d1_code_mc_llm / trainer_state.json
neginr's picture
End of training
9657f41 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 5.0,
"eval_steps": 500,
"global_step": 1235,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.004048582995951417,
"grad_norm": 6.380504327988584,
"learning_rate": 3.2258064516129035e-07,
"loss": 1.0371,
"step": 1
},
{
"epoch": 0.008097165991902834,
"grad_norm": 6.471978429571713,
"learning_rate": 6.451612903225807e-07,
"loss": 1.0465,
"step": 2
},
{
"epoch": 0.012145748987854251,
"grad_norm": 6.454196938940566,
"learning_rate": 9.67741935483871e-07,
"loss": 1.0318,
"step": 3
},
{
"epoch": 0.016194331983805668,
"grad_norm": 6.365578830018815,
"learning_rate": 1.2903225806451614e-06,
"loss": 1.0452,
"step": 4
},
{
"epoch": 0.020242914979757085,
"grad_norm": 5.959351226181745,
"learning_rate": 1.6129032258064516e-06,
"loss": 1.0296,
"step": 5
},
{
"epoch": 0.024291497975708502,
"grad_norm": 5.8091230209363225,
"learning_rate": 1.935483870967742e-06,
"loss": 1.038,
"step": 6
},
{
"epoch": 0.02834008097165992,
"grad_norm": 4.62804838136208,
"learning_rate": 2.2580645161290324e-06,
"loss": 1.0008,
"step": 7
},
{
"epoch": 0.032388663967611336,
"grad_norm": 4.178053585192368,
"learning_rate": 2.580645161290323e-06,
"loss": 0.993,
"step": 8
},
{
"epoch": 0.03643724696356275,
"grad_norm": 2.6782749846933016,
"learning_rate": 2.903225806451613e-06,
"loss": 0.958,
"step": 9
},
{
"epoch": 0.04048582995951417,
"grad_norm": 2.5844054085928376,
"learning_rate": 3.225806451612903e-06,
"loss": 0.9626,
"step": 10
},
{
"epoch": 0.044534412955465584,
"grad_norm": 2.4808762804440563,
"learning_rate": 3.548387096774194e-06,
"loss": 0.9578,
"step": 11
},
{
"epoch": 0.048582995951417005,
"grad_norm": 3.7352175046413274,
"learning_rate": 3.870967741935484e-06,
"loss": 0.93,
"step": 12
},
{
"epoch": 0.05263157894736842,
"grad_norm": 4.164278310008969,
"learning_rate": 4.193548387096774e-06,
"loss": 0.9346,
"step": 13
},
{
"epoch": 0.05668016194331984,
"grad_norm": 3.837787411292263,
"learning_rate": 4.516129032258065e-06,
"loss": 0.9322,
"step": 14
},
{
"epoch": 0.06072874493927125,
"grad_norm": 3.5873845078241366,
"learning_rate": 4.838709677419355e-06,
"loss": 0.9139,
"step": 15
},
{
"epoch": 0.06477732793522267,
"grad_norm": 2.6336301464486196,
"learning_rate": 5.161290322580646e-06,
"loss": 0.8763,
"step": 16
},
{
"epoch": 0.06882591093117409,
"grad_norm": 2.4552792457757198,
"learning_rate": 5.483870967741935e-06,
"loss": 0.8737,
"step": 17
},
{
"epoch": 0.0728744939271255,
"grad_norm": 2.091770350437549,
"learning_rate": 5.806451612903226e-06,
"loss": 0.8773,
"step": 18
},
{
"epoch": 0.07692307692307693,
"grad_norm": 1.5207008530722836,
"learning_rate": 6.129032258064517e-06,
"loss": 0.8473,
"step": 19
},
{
"epoch": 0.08097165991902834,
"grad_norm": 1.1500874344440868,
"learning_rate": 6.451612903225806e-06,
"loss": 0.8399,
"step": 20
},
{
"epoch": 0.08502024291497975,
"grad_norm": 0.9838623407162037,
"learning_rate": 6.774193548387097e-06,
"loss": 0.8339,
"step": 21
},
{
"epoch": 0.08906882591093117,
"grad_norm": 1.1435880143254613,
"learning_rate": 7.096774193548388e-06,
"loss": 0.805,
"step": 22
},
{
"epoch": 0.0931174089068826,
"grad_norm": 1.017951534182377,
"learning_rate": 7.4193548387096784e-06,
"loss": 0.8103,
"step": 23
},
{
"epoch": 0.09716599190283401,
"grad_norm": 0.8828678265467046,
"learning_rate": 7.741935483870968e-06,
"loss": 0.8026,
"step": 24
},
{
"epoch": 0.10121457489878542,
"grad_norm": 0.9244328041562546,
"learning_rate": 8.064516129032258e-06,
"loss": 0.7711,
"step": 25
},
{
"epoch": 0.10526315789473684,
"grad_norm": 0.8342469106156739,
"learning_rate": 8.387096774193549e-06,
"loss": 0.7847,
"step": 26
},
{
"epoch": 0.10931174089068826,
"grad_norm": 0.7638645536709161,
"learning_rate": 8.70967741935484e-06,
"loss": 0.7588,
"step": 27
},
{
"epoch": 0.11336032388663968,
"grad_norm": 0.7947044859175861,
"learning_rate": 9.03225806451613e-06,
"loss": 0.7731,
"step": 28
},
{
"epoch": 0.11740890688259109,
"grad_norm": 0.7211691415380782,
"learning_rate": 9.35483870967742e-06,
"loss": 0.7529,
"step": 29
},
{
"epoch": 0.1214574898785425,
"grad_norm": 0.7598444589486751,
"learning_rate": 9.67741935483871e-06,
"loss": 0.7618,
"step": 30
},
{
"epoch": 0.12550607287449392,
"grad_norm": 0.6640342978012149,
"learning_rate": 1e-05,
"loss": 0.7557,
"step": 31
},
{
"epoch": 0.12955465587044535,
"grad_norm": 0.6713568065378943,
"learning_rate": 1.0322580645161291e-05,
"loss": 0.7425,
"step": 32
},
{
"epoch": 0.13360323886639677,
"grad_norm": 0.7298172028086936,
"learning_rate": 1.0645161290322582e-05,
"loss": 0.7381,
"step": 33
},
{
"epoch": 0.13765182186234817,
"grad_norm": 0.8255187937413603,
"learning_rate": 1.096774193548387e-05,
"loss": 0.732,
"step": 34
},
{
"epoch": 0.1417004048582996,
"grad_norm": 0.5761396830156303,
"learning_rate": 1.1290322580645164e-05,
"loss": 0.7342,
"step": 35
},
{
"epoch": 0.145748987854251,
"grad_norm": 0.7305806834847355,
"learning_rate": 1.1612903225806453e-05,
"loss": 0.7402,
"step": 36
},
{
"epoch": 0.14979757085020243,
"grad_norm": 0.6914291084859883,
"learning_rate": 1.1935483870967743e-05,
"loss": 0.7297,
"step": 37
},
{
"epoch": 0.15384615384615385,
"grad_norm": 0.6000118377259037,
"learning_rate": 1.2258064516129034e-05,
"loss": 0.7061,
"step": 38
},
{
"epoch": 0.15789473684210525,
"grad_norm": 0.7033731634409024,
"learning_rate": 1.2580645161290324e-05,
"loss": 0.7058,
"step": 39
},
{
"epoch": 0.16194331983805668,
"grad_norm": 0.6643165169360806,
"learning_rate": 1.2903225806451613e-05,
"loss": 0.7031,
"step": 40
},
{
"epoch": 0.1659919028340081,
"grad_norm": 0.6284595819405678,
"learning_rate": 1.3225806451612903e-05,
"loss": 0.6996,
"step": 41
},
{
"epoch": 0.1700404858299595,
"grad_norm": 0.6506516555796614,
"learning_rate": 1.3548387096774194e-05,
"loss": 0.7172,
"step": 42
},
{
"epoch": 0.17408906882591094,
"grad_norm": 0.6143201524482045,
"learning_rate": 1.3870967741935486e-05,
"loss": 0.707,
"step": 43
},
{
"epoch": 0.17813765182186234,
"grad_norm": 0.7629737877147486,
"learning_rate": 1.4193548387096776e-05,
"loss": 0.6937,
"step": 44
},
{
"epoch": 0.18218623481781376,
"grad_norm": 0.5772378971728946,
"learning_rate": 1.4516129032258066e-05,
"loss": 0.6849,
"step": 45
},
{
"epoch": 0.1862348178137652,
"grad_norm": 0.6388523153412494,
"learning_rate": 1.4838709677419357e-05,
"loss": 0.6984,
"step": 46
},
{
"epoch": 0.1902834008097166,
"grad_norm": 0.5961680078598045,
"learning_rate": 1.5161290322580646e-05,
"loss": 0.6904,
"step": 47
},
{
"epoch": 0.19433198380566802,
"grad_norm": 0.47238911392413935,
"learning_rate": 1.5483870967741936e-05,
"loss": 0.6841,
"step": 48
},
{
"epoch": 0.19838056680161945,
"grad_norm": 0.6932777997929243,
"learning_rate": 1.5806451612903226e-05,
"loss": 0.6983,
"step": 49
},
{
"epoch": 0.20242914979757085,
"grad_norm": 0.45978058781221637,
"learning_rate": 1.6129032258064517e-05,
"loss": 0.6911,
"step": 50
},
{
"epoch": 0.20647773279352227,
"grad_norm": 0.6939880937869886,
"learning_rate": 1.6451612903225807e-05,
"loss": 0.684,
"step": 51
},
{
"epoch": 0.21052631578947367,
"grad_norm": 0.48467015928384455,
"learning_rate": 1.6774193548387098e-05,
"loss": 0.6974,
"step": 52
},
{
"epoch": 0.2145748987854251,
"grad_norm": 0.6400469910753417,
"learning_rate": 1.7096774193548388e-05,
"loss": 0.6891,
"step": 53
},
{
"epoch": 0.21862348178137653,
"grad_norm": 0.5568648599299353,
"learning_rate": 1.741935483870968e-05,
"loss": 0.6899,
"step": 54
},
{
"epoch": 0.22267206477732793,
"grad_norm": 0.4801227709206125,
"learning_rate": 1.774193548387097e-05,
"loss": 0.7032,
"step": 55
},
{
"epoch": 0.22672064777327935,
"grad_norm": 0.7500125385278652,
"learning_rate": 1.806451612903226e-05,
"loss": 0.6821,
"step": 56
},
{
"epoch": 0.23076923076923078,
"grad_norm": 0.7206840499834226,
"learning_rate": 1.838709677419355e-05,
"loss": 0.6855,
"step": 57
},
{
"epoch": 0.23481781376518218,
"grad_norm": 0.48894444816503563,
"learning_rate": 1.870967741935484e-05,
"loss": 0.6615,
"step": 58
},
{
"epoch": 0.2388663967611336,
"grad_norm": 0.5506470012401492,
"learning_rate": 1.903225806451613e-05,
"loss": 0.6567,
"step": 59
},
{
"epoch": 0.242914979757085,
"grad_norm": 0.6291672327081734,
"learning_rate": 1.935483870967742e-05,
"loss": 0.6772,
"step": 60
},
{
"epoch": 0.24696356275303644,
"grad_norm": 0.6647227084375813,
"learning_rate": 1.967741935483871e-05,
"loss": 0.6705,
"step": 61
},
{
"epoch": 0.25101214574898784,
"grad_norm": 0.5507355671063752,
"learning_rate": 2e-05,
"loss": 0.6695,
"step": 62
},
{
"epoch": 0.2550607287449393,
"grad_norm": 0.5751869325846238,
"learning_rate": 2.0322580645161292e-05,
"loss": 0.6766,
"step": 63
},
{
"epoch": 0.2591093117408907,
"grad_norm": 0.7261657860168421,
"learning_rate": 2.0645161290322582e-05,
"loss": 0.6609,
"step": 64
},
{
"epoch": 0.2631578947368421,
"grad_norm": 1.1256015503637904,
"learning_rate": 2.0967741935483873e-05,
"loss": 0.6719,
"step": 65
},
{
"epoch": 0.26720647773279355,
"grad_norm": 0.8254668504258054,
"learning_rate": 2.1290322580645163e-05,
"loss": 0.6672,
"step": 66
},
{
"epoch": 0.27125506072874495,
"grad_norm": 0.5195104622207534,
"learning_rate": 2.161290322580645e-05,
"loss": 0.6718,
"step": 67
},
{
"epoch": 0.27530364372469635,
"grad_norm": 0.6274107964197262,
"learning_rate": 2.193548387096774e-05,
"loss": 0.6696,
"step": 68
},
{
"epoch": 0.2793522267206478,
"grad_norm": 0.7974563543498501,
"learning_rate": 2.225806451612903e-05,
"loss": 0.6713,
"step": 69
},
{
"epoch": 0.2834008097165992,
"grad_norm": 0.7630953426511041,
"learning_rate": 2.2580645161290328e-05,
"loss": 0.6618,
"step": 70
},
{
"epoch": 0.2874493927125506,
"grad_norm": 0.6765803394048432,
"learning_rate": 2.2903225806451615e-05,
"loss": 0.6665,
"step": 71
},
{
"epoch": 0.291497975708502,
"grad_norm": 0.8710189523964542,
"learning_rate": 2.3225806451612906e-05,
"loss": 0.6587,
"step": 72
},
{
"epoch": 0.29554655870445345,
"grad_norm": 0.7479484680426236,
"learning_rate": 2.3548387096774196e-05,
"loss": 0.6747,
"step": 73
},
{
"epoch": 0.29959514170040485,
"grad_norm": 0.7289027780156538,
"learning_rate": 2.3870967741935486e-05,
"loss": 0.6707,
"step": 74
},
{
"epoch": 0.30364372469635625,
"grad_norm": 1.0068442583572375,
"learning_rate": 2.4193548387096777e-05,
"loss": 0.6812,
"step": 75
},
{
"epoch": 0.3076923076923077,
"grad_norm": 1.4095991212157215,
"learning_rate": 2.4516129032258067e-05,
"loss": 0.6624,
"step": 76
},
{
"epoch": 0.3117408906882591,
"grad_norm": 0.8056087119354614,
"learning_rate": 2.4838709677419358e-05,
"loss": 0.6492,
"step": 77
},
{
"epoch": 0.3157894736842105,
"grad_norm": 1.5442688348980358,
"learning_rate": 2.5161290322580648e-05,
"loss": 0.6577,
"step": 78
},
{
"epoch": 0.31983805668016196,
"grad_norm": 0.7896174958883257,
"learning_rate": 2.548387096774194e-05,
"loss": 0.6593,
"step": 79
},
{
"epoch": 0.32388663967611336,
"grad_norm": 1.5968753195192698,
"learning_rate": 2.5806451612903226e-05,
"loss": 0.6596,
"step": 80
},
{
"epoch": 0.32793522267206476,
"grad_norm": 0.7327968119891873,
"learning_rate": 2.6129032258064516e-05,
"loss": 0.6726,
"step": 81
},
{
"epoch": 0.3319838056680162,
"grad_norm": 1.5020501040485572,
"learning_rate": 2.6451612903225806e-05,
"loss": 0.6647,
"step": 82
},
{
"epoch": 0.3360323886639676,
"grad_norm": 0.9922004572409159,
"learning_rate": 2.6774193548387097e-05,
"loss": 0.6519,
"step": 83
},
{
"epoch": 0.340080971659919,
"grad_norm": 1.4783113355939315,
"learning_rate": 2.7096774193548387e-05,
"loss": 0.6457,
"step": 84
},
{
"epoch": 0.3441295546558704,
"grad_norm": 1.3224646982202917,
"learning_rate": 2.7419354838709678e-05,
"loss": 0.6529,
"step": 85
},
{
"epoch": 0.3481781376518219,
"grad_norm": 1.3698615349123318,
"learning_rate": 2.774193548387097e-05,
"loss": 0.6626,
"step": 86
},
{
"epoch": 0.3522267206477733,
"grad_norm": 1.4425485109691472,
"learning_rate": 2.8064516129032262e-05,
"loss": 0.6658,
"step": 87
},
{
"epoch": 0.3562753036437247,
"grad_norm": 1.4036676555425334,
"learning_rate": 2.8387096774193552e-05,
"loss": 0.6509,
"step": 88
},
{
"epoch": 0.3603238866396761,
"grad_norm": 1.0633467311875515,
"learning_rate": 2.8709677419354843e-05,
"loss": 0.654,
"step": 89
},
{
"epoch": 0.3643724696356275,
"grad_norm": 1.477489977000131,
"learning_rate": 2.9032258064516133e-05,
"loss": 0.6484,
"step": 90
},
{
"epoch": 0.3684210526315789,
"grad_norm": 1.0570556113371419,
"learning_rate": 2.9354838709677423e-05,
"loss": 0.6446,
"step": 91
},
{
"epoch": 0.3724696356275304,
"grad_norm": 1.061329622373411,
"learning_rate": 2.9677419354838714e-05,
"loss": 0.6657,
"step": 92
},
{
"epoch": 0.3765182186234818,
"grad_norm": 1.6066030318268782,
"learning_rate": 3.0000000000000004e-05,
"loss": 0.6632,
"step": 93
},
{
"epoch": 0.3805668016194332,
"grad_norm": 1.0233107852991696,
"learning_rate": 3.032258064516129e-05,
"loss": 0.6422,
"step": 94
},
{
"epoch": 0.38461538461538464,
"grad_norm": 1.2452828246428642,
"learning_rate": 3.0645161290322585e-05,
"loss": 0.6535,
"step": 95
},
{
"epoch": 0.38866396761133604,
"grad_norm": 1.3559642568868504,
"learning_rate": 3.096774193548387e-05,
"loss": 0.6521,
"step": 96
},
{
"epoch": 0.39271255060728744,
"grad_norm": 1.1505996505160931,
"learning_rate": 3.1290322580645166e-05,
"loss": 0.6508,
"step": 97
},
{
"epoch": 0.3967611336032389,
"grad_norm": 1.4744589710505556,
"learning_rate": 3.161290322580645e-05,
"loss": 0.6625,
"step": 98
},
{
"epoch": 0.4008097165991903,
"grad_norm": 0.9960259346948337,
"learning_rate": 3.193548387096774e-05,
"loss": 0.6518,
"step": 99
},
{
"epoch": 0.4048582995951417,
"grad_norm": 1.6993686373144559,
"learning_rate": 3.2258064516129034e-05,
"loss": 0.6461,
"step": 100
},
{
"epoch": 0.4089068825910931,
"grad_norm": 1.073204600107891,
"learning_rate": 3.258064516129033e-05,
"loss": 0.6445,
"step": 101
},
{
"epoch": 0.41295546558704455,
"grad_norm": 1.8224044428806039,
"learning_rate": 3.2903225806451614e-05,
"loss": 0.6597,
"step": 102
},
{
"epoch": 0.41700404858299595,
"grad_norm": 1.2041239818906626,
"learning_rate": 3.322580645161291e-05,
"loss": 0.6557,
"step": 103
},
{
"epoch": 0.42105263157894735,
"grad_norm": 1.493291188264804,
"learning_rate": 3.3548387096774195e-05,
"loss": 0.6419,
"step": 104
},
{
"epoch": 0.4251012145748988,
"grad_norm": 1.2447108284334414,
"learning_rate": 3.387096774193549e-05,
"loss": 0.6437,
"step": 105
},
{
"epoch": 0.4291497975708502,
"grad_norm": 1.125421232657515,
"learning_rate": 3.4193548387096776e-05,
"loss": 0.6471,
"step": 106
},
{
"epoch": 0.4331983805668016,
"grad_norm": 1.0002577394382406,
"learning_rate": 3.451612903225807e-05,
"loss": 0.6297,
"step": 107
},
{
"epoch": 0.43724696356275305,
"grad_norm": 1.2235858978961955,
"learning_rate": 3.483870967741936e-05,
"loss": 0.6496,
"step": 108
},
{
"epoch": 0.44129554655870445,
"grad_norm": 1.182947650238936,
"learning_rate": 3.516129032258065e-05,
"loss": 0.6405,
"step": 109
},
{
"epoch": 0.44534412955465585,
"grad_norm": 1.0242556986894125,
"learning_rate": 3.548387096774194e-05,
"loss": 0.6334,
"step": 110
},
{
"epoch": 0.4493927125506073,
"grad_norm": 1.3040280751123141,
"learning_rate": 3.5806451612903225e-05,
"loss": 0.6421,
"step": 111
},
{
"epoch": 0.4534412955465587,
"grad_norm": 0.9428970410030755,
"learning_rate": 3.612903225806452e-05,
"loss": 0.6378,
"step": 112
},
{
"epoch": 0.4574898785425101,
"grad_norm": 1.1149129695121178,
"learning_rate": 3.6451612903225805e-05,
"loss": 0.6404,
"step": 113
},
{
"epoch": 0.46153846153846156,
"grad_norm": 1.014597171755919,
"learning_rate": 3.67741935483871e-05,
"loss": 0.6436,
"step": 114
},
{
"epoch": 0.46558704453441296,
"grad_norm": 1.1664617229971261,
"learning_rate": 3.7096774193548386e-05,
"loss": 0.6487,
"step": 115
},
{
"epoch": 0.46963562753036436,
"grad_norm": 1.1564691259693904,
"learning_rate": 3.741935483870968e-05,
"loss": 0.6546,
"step": 116
},
{
"epoch": 0.47368421052631576,
"grad_norm": 0.9667340064275077,
"learning_rate": 3.7741935483870974e-05,
"loss": 0.637,
"step": 117
},
{
"epoch": 0.4777327935222672,
"grad_norm": 1.3781000310439084,
"learning_rate": 3.806451612903226e-05,
"loss": 0.6376,
"step": 118
},
{
"epoch": 0.4817813765182186,
"grad_norm": 0.8368408437014938,
"learning_rate": 3.8387096774193555e-05,
"loss": 0.6495,
"step": 119
},
{
"epoch": 0.48582995951417,
"grad_norm": 1.0300789986629295,
"learning_rate": 3.870967741935484e-05,
"loss": 0.6376,
"step": 120
},
{
"epoch": 0.4898785425101215,
"grad_norm": 0.9968681600782298,
"learning_rate": 3.9032258064516135e-05,
"loss": 0.638,
"step": 121
},
{
"epoch": 0.4939271255060729,
"grad_norm": 1.014460220207241,
"learning_rate": 3.935483870967742e-05,
"loss": 0.6404,
"step": 122
},
{
"epoch": 0.4979757085020243,
"grad_norm": 0.8783825493154576,
"learning_rate": 3.9677419354838716e-05,
"loss": 0.6482,
"step": 123
},
{
"epoch": 0.5020242914979757,
"grad_norm": 1.2092377287987128,
"learning_rate": 4e-05,
"loss": 0.6338,
"step": 124
},
{
"epoch": 0.5060728744939271,
"grad_norm": 1.508347238125606,
"learning_rate": 3.9999920040266474e-05,
"loss": 0.6398,
"step": 125
},
{
"epoch": 0.5101214574898786,
"grad_norm": 1.2017220949849239,
"learning_rate": 3.999968016170525e-05,
"loss": 0.6374,
"step": 126
},
{
"epoch": 0.5141700404858299,
"grad_norm": 1.3082799793737885,
"learning_rate": 3.9999280366234394e-05,
"loss": 0.643,
"step": 127
},
{
"epoch": 0.5182186234817814,
"grad_norm": 1.851714811050662,
"learning_rate": 3.999872065705065e-05,
"loss": 0.6518,
"step": 128
},
{
"epoch": 0.5222672064777328,
"grad_norm": 1.0349367701165293,
"learning_rate": 3.999800103862945e-05,
"loss": 0.6567,
"step": 129
},
{
"epoch": 0.5263157894736842,
"grad_norm": 2.3689694964831878,
"learning_rate": 3.9997121516724834e-05,
"loss": 0.6486,
"step": 130
},
{
"epoch": 0.5303643724696356,
"grad_norm": 1.7834778524562556,
"learning_rate": 3.999608209836944e-05,
"loss": 0.6499,
"step": 131
},
{
"epoch": 0.5344129554655871,
"grad_norm": 1.732609621043433,
"learning_rate": 3.999488279187443e-05,
"loss": 0.6375,
"step": 132
},
{
"epoch": 0.5384615384615384,
"grad_norm": 1.7509433022706227,
"learning_rate": 3.9993523606829425e-05,
"loss": 0.6524,
"step": 133
},
{
"epoch": 0.5425101214574899,
"grad_norm": 1.02168826940399,
"learning_rate": 3.9992004554102435e-05,
"loss": 0.6435,
"step": 134
},
{
"epoch": 0.5465587044534413,
"grad_norm": 1.2388979341511126,
"learning_rate": 3.999032564583976e-05,
"loss": 0.6478,
"step": 135
},
{
"epoch": 0.5506072874493927,
"grad_norm": 1.1811726464849823,
"learning_rate": 3.998848689546591e-05,
"loss": 0.6373,
"step": 136
},
{
"epoch": 0.5546558704453441,
"grad_norm": 0.9207939800296402,
"learning_rate": 3.9986488317683496e-05,
"loss": 0.6402,
"step": 137
},
{
"epoch": 0.5587044534412956,
"grad_norm": 1.3156282092887874,
"learning_rate": 3.9984329928473067e-05,
"loss": 0.6225,
"step": 138
},
{
"epoch": 0.562753036437247,
"grad_norm": 0.9244871613605044,
"learning_rate": 3.998201174509306e-05,
"loss": 0.6348,
"step": 139
},
{
"epoch": 0.5668016194331984,
"grad_norm": 1.3085816261443834,
"learning_rate": 3.997953378607961e-05,
"loss": 0.6181,
"step": 140
},
{
"epoch": 0.5708502024291497,
"grad_norm": 0.9429433052196275,
"learning_rate": 3.9976896071246403e-05,
"loss": 0.6328,
"step": 141
},
{
"epoch": 0.5748987854251012,
"grad_norm": 1.3486849161439345,
"learning_rate": 3.997409862168454e-05,
"loss": 0.6469,
"step": 142
},
{
"epoch": 0.5789473684210527,
"grad_norm": 1.0662624775972946,
"learning_rate": 3.997114145976237e-05,
"loss": 0.6466,
"step": 143
},
{
"epoch": 0.582995951417004,
"grad_norm": 1.2359883100996643,
"learning_rate": 3.9968024609125245e-05,
"loss": 0.6342,
"step": 144
},
{
"epoch": 0.5870445344129555,
"grad_norm": 0.9647542817269765,
"learning_rate": 3.996474809469545e-05,
"loss": 0.6362,
"step": 145
},
{
"epoch": 0.5910931174089069,
"grad_norm": 1.0962807729576134,
"learning_rate": 3.996131194267189e-05,
"loss": 0.6236,
"step": 146
},
{
"epoch": 0.5951417004048583,
"grad_norm": 1.087591309961643,
"learning_rate": 3.995771618052995e-05,
"loss": 0.6295,
"step": 147
},
{
"epoch": 0.5991902834008097,
"grad_norm": 0.9439036674861584,
"learning_rate": 3.9953960837021256e-05,
"loss": 0.6316,
"step": 148
},
{
"epoch": 0.6032388663967612,
"grad_norm": 0.8116103416802631,
"learning_rate": 3.995004594217343e-05,
"loss": 0.6318,
"step": 149
},
{
"epoch": 0.6072874493927125,
"grad_norm": 0.8158931125637131,
"learning_rate": 3.994597152728987e-05,
"loss": 0.6201,
"step": 150
},
{
"epoch": 0.611336032388664,
"grad_norm": 0.798264574059316,
"learning_rate": 3.994173762494947e-05,
"loss": 0.6455,
"step": 151
},
{
"epoch": 0.6153846153846154,
"grad_norm": 0.8616106626660048,
"learning_rate": 3.9937344269006424e-05,
"loss": 0.6377,
"step": 152
},
{
"epoch": 0.6194331983805668,
"grad_norm": 0.8657979313596202,
"learning_rate": 3.993279149458988e-05,
"loss": 0.6273,
"step": 153
},
{
"epoch": 0.6234817813765182,
"grad_norm": 0.6291319482933733,
"learning_rate": 3.9928079338103695e-05,
"loss": 0.6254,
"step": 154
},
{
"epoch": 0.6275303643724697,
"grad_norm": 0.6997462336111963,
"learning_rate": 3.992320783722616e-05,
"loss": 0.6135,
"step": 155
},
{
"epoch": 0.631578947368421,
"grad_norm": 0.5972763858090688,
"learning_rate": 3.991817703090965e-05,
"loss": 0.6161,
"step": 156
},
{
"epoch": 0.6356275303643725,
"grad_norm": 0.7056280586407752,
"learning_rate": 3.991298695938038e-05,
"loss": 0.6346,
"step": 157
},
{
"epoch": 0.6396761133603239,
"grad_norm": 0.6226156875032449,
"learning_rate": 3.9907637664138e-05,
"loss": 0.6374,
"step": 158
},
{
"epoch": 0.6437246963562753,
"grad_norm": 0.5064874451343923,
"learning_rate": 3.990212918795534e-05,
"loss": 0.6281,
"step": 159
},
{
"epoch": 0.6477732793522267,
"grad_norm": 0.5763757669097974,
"learning_rate": 3.9896461574878036e-05,
"loss": 0.6153,
"step": 160
},
{
"epoch": 0.6518218623481782,
"grad_norm": 0.8748254820303025,
"learning_rate": 3.989063487022416e-05,
"loss": 0.6146,
"step": 161
},
{
"epoch": 0.6558704453441295,
"grad_norm": 0.8765965498687993,
"learning_rate": 3.9884649120583903e-05,
"loss": 0.6138,
"step": 162
},
{
"epoch": 0.659919028340081,
"grad_norm": 0.5343345180150284,
"learning_rate": 3.987850437381915e-05,
"loss": 0.6182,
"step": 163
},
{
"epoch": 0.6639676113360324,
"grad_norm": 0.5039935940809683,
"learning_rate": 3.987220067906313e-05,
"loss": 0.6284,
"step": 164
},
{
"epoch": 0.6680161943319838,
"grad_norm": 0.5389014685289086,
"learning_rate": 3.986573808672002e-05,
"loss": 0.6229,
"step": 165
},
{
"epoch": 0.6720647773279352,
"grad_norm": 0.4739872485155583,
"learning_rate": 3.9859116648464534e-05,
"loss": 0.6123,
"step": 166
},
{
"epoch": 0.6761133603238867,
"grad_norm": 0.5564745171120434,
"learning_rate": 3.985233641724152e-05,
"loss": 0.6158,
"step": 167
},
{
"epoch": 0.680161943319838,
"grad_norm": 0.5011815818842251,
"learning_rate": 3.984539744726554e-05,
"loss": 0.6217,
"step": 168
},
{
"epoch": 0.6842105263157895,
"grad_norm": 0.4508124631154391,
"learning_rate": 3.983829979402038e-05,
"loss": 0.6289,
"step": 169
},
{
"epoch": 0.6882591093117408,
"grad_norm": 0.4330432939720494,
"learning_rate": 3.983104351425871e-05,
"loss": 0.637,
"step": 170
},
{
"epoch": 0.6923076923076923,
"grad_norm": 0.502800604372197,
"learning_rate": 3.982362866600154e-05,
"loss": 0.6136,
"step": 171
},
{
"epoch": 0.6963562753036437,
"grad_norm": 0.4813906280941234,
"learning_rate": 3.9816055308537806e-05,
"loss": 0.6089,
"step": 172
},
{
"epoch": 0.7004048582995951,
"grad_norm": 0.3877454461851501,
"learning_rate": 3.980832350242388e-05,
"loss": 0.6066,
"step": 173
},
{
"epoch": 0.7044534412955465,
"grad_norm": 0.5332669000150432,
"learning_rate": 3.980043330948306e-05,
"loss": 0.6232,
"step": 174
},
{
"epoch": 0.708502024291498,
"grad_norm": 0.4761153395672825,
"learning_rate": 3.9792384792805124e-05,
"loss": 0.6366,
"step": 175
},
{
"epoch": 0.7125506072874493,
"grad_norm": 0.4455439259101304,
"learning_rate": 3.978417801674579e-05,
"loss": 0.6179,
"step": 176
},
{
"epoch": 0.7165991902834008,
"grad_norm": 0.5668735423124472,
"learning_rate": 3.9775813046926244e-05,
"loss": 0.6245,
"step": 177
},
{
"epoch": 0.7206477732793523,
"grad_norm": 0.6227384298636165,
"learning_rate": 3.976728995023254e-05,
"loss": 0.6231,
"step": 178
},
{
"epoch": 0.7246963562753036,
"grad_norm": 0.7708879316590238,
"learning_rate": 3.975860879481514e-05,
"loss": 0.6108,
"step": 179
},
{
"epoch": 0.728744939271255,
"grad_norm": 0.7315265283045214,
"learning_rate": 3.974976965008833e-05,
"loss": 0.62,
"step": 180
},
{
"epoch": 0.7327935222672065,
"grad_norm": 0.6108000493103951,
"learning_rate": 3.9740772586729674e-05,
"loss": 0.6224,
"step": 181
},
{
"epoch": 0.7368421052631579,
"grad_norm": 0.5663331038280554,
"learning_rate": 3.973161767667946e-05,
"loss": 0.6174,
"step": 182
},
{
"epoch": 0.7408906882591093,
"grad_norm": 0.5599760852171393,
"learning_rate": 3.972230499314009e-05,
"loss": 0.6092,
"step": 183
},
{
"epoch": 0.7449392712550608,
"grad_norm": 0.42337862325586745,
"learning_rate": 3.9712834610575545e-05,
"loss": 0.6157,
"step": 184
},
{
"epoch": 0.7489878542510121,
"grad_norm": 0.5058339288170007,
"learning_rate": 3.9703206604710746e-05,
"loss": 0.6222,
"step": 185
},
{
"epoch": 0.7530364372469636,
"grad_norm": 0.5943333221517856,
"learning_rate": 3.969342105253098e-05,
"loss": 0.6315,
"step": 186
},
{
"epoch": 0.757085020242915,
"grad_norm": 0.5300319583645045,
"learning_rate": 3.9683478032281246e-05,
"loss": 0.6267,
"step": 187
},
{
"epoch": 0.7611336032388664,
"grad_norm": 0.4379695907799154,
"learning_rate": 3.967337762346568e-05,
"loss": 0.6154,
"step": 188
},
{
"epoch": 0.7651821862348178,
"grad_norm": 0.5591253733065618,
"learning_rate": 3.9663119906846885e-05,
"loss": 0.6104,
"step": 189
},
{
"epoch": 0.7692307692307693,
"grad_norm": 0.7136980908085325,
"learning_rate": 3.965270496444528e-05,
"loss": 0.6164,
"step": 190
},
{
"epoch": 0.7732793522267206,
"grad_norm": 0.5446047572820258,
"learning_rate": 3.9642132879538465e-05,
"loss": 0.6195,
"step": 191
},
{
"epoch": 0.7773279352226721,
"grad_norm": 0.48796633853444793,
"learning_rate": 3.9631403736660566e-05,
"loss": 0.6057,
"step": 192
},
{
"epoch": 0.7813765182186235,
"grad_norm": 0.6059125830150097,
"learning_rate": 3.962051762160151e-05,
"loss": 0.6181,
"step": 193
},
{
"epoch": 0.7854251012145749,
"grad_norm": 0.6023795767757892,
"learning_rate": 3.960947462140638e-05,
"loss": 0.6159,
"step": 194
},
{
"epoch": 0.7894736842105263,
"grad_norm": 0.5686237057784563,
"learning_rate": 3.959827482437472e-05,
"loss": 0.6229,
"step": 195
},
{
"epoch": 0.7935222672064778,
"grad_norm": 0.5179516367925641,
"learning_rate": 3.958691832005981e-05,
"loss": 0.604,
"step": 196
},
{
"epoch": 0.7975708502024291,
"grad_norm": 0.6076631120888005,
"learning_rate": 3.9575405199267944e-05,
"loss": 0.621,
"step": 197
},
{
"epoch": 0.8016194331983806,
"grad_norm": 0.7407824941143147,
"learning_rate": 3.956373555405774e-05,
"loss": 0.6089,
"step": 198
},
{
"epoch": 0.805668016194332,
"grad_norm": 0.7362797393874368,
"learning_rate": 3.955190947773937e-05,
"loss": 0.6152,
"step": 199
},
{
"epoch": 0.8097165991902834,
"grad_norm": 0.5546796829295568,
"learning_rate": 3.9539927064873824e-05,
"loss": 0.6072,
"step": 200
},
{
"epoch": 0.8137651821862348,
"grad_norm": 0.5363827183346322,
"learning_rate": 3.952778841127214e-05,
"loss": 0.6163,
"step": 201
},
{
"epoch": 0.8178137651821862,
"grad_norm": 0.5630574021256145,
"learning_rate": 3.95154936139947e-05,
"loss": 0.6027,
"step": 202
},
{
"epoch": 0.8218623481781376,
"grad_norm": 0.5730916678474456,
"learning_rate": 3.950304277135034e-05,
"loss": 0.6093,
"step": 203
},
{
"epoch": 0.8259109311740891,
"grad_norm": 0.5664379295522922,
"learning_rate": 3.94904359828957e-05,
"loss": 0.6118,
"step": 204
},
{
"epoch": 0.8299595141700404,
"grad_norm": 0.6952069941615893,
"learning_rate": 3.94776733494343e-05,
"loss": 0.6089,
"step": 205
},
{
"epoch": 0.8340080971659919,
"grad_norm": 0.5616229758171732,
"learning_rate": 3.9464754973015826e-05,
"loss": 0.6049,
"step": 206
},
{
"epoch": 0.8380566801619433,
"grad_norm": 0.7126339834366604,
"learning_rate": 3.9451680956935275e-05,
"loss": 0.6254,
"step": 207
},
{
"epoch": 0.8421052631578947,
"grad_norm": 0.952172740653009,
"learning_rate": 3.943845140573213e-05,
"loss": 0.608,
"step": 208
},
{
"epoch": 0.8461538461538461,
"grad_norm": 0.8635846790779812,
"learning_rate": 3.942506642518952e-05,
"loss": 0.613,
"step": 209
},
{
"epoch": 0.8502024291497976,
"grad_norm": 0.45810969744886737,
"learning_rate": 3.9411526122333396e-05,
"loss": 0.6078,
"step": 210
},
{
"epoch": 0.854251012145749,
"grad_norm": 0.7146902268235953,
"learning_rate": 3.939783060543167e-05,
"loss": 0.6002,
"step": 211
},
{
"epoch": 0.8582995951417004,
"grad_norm": 0.5691390876968877,
"learning_rate": 3.938397998399333e-05,
"loss": 0.6148,
"step": 212
},
{
"epoch": 0.8623481781376519,
"grad_norm": 0.382510968913962,
"learning_rate": 3.936997436876756e-05,
"loss": 0.6125,
"step": 213
},
{
"epoch": 0.8663967611336032,
"grad_norm": 0.5839219064529624,
"learning_rate": 3.935581387174289e-05,
"loss": 0.6248,
"step": 214
},
{
"epoch": 0.8704453441295547,
"grad_norm": 0.5182975118175669,
"learning_rate": 3.9341498606146295e-05,
"loss": 0.6029,
"step": 215
},
{
"epoch": 0.8744939271255061,
"grad_norm": 0.47346307599905607,
"learning_rate": 3.932702868644224e-05,
"loss": 0.5987,
"step": 216
},
{
"epoch": 0.8785425101214575,
"grad_norm": 0.4230067559569684,
"learning_rate": 3.931240422833181e-05,
"loss": 0.6011,
"step": 217
},
{
"epoch": 0.8825910931174089,
"grad_norm": 0.5712661837060536,
"learning_rate": 3.9297625348751806e-05,
"loss": 0.6085,
"step": 218
},
{
"epoch": 0.8866396761133604,
"grad_norm": 0.6250610276226568,
"learning_rate": 3.9282692165873744e-05,
"loss": 0.6108,
"step": 219
},
{
"epoch": 0.8906882591093117,
"grad_norm": 0.5620853355396411,
"learning_rate": 3.9267604799102945e-05,
"loss": 0.6174,
"step": 220
},
{
"epoch": 0.8947368421052632,
"grad_norm": 0.5787212604197847,
"learning_rate": 3.9252363369077615e-05,
"loss": 0.5967,
"step": 221
},
{
"epoch": 0.8987854251012146,
"grad_norm": 0.553697099751073,
"learning_rate": 3.9236967997667796e-05,
"loss": 0.5984,
"step": 222
},
{
"epoch": 0.902834008097166,
"grad_norm": 0.4569508508642684,
"learning_rate": 3.9221418807974496e-05,
"loss": 0.5983,
"step": 223
},
{
"epoch": 0.9068825910931174,
"grad_norm": 0.4560347379685896,
"learning_rate": 3.92057159243286e-05,
"loss": 0.6044,
"step": 224
},
{
"epoch": 0.9109311740890689,
"grad_norm": 0.4741896579486363,
"learning_rate": 3.9189859472289956e-05,
"loss": 0.5954,
"step": 225
},
{
"epoch": 0.9149797570850202,
"grad_norm": 0.4714472073611277,
"learning_rate": 3.9173849578646324e-05,
"loss": 0.5885,
"step": 226
},
{
"epoch": 0.9190283400809717,
"grad_norm": 0.5677671607771182,
"learning_rate": 3.9157686371412395e-05,
"loss": 0.6061,
"step": 227
},
{
"epoch": 0.9230769230769231,
"grad_norm": 0.3937124901766537,
"learning_rate": 3.914136997982874e-05,
"loss": 0.6038,
"step": 228
},
{
"epoch": 0.9271255060728745,
"grad_norm": 0.5545257082255058,
"learning_rate": 3.912490053436079e-05,
"loss": 0.6009,
"step": 229
},
{
"epoch": 0.9311740890688259,
"grad_norm": 0.4452844004523292,
"learning_rate": 3.9108278166697805e-05,
"loss": 0.5991,
"step": 230
},
{
"epoch": 0.9352226720647774,
"grad_norm": 0.4939377192207064,
"learning_rate": 3.909150300975178e-05,
"loss": 0.6059,
"step": 231
},
{
"epoch": 0.9392712550607287,
"grad_norm": 0.5555803580765023,
"learning_rate": 3.907457519765642e-05,
"loss": 0.6075,
"step": 232
},
{
"epoch": 0.9433198380566802,
"grad_norm": 0.5416403520108772,
"learning_rate": 3.905749486576607e-05,
"loss": 0.6178,
"step": 233
},
{
"epoch": 0.9473684210526315,
"grad_norm": 0.5210376597123697,
"learning_rate": 3.90402621506546e-05,
"loss": 0.592,
"step": 234
},
{
"epoch": 0.951417004048583,
"grad_norm": 0.4234310339822861,
"learning_rate": 3.902287719011435e-05,
"loss": 0.6176,
"step": 235
},
{
"epoch": 0.9554655870445344,
"grad_norm": 0.4174429430110697,
"learning_rate": 3.900534012315499e-05,
"loss": 0.5927,
"step": 236
},
{
"epoch": 0.9595141700404858,
"grad_norm": 0.5869177030063298,
"learning_rate": 3.8987651090002454e-05,
"loss": 0.6041,
"step": 237
},
{
"epoch": 0.9635627530364372,
"grad_norm": 0.48409888902689957,
"learning_rate": 3.896981023209777e-05,
"loss": 0.6127,
"step": 238
},
{
"epoch": 0.9676113360323887,
"grad_norm": 0.4468454676981497,
"learning_rate": 3.8951817692095966e-05,
"loss": 0.6139,
"step": 239
},
{
"epoch": 0.97165991902834,
"grad_norm": 0.41204130453916604,
"learning_rate": 3.893367361386491e-05,
"loss": 0.6003,
"step": 240
},
{
"epoch": 0.9757085020242915,
"grad_norm": 0.46310502937242665,
"learning_rate": 3.891537814248417e-05,
"loss": 0.6071,
"step": 241
},
{
"epoch": 0.979757085020243,
"grad_norm": 0.5065797970857927,
"learning_rate": 3.889693142424384e-05,
"loss": 0.5973,
"step": 242
},
{
"epoch": 0.9838056680161943,
"grad_norm": 0.5094440619361806,
"learning_rate": 3.8878333606643405e-05,
"loss": 0.5969,
"step": 243
},
{
"epoch": 0.9878542510121457,
"grad_norm": 0.34855118982935424,
"learning_rate": 3.885958483839049e-05,
"loss": 0.5971,
"step": 244
},
{
"epoch": 0.9919028340080972,
"grad_norm": 0.5169330056337861,
"learning_rate": 3.8840685269399786e-05,
"loss": 0.6109,
"step": 245
},
{
"epoch": 0.9959514170040485,
"grad_norm": 0.4699783965726958,
"learning_rate": 3.882163505079171e-05,
"loss": 0.6063,
"step": 246
},
{
"epoch": 1.0,
"grad_norm": 0.5018257828396283,
"learning_rate": 3.880243433489132e-05,
"loss": 0.6239,
"step": 247
},
{
"epoch": 1.0040485829959513,
"grad_norm": 0.6034542224341307,
"learning_rate": 3.878308327522702e-05,
"loss": 0.5743,
"step": 248
},
{
"epoch": 1.008097165991903,
"grad_norm": 0.4691474489017611,
"learning_rate": 3.876358202652936e-05,
"loss": 0.5547,
"step": 249
},
{
"epoch": 1.0121457489878543,
"grad_norm": 0.4844453271889736,
"learning_rate": 3.8743930744729835e-05,
"loss": 0.5581,
"step": 250
},
{
"epoch": 1.0161943319838056,
"grad_norm": 0.42608075012476415,
"learning_rate": 3.8724129586959535e-05,
"loss": 0.562,
"step": 251
},
{
"epoch": 1.0202429149797572,
"grad_norm": 0.49254021406074305,
"learning_rate": 3.870417871154801e-05,
"loss": 0.5532,
"step": 252
},
{
"epoch": 1.0242914979757085,
"grad_norm": 0.43418891168060825,
"learning_rate": 3.868407827802192e-05,
"loss": 0.552,
"step": 253
},
{
"epoch": 1.0283400809716599,
"grad_norm": 0.518695323867836,
"learning_rate": 3.8663828447103804e-05,
"loss": 0.5585,
"step": 254
},
{
"epoch": 1.0323886639676114,
"grad_norm": 0.5388391803678748,
"learning_rate": 3.8643429380710765e-05,
"loss": 0.5626,
"step": 255
},
{
"epoch": 1.0364372469635628,
"grad_norm": 0.6308691941371626,
"learning_rate": 3.862288124195319e-05,
"loss": 0.5534,
"step": 256
},
{
"epoch": 1.040485829959514,
"grad_norm": 0.7366117149058692,
"learning_rate": 3.8602184195133464e-05,
"loss": 0.5614,
"step": 257
},
{
"epoch": 1.0445344129554657,
"grad_norm": 0.7532849075824768,
"learning_rate": 3.8581338405744604e-05,
"loss": 0.5596,
"step": 258
},
{
"epoch": 1.048582995951417,
"grad_norm": 0.7103796677216142,
"learning_rate": 3.8560344040468996e-05,
"loss": 0.5446,
"step": 259
},
{
"epoch": 1.0526315789473684,
"grad_norm": 0.5848213470636262,
"learning_rate": 3.853920126717703e-05,
"loss": 0.5577,
"step": 260
},
{
"epoch": 1.05668016194332,
"grad_norm": 0.44560537621090507,
"learning_rate": 3.851791025492574e-05,
"loss": 0.5495,
"step": 261
},
{
"epoch": 1.0607287449392713,
"grad_norm": 0.41056905207284805,
"learning_rate": 3.8496471173957515e-05,
"loss": 0.5634,
"step": 262
},
{
"epoch": 1.0647773279352226,
"grad_norm": 0.5522807262012146,
"learning_rate": 3.847488419569866e-05,
"loss": 0.5666,
"step": 263
},
{
"epoch": 1.0688259109311742,
"grad_norm": 0.7205722464649814,
"learning_rate": 3.8453149492758077e-05,
"loss": 0.5369,
"step": 264
},
{
"epoch": 1.0728744939271255,
"grad_norm": 0.6542343658827969,
"learning_rate": 3.8431267238925886e-05,
"loss": 0.5524,
"step": 265
},
{
"epoch": 1.0769230769230769,
"grad_norm": 0.4310594762946326,
"learning_rate": 3.840923760917199e-05,
"loss": 0.5638,
"step": 266
},
{
"epoch": 1.0809716599190284,
"grad_norm": 0.4751713785030467,
"learning_rate": 3.838706077964473e-05,
"loss": 0.5435,
"step": 267
},
{
"epoch": 1.0850202429149798,
"grad_norm": 0.5193866478420406,
"learning_rate": 3.836473692766944e-05,
"loss": 0.5597,
"step": 268
},
{
"epoch": 1.0890688259109311,
"grad_norm": 0.6163320695643267,
"learning_rate": 3.834226623174705e-05,
"loss": 0.5488,
"step": 269
},
{
"epoch": 1.0931174089068827,
"grad_norm": 0.5346589956597334,
"learning_rate": 3.831964887155264e-05,
"loss": 0.5622,
"step": 270
},
{
"epoch": 1.097165991902834,
"grad_norm": 0.41123781009599275,
"learning_rate": 3.829688502793403e-05,
"loss": 0.5459,
"step": 271
},
{
"epoch": 1.1012145748987854,
"grad_norm": 0.39114679670054636,
"learning_rate": 3.827397488291029e-05,
"loss": 0.5622,
"step": 272
},
{
"epoch": 1.1052631578947367,
"grad_norm": 0.37458670677357453,
"learning_rate": 3.8250918619670343e-05,
"loss": 0.5561,
"step": 273
},
{
"epoch": 1.1093117408906883,
"grad_norm": 0.4118775319796883,
"learning_rate": 3.822771642257145e-05,
"loss": 0.5419,
"step": 274
},
{
"epoch": 1.1133603238866396,
"grad_norm": 0.4107321274967381,
"learning_rate": 3.820436847713776e-05,
"loss": 0.5527,
"step": 275
},
{
"epoch": 1.117408906882591,
"grad_norm": 0.4673286508850098,
"learning_rate": 3.8180874970058826e-05,
"loss": 0.554,
"step": 276
},
{
"epoch": 1.1214574898785425,
"grad_norm": 0.47063621288005403,
"learning_rate": 3.8157236089188103e-05,
"loss": 0.5459,
"step": 277
},
{
"epoch": 1.125506072874494,
"grad_norm": 0.5729452925870199,
"learning_rate": 3.813345202354145e-05,
"loss": 0.5579,
"step": 278
},
{
"epoch": 1.1295546558704452,
"grad_norm": 0.3937226895248115,
"learning_rate": 3.810952296329563e-05,
"loss": 0.5522,
"step": 279
},
{
"epoch": 1.1336032388663968,
"grad_norm": 0.4526482406794656,
"learning_rate": 3.808544909978676e-05,
"loss": 0.5478,
"step": 280
},
{
"epoch": 1.1376518218623481,
"grad_norm": 0.41183529039279504,
"learning_rate": 3.806123062550882e-05,
"loss": 0.5557,
"step": 281
},
{
"epoch": 1.1417004048582995,
"grad_norm": 0.43765609134271727,
"learning_rate": 3.803686773411208e-05,
"loss": 0.5419,
"step": 282
},
{
"epoch": 1.145748987854251,
"grad_norm": 0.45090739999832186,
"learning_rate": 3.801236062040158e-05,
"loss": 0.5528,
"step": 283
},
{
"epoch": 1.1497975708502024,
"grad_norm": 0.36725658071253475,
"learning_rate": 3.798770948033553e-05,
"loss": 0.5416,
"step": 284
},
{
"epoch": 1.1538461538461537,
"grad_norm": 0.49355116511869096,
"learning_rate": 3.796291451102381e-05,
"loss": 0.5592,
"step": 285
},
{
"epoch": 1.1578947368421053,
"grad_norm": 0.5287915889651591,
"learning_rate": 3.793797591072631e-05,
"loss": 0.5601,
"step": 286
},
{
"epoch": 1.1619433198380567,
"grad_norm": 0.40532720619500945,
"learning_rate": 3.791289387885144e-05,
"loss": 0.5625,
"step": 287
},
{
"epoch": 1.165991902834008,
"grad_norm": 0.4909527546974886,
"learning_rate": 3.788766861595443e-05,
"loss": 0.5555,
"step": 288
},
{
"epoch": 1.1700404858299596,
"grad_norm": 0.5422292778483524,
"learning_rate": 3.7862300323735835e-05,
"loss": 0.5554,
"step": 289
},
{
"epoch": 1.174089068825911,
"grad_norm": 0.4458828136374878,
"learning_rate": 3.783678920503983e-05,
"loss": 0.543,
"step": 290
},
{
"epoch": 1.1781376518218623,
"grad_norm": 0.4372519997235517,
"learning_rate": 3.781113546385265e-05,
"loss": 0.5411,
"step": 291
},
{
"epoch": 1.1821862348178138,
"grad_norm": 0.48167225351560505,
"learning_rate": 3.7785339305300916e-05,
"loss": 0.545,
"step": 292
},
{
"epoch": 1.1862348178137652,
"grad_norm": 0.39962450102883645,
"learning_rate": 3.7759400935650025e-05,
"loss": 0.5572,
"step": 293
},
{
"epoch": 1.1902834008097165,
"grad_norm": 0.4093948258708819,
"learning_rate": 3.77333205623025e-05,
"loss": 0.5709,
"step": 294
},
{
"epoch": 1.194331983805668,
"grad_norm": 0.5192056111615209,
"learning_rate": 3.7707098393796295e-05,
"loss": 0.5513,
"step": 295
},
{
"epoch": 1.1983805668016194,
"grad_norm": 0.4625822218508941,
"learning_rate": 3.7680734639803185e-05,
"loss": 0.5406,
"step": 296
},
{
"epoch": 1.2024291497975708,
"grad_norm": 0.44201331869128746,
"learning_rate": 3.765422951112704e-05,
"loss": 0.5492,
"step": 297
},
{
"epoch": 1.2064777327935223,
"grad_norm": 0.5380285993282039,
"learning_rate": 3.762758321970216e-05,
"loss": 0.5573,
"step": 298
},
{
"epoch": 1.2105263157894737,
"grad_norm": 0.42291389867534407,
"learning_rate": 3.7600795978591584e-05,
"loss": 0.5599,
"step": 299
},
{
"epoch": 1.214574898785425,
"grad_norm": 0.40395765378297416,
"learning_rate": 3.757386800198538e-05,
"loss": 0.5517,
"step": 300
},
{
"epoch": 1.2186234817813766,
"grad_norm": 0.4716326767078852,
"learning_rate": 3.7546799505198925e-05,
"loss": 0.5493,
"step": 301
},
{
"epoch": 1.222672064777328,
"grad_norm": 0.45923091850647735,
"learning_rate": 3.7519590704671197e-05,
"loss": 0.561,
"step": 302
},
{
"epoch": 1.2267206477732793,
"grad_norm": 0.3852248143824019,
"learning_rate": 3.749224181796305e-05,
"loss": 0.567,
"step": 303
},
{
"epoch": 1.2307692307692308,
"grad_norm": 0.4786680761954973,
"learning_rate": 3.746475306375545e-05,
"loss": 0.5598,
"step": 304
},
{
"epoch": 1.2348178137651822,
"grad_norm": 0.5315660460620871,
"learning_rate": 3.743712466184774e-05,
"loss": 0.5577,
"step": 305
},
{
"epoch": 1.2388663967611335,
"grad_norm": 0.3829534443251548,
"learning_rate": 3.7409356833155885e-05,
"loss": 0.551,
"step": 306
},
{
"epoch": 1.242914979757085,
"grad_norm": 0.4391623300682866,
"learning_rate": 3.73814497997107e-05,
"loss": 0.5371,
"step": 307
},
{
"epoch": 1.2469635627530364,
"grad_norm": 0.4545230661569007,
"learning_rate": 3.7353403784656095e-05,
"loss": 0.5462,
"step": 308
},
{
"epoch": 1.2510121457489878,
"grad_norm": 0.37106817866503555,
"learning_rate": 3.732521901224724e-05,
"loss": 0.5522,
"step": 309
},
{
"epoch": 1.2550607287449393,
"grad_norm": 0.41999002292327886,
"learning_rate": 3.7296895707848844e-05,
"loss": 0.5428,
"step": 310
},
{
"epoch": 1.2591093117408907,
"grad_norm": 0.37781453933873016,
"learning_rate": 3.7268434097933275e-05,
"loss": 0.5435,
"step": 311
},
{
"epoch": 1.263157894736842,
"grad_norm": 0.36993697187587243,
"learning_rate": 3.723983441007882e-05,
"loss": 0.5439,
"step": 312
},
{
"epoch": 1.2672064777327936,
"grad_norm": 0.44650765298745726,
"learning_rate": 3.721109687296781e-05,
"loss": 0.552,
"step": 313
},
{
"epoch": 1.271255060728745,
"grad_norm": 0.4946112709982116,
"learning_rate": 3.718222171638484e-05,
"loss": 0.5603,
"step": 314
},
{
"epoch": 1.2753036437246963,
"grad_norm": 0.468665055171741,
"learning_rate": 3.715320917121488e-05,
"loss": 0.5554,
"step": 315
},
{
"epoch": 1.2793522267206479,
"grad_norm": 0.4789696868873683,
"learning_rate": 3.712405946944149e-05,
"loss": 0.5494,
"step": 316
},
{
"epoch": 1.2834008097165992,
"grad_norm": 0.3603446803072295,
"learning_rate": 3.709477284414488e-05,
"loss": 0.5596,
"step": 317
},
{
"epoch": 1.2874493927125505,
"grad_norm": 0.4658165558886022,
"learning_rate": 3.7065349529500136e-05,
"loss": 0.5575,
"step": 318
},
{
"epoch": 1.291497975708502,
"grad_norm": 0.48174647090781725,
"learning_rate": 3.7035789760775305e-05,
"loss": 0.5418,
"step": 319
},
{
"epoch": 1.2955465587044535,
"grad_norm": 0.5044754584921156,
"learning_rate": 3.70060937743295e-05,
"loss": 0.5504,
"step": 320
},
{
"epoch": 1.2995951417004048,
"grad_norm": 0.5401473801865555,
"learning_rate": 3.6976261807611045e-05,
"loss": 0.5499,
"step": 321
},
{
"epoch": 1.3036437246963564,
"grad_norm": 0.49046036545627136,
"learning_rate": 3.694629409915555e-05,
"loss": 0.5456,
"step": 322
},
{
"epoch": 1.3076923076923077,
"grad_norm": 0.403866628435754,
"learning_rate": 3.691619088858401e-05,
"loss": 0.5618,
"step": 323
},
{
"epoch": 1.311740890688259,
"grad_norm": 0.45205860746270204,
"learning_rate": 3.68859524166009e-05,
"loss": 0.5633,
"step": 324
},
{
"epoch": 1.3157894736842106,
"grad_norm": 0.5298439820084963,
"learning_rate": 3.6855578924992226e-05,
"loss": 0.552,
"step": 325
},
{
"epoch": 1.319838056680162,
"grad_norm": 0.5394536034115333,
"learning_rate": 3.6825070656623626e-05,
"loss": 0.5404,
"step": 326
},
{
"epoch": 1.3238866396761133,
"grad_norm": 0.5070561809242364,
"learning_rate": 3.67944278554384e-05,
"loss": 0.5483,
"step": 327
},
{
"epoch": 1.3279352226720649,
"grad_norm": 0.47239051947461025,
"learning_rate": 3.676365076645557e-05,
"loss": 0.5617,
"step": 328
},
{
"epoch": 1.3319838056680162,
"grad_norm": 0.5192915843551382,
"learning_rate": 3.673273963576791e-05,
"loss": 0.5429,
"step": 329
},
{
"epoch": 1.3360323886639676,
"grad_norm": 0.4271927521645054,
"learning_rate": 3.670169471054001e-05,
"loss": 0.554,
"step": 330
},
{
"epoch": 1.3400809716599191,
"grad_norm": 0.517404021243629,
"learning_rate": 3.6670516239006254e-05,
"loss": 0.5477,
"step": 331
},
{
"epoch": 1.3441295546558705,
"grad_norm": 0.4946081809670534,
"learning_rate": 3.663920447046888e-05,
"loss": 0.5491,
"step": 332
},
{
"epoch": 1.3481781376518218,
"grad_norm": 0.41944016094746467,
"learning_rate": 3.6607759655295954e-05,
"loss": 0.5666,
"step": 333
},
{
"epoch": 1.3522267206477734,
"grad_norm": 0.46085470734550854,
"learning_rate": 3.657618204491937e-05,
"loss": 0.5559,
"step": 334
},
{
"epoch": 1.3562753036437247,
"grad_norm": 0.4708314849336806,
"learning_rate": 3.6544471891832875e-05,
"loss": 0.5533,
"step": 335
},
{
"epoch": 1.360323886639676,
"grad_norm": 0.4581915248887918,
"learning_rate": 3.6512629449589983e-05,
"loss": 0.5444,
"step": 336
},
{
"epoch": 1.3643724696356276,
"grad_norm": 0.5600946635069136,
"learning_rate": 3.6480654972802045e-05,
"loss": 0.552,
"step": 337
},
{
"epoch": 1.368421052631579,
"grad_norm": 0.653667303127481,
"learning_rate": 3.644854871713611e-05,
"loss": 0.5514,
"step": 338
},
{
"epoch": 1.3724696356275303,
"grad_norm": 0.45046980728148284,
"learning_rate": 3.6416310939312924e-05,
"loss": 0.5386,
"step": 339
},
{
"epoch": 1.376518218623482,
"grad_norm": 0.6862310051866388,
"learning_rate": 3.638394189710493e-05,
"loss": 0.5495,
"step": 340
},
{
"epoch": 1.3805668016194332,
"grad_norm": 0.5754798747207687,
"learning_rate": 3.635144184933412e-05,
"loss": 0.5628,
"step": 341
},
{
"epoch": 1.3846153846153846,
"grad_norm": 0.5532717910812307,
"learning_rate": 3.6318811055869994e-05,
"loss": 0.5675,
"step": 342
},
{
"epoch": 1.3886639676113361,
"grad_norm": 0.6227490127054163,
"learning_rate": 3.628604977762752e-05,
"loss": 0.5504,
"step": 343
},
{
"epoch": 1.3927125506072875,
"grad_norm": 0.43605984214145255,
"learning_rate": 3.6253158276565004e-05,
"loss": 0.5399,
"step": 344
},
{
"epoch": 1.3967611336032388,
"grad_norm": 0.4422018980891201,
"learning_rate": 3.6220136815682016e-05,
"loss": 0.5461,
"step": 345
},
{
"epoch": 1.4008097165991904,
"grad_norm": 0.5436423277822059,
"learning_rate": 3.618698565901727e-05,
"loss": 0.5516,
"step": 346
},
{
"epoch": 1.4048582995951417,
"grad_norm": 0.4864102094930917,
"learning_rate": 3.615370507164654e-05,
"loss": 0.5493,
"step": 347
},
{
"epoch": 1.408906882591093,
"grad_norm": 0.5471483800278454,
"learning_rate": 3.61202953196805e-05,
"loss": 0.5598,
"step": 348
},
{
"epoch": 1.4129554655870447,
"grad_norm": 0.7321021644234185,
"learning_rate": 3.6086756670262654e-05,
"loss": 0.5634,
"step": 349
},
{
"epoch": 1.417004048582996,
"grad_norm": 0.5188154534783666,
"learning_rate": 3.6053089391567135e-05,
"loss": 0.553,
"step": 350
},
{
"epoch": 1.4210526315789473,
"grad_norm": 0.5008938007718013,
"learning_rate": 3.601929375279662e-05,
"loss": 0.5525,
"step": 351
},
{
"epoch": 1.425101214574899,
"grad_norm": 0.5894298130855407,
"learning_rate": 3.598537002418012e-05,
"loss": 0.5522,
"step": 352
},
{
"epoch": 1.4291497975708503,
"grad_norm": 0.3540316856720585,
"learning_rate": 3.5951318476970885e-05,
"loss": 0.5472,
"step": 353
},
{
"epoch": 1.4331983805668016,
"grad_norm": 0.4358928187714944,
"learning_rate": 3.591713938344416e-05,
"loss": 0.5388,
"step": 354
},
{
"epoch": 1.4372469635627532,
"grad_norm": 0.3703307185085029,
"learning_rate": 3.588283301689507e-05,
"loss": 0.5532,
"step": 355
},
{
"epoch": 1.4412955465587045,
"grad_norm": 0.36507962145178297,
"learning_rate": 3.5848399651636424e-05,
"loss": 0.5539,
"step": 356
},
{
"epoch": 1.4453441295546559,
"grad_norm": 0.40247994158722167,
"learning_rate": 3.5813839562996473e-05,
"loss": 0.5474,
"step": 357
},
{
"epoch": 1.4493927125506074,
"grad_norm": 0.43075077103397547,
"learning_rate": 3.5779153027316775e-05,
"loss": 0.5378,
"step": 358
},
{
"epoch": 1.4534412955465588,
"grad_norm": 0.3559300809304106,
"learning_rate": 3.574434032194994e-05,
"loss": 0.5494,
"step": 359
},
{
"epoch": 1.45748987854251,
"grad_norm": 0.435987464253013,
"learning_rate": 3.5709401725257436e-05,
"loss": 0.5581,
"step": 360
},
{
"epoch": 1.4615384615384617,
"grad_norm": 0.46760442026904464,
"learning_rate": 3.5674337516607346e-05,
"loss": 0.5442,
"step": 361
},
{
"epoch": 1.465587044534413,
"grad_norm": 0.49183620878947026,
"learning_rate": 3.563914797637216e-05,
"loss": 0.5506,
"step": 362
},
{
"epoch": 1.4696356275303644,
"grad_norm": 0.6443822464795399,
"learning_rate": 3.560383338592649e-05,
"loss": 0.548,
"step": 363
},
{
"epoch": 1.4736842105263157,
"grad_norm": 0.47791966892882753,
"learning_rate": 3.556839402764487e-05,
"loss": 0.5434,
"step": 364
},
{
"epoch": 1.4777327935222673,
"grad_norm": 0.3413733526358684,
"learning_rate": 3.553283018489946e-05,
"loss": 0.5416,
"step": 365
},
{
"epoch": 1.4817813765182186,
"grad_norm": 0.4700090995751656,
"learning_rate": 3.54971421420578e-05,
"loss": 0.5553,
"step": 366
},
{
"epoch": 1.48582995951417,
"grad_norm": 0.3976485981986971,
"learning_rate": 3.5461330184480524e-05,
"loss": 0.5495,
"step": 367
},
{
"epoch": 1.4898785425101215,
"grad_norm": 0.3354570340739303,
"learning_rate": 3.54253945985191e-05,
"loss": 0.5459,
"step": 368
},
{
"epoch": 1.4939271255060729,
"grad_norm": 0.44105773749504856,
"learning_rate": 3.5389335671513524e-05,
"loss": 0.5492,
"step": 369
},
{
"epoch": 1.4979757085020242,
"grad_norm": 0.48811520480488574,
"learning_rate": 3.5353153691789986e-05,
"loss": 0.5562,
"step": 370
},
{
"epoch": 1.5020242914979756,
"grad_norm": 0.31427980341630823,
"learning_rate": 3.531684894865866e-05,
"loss": 0.5509,
"step": 371
},
{
"epoch": 1.5060728744939271,
"grad_norm": 0.3115464079471531,
"learning_rate": 3.528042173241128e-05,
"loss": 0.5491,
"step": 372
},
{
"epoch": 1.5101214574898787,
"grad_norm": 0.4123737021213841,
"learning_rate": 3.524387233431892e-05,
"loss": 0.5573,
"step": 373
},
{
"epoch": 1.5141700404858298,
"grad_norm": 0.47301077316338624,
"learning_rate": 3.520720104662958e-05,
"loss": 0.5505,
"step": 374
},
{
"epoch": 1.5182186234817814,
"grad_norm": 0.42767013777046636,
"learning_rate": 3.5170408162565904e-05,
"loss": 0.5452,
"step": 375
},
{
"epoch": 1.522267206477733,
"grad_norm": 0.3542520969791065,
"learning_rate": 3.513349397632281e-05,
"loss": 0.5611,
"step": 376
},
{
"epoch": 1.526315789473684,
"grad_norm": 0.3667483417207753,
"learning_rate": 3.5096458783065145e-05,
"loss": 0.5497,
"step": 377
},
{
"epoch": 1.5303643724696356,
"grad_norm": 0.3223774972007158,
"learning_rate": 3.505930287892533e-05,
"loss": 0.5603,
"step": 378
},
{
"epoch": 1.5344129554655872,
"grad_norm": 0.3124683892505031,
"learning_rate": 3.502202656100099e-05,
"loss": 0.5473,
"step": 379
},
{
"epoch": 1.5384615384615383,
"grad_norm": 0.32312010451811346,
"learning_rate": 3.498463012735256e-05,
"loss": 0.5634,
"step": 380
},
{
"epoch": 1.54251012145749,
"grad_norm": 0.30982529269561543,
"learning_rate": 3.494711387700094e-05,
"loss": 0.5549,
"step": 381
},
{
"epoch": 1.5465587044534415,
"grad_norm": 0.3043203013697682,
"learning_rate": 3.490947810992505e-05,
"loss": 0.5401,
"step": 382
},
{
"epoch": 1.5506072874493926,
"grad_norm": 0.3535748676929678,
"learning_rate": 3.48717231270595e-05,
"loss": 0.5534,
"step": 383
},
{
"epoch": 1.5546558704453441,
"grad_norm": 0.37532859770625554,
"learning_rate": 3.483384923029211e-05,
"loss": 0.5451,
"step": 384
},
{
"epoch": 1.5587044534412957,
"grad_norm": 0.31715582122954533,
"learning_rate": 3.479585672246156e-05,
"loss": 0.5422,
"step": 385
},
{
"epoch": 1.5627530364372468,
"grad_norm": 0.40815913390523195,
"learning_rate": 3.4757745907354924e-05,
"loss": 0.5386,
"step": 386
},
{
"epoch": 1.5668016194331984,
"grad_norm": 0.4650283473947194,
"learning_rate": 3.471951708970527e-05,
"loss": 0.5489,
"step": 387
},
{
"epoch": 1.5708502024291497,
"grad_norm": 0.29889279245296574,
"learning_rate": 3.468117057518921e-05,
"loss": 0.5459,
"step": 388
},
{
"epoch": 1.574898785425101,
"grad_norm": 0.3921342378924284,
"learning_rate": 3.4642706670424426e-05,
"loss": 0.5433,
"step": 389
},
{
"epoch": 1.5789473684210527,
"grad_norm": 0.3780032755798313,
"learning_rate": 3.460412568296731e-05,
"loss": 0.5492,
"step": 390
},
{
"epoch": 1.582995951417004,
"grad_norm": 0.3603263430475091,
"learning_rate": 3.456542792131039e-05,
"loss": 0.5482,
"step": 391
},
{
"epoch": 1.5870445344129553,
"grad_norm": 0.36162283915757043,
"learning_rate": 3.4526613694879936e-05,
"loss": 0.553,
"step": 392
},
{
"epoch": 1.591093117408907,
"grad_norm": 0.315444360422662,
"learning_rate": 3.448768331403348e-05,
"loss": 0.5538,
"step": 393
},
{
"epoch": 1.5951417004048583,
"grad_norm": 0.4141722336305682,
"learning_rate": 3.44486370900573e-05,
"loss": 0.5419,
"step": 394
},
{
"epoch": 1.5991902834008096,
"grad_norm": 0.38046442439352024,
"learning_rate": 3.440947533516396e-05,
"loss": 0.5474,
"step": 395
},
{
"epoch": 1.6032388663967612,
"grad_norm": 0.352839774386106,
"learning_rate": 3.4370198362489816e-05,
"loss": 0.5534,
"step": 396
},
{
"epoch": 1.6072874493927125,
"grad_norm": 0.3380961913163569,
"learning_rate": 3.43308064860925e-05,
"loss": 0.5432,
"step": 397
},
{
"epoch": 1.6113360323886639,
"grad_norm": 0.3288462351660842,
"learning_rate": 3.4291300020948393e-05,
"loss": 0.5482,
"step": 398
},
{
"epoch": 1.6153846153846154,
"grad_norm": 0.35930596928257313,
"learning_rate": 3.425167928295015e-05,
"loss": 0.5473,
"step": 399
},
{
"epoch": 1.6194331983805668,
"grad_norm": 0.34714944283446353,
"learning_rate": 3.421194458890411e-05,
"loss": 0.5472,
"step": 400
},
{
"epoch": 1.623481781376518,
"grad_norm": 0.3491248822289833,
"learning_rate": 3.4172096256527865e-05,
"loss": 0.5418,
"step": 401
},
{
"epoch": 1.6275303643724697,
"grad_norm": 0.35308140182398096,
"learning_rate": 3.4132134604447595e-05,
"loss": 0.5585,
"step": 402
},
{
"epoch": 1.631578947368421,
"grad_norm": 0.3367947716659763,
"learning_rate": 3.409205995219561e-05,
"loss": 0.5534,
"step": 403
},
{
"epoch": 1.6356275303643724,
"grad_norm": 0.31619958101911166,
"learning_rate": 3.4051872620207765e-05,
"loss": 0.5307,
"step": 404
},
{
"epoch": 1.639676113360324,
"grad_norm": 0.3774152124131668,
"learning_rate": 3.40115729298209e-05,
"loss": 0.544,
"step": 405
},
{
"epoch": 1.6437246963562753,
"grad_norm": 0.41554814530122947,
"learning_rate": 3.397116120327025e-05,
"loss": 0.5486,
"step": 406
},
{
"epoch": 1.6477732793522266,
"grad_norm": 0.3178786393587894,
"learning_rate": 3.3930637763686916e-05,
"loss": 0.5439,
"step": 407
},
{
"epoch": 1.6518218623481782,
"grad_norm": 0.29516237323945527,
"learning_rate": 3.389000293509524e-05,
"loss": 0.5446,
"step": 408
},
{
"epoch": 1.6558704453441295,
"grad_norm": 0.3321219986241057,
"learning_rate": 3.3849257042410225e-05,
"loss": 0.553,
"step": 409
},
{
"epoch": 1.6599190283400809,
"grad_norm": 0.3286653535432779,
"learning_rate": 3.380840041143494e-05,
"loss": 0.5381,
"step": 410
},
{
"epoch": 1.6639676113360324,
"grad_norm": 0.3620649699580795,
"learning_rate": 3.376743336885792e-05,
"loss": 0.5451,
"step": 411
},
{
"epoch": 1.6680161943319838,
"grad_norm": 0.37058007061593845,
"learning_rate": 3.372635624225055e-05,
"loss": 0.5393,
"step": 412
},
{
"epoch": 1.6720647773279351,
"grad_norm": 0.328644968413847,
"learning_rate": 3.3685169360064445e-05,
"loss": 0.547,
"step": 413
},
{
"epoch": 1.6761133603238867,
"grad_norm": 0.3801012320013721,
"learning_rate": 3.3643873051628804e-05,
"loss": 0.544,
"step": 414
},
{
"epoch": 1.680161943319838,
"grad_norm": 0.31422752805139614,
"learning_rate": 3.3602467647147814e-05,
"loss": 0.5628,
"step": 415
},
{
"epoch": 1.6842105263157894,
"grad_norm": 0.36505824748097365,
"learning_rate": 3.356095347769799e-05,
"loss": 0.5471,
"step": 416
},
{
"epoch": 1.688259109311741,
"grad_norm": 0.34449300663460375,
"learning_rate": 3.351933087522552e-05,
"loss": 0.5359,
"step": 417
},
{
"epoch": 1.6923076923076923,
"grad_norm": 0.33184992036351735,
"learning_rate": 3.3477600172543626e-05,
"loss": 0.5398,
"step": 418
},
{
"epoch": 1.6963562753036436,
"grad_norm": 0.4085906080457272,
"learning_rate": 3.3435761703329894e-05,
"loss": 0.5513,
"step": 419
},
{
"epoch": 1.7004048582995952,
"grad_norm": 0.5153346956881846,
"learning_rate": 3.3393815802123606e-05,
"loss": 0.5541,
"step": 420
},
{
"epoch": 1.7044534412955465,
"grad_norm": 0.36382935992167487,
"learning_rate": 3.335176280432308e-05,
"loss": 0.5439,
"step": 421
},
{
"epoch": 1.708502024291498,
"grad_norm": 0.4813876675584252,
"learning_rate": 3.330960304618295e-05,
"loss": 0.543,
"step": 422
},
{
"epoch": 1.7125506072874495,
"grad_norm": 0.443514850277477,
"learning_rate": 3.326733686481153e-05,
"loss": 0.5474,
"step": 423
},
{
"epoch": 1.7165991902834008,
"grad_norm": 0.5389738767214274,
"learning_rate": 3.3224964598168084e-05,
"loss": 0.5427,
"step": 424
},
{
"epoch": 1.7206477732793521,
"grad_norm": 0.5079891695012314,
"learning_rate": 3.318248658506012e-05,
"loss": 0.5508,
"step": 425
},
{
"epoch": 1.7246963562753037,
"grad_norm": 0.3687343136569852,
"learning_rate": 3.313990316514069e-05,
"loss": 0.5336,
"step": 426
},
{
"epoch": 1.728744939271255,
"grad_norm": 0.5010395046529087,
"learning_rate": 3.309721467890571e-05,
"loss": 0.5415,
"step": 427
},
{
"epoch": 1.7327935222672064,
"grad_norm": 0.5048393723426642,
"learning_rate": 3.305442146769115e-05,
"loss": 0.5518,
"step": 428
},
{
"epoch": 1.736842105263158,
"grad_norm": 0.3178379008285543,
"learning_rate": 3.30115238736704e-05,
"loss": 0.5496,
"step": 429
},
{
"epoch": 1.7408906882591093,
"grad_norm": 0.43130649837125073,
"learning_rate": 3.296852223985148e-05,
"loss": 0.5356,
"step": 430
},
{
"epoch": 1.7449392712550607,
"grad_norm": 0.3718918696388297,
"learning_rate": 3.2925416910074315e-05,
"loss": 0.5551,
"step": 431
},
{
"epoch": 1.7489878542510122,
"grad_norm": 0.3457696897135279,
"learning_rate": 3.288220822900796e-05,
"loss": 0.5338,
"step": 432
},
{
"epoch": 1.7530364372469636,
"grad_norm": 0.47347137778225296,
"learning_rate": 3.283889654214788e-05,
"loss": 0.5461,
"step": 433
},
{
"epoch": 1.757085020242915,
"grad_norm": 0.38312355746602045,
"learning_rate": 3.279548219581317e-05,
"loss": 0.5474,
"step": 434
},
{
"epoch": 1.7611336032388665,
"grad_norm": 0.376576597500733,
"learning_rate": 3.275196553714379e-05,
"loss": 0.5569,
"step": 435
},
{
"epoch": 1.7651821862348178,
"grad_norm": 0.4425813428398234,
"learning_rate": 3.270834691409778e-05,
"loss": 0.5526,
"step": 436
},
{
"epoch": 1.7692307692307692,
"grad_norm": 0.40467861458253523,
"learning_rate": 3.266462667544849e-05,
"loss": 0.5469,
"step": 437
},
{
"epoch": 1.7732793522267207,
"grad_norm": 0.4142345814956182,
"learning_rate": 3.262080517078178e-05,
"loss": 0.5512,
"step": 438
},
{
"epoch": 1.777327935222672,
"grad_norm": 0.5090122951118925,
"learning_rate": 3.257688275049323e-05,
"loss": 0.5572,
"step": 439
},
{
"epoch": 1.7813765182186234,
"grad_norm": 0.3668559381122426,
"learning_rate": 3.253285976578535e-05,
"loss": 0.5371,
"step": 440
},
{
"epoch": 1.785425101214575,
"grad_norm": 0.58017027766774,
"learning_rate": 3.2488736568664756e-05,
"loss": 0.5529,
"step": 441
},
{
"epoch": 1.7894736842105263,
"grad_norm": 0.47645229418231966,
"learning_rate": 3.244451351193935e-05,
"loss": 0.5402,
"step": 442
},
{
"epoch": 1.7935222672064777,
"grad_norm": 0.3918360316228305,
"learning_rate": 3.240019094921551e-05,
"loss": 0.5441,
"step": 443
},
{
"epoch": 1.7975708502024292,
"grad_norm": 0.4458024583922774,
"learning_rate": 3.235576923489527e-05,
"loss": 0.5456,
"step": 444
},
{
"epoch": 1.8016194331983806,
"grad_norm": 0.37829367150010773,
"learning_rate": 3.2311248724173476e-05,
"loss": 0.5423,
"step": 445
},
{
"epoch": 1.805668016194332,
"grad_norm": 0.42986449406432087,
"learning_rate": 3.226662977303494e-05,
"loss": 0.5501,
"step": 446
},
{
"epoch": 1.8097165991902835,
"grad_norm": 0.34123069368114434,
"learning_rate": 3.2221912738251626e-05,
"loss": 0.5563,
"step": 447
},
{
"epoch": 1.8137651821862348,
"grad_norm": 0.44320866206679127,
"learning_rate": 3.2177097977379736e-05,
"loss": 0.549,
"step": 448
},
{
"epoch": 1.8178137651821862,
"grad_norm": 0.3550705843284255,
"learning_rate": 3.21321858487569e-05,
"loss": 0.544,
"step": 449
},
{
"epoch": 1.8218623481781377,
"grad_norm": 0.3882516751545717,
"learning_rate": 3.208717671149931e-05,
"loss": 0.5416,
"step": 450
},
{
"epoch": 1.825910931174089,
"grad_norm": 0.4485770248489349,
"learning_rate": 3.204207092549882e-05,
"loss": 0.5495,
"step": 451
},
{
"epoch": 1.8299595141700404,
"grad_norm": 0.3616451201287925,
"learning_rate": 3.19968688514201e-05,
"loss": 0.5449,
"step": 452
},
{
"epoch": 1.834008097165992,
"grad_norm": 0.4375732019780631,
"learning_rate": 3.195157085069774e-05,
"loss": 0.5407,
"step": 453
},
{
"epoch": 1.8380566801619433,
"grad_norm": 0.3377461072392165,
"learning_rate": 3.190617728553332e-05,
"loss": 0.5443,
"step": 454
},
{
"epoch": 1.8421052631578947,
"grad_norm": 0.4350748474857319,
"learning_rate": 3.1860688518892606e-05,
"loss": 0.539,
"step": 455
},
{
"epoch": 1.8461538461538463,
"grad_norm": 0.4033938580672089,
"learning_rate": 3.181510491450254e-05,
"loss": 0.5526,
"step": 456
},
{
"epoch": 1.8502024291497976,
"grad_norm": 0.4165613392186398,
"learning_rate": 3.176942683684842e-05,
"loss": 0.5617,
"step": 457
},
{
"epoch": 1.854251012145749,
"grad_norm": 0.43948448966491577,
"learning_rate": 3.1723654651170934e-05,
"loss": 0.5433,
"step": 458
},
{
"epoch": 1.8582995951417005,
"grad_norm": 0.37286859042625403,
"learning_rate": 3.167778872346327e-05,
"loss": 0.524,
"step": 459
},
{
"epoch": 1.8623481781376519,
"grad_norm": 0.4057634806463296,
"learning_rate": 3.163182942046815e-05,
"loss": 0.5499,
"step": 460
},
{
"epoch": 1.8663967611336032,
"grad_norm": 0.4706529961321381,
"learning_rate": 3.158577710967494e-05,
"loss": 0.54,
"step": 461
},
{
"epoch": 1.8704453441295548,
"grad_norm": 0.3435100987983101,
"learning_rate": 3.153963215931669e-05,
"loss": 0.5354,
"step": 462
},
{
"epoch": 1.874493927125506,
"grad_norm": 0.3656646460470125,
"learning_rate": 3.14933949383672e-05,
"loss": 0.5388,
"step": 463
},
{
"epoch": 1.8785425101214575,
"grad_norm": 0.3744319332042174,
"learning_rate": 3.144706581653805e-05,
"loss": 0.5563,
"step": 464
},
{
"epoch": 1.882591093117409,
"grad_norm": 0.35315539714382327,
"learning_rate": 3.140064516427566e-05,
"loss": 0.5398,
"step": 465
},
{
"epoch": 1.8866396761133604,
"grad_norm": 0.34230531445417606,
"learning_rate": 3.135413335275833e-05,
"loss": 0.5406,
"step": 466
},
{
"epoch": 1.8906882591093117,
"grad_norm": 0.3400077466248577,
"learning_rate": 3.130753075389327e-05,
"loss": 0.5405,
"step": 467
},
{
"epoch": 1.8947368421052633,
"grad_norm": 0.36618552210400285,
"learning_rate": 3.1260837740313625e-05,
"loss": 0.5432,
"step": 468
},
{
"epoch": 1.8987854251012146,
"grad_norm": 0.3457675928464259,
"learning_rate": 3.121405468537547e-05,
"loss": 0.5454,
"step": 469
},
{
"epoch": 1.902834008097166,
"grad_norm": 0.36340793241064107,
"learning_rate": 3.116718196315488e-05,
"loss": 0.5382,
"step": 470
},
{
"epoch": 1.9068825910931175,
"grad_norm": 0.39756007185012354,
"learning_rate": 3.1120219948444894e-05,
"loss": 0.5363,
"step": 471
},
{
"epoch": 1.9109311740890689,
"grad_norm": 0.34068540495627975,
"learning_rate": 3.107316901675252e-05,
"loss": 0.5456,
"step": 472
},
{
"epoch": 1.9149797570850202,
"grad_norm": 0.40228575308644293,
"learning_rate": 3.1026029544295745e-05,
"loss": 0.5413,
"step": 473
},
{
"epoch": 1.9190283400809718,
"grad_norm": 0.3599411310095468,
"learning_rate": 3.097880190800057e-05,
"loss": 0.5538,
"step": 474
},
{
"epoch": 1.9230769230769231,
"grad_norm": 0.38575030452633347,
"learning_rate": 3.093148648549788e-05,
"loss": 0.5371,
"step": 475
},
{
"epoch": 1.9271255060728745,
"grad_norm": 0.31123906008437274,
"learning_rate": 3.088408365512055e-05,
"loss": 0.5514,
"step": 476
},
{
"epoch": 1.931174089068826,
"grad_norm": 0.36604652361800505,
"learning_rate": 3.083659379590034e-05,
"loss": 0.5589,
"step": 477
},
{
"epoch": 1.9352226720647774,
"grad_norm": 0.35409589624858884,
"learning_rate": 3.0789017287564913e-05,
"loss": 0.5454,
"step": 478
},
{
"epoch": 1.9392712550607287,
"grad_norm": 0.31660990123193217,
"learning_rate": 3.074135451053475e-05,
"loss": 0.5461,
"step": 479
},
{
"epoch": 1.9433198380566803,
"grad_norm": 0.3783912700696826,
"learning_rate": 3.069360584592015e-05,
"loss": 0.5387,
"step": 480
},
{
"epoch": 1.9473684210526314,
"grad_norm": 0.4222018267931458,
"learning_rate": 3.064577167551816e-05,
"loss": 0.5428,
"step": 481
},
{
"epoch": 1.951417004048583,
"grad_norm": 0.3671151500462562,
"learning_rate": 3.059785238180953e-05,
"loss": 0.5444,
"step": 482
},
{
"epoch": 1.9554655870445345,
"grad_norm": 0.3096011044723037,
"learning_rate": 3.054984834795566e-05,
"loss": 0.5461,
"step": 483
},
{
"epoch": 1.9595141700404857,
"grad_norm": 0.43522996620879,
"learning_rate": 3.0501759957795537e-05,
"loss": 0.5529,
"step": 484
},
{
"epoch": 1.9635627530364372,
"grad_norm": 0.3866763253283052,
"learning_rate": 3.0453587595842627e-05,
"loss": 0.5404,
"step": 485
},
{
"epoch": 1.9676113360323888,
"grad_norm": 0.4030378634934215,
"learning_rate": 3.0405331647281856e-05,
"loss": 0.5563,
"step": 486
},
{
"epoch": 1.97165991902834,
"grad_norm": 0.2949386468081334,
"learning_rate": 3.0356992497966508e-05,
"loss": 0.5394,
"step": 487
},
{
"epoch": 1.9757085020242915,
"grad_norm": 0.34270935649264433,
"learning_rate": 3.0308570534415137e-05,
"loss": 0.544,
"step": 488
},
{
"epoch": 1.979757085020243,
"grad_norm": 0.33550623706017285,
"learning_rate": 3.0260066143808467e-05,
"loss": 0.5597,
"step": 489
},
{
"epoch": 1.9838056680161942,
"grad_norm": 0.3612801181333005,
"learning_rate": 3.021147971398632e-05,
"loss": 0.5249,
"step": 490
},
{
"epoch": 1.9878542510121457,
"grad_norm": 0.304052822445638,
"learning_rate": 3.0162811633444478e-05,
"loss": 0.5487,
"step": 491
},
{
"epoch": 1.9919028340080973,
"grad_norm": 0.4247169097057753,
"learning_rate": 3.0114062291331635e-05,
"loss": 0.539,
"step": 492
},
{
"epoch": 1.9959514170040484,
"grad_norm": 0.37962712673852633,
"learning_rate": 3.0065232077446217e-05,
"loss": 0.5443,
"step": 493
},
{
"epoch": 2.0,
"grad_norm": 0.4072275216551432,
"learning_rate": 3.001632138223332e-05,
"loss": 0.5445,
"step": 494
},
{
"epoch": 2.0040485829959516,
"grad_norm": 0.5541832055506406,
"learning_rate": 2.9967330596781565e-05,
"loss": 0.4811,
"step": 495
},
{
"epoch": 2.0080971659919027,
"grad_norm": 0.5675067786981842,
"learning_rate": 2.9918260112819952e-05,
"loss": 0.4739,
"step": 496
},
{
"epoch": 2.0121457489878543,
"grad_norm": 0.6474474025263246,
"learning_rate": 2.9869110322714778e-05,
"loss": 0.4759,
"step": 497
},
{
"epoch": 2.016194331983806,
"grad_norm": 0.46630161919329144,
"learning_rate": 2.9819881619466447e-05,
"loss": 0.4825,
"step": 498
},
{
"epoch": 2.020242914979757,
"grad_norm": 0.5025464345514765,
"learning_rate": 2.9770574396706364e-05,
"loss": 0.4695,
"step": 499
},
{
"epoch": 2.0242914979757085,
"grad_norm": 0.45459030918987603,
"learning_rate": 2.972118904869376e-05,
"loss": 0.4745,
"step": 500
},
{
"epoch": 2.02834008097166,
"grad_norm": 0.5163210750527535,
"learning_rate": 2.9671725970312563e-05,
"loss": 0.4899,
"step": 501
},
{
"epoch": 2.032388663967611,
"grad_norm": 0.5065902721108726,
"learning_rate": 2.9622185557068238e-05,
"loss": 0.4779,
"step": 502
},
{
"epoch": 2.0364372469635628,
"grad_norm": 0.4759610616530296,
"learning_rate": 2.95725682050846e-05,
"loss": 0.4739,
"step": 503
},
{
"epoch": 2.0404858299595143,
"grad_norm": 0.4009504258286367,
"learning_rate": 2.9522874311100684e-05,
"loss": 0.4788,
"step": 504
},
{
"epoch": 2.0445344129554655,
"grad_norm": 0.5327181089655362,
"learning_rate": 2.9473104272467526e-05,
"loss": 0.4743,
"step": 505
},
{
"epoch": 2.048582995951417,
"grad_norm": 0.44407191205757607,
"learning_rate": 2.942325848714505e-05,
"loss": 0.4712,
"step": 506
},
{
"epoch": 2.0526315789473686,
"grad_norm": 0.4979583429755772,
"learning_rate": 2.9373337353698804e-05,
"loss": 0.4953,
"step": 507
},
{
"epoch": 2.0566801619433197,
"grad_norm": 0.566634809550106,
"learning_rate": 2.932334127129686e-05,
"loss": 0.4855,
"step": 508
},
{
"epoch": 2.0607287449392713,
"grad_norm": 0.4736506912281605,
"learning_rate": 2.9273270639706547e-05,
"loss": 0.48,
"step": 509
},
{
"epoch": 2.064777327935223,
"grad_norm": 0.5857206281531874,
"learning_rate": 2.922312585929131e-05,
"loss": 0.4835,
"step": 510
},
{
"epoch": 2.068825910931174,
"grad_norm": 0.46004106235853987,
"learning_rate": 2.9172907331007478e-05,
"loss": 0.4778,
"step": 511
},
{
"epoch": 2.0728744939271255,
"grad_norm": 0.42339565033406934,
"learning_rate": 2.9122615456401053e-05,
"loss": 0.471,
"step": 512
},
{
"epoch": 2.076923076923077,
"grad_norm": 0.5034689048007474,
"learning_rate": 2.9072250637604535e-05,
"loss": 0.4794,
"step": 513
},
{
"epoch": 2.080971659919028,
"grad_norm": 0.34221721871119004,
"learning_rate": 2.9021813277333667e-05,
"loss": 0.4652,
"step": 514
},
{
"epoch": 2.08502024291498,
"grad_norm": 0.5013033435242442,
"learning_rate": 2.8971303778884244e-05,
"loss": 0.4707,
"step": 515
},
{
"epoch": 2.0890688259109313,
"grad_norm": 0.3729785926270019,
"learning_rate": 2.8920722546128863e-05,
"loss": 0.4706,
"step": 516
},
{
"epoch": 2.0931174089068825,
"grad_norm": 0.3787058083315791,
"learning_rate": 2.887006998351372e-05,
"loss": 0.4735,
"step": 517
},
{
"epoch": 2.097165991902834,
"grad_norm": 0.45415499977283796,
"learning_rate": 2.881934649605535e-05,
"loss": 0.4848,
"step": 518
},
{
"epoch": 2.1012145748987856,
"grad_norm": 0.39704611720711,
"learning_rate": 2.8768552489337408e-05,
"loss": 0.5002,
"step": 519
},
{
"epoch": 2.1052631578947367,
"grad_norm": 1.2123549834703637,
"learning_rate": 2.8717688369507425e-05,
"loss": 0.483,
"step": 520
},
{
"epoch": 2.1093117408906883,
"grad_norm": 0.3363279086886143,
"learning_rate": 2.8666754543273535e-05,
"loss": 0.4794,
"step": 521
},
{
"epoch": 2.11336032388664,
"grad_norm": 0.44468868586530136,
"learning_rate": 2.8615751417901273e-05,
"loss": 0.4724,
"step": 522
},
{
"epoch": 2.117408906882591,
"grad_norm": 0.3464090331269905,
"learning_rate": 2.856467940121025e-05,
"loss": 0.4943,
"step": 523
},
{
"epoch": 2.1214574898785425,
"grad_norm": 0.33659140551085515,
"learning_rate": 2.8513538901570964e-05,
"loss": 0.4719,
"step": 524
},
{
"epoch": 2.125506072874494,
"grad_norm": 0.3515205208875961,
"learning_rate": 2.846233032790149e-05,
"loss": 0.5003,
"step": 525
},
{
"epoch": 2.1295546558704452,
"grad_norm": 0.33242834636532204,
"learning_rate": 2.84110540896642e-05,
"loss": 0.4866,
"step": 526
},
{
"epoch": 2.133603238866397,
"grad_norm": 0.3367385556539788,
"learning_rate": 2.8359710596862555e-05,
"loss": 0.4804,
"step": 527
},
{
"epoch": 2.1376518218623484,
"grad_norm": 0.35194443255550273,
"learning_rate": 2.8308300260037734e-05,
"loss": 0.4607,
"step": 528
},
{
"epoch": 2.1417004048582995,
"grad_norm": 0.34511321243995297,
"learning_rate": 2.8256823490265424e-05,
"loss": 0.473,
"step": 529
},
{
"epoch": 2.145748987854251,
"grad_norm": 0.39467018521475217,
"learning_rate": 2.8205280699152515e-05,
"loss": 0.4794,
"step": 530
},
{
"epoch": 2.1497975708502026,
"grad_norm": 0.41738535314713193,
"learning_rate": 2.8153672298833777e-05,
"loss": 0.4764,
"step": 531
},
{
"epoch": 2.1538461538461537,
"grad_norm": 0.3364047975708467,
"learning_rate": 2.8101998701968615e-05,
"loss": 0.4754,
"step": 532
},
{
"epoch": 2.1578947368421053,
"grad_norm": 0.5174884891420547,
"learning_rate": 2.805026032173773e-05,
"loss": 0.4673,
"step": 533
},
{
"epoch": 2.161943319838057,
"grad_norm": 0.32542273734372495,
"learning_rate": 2.7998457571839832e-05,
"loss": 0.4735,
"step": 534
},
{
"epoch": 2.165991902834008,
"grad_norm": 0.4111775284840265,
"learning_rate": 2.7946590866488322e-05,
"loss": 0.4944,
"step": 535
},
{
"epoch": 2.1700404858299596,
"grad_norm": 0.35039522682843827,
"learning_rate": 2.7894660620408e-05,
"loss": 0.4923,
"step": 536
},
{
"epoch": 2.174089068825911,
"grad_norm": 0.3376468668117972,
"learning_rate": 2.784266724883173e-05,
"loss": 0.4756,
"step": 537
},
{
"epoch": 2.1781376518218623,
"grad_norm": 0.7324155695083684,
"learning_rate": 2.7790611167497128e-05,
"loss": 0.492,
"step": 538
},
{
"epoch": 2.182186234817814,
"grad_norm": 0.3379414055311623,
"learning_rate": 2.7738492792643224e-05,
"loss": 0.4623,
"step": 539
},
{
"epoch": 2.1862348178137654,
"grad_norm": 0.29899105440411416,
"learning_rate": 2.768631254100716e-05,
"loss": 0.479,
"step": 540
},
{
"epoch": 2.1902834008097165,
"grad_norm": 0.31227018695625514,
"learning_rate": 2.7634070829820836e-05,
"loss": 0.4734,
"step": 541
},
{
"epoch": 2.194331983805668,
"grad_norm": 0.3142628041367449,
"learning_rate": 2.7581768076807592e-05,
"loss": 0.4901,
"step": 542
},
{
"epoch": 2.1983805668016196,
"grad_norm": 0.33147555076909097,
"learning_rate": 2.7529404700178832e-05,
"loss": 0.4647,
"step": 543
},
{
"epoch": 2.2024291497975708,
"grad_norm": 0.3185814193766743,
"learning_rate": 2.7476981118630735e-05,
"loss": 0.4882,
"step": 544
},
{
"epoch": 2.2064777327935223,
"grad_norm": 0.35521416516370763,
"learning_rate": 2.742449775134085e-05,
"loss": 0.4772,
"step": 545
},
{
"epoch": 2.2105263157894735,
"grad_norm": 0.31556681797505165,
"learning_rate": 2.7371955017964788e-05,
"loss": 0.4766,
"step": 546
},
{
"epoch": 2.214574898785425,
"grad_norm": 0.33698443874297485,
"learning_rate": 2.7319353338632845e-05,
"loss": 0.4758,
"step": 547
},
{
"epoch": 2.2186234817813766,
"grad_norm": 0.34817730413489273,
"learning_rate": 2.7266693133946645e-05,
"loss": 0.4728,
"step": 548
},
{
"epoch": 2.2226720647773277,
"grad_norm": 0.4044409031615604,
"learning_rate": 2.7213974824975785e-05,
"loss": 0.4703,
"step": 549
},
{
"epoch": 2.2267206477732793,
"grad_norm": 0.31625961000239505,
"learning_rate": 2.7161198833254464e-05,
"loss": 0.4741,
"step": 550
},
{
"epoch": 2.230769230769231,
"grad_norm": 0.3058291494775051,
"learning_rate": 2.7108365580778095e-05,
"loss": 0.4776,
"step": 551
},
{
"epoch": 2.234817813765182,
"grad_norm": 0.3684975634349,
"learning_rate": 2.705547548999996e-05,
"loss": 0.4648,
"step": 552
},
{
"epoch": 2.2388663967611335,
"grad_norm": 0.2999807708318528,
"learning_rate": 2.7002528983827817e-05,
"loss": 0.4773,
"step": 553
},
{
"epoch": 2.242914979757085,
"grad_norm": 0.32843997451511336,
"learning_rate": 2.6949526485620517e-05,
"loss": 0.4834,
"step": 554
},
{
"epoch": 2.246963562753036,
"grad_norm": 0.30639146336790485,
"learning_rate": 2.6896468419184623e-05,
"loss": 0.4765,
"step": 555
},
{
"epoch": 2.251012145748988,
"grad_norm": 0.3236643194403029,
"learning_rate": 2.6843355208771024e-05,
"loss": 0.4749,
"step": 556
},
{
"epoch": 2.2550607287449393,
"grad_norm": 0.31304768023580326,
"learning_rate": 2.6790187279071536e-05,
"loss": 0.4686,
"step": 557
},
{
"epoch": 2.2591093117408905,
"grad_norm": 0.3384799879345065,
"learning_rate": 2.67369650552155e-05,
"loss": 0.4809,
"step": 558
},
{
"epoch": 2.263157894736842,
"grad_norm": 0.3682385156682879,
"learning_rate": 2.6683688962766407e-05,
"loss": 0.4792,
"step": 559
},
{
"epoch": 2.2672064777327936,
"grad_norm": 0.3318715834175961,
"learning_rate": 2.6630359427718468e-05,
"loss": 0.4808,
"step": 560
},
{
"epoch": 2.2712550607287447,
"grad_norm": 0.36405715968918495,
"learning_rate": 2.6576976876493225e-05,
"loss": 0.4857,
"step": 561
},
{
"epoch": 2.2753036437246963,
"grad_norm": 0.3206189185302972,
"learning_rate": 2.652354173593614e-05,
"loss": 0.4821,
"step": 562
},
{
"epoch": 2.279352226720648,
"grad_norm": 0.38819704695632745,
"learning_rate": 2.647005443331317e-05,
"loss": 0.4737,
"step": 563
},
{
"epoch": 2.283400809716599,
"grad_norm": 0.2905974513948015,
"learning_rate": 2.6416515396307354e-05,
"loss": 0.4819,
"step": 564
},
{
"epoch": 2.2874493927125505,
"grad_norm": 0.3877001266589502,
"learning_rate": 2.636292505301542e-05,
"loss": 0.479,
"step": 565
},
{
"epoch": 2.291497975708502,
"grad_norm": 0.35390298636423956,
"learning_rate": 2.630928383194431e-05,
"loss": 0.4774,
"step": 566
},
{
"epoch": 2.2955465587044532,
"grad_norm": 0.3763255963202235,
"learning_rate": 2.625559216200781e-05,
"loss": 0.4724,
"step": 567
},
{
"epoch": 2.299595141700405,
"grad_norm": 0.3200169917118948,
"learning_rate": 2.620185047252308e-05,
"loss": 0.4827,
"step": 568
},
{
"epoch": 2.3036437246963564,
"grad_norm": 0.3642741158801164,
"learning_rate": 2.614805919320723e-05,
"loss": 0.4689,
"step": 569
},
{
"epoch": 2.3076923076923075,
"grad_norm": 0.33684920676840346,
"learning_rate": 2.6094218754173895e-05,
"loss": 0.4757,
"step": 570
},
{
"epoch": 2.311740890688259,
"grad_norm": 0.45540514145794675,
"learning_rate": 2.60403295859298e-05,
"loss": 0.4745,
"step": 571
},
{
"epoch": 2.3157894736842106,
"grad_norm": 0.35092176284221543,
"learning_rate": 2.5986392119371297e-05,
"loss": 0.4715,
"step": 572
},
{
"epoch": 2.3198380566801617,
"grad_norm": 0.37995080200679143,
"learning_rate": 2.5932406785780928e-05,
"loss": 0.4763,
"step": 573
},
{
"epoch": 2.3238866396761133,
"grad_norm": 0.34666050815308247,
"learning_rate": 2.5878374016823987e-05,
"loss": 0.4844,
"step": 574
},
{
"epoch": 2.327935222672065,
"grad_norm": 0.3429713858576099,
"learning_rate": 2.5824294244545045e-05,
"loss": 0.4725,
"step": 575
},
{
"epoch": 2.331983805668016,
"grad_norm": 0.43246686212545277,
"learning_rate": 2.577016790136453e-05,
"loss": 0.4988,
"step": 576
},
{
"epoch": 2.3360323886639676,
"grad_norm": 0.42446721418156974,
"learning_rate": 2.5715995420075225e-05,
"loss": 0.4845,
"step": 577
},
{
"epoch": 2.340080971659919,
"grad_norm": 0.3129089378409444,
"learning_rate": 2.5661777233838864e-05,
"loss": 0.4832,
"step": 578
},
{
"epoch": 2.3441295546558703,
"grad_norm": 0.3739581450285314,
"learning_rate": 2.5607513776182616e-05,
"loss": 0.4825,
"step": 579
},
{
"epoch": 2.348178137651822,
"grad_norm": 0.31956479196438164,
"learning_rate": 2.5553205480995633e-05,
"loss": 0.4816,
"step": 580
},
{
"epoch": 2.3522267206477734,
"grad_norm": 0.42112999638161236,
"learning_rate": 2.5498852782525604e-05,
"loss": 0.4906,
"step": 581
},
{
"epoch": 2.3562753036437245,
"grad_norm": 0.3052230619437698,
"learning_rate": 2.5444456115375252e-05,
"loss": 0.4832,
"step": 582
},
{
"epoch": 2.360323886639676,
"grad_norm": 0.3730606927038924,
"learning_rate": 2.539001591449889e-05,
"loss": 0.4736,
"step": 583
},
{
"epoch": 2.3643724696356276,
"grad_norm": 0.3013016747631881,
"learning_rate": 2.53355326151989e-05,
"loss": 0.4855,
"step": 584
},
{
"epoch": 2.3684210526315788,
"grad_norm": 0.3037913465551971,
"learning_rate": 2.5281006653122288e-05,
"loss": 0.4788,
"step": 585
},
{
"epoch": 2.3724696356275303,
"grad_norm": 0.3032591671253608,
"learning_rate": 2.5226438464257217e-05,
"loss": 0.4866,
"step": 586
},
{
"epoch": 2.376518218623482,
"grad_norm": 0.3020327994240721,
"learning_rate": 2.5171828484929447e-05,
"loss": 0.4728,
"step": 587
},
{
"epoch": 2.380566801619433,
"grad_norm": 0.3640788741717116,
"learning_rate": 2.5117177151798934e-05,
"loss": 0.4804,
"step": 588
},
{
"epoch": 2.3846153846153846,
"grad_norm": 0.3066411117395843,
"learning_rate": 2.506248490185627e-05,
"loss": 0.4833,
"step": 589
},
{
"epoch": 2.388663967611336,
"grad_norm": 0.339039016927724,
"learning_rate": 2.500775217241923e-05,
"loss": 0.4948,
"step": 590
},
{
"epoch": 2.3927125506072873,
"grad_norm": 0.34986474676772283,
"learning_rate": 2.4952979401129266e-05,
"loss": 0.4786,
"step": 591
},
{
"epoch": 2.396761133603239,
"grad_norm": 0.3208500530240012,
"learning_rate": 2.4898167025948e-05,
"loss": 0.4839,
"step": 592
},
{
"epoch": 2.4008097165991904,
"grad_norm": 0.3062642334167518,
"learning_rate": 2.4843315485153703e-05,
"loss": 0.4848,
"step": 593
},
{
"epoch": 2.4048582995951415,
"grad_norm": 0.3286638357050328,
"learning_rate": 2.4788425217337856e-05,
"loss": 0.482,
"step": 594
},
{
"epoch": 2.408906882591093,
"grad_norm": 0.35548163329421706,
"learning_rate": 2.4733496661401572e-05,
"loss": 0.4733,
"step": 595
},
{
"epoch": 2.4129554655870447,
"grad_norm": 0.2894754167177968,
"learning_rate": 2.4678530256552114e-05,
"loss": 0.4821,
"step": 596
},
{
"epoch": 2.417004048582996,
"grad_norm": 0.34973173497104054,
"learning_rate": 2.46235264422994e-05,
"loss": 0.4879,
"step": 597
},
{
"epoch": 2.4210526315789473,
"grad_norm": 0.30235869601305165,
"learning_rate": 2.4568485658452452e-05,
"loss": 0.4763,
"step": 598
},
{
"epoch": 2.425101214574899,
"grad_norm": 0.3455536691700398,
"learning_rate": 2.451340834511592e-05,
"loss": 0.4927,
"step": 599
},
{
"epoch": 2.42914979757085,
"grad_norm": 0.28885434850948766,
"learning_rate": 2.4458294942686528e-05,
"loss": 0.4889,
"step": 600
},
{
"epoch": 2.4331983805668016,
"grad_norm": 0.3668455195567837,
"learning_rate": 2.440314589184958e-05,
"loss": 0.4714,
"step": 601
},
{
"epoch": 2.437246963562753,
"grad_norm": 0.29027116589942287,
"learning_rate": 2.434796163357541e-05,
"loss": 0.4796,
"step": 602
},
{
"epoch": 2.4412955465587043,
"grad_norm": 0.33873936505248686,
"learning_rate": 2.429274260911588e-05,
"loss": 0.4838,
"step": 603
},
{
"epoch": 2.445344129554656,
"grad_norm": 0.29028064939538706,
"learning_rate": 2.423748926000084e-05,
"loss": 0.4766,
"step": 604
},
{
"epoch": 2.4493927125506074,
"grad_norm": 0.29802403207189465,
"learning_rate": 2.418220202803459e-05,
"loss": 0.4813,
"step": 605
},
{
"epoch": 2.4534412955465585,
"grad_norm": 0.2791056355462502,
"learning_rate": 2.412688135529237e-05,
"loss": 0.4707,
"step": 606
},
{
"epoch": 2.45748987854251,
"grad_norm": 0.2994506588867178,
"learning_rate": 2.4071527684116807e-05,
"loss": 0.4825,
"step": 607
},
{
"epoch": 2.4615384615384617,
"grad_norm": 0.24752653896881993,
"learning_rate": 2.4016141457114372e-05,
"loss": 0.4832,
"step": 608
},
{
"epoch": 2.465587044534413,
"grad_norm": 0.33594209170450495,
"learning_rate": 2.396072311715187e-05,
"loss": 0.4717,
"step": 609
},
{
"epoch": 2.4696356275303644,
"grad_norm": 0.2940552064001648,
"learning_rate": 2.3905273107352863e-05,
"loss": 0.4875,
"step": 610
},
{
"epoch": 2.473684210526316,
"grad_norm": 0.32701197002266985,
"learning_rate": 2.3849791871094157e-05,
"loss": 0.4849,
"step": 611
},
{
"epoch": 2.477732793522267,
"grad_norm": 0.33760413136467515,
"learning_rate": 2.379427985200223e-05,
"loss": 0.4715,
"step": 612
},
{
"epoch": 2.4817813765182186,
"grad_norm": 0.3348627317308485,
"learning_rate": 2.373873749394972e-05,
"loss": 0.4768,
"step": 613
},
{
"epoch": 2.48582995951417,
"grad_norm": 0.36306606116692075,
"learning_rate": 2.3683165241051836e-05,
"loss": 0.488,
"step": 614
},
{
"epoch": 2.4898785425101213,
"grad_norm": 0.34182570688354746,
"learning_rate": 2.3627563537662823e-05,
"loss": 0.4903,
"step": 615
},
{
"epoch": 2.493927125506073,
"grad_norm": 0.3467435773507423,
"learning_rate": 2.3571932828372433e-05,
"loss": 0.4903,
"step": 616
},
{
"epoch": 2.4979757085020244,
"grad_norm": 0.3130634986882777,
"learning_rate": 2.351627355800233e-05,
"loss": 0.4811,
"step": 617
},
{
"epoch": 2.5020242914979756,
"grad_norm": 0.34040469846522836,
"learning_rate": 2.346058617160256e-05,
"loss": 0.4766,
"step": 618
},
{
"epoch": 2.506072874493927,
"grad_norm": 0.3084197640124951,
"learning_rate": 2.3404871114447976e-05,
"loss": 0.4723,
"step": 619
},
{
"epoch": 2.5101214574898787,
"grad_norm": 0.2914174347212612,
"learning_rate": 2.334912883203469e-05,
"loss": 0.4832,
"step": 620
},
{
"epoch": 2.51417004048583,
"grad_norm": 0.31511072293515013,
"learning_rate": 2.3293359770076505e-05,
"loss": 0.4736,
"step": 621
},
{
"epoch": 2.5182186234817814,
"grad_norm": 0.26652946949699496,
"learning_rate": 2.323756437450136e-05,
"loss": 0.4797,
"step": 622
},
{
"epoch": 2.522267206477733,
"grad_norm": 0.3356713117775427,
"learning_rate": 2.3181743091447754e-05,
"loss": 0.4786,
"step": 623
},
{
"epoch": 2.526315789473684,
"grad_norm": 0.26142090181634,
"learning_rate": 2.3125896367261164e-05,
"loss": 0.4837,
"step": 624
},
{
"epoch": 2.5303643724696356,
"grad_norm": 0.35366970222949135,
"learning_rate": 2.307002464849052e-05,
"loss": 0.4885,
"step": 625
},
{
"epoch": 2.534412955465587,
"grad_norm": 0.29161153233596876,
"learning_rate": 2.30141283818846e-05,
"loss": 0.4897,
"step": 626
},
{
"epoch": 2.5384615384615383,
"grad_norm": 0.31594568947656165,
"learning_rate": 2.2958208014388462e-05,
"loss": 0.4868,
"step": 627
},
{
"epoch": 2.54251012145749,
"grad_norm": 0.3067302017756561,
"learning_rate": 2.2902263993139857e-05,
"loss": 0.4782,
"step": 628
},
{
"epoch": 2.5465587044534415,
"grad_norm": 0.3009641300560333,
"learning_rate": 2.2846296765465708e-05,
"loss": 0.4823,
"step": 629
},
{
"epoch": 2.5506072874493926,
"grad_norm": 0.3150237303610242,
"learning_rate": 2.2790306778878465e-05,
"loss": 0.4704,
"step": 630
},
{
"epoch": 2.554655870445344,
"grad_norm": 0.3138614572109611,
"learning_rate": 2.273429448107257e-05,
"loss": 0.475,
"step": 631
},
{
"epoch": 2.5587044534412957,
"grad_norm": 0.27651823044722423,
"learning_rate": 2.2678260319920868e-05,
"loss": 0.478,
"step": 632
},
{
"epoch": 2.562753036437247,
"grad_norm": 0.3563194306221572,
"learning_rate": 2.262220474347102e-05,
"loss": 0.4762,
"step": 633
},
{
"epoch": 2.5668016194331984,
"grad_norm": 0.2958462457686295,
"learning_rate": 2.2566128199941918e-05,
"loss": 0.4898,
"step": 634
},
{
"epoch": 2.57085020242915,
"grad_norm": 0.2846455631678649,
"learning_rate": 2.2510031137720104e-05,
"loss": 0.4821,
"step": 635
},
{
"epoch": 2.574898785425101,
"grad_norm": 0.29932015045012517,
"learning_rate": 2.2453914005356206e-05,
"loss": 0.4727,
"step": 636
},
{
"epoch": 2.5789473684210527,
"grad_norm": 0.30044784488182424,
"learning_rate": 2.2397777251561302e-05,
"loss": 0.4785,
"step": 637
},
{
"epoch": 2.582995951417004,
"grad_norm": 0.3742498215687435,
"learning_rate": 2.2341621325203393e-05,
"loss": 0.4862,
"step": 638
},
{
"epoch": 2.5870445344129553,
"grad_norm": 0.3136473885653711,
"learning_rate": 2.2285446675303764e-05,
"loss": 0.4841,
"step": 639
},
{
"epoch": 2.591093117408907,
"grad_norm": 0.31447584937719925,
"learning_rate": 2.2229253751033423e-05,
"loss": 0.4767,
"step": 640
},
{
"epoch": 2.5951417004048585,
"grad_norm": 0.3272628209204911,
"learning_rate": 2.2173043001709492e-05,
"loss": 0.4926,
"step": 641
},
{
"epoch": 2.5991902834008096,
"grad_norm": 0.27891324598113554,
"learning_rate": 2.2116814876791627e-05,
"loss": 0.4768,
"step": 642
},
{
"epoch": 2.603238866396761,
"grad_norm": 0.3671849574986216,
"learning_rate": 2.2060569825878413e-05,
"loss": 0.4762,
"step": 643
},
{
"epoch": 2.6072874493927127,
"grad_norm": 0.26358485604690884,
"learning_rate": 2.2004308298703777e-05,
"loss": 0.4826,
"step": 644
},
{
"epoch": 2.611336032388664,
"grad_norm": 0.34520301015534155,
"learning_rate": 2.1948030745133396e-05,
"loss": 0.475,
"step": 645
},
{
"epoch": 2.6153846153846154,
"grad_norm": 0.31018274754387243,
"learning_rate": 2.1891737615161083e-05,
"loss": 0.4752,
"step": 646
},
{
"epoch": 2.619433198380567,
"grad_norm": 0.2915787213325987,
"learning_rate": 2.1835429358905218e-05,
"loss": 0.4776,
"step": 647
},
{
"epoch": 2.623481781376518,
"grad_norm": 0.32534117122660233,
"learning_rate": 2.17791064266051e-05,
"loss": 0.4829,
"step": 648
},
{
"epoch": 2.6275303643724697,
"grad_norm": 0.3123739282126088,
"learning_rate": 2.172276926861741e-05,
"loss": 0.475,
"step": 649
},
{
"epoch": 2.6315789473684212,
"grad_norm": 0.295935936235898,
"learning_rate": 2.166641833541255e-05,
"loss": 0.4762,
"step": 650
},
{
"epoch": 2.6356275303643724,
"grad_norm": 0.30941629562635886,
"learning_rate": 2.161005407757109e-05,
"loss": 0.4804,
"step": 651
},
{
"epoch": 2.639676113360324,
"grad_norm": 0.2917409812213919,
"learning_rate": 2.155367694578013e-05,
"loss": 0.4864,
"step": 652
},
{
"epoch": 2.6437246963562755,
"grad_norm": 0.29897567321560997,
"learning_rate": 2.1497287390829712e-05,
"loss": 0.4688,
"step": 653
},
{
"epoch": 2.6477732793522266,
"grad_norm": 0.2862323173319777,
"learning_rate": 2.1440885863609208e-05,
"loss": 0.4659,
"step": 654
},
{
"epoch": 2.651821862348178,
"grad_norm": 0.2806616361720781,
"learning_rate": 2.1384472815103737e-05,
"loss": 0.4797,
"step": 655
},
{
"epoch": 2.6558704453441297,
"grad_norm": 0.2887000116704274,
"learning_rate": 2.1328048696390534e-05,
"loss": 0.4735,
"step": 656
},
{
"epoch": 2.659919028340081,
"grad_norm": 0.2536191961905323,
"learning_rate": 2.1271613958635342e-05,
"loss": 0.4891,
"step": 657
},
{
"epoch": 2.6639676113360324,
"grad_norm": 0.2822281505367809,
"learning_rate": 2.1215169053088825e-05,
"loss": 0.4681,
"step": 658
},
{
"epoch": 2.668016194331984,
"grad_norm": 0.2679029436713861,
"learning_rate": 2.1158714431082937e-05,
"loss": 0.467,
"step": 659
},
{
"epoch": 2.672064777327935,
"grad_norm": 0.2638072603490589,
"learning_rate": 2.110225054402733e-05,
"loss": 0.4727,
"step": 660
},
{
"epoch": 2.6761133603238867,
"grad_norm": 0.30536838771684255,
"learning_rate": 2.1045777843405747e-05,
"loss": 0.4863,
"step": 661
},
{
"epoch": 2.6801619433198383,
"grad_norm": 0.2697141775060618,
"learning_rate": 2.09892967807724e-05,
"loss": 0.4905,
"step": 662
},
{
"epoch": 2.6842105263157894,
"grad_norm": 0.3331867761005994,
"learning_rate": 2.093280780774835e-05,
"loss": 0.483,
"step": 663
},
{
"epoch": 2.688259109311741,
"grad_norm": 0.2687776680167029,
"learning_rate": 2.087631137601793e-05,
"loss": 0.4816,
"step": 664
},
{
"epoch": 2.6923076923076925,
"grad_norm": 0.3321100282951798,
"learning_rate": 2.0819807937325104e-05,
"loss": 0.4825,
"step": 665
},
{
"epoch": 2.6963562753036436,
"grad_norm": 0.27967438204981354,
"learning_rate": 2.0763297943469847e-05,
"loss": 0.4762,
"step": 666
},
{
"epoch": 2.700404858299595,
"grad_norm": 0.29847051929752905,
"learning_rate": 2.070678184630458e-05,
"loss": 0.4854,
"step": 667
},
{
"epoch": 2.7044534412955468,
"grad_norm": 0.2689597889057596,
"learning_rate": 2.06502600977305e-05,
"loss": 0.4875,
"step": 668
},
{
"epoch": 2.708502024291498,
"grad_norm": 0.28527762586857996,
"learning_rate": 2.0593733149694008e-05,
"loss": 0.4725,
"step": 669
},
{
"epoch": 2.7125506072874495,
"grad_norm": 0.26527778797282436,
"learning_rate": 2.0537201454183074e-05,
"loss": 0.4851,
"step": 670
},
{
"epoch": 2.716599190283401,
"grad_norm": 0.28040844982867025,
"learning_rate": 2.0480665463223626e-05,
"loss": 0.4734,
"step": 671
},
{
"epoch": 2.720647773279352,
"grad_norm": 0.27821965537886756,
"learning_rate": 2.042412562887594e-05,
"loss": 0.5019,
"step": 672
},
{
"epoch": 2.7246963562753037,
"grad_norm": 0.28577057533054334,
"learning_rate": 2.0367582403231033e-05,
"loss": 0.4858,
"step": 673
},
{
"epoch": 2.7287449392712553,
"grad_norm": 0.2871319682235321,
"learning_rate": 2.0311036238407023e-05,
"loss": 0.4851,
"step": 674
},
{
"epoch": 2.7327935222672064,
"grad_norm": 0.37116832639774516,
"learning_rate": 2.0254487586545534e-05,
"loss": 0.4791,
"step": 675
},
{
"epoch": 2.736842105263158,
"grad_norm": 0.2990636574828412,
"learning_rate": 2.0197936899808088e-05,
"loss": 0.4872,
"step": 676
},
{
"epoch": 2.7408906882591095,
"grad_norm": 0.3358376246916564,
"learning_rate": 2.0141384630372466e-05,
"loss": 0.4626,
"step": 677
},
{
"epoch": 2.7449392712550607,
"grad_norm": 0.30527981895142137,
"learning_rate": 2.0084831230429106e-05,
"loss": 0.486,
"step": 678
},
{
"epoch": 2.748987854251012,
"grad_norm": 0.3267269295629963,
"learning_rate": 2.0028277152177492e-05,
"loss": 0.4702,
"step": 679
},
{
"epoch": 2.753036437246964,
"grad_norm": 0.30748792707525313,
"learning_rate": 1.9971722847822518e-05,
"loss": 0.4955,
"step": 680
},
{
"epoch": 2.757085020242915,
"grad_norm": 0.343883996633968,
"learning_rate": 1.99151687695709e-05,
"loss": 0.4878,
"step": 681
},
{
"epoch": 2.7611336032388665,
"grad_norm": 0.3120994238541462,
"learning_rate": 1.9858615369627537e-05,
"loss": 0.4803,
"step": 682
},
{
"epoch": 2.765182186234818,
"grad_norm": 0.33901712223185687,
"learning_rate": 1.9802063100191916e-05,
"loss": 0.4839,
"step": 683
},
{
"epoch": 2.769230769230769,
"grad_norm": 0.30936487490227943,
"learning_rate": 1.974551241345447e-05,
"loss": 0.479,
"step": 684
},
{
"epoch": 2.7732793522267207,
"grad_norm": 0.31930180157449123,
"learning_rate": 1.9688963761592984e-05,
"loss": 0.4776,
"step": 685
},
{
"epoch": 2.7773279352226723,
"grad_norm": 0.25589040110426686,
"learning_rate": 1.963241759676897e-05,
"loss": 0.4718,
"step": 686
},
{
"epoch": 2.7813765182186234,
"grad_norm": 0.30505444578943974,
"learning_rate": 1.9575874371124066e-05,
"loss": 0.4838,
"step": 687
},
{
"epoch": 2.785425101214575,
"grad_norm": 0.2836512723158457,
"learning_rate": 1.951933453677638e-05,
"loss": 0.4766,
"step": 688
},
{
"epoch": 2.7894736842105265,
"grad_norm": 0.3162298752770068,
"learning_rate": 1.9462798545816932e-05,
"loss": 0.4747,
"step": 689
},
{
"epoch": 2.7935222672064777,
"grad_norm": 0.31663404500116604,
"learning_rate": 1.9406266850305996e-05,
"loss": 0.4744,
"step": 690
},
{
"epoch": 2.7975708502024292,
"grad_norm": 0.34162917719530883,
"learning_rate": 1.9349739902269503e-05,
"loss": 0.4735,
"step": 691
},
{
"epoch": 2.801619433198381,
"grad_norm": 0.3181853981340164,
"learning_rate": 1.9293218153695425e-05,
"loss": 0.4688,
"step": 692
},
{
"epoch": 2.805668016194332,
"grad_norm": 0.3865013751285766,
"learning_rate": 1.9236702056530153e-05,
"loss": 0.481,
"step": 693
},
{
"epoch": 2.8097165991902835,
"grad_norm": 0.280750514053072,
"learning_rate": 1.9180192062674907e-05,
"loss": 0.4719,
"step": 694
},
{
"epoch": 2.813765182186235,
"grad_norm": 0.399556584202146,
"learning_rate": 1.9123688623982076e-05,
"loss": 0.475,
"step": 695
},
{
"epoch": 2.817813765182186,
"grad_norm": 0.26304666375336117,
"learning_rate": 1.906719219225165e-05,
"loss": 0.4667,
"step": 696
},
{
"epoch": 2.8218623481781377,
"grad_norm": 0.311171827385552,
"learning_rate": 1.9010703219227603e-05,
"loss": 0.4696,
"step": 697
},
{
"epoch": 2.8259109311740893,
"grad_norm": 0.294707093723884,
"learning_rate": 1.8954222156594256e-05,
"loss": 0.4936,
"step": 698
},
{
"epoch": 2.8299595141700404,
"grad_norm": 0.3157257352488056,
"learning_rate": 1.8897749455972673e-05,
"loss": 0.4736,
"step": 699
},
{
"epoch": 2.834008097165992,
"grad_norm": 0.29490049700292,
"learning_rate": 1.884128556891707e-05,
"loss": 0.4761,
"step": 700
},
{
"epoch": 2.8380566801619436,
"grad_norm": 0.286742579257207,
"learning_rate": 1.878483094691119e-05,
"loss": 0.4791,
"step": 701
},
{
"epoch": 2.8421052631578947,
"grad_norm": 0.3012069862313598,
"learning_rate": 1.8728386041364664e-05,
"loss": 0.4769,
"step": 702
},
{
"epoch": 2.8461538461538463,
"grad_norm": 0.26251595965495184,
"learning_rate": 1.867195130360947e-05,
"loss": 0.4787,
"step": 703
},
{
"epoch": 2.850202429149798,
"grad_norm": 0.30641593007072887,
"learning_rate": 1.8615527184896263e-05,
"loss": 0.4859,
"step": 704
},
{
"epoch": 2.854251012145749,
"grad_norm": 0.25858616959524017,
"learning_rate": 1.8559114136390795e-05,
"loss": 0.4906,
"step": 705
},
{
"epoch": 2.8582995951417005,
"grad_norm": 0.2981298678231562,
"learning_rate": 1.8502712609170298e-05,
"loss": 0.4873,
"step": 706
},
{
"epoch": 2.862348178137652,
"grad_norm": 0.2625407804188751,
"learning_rate": 1.8446323054219876e-05,
"loss": 0.4774,
"step": 707
},
{
"epoch": 2.866396761133603,
"grad_norm": 0.29128926883834416,
"learning_rate": 1.838994592242891e-05,
"loss": 0.4757,
"step": 708
},
{
"epoch": 2.8704453441295548,
"grad_norm": 0.2934411060664796,
"learning_rate": 1.8333581664587453e-05,
"loss": 0.4699,
"step": 709
},
{
"epoch": 2.8744939271255063,
"grad_norm": 0.2778360710387577,
"learning_rate": 1.8277230731382593e-05,
"loss": 0.4868,
"step": 710
},
{
"epoch": 2.8785425101214575,
"grad_norm": 0.30084612576565606,
"learning_rate": 1.82208935733949e-05,
"loss": 0.4739,
"step": 711
},
{
"epoch": 2.882591093117409,
"grad_norm": 0.2808938562299745,
"learning_rate": 1.8164570641094793e-05,
"loss": 0.4783,
"step": 712
},
{
"epoch": 2.8866396761133606,
"grad_norm": 0.29361258693845094,
"learning_rate": 1.810826238483892e-05,
"loss": 0.4631,
"step": 713
},
{
"epoch": 2.8906882591093117,
"grad_norm": 0.29708447950599526,
"learning_rate": 1.8051969254866608e-05,
"loss": 0.472,
"step": 714
},
{
"epoch": 2.8947368421052633,
"grad_norm": 0.2738523171422319,
"learning_rate": 1.7995691701296226e-05,
"loss": 0.4817,
"step": 715
},
{
"epoch": 2.898785425101215,
"grad_norm": 0.31300752619455485,
"learning_rate": 1.7939430174121593e-05,
"loss": 0.4833,
"step": 716
},
{
"epoch": 2.902834008097166,
"grad_norm": 0.2753404168941184,
"learning_rate": 1.788318512320838e-05,
"loss": 0.4715,
"step": 717
},
{
"epoch": 2.9068825910931175,
"grad_norm": 0.29964683642321327,
"learning_rate": 1.782695699829051e-05,
"loss": 0.4832,
"step": 718
},
{
"epoch": 2.910931174089069,
"grad_norm": 0.25811312316486534,
"learning_rate": 1.7770746248966587e-05,
"loss": 0.4682,
"step": 719
},
{
"epoch": 2.91497975708502,
"grad_norm": 0.3173601262454415,
"learning_rate": 1.7714553324696243e-05,
"loss": 0.4811,
"step": 720
},
{
"epoch": 2.919028340080972,
"grad_norm": 0.27277883570732314,
"learning_rate": 1.7658378674796614e-05,
"loss": 0.4716,
"step": 721
},
{
"epoch": 2.9230769230769234,
"grad_norm": 0.2672956639831515,
"learning_rate": 1.76022227484387e-05,
"loss": 0.4885,
"step": 722
},
{
"epoch": 2.9271255060728745,
"grad_norm": 0.30727035142676984,
"learning_rate": 1.7546085994643807e-05,
"loss": 0.4851,
"step": 723
},
{
"epoch": 2.931174089068826,
"grad_norm": 0.24531056777584684,
"learning_rate": 1.7489968862279902e-05,
"loss": 0.4688,
"step": 724
},
{
"epoch": 2.9352226720647776,
"grad_norm": 0.27047774485188175,
"learning_rate": 1.743387180005809e-05,
"loss": 0.4758,
"step": 725
},
{
"epoch": 2.9392712550607287,
"grad_norm": 0.30156003993925423,
"learning_rate": 1.737779525652899e-05,
"loss": 0.4858,
"step": 726
},
{
"epoch": 2.9433198380566803,
"grad_norm": 0.24845461814262843,
"learning_rate": 1.732173968007914e-05,
"loss": 0.4776,
"step": 727
},
{
"epoch": 2.9473684210526314,
"grad_norm": 0.28938834375485045,
"learning_rate": 1.7265705518927437e-05,
"loss": 0.4769,
"step": 728
},
{
"epoch": 2.951417004048583,
"grad_norm": 0.26718691136362505,
"learning_rate": 1.7209693221121542e-05,
"loss": 0.4781,
"step": 729
},
{
"epoch": 2.9554655870445345,
"grad_norm": 0.2819000442797643,
"learning_rate": 1.7153703234534302e-05,
"loss": 0.478,
"step": 730
},
{
"epoch": 2.9595141700404857,
"grad_norm": 0.2692135927535854,
"learning_rate": 1.709773600686015e-05,
"loss": 0.4874,
"step": 731
},
{
"epoch": 2.9635627530364372,
"grad_norm": 0.30296001969716185,
"learning_rate": 1.7041791985611548e-05,
"loss": 0.4889,
"step": 732
},
{
"epoch": 2.967611336032389,
"grad_norm": 0.26682166715128686,
"learning_rate": 1.69858716181154e-05,
"loss": 0.4843,
"step": 733
},
{
"epoch": 2.97165991902834,
"grad_norm": 0.2568273391056995,
"learning_rate": 1.692997535150948e-05,
"loss": 0.477,
"step": 734
},
{
"epoch": 2.9757085020242915,
"grad_norm": 0.24948728308035617,
"learning_rate": 1.687410363273884e-05,
"loss": 0.4672,
"step": 735
},
{
"epoch": 2.979757085020243,
"grad_norm": 0.2740279014265944,
"learning_rate": 1.6818256908552257e-05,
"loss": 0.477,
"step": 736
},
{
"epoch": 2.983805668016194,
"grad_norm": 0.2826451980076029,
"learning_rate": 1.6762435625498646e-05,
"loss": 0.4806,
"step": 737
},
{
"epoch": 2.9878542510121457,
"grad_norm": 0.2857029739818436,
"learning_rate": 1.67066402299235e-05,
"loss": 0.4756,
"step": 738
},
{
"epoch": 2.9919028340080973,
"grad_norm": 0.2992887767350506,
"learning_rate": 1.6650871167965313e-05,
"loss": 0.48,
"step": 739
},
{
"epoch": 2.9959514170040484,
"grad_norm": 0.25298245315807055,
"learning_rate": 1.6595128885552028e-05,
"loss": 0.4884,
"step": 740
},
{
"epoch": 3.0,
"grad_norm": 0.3311319133002806,
"learning_rate": 1.653941382839745e-05,
"loss": 0.4797,
"step": 741
},
{
"epoch": 3.0040485829959516,
"grad_norm": 0.4366576101664064,
"learning_rate": 1.6483726441997673e-05,
"loss": 0.4168,
"step": 742
},
{
"epoch": 3.0080971659919027,
"grad_norm": 0.5359397306484873,
"learning_rate": 1.642806717162757e-05,
"loss": 0.4159,
"step": 743
},
{
"epoch": 3.0121457489878543,
"grad_norm": 0.3262653299494195,
"learning_rate": 1.637243646233718e-05,
"loss": 0.4201,
"step": 744
},
{
"epoch": 3.016194331983806,
"grad_norm": 0.49705842482517887,
"learning_rate": 1.6316834758948174e-05,
"loss": 0.4344,
"step": 745
},
{
"epoch": 3.020242914979757,
"grad_norm": 0.3372125757750969,
"learning_rate": 1.6261262506050282e-05,
"loss": 0.4136,
"step": 746
},
{
"epoch": 3.0242914979757085,
"grad_norm": 0.35890558988508664,
"learning_rate": 1.620572014799777e-05,
"loss": 0.4219,
"step": 747
},
{
"epoch": 3.02834008097166,
"grad_norm": 0.354489300204144,
"learning_rate": 1.6150208128905857e-05,
"loss": 0.4204,
"step": 748
},
{
"epoch": 3.032388663967611,
"grad_norm": 0.35673578127598027,
"learning_rate": 1.6094726892647147e-05,
"loss": 0.4156,
"step": 749
},
{
"epoch": 3.0364372469635628,
"grad_norm": 0.3335019077049019,
"learning_rate": 1.6039276882848135e-05,
"loss": 0.4189,
"step": 750
},
{
"epoch": 3.0404858299595143,
"grad_norm": 0.33979148814081733,
"learning_rate": 1.5983858542885635e-05,
"loss": 0.4212,
"step": 751
},
{
"epoch": 3.0445344129554655,
"grad_norm": 0.32605146251940037,
"learning_rate": 1.5928472315883203e-05,
"loss": 0.4221,
"step": 752
},
{
"epoch": 3.048582995951417,
"grad_norm": 0.31600942564726936,
"learning_rate": 1.5873118644707633e-05,
"loss": 0.4062,
"step": 753
},
{
"epoch": 3.0526315789473686,
"grad_norm": 0.2938228164441109,
"learning_rate": 1.5817797971965413e-05,
"loss": 0.4019,
"step": 754
},
{
"epoch": 3.0566801619433197,
"grad_norm": 0.2860082273816139,
"learning_rate": 1.576251073999917e-05,
"loss": 0.4268,
"step": 755
},
{
"epoch": 3.0607287449392713,
"grad_norm": 0.30203727674211045,
"learning_rate": 1.5707257390884126e-05,
"loss": 0.4008,
"step": 756
},
{
"epoch": 3.064777327935223,
"grad_norm": 0.27166321709629027,
"learning_rate": 1.5652038366424595e-05,
"loss": 0.4082,
"step": 757
},
{
"epoch": 3.068825910931174,
"grad_norm": 0.29202315783587607,
"learning_rate": 1.5596854108150424e-05,
"loss": 0.4197,
"step": 758
},
{
"epoch": 3.0728744939271255,
"grad_norm": 0.2800073492069457,
"learning_rate": 1.5541705057313476e-05,
"loss": 0.4152,
"step": 759
},
{
"epoch": 3.076923076923077,
"grad_norm": 0.2815478044908468,
"learning_rate": 1.5486591654884086e-05,
"loss": 0.4268,
"step": 760
},
{
"epoch": 3.080971659919028,
"grad_norm": 0.29987630325243464,
"learning_rate": 1.543151434154755e-05,
"loss": 0.4226,
"step": 761
},
{
"epoch": 3.08502024291498,
"grad_norm": 0.28226466706182046,
"learning_rate": 1.5376473557700615e-05,
"loss": 0.4063,
"step": 762
},
{
"epoch": 3.0890688259109313,
"grad_norm": 0.27864781115832826,
"learning_rate": 1.5321469743447893e-05,
"loss": 0.4031,
"step": 763
},
{
"epoch": 3.0931174089068825,
"grad_norm": 0.26785739544638526,
"learning_rate": 1.5266503338598434e-05,
"loss": 0.4199,
"step": 764
},
{
"epoch": 3.097165991902834,
"grad_norm": 0.28326304562732707,
"learning_rate": 1.5211574782662149e-05,
"loss": 0.4098,
"step": 765
},
{
"epoch": 3.1012145748987856,
"grad_norm": 0.2702009085656908,
"learning_rate": 1.5156684514846307e-05,
"loss": 0.3994,
"step": 766
},
{
"epoch": 3.1052631578947367,
"grad_norm": 0.27232552187973696,
"learning_rate": 1.5101832974052012e-05,
"loss": 0.4117,
"step": 767
},
{
"epoch": 3.1093117408906883,
"grad_norm": 0.31244704160820574,
"learning_rate": 1.5047020598870737e-05,
"loss": 0.4245,
"step": 768
},
{
"epoch": 3.11336032388664,
"grad_norm": 0.27018836141592434,
"learning_rate": 1.4992247827580778e-05,
"loss": 0.412,
"step": 769
},
{
"epoch": 3.117408906882591,
"grad_norm": 0.29774078849494834,
"learning_rate": 1.493751509814374e-05,
"loss": 0.4144,
"step": 770
},
{
"epoch": 3.1214574898785425,
"grad_norm": 0.23402943873129978,
"learning_rate": 1.4882822848201073e-05,
"loss": 0.4075,
"step": 771
},
{
"epoch": 3.125506072874494,
"grad_norm": 0.2849681070404747,
"learning_rate": 1.4828171515070553e-05,
"loss": 0.4063,
"step": 772
},
{
"epoch": 3.1295546558704452,
"grad_norm": 0.2599257559557787,
"learning_rate": 1.4773561535742793e-05,
"loss": 0.3898,
"step": 773
},
{
"epoch": 3.133603238866397,
"grad_norm": 0.2710254402497106,
"learning_rate": 1.4718993346877715e-05,
"loss": 0.4069,
"step": 774
},
{
"epoch": 3.1376518218623484,
"grad_norm": 0.28294166784448466,
"learning_rate": 1.466446738480111e-05,
"loss": 0.4084,
"step": 775
},
{
"epoch": 3.1417004048582995,
"grad_norm": 0.2634162314596436,
"learning_rate": 1.460998408550112e-05,
"loss": 0.4047,
"step": 776
},
{
"epoch": 3.145748987854251,
"grad_norm": 0.2851688347565253,
"learning_rate": 1.4555543884624751e-05,
"loss": 0.4272,
"step": 777
},
{
"epoch": 3.1497975708502026,
"grad_norm": 0.273357696455234,
"learning_rate": 1.4501147217474402e-05,
"loss": 0.4105,
"step": 778
},
{
"epoch": 3.1538461538461537,
"grad_norm": 0.2453526275344455,
"learning_rate": 1.444679451900437e-05,
"loss": 0.4208,
"step": 779
},
{
"epoch": 3.1578947368421053,
"grad_norm": 0.27393367857795164,
"learning_rate": 1.4392486223817397e-05,
"loss": 0.4039,
"step": 780
},
{
"epoch": 3.161943319838057,
"grad_norm": 0.23918139183397133,
"learning_rate": 1.4338222766161143e-05,
"loss": 0.4089,
"step": 781
},
{
"epoch": 3.165991902834008,
"grad_norm": 0.2746269487209849,
"learning_rate": 1.4284004579924777e-05,
"loss": 0.4129,
"step": 782
},
{
"epoch": 3.1700404858299596,
"grad_norm": 0.25551327888037234,
"learning_rate": 1.4229832098635479e-05,
"loss": 0.4032,
"step": 783
},
{
"epoch": 3.174089068825911,
"grad_norm": 0.26255309489928974,
"learning_rate": 1.4175705755454963e-05,
"loss": 0.3953,
"step": 784
},
{
"epoch": 3.1781376518218623,
"grad_norm": 0.2599742727669985,
"learning_rate": 1.412162598317602e-05,
"loss": 0.4001,
"step": 785
},
{
"epoch": 3.182186234817814,
"grad_norm": 0.2666361234777174,
"learning_rate": 1.406759321421907e-05,
"loss": 0.401,
"step": 786
},
{
"epoch": 3.1862348178137654,
"grad_norm": 0.2603092336612632,
"learning_rate": 1.401360788062871e-05,
"loss": 0.4172,
"step": 787
},
{
"epoch": 3.1902834008097165,
"grad_norm": 0.25593470358248216,
"learning_rate": 1.3959670414070204e-05,
"loss": 0.4265,
"step": 788
},
{
"epoch": 3.194331983805668,
"grad_norm": 0.27379743334828555,
"learning_rate": 1.3905781245826108e-05,
"loss": 0.402,
"step": 789
},
{
"epoch": 3.1983805668016196,
"grad_norm": 0.27148196893109966,
"learning_rate": 1.3851940806792778e-05,
"loss": 0.4244,
"step": 790
},
{
"epoch": 3.2024291497975708,
"grad_norm": 0.2881331959181754,
"learning_rate": 1.379814952747693e-05,
"loss": 0.4147,
"step": 791
},
{
"epoch": 3.2064777327935223,
"grad_norm": 0.24430638539274413,
"learning_rate": 1.3744407837992193e-05,
"loss": 0.4091,
"step": 792
},
{
"epoch": 3.2105263157894735,
"grad_norm": 0.27060234549866247,
"learning_rate": 1.3690716168055692e-05,
"loss": 0.4255,
"step": 793
},
{
"epoch": 3.214574898785425,
"grad_norm": 0.25416106422307927,
"learning_rate": 1.363707494698459e-05,
"loss": 0.4188,
"step": 794
},
{
"epoch": 3.2186234817813766,
"grad_norm": 0.25053130741675245,
"learning_rate": 1.358348460369265e-05,
"loss": 0.4145,
"step": 795
},
{
"epoch": 3.2226720647773277,
"grad_norm": 0.250319876565409,
"learning_rate": 1.352994556668684e-05,
"loss": 0.4224,
"step": 796
},
{
"epoch": 3.2267206477732793,
"grad_norm": 0.2606883846467937,
"learning_rate": 1.3476458264063865e-05,
"loss": 0.423,
"step": 797
},
{
"epoch": 3.230769230769231,
"grad_norm": 0.2535839758248415,
"learning_rate": 1.342302312350678e-05,
"loss": 0.4164,
"step": 798
},
{
"epoch": 3.234817813765182,
"grad_norm": 0.25729356001040427,
"learning_rate": 1.3369640572281537e-05,
"loss": 0.411,
"step": 799
},
{
"epoch": 3.2388663967611335,
"grad_norm": 0.24006221229165534,
"learning_rate": 1.3316311037233596e-05,
"loss": 0.4107,
"step": 800
},
{
"epoch": 3.242914979757085,
"grad_norm": 0.2569140848263187,
"learning_rate": 1.326303494478451e-05,
"loss": 0.4089,
"step": 801
},
{
"epoch": 3.246963562753036,
"grad_norm": 0.2524900544623315,
"learning_rate": 1.3209812720928472e-05,
"loss": 0.4198,
"step": 802
},
{
"epoch": 3.251012145748988,
"grad_norm": 0.24938026523541104,
"learning_rate": 1.315664479122898e-05,
"loss": 0.4081,
"step": 803
},
{
"epoch": 3.2550607287449393,
"grad_norm": 0.24124741170276762,
"learning_rate": 1.3103531580815378e-05,
"loss": 0.4177,
"step": 804
},
{
"epoch": 3.2591093117408905,
"grad_norm": 0.2396344285001214,
"learning_rate": 1.3050473514379488e-05,
"loss": 0.398,
"step": 805
},
{
"epoch": 3.263157894736842,
"grad_norm": 0.24512274723655114,
"learning_rate": 1.2997471016172188e-05,
"loss": 0.4288,
"step": 806
},
{
"epoch": 3.2672064777327936,
"grad_norm": 0.22922586603686473,
"learning_rate": 1.2944524510000042e-05,
"loss": 0.435,
"step": 807
},
{
"epoch": 3.2712550607287447,
"grad_norm": 0.2396058866951406,
"learning_rate": 1.2891634419221908e-05,
"loss": 0.4217,
"step": 808
},
{
"epoch": 3.2753036437246963,
"grad_norm": 0.2625023062432148,
"learning_rate": 1.2838801166745545e-05,
"loss": 0.426,
"step": 809
},
{
"epoch": 3.279352226720648,
"grad_norm": 0.2487509200305084,
"learning_rate": 1.2786025175024216e-05,
"loss": 0.4211,
"step": 810
},
{
"epoch": 3.283400809716599,
"grad_norm": 0.23670582892950354,
"learning_rate": 1.2733306866053357e-05,
"loss": 0.4134,
"step": 811
},
{
"epoch": 3.2874493927125505,
"grad_norm": 0.24586848286733431,
"learning_rate": 1.2680646661367163e-05,
"loss": 0.399,
"step": 812
},
{
"epoch": 3.291497975708502,
"grad_norm": 0.27054769576856635,
"learning_rate": 1.2628044982035219e-05,
"loss": 0.4136,
"step": 813
},
{
"epoch": 3.2955465587044532,
"grad_norm": 0.24530813516936556,
"learning_rate": 1.2575502248659155e-05,
"loss": 0.4244,
"step": 814
},
{
"epoch": 3.299595141700405,
"grad_norm": 0.26422245514970555,
"learning_rate": 1.2523018881369269e-05,
"loss": 0.4241,
"step": 815
},
{
"epoch": 3.3036437246963564,
"grad_norm": 0.22776709395915512,
"learning_rate": 1.2470595299821176e-05,
"loss": 0.4258,
"step": 816
},
{
"epoch": 3.3076923076923075,
"grad_norm": 0.26659998245659755,
"learning_rate": 1.2418231923192416e-05,
"loss": 0.4078,
"step": 817
},
{
"epoch": 3.311740890688259,
"grad_norm": 0.25686229301737873,
"learning_rate": 1.2365929170179165e-05,
"loss": 0.3981,
"step": 818
},
{
"epoch": 3.3157894736842106,
"grad_norm": 0.25619823140440323,
"learning_rate": 1.2313687458992848e-05,
"loss": 0.3998,
"step": 819
},
{
"epoch": 3.3198380566801617,
"grad_norm": 0.24895135339632105,
"learning_rate": 1.2261507207356785e-05,
"loss": 0.43,
"step": 820
},
{
"epoch": 3.3238866396761133,
"grad_norm": 0.2538722242739509,
"learning_rate": 1.220938883250288e-05,
"loss": 0.427,
"step": 821
},
{
"epoch": 3.327935222672065,
"grad_norm": 0.24510217480171728,
"learning_rate": 1.215733275116827e-05,
"loss": 0.4075,
"step": 822
},
{
"epoch": 3.331983805668016,
"grad_norm": 0.2740287085637976,
"learning_rate": 1.2105339379592005e-05,
"loss": 0.4224,
"step": 823
},
{
"epoch": 3.3360323886639676,
"grad_norm": 0.27949318526774897,
"learning_rate": 1.2053409133511681e-05,
"loss": 0.4098,
"step": 824
},
{
"epoch": 3.340080971659919,
"grad_norm": 0.28010242600627566,
"learning_rate": 1.2001542428160173e-05,
"loss": 0.4078,
"step": 825
},
{
"epoch": 3.3441295546558703,
"grad_norm": 0.25302882784656755,
"learning_rate": 1.1949739678262278e-05,
"loss": 0.4151,
"step": 826
},
{
"epoch": 3.348178137651822,
"grad_norm": 0.29054996450024245,
"learning_rate": 1.189800129803139e-05,
"loss": 0.4163,
"step": 827
},
{
"epoch": 3.3522267206477734,
"grad_norm": 0.2531849978614158,
"learning_rate": 1.1846327701166228e-05,
"loss": 0.4222,
"step": 828
},
{
"epoch": 3.3562753036437245,
"grad_norm": 0.2701032130261065,
"learning_rate": 1.1794719300847493e-05,
"loss": 0.4147,
"step": 829
},
{
"epoch": 3.360323886639676,
"grad_norm": 0.2651747011118883,
"learning_rate": 1.1743176509734582e-05,
"loss": 0.4255,
"step": 830
},
{
"epoch": 3.3643724696356276,
"grad_norm": 0.24430483689083252,
"learning_rate": 1.1691699739962275e-05,
"loss": 0.4077,
"step": 831
},
{
"epoch": 3.3684210526315788,
"grad_norm": 0.24753902532612707,
"learning_rate": 1.1640289403137456e-05,
"loss": 0.4157,
"step": 832
},
{
"epoch": 3.3724696356275303,
"grad_norm": 0.23245111154622744,
"learning_rate": 1.1588945910335802e-05,
"loss": 0.4199,
"step": 833
},
{
"epoch": 3.376518218623482,
"grad_norm": 0.24251651592689563,
"learning_rate": 1.1537669672098517e-05,
"loss": 0.4183,
"step": 834
},
{
"epoch": 3.380566801619433,
"grad_norm": 0.24734557868660745,
"learning_rate": 1.1486461098429036e-05,
"loss": 0.4221,
"step": 835
},
{
"epoch": 3.3846153846153846,
"grad_norm": 0.24571766142272575,
"learning_rate": 1.1435320598789747e-05,
"loss": 0.4115,
"step": 836
},
{
"epoch": 3.388663967611336,
"grad_norm": 0.26030553759414854,
"learning_rate": 1.1384248582098736e-05,
"loss": 0.4143,
"step": 837
},
{
"epoch": 3.3927125506072873,
"grad_norm": 0.2320475266510005,
"learning_rate": 1.1333245456726467e-05,
"loss": 0.4286,
"step": 838
},
{
"epoch": 3.396761133603239,
"grad_norm": 0.25497494630184314,
"learning_rate": 1.128231163049258e-05,
"loss": 0.4394,
"step": 839
},
{
"epoch": 3.4008097165991904,
"grad_norm": 0.22753512227171388,
"learning_rate": 1.123144751066259e-05,
"loss": 0.4136,
"step": 840
},
{
"epoch": 3.4048582995951415,
"grad_norm": 0.23816240560422994,
"learning_rate": 1.1180653503944657e-05,
"loss": 0.4122,
"step": 841
},
{
"epoch": 3.408906882591093,
"grad_norm": 0.2607669142701146,
"learning_rate": 1.1129930016486287e-05,
"loss": 0.4083,
"step": 842
},
{
"epoch": 3.4129554655870447,
"grad_norm": 0.24902145651962673,
"learning_rate": 1.107927745387114e-05,
"loss": 0.4255,
"step": 843
},
{
"epoch": 3.417004048582996,
"grad_norm": 0.251390455723634,
"learning_rate": 1.1028696221115766e-05,
"loss": 0.4173,
"step": 844
},
{
"epoch": 3.4210526315789473,
"grad_norm": 0.2896314557461618,
"learning_rate": 1.0978186722666341e-05,
"loss": 0.3983,
"step": 845
},
{
"epoch": 3.425101214574899,
"grad_norm": 0.2547424248311804,
"learning_rate": 1.0927749362395475e-05,
"loss": 0.4078,
"step": 846
},
{
"epoch": 3.42914979757085,
"grad_norm": 0.2626210379805052,
"learning_rate": 1.0877384543598954e-05,
"loss": 0.4171,
"step": 847
},
{
"epoch": 3.4331983805668016,
"grad_norm": 0.25973258419134393,
"learning_rate": 1.0827092668992529e-05,
"loss": 0.4155,
"step": 848
},
{
"epoch": 3.437246963562753,
"grad_norm": 0.23079112307985866,
"learning_rate": 1.077687414070869e-05,
"loss": 0.4303,
"step": 849
},
{
"epoch": 3.4412955465587043,
"grad_norm": 0.24669016201905494,
"learning_rate": 1.0726729360293451e-05,
"loss": 0.4238,
"step": 850
},
{
"epoch": 3.445344129554656,
"grad_norm": 0.24841210248718254,
"learning_rate": 1.0676658728703147e-05,
"loss": 0.4154,
"step": 851
},
{
"epoch": 3.4493927125506074,
"grad_norm": 0.23633634267260414,
"learning_rate": 1.0626662646301198e-05,
"loss": 0.4237,
"step": 852
},
{
"epoch": 3.4534412955465585,
"grad_norm": 0.2512780528956535,
"learning_rate": 1.0576741512854958e-05,
"loss": 0.4084,
"step": 853
},
{
"epoch": 3.45748987854251,
"grad_norm": 0.23868439459918567,
"learning_rate": 1.052689572753247e-05,
"loss": 0.4135,
"step": 854
},
{
"epoch": 3.4615384615384617,
"grad_norm": 0.2595284777930487,
"learning_rate": 1.0477125688899328e-05,
"loss": 0.4256,
"step": 855
},
{
"epoch": 3.465587044534413,
"grad_norm": 0.2441679982591282,
"learning_rate": 1.0427431794915406e-05,
"loss": 0.4183,
"step": 856
},
{
"epoch": 3.4696356275303644,
"grad_norm": 0.24609652751432382,
"learning_rate": 1.0377814442931769e-05,
"loss": 0.4223,
"step": 857
},
{
"epoch": 3.473684210526316,
"grad_norm": 0.2596433687445063,
"learning_rate": 1.0328274029687439e-05,
"loss": 0.4195,
"step": 858
},
{
"epoch": 3.477732793522267,
"grad_norm": 0.24861451301251805,
"learning_rate": 1.0278810951306253e-05,
"loss": 0.4142,
"step": 859
},
{
"epoch": 3.4817813765182186,
"grad_norm": 0.26230753769013326,
"learning_rate": 1.0229425603293647e-05,
"loss": 0.4179,
"step": 860
},
{
"epoch": 3.48582995951417,
"grad_norm": 0.2605065705057161,
"learning_rate": 1.0180118380533558e-05,
"loss": 0.4241,
"step": 861
},
{
"epoch": 3.4898785425101213,
"grad_norm": 0.27113654084749794,
"learning_rate": 1.0130889677285225e-05,
"loss": 0.431,
"step": 862
},
{
"epoch": 3.493927125506073,
"grad_norm": 0.24455370182990083,
"learning_rate": 1.008173988718005e-05,
"loss": 0.4035,
"step": 863
},
{
"epoch": 3.4979757085020244,
"grad_norm": 0.25405502133564856,
"learning_rate": 1.0032669403218439e-05,
"loss": 0.4126,
"step": 864
},
{
"epoch": 3.5020242914979756,
"grad_norm": 0.26927545746079323,
"learning_rate": 9.983678617766677e-06,
"loss": 0.4193,
"step": 865
},
{
"epoch": 3.506072874493927,
"grad_norm": 0.24946720587455254,
"learning_rate": 9.934767922553788e-06,
"loss": 0.411,
"step": 866
},
{
"epoch": 3.5101214574898787,
"grad_norm": 0.271534167918417,
"learning_rate": 9.885937708668371e-06,
"loss": 0.4207,
"step": 867
},
{
"epoch": 3.51417004048583,
"grad_norm": 0.24623322248992674,
"learning_rate": 9.837188366555524e-06,
"loss": 0.4166,
"step": 868
},
{
"epoch": 3.5182186234817814,
"grad_norm": 0.27103807569234306,
"learning_rate": 9.788520286013691e-06,
"loss": 0.4117,
"step": 869
},
{
"epoch": 3.522267206477733,
"grad_norm": 0.2450984456677342,
"learning_rate": 9.739933856191537e-06,
"loss": 0.4063,
"step": 870
},
{
"epoch": 3.526315789473684,
"grad_norm": 0.25699461630977893,
"learning_rate": 9.691429465584866e-06,
"loss": 0.4222,
"step": 871
},
{
"epoch": 3.5303643724696356,
"grad_norm": 0.2420319715906887,
"learning_rate": 9.643007502033492e-06,
"loss": 0.429,
"step": 872
},
{
"epoch": 3.534412955465587,
"grad_norm": 0.2515732118598414,
"learning_rate": 9.594668352718152e-06,
"loss": 0.4044,
"step": 873
},
{
"epoch": 3.5384615384615383,
"grad_norm": 0.24391355312865912,
"learning_rate": 9.546412404157383e-06,
"loss": 0.4145,
"step": 874
},
{
"epoch": 3.54251012145749,
"grad_norm": 0.23114311814486418,
"learning_rate": 9.498240042204471e-06,
"loss": 0.4178,
"step": 875
},
{
"epoch": 3.5465587044534415,
"grad_norm": 0.23531970279842132,
"learning_rate": 9.450151652044339e-06,
"loss": 0.4167,
"step": 876
},
{
"epoch": 3.5506072874493926,
"grad_norm": 0.2539743672382336,
"learning_rate": 9.402147618190472e-06,
"loss": 0.3882,
"step": 877
},
{
"epoch": 3.554655870445344,
"grad_norm": 0.23971822246528093,
"learning_rate": 9.354228324481841e-06,
"loss": 0.4288,
"step": 878
},
{
"epoch": 3.5587044534412957,
"grad_norm": 0.24675601567590266,
"learning_rate": 9.306394154079852e-06,
"loss": 0.4293,
"step": 879
},
{
"epoch": 3.562753036437247,
"grad_norm": 0.23668154005748182,
"learning_rate": 9.258645489465254e-06,
"loss": 0.4141,
"step": 880
},
{
"epoch": 3.5668016194331984,
"grad_norm": 0.23078697851393587,
"learning_rate": 9.21098271243509e-06,
"loss": 0.4257,
"step": 881
},
{
"epoch": 3.57085020242915,
"grad_norm": 0.23336590598393003,
"learning_rate": 9.163406204099659e-06,
"loss": 0.4174,
"step": 882
},
{
"epoch": 3.574898785425101,
"grad_norm": 0.2441633633179718,
"learning_rate": 9.115916344879452e-06,
"loss": 0.4138,
"step": 883
},
{
"epoch": 3.5789473684210527,
"grad_norm": 0.22643630413744958,
"learning_rate": 9.068513514502126e-06,
"loss": 0.4048,
"step": 884
},
{
"epoch": 3.582995951417004,
"grad_norm": 0.24255427762570964,
"learning_rate": 9.021198091999439e-06,
"loss": 0.419,
"step": 885
},
{
"epoch": 3.5870445344129553,
"grad_norm": 0.2592368038358181,
"learning_rate": 8.973970455704251e-06,
"loss": 0.4127,
"step": 886
},
{
"epoch": 3.591093117408907,
"grad_norm": 0.22866576277222386,
"learning_rate": 8.926830983247495e-06,
"loss": 0.4089,
"step": 887
},
{
"epoch": 3.5951417004048585,
"grad_norm": 0.2559305078010084,
"learning_rate": 8.879780051555116e-06,
"loss": 0.4263,
"step": 888
},
{
"epoch": 3.5991902834008096,
"grad_norm": 0.2655117959801234,
"learning_rate": 8.83281803684512e-06,
"loss": 0.4289,
"step": 889
},
{
"epoch": 3.603238866396761,
"grad_norm": 0.23463836285416753,
"learning_rate": 8.785945314624532e-06,
"loss": 0.4249,
"step": 890
},
{
"epoch": 3.6072874493927127,
"grad_norm": 0.2546249381024098,
"learning_rate": 8.739162259686385e-06,
"loss": 0.4189,
"step": 891
},
{
"epoch": 3.611336032388664,
"grad_norm": 0.2429389429772761,
"learning_rate": 8.692469246106734e-06,
"loss": 0.4242,
"step": 892
},
{
"epoch": 3.6153846153846154,
"grad_norm": 0.24151162494893308,
"learning_rate": 8.645866647241677e-06,
"loss": 0.4082,
"step": 893
},
{
"epoch": 3.619433198380567,
"grad_norm": 0.22316119819656713,
"learning_rate": 8.599354835724348e-06,
"loss": 0.4134,
"step": 894
},
{
"epoch": 3.623481781376518,
"grad_norm": 0.24186279763904744,
"learning_rate": 8.552934183461958e-06,
"loss": 0.4086,
"step": 895
},
{
"epoch": 3.6275303643724697,
"grad_norm": 0.2364896518846595,
"learning_rate": 8.506605061632805e-06,
"loss": 0.4233,
"step": 896
},
{
"epoch": 3.6315789473684212,
"grad_norm": 0.23556063587034262,
"learning_rate": 8.460367840683308e-06,
"loss": 0.4208,
"step": 897
},
{
"epoch": 3.6356275303643724,
"grad_norm": 0.228912302849149,
"learning_rate": 8.414222890325065e-06,
"loss": 0.4178,
"step": 898
},
{
"epoch": 3.639676113360324,
"grad_norm": 0.2156053988945064,
"learning_rate": 8.368170579531855e-06,
"loss": 0.4322,
"step": 899
},
{
"epoch": 3.6437246963562755,
"grad_norm": 0.24446748254354342,
"learning_rate": 8.322211276536733e-06,
"loss": 0.4092,
"step": 900
},
{
"epoch": 3.6477732793522266,
"grad_norm": 0.24370635726571696,
"learning_rate": 8.27634534882907e-06,
"loss": 0.399,
"step": 901
},
{
"epoch": 3.651821862348178,
"grad_norm": 0.22290179902120083,
"learning_rate": 8.230573163151585e-06,
"loss": 0.4197,
"step": 902
},
{
"epoch": 3.6558704453441297,
"grad_norm": 0.22655339790224482,
"learning_rate": 8.184895085497466e-06,
"loss": 0.4137,
"step": 903
},
{
"epoch": 3.659919028340081,
"grad_norm": 0.234890492678875,
"learning_rate": 8.139311481107401e-06,
"loss": 0.4069,
"step": 904
},
{
"epoch": 3.6639676113360324,
"grad_norm": 0.22718601503235492,
"learning_rate": 8.093822714466686e-06,
"loss": 0.4189,
"step": 905
},
{
"epoch": 3.668016194331984,
"grad_norm": 0.2252150173249848,
"learning_rate": 8.048429149302271e-06,
"loss": 0.4218,
"step": 906
},
{
"epoch": 3.672064777327935,
"grad_norm": 0.21875176697682647,
"learning_rate": 8.003131148579902e-06,
"loss": 0.4157,
"step": 907
},
{
"epoch": 3.6761133603238867,
"grad_norm": 0.2237771797117402,
"learning_rate": 7.957929074501182e-06,
"loss": 0.4173,
"step": 908
},
{
"epoch": 3.6801619433198383,
"grad_norm": 0.23398876786775097,
"learning_rate": 7.912823288500695e-06,
"loss": 0.4104,
"step": 909
},
{
"epoch": 3.6842105263157894,
"grad_norm": 0.22744484060648107,
"learning_rate": 7.867814151243102e-06,
"loss": 0.4209,
"step": 910
},
{
"epoch": 3.688259109311741,
"grad_norm": 0.21925804398935855,
"learning_rate": 7.822902022620267e-06,
"loss": 0.4167,
"step": 911
},
{
"epoch": 3.6923076923076925,
"grad_norm": 0.23274433689396026,
"learning_rate": 7.778087261748377e-06,
"loss": 0.4089,
"step": 912
},
{
"epoch": 3.6963562753036436,
"grad_norm": 0.21977339032743595,
"learning_rate": 7.733370226965056e-06,
"loss": 0.4086,
"step": 913
},
{
"epoch": 3.700404858299595,
"grad_norm": 0.21530404749470256,
"learning_rate": 7.688751275826527e-06,
"loss": 0.417,
"step": 914
},
{
"epoch": 3.7044534412955468,
"grad_norm": 0.23336937995207252,
"learning_rate": 7.644230765104735e-06,
"loss": 0.4181,
"step": 915
},
{
"epoch": 3.708502024291498,
"grad_norm": 0.21930050730380016,
"learning_rate": 7.599809050784502e-06,
"loss": 0.4159,
"step": 916
},
{
"epoch": 3.7125506072874495,
"grad_norm": 0.21906598589156484,
"learning_rate": 7.555486488060661e-06,
"loss": 0.4109,
"step": 917
},
{
"epoch": 3.716599190283401,
"grad_norm": 0.2353333218729079,
"learning_rate": 7.511263431335249e-06,
"loss": 0.4232,
"step": 918
},
{
"epoch": 3.720647773279352,
"grad_norm": 0.21826066350064355,
"learning_rate": 7.467140234214656e-06,
"loss": 0.4077,
"step": 919
},
{
"epoch": 3.7246963562753037,
"grad_norm": 0.22565730330154843,
"learning_rate": 7.423117249506777e-06,
"loss": 0.4234,
"step": 920
},
{
"epoch": 3.7287449392712553,
"grad_norm": 0.22988768699490583,
"learning_rate": 7.379194829218231e-06,
"loss": 0.42,
"step": 921
},
{
"epoch": 3.7327935222672064,
"grad_norm": 0.22715262953640575,
"learning_rate": 7.335373324551518e-06,
"loss": 0.4173,
"step": 922
},
{
"epoch": 3.736842105263158,
"grad_norm": 0.2238530213460016,
"learning_rate": 7.291653085902224e-06,
"loss": 0.4042,
"step": 923
},
{
"epoch": 3.7408906882591095,
"grad_norm": 0.22653489739593397,
"learning_rate": 7.248034462856213e-06,
"loss": 0.4259,
"step": 924
},
{
"epoch": 3.7449392712550607,
"grad_norm": 0.23357142312516424,
"learning_rate": 7.20451780418683e-06,
"loss": 0.4121,
"step": 925
},
{
"epoch": 3.748987854251012,
"grad_norm": 0.21934953067693214,
"learning_rate": 7.161103457852128e-06,
"loss": 0.4266,
"step": 926
},
{
"epoch": 3.753036437246964,
"grad_norm": 0.22259575632880976,
"learning_rate": 7.117791770992046e-06,
"loss": 0.4082,
"step": 927
},
{
"epoch": 3.757085020242915,
"grad_norm": 0.22916374295577277,
"learning_rate": 7.074583089925688e-06,
"loss": 0.417,
"step": 928
},
{
"epoch": 3.7611336032388665,
"grad_norm": 0.22386793969848015,
"learning_rate": 7.0314777601485155e-06,
"loss": 0.4088,
"step": 929
},
{
"epoch": 3.765182186234818,
"grad_norm": 0.23086095310467006,
"learning_rate": 6.988476126329604e-06,
"loss": 0.4172,
"step": 930
},
{
"epoch": 3.769230769230769,
"grad_norm": 0.2344471096288637,
"learning_rate": 6.945578532308857e-06,
"loss": 0.4003,
"step": 931
},
{
"epoch": 3.7732793522267207,
"grad_norm": 0.2252777755821454,
"learning_rate": 6.902785321094301e-06,
"loss": 0.4039,
"step": 932
},
{
"epoch": 3.7773279352226723,
"grad_norm": 0.2150699343096909,
"learning_rate": 6.8600968348593115e-06,
"loss": 0.4004,
"step": 933
},
{
"epoch": 3.7813765182186234,
"grad_norm": 0.22099938391037616,
"learning_rate": 6.817513414939892e-06,
"loss": 0.4208,
"step": 934
},
{
"epoch": 3.785425101214575,
"grad_norm": 0.22158865420435883,
"learning_rate": 6.775035401831924e-06,
"loss": 0.42,
"step": 935
},
{
"epoch": 3.7894736842105265,
"grad_norm": 0.2169384255726406,
"learning_rate": 6.732663135188473e-06,
"loss": 0.4227,
"step": 936
},
{
"epoch": 3.7935222672064777,
"grad_norm": 0.2259364306985713,
"learning_rate": 6.690396953817055e-06,
"loss": 0.4246,
"step": 937
},
{
"epoch": 3.7975708502024292,
"grad_norm": 0.22221898270211024,
"learning_rate": 6.648237195676927e-06,
"loss": 0.4234,
"step": 938
},
{
"epoch": 3.801619433198381,
"grad_norm": 0.22390814790543945,
"learning_rate": 6.606184197876395e-06,
"loss": 0.3982,
"step": 939
},
{
"epoch": 3.805668016194332,
"grad_norm": 0.2263013377067371,
"learning_rate": 6.564238296670109e-06,
"loss": 0.4085,
"step": 940
},
{
"epoch": 3.8097165991902835,
"grad_norm": 0.24261270211581606,
"learning_rate": 6.52239982745638e-06,
"loss": 0.4261,
"step": 941
},
{
"epoch": 3.813765182186235,
"grad_norm": 0.21994251813293914,
"learning_rate": 6.480669124774484e-06,
"loss": 0.438,
"step": 942
},
{
"epoch": 3.817813765182186,
"grad_norm": 0.2522460235892635,
"learning_rate": 6.439046522302013e-06,
"loss": 0.4234,
"step": 943
},
{
"epoch": 3.8218623481781377,
"grad_norm": 0.25547503521348625,
"learning_rate": 6.397532352852191e-06,
"loss": 0.4189,
"step": 944
},
{
"epoch": 3.8259109311740893,
"grad_norm": 0.22909601522063638,
"learning_rate": 6.356126948371202e-06,
"loss": 0.4008,
"step": 945
},
{
"epoch": 3.8299595141700404,
"grad_norm": 0.22818754847891534,
"learning_rate": 6.314830639935559e-06,
"loss": 0.4086,
"step": 946
},
{
"epoch": 3.834008097165992,
"grad_norm": 0.23873560253941128,
"learning_rate": 6.273643757749448e-06,
"loss": 0.4176,
"step": 947
},
{
"epoch": 3.8380566801619436,
"grad_norm": 0.2290819940649273,
"learning_rate": 6.2325666311420856e-06,
"loss": 0.4124,
"step": 948
},
{
"epoch": 3.8421052631578947,
"grad_norm": 0.2277701915211981,
"learning_rate": 6.1915995885650675e-06,
"loss": 0.4002,
"step": 949
},
{
"epoch": 3.8461538461538463,
"grad_norm": 0.22666308456491996,
"learning_rate": 6.150742957589784e-06,
"loss": 0.4116,
"step": 950
},
{
"epoch": 3.850202429149798,
"grad_norm": 0.2718317474355078,
"learning_rate": 6.109997064904774e-06,
"loss": 0.4161,
"step": 951
},
{
"epoch": 3.854251012145749,
"grad_norm": 0.23166272149990216,
"learning_rate": 6.069362236313092e-06,
"loss": 0.4237,
"step": 952
},
{
"epoch": 3.8582995951417005,
"grad_norm": 0.2199305876486247,
"learning_rate": 6.0288387967297565e-06,
"loss": 0.4157,
"step": 953
},
{
"epoch": 3.862348178137652,
"grad_norm": 0.22741749158442268,
"learning_rate": 5.988427070179108e-06,
"loss": 0.4132,
"step": 954
},
{
"epoch": 3.866396761133603,
"grad_norm": 0.21702038867269932,
"learning_rate": 5.948127379792234e-06,
"loss": 0.3971,
"step": 955
},
{
"epoch": 3.8704453441295548,
"grad_norm": 0.2227842282062591,
"learning_rate": 5.90794004780439e-06,
"loss": 0.4254,
"step": 956
},
{
"epoch": 3.8744939271255063,
"grad_norm": 0.21600381453286988,
"learning_rate": 5.867865395552405e-06,
"loss": 0.4092,
"step": 957
},
{
"epoch": 3.8785425101214575,
"grad_norm": 0.22223510413561434,
"learning_rate": 5.8279037434721364e-06,
"loss": 0.4109,
"step": 958
},
{
"epoch": 3.882591093117409,
"grad_norm": 0.22250058444737714,
"learning_rate": 5.788055411095892e-06,
"loss": 0.4087,
"step": 959
},
{
"epoch": 3.8866396761133606,
"grad_norm": 0.22306949659865075,
"learning_rate": 5.7483207170498625e-06,
"loss": 0.4031,
"step": 960
},
{
"epoch": 3.8906882591093117,
"grad_norm": 0.24807363910034688,
"learning_rate": 5.708699979051613e-06,
"loss": 0.4011,
"step": 961
},
{
"epoch": 3.8947368421052633,
"grad_norm": 0.22385404644605364,
"learning_rate": 5.66919351390751e-06,
"loss": 0.4111,
"step": 962
},
{
"epoch": 3.898785425101215,
"grad_norm": 0.22636033767687344,
"learning_rate": 5.629801637510188e-06,
"loss": 0.3955,
"step": 963
},
{
"epoch": 3.902834008097166,
"grad_norm": 0.2406277808744602,
"learning_rate": 5.590524664836043e-06,
"loss": 0.4066,
"step": 964
},
{
"epoch": 3.9068825910931175,
"grad_norm": 0.2260732784861527,
"learning_rate": 5.551362909942708e-06,
"loss": 0.4012,
"step": 965
},
{
"epoch": 3.910931174089069,
"grad_norm": 0.21698654610229692,
"learning_rate": 5.51231668596653e-06,
"loss": 0.4002,
"step": 966
},
{
"epoch": 3.91497975708502,
"grad_norm": 0.22254337239614422,
"learning_rate": 5.47338630512007e-06,
"loss": 0.4172,
"step": 967
},
{
"epoch": 3.919028340080972,
"grad_norm": 0.24444669047442844,
"learning_rate": 5.434572078689617e-06,
"loss": 0.4177,
"step": 968
},
{
"epoch": 3.9230769230769234,
"grad_norm": 0.22003098737933954,
"learning_rate": 5.395874317032694e-06,
"loss": 0.4206,
"step": 969
},
{
"epoch": 3.9271255060728745,
"grad_norm": 0.20572455367865158,
"learning_rate": 5.357293329575573e-06,
"loss": 0.4152,
"step": 970
},
{
"epoch": 3.931174089068826,
"grad_norm": 0.2172037324046676,
"learning_rate": 5.318829424810797e-06,
"loss": 0.4095,
"step": 971
},
{
"epoch": 3.9352226720647776,
"grad_norm": 0.2372517571358378,
"learning_rate": 5.280482910294728e-06,
"loss": 0.4188,
"step": 972
},
{
"epoch": 3.9392712550607287,
"grad_norm": 0.21702693485466934,
"learning_rate": 5.242254092645078e-06,
"loss": 0.414,
"step": 973
},
{
"epoch": 3.9433198380566803,
"grad_norm": 0.21112340089676215,
"learning_rate": 5.2041432775384445e-06,
"loss": 0.4173,
"step": 974
},
{
"epoch": 3.9473684210526314,
"grad_norm": 0.21885483979428774,
"learning_rate": 5.166150769707892e-06,
"loss": 0.4269,
"step": 975
},
{
"epoch": 3.951417004048583,
"grad_norm": 0.2143540363413779,
"learning_rate": 5.128276872940503e-06,
"loss": 0.4155,
"step": 976
},
{
"epoch": 3.9554655870445345,
"grad_norm": 0.20872517397500437,
"learning_rate": 5.090521890074953e-06,
"loss": 0.4135,
"step": 977
},
{
"epoch": 3.9595141700404857,
"grad_norm": 0.2211568011718654,
"learning_rate": 5.0528861229990675e-06,
"loss": 0.4295,
"step": 978
},
{
"epoch": 3.9635627530364372,
"grad_norm": 0.21610671385045094,
"learning_rate": 5.015369872647442e-06,
"loss": 0.4227,
"step": 979
},
{
"epoch": 3.967611336032389,
"grad_norm": 0.2115769237808509,
"learning_rate": 4.9779734389990195e-06,
"loss": 0.423,
"step": 980
},
{
"epoch": 3.97165991902834,
"grad_norm": 0.21068879668095236,
"learning_rate": 4.940697121074675e-06,
"loss": 0.4023,
"step": 981
},
{
"epoch": 3.9757085020242915,
"grad_norm": 0.21391644592651823,
"learning_rate": 4.903541216934864e-06,
"loss": 0.4154,
"step": 982
},
{
"epoch": 3.979757085020243,
"grad_norm": 0.21026184043338414,
"learning_rate": 4.8665060236772e-06,
"loss": 0.4228,
"step": 983
},
{
"epoch": 3.983805668016194,
"grad_norm": 0.21582310694226428,
"learning_rate": 4.829591837434102e-06,
"loss": 0.4116,
"step": 984
},
{
"epoch": 3.9878542510121457,
"grad_norm": 0.2057023776800959,
"learning_rate": 4.792798953370424e-06,
"loss": 0.4119,
"step": 985
},
{
"epoch": 3.9919028340080973,
"grad_norm": 0.19615416800933466,
"learning_rate": 4.756127665681083e-06,
"loss": 0.423,
"step": 986
},
{
"epoch": 3.9959514170040484,
"grad_norm": 0.21086810311125248,
"learning_rate": 4.719578267588725e-06,
"loss": 0.4175,
"step": 987
},
{
"epoch": 4.0,
"grad_norm": 0.20876298379967884,
"learning_rate": 4.6831510513413525e-06,
"loss": 0.4099,
"step": 988
},
{
"epoch": 4.004048582995951,
"grad_norm": 0.42382247528907274,
"learning_rate": 4.6468463082100176e-06,
"loss": 0.3614,
"step": 989
},
{
"epoch": 4.008097165991903,
"grad_norm": 0.2647980583538397,
"learning_rate": 4.610664328486485e-06,
"loss": 0.3813,
"step": 990
},
{
"epoch": 4.012145748987854,
"grad_norm": 0.36125886294749343,
"learning_rate": 4.5746054014809e-06,
"loss": 0.3771,
"step": 991
},
{
"epoch": 4.016194331983805,
"grad_norm": 0.4601730296810761,
"learning_rate": 4.5386698155194766e-06,
"loss": 0.3705,
"step": 992
},
{
"epoch": 4.020242914979757,
"grad_norm": 0.299564887192682,
"learning_rate": 4.502857857942204e-06,
"loss": 0.3682,
"step": 993
},
{
"epoch": 4.0242914979757085,
"grad_norm": 0.36801470143355247,
"learning_rate": 4.467169815100549e-06,
"loss": 0.3735,
"step": 994
},
{
"epoch": 4.02834008097166,
"grad_norm": 0.4175963633232708,
"learning_rate": 4.431605972355137e-06,
"loss": 0.3625,
"step": 995
},
{
"epoch": 4.032388663967612,
"grad_norm": 0.2816130762808621,
"learning_rate": 4.396166614073516e-06,
"loss": 0.3819,
"step": 996
},
{
"epoch": 4.036437246963563,
"grad_norm": 0.24754716835190513,
"learning_rate": 4.360852023627846e-06,
"loss": 0.3656,
"step": 997
},
{
"epoch": 4.040485829959514,
"grad_norm": 0.3720742814661759,
"learning_rate": 4.325662483392656e-06,
"loss": 0.3896,
"step": 998
},
{
"epoch": 4.044534412955466,
"grad_norm": 0.3429438984885598,
"learning_rate": 4.29059827474257e-06,
"loss": 0.3633,
"step": 999
},
{
"epoch": 4.048582995951417,
"grad_norm": 0.2557075303682421,
"learning_rate": 4.255659678050063e-06,
"loss": 0.3647,
"step": 1000
},
{
"epoch": 4.052631578947368,
"grad_norm": 0.24054650429409322,
"learning_rate": 4.22084697268323e-06,
"loss": 0.377,
"step": 1001
},
{
"epoch": 4.05668016194332,
"grad_norm": 0.270404468025555,
"learning_rate": 4.186160437003535e-06,
"loss": 0.3703,
"step": 1002
},
{
"epoch": 4.060728744939271,
"grad_norm": 0.28994825893605725,
"learning_rate": 4.151600348363582e-06,
"loss": 0.3772,
"step": 1003
},
{
"epoch": 4.064777327935222,
"grad_norm": 0.24417280007240452,
"learning_rate": 4.11716698310493e-06,
"loss": 0.3635,
"step": 1004
},
{
"epoch": 4.068825910931174,
"grad_norm": 0.23377737251520686,
"learning_rate": 4.082860616555848e-06,
"loss": 0.3757,
"step": 1005
},
{
"epoch": 4.0728744939271255,
"grad_norm": 0.24876366263254315,
"learning_rate": 4.048681523029125e-06,
"loss": 0.3573,
"step": 1006
},
{
"epoch": 4.076923076923077,
"grad_norm": 0.26195395084863266,
"learning_rate": 4.014629975819882e-06,
"loss": 0.3747,
"step": 1007
},
{
"epoch": 4.080971659919029,
"grad_norm": 0.23407737127651548,
"learning_rate": 3.9807062472033855e-06,
"loss": 0.3719,
"step": 1008
},
{
"epoch": 4.08502024291498,
"grad_norm": 0.22039168777198792,
"learning_rate": 3.94691060843287e-06,
"loss": 0.3891,
"step": 1009
},
{
"epoch": 4.089068825910931,
"grad_norm": 0.21832379675562688,
"learning_rate": 3.913243329737353e-06,
"loss": 0.3776,
"step": 1010
},
{
"epoch": 4.093117408906883,
"grad_norm": 0.2237024215845808,
"learning_rate": 3.879704680319503e-06,
"loss": 0.3673,
"step": 1011
},
{
"epoch": 4.097165991902834,
"grad_norm": 0.2295103940503463,
"learning_rate": 3.846294928353471e-06,
"loss": 0.3693,
"step": 1012
},
{
"epoch": 4.101214574898785,
"grad_norm": 0.22364282977533975,
"learning_rate": 3.8130143409827345e-06,
"loss": 0.3565,
"step": 1013
},
{
"epoch": 4.105263157894737,
"grad_norm": 0.21926681810670198,
"learning_rate": 3.7798631843179887e-06,
"loss": 0.3618,
"step": 1014
},
{
"epoch": 4.109311740890688,
"grad_norm": 0.2138036870356754,
"learning_rate": 3.7468417234349973e-06,
"loss": 0.3728,
"step": 1015
},
{
"epoch": 4.113360323886639,
"grad_norm": 0.23531286843969537,
"learning_rate": 3.7139502223724823e-06,
"loss": 0.3816,
"step": 1016
},
{
"epoch": 4.117408906882591,
"grad_norm": 0.21749649287337416,
"learning_rate": 3.6811889441300095e-06,
"loss": 0.3672,
"step": 1017
},
{
"epoch": 4.1214574898785425,
"grad_norm": 0.21617019424568912,
"learning_rate": 3.648558150665886e-06,
"loss": 0.3836,
"step": 1018
},
{
"epoch": 4.125506072874494,
"grad_norm": 0.2186732571017664,
"learning_rate": 3.616058102895075e-06,
"loss": 0.3696,
"step": 1019
},
{
"epoch": 4.129554655870446,
"grad_norm": 0.22278750504420372,
"learning_rate": 3.583689060687079e-06,
"loss": 0.3766,
"step": 1020
},
{
"epoch": 4.133603238866397,
"grad_norm": 0.21368103310597453,
"learning_rate": 3.5514512828639023e-06,
"loss": 0.3756,
"step": 1021
},
{
"epoch": 4.137651821862348,
"grad_norm": 0.21713720535084663,
"learning_rate": 3.51934502719796e-06,
"loss": 0.3606,
"step": 1022
},
{
"epoch": 4.1417004048583,
"grad_norm": 0.20890779355741607,
"learning_rate": 3.4873705504100185e-06,
"loss": 0.3729,
"step": 1023
},
{
"epoch": 4.145748987854251,
"grad_norm": 0.21846086443669116,
"learning_rate": 3.4555281081671345e-06,
"loss": 0.3808,
"step": 1024
},
{
"epoch": 4.149797570850202,
"grad_norm": 0.21187133007153508,
"learning_rate": 3.4238179550806326e-06,
"loss": 0.3513,
"step": 1025
},
{
"epoch": 4.153846153846154,
"grad_norm": 0.20143862909236357,
"learning_rate": 3.3922403447040497e-06,
"loss": 0.3776,
"step": 1026
},
{
"epoch": 4.157894736842105,
"grad_norm": 0.2094187801511389,
"learning_rate": 3.3607955295311247e-06,
"loss": 0.377,
"step": 1027
},
{
"epoch": 4.161943319838056,
"grad_norm": 0.2064825292384927,
"learning_rate": 3.32948376099375e-06,
"loss": 0.3713,
"step": 1028
},
{
"epoch": 4.165991902834008,
"grad_norm": 0.20438977306432068,
"learning_rate": 3.2983052894599976e-06,
"loss": 0.3586,
"step": 1029
},
{
"epoch": 4.17004048582996,
"grad_norm": 0.20630811602404753,
"learning_rate": 3.2672603642320945e-06,
"loss": 0.3669,
"step": 1030
},
{
"epoch": 4.174089068825911,
"grad_norm": 0.2105977609797155,
"learning_rate": 3.236349233544438e-06,
"loss": 0.3762,
"step": 1031
},
{
"epoch": 4.178137651821863,
"grad_norm": 0.20750257272299724,
"learning_rate": 3.205572144561604e-06,
"loss": 0.3898,
"step": 1032
},
{
"epoch": 4.182186234817814,
"grad_norm": 0.21966572177527005,
"learning_rate": 3.174929343376374e-06,
"loss": 0.3675,
"step": 1033
},
{
"epoch": 4.186234817813765,
"grad_norm": 0.20730246669650135,
"learning_rate": 3.1444210750077777e-06,
"loss": 0.3665,
"step": 1034
},
{
"epoch": 4.190283400809717,
"grad_norm": 0.21520941301433122,
"learning_rate": 3.1140475833991046e-06,
"loss": 0.3578,
"step": 1035
},
{
"epoch": 4.194331983805668,
"grad_norm": 0.21398109580214383,
"learning_rate": 3.0838091114159895e-06,
"loss": 0.3772,
"step": 1036
},
{
"epoch": 4.198380566801619,
"grad_norm": 0.20938152982716873,
"learning_rate": 3.0537059008444524e-06,
"loss": 0.3664,
"step": 1037
},
{
"epoch": 4.202429149797571,
"grad_norm": 0.2041006016152618,
"learning_rate": 3.0237381923889563e-06,
"loss": 0.38,
"step": 1038
},
{
"epoch": 4.206477732793522,
"grad_norm": 0.2032398833028182,
"learning_rate": 2.9939062256704997e-06,
"loss": 0.3661,
"step": 1039
},
{
"epoch": 4.2105263157894735,
"grad_norm": 0.21238392774791082,
"learning_rate": 2.964210239224696e-06,
"loss": 0.3823,
"step": 1040
},
{
"epoch": 4.2145748987854255,
"grad_norm": 0.20630881660452502,
"learning_rate": 2.934650470499869e-06,
"loss": 0.3761,
"step": 1041
},
{
"epoch": 4.218623481781377,
"grad_norm": 0.21368324684698514,
"learning_rate": 2.905227155855128e-06,
"loss": 0.3657,
"step": 1042
},
{
"epoch": 4.222672064777328,
"grad_norm": 0.20564635285800928,
"learning_rate": 2.875940530558519e-06,
"loss": 0.3714,
"step": 1043
},
{
"epoch": 4.22672064777328,
"grad_norm": 0.20198084002405725,
"learning_rate": 2.8467908287851154e-06,
"loss": 0.372,
"step": 1044
},
{
"epoch": 4.230769230769231,
"grad_norm": 0.2066609075661131,
"learning_rate": 2.8177782836151603e-06,
"loss": 0.3654,
"step": 1045
},
{
"epoch": 4.234817813765182,
"grad_norm": 0.2072331941642048,
"learning_rate": 2.7889031270321874e-06,
"loss": 0.3663,
"step": 1046
},
{
"epoch": 4.238866396761134,
"grad_norm": 0.20406212990565367,
"learning_rate": 2.760165589921184e-06,
"loss": 0.3725,
"step": 1047
},
{
"epoch": 4.242914979757085,
"grad_norm": 0.22038935141402863,
"learning_rate": 2.7315659020667305e-06,
"loss": 0.3695,
"step": 1048
},
{
"epoch": 4.246963562753036,
"grad_norm": 0.20760045002599922,
"learning_rate": 2.703104292151164e-06,
"loss": 0.3697,
"step": 1049
},
{
"epoch": 4.251012145748988,
"grad_norm": 0.20948623428306098,
"learning_rate": 2.6747809877527607e-06,
"loss": 0.3727,
"step": 1050
},
{
"epoch": 4.255060728744939,
"grad_norm": 0.21232408862614754,
"learning_rate": 2.6465962153439082e-06,
"loss": 0.3561,
"step": 1051
},
{
"epoch": 4.2591093117408905,
"grad_norm": 0.20256422843816121,
"learning_rate": 2.618550200289303e-06,
"loss": 0.37,
"step": 1052
},
{
"epoch": 4.2631578947368425,
"grad_norm": 0.206951408660255,
"learning_rate": 2.59064316684412e-06,
"loss": 0.3561,
"step": 1053
},
{
"epoch": 4.267206477732794,
"grad_norm": 0.2025687312345178,
"learning_rate": 2.562875338152264e-06,
"loss": 0.3716,
"step": 1054
},
{
"epoch": 4.271255060728745,
"grad_norm": 0.20504856806724667,
"learning_rate": 2.5352469362445552e-06,
"loss": 0.3702,
"step": 1055
},
{
"epoch": 4.275303643724697,
"grad_norm": 0.20451532097638975,
"learning_rate": 2.5077581820369523e-06,
"loss": 0.3761,
"step": 1056
},
{
"epoch": 4.279352226720648,
"grad_norm": 0.23845007560415177,
"learning_rate": 2.480409295328803e-06,
"loss": 0.3498,
"step": 1057
},
{
"epoch": 4.283400809716599,
"grad_norm": 0.2003773559445591,
"learning_rate": 2.45320049480108e-06,
"loss": 0.3762,
"step": 1058
},
{
"epoch": 4.287449392712551,
"grad_norm": 0.21029138864425184,
"learning_rate": 2.4261319980146293e-06,
"loss": 0.3681,
"step": 1059
},
{
"epoch": 4.291497975708502,
"grad_norm": 0.20686763545952264,
"learning_rate": 2.399204021408421e-06,
"loss": 0.3668,
"step": 1060
},
{
"epoch": 4.295546558704453,
"grad_norm": 0.20686400907499028,
"learning_rate": 2.3724167802978395e-06,
"loss": 0.3746,
"step": 1061
},
{
"epoch": 4.299595141700405,
"grad_norm": 0.2084600265226674,
"learning_rate": 2.3457704888729625e-06,
"loss": 0.3632,
"step": 1062
},
{
"epoch": 4.303643724696356,
"grad_norm": 0.20954641513596403,
"learning_rate": 2.3192653601968162e-06,
"loss": 0.3669,
"step": 1063
},
{
"epoch": 4.3076923076923075,
"grad_norm": 0.2048518581164557,
"learning_rate": 2.2929016062037057e-06,
"loss": 0.3675,
"step": 1064
},
{
"epoch": 4.3117408906882595,
"grad_norm": 0.20430534556564137,
"learning_rate": 2.2666794376975055e-06,
"loss": 0.3633,
"step": 1065
},
{
"epoch": 4.315789473684211,
"grad_norm": 0.20882795278132418,
"learning_rate": 2.24059906434998e-06,
"loss": 0.3827,
"step": 1066
},
{
"epoch": 4.319838056680162,
"grad_norm": 0.2137339202844619,
"learning_rate": 2.214660694699091e-06,
"loss": 0.3729,
"step": 1067
},
{
"epoch": 4.323886639676114,
"grad_norm": 0.20506916948411802,
"learning_rate": 2.188864536147355e-06,
"loss": 0.3662,
"step": 1068
},
{
"epoch": 4.327935222672065,
"grad_norm": 0.20800124850547388,
"learning_rate": 2.1632107949601757e-06,
"loss": 0.3825,
"step": 1069
},
{
"epoch": 4.331983805668016,
"grad_norm": 0.20377720268928654,
"learning_rate": 2.137699676264171e-06,
"loss": 0.3756,
"step": 1070
},
{
"epoch": 4.336032388663968,
"grad_norm": 0.2065948545640369,
"learning_rate": 2.1123313840455738e-06,
"loss": 0.367,
"step": 1071
},
{
"epoch": 4.340080971659919,
"grad_norm": 0.21721075884986127,
"learning_rate": 2.08710612114857e-06,
"loss": 0.3627,
"step": 1072
},
{
"epoch": 4.34412955465587,
"grad_norm": 0.2232235525300807,
"learning_rate": 2.0620240892736953e-06,
"loss": 0.3757,
"step": 1073
},
{
"epoch": 4.348178137651822,
"grad_norm": 0.2042831105985488,
"learning_rate": 2.0370854889762e-06,
"loss": 0.3725,
"step": 1074
},
{
"epoch": 4.352226720647773,
"grad_norm": 0.2047474133858804,
"learning_rate": 2.0122905196644726e-06,
"loss": 0.3584,
"step": 1075
},
{
"epoch": 4.3562753036437245,
"grad_norm": 0.21085052993750636,
"learning_rate": 1.987639379598427e-06,
"loss": 0.3568,
"step": 1076
},
{
"epoch": 4.3603238866396765,
"grad_norm": 0.21299686688798514,
"learning_rate": 1.963132265887919e-06,
"loss": 0.3652,
"step": 1077
},
{
"epoch": 4.364372469635628,
"grad_norm": 0.2127514811464479,
"learning_rate": 1.9387693744911784e-06,
"loss": 0.3632,
"step": 1078
},
{
"epoch": 4.368421052631579,
"grad_norm": 0.1995076350229933,
"learning_rate": 1.9145509002132368e-06,
"loss": 0.3645,
"step": 1079
},
{
"epoch": 4.372469635627531,
"grad_norm": 0.20232614776589647,
"learning_rate": 1.8904770367043723e-06,
"loss": 0.3683,
"step": 1080
},
{
"epoch": 4.376518218623482,
"grad_norm": 0.20788700418873032,
"learning_rate": 1.8665479764585504e-06,
"loss": 0.3722,
"step": 1081
},
{
"epoch": 4.380566801619433,
"grad_norm": 0.19942474562975884,
"learning_rate": 1.8427639108118977e-06,
"loss": 0.3641,
"step": 1082
},
{
"epoch": 4.384615384615385,
"grad_norm": 0.20187318123239223,
"learning_rate": 1.8191250299411755e-06,
"loss": 0.3597,
"step": 1083
},
{
"epoch": 4.388663967611336,
"grad_norm": 0.20226567536790685,
"learning_rate": 1.7956315228622446e-06,
"loss": 0.3761,
"step": 1084
},
{
"epoch": 4.392712550607287,
"grad_norm": 0.20621478509748,
"learning_rate": 1.7722835774285553e-06,
"loss": 0.3719,
"step": 1085
},
{
"epoch": 4.396761133603239,
"grad_norm": 0.20890644802775507,
"learning_rate": 1.7490813803296624e-06,
"loss": 0.3495,
"step": 1086
},
{
"epoch": 4.40080971659919,
"grad_norm": 0.20348689882791837,
"learning_rate": 1.726025117089718e-06,
"loss": 0.3688,
"step": 1087
},
{
"epoch": 4.4048582995951415,
"grad_norm": 0.2016105775992098,
"learning_rate": 1.70311497206598e-06,
"loss": 0.3497,
"step": 1088
},
{
"epoch": 4.4089068825910935,
"grad_norm": 0.20870551971002985,
"learning_rate": 1.6803511284473638e-06,
"loss": 0.3711,
"step": 1089
},
{
"epoch": 4.412955465587045,
"grad_norm": 0.2065618811533438,
"learning_rate": 1.6577337682529538e-06,
"loss": 0.3629,
"step": 1090
},
{
"epoch": 4.417004048582996,
"grad_norm": 0.20895064642763853,
"learning_rate": 1.6352630723305617e-06,
"loss": 0.3614,
"step": 1091
},
{
"epoch": 4.421052631578947,
"grad_norm": 0.2061399623074591,
"learning_rate": 1.612939220355272e-06,
"loss": 0.3767,
"step": 1092
},
{
"epoch": 4.425101214574899,
"grad_norm": 0.20813914914922543,
"learning_rate": 1.5907623908280112e-06,
"loss": 0.3575,
"step": 1093
},
{
"epoch": 4.42914979757085,
"grad_norm": 0.20513842696161744,
"learning_rate": 1.568732761074121e-06,
"loss": 0.3778,
"step": 1094
},
{
"epoch": 4.433198380566802,
"grad_norm": 0.20169331780446426,
"learning_rate": 1.5468505072419237e-06,
"loss": 0.3658,
"step": 1095
},
{
"epoch": 4.437246963562753,
"grad_norm": 0.19890847056788485,
"learning_rate": 1.5251158043013448e-06,
"loss": 0.3679,
"step": 1096
},
{
"epoch": 4.441295546558704,
"grad_norm": 0.1969319065772915,
"learning_rate": 1.5035288260424885e-06,
"loss": 0.3644,
"step": 1097
},
{
"epoch": 4.445344129554655,
"grad_norm": 0.19436630501386296,
"learning_rate": 1.4820897450742623e-06,
"loss": 0.3727,
"step": 1098
},
{
"epoch": 4.449392712550607,
"grad_norm": 0.20730919871323428,
"learning_rate": 1.4607987328229767e-06,
"loss": 0.3732,
"step": 1099
},
{
"epoch": 4.4534412955465585,
"grad_norm": 0.19877791636374342,
"learning_rate": 1.4396559595310056e-06,
"loss": 0.3802,
"step": 1100
},
{
"epoch": 4.4574898785425106,
"grad_norm": 0.1975170198699071,
"learning_rate": 1.4186615942553973e-06,
"loss": 0.3819,
"step": 1101
},
{
"epoch": 4.461538461538462,
"grad_norm": 0.20798746144847985,
"learning_rate": 1.3978158048665424e-06,
"loss": 0.3773,
"step": 1102
},
{
"epoch": 4.465587044534413,
"grad_norm": 0.20794319931121036,
"learning_rate": 1.3771187580468115e-06,
"loss": 0.3757,
"step": 1103
},
{
"epoch": 4.469635627530364,
"grad_norm": 0.19622399921807476,
"learning_rate": 1.3565706192892392e-06,
"loss": 0.3778,
"step": 1104
},
{
"epoch": 4.473684210526316,
"grad_norm": 0.1935260218084233,
"learning_rate": 1.3361715528961993e-06,
"loss": 0.3739,
"step": 1105
},
{
"epoch": 4.477732793522267,
"grad_norm": 0.20270573326588914,
"learning_rate": 1.3159217219780817e-06,
"loss": 0.3787,
"step": 1106
},
{
"epoch": 4.481781376518219,
"grad_norm": 0.19795460137302578,
"learning_rate": 1.2958212884519949e-06,
"loss": 0.3569,
"step": 1107
},
{
"epoch": 4.48582995951417,
"grad_norm": 0.20134149652732045,
"learning_rate": 1.2758704130404675e-06,
"loss": 0.3674,
"step": 1108
},
{
"epoch": 4.489878542510121,
"grad_norm": 0.19860645051076123,
"learning_rate": 1.2560692552701738e-06,
"loss": 0.371,
"step": 1109
},
{
"epoch": 4.493927125506072,
"grad_norm": 0.2018742734722571,
"learning_rate": 1.2364179734706382e-06,
"loss": 0.3797,
"step": 1110
},
{
"epoch": 4.497975708502024,
"grad_norm": 0.20188479900313266,
"learning_rate": 1.2169167247729852e-06,
"loss": 0.3709,
"step": 1111
},
{
"epoch": 4.502024291497976,
"grad_norm": 0.20590209861879818,
"learning_rate": 1.1975656651086864e-06,
"loss": 0.3692,
"step": 1112
},
{
"epoch": 4.506072874493928,
"grad_norm": 0.19341124487081332,
"learning_rate": 1.1783649492082926e-06,
"loss": 0.3696,
"step": 1113
},
{
"epoch": 4.510121457489879,
"grad_norm": 0.20067311753908282,
"learning_rate": 1.1593147306002183e-06,
"loss": 0.3717,
"step": 1114
},
{
"epoch": 4.51417004048583,
"grad_norm": 0.203067374368333,
"learning_rate": 1.140415161609505e-06,
"loss": 0.3723,
"step": 1115
},
{
"epoch": 4.518218623481781,
"grad_norm": 0.20409726290616365,
"learning_rate": 1.121666393356604e-06,
"loss": 0.3804,
"step": 1116
},
{
"epoch": 4.522267206477733,
"grad_norm": 0.19661011327050654,
"learning_rate": 1.1030685757561632e-06,
"loss": 0.3761,
"step": 1117
},
{
"epoch": 4.526315789473684,
"grad_norm": 0.20463689524002315,
"learning_rate": 1.0846218575158373e-06,
"loss": 0.372,
"step": 1118
},
{
"epoch": 4.530364372469636,
"grad_norm": 0.19869580936073372,
"learning_rate": 1.066326386135097e-06,
"loss": 0.3917,
"step": 1119
},
{
"epoch": 4.534412955465587,
"grad_norm": 0.19791246552900404,
"learning_rate": 1.0481823079040398e-06,
"loss": 0.3797,
"step": 1120
},
{
"epoch": 4.538461538461538,
"grad_norm": 0.1961398227217507,
"learning_rate": 1.0301897679022344e-06,
"loss": 0.3738,
"step": 1121
},
{
"epoch": 4.5425101214574894,
"grad_norm": 0.20065937496323938,
"learning_rate": 1.0123489099975491e-06,
"loss": 0.3618,
"step": 1122
},
{
"epoch": 4.5465587044534415,
"grad_norm": 0.19581569963569823,
"learning_rate": 9.9465987684501e-07,
"loss": 0.3649,
"step": 1123
},
{
"epoch": 4.550607287449393,
"grad_norm": 0.19621250388341596,
"learning_rate": 9.771228098856534e-07,
"loss": 0.381,
"step": 1124
},
{
"epoch": 4.554655870445345,
"grad_norm": 0.19648887313044078,
"learning_rate": 9.597378493454013e-07,
"loss": 0.3623,
"step": 1125
},
{
"epoch": 4.558704453441296,
"grad_norm": 0.2103563448856451,
"learning_rate": 9.425051342339353e-07,
"loss": 0.3618,
"step": 1126
},
{
"epoch": 4.562753036437247,
"grad_norm": 0.20869335917303694,
"learning_rate": 9.254248023435864e-07,
"loss": 0.3651,
"step": 1127
},
{
"epoch": 4.566801619433198,
"grad_norm": 0.1914346567096459,
"learning_rate": 9.084969902482288e-07,
"loss": 0.3692,
"step": 1128
},
{
"epoch": 4.57085020242915,
"grad_norm": 0.20438623077567808,
"learning_rate": 8.917218333021993e-07,
"loss": 0.3944,
"step": 1129
},
{
"epoch": 4.574898785425101,
"grad_norm": 0.19003351212312056,
"learning_rate": 8.750994656392109e-07,
"loss": 0.375,
"step": 1130
},
{
"epoch": 4.578947368421053,
"grad_norm": 0.19566096743856556,
"learning_rate": 8.58630020171265e-07,
"loss": 0.3678,
"step": 1131
},
{
"epoch": 4.582995951417004,
"grad_norm": 0.20098579760445134,
"learning_rate": 8.423136285876099e-07,
"loss": 0.3627,
"step": 1132
},
{
"epoch": 4.587044534412955,
"grad_norm": 0.20254058965697289,
"learning_rate": 8.26150421353682e-07,
"loss": 0.367,
"step": 1133
},
{
"epoch": 4.5910931174089065,
"grad_norm": 0.19178356784653985,
"learning_rate": 8.101405277100549e-07,
"loss": 0.3755,
"step": 1134
},
{
"epoch": 4.5951417004048585,
"grad_norm": 0.1955989703859928,
"learning_rate": 7.942840756714077e-07,
"loss": 0.3598,
"step": 1135
},
{
"epoch": 4.59919028340081,
"grad_norm": 0.20122524026712538,
"learning_rate": 7.785811920255093e-07,
"loss": 0.3512,
"step": 1136
},
{
"epoch": 4.603238866396762,
"grad_norm": 0.1988983735050837,
"learning_rate": 7.630320023322024e-07,
"loss": 0.3552,
"step": 1137
},
{
"epoch": 4.607287449392713,
"grad_norm": 0.19284369028322598,
"learning_rate": 7.476366309223903e-07,
"loss": 0.3839,
"step": 1138
},
{
"epoch": 4.611336032388664,
"grad_norm": 0.19801647938746694,
"learning_rate": 7.323952008970537e-07,
"loss": 0.3564,
"step": 1139
},
{
"epoch": 4.615384615384615,
"grad_norm": 0.2005190124600589,
"learning_rate": 7.173078341262618e-07,
"loss": 0.3688,
"step": 1140
},
{
"epoch": 4.619433198380567,
"grad_norm": 0.19765861988870062,
"learning_rate": 7.023746512481989e-07,
"loss": 0.3545,
"step": 1141
},
{
"epoch": 4.623481781376518,
"grad_norm": 0.20184642036613484,
"learning_rate": 6.875957716681902e-07,
"loss": 0.3633,
"step": 1142
},
{
"epoch": 4.62753036437247,
"grad_norm": 0.19710332400647712,
"learning_rate": 6.729713135577687e-07,
"loss": 0.3693,
"step": 1143
},
{
"epoch": 4.631578947368421,
"grad_norm": 0.20356345349067242,
"learning_rate": 6.585013938537143e-07,
"loss": 0.3691,
"step": 1144
},
{
"epoch": 4.635627530364372,
"grad_norm": 0.19201082715412704,
"learning_rate": 6.441861282571115e-07,
"loss": 0.3724,
"step": 1145
},
{
"epoch": 4.6396761133603235,
"grad_norm": 0.1873390440876401,
"learning_rate": 6.300256312324471e-07,
"loss": 0.3764,
"step": 1146
},
{
"epoch": 4.6437246963562755,
"grad_norm": 0.19232568538999753,
"learning_rate": 6.160200160066776e-07,
"loss": 0.3832,
"step": 1147
},
{
"epoch": 4.647773279352227,
"grad_norm": 0.1974994003105234,
"learning_rate": 6.021693945683327e-07,
"loss": 0.3769,
"step": 1148
},
{
"epoch": 4.651821862348179,
"grad_norm": 0.2026352849556783,
"learning_rate": 5.884738776666044e-07,
"loss": 0.3608,
"step": 1149
},
{
"epoch": 4.65587044534413,
"grad_norm": 0.19744929818135215,
"learning_rate": 5.749335748104878e-07,
"loss": 0.3605,
"step": 1150
},
{
"epoch": 4.659919028340081,
"grad_norm": 0.19399785425317317,
"learning_rate": 5.615485942678778e-07,
"loss": 0.3742,
"step": 1151
},
{
"epoch": 4.663967611336032,
"grad_norm": 0.19858491594228006,
"learning_rate": 5.483190430647289e-07,
"loss": 0.3745,
"step": 1152
},
{
"epoch": 4.668016194331984,
"grad_norm": 0.19406594284388784,
"learning_rate": 5.352450269841747e-07,
"loss": 0.377,
"step": 1153
},
{
"epoch": 4.672064777327935,
"grad_norm": 0.2067308270239739,
"learning_rate": 5.22326650565701e-07,
"loss": 0.3667,
"step": 1154
},
{
"epoch": 4.676113360323887,
"grad_norm": 0.19678871912636448,
"learning_rate": 5.095640171043048e-07,
"loss": 0.38,
"step": 1155
},
{
"epoch": 4.680161943319838,
"grad_norm": 0.20362379342119655,
"learning_rate": 4.969572286496594e-07,
"loss": 0.3631,
"step": 1156
},
{
"epoch": 4.684210526315789,
"grad_norm": 0.1942510147845137,
"learning_rate": 4.845063860053079e-07,
"loss": 0.3829,
"step": 1157
},
{
"epoch": 4.6882591093117405,
"grad_norm": 0.19264596479782264,
"learning_rate": 4.7221158872785997e-07,
"loss": 0.3648,
"step": 1158
},
{
"epoch": 4.6923076923076925,
"grad_norm": 0.19395442848659555,
"learning_rate": 4.600729351261857e-07,
"loss": 0.3528,
"step": 1159
},
{
"epoch": 4.696356275303644,
"grad_norm": 0.19568984170686335,
"learning_rate": 4.4809052226063355e-07,
"loss": 0.3678,
"step": 1160
},
{
"epoch": 4.700404858299595,
"grad_norm": 0.19519652286873213,
"learning_rate": 4.3626444594226046e-07,
"loss": 0.3716,
"step": 1161
},
{
"epoch": 4.704453441295547,
"grad_norm": 0.19709870209900138,
"learning_rate": 4.245948007320588e-07,
"loss": 0.369,
"step": 1162
},
{
"epoch": 4.708502024291498,
"grad_norm": 0.218507296507631,
"learning_rate": 4.130816799401971e-07,
"loss": 0.3579,
"step": 1163
},
{
"epoch": 4.712550607287449,
"grad_norm": 0.19747215842193314,
"learning_rate": 4.017251756252827e-07,
"loss": 0.3633,
"step": 1164
},
{
"epoch": 4.716599190283401,
"grad_norm": 0.2020394363995586,
"learning_rate": 3.9052537859362473e-07,
"loss": 0.3571,
"step": 1165
},
{
"epoch": 4.720647773279352,
"grad_norm": 0.1958476926475864,
"learning_rate": 3.7948237839849687e-07,
"loss": 0.3791,
"step": 1166
},
{
"epoch": 4.724696356275303,
"grad_norm": 0.19964575343952434,
"learning_rate": 3.6859626333943797e-07,
"loss": 0.3862,
"step": 1167
},
{
"epoch": 4.728744939271255,
"grad_norm": 0.21296152234562446,
"learning_rate": 3.578671204615347e-07,
"loss": 0.3636,
"step": 1168
},
{
"epoch": 4.732793522267206,
"grad_norm": 0.19355355673979535,
"learning_rate": 3.472950355547267e-07,
"loss": 0.366,
"step": 1169
},
{
"epoch": 4.7368421052631575,
"grad_norm": 0.19931752101170477,
"learning_rate": 3.3688009315312244e-07,
"loss": 0.3732,
"step": 1170
},
{
"epoch": 4.7408906882591095,
"grad_norm": 0.1956768946587975,
"learning_rate": 3.266223765343224e-07,
"loss": 0.375,
"step": 1171
},
{
"epoch": 4.744939271255061,
"grad_norm": 0.19366262944720947,
"learning_rate": 3.16521967718757e-07,
"loss": 0.3793,
"step": 1172
},
{
"epoch": 4.748987854251012,
"grad_norm": 0.19208180646565629,
"learning_rate": 3.0657894746902726e-07,
"loss": 0.3795,
"step": 1173
},
{
"epoch": 4.753036437246964,
"grad_norm": 0.19251064951447538,
"learning_rate": 2.9679339528925654e-07,
"loss": 0.3769,
"step": 1174
},
{
"epoch": 4.757085020242915,
"grad_norm": 0.18706339911693598,
"learning_rate": 2.871653894244597e-07,
"loss": 0.3724,
"step": 1175
},
{
"epoch": 4.761133603238866,
"grad_norm": 0.19532675997898416,
"learning_rate": 2.7769500685991266e-07,
"loss": 0.3661,
"step": 1176
},
{
"epoch": 4.765182186234818,
"grad_norm": 0.20508169172379803,
"learning_rate": 2.683823233205485e-07,
"loss": 0.3589,
"step": 1177
},
{
"epoch": 4.769230769230769,
"grad_norm": 0.20617289993089233,
"learning_rate": 2.592274132703287e-07,
"loss": 0.3404,
"step": 1178
},
{
"epoch": 4.77327935222672,
"grad_norm": 0.1981575920969895,
"learning_rate": 2.502303499116754e-07,
"loss": 0.3702,
"step": 1179
},
{
"epoch": 4.777327935222672,
"grad_norm": 0.1994822112786611,
"learning_rate": 2.413912051848666e-07,
"loss": 0.359,
"step": 1180
},
{
"epoch": 4.781376518218623,
"grad_norm": 0.19495734614484614,
"learning_rate": 2.3271004976746613e-07,
"loss": 0.3704,
"step": 1181
},
{
"epoch": 4.7854251012145745,
"grad_norm": 0.19816049467840358,
"learning_rate": 2.2418695307376393e-07,
"loss": 0.3731,
"step": 1182
},
{
"epoch": 4.7894736842105265,
"grad_norm": 0.19092744729356495,
"learning_rate": 2.1582198325420968e-07,
"loss": 0.3622,
"step": 1183
},
{
"epoch": 4.793522267206478,
"grad_norm": 0.1965915358954332,
"learning_rate": 2.0761520719488447e-07,
"loss": 0.3724,
"step": 1184
},
{
"epoch": 4.797570850202429,
"grad_norm": 0.19139307473287062,
"learning_rate": 1.9956669051694798e-07,
"loss": 0.3809,
"step": 1185
},
{
"epoch": 4.801619433198381,
"grad_norm": 0.2098425737568775,
"learning_rate": 1.916764975761254e-07,
"loss": 0.3548,
"step": 1186
},
{
"epoch": 4.805668016194332,
"grad_norm": 0.20297660954735577,
"learning_rate": 1.839446914621923e-07,
"loss": 0.3741,
"step": 1187
},
{
"epoch": 4.809716599190283,
"grad_norm": 0.19898259017118627,
"learning_rate": 1.7637133399845962e-07,
"loss": 0.3506,
"step": 1188
},
{
"epoch": 4.813765182186235,
"grad_norm": 0.1920981150858864,
"learning_rate": 1.6895648574129397e-07,
"loss": 0.3721,
"step": 1189
},
{
"epoch": 4.817813765182186,
"grad_norm": 0.1920893806219909,
"learning_rate": 1.6170020597962465e-07,
"loss": 0.3598,
"step": 1190
},
{
"epoch": 4.821862348178137,
"grad_norm": 0.2014662207035411,
"learning_rate": 1.5460255273447078e-07,
"loss": 0.3719,
"step": 1191
},
{
"epoch": 4.825910931174089,
"grad_norm": 0.1929117331204443,
"learning_rate": 1.4766358275847715e-07,
"loss": 0.3578,
"step": 1192
},
{
"epoch": 4.82995951417004,
"grad_norm": 0.1934181109620901,
"learning_rate": 1.4088335153546573e-07,
"loss": 0.3598,
"step": 1193
},
{
"epoch": 4.834008097165992,
"grad_norm": 0.19199246636016315,
"learning_rate": 1.3426191327998496e-07,
"loss": 0.3653,
"step": 1194
},
{
"epoch": 4.838056680161944,
"grad_norm": 0.18900013576628835,
"learning_rate": 1.277993209368744e-07,
"loss": 0.363,
"step": 1195
},
{
"epoch": 4.842105263157895,
"grad_norm": 0.19587800584473747,
"learning_rate": 1.214956261808542e-07,
"loss": 0.354,
"step": 1196
},
{
"epoch": 4.846153846153846,
"grad_norm": 0.2006841177277032,
"learning_rate": 1.1535087941609846e-07,
"loss": 0.3716,
"step": 1197
},
{
"epoch": 4.850202429149798,
"grad_norm": 0.1936553128558643,
"learning_rate": 1.0936512977583802e-07,
"loss": 0.3772,
"step": 1198
},
{
"epoch": 4.854251012145749,
"grad_norm": 0.19214093069102606,
"learning_rate": 1.0353842512196732e-07,
"loss": 0.37,
"step": 1199
},
{
"epoch": 4.8582995951417,
"grad_norm": 0.18906140174212893,
"learning_rate": 9.787081204466253e-08,
"loss": 0.3757,
"step": 1200
},
{
"epoch": 4.862348178137652,
"grad_norm": 0.2048180006927712,
"learning_rate": 9.236233586200627e-08,
"loss": 0.355,
"step": 1201
},
{
"epoch": 4.866396761133603,
"grad_norm": 0.19140390940708693,
"learning_rate": 8.70130406196279e-08,
"loss": 0.3792,
"step": 1202
},
{
"epoch": 4.870445344129554,
"grad_norm": 0.1907015625400169,
"learning_rate": 8.182296909035047e-08,
"loss": 0.3815,
"step": 1203
},
{
"epoch": 4.874493927125506,
"grad_norm": 0.19084531526423584,
"learning_rate": 7.679216277384439e-08,
"loss": 0.3714,
"step": 1204
},
{
"epoch": 4.8785425101214575,
"grad_norm": 0.19659654260975726,
"learning_rate": 7.19206618963053e-08,
"loss": 0.3722,
"step": 1205
},
{
"epoch": 4.882591093117409,
"grad_norm": 0.1936782521161512,
"learning_rate": 6.720850541012347e-08,
"loss": 0.3591,
"step": 1206
},
{
"epoch": 4.886639676113361,
"grad_norm": 0.19628024337106623,
"learning_rate": 6.265573099357714e-08,
"loss": 0.3528,
"step": 1207
},
{
"epoch": 4.890688259109312,
"grad_norm": 0.21410954379406474,
"learning_rate": 5.826237505053067e-08,
"loss": 0.3822,
"step": 1208
},
{
"epoch": 4.894736842105263,
"grad_norm": 0.19588495198406605,
"learning_rate": 5.4028472710137e-08,
"loss": 0.3775,
"step": 1209
},
{
"epoch": 4.898785425101215,
"grad_norm": 0.20322891501303425,
"learning_rate": 4.9954057826571146e-08,
"loss": 0.36,
"step": 1210
},
{
"epoch": 4.902834008097166,
"grad_norm": 0.18914648305863327,
"learning_rate": 4.6039162978743776e-08,
"loss": 0.3851,
"step": 1211
},
{
"epoch": 4.906882591093117,
"grad_norm": 0.19816692352704074,
"learning_rate": 4.228381947004812e-08,
"loss": 0.3596,
"step": 1212
},
{
"epoch": 4.910931174089069,
"grad_norm": 0.196771030804237,
"learning_rate": 3.868805732811343e-08,
"loss": 0.3733,
"step": 1213
},
{
"epoch": 4.91497975708502,
"grad_norm": 0.19077579266048533,
"learning_rate": 3.525190530455636e-08,
"loss": 0.3842,
"step": 1214
},
{
"epoch": 4.919028340080971,
"grad_norm": 0.19763262924702132,
"learning_rate": 3.197539087475887e-08,
"loss": 0.3721,
"step": 1215
},
{
"epoch": 4.923076923076923,
"grad_norm": 0.19129829158851547,
"learning_rate": 2.8858540237641786e-08,
"loss": 0.3777,
"step": 1216
},
{
"epoch": 4.9271255060728745,
"grad_norm": 0.19006854997005546,
"learning_rate": 2.5901378315458247e-08,
"loss": 0.3739,
"step": 1217
},
{
"epoch": 4.931174089068826,
"grad_norm": 0.19406946216244375,
"learning_rate": 2.3103928753598347e-08,
"loss": 0.3791,
"step": 1218
},
{
"epoch": 4.935222672064778,
"grad_norm": 0.1957330048010809,
"learning_rate": 2.0466213920393718e-08,
"loss": 0.3729,
"step": 1219
},
{
"epoch": 4.939271255060729,
"grad_norm": 0.19749934046581955,
"learning_rate": 1.7988254906942115e-08,
"loss": 0.355,
"step": 1220
},
{
"epoch": 4.94331983805668,
"grad_norm": 0.19229535001888162,
"learning_rate": 1.5670071526936447e-08,
"loss": 0.3693,
"step": 1221
},
{
"epoch": 4.947368421052632,
"grad_norm": 0.1926571054365805,
"learning_rate": 1.3511682316509344e-08,
"loss": 0.3724,
"step": 1222
},
{
"epoch": 4.951417004048583,
"grad_norm": 0.1999301812883294,
"learning_rate": 1.1513104534086605e-08,
"loss": 0.3822,
"step": 1223
},
{
"epoch": 4.955465587044534,
"grad_norm": 0.18890173970690227,
"learning_rate": 9.67435416023843e-09,
"loss": 0.3651,
"step": 1224
},
{
"epoch": 4.959514170040486,
"grad_norm": 0.18861778939134502,
"learning_rate": 7.995445897566178e-09,
"loss": 0.3736,
"step": 1225
},
{
"epoch": 4.963562753036437,
"grad_norm": 0.19262581133791762,
"learning_rate": 6.476393170578021e-09,
"loss": 0.3628,
"step": 1226
},
{
"epoch": 4.967611336032388,
"grad_norm": 0.20665075032219152,
"learning_rate": 5.11720812557348e-09,
"loss": 0.3608,
"step": 1227
},
{
"epoch": 4.97165991902834,
"grad_norm": 0.19674315697100886,
"learning_rate": 3.917901630563492e-09,
"loss": 0.3606,
"step": 1228
},
{
"epoch": 4.9757085020242915,
"grad_norm": 0.19915981854467507,
"learning_rate": 2.878483275170485e-09,
"loss": 0.3642,
"step": 1229
},
{
"epoch": 4.979757085020243,
"grad_norm": 0.1957918958060553,
"learning_rate": 1.998961370555108e-09,
"loss": 0.3709,
"step": 1230
},
{
"epoch": 4.983805668016195,
"grad_norm": 0.19472870610287385,
"learning_rate": 1.2793429493518361e-09,
"loss": 0.3824,
"step": 1231
},
{
"epoch": 4.987854251012146,
"grad_norm": 0.20300721920047407,
"learning_rate": 7.196337656112384e-10,
"loss": 0.3579,
"step": 1232
},
{
"epoch": 4.991902834008097,
"grad_norm": 0.19863037256828017,
"learning_rate": 3.198382947511292e-10,
"loss": 0.3764,
"step": 1233
},
{
"epoch": 4.995951417004049,
"grad_norm": 0.19852859534470801,
"learning_rate": 7.995973352770137e-11,
"loss": 0.3689,
"step": 1234
},
{
"epoch": 5.0,
"grad_norm": 0.19428581762480948,
"learning_rate": 0.0,
"loss": 0.3535,
"step": 1235
},
{
"epoch": 5.0,
"step": 1235,
"total_flos": 5.150940430581367e+18,
"train_loss": 0.4971147541816418,
"train_runtime": 29485.7262,
"train_samples_per_second": 5.359,
"train_steps_per_second": 0.042
}
],
"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": 5.150940430581367e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}