d1_math_gpt / trainer_state.json
neginr's picture
End of training
696fb4c verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 5.0,
"eval_steps": 500,
"global_step": 1235,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.004048582995951417,
"grad_norm": 6.401622941849267,
"learning_rate": 3.2258064516129035e-07,
"loss": 0.8141,
"step": 1
},
{
"epoch": 0.008097165991902834,
"grad_norm": 6.403231563321921,
"learning_rate": 6.451612903225807e-07,
"loss": 0.8239,
"step": 2
},
{
"epoch": 0.012145748987854251,
"grad_norm": 6.192814078910405,
"learning_rate": 9.67741935483871e-07,
"loss": 0.7913,
"step": 3
},
{
"epoch": 0.016194331983805668,
"grad_norm": 6.145781709503103,
"learning_rate": 1.2903225806451614e-06,
"loss": 0.8017,
"step": 4
},
{
"epoch": 0.020242914979757085,
"grad_norm": 5.8171690501088555,
"learning_rate": 1.6129032258064516e-06,
"loss": 0.788,
"step": 5
},
{
"epoch": 0.024291497975708502,
"grad_norm": 5.714796206407175,
"learning_rate": 1.935483870967742e-06,
"loss": 0.8017,
"step": 6
},
{
"epoch": 0.02834008097165992,
"grad_norm": 4.400311616098369,
"learning_rate": 2.2580645161290324e-06,
"loss": 0.7678,
"step": 7
},
{
"epoch": 0.032388663967611336,
"grad_norm": 4.105011258679847,
"learning_rate": 2.580645161290323e-06,
"loss": 0.7537,
"step": 8
},
{
"epoch": 0.03643724696356275,
"grad_norm": 2.3120376675260066,
"learning_rate": 2.903225806451613e-06,
"loss": 0.7281,
"step": 9
},
{
"epoch": 0.04048582995951417,
"grad_norm": 2.018479638584789,
"learning_rate": 3.225806451612903e-06,
"loss": 0.7121,
"step": 10
},
{
"epoch": 0.044534412955465584,
"grad_norm": 1.934951698649959,
"learning_rate": 3.548387096774194e-06,
"loss": 0.7197,
"step": 11
},
{
"epoch": 0.048582995951417005,
"grad_norm": 3.7243348939868217,
"learning_rate": 3.870967741935484e-06,
"loss": 0.7085,
"step": 12
},
{
"epoch": 0.05263157894736842,
"grad_norm": 4.12076556192734,
"learning_rate": 4.193548387096774e-06,
"loss": 0.7137,
"step": 13
},
{
"epoch": 0.05668016194331984,
"grad_norm": 4.126656073192889,
"learning_rate": 4.516129032258065e-06,
"loss": 0.7137,
"step": 14
},
{
"epoch": 0.06072874493927125,
"grad_norm": 3.919134146616194,
"learning_rate": 4.838709677419355e-06,
"loss": 0.7003,
"step": 15
},
{
"epoch": 0.06477732793522267,
"grad_norm": 3.1536981603255176,
"learning_rate": 5.161290322580646e-06,
"loss": 0.6656,
"step": 16
},
{
"epoch": 0.06882591093117409,
"grad_norm": 2.9425454939481916,
"learning_rate": 5.483870967741935e-06,
"loss": 0.6728,
"step": 17
},
{
"epoch": 0.0728744939271255,
"grad_norm": 2.166223924420372,
"learning_rate": 5.806451612903226e-06,
"loss": 0.6323,
"step": 18
},
{
"epoch": 0.07692307692307693,
"grad_norm": 1.4083934452332474,
"learning_rate": 6.129032258064517e-06,
"loss": 0.616,
"step": 19
},
{
"epoch": 0.08097165991902834,
"grad_norm": 1.2378533006570058,
"learning_rate": 6.451612903225806e-06,
"loss": 0.6003,
"step": 20
},
{
"epoch": 0.08502024291497975,
"grad_norm": 1.4582725060572614,
"learning_rate": 6.774193548387097e-06,
"loss": 0.6111,
"step": 21
},
{
"epoch": 0.08906882591093117,
"grad_norm": 1.51597286484319,
"learning_rate": 7.096774193548388e-06,
"loss": 0.5964,
"step": 22
},
{
"epoch": 0.0931174089068826,
"grad_norm": 1.2506613045308521,
"learning_rate": 7.4193548387096784e-06,
"loss": 0.601,
"step": 23
},
{
"epoch": 0.09716599190283401,
"grad_norm": 0.9151924672189861,
"learning_rate": 7.741935483870968e-06,
"loss": 0.5888,
"step": 24
},
{
"epoch": 0.10121457489878542,
"grad_norm": 0.9246690751784659,
"learning_rate": 8.064516129032258e-06,
"loss": 0.5722,
"step": 25
},
{
"epoch": 0.10526315789473684,
"grad_norm": 0.9223572490572591,
"learning_rate": 8.387096774193549e-06,
"loss": 0.5789,
"step": 26
},
{
"epoch": 0.10931174089068826,
"grad_norm": 0.7011144938226701,
"learning_rate": 8.70967741935484e-06,
"loss": 0.5801,
"step": 27
},
{
"epoch": 0.11336032388663968,
"grad_norm": 0.8266192273876617,
"learning_rate": 9.03225806451613e-06,
"loss": 0.5611,
"step": 28
},
{
"epoch": 0.11740890688259109,
"grad_norm": 0.7953375210300712,
"learning_rate": 9.35483870967742e-06,
"loss": 0.541,
"step": 29
},
{
"epoch": 0.1214574898785425,
"grad_norm": 0.6187887919327005,
"learning_rate": 9.67741935483871e-06,
"loss": 0.5484,
"step": 30
},
{
"epoch": 0.12550607287449392,
"grad_norm": 0.5621456488377176,
"learning_rate": 1e-05,
"loss": 0.5384,
"step": 31
},
{
"epoch": 0.12955465587044535,
"grad_norm": 0.6728929815382415,
"learning_rate": 1.0322580645161291e-05,
"loss": 0.5308,
"step": 32
},
{
"epoch": 0.13360323886639677,
"grad_norm": 0.7828560706667843,
"learning_rate": 1.0645161290322582e-05,
"loss": 0.5475,
"step": 33
},
{
"epoch": 0.13765182186234817,
"grad_norm": 0.6078823834218207,
"learning_rate": 1.096774193548387e-05,
"loss": 0.5203,
"step": 34
},
{
"epoch": 0.1417004048582996,
"grad_norm": 0.5120690431757132,
"learning_rate": 1.1290322580645164e-05,
"loss": 0.532,
"step": 35
},
{
"epoch": 0.145748987854251,
"grad_norm": 0.6551891278361649,
"learning_rate": 1.1612903225806453e-05,
"loss": 0.5278,
"step": 36
},
{
"epoch": 0.14979757085020243,
"grad_norm": 0.6083388707668052,
"learning_rate": 1.1935483870967743e-05,
"loss": 0.5307,
"step": 37
},
{
"epoch": 0.15384615384615385,
"grad_norm": 0.591964588608808,
"learning_rate": 1.2258064516129034e-05,
"loss": 0.5309,
"step": 38
},
{
"epoch": 0.15789473684210525,
"grad_norm": 0.46434217630518265,
"learning_rate": 1.2580645161290324e-05,
"loss": 0.5259,
"step": 39
},
{
"epoch": 0.16194331983805668,
"grad_norm": 0.5304409473391041,
"learning_rate": 1.2903225806451613e-05,
"loss": 0.5101,
"step": 40
},
{
"epoch": 0.1659919028340081,
"grad_norm": 0.6024453017545047,
"learning_rate": 1.3225806451612903e-05,
"loss": 0.5166,
"step": 41
},
{
"epoch": 0.1700404858299595,
"grad_norm": 0.4961694592679978,
"learning_rate": 1.3548387096774194e-05,
"loss": 0.5179,
"step": 42
},
{
"epoch": 0.17408906882591094,
"grad_norm": 0.4408944185591605,
"learning_rate": 1.3870967741935486e-05,
"loss": 0.5155,
"step": 43
},
{
"epoch": 0.17813765182186234,
"grad_norm": 0.4258967398477333,
"learning_rate": 1.4193548387096776e-05,
"loss": 0.5084,
"step": 44
},
{
"epoch": 0.18218623481781376,
"grad_norm": 0.4771482815147509,
"learning_rate": 1.4516129032258066e-05,
"loss": 0.5099,
"step": 45
},
{
"epoch": 0.1862348178137652,
"grad_norm": 0.4313023062976589,
"learning_rate": 1.4838709677419357e-05,
"loss": 0.4993,
"step": 46
},
{
"epoch": 0.1902834008097166,
"grad_norm": 0.37208474715721407,
"learning_rate": 1.5161290322580646e-05,
"loss": 0.4981,
"step": 47
},
{
"epoch": 0.19433198380566802,
"grad_norm": 0.4288122625364947,
"learning_rate": 1.5483870967741936e-05,
"loss": 0.5145,
"step": 48
},
{
"epoch": 0.19838056680161945,
"grad_norm": 0.4463227524891996,
"learning_rate": 1.5806451612903226e-05,
"loss": 0.5021,
"step": 49
},
{
"epoch": 0.20242914979757085,
"grad_norm": 0.4068619208847059,
"learning_rate": 1.6129032258064517e-05,
"loss": 0.4874,
"step": 50
},
{
"epoch": 0.20647773279352227,
"grad_norm": 0.4992651692527317,
"learning_rate": 1.6451612903225807e-05,
"loss": 0.5116,
"step": 51
},
{
"epoch": 0.21052631578947367,
"grad_norm": 0.4857198971045037,
"learning_rate": 1.6774193548387098e-05,
"loss": 0.5022,
"step": 52
},
{
"epoch": 0.2145748987854251,
"grad_norm": 0.4486043409392543,
"learning_rate": 1.7096774193548388e-05,
"loss": 0.5065,
"step": 53
},
{
"epoch": 0.21862348178137653,
"grad_norm": 0.3870754338993052,
"learning_rate": 1.741935483870968e-05,
"loss": 0.4935,
"step": 54
},
{
"epoch": 0.22267206477732793,
"grad_norm": 0.4281905902138728,
"learning_rate": 1.774193548387097e-05,
"loss": 0.4942,
"step": 55
},
{
"epoch": 0.22672064777327935,
"grad_norm": 0.41594038310000697,
"learning_rate": 1.806451612903226e-05,
"loss": 0.4787,
"step": 56
},
{
"epoch": 0.23076923076923078,
"grad_norm": 0.4259934554377301,
"learning_rate": 1.838709677419355e-05,
"loss": 0.4718,
"step": 57
},
{
"epoch": 0.23481781376518218,
"grad_norm": 0.4148276163317532,
"learning_rate": 1.870967741935484e-05,
"loss": 0.5015,
"step": 58
},
{
"epoch": 0.2388663967611336,
"grad_norm": 0.46104367733327245,
"learning_rate": 1.903225806451613e-05,
"loss": 0.4999,
"step": 59
},
{
"epoch": 0.242914979757085,
"grad_norm": 0.5429155991750878,
"learning_rate": 1.935483870967742e-05,
"loss": 0.4941,
"step": 60
},
{
"epoch": 0.24696356275303644,
"grad_norm": 0.6816415321188358,
"learning_rate": 1.967741935483871e-05,
"loss": 0.483,
"step": 61
},
{
"epoch": 0.25101214574898784,
"grad_norm": 0.49589263276156686,
"learning_rate": 2e-05,
"loss": 0.4849,
"step": 62
},
{
"epoch": 0.2550607287449393,
"grad_norm": 0.5100686381735254,
"learning_rate": 2.0322580645161292e-05,
"loss": 0.4792,
"step": 63
},
{
"epoch": 0.2591093117408907,
"grad_norm": 0.4985593229382761,
"learning_rate": 2.0645161290322582e-05,
"loss": 0.4726,
"step": 64
},
{
"epoch": 0.2631578947368421,
"grad_norm": 0.4657831564670269,
"learning_rate": 2.0967741935483873e-05,
"loss": 0.4786,
"step": 65
},
{
"epoch": 0.26720647773279355,
"grad_norm": 0.5184026401751588,
"learning_rate": 2.1290322580645163e-05,
"loss": 0.4781,
"step": 66
},
{
"epoch": 0.27125506072874495,
"grad_norm": 0.5989603675468221,
"learning_rate": 2.161290322580645e-05,
"loss": 0.4847,
"step": 67
},
{
"epoch": 0.27530364372469635,
"grad_norm": 0.629518219849865,
"learning_rate": 2.193548387096774e-05,
"loss": 0.4897,
"step": 68
},
{
"epoch": 0.2793522267206478,
"grad_norm": 0.4517350383980552,
"learning_rate": 2.225806451612903e-05,
"loss": 0.4773,
"step": 69
},
{
"epoch": 0.2834008097165992,
"grad_norm": 0.5527531782645612,
"learning_rate": 2.2580645161290328e-05,
"loss": 0.4731,
"step": 70
},
{
"epoch": 0.2874493927125506,
"grad_norm": 0.41132319855131116,
"learning_rate": 2.2903225806451615e-05,
"loss": 0.4713,
"step": 71
},
{
"epoch": 0.291497975708502,
"grad_norm": 0.4997227886575602,
"learning_rate": 2.3225806451612906e-05,
"loss": 0.4603,
"step": 72
},
{
"epoch": 0.29554655870445345,
"grad_norm": 0.6325762695382688,
"learning_rate": 2.3548387096774196e-05,
"loss": 0.464,
"step": 73
},
{
"epoch": 0.29959514170040485,
"grad_norm": 0.715009673006852,
"learning_rate": 2.3870967741935486e-05,
"loss": 0.4731,
"step": 74
},
{
"epoch": 0.30364372469635625,
"grad_norm": 0.8293452513354779,
"learning_rate": 2.4193548387096777e-05,
"loss": 0.4753,
"step": 75
},
{
"epoch": 0.3076923076923077,
"grad_norm": 1.4063529531017553,
"learning_rate": 2.4516129032258067e-05,
"loss": 0.4856,
"step": 76
},
{
"epoch": 0.3117408906882591,
"grad_norm": 0.6461484755172606,
"learning_rate": 2.4838709677419358e-05,
"loss": 0.4619,
"step": 77
},
{
"epoch": 0.3157894736842105,
"grad_norm": 1.2608251817226845,
"learning_rate": 2.5161290322580648e-05,
"loss": 0.4783,
"step": 78
},
{
"epoch": 0.31983805668016196,
"grad_norm": 0.7758145755408175,
"learning_rate": 2.548387096774194e-05,
"loss": 0.4621,
"step": 79
},
{
"epoch": 0.32388663967611336,
"grad_norm": 0.9095792928549379,
"learning_rate": 2.5806451612903226e-05,
"loss": 0.4864,
"step": 80
},
{
"epoch": 0.32793522267206476,
"grad_norm": 1.1190959619416552,
"learning_rate": 2.6129032258064516e-05,
"loss": 0.4541,
"step": 81
},
{
"epoch": 0.3319838056680162,
"grad_norm": 0.7573393398766106,
"learning_rate": 2.6451612903225806e-05,
"loss": 0.4836,
"step": 82
},
{
"epoch": 0.3360323886639676,
"grad_norm": 0.9019958801403516,
"learning_rate": 2.6774193548387097e-05,
"loss": 0.4658,
"step": 83
},
{
"epoch": 0.340080971659919,
"grad_norm": 1.02377227428937,
"learning_rate": 2.7096774193548387e-05,
"loss": 0.4717,
"step": 84
},
{
"epoch": 0.3441295546558704,
"grad_norm": 0.7213905685988595,
"learning_rate": 2.7419354838709678e-05,
"loss": 0.4545,
"step": 85
},
{
"epoch": 0.3481781376518219,
"grad_norm": 0.57348989998002,
"learning_rate": 2.774193548387097e-05,
"loss": 0.4683,
"step": 86
},
{
"epoch": 0.3522267206477733,
"grad_norm": 0.8532958544823128,
"learning_rate": 2.8064516129032262e-05,
"loss": 0.4639,
"step": 87
},
{
"epoch": 0.3562753036437247,
"grad_norm": 0.7378198106251977,
"learning_rate": 2.8387096774193552e-05,
"loss": 0.4649,
"step": 88
},
{
"epoch": 0.3603238866396761,
"grad_norm": 0.5108722033732581,
"learning_rate": 2.8709677419354843e-05,
"loss": 0.464,
"step": 89
},
{
"epoch": 0.3643724696356275,
"grad_norm": 0.7724741973130086,
"learning_rate": 2.9032258064516133e-05,
"loss": 0.4559,
"step": 90
},
{
"epoch": 0.3684210526315789,
"grad_norm": 0.880032069897178,
"learning_rate": 2.9354838709677423e-05,
"loss": 0.4605,
"step": 91
},
{
"epoch": 0.3724696356275304,
"grad_norm": 0.9434035863193581,
"learning_rate": 2.9677419354838714e-05,
"loss": 0.466,
"step": 92
},
{
"epoch": 0.3765182186234818,
"grad_norm": 1.1450912069083654,
"learning_rate": 3.0000000000000004e-05,
"loss": 0.4656,
"step": 93
},
{
"epoch": 0.3805668016194332,
"grad_norm": 0.6060693406077229,
"learning_rate": 3.032258064516129e-05,
"loss": 0.4449,
"step": 94
},
{
"epoch": 0.38461538461538464,
"grad_norm": 0.6677671092410973,
"learning_rate": 3.0645161290322585e-05,
"loss": 0.4618,
"step": 95
},
{
"epoch": 0.38866396761133604,
"grad_norm": 0.9496221699990917,
"learning_rate": 3.096774193548387e-05,
"loss": 0.4544,
"step": 96
},
{
"epoch": 0.39271255060728744,
"grad_norm": 0.8944363782097431,
"learning_rate": 3.1290322580645166e-05,
"loss": 0.4553,
"step": 97
},
{
"epoch": 0.3967611336032389,
"grad_norm": 0.6465963382477403,
"learning_rate": 3.161290322580645e-05,
"loss": 0.4417,
"step": 98
},
{
"epoch": 0.4008097165991903,
"grad_norm": 0.9068512835231436,
"learning_rate": 3.193548387096774e-05,
"loss": 0.4713,
"step": 99
},
{
"epoch": 0.4048582995951417,
"grad_norm": 0.969038830437448,
"learning_rate": 3.2258064516129034e-05,
"loss": 0.4651,
"step": 100
},
{
"epoch": 0.4089068825910931,
"grad_norm": 0.6600245172332877,
"learning_rate": 3.258064516129033e-05,
"loss": 0.47,
"step": 101
},
{
"epoch": 0.41295546558704455,
"grad_norm": 1.0004004254863514,
"learning_rate": 3.2903225806451614e-05,
"loss": 0.4475,
"step": 102
},
{
"epoch": 0.41700404858299595,
"grad_norm": 0.9713127546001857,
"learning_rate": 3.322580645161291e-05,
"loss": 0.462,
"step": 103
},
{
"epoch": 0.42105263157894735,
"grad_norm": 0.6025071411875783,
"learning_rate": 3.3548387096774195e-05,
"loss": 0.4546,
"step": 104
},
{
"epoch": 0.4251012145748988,
"grad_norm": 0.6935251195289093,
"learning_rate": 3.387096774193549e-05,
"loss": 0.4526,
"step": 105
},
{
"epoch": 0.4291497975708502,
"grad_norm": 0.8002953320470126,
"learning_rate": 3.4193548387096776e-05,
"loss": 0.4528,
"step": 106
},
{
"epoch": 0.4331983805668016,
"grad_norm": 0.8607679421710911,
"learning_rate": 3.451612903225807e-05,
"loss": 0.4576,
"step": 107
},
{
"epoch": 0.43724696356275305,
"grad_norm": 0.9162023224135496,
"learning_rate": 3.483870967741936e-05,
"loss": 0.463,
"step": 108
},
{
"epoch": 0.44129554655870445,
"grad_norm": 0.8209911697467708,
"learning_rate": 3.516129032258065e-05,
"loss": 0.4629,
"step": 109
},
{
"epoch": 0.44534412955465585,
"grad_norm": 0.8394123026208852,
"learning_rate": 3.548387096774194e-05,
"loss": 0.4612,
"step": 110
},
{
"epoch": 0.4493927125506073,
"grad_norm": 0.9182588755216764,
"learning_rate": 3.5806451612903225e-05,
"loss": 0.4659,
"step": 111
},
{
"epoch": 0.4534412955465587,
"grad_norm": 0.7919025764811939,
"learning_rate": 3.612903225806452e-05,
"loss": 0.4719,
"step": 112
},
{
"epoch": 0.4574898785425101,
"grad_norm": 1.2847398799481005,
"learning_rate": 3.6451612903225805e-05,
"loss": 0.4461,
"step": 113
},
{
"epoch": 0.46153846153846156,
"grad_norm": 1.8624625222006586,
"learning_rate": 3.67741935483871e-05,
"loss": 0.4638,
"step": 114
},
{
"epoch": 0.46558704453441296,
"grad_norm": 1.1868497377722065,
"learning_rate": 3.7096774193548386e-05,
"loss": 0.4508,
"step": 115
},
{
"epoch": 0.46963562753036436,
"grad_norm": 0.9264049781673939,
"learning_rate": 3.741935483870968e-05,
"loss": 0.4498,
"step": 116
},
{
"epoch": 0.47368421052631576,
"grad_norm": 0.5850593513439407,
"learning_rate": 3.7741935483870974e-05,
"loss": 0.4476,
"step": 117
},
{
"epoch": 0.4777327935222672,
"grad_norm": 0.8246251878771929,
"learning_rate": 3.806451612903226e-05,
"loss": 0.4533,
"step": 118
},
{
"epoch": 0.4817813765182186,
"grad_norm": 0.8165717357208206,
"learning_rate": 3.8387096774193555e-05,
"loss": 0.4513,
"step": 119
},
{
"epoch": 0.48582995951417,
"grad_norm": 0.792629901465678,
"learning_rate": 3.870967741935484e-05,
"loss": 0.4589,
"step": 120
},
{
"epoch": 0.4898785425101215,
"grad_norm": 0.7531951211530213,
"learning_rate": 3.9032258064516135e-05,
"loss": 0.4453,
"step": 121
},
{
"epoch": 0.4939271255060729,
"grad_norm": 0.8304925287358423,
"learning_rate": 3.935483870967742e-05,
"loss": 0.4497,
"step": 122
},
{
"epoch": 0.4979757085020243,
"grad_norm": 0.6715210087117058,
"learning_rate": 3.9677419354838716e-05,
"loss": 0.4408,
"step": 123
},
{
"epoch": 0.5020242914979757,
"grad_norm": 0.594337128179534,
"learning_rate": 4e-05,
"loss": 0.4523,
"step": 124
},
{
"epoch": 0.5060728744939271,
"grad_norm": 0.7109560288569454,
"learning_rate": 3.9999920040266474e-05,
"loss": 0.4427,
"step": 125
},
{
"epoch": 0.5101214574898786,
"grad_norm": 0.7688542711676963,
"learning_rate": 3.999968016170525e-05,
"loss": 0.4525,
"step": 126
},
{
"epoch": 0.5141700404858299,
"grad_norm": 0.6485897499216077,
"learning_rate": 3.9999280366234394e-05,
"loss": 0.4521,
"step": 127
},
{
"epoch": 0.5182186234817814,
"grad_norm": 0.6313209795593597,
"learning_rate": 3.999872065705065e-05,
"loss": 0.4511,
"step": 128
},
{
"epoch": 0.5222672064777328,
"grad_norm": 0.5978487899675834,
"learning_rate": 3.999800103862945e-05,
"loss": 0.4315,
"step": 129
},
{
"epoch": 0.5263157894736842,
"grad_norm": 0.6570959775262086,
"learning_rate": 3.9997121516724834e-05,
"loss": 0.4545,
"step": 130
},
{
"epoch": 0.5303643724696356,
"grad_norm": 0.678087647802001,
"learning_rate": 3.999608209836944e-05,
"loss": 0.4374,
"step": 131
},
{
"epoch": 0.5344129554655871,
"grad_norm": 0.7751612407114034,
"learning_rate": 3.999488279187443e-05,
"loss": 0.4331,
"step": 132
},
{
"epoch": 0.5384615384615384,
"grad_norm": 0.5679171871029915,
"learning_rate": 3.9993523606829425e-05,
"loss": 0.4428,
"step": 133
},
{
"epoch": 0.5425101214574899,
"grad_norm": 0.5527454673414907,
"learning_rate": 3.9992004554102435e-05,
"loss": 0.4314,
"step": 134
},
{
"epoch": 0.5465587044534413,
"grad_norm": 0.8106603827549257,
"learning_rate": 3.999032564583976e-05,
"loss": 0.4427,
"step": 135
},
{
"epoch": 0.5506072874493927,
"grad_norm": 0.9356985281685941,
"learning_rate": 3.998848689546591e-05,
"loss": 0.4569,
"step": 136
},
{
"epoch": 0.5546558704453441,
"grad_norm": 0.5130636358911784,
"learning_rate": 3.9986488317683496e-05,
"loss": 0.4511,
"step": 137
},
{
"epoch": 0.5587044534412956,
"grad_norm": 0.5929424998810572,
"learning_rate": 3.9984329928473067e-05,
"loss": 0.4453,
"step": 138
},
{
"epoch": 0.562753036437247,
"grad_norm": 0.6338076960311926,
"learning_rate": 3.998201174509306e-05,
"loss": 0.4332,
"step": 139
},
{
"epoch": 0.5668016194331984,
"grad_norm": 0.5420277804496704,
"learning_rate": 3.997953378607961e-05,
"loss": 0.4464,
"step": 140
},
{
"epoch": 0.5708502024291497,
"grad_norm": 0.5696312381626168,
"learning_rate": 3.9976896071246403e-05,
"loss": 0.4305,
"step": 141
},
{
"epoch": 0.5748987854251012,
"grad_norm": 0.9270172890848228,
"learning_rate": 3.997409862168454e-05,
"loss": 0.45,
"step": 142
},
{
"epoch": 0.5789473684210527,
"grad_norm": 0.9425431671955719,
"learning_rate": 3.997114145976237e-05,
"loss": 0.4426,
"step": 143
},
{
"epoch": 0.582995951417004,
"grad_norm": 0.5915668610689433,
"learning_rate": 3.9968024609125245e-05,
"loss": 0.4281,
"step": 144
},
{
"epoch": 0.5870445344129555,
"grad_norm": 0.5122082509402403,
"learning_rate": 3.996474809469545e-05,
"loss": 0.4364,
"step": 145
},
{
"epoch": 0.5910931174089069,
"grad_norm": 0.5556236997908491,
"learning_rate": 3.996131194267189e-05,
"loss": 0.4377,
"step": 146
},
{
"epoch": 0.5951417004048583,
"grad_norm": 0.5187582195062321,
"learning_rate": 3.995771618052995e-05,
"loss": 0.4301,
"step": 147
},
{
"epoch": 0.5991902834008097,
"grad_norm": 0.6143552968496623,
"learning_rate": 3.9953960837021256e-05,
"loss": 0.4453,
"step": 148
},
{
"epoch": 0.6032388663967612,
"grad_norm": 0.5714116074692257,
"learning_rate": 3.995004594217343e-05,
"loss": 0.4463,
"step": 149
},
{
"epoch": 0.6072874493927125,
"grad_norm": 0.5112478862959786,
"learning_rate": 3.994597152728987e-05,
"loss": 0.4338,
"step": 150
},
{
"epoch": 0.611336032388664,
"grad_norm": 0.5783896148135652,
"learning_rate": 3.994173762494947e-05,
"loss": 0.4498,
"step": 151
},
{
"epoch": 0.6153846153846154,
"grad_norm": 0.4577117846455844,
"learning_rate": 3.9937344269006424e-05,
"loss": 0.4363,
"step": 152
},
{
"epoch": 0.6194331983805668,
"grad_norm": 0.5844178195905761,
"learning_rate": 3.993279149458988e-05,
"loss": 0.4397,
"step": 153
},
{
"epoch": 0.6234817813765182,
"grad_norm": 0.6356496130929709,
"learning_rate": 3.9928079338103695e-05,
"loss": 0.4428,
"step": 154
},
{
"epoch": 0.6275303643724697,
"grad_norm": 0.4495852128928547,
"learning_rate": 3.992320783722616e-05,
"loss": 0.4252,
"step": 155
},
{
"epoch": 0.631578947368421,
"grad_norm": 0.44161247140134197,
"learning_rate": 3.991817703090965e-05,
"loss": 0.433,
"step": 156
},
{
"epoch": 0.6356275303643725,
"grad_norm": 0.4933001361262391,
"learning_rate": 3.991298695938038e-05,
"loss": 0.4391,
"step": 157
},
{
"epoch": 0.6396761133603239,
"grad_norm": 0.46127780772648236,
"learning_rate": 3.9907637664138e-05,
"loss": 0.4318,
"step": 158
},
{
"epoch": 0.6437246963562753,
"grad_norm": 0.40261284898177824,
"learning_rate": 3.990212918795534e-05,
"loss": 0.4519,
"step": 159
},
{
"epoch": 0.6477732793522267,
"grad_norm": 0.49117850044482103,
"learning_rate": 3.9896461574878036e-05,
"loss": 0.4157,
"step": 160
},
{
"epoch": 0.6518218623481782,
"grad_norm": 0.516124633597435,
"learning_rate": 3.989063487022416e-05,
"loss": 0.446,
"step": 161
},
{
"epoch": 0.6558704453441295,
"grad_norm": 0.5481111355254353,
"learning_rate": 3.9884649120583903e-05,
"loss": 0.4378,
"step": 162
},
{
"epoch": 0.659919028340081,
"grad_norm": 0.686323241096072,
"learning_rate": 3.987850437381915e-05,
"loss": 0.4236,
"step": 163
},
{
"epoch": 0.6639676113360324,
"grad_norm": 0.7922853282350404,
"learning_rate": 3.987220067906313e-05,
"loss": 0.436,
"step": 164
},
{
"epoch": 0.6680161943319838,
"grad_norm": 0.6440594186172385,
"learning_rate": 3.986573808672002e-05,
"loss": 0.4371,
"step": 165
},
{
"epoch": 0.6720647773279352,
"grad_norm": 0.6685079614890168,
"learning_rate": 3.9859116648464534e-05,
"loss": 0.4243,
"step": 166
},
{
"epoch": 0.6761133603238867,
"grad_norm": 0.809724533487586,
"learning_rate": 3.985233641724152e-05,
"loss": 0.4284,
"step": 167
},
{
"epoch": 0.680161943319838,
"grad_norm": 0.6233647686461461,
"learning_rate": 3.984539744726554e-05,
"loss": 0.4381,
"step": 168
},
{
"epoch": 0.6842105263157895,
"grad_norm": 0.6497890288614713,
"learning_rate": 3.983829979402038e-05,
"loss": 0.42,
"step": 169
},
{
"epoch": 0.6882591093117408,
"grad_norm": 0.7273580401934112,
"learning_rate": 3.983104351425871e-05,
"loss": 0.4514,
"step": 170
},
{
"epoch": 0.6923076923076923,
"grad_norm": 0.5866134305509796,
"learning_rate": 3.982362866600154e-05,
"loss": 0.4267,
"step": 171
},
{
"epoch": 0.6963562753036437,
"grad_norm": 0.5322202628480114,
"learning_rate": 3.9816055308537806e-05,
"loss": 0.4273,
"step": 172
},
{
"epoch": 0.7004048582995951,
"grad_norm": 0.5434711386898373,
"learning_rate": 3.980832350242388e-05,
"loss": 0.4385,
"step": 173
},
{
"epoch": 0.7044534412955465,
"grad_norm": 0.5798217005602863,
"learning_rate": 3.980043330948306e-05,
"loss": 0.4465,
"step": 174
},
{
"epoch": 0.708502024291498,
"grad_norm": 0.45197378808153277,
"learning_rate": 3.9792384792805124e-05,
"loss": 0.4308,
"step": 175
},
{
"epoch": 0.7125506072874493,
"grad_norm": 0.40433285967761307,
"learning_rate": 3.978417801674579e-05,
"loss": 0.4308,
"step": 176
},
{
"epoch": 0.7165991902834008,
"grad_norm": 0.45428375320242814,
"learning_rate": 3.9775813046926244e-05,
"loss": 0.4172,
"step": 177
},
{
"epoch": 0.7206477732793523,
"grad_norm": 0.45310505703251164,
"learning_rate": 3.976728995023254e-05,
"loss": 0.4263,
"step": 178
},
{
"epoch": 0.7246963562753036,
"grad_norm": 0.4864963654015599,
"learning_rate": 3.975860879481514e-05,
"loss": 0.4259,
"step": 179
},
{
"epoch": 0.728744939271255,
"grad_norm": 0.5864008414733007,
"learning_rate": 3.974976965008833e-05,
"loss": 0.4238,
"step": 180
},
{
"epoch": 0.7327935222672065,
"grad_norm": 0.4962475978705514,
"learning_rate": 3.9740772586729674e-05,
"loss": 0.4374,
"step": 181
},
{
"epoch": 0.7368421052631579,
"grad_norm": 0.47813122120789164,
"learning_rate": 3.973161767667946e-05,
"loss": 0.4242,
"step": 182
},
{
"epoch": 0.7408906882591093,
"grad_norm": 0.5604813984294222,
"learning_rate": 3.972230499314009e-05,
"loss": 0.4417,
"step": 183
},
{
"epoch": 0.7449392712550608,
"grad_norm": 0.5720461964633011,
"learning_rate": 3.9712834610575545e-05,
"loss": 0.4254,
"step": 184
},
{
"epoch": 0.7489878542510121,
"grad_norm": 0.4688818098461263,
"learning_rate": 3.9703206604710746e-05,
"loss": 0.4269,
"step": 185
},
{
"epoch": 0.7530364372469636,
"grad_norm": 0.5510274126400005,
"learning_rate": 3.969342105253098e-05,
"loss": 0.4255,
"step": 186
},
{
"epoch": 0.757085020242915,
"grad_norm": 0.4966258509300974,
"learning_rate": 3.9683478032281246e-05,
"loss": 0.4204,
"step": 187
},
{
"epoch": 0.7611336032388664,
"grad_norm": 0.5072686380879243,
"learning_rate": 3.967337762346568e-05,
"loss": 0.4316,
"step": 188
},
{
"epoch": 0.7651821862348178,
"grad_norm": 0.5962262249039644,
"learning_rate": 3.9663119906846885e-05,
"loss": 0.4219,
"step": 189
},
{
"epoch": 0.7692307692307693,
"grad_norm": 0.602356319863623,
"learning_rate": 3.965270496444528e-05,
"loss": 0.437,
"step": 190
},
{
"epoch": 0.7732793522267206,
"grad_norm": 0.5488366546944169,
"learning_rate": 3.9642132879538465e-05,
"loss": 0.4267,
"step": 191
},
{
"epoch": 0.7773279352226721,
"grad_norm": 0.6146587827766175,
"learning_rate": 3.9631403736660566e-05,
"loss": 0.4356,
"step": 192
},
{
"epoch": 0.7813765182186235,
"grad_norm": 0.5068598734988092,
"learning_rate": 3.962051762160151e-05,
"loss": 0.4309,
"step": 193
},
{
"epoch": 0.7854251012145749,
"grad_norm": 0.4105877726934442,
"learning_rate": 3.960947462140638e-05,
"loss": 0.4305,
"step": 194
},
{
"epoch": 0.7894736842105263,
"grad_norm": 0.4826621564842931,
"learning_rate": 3.959827482437472e-05,
"loss": 0.4076,
"step": 195
},
{
"epoch": 0.7935222672064778,
"grad_norm": 0.4777693493523468,
"learning_rate": 3.958691832005981e-05,
"loss": 0.4258,
"step": 196
},
{
"epoch": 0.7975708502024291,
"grad_norm": 0.3816507129483322,
"learning_rate": 3.9575405199267944e-05,
"loss": 0.4215,
"step": 197
},
{
"epoch": 0.8016194331983806,
"grad_norm": 0.472676020301176,
"learning_rate": 3.956373555405774e-05,
"loss": 0.4251,
"step": 198
},
{
"epoch": 0.805668016194332,
"grad_norm": 0.44022439115821327,
"learning_rate": 3.955190947773937e-05,
"loss": 0.4321,
"step": 199
},
{
"epoch": 0.8097165991902834,
"grad_norm": 0.5387862896983523,
"learning_rate": 3.9539927064873824e-05,
"loss": 0.4216,
"step": 200
},
{
"epoch": 0.8137651821862348,
"grad_norm": 0.4564626308492532,
"learning_rate": 3.952778841127214e-05,
"loss": 0.4229,
"step": 201
},
{
"epoch": 0.8178137651821862,
"grad_norm": 0.47988825457578554,
"learning_rate": 3.95154936139947e-05,
"loss": 0.4286,
"step": 202
},
{
"epoch": 0.8218623481781376,
"grad_norm": 0.4617418102486489,
"learning_rate": 3.950304277135034e-05,
"loss": 0.4255,
"step": 203
},
{
"epoch": 0.8259109311740891,
"grad_norm": 0.4658078655886394,
"learning_rate": 3.94904359828957e-05,
"loss": 0.4207,
"step": 204
},
{
"epoch": 0.8299595141700404,
"grad_norm": 0.39730544272216034,
"learning_rate": 3.94776733494343e-05,
"loss": 0.4078,
"step": 205
},
{
"epoch": 0.8340080971659919,
"grad_norm": 0.5200523697699024,
"learning_rate": 3.9464754973015826e-05,
"loss": 0.4302,
"step": 206
},
{
"epoch": 0.8380566801619433,
"grad_norm": 0.45779860854291654,
"learning_rate": 3.9451680956935275e-05,
"loss": 0.4267,
"step": 207
},
{
"epoch": 0.8421052631578947,
"grad_norm": 0.399240864679267,
"learning_rate": 3.943845140573213e-05,
"loss": 0.4333,
"step": 208
},
{
"epoch": 0.8461538461538461,
"grad_norm": 0.413345847054863,
"learning_rate": 3.942506642518952e-05,
"loss": 0.4151,
"step": 209
},
{
"epoch": 0.8502024291497976,
"grad_norm": 0.45157676035243116,
"learning_rate": 3.9411526122333396e-05,
"loss": 0.4271,
"step": 210
},
{
"epoch": 0.854251012145749,
"grad_norm": 0.4857455878207148,
"learning_rate": 3.939783060543167e-05,
"loss": 0.415,
"step": 211
},
{
"epoch": 0.8582995951417004,
"grad_norm": 0.5965576716235503,
"learning_rate": 3.938397998399333e-05,
"loss": 0.4156,
"step": 212
},
{
"epoch": 0.8623481781376519,
"grad_norm": 0.46987236467290516,
"learning_rate": 3.936997436876756e-05,
"loss": 0.4285,
"step": 213
},
{
"epoch": 0.8663967611336032,
"grad_norm": 0.4726272722635668,
"learning_rate": 3.935581387174289e-05,
"loss": 0.427,
"step": 214
},
{
"epoch": 0.8704453441295547,
"grad_norm": 0.6477032599047896,
"learning_rate": 3.9341498606146295e-05,
"loss": 0.4233,
"step": 215
},
{
"epoch": 0.8744939271255061,
"grad_norm": 0.5158335665033074,
"learning_rate": 3.932702868644224e-05,
"loss": 0.4042,
"step": 216
},
{
"epoch": 0.8785425101214575,
"grad_norm": 0.5377471425528877,
"learning_rate": 3.931240422833181e-05,
"loss": 0.4106,
"step": 217
},
{
"epoch": 0.8825910931174089,
"grad_norm": 0.4881490666739434,
"learning_rate": 3.9297625348751806e-05,
"loss": 0.434,
"step": 218
},
{
"epoch": 0.8866396761133604,
"grad_norm": 0.38359331951899395,
"learning_rate": 3.9282692165873744e-05,
"loss": 0.4183,
"step": 219
},
{
"epoch": 0.8906882591093117,
"grad_norm": 0.4803450453867985,
"learning_rate": 3.9267604799102945e-05,
"loss": 0.4107,
"step": 220
},
{
"epoch": 0.8947368421052632,
"grad_norm": 0.45962461602416227,
"learning_rate": 3.9252363369077615e-05,
"loss": 0.4122,
"step": 221
},
{
"epoch": 0.8987854251012146,
"grad_norm": 0.5135394481619739,
"learning_rate": 3.9236967997667796e-05,
"loss": 0.4176,
"step": 222
},
{
"epoch": 0.902834008097166,
"grad_norm": 0.5035679101318152,
"learning_rate": 3.9221418807974496e-05,
"loss": 0.4202,
"step": 223
},
{
"epoch": 0.9068825910931174,
"grad_norm": 0.4309195262180922,
"learning_rate": 3.92057159243286e-05,
"loss": 0.4249,
"step": 224
},
{
"epoch": 0.9109311740890689,
"grad_norm": 0.5174358855797475,
"learning_rate": 3.9189859472289956e-05,
"loss": 0.4157,
"step": 225
},
{
"epoch": 0.9149797570850202,
"grad_norm": 0.4714792315537687,
"learning_rate": 3.9173849578646324e-05,
"loss": 0.4191,
"step": 226
},
{
"epoch": 0.9190283400809717,
"grad_norm": 0.5833808284626206,
"learning_rate": 3.9157686371412395e-05,
"loss": 0.4214,
"step": 227
},
{
"epoch": 0.9230769230769231,
"grad_norm": 0.5131437367143241,
"learning_rate": 3.914136997982874e-05,
"loss": 0.4207,
"step": 228
},
{
"epoch": 0.9271255060728745,
"grad_norm": 0.40564974309691876,
"learning_rate": 3.912490053436079e-05,
"loss": 0.4138,
"step": 229
},
{
"epoch": 0.9311740890688259,
"grad_norm": 0.4189299795646502,
"learning_rate": 3.9108278166697805e-05,
"loss": 0.3994,
"step": 230
},
{
"epoch": 0.9352226720647774,
"grad_norm": 0.4904409272882797,
"learning_rate": 3.909150300975178e-05,
"loss": 0.4106,
"step": 231
},
{
"epoch": 0.9392712550607287,
"grad_norm": 0.43850602402642874,
"learning_rate": 3.907457519765642e-05,
"loss": 0.4142,
"step": 232
},
{
"epoch": 0.9433198380566802,
"grad_norm": 0.47778307951054744,
"learning_rate": 3.905749486576607e-05,
"loss": 0.4146,
"step": 233
},
{
"epoch": 0.9473684210526315,
"grad_norm": 0.3767454576636596,
"learning_rate": 3.90402621506546e-05,
"loss": 0.4238,
"step": 234
},
{
"epoch": 0.951417004048583,
"grad_norm": 0.4825776581549396,
"learning_rate": 3.902287719011435e-05,
"loss": 0.4112,
"step": 235
},
{
"epoch": 0.9554655870445344,
"grad_norm": 0.4980353060094576,
"learning_rate": 3.900534012315499e-05,
"loss": 0.4169,
"step": 236
},
{
"epoch": 0.9595141700404858,
"grad_norm": 0.5087086181176929,
"learning_rate": 3.8987651090002454e-05,
"loss": 0.4234,
"step": 237
},
{
"epoch": 0.9635627530364372,
"grad_norm": 0.5452685968066828,
"learning_rate": 3.896981023209777e-05,
"loss": 0.425,
"step": 238
},
{
"epoch": 0.9676113360323887,
"grad_norm": 0.4290817090074161,
"learning_rate": 3.8951817692095966e-05,
"loss": 0.4035,
"step": 239
},
{
"epoch": 0.97165991902834,
"grad_norm": 0.4547385427852664,
"learning_rate": 3.893367361386491e-05,
"loss": 0.4116,
"step": 240
},
{
"epoch": 0.9757085020242915,
"grad_norm": 0.5064929121121033,
"learning_rate": 3.891537814248417e-05,
"loss": 0.411,
"step": 241
},
{
"epoch": 0.979757085020243,
"grad_norm": 0.48249463203797116,
"learning_rate": 3.889693142424384e-05,
"loss": 0.4053,
"step": 242
},
{
"epoch": 0.9838056680161943,
"grad_norm": 0.43701320527191384,
"learning_rate": 3.8878333606643405e-05,
"loss": 0.4046,
"step": 243
},
{
"epoch": 0.9878542510121457,
"grad_norm": 0.4279238441479248,
"learning_rate": 3.885958483839049e-05,
"loss": 0.4281,
"step": 244
},
{
"epoch": 0.9919028340080972,
"grad_norm": 0.3489035059997058,
"learning_rate": 3.8840685269399786e-05,
"loss": 0.4067,
"step": 245
},
{
"epoch": 0.9959514170040485,
"grad_norm": 0.44304859817283937,
"learning_rate": 3.882163505079171e-05,
"loss": 0.4123,
"step": 246
},
{
"epoch": 1.0,
"grad_norm": 0.4767115766435931,
"learning_rate": 3.880243433489132e-05,
"loss": 0.4039,
"step": 247
},
{
"epoch": 1.0040485829959513,
"grad_norm": 0.4525740139958258,
"learning_rate": 3.878308327522702e-05,
"loss": 0.3705,
"step": 248
},
{
"epoch": 1.008097165991903,
"grad_norm": 0.5938907983931794,
"learning_rate": 3.876358202652936e-05,
"loss": 0.3773,
"step": 249
},
{
"epoch": 1.0121457489878543,
"grad_norm": 0.6128600059885416,
"learning_rate": 3.8743930744729835e-05,
"loss": 0.3838,
"step": 250
},
{
"epoch": 1.0161943319838056,
"grad_norm": 0.4566032861544371,
"learning_rate": 3.8724129586959535e-05,
"loss": 0.3661,
"step": 251
},
{
"epoch": 1.0202429149797572,
"grad_norm": 0.5113199679856734,
"learning_rate": 3.870417871154801e-05,
"loss": 0.3686,
"step": 252
},
{
"epoch": 1.0242914979757085,
"grad_norm": 0.6324268407212755,
"learning_rate": 3.868407827802192e-05,
"loss": 0.3591,
"step": 253
},
{
"epoch": 1.0283400809716599,
"grad_norm": 0.6187620544553952,
"learning_rate": 3.8663828447103804e-05,
"loss": 0.3789,
"step": 254
},
{
"epoch": 1.0323886639676114,
"grad_norm": 0.5151873830885907,
"learning_rate": 3.8643429380710765e-05,
"loss": 0.384,
"step": 255
},
{
"epoch": 1.0364372469635628,
"grad_norm": 0.48304513477224464,
"learning_rate": 3.862288124195319e-05,
"loss": 0.3746,
"step": 256
},
{
"epoch": 1.040485829959514,
"grad_norm": 0.48060792774091604,
"learning_rate": 3.8602184195133464e-05,
"loss": 0.3674,
"step": 257
},
{
"epoch": 1.0445344129554657,
"grad_norm": 0.5817685470911269,
"learning_rate": 3.8581338405744604e-05,
"loss": 0.368,
"step": 258
},
{
"epoch": 1.048582995951417,
"grad_norm": 0.5020834291105305,
"learning_rate": 3.8560344040468996e-05,
"loss": 0.3814,
"step": 259
},
{
"epoch": 1.0526315789473684,
"grad_norm": 0.6937241875848924,
"learning_rate": 3.853920126717703e-05,
"loss": 0.3584,
"step": 260
},
{
"epoch": 1.05668016194332,
"grad_norm": 0.5609824460288562,
"learning_rate": 3.851791025492574e-05,
"loss": 0.3775,
"step": 261
},
{
"epoch": 1.0607287449392713,
"grad_norm": 0.45758995107771405,
"learning_rate": 3.8496471173957515e-05,
"loss": 0.3721,
"step": 262
},
{
"epoch": 1.0647773279352226,
"grad_norm": 0.5579825278589824,
"learning_rate": 3.847488419569866e-05,
"loss": 0.3817,
"step": 263
},
{
"epoch": 1.0688259109311742,
"grad_norm": 0.5470270787216607,
"learning_rate": 3.8453149492758077e-05,
"loss": 0.383,
"step": 264
},
{
"epoch": 1.0728744939271255,
"grad_norm": 0.4463759398245354,
"learning_rate": 3.8431267238925886e-05,
"loss": 0.3674,
"step": 265
},
{
"epoch": 1.0769230769230769,
"grad_norm": 0.4673351217519132,
"learning_rate": 3.840923760917199e-05,
"loss": 0.3705,
"step": 266
},
{
"epoch": 1.0809716599190284,
"grad_norm": 0.46841183690802823,
"learning_rate": 3.838706077964473e-05,
"loss": 0.3824,
"step": 267
},
{
"epoch": 1.0850202429149798,
"grad_norm": 0.3978158292787416,
"learning_rate": 3.836473692766944e-05,
"loss": 0.3612,
"step": 268
},
{
"epoch": 1.0890688259109311,
"grad_norm": 0.4826076270913436,
"learning_rate": 3.834226623174705e-05,
"loss": 0.3601,
"step": 269
},
{
"epoch": 1.0931174089068827,
"grad_norm": 0.48371239763246654,
"learning_rate": 3.831964887155264e-05,
"loss": 0.3782,
"step": 270
},
{
"epoch": 1.097165991902834,
"grad_norm": 0.45847098290658006,
"learning_rate": 3.829688502793403e-05,
"loss": 0.3697,
"step": 271
},
{
"epoch": 1.1012145748987854,
"grad_norm": 0.5484109652418133,
"learning_rate": 3.827397488291029e-05,
"loss": 0.363,
"step": 272
},
{
"epoch": 1.1052631578947367,
"grad_norm": 0.5049581659728827,
"learning_rate": 3.8250918619670343e-05,
"loss": 0.3732,
"step": 273
},
{
"epoch": 1.1093117408906883,
"grad_norm": 0.4501375567389312,
"learning_rate": 3.822771642257145e-05,
"loss": 0.3645,
"step": 274
},
{
"epoch": 1.1133603238866396,
"grad_norm": 0.3760067553461168,
"learning_rate": 3.820436847713776e-05,
"loss": 0.3714,
"step": 275
},
{
"epoch": 1.117408906882591,
"grad_norm": 0.4237983418193164,
"learning_rate": 3.8180874970058826e-05,
"loss": 0.379,
"step": 276
},
{
"epoch": 1.1214574898785425,
"grad_norm": 0.430337640902461,
"learning_rate": 3.8157236089188103e-05,
"loss": 0.3755,
"step": 277
},
{
"epoch": 1.125506072874494,
"grad_norm": 0.35090294447723225,
"learning_rate": 3.813345202354145e-05,
"loss": 0.3715,
"step": 278
},
{
"epoch": 1.1295546558704452,
"grad_norm": 0.3664898414209936,
"learning_rate": 3.810952296329563e-05,
"loss": 0.3627,
"step": 279
},
{
"epoch": 1.1336032388663968,
"grad_norm": 0.39813619877231465,
"learning_rate": 3.808544909978676e-05,
"loss": 0.3708,
"step": 280
},
{
"epoch": 1.1376518218623481,
"grad_norm": 0.4652516329517251,
"learning_rate": 3.806123062550882e-05,
"loss": 0.3677,
"step": 281
},
{
"epoch": 1.1417004048582995,
"grad_norm": 0.4236986193131142,
"learning_rate": 3.803686773411208e-05,
"loss": 0.3701,
"step": 282
},
{
"epoch": 1.145748987854251,
"grad_norm": 0.4140537531523329,
"learning_rate": 3.801236062040158e-05,
"loss": 0.3627,
"step": 283
},
{
"epoch": 1.1497975708502024,
"grad_norm": 0.41642540509592396,
"learning_rate": 3.798770948033553e-05,
"loss": 0.3615,
"step": 284
},
{
"epoch": 1.1538461538461537,
"grad_norm": 0.43569480293913904,
"learning_rate": 3.796291451102381e-05,
"loss": 0.366,
"step": 285
},
{
"epoch": 1.1578947368421053,
"grad_norm": 0.406608000049377,
"learning_rate": 3.793797591072631e-05,
"loss": 0.3655,
"step": 286
},
{
"epoch": 1.1619433198380567,
"grad_norm": 0.4360888806160911,
"learning_rate": 3.791289387885144e-05,
"loss": 0.3662,
"step": 287
},
{
"epoch": 1.165991902834008,
"grad_norm": 0.3696133354678672,
"learning_rate": 3.788766861595443e-05,
"loss": 0.3647,
"step": 288
},
{
"epoch": 1.1700404858299596,
"grad_norm": 0.41377423387629014,
"learning_rate": 3.7862300323735835e-05,
"loss": 0.3843,
"step": 289
},
{
"epoch": 1.174089068825911,
"grad_norm": 0.3804174820971082,
"learning_rate": 3.783678920503983e-05,
"loss": 0.3737,
"step": 290
},
{
"epoch": 1.1781376518218623,
"grad_norm": 0.41258345122670964,
"learning_rate": 3.781113546385265e-05,
"loss": 0.3712,
"step": 291
},
{
"epoch": 1.1821862348178138,
"grad_norm": 0.3825134489540836,
"learning_rate": 3.7785339305300916e-05,
"loss": 0.3738,
"step": 292
},
{
"epoch": 1.1862348178137652,
"grad_norm": 0.3576840255460259,
"learning_rate": 3.7759400935650025e-05,
"loss": 0.3717,
"step": 293
},
{
"epoch": 1.1902834008097165,
"grad_norm": 0.3472048840436469,
"learning_rate": 3.77333205623025e-05,
"loss": 0.3594,
"step": 294
},
{
"epoch": 1.194331983805668,
"grad_norm": 0.38728481298838113,
"learning_rate": 3.7707098393796295e-05,
"loss": 0.3708,
"step": 295
},
{
"epoch": 1.1983805668016194,
"grad_norm": 0.44679639105121766,
"learning_rate": 3.7680734639803185e-05,
"loss": 0.3716,
"step": 296
},
{
"epoch": 1.2024291497975708,
"grad_norm": 0.3755740031483349,
"learning_rate": 3.765422951112704e-05,
"loss": 0.3718,
"step": 297
},
{
"epoch": 1.2064777327935223,
"grad_norm": 0.3983206483365791,
"learning_rate": 3.762758321970216e-05,
"loss": 0.3612,
"step": 298
},
{
"epoch": 1.2105263157894737,
"grad_norm": 0.3766974432925036,
"learning_rate": 3.7600795978591584e-05,
"loss": 0.378,
"step": 299
},
{
"epoch": 1.214574898785425,
"grad_norm": 0.36890775596059333,
"learning_rate": 3.757386800198538e-05,
"loss": 0.3784,
"step": 300
},
{
"epoch": 1.2186234817813766,
"grad_norm": 0.4283627762496141,
"learning_rate": 3.7546799505198925e-05,
"loss": 0.3632,
"step": 301
},
{
"epoch": 1.222672064777328,
"grad_norm": 0.4389474531231389,
"learning_rate": 3.7519590704671197e-05,
"loss": 0.3698,
"step": 302
},
{
"epoch": 1.2267206477732793,
"grad_norm": 0.4323646857917935,
"learning_rate": 3.749224181796305e-05,
"loss": 0.3712,
"step": 303
},
{
"epoch": 1.2307692307692308,
"grad_norm": 0.32751044417607394,
"learning_rate": 3.746475306375545e-05,
"loss": 0.3701,
"step": 304
},
{
"epoch": 1.2348178137651822,
"grad_norm": 0.4169368810116266,
"learning_rate": 3.743712466184774e-05,
"loss": 0.3659,
"step": 305
},
{
"epoch": 1.2388663967611335,
"grad_norm": 0.42090857633126855,
"learning_rate": 3.7409356833155885e-05,
"loss": 0.3597,
"step": 306
},
{
"epoch": 1.242914979757085,
"grad_norm": 0.46457571324451785,
"learning_rate": 3.73814497997107e-05,
"loss": 0.3767,
"step": 307
},
{
"epoch": 1.2469635627530364,
"grad_norm": 0.3831245905601353,
"learning_rate": 3.7353403784656095e-05,
"loss": 0.3804,
"step": 308
},
{
"epoch": 1.2510121457489878,
"grad_norm": 0.31971868065637965,
"learning_rate": 3.732521901224724e-05,
"loss": 0.3679,
"step": 309
},
{
"epoch": 1.2550607287449393,
"grad_norm": 0.43254003179233214,
"learning_rate": 3.7296895707848844e-05,
"loss": 0.3778,
"step": 310
},
{
"epoch": 1.2591093117408907,
"grad_norm": 0.4263513984467474,
"learning_rate": 3.7268434097933275e-05,
"loss": 0.3783,
"step": 311
},
{
"epoch": 1.263157894736842,
"grad_norm": 0.34806420586124237,
"learning_rate": 3.723983441007882e-05,
"loss": 0.354,
"step": 312
},
{
"epoch": 1.2672064777327936,
"grad_norm": 0.3847656650465733,
"learning_rate": 3.721109687296781e-05,
"loss": 0.3605,
"step": 313
},
{
"epoch": 1.271255060728745,
"grad_norm": 0.4157311774954319,
"learning_rate": 3.718222171638484e-05,
"loss": 0.3751,
"step": 314
},
{
"epoch": 1.2753036437246963,
"grad_norm": 0.362946018079254,
"learning_rate": 3.715320917121488e-05,
"loss": 0.3683,
"step": 315
},
{
"epoch": 1.2793522267206479,
"grad_norm": 0.32210709187840675,
"learning_rate": 3.712405946944149e-05,
"loss": 0.3538,
"step": 316
},
{
"epoch": 1.2834008097165992,
"grad_norm": 0.3445659937440463,
"learning_rate": 3.709477284414488e-05,
"loss": 0.3585,
"step": 317
},
{
"epoch": 1.2874493927125505,
"grad_norm": 0.373761179610803,
"learning_rate": 3.7065349529500136e-05,
"loss": 0.361,
"step": 318
},
{
"epoch": 1.291497975708502,
"grad_norm": 0.4638532938783912,
"learning_rate": 3.7035789760775305e-05,
"loss": 0.3724,
"step": 319
},
{
"epoch": 1.2955465587044535,
"grad_norm": 0.3285074006937118,
"learning_rate": 3.70060937743295e-05,
"loss": 0.3727,
"step": 320
},
{
"epoch": 1.2995951417004048,
"grad_norm": 0.447359452794575,
"learning_rate": 3.6976261807611045e-05,
"loss": 0.3788,
"step": 321
},
{
"epoch": 1.3036437246963564,
"grad_norm": 0.4860481721590511,
"learning_rate": 3.694629409915555e-05,
"loss": 0.3758,
"step": 322
},
{
"epoch": 1.3076923076923077,
"grad_norm": 0.3937962634069424,
"learning_rate": 3.691619088858401e-05,
"loss": 0.37,
"step": 323
},
{
"epoch": 1.311740890688259,
"grad_norm": 0.3197856014673181,
"learning_rate": 3.68859524166009e-05,
"loss": 0.3547,
"step": 324
},
{
"epoch": 1.3157894736842106,
"grad_norm": 0.43517095329347566,
"learning_rate": 3.6855578924992226e-05,
"loss": 0.3532,
"step": 325
},
{
"epoch": 1.319838056680162,
"grad_norm": 0.4018628375502082,
"learning_rate": 3.6825070656623626e-05,
"loss": 0.3641,
"step": 326
},
{
"epoch": 1.3238866396761133,
"grad_norm": 0.5337020427582904,
"learning_rate": 3.67944278554384e-05,
"loss": 0.3751,
"step": 327
},
{
"epoch": 1.3279352226720649,
"grad_norm": 0.3537918283469127,
"learning_rate": 3.676365076645557e-05,
"loss": 0.3644,
"step": 328
},
{
"epoch": 1.3319838056680162,
"grad_norm": 0.46084107475388375,
"learning_rate": 3.673273963576791e-05,
"loss": 0.3626,
"step": 329
},
{
"epoch": 1.3360323886639676,
"grad_norm": 0.3688237740969206,
"learning_rate": 3.670169471054001e-05,
"loss": 0.3686,
"step": 330
},
{
"epoch": 1.3400809716599191,
"grad_norm": 0.41848249603768284,
"learning_rate": 3.6670516239006254e-05,
"loss": 0.3617,
"step": 331
},
{
"epoch": 1.3441295546558705,
"grad_norm": 0.39011121291637973,
"learning_rate": 3.663920447046888e-05,
"loss": 0.3821,
"step": 332
},
{
"epoch": 1.3481781376518218,
"grad_norm": 0.42354665469252195,
"learning_rate": 3.6607759655295954e-05,
"loss": 0.3582,
"step": 333
},
{
"epoch": 1.3522267206477734,
"grad_norm": 0.43507141402403277,
"learning_rate": 3.657618204491937e-05,
"loss": 0.3697,
"step": 334
},
{
"epoch": 1.3562753036437247,
"grad_norm": 0.3847556887875649,
"learning_rate": 3.6544471891832875e-05,
"loss": 0.3675,
"step": 335
},
{
"epoch": 1.360323886639676,
"grad_norm": 0.423548429669823,
"learning_rate": 3.6512629449589983e-05,
"loss": 0.3624,
"step": 336
},
{
"epoch": 1.3643724696356276,
"grad_norm": 0.3248751478076679,
"learning_rate": 3.6480654972802045e-05,
"loss": 0.3692,
"step": 337
},
{
"epoch": 1.368421052631579,
"grad_norm": 0.4161391888996143,
"learning_rate": 3.644854871713611e-05,
"loss": 0.3665,
"step": 338
},
{
"epoch": 1.3724696356275303,
"grad_norm": 0.43669989085965893,
"learning_rate": 3.6416310939312924e-05,
"loss": 0.3613,
"step": 339
},
{
"epoch": 1.376518218623482,
"grad_norm": 0.36325879398956995,
"learning_rate": 3.638394189710493e-05,
"loss": 0.3694,
"step": 340
},
{
"epoch": 1.3805668016194332,
"grad_norm": 0.41810318165812393,
"learning_rate": 3.635144184933412e-05,
"loss": 0.3636,
"step": 341
},
{
"epoch": 1.3846153846153846,
"grad_norm": 0.31041430570027534,
"learning_rate": 3.6318811055869994e-05,
"loss": 0.3653,
"step": 342
},
{
"epoch": 1.3886639676113361,
"grad_norm": 0.3975632884602416,
"learning_rate": 3.628604977762752e-05,
"loss": 0.371,
"step": 343
},
{
"epoch": 1.3927125506072875,
"grad_norm": 0.3972565095822435,
"learning_rate": 3.6253158276565004e-05,
"loss": 0.3737,
"step": 344
},
{
"epoch": 1.3967611336032388,
"grad_norm": 0.42480254739932233,
"learning_rate": 3.6220136815682016e-05,
"loss": 0.3798,
"step": 345
},
{
"epoch": 1.4008097165991904,
"grad_norm": 0.39356522592127563,
"learning_rate": 3.618698565901727e-05,
"loss": 0.3649,
"step": 346
},
{
"epoch": 1.4048582995951417,
"grad_norm": 0.33964087219204664,
"learning_rate": 3.615370507164654e-05,
"loss": 0.3718,
"step": 347
},
{
"epoch": 1.408906882591093,
"grad_norm": 0.3877301589998507,
"learning_rate": 3.61202953196805e-05,
"loss": 0.3655,
"step": 348
},
{
"epoch": 1.4129554655870447,
"grad_norm": 0.3764696563978075,
"learning_rate": 3.6086756670262654e-05,
"loss": 0.3574,
"step": 349
},
{
"epoch": 1.417004048582996,
"grad_norm": 0.3603569523160451,
"learning_rate": 3.6053089391567135e-05,
"loss": 0.3607,
"step": 350
},
{
"epoch": 1.4210526315789473,
"grad_norm": 0.3563570725158935,
"learning_rate": 3.601929375279662e-05,
"loss": 0.3631,
"step": 351
},
{
"epoch": 1.425101214574899,
"grad_norm": 0.34030429066489387,
"learning_rate": 3.598537002418012e-05,
"loss": 0.3586,
"step": 352
},
{
"epoch": 1.4291497975708503,
"grad_norm": 0.32450083153135667,
"learning_rate": 3.5951318476970885e-05,
"loss": 0.3712,
"step": 353
},
{
"epoch": 1.4331983805668016,
"grad_norm": 0.33054724240262223,
"learning_rate": 3.591713938344416e-05,
"loss": 0.3642,
"step": 354
},
{
"epoch": 1.4372469635627532,
"grad_norm": 0.3194563560790769,
"learning_rate": 3.588283301689507e-05,
"loss": 0.3692,
"step": 355
},
{
"epoch": 1.4412955465587045,
"grad_norm": 6.105354473181902,
"learning_rate": 3.5848399651636424e-05,
"loss": 0.4457,
"step": 356
},
{
"epoch": 1.4453441295546559,
"grad_norm": 0.5691386876627894,
"learning_rate": 3.5813839562996473e-05,
"loss": 0.3763,
"step": 357
},
{
"epoch": 1.4493927125506074,
"grad_norm": 0.5870064283396903,
"learning_rate": 3.5779153027316775e-05,
"loss": 0.3584,
"step": 358
},
{
"epoch": 1.4534412955465588,
"grad_norm": 0.7238054324227389,
"learning_rate": 3.574434032194994e-05,
"loss": 0.3671,
"step": 359
},
{
"epoch": 1.45748987854251,
"grad_norm": 0.6401209910083444,
"learning_rate": 3.5709401725257436e-05,
"loss": 0.3639,
"step": 360
},
{
"epoch": 1.4615384615384617,
"grad_norm": 0.3950153707135682,
"learning_rate": 3.5674337516607346e-05,
"loss": 0.364,
"step": 361
},
{
"epoch": 1.465587044534413,
"grad_norm": 0.8522275167982308,
"learning_rate": 3.563914797637216e-05,
"loss": 0.3549,
"step": 362
},
{
"epoch": 1.4696356275303644,
"grad_norm": 0.8338044498279499,
"learning_rate": 3.560383338592649e-05,
"loss": 0.375,
"step": 363
},
{
"epoch": 1.4736842105263157,
"grad_norm": 0.4046203426817598,
"learning_rate": 3.556839402764487e-05,
"loss": 0.3721,
"step": 364
},
{
"epoch": 1.4777327935222673,
"grad_norm": 0.4786534340532183,
"learning_rate": 3.553283018489946e-05,
"loss": 0.3629,
"step": 365
},
{
"epoch": 1.4817813765182186,
"grad_norm": 0.3483379542801942,
"learning_rate": 3.54971421420578e-05,
"loss": 0.3774,
"step": 366
},
{
"epoch": 1.48582995951417,
"grad_norm": 0.44437595910281746,
"learning_rate": 3.5461330184480524e-05,
"loss": 0.361,
"step": 367
},
{
"epoch": 1.4898785425101215,
"grad_norm": 0.40908041721382044,
"learning_rate": 3.54253945985191e-05,
"loss": 0.3677,
"step": 368
},
{
"epoch": 1.4939271255060729,
"grad_norm": 0.40559883319455076,
"learning_rate": 3.5389335671513524e-05,
"loss": 0.3615,
"step": 369
},
{
"epoch": 1.4979757085020242,
"grad_norm": 0.37907923855982767,
"learning_rate": 3.5353153691789986e-05,
"loss": 0.3636,
"step": 370
},
{
"epoch": 1.5020242914979756,
"grad_norm": 0.3995856043468485,
"learning_rate": 3.531684894865866e-05,
"loss": 0.3658,
"step": 371
},
{
"epoch": 1.5060728744939271,
"grad_norm": 0.3793161920950771,
"learning_rate": 3.528042173241128e-05,
"loss": 0.3707,
"step": 372
},
{
"epoch": 1.5101214574898787,
"grad_norm": 0.4365752356756936,
"learning_rate": 3.524387233431892e-05,
"loss": 0.3693,
"step": 373
},
{
"epoch": 1.5141700404858298,
"grad_norm": 0.322594636720192,
"learning_rate": 3.520720104662958e-05,
"loss": 0.3585,
"step": 374
},
{
"epoch": 1.5182186234817814,
"grad_norm": 0.3773691441302527,
"learning_rate": 3.5170408162565904e-05,
"loss": 0.3722,
"step": 375
},
{
"epoch": 1.522267206477733,
"grad_norm": 0.39526257847018353,
"learning_rate": 3.513349397632281e-05,
"loss": 0.358,
"step": 376
},
{
"epoch": 1.526315789473684,
"grad_norm": 0.38477967054947715,
"learning_rate": 3.5096458783065145e-05,
"loss": 0.3617,
"step": 377
},
{
"epoch": 1.5303643724696356,
"grad_norm": 0.44994064483712143,
"learning_rate": 3.505930287892533e-05,
"loss": 0.3756,
"step": 378
},
{
"epoch": 1.5344129554655872,
"grad_norm": 0.3896879321208562,
"learning_rate": 3.502202656100099e-05,
"loss": 0.3588,
"step": 379
},
{
"epoch": 1.5384615384615383,
"grad_norm": 0.400379075023613,
"learning_rate": 3.498463012735256e-05,
"loss": 0.357,
"step": 380
},
{
"epoch": 1.54251012145749,
"grad_norm": 0.3980489077093668,
"learning_rate": 3.494711387700094e-05,
"loss": 0.3679,
"step": 381
},
{
"epoch": 1.5465587044534415,
"grad_norm": 0.3005797708339844,
"learning_rate": 3.490947810992505e-05,
"loss": 0.3654,
"step": 382
},
{
"epoch": 1.5506072874493926,
"grad_norm": 0.4376225104980973,
"learning_rate": 3.48717231270595e-05,
"loss": 0.3728,
"step": 383
},
{
"epoch": 1.5546558704453441,
"grad_norm": 0.4339156269815808,
"learning_rate": 3.483384923029211e-05,
"loss": 0.3803,
"step": 384
},
{
"epoch": 1.5587044534412957,
"grad_norm": 0.3604827006210426,
"learning_rate": 3.479585672246156e-05,
"loss": 0.3694,
"step": 385
},
{
"epoch": 1.5627530364372468,
"grad_norm": 0.33760017076773674,
"learning_rate": 3.4757745907354924e-05,
"loss": 0.3606,
"step": 386
},
{
"epoch": 1.5668016194331984,
"grad_norm": 0.38661038893487587,
"learning_rate": 3.471951708970527e-05,
"loss": 0.3702,
"step": 387
},
{
"epoch": 1.5708502024291497,
"grad_norm": 0.37722431818750185,
"learning_rate": 3.468117057518921e-05,
"loss": 0.3773,
"step": 388
},
{
"epoch": 1.574898785425101,
"grad_norm": 0.38413900938610956,
"learning_rate": 3.4642706670424426e-05,
"loss": 0.3774,
"step": 389
},
{
"epoch": 1.5789473684210527,
"grad_norm": 0.39635663359627016,
"learning_rate": 3.460412568296731e-05,
"loss": 0.3479,
"step": 390
},
{
"epoch": 1.582995951417004,
"grad_norm": 0.3993876210070804,
"learning_rate": 3.456542792131039e-05,
"loss": 0.3618,
"step": 391
},
{
"epoch": 1.5870445344129553,
"grad_norm": 0.33989700748762447,
"learning_rate": 3.4526613694879936e-05,
"loss": 0.3724,
"step": 392
},
{
"epoch": 1.591093117408907,
"grad_norm": 0.39749863398496466,
"learning_rate": 3.448768331403348e-05,
"loss": 0.3671,
"step": 393
},
{
"epoch": 1.5951417004048583,
"grad_norm": 0.42530451655227147,
"learning_rate": 3.44486370900573e-05,
"loss": 0.3696,
"step": 394
},
{
"epoch": 1.5991902834008096,
"grad_norm": 0.33103808704830834,
"learning_rate": 3.440947533516396e-05,
"loss": 0.35,
"step": 395
},
{
"epoch": 1.6032388663967612,
"grad_norm": 0.4115737976933898,
"learning_rate": 3.4370198362489816e-05,
"loss": 0.3653,
"step": 396
},
{
"epoch": 1.6072874493927125,
"grad_norm": 0.3501754484910499,
"learning_rate": 3.43308064860925e-05,
"loss": 0.3836,
"step": 397
},
{
"epoch": 1.6113360323886639,
"grad_norm": 0.41492762383430976,
"learning_rate": 3.4291300020948393e-05,
"loss": 0.3676,
"step": 398
},
{
"epoch": 1.6153846153846154,
"grad_norm": 0.3764425543033012,
"learning_rate": 3.425167928295015e-05,
"loss": 0.3777,
"step": 399
},
{
"epoch": 1.6194331983805668,
"grad_norm": 0.9219090666869455,
"learning_rate": 3.421194458890411e-05,
"loss": 0.376,
"step": 400
},
{
"epoch": 1.623481781376518,
"grad_norm": 0.32300052115788586,
"learning_rate": 3.4172096256527865e-05,
"loss": 0.352,
"step": 401
},
{
"epoch": 1.6275303643724697,
"grad_norm": 0.47815201736556434,
"learning_rate": 3.4132134604447595e-05,
"loss": 0.3823,
"step": 402
},
{
"epoch": 1.631578947368421,
"grad_norm": 0.42946572549184436,
"learning_rate": 3.409205995219561e-05,
"loss": 0.3653,
"step": 403
},
{
"epoch": 1.6356275303643724,
"grad_norm": 0.29438071575775465,
"learning_rate": 3.4051872620207765e-05,
"loss": 0.36,
"step": 404
},
{
"epoch": 1.639676113360324,
"grad_norm": 0.42123952884308363,
"learning_rate": 3.40115729298209e-05,
"loss": 0.3691,
"step": 405
},
{
"epoch": 1.6437246963562753,
"grad_norm": 0.4980270164423123,
"learning_rate": 3.397116120327025e-05,
"loss": 0.3475,
"step": 406
},
{
"epoch": 1.6477732793522266,
"grad_norm": 0.4353271016045528,
"learning_rate": 3.3930637763686916e-05,
"loss": 0.3591,
"step": 407
},
{
"epoch": 1.6518218623481782,
"grad_norm": 0.36020362067827827,
"learning_rate": 3.389000293509524e-05,
"loss": 0.3637,
"step": 408
},
{
"epoch": 1.6558704453441295,
"grad_norm": 0.43883254430065693,
"learning_rate": 3.3849257042410225e-05,
"loss": 0.3756,
"step": 409
},
{
"epoch": 1.6599190283400809,
"grad_norm": 0.36305614793413793,
"learning_rate": 3.380840041143494e-05,
"loss": 0.3772,
"step": 410
},
{
"epoch": 1.6639676113360324,
"grad_norm": 0.3566350481744007,
"learning_rate": 3.376743336885792e-05,
"loss": 0.3627,
"step": 411
},
{
"epoch": 1.6680161943319838,
"grad_norm": 0.3393237990063532,
"learning_rate": 3.372635624225055e-05,
"loss": 0.3716,
"step": 412
},
{
"epoch": 1.6720647773279351,
"grad_norm": 0.46241391498354106,
"learning_rate": 3.3685169360064445e-05,
"loss": 0.3568,
"step": 413
},
{
"epoch": 1.6761133603238867,
"grad_norm": 0.36235941003401034,
"learning_rate": 3.3643873051628804e-05,
"loss": 0.3551,
"step": 414
},
{
"epoch": 1.680161943319838,
"grad_norm": 0.415646239190765,
"learning_rate": 3.3602467647147814e-05,
"loss": 0.3576,
"step": 415
},
{
"epoch": 1.6842105263157894,
"grad_norm": 0.38611359185106614,
"learning_rate": 3.356095347769799e-05,
"loss": 0.3673,
"step": 416
},
{
"epoch": 1.688259109311741,
"grad_norm": 0.32447518215596377,
"learning_rate": 3.351933087522552e-05,
"loss": 0.3694,
"step": 417
},
{
"epoch": 1.6923076923076923,
"grad_norm": 0.3569355688247402,
"learning_rate": 3.3477600172543626e-05,
"loss": 0.3711,
"step": 418
},
{
"epoch": 1.6963562753036436,
"grad_norm": 0.3980203474764936,
"learning_rate": 3.3435761703329894e-05,
"loss": 0.3688,
"step": 419
},
{
"epoch": 1.7004048582995952,
"grad_norm": 0.3102356532213161,
"learning_rate": 3.3393815802123606e-05,
"loss": 0.3675,
"step": 420
},
{
"epoch": 1.7044534412955465,
"grad_norm": 0.39938499033579966,
"learning_rate": 3.335176280432308e-05,
"loss": 0.3645,
"step": 421
},
{
"epoch": 1.708502024291498,
"grad_norm": 0.3083580922542779,
"learning_rate": 3.330960304618295e-05,
"loss": 0.3598,
"step": 422
},
{
"epoch": 1.7125506072874495,
"grad_norm": 0.3030983943567573,
"learning_rate": 3.326733686481153e-05,
"loss": 0.3663,
"step": 423
},
{
"epoch": 1.7165991902834008,
"grad_norm": 0.32469683247886844,
"learning_rate": 3.3224964598168084e-05,
"loss": 0.3665,
"step": 424
},
{
"epoch": 1.7206477732793521,
"grad_norm": 0.3264855378845488,
"learning_rate": 3.318248658506012e-05,
"loss": 0.3606,
"step": 425
},
{
"epoch": 1.7246963562753037,
"grad_norm": 0.39642167255055605,
"learning_rate": 3.313990316514069e-05,
"loss": 0.353,
"step": 426
},
{
"epoch": 1.728744939271255,
"grad_norm": 0.37936484879457816,
"learning_rate": 3.309721467890571e-05,
"loss": 0.3522,
"step": 427
},
{
"epoch": 1.7327935222672064,
"grad_norm": 0.3366328377286582,
"learning_rate": 3.305442146769115e-05,
"loss": 0.3754,
"step": 428
},
{
"epoch": 1.736842105263158,
"grad_norm": 0.43115613142218145,
"learning_rate": 3.30115238736704e-05,
"loss": 0.368,
"step": 429
},
{
"epoch": 1.7408906882591093,
"grad_norm": 0.35432205136420264,
"learning_rate": 3.296852223985148e-05,
"loss": 0.3583,
"step": 430
},
{
"epoch": 1.7449392712550607,
"grad_norm": 0.33963772180874735,
"learning_rate": 3.2925416910074315e-05,
"loss": 0.3587,
"step": 431
},
{
"epoch": 1.7489878542510122,
"grad_norm": 0.36659005255282834,
"learning_rate": 3.288220822900796e-05,
"loss": 0.3572,
"step": 432
},
{
"epoch": 1.7530364372469636,
"grad_norm": 0.31666126316808924,
"learning_rate": 3.283889654214788e-05,
"loss": 0.3496,
"step": 433
},
{
"epoch": 1.757085020242915,
"grad_norm": 0.3882164127896196,
"learning_rate": 3.279548219581317e-05,
"loss": 0.3603,
"step": 434
},
{
"epoch": 1.7611336032388665,
"grad_norm": 0.3838887878030335,
"learning_rate": 3.275196553714379e-05,
"loss": 0.3718,
"step": 435
},
{
"epoch": 1.7651821862348178,
"grad_norm": 0.33249303645181805,
"learning_rate": 3.270834691409778e-05,
"loss": 0.3709,
"step": 436
},
{
"epoch": 1.7692307692307692,
"grad_norm": 0.3734363629880116,
"learning_rate": 3.266462667544849e-05,
"loss": 0.3456,
"step": 437
},
{
"epoch": 1.7732793522267207,
"grad_norm": 0.3535794776438341,
"learning_rate": 3.262080517078178e-05,
"loss": 0.3718,
"step": 438
},
{
"epoch": 1.777327935222672,
"grad_norm": 0.38749688030808344,
"learning_rate": 3.257688275049323e-05,
"loss": 0.3565,
"step": 439
},
{
"epoch": 1.7813765182186234,
"grad_norm": 0.30209892119253656,
"learning_rate": 3.253285976578535e-05,
"loss": 0.3627,
"step": 440
},
{
"epoch": 1.785425101214575,
"grad_norm": 0.4558606317362678,
"learning_rate": 3.2488736568664756e-05,
"loss": 0.361,
"step": 441
},
{
"epoch": 1.7894736842105263,
"grad_norm": 0.30447813284923186,
"learning_rate": 3.244451351193935e-05,
"loss": 0.3621,
"step": 442
},
{
"epoch": 1.7935222672064777,
"grad_norm": 0.4183063296985204,
"learning_rate": 3.240019094921551e-05,
"loss": 0.3576,
"step": 443
},
{
"epoch": 1.7975708502024292,
"grad_norm": 0.30480230660627394,
"learning_rate": 3.235576923489527e-05,
"loss": 0.3567,
"step": 444
},
{
"epoch": 1.8016194331983806,
"grad_norm": 0.33450645029758186,
"learning_rate": 3.2311248724173476e-05,
"loss": 0.3505,
"step": 445
},
{
"epoch": 1.805668016194332,
"grad_norm": 0.2987102396034446,
"learning_rate": 3.226662977303494e-05,
"loss": 0.3724,
"step": 446
},
{
"epoch": 1.8097165991902835,
"grad_norm": 0.31748487233347455,
"learning_rate": 3.2221912738251626e-05,
"loss": 0.3689,
"step": 447
},
{
"epoch": 1.8137651821862348,
"grad_norm": 0.3181711995281995,
"learning_rate": 3.2177097977379736e-05,
"loss": 0.3637,
"step": 448
},
{
"epoch": 1.8178137651821862,
"grad_norm": 0.35489963049951434,
"learning_rate": 3.21321858487569e-05,
"loss": 0.3586,
"step": 449
},
{
"epoch": 1.8218623481781377,
"grad_norm": 0.32129316731578844,
"learning_rate": 3.208717671149931e-05,
"loss": 0.3534,
"step": 450
},
{
"epoch": 1.825910931174089,
"grad_norm": 0.33610405373861246,
"learning_rate": 3.204207092549882e-05,
"loss": 0.3635,
"step": 451
},
{
"epoch": 1.8299595141700404,
"grad_norm": 17.193823813869567,
"learning_rate": 3.19968688514201e-05,
"loss": 0.3819,
"step": 452
},
{
"epoch": 1.834008097165992,
"grad_norm": 0.5121298428339934,
"learning_rate": 3.195157085069774e-05,
"loss": 0.3613,
"step": 453
},
{
"epoch": 1.8380566801619433,
"grad_norm": 0.42624512107379886,
"learning_rate": 3.190617728553332e-05,
"loss": 0.3731,
"step": 454
},
{
"epoch": 1.8421052631578947,
"grad_norm": 0.34731071666972746,
"learning_rate": 3.1860688518892606e-05,
"loss": 0.3496,
"step": 455
},
{
"epoch": 1.8461538461538463,
"grad_norm": 0.46753407315324313,
"learning_rate": 3.181510491450254e-05,
"loss": 0.3593,
"step": 456
},
{
"epoch": 1.8502024291497976,
"grad_norm": 0.4431533119356013,
"learning_rate": 3.176942683684842e-05,
"loss": 0.3512,
"step": 457
},
{
"epoch": 1.854251012145749,
"grad_norm": 0.4313772496601665,
"learning_rate": 3.1723654651170934e-05,
"loss": 0.3722,
"step": 458
},
{
"epoch": 1.8582995951417005,
"grad_norm": 0.4961260695963409,
"learning_rate": 3.167778872346327e-05,
"loss": 0.3565,
"step": 459
},
{
"epoch": 1.8623481781376519,
"grad_norm": 0.45561731995467103,
"learning_rate": 3.163182942046815e-05,
"loss": 0.3599,
"step": 460
},
{
"epoch": 1.8663967611336032,
"grad_norm": 0.3777683167794588,
"learning_rate": 3.158577710967494e-05,
"loss": 0.3743,
"step": 461
},
{
"epoch": 1.8704453441295548,
"grad_norm": 0.4624143395579991,
"learning_rate": 3.153963215931669e-05,
"loss": 0.3594,
"step": 462
},
{
"epoch": 1.874493927125506,
"grad_norm": 0.33098613789715503,
"learning_rate": 3.14933949383672e-05,
"loss": 0.3524,
"step": 463
},
{
"epoch": 1.8785425101214575,
"grad_norm": 0.33665217760193067,
"learning_rate": 3.144706581653805e-05,
"loss": 0.3418,
"step": 464
},
{
"epoch": 1.882591093117409,
"grad_norm": 0.416530596012651,
"learning_rate": 3.140064516427566e-05,
"loss": 0.3553,
"step": 465
},
{
"epoch": 1.8866396761133604,
"grad_norm": 0.3406700401661233,
"learning_rate": 3.135413335275833e-05,
"loss": 0.3677,
"step": 466
},
{
"epoch": 1.8906882591093117,
"grad_norm": 0.40660983497369724,
"learning_rate": 3.130753075389327e-05,
"loss": 0.3573,
"step": 467
},
{
"epoch": 1.8947368421052633,
"grad_norm": 0.3346955892723319,
"learning_rate": 3.1260837740313625e-05,
"loss": 0.3575,
"step": 468
},
{
"epoch": 1.8987854251012146,
"grad_norm": 0.3701645715897484,
"learning_rate": 3.121405468537547e-05,
"loss": 0.3744,
"step": 469
},
{
"epoch": 1.902834008097166,
"grad_norm": 0.371240090603289,
"learning_rate": 3.116718196315488e-05,
"loss": 0.3612,
"step": 470
},
{
"epoch": 1.9068825910931175,
"grad_norm": 0.40230262232452685,
"learning_rate": 3.1120219948444894e-05,
"loss": 0.3625,
"step": 471
},
{
"epoch": 1.9109311740890689,
"grad_norm": 0.29983340524751256,
"learning_rate": 3.107316901675252e-05,
"loss": 0.3602,
"step": 472
},
{
"epoch": 1.9149797570850202,
"grad_norm": 0.3233481778846084,
"learning_rate": 3.1026029544295745e-05,
"loss": 0.3635,
"step": 473
},
{
"epoch": 1.9190283400809718,
"grad_norm": 0.2797945807336666,
"learning_rate": 3.097880190800057e-05,
"loss": 0.3671,
"step": 474
},
{
"epoch": 1.9230769230769231,
"grad_norm": 0.3304943728104111,
"learning_rate": 3.093148648549788e-05,
"loss": 0.3545,
"step": 475
},
{
"epoch": 1.9271255060728745,
"grad_norm": 0.29820433460985235,
"learning_rate": 3.088408365512055e-05,
"loss": 0.3465,
"step": 476
},
{
"epoch": 1.931174089068826,
"grad_norm": 0.3224538705420348,
"learning_rate": 3.083659379590034e-05,
"loss": 0.3717,
"step": 477
},
{
"epoch": 1.9352226720647774,
"grad_norm": 0.3334732430203597,
"learning_rate": 3.0789017287564913e-05,
"loss": 0.3572,
"step": 478
},
{
"epoch": 1.9392712550607287,
"grad_norm": 0.28550908462585106,
"learning_rate": 3.074135451053475e-05,
"loss": 0.3542,
"step": 479
},
{
"epoch": 1.9433198380566803,
"grad_norm": 0.31258575120073806,
"learning_rate": 3.069360584592015e-05,
"loss": 0.367,
"step": 480
},
{
"epoch": 1.9473684210526314,
"grad_norm": 0.3541585789700986,
"learning_rate": 3.064577167551816e-05,
"loss": 0.3645,
"step": 481
},
{
"epoch": 1.951417004048583,
"grad_norm": 0.29255303289781,
"learning_rate": 3.059785238180953e-05,
"loss": 0.349,
"step": 482
},
{
"epoch": 1.9554655870445345,
"grad_norm": 0.3578815952794331,
"learning_rate": 3.054984834795566e-05,
"loss": 0.3596,
"step": 483
},
{
"epoch": 1.9595141700404857,
"grad_norm": 0.29987812993473534,
"learning_rate": 3.0501759957795537e-05,
"loss": 0.3565,
"step": 484
},
{
"epoch": 1.9635627530364372,
"grad_norm": 0.31629292817323273,
"learning_rate": 3.0453587595842627e-05,
"loss": 0.3615,
"step": 485
},
{
"epoch": 1.9676113360323888,
"grad_norm": 0.341391801910758,
"learning_rate": 3.0405331647281856e-05,
"loss": 0.3572,
"step": 486
},
{
"epoch": 1.97165991902834,
"grad_norm": 0.29515463758915894,
"learning_rate": 3.0356992497966508e-05,
"loss": 0.3415,
"step": 487
},
{
"epoch": 1.9757085020242915,
"grad_norm": 0.40192553035614015,
"learning_rate": 3.0308570534415137e-05,
"loss": 0.3508,
"step": 488
},
{
"epoch": 1.979757085020243,
"grad_norm": 0.31374235575267323,
"learning_rate": 3.0260066143808467e-05,
"loss": 0.3501,
"step": 489
},
{
"epoch": 1.9838056680161942,
"grad_norm": 0.38613792627714266,
"learning_rate": 3.021147971398632e-05,
"loss": 0.3453,
"step": 490
},
{
"epoch": 1.9878542510121457,
"grad_norm": 0.3579175630612423,
"learning_rate": 3.0162811633444478e-05,
"loss": 0.3575,
"step": 491
},
{
"epoch": 1.9919028340080973,
"grad_norm": 0.3658549919918361,
"learning_rate": 3.0114062291331635e-05,
"loss": 0.3562,
"step": 492
},
{
"epoch": 1.9959514170040484,
"grad_norm": 0.33111221051814815,
"learning_rate": 3.0065232077446217e-05,
"loss": 0.36,
"step": 493
},
{
"epoch": 2.0,
"grad_norm": 0.361343713958839,
"learning_rate": 3.001632138223332e-05,
"loss": 0.3525,
"step": 494
},
{
"epoch": 2.0040485829959516,
"grad_norm": 0.5063595462838176,
"learning_rate": 2.9967330596781565e-05,
"loss": 0.3091,
"step": 495
},
{
"epoch": 2.0080971659919027,
"grad_norm": 0.6322236157401097,
"learning_rate": 2.9918260112819952e-05,
"loss": 0.3019,
"step": 496
},
{
"epoch": 2.0121457489878543,
"grad_norm": 0.5397548141645832,
"learning_rate": 2.9869110322714778e-05,
"loss": 0.2933,
"step": 497
},
{
"epoch": 2.016194331983806,
"grad_norm": 0.35807232051534543,
"learning_rate": 2.9819881619466447e-05,
"loss": 0.3032,
"step": 498
},
{
"epoch": 2.020242914979757,
"grad_norm": 0.48785018282229575,
"learning_rate": 2.9770574396706364e-05,
"loss": 0.2828,
"step": 499
},
{
"epoch": 2.0242914979757085,
"grad_norm": 0.3609134345123077,
"learning_rate": 2.972118904869376e-05,
"loss": 0.2801,
"step": 500
},
{
"epoch": 2.02834008097166,
"grad_norm": 0.4098214069480718,
"learning_rate": 2.9671725970312563e-05,
"loss": 0.2878,
"step": 501
},
{
"epoch": 2.032388663967611,
"grad_norm": 0.38117372527998017,
"learning_rate": 2.9622185557068238e-05,
"loss": 0.2932,
"step": 502
},
{
"epoch": 2.0364372469635628,
"grad_norm": 0.3481944514740851,
"learning_rate": 2.95725682050846e-05,
"loss": 0.2985,
"step": 503
},
{
"epoch": 2.0404858299595143,
"grad_norm": 0.39866999738874637,
"learning_rate": 2.9522874311100684e-05,
"loss": 0.2957,
"step": 504
},
{
"epoch": 2.0445344129554655,
"grad_norm": 0.38812221984231027,
"learning_rate": 2.9473104272467526e-05,
"loss": 0.2866,
"step": 505
},
{
"epoch": 2.048582995951417,
"grad_norm": 0.3609909974483178,
"learning_rate": 2.942325848714505e-05,
"loss": 0.2965,
"step": 506
},
{
"epoch": 2.0526315789473686,
"grad_norm": 0.3686625856506501,
"learning_rate": 2.9373337353698804e-05,
"loss": 0.2934,
"step": 507
},
{
"epoch": 2.0566801619433197,
"grad_norm": 0.31073276368936187,
"learning_rate": 2.932334127129686e-05,
"loss": 0.2887,
"step": 508
},
{
"epoch": 2.0607287449392713,
"grad_norm": 0.41303342022980993,
"learning_rate": 2.9273270639706547e-05,
"loss": 0.3006,
"step": 509
},
{
"epoch": 2.064777327935223,
"grad_norm": 0.36867081583356687,
"learning_rate": 2.922312585929131e-05,
"loss": 0.2956,
"step": 510
},
{
"epoch": 2.068825910931174,
"grad_norm": 0.30485742402433963,
"learning_rate": 2.9172907331007478e-05,
"loss": 0.2948,
"step": 511
},
{
"epoch": 2.0728744939271255,
"grad_norm": 0.4033815823998427,
"learning_rate": 2.9122615456401053e-05,
"loss": 0.3012,
"step": 512
},
{
"epoch": 2.076923076923077,
"grad_norm": 0.326496406467837,
"learning_rate": 2.9072250637604535e-05,
"loss": 0.2943,
"step": 513
},
{
"epoch": 2.080971659919028,
"grad_norm": 0.37694270529839924,
"learning_rate": 2.9021813277333667e-05,
"loss": 0.2945,
"step": 514
},
{
"epoch": 2.08502024291498,
"grad_norm": 0.3418615909273026,
"learning_rate": 2.8971303778884244e-05,
"loss": 0.2894,
"step": 515
},
{
"epoch": 2.0890688259109313,
"grad_norm": 0.3678732315123243,
"learning_rate": 2.8920722546128863e-05,
"loss": 0.2924,
"step": 516
},
{
"epoch": 2.0931174089068825,
"grad_norm": 0.39053552371517963,
"learning_rate": 2.887006998351372e-05,
"loss": 0.2981,
"step": 517
},
{
"epoch": 2.097165991902834,
"grad_norm": 0.3097232003150849,
"learning_rate": 2.881934649605535e-05,
"loss": 0.2985,
"step": 518
},
{
"epoch": 2.1012145748987856,
"grad_norm": 0.43468461630625504,
"learning_rate": 2.8768552489337408e-05,
"loss": 0.293,
"step": 519
},
{
"epoch": 2.1052631578947367,
"grad_norm": 0.4729526741919419,
"learning_rate": 2.8717688369507425e-05,
"loss": 0.2991,
"step": 520
},
{
"epoch": 2.1093117408906883,
"grad_norm": 0.3215648758937648,
"learning_rate": 2.8666754543273535e-05,
"loss": 0.2963,
"step": 521
},
{
"epoch": 2.11336032388664,
"grad_norm": 0.3842501386200612,
"learning_rate": 2.8615751417901273e-05,
"loss": 0.2965,
"step": 522
},
{
"epoch": 2.117408906882591,
"grad_norm": 0.33267662709156837,
"learning_rate": 2.856467940121025e-05,
"loss": 0.3053,
"step": 523
},
{
"epoch": 2.1214574898785425,
"grad_norm": 0.39884625485847414,
"learning_rate": 2.8513538901570964e-05,
"loss": 0.2908,
"step": 524
},
{
"epoch": 2.125506072874494,
"grad_norm": 0.36075314052236956,
"learning_rate": 2.846233032790149e-05,
"loss": 0.3047,
"step": 525
},
{
"epoch": 2.1295546558704452,
"grad_norm": 0.3843825947107099,
"learning_rate": 2.84110540896642e-05,
"loss": 0.2949,
"step": 526
},
{
"epoch": 2.133603238866397,
"grad_norm": 0.32532735407211644,
"learning_rate": 2.8359710596862555e-05,
"loss": 0.2985,
"step": 527
},
{
"epoch": 2.1376518218623484,
"grad_norm": 0.35471188549779753,
"learning_rate": 2.8308300260037734e-05,
"loss": 0.2976,
"step": 528
},
{
"epoch": 2.1417004048582995,
"grad_norm": 0.33431268578109025,
"learning_rate": 2.8256823490265424e-05,
"loss": 0.2911,
"step": 529
},
{
"epoch": 2.145748987854251,
"grad_norm": 0.37270185257924393,
"learning_rate": 2.8205280699152515e-05,
"loss": 0.3016,
"step": 530
},
{
"epoch": 2.1497975708502026,
"grad_norm": 0.4002733380701121,
"learning_rate": 2.8153672298833777e-05,
"loss": 0.3085,
"step": 531
},
{
"epoch": 2.1538461538461537,
"grad_norm": 0.33004321493038824,
"learning_rate": 2.8101998701968615e-05,
"loss": 0.2959,
"step": 532
},
{
"epoch": 2.1578947368421053,
"grad_norm": 0.44774484764321665,
"learning_rate": 2.805026032173773e-05,
"loss": 0.3036,
"step": 533
},
{
"epoch": 2.161943319838057,
"grad_norm": 0.3024614519447385,
"learning_rate": 2.7998457571839832e-05,
"loss": 0.3013,
"step": 534
},
{
"epoch": 2.165991902834008,
"grad_norm": 0.44336240640919405,
"learning_rate": 2.7946590866488322e-05,
"loss": 0.2935,
"step": 535
},
{
"epoch": 2.1700404858299596,
"grad_norm": 0.33289712016630013,
"learning_rate": 2.7894660620408e-05,
"loss": 0.2854,
"step": 536
},
{
"epoch": 2.174089068825911,
"grad_norm": 0.35745645546558374,
"learning_rate": 2.784266724883173e-05,
"loss": 0.3004,
"step": 537
},
{
"epoch": 2.1781376518218623,
"grad_norm": 0.34234179639061185,
"learning_rate": 2.7790611167497128e-05,
"loss": 0.3068,
"step": 538
},
{
"epoch": 2.182186234817814,
"grad_norm": 0.3669441506194541,
"learning_rate": 2.7738492792643224e-05,
"loss": 0.2896,
"step": 539
},
{
"epoch": 2.1862348178137654,
"grad_norm": 0.27909765640303874,
"learning_rate": 2.768631254100716e-05,
"loss": 0.2924,
"step": 540
},
{
"epoch": 2.1902834008097165,
"grad_norm": 0.3534259000410571,
"learning_rate": 2.7634070829820836e-05,
"loss": 0.283,
"step": 541
},
{
"epoch": 2.194331983805668,
"grad_norm": 0.32776122538637437,
"learning_rate": 2.7581768076807592e-05,
"loss": 0.289,
"step": 542
},
{
"epoch": 2.1983805668016196,
"grad_norm": 0.30697024451534277,
"learning_rate": 2.7529404700178832e-05,
"loss": 0.2911,
"step": 543
},
{
"epoch": 2.2024291497975708,
"grad_norm": 0.3734861954455827,
"learning_rate": 2.7476981118630735e-05,
"loss": 0.2881,
"step": 544
},
{
"epoch": 2.2064777327935223,
"grad_norm": 0.2973561422337816,
"learning_rate": 2.742449775134085e-05,
"loss": 0.2931,
"step": 545
},
{
"epoch": 2.2105263157894735,
"grad_norm": 0.2997900099659482,
"learning_rate": 2.7371955017964788e-05,
"loss": 0.2916,
"step": 546
},
{
"epoch": 2.214574898785425,
"grad_norm": 0.31192011672576264,
"learning_rate": 2.7319353338632845e-05,
"loss": 0.3021,
"step": 547
},
{
"epoch": 2.2186234817813766,
"grad_norm": 0.28253635154725926,
"learning_rate": 2.7266693133946645e-05,
"loss": 0.2858,
"step": 548
},
{
"epoch": 2.2226720647773277,
"grad_norm": 0.279843716109016,
"learning_rate": 2.7213974824975785e-05,
"loss": 0.2951,
"step": 549
},
{
"epoch": 2.2267206477732793,
"grad_norm": 0.261057166000334,
"learning_rate": 2.7161198833254464e-05,
"loss": 0.2799,
"step": 550
},
{
"epoch": 2.230769230769231,
"grad_norm": 0.27293910229556634,
"learning_rate": 2.7108365580778095e-05,
"loss": 0.3081,
"step": 551
},
{
"epoch": 2.234817813765182,
"grad_norm": 0.27635695779793334,
"learning_rate": 2.705547548999996e-05,
"loss": 0.3043,
"step": 552
},
{
"epoch": 2.2388663967611335,
"grad_norm": 0.281941148817239,
"learning_rate": 2.7002528983827817e-05,
"loss": 0.2852,
"step": 553
},
{
"epoch": 2.242914979757085,
"grad_norm": 0.3543707878034661,
"learning_rate": 2.6949526485620517e-05,
"loss": 0.2992,
"step": 554
},
{
"epoch": 2.246963562753036,
"grad_norm": 0.2762683740081704,
"learning_rate": 2.6896468419184623e-05,
"loss": 0.2954,
"step": 555
},
{
"epoch": 2.251012145748988,
"grad_norm": 0.3434293043785927,
"learning_rate": 2.6843355208771024e-05,
"loss": 0.302,
"step": 556
},
{
"epoch": 2.2550607287449393,
"grad_norm": 0.3252059128816185,
"learning_rate": 2.6790187279071536e-05,
"loss": 0.2948,
"step": 557
},
{
"epoch": 2.2591093117408905,
"grad_norm": 0.27696157981383585,
"learning_rate": 2.67369650552155e-05,
"loss": 0.2958,
"step": 558
},
{
"epoch": 2.263157894736842,
"grad_norm": 0.32966913660827396,
"learning_rate": 2.6683688962766407e-05,
"loss": 0.3032,
"step": 559
},
{
"epoch": 2.2672064777327936,
"grad_norm": 0.308747748121563,
"learning_rate": 2.6630359427718468e-05,
"loss": 0.298,
"step": 560
},
{
"epoch": 2.2712550607287447,
"grad_norm": 0.28672475046204055,
"learning_rate": 2.6576976876493225e-05,
"loss": 0.2961,
"step": 561
},
{
"epoch": 2.2753036437246963,
"grad_norm": 0.32341375754603147,
"learning_rate": 2.652354173593614e-05,
"loss": 0.2878,
"step": 562
},
{
"epoch": 2.279352226720648,
"grad_norm": 0.2760979230001483,
"learning_rate": 2.647005443331317e-05,
"loss": 0.3013,
"step": 563
},
{
"epoch": 2.283400809716599,
"grad_norm": 0.3267093523694099,
"learning_rate": 2.6416515396307354e-05,
"loss": 0.293,
"step": 564
},
{
"epoch": 2.2874493927125505,
"grad_norm": 0.2600199347770836,
"learning_rate": 2.636292505301542e-05,
"loss": 0.2912,
"step": 565
},
{
"epoch": 2.291497975708502,
"grad_norm": 0.2914083665247395,
"learning_rate": 2.630928383194431e-05,
"loss": 0.2797,
"step": 566
},
{
"epoch": 2.2955465587044532,
"grad_norm": 0.30014335670628434,
"learning_rate": 2.625559216200781e-05,
"loss": 0.3014,
"step": 567
},
{
"epoch": 2.299595141700405,
"grad_norm": 0.2827384560274127,
"learning_rate": 2.620185047252308e-05,
"loss": 0.2936,
"step": 568
},
{
"epoch": 2.3036437246963564,
"grad_norm": 0.27844758158477373,
"learning_rate": 2.614805919320723e-05,
"loss": 0.2985,
"step": 569
},
{
"epoch": 2.3076923076923075,
"grad_norm": 0.31900623423156127,
"learning_rate": 2.6094218754173895e-05,
"loss": 0.3028,
"step": 570
},
{
"epoch": 2.311740890688259,
"grad_norm": 0.2827992265314502,
"learning_rate": 2.60403295859298e-05,
"loss": 0.2796,
"step": 571
},
{
"epoch": 2.3157894736842106,
"grad_norm": 0.2988266464285959,
"learning_rate": 2.5986392119371297e-05,
"loss": 0.2825,
"step": 572
},
{
"epoch": 2.3198380566801617,
"grad_norm": 0.30393163916748905,
"learning_rate": 2.5932406785780928e-05,
"loss": 0.2945,
"step": 573
},
{
"epoch": 2.3238866396761133,
"grad_norm": 0.28151710554102494,
"learning_rate": 2.5878374016823987e-05,
"loss": 0.2977,
"step": 574
},
{
"epoch": 2.327935222672065,
"grad_norm": 0.30560961243653667,
"learning_rate": 2.5824294244545045e-05,
"loss": 0.2955,
"step": 575
},
{
"epoch": 2.331983805668016,
"grad_norm": 0.27727207458377934,
"learning_rate": 2.577016790136453e-05,
"loss": 0.2882,
"step": 576
},
{
"epoch": 2.3360323886639676,
"grad_norm": 0.3120603411231914,
"learning_rate": 2.5715995420075225e-05,
"loss": 0.2999,
"step": 577
},
{
"epoch": 2.340080971659919,
"grad_norm": 0.2879710490622399,
"learning_rate": 2.5661777233838864e-05,
"loss": 0.2991,
"step": 578
},
{
"epoch": 2.3441295546558703,
"grad_norm": 0.27934458179610794,
"learning_rate": 2.5607513776182616e-05,
"loss": 0.3029,
"step": 579
},
{
"epoch": 2.348178137651822,
"grad_norm": 0.29587431636755157,
"learning_rate": 2.5553205480995633e-05,
"loss": 0.307,
"step": 580
},
{
"epoch": 2.3522267206477734,
"grad_norm": 0.2762481514570396,
"learning_rate": 2.5498852782525604e-05,
"loss": 0.2804,
"step": 581
},
{
"epoch": 2.3562753036437245,
"grad_norm": 0.2993486213719945,
"learning_rate": 2.5444456115375252e-05,
"loss": 0.2899,
"step": 582
},
{
"epoch": 2.360323886639676,
"grad_norm": 0.2690581364295306,
"learning_rate": 2.539001591449889e-05,
"loss": 0.2997,
"step": 583
},
{
"epoch": 2.3643724696356276,
"grad_norm": 0.28688693983455904,
"learning_rate": 2.53355326151989e-05,
"loss": 0.299,
"step": 584
},
{
"epoch": 2.3684210526315788,
"grad_norm": 0.26302977285152196,
"learning_rate": 2.5281006653122288e-05,
"loss": 0.2952,
"step": 585
},
{
"epoch": 2.3724696356275303,
"grad_norm": 0.30339328925426834,
"learning_rate": 2.5226438464257217e-05,
"loss": 0.2934,
"step": 586
},
{
"epoch": 2.376518218623482,
"grad_norm": 0.2529684251102793,
"learning_rate": 2.5171828484929447e-05,
"loss": 0.2984,
"step": 587
},
{
"epoch": 2.380566801619433,
"grad_norm": 0.2936709365658734,
"learning_rate": 2.5117177151798934e-05,
"loss": 0.2977,
"step": 588
},
{
"epoch": 2.3846153846153846,
"grad_norm": 0.2661081924578192,
"learning_rate": 2.506248490185627e-05,
"loss": 0.2888,
"step": 589
},
{
"epoch": 2.388663967611336,
"grad_norm": 0.297099353592018,
"learning_rate": 2.500775217241923e-05,
"loss": 0.2965,
"step": 590
},
{
"epoch": 2.3927125506072873,
"grad_norm": 0.2899533043437716,
"learning_rate": 2.4952979401129266e-05,
"loss": 0.3063,
"step": 591
},
{
"epoch": 2.396761133603239,
"grad_norm": 0.31664949413798804,
"learning_rate": 2.4898167025948e-05,
"loss": 0.2941,
"step": 592
},
{
"epoch": 2.4008097165991904,
"grad_norm": 0.2754440971407662,
"learning_rate": 2.4843315485153703e-05,
"loss": 0.2915,
"step": 593
},
{
"epoch": 2.4048582995951415,
"grad_norm": 0.35103229072942577,
"learning_rate": 2.4788425217337856e-05,
"loss": 0.2907,
"step": 594
},
{
"epoch": 2.408906882591093,
"grad_norm": 0.2948628700077619,
"learning_rate": 2.4733496661401572e-05,
"loss": 0.2862,
"step": 595
},
{
"epoch": 2.4129554655870447,
"grad_norm": 0.30809556095136115,
"learning_rate": 2.4678530256552114e-05,
"loss": 0.2887,
"step": 596
},
{
"epoch": 2.417004048582996,
"grad_norm": 0.30664734377030933,
"learning_rate": 2.46235264422994e-05,
"loss": 0.2912,
"step": 597
},
{
"epoch": 2.4210526315789473,
"grad_norm": 0.27915682560875954,
"learning_rate": 2.4568485658452452e-05,
"loss": 0.2982,
"step": 598
},
{
"epoch": 2.425101214574899,
"grad_norm": 0.36238420718711123,
"learning_rate": 2.451340834511592e-05,
"loss": 0.2998,
"step": 599
},
{
"epoch": 2.42914979757085,
"grad_norm": 0.2605353425969213,
"learning_rate": 2.4458294942686528e-05,
"loss": 0.2851,
"step": 600
},
{
"epoch": 2.4331983805668016,
"grad_norm": 0.2922866789241617,
"learning_rate": 2.440314589184958e-05,
"loss": 0.3012,
"step": 601
},
{
"epoch": 2.437246963562753,
"grad_norm": 0.2919446025257949,
"learning_rate": 2.434796163357541e-05,
"loss": 0.2991,
"step": 602
},
{
"epoch": 2.4412955465587043,
"grad_norm": 0.30278568756007623,
"learning_rate": 2.429274260911588e-05,
"loss": 0.2943,
"step": 603
},
{
"epoch": 2.445344129554656,
"grad_norm": 0.28040188313718667,
"learning_rate": 2.423748926000084e-05,
"loss": 0.2953,
"step": 604
},
{
"epoch": 2.4493927125506074,
"grad_norm": 0.2514748477584516,
"learning_rate": 2.418220202803459e-05,
"loss": 0.2922,
"step": 605
},
{
"epoch": 2.4534412955465585,
"grad_norm": 0.28160595495562907,
"learning_rate": 2.412688135529237e-05,
"loss": 0.2944,
"step": 606
},
{
"epoch": 2.45748987854251,
"grad_norm": 0.282033760166221,
"learning_rate": 2.4071527684116807e-05,
"loss": 0.2874,
"step": 607
},
{
"epoch": 2.4615384615384617,
"grad_norm": 0.2815940646438026,
"learning_rate": 2.4016141457114372e-05,
"loss": 0.2891,
"step": 608
},
{
"epoch": 2.465587044534413,
"grad_norm": 0.32429473938321873,
"learning_rate": 2.396072311715187e-05,
"loss": 0.2943,
"step": 609
},
{
"epoch": 2.4696356275303644,
"grad_norm": 0.2735262460222353,
"learning_rate": 2.3905273107352863e-05,
"loss": 0.2829,
"step": 610
},
{
"epoch": 2.473684210526316,
"grad_norm": 0.30606804021842793,
"learning_rate": 2.3849791871094157e-05,
"loss": 0.2944,
"step": 611
},
{
"epoch": 2.477732793522267,
"grad_norm": 0.2560076191635686,
"learning_rate": 2.379427985200223e-05,
"loss": 0.2879,
"step": 612
},
{
"epoch": 2.4817813765182186,
"grad_norm": 0.27238803791967564,
"learning_rate": 2.373873749394972e-05,
"loss": 0.2939,
"step": 613
},
{
"epoch": 2.48582995951417,
"grad_norm": 0.26613065344860254,
"learning_rate": 2.3683165241051836e-05,
"loss": 0.2902,
"step": 614
},
{
"epoch": 2.4898785425101213,
"grad_norm": 0.2545414066596272,
"learning_rate": 2.3627563537662823e-05,
"loss": 0.2965,
"step": 615
},
{
"epoch": 2.493927125506073,
"grad_norm": 0.31910206291798804,
"learning_rate": 2.3571932828372433e-05,
"loss": 0.2966,
"step": 616
},
{
"epoch": 2.4979757085020244,
"grad_norm": 0.29773127037260005,
"learning_rate": 2.351627355800233e-05,
"loss": 0.2928,
"step": 617
},
{
"epoch": 2.5020242914979756,
"grad_norm": 0.3090919014981065,
"learning_rate": 2.346058617160256e-05,
"loss": 0.2836,
"step": 618
},
{
"epoch": 2.506072874493927,
"grad_norm": 0.2773091164747954,
"learning_rate": 2.3404871114447976e-05,
"loss": 0.2967,
"step": 619
},
{
"epoch": 2.5101214574898787,
"grad_norm": 0.31718839098411294,
"learning_rate": 2.334912883203469e-05,
"loss": 0.2884,
"step": 620
},
{
"epoch": 2.51417004048583,
"grad_norm": 0.3023170102437494,
"learning_rate": 2.3293359770076505e-05,
"loss": 0.2876,
"step": 621
},
{
"epoch": 2.5182186234817814,
"grad_norm": 0.3093961498054293,
"learning_rate": 2.323756437450136e-05,
"loss": 0.2961,
"step": 622
},
{
"epoch": 2.522267206477733,
"grad_norm": 0.2651112257245419,
"learning_rate": 2.3181743091447754e-05,
"loss": 0.2915,
"step": 623
},
{
"epoch": 2.526315789473684,
"grad_norm": 0.3424267625419925,
"learning_rate": 2.3125896367261164e-05,
"loss": 0.2885,
"step": 624
},
{
"epoch": 2.5303643724696356,
"grad_norm": 0.2804435188048513,
"learning_rate": 2.307002464849052e-05,
"loss": 0.2782,
"step": 625
},
{
"epoch": 2.534412955465587,
"grad_norm": 0.30270437358315844,
"learning_rate": 2.30141283818846e-05,
"loss": 0.3074,
"step": 626
},
{
"epoch": 2.5384615384615383,
"grad_norm": 0.2749859350029136,
"learning_rate": 2.2958208014388462e-05,
"loss": 0.2938,
"step": 627
},
{
"epoch": 2.54251012145749,
"grad_norm": 0.2704497883241557,
"learning_rate": 2.2902263993139857e-05,
"loss": 0.2738,
"step": 628
},
{
"epoch": 2.5465587044534415,
"grad_norm": 0.251685885539824,
"learning_rate": 2.2846296765465708e-05,
"loss": 0.2976,
"step": 629
},
{
"epoch": 2.5506072874493926,
"grad_norm": 0.2779406247032837,
"learning_rate": 2.2790306778878465e-05,
"loss": 0.2859,
"step": 630
},
{
"epoch": 2.554655870445344,
"grad_norm": 0.24137264145782744,
"learning_rate": 2.273429448107257e-05,
"loss": 0.2773,
"step": 631
},
{
"epoch": 2.5587044534412957,
"grad_norm": 0.2816307385674295,
"learning_rate": 2.2678260319920868e-05,
"loss": 0.2921,
"step": 632
},
{
"epoch": 2.562753036437247,
"grad_norm": 0.24812021829818853,
"learning_rate": 2.262220474347102e-05,
"loss": 0.2935,
"step": 633
},
{
"epoch": 2.5668016194331984,
"grad_norm": 0.2905463850538257,
"learning_rate": 2.2566128199941918e-05,
"loss": 0.2821,
"step": 634
},
{
"epoch": 2.57085020242915,
"grad_norm": 0.2653271007552616,
"learning_rate": 2.2510031137720104e-05,
"loss": 0.2885,
"step": 635
},
{
"epoch": 2.574898785425101,
"grad_norm": 0.2909360205696752,
"learning_rate": 2.2453914005356206e-05,
"loss": 0.3045,
"step": 636
},
{
"epoch": 2.5789473684210527,
"grad_norm": 0.2759853981450979,
"learning_rate": 2.2397777251561302e-05,
"loss": 0.2893,
"step": 637
},
{
"epoch": 2.582995951417004,
"grad_norm": 0.2921669481920817,
"learning_rate": 2.2341621325203393e-05,
"loss": 0.2865,
"step": 638
},
{
"epoch": 2.5870445344129553,
"grad_norm": 0.3501399786885425,
"learning_rate": 2.2285446675303764e-05,
"loss": 0.2977,
"step": 639
},
{
"epoch": 2.591093117408907,
"grad_norm": 0.33208908001047993,
"learning_rate": 2.2229253751033423e-05,
"loss": 0.2872,
"step": 640
},
{
"epoch": 2.5951417004048585,
"grad_norm": 0.26664453566017393,
"learning_rate": 2.2173043001709492e-05,
"loss": 0.3079,
"step": 641
},
{
"epoch": 2.5991902834008096,
"grad_norm": 0.27835999899857483,
"learning_rate": 2.2116814876791627e-05,
"loss": 0.283,
"step": 642
},
{
"epoch": 2.603238866396761,
"grad_norm": 0.2729892689837168,
"learning_rate": 2.2060569825878413e-05,
"loss": 0.2979,
"step": 643
},
{
"epoch": 2.6072874493927127,
"grad_norm": 0.23080580992219152,
"learning_rate": 2.2004308298703777e-05,
"loss": 0.287,
"step": 644
},
{
"epoch": 2.611336032388664,
"grad_norm": 0.2610148507176073,
"learning_rate": 2.1948030745133396e-05,
"loss": 0.2934,
"step": 645
},
{
"epoch": 2.6153846153846154,
"grad_norm": 0.25825912759700603,
"learning_rate": 2.1891737615161083e-05,
"loss": 0.282,
"step": 646
},
{
"epoch": 2.619433198380567,
"grad_norm": 0.2706791332277751,
"learning_rate": 2.1835429358905218e-05,
"loss": 0.2952,
"step": 647
},
{
"epoch": 2.623481781376518,
"grad_norm": 0.2704644278780107,
"learning_rate": 2.17791064266051e-05,
"loss": 0.2948,
"step": 648
},
{
"epoch": 2.6275303643724697,
"grad_norm": 0.24062040902223414,
"learning_rate": 2.172276926861741e-05,
"loss": 0.3048,
"step": 649
},
{
"epoch": 2.6315789473684212,
"grad_norm": 0.2727919221723851,
"learning_rate": 2.166641833541255e-05,
"loss": 0.286,
"step": 650
},
{
"epoch": 2.6356275303643724,
"grad_norm": 0.24399546228363034,
"learning_rate": 2.161005407757109e-05,
"loss": 0.2873,
"step": 651
},
{
"epoch": 2.639676113360324,
"grad_norm": 0.2931312981647137,
"learning_rate": 2.155367694578013e-05,
"loss": 0.2951,
"step": 652
},
{
"epoch": 2.6437246963562755,
"grad_norm": 0.2842202879438101,
"learning_rate": 2.1497287390829712e-05,
"loss": 0.2925,
"step": 653
},
{
"epoch": 2.6477732793522266,
"grad_norm": 0.2882325939100774,
"learning_rate": 2.1440885863609208e-05,
"loss": 0.305,
"step": 654
},
{
"epoch": 2.651821862348178,
"grad_norm": 0.30495691988249674,
"learning_rate": 2.1384472815103737e-05,
"loss": 0.2915,
"step": 655
},
{
"epoch": 2.6558704453441297,
"grad_norm": 0.296526516744605,
"learning_rate": 2.1328048696390534e-05,
"loss": 0.2926,
"step": 656
},
{
"epoch": 2.659919028340081,
"grad_norm": 0.32932824486179146,
"learning_rate": 2.1271613958635342e-05,
"loss": 0.3141,
"step": 657
},
{
"epoch": 2.6639676113360324,
"grad_norm": 0.3227121923433077,
"learning_rate": 2.1215169053088825e-05,
"loss": 0.2868,
"step": 658
},
{
"epoch": 2.668016194331984,
"grad_norm": 0.295160311659063,
"learning_rate": 2.1158714431082937e-05,
"loss": 0.2868,
"step": 659
},
{
"epoch": 2.672064777327935,
"grad_norm": 0.3073323230041145,
"learning_rate": 2.110225054402733e-05,
"loss": 0.2936,
"step": 660
},
{
"epoch": 2.6761133603238867,
"grad_norm": 0.30004691061369343,
"learning_rate": 2.1045777843405747e-05,
"loss": 0.2973,
"step": 661
},
{
"epoch": 2.6801619433198383,
"grad_norm": 0.3152957981796781,
"learning_rate": 2.09892967807724e-05,
"loss": 0.3077,
"step": 662
},
{
"epoch": 2.6842105263157894,
"grad_norm": 0.24709516791681055,
"learning_rate": 2.093280780774835e-05,
"loss": 0.2858,
"step": 663
},
{
"epoch": 2.688259109311741,
"grad_norm": 0.31057432841028104,
"learning_rate": 2.087631137601793e-05,
"loss": 0.2881,
"step": 664
},
{
"epoch": 2.6923076923076925,
"grad_norm": 0.2904944382850072,
"learning_rate": 2.0819807937325104e-05,
"loss": 0.2984,
"step": 665
},
{
"epoch": 2.6963562753036436,
"grad_norm": 0.2962421378120808,
"learning_rate": 2.0763297943469847e-05,
"loss": 0.2825,
"step": 666
},
{
"epoch": 2.700404858299595,
"grad_norm": 0.2941359270812712,
"learning_rate": 2.070678184630458e-05,
"loss": 0.2785,
"step": 667
},
{
"epoch": 2.7044534412955468,
"grad_norm": 0.36163315540213736,
"learning_rate": 2.06502600977305e-05,
"loss": 0.297,
"step": 668
},
{
"epoch": 2.708502024291498,
"grad_norm": 0.302599111484528,
"learning_rate": 2.0593733149694008e-05,
"loss": 0.2932,
"step": 669
},
{
"epoch": 2.7125506072874495,
"grad_norm": 0.30029726812304813,
"learning_rate": 2.0537201454183074e-05,
"loss": 0.283,
"step": 670
},
{
"epoch": 2.716599190283401,
"grad_norm": 0.28610204771170805,
"learning_rate": 2.0480665463223626e-05,
"loss": 0.2929,
"step": 671
},
{
"epoch": 2.720647773279352,
"grad_norm": 0.2830418219000095,
"learning_rate": 2.042412562887594e-05,
"loss": 0.2878,
"step": 672
},
{
"epoch": 2.7246963562753037,
"grad_norm": 0.24892570356830707,
"learning_rate": 2.0367582403231033e-05,
"loss": 0.3028,
"step": 673
},
{
"epoch": 2.7287449392712553,
"grad_norm": 0.282040743849998,
"learning_rate": 2.0311036238407023e-05,
"loss": 0.2966,
"step": 674
},
{
"epoch": 2.7327935222672064,
"grad_norm": 0.23735615898529744,
"learning_rate": 2.0254487586545534e-05,
"loss": 0.286,
"step": 675
},
{
"epoch": 2.736842105263158,
"grad_norm": 0.2934208056077396,
"learning_rate": 2.0197936899808088e-05,
"loss": 0.3001,
"step": 676
},
{
"epoch": 2.7408906882591095,
"grad_norm": 0.2563243117875018,
"learning_rate": 2.0141384630372466e-05,
"loss": 0.2976,
"step": 677
},
{
"epoch": 2.7449392712550607,
"grad_norm": 0.27375494183364696,
"learning_rate": 2.0084831230429106e-05,
"loss": 0.2941,
"step": 678
},
{
"epoch": 2.748987854251012,
"grad_norm": 0.27501548381923924,
"learning_rate": 2.0028277152177492e-05,
"loss": 0.2806,
"step": 679
},
{
"epoch": 2.753036437246964,
"grad_norm": 0.2646496327509352,
"learning_rate": 1.9971722847822518e-05,
"loss": 0.302,
"step": 680
},
{
"epoch": 2.757085020242915,
"grad_norm": 0.28308000389420834,
"learning_rate": 1.99151687695709e-05,
"loss": 0.2927,
"step": 681
},
{
"epoch": 2.7611336032388665,
"grad_norm": 0.2524625322102024,
"learning_rate": 1.9858615369627537e-05,
"loss": 0.3005,
"step": 682
},
{
"epoch": 2.765182186234818,
"grad_norm": 0.28959576396009645,
"learning_rate": 1.9802063100191916e-05,
"loss": 0.3045,
"step": 683
},
{
"epoch": 2.769230769230769,
"grad_norm": 0.25920370752723715,
"learning_rate": 1.974551241345447e-05,
"loss": 0.2996,
"step": 684
},
{
"epoch": 2.7732793522267207,
"grad_norm": 0.2682310624352893,
"learning_rate": 1.9688963761592984e-05,
"loss": 0.2794,
"step": 685
},
{
"epoch": 2.7773279352226723,
"grad_norm": 0.2717187793469048,
"learning_rate": 1.963241759676897e-05,
"loss": 0.2835,
"step": 686
},
{
"epoch": 2.7813765182186234,
"grad_norm": 0.2787941387493926,
"learning_rate": 1.9575874371124066e-05,
"loss": 0.285,
"step": 687
},
{
"epoch": 2.785425101214575,
"grad_norm": 0.2662362424650215,
"learning_rate": 1.951933453677638e-05,
"loss": 0.284,
"step": 688
},
{
"epoch": 2.7894736842105265,
"grad_norm": 0.2746311495338873,
"learning_rate": 1.9462798545816932e-05,
"loss": 0.2852,
"step": 689
},
{
"epoch": 2.7935222672064777,
"grad_norm": 0.23438308121079285,
"learning_rate": 1.9406266850305996e-05,
"loss": 0.286,
"step": 690
},
{
"epoch": 2.7975708502024292,
"grad_norm": 0.24123123061897947,
"learning_rate": 1.9349739902269503e-05,
"loss": 0.2858,
"step": 691
},
{
"epoch": 2.801619433198381,
"grad_norm": 0.24596101087628408,
"learning_rate": 1.9293218153695425e-05,
"loss": 0.2756,
"step": 692
},
{
"epoch": 2.805668016194332,
"grad_norm": 0.26592586709122884,
"learning_rate": 1.9236702056530153e-05,
"loss": 0.2808,
"step": 693
},
{
"epoch": 2.8097165991902835,
"grad_norm": 0.25383706102950154,
"learning_rate": 1.9180192062674907e-05,
"loss": 0.2906,
"step": 694
},
{
"epoch": 2.813765182186235,
"grad_norm": 0.27025147946973965,
"learning_rate": 1.9123688623982076e-05,
"loss": 0.296,
"step": 695
},
{
"epoch": 2.817813765182186,
"grad_norm": 0.26241399708117435,
"learning_rate": 1.906719219225165e-05,
"loss": 0.2779,
"step": 696
},
{
"epoch": 2.8218623481781377,
"grad_norm": 0.24597608446410474,
"learning_rate": 1.9010703219227603e-05,
"loss": 0.3014,
"step": 697
},
{
"epoch": 2.8259109311740893,
"grad_norm": 0.2716017286301881,
"learning_rate": 1.8954222156594256e-05,
"loss": 0.3005,
"step": 698
},
{
"epoch": 2.8299595141700404,
"grad_norm": 0.27166865578540644,
"learning_rate": 1.8897749455972673e-05,
"loss": 0.3017,
"step": 699
},
{
"epoch": 2.834008097165992,
"grad_norm": 0.22416871670484415,
"learning_rate": 1.884128556891707e-05,
"loss": 0.2894,
"step": 700
},
{
"epoch": 2.8380566801619436,
"grad_norm": 0.3068493869731707,
"learning_rate": 1.878483094691119e-05,
"loss": 0.2686,
"step": 701
},
{
"epoch": 2.8421052631578947,
"grad_norm": 0.22543595380314205,
"learning_rate": 1.8728386041364664e-05,
"loss": 0.2855,
"step": 702
},
{
"epoch": 2.8461538461538463,
"grad_norm": 0.2737027824353838,
"learning_rate": 1.867195130360947e-05,
"loss": 0.2802,
"step": 703
},
{
"epoch": 2.850202429149798,
"grad_norm": 0.2328343567118497,
"learning_rate": 1.8615527184896263e-05,
"loss": 0.2851,
"step": 704
},
{
"epoch": 2.854251012145749,
"grad_norm": 0.30265779642187407,
"learning_rate": 1.8559114136390795e-05,
"loss": 0.2899,
"step": 705
},
{
"epoch": 2.8582995951417005,
"grad_norm": 0.22497884203710208,
"learning_rate": 1.8502712609170298e-05,
"loss": 0.2748,
"step": 706
},
{
"epoch": 2.862348178137652,
"grad_norm": 0.2912583332407403,
"learning_rate": 1.8446323054219876e-05,
"loss": 0.2872,
"step": 707
},
{
"epoch": 2.866396761133603,
"grad_norm": 0.24153173040189252,
"learning_rate": 1.838994592242891e-05,
"loss": 0.3026,
"step": 708
},
{
"epoch": 2.8704453441295548,
"grad_norm": 0.2510514364286701,
"learning_rate": 1.8333581664587453e-05,
"loss": 0.2926,
"step": 709
},
{
"epoch": 2.8744939271255063,
"grad_norm": 0.23078158829449044,
"learning_rate": 1.8277230731382593e-05,
"loss": 0.2908,
"step": 710
},
{
"epoch": 2.8785425101214575,
"grad_norm": 0.2570287527556277,
"learning_rate": 1.82208935733949e-05,
"loss": 0.3018,
"step": 711
},
{
"epoch": 2.882591093117409,
"grad_norm": 0.2804802581944658,
"learning_rate": 1.8164570641094793e-05,
"loss": 0.2784,
"step": 712
},
{
"epoch": 2.8866396761133606,
"grad_norm": 0.24919834990704318,
"learning_rate": 1.810826238483892e-05,
"loss": 0.3011,
"step": 713
},
{
"epoch": 2.8906882591093117,
"grad_norm": 0.24818732969985663,
"learning_rate": 1.8051969254866608e-05,
"loss": 0.2988,
"step": 714
},
{
"epoch": 2.8947368421052633,
"grad_norm": 0.24616085059538875,
"learning_rate": 1.7995691701296226e-05,
"loss": 0.2982,
"step": 715
},
{
"epoch": 2.898785425101215,
"grad_norm": 0.26651939043504463,
"learning_rate": 1.7939430174121593e-05,
"loss": 0.3071,
"step": 716
},
{
"epoch": 2.902834008097166,
"grad_norm": 0.22536784307062377,
"learning_rate": 1.788318512320838e-05,
"loss": 0.275,
"step": 717
},
{
"epoch": 2.9068825910931175,
"grad_norm": 0.30078802564244334,
"learning_rate": 1.782695699829051e-05,
"loss": 0.2872,
"step": 718
},
{
"epoch": 2.910931174089069,
"grad_norm": 0.24396491071755858,
"learning_rate": 1.7770746248966587e-05,
"loss": 0.2749,
"step": 719
},
{
"epoch": 2.91497975708502,
"grad_norm": 0.31955504764156795,
"learning_rate": 1.7714553324696243e-05,
"loss": 0.29,
"step": 720
},
{
"epoch": 2.919028340080972,
"grad_norm": 0.22552131614127305,
"learning_rate": 1.7658378674796614e-05,
"loss": 0.2722,
"step": 721
},
{
"epoch": 2.9230769230769234,
"grad_norm": 0.30747018733276765,
"learning_rate": 1.76022227484387e-05,
"loss": 0.3073,
"step": 722
},
{
"epoch": 2.9271255060728745,
"grad_norm": 0.2365544854633195,
"learning_rate": 1.7546085994643807e-05,
"loss": 0.2911,
"step": 723
},
{
"epoch": 2.931174089068826,
"grad_norm": 0.2762304774859214,
"learning_rate": 1.7489968862279902e-05,
"loss": 0.2998,
"step": 724
},
{
"epoch": 2.9352226720647776,
"grad_norm": 0.265905093353331,
"learning_rate": 1.743387180005809e-05,
"loss": 0.2895,
"step": 725
},
{
"epoch": 2.9392712550607287,
"grad_norm": 0.26059495195222926,
"learning_rate": 1.737779525652899e-05,
"loss": 0.286,
"step": 726
},
{
"epoch": 2.9433198380566803,
"grad_norm": 0.2878359640396529,
"learning_rate": 1.732173968007914e-05,
"loss": 0.2862,
"step": 727
},
{
"epoch": 2.9473684210526314,
"grad_norm": 0.30486937611626747,
"learning_rate": 1.7265705518927437e-05,
"loss": 0.2915,
"step": 728
},
{
"epoch": 2.951417004048583,
"grad_norm": 0.292717248605955,
"learning_rate": 1.7209693221121542e-05,
"loss": 0.2891,
"step": 729
},
{
"epoch": 2.9554655870445345,
"grad_norm": 0.2667775831349283,
"learning_rate": 1.7153703234534302e-05,
"loss": 0.2766,
"step": 730
},
{
"epoch": 2.9595141700404857,
"grad_norm": 0.35248066420679575,
"learning_rate": 1.709773600686015e-05,
"loss": 0.2811,
"step": 731
},
{
"epoch": 2.9635627530364372,
"grad_norm": 0.2579560789616744,
"learning_rate": 1.7041791985611548e-05,
"loss": 0.2988,
"step": 732
},
{
"epoch": 2.967611336032389,
"grad_norm": 0.30141893360804944,
"learning_rate": 1.69858716181154e-05,
"loss": 0.2898,
"step": 733
},
{
"epoch": 2.97165991902834,
"grad_norm": 0.24230667497420216,
"learning_rate": 1.692997535150948e-05,
"loss": 0.2825,
"step": 734
},
{
"epoch": 2.9757085020242915,
"grad_norm": 0.27843868061304733,
"learning_rate": 1.687410363273884e-05,
"loss": 0.2822,
"step": 735
},
{
"epoch": 2.979757085020243,
"grad_norm": 0.24637497543749354,
"learning_rate": 1.6818256908552257e-05,
"loss": 0.2762,
"step": 736
},
{
"epoch": 2.983805668016194,
"grad_norm": 0.2615248540416007,
"learning_rate": 1.6762435625498646e-05,
"loss": 0.2964,
"step": 737
},
{
"epoch": 2.9878542510121457,
"grad_norm": 0.2686931622746716,
"learning_rate": 1.67066402299235e-05,
"loss": 0.2936,
"step": 738
},
{
"epoch": 2.9919028340080973,
"grad_norm": 0.25180723191697296,
"learning_rate": 1.6650871167965313e-05,
"loss": 0.291,
"step": 739
},
{
"epoch": 2.9959514170040484,
"grad_norm": 0.2567609769884492,
"learning_rate": 1.6595128885552028e-05,
"loss": 0.2857,
"step": 740
},
{
"epoch": 3.0,
"grad_norm": 0.2471970505725342,
"learning_rate": 1.653941382839745e-05,
"loss": 0.2824,
"step": 741
},
{
"epoch": 3.0040485829959516,
"grad_norm": 0.3709069179035863,
"learning_rate": 1.6483726441997673e-05,
"loss": 0.2332,
"step": 742
},
{
"epoch": 3.0080971659919027,
"grad_norm": 0.25523522347249733,
"learning_rate": 1.642806717162757e-05,
"loss": 0.2156,
"step": 743
},
{
"epoch": 3.0121457489878543,
"grad_norm": 0.5162317013470084,
"learning_rate": 1.637243646233718e-05,
"loss": 0.235,
"step": 744
},
{
"epoch": 3.016194331983806,
"grad_norm": 0.4159852137411412,
"learning_rate": 1.6316834758948174e-05,
"loss": 0.2291,
"step": 745
},
{
"epoch": 3.020242914979757,
"grad_norm": 0.33748928752941876,
"learning_rate": 1.6261262506050282e-05,
"loss": 0.216,
"step": 746
},
{
"epoch": 3.0242914979757085,
"grad_norm": 0.31212309861243964,
"learning_rate": 1.620572014799777e-05,
"loss": 0.2496,
"step": 747
},
{
"epoch": 3.02834008097166,
"grad_norm": 0.33630689403067726,
"learning_rate": 1.6150208128905857e-05,
"loss": 0.2347,
"step": 748
},
{
"epoch": 3.032388663967611,
"grad_norm": 0.2802217727363067,
"learning_rate": 1.6094726892647147e-05,
"loss": 0.217,
"step": 749
},
{
"epoch": 3.0364372469635628,
"grad_norm": 0.29732460106548203,
"learning_rate": 1.6039276882848135e-05,
"loss": 0.2237,
"step": 750
},
{
"epoch": 3.0404858299595143,
"grad_norm": 0.246571635449631,
"learning_rate": 1.5983858542885635e-05,
"loss": 0.244,
"step": 751
},
{
"epoch": 3.0445344129554655,
"grad_norm": 0.28210867071422546,
"learning_rate": 1.5928472315883203e-05,
"loss": 0.2264,
"step": 752
},
{
"epoch": 3.048582995951417,
"grad_norm": 0.25406953148112815,
"learning_rate": 1.5873118644707633e-05,
"loss": 0.2284,
"step": 753
},
{
"epoch": 3.0526315789473686,
"grad_norm": 0.2860509510319946,
"learning_rate": 1.5817797971965413e-05,
"loss": 0.2338,
"step": 754
},
{
"epoch": 3.0566801619433197,
"grad_norm": 0.2558852591627757,
"learning_rate": 1.576251073999917e-05,
"loss": 0.2356,
"step": 755
},
{
"epoch": 3.0607287449392713,
"grad_norm": 0.2742301451141068,
"learning_rate": 1.5707257390884126e-05,
"loss": 0.2291,
"step": 756
},
{
"epoch": 3.064777327935223,
"grad_norm": 0.23545896707935618,
"learning_rate": 1.5652038366424595e-05,
"loss": 0.2195,
"step": 757
},
{
"epoch": 3.068825910931174,
"grad_norm": 0.26077106124231064,
"learning_rate": 1.5596854108150424e-05,
"loss": 0.2268,
"step": 758
},
{
"epoch": 3.0728744939271255,
"grad_norm": 0.2417961282615855,
"learning_rate": 1.5541705057313476e-05,
"loss": 0.2109,
"step": 759
},
{
"epoch": 3.076923076923077,
"grad_norm": 0.24289597989914385,
"learning_rate": 1.5486591654884086e-05,
"loss": 0.2337,
"step": 760
},
{
"epoch": 3.080971659919028,
"grad_norm": 0.251356586657077,
"learning_rate": 1.543151434154755e-05,
"loss": 0.2245,
"step": 761
},
{
"epoch": 3.08502024291498,
"grad_norm": 0.2398602752783331,
"learning_rate": 1.5376473557700615e-05,
"loss": 0.2293,
"step": 762
},
{
"epoch": 3.0890688259109313,
"grad_norm": 0.2739041317347452,
"learning_rate": 1.5321469743447893e-05,
"loss": 0.2117,
"step": 763
},
{
"epoch": 3.0931174089068825,
"grad_norm": 0.23066978515870934,
"learning_rate": 1.5266503338598434e-05,
"loss": 0.2177,
"step": 764
},
{
"epoch": 3.097165991902834,
"grad_norm": 0.2717624011499155,
"learning_rate": 1.5211574782662149e-05,
"loss": 0.2376,
"step": 765
},
{
"epoch": 3.1012145748987856,
"grad_norm": 0.23458865444001553,
"learning_rate": 1.5156684514846307e-05,
"loss": 0.2236,
"step": 766
},
{
"epoch": 3.1052631578947367,
"grad_norm": 0.2562957334252301,
"learning_rate": 1.5101832974052012e-05,
"loss": 0.2199,
"step": 767
},
{
"epoch": 3.1093117408906883,
"grad_norm": 0.2263188521148361,
"learning_rate": 1.5047020598870737e-05,
"loss": 0.2278,
"step": 768
},
{
"epoch": 3.11336032388664,
"grad_norm": 0.23088684097888565,
"learning_rate": 1.4992247827580778e-05,
"loss": 0.2333,
"step": 769
},
{
"epoch": 3.117408906882591,
"grad_norm": 0.22181983944825232,
"learning_rate": 1.493751509814374e-05,
"loss": 0.2203,
"step": 770
},
{
"epoch": 3.1214574898785425,
"grad_norm": 0.254958283635696,
"learning_rate": 1.4882822848201073e-05,
"loss": 0.2297,
"step": 771
},
{
"epoch": 3.125506072874494,
"grad_norm": 0.2517029711348442,
"learning_rate": 1.4828171515070553e-05,
"loss": 0.2279,
"step": 772
},
{
"epoch": 3.1295546558704452,
"grad_norm": 0.24707179767792015,
"learning_rate": 1.4773561535742793e-05,
"loss": 0.2361,
"step": 773
},
{
"epoch": 3.133603238866397,
"grad_norm": 0.241995113621964,
"learning_rate": 1.4718993346877715e-05,
"loss": 0.2294,
"step": 774
},
{
"epoch": 3.1376518218623484,
"grad_norm": 0.2681284911658161,
"learning_rate": 1.466446738480111e-05,
"loss": 0.2307,
"step": 775
},
{
"epoch": 3.1417004048582995,
"grad_norm": 0.2383863425262094,
"learning_rate": 1.460998408550112e-05,
"loss": 0.243,
"step": 776
},
{
"epoch": 3.145748987854251,
"grad_norm": 0.26023734466933307,
"learning_rate": 1.4555543884624751e-05,
"loss": 0.2315,
"step": 777
},
{
"epoch": 3.1497975708502026,
"grad_norm": 0.22072787595024596,
"learning_rate": 1.4501147217474402e-05,
"loss": 0.2312,
"step": 778
},
{
"epoch": 3.1538461538461537,
"grad_norm": 0.257643295388198,
"learning_rate": 1.444679451900437e-05,
"loss": 0.2232,
"step": 779
},
{
"epoch": 3.1578947368421053,
"grad_norm": 0.22912546581359597,
"learning_rate": 1.4392486223817397e-05,
"loss": 0.2221,
"step": 780
},
{
"epoch": 3.161943319838057,
"grad_norm": 0.27393597792907387,
"learning_rate": 1.4338222766161143e-05,
"loss": 0.2229,
"step": 781
},
{
"epoch": 3.165991902834008,
"grad_norm": 0.2637765150993913,
"learning_rate": 1.4284004579924777e-05,
"loss": 0.2252,
"step": 782
},
{
"epoch": 3.1700404858299596,
"grad_norm": 0.2459654384327267,
"learning_rate": 1.4229832098635479e-05,
"loss": 0.2151,
"step": 783
},
{
"epoch": 3.174089068825911,
"grad_norm": 0.2581029396720227,
"learning_rate": 1.4175705755454963e-05,
"loss": 0.2244,
"step": 784
},
{
"epoch": 3.1781376518218623,
"grad_norm": 0.2610929764623126,
"learning_rate": 1.412162598317602e-05,
"loss": 0.2304,
"step": 785
},
{
"epoch": 3.182186234817814,
"grad_norm": 0.25983274286374136,
"learning_rate": 1.406759321421907e-05,
"loss": 0.2184,
"step": 786
},
{
"epoch": 3.1862348178137654,
"grad_norm": 0.25897665717472507,
"learning_rate": 1.401360788062871e-05,
"loss": 0.2254,
"step": 787
},
{
"epoch": 3.1902834008097165,
"grad_norm": 0.2549882887450048,
"learning_rate": 1.3959670414070204e-05,
"loss": 0.2324,
"step": 788
},
{
"epoch": 3.194331983805668,
"grad_norm": 0.2679498755842375,
"learning_rate": 1.3905781245826108e-05,
"loss": 0.2222,
"step": 789
},
{
"epoch": 3.1983805668016196,
"grad_norm": 0.2405921358089652,
"learning_rate": 1.3851940806792778e-05,
"loss": 0.2259,
"step": 790
},
{
"epoch": 3.2024291497975708,
"grad_norm": 0.2534288589320947,
"learning_rate": 1.379814952747693e-05,
"loss": 0.2304,
"step": 791
},
{
"epoch": 3.2064777327935223,
"grad_norm": 0.22560512232966054,
"learning_rate": 1.3744407837992193e-05,
"loss": 0.2195,
"step": 792
},
{
"epoch": 3.2105263157894735,
"grad_norm": 0.26350339271035667,
"learning_rate": 1.3690716168055692e-05,
"loss": 0.2239,
"step": 793
},
{
"epoch": 3.214574898785425,
"grad_norm": 0.23890930878723485,
"learning_rate": 1.363707494698459e-05,
"loss": 0.2239,
"step": 794
},
{
"epoch": 3.2186234817813766,
"grad_norm": 0.23664074346807876,
"learning_rate": 1.358348460369265e-05,
"loss": 0.2231,
"step": 795
},
{
"epoch": 3.2226720647773277,
"grad_norm": 0.26490322316144255,
"learning_rate": 1.352994556668684e-05,
"loss": 0.2298,
"step": 796
},
{
"epoch": 3.2267206477732793,
"grad_norm": 0.2386495891565952,
"learning_rate": 1.3476458264063865e-05,
"loss": 0.225,
"step": 797
},
{
"epoch": 3.230769230769231,
"grad_norm": 0.2741986738674334,
"learning_rate": 1.342302312350678e-05,
"loss": 0.2233,
"step": 798
},
{
"epoch": 3.234817813765182,
"grad_norm": 0.24747838487805177,
"learning_rate": 1.3369640572281537e-05,
"loss": 0.2337,
"step": 799
},
{
"epoch": 3.2388663967611335,
"grad_norm": 0.2357032627332941,
"learning_rate": 1.3316311037233596e-05,
"loss": 0.2192,
"step": 800
},
{
"epoch": 3.242914979757085,
"grad_norm": 0.23252419432830523,
"learning_rate": 1.326303494478451e-05,
"loss": 0.2269,
"step": 801
},
{
"epoch": 3.246963562753036,
"grad_norm": 0.23272937851070763,
"learning_rate": 1.3209812720928472e-05,
"loss": 0.2355,
"step": 802
},
{
"epoch": 3.251012145748988,
"grad_norm": 0.23003831882006423,
"learning_rate": 1.315664479122898e-05,
"loss": 0.2327,
"step": 803
},
{
"epoch": 3.2550607287449393,
"grad_norm": 0.22869923494655459,
"learning_rate": 1.3103531580815378e-05,
"loss": 0.236,
"step": 804
},
{
"epoch": 3.2591093117408905,
"grad_norm": 0.2271878476021522,
"learning_rate": 1.3050473514379488e-05,
"loss": 0.2219,
"step": 805
},
{
"epoch": 3.263157894736842,
"grad_norm": 0.24335791068748308,
"learning_rate": 1.2997471016172188e-05,
"loss": 0.2143,
"step": 806
},
{
"epoch": 3.2672064777327936,
"grad_norm": 0.23658509726505597,
"learning_rate": 1.2944524510000042e-05,
"loss": 0.238,
"step": 807
},
{
"epoch": 3.2712550607287447,
"grad_norm": 0.24419386966295617,
"learning_rate": 1.2891634419221908e-05,
"loss": 0.2284,
"step": 808
},
{
"epoch": 3.2753036437246963,
"grad_norm": 0.23234183268017305,
"learning_rate": 1.2838801166745545e-05,
"loss": 0.2202,
"step": 809
},
{
"epoch": 3.279352226720648,
"grad_norm": 0.2574686333744161,
"learning_rate": 1.2786025175024216e-05,
"loss": 0.23,
"step": 810
},
{
"epoch": 3.283400809716599,
"grad_norm": 0.2521496644732963,
"learning_rate": 1.2733306866053357e-05,
"loss": 0.2193,
"step": 811
},
{
"epoch": 3.2874493927125505,
"grad_norm": 0.24281629120073883,
"learning_rate": 1.2680646661367163e-05,
"loss": 0.2422,
"step": 812
},
{
"epoch": 3.291497975708502,
"grad_norm": 0.28433656145724495,
"learning_rate": 1.2628044982035219e-05,
"loss": 0.237,
"step": 813
},
{
"epoch": 3.2955465587044532,
"grad_norm": 0.23825528218557304,
"learning_rate": 1.2575502248659155e-05,
"loss": 0.2212,
"step": 814
},
{
"epoch": 3.299595141700405,
"grad_norm": 0.2593994524755086,
"learning_rate": 1.2523018881369269e-05,
"loss": 0.2223,
"step": 815
},
{
"epoch": 3.3036437246963564,
"grad_norm": 0.2592653615507496,
"learning_rate": 1.2470595299821176e-05,
"loss": 0.2222,
"step": 816
},
{
"epoch": 3.3076923076923075,
"grad_norm": 0.23018310714310244,
"learning_rate": 1.2418231923192416e-05,
"loss": 0.2254,
"step": 817
},
{
"epoch": 3.311740890688259,
"grad_norm": 0.26663840726387594,
"learning_rate": 1.2365929170179165e-05,
"loss": 0.2261,
"step": 818
},
{
"epoch": 3.3157894736842106,
"grad_norm": 0.24840103758636525,
"learning_rate": 1.2313687458992848e-05,
"loss": 0.2247,
"step": 819
},
{
"epoch": 3.3198380566801617,
"grad_norm": 0.2370507815659776,
"learning_rate": 1.2261507207356785e-05,
"loss": 0.2156,
"step": 820
},
{
"epoch": 3.3238866396761133,
"grad_norm": 0.29140670595460644,
"learning_rate": 1.220938883250288e-05,
"loss": 0.2239,
"step": 821
},
{
"epoch": 3.327935222672065,
"grad_norm": 0.24106578275691484,
"learning_rate": 1.215733275116827e-05,
"loss": 0.2295,
"step": 822
},
{
"epoch": 3.331983805668016,
"grad_norm": 0.24337296646751605,
"learning_rate": 1.2105339379592005e-05,
"loss": 0.214,
"step": 823
},
{
"epoch": 3.3360323886639676,
"grad_norm": 0.2533505834540974,
"learning_rate": 1.2053409133511681e-05,
"loss": 0.2433,
"step": 824
},
{
"epoch": 3.340080971659919,
"grad_norm": 0.22528433083662855,
"learning_rate": 1.2001542428160173e-05,
"loss": 0.2037,
"step": 825
},
{
"epoch": 3.3441295546558703,
"grad_norm": 0.2710618731905354,
"learning_rate": 1.1949739678262278e-05,
"loss": 0.2157,
"step": 826
},
{
"epoch": 3.348178137651822,
"grad_norm": 0.2405341286904844,
"learning_rate": 1.189800129803139e-05,
"loss": 0.2292,
"step": 827
},
{
"epoch": 3.3522267206477734,
"grad_norm": 0.24727061458726843,
"learning_rate": 1.1846327701166228e-05,
"loss": 0.2205,
"step": 828
},
{
"epoch": 3.3562753036437245,
"grad_norm": 0.2309211202465577,
"learning_rate": 1.1794719300847493e-05,
"loss": 0.2379,
"step": 829
},
{
"epoch": 3.360323886639676,
"grad_norm": 0.23650878652979132,
"learning_rate": 1.1743176509734582e-05,
"loss": 0.2195,
"step": 830
},
{
"epoch": 3.3643724696356276,
"grad_norm": 0.241431055266019,
"learning_rate": 1.1691699739962275e-05,
"loss": 0.2144,
"step": 831
},
{
"epoch": 3.3684210526315788,
"grad_norm": 0.2778718113279376,
"learning_rate": 1.1640289403137456e-05,
"loss": 0.2179,
"step": 832
},
{
"epoch": 3.3724696356275303,
"grad_norm": 0.24394564725512688,
"learning_rate": 1.1588945910335802e-05,
"loss": 0.2256,
"step": 833
},
{
"epoch": 3.376518218623482,
"grad_norm": 0.24133884326058433,
"learning_rate": 1.1537669672098517e-05,
"loss": 0.2235,
"step": 834
},
{
"epoch": 3.380566801619433,
"grad_norm": 0.2322635185049532,
"learning_rate": 1.1486461098429036e-05,
"loss": 0.2263,
"step": 835
},
{
"epoch": 3.3846153846153846,
"grad_norm": 0.25065662627060187,
"learning_rate": 1.1435320598789747e-05,
"loss": 0.2064,
"step": 836
},
{
"epoch": 3.388663967611336,
"grad_norm": 0.2631074957263954,
"learning_rate": 1.1384248582098736e-05,
"loss": 0.2322,
"step": 837
},
{
"epoch": 3.3927125506072873,
"grad_norm": 0.25075180915587614,
"learning_rate": 1.1333245456726467e-05,
"loss": 0.2378,
"step": 838
},
{
"epoch": 3.396761133603239,
"grad_norm": 0.2705432663027698,
"learning_rate": 1.128231163049258e-05,
"loss": 0.2114,
"step": 839
},
{
"epoch": 3.4008097165991904,
"grad_norm": 0.21596094113183062,
"learning_rate": 1.123144751066259e-05,
"loss": 0.2295,
"step": 840
},
{
"epoch": 3.4048582995951415,
"grad_norm": 0.2651038597425194,
"learning_rate": 1.1180653503944657e-05,
"loss": 0.2237,
"step": 841
},
{
"epoch": 3.408906882591093,
"grad_norm": 0.23990809384709233,
"learning_rate": 1.1129930016486287e-05,
"loss": 0.2214,
"step": 842
},
{
"epoch": 3.4129554655870447,
"grad_norm": 0.26084561454850064,
"learning_rate": 1.107927745387114e-05,
"loss": 0.2228,
"step": 843
},
{
"epoch": 3.417004048582996,
"grad_norm": 0.25289228464534314,
"learning_rate": 1.1028696221115766e-05,
"loss": 0.2273,
"step": 844
},
{
"epoch": 3.4210526315789473,
"grad_norm": 0.2272668323518232,
"learning_rate": 1.0978186722666341e-05,
"loss": 0.2389,
"step": 845
},
{
"epoch": 3.425101214574899,
"grad_norm": 0.2237365841199938,
"learning_rate": 1.0927749362395475e-05,
"loss": 0.2097,
"step": 846
},
{
"epoch": 3.42914979757085,
"grad_norm": 0.23140320546917645,
"learning_rate": 1.0877384543598954e-05,
"loss": 0.24,
"step": 847
},
{
"epoch": 3.4331983805668016,
"grad_norm": 0.21319222814117514,
"learning_rate": 1.0827092668992529e-05,
"loss": 0.2315,
"step": 848
},
{
"epoch": 3.437246963562753,
"grad_norm": 0.24886314737979917,
"learning_rate": 1.077687414070869e-05,
"loss": 0.2144,
"step": 849
},
{
"epoch": 3.4412955465587043,
"grad_norm": 0.20614843731366525,
"learning_rate": 1.0726729360293451e-05,
"loss": 0.2266,
"step": 850
},
{
"epoch": 3.445344129554656,
"grad_norm": 0.24859276890613968,
"learning_rate": 1.0676658728703147e-05,
"loss": 0.2294,
"step": 851
},
{
"epoch": 3.4493927125506074,
"grad_norm": 0.2383413753754853,
"learning_rate": 1.0626662646301198e-05,
"loss": 0.2135,
"step": 852
},
{
"epoch": 3.4534412955465585,
"grad_norm": 0.24307327045106364,
"learning_rate": 1.0576741512854958e-05,
"loss": 0.2079,
"step": 853
},
{
"epoch": 3.45748987854251,
"grad_norm": 0.23446287437658164,
"learning_rate": 1.052689572753247e-05,
"loss": 0.2212,
"step": 854
},
{
"epoch": 3.4615384615384617,
"grad_norm": 0.2163046806010593,
"learning_rate": 1.0477125688899328e-05,
"loss": 0.2306,
"step": 855
},
{
"epoch": 3.465587044534413,
"grad_norm": 0.22672280868484235,
"learning_rate": 1.0427431794915406e-05,
"loss": 0.2202,
"step": 856
},
{
"epoch": 3.4696356275303644,
"grad_norm": 0.21995931792441548,
"learning_rate": 1.0377814442931769e-05,
"loss": 0.219,
"step": 857
},
{
"epoch": 3.473684210526316,
"grad_norm": 0.21042934065474025,
"learning_rate": 1.0328274029687439e-05,
"loss": 0.2141,
"step": 858
},
{
"epoch": 3.477732793522267,
"grad_norm": 0.23490436342585963,
"learning_rate": 1.0278810951306253e-05,
"loss": 0.235,
"step": 859
},
{
"epoch": 3.4817813765182186,
"grad_norm": 0.23201019992785735,
"learning_rate": 1.0229425603293647e-05,
"loss": 0.2323,
"step": 860
},
{
"epoch": 3.48582995951417,
"grad_norm": 0.21119238090067988,
"learning_rate": 1.0180118380533558e-05,
"loss": 0.2399,
"step": 861
},
{
"epoch": 3.4898785425101213,
"grad_norm": 0.2281792837068995,
"learning_rate": 1.0130889677285225e-05,
"loss": 0.2192,
"step": 862
},
{
"epoch": 3.493927125506073,
"grad_norm": 0.23516161172468494,
"learning_rate": 1.008173988718005e-05,
"loss": 0.2253,
"step": 863
},
{
"epoch": 3.4979757085020244,
"grad_norm": 0.21300755128789622,
"learning_rate": 1.0032669403218439e-05,
"loss": 0.2287,
"step": 864
},
{
"epoch": 3.5020242914979756,
"grad_norm": 0.21299438044537783,
"learning_rate": 9.983678617766677e-06,
"loss": 0.2214,
"step": 865
},
{
"epoch": 3.506072874493927,
"grad_norm": 0.20755710753452564,
"learning_rate": 9.934767922553788e-06,
"loss": 0.2204,
"step": 866
},
{
"epoch": 3.5101214574898787,
"grad_norm": 0.2071057562309898,
"learning_rate": 9.885937708668371e-06,
"loss": 0.2353,
"step": 867
},
{
"epoch": 3.51417004048583,
"grad_norm": 0.23590529151086315,
"learning_rate": 9.837188366555524e-06,
"loss": 0.2261,
"step": 868
},
{
"epoch": 3.5182186234817814,
"grad_norm": 0.23773844751946202,
"learning_rate": 9.788520286013691e-06,
"loss": 0.2113,
"step": 869
},
{
"epoch": 3.522267206477733,
"grad_norm": 0.2187780156221126,
"learning_rate": 9.739933856191537e-06,
"loss": 0.2239,
"step": 870
},
{
"epoch": 3.526315789473684,
"grad_norm": 0.2089844149951792,
"learning_rate": 9.691429465584866e-06,
"loss": 0.2302,
"step": 871
},
{
"epoch": 3.5303643724696356,
"grad_norm": 0.2305523300573757,
"learning_rate": 9.643007502033492e-06,
"loss": 0.2145,
"step": 872
},
{
"epoch": 3.534412955465587,
"grad_norm": 0.21752773081932364,
"learning_rate": 9.594668352718152e-06,
"loss": 0.2362,
"step": 873
},
{
"epoch": 3.5384615384615383,
"grad_norm": 0.2220434759618642,
"learning_rate": 9.546412404157383e-06,
"loss": 0.23,
"step": 874
},
{
"epoch": 3.54251012145749,
"grad_norm": 0.2421651237469766,
"learning_rate": 9.498240042204471e-06,
"loss": 0.2222,
"step": 875
},
{
"epoch": 3.5465587044534415,
"grad_norm": 0.21782473715075393,
"learning_rate": 9.450151652044339e-06,
"loss": 0.208,
"step": 876
},
{
"epoch": 3.5506072874493926,
"grad_norm": 0.20981569630987507,
"learning_rate": 9.402147618190472e-06,
"loss": 0.2191,
"step": 877
},
{
"epoch": 3.554655870445344,
"grad_norm": 0.2315288853131941,
"learning_rate": 9.354228324481841e-06,
"loss": 0.2149,
"step": 878
},
{
"epoch": 3.5587044534412957,
"grad_norm": 0.22038891668206562,
"learning_rate": 9.306394154079852e-06,
"loss": 0.2379,
"step": 879
},
{
"epoch": 3.562753036437247,
"grad_norm": 0.2203895233186104,
"learning_rate": 9.258645489465254e-06,
"loss": 0.2292,
"step": 880
},
{
"epoch": 3.5668016194331984,
"grad_norm": 0.2147659818201551,
"learning_rate": 9.21098271243509e-06,
"loss": 0.229,
"step": 881
},
{
"epoch": 3.57085020242915,
"grad_norm": 0.22342456827385707,
"learning_rate": 9.163406204099659e-06,
"loss": 0.2316,
"step": 882
},
{
"epoch": 3.574898785425101,
"grad_norm": 0.20724168301544812,
"learning_rate": 9.115916344879452e-06,
"loss": 0.2131,
"step": 883
},
{
"epoch": 3.5789473684210527,
"grad_norm": 0.22644469189595037,
"learning_rate": 9.068513514502126e-06,
"loss": 0.2076,
"step": 884
},
{
"epoch": 3.582995951417004,
"grad_norm": 0.2569316936825209,
"learning_rate": 9.021198091999439e-06,
"loss": 0.2327,
"step": 885
},
{
"epoch": 3.5870445344129553,
"grad_norm": 0.20309308708649895,
"learning_rate": 8.973970455704251e-06,
"loss": 0.2241,
"step": 886
},
{
"epoch": 3.591093117408907,
"grad_norm": 0.1999614639487628,
"learning_rate": 8.926830983247495e-06,
"loss": 0.2125,
"step": 887
},
{
"epoch": 3.5951417004048585,
"grad_norm": 0.22954591400849497,
"learning_rate": 8.879780051555116e-06,
"loss": 0.2179,
"step": 888
},
{
"epoch": 3.5991902834008096,
"grad_norm": 0.21052789950116216,
"learning_rate": 8.83281803684512e-06,
"loss": 0.2159,
"step": 889
},
{
"epoch": 3.603238866396761,
"grad_norm": 0.23104744713492828,
"learning_rate": 8.785945314624532e-06,
"loss": 0.2155,
"step": 890
},
{
"epoch": 3.6072874493927127,
"grad_norm": 0.22647623670826816,
"learning_rate": 8.739162259686385e-06,
"loss": 0.2176,
"step": 891
},
{
"epoch": 3.611336032388664,
"grad_norm": 0.21544841958377503,
"learning_rate": 8.692469246106734e-06,
"loss": 0.2235,
"step": 892
},
{
"epoch": 3.6153846153846154,
"grad_norm": 0.21159470957466642,
"learning_rate": 8.645866647241677e-06,
"loss": 0.2235,
"step": 893
},
{
"epoch": 3.619433198380567,
"grad_norm": 0.20737285263536137,
"learning_rate": 8.599354835724348e-06,
"loss": 0.2319,
"step": 894
},
{
"epoch": 3.623481781376518,
"grad_norm": 0.21351392823160625,
"learning_rate": 8.552934183461958e-06,
"loss": 0.2269,
"step": 895
},
{
"epoch": 3.6275303643724697,
"grad_norm": 0.21629481457086852,
"learning_rate": 8.506605061632805e-06,
"loss": 0.2367,
"step": 896
},
{
"epoch": 3.6315789473684212,
"grad_norm": 0.23492274390252676,
"learning_rate": 8.460367840683308e-06,
"loss": 0.2263,
"step": 897
},
{
"epoch": 3.6356275303643724,
"grad_norm": 0.20328560566675857,
"learning_rate": 8.414222890325065e-06,
"loss": 0.2012,
"step": 898
},
{
"epoch": 3.639676113360324,
"grad_norm": 0.23080369312124324,
"learning_rate": 8.368170579531855e-06,
"loss": 0.2324,
"step": 899
},
{
"epoch": 3.6437246963562755,
"grad_norm": 0.20747490251577516,
"learning_rate": 8.322211276536733e-06,
"loss": 0.2185,
"step": 900
},
{
"epoch": 3.6477732793522266,
"grad_norm": 0.2067778666212408,
"learning_rate": 8.27634534882907e-06,
"loss": 0.2194,
"step": 901
},
{
"epoch": 3.651821862348178,
"grad_norm": 0.19286911042538726,
"learning_rate": 8.230573163151585e-06,
"loss": 0.2031,
"step": 902
},
{
"epoch": 3.6558704453441297,
"grad_norm": 0.20173807732854662,
"learning_rate": 8.184895085497466e-06,
"loss": 0.2213,
"step": 903
},
{
"epoch": 3.659919028340081,
"grad_norm": 0.19522020422146164,
"learning_rate": 8.139311481107401e-06,
"loss": 0.2046,
"step": 904
},
{
"epoch": 3.6639676113360324,
"grad_norm": 0.20680915929564114,
"learning_rate": 8.093822714466686e-06,
"loss": 0.2242,
"step": 905
},
{
"epoch": 3.668016194331984,
"grad_norm": 0.21006473434955084,
"learning_rate": 8.048429149302271e-06,
"loss": 0.2207,
"step": 906
},
{
"epoch": 3.672064777327935,
"grad_norm": 0.19693236961230254,
"learning_rate": 8.003131148579902e-06,
"loss": 0.2151,
"step": 907
},
{
"epoch": 3.6761133603238867,
"grad_norm": 0.21001354506413145,
"learning_rate": 7.957929074501182e-06,
"loss": 0.2279,
"step": 908
},
{
"epoch": 3.6801619433198383,
"grad_norm": 0.21164196064169913,
"learning_rate": 7.912823288500695e-06,
"loss": 0.2069,
"step": 909
},
{
"epoch": 3.6842105263157894,
"grad_norm": 0.2124841825227164,
"learning_rate": 7.867814151243102e-06,
"loss": 0.2086,
"step": 910
},
{
"epoch": 3.688259109311741,
"grad_norm": 0.22065727817859668,
"learning_rate": 7.822902022620267e-06,
"loss": 0.2206,
"step": 911
},
{
"epoch": 3.6923076923076925,
"grad_norm": 0.22112041668956597,
"learning_rate": 7.778087261748377e-06,
"loss": 0.2212,
"step": 912
},
{
"epoch": 3.6963562753036436,
"grad_norm": 0.23403357800893074,
"learning_rate": 7.733370226965056e-06,
"loss": 0.2128,
"step": 913
},
{
"epoch": 3.700404858299595,
"grad_norm": 0.2265205433698547,
"learning_rate": 7.688751275826527e-06,
"loss": 0.2189,
"step": 914
},
{
"epoch": 3.7044534412955468,
"grad_norm": 0.2176755083915206,
"learning_rate": 7.644230765104735e-06,
"loss": 0.2213,
"step": 915
},
{
"epoch": 3.708502024291498,
"grad_norm": 0.23085707021143434,
"learning_rate": 7.599809050784502e-06,
"loss": 0.2327,
"step": 916
},
{
"epoch": 3.7125506072874495,
"grad_norm": 0.19876581089627868,
"learning_rate": 7.555486488060661e-06,
"loss": 0.2075,
"step": 917
},
{
"epoch": 3.716599190283401,
"grad_norm": 0.2026737655537992,
"learning_rate": 7.511263431335249e-06,
"loss": 0.2318,
"step": 918
},
{
"epoch": 3.720647773279352,
"grad_norm": 0.22588852887072108,
"learning_rate": 7.467140234214656e-06,
"loss": 0.2233,
"step": 919
},
{
"epoch": 3.7246963562753037,
"grad_norm": 0.20990764545066498,
"learning_rate": 7.423117249506777e-06,
"loss": 0.2232,
"step": 920
},
{
"epoch": 3.7287449392712553,
"grad_norm": 0.19692913183694546,
"learning_rate": 7.379194829218231e-06,
"loss": 0.2193,
"step": 921
},
{
"epoch": 3.7327935222672064,
"grad_norm": 0.2306994980804073,
"learning_rate": 7.335373324551518e-06,
"loss": 0.2119,
"step": 922
},
{
"epoch": 3.736842105263158,
"grad_norm": 0.2062629022020476,
"learning_rate": 7.291653085902224e-06,
"loss": 0.2238,
"step": 923
},
{
"epoch": 3.7408906882591095,
"grad_norm": 0.20909467410565435,
"learning_rate": 7.248034462856213e-06,
"loss": 0.2243,
"step": 924
},
{
"epoch": 3.7449392712550607,
"grad_norm": 0.22395720221467758,
"learning_rate": 7.20451780418683e-06,
"loss": 0.2237,
"step": 925
},
{
"epoch": 3.748987854251012,
"grad_norm": 0.19640048696517273,
"learning_rate": 7.161103457852128e-06,
"loss": 0.2255,
"step": 926
},
{
"epoch": 3.753036437246964,
"grad_norm": 0.2022187476884077,
"learning_rate": 7.117791770992046e-06,
"loss": 0.2177,
"step": 927
},
{
"epoch": 3.757085020242915,
"grad_norm": 0.23545858684015647,
"learning_rate": 7.074583089925688e-06,
"loss": 0.2263,
"step": 928
},
{
"epoch": 3.7611336032388665,
"grad_norm": 0.20099513102798838,
"learning_rate": 7.0314777601485155e-06,
"loss": 0.2214,
"step": 929
},
{
"epoch": 3.765182186234818,
"grad_norm": 0.20467963659502528,
"learning_rate": 6.988476126329604e-06,
"loss": 0.2171,
"step": 930
},
{
"epoch": 3.769230769230769,
"grad_norm": 0.20435195034013268,
"learning_rate": 6.945578532308857e-06,
"loss": 0.2187,
"step": 931
},
{
"epoch": 3.7732793522267207,
"grad_norm": 0.19776161319845387,
"learning_rate": 6.902785321094301e-06,
"loss": 0.2246,
"step": 932
},
{
"epoch": 3.7773279352226723,
"grad_norm": 0.2001960277950217,
"learning_rate": 6.8600968348593115e-06,
"loss": 0.221,
"step": 933
},
{
"epoch": 3.7813765182186234,
"grad_norm": 0.20487263175967546,
"learning_rate": 6.817513414939892e-06,
"loss": 0.2207,
"step": 934
},
{
"epoch": 3.785425101214575,
"grad_norm": 0.21451249950340892,
"learning_rate": 6.775035401831924e-06,
"loss": 0.2172,
"step": 935
},
{
"epoch": 3.7894736842105265,
"grad_norm": 0.20971744802420747,
"learning_rate": 6.732663135188473e-06,
"loss": 0.2085,
"step": 936
},
{
"epoch": 3.7935222672064777,
"grad_norm": 0.20367496287418305,
"learning_rate": 6.690396953817055e-06,
"loss": 0.2182,
"step": 937
},
{
"epoch": 3.7975708502024292,
"grad_norm": 0.19925972480588106,
"learning_rate": 6.648237195676927e-06,
"loss": 0.2123,
"step": 938
},
{
"epoch": 3.801619433198381,
"grad_norm": 0.20706726814005255,
"learning_rate": 6.606184197876395e-06,
"loss": 0.2261,
"step": 939
},
{
"epoch": 3.805668016194332,
"grad_norm": 0.2186952507927138,
"learning_rate": 6.564238296670109e-06,
"loss": 0.2289,
"step": 940
},
{
"epoch": 3.8097165991902835,
"grad_norm": 0.21138866241289803,
"learning_rate": 6.52239982745638e-06,
"loss": 0.2244,
"step": 941
},
{
"epoch": 3.813765182186235,
"grad_norm": 0.20153285550303834,
"learning_rate": 6.480669124774484e-06,
"loss": 0.2237,
"step": 942
},
{
"epoch": 3.817813765182186,
"grad_norm": 0.21973870925943503,
"learning_rate": 6.439046522302013e-06,
"loss": 0.2293,
"step": 943
},
{
"epoch": 3.8218623481781377,
"grad_norm": 0.22759542460137638,
"learning_rate": 6.397532352852191e-06,
"loss": 0.2146,
"step": 944
},
{
"epoch": 3.8259109311740893,
"grad_norm": 0.21056862503963925,
"learning_rate": 6.356126948371202e-06,
"loss": 0.2127,
"step": 945
},
{
"epoch": 3.8299595141700404,
"grad_norm": 0.20106740056208058,
"learning_rate": 6.314830639935559e-06,
"loss": 0.236,
"step": 946
},
{
"epoch": 3.834008097165992,
"grad_norm": 0.2015063351213087,
"learning_rate": 6.273643757749448e-06,
"loss": 0.2105,
"step": 947
},
{
"epoch": 3.8380566801619436,
"grad_norm": 0.20171959559627137,
"learning_rate": 6.2325666311420856e-06,
"loss": 0.2236,
"step": 948
},
{
"epoch": 3.8421052631578947,
"grad_norm": 0.21242722723345345,
"learning_rate": 6.1915995885650675e-06,
"loss": 0.2166,
"step": 949
},
{
"epoch": 3.8461538461538463,
"grad_norm": 0.20235148884026255,
"learning_rate": 6.150742957589784e-06,
"loss": 0.2223,
"step": 950
},
{
"epoch": 3.850202429149798,
"grad_norm": 0.20243087863525214,
"learning_rate": 6.109997064904774e-06,
"loss": 0.2474,
"step": 951
},
{
"epoch": 3.854251012145749,
"grad_norm": 0.2162579724133153,
"learning_rate": 6.069362236313092e-06,
"loss": 0.1993,
"step": 952
},
{
"epoch": 3.8582995951417005,
"grad_norm": 0.2021237877114475,
"learning_rate": 6.0288387967297565e-06,
"loss": 0.2227,
"step": 953
},
{
"epoch": 3.862348178137652,
"grad_norm": 0.2005935462570149,
"learning_rate": 5.988427070179108e-06,
"loss": 0.2162,
"step": 954
},
{
"epoch": 3.866396761133603,
"grad_norm": 0.20194181841581135,
"learning_rate": 5.948127379792234e-06,
"loss": 0.2248,
"step": 955
},
{
"epoch": 3.8704453441295548,
"grad_norm": 0.20451455715830724,
"learning_rate": 5.90794004780439e-06,
"loss": 0.2239,
"step": 956
},
{
"epoch": 3.8744939271255063,
"grad_norm": 0.1985163066197832,
"learning_rate": 5.867865395552405e-06,
"loss": 0.2288,
"step": 957
},
{
"epoch": 3.8785425101214575,
"grad_norm": 0.20731711562220248,
"learning_rate": 5.8279037434721364e-06,
"loss": 0.2158,
"step": 958
},
{
"epoch": 3.882591093117409,
"grad_norm": 0.2306485719691753,
"learning_rate": 5.788055411095892e-06,
"loss": 0.2179,
"step": 959
},
{
"epoch": 3.8866396761133606,
"grad_norm": 0.21800758389827296,
"learning_rate": 5.7483207170498625e-06,
"loss": 0.2232,
"step": 960
},
{
"epoch": 3.8906882591093117,
"grad_norm": 0.21499269733137424,
"learning_rate": 5.708699979051613e-06,
"loss": 0.2254,
"step": 961
},
{
"epoch": 3.8947368421052633,
"grad_norm": 0.2019838321264202,
"learning_rate": 5.66919351390751e-06,
"loss": 0.2083,
"step": 962
},
{
"epoch": 3.898785425101215,
"grad_norm": 0.23394301493604178,
"learning_rate": 5.629801637510188e-06,
"loss": 0.2214,
"step": 963
},
{
"epoch": 3.902834008097166,
"grad_norm": 0.21168852811710115,
"learning_rate": 5.590524664836043e-06,
"loss": 0.218,
"step": 964
},
{
"epoch": 3.9068825910931175,
"grad_norm": 0.21943031311330885,
"learning_rate": 5.551362909942708e-06,
"loss": 0.2108,
"step": 965
},
{
"epoch": 3.910931174089069,
"grad_norm": 0.2232321631907178,
"learning_rate": 5.51231668596653e-06,
"loss": 0.2145,
"step": 966
},
{
"epoch": 3.91497975708502,
"grad_norm": 0.22795627752124528,
"learning_rate": 5.47338630512007e-06,
"loss": 0.2117,
"step": 967
},
{
"epoch": 3.919028340080972,
"grad_norm": 0.20498812352014678,
"learning_rate": 5.434572078689617e-06,
"loss": 0.2362,
"step": 968
},
{
"epoch": 3.9230769230769234,
"grad_norm": 0.2396624862376877,
"learning_rate": 5.395874317032694e-06,
"loss": 0.2192,
"step": 969
},
{
"epoch": 3.9271255060728745,
"grad_norm": 0.1987259420655526,
"learning_rate": 5.357293329575573e-06,
"loss": 0.2139,
"step": 970
},
{
"epoch": 3.931174089068826,
"grad_norm": 0.2006873187623747,
"learning_rate": 5.318829424810797e-06,
"loss": 0.2254,
"step": 971
},
{
"epoch": 3.9352226720647776,
"grad_norm": 0.20142901351153442,
"learning_rate": 5.280482910294728e-06,
"loss": 0.2279,
"step": 972
},
{
"epoch": 3.9392712550607287,
"grad_norm": 0.22147524104784244,
"learning_rate": 5.242254092645078e-06,
"loss": 0.2196,
"step": 973
},
{
"epoch": 3.9433198380566803,
"grad_norm": 0.2165915535212695,
"learning_rate": 5.2041432775384445e-06,
"loss": 0.2172,
"step": 974
},
{
"epoch": 3.9473684210526314,
"grad_norm": 0.2085753586805864,
"learning_rate": 5.166150769707892e-06,
"loss": 0.2076,
"step": 975
},
{
"epoch": 3.951417004048583,
"grad_norm": 0.24598433592774913,
"learning_rate": 5.128276872940503e-06,
"loss": 0.2282,
"step": 976
},
{
"epoch": 3.9554655870445345,
"grad_norm": 0.21186844780170588,
"learning_rate": 5.090521890074953e-06,
"loss": 0.2228,
"step": 977
},
{
"epoch": 3.9595141700404857,
"grad_norm": 0.20661569107934433,
"learning_rate": 5.0528861229990675e-06,
"loss": 0.2031,
"step": 978
},
{
"epoch": 3.9635627530364372,
"grad_norm": 0.2140295996445021,
"learning_rate": 5.015369872647442e-06,
"loss": 0.215,
"step": 979
},
{
"epoch": 3.967611336032389,
"grad_norm": 0.21100310228394412,
"learning_rate": 4.9779734389990195e-06,
"loss": 0.2292,
"step": 980
},
{
"epoch": 3.97165991902834,
"grad_norm": 0.20103235584104573,
"learning_rate": 4.940697121074675e-06,
"loss": 0.2162,
"step": 981
},
{
"epoch": 3.9757085020242915,
"grad_norm": 0.19451098009911247,
"learning_rate": 4.903541216934864e-06,
"loss": 0.2053,
"step": 982
},
{
"epoch": 3.979757085020243,
"grad_norm": 0.20613633322524433,
"learning_rate": 4.8665060236772e-06,
"loss": 0.2249,
"step": 983
},
{
"epoch": 3.983805668016194,
"grad_norm": 0.19269834992800397,
"learning_rate": 4.829591837434102e-06,
"loss": 0.2101,
"step": 984
},
{
"epoch": 3.9878542510121457,
"grad_norm": 0.19003977582573672,
"learning_rate": 4.792798953370424e-06,
"loss": 0.2233,
"step": 985
},
{
"epoch": 3.9919028340080973,
"grad_norm": 0.19349158813970999,
"learning_rate": 4.756127665681083e-06,
"loss": 0.2102,
"step": 986
},
{
"epoch": 3.9959514170040484,
"grad_norm": 0.2012353694173147,
"learning_rate": 4.719578267588725e-06,
"loss": 0.2284,
"step": 987
},
{
"epoch": 4.0,
"grad_norm": 0.20245676534702506,
"learning_rate": 4.6831510513413525e-06,
"loss": 0.2027,
"step": 988
},
{
"epoch": 4.004048582995951,
"grad_norm": 0.3805279810265701,
"learning_rate": 4.6468463082100176e-06,
"loss": 0.1785,
"step": 989
},
{
"epoch": 4.008097165991903,
"grad_norm": 0.2542549927993061,
"learning_rate": 4.610664328486485e-06,
"loss": 0.177,
"step": 990
},
{
"epoch": 4.012145748987854,
"grad_norm": 0.32989097795666494,
"learning_rate": 4.5746054014809e-06,
"loss": 0.1806,
"step": 991
},
{
"epoch": 4.016194331983805,
"grad_norm": 0.3576743076752589,
"learning_rate": 4.5386698155194766e-06,
"loss": 0.1784,
"step": 992
},
{
"epoch": 4.020242914979757,
"grad_norm": 0.2335192050613002,
"learning_rate": 4.502857857942204e-06,
"loss": 0.1659,
"step": 993
},
{
"epoch": 4.0242914979757085,
"grad_norm": 0.28177406396130755,
"learning_rate": 4.467169815100549e-06,
"loss": 0.1701,
"step": 994
},
{
"epoch": 4.02834008097166,
"grad_norm": 0.3262144514045581,
"learning_rate": 4.431605972355137e-06,
"loss": 0.1819,
"step": 995
},
{
"epoch": 4.032388663967612,
"grad_norm": 0.24566423030453793,
"learning_rate": 4.396166614073516e-06,
"loss": 0.1609,
"step": 996
},
{
"epoch": 4.036437246963563,
"grad_norm": 0.24918499077637285,
"learning_rate": 4.360852023627846e-06,
"loss": 0.1816,
"step": 997
},
{
"epoch": 4.040485829959514,
"grad_norm": 0.3368458973168393,
"learning_rate": 4.325662483392656e-06,
"loss": 0.1858,
"step": 998
},
{
"epoch": 4.044534412955466,
"grad_norm": 0.2519230351226579,
"learning_rate": 4.29059827474257e-06,
"loss": 0.1831,
"step": 999
},
{
"epoch": 4.048582995951417,
"grad_norm": 0.21976884054301227,
"learning_rate": 4.255659678050063e-06,
"loss": 0.1954,
"step": 1000
},
{
"epoch": 4.052631578947368,
"grad_norm": 0.24742239590129428,
"learning_rate": 4.22084697268323e-06,
"loss": 0.1787,
"step": 1001
},
{
"epoch": 4.05668016194332,
"grad_norm": 0.28071244197774875,
"learning_rate": 4.186160437003535e-06,
"loss": 0.1892,
"step": 1002
},
{
"epoch": 4.060728744939271,
"grad_norm": 0.21762129360108007,
"learning_rate": 4.151600348363582e-06,
"loss": 0.1711,
"step": 1003
},
{
"epoch": 4.064777327935222,
"grad_norm": 0.20310084165773692,
"learning_rate": 4.11716698310493e-06,
"loss": 0.1784,
"step": 1004
},
{
"epoch": 4.068825910931174,
"grad_norm": 0.24542505328337724,
"learning_rate": 4.082860616555848e-06,
"loss": 0.1744,
"step": 1005
},
{
"epoch": 4.0728744939271255,
"grad_norm": 0.266028695835179,
"learning_rate": 4.048681523029125e-06,
"loss": 0.1725,
"step": 1006
},
{
"epoch": 4.076923076923077,
"grad_norm": 0.23350828724640674,
"learning_rate": 4.014629975819882e-06,
"loss": 0.1864,
"step": 1007
},
{
"epoch": 4.080971659919029,
"grad_norm": 0.20114070782836574,
"learning_rate": 3.9807062472033855e-06,
"loss": 0.1788,
"step": 1008
},
{
"epoch": 4.08502024291498,
"grad_norm": 0.23622893954792348,
"learning_rate": 3.94691060843287e-06,
"loss": 0.1809,
"step": 1009
},
{
"epoch": 4.089068825910931,
"grad_norm": 0.2238088113073252,
"learning_rate": 3.913243329737353e-06,
"loss": 0.1805,
"step": 1010
},
{
"epoch": 4.093117408906883,
"grad_norm": 0.22091581732255736,
"learning_rate": 3.879704680319503e-06,
"loss": 0.1871,
"step": 1011
},
{
"epoch": 4.097165991902834,
"grad_norm": 0.1949051052503906,
"learning_rate": 3.846294928353471e-06,
"loss": 0.1695,
"step": 1012
},
{
"epoch": 4.101214574898785,
"grad_norm": 0.21045228866154955,
"learning_rate": 3.8130143409827345e-06,
"loss": 0.1802,
"step": 1013
},
{
"epoch": 4.105263157894737,
"grad_norm": 0.23721709478381572,
"learning_rate": 3.7798631843179887e-06,
"loss": 0.1842,
"step": 1014
},
{
"epoch": 4.109311740890688,
"grad_norm": 0.19982446544943194,
"learning_rate": 3.7468417234349973e-06,
"loss": 0.1683,
"step": 1015
},
{
"epoch": 4.113360323886639,
"grad_norm": 0.188487961909984,
"learning_rate": 3.7139502223724823e-06,
"loss": 0.1765,
"step": 1016
},
{
"epoch": 4.117408906882591,
"grad_norm": 0.20147913659424524,
"learning_rate": 3.6811889441300095e-06,
"loss": 0.1641,
"step": 1017
},
{
"epoch": 4.1214574898785425,
"grad_norm": 0.2036328093466251,
"learning_rate": 3.648558150665886e-06,
"loss": 0.1798,
"step": 1018
},
{
"epoch": 4.125506072874494,
"grad_norm": 0.20321624793912244,
"learning_rate": 3.616058102895075e-06,
"loss": 0.1758,
"step": 1019
},
{
"epoch": 4.129554655870446,
"grad_norm": 0.1967763531501023,
"learning_rate": 3.583689060687079e-06,
"loss": 0.1781,
"step": 1020
},
{
"epoch": 4.133603238866397,
"grad_norm": 0.2122105663744907,
"learning_rate": 3.5514512828639023e-06,
"loss": 0.1791,
"step": 1021
},
{
"epoch": 4.137651821862348,
"grad_norm": 0.21182398985264808,
"learning_rate": 3.51934502719796e-06,
"loss": 0.1757,
"step": 1022
},
{
"epoch": 4.1417004048583,
"grad_norm": 0.18438875320098533,
"learning_rate": 3.4873705504100185e-06,
"loss": 0.1785,
"step": 1023
},
{
"epoch": 4.145748987854251,
"grad_norm": 0.2035320622437151,
"learning_rate": 3.4555281081671345e-06,
"loss": 0.1828,
"step": 1024
},
{
"epoch": 4.149797570850202,
"grad_norm": 0.2002698080471503,
"learning_rate": 3.4238179550806326e-06,
"loss": 0.1824,
"step": 1025
},
{
"epoch": 4.153846153846154,
"grad_norm": 0.1877872008521769,
"learning_rate": 3.3922403447040497e-06,
"loss": 0.1735,
"step": 1026
},
{
"epoch": 4.157894736842105,
"grad_norm": 0.18870468772910423,
"learning_rate": 3.3607955295311247e-06,
"loss": 0.1813,
"step": 1027
},
{
"epoch": 4.161943319838056,
"grad_norm": 0.18702332196046578,
"learning_rate": 3.32948376099375e-06,
"loss": 0.1772,
"step": 1028
},
{
"epoch": 4.165991902834008,
"grad_norm": 0.1975590173596765,
"learning_rate": 3.2983052894599976e-06,
"loss": 0.1753,
"step": 1029
},
{
"epoch": 4.17004048582996,
"grad_norm": 0.18281936574417554,
"learning_rate": 3.2672603642320945e-06,
"loss": 0.1796,
"step": 1030
},
{
"epoch": 4.174089068825911,
"grad_norm": 0.18498855531103095,
"learning_rate": 3.236349233544438e-06,
"loss": 0.1934,
"step": 1031
},
{
"epoch": 4.178137651821863,
"grad_norm": 0.18642912855421437,
"learning_rate": 3.205572144561604e-06,
"loss": 0.1783,
"step": 1032
},
{
"epoch": 4.182186234817814,
"grad_norm": 0.19013639558694698,
"learning_rate": 3.174929343376374e-06,
"loss": 0.1826,
"step": 1033
},
{
"epoch": 4.186234817813765,
"grad_norm": 0.18808986404436284,
"learning_rate": 3.1444210750077777e-06,
"loss": 0.1777,
"step": 1034
},
{
"epoch": 4.190283400809717,
"grad_norm": 0.19335126007188913,
"learning_rate": 3.1140475833991046e-06,
"loss": 0.1883,
"step": 1035
},
{
"epoch": 4.194331983805668,
"grad_norm": 0.18494009933586872,
"learning_rate": 3.0838091114159895e-06,
"loss": 0.1746,
"step": 1036
},
{
"epoch": 4.198380566801619,
"grad_norm": 0.18974753571730718,
"learning_rate": 3.0537059008444524e-06,
"loss": 0.164,
"step": 1037
},
{
"epoch": 4.202429149797571,
"grad_norm": 0.1836400984113335,
"learning_rate": 3.0237381923889563e-06,
"loss": 0.1729,
"step": 1038
},
{
"epoch": 4.206477732793522,
"grad_norm": 0.19245040158769308,
"learning_rate": 2.9939062256704997e-06,
"loss": 0.1758,
"step": 1039
},
{
"epoch": 4.2105263157894735,
"grad_norm": 0.19085384118154256,
"learning_rate": 2.964210239224696e-06,
"loss": 0.1814,
"step": 1040
},
{
"epoch": 4.2145748987854255,
"grad_norm": 0.18275225930588943,
"learning_rate": 2.934650470499869e-06,
"loss": 0.1809,
"step": 1041
},
{
"epoch": 4.218623481781377,
"grad_norm": 0.1924355608966222,
"learning_rate": 2.905227155855128e-06,
"loss": 0.1844,
"step": 1042
},
{
"epoch": 4.222672064777328,
"grad_norm": 0.1884150360400434,
"learning_rate": 2.875940530558519e-06,
"loss": 0.1811,
"step": 1043
},
{
"epoch": 4.22672064777328,
"grad_norm": 0.1864640492880474,
"learning_rate": 2.8467908287851154e-06,
"loss": 0.1643,
"step": 1044
},
{
"epoch": 4.230769230769231,
"grad_norm": 0.18871697513533084,
"learning_rate": 2.8177782836151603e-06,
"loss": 0.1815,
"step": 1045
},
{
"epoch": 4.234817813765182,
"grad_norm": 0.18594381063049512,
"learning_rate": 2.7889031270321874e-06,
"loss": 0.1839,
"step": 1046
},
{
"epoch": 4.238866396761134,
"grad_norm": 0.18701692681512697,
"learning_rate": 2.760165589921184e-06,
"loss": 0.1844,
"step": 1047
},
{
"epoch": 4.242914979757085,
"grad_norm": 0.18924675488120435,
"learning_rate": 2.7315659020667305e-06,
"loss": 0.1826,
"step": 1048
},
{
"epoch": 4.246963562753036,
"grad_norm": 0.1904244088063013,
"learning_rate": 2.703104292151164e-06,
"loss": 0.1782,
"step": 1049
},
{
"epoch": 4.251012145748988,
"grad_norm": 0.20208336490891105,
"learning_rate": 2.6747809877527607e-06,
"loss": 0.1838,
"step": 1050
},
{
"epoch": 4.255060728744939,
"grad_norm": 0.19888617524283803,
"learning_rate": 2.6465962153439082e-06,
"loss": 0.1795,
"step": 1051
},
{
"epoch": 4.2591093117408905,
"grad_norm": 0.18957619791105412,
"learning_rate": 2.618550200289303e-06,
"loss": 0.1797,
"step": 1052
},
{
"epoch": 4.2631578947368425,
"grad_norm": 0.18759240276954775,
"learning_rate": 2.59064316684412e-06,
"loss": 0.1733,
"step": 1053
},
{
"epoch": 4.267206477732794,
"grad_norm": 0.194636661058885,
"learning_rate": 2.562875338152264e-06,
"loss": 0.1768,
"step": 1054
},
{
"epoch": 4.271255060728745,
"grad_norm": 0.18125209735488826,
"learning_rate": 2.5352469362445552e-06,
"loss": 0.1923,
"step": 1055
},
{
"epoch": 4.275303643724697,
"grad_norm": 0.17675310887641366,
"learning_rate": 2.5077581820369523e-06,
"loss": 0.1694,
"step": 1056
},
{
"epoch": 4.279352226720648,
"grad_norm": 0.18342081278263273,
"learning_rate": 2.480409295328803e-06,
"loss": 0.1801,
"step": 1057
},
{
"epoch": 4.283400809716599,
"grad_norm": 0.18577098276034135,
"learning_rate": 2.45320049480108e-06,
"loss": 0.1676,
"step": 1058
},
{
"epoch": 4.287449392712551,
"grad_norm": 0.18475371446107502,
"learning_rate": 2.4261319980146293e-06,
"loss": 0.1729,
"step": 1059
},
{
"epoch": 4.291497975708502,
"grad_norm": 0.19998931294937475,
"learning_rate": 2.399204021408421e-06,
"loss": 0.1753,
"step": 1060
},
{
"epoch": 4.295546558704453,
"grad_norm": 0.18282606089397432,
"learning_rate": 2.3724167802978395e-06,
"loss": 0.1776,
"step": 1061
},
{
"epoch": 4.299595141700405,
"grad_norm": 0.18234362027623108,
"learning_rate": 2.3457704888729625e-06,
"loss": 0.1734,
"step": 1062
},
{
"epoch": 4.303643724696356,
"grad_norm": 0.18520892429869637,
"learning_rate": 2.3192653601968162e-06,
"loss": 0.1804,
"step": 1063
},
{
"epoch": 4.3076923076923075,
"grad_norm": 0.18667954262361397,
"learning_rate": 2.2929016062037057e-06,
"loss": 0.1773,
"step": 1064
},
{
"epoch": 4.3117408906882595,
"grad_norm": 0.17869590645989336,
"learning_rate": 2.2666794376975055e-06,
"loss": 0.1796,
"step": 1065
},
{
"epoch": 4.315789473684211,
"grad_norm": 0.17849529180204968,
"learning_rate": 2.24059906434998e-06,
"loss": 0.17,
"step": 1066
},
{
"epoch": 4.319838056680162,
"grad_norm": 0.17514166304620019,
"learning_rate": 2.214660694699091e-06,
"loss": 0.1806,
"step": 1067
},
{
"epoch": 4.323886639676114,
"grad_norm": 0.19018893751635205,
"learning_rate": 2.188864536147355e-06,
"loss": 0.1807,
"step": 1068
},
{
"epoch": 4.327935222672065,
"grad_norm": 0.2011330296947298,
"learning_rate": 2.1632107949601757e-06,
"loss": 0.1943,
"step": 1069
},
{
"epoch": 4.331983805668016,
"grad_norm": 0.19104106857335543,
"learning_rate": 2.137699676264171e-06,
"loss": 0.1668,
"step": 1070
},
{
"epoch": 4.336032388663968,
"grad_norm": 0.1882615446958739,
"learning_rate": 2.1123313840455738e-06,
"loss": 0.1912,
"step": 1071
},
{
"epoch": 4.340080971659919,
"grad_norm": 0.1862119682088457,
"learning_rate": 2.08710612114857e-06,
"loss": 0.1851,
"step": 1072
},
{
"epoch": 4.34412955465587,
"grad_norm": 0.18883639368064845,
"learning_rate": 2.0620240892736953e-06,
"loss": 0.1803,
"step": 1073
},
{
"epoch": 4.348178137651822,
"grad_norm": 0.18667351033396445,
"learning_rate": 2.0370854889762e-06,
"loss": 0.1795,
"step": 1074
},
{
"epoch": 4.352226720647773,
"grad_norm": 0.17901434293534668,
"learning_rate": 2.0122905196644726e-06,
"loss": 0.1704,
"step": 1075
},
{
"epoch": 4.3562753036437245,
"grad_norm": 0.18381652739651944,
"learning_rate": 1.987639379598427e-06,
"loss": 0.1727,
"step": 1076
},
{
"epoch": 4.3603238866396765,
"grad_norm": 0.19102818180919645,
"learning_rate": 1.963132265887919e-06,
"loss": 0.1884,
"step": 1077
},
{
"epoch": 4.364372469635628,
"grad_norm": 0.18192966557185974,
"learning_rate": 1.9387693744911784e-06,
"loss": 0.1725,
"step": 1078
},
{
"epoch": 4.368421052631579,
"grad_norm": 0.18805041347020784,
"learning_rate": 1.9145509002132368e-06,
"loss": 0.1843,
"step": 1079
},
{
"epoch": 4.372469635627531,
"grad_norm": 0.1884941299320822,
"learning_rate": 1.8904770367043723e-06,
"loss": 0.1592,
"step": 1080
},
{
"epoch": 4.376518218623482,
"grad_norm": 0.192897776514045,
"learning_rate": 1.8665479764585504e-06,
"loss": 0.1837,
"step": 1081
},
{
"epoch": 4.380566801619433,
"grad_norm": 0.19293622380340378,
"learning_rate": 1.8427639108118977e-06,
"loss": 0.1896,
"step": 1082
},
{
"epoch": 4.384615384615385,
"grad_norm": 0.2004471967681598,
"learning_rate": 1.8191250299411755e-06,
"loss": 0.1809,
"step": 1083
},
{
"epoch": 4.388663967611336,
"grad_norm": 0.1855240515716235,
"learning_rate": 1.7956315228622446e-06,
"loss": 0.1631,
"step": 1084
},
{
"epoch": 4.392712550607287,
"grad_norm": 0.18483933426886037,
"learning_rate": 1.7722835774285553e-06,
"loss": 0.1764,
"step": 1085
},
{
"epoch": 4.396761133603239,
"grad_norm": 0.18609717998922395,
"learning_rate": 1.7490813803296624e-06,
"loss": 0.1749,
"step": 1086
},
{
"epoch": 4.40080971659919,
"grad_norm": 0.18804517427729464,
"learning_rate": 1.726025117089718e-06,
"loss": 0.1867,
"step": 1087
},
{
"epoch": 4.4048582995951415,
"grad_norm": 0.1864953085592189,
"learning_rate": 1.70311497206598e-06,
"loss": 0.1703,
"step": 1088
},
{
"epoch": 4.4089068825910935,
"grad_norm": 0.21581087751434702,
"learning_rate": 1.6803511284473638e-06,
"loss": 0.1779,
"step": 1089
},
{
"epoch": 4.412955465587045,
"grad_norm": 0.18363970501675858,
"learning_rate": 1.6577337682529538e-06,
"loss": 0.1821,
"step": 1090
},
{
"epoch": 4.417004048582996,
"grad_norm": 0.17691623532442965,
"learning_rate": 1.6352630723305617e-06,
"loss": 0.1674,
"step": 1091
},
{
"epoch": 4.421052631578947,
"grad_norm": 0.18890199874157867,
"learning_rate": 1.612939220355272e-06,
"loss": 0.1749,
"step": 1092
},
{
"epoch": 4.425101214574899,
"grad_norm": 0.18630612937065502,
"learning_rate": 1.5907623908280112e-06,
"loss": 0.1775,
"step": 1093
},
{
"epoch": 4.42914979757085,
"grad_norm": 0.17951500042052068,
"learning_rate": 1.568732761074121e-06,
"loss": 0.1742,
"step": 1094
},
{
"epoch": 4.433198380566802,
"grad_norm": 0.19130973167916618,
"learning_rate": 1.5468505072419237e-06,
"loss": 0.1792,
"step": 1095
},
{
"epoch": 4.437246963562753,
"grad_norm": 0.18293419842307776,
"learning_rate": 1.5251158043013448e-06,
"loss": 0.1882,
"step": 1096
},
{
"epoch": 4.441295546558704,
"grad_norm": 0.19175762843272368,
"learning_rate": 1.5035288260424885e-06,
"loss": 0.1823,
"step": 1097
},
{
"epoch": 4.445344129554655,
"grad_norm": 0.1825266004250967,
"learning_rate": 1.4820897450742623e-06,
"loss": 0.177,
"step": 1098
},
{
"epoch": 4.449392712550607,
"grad_norm": 0.1891359360806475,
"learning_rate": 1.4607987328229767e-06,
"loss": 0.1865,
"step": 1099
},
{
"epoch": 4.4534412955465585,
"grad_norm": 0.1892500629410147,
"learning_rate": 1.4396559595310056e-06,
"loss": 0.178,
"step": 1100
},
{
"epoch": 4.4574898785425106,
"grad_norm": 0.18077787992945557,
"learning_rate": 1.4186615942553973e-06,
"loss": 0.1755,
"step": 1101
},
{
"epoch": 4.461538461538462,
"grad_norm": 0.18101151732245432,
"learning_rate": 1.3978158048665424e-06,
"loss": 0.174,
"step": 1102
},
{
"epoch": 4.465587044534413,
"grad_norm": 0.18908375034885072,
"learning_rate": 1.3771187580468115e-06,
"loss": 0.1796,
"step": 1103
},
{
"epoch": 4.469635627530364,
"grad_norm": 0.19551150592976638,
"learning_rate": 1.3565706192892392e-06,
"loss": 0.188,
"step": 1104
},
{
"epoch": 4.473684210526316,
"grad_norm": 0.18261242641534828,
"learning_rate": 1.3361715528961993e-06,
"loss": 0.1728,
"step": 1105
},
{
"epoch": 4.477732793522267,
"grad_norm": 0.18436494952887497,
"learning_rate": 1.3159217219780817e-06,
"loss": 0.1877,
"step": 1106
},
{
"epoch": 4.481781376518219,
"grad_norm": 0.1869511848401393,
"learning_rate": 1.2958212884519949e-06,
"loss": 0.1776,
"step": 1107
},
{
"epoch": 4.48582995951417,
"grad_norm": 0.17501589286343822,
"learning_rate": 1.2758704130404675e-06,
"loss": 0.1842,
"step": 1108
},
{
"epoch": 4.489878542510121,
"grad_norm": 0.17681274796909935,
"learning_rate": 1.2560692552701738e-06,
"loss": 0.1946,
"step": 1109
},
{
"epoch": 4.493927125506072,
"grad_norm": 0.17972341307183184,
"learning_rate": 1.2364179734706382e-06,
"loss": 0.1675,
"step": 1110
},
{
"epoch": 4.497975708502024,
"grad_norm": 0.1983003964994461,
"learning_rate": 1.2169167247729852e-06,
"loss": 0.1726,
"step": 1111
},
{
"epoch": 4.502024291497976,
"grad_norm": 0.17803963447565055,
"learning_rate": 1.1975656651086864e-06,
"loss": 0.1756,
"step": 1112
},
{
"epoch": 4.506072874493928,
"grad_norm": 0.18015974116553807,
"learning_rate": 1.1783649492082926e-06,
"loss": 0.1801,
"step": 1113
},
{
"epoch": 4.510121457489879,
"grad_norm": 0.17958146801057018,
"learning_rate": 1.1593147306002183e-06,
"loss": 0.1736,
"step": 1114
},
{
"epoch": 4.51417004048583,
"grad_norm": 0.19599181758847362,
"learning_rate": 1.140415161609505e-06,
"loss": 0.1733,
"step": 1115
},
{
"epoch": 4.518218623481781,
"grad_norm": 0.17307857733055168,
"learning_rate": 1.121666393356604e-06,
"loss": 0.1779,
"step": 1116
},
{
"epoch": 4.522267206477733,
"grad_norm": 0.1845030637068005,
"learning_rate": 1.1030685757561632e-06,
"loss": 0.183,
"step": 1117
},
{
"epoch": 4.526315789473684,
"grad_norm": 0.18514338364174698,
"learning_rate": 1.0846218575158373e-06,
"loss": 0.188,
"step": 1118
},
{
"epoch": 4.530364372469636,
"grad_norm": 0.18600824739028632,
"learning_rate": 1.066326386135097e-06,
"loss": 0.1878,
"step": 1119
},
{
"epoch": 4.534412955465587,
"grad_norm": 0.17867997830431898,
"learning_rate": 1.0481823079040398e-06,
"loss": 0.1833,
"step": 1120
},
{
"epoch": 4.538461538461538,
"grad_norm": 0.18396837666060992,
"learning_rate": 1.0301897679022344e-06,
"loss": 0.1768,
"step": 1121
},
{
"epoch": 4.5425101214574894,
"grad_norm": 0.17500851622567573,
"learning_rate": 1.0123489099975491e-06,
"loss": 0.1836,
"step": 1122
},
{
"epoch": 4.5465587044534415,
"grad_norm": 0.17318102481905628,
"learning_rate": 9.9465987684501e-07,
"loss": 0.176,
"step": 1123
},
{
"epoch": 4.550607287449393,
"grad_norm": 0.18617888395063678,
"learning_rate": 9.771228098856534e-07,
"loss": 0.1834,
"step": 1124
},
{
"epoch": 4.554655870445345,
"grad_norm": 0.1834412787058778,
"learning_rate": 9.597378493454013e-07,
"loss": 0.1772,
"step": 1125
},
{
"epoch": 4.558704453441296,
"grad_norm": 0.17732660501682504,
"learning_rate": 9.425051342339353e-07,
"loss": 0.1869,
"step": 1126
},
{
"epoch": 4.562753036437247,
"grad_norm": 0.18437657562766652,
"learning_rate": 9.254248023435864e-07,
"loss": 0.1839,
"step": 1127
},
{
"epoch": 4.566801619433198,
"grad_norm": 0.1741345158332422,
"learning_rate": 9.084969902482288e-07,
"loss": 0.1869,
"step": 1128
},
{
"epoch": 4.57085020242915,
"grad_norm": 0.18051865150940868,
"learning_rate": 8.917218333021993e-07,
"loss": 0.1702,
"step": 1129
},
{
"epoch": 4.574898785425101,
"grad_norm": 0.17045442349875015,
"learning_rate": 8.750994656392109e-07,
"loss": 0.1629,
"step": 1130
},
{
"epoch": 4.578947368421053,
"grad_norm": 0.17568973499520812,
"learning_rate": 8.58630020171265e-07,
"loss": 0.184,
"step": 1131
},
{
"epoch": 4.582995951417004,
"grad_norm": 0.19243619976476983,
"learning_rate": 8.423136285876099e-07,
"loss": 0.1774,
"step": 1132
},
{
"epoch": 4.587044534412955,
"grad_norm": 0.17992139944141713,
"learning_rate": 8.26150421353682e-07,
"loss": 0.1776,
"step": 1133
},
{
"epoch": 4.5910931174089065,
"grad_norm": 0.2007801681124257,
"learning_rate": 8.101405277100549e-07,
"loss": 0.177,
"step": 1134
},
{
"epoch": 4.5951417004048585,
"grad_norm": 0.18121529234022477,
"learning_rate": 7.942840756714077e-07,
"loss": 0.1861,
"step": 1135
},
{
"epoch": 4.59919028340081,
"grad_norm": 0.18310842041940217,
"learning_rate": 7.785811920255093e-07,
"loss": 0.1786,
"step": 1136
},
{
"epoch": 4.603238866396762,
"grad_norm": 0.18236467198555836,
"learning_rate": 7.630320023322024e-07,
"loss": 0.1598,
"step": 1137
},
{
"epoch": 4.607287449392713,
"grad_norm": 0.18428463771804515,
"learning_rate": 7.476366309223903e-07,
"loss": 0.1741,
"step": 1138
},
{
"epoch": 4.611336032388664,
"grad_norm": 0.1846780803932637,
"learning_rate": 7.323952008970537e-07,
"loss": 0.1706,
"step": 1139
},
{
"epoch": 4.615384615384615,
"grad_norm": 0.17812811524449382,
"learning_rate": 7.173078341262618e-07,
"loss": 0.1825,
"step": 1140
},
{
"epoch": 4.619433198380567,
"grad_norm": 0.18783652811427481,
"learning_rate": 7.023746512481989e-07,
"loss": 0.1836,
"step": 1141
},
{
"epoch": 4.623481781376518,
"grad_norm": 0.1927838535365029,
"learning_rate": 6.875957716681902e-07,
"loss": 0.1587,
"step": 1142
},
{
"epoch": 4.62753036437247,
"grad_norm": 0.18656691979849319,
"learning_rate": 6.729713135577687e-07,
"loss": 0.1645,
"step": 1143
},
{
"epoch": 4.631578947368421,
"grad_norm": 0.18199113808767528,
"learning_rate": 6.585013938537143e-07,
"loss": 0.1867,
"step": 1144
},
{
"epoch": 4.635627530364372,
"grad_norm": 0.17775273278607032,
"learning_rate": 6.441861282571115e-07,
"loss": 0.162,
"step": 1145
},
{
"epoch": 4.6396761133603235,
"grad_norm": 0.18357787317196597,
"learning_rate": 6.300256312324471e-07,
"loss": 0.1772,
"step": 1146
},
{
"epoch": 4.6437246963562755,
"grad_norm": 0.19218806805098668,
"learning_rate": 6.160200160066776e-07,
"loss": 0.176,
"step": 1147
},
{
"epoch": 4.647773279352227,
"grad_norm": 0.17449484834540205,
"learning_rate": 6.021693945683327e-07,
"loss": 0.1691,
"step": 1148
},
{
"epoch": 4.651821862348179,
"grad_norm": 0.19410764077979917,
"learning_rate": 5.884738776666044e-07,
"loss": 0.1708,
"step": 1149
},
{
"epoch": 4.65587044534413,
"grad_norm": 0.1859712066057742,
"learning_rate": 5.749335748104878e-07,
"loss": 0.1775,
"step": 1150
},
{
"epoch": 4.659919028340081,
"grad_norm": 0.1795955571240998,
"learning_rate": 5.615485942678778e-07,
"loss": 0.1835,
"step": 1151
},
{
"epoch": 4.663967611336032,
"grad_norm": 0.18108769809247074,
"learning_rate": 5.483190430647289e-07,
"loss": 0.1599,
"step": 1152
},
{
"epoch": 4.668016194331984,
"grad_norm": 0.16947687545383314,
"learning_rate": 5.352450269841747e-07,
"loss": 0.1661,
"step": 1153
},
{
"epoch": 4.672064777327935,
"grad_norm": 0.16973774205982853,
"learning_rate": 5.22326650565701e-07,
"loss": 0.1724,
"step": 1154
},
{
"epoch": 4.676113360323887,
"grad_norm": 0.18497133772518962,
"learning_rate": 5.095640171043048e-07,
"loss": 0.1642,
"step": 1155
},
{
"epoch": 4.680161943319838,
"grad_norm": 0.17876049855550102,
"learning_rate": 4.969572286496594e-07,
"loss": 0.18,
"step": 1156
},
{
"epoch": 4.684210526315789,
"grad_norm": 0.1796090995261387,
"learning_rate": 4.845063860053079e-07,
"loss": 0.1778,
"step": 1157
},
{
"epoch": 4.6882591093117405,
"grad_norm": 0.18463654696691986,
"learning_rate": 4.7221158872785997e-07,
"loss": 0.1777,
"step": 1158
},
{
"epoch": 4.6923076923076925,
"grad_norm": 0.18383825429098924,
"learning_rate": 4.600729351261857e-07,
"loss": 0.1831,
"step": 1159
},
{
"epoch": 4.696356275303644,
"grad_norm": 0.18712914261511485,
"learning_rate": 4.4809052226063355e-07,
"loss": 0.1743,
"step": 1160
},
{
"epoch": 4.700404858299595,
"grad_norm": 0.17156376031692427,
"learning_rate": 4.3626444594226046e-07,
"loss": 0.1836,
"step": 1161
},
{
"epoch": 4.704453441295547,
"grad_norm": 0.18940290677015248,
"learning_rate": 4.245948007320588e-07,
"loss": 0.1724,
"step": 1162
},
{
"epoch": 4.708502024291498,
"grad_norm": 0.17758038296674122,
"learning_rate": 4.130816799401971e-07,
"loss": 0.1642,
"step": 1163
},
{
"epoch": 4.712550607287449,
"grad_norm": 0.18558495154184917,
"learning_rate": 4.017251756252827e-07,
"loss": 0.175,
"step": 1164
},
{
"epoch": 4.716599190283401,
"grad_norm": 0.17460952762273768,
"learning_rate": 3.9052537859362473e-07,
"loss": 0.1841,
"step": 1165
},
{
"epoch": 4.720647773279352,
"grad_norm": 0.1779047135634264,
"learning_rate": 3.7948237839849687e-07,
"loss": 0.1924,
"step": 1166
},
{
"epoch": 4.724696356275303,
"grad_norm": 0.17606334478111846,
"learning_rate": 3.6859626333943797e-07,
"loss": 0.1801,
"step": 1167
},
{
"epoch": 4.728744939271255,
"grad_norm": 0.18280246038392914,
"learning_rate": 3.578671204615347e-07,
"loss": 0.1717,
"step": 1168
},
{
"epoch": 4.732793522267206,
"grad_norm": 0.17597630406119336,
"learning_rate": 3.472950355547267e-07,
"loss": 0.1694,
"step": 1169
},
{
"epoch": 4.7368421052631575,
"grad_norm": 0.18073160652127168,
"learning_rate": 3.3688009315312244e-07,
"loss": 0.1821,
"step": 1170
},
{
"epoch": 4.7408906882591095,
"grad_norm": 0.1887391647419256,
"learning_rate": 3.266223765343224e-07,
"loss": 0.157,
"step": 1171
},
{
"epoch": 4.744939271255061,
"grad_norm": 0.17787068055124014,
"learning_rate": 3.16521967718757e-07,
"loss": 0.1788,
"step": 1172
},
{
"epoch": 4.748987854251012,
"grad_norm": 0.1774374479492718,
"learning_rate": 3.0657894746902726e-07,
"loss": 0.1814,
"step": 1173
},
{
"epoch": 4.753036437246964,
"grad_norm": 0.1681367740159975,
"learning_rate": 2.9679339528925654e-07,
"loss": 0.1725,
"step": 1174
},
{
"epoch": 4.757085020242915,
"grad_norm": 0.1785289846518124,
"learning_rate": 2.871653894244597e-07,
"loss": 0.1717,
"step": 1175
},
{
"epoch": 4.761133603238866,
"grad_norm": 0.18345963725195288,
"learning_rate": 2.7769500685991266e-07,
"loss": 0.1774,
"step": 1176
},
{
"epoch": 4.765182186234818,
"grad_norm": 0.17675557483152973,
"learning_rate": 2.683823233205485e-07,
"loss": 0.1826,
"step": 1177
},
{
"epoch": 4.769230769230769,
"grad_norm": 0.1838159963955485,
"learning_rate": 2.592274132703287e-07,
"loss": 0.1777,
"step": 1178
},
{
"epoch": 4.77327935222672,
"grad_norm": 0.18448115907575743,
"learning_rate": 2.502303499116754e-07,
"loss": 0.1755,
"step": 1179
},
{
"epoch": 4.777327935222672,
"grad_norm": 0.17417676408771857,
"learning_rate": 2.413912051848666e-07,
"loss": 0.19,
"step": 1180
},
{
"epoch": 4.781376518218623,
"grad_norm": 0.18425468967913183,
"learning_rate": 2.3271004976746613e-07,
"loss": 0.1736,
"step": 1181
},
{
"epoch": 4.7854251012145745,
"grad_norm": 0.17210716719488703,
"learning_rate": 2.2418695307376393e-07,
"loss": 0.1638,
"step": 1182
},
{
"epoch": 4.7894736842105265,
"grad_norm": 0.1859303132596332,
"learning_rate": 2.1582198325420968e-07,
"loss": 0.1809,
"step": 1183
},
{
"epoch": 4.793522267206478,
"grad_norm": 0.1729724189640435,
"learning_rate": 2.0761520719488447e-07,
"loss": 0.1743,
"step": 1184
},
{
"epoch": 4.797570850202429,
"grad_norm": 0.17917846909285304,
"learning_rate": 1.9956669051694798e-07,
"loss": 0.1836,
"step": 1185
},
{
"epoch": 4.801619433198381,
"grad_norm": 0.18115230392231071,
"learning_rate": 1.916764975761254e-07,
"loss": 0.1714,
"step": 1186
},
{
"epoch": 4.805668016194332,
"grad_norm": 0.1754023552306836,
"learning_rate": 1.839446914621923e-07,
"loss": 0.1759,
"step": 1187
},
{
"epoch": 4.809716599190283,
"grad_norm": 0.17677624349626006,
"learning_rate": 1.7637133399845962e-07,
"loss": 0.1745,
"step": 1188
},
{
"epoch": 4.813765182186235,
"grad_norm": 0.17218634006244016,
"learning_rate": 1.6895648574129397e-07,
"loss": 0.184,
"step": 1189
},
{
"epoch": 4.817813765182186,
"grad_norm": 0.17392025941922454,
"learning_rate": 1.6170020597962465e-07,
"loss": 0.1727,
"step": 1190
},
{
"epoch": 4.821862348178137,
"grad_norm": 0.1815953006577717,
"learning_rate": 1.5460255273447078e-07,
"loss": 0.1854,
"step": 1191
},
{
"epoch": 4.825910931174089,
"grad_norm": 0.1843329972140304,
"learning_rate": 1.4766358275847715e-07,
"loss": 0.1636,
"step": 1192
},
{
"epoch": 4.82995951417004,
"grad_norm": 0.181853051189413,
"learning_rate": 1.4088335153546573e-07,
"loss": 0.1856,
"step": 1193
},
{
"epoch": 4.834008097165992,
"grad_norm": 0.18411456084619884,
"learning_rate": 1.3426191327998496e-07,
"loss": 0.1795,
"step": 1194
},
{
"epoch": 4.838056680161944,
"grad_norm": 0.1816658517058747,
"learning_rate": 1.277993209368744e-07,
"loss": 0.1857,
"step": 1195
},
{
"epoch": 4.842105263157895,
"grad_norm": 0.17714016394098572,
"learning_rate": 1.214956261808542e-07,
"loss": 0.1822,
"step": 1196
},
{
"epoch": 4.846153846153846,
"grad_norm": 0.1765321173786188,
"learning_rate": 1.1535087941609846e-07,
"loss": 0.185,
"step": 1197
},
{
"epoch": 4.850202429149798,
"grad_norm": 0.18511320212921695,
"learning_rate": 1.0936512977583802e-07,
"loss": 0.1938,
"step": 1198
},
{
"epoch": 4.854251012145749,
"grad_norm": 0.1892471577158446,
"learning_rate": 1.0353842512196732e-07,
"loss": 0.1739,
"step": 1199
},
{
"epoch": 4.8582995951417,
"grad_norm": 0.18177645182883104,
"learning_rate": 9.787081204466253e-08,
"loss": 0.1724,
"step": 1200
},
{
"epoch": 4.862348178137652,
"grad_norm": 0.17345621426753796,
"learning_rate": 9.236233586200627e-08,
"loss": 0.1648,
"step": 1201
},
{
"epoch": 4.866396761133603,
"grad_norm": 0.17704126442506027,
"learning_rate": 8.70130406196279e-08,
"loss": 0.1795,
"step": 1202
},
{
"epoch": 4.870445344129554,
"grad_norm": 0.17468772654057474,
"learning_rate": 8.182296909035047e-08,
"loss": 0.1763,
"step": 1203
},
{
"epoch": 4.874493927125506,
"grad_norm": 0.17811412000695248,
"learning_rate": 7.679216277384439e-08,
"loss": 0.1794,
"step": 1204
},
{
"epoch": 4.8785425101214575,
"grad_norm": 0.16784437107504968,
"learning_rate": 7.19206618963053e-08,
"loss": 0.1734,
"step": 1205
},
{
"epoch": 4.882591093117409,
"grad_norm": 0.1827524895544696,
"learning_rate": 6.720850541012347e-08,
"loss": 0.1745,
"step": 1206
},
{
"epoch": 4.886639676113361,
"grad_norm": 0.17670618847113131,
"learning_rate": 6.265573099357714e-08,
"loss": 0.1773,
"step": 1207
},
{
"epoch": 4.890688259109312,
"grad_norm": 0.17450550795459124,
"learning_rate": 5.826237505053067e-08,
"loss": 0.1644,
"step": 1208
},
{
"epoch": 4.894736842105263,
"grad_norm": 0.18223329230818436,
"learning_rate": 5.4028472710137e-08,
"loss": 0.1818,
"step": 1209
},
{
"epoch": 4.898785425101215,
"grad_norm": 0.17671551955423173,
"learning_rate": 4.9954057826571146e-08,
"loss": 0.1636,
"step": 1210
},
{
"epoch": 4.902834008097166,
"grad_norm": 0.1818459180855819,
"learning_rate": 4.6039162978743776e-08,
"loss": 0.1731,
"step": 1211
},
{
"epoch": 4.906882591093117,
"grad_norm": 0.17670944858872117,
"learning_rate": 4.228381947004812e-08,
"loss": 0.1837,
"step": 1212
},
{
"epoch": 4.910931174089069,
"grad_norm": 0.18461157423900867,
"learning_rate": 3.868805732811343e-08,
"loss": 0.1694,
"step": 1213
},
{
"epoch": 4.91497975708502,
"grad_norm": 0.1739076089415337,
"learning_rate": 3.525190530455636e-08,
"loss": 0.1744,
"step": 1214
},
{
"epoch": 4.919028340080971,
"grad_norm": 0.17870089815012077,
"learning_rate": 3.197539087475887e-08,
"loss": 0.1943,
"step": 1215
},
{
"epoch": 4.923076923076923,
"grad_norm": 0.17382043034448794,
"learning_rate": 2.8858540237641786e-08,
"loss": 0.1923,
"step": 1216
},
{
"epoch": 4.9271255060728745,
"grad_norm": 0.174162006249288,
"learning_rate": 2.5901378315458247e-08,
"loss": 0.1744,
"step": 1217
},
{
"epoch": 4.931174089068826,
"grad_norm": 0.17204373494292588,
"learning_rate": 2.3103928753598347e-08,
"loss": 0.179,
"step": 1218
},
{
"epoch": 4.935222672064778,
"grad_norm": 0.1730930266726619,
"learning_rate": 2.0466213920393718e-08,
"loss": 0.1703,
"step": 1219
},
{
"epoch": 4.939271255060729,
"grad_norm": 0.18100051946144902,
"learning_rate": 1.7988254906942115e-08,
"loss": 0.1686,
"step": 1220
},
{
"epoch": 4.94331983805668,
"grad_norm": 0.17699738360028802,
"learning_rate": 1.5670071526936447e-08,
"loss": 0.1855,
"step": 1221
},
{
"epoch": 4.947368421052632,
"grad_norm": 0.17462411752201706,
"learning_rate": 1.3511682316509344e-08,
"loss": 0.1948,
"step": 1222
},
{
"epoch": 4.951417004048583,
"grad_norm": 0.18477583523514343,
"learning_rate": 1.1513104534086605e-08,
"loss": 0.1869,
"step": 1223
},
{
"epoch": 4.955465587044534,
"grad_norm": 0.17674869251666472,
"learning_rate": 9.67435416023843e-09,
"loss": 0.1778,
"step": 1224
},
{
"epoch": 4.959514170040486,
"grad_norm": 0.1806300223222456,
"learning_rate": 7.995445897566178e-09,
"loss": 0.173,
"step": 1225
},
{
"epoch": 4.963562753036437,
"grad_norm": 0.18877773718439467,
"learning_rate": 6.476393170578021e-09,
"loss": 0.1713,
"step": 1226
},
{
"epoch": 4.967611336032388,
"grad_norm": 0.18143674569749024,
"learning_rate": 5.11720812557348e-09,
"loss": 0.1722,
"step": 1227
},
{
"epoch": 4.97165991902834,
"grad_norm": 0.18015566565259264,
"learning_rate": 3.917901630563492e-09,
"loss": 0.1772,
"step": 1228
},
{
"epoch": 4.9757085020242915,
"grad_norm": 0.1822518377421736,
"learning_rate": 2.878483275170485e-09,
"loss": 0.1815,
"step": 1229
},
{
"epoch": 4.979757085020243,
"grad_norm": 0.19513490354109136,
"learning_rate": 1.998961370555108e-09,
"loss": 0.1769,
"step": 1230
},
{
"epoch": 4.983805668016195,
"grad_norm": 0.16463013548412456,
"learning_rate": 1.2793429493518361e-09,
"loss": 0.1726,
"step": 1231
},
{
"epoch": 4.987854251012146,
"grad_norm": 0.17854172228431875,
"learning_rate": 7.196337656112384e-10,
"loss": 0.1704,
"step": 1232
},
{
"epoch": 4.991902834008097,
"grad_norm": 0.1699085711685873,
"learning_rate": 3.198382947511292e-10,
"loss": 0.1849,
"step": 1233
},
{
"epoch": 4.995951417004049,
"grad_norm": 0.18044380582961425,
"learning_rate": 7.995973352770137e-11,
"loss": 0.1871,
"step": 1234
},
{
"epoch": 5.0,
"grad_norm": 0.17580441644122466,
"learning_rate": 0.0,
"loss": 0.1693,
"step": 1235
},
{
"epoch": 5.0,
"step": 1235,
"total_flos": 4.66034976231732e+18,
"train_loss": 0.30737045316802347,
"train_runtime": 28730.5073,
"train_samples_per_second": 5.499,
"train_steps_per_second": 0.043
}
],
"logging_steps": 1,
"max_steps": 1235,
"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": 4.66034976231732e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}