meta_chat_reasoning_75_25 / trainer_state.json
ryanmarten's picture
Upload model
18b4e46 verified
raw
history blame
215 kB
{
"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": 5.206023149846947,
"learning_rate": 3.2258064516129035e-07,
"loss": 0.9021,
"step": 1
},
{
"epoch": 0.008097165991902834,
"grad_norm": 5.127367715822021,
"learning_rate": 6.451612903225807e-07,
"loss": 0.9144,
"step": 2
},
{
"epoch": 0.012145748987854251,
"grad_norm": 4.828289238902112,
"learning_rate": 9.67741935483871e-07,
"loss": 0.8812,
"step": 3
},
{
"epoch": 0.016194331983805668,
"grad_norm": 4.701742410257861,
"learning_rate": 1.2903225806451614e-06,
"loss": 0.8614,
"step": 4
},
{
"epoch": 0.020242914979757085,
"grad_norm": 4.744033235034954,
"learning_rate": 1.6129032258064516e-06,
"loss": 0.9527,
"step": 5
},
{
"epoch": 0.024291497975708502,
"grad_norm": 4.572979602568634,
"learning_rate": 1.935483870967742e-06,
"loss": 0.8866,
"step": 6
},
{
"epoch": 0.02834008097165992,
"grad_norm": 3.6445319198113,
"learning_rate": 2.2580645161290324e-06,
"loss": 0.8441,
"step": 7
},
{
"epoch": 0.032388663967611336,
"grad_norm": 3.6225311596571004,
"learning_rate": 2.580645161290323e-06,
"loss": 0.9287,
"step": 8
},
{
"epoch": 0.03643724696356275,
"grad_norm": 2.3523630172464967,
"learning_rate": 2.903225806451613e-06,
"loss": 0.839,
"step": 9
},
{
"epoch": 0.04048582995951417,
"grad_norm": 2.0403424596455575,
"learning_rate": 3.225806451612903e-06,
"loss": 0.817,
"step": 10
},
{
"epoch": 0.044534412955465584,
"grad_norm": 2.0051912490704034,
"learning_rate": 3.548387096774194e-06,
"loss": 0.7662,
"step": 11
},
{
"epoch": 0.048582995951417005,
"grad_norm": 2.0218268545207043,
"learning_rate": 3.870967741935484e-06,
"loss": 0.7739,
"step": 12
},
{
"epoch": 0.05263157894736842,
"grad_norm": 2.1103755458147573,
"learning_rate": 4.193548387096774e-06,
"loss": 0.7811,
"step": 13
},
{
"epoch": 0.05668016194331984,
"grad_norm": 2.55746157524899,
"learning_rate": 4.516129032258065e-06,
"loss": 0.7986,
"step": 14
},
{
"epoch": 0.06072874493927125,
"grad_norm": 2.507364348536475,
"learning_rate": 4.838709677419355e-06,
"loss": 0.965,
"step": 15
},
{
"epoch": 0.06477732793522267,
"grad_norm": 1.8709089349452113,
"learning_rate": 5.161290322580646e-06,
"loss": 0.6845,
"step": 16
},
{
"epoch": 0.06882591093117409,
"grad_norm": 2.413761499952424,
"learning_rate": 5.483870967741935e-06,
"loss": 0.7867,
"step": 17
},
{
"epoch": 0.0728744939271255,
"grad_norm": 2.239208909483675,
"learning_rate": 5.806451612903226e-06,
"loss": 0.7513,
"step": 18
},
{
"epoch": 0.07692307692307693,
"grad_norm": 1.8363413762734406,
"learning_rate": 6.129032258064517e-06,
"loss": 0.7085,
"step": 19
},
{
"epoch": 0.08097165991902834,
"grad_norm": 1.658087099140405,
"learning_rate": 6.451612903225806e-06,
"loss": 0.8145,
"step": 20
},
{
"epoch": 0.08502024291497975,
"grad_norm": 1.420433180712179,
"learning_rate": 6.774193548387097e-06,
"loss": 0.7138,
"step": 21
},
{
"epoch": 0.08906882591093117,
"grad_norm": 1.4582659173289008,
"learning_rate": 7.096774193548388e-06,
"loss": 0.6734,
"step": 22
},
{
"epoch": 0.0931174089068826,
"grad_norm": 1.3870994782801094,
"learning_rate": 7.4193548387096784e-06,
"loss": 0.7435,
"step": 23
},
{
"epoch": 0.09716599190283401,
"grad_norm": 1.3206234020790917,
"learning_rate": 7.741935483870968e-06,
"loss": 0.6969,
"step": 24
},
{
"epoch": 0.10121457489878542,
"grad_norm": 1.3154319205370968,
"learning_rate": 8.064516129032258e-06,
"loss": 0.7213,
"step": 25
},
{
"epoch": 0.10526315789473684,
"grad_norm": 1.142418739913944,
"learning_rate": 8.387096774193549e-06,
"loss": 0.6342,
"step": 26
},
{
"epoch": 0.10931174089068826,
"grad_norm": 1.0354135714768884,
"learning_rate": 8.70967741935484e-06,
"loss": 0.6479,
"step": 27
},
{
"epoch": 0.11336032388663968,
"grad_norm": 1.2941051053283674,
"learning_rate": 9.03225806451613e-06,
"loss": 0.6723,
"step": 28
},
{
"epoch": 0.11740890688259109,
"grad_norm": 1.0548366073829432,
"learning_rate": 9.35483870967742e-06,
"loss": 0.5922,
"step": 29
},
{
"epoch": 0.1214574898785425,
"grad_norm": 1.1678923131668144,
"learning_rate": 9.67741935483871e-06,
"loss": 0.6809,
"step": 30
},
{
"epoch": 0.12550607287449392,
"grad_norm": 1.2292848704070616,
"learning_rate": 1e-05,
"loss": 0.6366,
"step": 31
},
{
"epoch": 0.12955465587044535,
"grad_norm": 1.012278420212709,
"learning_rate": 1.0322580645161291e-05,
"loss": 0.6056,
"step": 32
},
{
"epoch": 0.13360323886639677,
"grad_norm": 1.1868899829781985,
"learning_rate": 1.0645161290322582e-05,
"loss": 0.6606,
"step": 33
},
{
"epoch": 0.13765182186234817,
"grad_norm": 1.0164820214401544,
"learning_rate": 1.096774193548387e-05,
"loss": 0.6298,
"step": 34
},
{
"epoch": 0.1417004048582996,
"grad_norm": 1.0115627635175424,
"learning_rate": 1.1290322580645164e-05,
"loss": 0.6552,
"step": 35
},
{
"epoch": 0.145748987854251,
"grad_norm": 0.960531739520444,
"learning_rate": 1.1612903225806453e-05,
"loss": 0.6303,
"step": 36
},
{
"epoch": 0.14979757085020243,
"grad_norm": 0.8640723153398834,
"learning_rate": 1.1935483870967743e-05,
"loss": 0.6546,
"step": 37
},
{
"epoch": 0.15384615384615385,
"grad_norm": 0.9363085246065008,
"learning_rate": 1.2258064516129034e-05,
"loss": 0.6478,
"step": 38
},
{
"epoch": 0.15789473684210525,
"grad_norm": 0.8746625314740428,
"learning_rate": 1.2580645161290324e-05,
"loss": 0.6073,
"step": 39
},
{
"epoch": 0.16194331983805668,
"grad_norm": 0.9539654913315541,
"learning_rate": 1.2903225806451613e-05,
"loss": 0.6562,
"step": 40
},
{
"epoch": 0.1659919028340081,
"grad_norm": 0.9150520409527293,
"learning_rate": 1.3225806451612903e-05,
"loss": 0.6114,
"step": 41
},
{
"epoch": 0.1700404858299595,
"grad_norm": 0.8538199827674505,
"learning_rate": 1.3548387096774194e-05,
"loss": 0.649,
"step": 42
},
{
"epoch": 0.17408906882591094,
"grad_norm": 0.9850854438001442,
"learning_rate": 1.3870967741935486e-05,
"loss": 0.585,
"step": 43
},
{
"epoch": 0.17813765182186234,
"grad_norm": 0.8909996256158792,
"learning_rate": 1.4193548387096776e-05,
"loss": 0.5976,
"step": 44
},
{
"epoch": 0.18218623481781376,
"grad_norm": 0.7640564076376722,
"learning_rate": 1.4516129032258066e-05,
"loss": 0.6046,
"step": 45
},
{
"epoch": 0.1862348178137652,
"grad_norm": 0.7682899688514077,
"learning_rate": 1.4838709677419357e-05,
"loss": 0.6237,
"step": 46
},
{
"epoch": 0.1902834008097166,
"grad_norm": 0.7484815688030636,
"learning_rate": 1.5161290322580646e-05,
"loss": 0.6029,
"step": 47
},
{
"epoch": 0.19433198380566802,
"grad_norm": 0.809295629690969,
"learning_rate": 1.5483870967741936e-05,
"loss": 0.6382,
"step": 48
},
{
"epoch": 0.19838056680161945,
"grad_norm": 0.7624819713949866,
"learning_rate": 1.5806451612903226e-05,
"loss": 0.6009,
"step": 49
},
{
"epoch": 0.20242914979757085,
"grad_norm": 1.5608435065576354,
"learning_rate": 1.6129032258064517e-05,
"loss": 0.6769,
"step": 50
},
{
"epoch": 0.20647773279352227,
"grad_norm": 18.28333542365194,
"learning_rate": 1.6451612903225807e-05,
"loss": 0.6139,
"step": 51
},
{
"epoch": 0.21052631578947367,
"grad_norm": 1.6643731792294951,
"learning_rate": 1.6774193548387098e-05,
"loss": 0.6475,
"step": 52
},
{
"epoch": 0.2145748987854251,
"grad_norm": 1.0495317785856653,
"learning_rate": 1.7096774193548388e-05,
"loss": 0.5922,
"step": 53
},
{
"epoch": 0.21862348178137653,
"grad_norm": 0.8119193272837752,
"learning_rate": 1.741935483870968e-05,
"loss": 0.5766,
"step": 54
},
{
"epoch": 0.22267206477732793,
"grad_norm": 0.9308809497130612,
"learning_rate": 1.774193548387097e-05,
"loss": 0.6865,
"step": 55
},
{
"epoch": 0.22672064777327935,
"grad_norm": 0.8900714627178067,
"learning_rate": 1.806451612903226e-05,
"loss": 0.6731,
"step": 56
},
{
"epoch": 0.23076923076923078,
"grad_norm": 5.133242762553566,
"learning_rate": 1.838709677419355e-05,
"loss": 0.6428,
"step": 57
},
{
"epoch": 0.23481781376518218,
"grad_norm": 1.0696992385051503,
"learning_rate": 1.870967741935484e-05,
"loss": 0.644,
"step": 58
},
{
"epoch": 0.2388663967611336,
"grad_norm": 0.9378023346872653,
"learning_rate": 1.903225806451613e-05,
"loss": 0.5853,
"step": 59
},
{
"epoch": 0.242914979757085,
"grad_norm": 1.1900612248355493,
"learning_rate": 1.935483870967742e-05,
"loss": 0.6695,
"step": 60
},
{
"epoch": 0.24696356275303644,
"grad_norm": 0.9118089474391197,
"learning_rate": 1.967741935483871e-05,
"loss": 0.6732,
"step": 61
},
{
"epoch": 0.25101214574898784,
"grad_norm": 0.9693253297015716,
"learning_rate": 2e-05,
"loss": 0.5871,
"step": 62
},
{
"epoch": 0.2550607287449393,
"grad_norm": 0.8430463938649798,
"learning_rate": 2.0322580645161292e-05,
"loss": 0.7628,
"step": 63
},
{
"epoch": 0.2591093117408907,
"grad_norm": 0.8211029435640651,
"learning_rate": 2.0645161290322582e-05,
"loss": 0.6137,
"step": 64
},
{
"epoch": 0.2631578947368421,
"grad_norm": 0.8115215916237623,
"learning_rate": 2.0967741935483873e-05,
"loss": 0.5956,
"step": 65
},
{
"epoch": 0.26720647773279355,
"grad_norm": 0.9541354490250532,
"learning_rate": 2.1290322580645163e-05,
"loss": 0.5787,
"step": 66
},
{
"epoch": 0.27125506072874495,
"grad_norm": 1.2729335186353792,
"learning_rate": 2.161290322580645e-05,
"loss": 0.6145,
"step": 67
},
{
"epoch": 0.27530364372469635,
"grad_norm": 0.7867643339499647,
"learning_rate": 2.193548387096774e-05,
"loss": 0.6,
"step": 68
},
{
"epoch": 0.2793522267206478,
"grad_norm": 0.9358826829974097,
"learning_rate": 2.225806451612903e-05,
"loss": 0.626,
"step": 69
},
{
"epoch": 0.2834008097165992,
"grad_norm": 0.9877936283611368,
"learning_rate": 2.2580645161290328e-05,
"loss": 0.6181,
"step": 70
},
{
"epoch": 0.2874493927125506,
"grad_norm": 0.8942604228127915,
"learning_rate": 2.2903225806451615e-05,
"loss": 0.6103,
"step": 71
},
{
"epoch": 0.291497975708502,
"grad_norm": 0.7937056092612084,
"learning_rate": 2.3225806451612906e-05,
"loss": 0.5853,
"step": 72
},
{
"epoch": 0.29554655870445345,
"grad_norm": 0.8558849050404708,
"learning_rate": 2.3548387096774196e-05,
"loss": 0.6285,
"step": 73
},
{
"epoch": 0.29959514170040485,
"grad_norm": 0.9324562700792617,
"learning_rate": 2.3870967741935486e-05,
"loss": 0.6392,
"step": 74
},
{
"epoch": 0.30364372469635625,
"grad_norm": 0.814603268471525,
"learning_rate": 2.4193548387096777e-05,
"loss": 0.5718,
"step": 75
},
{
"epoch": 0.3076923076923077,
"grad_norm": 0.7507570368768511,
"learning_rate": 2.4516129032258067e-05,
"loss": 0.5455,
"step": 76
},
{
"epoch": 0.3117408906882591,
"grad_norm": 0.887872154291547,
"learning_rate": 2.4838709677419358e-05,
"loss": 0.5917,
"step": 77
},
{
"epoch": 0.3157894736842105,
"grad_norm": 0.8992644606668067,
"learning_rate": 2.5161290322580648e-05,
"loss": 0.5906,
"step": 78
},
{
"epoch": 0.31983805668016196,
"grad_norm": 0.776751559500131,
"learning_rate": 2.548387096774194e-05,
"loss": 0.5661,
"step": 79
},
{
"epoch": 0.32388663967611336,
"grad_norm": 0.9273071663863938,
"learning_rate": 2.5806451612903226e-05,
"loss": 0.6615,
"step": 80
},
{
"epoch": 0.32793522267206476,
"grad_norm": 0.9538318269973226,
"learning_rate": 2.6129032258064516e-05,
"loss": 0.5973,
"step": 81
},
{
"epoch": 0.3319838056680162,
"grad_norm": 0.7785111683070837,
"learning_rate": 2.6451612903225806e-05,
"loss": 0.5581,
"step": 82
},
{
"epoch": 0.3360323886639676,
"grad_norm": 0.9448262101413134,
"learning_rate": 2.6774193548387097e-05,
"loss": 0.6403,
"step": 83
},
{
"epoch": 0.340080971659919,
"grad_norm": 0.9979259333855657,
"learning_rate": 2.7096774193548387e-05,
"loss": 0.5869,
"step": 84
},
{
"epoch": 0.3441295546558704,
"grad_norm": 0.7738233843896348,
"learning_rate": 2.7419354838709678e-05,
"loss": 0.5524,
"step": 85
},
{
"epoch": 0.3481781376518219,
"grad_norm": 0.9953910487554702,
"learning_rate": 2.774193548387097e-05,
"loss": 0.5783,
"step": 86
},
{
"epoch": 0.3522267206477733,
"grad_norm": 0.8756364998210645,
"learning_rate": 2.8064516129032262e-05,
"loss": 0.5781,
"step": 87
},
{
"epoch": 0.3562753036437247,
"grad_norm": 1.4190611848646701,
"learning_rate": 2.8387096774193552e-05,
"loss": 0.6407,
"step": 88
},
{
"epoch": 0.3603238866396761,
"grad_norm": 0.8863857845379071,
"learning_rate": 2.8709677419354843e-05,
"loss": 0.589,
"step": 89
},
{
"epoch": 0.3643724696356275,
"grad_norm": 0.7677018147787863,
"learning_rate": 2.9032258064516133e-05,
"loss": 0.7184,
"step": 90
},
{
"epoch": 0.3684210526315789,
"grad_norm": 0.9445159377657347,
"learning_rate": 2.9354838709677423e-05,
"loss": 0.6374,
"step": 91
},
{
"epoch": 0.3724696356275304,
"grad_norm": 0.8569084759583876,
"learning_rate": 2.9677419354838714e-05,
"loss": 0.6267,
"step": 92
},
{
"epoch": 0.3765182186234818,
"grad_norm": 0.8675849207737646,
"learning_rate": 3.0000000000000004e-05,
"loss": 0.657,
"step": 93
},
{
"epoch": 0.3805668016194332,
"grad_norm": 0.8531383445988553,
"learning_rate": 3.032258064516129e-05,
"loss": 0.5646,
"step": 94
},
{
"epoch": 0.38461538461538464,
"grad_norm": 0.8878441995705032,
"learning_rate": 3.0645161290322585e-05,
"loss": 0.7974,
"step": 95
},
{
"epoch": 0.38866396761133604,
"grad_norm": 0.8241822736104045,
"learning_rate": 3.096774193548387e-05,
"loss": 0.6104,
"step": 96
},
{
"epoch": 0.39271255060728744,
"grad_norm": 0.943558653250568,
"learning_rate": 3.1290322580645166e-05,
"loss": 0.606,
"step": 97
},
{
"epoch": 0.3967611336032389,
"grad_norm": 0.8598040752818146,
"learning_rate": 3.161290322580645e-05,
"loss": 0.6125,
"step": 98
},
{
"epoch": 0.4008097165991903,
"grad_norm": 0.995089353605063,
"learning_rate": 3.193548387096774e-05,
"loss": 0.6,
"step": 99
},
{
"epoch": 0.4048582995951417,
"grad_norm": 0.825966867619722,
"learning_rate": 3.2258064516129034e-05,
"loss": 0.5748,
"step": 100
},
{
"epoch": 0.4089068825910931,
"grad_norm": 0.8217507662613318,
"learning_rate": 3.258064516129033e-05,
"loss": 0.5988,
"step": 101
},
{
"epoch": 0.41295546558704455,
"grad_norm": 0.9151564735577564,
"learning_rate": 3.2903225806451614e-05,
"loss": 0.5922,
"step": 102
},
{
"epoch": 0.41700404858299595,
"grad_norm": 0.8233434739967384,
"learning_rate": 3.322580645161291e-05,
"loss": 0.6075,
"step": 103
},
{
"epoch": 0.42105263157894735,
"grad_norm": 0.9688676667960263,
"learning_rate": 3.3548387096774195e-05,
"loss": 0.7848,
"step": 104
},
{
"epoch": 0.4251012145748988,
"grad_norm": 0.9784649776024908,
"learning_rate": 3.387096774193549e-05,
"loss": 0.6461,
"step": 105
},
{
"epoch": 0.4291497975708502,
"grad_norm": 0.803286623336531,
"learning_rate": 3.4193548387096776e-05,
"loss": 0.5819,
"step": 106
},
{
"epoch": 0.4331983805668016,
"grad_norm": 0.7976017409919359,
"learning_rate": 3.451612903225807e-05,
"loss": 0.5792,
"step": 107
},
{
"epoch": 0.43724696356275305,
"grad_norm": 0.8665279079550425,
"learning_rate": 3.483870967741936e-05,
"loss": 0.6036,
"step": 108
},
{
"epoch": 0.44129554655870445,
"grad_norm": 0.8353560952699988,
"learning_rate": 3.516129032258065e-05,
"loss": 0.637,
"step": 109
},
{
"epoch": 0.44534412955465585,
"grad_norm": 0.9597443186128268,
"learning_rate": 3.548387096774194e-05,
"loss": 0.5915,
"step": 110
},
{
"epoch": 0.4493927125506073,
"grad_norm": 0.7882955948759052,
"learning_rate": 3.5806451612903225e-05,
"loss": 0.6535,
"step": 111
},
{
"epoch": 0.4534412955465587,
"grad_norm": 0.8170506099849301,
"learning_rate": 3.612903225806452e-05,
"loss": 0.6385,
"step": 112
},
{
"epoch": 0.4574898785425101,
"grad_norm": 0.7981397871476831,
"learning_rate": 3.6451612903225805e-05,
"loss": 0.7634,
"step": 113
},
{
"epoch": 0.46153846153846156,
"grad_norm": 0.8310444035332072,
"learning_rate": 3.67741935483871e-05,
"loss": 0.6073,
"step": 114
},
{
"epoch": 0.46558704453441296,
"grad_norm": 0.8682313221248011,
"learning_rate": 3.7096774193548386e-05,
"loss": 0.6363,
"step": 115
},
{
"epoch": 0.46963562753036436,
"grad_norm": 0.7999782365488793,
"learning_rate": 3.741935483870968e-05,
"loss": 0.6337,
"step": 116
},
{
"epoch": 0.47368421052631576,
"grad_norm": 0.7202975365926318,
"learning_rate": 3.7741935483870974e-05,
"loss": 0.5638,
"step": 117
},
{
"epoch": 0.4777327935222672,
"grad_norm": 0.722369448763757,
"learning_rate": 3.806451612903226e-05,
"loss": 0.5855,
"step": 118
},
{
"epoch": 0.4817813765182186,
"grad_norm": 0.7448098988441882,
"learning_rate": 3.8387096774193555e-05,
"loss": 0.6141,
"step": 119
},
{
"epoch": 0.48582995951417,
"grad_norm": 0.7485316082591391,
"learning_rate": 3.870967741935484e-05,
"loss": 0.6387,
"step": 120
},
{
"epoch": 0.4898785425101215,
"grad_norm": 0.7558142649368498,
"learning_rate": 3.9032258064516135e-05,
"loss": 0.6447,
"step": 121
},
{
"epoch": 0.4939271255060729,
"grad_norm": 0.6500680385968292,
"learning_rate": 3.935483870967742e-05,
"loss": 0.5973,
"step": 122
},
{
"epoch": 0.4979757085020243,
"grad_norm": 0.7774301379855769,
"learning_rate": 3.9677419354838716e-05,
"loss": 0.83,
"step": 123
},
{
"epoch": 0.5020242914979757,
"grad_norm": 0.7406010976620854,
"learning_rate": 4e-05,
"loss": 0.6075,
"step": 124
},
{
"epoch": 0.5060728744939271,
"grad_norm": 0.7356677598097976,
"learning_rate": 3.9999920040266474e-05,
"loss": 0.6227,
"step": 125
},
{
"epoch": 0.5101214574898786,
"grad_norm": 0.6801537708549805,
"learning_rate": 3.999968016170525e-05,
"loss": 0.5892,
"step": 126
},
{
"epoch": 0.5141700404858299,
"grad_norm": 0.6826901186135467,
"learning_rate": 3.9999280366234394e-05,
"loss": 0.5841,
"step": 127
},
{
"epoch": 0.5182186234817814,
"grad_norm": 0.8118023346591131,
"learning_rate": 3.999872065705065e-05,
"loss": 0.597,
"step": 128
},
{
"epoch": 0.5222672064777328,
"grad_norm": 0.7078328462204903,
"learning_rate": 3.999800103862945e-05,
"loss": 0.6542,
"step": 129
},
{
"epoch": 0.5263157894736842,
"grad_norm": 0.7762894882107685,
"learning_rate": 3.9997121516724834e-05,
"loss": 0.5894,
"step": 130
},
{
"epoch": 0.5303643724696356,
"grad_norm": 0.713765247421285,
"learning_rate": 3.999608209836944e-05,
"loss": 0.6117,
"step": 131
},
{
"epoch": 0.5344129554655871,
"grad_norm": 0.7927658840991731,
"learning_rate": 3.999488279187443e-05,
"loss": 0.5818,
"step": 132
},
{
"epoch": 0.5384615384615384,
"grad_norm": 0.7579944937620262,
"learning_rate": 3.9993523606829425e-05,
"loss": 0.6363,
"step": 133
},
{
"epoch": 0.5425101214574899,
"grad_norm": 0.846152622686558,
"learning_rate": 3.9992004554102435e-05,
"loss": 0.6291,
"step": 134
},
{
"epoch": 0.5465587044534413,
"grad_norm": 0.8206106375546257,
"learning_rate": 3.999032564583976e-05,
"loss": 0.6529,
"step": 135
},
{
"epoch": 0.5506072874493927,
"grad_norm": 0.7952272165404498,
"learning_rate": 3.998848689546591e-05,
"loss": 0.6241,
"step": 136
},
{
"epoch": 0.5546558704453441,
"grad_norm": 0.7875636600210746,
"learning_rate": 3.9986488317683496e-05,
"loss": 0.6064,
"step": 137
},
{
"epoch": 0.5587044534412956,
"grad_norm": 0.8151687353934215,
"learning_rate": 3.9984329928473067e-05,
"loss": 0.6508,
"step": 138
},
{
"epoch": 0.562753036437247,
"grad_norm": 0.7771642418536502,
"learning_rate": 3.998201174509306e-05,
"loss": 0.6274,
"step": 139
},
{
"epoch": 0.5668016194331984,
"grad_norm": 0.7218110433125675,
"learning_rate": 3.997953378607961e-05,
"loss": 0.579,
"step": 140
},
{
"epoch": 0.5708502024291497,
"grad_norm": 0.8367731709533636,
"learning_rate": 3.9976896071246403e-05,
"loss": 0.6211,
"step": 141
},
{
"epoch": 0.5748987854251012,
"grad_norm": 0.775199629856055,
"learning_rate": 3.997409862168454e-05,
"loss": 0.5971,
"step": 142
},
{
"epoch": 0.5789473684210527,
"grad_norm": 1.053243981506196,
"learning_rate": 3.997114145976237e-05,
"loss": 0.7237,
"step": 143
},
{
"epoch": 0.582995951417004,
"grad_norm": 0.7670077965260371,
"learning_rate": 3.9968024609125245e-05,
"loss": 0.59,
"step": 144
},
{
"epoch": 0.5870445344129555,
"grad_norm": 0.804986475325952,
"learning_rate": 3.996474809469545e-05,
"loss": 0.6276,
"step": 145
},
{
"epoch": 0.5910931174089069,
"grad_norm": 0.6984976897842319,
"learning_rate": 3.996131194267189e-05,
"loss": 0.6055,
"step": 146
},
{
"epoch": 0.5951417004048583,
"grad_norm": 0.832998629693295,
"learning_rate": 3.995771618052995e-05,
"loss": 0.6329,
"step": 147
},
{
"epoch": 0.5991902834008097,
"grad_norm": 0.7031312055174446,
"learning_rate": 3.9953960837021256e-05,
"loss": 0.6054,
"step": 148
},
{
"epoch": 0.6032388663967612,
"grad_norm": 0.7744596295942834,
"learning_rate": 3.995004594217343e-05,
"loss": 0.6789,
"step": 149
},
{
"epoch": 0.6072874493927125,
"grad_norm": 0.6628610698809362,
"learning_rate": 3.994597152728987e-05,
"loss": 0.5915,
"step": 150
},
{
"epoch": 0.611336032388664,
"grad_norm": 0.7346728275972338,
"learning_rate": 3.994173762494947e-05,
"loss": 0.5763,
"step": 151
},
{
"epoch": 0.6153846153846154,
"grad_norm": 0.6907783760637236,
"learning_rate": 3.9937344269006424e-05,
"loss": 0.5895,
"step": 152
},
{
"epoch": 0.6194331983805668,
"grad_norm": 0.7419838216565653,
"learning_rate": 3.993279149458988e-05,
"loss": 0.6378,
"step": 153
},
{
"epoch": 0.6234817813765182,
"grad_norm": 0.6681696597221729,
"learning_rate": 3.9928079338103695e-05,
"loss": 0.5792,
"step": 154
},
{
"epoch": 0.6275303643724697,
"grad_norm": 0.6316319242708208,
"learning_rate": 3.992320783722616e-05,
"loss": 0.6059,
"step": 155
},
{
"epoch": 0.631578947368421,
"grad_norm": 0.7477496534840036,
"learning_rate": 3.991817703090965e-05,
"loss": 0.5982,
"step": 156
},
{
"epoch": 0.6356275303643725,
"grad_norm": 0.7336267061038764,
"learning_rate": 3.991298695938038e-05,
"loss": 0.6261,
"step": 157
},
{
"epoch": 0.6396761133603239,
"grad_norm": 0.6527653514761087,
"learning_rate": 3.9907637664138e-05,
"loss": 0.6064,
"step": 158
},
{
"epoch": 0.6437246963562753,
"grad_norm": 0.6732215895869765,
"learning_rate": 3.990212918795534e-05,
"loss": 0.5741,
"step": 159
},
{
"epoch": 0.6477732793522267,
"grad_norm": 0.8137843303616258,
"learning_rate": 3.9896461574878036e-05,
"loss": 0.632,
"step": 160
},
{
"epoch": 0.6518218623481782,
"grad_norm": 0.6288118921822599,
"learning_rate": 3.989063487022416e-05,
"loss": 0.6236,
"step": 161
},
{
"epoch": 0.6558704453441295,
"grad_norm": 0.7031030225690456,
"learning_rate": 3.9884649120583903e-05,
"loss": 0.7433,
"step": 162
},
{
"epoch": 0.659919028340081,
"grad_norm": 0.6570174541917876,
"learning_rate": 3.987850437381915e-05,
"loss": 0.6381,
"step": 163
},
{
"epoch": 0.6639676113360324,
"grad_norm": 0.7442066862297487,
"learning_rate": 3.987220067906313e-05,
"loss": 0.6382,
"step": 164
},
{
"epoch": 0.6680161943319838,
"grad_norm": 0.7859571467290338,
"learning_rate": 3.986573808672002e-05,
"loss": 0.6311,
"step": 165
},
{
"epoch": 0.6720647773279352,
"grad_norm": 0.6799591011608732,
"learning_rate": 3.9859116648464534e-05,
"loss": 0.631,
"step": 166
},
{
"epoch": 0.6761133603238867,
"grad_norm": 0.7393683325896692,
"learning_rate": 3.985233641724152e-05,
"loss": 0.6248,
"step": 167
},
{
"epoch": 0.680161943319838,
"grad_norm": 0.7230703191896984,
"learning_rate": 3.984539744726554e-05,
"loss": 0.6247,
"step": 168
},
{
"epoch": 0.6842105263157895,
"grad_norm": 0.6794167090431038,
"learning_rate": 3.983829979402038e-05,
"loss": 0.6182,
"step": 169
},
{
"epoch": 0.6882591093117408,
"grad_norm": 0.6832691830866277,
"learning_rate": 3.983104351425871e-05,
"loss": 0.6187,
"step": 170
},
{
"epoch": 0.6923076923076923,
"grad_norm": 0.7204839785539684,
"learning_rate": 3.982362866600154e-05,
"loss": 0.614,
"step": 171
},
{
"epoch": 0.6963562753036437,
"grad_norm": 0.6085384409929323,
"learning_rate": 3.9816055308537806e-05,
"loss": 0.5657,
"step": 172
},
{
"epoch": 0.7004048582995951,
"grad_norm": 0.7049381418712354,
"learning_rate": 3.980832350242388e-05,
"loss": 0.6193,
"step": 173
},
{
"epoch": 0.7044534412955465,
"grad_norm": 0.735103893499536,
"learning_rate": 3.980043330948306e-05,
"loss": 0.6299,
"step": 174
},
{
"epoch": 0.708502024291498,
"grad_norm": 0.6650591421182982,
"learning_rate": 3.9792384792805124e-05,
"loss": 0.6124,
"step": 175
},
{
"epoch": 0.7125506072874493,
"grad_norm": 0.7343574896896063,
"learning_rate": 3.978417801674579e-05,
"loss": 0.6895,
"step": 176
},
{
"epoch": 0.7165991902834008,
"grad_norm": 0.6355312056019733,
"learning_rate": 3.9775813046926244e-05,
"loss": 0.5965,
"step": 177
},
{
"epoch": 0.7206477732793523,
"grad_norm": 0.6461493048755953,
"learning_rate": 3.976728995023254e-05,
"loss": 0.5791,
"step": 178
},
{
"epoch": 0.7246963562753036,
"grad_norm": 0.6398303588627815,
"learning_rate": 3.975860879481514e-05,
"loss": 0.612,
"step": 179
},
{
"epoch": 0.728744939271255,
"grad_norm": 0.645850341888101,
"learning_rate": 3.974976965008833e-05,
"loss": 0.6326,
"step": 180
},
{
"epoch": 0.7327935222672065,
"grad_norm": 0.7219046501013746,
"learning_rate": 3.9740772586729674e-05,
"loss": 0.6093,
"step": 181
},
{
"epoch": 0.7368421052631579,
"grad_norm": 0.633144781966911,
"learning_rate": 3.973161767667946e-05,
"loss": 0.5785,
"step": 182
},
{
"epoch": 0.7408906882591093,
"grad_norm": 0.6795012854414282,
"learning_rate": 3.972230499314009e-05,
"loss": 0.6455,
"step": 183
},
{
"epoch": 0.7449392712550608,
"grad_norm": 0.6847967809984112,
"learning_rate": 3.9712834610575545e-05,
"loss": 0.6025,
"step": 184
},
{
"epoch": 0.7489878542510121,
"grad_norm": 0.6644753563048008,
"learning_rate": 3.9703206604710746e-05,
"loss": 0.5745,
"step": 185
},
{
"epoch": 0.7530364372469636,
"grad_norm": 0.5862993193930612,
"learning_rate": 3.969342105253098e-05,
"loss": 0.5644,
"step": 186
},
{
"epoch": 0.757085020242915,
"grad_norm": 0.626153420743745,
"learning_rate": 3.9683478032281246e-05,
"loss": 0.5927,
"step": 187
},
{
"epoch": 0.7611336032388664,
"grad_norm": 0.7651615596998178,
"learning_rate": 3.967337762346568e-05,
"loss": 0.6297,
"step": 188
},
{
"epoch": 0.7651821862348178,
"grad_norm": 0.5696505301132955,
"learning_rate": 3.9663119906846885e-05,
"loss": 0.5775,
"step": 189
},
{
"epoch": 0.7692307692307693,
"grad_norm": 0.6267653376638201,
"learning_rate": 3.965270496444528e-05,
"loss": 0.6367,
"step": 190
},
{
"epoch": 0.7732793522267206,
"grad_norm": 0.7456597447422313,
"learning_rate": 3.9642132879538465e-05,
"loss": 0.6165,
"step": 191
},
{
"epoch": 0.7773279352226721,
"grad_norm": 0.6525294948660847,
"learning_rate": 3.9631403736660566e-05,
"loss": 0.6409,
"step": 192
},
{
"epoch": 0.7813765182186235,
"grad_norm": 0.6591580587003882,
"learning_rate": 3.962051762160151e-05,
"loss": 0.6306,
"step": 193
},
{
"epoch": 0.7854251012145749,
"grad_norm": 0.6406039938523258,
"learning_rate": 3.960947462140638e-05,
"loss": 0.5971,
"step": 194
},
{
"epoch": 0.7894736842105263,
"grad_norm": 0.7261078560370556,
"learning_rate": 3.959827482437472e-05,
"loss": 0.6058,
"step": 195
},
{
"epoch": 0.7935222672064778,
"grad_norm": 0.6389155468362202,
"learning_rate": 3.958691832005981e-05,
"loss": 0.6165,
"step": 196
},
{
"epoch": 0.7975708502024291,
"grad_norm": 0.6723570032360663,
"learning_rate": 3.9575405199267944e-05,
"loss": 0.6099,
"step": 197
},
{
"epoch": 0.8016194331983806,
"grad_norm": 0.7073117340589495,
"learning_rate": 3.956373555405774e-05,
"loss": 0.6595,
"step": 198
},
{
"epoch": 0.805668016194332,
"grad_norm": 0.698603988831562,
"learning_rate": 3.955190947773937e-05,
"loss": 0.6412,
"step": 199
},
{
"epoch": 0.8097165991902834,
"grad_norm": 0.6713207816683973,
"learning_rate": 3.9539927064873824e-05,
"loss": 0.5912,
"step": 200
},
{
"epoch": 0.8137651821862348,
"grad_norm": 0.6942558869670397,
"learning_rate": 3.952778841127214e-05,
"loss": 0.6133,
"step": 201
},
{
"epoch": 0.8178137651821862,
"grad_norm": 0.725079807328886,
"learning_rate": 3.95154936139947e-05,
"loss": 0.6218,
"step": 202
},
{
"epoch": 0.8218623481781376,
"grad_norm": 0.8055552729436208,
"learning_rate": 3.950304277135034e-05,
"loss": 0.6572,
"step": 203
},
{
"epoch": 0.8259109311740891,
"grad_norm": 0.678419656738104,
"learning_rate": 3.94904359828957e-05,
"loss": 0.6814,
"step": 204
},
{
"epoch": 0.8299595141700404,
"grad_norm": 0.6860121930002119,
"learning_rate": 3.94776733494343e-05,
"loss": 0.6533,
"step": 205
},
{
"epoch": 0.8340080971659919,
"grad_norm": 0.617351571649231,
"learning_rate": 3.9464754973015826e-05,
"loss": 0.5772,
"step": 206
},
{
"epoch": 0.8380566801619433,
"grad_norm": 0.6725922148024599,
"learning_rate": 3.9451680956935275e-05,
"loss": 0.5606,
"step": 207
},
{
"epoch": 0.8421052631578947,
"grad_norm": 0.7383205836177397,
"learning_rate": 3.943845140573213e-05,
"loss": 0.6145,
"step": 208
},
{
"epoch": 0.8461538461538461,
"grad_norm": 0.7398200045457529,
"learning_rate": 3.942506642518952e-05,
"loss": 0.623,
"step": 209
},
{
"epoch": 0.8502024291497976,
"grad_norm": 0.7556412787873144,
"learning_rate": 3.9411526122333396e-05,
"loss": 0.6728,
"step": 210
},
{
"epoch": 0.854251012145749,
"grad_norm": 0.7077455029515518,
"learning_rate": 3.939783060543167e-05,
"loss": 0.6376,
"step": 211
},
{
"epoch": 0.8582995951417004,
"grad_norm": 0.6759558194674467,
"learning_rate": 3.938397998399333e-05,
"loss": 0.6466,
"step": 212
},
{
"epoch": 0.8623481781376519,
"grad_norm": 0.7044189564699292,
"learning_rate": 3.936997436876756e-05,
"loss": 0.5977,
"step": 213
},
{
"epoch": 0.8663967611336032,
"grad_norm": 0.662462252413443,
"learning_rate": 3.935581387174289e-05,
"loss": 0.6102,
"step": 214
},
{
"epoch": 0.8704453441295547,
"grad_norm": 0.6869258133169127,
"learning_rate": 3.9341498606146295e-05,
"loss": 0.6004,
"step": 215
},
{
"epoch": 0.8744939271255061,
"grad_norm": 0.6374383443048676,
"learning_rate": 3.932702868644224e-05,
"loss": 0.6406,
"step": 216
},
{
"epoch": 0.8785425101214575,
"grad_norm": 0.7493795047084517,
"learning_rate": 3.931240422833181e-05,
"loss": 0.6163,
"step": 217
},
{
"epoch": 0.8825910931174089,
"grad_norm": 0.7160563409951953,
"learning_rate": 3.9297625348751806e-05,
"loss": 0.6166,
"step": 218
},
{
"epoch": 0.8866396761133604,
"grad_norm": 0.6576966655531219,
"learning_rate": 3.9282692165873744e-05,
"loss": 0.6344,
"step": 219
},
{
"epoch": 0.8906882591093117,
"grad_norm": 0.6659333151237702,
"learning_rate": 3.9267604799102945e-05,
"loss": 0.574,
"step": 220
},
{
"epoch": 0.8947368421052632,
"grad_norm": 0.6750132663462449,
"learning_rate": 3.9252363369077615e-05,
"loss": 0.6481,
"step": 221
},
{
"epoch": 0.8987854251012146,
"grad_norm": 0.7272103893114119,
"learning_rate": 3.9236967997667796e-05,
"loss": 0.6826,
"step": 222
},
{
"epoch": 0.902834008097166,
"grad_norm": 0.6843785692346549,
"learning_rate": 3.9221418807974496e-05,
"loss": 0.5991,
"step": 223
},
{
"epoch": 0.9068825910931174,
"grad_norm": 0.758411367420483,
"learning_rate": 3.92057159243286e-05,
"loss": 0.6115,
"step": 224
},
{
"epoch": 0.9109311740890689,
"grad_norm": 0.7428420699276798,
"learning_rate": 3.9189859472289956e-05,
"loss": 0.6194,
"step": 225
},
{
"epoch": 0.9149797570850202,
"grad_norm": 0.6625822452778825,
"learning_rate": 3.9173849578646324e-05,
"loss": 0.5764,
"step": 226
},
{
"epoch": 0.9190283400809717,
"grad_norm": 0.6388855922631688,
"learning_rate": 3.9157686371412395e-05,
"loss": 0.6107,
"step": 227
},
{
"epoch": 0.9230769230769231,
"grad_norm": 0.6607024050448685,
"learning_rate": 3.914136997982874e-05,
"loss": 0.6156,
"step": 228
},
{
"epoch": 0.9271255060728745,
"grad_norm": 0.698482910455574,
"learning_rate": 3.912490053436079e-05,
"loss": 0.6069,
"step": 229
},
{
"epoch": 0.9311740890688259,
"grad_norm": 0.5953327803237067,
"learning_rate": 3.9108278166697805e-05,
"loss": 0.6481,
"step": 230
},
{
"epoch": 0.9352226720647774,
"grad_norm": 0.6678620293273981,
"learning_rate": 3.909150300975178e-05,
"loss": 0.8301,
"step": 231
},
{
"epoch": 0.9392712550607287,
"grad_norm": 0.6561623836974516,
"learning_rate": 3.907457519765642e-05,
"loss": 0.6178,
"step": 232
},
{
"epoch": 0.9433198380566802,
"grad_norm": 0.5855394612534037,
"learning_rate": 3.905749486576607e-05,
"loss": 0.5632,
"step": 233
},
{
"epoch": 0.9473684210526315,
"grad_norm": 0.7074028259807742,
"learning_rate": 3.90402621506546e-05,
"loss": 0.6211,
"step": 234
},
{
"epoch": 0.951417004048583,
"grad_norm": 0.7016757841357273,
"learning_rate": 3.902287719011435e-05,
"loss": 0.6512,
"step": 235
},
{
"epoch": 0.9554655870445344,
"grad_norm": 0.678760601962916,
"learning_rate": 3.900534012315499e-05,
"loss": 0.5812,
"step": 236
},
{
"epoch": 0.9595141700404858,
"grad_norm": 0.7081999449379383,
"learning_rate": 3.8987651090002454e-05,
"loss": 0.6207,
"step": 237
},
{
"epoch": 0.9635627530364372,
"grad_norm": 0.7892675326403209,
"learning_rate": 3.896981023209777e-05,
"loss": 0.6538,
"step": 238
},
{
"epoch": 0.9676113360323887,
"grad_norm": 0.6881652418630346,
"learning_rate": 3.8951817692095966e-05,
"loss": 0.6104,
"step": 239
},
{
"epoch": 0.97165991902834,
"grad_norm": 0.6984942957277719,
"learning_rate": 3.893367361386491e-05,
"loss": 0.6316,
"step": 240
},
{
"epoch": 0.9757085020242915,
"grad_norm": 0.7182209581627654,
"learning_rate": 3.891537814248417e-05,
"loss": 0.5934,
"step": 241
},
{
"epoch": 0.979757085020243,
"grad_norm": 0.6176977465400781,
"learning_rate": 3.889693142424384e-05,
"loss": 0.5917,
"step": 242
},
{
"epoch": 0.9838056680161943,
"grad_norm": 0.6904491459118738,
"learning_rate": 3.8878333606643405e-05,
"loss": 0.7174,
"step": 243
},
{
"epoch": 0.9878542510121457,
"grad_norm": 0.6216163968681959,
"learning_rate": 3.885958483839049e-05,
"loss": 0.7696,
"step": 244
},
{
"epoch": 0.9919028340080972,
"grad_norm": 0.735834336623467,
"learning_rate": 3.8840685269399786e-05,
"loss": 0.6181,
"step": 245
},
{
"epoch": 0.9959514170040485,
"grad_norm": 0.6553789475417823,
"learning_rate": 3.882163505079171e-05,
"loss": 0.6212,
"step": 246
},
{
"epoch": 1.0,
"grad_norm": 0.7315701557535914,
"learning_rate": 3.880243433489132e-05,
"loss": 0.6185,
"step": 247
},
{
"epoch": 1.0040485829959513,
"grad_norm": 0.759102965051288,
"learning_rate": 3.878308327522702e-05,
"loss": 0.4889,
"step": 248
},
{
"epoch": 1.008097165991903,
"grad_norm": 0.6013038546568937,
"learning_rate": 3.876358202652936e-05,
"loss": 0.4484,
"step": 249
},
{
"epoch": 1.0121457489878543,
"grad_norm": 0.7855037341199262,
"learning_rate": 3.8743930744729835e-05,
"loss": 0.4671,
"step": 250
},
{
"epoch": 1.0161943319838056,
"grad_norm": 0.7611286541775798,
"learning_rate": 3.8724129586959535e-05,
"loss": 0.4675,
"step": 251
},
{
"epoch": 1.0202429149797572,
"grad_norm": 0.7526146929241442,
"learning_rate": 3.870417871154801e-05,
"loss": 0.4415,
"step": 252
},
{
"epoch": 1.0242914979757085,
"grad_norm": 0.6711531826688504,
"learning_rate": 3.868407827802192e-05,
"loss": 0.5043,
"step": 253
},
{
"epoch": 1.0283400809716599,
"grad_norm": 0.6703022445612726,
"learning_rate": 3.8663828447103804e-05,
"loss": 0.4623,
"step": 254
},
{
"epoch": 1.0323886639676114,
"grad_norm": 0.7324808753455638,
"learning_rate": 3.8643429380710765e-05,
"loss": 0.4332,
"step": 255
},
{
"epoch": 1.0364372469635628,
"grad_norm": 0.7375038545987127,
"learning_rate": 3.862288124195319e-05,
"loss": 0.4269,
"step": 256
},
{
"epoch": 1.040485829959514,
"grad_norm": 0.6359959876982106,
"learning_rate": 3.8602184195133464e-05,
"loss": 0.4496,
"step": 257
},
{
"epoch": 1.0445344129554657,
"grad_norm": 0.6381820875107513,
"learning_rate": 3.8581338405744604e-05,
"loss": 0.4204,
"step": 258
},
{
"epoch": 1.048582995951417,
"grad_norm": 0.7365271554308105,
"learning_rate": 3.8560344040468996e-05,
"loss": 0.4577,
"step": 259
},
{
"epoch": 1.0526315789473684,
"grad_norm": 0.7261151911748543,
"learning_rate": 3.853920126717703e-05,
"loss": 0.4717,
"step": 260
},
{
"epoch": 1.05668016194332,
"grad_norm": 0.6631171002579896,
"learning_rate": 3.851791025492574e-05,
"loss": 0.4525,
"step": 261
},
{
"epoch": 1.0607287449392713,
"grad_norm": 0.6334321981566126,
"learning_rate": 3.8496471173957515e-05,
"loss": 0.4232,
"step": 262
},
{
"epoch": 1.0647773279352226,
"grad_norm": 0.719191762598135,
"learning_rate": 3.847488419569866e-05,
"loss": 0.4766,
"step": 263
},
{
"epoch": 1.0688259109311742,
"grad_norm": 0.7539204207068246,
"learning_rate": 3.8453149492758077e-05,
"loss": 0.4475,
"step": 264
},
{
"epoch": 1.0728744939271255,
"grad_norm": 1.1588541642874723,
"learning_rate": 3.8431267238925886e-05,
"loss": 0.4614,
"step": 265
},
{
"epoch": 1.0769230769230769,
"grad_norm": 0.7500061802945155,
"learning_rate": 3.840923760917199e-05,
"loss": 0.4416,
"step": 266
},
{
"epoch": 1.0809716599190284,
"grad_norm": 0.8070754389604308,
"learning_rate": 3.838706077964473e-05,
"loss": 0.4671,
"step": 267
},
{
"epoch": 1.0850202429149798,
"grad_norm": 0.7928651889954846,
"learning_rate": 3.836473692766944e-05,
"loss": 0.5022,
"step": 268
},
{
"epoch": 1.0890688259109311,
"grad_norm": 0.6240334985831181,
"learning_rate": 3.834226623174705e-05,
"loss": 0.4524,
"step": 269
},
{
"epoch": 1.0931174089068827,
"grad_norm": 0.7272592510308632,
"learning_rate": 3.831964887155264e-05,
"loss": 0.48,
"step": 270
},
{
"epoch": 1.097165991902834,
"grad_norm": 0.6103578442262221,
"learning_rate": 3.829688502793403e-05,
"loss": 0.5512,
"step": 271
},
{
"epoch": 1.1012145748987854,
"grad_norm": 0.7236032895649986,
"learning_rate": 3.827397488291029e-05,
"loss": 0.6474,
"step": 272
},
{
"epoch": 1.1052631578947367,
"grad_norm": 0.5886460545642788,
"learning_rate": 3.8250918619670343e-05,
"loss": 0.448,
"step": 273
},
{
"epoch": 1.1093117408906883,
"grad_norm": 0.6244390600486447,
"learning_rate": 3.822771642257145e-05,
"loss": 0.4671,
"step": 274
},
{
"epoch": 1.1133603238866396,
"grad_norm": 0.6118075783155814,
"learning_rate": 3.820436847713776e-05,
"loss": 0.5707,
"step": 275
},
{
"epoch": 1.117408906882591,
"grad_norm": 0.6423205412102043,
"learning_rate": 3.8180874970058826e-05,
"loss": 0.4464,
"step": 276
},
{
"epoch": 1.1214574898785425,
"grad_norm": 0.7527604781438215,
"learning_rate": 3.8157236089188103e-05,
"loss": 0.4952,
"step": 277
},
{
"epoch": 1.125506072874494,
"grad_norm": 0.7362236080397708,
"learning_rate": 3.813345202354145e-05,
"loss": 0.4298,
"step": 278
},
{
"epoch": 1.1295546558704452,
"grad_norm": 0.7633034602021315,
"learning_rate": 3.810952296329563e-05,
"loss": 0.4539,
"step": 279
},
{
"epoch": 1.1336032388663968,
"grad_norm": 0.6612327352147993,
"learning_rate": 3.808544909978676e-05,
"loss": 0.4392,
"step": 280
},
{
"epoch": 1.1376518218623481,
"grad_norm": 0.891044314079862,
"learning_rate": 3.806123062550882e-05,
"loss": 0.5314,
"step": 281
},
{
"epoch": 1.1417004048582995,
"grad_norm": 0.6589231731767884,
"learning_rate": 3.803686773411208e-05,
"loss": 0.4947,
"step": 282
},
{
"epoch": 1.145748987854251,
"grad_norm": 0.6746239811694618,
"learning_rate": 3.801236062040158e-05,
"loss": 0.451,
"step": 283
},
{
"epoch": 1.1497975708502024,
"grad_norm": 0.7408932104101391,
"learning_rate": 3.798770948033553e-05,
"loss": 0.4385,
"step": 284
},
{
"epoch": 1.1538461538461537,
"grad_norm": 0.5757936433478343,
"learning_rate": 3.796291451102381e-05,
"loss": 0.4504,
"step": 285
},
{
"epoch": 1.1578947368421053,
"grad_norm": 0.6552491155393835,
"learning_rate": 3.793797591072631e-05,
"loss": 0.4236,
"step": 286
},
{
"epoch": 1.1619433198380567,
"grad_norm": 0.6988375191849772,
"learning_rate": 3.791289387885144e-05,
"loss": 0.4317,
"step": 287
},
{
"epoch": 1.165991902834008,
"grad_norm": 0.6670642310410722,
"learning_rate": 3.788766861595443e-05,
"loss": 0.4399,
"step": 288
},
{
"epoch": 1.1700404858299596,
"grad_norm": 0.6980400331923179,
"learning_rate": 3.7862300323735835e-05,
"loss": 0.4615,
"step": 289
},
{
"epoch": 1.174089068825911,
"grad_norm": 0.7314041349002252,
"learning_rate": 3.783678920503983e-05,
"loss": 0.4642,
"step": 290
},
{
"epoch": 1.1781376518218623,
"grad_norm": 0.639603489831615,
"learning_rate": 3.781113546385265e-05,
"loss": 0.4491,
"step": 291
},
{
"epoch": 1.1821862348178138,
"grad_norm": 0.6661923902861597,
"learning_rate": 3.7785339305300916e-05,
"loss": 0.4476,
"step": 292
},
{
"epoch": 1.1862348178137652,
"grad_norm": 0.6582221626299164,
"learning_rate": 3.7759400935650025e-05,
"loss": 0.6297,
"step": 293
},
{
"epoch": 1.1902834008097165,
"grad_norm": 0.6694376018941177,
"learning_rate": 3.77333205623025e-05,
"loss": 0.4446,
"step": 294
},
{
"epoch": 1.194331983805668,
"grad_norm": 0.6640907860797097,
"learning_rate": 3.7707098393796295e-05,
"loss": 0.4252,
"step": 295
},
{
"epoch": 1.1983805668016194,
"grad_norm": 0.6722255476299922,
"learning_rate": 3.7680734639803185e-05,
"loss": 0.4537,
"step": 296
},
{
"epoch": 1.2024291497975708,
"grad_norm": 1.5156302428512753,
"learning_rate": 3.765422951112704e-05,
"loss": 0.4827,
"step": 297
},
{
"epoch": 1.2064777327935223,
"grad_norm": 0.7507396692338515,
"learning_rate": 3.762758321970216e-05,
"loss": 0.8273,
"step": 298
},
{
"epoch": 1.2105263157894737,
"grad_norm": 0.618081078115142,
"learning_rate": 3.7600795978591584e-05,
"loss": 0.423,
"step": 299
},
{
"epoch": 1.214574898785425,
"grad_norm": 1.2123396187454636,
"learning_rate": 3.757386800198538e-05,
"loss": 0.5461,
"step": 300
},
{
"epoch": 1.2186234817813766,
"grad_norm": 0.6795211719270041,
"learning_rate": 3.7546799505198925e-05,
"loss": 0.4093,
"step": 301
},
{
"epoch": 1.222672064777328,
"grad_norm": 0.728342626237238,
"learning_rate": 3.7519590704671197e-05,
"loss": 0.461,
"step": 302
},
{
"epoch": 1.2267206477732793,
"grad_norm": 0.6013129927347952,
"learning_rate": 3.749224181796305e-05,
"loss": 0.4994,
"step": 303
},
{
"epoch": 1.2307692307692308,
"grad_norm": 0.5849351192150203,
"learning_rate": 3.746475306375545e-05,
"loss": 0.4178,
"step": 304
},
{
"epoch": 1.2348178137651822,
"grad_norm": 0.7039950927160753,
"learning_rate": 3.743712466184774e-05,
"loss": 0.3939,
"step": 305
},
{
"epoch": 1.2388663967611335,
"grad_norm": 0.6522249216047952,
"learning_rate": 3.7409356833155885e-05,
"loss": 0.4637,
"step": 306
},
{
"epoch": 1.242914979757085,
"grad_norm": 0.72491031441313,
"learning_rate": 3.73814497997107e-05,
"loss": 0.4672,
"step": 307
},
{
"epoch": 1.2469635627530364,
"grad_norm": 0.7019436047777963,
"learning_rate": 3.7353403784656095e-05,
"loss": 0.428,
"step": 308
},
{
"epoch": 1.2510121457489878,
"grad_norm": 0.5866221262120821,
"learning_rate": 3.732521901224724e-05,
"loss": 0.4245,
"step": 309
},
{
"epoch": 1.2550607287449393,
"grad_norm": 0.6301049554454239,
"learning_rate": 3.7296895707848844e-05,
"loss": 0.444,
"step": 310
},
{
"epoch": 1.2591093117408907,
"grad_norm": 0.6774969560380908,
"learning_rate": 3.7268434097933275e-05,
"loss": 0.4605,
"step": 311
},
{
"epoch": 1.263157894736842,
"grad_norm": 0.7383954779910289,
"learning_rate": 3.723983441007882e-05,
"loss": 0.4654,
"step": 312
},
{
"epoch": 1.2672064777327936,
"grad_norm": 0.6741397209136974,
"learning_rate": 3.721109687296781e-05,
"loss": 0.4953,
"step": 313
},
{
"epoch": 1.271255060728745,
"grad_norm": 1.7370605228096343,
"learning_rate": 3.718222171638484e-05,
"loss": 0.6126,
"step": 314
},
{
"epoch": 1.2753036437246963,
"grad_norm": 1.1148081599479835,
"learning_rate": 3.715320917121488e-05,
"loss": 0.5716,
"step": 315
},
{
"epoch": 1.2793522267206479,
"grad_norm": 0.6160587925645875,
"learning_rate": 3.712405946944149e-05,
"loss": 0.4449,
"step": 316
},
{
"epoch": 1.2834008097165992,
"grad_norm": 0.6553344452439642,
"learning_rate": 3.709477284414488e-05,
"loss": 0.4478,
"step": 317
},
{
"epoch": 1.2874493927125505,
"grad_norm": 0.6116034186990974,
"learning_rate": 3.7065349529500136e-05,
"loss": 0.4568,
"step": 318
},
{
"epoch": 1.291497975708502,
"grad_norm": 0.6473770144556205,
"learning_rate": 3.7035789760775305e-05,
"loss": 0.4265,
"step": 319
},
{
"epoch": 1.2955465587044535,
"grad_norm": 0.6178103193245419,
"learning_rate": 3.70060937743295e-05,
"loss": 0.4674,
"step": 320
},
{
"epoch": 1.2995951417004048,
"grad_norm": 0.6740521135791829,
"learning_rate": 3.6976261807611045e-05,
"loss": 0.4968,
"step": 321
},
{
"epoch": 1.3036437246963564,
"grad_norm": 0.6981821931311346,
"learning_rate": 3.694629409915555e-05,
"loss": 0.4639,
"step": 322
},
{
"epoch": 1.3076923076923077,
"grad_norm": 0.6334510135499043,
"learning_rate": 3.691619088858401e-05,
"loss": 0.4751,
"step": 323
},
{
"epoch": 1.311740890688259,
"grad_norm": 0.6086490021733854,
"learning_rate": 3.68859524166009e-05,
"loss": 0.4128,
"step": 324
},
{
"epoch": 1.3157894736842106,
"grad_norm": 0.745055098865499,
"learning_rate": 3.6855578924992226e-05,
"loss": 0.4668,
"step": 325
},
{
"epoch": 1.319838056680162,
"grad_norm": 0.5875301794760396,
"learning_rate": 3.6825070656623626e-05,
"loss": 0.4304,
"step": 326
},
{
"epoch": 1.3238866396761133,
"grad_norm": 0.6883848441833843,
"learning_rate": 3.67944278554384e-05,
"loss": 0.4496,
"step": 327
},
{
"epoch": 1.3279352226720649,
"grad_norm": 0.6360510237247594,
"learning_rate": 3.676365076645557e-05,
"loss": 0.4388,
"step": 328
},
{
"epoch": 1.3319838056680162,
"grad_norm": 0.5325578767421302,
"learning_rate": 3.673273963576791e-05,
"loss": 0.4133,
"step": 329
},
{
"epoch": 1.3360323886639676,
"grad_norm": 0.6029699139458182,
"learning_rate": 3.670169471054001e-05,
"loss": 0.4989,
"step": 330
},
{
"epoch": 1.3400809716599191,
"grad_norm": 0.6457831637077941,
"learning_rate": 3.6670516239006254e-05,
"loss": 0.4709,
"step": 331
},
{
"epoch": 1.3441295546558705,
"grad_norm": 0.6007096707492393,
"learning_rate": 3.663920447046888e-05,
"loss": 0.4441,
"step": 332
},
{
"epoch": 1.3481781376518218,
"grad_norm": 0.6327290611328664,
"learning_rate": 3.6607759655295954e-05,
"loss": 0.4507,
"step": 333
},
{
"epoch": 1.3522267206477734,
"grad_norm": 0.62188343632535,
"learning_rate": 3.657618204491937e-05,
"loss": 0.4841,
"step": 334
},
{
"epoch": 1.3562753036437247,
"grad_norm": 0.587358962892816,
"learning_rate": 3.6544471891832875e-05,
"loss": 0.4129,
"step": 335
},
{
"epoch": 1.360323886639676,
"grad_norm": 0.6401414232586959,
"learning_rate": 3.6512629449589983e-05,
"loss": 0.4774,
"step": 336
},
{
"epoch": 1.3643724696356276,
"grad_norm": 0.6892163769636455,
"learning_rate": 3.6480654972802045e-05,
"loss": 0.5494,
"step": 337
},
{
"epoch": 1.368421052631579,
"grad_norm": 0.6022855108458486,
"learning_rate": 3.644854871713611e-05,
"loss": 0.4399,
"step": 338
},
{
"epoch": 1.3724696356275303,
"grad_norm": 0.5897994910652277,
"learning_rate": 3.6416310939312924e-05,
"loss": 0.4698,
"step": 339
},
{
"epoch": 1.376518218623482,
"grad_norm": 0.7212652712573405,
"learning_rate": 3.638394189710493e-05,
"loss": 0.4621,
"step": 340
},
{
"epoch": 1.3805668016194332,
"grad_norm": 0.5869969523242061,
"learning_rate": 3.635144184933412e-05,
"loss": 0.4309,
"step": 341
},
{
"epoch": 1.3846153846153846,
"grad_norm": 0.7203165075356548,
"learning_rate": 3.6318811055869994e-05,
"loss": 0.4989,
"step": 342
},
{
"epoch": 1.3886639676113361,
"grad_norm": 0.5985716483699671,
"learning_rate": 3.628604977762752e-05,
"loss": 0.4886,
"step": 343
},
{
"epoch": 1.3927125506072875,
"grad_norm": 0.7509703868715388,
"learning_rate": 3.6253158276565004e-05,
"loss": 0.4321,
"step": 344
},
{
"epoch": 1.3967611336032388,
"grad_norm": 0.5739335470904157,
"learning_rate": 3.6220136815682016e-05,
"loss": 0.4213,
"step": 345
},
{
"epoch": 1.4008097165991904,
"grad_norm": 0.6018949962621126,
"learning_rate": 3.618698565901727e-05,
"loss": 0.4513,
"step": 346
},
{
"epoch": 1.4048582995951417,
"grad_norm": 0.6756027996858932,
"learning_rate": 3.615370507164654e-05,
"loss": 0.4201,
"step": 347
},
{
"epoch": 1.408906882591093,
"grad_norm": 0.7146290124588583,
"learning_rate": 3.61202953196805e-05,
"loss": 0.4288,
"step": 348
},
{
"epoch": 1.4129554655870447,
"grad_norm": 0.7926876411707231,
"learning_rate": 3.6086756670262654e-05,
"loss": 0.4621,
"step": 349
},
{
"epoch": 1.417004048582996,
"grad_norm": 0.6691248471463158,
"learning_rate": 3.6053089391567135e-05,
"loss": 0.4968,
"step": 350
},
{
"epoch": 1.4210526315789473,
"grad_norm": 0.5461631531216679,
"learning_rate": 3.601929375279662e-05,
"loss": 0.4272,
"step": 351
},
{
"epoch": 1.425101214574899,
"grad_norm": 0.5970493901167488,
"learning_rate": 3.598537002418012e-05,
"loss": 0.4721,
"step": 352
},
{
"epoch": 1.4291497975708503,
"grad_norm": 0.6585938748557942,
"learning_rate": 3.5951318476970885e-05,
"loss": 0.4775,
"step": 353
},
{
"epoch": 1.4331983805668016,
"grad_norm": 0.5469321762013973,
"learning_rate": 3.591713938344416e-05,
"loss": 0.4426,
"step": 354
},
{
"epoch": 1.4372469635627532,
"grad_norm": 0.5328618866162941,
"learning_rate": 3.588283301689507e-05,
"loss": 0.4034,
"step": 355
},
{
"epoch": 1.4412955465587045,
"grad_norm": 0.6319466400800867,
"learning_rate": 3.5848399651636424e-05,
"loss": 0.3963,
"step": 356
},
{
"epoch": 1.4453441295546559,
"grad_norm": 0.6570797653807263,
"learning_rate": 3.5813839562996473e-05,
"loss": 0.4159,
"step": 357
},
{
"epoch": 1.4493927125506074,
"grad_norm": 0.5376573593269107,
"learning_rate": 3.5779153027316775e-05,
"loss": 0.4012,
"step": 358
},
{
"epoch": 1.4534412955465588,
"grad_norm": 0.6587638689125079,
"learning_rate": 3.574434032194994e-05,
"loss": 0.603,
"step": 359
},
{
"epoch": 1.45748987854251,
"grad_norm": 0.635315970894013,
"learning_rate": 3.5709401725257436e-05,
"loss": 0.4447,
"step": 360
},
{
"epoch": 1.4615384615384617,
"grad_norm": 0.5636355570517985,
"learning_rate": 3.5674337516607346e-05,
"loss": 0.4378,
"step": 361
},
{
"epoch": 1.465587044534413,
"grad_norm": 0.6556549859862058,
"learning_rate": 3.563914797637216e-05,
"loss": 0.6674,
"step": 362
},
{
"epoch": 1.4696356275303644,
"grad_norm": 0.5656784647150669,
"learning_rate": 3.560383338592649e-05,
"loss": 0.4713,
"step": 363
},
{
"epoch": 1.4736842105263157,
"grad_norm": 0.6921488187245752,
"learning_rate": 3.556839402764487e-05,
"loss": 0.5067,
"step": 364
},
{
"epoch": 1.4777327935222673,
"grad_norm": 0.6280082768703517,
"learning_rate": 3.553283018489946e-05,
"loss": 0.5773,
"step": 365
},
{
"epoch": 1.4817813765182186,
"grad_norm": 0.6418181694777061,
"learning_rate": 3.54971421420578e-05,
"loss": 0.4184,
"step": 366
},
{
"epoch": 1.48582995951417,
"grad_norm": 0.5951911152509046,
"learning_rate": 3.5461330184480524e-05,
"loss": 0.4411,
"step": 367
},
{
"epoch": 1.4898785425101215,
"grad_norm": 0.5674287906119505,
"learning_rate": 3.54253945985191e-05,
"loss": 0.4399,
"step": 368
},
{
"epoch": 1.4939271255060729,
"grad_norm": 0.5810795608028063,
"learning_rate": 3.5389335671513524e-05,
"loss": 0.4669,
"step": 369
},
{
"epoch": 1.4979757085020242,
"grad_norm": 0.5905109572804165,
"learning_rate": 3.5353153691789986e-05,
"loss": 0.4605,
"step": 370
},
{
"epoch": 1.5020242914979756,
"grad_norm": 0.6580871344177097,
"learning_rate": 3.531684894865866e-05,
"loss": 0.429,
"step": 371
},
{
"epoch": 1.5060728744939271,
"grad_norm": 0.6724017401795713,
"learning_rate": 3.528042173241128e-05,
"loss": 0.4608,
"step": 372
},
{
"epoch": 1.5101214574898787,
"grad_norm": 0.6181945864775077,
"learning_rate": 3.524387233431892e-05,
"loss": 0.435,
"step": 373
},
{
"epoch": 1.5141700404858298,
"grad_norm": 0.7595877062163682,
"learning_rate": 3.520720104662958e-05,
"loss": 0.4976,
"step": 374
},
{
"epoch": 1.5182186234817814,
"grad_norm": 0.7031257042733826,
"learning_rate": 3.5170408162565904e-05,
"loss": 0.4715,
"step": 375
},
{
"epoch": 1.522267206477733,
"grad_norm": 0.7671232533410913,
"learning_rate": 3.513349397632281e-05,
"loss": 0.4423,
"step": 376
},
{
"epoch": 1.526315789473684,
"grad_norm": 0.6296696182118507,
"learning_rate": 3.5096458783065145e-05,
"loss": 0.4484,
"step": 377
},
{
"epoch": 1.5303643724696356,
"grad_norm": 0.6985083390928943,
"learning_rate": 3.505930287892533e-05,
"loss": 0.4701,
"step": 378
},
{
"epoch": 1.5344129554655872,
"grad_norm": 0.7086749882238011,
"learning_rate": 3.502202656100099e-05,
"loss": 0.4814,
"step": 379
},
{
"epoch": 1.5384615384615383,
"grad_norm": 0.6631503807776257,
"learning_rate": 3.498463012735256e-05,
"loss": 0.4638,
"step": 380
},
{
"epoch": 1.54251012145749,
"grad_norm": 0.755087734242397,
"learning_rate": 3.494711387700094e-05,
"loss": 0.4677,
"step": 381
},
{
"epoch": 1.5465587044534415,
"grad_norm": 0.58454630447729,
"learning_rate": 3.490947810992505e-05,
"loss": 0.4289,
"step": 382
},
{
"epoch": 1.5506072874493926,
"grad_norm": 0.6768030997093785,
"learning_rate": 3.48717231270595e-05,
"loss": 0.4414,
"step": 383
},
{
"epoch": 1.5546558704453441,
"grad_norm": 0.6486712648416931,
"learning_rate": 3.483384923029211e-05,
"loss": 0.4477,
"step": 384
},
{
"epoch": 1.5587044534412957,
"grad_norm": 0.6293295551835076,
"learning_rate": 3.479585672246156e-05,
"loss": 0.4593,
"step": 385
},
{
"epoch": 1.5627530364372468,
"grad_norm": 0.5959434214874487,
"learning_rate": 3.4757745907354924e-05,
"loss": 0.4712,
"step": 386
},
{
"epoch": 1.5668016194331984,
"grad_norm": 0.8760741981012458,
"learning_rate": 3.471951708970527e-05,
"loss": 0.5116,
"step": 387
},
{
"epoch": 1.5708502024291497,
"grad_norm": 0.7288221732364802,
"learning_rate": 3.468117057518921e-05,
"loss": 0.4709,
"step": 388
},
{
"epoch": 1.574898785425101,
"grad_norm": 0.711095065454808,
"learning_rate": 3.4642706670424426e-05,
"loss": 0.5027,
"step": 389
},
{
"epoch": 1.5789473684210527,
"grad_norm": 0.6173724159209037,
"learning_rate": 3.460412568296731e-05,
"loss": 0.4583,
"step": 390
},
{
"epoch": 1.582995951417004,
"grad_norm": 0.706562524735311,
"learning_rate": 3.456542792131039e-05,
"loss": 0.4805,
"step": 391
},
{
"epoch": 1.5870445344129553,
"grad_norm": 0.6711104806075694,
"learning_rate": 3.4526613694879936e-05,
"loss": 0.4597,
"step": 392
},
{
"epoch": 1.591093117408907,
"grad_norm": 0.6022869922669706,
"learning_rate": 3.448768331403348e-05,
"loss": 0.4494,
"step": 393
},
{
"epoch": 1.5951417004048583,
"grad_norm": 0.6254044147329975,
"learning_rate": 3.44486370900573e-05,
"loss": 0.4263,
"step": 394
},
{
"epoch": 1.5991902834008096,
"grad_norm": 0.6993324270024976,
"learning_rate": 3.440947533516396e-05,
"loss": 0.4705,
"step": 395
},
{
"epoch": 1.6032388663967612,
"grad_norm": 0.6619440502965959,
"learning_rate": 3.4370198362489816e-05,
"loss": 0.5376,
"step": 396
},
{
"epoch": 1.6072874493927125,
"grad_norm": 0.6604028434255416,
"learning_rate": 3.43308064860925e-05,
"loss": 0.4636,
"step": 397
},
{
"epoch": 1.6113360323886639,
"grad_norm": 0.6287000906351251,
"learning_rate": 3.4291300020948393e-05,
"loss": 0.4328,
"step": 398
},
{
"epoch": 1.6153846153846154,
"grad_norm": 0.5285456801710722,
"learning_rate": 3.425167928295015e-05,
"loss": 0.4473,
"step": 399
},
{
"epoch": 1.6194331983805668,
"grad_norm": 0.6406365778630155,
"learning_rate": 3.421194458890411e-05,
"loss": 0.6338,
"step": 400
},
{
"epoch": 1.623481781376518,
"grad_norm": 0.6742993061606378,
"learning_rate": 3.4172096256527865e-05,
"loss": 0.477,
"step": 401
},
{
"epoch": 1.6275303643724697,
"grad_norm": 0.5688605183259653,
"learning_rate": 3.4132134604447595e-05,
"loss": 0.4509,
"step": 402
},
{
"epoch": 1.631578947368421,
"grad_norm": 0.6429030610754473,
"learning_rate": 3.409205995219561e-05,
"loss": 0.4325,
"step": 403
},
{
"epoch": 1.6356275303643724,
"grad_norm": 0.6694434923869113,
"learning_rate": 3.4051872620207765e-05,
"loss": 0.4674,
"step": 404
},
{
"epoch": 1.639676113360324,
"grad_norm": 0.631096244347631,
"learning_rate": 3.40115729298209e-05,
"loss": 0.4668,
"step": 405
},
{
"epoch": 1.6437246963562753,
"grad_norm": 0.6889673831777977,
"learning_rate": 3.397116120327025e-05,
"loss": 0.4369,
"step": 406
},
{
"epoch": 1.6477732793522266,
"grad_norm": 0.5955658930339853,
"learning_rate": 3.3930637763686916e-05,
"loss": 0.4471,
"step": 407
},
{
"epoch": 1.6518218623481782,
"grad_norm": 0.5732543248980695,
"learning_rate": 3.389000293509524e-05,
"loss": 0.4637,
"step": 408
},
{
"epoch": 1.6558704453441295,
"grad_norm": 0.6348086751754315,
"learning_rate": 3.3849257042410225e-05,
"loss": 0.4323,
"step": 409
},
{
"epoch": 1.6599190283400809,
"grad_norm": 0.5797431828266197,
"learning_rate": 3.380840041143494e-05,
"loss": 0.4599,
"step": 410
},
{
"epoch": 1.6639676113360324,
"grad_norm": 0.7213559312319214,
"learning_rate": 3.376743336885792e-05,
"loss": 0.4648,
"step": 411
},
{
"epoch": 1.6680161943319838,
"grad_norm": 0.6358739308796337,
"learning_rate": 3.372635624225055e-05,
"loss": 0.4483,
"step": 412
},
{
"epoch": 1.6720647773279351,
"grad_norm": 0.7032270734895297,
"learning_rate": 3.3685169360064445e-05,
"loss": 0.4688,
"step": 413
},
{
"epoch": 1.6761133603238867,
"grad_norm": 0.5836653114781514,
"learning_rate": 3.3643873051628804e-05,
"loss": 0.4567,
"step": 414
},
{
"epoch": 1.680161943319838,
"grad_norm": 0.5726102238679637,
"learning_rate": 3.3602467647147814e-05,
"loss": 0.4568,
"step": 415
},
{
"epoch": 1.6842105263157894,
"grad_norm": 0.6430449433670126,
"learning_rate": 3.356095347769799e-05,
"loss": 0.4825,
"step": 416
},
{
"epoch": 1.688259109311741,
"grad_norm": 0.7487685601133842,
"learning_rate": 3.351933087522552e-05,
"loss": 0.4516,
"step": 417
},
{
"epoch": 1.6923076923076923,
"grad_norm": 0.680757145832017,
"learning_rate": 3.3477600172543626e-05,
"loss": 0.4868,
"step": 418
},
{
"epoch": 1.6963562753036436,
"grad_norm": 0.6642236663827918,
"learning_rate": 3.3435761703329894e-05,
"loss": 0.4712,
"step": 419
},
{
"epoch": 1.7004048582995952,
"grad_norm": 0.5856764297052381,
"learning_rate": 3.3393815802123606e-05,
"loss": 0.4181,
"step": 420
},
{
"epoch": 1.7044534412955465,
"grad_norm": 0.6196114841721043,
"learning_rate": 3.335176280432308e-05,
"loss": 0.4341,
"step": 421
},
{
"epoch": 1.708502024291498,
"grad_norm": 0.6538272863849872,
"learning_rate": 3.330960304618295e-05,
"loss": 0.4321,
"step": 422
},
{
"epoch": 1.7125506072874495,
"grad_norm": 0.6353336896807512,
"learning_rate": 3.326733686481153e-05,
"loss": 0.4553,
"step": 423
},
{
"epoch": 1.7165991902834008,
"grad_norm": 0.595765769106745,
"learning_rate": 3.3224964598168084e-05,
"loss": 0.4178,
"step": 424
},
{
"epoch": 1.7206477732793521,
"grad_norm": 0.6636070988348705,
"learning_rate": 3.318248658506012e-05,
"loss": 0.4386,
"step": 425
},
{
"epoch": 1.7246963562753037,
"grad_norm": 0.6592533630734908,
"learning_rate": 3.313990316514069e-05,
"loss": 0.4965,
"step": 426
},
{
"epoch": 1.728744939271255,
"grad_norm": 0.6745146871769158,
"learning_rate": 3.309721467890571e-05,
"loss": 0.4474,
"step": 427
},
{
"epoch": 1.7327935222672064,
"grad_norm": 0.7258223222772073,
"learning_rate": 3.305442146769115e-05,
"loss": 0.4836,
"step": 428
},
{
"epoch": 1.736842105263158,
"grad_norm": 0.5676325239198163,
"learning_rate": 3.30115238736704e-05,
"loss": 0.4291,
"step": 429
},
{
"epoch": 1.7408906882591093,
"grad_norm": 0.6249549593605349,
"learning_rate": 3.296852223985148e-05,
"loss": 0.4726,
"step": 430
},
{
"epoch": 1.7449392712550607,
"grad_norm": 0.6370028761287438,
"learning_rate": 3.2925416910074315e-05,
"loss": 0.4291,
"step": 431
},
{
"epoch": 1.7489878542510122,
"grad_norm": 0.6848836069807216,
"learning_rate": 3.288220822900796e-05,
"loss": 0.4432,
"step": 432
},
{
"epoch": 1.7530364372469636,
"grad_norm": 0.6012170993103692,
"learning_rate": 3.283889654214788e-05,
"loss": 0.434,
"step": 433
},
{
"epoch": 1.757085020242915,
"grad_norm": 0.7015545092215747,
"learning_rate": 3.279548219581317e-05,
"loss": 0.4535,
"step": 434
},
{
"epoch": 1.7611336032388665,
"grad_norm": 0.6711280161320416,
"learning_rate": 3.275196553714379e-05,
"loss": 0.4583,
"step": 435
},
{
"epoch": 1.7651821862348178,
"grad_norm": 0.680843043032745,
"learning_rate": 3.270834691409778e-05,
"loss": 0.466,
"step": 436
},
{
"epoch": 1.7692307692307692,
"grad_norm": 0.5884899137953711,
"learning_rate": 3.266462667544849e-05,
"loss": 0.4482,
"step": 437
},
{
"epoch": 1.7732793522267207,
"grad_norm": 0.6359107519006382,
"learning_rate": 3.262080517078178e-05,
"loss": 0.6471,
"step": 438
},
{
"epoch": 1.777327935222672,
"grad_norm": 0.5398601455130713,
"learning_rate": 3.257688275049323e-05,
"loss": 0.4317,
"step": 439
},
{
"epoch": 1.7813765182186234,
"grad_norm": 0.8325941116297061,
"learning_rate": 3.253285976578535e-05,
"loss": 0.4662,
"step": 440
},
{
"epoch": 1.785425101214575,
"grad_norm": 0.6462671875248893,
"learning_rate": 3.2488736568664756e-05,
"loss": 0.4838,
"step": 441
},
{
"epoch": 1.7894736842105263,
"grad_norm": 0.7457478992248949,
"learning_rate": 3.244451351193935e-05,
"loss": 0.4687,
"step": 442
},
{
"epoch": 1.7935222672064777,
"grad_norm": 0.6437340500526826,
"learning_rate": 3.240019094921551e-05,
"loss": 0.4093,
"step": 443
},
{
"epoch": 1.7975708502024292,
"grad_norm": 0.6764790285252201,
"learning_rate": 3.235576923489527e-05,
"loss": 0.4683,
"step": 444
},
{
"epoch": 1.8016194331983806,
"grad_norm": 0.6797334729262676,
"learning_rate": 3.2311248724173476e-05,
"loss": 0.5381,
"step": 445
},
{
"epoch": 1.805668016194332,
"grad_norm": 0.5847565472331825,
"learning_rate": 3.226662977303494e-05,
"loss": 0.4689,
"step": 446
},
{
"epoch": 1.8097165991902835,
"grad_norm": 0.6869348964669191,
"learning_rate": 3.2221912738251626e-05,
"loss": 0.4645,
"step": 447
},
{
"epoch": 1.8137651821862348,
"grad_norm": 0.601214089750986,
"learning_rate": 3.2177097977379736e-05,
"loss": 0.4626,
"step": 448
},
{
"epoch": 1.8178137651821862,
"grad_norm": 0.5882836287945677,
"learning_rate": 3.21321858487569e-05,
"loss": 0.4773,
"step": 449
},
{
"epoch": 1.8218623481781377,
"grad_norm": 0.6396997234047858,
"learning_rate": 3.208717671149931e-05,
"loss": 0.4164,
"step": 450
},
{
"epoch": 1.825910931174089,
"grad_norm": 0.7348800666946879,
"learning_rate": 3.204207092549882e-05,
"loss": 0.4428,
"step": 451
},
{
"epoch": 1.8299595141700404,
"grad_norm": 0.5901327242128245,
"learning_rate": 3.19968688514201e-05,
"loss": 0.4515,
"step": 452
},
{
"epoch": 1.834008097165992,
"grad_norm": 0.6871822267627641,
"learning_rate": 3.195157085069774e-05,
"loss": 0.4698,
"step": 453
},
{
"epoch": 1.8380566801619433,
"grad_norm": 0.7680561904964247,
"learning_rate": 3.190617728553332e-05,
"loss": 0.4893,
"step": 454
},
{
"epoch": 1.8421052631578947,
"grad_norm": 0.6655856621996441,
"learning_rate": 3.1860688518892606e-05,
"loss": 0.4607,
"step": 455
},
{
"epoch": 1.8461538461538463,
"grad_norm": 0.5060606735314179,
"learning_rate": 3.181510491450254e-05,
"loss": 0.4366,
"step": 456
},
{
"epoch": 1.8502024291497976,
"grad_norm": 0.5773907350151699,
"learning_rate": 3.176942683684842e-05,
"loss": 0.4179,
"step": 457
},
{
"epoch": 1.854251012145749,
"grad_norm": 0.5735905443738153,
"learning_rate": 3.1723654651170934e-05,
"loss": 0.4411,
"step": 458
},
{
"epoch": 1.8582995951417005,
"grad_norm": 0.5854048367136253,
"learning_rate": 3.167778872346327e-05,
"loss": 0.4005,
"step": 459
},
{
"epoch": 1.8623481781376519,
"grad_norm": 0.5417185584299404,
"learning_rate": 3.163182942046815e-05,
"loss": 0.4312,
"step": 460
},
{
"epoch": 1.8663967611336032,
"grad_norm": 0.5145608832234141,
"learning_rate": 3.158577710967494e-05,
"loss": 0.4369,
"step": 461
},
{
"epoch": 1.8704453441295548,
"grad_norm": 0.5981729196525216,
"learning_rate": 3.153963215931669e-05,
"loss": 0.4364,
"step": 462
},
{
"epoch": 1.874493927125506,
"grad_norm": 0.5952850819484111,
"learning_rate": 3.14933949383672e-05,
"loss": 0.456,
"step": 463
},
{
"epoch": 1.8785425101214575,
"grad_norm": 0.5317071862101128,
"learning_rate": 3.144706581653805e-05,
"loss": 0.4655,
"step": 464
},
{
"epoch": 1.882591093117409,
"grad_norm": 0.5377094592159462,
"learning_rate": 3.140064516427566e-05,
"loss": 0.5491,
"step": 465
},
{
"epoch": 1.8866396761133604,
"grad_norm": 0.6249706374049208,
"learning_rate": 3.135413335275833e-05,
"loss": 0.4942,
"step": 466
},
{
"epoch": 1.8906882591093117,
"grad_norm": 0.7302450889930377,
"learning_rate": 3.130753075389327e-05,
"loss": 0.4565,
"step": 467
},
{
"epoch": 1.8947368421052633,
"grad_norm": 0.6025036752610795,
"learning_rate": 3.1260837740313625e-05,
"loss": 0.5929,
"step": 468
},
{
"epoch": 1.8987854251012146,
"grad_norm": 0.5981487412063329,
"learning_rate": 3.121405468537547e-05,
"loss": 0.4199,
"step": 469
},
{
"epoch": 1.902834008097166,
"grad_norm": 0.5504257770365214,
"learning_rate": 3.116718196315488e-05,
"loss": 0.4286,
"step": 470
},
{
"epoch": 1.9068825910931175,
"grad_norm": 0.7012872670421934,
"learning_rate": 3.1120219948444894e-05,
"loss": 0.4615,
"step": 471
},
{
"epoch": 1.9109311740890689,
"grad_norm": 0.5619820367002932,
"learning_rate": 3.107316901675252e-05,
"loss": 0.4488,
"step": 472
},
{
"epoch": 1.9149797570850202,
"grad_norm": 0.6224572985647785,
"learning_rate": 3.1026029544295745e-05,
"loss": 0.4462,
"step": 473
},
{
"epoch": 1.9190283400809718,
"grad_norm": 0.5866898775129503,
"learning_rate": 3.097880190800057e-05,
"loss": 0.4281,
"step": 474
},
{
"epoch": 1.9230769230769231,
"grad_norm": 0.6166255437917166,
"learning_rate": 3.093148648549788e-05,
"loss": 0.4831,
"step": 475
},
{
"epoch": 1.9271255060728745,
"grad_norm": 0.6240777824104258,
"learning_rate": 3.088408365512055e-05,
"loss": 0.5622,
"step": 476
},
{
"epoch": 1.931174089068826,
"grad_norm": 0.62146396371291,
"learning_rate": 3.083659379590034e-05,
"loss": 0.5096,
"step": 477
},
{
"epoch": 1.9352226720647774,
"grad_norm": 0.5518457602596715,
"learning_rate": 3.0789017287564913e-05,
"loss": 0.4599,
"step": 478
},
{
"epoch": 1.9392712550607287,
"grad_norm": 0.6765327236259855,
"learning_rate": 3.074135451053475e-05,
"loss": 0.4336,
"step": 479
},
{
"epoch": 1.9433198380566803,
"grad_norm": 0.5936016999963062,
"learning_rate": 3.069360584592015e-05,
"loss": 0.4398,
"step": 480
},
{
"epoch": 1.9473684210526314,
"grad_norm": 0.5679409239888074,
"learning_rate": 3.064577167551816e-05,
"loss": 0.4047,
"step": 481
},
{
"epoch": 1.951417004048583,
"grad_norm": 0.4716537233517171,
"learning_rate": 3.059785238180953e-05,
"loss": 0.4411,
"step": 482
},
{
"epoch": 1.9554655870445345,
"grad_norm": 0.5888624002080831,
"learning_rate": 3.054984834795566e-05,
"loss": 0.4478,
"step": 483
},
{
"epoch": 1.9595141700404857,
"grad_norm": 0.5958215833104017,
"learning_rate": 3.0501759957795537e-05,
"loss": 0.4339,
"step": 484
},
{
"epoch": 1.9635627530364372,
"grad_norm": 0.6723468858952433,
"learning_rate": 3.0453587595842627e-05,
"loss": 0.462,
"step": 485
},
{
"epoch": 1.9676113360323888,
"grad_norm": 0.5657049622574739,
"learning_rate": 3.0405331647281856e-05,
"loss": 0.4766,
"step": 486
},
{
"epoch": 1.97165991902834,
"grad_norm": 0.5881959536793905,
"learning_rate": 3.0356992497966508e-05,
"loss": 0.6051,
"step": 487
},
{
"epoch": 1.9757085020242915,
"grad_norm": 0.5135669730845456,
"learning_rate": 3.0308570534415137e-05,
"loss": 0.4484,
"step": 488
},
{
"epoch": 1.979757085020243,
"grad_norm": 0.5404859586298284,
"learning_rate": 3.0260066143808467e-05,
"loss": 0.4267,
"step": 489
},
{
"epoch": 1.9838056680161942,
"grad_norm": 0.8095295446094322,
"learning_rate": 3.021147971398632e-05,
"loss": 0.4458,
"step": 490
},
{
"epoch": 1.9878542510121457,
"grad_norm": 0.6482797482442715,
"learning_rate": 3.0162811633444478e-05,
"loss": 0.485,
"step": 491
},
{
"epoch": 1.9919028340080973,
"grad_norm": 0.5973683126764877,
"learning_rate": 3.0114062291331635e-05,
"loss": 0.4345,
"step": 492
},
{
"epoch": 1.9959514170040484,
"grad_norm": 0.5779070274429924,
"learning_rate": 3.0065232077446217e-05,
"loss": 0.4485,
"step": 493
},
{
"epoch": 2.0,
"grad_norm": 0.6896477877892545,
"learning_rate": 3.001632138223332e-05,
"loss": 0.459,
"step": 494
},
{
"epoch": 2.0040485829959516,
"grad_norm": 0.6863075415232043,
"learning_rate": 2.9967330596781565e-05,
"loss": 0.3702,
"step": 495
},
{
"epoch": 2.0080971659919027,
"grad_norm": 0.5377278554129031,
"learning_rate": 2.9918260112819952e-05,
"loss": 0.3187,
"step": 496
},
{
"epoch": 2.0121457489878543,
"grad_norm": 0.8419772301419999,
"learning_rate": 2.9869110322714778e-05,
"loss": 0.3235,
"step": 497
},
{
"epoch": 2.016194331983806,
"grad_norm": 0.7489827780031193,
"learning_rate": 2.9819881619466447e-05,
"loss": 0.359,
"step": 498
},
{
"epoch": 2.020242914979757,
"grad_norm": 0.7005179663477468,
"learning_rate": 2.9770574396706364e-05,
"loss": 0.3025,
"step": 499
},
{
"epoch": 2.0242914979757085,
"grad_norm": 0.5785561725866095,
"learning_rate": 2.972118904869376e-05,
"loss": 0.3367,
"step": 500
},
{
"epoch": 2.02834008097166,
"grad_norm": 0.590048677671721,
"learning_rate": 2.9671725970312563e-05,
"loss": 0.2794,
"step": 501
},
{
"epoch": 2.032388663967611,
"grad_norm": 0.6143852270331712,
"learning_rate": 2.9622185557068238e-05,
"loss": 0.3165,
"step": 502
},
{
"epoch": 2.0364372469635628,
"grad_norm": 0.6620332997889308,
"learning_rate": 2.95725682050846e-05,
"loss": 0.2847,
"step": 503
},
{
"epoch": 2.0404858299595143,
"grad_norm": 0.5994405553327697,
"learning_rate": 2.9522874311100684e-05,
"loss": 0.3177,
"step": 504
},
{
"epoch": 2.0445344129554655,
"grad_norm": 0.6436005461663327,
"learning_rate": 2.9473104272467526e-05,
"loss": 0.2969,
"step": 505
},
{
"epoch": 2.048582995951417,
"grad_norm": 0.6670603788434294,
"learning_rate": 2.942325848714505e-05,
"loss": 0.2852,
"step": 506
},
{
"epoch": 2.0526315789473686,
"grad_norm": 0.5805245090556064,
"learning_rate": 2.9373337353698804e-05,
"loss": 0.2753,
"step": 507
},
{
"epoch": 2.0566801619433197,
"grad_norm": 0.5855364329202718,
"learning_rate": 2.932334127129686e-05,
"loss": 0.4217,
"step": 508
},
{
"epoch": 2.0607287449392713,
"grad_norm": 0.6144316174770214,
"learning_rate": 2.9273270639706547e-05,
"loss": 0.2692,
"step": 509
},
{
"epoch": 2.064777327935223,
"grad_norm": 0.5614574046733025,
"learning_rate": 2.922312585929131e-05,
"loss": 0.2596,
"step": 510
},
{
"epoch": 2.068825910931174,
"grad_norm": 0.5279891092188778,
"learning_rate": 2.9172907331007478e-05,
"loss": 0.3419,
"step": 511
},
{
"epoch": 2.0728744939271255,
"grad_norm": 0.5089432324383112,
"learning_rate": 2.9122615456401053e-05,
"loss": 0.4052,
"step": 512
},
{
"epoch": 2.076923076923077,
"grad_norm": 0.5374679483074964,
"learning_rate": 2.9072250637604535e-05,
"loss": 0.3787,
"step": 513
},
{
"epoch": 2.080971659919028,
"grad_norm": 0.5806848279640677,
"learning_rate": 2.9021813277333667e-05,
"loss": 0.2886,
"step": 514
},
{
"epoch": 2.08502024291498,
"grad_norm": 0.6081142655384225,
"learning_rate": 2.8971303778884244e-05,
"loss": 0.2787,
"step": 515
},
{
"epoch": 2.0890688259109313,
"grad_norm": 0.5526248525179729,
"learning_rate": 2.8920722546128863e-05,
"loss": 0.2732,
"step": 516
},
{
"epoch": 2.0931174089068825,
"grad_norm": 0.6032814972378581,
"learning_rate": 2.887006998351372e-05,
"loss": 0.4217,
"step": 517
},
{
"epoch": 2.097165991902834,
"grad_norm": 0.5366234862902751,
"learning_rate": 2.881934649605535e-05,
"loss": 0.3072,
"step": 518
},
{
"epoch": 2.1012145748987856,
"grad_norm": 0.5175041934583452,
"learning_rate": 2.8768552489337408e-05,
"loss": 0.4,
"step": 519
},
{
"epoch": 2.1052631578947367,
"grad_norm": 0.586494477532936,
"learning_rate": 2.8717688369507425e-05,
"loss": 0.292,
"step": 520
},
{
"epoch": 2.1093117408906883,
"grad_norm": 0.5081949431780672,
"learning_rate": 2.8666754543273535e-05,
"loss": 0.2852,
"step": 521
},
{
"epoch": 2.11336032388664,
"grad_norm": 0.5264460733776608,
"learning_rate": 2.8615751417901273e-05,
"loss": 0.3029,
"step": 522
},
{
"epoch": 2.117408906882591,
"grad_norm": 0.5862188532136131,
"learning_rate": 2.856467940121025e-05,
"loss": 0.3119,
"step": 523
},
{
"epoch": 2.1214574898785425,
"grad_norm": 0.70191442903375,
"learning_rate": 2.8513538901570964e-05,
"loss": 0.2855,
"step": 524
},
{
"epoch": 2.125506072874494,
"grad_norm": 0.5536453418145727,
"learning_rate": 2.846233032790149e-05,
"loss": 0.2857,
"step": 525
},
{
"epoch": 2.1295546558704452,
"grad_norm": 0.5621652006854014,
"learning_rate": 2.84110540896642e-05,
"loss": 0.2753,
"step": 526
},
{
"epoch": 2.133603238866397,
"grad_norm": 0.5497080290253992,
"learning_rate": 2.8359710596862555e-05,
"loss": 0.3103,
"step": 527
},
{
"epoch": 2.1376518218623484,
"grad_norm": 0.5876215772607103,
"learning_rate": 2.8308300260037734e-05,
"loss": 0.2704,
"step": 528
},
{
"epoch": 2.1417004048582995,
"grad_norm": 0.5805257904415427,
"learning_rate": 2.8256823490265424e-05,
"loss": 0.3307,
"step": 529
},
{
"epoch": 2.145748987854251,
"grad_norm": 0.4911349423701492,
"learning_rate": 2.8205280699152515e-05,
"loss": 0.3026,
"step": 530
},
{
"epoch": 2.1497975708502026,
"grad_norm": 0.7097912475555438,
"learning_rate": 2.8153672298833777e-05,
"loss": 0.3037,
"step": 531
},
{
"epoch": 2.1538461538461537,
"grad_norm": 0.5441969901190017,
"learning_rate": 2.8101998701968615e-05,
"loss": 0.3084,
"step": 532
},
{
"epoch": 2.1578947368421053,
"grad_norm": 0.5839950365634523,
"learning_rate": 2.805026032173773e-05,
"loss": 0.3121,
"step": 533
},
{
"epoch": 2.161943319838057,
"grad_norm": 0.5386457872726763,
"learning_rate": 2.7998457571839832e-05,
"loss": 0.3016,
"step": 534
},
{
"epoch": 2.165991902834008,
"grad_norm": 0.6725591074956714,
"learning_rate": 2.7946590866488322e-05,
"loss": 0.2761,
"step": 535
},
{
"epoch": 2.1700404858299596,
"grad_norm": 0.5505613360454471,
"learning_rate": 2.7894660620408e-05,
"loss": 0.3497,
"step": 536
},
{
"epoch": 2.174089068825911,
"grad_norm": 0.7317832002000151,
"learning_rate": 2.784266724883173e-05,
"loss": 0.2847,
"step": 537
},
{
"epoch": 2.1781376518218623,
"grad_norm": 0.529693697313566,
"learning_rate": 2.7790611167497128e-05,
"loss": 0.2679,
"step": 538
},
{
"epoch": 2.182186234817814,
"grad_norm": 0.5305975280525371,
"learning_rate": 2.7738492792643224e-05,
"loss": 0.2553,
"step": 539
},
{
"epoch": 2.1862348178137654,
"grad_norm": 0.6079263615167976,
"learning_rate": 2.768631254100716e-05,
"loss": 0.2666,
"step": 540
},
{
"epoch": 2.1902834008097165,
"grad_norm": 0.5466608170622903,
"learning_rate": 2.7634070829820836e-05,
"loss": 0.2355,
"step": 541
},
{
"epoch": 2.194331983805668,
"grad_norm": 0.5636144001569664,
"learning_rate": 2.7581768076807592e-05,
"loss": 0.3144,
"step": 542
},
{
"epoch": 2.1983805668016196,
"grad_norm": 0.5657260320728211,
"learning_rate": 2.7529404700178832e-05,
"loss": 0.2908,
"step": 543
},
{
"epoch": 2.2024291497975708,
"grad_norm": 0.6027949428572852,
"learning_rate": 2.7476981118630735e-05,
"loss": 0.2946,
"step": 544
},
{
"epoch": 2.2064777327935223,
"grad_norm": 0.5816837484120482,
"learning_rate": 2.742449775134085e-05,
"loss": 0.2844,
"step": 545
},
{
"epoch": 2.2105263157894735,
"grad_norm": 0.49698745704387876,
"learning_rate": 2.7371955017964788e-05,
"loss": 0.3173,
"step": 546
},
{
"epoch": 2.214574898785425,
"grad_norm": 0.5183442273821874,
"learning_rate": 2.7319353338632845e-05,
"loss": 0.2946,
"step": 547
},
{
"epoch": 2.2186234817813766,
"grad_norm": 0.5578950203708293,
"learning_rate": 2.7266693133946645e-05,
"loss": 0.299,
"step": 548
},
{
"epoch": 2.2226720647773277,
"grad_norm": 0.5707774384297403,
"learning_rate": 2.7213974824975785e-05,
"loss": 0.2836,
"step": 549
},
{
"epoch": 2.2267206477732793,
"grad_norm": 0.5189748036299769,
"learning_rate": 2.7161198833254464e-05,
"loss": 0.2833,
"step": 550
},
{
"epoch": 2.230769230769231,
"grad_norm": 0.5104605404415739,
"learning_rate": 2.7108365580778095e-05,
"loss": 0.3137,
"step": 551
},
{
"epoch": 2.234817813765182,
"grad_norm": 0.5185701462733111,
"learning_rate": 2.705547548999996e-05,
"loss": 0.2838,
"step": 552
},
{
"epoch": 2.2388663967611335,
"grad_norm": 0.5297555417684358,
"learning_rate": 2.7002528983827817e-05,
"loss": 0.3207,
"step": 553
},
{
"epoch": 2.242914979757085,
"grad_norm": 0.6179076293582583,
"learning_rate": 2.6949526485620517e-05,
"loss": 0.3633,
"step": 554
},
{
"epoch": 2.246963562753036,
"grad_norm": 0.6458098786746872,
"learning_rate": 2.6896468419184623e-05,
"loss": 0.328,
"step": 555
},
{
"epoch": 2.251012145748988,
"grad_norm": 0.5723020097888641,
"learning_rate": 2.6843355208771024e-05,
"loss": 0.2579,
"step": 556
},
{
"epoch": 2.2550607287449393,
"grad_norm": 0.5375049685858626,
"learning_rate": 2.6790187279071536e-05,
"loss": 0.2443,
"step": 557
},
{
"epoch": 2.2591093117408905,
"grad_norm": 0.5149558600830384,
"learning_rate": 2.67369650552155e-05,
"loss": 0.5526,
"step": 558
},
{
"epoch": 2.263157894736842,
"grad_norm": 0.5895630386982877,
"learning_rate": 2.6683688962766407e-05,
"loss": 0.3201,
"step": 559
},
{
"epoch": 2.2672064777327936,
"grad_norm": 0.5689276252339591,
"learning_rate": 2.6630359427718468e-05,
"loss": 0.2972,
"step": 560
},
{
"epoch": 2.2712550607287447,
"grad_norm": 0.5479033552410301,
"learning_rate": 2.6576976876493225e-05,
"loss": 0.2495,
"step": 561
},
{
"epoch": 2.2753036437246963,
"grad_norm": 0.5217680337459303,
"learning_rate": 2.652354173593614e-05,
"loss": 0.3291,
"step": 562
},
{
"epoch": 2.279352226720648,
"grad_norm": 0.5965954320504196,
"learning_rate": 2.647005443331317e-05,
"loss": 0.3202,
"step": 563
},
{
"epoch": 2.283400809716599,
"grad_norm": 0.5749986761331227,
"learning_rate": 2.6416515396307354e-05,
"loss": 0.4919,
"step": 564
},
{
"epoch": 2.2874493927125505,
"grad_norm": 0.5698543284763582,
"learning_rate": 2.636292505301542e-05,
"loss": 0.3428,
"step": 565
},
{
"epoch": 2.291497975708502,
"grad_norm": 0.52988181253324,
"learning_rate": 2.630928383194431e-05,
"loss": 0.3177,
"step": 566
},
{
"epoch": 2.2955465587044532,
"grad_norm": 0.5258720264682484,
"learning_rate": 2.625559216200781e-05,
"loss": 0.3258,
"step": 567
},
{
"epoch": 2.299595141700405,
"grad_norm": 0.5615531270979504,
"learning_rate": 2.620185047252308e-05,
"loss": 0.2856,
"step": 568
},
{
"epoch": 2.3036437246963564,
"grad_norm": 0.6048226111374874,
"learning_rate": 2.614805919320723e-05,
"loss": 0.3122,
"step": 569
},
{
"epoch": 2.3076923076923075,
"grad_norm": 0.5299650620937657,
"learning_rate": 2.6094218754173895e-05,
"loss": 0.2932,
"step": 570
},
{
"epoch": 2.311740890688259,
"grad_norm": 0.562503904125911,
"learning_rate": 2.60403295859298e-05,
"loss": 0.3764,
"step": 571
},
{
"epoch": 2.3157894736842106,
"grad_norm": 0.5428827857389575,
"learning_rate": 2.5986392119371297e-05,
"loss": 0.2722,
"step": 572
},
{
"epoch": 2.3198380566801617,
"grad_norm": 0.49502266721573635,
"learning_rate": 2.5932406785780928e-05,
"loss": 0.3269,
"step": 573
},
{
"epoch": 2.3238866396761133,
"grad_norm": 0.5508138602552168,
"learning_rate": 2.5878374016823987e-05,
"loss": 0.2844,
"step": 574
},
{
"epoch": 2.327935222672065,
"grad_norm": 0.5283244657265526,
"learning_rate": 2.5824294244545045e-05,
"loss": 0.2826,
"step": 575
},
{
"epoch": 2.331983805668016,
"grad_norm": 0.6314236188407515,
"learning_rate": 2.577016790136453e-05,
"loss": 0.2928,
"step": 576
},
{
"epoch": 2.3360323886639676,
"grad_norm": 0.6058111170627218,
"learning_rate": 2.5715995420075225e-05,
"loss": 0.3252,
"step": 577
},
{
"epoch": 2.340080971659919,
"grad_norm": 0.6386996956302337,
"learning_rate": 2.5661777233838864e-05,
"loss": 0.2437,
"step": 578
},
{
"epoch": 2.3441295546558703,
"grad_norm": 0.684047438329527,
"learning_rate": 2.5607513776182616e-05,
"loss": 0.2908,
"step": 579
},
{
"epoch": 2.348178137651822,
"grad_norm": 0.5801334488139187,
"learning_rate": 2.5553205480995633e-05,
"loss": 0.2713,
"step": 580
},
{
"epoch": 2.3522267206477734,
"grad_norm": 0.5308088313982777,
"learning_rate": 2.5498852782525604e-05,
"loss": 0.2965,
"step": 581
},
{
"epoch": 2.3562753036437245,
"grad_norm": 0.5729841766319898,
"learning_rate": 2.5444456115375252e-05,
"loss": 0.3003,
"step": 582
},
{
"epoch": 2.360323886639676,
"grad_norm": 0.6229848420513753,
"learning_rate": 2.539001591449889e-05,
"loss": 0.308,
"step": 583
},
{
"epoch": 2.3643724696356276,
"grad_norm": 0.5713909524057854,
"learning_rate": 2.53355326151989e-05,
"loss": 0.2774,
"step": 584
},
{
"epoch": 2.3684210526315788,
"grad_norm": 0.5259067482688524,
"learning_rate": 2.5281006653122288e-05,
"loss": 0.2977,
"step": 585
},
{
"epoch": 2.3724696356275303,
"grad_norm": 0.5279486559388696,
"learning_rate": 2.5226438464257217e-05,
"loss": 0.2965,
"step": 586
},
{
"epoch": 2.376518218623482,
"grad_norm": 0.5220717873236361,
"learning_rate": 2.5171828484929447e-05,
"loss": 0.2711,
"step": 587
},
{
"epoch": 2.380566801619433,
"grad_norm": 0.5586032453860489,
"learning_rate": 2.5117177151798934e-05,
"loss": 0.2927,
"step": 588
},
{
"epoch": 2.3846153846153846,
"grad_norm": 0.5842807663396684,
"learning_rate": 2.506248490185627e-05,
"loss": 0.2865,
"step": 589
},
{
"epoch": 2.388663967611336,
"grad_norm": 0.6358243044152757,
"learning_rate": 2.500775217241923e-05,
"loss": 0.3133,
"step": 590
},
{
"epoch": 2.3927125506072873,
"grad_norm": 0.5247371726661287,
"learning_rate": 2.4952979401129266e-05,
"loss": 0.3094,
"step": 591
},
{
"epoch": 2.396761133603239,
"grad_norm": 0.49128552786418384,
"learning_rate": 2.4898167025948e-05,
"loss": 0.3018,
"step": 592
},
{
"epoch": 2.4008097165991904,
"grad_norm": 0.6107014329993762,
"learning_rate": 2.4843315485153703e-05,
"loss": 0.33,
"step": 593
},
{
"epoch": 2.4048582995951415,
"grad_norm": 0.5983680167968852,
"learning_rate": 2.4788425217337856e-05,
"loss": 0.2901,
"step": 594
},
{
"epoch": 2.408906882591093,
"grad_norm": 0.5431101430814217,
"learning_rate": 2.4733496661401572e-05,
"loss": 0.29,
"step": 595
},
{
"epoch": 2.4129554655870447,
"grad_norm": 0.5442893180086952,
"learning_rate": 2.4678530256552114e-05,
"loss": 0.2881,
"step": 596
},
{
"epoch": 2.417004048582996,
"grad_norm": 0.5745234662038539,
"learning_rate": 2.46235264422994e-05,
"loss": 0.2904,
"step": 597
},
{
"epoch": 2.4210526315789473,
"grad_norm": 0.6125002932951534,
"learning_rate": 2.4568485658452452e-05,
"loss": 0.3078,
"step": 598
},
{
"epoch": 2.425101214574899,
"grad_norm": 0.5460955777063871,
"learning_rate": 2.451340834511592e-05,
"loss": 0.3137,
"step": 599
},
{
"epoch": 2.42914979757085,
"grad_norm": 0.47557874771382197,
"learning_rate": 2.4458294942686528e-05,
"loss": 0.2594,
"step": 600
},
{
"epoch": 2.4331983805668016,
"grad_norm": 0.5027364293043021,
"learning_rate": 2.440314589184958e-05,
"loss": 0.3081,
"step": 601
},
{
"epoch": 2.437246963562753,
"grad_norm": 0.5360807274899916,
"learning_rate": 2.434796163357541e-05,
"loss": 0.2806,
"step": 602
},
{
"epoch": 2.4412955465587043,
"grad_norm": 0.634090560101463,
"learning_rate": 2.429274260911588e-05,
"loss": 0.3037,
"step": 603
},
{
"epoch": 2.445344129554656,
"grad_norm": 0.5435070964840196,
"learning_rate": 2.423748926000084e-05,
"loss": 0.2385,
"step": 604
},
{
"epoch": 2.4493927125506074,
"grad_norm": 0.5278778196256292,
"learning_rate": 2.418220202803459e-05,
"loss": 0.2989,
"step": 605
},
{
"epoch": 2.4534412955465585,
"grad_norm": 0.5139915011503452,
"learning_rate": 2.412688135529237e-05,
"loss": 0.2706,
"step": 606
},
{
"epoch": 2.45748987854251,
"grad_norm": 0.5933109243361858,
"learning_rate": 2.4071527684116807e-05,
"loss": 0.2744,
"step": 607
},
{
"epoch": 2.4615384615384617,
"grad_norm": 0.6197052845695492,
"learning_rate": 2.4016141457114372e-05,
"loss": 0.2869,
"step": 608
},
{
"epoch": 2.465587044534413,
"grad_norm": 0.5093972558513972,
"learning_rate": 2.396072311715187e-05,
"loss": 0.3849,
"step": 609
},
{
"epoch": 2.4696356275303644,
"grad_norm": 0.5716301683189176,
"learning_rate": 2.3905273107352863e-05,
"loss": 0.2921,
"step": 610
},
{
"epoch": 2.473684210526316,
"grad_norm": 0.5149352520887946,
"learning_rate": 2.3849791871094157e-05,
"loss": 0.2823,
"step": 611
},
{
"epoch": 2.477732793522267,
"grad_norm": 0.5163903459347378,
"learning_rate": 2.379427985200223e-05,
"loss": 0.2783,
"step": 612
},
{
"epoch": 2.4817813765182186,
"grad_norm": 0.6104742949126789,
"learning_rate": 2.373873749394972e-05,
"loss": 0.301,
"step": 613
},
{
"epoch": 2.48582995951417,
"grad_norm": 0.5207925524021771,
"learning_rate": 2.3683165241051836e-05,
"loss": 0.29,
"step": 614
},
{
"epoch": 2.4898785425101213,
"grad_norm": 0.5514031879066985,
"learning_rate": 2.3627563537662823e-05,
"loss": 0.3735,
"step": 615
},
{
"epoch": 2.493927125506073,
"grad_norm": 0.5489605752611016,
"learning_rate": 2.3571932828372433e-05,
"loss": 0.2516,
"step": 616
},
{
"epoch": 2.4979757085020244,
"grad_norm": 0.5678809063083555,
"learning_rate": 2.351627355800233e-05,
"loss": 0.3145,
"step": 617
},
{
"epoch": 2.5020242914979756,
"grad_norm": 0.5084236891824778,
"learning_rate": 2.346058617160256e-05,
"loss": 0.3062,
"step": 618
},
{
"epoch": 2.506072874493927,
"grad_norm": 0.4944055548092932,
"learning_rate": 2.3404871114447976e-05,
"loss": 0.2863,
"step": 619
},
{
"epoch": 2.5101214574898787,
"grad_norm": 0.48648585290670493,
"learning_rate": 2.334912883203469e-05,
"loss": 0.3074,
"step": 620
},
{
"epoch": 2.51417004048583,
"grad_norm": 0.5562795829023484,
"learning_rate": 2.3293359770076505e-05,
"loss": 0.2992,
"step": 621
},
{
"epoch": 2.5182186234817814,
"grad_norm": 0.5152809085296434,
"learning_rate": 2.323756437450136e-05,
"loss": 0.2989,
"step": 622
},
{
"epoch": 2.522267206477733,
"grad_norm": 0.5444697190692834,
"learning_rate": 2.3181743091447754e-05,
"loss": 0.3018,
"step": 623
},
{
"epoch": 2.526315789473684,
"grad_norm": 0.48798674030889344,
"learning_rate": 2.3125896367261164e-05,
"loss": 0.2785,
"step": 624
},
{
"epoch": 2.5303643724696356,
"grad_norm": 0.5280259082895177,
"learning_rate": 2.307002464849052e-05,
"loss": 0.2706,
"step": 625
},
{
"epoch": 2.534412955465587,
"grad_norm": 0.5427013438413514,
"learning_rate": 2.30141283818846e-05,
"loss": 0.2692,
"step": 626
},
{
"epoch": 2.5384615384615383,
"grad_norm": 0.5916694250229152,
"learning_rate": 2.2958208014388462e-05,
"loss": 0.3068,
"step": 627
},
{
"epoch": 2.54251012145749,
"grad_norm": 0.5481187731304398,
"learning_rate": 2.2902263993139857e-05,
"loss": 0.3275,
"step": 628
},
{
"epoch": 2.5465587044534415,
"grad_norm": 0.5512033670944808,
"learning_rate": 2.2846296765465708e-05,
"loss": 0.2551,
"step": 629
},
{
"epoch": 2.5506072874493926,
"grad_norm": 0.5773550115199773,
"learning_rate": 2.2790306778878465e-05,
"loss": 0.2492,
"step": 630
},
{
"epoch": 2.554655870445344,
"grad_norm": 0.5629785720100552,
"learning_rate": 2.273429448107257e-05,
"loss": 0.2901,
"step": 631
},
{
"epoch": 2.5587044534412957,
"grad_norm": 0.6098747940740965,
"learning_rate": 2.2678260319920868e-05,
"loss": 0.3117,
"step": 632
},
{
"epoch": 2.562753036437247,
"grad_norm": 0.5816524762220543,
"learning_rate": 2.262220474347102e-05,
"loss": 0.2884,
"step": 633
},
{
"epoch": 2.5668016194331984,
"grad_norm": 0.6125340303930805,
"learning_rate": 2.2566128199941918e-05,
"loss": 0.3005,
"step": 634
},
{
"epoch": 2.57085020242915,
"grad_norm": 0.6246072739989248,
"learning_rate": 2.2510031137720104e-05,
"loss": 0.3251,
"step": 635
},
{
"epoch": 2.574898785425101,
"grad_norm": 0.5568188991582472,
"learning_rate": 2.2453914005356206e-05,
"loss": 0.2886,
"step": 636
},
{
"epoch": 2.5789473684210527,
"grad_norm": 0.5767793467914325,
"learning_rate": 2.2397777251561302e-05,
"loss": 0.3047,
"step": 637
},
{
"epoch": 2.582995951417004,
"grad_norm": 0.613694407727536,
"learning_rate": 2.2341621325203393e-05,
"loss": 0.2835,
"step": 638
},
{
"epoch": 2.5870445344129553,
"grad_norm": 0.5643994573358646,
"learning_rate": 2.2285446675303764e-05,
"loss": 0.2918,
"step": 639
},
{
"epoch": 2.591093117408907,
"grad_norm": 0.49406350931202514,
"learning_rate": 2.2229253751033423e-05,
"loss": 0.2959,
"step": 640
},
{
"epoch": 2.5951417004048585,
"grad_norm": 0.57020642418623,
"learning_rate": 2.2173043001709492e-05,
"loss": 0.3336,
"step": 641
},
{
"epoch": 2.5991902834008096,
"grad_norm": 0.6355654620355419,
"learning_rate": 2.2116814876791627e-05,
"loss": 0.2852,
"step": 642
},
{
"epoch": 2.603238866396761,
"grad_norm": 0.5465224087967067,
"learning_rate": 2.2060569825878413e-05,
"loss": 0.302,
"step": 643
},
{
"epoch": 2.6072874493927127,
"grad_norm": 0.5399395784134771,
"learning_rate": 2.2004308298703777e-05,
"loss": 0.2601,
"step": 644
},
{
"epoch": 2.611336032388664,
"grad_norm": 0.5844166725661964,
"learning_rate": 2.1948030745133396e-05,
"loss": 0.2524,
"step": 645
},
{
"epoch": 2.6153846153846154,
"grad_norm": 0.6653304529213336,
"learning_rate": 2.1891737615161083e-05,
"loss": 0.2884,
"step": 646
},
{
"epoch": 2.619433198380567,
"grad_norm": 0.5858923687497104,
"learning_rate": 2.1835429358905218e-05,
"loss": 0.2876,
"step": 647
},
{
"epoch": 2.623481781376518,
"grad_norm": 0.5538931719815615,
"learning_rate": 2.17791064266051e-05,
"loss": 0.2893,
"step": 648
},
{
"epoch": 2.6275303643724697,
"grad_norm": 0.5066814603018065,
"learning_rate": 2.172276926861741e-05,
"loss": 0.3209,
"step": 649
},
{
"epoch": 2.6315789473684212,
"grad_norm": 0.4945298599655192,
"learning_rate": 2.166641833541255e-05,
"loss": 0.3072,
"step": 650
},
{
"epoch": 2.6356275303643724,
"grad_norm": 0.5632369607494266,
"learning_rate": 2.161005407757109e-05,
"loss": 0.2924,
"step": 651
},
{
"epoch": 2.639676113360324,
"grad_norm": 0.6008408043157261,
"learning_rate": 2.155367694578013e-05,
"loss": 0.2978,
"step": 652
},
{
"epoch": 2.6437246963562755,
"grad_norm": 0.5801709845297643,
"learning_rate": 2.1497287390829712e-05,
"loss": 0.2805,
"step": 653
},
{
"epoch": 2.6477732793522266,
"grad_norm": 0.5675232066332587,
"learning_rate": 2.1440885863609208e-05,
"loss": 0.2945,
"step": 654
},
{
"epoch": 2.651821862348178,
"grad_norm": 0.6616434124358128,
"learning_rate": 2.1384472815103737e-05,
"loss": 0.3213,
"step": 655
},
{
"epoch": 2.6558704453441297,
"grad_norm": 0.6182666914028374,
"learning_rate": 2.1328048696390534e-05,
"loss": 0.3015,
"step": 656
},
{
"epoch": 2.659919028340081,
"grad_norm": 0.4932367745290688,
"learning_rate": 2.1271613958635342e-05,
"loss": 0.3159,
"step": 657
},
{
"epoch": 2.6639676113360324,
"grad_norm": 0.511112479639173,
"learning_rate": 2.1215169053088825e-05,
"loss": 0.302,
"step": 658
},
{
"epoch": 2.668016194331984,
"grad_norm": 0.5189430188949875,
"learning_rate": 2.1158714431082937e-05,
"loss": 0.3147,
"step": 659
},
{
"epoch": 2.672064777327935,
"grad_norm": 0.5257150824361613,
"learning_rate": 2.110225054402733e-05,
"loss": 0.2679,
"step": 660
},
{
"epoch": 2.6761133603238867,
"grad_norm": 0.5476643131392643,
"learning_rate": 2.1045777843405747e-05,
"loss": 0.3098,
"step": 661
},
{
"epoch": 2.6801619433198383,
"grad_norm": 0.5629284237151541,
"learning_rate": 2.09892967807724e-05,
"loss": 0.3085,
"step": 662
},
{
"epoch": 2.6842105263157894,
"grad_norm": 0.5457688540653842,
"learning_rate": 2.093280780774835e-05,
"loss": 0.3096,
"step": 663
},
{
"epoch": 2.688259109311741,
"grad_norm": 0.6196009318405132,
"learning_rate": 2.087631137601793e-05,
"loss": 0.3215,
"step": 664
},
{
"epoch": 2.6923076923076925,
"grad_norm": 0.4937251442848098,
"learning_rate": 2.0819807937325104e-05,
"loss": 0.356,
"step": 665
},
{
"epoch": 2.6963562753036436,
"grad_norm": 0.4896875351348081,
"learning_rate": 2.0763297943469847e-05,
"loss": 0.4753,
"step": 666
},
{
"epoch": 2.700404858299595,
"grad_norm": 0.5513177352625394,
"learning_rate": 2.070678184630458e-05,
"loss": 0.2396,
"step": 667
},
{
"epoch": 2.7044534412955468,
"grad_norm": 0.5830063895027247,
"learning_rate": 2.06502600977305e-05,
"loss": 0.2511,
"step": 668
},
{
"epoch": 2.708502024291498,
"grad_norm": 0.48220310347427314,
"learning_rate": 2.0593733149694008e-05,
"loss": 0.2722,
"step": 669
},
{
"epoch": 2.7125506072874495,
"grad_norm": 0.49300888128314085,
"learning_rate": 2.0537201454183074e-05,
"loss": 0.2686,
"step": 670
},
{
"epoch": 2.716599190283401,
"grad_norm": 0.6185481374353008,
"learning_rate": 2.0480665463223626e-05,
"loss": 0.3168,
"step": 671
},
{
"epoch": 2.720647773279352,
"grad_norm": 0.5301052044704373,
"learning_rate": 2.042412562887594e-05,
"loss": 0.2657,
"step": 672
},
{
"epoch": 2.7246963562753037,
"grad_norm": 0.5363580018046086,
"learning_rate": 2.0367582403231033e-05,
"loss": 0.3294,
"step": 673
},
{
"epoch": 2.7287449392712553,
"grad_norm": 0.5315565775626662,
"learning_rate": 2.0311036238407023e-05,
"loss": 0.3041,
"step": 674
},
{
"epoch": 2.7327935222672064,
"grad_norm": 0.5602874237176908,
"learning_rate": 2.0254487586545534e-05,
"loss": 0.2917,
"step": 675
},
{
"epoch": 2.736842105263158,
"grad_norm": 0.5713773750060585,
"learning_rate": 2.0197936899808088e-05,
"loss": 0.2985,
"step": 676
},
{
"epoch": 2.7408906882591095,
"grad_norm": 0.5606336950712749,
"learning_rate": 2.0141384630372466e-05,
"loss": 0.3592,
"step": 677
},
{
"epoch": 2.7449392712550607,
"grad_norm": 0.5122889544784002,
"learning_rate": 2.0084831230429106e-05,
"loss": 0.3084,
"step": 678
},
{
"epoch": 2.748987854251012,
"grad_norm": 0.5672215686009403,
"learning_rate": 2.0028277152177492e-05,
"loss": 0.309,
"step": 679
},
{
"epoch": 2.753036437246964,
"grad_norm": 0.6439426152376158,
"learning_rate": 1.9971722847822518e-05,
"loss": 0.3099,
"step": 680
},
{
"epoch": 2.757085020242915,
"grad_norm": 0.5369673918596392,
"learning_rate": 1.99151687695709e-05,
"loss": 0.3138,
"step": 681
},
{
"epoch": 2.7611336032388665,
"grad_norm": 0.5465418322209264,
"learning_rate": 1.9858615369627537e-05,
"loss": 0.2633,
"step": 682
},
{
"epoch": 2.765182186234818,
"grad_norm": 0.5041863765028758,
"learning_rate": 1.9802063100191916e-05,
"loss": 0.272,
"step": 683
},
{
"epoch": 2.769230769230769,
"grad_norm": 0.5663515202614272,
"learning_rate": 1.974551241345447e-05,
"loss": 0.3062,
"step": 684
},
{
"epoch": 2.7732793522267207,
"grad_norm": 0.540873290390128,
"learning_rate": 1.9688963761592984e-05,
"loss": 0.2849,
"step": 685
},
{
"epoch": 2.7773279352226723,
"grad_norm": 0.49482669761354264,
"learning_rate": 1.963241759676897e-05,
"loss": 0.2709,
"step": 686
},
{
"epoch": 2.7813765182186234,
"grad_norm": 0.543867631752503,
"learning_rate": 1.9575874371124066e-05,
"loss": 0.3274,
"step": 687
},
{
"epoch": 2.785425101214575,
"grad_norm": 0.5441434390104223,
"learning_rate": 1.951933453677638e-05,
"loss": 0.3478,
"step": 688
},
{
"epoch": 2.7894736842105265,
"grad_norm": 0.5679395447858422,
"learning_rate": 1.9462798545816932e-05,
"loss": 0.2841,
"step": 689
},
{
"epoch": 2.7935222672064777,
"grad_norm": 0.6097684780449902,
"learning_rate": 1.9406266850305996e-05,
"loss": 0.2898,
"step": 690
},
{
"epoch": 2.7975708502024292,
"grad_norm": 0.49188774753875303,
"learning_rate": 1.9349739902269503e-05,
"loss": 0.2896,
"step": 691
},
{
"epoch": 2.801619433198381,
"grad_norm": 0.6330440299558534,
"learning_rate": 1.9293218153695425e-05,
"loss": 0.2724,
"step": 692
},
{
"epoch": 2.805668016194332,
"grad_norm": 0.5832439305288611,
"learning_rate": 1.9236702056530153e-05,
"loss": 0.278,
"step": 693
},
{
"epoch": 2.8097165991902835,
"grad_norm": 0.5543391954113941,
"learning_rate": 1.9180192062674907e-05,
"loss": 0.2783,
"step": 694
},
{
"epoch": 2.813765182186235,
"grad_norm": 0.528703706349412,
"learning_rate": 1.9123688623982076e-05,
"loss": 0.2853,
"step": 695
},
{
"epoch": 2.817813765182186,
"grad_norm": 0.5105889524990634,
"learning_rate": 1.906719219225165e-05,
"loss": 0.4498,
"step": 696
},
{
"epoch": 2.8218623481781377,
"grad_norm": 0.5064289582365014,
"learning_rate": 1.9010703219227603e-05,
"loss": 0.2955,
"step": 697
},
{
"epoch": 2.8259109311740893,
"grad_norm": 0.5280982614758641,
"learning_rate": 1.8954222156594256e-05,
"loss": 0.2894,
"step": 698
},
{
"epoch": 2.8299595141700404,
"grad_norm": 0.5450907591802948,
"learning_rate": 1.8897749455972673e-05,
"loss": 0.4174,
"step": 699
},
{
"epoch": 2.834008097165992,
"grad_norm": 0.5111445705643317,
"learning_rate": 1.884128556891707e-05,
"loss": 0.274,
"step": 700
},
{
"epoch": 2.8380566801619436,
"grad_norm": 0.5135570317550712,
"learning_rate": 1.878483094691119e-05,
"loss": 0.2597,
"step": 701
},
{
"epoch": 2.8421052631578947,
"grad_norm": 0.527417905374612,
"learning_rate": 1.8728386041364664e-05,
"loss": 0.3021,
"step": 702
},
{
"epoch": 2.8461538461538463,
"grad_norm": 0.4884679635632646,
"learning_rate": 1.867195130360947e-05,
"loss": 0.2974,
"step": 703
},
{
"epoch": 2.850202429149798,
"grad_norm": 0.4914919622492545,
"learning_rate": 1.8615527184896263e-05,
"loss": 0.2871,
"step": 704
},
{
"epoch": 2.854251012145749,
"grad_norm": 0.58337987992248,
"learning_rate": 1.8559114136390795e-05,
"loss": 0.2645,
"step": 705
},
{
"epoch": 2.8582995951417005,
"grad_norm": 0.535088689274119,
"learning_rate": 1.8502712609170298e-05,
"loss": 0.28,
"step": 706
},
{
"epoch": 2.862348178137652,
"grad_norm": 0.5650829953957726,
"learning_rate": 1.8446323054219876e-05,
"loss": 0.3078,
"step": 707
},
{
"epoch": 2.866396761133603,
"grad_norm": 0.5796818534366384,
"learning_rate": 1.838994592242891e-05,
"loss": 0.3011,
"step": 708
},
{
"epoch": 2.8704453441295548,
"grad_norm": 0.47446291787540007,
"learning_rate": 1.8333581664587453e-05,
"loss": 0.3271,
"step": 709
},
{
"epoch": 2.8744939271255063,
"grad_norm": 0.5492157271219826,
"learning_rate": 1.8277230731382593e-05,
"loss": 0.3002,
"step": 710
},
{
"epoch": 2.8785425101214575,
"grad_norm": 0.5899539528272418,
"learning_rate": 1.82208935733949e-05,
"loss": 0.3091,
"step": 711
},
{
"epoch": 2.882591093117409,
"grad_norm": 0.5352154341375388,
"learning_rate": 1.8164570641094793e-05,
"loss": 0.3024,
"step": 712
},
{
"epoch": 2.8866396761133606,
"grad_norm": 0.5659697108184499,
"learning_rate": 1.810826238483892e-05,
"loss": 0.2751,
"step": 713
},
{
"epoch": 2.8906882591093117,
"grad_norm": 0.5047075291045129,
"learning_rate": 1.8051969254866608e-05,
"loss": 0.3029,
"step": 714
},
{
"epoch": 2.8947368421052633,
"grad_norm": 0.5300932495632037,
"learning_rate": 1.7995691701296226e-05,
"loss": 0.2987,
"step": 715
},
{
"epoch": 2.898785425101215,
"grad_norm": 0.5704375312558276,
"learning_rate": 1.7939430174121593e-05,
"loss": 0.2988,
"step": 716
},
{
"epoch": 2.902834008097166,
"grad_norm": 0.5580021289745067,
"learning_rate": 1.788318512320838e-05,
"loss": 0.321,
"step": 717
},
{
"epoch": 2.9068825910931175,
"grad_norm": 0.5134138365344855,
"learning_rate": 1.782695699829051e-05,
"loss": 0.2987,
"step": 718
},
{
"epoch": 2.910931174089069,
"grad_norm": 0.5242517239809984,
"learning_rate": 1.7770746248966587e-05,
"loss": 0.2977,
"step": 719
},
{
"epoch": 2.91497975708502,
"grad_norm": 0.5623890234229928,
"learning_rate": 1.7714553324696243e-05,
"loss": 0.2867,
"step": 720
},
{
"epoch": 2.919028340080972,
"grad_norm": 0.5670962297738459,
"learning_rate": 1.7658378674796614e-05,
"loss": 0.2625,
"step": 721
},
{
"epoch": 2.9230769230769234,
"grad_norm": 0.6442344913289499,
"learning_rate": 1.76022227484387e-05,
"loss": 0.3377,
"step": 722
},
{
"epoch": 2.9271255060728745,
"grad_norm": 0.5659561349719078,
"learning_rate": 1.7546085994643807e-05,
"loss": 0.2993,
"step": 723
},
{
"epoch": 2.931174089068826,
"grad_norm": 0.5450944016071032,
"learning_rate": 1.7489968862279902e-05,
"loss": 0.2966,
"step": 724
},
{
"epoch": 2.9352226720647776,
"grad_norm": 0.5449980003848247,
"learning_rate": 1.743387180005809e-05,
"loss": 0.302,
"step": 725
},
{
"epoch": 2.9392712550607287,
"grad_norm": 0.5766321663134094,
"learning_rate": 1.737779525652899e-05,
"loss": 0.2878,
"step": 726
},
{
"epoch": 2.9433198380566803,
"grad_norm": 0.534457749724714,
"learning_rate": 1.732173968007914e-05,
"loss": 0.2807,
"step": 727
},
{
"epoch": 2.9473684210526314,
"grad_norm": 0.5117666103896085,
"learning_rate": 1.7265705518927437e-05,
"loss": 0.2642,
"step": 728
},
{
"epoch": 2.951417004048583,
"grad_norm": 0.5641731431816115,
"learning_rate": 1.7209693221121542e-05,
"loss": 0.2782,
"step": 729
},
{
"epoch": 2.9554655870445345,
"grad_norm": 0.5647595530393342,
"learning_rate": 1.7153703234534302e-05,
"loss": 0.2816,
"step": 730
},
{
"epoch": 2.9595141700404857,
"grad_norm": 0.4503310018645019,
"learning_rate": 1.709773600686015e-05,
"loss": 0.3272,
"step": 731
},
{
"epoch": 2.9635627530364372,
"grad_norm": 0.5979304531886177,
"learning_rate": 1.7041791985611548e-05,
"loss": 0.3168,
"step": 732
},
{
"epoch": 2.967611336032389,
"grad_norm": 0.5134738118290566,
"learning_rate": 1.69858716181154e-05,
"loss": 0.2928,
"step": 733
},
{
"epoch": 2.97165991902834,
"grad_norm": 0.6146061652482667,
"learning_rate": 1.692997535150948e-05,
"loss": 0.2916,
"step": 734
},
{
"epoch": 2.9757085020242915,
"grad_norm": 0.49968667556653285,
"learning_rate": 1.687410363273884e-05,
"loss": 0.3066,
"step": 735
},
{
"epoch": 2.979757085020243,
"grad_norm": 0.5076864769638904,
"learning_rate": 1.6818256908552257e-05,
"loss": 0.2793,
"step": 736
},
{
"epoch": 2.983805668016194,
"grad_norm": 0.5079136270306946,
"learning_rate": 1.6762435625498646e-05,
"loss": 0.2891,
"step": 737
},
{
"epoch": 2.9878542510121457,
"grad_norm": 0.5371460210640968,
"learning_rate": 1.67066402299235e-05,
"loss": 0.2562,
"step": 738
},
{
"epoch": 2.9919028340080973,
"grad_norm": 0.5439422077565929,
"learning_rate": 1.6650871167965313e-05,
"loss": 0.2972,
"step": 739
},
{
"epoch": 2.9959514170040484,
"grad_norm": 0.5472769491370766,
"learning_rate": 1.6595128885552028e-05,
"loss": 0.2692,
"step": 740
},
{
"epoch": 3.0,
"grad_norm": 0.5231904970662975,
"learning_rate": 1.653941382839745e-05,
"loss": 0.2623,
"step": 741
},
{
"epoch": 3.0040485829959516,
"grad_norm": 0.4804366553674705,
"learning_rate": 1.6483726441997673e-05,
"loss": 0.1497,
"step": 742
},
{
"epoch": 3.0080971659919027,
"grad_norm": 0.4978347646286166,
"learning_rate": 1.642806717162757e-05,
"loss": 0.1718,
"step": 743
},
{
"epoch": 3.0121457489878543,
"grad_norm": 0.440964044336915,
"learning_rate": 1.637243646233718e-05,
"loss": 0.1766,
"step": 744
},
{
"epoch": 3.016194331983806,
"grad_norm": 0.4638357058900949,
"learning_rate": 1.6316834758948174e-05,
"loss": 0.2342,
"step": 745
},
{
"epoch": 3.020242914979757,
"grad_norm": 0.5373484018680199,
"learning_rate": 1.6261262506050282e-05,
"loss": 0.1544,
"step": 746
},
{
"epoch": 3.0242914979757085,
"grad_norm": 0.5637974749166524,
"learning_rate": 1.620572014799777e-05,
"loss": 0.2089,
"step": 747
},
{
"epoch": 3.02834008097166,
"grad_norm": 0.505192435984692,
"learning_rate": 1.6150208128905857e-05,
"loss": 0.2368,
"step": 748
},
{
"epoch": 3.032388663967611,
"grad_norm": 0.5215385663766076,
"learning_rate": 1.6094726892647147e-05,
"loss": 0.1568,
"step": 749
},
{
"epoch": 3.0364372469635628,
"grad_norm": 0.5263885845432145,
"learning_rate": 1.6039276882848135e-05,
"loss": 0.1517,
"step": 750
},
{
"epoch": 3.0404858299595143,
"grad_norm": 0.5285153863535246,
"learning_rate": 1.5983858542885635e-05,
"loss": 0.2205,
"step": 751
},
{
"epoch": 3.0445344129554655,
"grad_norm": 0.5287737883781297,
"learning_rate": 1.5928472315883203e-05,
"loss": 0.1959,
"step": 752
},
{
"epoch": 3.048582995951417,
"grad_norm": 0.48247270061296715,
"learning_rate": 1.5873118644707633e-05,
"loss": 0.1608,
"step": 753
},
{
"epoch": 3.0526315789473686,
"grad_norm": 0.43757441353512155,
"learning_rate": 1.5817797971965413e-05,
"loss": 0.1585,
"step": 754
},
{
"epoch": 3.0566801619433197,
"grad_norm": 0.49382369129792486,
"learning_rate": 1.576251073999917e-05,
"loss": 0.1908,
"step": 755
},
{
"epoch": 3.0607287449392713,
"grad_norm": 0.48903085642866834,
"learning_rate": 1.5707257390884126e-05,
"loss": 0.1725,
"step": 756
},
{
"epoch": 3.064777327935223,
"grad_norm": 0.4493118020962484,
"learning_rate": 1.5652038366424595e-05,
"loss": 0.1634,
"step": 757
},
{
"epoch": 3.068825910931174,
"grad_norm": 0.49831277905950466,
"learning_rate": 1.5596854108150424e-05,
"loss": 0.1952,
"step": 758
},
{
"epoch": 3.0728744939271255,
"grad_norm": 0.4698327125584696,
"learning_rate": 1.5541705057313476e-05,
"loss": 0.1739,
"step": 759
},
{
"epoch": 3.076923076923077,
"grad_norm": 0.4789726256413967,
"learning_rate": 1.5486591654884086e-05,
"loss": 0.171,
"step": 760
},
{
"epoch": 3.080971659919028,
"grad_norm": 0.46206335185623615,
"learning_rate": 1.543151434154755e-05,
"loss": 0.197,
"step": 761
},
{
"epoch": 3.08502024291498,
"grad_norm": 0.4842407809146105,
"learning_rate": 1.5376473557700615e-05,
"loss": 0.1618,
"step": 762
},
{
"epoch": 3.0890688259109313,
"grad_norm": 0.4498050041680538,
"learning_rate": 1.5321469743447893e-05,
"loss": 0.1654,
"step": 763
},
{
"epoch": 3.0931174089068825,
"grad_norm": 0.46838573601408834,
"learning_rate": 1.5266503338598434e-05,
"loss": 0.1985,
"step": 764
},
{
"epoch": 3.097165991902834,
"grad_norm": 0.4630412331524798,
"learning_rate": 1.5211574782662149e-05,
"loss": 0.1758,
"step": 765
},
{
"epoch": 3.1012145748987856,
"grad_norm": 0.45894444169873694,
"learning_rate": 1.5156684514846307e-05,
"loss": 0.2054,
"step": 766
},
{
"epoch": 3.1052631578947367,
"grad_norm": 0.43228754683238946,
"learning_rate": 1.5101832974052012e-05,
"loss": 0.1766,
"step": 767
},
{
"epoch": 3.1093117408906883,
"grad_norm": 0.3665207872338171,
"learning_rate": 1.5047020598870737e-05,
"loss": 0.2036,
"step": 768
},
{
"epoch": 3.11336032388664,
"grad_norm": 0.42096549937936045,
"learning_rate": 1.4992247827580778e-05,
"loss": 0.1963,
"step": 769
},
{
"epoch": 3.117408906882591,
"grad_norm": 0.5112916197076057,
"learning_rate": 1.493751509814374e-05,
"loss": 0.1862,
"step": 770
},
{
"epoch": 3.1214574898785425,
"grad_norm": 0.4602149479798082,
"learning_rate": 1.4882822848201073e-05,
"loss": 0.1831,
"step": 771
},
{
"epoch": 3.125506072874494,
"grad_norm": 0.41444760963038824,
"learning_rate": 1.4828171515070553e-05,
"loss": 0.2141,
"step": 772
},
{
"epoch": 3.1295546558704452,
"grad_norm": 0.4494744048596974,
"learning_rate": 1.4773561535742793e-05,
"loss": 0.1773,
"step": 773
},
{
"epoch": 3.133603238866397,
"grad_norm": 0.43417749536290073,
"learning_rate": 1.4718993346877715e-05,
"loss": 0.3526,
"step": 774
},
{
"epoch": 3.1376518218623484,
"grad_norm": 0.5773995305401083,
"learning_rate": 1.466446738480111e-05,
"loss": 0.1727,
"step": 775
},
{
"epoch": 3.1417004048582995,
"grad_norm": 0.46906724961692653,
"learning_rate": 1.460998408550112e-05,
"loss": 0.1806,
"step": 776
},
{
"epoch": 3.145748987854251,
"grad_norm": 0.46084973207540586,
"learning_rate": 1.4555543884624751e-05,
"loss": 0.1826,
"step": 777
},
{
"epoch": 3.1497975708502026,
"grad_norm": 0.4726031436035235,
"learning_rate": 1.4501147217474402e-05,
"loss": 0.172,
"step": 778
},
{
"epoch": 3.1538461538461537,
"grad_norm": 0.458435497291283,
"learning_rate": 1.444679451900437e-05,
"loss": 0.1894,
"step": 779
},
{
"epoch": 3.1578947368421053,
"grad_norm": 0.4480179770661917,
"learning_rate": 1.4392486223817397e-05,
"loss": 0.1397,
"step": 780
},
{
"epoch": 3.161943319838057,
"grad_norm": 0.4219132573294055,
"learning_rate": 1.4338222766161143e-05,
"loss": 0.1772,
"step": 781
},
{
"epoch": 3.165991902834008,
"grad_norm": 0.42131370229368936,
"learning_rate": 1.4284004579924777e-05,
"loss": 0.1758,
"step": 782
},
{
"epoch": 3.1700404858299596,
"grad_norm": 0.44930921436677407,
"learning_rate": 1.4229832098635479e-05,
"loss": 0.2329,
"step": 783
},
{
"epoch": 3.174089068825911,
"grad_norm": 0.47597228431631666,
"learning_rate": 1.4175705755454963e-05,
"loss": 0.1688,
"step": 784
},
{
"epoch": 3.1781376518218623,
"grad_norm": 0.4433032485845564,
"learning_rate": 1.412162598317602e-05,
"loss": 0.2246,
"step": 785
},
{
"epoch": 3.182186234817814,
"grad_norm": 0.47689240754164325,
"learning_rate": 1.406759321421907e-05,
"loss": 0.1695,
"step": 786
},
{
"epoch": 3.1862348178137654,
"grad_norm": 0.471323603419058,
"learning_rate": 1.401360788062871e-05,
"loss": 0.1669,
"step": 787
},
{
"epoch": 3.1902834008097165,
"grad_norm": 0.4718985272625318,
"learning_rate": 1.3959670414070204e-05,
"loss": 0.2196,
"step": 788
},
{
"epoch": 3.194331983805668,
"grad_norm": 0.4333918532827137,
"learning_rate": 1.3905781245826108e-05,
"loss": 0.2001,
"step": 789
},
{
"epoch": 3.1983805668016196,
"grad_norm": 0.5452817009093589,
"learning_rate": 1.3851940806792778e-05,
"loss": 0.2908,
"step": 790
},
{
"epoch": 3.2024291497975708,
"grad_norm": 0.4353047479484203,
"learning_rate": 1.379814952747693e-05,
"loss": 0.1613,
"step": 791
},
{
"epoch": 3.2064777327935223,
"grad_norm": 0.45964710631413525,
"learning_rate": 1.3744407837992193e-05,
"loss": 0.177,
"step": 792
},
{
"epoch": 3.2105263157894735,
"grad_norm": 0.44034726799085455,
"learning_rate": 1.3690716168055692e-05,
"loss": 0.1895,
"step": 793
},
{
"epoch": 3.214574898785425,
"grad_norm": 0.44274824300973104,
"learning_rate": 1.363707494698459e-05,
"loss": 0.1911,
"step": 794
},
{
"epoch": 3.2186234817813766,
"grad_norm": 0.4623154824719728,
"learning_rate": 1.358348460369265e-05,
"loss": 0.2017,
"step": 795
},
{
"epoch": 3.2226720647773277,
"grad_norm": 0.48628006963248044,
"learning_rate": 1.352994556668684e-05,
"loss": 0.153,
"step": 796
},
{
"epoch": 3.2267206477732793,
"grad_norm": 0.4474688828556001,
"learning_rate": 1.3476458264063865e-05,
"loss": 0.1734,
"step": 797
},
{
"epoch": 3.230769230769231,
"grad_norm": 0.48660490019616903,
"learning_rate": 1.342302312350678e-05,
"loss": 0.2139,
"step": 798
},
{
"epoch": 3.234817813765182,
"grad_norm": 0.49043789448674674,
"learning_rate": 1.3369640572281537e-05,
"loss": 0.2345,
"step": 799
},
{
"epoch": 3.2388663967611335,
"grad_norm": 0.42925599322529895,
"learning_rate": 1.3316311037233596e-05,
"loss": 0.1637,
"step": 800
},
{
"epoch": 3.242914979757085,
"grad_norm": 0.48980722584838227,
"learning_rate": 1.326303494478451e-05,
"loss": 0.2487,
"step": 801
},
{
"epoch": 3.246963562753036,
"grad_norm": 0.4526954136107345,
"learning_rate": 1.3209812720928472e-05,
"loss": 0.1726,
"step": 802
},
{
"epoch": 3.251012145748988,
"grad_norm": 0.4754451104361133,
"learning_rate": 1.315664479122898e-05,
"loss": 0.1981,
"step": 803
},
{
"epoch": 3.2550607287449393,
"grad_norm": 0.4397933016434085,
"learning_rate": 1.3103531580815378e-05,
"loss": 0.2057,
"step": 804
},
{
"epoch": 3.2591093117408905,
"grad_norm": 0.47807927138159695,
"learning_rate": 1.3050473514379488e-05,
"loss": 0.182,
"step": 805
},
{
"epoch": 3.263157894736842,
"grad_norm": 0.5317945571256694,
"learning_rate": 1.2997471016172188e-05,
"loss": 0.209,
"step": 806
},
{
"epoch": 3.2672064777327936,
"grad_norm": 0.4997074302503826,
"learning_rate": 1.2944524510000042e-05,
"loss": 0.2145,
"step": 807
},
{
"epoch": 3.2712550607287447,
"grad_norm": 0.4258161167561619,
"learning_rate": 1.2891634419221908e-05,
"loss": 0.292,
"step": 808
},
{
"epoch": 3.2753036437246963,
"grad_norm": 0.4834929783133769,
"learning_rate": 1.2838801166745545e-05,
"loss": 0.1831,
"step": 809
},
{
"epoch": 3.279352226720648,
"grad_norm": 0.4706233051190158,
"learning_rate": 1.2786025175024216e-05,
"loss": 0.1645,
"step": 810
},
{
"epoch": 3.283400809716599,
"grad_norm": 0.4287529158946973,
"learning_rate": 1.2733306866053357e-05,
"loss": 0.2255,
"step": 811
},
{
"epoch": 3.2874493927125505,
"grad_norm": 0.5700756433894127,
"learning_rate": 1.2680646661367163e-05,
"loss": 0.1779,
"step": 812
},
{
"epoch": 3.291497975708502,
"grad_norm": 0.45751614413144837,
"learning_rate": 1.2628044982035219e-05,
"loss": 0.1831,
"step": 813
},
{
"epoch": 3.2955465587044532,
"grad_norm": 0.5247336829400544,
"learning_rate": 1.2575502248659155e-05,
"loss": 0.3669,
"step": 814
},
{
"epoch": 3.299595141700405,
"grad_norm": 0.4088210763636888,
"learning_rate": 1.2523018881369269e-05,
"loss": 0.1979,
"step": 815
},
{
"epoch": 3.3036437246963564,
"grad_norm": 0.555807752117483,
"learning_rate": 1.2470595299821176e-05,
"loss": 0.1801,
"step": 816
},
{
"epoch": 3.3076923076923075,
"grad_norm": 0.4374606941389995,
"learning_rate": 1.2418231923192416e-05,
"loss": 0.1831,
"step": 817
},
{
"epoch": 3.311740890688259,
"grad_norm": 0.42052114404454544,
"learning_rate": 1.2365929170179165e-05,
"loss": 0.1995,
"step": 818
},
{
"epoch": 3.3157894736842106,
"grad_norm": 0.43643362446851475,
"learning_rate": 1.2313687458992848e-05,
"loss": 0.19,
"step": 819
},
{
"epoch": 3.3198380566801617,
"grad_norm": 0.42468101361880967,
"learning_rate": 1.2261507207356785e-05,
"loss": 0.1522,
"step": 820
},
{
"epoch": 3.3238866396761133,
"grad_norm": 0.47230453515836546,
"learning_rate": 1.220938883250288e-05,
"loss": 0.1423,
"step": 821
},
{
"epoch": 3.327935222672065,
"grad_norm": 0.43699445493537853,
"learning_rate": 1.215733275116827e-05,
"loss": 0.1687,
"step": 822
},
{
"epoch": 3.331983805668016,
"grad_norm": 0.5096952015286916,
"learning_rate": 1.2105339379592005e-05,
"loss": 0.231,
"step": 823
},
{
"epoch": 3.3360323886639676,
"grad_norm": 0.5045210995229346,
"learning_rate": 1.2053409133511681e-05,
"loss": 0.194,
"step": 824
},
{
"epoch": 3.340080971659919,
"grad_norm": 0.5120620249188621,
"learning_rate": 1.2001542428160173e-05,
"loss": 0.2901,
"step": 825
},
{
"epoch": 3.3441295546558703,
"grad_norm": 0.44742081548370793,
"learning_rate": 1.1949739678262278e-05,
"loss": 0.1782,
"step": 826
},
{
"epoch": 3.348178137651822,
"grad_norm": 0.41805933676270446,
"learning_rate": 1.189800129803139e-05,
"loss": 0.1523,
"step": 827
},
{
"epoch": 3.3522267206477734,
"grad_norm": 0.4524261326711509,
"learning_rate": 1.1846327701166228e-05,
"loss": 0.1631,
"step": 828
},
{
"epoch": 3.3562753036437245,
"grad_norm": 0.4531240227223847,
"learning_rate": 1.1794719300847493e-05,
"loss": 0.223,
"step": 829
},
{
"epoch": 3.360323886639676,
"grad_norm": 0.4483226137283861,
"learning_rate": 1.1743176509734582e-05,
"loss": 0.1994,
"step": 830
},
{
"epoch": 3.3643724696356276,
"grad_norm": 0.5301403798098551,
"learning_rate": 1.1691699739962275e-05,
"loss": 0.3397,
"step": 831
},
{
"epoch": 3.3684210526315788,
"grad_norm": 0.5293723155401686,
"learning_rate": 1.1640289403137456e-05,
"loss": 0.1774,
"step": 832
},
{
"epoch": 3.3724696356275303,
"grad_norm": 0.4210388756667714,
"learning_rate": 1.1588945910335802e-05,
"loss": 0.1846,
"step": 833
},
{
"epoch": 3.376518218623482,
"grad_norm": 0.45403918897914397,
"learning_rate": 1.1537669672098517e-05,
"loss": 0.2306,
"step": 834
},
{
"epoch": 3.380566801619433,
"grad_norm": 0.4494354278945856,
"learning_rate": 1.1486461098429036e-05,
"loss": 0.157,
"step": 835
},
{
"epoch": 3.3846153846153846,
"grad_norm": 0.4224314050145944,
"learning_rate": 1.1435320598789747e-05,
"loss": 0.1676,
"step": 836
},
{
"epoch": 3.388663967611336,
"grad_norm": 0.540722302044397,
"learning_rate": 1.1384248582098736e-05,
"loss": 0.189,
"step": 837
},
{
"epoch": 3.3927125506072873,
"grad_norm": 0.5338253597856661,
"learning_rate": 1.1333245456726467e-05,
"loss": 0.1664,
"step": 838
},
{
"epoch": 3.396761133603239,
"grad_norm": 0.43673031013941477,
"learning_rate": 1.128231163049258e-05,
"loss": 0.1545,
"step": 839
},
{
"epoch": 3.4008097165991904,
"grad_norm": 0.47296534918955097,
"learning_rate": 1.123144751066259e-05,
"loss": 0.1991,
"step": 840
},
{
"epoch": 3.4048582995951415,
"grad_norm": 0.4787918532963703,
"learning_rate": 1.1180653503944657e-05,
"loss": 0.1603,
"step": 841
},
{
"epoch": 3.408906882591093,
"grad_norm": 0.43399006602940854,
"learning_rate": 1.1129930016486287e-05,
"loss": 0.1943,
"step": 842
},
{
"epoch": 3.4129554655870447,
"grad_norm": 0.4191771129041629,
"learning_rate": 1.107927745387114e-05,
"loss": 0.1979,
"step": 843
},
{
"epoch": 3.417004048582996,
"grad_norm": 0.45754852099866844,
"learning_rate": 1.1028696221115766e-05,
"loss": 0.2027,
"step": 844
},
{
"epoch": 3.4210526315789473,
"grad_norm": 0.4426977781940336,
"learning_rate": 1.0978186722666341e-05,
"loss": 0.1948,
"step": 845
},
{
"epoch": 3.425101214574899,
"grad_norm": 0.4569800707233501,
"learning_rate": 1.0927749362395475e-05,
"loss": 0.1813,
"step": 846
},
{
"epoch": 3.42914979757085,
"grad_norm": 0.4680562862918238,
"learning_rate": 1.0877384543598954e-05,
"loss": 0.2122,
"step": 847
},
{
"epoch": 3.4331983805668016,
"grad_norm": 0.4725674659351346,
"learning_rate": 1.0827092668992529e-05,
"loss": 0.2096,
"step": 848
},
{
"epoch": 3.437246963562753,
"grad_norm": 0.4499963329612661,
"learning_rate": 1.077687414070869e-05,
"loss": 0.1829,
"step": 849
},
{
"epoch": 3.4412955465587043,
"grad_norm": 0.3844131965298602,
"learning_rate": 1.0726729360293451e-05,
"loss": 0.1814,
"step": 850
},
{
"epoch": 3.445344129554656,
"grad_norm": 0.43684107750447,
"learning_rate": 1.0676658728703147e-05,
"loss": 0.1957,
"step": 851
},
{
"epoch": 3.4493927125506074,
"grad_norm": 0.538004078079512,
"learning_rate": 1.0626662646301198e-05,
"loss": 0.1662,
"step": 852
},
{
"epoch": 3.4534412955465585,
"grad_norm": 0.4607373082231513,
"learning_rate": 1.0576741512854958e-05,
"loss": 0.1969,
"step": 853
},
{
"epoch": 3.45748987854251,
"grad_norm": 0.5233732445279864,
"learning_rate": 1.052689572753247e-05,
"loss": 0.213,
"step": 854
},
{
"epoch": 3.4615384615384617,
"grad_norm": 0.4087573211998746,
"learning_rate": 1.0477125688899328e-05,
"loss": 0.1356,
"step": 855
},
{
"epoch": 3.465587044534413,
"grad_norm": 0.4460283769783395,
"learning_rate": 1.0427431794915406e-05,
"loss": 0.1892,
"step": 856
},
{
"epoch": 3.4696356275303644,
"grad_norm": 0.4421869849019569,
"learning_rate": 1.0377814442931769e-05,
"loss": 0.1582,
"step": 857
},
{
"epoch": 3.473684210526316,
"grad_norm": 0.4653562250611419,
"learning_rate": 1.0328274029687439e-05,
"loss": 0.1983,
"step": 858
},
{
"epoch": 3.477732793522267,
"grad_norm": 0.45067799594021984,
"learning_rate": 1.0278810951306253e-05,
"loss": 0.1949,
"step": 859
},
{
"epoch": 3.4817813765182186,
"grad_norm": 0.4600667198531214,
"learning_rate": 1.0229425603293647e-05,
"loss": 0.2108,
"step": 860
},
{
"epoch": 3.48582995951417,
"grad_norm": 0.5080765798680222,
"learning_rate": 1.0180118380533558e-05,
"loss": 0.2742,
"step": 861
},
{
"epoch": 3.4898785425101213,
"grad_norm": 0.4324835195240619,
"learning_rate": 1.0130889677285225e-05,
"loss": 0.2296,
"step": 862
},
{
"epoch": 3.493927125506073,
"grad_norm": 0.4551182064020533,
"learning_rate": 1.008173988718005e-05,
"loss": 0.1514,
"step": 863
},
{
"epoch": 3.4979757085020244,
"grad_norm": 0.4656615396420351,
"learning_rate": 1.0032669403218439e-05,
"loss": 0.1874,
"step": 864
},
{
"epoch": 3.5020242914979756,
"grad_norm": 0.5027873524986649,
"learning_rate": 9.983678617766677e-06,
"loss": 0.183,
"step": 865
},
{
"epoch": 3.506072874493927,
"grad_norm": 0.4299534334995737,
"learning_rate": 9.934767922553788e-06,
"loss": 0.2097,
"step": 866
},
{
"epoch": 3.5101214574898787,
"grad_norm": 0.5054697083953896,
"learning_rate": 9.885937708668371e-06,
"loss": 0.1813,
"step": 867
},
{
"epoch": 3.51417004048583,
"grad_norm": 0.4368377520468624,
"learning_rate": 9.837188366555524e-06,
"loss": 0.1841,
"step": 868
},
{
"epoch": 3.5182186234817814,
"grad_norm": 0.46083120727533017,
"learning_rate": 9.788520286013691e-06,
"loss": 0.1391,
"step": 869
},
{
"epoch": 3.522267206477733,
"grad_norm": 0.42674015479065747,
"learning_rate": 9.739933856191537e-06,
"loss": 0.1929,
"step": 870
},
{
"epoch": 3.526315789473684,
"grad_norm": 0.46368367338159316,
"learning_rate": 9.691429465584866e-06,
"loss": 0.287,
"step": 871
},
{
"epoch": 3.5303643724696356,
"grad_norm": 0.42167986094140475,
"learning_rate": 9.643007502033492e-06,
"loss": 0.1978,
"step": 872
},
{
"epoch": 3.534412955465587,
"grad_norm": 0.5493308998787867,
"learning_rate": 9.594668352718152e-06,
"loss": 0.1752,
"step": 873
},
{
"epoch": 3.5384615384615383,
"grad_norm": 0.507820068740953,
"learning_rate": 9.546412404157383e-06,
"loss": 0.1855,
"step": 874
},
{
"epoch": 3.54251012145749,
"grad_norm": 0.49451355896716576,
"learning_rate": 9.498240042204471e-06,
"loss": 0.313,
"step": 875
},
{
"epoch": 3.5465587044534415,
"grad_norm": 0.5318745748074805,
"learning_rate": 9.450151652044339e-06,
"loss": 0.2138,
"step": 876
},
{
"epoch": 3.5506072874493926,
"grad_norm": 0.4829614392651031,
"learning_rate": 9.402147618190472e-06,
"loss": 0.1894,
"step": 877
},
{
"epoch": 3.554655870445344,
"grad_norm": 0.4643428401775076,
"learning_rate": 9.354228324481841e-06,
"loss": 0.1586,
"step": 878
},
{
"epoch": 3.5587044534412957,
"grad_norm": 0.3753419452012955,
"learning_rate": 9.306394154079852e-06,
"loss": 0.3186,
"step": 879
},
{
"epoch": 3.562753036437247,
"grad_norm": 0.43283840341138946,
"learning_rate": 9.258645489465254e-06,
"loss": 0.2378,
"step": 880
},
{
"epoch": 3.5668016194331984,
"grad_norm": 0.44457269381605863,
"learning_rate": 9.21098271243509e-06,
"loss": 0.2589,
"step": 881
},
{
"epoch": 3.57085020242915,
"grad_norm": 0.5114800490653759,
"learning_rate": 9.163406204099659e-06,
"loss": 0.1732,
"step": 882
},
{
"epoch": 3.574898785425101,
"grad_norm": 0.4129661476676854,
"learning_rate": 9.115916344879452e-06,
"loss": 0.1859,
"step": 883
},
{
"epoch": 3.5789473684210527,
"grad_norm": 0.42757300680368726,
"learning_rate": 9.068513514502126e-06,
"loss": 0.1776,
"step": 884
},
{
"epoch": 3.582995951417004,
"grad_norm": 0.43998846237352646,
"learning_rate": 9.021198091999439e-06,
"loss": 0.2237,
"step": 885
},
{
"epoch": 3.5870445344129553,
"grad_norm": 0.45745596374550185,
"learning_rate": 8.973970455704251e-06,
"loss": 0.1756,
"step": 886
},
{
"epoch": 3.591093117408907,
"grad_norm": 0.41407069051114015,
"learning_rate": 8.926830983247495e-06,
"loss": 0.2057,
"step": 887
},
{
"epoch": 3.5951417004048585,
"grad_norm": 0.42468069732378033,
"learning_rate": 8.879780051555116e-06,
"loss": 0.1522,
"step": 888
},
{
"epoch": 3.5991902834008096,
"grad_norm": 0.4569722485790906,
"learning_rate": 8.83281803684512e-06,
"loss": 0.2112,
"step": 889
},
{
"epoch": 3.603238866396761,
"grad_norm": 0.5621765151509991,
"learning_rate": 8.785945314624532e-06,
"loss": 0.1522,
"step": 890
},
{
"epoch": 3.6072874493927127,
"grad_norm": 0.4304810512493557,
"learning_rate": 8.739162259686385e-06,
"loss": 0.193,
"step": 891
},
{
"epoch": 3.611336032388664,
"grad_norm": 0.41188413413071284,
"learning_rate": 8.692469246106734e-06,
"loss": 0.1756,
"step": 892
},
{
"epoch": 3.6153846153846154,
"grad_norm": 0.45368017881685485,
"learning_rate": 8.645866647241677e-06,
"loss": 0.1905,
"step": 893
},
{
"epoch": 3.619433198380567,
"grad_norm": 0.439369840854364,
"learning_rate": 8.599354835724348e-06,
"loss": 0.2251,
"step": 894
},
{
"epoch": 3.623481781376518,
"grad_norm": 0.442693802908622,
"learning_rate": 8.552934183461958e-06,
"loss": 0.194,
"step": 895
},
{
"epoch": 3.6275303643724697,
"grad_norm": 0.4336096856970324,
"learning_rate": 8.506605061632805e-06,
"loss": 0.1926,
"step": 896
},
{
"epoch": 3.6315789473684212,
"grad_norm": 0.4697959668799715,
"learning_rate": 8.460367840683308e-06,
"loss": 0.2269,
"step": 897
},
{
"epoch": 3.6356275303643724,
"grad_norm": 0.4707671377900855,
"learning_rate": 8.414222890325065e-06,
"loss": 0.1869,
"step": 898
},
{
"epoch": 3.639676113360324,
"grad_norm": 0.46513264244995584,
"learning_rate": 8.368170579531855e-06,
"loss": 0.254,
"step": 899
},
{
"epoch": 3.6437246963562755,
"grad_norm": 0.43959116710192486,
"learning_rate": 8.322211276536733e-06,
"loss": 0.1805,
"step": 900
},
{
"epoch": 3.6477732793522266,
"grad_norm": 0.46673735658513693,
"learning_rate": 8.27634534882907e-06,
"loss": 0.3118,
"step": 901
},
{
"epoch": 3.651821862348178,
"grad_norm": 0.4500609134724881,
"learning_rate": 8.230573163151585e-06,
"loss": 0.1401,
"step": 902
},
{
"epoch": 3.6558704453441297,
"grad_norm": 0.5512102577335931,
"learning_rate": 8.184895085497466e-06,
"loss": 0.1834,
"step": 903
},
{
"epoch": 3.659919028340081,
"grad_norm": 0.5069253827645228,
"learning_rate": 8.139311481107401e-06,
"loss": 0.1881,
"step": 904
},
{
"epoch": 3.6639676113360324,
"grad_norm": 0.44205587615065617,
"learning_rate": 8.093822714466686e-06,
"loss": 0.3022,
"step": 905
},
{
"epoch": 3.668016194331984,
"grad_norm": 0.4290237394868093,
"learning_rate": 8.048429149302271e-06,
"loss": 0.1853,
"step": 906
},
{
"epoch": 3.672064777327935,
"grad_norm": 0.45807278167345306,
"learning_rate": 8.003131148579902e-06,
"loss": 0.2064,
"step": 907
},
{
"epoch": 3.6761133603238867,
"grad_norm": 0.411454588087553,
"learning_rate": 7.957929074501182e-06,
"loss": 0.1761,
"step": 908
},
{
"epoch": 3.6801619433198383,
"grad_norm": 0.46973962612248366,
"learning_rate": 7.912823288500695e-06,
"loss": 0.1579,
"step": 909
},
{
"epoch": 3.6842105263157894,
"grad_norm": 0.41163269960553295,
"learning_rate": 7.867814151243102e-06,
"loss": 0.1563,
"step": 910
},
{
"epoch": 3.688259109311741,
"grad_norm": 0.45021338205881756,
"learning_rate": 7.822902022620267e-06,
"loss": 0.1969,
"step": 911
},
{
"epoch": 3.6923076923076925,
"grad_norm": 0.46588045722097865,
"learning_rate": 7.778087261748377e-06,
"loss": 0.189,
"step": 912
},
{
"epoch": 3.6963562753036436,
"grad_norm": 0.43306528727331095,
"learning_rate": 7.733370226965056e-06,
"loss": 0.1626,
"step": 913
},
{
"epoch": 3.700404858299595,
"grad_norm": 0.4284758093648557,
"learning_rate": 7.688751275826527e-06,
"loss": 0.175,
"step": 914
},
{
"epoch": 3.7044534412955468,
"grad_norm": 0.4881066984973405,
"learning_rate": 7.644230765104735e-06,
"loss": 0.1394,
"step": 915
},
{
"epoch": 3.708502024291498,
"grad_norm": 0.41133014722448213,
"learning_rate": 7.599809050784502e-06,
"loss": 0.1687,
"step": 916
},
{
"epoch": 3.7125506072874495,
"grad_norm": 0.4244476118258955,
"learning_rate": 7.555486488060661e-06,
"loss": 0.1716,
"step": 917
},
{
"epoch": 3.716599190283401,
"grad_norm": 0.4285180457130212,
"learning_rate": 7.511263431335249e-06,
"loss": 0.1996,
"step": 918
},
{
"epoch": 3.720647773279352,
"grad_norm": 0.49214804520204675,
"learning_rate": 7.467140234214656e-06,
"loss": 0.1711,
"step": 919
},
{
"epoch": 3.7246963562753037,
"grad_norm": 0.43087661773660313,
"learning_rate": 7.423117249506777e-06,
"loss": 0.1427,
"step": 920
},
{
"epoch": 3.7287449392712553,
"grad_norm": 0.4687353245073113,
"learning_rate": 7.379194829218231e-06,
"loss": 0.2069,
"step": 921
},
{
"epoch": 3.7327935222672064,
"grad_norm": 0.41065007019061095,
"learning_rate": 7.335373324551518e-06,
"loss": 0.191,
"step": 922
},
{
"epoch": 3.736842105263158,
"grad_norm": 0.4349763461621196,
"learning_rate": 7.291653085902224e-06,
"loss": 0.1768,
"step": 923
},
{
"epoch": 3.7408906882591095,
"grad_norm": 0.949693700266149,
"learning_rate": 7.248034462856213e-06,
"loss": 0.1781,
"step": 924
},
{
"epoch": 3.7449392712550607,
"grad_norm": 0.4540251207644304,
"learning_rate": 7.20451780418683e-06,
"loss": 0.1857,
"step": 925
},
{
"epoch": 3.748987854251012,
"grad_norm": 0.4704446564610619,
"learning_rate": 7.161103457852128e-06,
"loss": 0.1488,
"step": 926
},
{
"epoch": 3.753036437246964,
"grad_norm": 0.4909620557204366,
"learning_rate": 7.117791770992046e-06,
"loss": 0.1902,
"step": 927
},
{
"epoch": 3.757085020242915,
"grad_norm": 0.4196836130175124,
"learning_rate": 7.074583089925688e-06,
"loss": 0.2032,
"step": 928
},
{
"epoch": 3.7611336032388665,
"grad_norm": 0.4924979303667847,
"learning_rate": 7.0314777601485155e-06,
"loss": 0.1912,
"step": 929
},
{
"epoch": 3.765182186234818,
"grad_norm": 0.4342550210419619,
"learning_rate": 6.988476126329604e-06,
"loss": 0.197,
"step": 930
},
{
"epoch": 3.769230769230769,
"grad_norm": 0.4525585607392695,
"learning_rate": 6.945578532308857e-06,
"loss": 0.1918,
"step": 931
},
{
"epoch": 3.7732793522267207,
"grad_norm": 0.39276987595545454,
"learning_rate": 6.902785321094301e-06,
"loss": 0.1708,
"step": 932
},
{
"epoch": 3.7773279352226723,
"grad_norm": 0.5005638325928964,
"learning_rate": 6.8600968348593115e-06,
"loss": 0.231,
"step": 933
},
{
"epoch": 3.7813765182186234,
"grad_norm": 0.4366100479349994,
"learning_rate": 6.817513414939892e-06,
"loss": 0.2149,
"step": 934
},
{
"epoch": 3.785425101214575,
"grad_norm": 0.40838933292365587,
"learning_rate": 6.775035401831924e-06,
"loss": 0.1674,
"step": 935
},
{
"epoch": 3.7894736842105265,
"grad_norm": 0.4589974621815031,
"learning_rate": 6.732663135188473e-06,
"loss": 0.1724,
"step": 936
},
{
"epoch": 3.7935222672064777,
"grad_norm": 0.4108661586689894,
"learning_rate": 6.690396953817055e-06,
"loss": 0.2271,
"step": 937
},
{
"epoch": 3.7975708502024292,
"grad_norm": 0.4062021988220444,
"learning_rate": 6.648237195676927e-06,
"loss": 0.153,
"step": 938
},
{
"epoch": 3.801619433198381,
"grad_norm": 0.44250026553751465,
"learning_rate": 6.606184197876395e-06,
"loss": 0.1773,
"step": 939
},
{
"epoch": 3.805668016194332,
"grad_norm": 0.46575220861565125,
"learning_rate": 6.564238296670109e-06,
"loss": 0.1902,
"step": 940
},
{
"epoch": 3.8097165991902835,
"grad_norm": 0.44896930138531377,
"learning_rate": 6.52239982745638e-06,
"loss": 0.1911,
"step": 941
},
{
"epoch": 3.813765182186235,
"grad_norm": 0.44389786771612855,
"learning_rate": 6.480669124774484e-06,
"loss": 0.2034,
"step": 942
},
{
"epoch": 3.817813765182186,
"grad_norm": 0.5040165578339232,
"learning_rate": 6.439046522302013e-06,
"loss": 0.2023,
"step": 943
},
{
"epoch": 3.8218623481781377,
"grad_norm": 0.517428148645111,
"learning_rate": 6.397532352852191e-06,
"loss": 0.1776,
"step": 944
},
{
"epoch": 3.8259109311740893,
"grad_norm": 0.45442494818466855,
"learning_rate": 6.356126948371202e-06,
"loss": 0.2097,
"step": 945
},
{
"epoch": 3.8299595141700404,
"grad_norm": 0.42061265992896074,
"learning_rate": 6.314830639935559e-06,
"loss": 0.1936,
"step": 946
},
{
"epoch": 3.834008097165992,
"grad_norm": 0.4506002268042962,
"learning_rate": 6.273643757749448e-06,
"loss": 0.1729,
"step": 947
},
{
"epoch": 3.8380566801619436,
"grad_norm": 0.4551491416226142,
"learning_rate": 6.2325666311420856e-06,
"loss": 0.1735,
"step": 948
},
{
"epoch": 3.8421052631578947,
"grad_norm": 0.4551909175661564,
"learning_rate": 6.1915995885650675e-06,
"loss": 0.1837,
"step": 949
},
{
"epoch": 3.8461538461538463,
"grad_norm": 0.40657271238743464,
"learning_rate": 6.150742957589784e-06,
"loss": 0.1836,
"step": 950
},
{
"epoch": 3.850202429149798,
"grad_norm": 0.4399226295867325,
"learning_rate": 6.109997064904774e-06,
"loss": 0.2248,
"step": 951
},
{
"epoch": 3.854251012145749,
"grad_norm": 0.4001252996918671,
"learning_rate": 6.069362236313092e-06,
"loss": 0.1668,
"step": 952
},
{
"epoch": 3.8582995951417005,
"grad_norm": 0.45384938048840373,
"learning_rate": 6.0288387967297565e-06,
"loss": 0.1444,
"step": 953
},
{
"epoch": 3.862348178137652,
"grad_norm": 0.43754037214952113,
"learning_rate": 5.988427070179108e-06,
"loss": 0.1641,
"step": 954
},
{
"epoch": 3.866396761133603,
"grad_norm": 0.43433771143238564,
"learning_rate": 5.948127379792234e-06,
"loss": 0.1699,
"step": 955
},
{
"epoch": 3.8704453441295548,
"grad_norm": 0.46184617444214937,
"learning_rate": 5.90794004780439e-06,
"loss": 0.1521,
"step": 956
},
{
"epoch": 3.8744939271255063,
"grad_norm": 0.5084171179830917,
"learning_rate": 5.867865395552405e-06,
"loss": 0.1695,
"step": 957
},
{
"epoch": 3.8785425101214575,
"grad_norm": 0.4385169614778726,
"learning_rate": 5.8279037434721364e-06,
"loss": 0.1797,
"step": 958
},
{
"epoch": 3.882591093117409,
"grad_norm": 0.4334545723270758,
"learning_rate": 5.788055411095892e-06,
"loss": 0.1598,
"step": 959
},
{
"epoch": 3.8866396761133606,
"grad_norm": 0.500288052425768,
"learning_rate": 5.7483207170498625e-06,
"loss": 0.2568,
"step": 960
},
{
"epoch": 3.8906882591093117,
"grad_norm": 0.44681280767853715,
"learning_rate": 5.708699979051613e-06,
"loss": 0.1468,
"step": 961
},
{
"epoch": 3.8947368421052633,
"grad_norm": 0.4364418188346784,
"learning_rate": 5.66919351390751e-06,
"loss": 0.1802,
"step": 962
},
{
"epoch": 3.898785425101215,
"grad_norm": 0.4900906258866087,
"learning_rate": 5.629801637510188e-06,
"loss": 0.1834,
"step": 963
},
{
"epoch": 3.902834008097166,
"grad_norm": 0.47754824182548394,
"learning_rate": 5.590524664836043e-06,
"loss": 0.1525,
"step": 964
},
{
"epoch": 3.9068825910931175,
"grad_norm": 0.42806087235563484,
"learning_rate": 5.551362909942708e-06,
"loss": 0.1889,
"step": 965
},
{
"epoch": 3.910931174089069,
"grad_norm": 0.44641577954462247,
"learning_rate": 5.51231668596653e-06,
"loss": 0.176,
"step": 966
},
{
"epoch": 3.91497975708502,
"grad_norm": 0.4371568278794365,
"learning_rate": 5.47338630512007e-06,
"loss": 0.1726,
"step": 967
},
{
"epoch": 3.919028340080972,
"grad_norm": 0.4502148686674381,
"learning_rate": 5.434572078689617e-06,
"loss": 0.1838,
"step": 968
},
{
"epoch": 3.9230769230769234,
"grad_norm": 0.42029708247265973,
"learning_rate": 5.395874317032694e-06,
"loss": 0.1655,
"step": 969
},
{
"epoch": 3.9271255060728745,
"grad_norm": 0.48915578333341586,
"learning_rate": 5.357293329575573e-06,
"loss": 0.1236,
"step": 970
},
{
"epoch": 3.931174089068826,
"grad_norm": 0.4015289090378644,
"learning_rate": 5.318829424810797e-06,
"loss": 0.2222,
"step": 971
},
{
"epoch": 3.9352226720647776,
"grad_norm": 0.3852514754962283,
"learning_rate": 5.280482910294728e-06,
"loss": 0.1645,
"step": 972
},
{
"epoch": 3.9392712550607287,
"grad_norm": 0.41553665958024494,
"learning_rate": 5.242254092645078e-06,
"loss": 0.1728,
"step": 973
},
{
"epoch": 3.9433198380566803,
"grad_norm": 0.4379763646781338,
"learning_rate": 5.2041432775384445e-06,
"loss": 0.1524,
"step": 974
},
{
"epoch": 3.9473684210526314,
"grad_norm": 0.4270925246531206,
"learning_rate": 5.166150769707892e-06,
"loss": 0.1408,
"step": 975
},
{
"epoch": 3.951417004048583,
"grad_norm": 0.39740276231569366,
"learning_rate": 5.128276872940503e-06,
"loss": 0.2165,
"step": 976
},
{
"epoch": 3.9554655870445345,
"grad_norm": 0.44752374993901134,
"learning_rate": 5.090521890074953e-06,
"loss": 0.168,
"step": 977
},
{
"epoch": 3.9595141700404857,
"grad_norm": 0.4305934294372982,
"learning_rate": 5.0528861229990675e-06,
"loss": 0.1963,
"step": 978
},
{
"epoch": 3.9635627530364372,
"grad_norm": 0.45813709517286033,
"learning_rate": 5.015369872647442e-06,
"loss": 0.1196,
"step": 979
},
{
"epoch": 3.967611336032389,
"grad_norm": 0.4408079178476693,
"learning_rate": 4.9779734389990195e-06,
"loss": 0.1351,
"step": 980
},
{
"epoch": 3.97165991902834,
"grad_norm": 0.522299492841269,
"learning_rate": 4.940697121074675e-06,
"loss": 0.3228,
"step": 981
},
{
"epoch": 3.9757085020242915,
"grad_norm": 0.4313184332173899,
"learning_rate": 4.903541216934864e-06,
"loss": 0.1628,
"step": 982
},
{
"epoch": 3.979757085020243,
"grad_norm": 0.45924939508066426,
"learning_rate": 4.8665060236772e-06,
"loss": 0.2155,
"step": 983
},
{
"epoch": 3.983805668016194,
"grad_norm": 0.5162821255707305,
"learning_rate": 4.829591837434102e-06,
"loss": 0.2092,
"step": 984
},
{
"epoch": 3.9878542510121457,
"grad_norm": 0.42455841474824924,
"learning_rate": 4.792798953370424e-06,
"loss": 0.1875,
"step": 985
},
{
"epoch": 3.9919028340080973,
"grad_norm": 0.5166445251082739,
"learning_rate": 4.756127665681083e-06,
"loss": 0.1904,
"step": 986
},
{
"epoch": 3.9959514170040484,
"grad_norm": 0.45589438878386124,
"learning_rate": 4.719578267588725e-06,
"loss": 0.1827,
"step": 987
},
{
"epoch": 4.0,
"grad_norm": 0.45546255664695257,
"learning_rate": 4.6831510513413525e-06,
"loss": 0.1203,
"step": 988
},
{
"epoch": 4.004048582995951,
"grad_norm": 0.48680655303039133,
"learning_rate": 4.6468463082100176e-06,
"loss": 0.1191,
"step": 989
},
{
"epoch": 4.008097165991903,
"grad_norm": 0.4236554171260114,
"learning_rate": 4.610664328486485e-06,
"loss": 0.1626,
"step": 990
},
{
"epoch": 4.012145748987854,
"grad_norm": 0.35617242216218864,
"learning_rate": 4.5746054014809e-06,
"loss": 0.1508,
"step": 991
},
{
"epoch": 4.016194331983805,
"grad_norm": 0.3814544069509256,
"learning_rate": 4.5386698155194766e-06,
"loss": 0.0966,
"step": 992
},
{
"epoch": 4.020242914979757,
"grad_norm": 0.335275869002279,
"learning_rate": 4.502857857942204e-06,
"loss": 0.1284,
"step": 993
},
{
"epoch": 4.0242914979757085,
"grad_norm": 0.39009398107662496,
"learning_rate": 4.467169815100549e-06,
"loss": 0.1211,
"step": 994
},
{
"epoch": 4.02834008097166,
"grad_norm": 0.34984499951579284,
"learning_rate": 4.431605972355137e-06,
"loss": 0.0944,
"step": 995
},
{
"epoch": 4.032388663967612,
"grad_norm": 0.4179630895233436,
"learning_rate": 4.396166614073516e-06,
"loss": 0.1355,
"step": 996
},
{
"epoch": 4.036437246963563,
"grad_norm": 0.4606445089638128,
"learning_rate": 4.360852023627846e-06,
"loss": 0.1349,
"step": 997
},
{
"epoch": 4.040485829959514,
"grad_norm": 0.470103624718103,
"learning_rate": 4.325662483392656e-06,
"loss": 0.1424,
"step": 998
},
{
"epoch": 4.044534412955466,
"grad_norm": 0.38010769599050476,
"learning_rate": 4.29059827474257e-06,
"loss": 0.1007,
"step": 999
},
{
"epoch": 4.048582995951417,
"grad_norm": 0.3381242352182827,
"learning_rate": 4.255659678050063e-06,
"loss": 0.0935,
"step": 1000
},
{
"epoch": 4.052631578947368,
"grad_norm": 0.37164415741841766,
"learning_rate": 4.22084697268323e-06,
"loss": 0.0805,
"step": 1001
},
{
"epoch": 4.05668016194332,
"grad_norm": 0.38567292361567074,
"learning_rate": 4.186160437003535e-06,
"loss": 0.1212,
"step": 1002
},
{
"epoch": 4.060728744939271,
"grad_norm": 0.4544646753001259,
"learning_rate": 4.151600348363582e-06,
"loss": 0.1345,
"step": 1003
},
{
"epoch": 4.064777327935222,
"grad_norm": 0.3736311754613385,
"learning_rate": 4.11716698310493e-06,
"loss": 0.0954,
"step": 1004
},
{
"epoch": 4.068825910931174,
"grad_norm": 0.4001057065736088,
"learning_rate": 4.082860616555848e-06,
"loss": 0.1467,
"step": 1005
},
{
"epoch": 4.0728744939271255,
"grad_norm": 0.38035799327325737,
"learning_rate": 4.048681523029125e-06,
"loss": 0.1603,
"step": 1006
},
{
"epoch": 4.076923076923077,
"grad_norm": 0.3714413486467079,
"learning_rate": 4.014629975819882e-06,
"loss": 0.1132,
"step": 1007
},
{
"epoch": 4.080971659919029,
"grad_norm": 0.40687801122524136,
"learning_rate": 3.9807062472033855e-06,
"loss": 0.1156,
"step": 1008
},
{
"epoch": 4.08502024291498,
"grad_norm": 0.38628068211070893,
"learning_rate": 3.94691060843287e-06,
"loss": 0.1351,
"step": 1009
},
{
"epoch": 4.089068825910931,
"grad_norm": 0.3891381690479926,
"learning_rate": 3.913243329737353e-06,
"loss": 0.1501,
"step": 1010
},
{
"epoch": 4.093117408906883,
"grad_norm": 0.41788406015407575,
"learning_rate": 3.879704680319503e-06,
"loss": 0.129,
"step": 1011
},
{
"epoch": 4.097165991902834,
"grad_norm": 0.421265264451386,
"learning_rate": 3.846294928353471e-06,
"loss": 0.1322,
"step": 1012
},
{
"epoch": 4.101214574898785,
"grad_norm": 0.42654714747473477,
"learning_rate": 3.8130143409827345e-06,
"loss": 0.1125,
"step": 1013
},
{
"epoch": 4.105263157894737,
"grad_norm": 0.3766590081394487,
"learning_rate": 3.7798631843179887e-06,
"loss": 0.1343,
"step": 1014
},
{
"epoch": 4.109311740890688,
"grad_norm": 0.46868224395704805,
"learning_rate": 3.7468417234349973e-06,
"loss": 0.2846,
"step": 1015
},
{
"epoch": 4.113360323886639,
"grad_norm": 0.33590681089704055,
"learning_rate": 3.7139502223724823e-06,
"loss": 0.1376,
"step": 1016
},
{
"epoch": 4.117408906882591,
"grad_norm": 0.419988389403499,
"learning_rate": 3.6811889441300095e-06,
"loss": 0.1363,
"step": 1017
},
{
"epoch": 4.1214574898785425,
"grad_norm": 0.3753503000021104,
"learning_rate": 3.648558150665886e-06,
"loss": 0.1341,
"step": 1018
},
{
"epoch": 4.125506072874494,
"grad_norm": 0.3171436934614244,
"learning_rate": 3.616058102895075e-06,
"loss": 0.1022,
"step": 1019
},
{
"epoch": 4.129554655870446,
"grad_norm": 0.3763523623856292,
"learning_rate": 3.583689060687079e-06,
"loss": 0.1438,
"step": 1020
},
{
"epoch": 4.133603238866397,
"grad_norm": 0.33275541040171924,
"learning_rate": 3.5514512828639023e-06,
"loss": 0.1356,
"step": 1021
},
{
"epoch": 4.137651821862348,
"grad_norm": 0.37067359813317563,
"learning_rate": 3.51934502719796e-06,
"loss": 0.1328,
"step": 1022
},
{
"epoch": 4.1417004048583,
"grad_norm": 0.3596379659346746,
"learning_rate": 3.4873705504100185e-06,
"loss": 0.2181,
"step": 1023
},
{
"epoch": 4.145748987854251,
"grad_norm": 0.4027943490316193,
"learning_rate": 3.4555281081671345e-06,
"loss": 0.1271,
"step": 1024
},
{
"epoch": 4.149797570850202,
"grad_norm": 0.32590932140301065,
"learning_rate": 3.4238179550806326e-06,
"loss": 0.09,
"step": 1025
},
{
"epoch": 4.153846153846154,
"grad_norm": 0.33896891873507945,
"learning_rate": 3.3922403447040497e-06,
"loss": 0.0921,
"step": 1026
},
{
"epoch": 4.157894736842105,
"grad_norm": 0.3449706545054024,
"learning_rate": 3.3607955295311247e-06,
"loss": 0.2123,
"step": 1027
},
{
"epoch": 4.161943319838056,
"grad_norm": 0.46240103747817735,
"learning_rate": 3.32948376099375e-06,
"loss": 0.1337,
"step": 1028
},
{
"epoch": 4.165991902834008,
"grad_norm": 0.416310639783333,
"learning_rate": 3.2983052894599976e-06,
"loss": 0.104,
"step": 1029
},
{
"epoch": 4.17004048582996,
"grad_norm": 0.32985058441159376,
"learning_rate": 3.2672603642320945e-06,
"loss": 0.1447,
"step": 1030
},
{
"epoch": 4.174089068825911,
"grad_norm": 0.351422598656761,
"learning_rate": 3.236349233544438e-06,
"loss": 0.1136,
"step": 1031
},
{
"epoch": 4.178137651821863,
"grad_norm": 0.33777538875042057,
"learning_rate": 3.205572144561604e-06,
"loss": 0.1875,
"step": 1032
},
{
"epoch": 4.182186234817814,
"grad_norm": 0.34167310924750033,
"learning_rate": 3.174929343376374e-06,
"loss": 0.1512,
"step": 1033
},
{
"epoch": 4.186234817813765,
"grad_norm": 0.34535449518690386,
"learning_rate": 3.1444210750077777e-06,
"loss": 0.1037,
"step": 1034
},
{
"epoch": 4.190283400809717,
"grad_norm": 0.3416670727165505,
"learning_rate": 3.1140475833991046e-06,
"loss": 0.0967,
"step": 1035
},
{
"epoch": 4.194331983805668,
"grad_norm": 0.4049503401150219,
"learning_rate": 3.0838091114159895e-06,
"loss": 0.1374,
"step": 1036
},
{
"epoch": 4.198380566801619,
"grad_norm": 0.33244825372094483,
"learning_rate": 3.0537059008444524e-06,
"loss": 0.1428,
"step": 1037
},
{
"epoch": 4.202429149797571,
"grad_norm": 0.4103012606550283,
"learning_rate": 3.0237381923889563e-06,
"loss": 0.1415,
"step": 1038
},
{
"epoch": 4.206477732793522,
"grad_norm": 0.359448921719285,
"learning_rate": 2.9939062256704997e-06,
"loss": 0.1271,
"step": 1039
},
{
"epoch": 4.2105263157894735,
"grad_norm": 0.3676107190717254,
"learning_rate": 2.964210239224696e-06,
"loss": 0.1144,
"step": 1040
},
{
"epoch": 4.2145748987854255,
"grad_norm": 0.39680495747049715,
"learning_rate": 2.934650470499869e-06,
"loss": 0.1243,
"step": 1041
},
{
"epoch": 4.218623481781377,
"grad_norm": 0.3652205399521062,
"learning_rate": 2.905227155855128e-06,
"loss": 0.2193,
"step": 1042
},
{
"epoch": 4.222672064777328,
"grad_norm": 0.36631756924524383,
"learning_rate": 2.875940530558519e-06,
"loss": 0.1318,
"step": 1043
},
{
"epoch": 4.22672064777328,
"grad_norm": 0.3671775166950895,
"learning_rate": 2.8467908287851154e-06,
"loss": 0.179,
"step": 1044
},
{
"epoch": 4.230769230769231,
"grad_norm": 0.3538517087553492,
"learning_rate": 2.8177782836151603e-06,
"loss": 0.1386,
"step": 1045
},
{
"epoch": 4.234817813765182,
"grad_norm": 0.3287261547041422,
"learning_rate": 2.7889031270321874e-06,
"loss": 0.1153,
"step": 1046
},
{
"epoch": 4.238866396761134,
"grad_norm": 0.35432970091772553,
"learning_rate": 2.760165589921184e-06,
"loss": 0.1502,
"step": 1047
},
{
"epoch": 4.242914979757085,
"grad_norm": 0.3400480818176099,
"learning_rate": 2.7315659020667305e-06,
"loss": 0.1578,
"step": 1048
},
{
"epoch": 4.246963562753036,
"grad_norm": 0.34140781849879354,
"learning_rate": 2.703104292151164e-06,
"loss": 0.1576,
"step": 1049
},
{
"epoch": 4.251012145748988,
"grad_norm": 0.33710020245751426,
"learning_rate": 2.6747809877527607e-06,
"loss": 0.125,
"step": 1050
},
{
"epoch": 4.255060728744939,
"grad_norm": 0.36331111668653326,
"learning_rate": 2.6465962153439082e-06,
"loss": 0.1686,
"step": 1051
},
{
"epoch": 4.2591093117408905,
"grad_norm": 0.3800074594835834,
"learning_rate": 2.618550200289303e-06,
"loss": 0.1039,
"step": 1052
},
{
"epoch": 4.2631578947368425,
"grad_norm": 0.33878558595686714,
"learning_rate": 2.59064316684412e-06,
"loss": 0.1157,
"step": 1053
},
{
"epoch": 4.267206477732794,
"grad_norm": 0.3339513645428499,
"learning_rate": 2.562875338152264e-06,
"loss": 0.1158,
"step": 1054
},
{
"epoch": 4.271255060728745,
"grad_norm": 0.3622172747841323,
"learning_rate": 2.5352469362445552e-06,
"loss": 0.1404,
"step": 1055
},
{
"epoch": 4.275303643724697,
"grad_norm": 0.35644930628812427,
"learning_rate": 2.5077581820369523e-06,
"loss": 0.1216,
"step": 1056
},
{
"epoch": 4.279352226720648,
"grad_norm": 0.37449868550236987,
"learning_rate": 2.480409295328803e-06,
"loss": 0.094,
"step": 1057
},
{
"epoch": 4.283400809716599,
"grad_norm": 0.3701780318839263,
"learning_rate": 2.45320049480108e-06,
"loss": 0.2749,
"step": 1058
},
{
"epoch": 4.287449392712551,
"grad_norm": 0.3441684201080886,
"learning_rate": 2.4261319980146293e-06,
"loss": 0.116,
"step": 1059
},
{
"epoch": 4.291497975708502,
"grad_norm": 0.34050547751948845,
"learning_rate": 2.399204021408421e-06,
"loss": 0.1208,
"step": 1060
},
{
"epoch": 4.295546558704453,
"grad_norm": 0.36241461292062854,
"learning_rate": 2.3724167802978395e-06,
"loss": 0.133,
"step": 1061
},
{
"epoch": 4.299595141700405,
"grad_norm": 0.3894708706334087,
"learning_rate": 2.3457704888729625e-06,
"loss": 0.1334,
"step": 1062
},
{
"epoch": 4.303643724696356,
"grad_norm": 0.37339110733007824,
"learning_rate": 2.3192653601968162e-06,
"loss": 0.1246,
"step": 1063
},
{
"epoch": 4.3076923076923075,
"grad_norm": 0.3301954708718111,
"learning_rate": 2.2929016062037057e-06,
"loss": 0.0986,
"step": 1064
},
{
"epoch": 4.3117408906882595,
"grad_norm": 0.3121626769151538,
"learning_rate": 2.2666794376975055e-06,
"loss": 0.125,
"step": 1065
},
{
"epoch": 4.315789473684211,
"grad_norm": 0.3519925462154742,
"learning_rate": 2.24059906434998e-06,
"loss": 0.1442,
"step": 1066
},
{
"epoch": 4.319838056680162,
"grad_norm": 0.4129407102340751,
"learning_rate": 2.214660694699091e-06,
"loss": 0.2154,
"step": 1067
},
{
"epoch": 4.323886639676114,
"grad_norm": 0.3657774094399879,
"learning_rate": 2.188864536147355e-06,
"loss": 0.1149,
"step": 1068
},
{
"epoch": 4.327935222672065,
"grad_norm": 0.37067694663148737,
"learning_rate": 2.1632107949601757e-06,
"loss": 0.1052,
"step": 1069
},
{
"epoch": 4.331983805668016,
"grad_norm": 0.3656479377804639,
"learning_rate": 2.137699676264171e-06,
"loss": 0.1423,
"step": 1070
},
{
"epoch": 4.336032388663968,
"grad_norm": 0.4095584993018547,
"learning_rate": 2.1123313840455738e-06,
"loss": 0.1351,
"step": 1071
},
{
"epoch": 4.340080971659919,
"grad_norm": 0.3403249432394174,
"learning_rate": 2.08710612114857e-06,
"loss": 0.0937,
"step": 1072
},
{
"epoch": 4.34412955465587,
"grad_norm": 0.39728939604174823,
"learning_rate": 2.0620240892736953e-06,
"loss": 0.1382,
"step": 1073
},
{
"epoch": 4.348178137651822,
"grad_norm": 0.3491139450204542,
"learning_rate": 2.0370854889762e-06,
"loss": 0.0894,
"step": 1074
},
{
"epoch": 4.352226720647773,
"grad_norm": 0.36364299042941195,
"learning_rate": 2.0122905196644726e-06,
"loss": 0.0883,
"step": 1075
},
{
"epoch": 4.3562753036437245,
"grad_norm": 0.33837840406112435,
"learning_rate": 1.987639379598427e-06,
"loss": 0.0737,
"step": 1076
},
{
"epoch": 4.3603238866396765,
"grad_norm": 0.37714369890777305,
"learning_rate": 1.963132265887919e-06,
"loss": 0.1373,
"step": 1077
},
{
"epoch": 4.364372469635628,
"grad_norm": 0.34931542217967404,
"learning_rate": 1.9387693744911784e-06,
"loss": 0.1184,
"step": 1078
},
{
"epoch": 4.368421052631579,
"grad_norm": 0.3454112574687263,
"learning_rate": 1.9145509002132368e-06,
"loss": 0.1033,
"step": 1079
},
{
"epoch": 4.372469635627531,
"grad_norm": 0.34350174131247574,
"learning_rate": 1.8904770367043723e-06,
"loss": 0.1165,
"step": 1080
},
{
"epoch": 4.376518218623482,
"grad_norm": 0.32861860913799884,
"learning_rate": 1.8665479764585504e-06,
"loss": 0.112,
"step": 1081
},
{
"epoch": 4.380566801619433,
"grad_norm": 0.3573142617911389,
"learning_rate": 1.8427639108118977e-06,
"loss": 0.1096,
"step": 1082
},
{
"epoch": 4.384615384615385,
"grad_norm": 0.32150828637764506,
"learning_rate": 1.8191250299411755e-06,
"loss": 0.1041,
"step": 1083
},
{
"epoch": 4.388663967611336,
"grad_norm": 0.3236673063529711,
"learning_rate": 1.7956315228622446e-06,
"loss": 0.1025,
"step": 1084
},
{
"epoch": 4.392712550607287,
"grad_norm": 0.3796818496323241,
"learning_rate": 1.7722835774285553e-06,
"loss": 0.0963,
"step": 1085
},
{
"epoch": 4.396761133603239,
"grad_norm": 0.4393524290819077,
"learning_rate": 1.7490813803296624e-06,
"loss": 0.2077,
"step": 1086
},
{
"epoch": 4.40080971659919,
"grad_norm": 0.3599308726117657,
"learning_rate": 1.726025117089718e-06,
"loss": 0.1123,
"step": 1087
},
{
"epoch": 4.4048582995951415,
"grad_norm": 0.39240552743773954,
"learning_rate": 1.70311497206598e-06,
"loss": 0.1117,
"step": 1088
},
{
"epoch": 4.4089068825910935,
"grad_norm": 0.3702754841199525,
"learning_rate": 1.6803511284473638e-06,
"loss": 0.114,
"step": 1089
},
{
"epoch": 4.412955465587045,
"grad_norm": 0.3685641743011286,
"learning_rate": 1.6577337682529538e-06,
"loss": 0.1477,
"step": 1090
},
{
"epoch": 4.417004048582996,
"grad_norm": 0.360432806719021,
"learning_rate": 1.6352630723305617e-06,
"loss": 0.1285,
"step": 1091
},
{
"epoch": 4.421052631578947,
"grad_norm": 0.3491952205573264,
"learning_rate": 1.612939220355272e-06,
"loss": 0.1516,
"step": 1092
},
{
"epoch": 4.425101214574899,
"grad_norm": 0.3590581850665295,
"learning_rate": 1.5907623908280112e-06,
"loss": 0.1246,
"step": 1093
},
{
"epoch": 4.42914979757085,
"grad_norm": 0.35055423605555885,
"learning_rate": 1.568732761074121e-06,
"loss": 0.1316,
"step": 1094
},
{
"epoch": 4.433198380566802,
"grad_norm": 0.40436316852426135,
"learning_rate": 1.5468505072419237e-06,
"loss": 0.1858,
"step": 1095
},
{
"epoch": 4.437246963562753,
"grad_norm": 0.3931800762656328,
"learning_rate": 1.5251158043013448e-06,
"loss": 0.1466,
"step": 1096
},
{
"epoch": 4.441295546558704,
"grad_norm": 0.38327755488391846,
"learning_rate": 1.5035288260424885e-06,
"loss": 0.0972,
"step": 1097
},
{
"epoch": 4.445344129554655,
"grad_norm": 0.3349162857736668,
"learning_rate": 1.4820897450742623e-06,
"loss": 0.1193,
"step": 1098
},
{
"epoch": 4.449392712550607,
"grad_norm": 0.39451097641634697,
"learning_rate": 1.4607987328229767e-06,
"loss": 0.1277,
"step": 1099
},
{
"epoch": 4.4534412955465585,
"grad_norm": 0.3706358147244772,
"learning_rate": 1.4396559595310056e-06,
"loss": 0.1661,
"step": 1100
},
{
"epoch": 4.4574898785425106,
"grad_norm": 0.35937843701625616,
"learning_rate": 1.4186615942553973e-06,
"loss": 0.1199,
"step": 1101
},
{
"epoch": 4.461538461538462,
"grad_norm": 0.3741412766313026,
"learning_rate": 1.3978158048665424e-06,
"loss": 0.1425,
"step": 1102
},
{
"epoch": 4.465587044534413,
"grad_norm": 0.3616047533218161,
"learning_rate": 1.3771187580468115e-06,
"loss": 0.1188,
"step": 1103
},
{
"epoch": 4.469635627530364,
"grad_norm": 0.33613983195528435,
"learning_rate": 1.3565706192892392e-06,
"loss": 0.1533,
"step": 1104
},
{
"epoch": 4.473684210526316,
"grad_norm": 0.3271377087403872,
"learning_rate": 1.3361715528961993e-06,
"loss": 0.1338,
"step": 1105
},
{
"epoch": 4.477732793522267,
"grad_norm": 0.3697955468500856,
"learning_rate": 1.3159217219780817e-06,
"loss": 0.1267,
"step": 1106
},
{
"epoch": 4.481781376518219,
"grad_norm": 0.35415180211518854,
"learning_rate": 1.2958212884519949e-06,
"loss": 0.0936,
"step": 1107
},
{
"epoch": 4.48582995951417,
"grad_norm": 0.36951337963182873,
"learning_rate": 1.2758704130404675e-06,
"loss": 0.1104,
"step": 1108
},
{
"epoch": 4.489878542510121,
"grad_norm": 0.34611473912492524,
"learning_rate": 1.2560692552701738e-06,
"loss": 0.1224,
"step": 1109
},
{
"epoch": 4.493927125506072,
"grad_norm": 0.3363549724427907,
"learning_rate": 1.2364179734706382e-06,
"loss": 0.1389,
"step": 1110
},
{
"epoch": 4.497975708502024,
"grad_norm": 0.32445899803371175,
"learning_rate": 1.2169167247729852e-06,
"loss": 0.1232,
"step": 1111
},
{
"epoch": 4.502024291497976,
"grad_norm": 0.3694690142545082,
"learning_rate": 1.1975656651086864e-06,
"loss": 0.1261,
"step": 1112
},
{
"epoch": 4.506072874493928,
"grad_norm": 17.160548836721244,
"learning_rate": 1.1783649492082926e-06,
"loss": 0.2691,
"step": 1113
},
{
"epoch": 4.510121457489879,
"grad_norm": 0.38117444603939205,
"learning_rate": 1.1593147306002183e-06,
"loss": 0.0908,
"step": 1114
},
{
"epoch": 4.51417004048583,
"grad_norm": 0.34178819040955705,
"learning_rate": 1.140415161609505e-06,
"loss": 0.1291,
"step": 1115
},
{
"epoch": 4.518218623481781,
"grad_norm": 0.4037828921707418,
"learning_rate": 1.121666393356604e-06,
"loss": 0.1323,
"step": 1116
},
{
"epoch": 4.522267206477733,
"grad_norm": 0.4021449443966613,
"learning_rate": 1.1030685757561632e-06,
"loss": 0.2081,
"step": 1117
},
{
"epoch": 4.526315789473684,
"grad_norm": 0.3562018143341334,
"learning_rate": 1.0846218575158373e-06,
"loss": 0.1098,
"step": 1118
},
{
"epoch": 4.530364372469636,
"grad_norm": 0.34605279878682615,
"learning_rate": 1.066326386135097e-06,
"loss": 0.1189,
"step": 1119
},
{
"epoch": 4.534412955465587,
"grad_norm": 0.3357460576165338,
"learning_rate": 1.0481823079040398e-06,
"loss": 0.1058,
"step": 1120
},
{
"epoch": 4.538461538461538,
"grad_norm": 0.34959119224702345,
"learning_rate": 1.0301897679022344e-06,
"loss": 0.1332,
"step": 1121
},
{
"epoch": 4.5425101214574894,
"grad_norm": 0.34271972450215765,
"learning_rate": 1.0123489099975491e-06,
"loss": 0.1154,
"step": 1122
},
{
"epoch": 4.5465587044534415,
"grad_norm": 0.34031979419875374,
"learning_rate": 9.9465987684501e-07,
"loss": 0.1295,
"step": 1123
},
{
"epoch": 4.550607287449393,
"grad_norm": 0.33556508784467437,
"learning_rate": 9.771228098856534e-07,
"loss": 0.1267,
"step": 1124
},
{
"epoch": 4.554655870445345,
"grad_norm": 0.35779497963682255,
"learning_rate": 9.597378493454013e-07,
"loss": 0.1595,
"step": 1125
},
{
"epoch": 4.558704453441296,
"grad_norm": 0.31909621921227016,
"learning_rate": 9.425051342339353e-07,
"loss": 0.1195,
"step": 1126
},
{
"epoch": 4.562753036437247,
"grad_norm": 0.37169423759246184,
"learning_rate": 9.254248023435864e-07,
"loss": 0.0978,
"step": 1127
},
{
"epoch": 4.566801619433198,
"grad_norm": 0.35931581282505515,
"learning_rate": 9.084969902482288e-07,
"loss": 0.1132,
"step": 1128
},
{
"epoch": 4.57085020242915,
"grad_norm": 0.3575231016793557,
"learning_rate": 8.917218333021993e-07,
"loss": 0.1432,
"step": 1129
},
{
"epoch": 4.574898785425101,
"grad_norm": 0.3570171738170682,
"learning_rate": 8.750994656392109e-07,
"loss": 0.3219,
"step": 1130
},
{
"epoch": 4.578947368421053,
"grad_norm": 0.38309968458843485,
"learning_rate": 8.58630020171265e-07,
"loss": 0.1008,
"step": 1131
},
{
"epoch": 4.582995951417004,
"grad_norm": 0.37453546034346785,
"learning_rate": 8.423136285876099e-07,
"loss": 0.1508,
"step": 1132
},
{
"epoch": 4.587044534412955,
"grad_norm": 0.35806439146085206,
"learning_rate": 8.26150421353682e-07,
"loss": 0.091,
"step": 1133
},
{
"epoch": 4.5910931174089065,
"grad_norm": 0.35268989869469297,
"learning_rate": 8.101405277100549e-07,
"loss": 0.1155,
"step": 1134
},
{
"epoch": 4.5951417004048585,
"grad_norm": 0.35673818523177203,
"learning_rate": 7.942840756714077e-07,
"loss": 0.1431,
"step": 1135
},
{
"epoch": 4.59919028340081,
"grad_norm": 0.3351487088091794,
"learning_rate": 7.785811920255093e-07,
"loss": 0.1048,
"step": 1136
},
{
"epoch": 4.603238866396762,
"grad_norm": 0.3179921542841759,
"learning_rate": 7.630320023322024e-07,
"loss": 0.1454,
"step": 1137
},
{
"epoch": 4.607287449392713,
"grad_norm": 0.3846198436864008,
"learning_rate": 7.476366309223903e-07,
"loss": 0.1532,
"step": 1138
},
{
"epoch": 4.611336032388664,
"grad_norm": 0.3675325523065151,
"learning_rate": 7.323952008970537e-07,
"loss": 0.1468,
"step": 1139
},
{
"epoch": 4.615384615384615,
"grad_norm": 0.3501464529805811,
"learning_rate": 7.173078341262618e-07,
"loss": 0.1298,
"step": 1140
},
{
"epoch": 4.619433198380567,
"grad_norm": 0.3140101947889388,
"learning_rate": 7.023746512481989e-07,
"loss": 0.1289,
"step": 1141
},
{
"epoch": 4.623481781376518,
"grad_norm": 0.4136393549693248,
"learning_rate": 6.875957716681902e-07,
"loss": 0.1328,
"step": 1142
},
{
"epoch": 4.62753036437247,
"grad_norm": 0.3789036634665248,
"learning_rate": 6.729713135577687e-07,
"loss": 0.115,
"step": 1143
},
{
"epoch": 4.631578947368421,
"grad_norm": 0.3520431935832073,
"learning_rate": 6.585013938537143e-07,
"loss": 0.1736,
"step": 1144
},
{
"epoch": 4.635627530364372,
"grad_norm": 0.37684213019886514,
"learning_rate": 6.441861282571115e-07,
"loss": 0.121,
"step": 1145
},
{
"epoch": 4.6396761133603235,
"grad_norm": 0.3191643443386397,
"learning_rate": 6.300256312324471e-07,
"loss": 0.1492,
"step": 1146
},
{
"epoch": 4.6437246963562755,
"grad_norm": 0.32690680177704884,
"learning_rate": 6.160200160066776e-07,
"loss": 0.1216,
"step": 1147
},
{
"epoch": 4.647773279352227,
"grad_norm": 0.3479529256272357,
"learning_rate": 6.021693945683327e-07,
"loss": 0.1749,
"step": 1148
},
{
"epoch": 4.651821862348179,
"grad_norm": 0.3506878576485915,
"learning_rate": 5.884738776666044e-07,
"loss": 0.1214,
"step": 1149
},
{
"epoch": 4.65587044534413,
"grad_norm": 0.3403142993209038,
"learning_rate": 5.749335748104878e-07,
"loss": 0.0984,
"step": 1150
},
{
"epoch": 4.659919028340081,
"grad_norm": 0.38453343875029444,
"learning_rate": 5.615485942678778e-07,
"loss": 0.1289,
"step": 1151
},
{
"epoch": 4.663967611336032,
"grad_norm": 0.35919126427072307,
"learning_rate": 5.483190430647289e-07,
"loss": 0.1301,
"step": 1152
},
{
"epoch": 4.668016194331984,
"grad_norm": 0.39314483620372714,
"learning_rate": 5.352450269841747e-07,
"loss": 0.1357,
"step": 1153
},
{
"epoch": 4.672064777327935,
"grad_norm": 0.33986123117470163,
"learning_rate": 5.22326650565701e-07,
"loss": 0.1305,
"step": 1154
},
{
"epoch": 4.676113360323887,
"grad_norm": 0.3855403178409885,
"learning_rate": 5.095640171043048e-07,
"loss": 0.1626,
"step": 1155
},
{
"epoch": 4.680161943319838,
"grad_norm": 0.3315089740399664,
"learning_rate": 4.969572286496594e-07,
"loss": 0.1485,
"step": 1156
},
{
"epoch": 4.684210526315789,
"grad_norm": 0.3338481950324901,
"learning_rate": 4.845063860053079e-07,
"loss": 0.1341,
"step": 1157
},
{
"epoch": 4.6882591093117405,
"grad_norm": 0.35062666897781425,
"learning_rate": 4.7221158872785997e-07,
"loss": 0.1198,
"step": 1158
},
{
"epoch": 4.6923076923076925,
"grad_norm": 0.33630353762752635,
"learning_rate": 4.600729351261857e-07,
"loss": 0.1336,
"step": 1159
},
{
"epoch": 4.696356275303644,
"grad_norm": 0.315491510475028,
"learning_rate": 4.4809052226063355e-07,
"loss": 0.1393,
"step": 1160
},
{
"epoch": 4.700404858299595,
"grad_norm": 0.4170650734543035,
"learning_rate": 4.3626444594226046e-07,
"loss": 0.1561,
"step": 1161
},
{
"epoch": 4.704453441295547,
"grad_norm": 0.41406515310645703,
"learning_rate": 4.245948007320588e-07,
"loss": 0.1133,
"step": 1162
},
{
"epoch": 4.708502024291498,
"grad_norm": 0.3383003444610845,
"learning_rate": 4.130816799401971e-07,
"loss": 0.1522,
"step": 1163
},
{
"epoch": 4.712550607287449,
"grad_norm": 0.3395939878808812,
"learning_rate": 4.017251756252827e-07,
"loss": 0.1231,
"step": 1164
},
{
"epoch": 4.716599190283401,
"grad_norm": 0.3445092242294066,
"learning_rate": 3.9052537859362473e-07,
"loss": 0.141,
"step": 1165
},
{
"epoch": 4.720647773279352,
"grad_norm": 0.3394070995613691,
"learning_rate": 3.7948237839849687e-07,
"loss": 0.1331,
"step": 1166
},
{
"epoch": 4.724696356275303,
"grad_norm": 0.3518888240370481,
"learning_rate": 3.6859626333943797e-07,
"loss": 0.1306,
"step": 1167
},
{
"epoch": 4.728744939271255,
"grad_norm": 0.3570198850006878,
"learning_rate": 3.578671204615347e-07,
"loss": 0.1368,
"step": 1168
},
{
"epoch": 4.732793522267206,
"grad_norm": 0.4163942755326552,
"learning_rate": 3.472950355547267e-07,
"loss": 0.213,
"step": 1169
},
{
"epoch": 4.7368421052631575,
"grad_norm": 0.32154732975847505,
"learning_rate": 3.3688009315312244e-07,
"loss": 0.1172,
"step": 1170
},
{
"epoch": 4.7408906882591095,
"grad_norm": 0.33690397574822584,
"learning_rate": 3.266223765343224e-07,
"loss": 0.1345,
"step": 1171
},
{
"epoch": 4.744939271255061,
"grad_norm": 0.35755814171841405,
"learning_rate": 3.16521967718757e-07,
"loss": 0.13,
"step": 1172
},
{
"epoch": 4.748987854251012,
"grad_norm": 0.3345106766873073,
"learning_rate": 3.0657894746902726e-07,
"loss": 0.114,
"step": 1173
},
{
"epoch": 4.753036437246964,
"grad_norm": 0.3227754764111626,
"learning_rate": 2.9679339528925654e-07,
"loss": 0.1506,
"step": 1174
},
{
"epoch": 4.757085020242915,
"grad_norm": 0.33466904433378203,
"learning_rate": 2.871653894244597e-07,
"loss": 0.083,
"step": 1175
},
{
"epoch": 4.761133603238866,
"grad_norm": 0.3927396949584569,
"learning_rate": 2.7769500685991266e-07,
"loss": 0.0958,
"step": 1176
},
{
"epoch": 4.765182186234818,
"grad_norm": 0.3576956586985309,
"learning_rate": 2.683823233205485e-07,
"loss": 0.1299,
"step": 1177
},
{
"epoch": 4.769230769230769,
"grad_norm": 0.37257188414710035,
"learning_rate": 2.592274132703287e-07,
"loss": 0.1313,
"step": 1178
},
{
"epoch": 4.77327935222672,
"grad_norm": 0.36515851494793405,
"learning_rate": 2.502303499116754e-07,
"loss": 0.1318,
"step": 1179
},
{
"epoch": 4.777327935222672,
"grad_norm": 0.3659241514211836,
"learning_rate": 2.413912051848666e-07,
"loss": 0.1065,
"step": 1180
},
{
"epoch": 4.781376518218623,
"grad_norm": 0.3151390241854734,
"learning_rate": 2.3271004976746613e-07,
"loss": 0.1322,
"step": 1181
},
{
"epoch": 4.7854251012145745,
"grad_norm": 0.31619681176538433,
"learning_rate": 2.2418695307376393e-07,
"loss": 0.1195,
"step": 1182
},
{
"epoch": 4.7894736842105265,
"grad_norm": 0.3415560276914603,
"learning_rate": 2.1582198325420968e-07,
"loss": 0.1096,
"step": 1183
},
{
"epoch": 4.793522267206478,
"grad_norm": 0.3373124807133392,
"learning_rate": 2.0761520719488447e-07,
"loss": 0.1101,
"step": 1184
},
{
"epoch": 4.797570850202429,
"grad_norm": 0.4043862254658608,
"learning_rate": 1.9956669051694798e-07,
"loss": 0.138,
"step": 1185
},
{
"epoch": 4.801619433198381,
"grad_norm": 0.35318234857966235,
"learning_rate": 1.916764975761254e-07,
"loss": 0.1206,
"step": 1186
},
{
"epoch": 4.805668016194332,
"grad_norm": 0.3674297345556044,
"learning_rate": 1.839446914621923e-07,
"loss": 0.1472,
"step": 1187
},
{
"epoch": 4.809716599190283,
"grad_norm": 0.3620502125928328,
"learning_rate": 1.7637133399845962e-07,
"loss": 0.1357,
"step": 1188
},
{
"epoch": 4.813765182186235,
"grad_norm": 0.3655219615154056,
"learning_rate": 1.6895648574129397e-07,
"loss": 0.3002,
"step": 1189
},
{
"epoch": 4.817813765182186,
"grad_norm": 0.35902729446656373,
"learning_rate": 1.6170020597962465e-07,
"loss": 0.1025,
"step": 1190
},
{
"epoch": 4.821862348178137,
"grad_norm": 0.33608916528176724,
"learning_rate": 1.5460255273447078e-07,
"loss": 0.1548,
"step": 1191
},
{
"epoch": 4.825910931174089,
"grad_norm": 0.3558522895399451,
"learning_rate": 1.4766358275847715e-07,
"loss": 0.1011,
"step": 1192
},
{
"epoch": 4.82995951417004,
"grad_norm": 0.40663922878189834,
"learning_rate": 1.4088335153546573e-07,
"loss": 0.1335,
"step": 1193
},
{
"epoch": 4.834008097165992,
"grad_norm": 0.3024688157683703,
"learning_rate": 1.3426191327998496e-07,
"loss": 0.1147,
"step": 1194
},
{
"epoch": 4.838056680161944,
"grad_norm": 0.320832822085655,
"learning_rate": 1.277993209368744e-07,
"loss": 0.1279,
"step": 1195
},
{
"epoch": 4.842105263157895,
"grad_norm": 0.32829109496761805,
"learning_rate": 1.214956261808542e-07,
"loss": 0.1381,
"step": 1196
},
{
"epoch": 4.846153846153846,
"grad_norm": 0.39180955968365794,
"learning_rate": 1.1535087941609846e-07,
"loss": 0.1427,
"step": 1197
},
{
"epoch": 4.850202429149798,
"grad_norm": 0.32662876595569923,
"learning_rate": 1.0936512977583802e-07,
"loss": 0.1561,
"step": 1198
},
{
"epoch": 4.854251012145749,
"grad_norm": 0.34224347982279,
"learning_rate": 1.0353842512196732e-07,
"loss": 0.3337,
"step": 1199
},
{
"epoch": 4.8582995951417,
"grad_norm": 0.38043614851623625,
"learning_rate": 9.787081204466253e-08,
"loss": 0.2004,
"step": 1200
},
{
"epoch": 4.862348178137652,
"grad_norm": 0.4158836280297478,
"learning_rate": 9.236233586200627e-08,
"loss": 0.1256,
"step": 1201
},
{
"epoch": 4.866396761133603,
"grad_norm": 0.3737471881134617,
"learning_rate": 8.70130406196279e-08,
"loss": 0.1336,
"step": 1202
},
{
"epoch": 4.870445344129554,
"grad_norm": 0.3592440511207493,
"learning_rate": 8.182296909035047e-08,
"loss": 0.1152,
"step": 1203
},
{
"epoch": 4.874493927125506,
"grad_norm": 0.3468040281771846,
"learning_rate": 7.679216277384439e-08,
"loss": 0.096,
"step": 1204
},
{
"epoch": 4.8785425101214575,
"grad_norm": 0.3532873843711534,
"learning_rate": 7.19206618963053e-08,
"loss": 0.0997,
"step": 1205
},
{
"epoch": 4.882591093117409,
"grad_norm": 0.3437386951644026,
"learning_rate": 6.720850541012347e-08,
"loss": 0.1203,
"step": 1206
},
{
"epoch": 4.886639676113361,
"grad_norm": 0.4458600038904049,
"learning_rate": 6.265573099357714e-08,
"loss": 0.1345,
"step": 1207
},
{
"epoch": 4.890688259109312,
"grad_norm": 0.3205111227228922,
"learning_rate": 5.826237505053067e-08,
"loss": 0.14,
"step": 1208
},
{
"epoch": 4.894736842105263,
"grad_norm": 0.3045062941161537,
"learning_rate": 5.4028472710137e-08,
"loss": 0.1029,
"step": 1209
},
{
"epoch": 4.898785425101215,
"grad_norm": 0.3949587164183892,
"learning_rate": 4.9954057826571146e-08,
"loss": 0.1114,
"step": 1210
},
{
"epoch": 4.902834008097166,
"grad_norm": 0.3866252111735923,
"learning_rate": 4.6039162978743776e-08,
"loss": 0.1347,
"step": 1211
},
{
"epoch": 4.906882591093117,
"grad_norm": 0.30910668384565354,
"learning_rate": 4.228381947004812e-08,
"loss": 0.1123,
"step": 1212
},
{
"epoch": 4.910931174089069,
"grad_norm": 0.32103972812846626,
"learning_rate": 3.868805732811343e-08,
"loss": 0.1226,
"step": 1213
},
{
"epoch": 4.91497975708502,
"grad_norm": 0.3441455449212831,
"learning_rate": 3.525190530455636e-08,
"loss": 0.0964,
"step": 1214
},
{
"epoch": 4.919028340080971,
"grad_norm": 0.35785319420218337,
"learning_rate": 3.197539087475887e-08,
"loss": 0.1518,
"step": 1215
},
{
"epoch": 4.923076923076923,
"grad_norm": 0.32083768199960466,
"learning_rate": 2.8858540237641786e-08,
"loss": 0.1032,
"step": 1216
},
{
"epoch": 4.9271255060728745,
"grad_norm": 0.3829511664327532,
"learning_rate": 2.5901378315458247e-08,
"loss": 0.0991,
"step": 1217
},
{
"epoch": 4.931174089068826,
"grad_norm": 0.3485844055756932,
"learning_rate": 2.3103928753598347e-08,
"loss": 0.1367,
"step": 1218
},
{
"epoch": 4.935222672064778,
"grad_norm": 0.3385715542038344,
"learning_rate": 2.0466213920393718e-08,
"loss": 0.2183,
"step": 1219
},
{
"epoch": 4.939271255060729,
"grad_norm": 0.36404243011989373,
"learning_rate": 1.7988254906942115e-08,
"loss": 0.1205,
"step": 1220
},
{
"epoch": 4.94331983805668,
"grad_norm": 0.34573721944274277,
"learning_rate": 1.5670071526936447e-08,
"loss": 0.2423,
"step": 1221
},
{
"epoch": 4.947368421052632,
"grad_norm": 0.3496238641806388,
"learning_rate": 1.3511682316509344e-08,
"loss": 0.1081,
"step": 1222
},
{
"epoch": 4.951417004048583,
"grad_norm": 0.34324737841915026,
"learning_rate": 1.1513104534086605e-08,
"loss": 0.1041,
"step": 1223
},
{
"epoch": 4.955465587044534,
"grad_norm": 0.41677765316185095,
"learning_rate": 9.67435416023843e-09,
"loss": 0.1994,
"step": 1224
},
{
"epoch": 4.959514170040486,
"grad_norm": 0.39099387374681077,
"learning_rate": 7.995445897566178e-09,
"loss": 0.109,
"step": 1225
},
{
"epoch": 4.963562753036437,
"grad_norm": 0.3536923903809609,
"learning_rate": 6.476393170578021e-09,
"loss": 0.1269,
"step": 1226
},
{
"epoch": 4.967611336032388,
"grad_norm": 0.34277817492924406,
"learning_rate": 5.11720812557348e-09,
"loss": 0.1436,
"step": 1227
},
{
"epoch": 4.97165991902834,
"grad_norm": 0.38301349466796397,
"learning_rate": 3.917901630563492e-09,
"loss": 0.1174,
"step": 1228
},
{
"epoch": 4.9757085020242915,
"grad_norm": 0.31597638874449263,
"learning_rate": 2.878483275170485e-09,
"loss": 0.1196,
"step": 1229
},
{
"epoch": 4.979757085020243,
"grad_norm": 0.36257417516927987,
"learning_rate": 1.998961370555108e-09,
"loss": 0.1496,
"step": 1230
},
{
"epoch": 4.983805668016195,
"grad_norm": 0.4018191091421946,
"learning_rate": 1.2793429493518361e-09,
"loss": 0.2519,
"step": 1231
},
{
"epoch": 4.987854251012146,
"grad_norm": 0.36600998309944394,
"learning_rate": 7.196337656112384e-10,
"loss": 0.1216,
"step": 1232
},
{
"epoch": 4.991902834008097,
"grad_norm": 0.34853677220869234,
"learning_rate": 3.198382947511292e-10,
"loss": 0.1065,
"step": 1233
},
{
"epoch": 4.995951417004049,
"grad_norm": 0.34154235492761387,
"learning_rate": 7.995973352770137e-11,
"loss": 0.134,
"step": 1234
},
{
"epoch": 5.0,
"grad_norm": 0.4363070362365012,
"learning_rate": 0.0,
"loss": 0.1316,
"step": 1235
},
{
"epoch": 5.0,
"step": 1235,
"total_flos": 8.366862746339246e+17,
"train_loss": 0.3469991305277415,
"train_runtime": 20404.7802,
"train_samples_per_second": 7.743,
"train_steps_per_second": 0.061
}
],
"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": 8.366862746339246e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}