      "loss": 0.1648,
+      "step": 2985
+    },
+    {
+      "epoch": 4.810310108739428,
+      "grad_norm": 0.21974675453737633,
+      "learning_rate": 1.6455254578962243e-07,
+      "loss": 0.1332,
+      "step": 2986
+    },
+    {
+      "epoch": 4.811921063229963,
+      "grad_norm": 0.21753406916106324,
+      "learning_rate": 1.6168219972890087e-07,
+      "loss": 0.146,
+      "step": 2987
+    },
+    {
+      "epoch": 4.813532017720499,
+      "grad_norm": 0.20824467291862445,
+      "learning_rate": 1.5883700703568373e-07,
+      "loss": 0.1513,
+      "step": 2988
+    },
+    {
+      "epoch": 4.815142972211035,
+      "grad_norm": 0.21940558640107516,
+      "learning_rate": 1.5601697131744308e-07,
+      "loss": 0.1349,
+      "step": 2989
+    },
+    {
+      "epoch": 4.816753926701571,
+      "grad_norm": 0.2169053830515682,
+      "learning_rate": 1.5322209614975214e-07,
+      "loss": 0.1427,
+      "step": 2990
+    },
+    {
+      "epoch": 4.818364881192107,
+      "grad_norm": 0.20841544497570255,
+      "learning_rate": 1.5045238507628513e-07,
+      "loss": 0.1293,
+      "step": 2991
+    },
+    {
+      "epoch": 4.819975835682642,
+      "grad_norm": 0.22675681327150446,
+      "learning_rate": 1.477078416088107e-07,
+      "loss": 0.1549,
+      "step": 2992
+    },
+    {
+      "epoch": 4.8215867901731775,
+      "grad_norm": 0.2147630928023925,
+      "learning_rate": 1.44988469227183e-07,
+      "loss": 0.1446,
+      "step": 2993
+    },
+    {
+      "epoch": 4.823197744663713,
+      "grad_norm": 0.22130327588900567,
+      "learning_rate": 1.422942713793485e-07,
+      "loss": 0.1624,
+      "step": 2994
+    },
+    {
+      "epoch": 4.824808699154249,
+      "grad_norm": 0.21320882618830075,
+      "learning_rate": 1.396252514813279e-07,
+      "loss": 0.1432,
+      "step": 2995
+    },
+    {
+      "epoch": 4.826419653644784,
+      "grad_norm": 0.2086230672911769,
+      "learning_rate": 1.369814129172209e-07,
+      "loss": 0.144,
+      "step": 2996
+    },
+    {
+      "epoch": 4.82803060813532,
+      "grad_norm": 0.2134036368549228,
+      "learning_rate": 1.3436275903919716e-07,
+      "loss": 0.1468,
+      "step": 2997
+    },
+    {
+      "epoch": 4.829641562625856,
+      "grad_norm": 0.21195574971999337,
+      "learning_rate": 1.3176929316749632e-07,
+      "loss": 0.131,
+      "step": 2998
+    },
+    {
+      "epoch": 4.831252517116392,
+      "grad_norm": 0.21601750704641073,
+      "learning_rate": 1.2920101859042578e-07,
+      "loss": 0.1539,
+      "step": 2999
+    },
+    {
+      "epoch": 4.832863471606927,
+      "grad_norm": 0.22404313051428473,
+      "learning_rate": 1.2665793856434516e-07,
+      "loss": 0.1334,
+      "step": 3000
+    },
+    {
+      "epoch": 4.834474426097462,
+      "grad_norm": 0.21836677236601115,
+      "learning_rate": 1.2414005631366855e-07,
+      "loss": 0.1498,
+      "step": 3001
+    },
+    {
+      "epoch": 4.836085380587998,
+      "grad_norm": 0.21793674127632842,
+      "learning_rate": 1.2164737503087108e-07,
+      "loss": 0.1456,
+      "step": 3002
+    },
+    {
+      "epoch": 4.837696335078534,
+      "grad_norm": 0.20502370809178963,
+      "learning_rate": 1.1917989787646689e-07,
+      "loss": 0.145,
+      "step": 3003
+    },
+    {
+      "epoch": 4.83930728956907,
+      "grad_norm": 0.21438747414467008,
+      "learning_rate": 1.1673762797901334e-07,
+      "loss": 0.1404,
+      "step": 3004
+    },
+    {
+      "epoch": 4.840918244059606,
+      "grad_norm": 0.21658308626209727,
+      "learning_rate": 1.1432056843511342e-07,
+      "loss": 0.1366,
+      "step": 3005
+    },
+    {
+      "epoch": 4.842529198550141,
+      "grad_norm": 0.2146719093301986,
+      "learning_rate": 1.1192872230939789e-07,
+      "loss": 0.1513,
+      "step": 3006
+    },
+    {
+      "epoch": 4.8441401530406765,
+      "grad_norm": 0.24537221608744508,
+      "learning_rate": 1.0956209263453421e-07,
+      "loss": 0.1413,
+      "step": 3007
+    },
+    {
+      "epoch": 4.845751107531212,
+      "grad_norm": 0.22340926795790925,
+      "learning_rate": 1.0722068241121319e-07,
+      "loss": 0.1458,
+      "step": 3008
+    },
+    {
+      "epoch": 4.847362062021748,
+      "grad_norm": 0.22379919725594086,
+      "learning_rate": 1.0490449460815788e-07,
+      "loss": 0.1499,
+      "step": 3009
+    },
+    {
+      "epoch": 4.848973016512284,
+      "grad_norm": 0.2221902130763464,
+      "learning_rate": 1.0261353216209691e-07,
+      "loss": 0.1473,
+      "step": 3010
+    },
+    {
+      "epoch": 4.850583971002819,
+      "grad_norm": 0.22992060085331859,
+      "learning_rate": 1.0034779797778893e-07,
+      "loss": 0.1421,
+      "step": 3011
+    },
+    {
+      "epoch": 4.852194925493355,
+      "grad_norm": 0.22018812958227876,
+      "learning_rate": 9.810729492800042e-08,
+      "loss": 0.1456,
+      "step": 3012
+    },
+    {
+      "epoch": 4.853805879983891,
+      "grad_norm": 0.21391699346938456,
+      "learning_rate": 9.589202585350565e-08,
+      "loss": 0.1488,
+      "step": 3013
+    },
+    {
+      "epoch": 4.855416834474426,
+      "grad_norm": 0.21652663225257696,
+      "learning_rate": 9.370199356308229e-08,
+      "loss": 0.1547,
+      "step": 3014
+    },
+    {
+      "epoch": 4.857027788964961,
+      "grad_norm": 0.2075965000591131,
+      "learning_rate": 9.153720083351358e-08,
+      "loss": 0.1432,
+      "step": 3015
+    },
+    {
+      "epoch": 4.858638743455497,
+      "grad_norm": 0.2162147197513201,
+      "learning_rate": 8.939765040958392e-08,
+      "loss": 0.1352,
+      "step": 3016
+    },
+    {
+      "epoch": 4.860249697946033,
+      "grad_norm": 0.21261429233585386,
+      "learning_rate": 8.728334500406332e-08,
+      "loss": 0.1489,
+      "step": 3017
+    },
+    {
+      "epoch": 4.861860652436569,
+      "grad_norm": 0.22990524788122235,
+      "learning_rate": 8.519428729772072e-08,
+      "loss": 0.1419,
+      "step": 3018
+    },
+    {
+      "epoch": 4.863471606927105,
+      "grad_norm": 0.22341316580801132,
+      "learning_rate": 8.313047993931067e-08,
+      "loss": 0.1434,
+      "step": 3019
+    },
+    {
+      "epoch": 4.86508256141764,
+      "grad_norm": 0.22386199673054813,
+      "learning_rate": 8.109192554557333e-08,
+      "loss": 0.1401,
+      "step": 3020
+    },
+    {
+      "epoch": 4.8666935159081754,
+      "grad_norm": 0.22098592505813894,
+      "learning_rate": 7.907862670122557e-08,
+      "loss": 0.1294,
+      "step": 3021
+    },
+    {
+      "epoch": 4.868304470398711,
+      "grad_norm": 0.21149506248274522,
+      "learning_rate": 7.709058595897213e-08,
+      "loss": 0.146,
+      "step": 3022
+    },
+    {
+      "epoch": 4.869915424889247,
+      "grad_norm": 0.2120571099622338,
+      "learning_rate": 7.51278058394811e-08,
+      "loss": 0.1644,
+      "step": 3023
+    },
+    {
+      "epoch": 4.871526379379782,
+      "grad_norm": 0.21984813243178256,
+      "learning_rate": 7.319028883139956e-08,
+      "loss": 0.1435,
+      "step": 3024
+    },
+    {
+      "epoch": 4.873137333870318,
+      "grad_norm": 0.2046339895959685,
+      "learning_rate": 7.12780373913402e-08,
+      "loss": 0.1479,
+      "step": 3025
+    },
+    {
+      "epoch": 4.874748288360854,
+      "grad_norm": 0.21213665620935812,
+      "learning_rate": 6.939105394388356e-08,
+      "loss": 0.1558,
+      "step": 3026
+    },
+    {
+      "epoch": 4.8763592428513896,
+      "grad_norm": 0.20974836104729822,
+      "learning_rate": 6.752934088156693e-08,
+      "loss": 0.1607,
+      "step": 3027
+    },
+    {
+      "epoch": 4.877970197341925,
+      "grad_norm": 0.21367805851277302,
+      "learning_rate": 6.569290056489542e-08,
+      "loss": 0.1412,
+      "step": 3028
+    },
+    {
+      "epoch": 4.879581151832461,
+      "grad_norm": 0.21328781532029567,
+      "learning_rate": 6.3881735322322e-08,
+      "loss": 0.1393,
+      "step": 3029
+    },
+    {
+      "epoch": 4.881192106322996,
+      "grad_norm": 0.2147863974760739,
+      "learning_rate": 6.209584745025643e-08,
+      "loss": 0.1507,
+      "step": 3030
+    },
+    {
+      "epoch": 4.882803060813532,
+      "grad_norm": 0.2116750037903575,
+      "learning_rate": 6.033523921306072e-08,
+      "loss": 0.1423,
+      "step": 3031
+    },
+    {
+      "epoch": 4.884414015304068,
+      "grad_norm": 0.21829717085107075,
+      "learning_rate": 5.859991284303812e-08,
+      "loss": 0.1416,
+      "step": 3032
+    },
+    {
+      "epoch": 4.886024969794604,
+      "grad_norm": 0.2175366973660451,
+      "learning_rate": 5.688987054044637e-08,
+      "loss": 0.1472,
+      "step": 3033
+    },
+    {
+      "epoch": 4.887635924285139,
+      "grad_norm": 0.20843713478870315,
+      "learning_rate": 5.520511447347776e-08,
+      "loss": 0.1381,
+      "step": 3034
+    },
+    {
+      "epoch": 4.889246878775674,
+      "grad_norm": 0.2123862342443856,
+      "learning_rate": 5.3545646778263575e-08,
+      "loss": 0.1511,
+      "step": 3035
+    },
+    {
+      "epoch": 4.89085783326621,
+      "grad_norm": 0.2219417417741511,
+      "learning_rate": 5.191146955887405e-08,
+      "loss": 0.1403,
+      "step": 3036
+    },
+    {
+      "epoch": 4.892468787756746,
+      "grad_norm": 0.21174959404262084,
+      "learning_rate": 5.0302584887313986e-08,
+      "loss": 0.1451,
+      "step": 3037
+    },
+    {
+      "epoch": 4.894079742247282,
+      "grad_norm": 0.21072496789865433,
+      "learning_rate": 4.871899480351605e-08,
+      "loss": 0.1449,
+      "step": 3038
+    },
+    {
+      "epoch": 4.895690696737817,
+      "grad_norm": 0.22806728378504748,
+      "learning_rate": 4.7160701315343e-08,
+      "loss": 0.1472,
+      "step": 3039
+    },
+    {
+      "epoch": 4.897301651228353,
+      "grad_norm": 0.2168054676106727,
+      "learning_rate": 4.562770639858549e-08,
+      "loss": 0.1575,
+      "step": 3040
+    },
+    {
+      "epoch": 4.8989126057188885,
+      "grad_norm": 0.2168197173485377,
+      "learning_rate": 4.412001199695537e-08,
+      "loss": 0.1482,
+      "step": 3041
+    },
+    {
+      "epoch": 4.900523560209424,
+      "grad_norm": 0.2250248057225666,
+      "learning_rate": 4.2637620022085715e-08,
+      "loss": 0.1377,
+      "step": 3042
+    },
+    {
+      "epoch": 4.902134514699959,
+      "grad_norm": 0.21418687912333567,
+      "learning_rate": 4.118053235352859e-08,
+      "loss": 0.135,
+      "step": 3043
+    },
+    {
+      "epoch": 4.903745469190495,
+      "grad_norm": 0.21982028957248173,
+      "learning_rate": 3.974875083875285e-08,
+      "loss": 0.1459,
+      "step": 3044
+    },
+    {
+      "epoch": 4.905356423681031,
+      "grad_norm": 0.21288538105643232,
+      "learning_rate": 3.834227729313966e-08,
+      "loss": 0.1442,
+      "step": 3045
+    },
+    {
+      "epoch": 4.906967378171567,
+      "grad_norm": 0.24690110457527772,
+      "learning_rate": 3.696111349998255e-08,
+      "loss": 0.1438,
+      "step": 3046
+    },
+    {
+      "epoch": 4.908578332662103,
+      "grad_norm": 0.21048986451072557,
+      "learning_rate": 3.5605261210485134e-08,
+      "loss": 0.1472,
+      "step": 3047
+    },
+    {
+      "epoch": 4.910189287152638,
+      "grad_norm": 0.21299506119550873,
+      "learning_rate": 3.427472214375671e-08,
+      "loss": 0.1595,
+      "step": 3048
+    },
+    {
+      "epoch": 4.911800241643173,
+      "grad_norm": 0.2098004986312955,
+      "learning_rate": 3.296949798681226e-08,
+      "loss": 0.1418,
+      "step": 3049
+    },
+    {
+      "epoch": 4.913411196133709,
+      "grad_norm": 0.2111424848272005,
+      "learning_rate": 3.1689590394570204e-08,
+      "loss": 0.1478,
+      "step": 3050
+    },
+    {
+      "epoch": 4.915022150624245,
+      "grad_norm": 0.2056760445046121,
+      "learning_rate": 3.0435000989850194e-08,
+      "loss": 0.1522,
+      "step": 3051
+    },
+    {
+      "epoch": 4.91663310511478,
+      "grad_norm": 0.22222100807383482,
+      "learning_rate": 2.9205731363364244e-08,
+      "loss": 0.1467,
+      "step": 3052
+    },
+    {
+      "epoch": 4.918244059605316,
+      "grad_norm": 0.22694945838983843,
+      "learning_rate": 2.8001783073732248e-08,
+      "loss": 0.1368,
+      "step": 3053
+    },
+    {
+      "epoch": 4.919855014095852,
+      "grad_norm": 0.20514202688175226,
+      "learning_rate": 2.6823157647457577e-08,
+      "loss": 0.1377,
+      "step": 3054
+    },
+    {
+      "epoch": 4.9214659685863875,
+      "grad_norm": 0.21345732647144894,
+      "learning_rate": 2.566985657894483e-08,
+      "loss": 0.1485,
+      "step": 3055
+    },
+    {
+      "epoch": 4.923076923076923,
+      "grad_norm": 0.21597379148933313,
+      "learning_rate": 2.4541881330482075e-08,
+      "loss": 0.1424,
+      "step": 3056
+    },
+    {
+      "epoch": 4.924687877567459,
+      "grad_norm": 0.20890314593204515,
+      "learning_rate": 2.3439233332251953e-08,
+      "loss": 0.1628,
+      "step": 3057
+    },
+    {
+      "epoch": 4.926298832057994,
+      "grad_norm": 0.21635021856559442,
+      "learning_rate": 2.236191398232057e-08,
+      "loss": 0.1445,
+      "step": 3058
+    },
+    {
+      "epoch": 4.92790978654853,
+      "grad_norm": 0.2202683990747119,
+      "learning_rate": 2.1309924646641945e-08,
+      "loss": 0.1482,
+      "step": 3059
+    },
+    {
+      "epoch": 4.929520741039066,
+      "grad_norm": 0.21390221711743562,
+      "learning_rate": 2.0283266659051338e-08,
+      "loss": 0.1388,
+      "step": 3060
+    },
+    {
+      "epoch": 4.931131695529602,
+      "grad_norm": 0.22450839818366922,
+      "learning_rate": 1.9281941321271925e-08,
+      "loss": 0.1565,
+      "step": 3061
+    },
+    {
+      "epoch": 4.932742650020137,
+      "grad_norm": 0.223216962644783,
+      "learning_rate": 1.8305949902897026e-08,
+      "loss": 0.1492,
+      "step": 3062
+    },
+    {
+      "epoch": 4.934353604510672,
+      "grad_norm": 0.2005760944353557,
+      "learning_rate": 1.7355293641405647e-08,
+      "loss": 0.1545,
+      "step": 3063
+    },
+    {
+      "epoch": 4.935964559001208,
+      "grad_norm": 0.22213102975484492,
+      "learning_rate": 1.6429973742153606e-08,
+      "loss": 0.1418,
+      "step": 3064
+    },
+    {
+      "epoch": 4.937575513491744,
+      "grad_norm": 0.22253914887075765,
+      "learning_rate": 1.552999137836908e-08,
+      "loss": 0.136,
+      "step": 3065
+    },
+    {
+      "epoch": 4.93918646798228,
+      "grad_norm": 0.22754585273340733,
+      "learning_rate": 1.4655347691159283e-08,
+      "loss": 0.1514,
+      "step": 3066
+    },
+    {
+      "epoch": 4.940797422472815,
+      "grad_norm": 0.20712918988601425,
+      "learning_rate": 1.3806043789497126e-08,
+      "loss": 0.143,
+      "step": 3067
+    },
+    {
+      "epoch": 4.942408376963351,
+      "grad_norm": 0.22069265881378214,
+      "learning_rate": 1.2982080750234549e-08,
+      "loss": 0.1491,
+      "step": 3068
+    },
+    {
+      "epoch": 4.9440193314538865,
+      "grad_norm": 0.20959463552056892,
+      "learning_rate": 1.2183459618084759e-08,
+      "loss": 0.1523,
+      "step": 3069
+    },
+    {
+      "epoch": 4.945630285944422,
+      "grad_norm": 0.21612258928364242,
+      "learning_rate": 1.1410181405639986e-08,
+      "loss": 0.1442,
+      "step": 3070
+    },
+    {
+      "epoch": 4.947241240434957,
+      "grad_norm": 0.2162006333759751,
+      "learning_rate": 1.0662247093349287e-08,
+      "loss": 0.1244,
+      "step": 3071
+    },
+    {
+      "epoch": 4.948852194925493,
+      "grad_norm": 0.22621189321655785,
+      "learning_rate": 9.939657629534083e-09,
+      "loss": 0.1624,
+      "step": 3072
+    },
+    {
+      "epoch": 4.950463149416029,
+      "grad_norm": 0.22325164035423106,
+      "learning_rate": 9.242413930377059e-09,
+      "loss": 0.1501,
+      "step": 3073
+    },
+    {
+      "epoch": 4.952074103906565,
+      "grad_norm": 0.20908742596922356,
+      "learning_rate": 8.570516879928826e-09,
+      "loss": 0.1313,
+      "step": 3074
+    },
+    {
+      "epoch": 4.953685058397101,
+      "grad_norm": 0.22585183848954468,
+      "learning_rate": 7.923967330099036e-09,
+      "loss": 0.1474,
+      "step": 3075
+    },
+    {
+      "epoch": 4.9552960128876355,
+      "grad_norm": 0.21204595764120757,
+      "learning_rate": 7.3027661006586095e-09,
+      "loss": 0.1531,
+      "step": 3076
+    },
+    {
+      "epoch": 4.956906967378171,
+      "grad_norm": 0.20138290136485443,
+      "learning_rate": 6.706913979241947e-09,
+      "loss": 0.1603,
+      "step": 3077
+    },
+    {
+      "epoch": 4.958517921868707,
+      "grad_norm": 0.21351586165159536,
+      "learning_rate": 6.136411721340274e-09,
+      "loss": 0.1383,
+      "step": 3078
+    },
+    {
+      "epoch": 4.960128876359243,
+      "grad_norm": 0.21376007659935434,
+      "learning_rate": 5.5912600503038594e-09,
+      "loss": 0.1445,
+      "step": 3079
+    },
+    {
+      "epoch": 4.961739830849779,
+      "grad_norm": 0.21489003486773706,
+      "learning_rate": 5.071459657339794e-09,
+      "loss": 0.1388,
+      "step": 3080
+    },
+    {
+      "epoch": 4.963350785340314,
+      "grad_norm": 0.21755924068751195,
+      "learning_rate": 4.577011201511994e-09,
+      "loss": 0.1456,
+      "step": 3081
+    },
+    {
+      "epoch": 4.96496173983085,
+      "grad_norm": 0.20180190976203755,
+      "learning_rate": 4.107915309743416e-09,
+      "loss": 0.146,
+      "step": 3082
+    },
+    {
+      "epoch": 4.9665726943213855,
+      "grad_norm": 0.21855543222434343,
+      "learning_rate": 3.664172576807179e-09,
+      "loss": 0.1405,
+      "step": 3083
+    },
+    {
+      "epoch": 4.968183648811921,
+      "grad_norm": 0.215264107680761,
+      "learning_rate": 3.2457835653332272e-09,
+      "loss": 0.1456,
+      "step": 3084
+    },
+    {
+      "epoch": 4.969794603302457,
+      "grad_norm": 0.203470431035806,
+      "learning_rate": 2.8527488058038844e-09,
+      "loss": 0.1432,
+      "step": 3085
+    },
+    {
+      "epoch": 4.971405557792992,
+      "grad_norm": 0.21276675372989584,
+      "learning_rate": 2.485068796556078e-09,
+      "loss": 0.1577,
+      "step": 3086
+    },
+    {
+      "epoch": 4.973016512283528,
+      "grad_norm": 0.21088797347840574,
+      "learning_rate": 2.142744003779118e-09,
+      "loss": 0.1421,
+      "step": 3087
+    },
+    {
+      "epoch": 4.974627466774064,
+      "grad_norm": 0.2246105636966214,
+      "learning_rate": 1.8257748615102545e-09,
+      "loss": 0.1351,
+      "step": 3088
+    },
+    {
+      "epoch": 4.9762384212646,
+      "grad_norm": 0.21752395162393956,
+      "learning_rate": 1.5341617716435608e-09,
+      "loss": 0.1498,
+      "step": 3089
+    },
+    {
+      "epoch": 4.9778493757551345,
+      "grad_norm": 0.21715767148751064,
+      "learning_rate": 1.2679051039188317e-09,
+      "loss": 0.1431,
+      "step": 3090
+    },
+    {
+      "epoch": 4.97946033024567,
+      "grad_norm": 0.200604318655284,
+      "learning_rate": 1.0270051959282434e-09,
+      "loss": 0.1356,
+      "step": 3091
+    },
+    {
+      "epoch": 4.981071284736206,
+      "grad_norm": 0.22114376968996727,
+      "learning_rate": 8.114623531119137e-10,
+      "loss": 0.1494,
+      "step": 3092
+    },
+    {
+      "epoch": 4.982682239226742,
+      "grad_norm": 0.21912817007954846,
+      "learning_rate": 6.212768487623422e-10,
+      "loss": 0.147,
+      "step": 3093
+    },
+    {
+      "epoch": 4.984293193717278,
+      "grad_norm": 0.21734907735157796,
+      "learning_rate": 4.564489240177494e-10,
+      "loss": 0.1427,
+      "step": 3094
+    },
+    {
+      "epoch": 4.985904148207813,
+      "grad_norm": 0.2122215913186033,
+      "learning_rate": 3.1697878786873804e-10,
+      "loss": 0.1379,
+      "step": 3095
+    },
+    {
+      "epoch": 4.987515102698349,
+      "grad_norm": 0.22045381925803034,
+      "learning_rate": 2.0286661714941092e-10,
+      "loss": 0.154,
+      "step": 3096
+    },
+    {
+      "epoch": 4.9891260571888845,
+      "grad_norm": 0.21230236722633794,
+      "learning_rate": 1.1411255654625309e-10,
+      "loss": 0.1398,
+      "step": 3097
+    },
+    {
+      "epoch": 4.99073701167942,
+      "grad_norm": 0.2155863280622798,
+      "learning_rate": 5.0716718591470313e-11,
+      "loss": 0.1415,
+      "step": 3098
+    },
+    {
+      "epoch": 4.992347966169955,
+      "grad_norm": 0.21790703900678665,
+      "learning_rate": 1.267918366743004e-11,
+      "loss": 0.1517,
+      "step": 3099
+    },
+    {
+      "epoch": 4.993958920660491,
+      "grad_norm": 0.21710189361301535,
+      "learning_rate": 0.0,
+      "loss": 0.1494,
+      "step": 3100
+    },
+    {
+      "epoch": 4.993958920660491,
+      "step": 3100,
+      "total_flos": 6.310856434702221e+18,
+      "train_loss": 0.2839509548799646,
+      "train_runtime": 73731.2238,
+      "train_samples_per_second": 5.387,
+      "train_steps_per_second": 0.042
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 3100,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 6.310856434702221e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}