e1_science_longest_phi / trainer_state.json
neginr's picture
End of training
da5f580 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": 6.956859587512121,
"learning_rate": 3.2258064516129035e-07,
"loss": 1.8148,
"step": 1
},
{
"epoch": 0.008097165991902834,
"grad_norm": 7.060793510413111,
"learning_rate": 6.451612903225807e-07,
"loss": 1.8261,
"step": 2
},
{
"epoch": 0.012145748987854251,
"grad_norm": 7.215767445333279,
"learning_rate": 9.67741935483871e-07,
"loss": 1.8512,
"step": 3
},
{
"epoch": 0.016194331983805668,
"grad_norm": 7.048626957148214,
"learning_rate": 1.2903225806451614e-06,
"loss": 1.8513,
"step": 4
},
{
"epoch": 0.020242914979757085,
"grad_norm": 6.663990610624765,
"learning_rate": 1.6129032258064516e-06,
"loss": 1.863,
"step": 5
},
{
"epoch": 0.024291497975708502,
"grad_norm": 6.36864434388696,
"learning_rate": 1.935483870967742e-06,
"loss": 1.8127,
"step": 6
},
{
"epoch": 0.02834008097165992,
"grad_norm": 5.11966830711861,
"learning_rate": 2.2580645161290324e-06,
"loss": 1.7847,
"step": 7
},
{
"epoch": 0.032388663967611336,
"grad_norm": 4.6538132654596565,
"learning_rate": 2.580645161290323e-06,
"loss": 1.7768,
"step": 8
},
{
"epoch": 0.03643724696356275,
"grad_norm": 2.486578977913623,
"learning_rate": 2.903225806451613e-06,
"loss": 1.752,
"step": 9
},
{
"epoch": 0.04048582995951417,
"grad_norm": 2.279483768281012,
"learning_rate": 3.225806451612903e-06,
"loss": 1.7295,
"step": 10
},
{
"epoch": 0.044534412955465584,
"grad_norm": 2.1336445057745363,
"learning_rate": 3.548387096774194e-06,
"loss": 1.7149,
"step": 11
},
{
"epoch": 0.048582995951417005,
"grad_norm": 4.421847478952836,
"learning_rate": 3.870967741935484e-06,
"loss": 1.7109,
"step": 12
},
{
"epoch": 0.05263157894736842,
"grad_norm": 4.286597269811606,
"learning_rate": 4.193548387096774e-06,
"loss": 1.7143,
"step": 13
},
{
"epoch": 0.05668016194331984,
"grad_norm": 4.130539929178081,
"learning_rate": 4.516129032258065e-06,
"loss": 1.7036,
"step": 14
},
{
"epoch": 0.06072874493927125,
"grad_norm": 3.9093968153472316,
"learning_rate": 4.838709677419355e-06,
"loss": 1.7243,
"step": 15
},
{
"epoch": 0.06477732793522267,
"grad_norm": 2.4399950237954813,
"learning_rate": 5.161290322580646e-06,
"loss": 1.6593,
"step": 16
},
{
"epoch": 0.06882591093117409,
"grad_norm": 2.105489275078297,
"learning_rate": 5.483870967741935e-06,
"loss": 1.6642,
"step": 17
},
{
"epoch": 0.0728744939271255,
"grad_norm": 1.80192553797292,
"learning_rate": 5.806451612903226e-06,
"loss": 1.6556,
"step": 18
},
{
"epoch": 0.07692307692307693,
"grad_norm": 1.494597878209307,
"learning_rate": 6.129032258064517e-06,
"loss": 1.5945,
"step": 19
},
{
"epoch": 0.08097165991902834,
"grad_norm": 1.205067068469863,
"learning_rate": 6.451612903225806e-06,
"loss": 1.6013,
"step": 20
},
{
"epoch": 0.08502024291497975,
"grad_norm": 1.1829785062838,
"learning_rate": 6.774193548387097e-06,
"loss": 1.5957,
"step": 21
},
{
"epoch": 0.08906882591093117,
"grad_norm": 1.3126451966249633,
"learning_rate": 7.096774193548388e-06,
"loss": 1.6091,
"step": 22
},
{
"epoch": 0.0931174089068826,
"grad_norm": 1.1472671715087541,
"learning_rate": 7.4193548387096784e-06,
"loss": 1.6,
"step": 23
},
{
"epoch": 0.09716599190283401,
"grad_norm": 1.0054631475399167,
"learning_rate": 7.741935483870968e-06,
"loss": 1.5468,
"step": 24
},
{
"epoch": 0.10121457489878542,
"grad_norm": 1.076646014901512,
"learning_rate": 8.064516129032258e-06,
"loss": 1.6108,
"step": 25
},
{
"epoch": 0.10526315789473684,
"grad_norm": 1.0556965341098696,
"learning_rate": 8.387096774193549e-06,
"loss": 1.5927,
"step": 26
},
{
"epoch": 0.10931174089068826,
"grad_norm": 1.2294162219663367,
"learning_rate": 8.70967741935484e-06,
"loss": 1.5637,
"step": 27
},
{
"epoch": 0.11336032388663968,
"grad_norm": 1.0783215073233208,
"learning_rate": 9.03225806451613e-06,
"loss": 1.5022,
"step": 28
},
{
"epoch": 0.11740890688259109,
"grad_norm": 0.8927443063970538,
"learning_rate": 9.35483870967742e-06,
"loss": 1.5714,
"step": 29
},
{
"epoch": 0.1214574898785425,
"grad_norm": 1.0858465822787713,
"learning_rate": 9.67741935483871e-06,
"loss": 1.558,
"step": 30
},
{
"epoch": 0.12550607287449392,
"grad_norm": 0.9734124809141845,
"learning_rate": 1e-05,
"loss": 1.535,
"step": 31
},
{
"epoch": 0.12955465587044535,
"grad_norm": 0.8346812839666283,
"learning_rate": 1.0322580645161291e-05,
"loss": 1.5181,
"step": 32
},
{
"epoch": 0.13360323886639677,
"grad_norm": 0.71205244222026,
"learning_rate": 1.0645161290322582e-05,
"loss": 1.531,
"step": 33
},
{
"epoch": 0.13765182186234817,
"grad_norm": 0.8849676047790562,
"learning_rate": 1.096774193548387e-05,
"loss": 1.5058,
"step": 34
},
{
"epoch": 0.1417004048582996,
"grad_norm": 0.8675322195240643,
"learning_rate": 1.1290322580645164e-05,
"loss": 1.4684,
"step": 35
},
{
"epoch": 0.145748987854251,
"grad_norm": 0.707844870810787,
"learning_rate": 1.1612903225806453e-05,
"loss": 1.5288,
"step": 36
},
{
"epoch": 0.14979757085020243,
"grad_norm": 0.7926652797866107,
"learning_rate": 1.1935483870967743e-05,
"loss": 1.4971,
"step": 37
},
{
"epoch": 0.15384615384615385,
"grad_norm": 0.8989613992935699,
"learning_rate": 1.2258064516129034e-05,
"loss": 1.521,
"step": 38
},
{
"epoch": 0.15789473684210525,
"grad_norm": 0.7096177665336705,
"learning_rate": 1.2580645161290324e-05,
"loss": 1.5604,
"step": 39
},
{
"epoch": 0.16194331983805668,
"grad_norm": 0.7177226086762879,
"learning_rate": 1.2903225806451613e-05,
"loss": 1.5117,
"step": 40
},
{
"epoch": 0.1659919028340081,
"grad_norm": 0.8195006177256143,
"learning_rate": 1.3225806451612903e-05,
"loss": 1.5245,
"step": 41
},
{
"epoch": 0.1700404858299595,
"grad_norm": 0.7162980071460926,
"learning_rate": 1.3548387096774194e-05,
"loss": 1.5069,
"step": 42
},
{
"epoch": 0.17408906882591094,
"grad_norm": 0.8663719257051337,
"learning_rate": 1.3870967741935486e-05,
"loss": 1.5197,
"step": 43
},
{
"epoch": 0.17813765182186234,
"grad_norm": 0.6577024537941576,
"learning_rate": 1.4193548387096776e-05,
"loss": 1.4687,
"step": 44
},
{
"epoch": 0.18218623481781376,
"grad_norm": 0.9361573165787114,
"learning_rate": 1.4516129032258066e-05,
"loss": 1.4851,
"step": 45
},
{
"epoch": 0.1862348178137652,
"grad_norm": 0.6781685937245396,
"learning_rate": 1.4838709677419357e-05,
"loss": 1.5004,
"step": 46
},
{
"epoch": 0.1902834008097166,
"grad_norm": 0.8818349752394763,
"learning_rate": 1.5161290322580646e-05,
"loss": 1.4905,
"step": 47
},
{
"epoch": 0.19433198380566802,
"grad_norm": 0.7159677779294384,
"learning_rate": 1.5483870967741936e-05,
"loss": 1.4854,
"step": 48
},
{
"epoch": 0.19838056680161945,
"grad_norm": 0.6454198511753046,
"learning_rate": 1.5806451612903226e-05,
"loss": 1.5198,
"step": 49
},
{
"epoch": 0.20242914979757085,
"grad_norm": 0.5970834832331787,
"learning_rate": 1.6129032258064517e-05,
"loss": 1.4602,
"step": 50
},
{
"epoch": 0.20647773279352227,
"grad_norm": 0.6677523359852976,
"learning_rate": 1.6451612903225807e-05,
"loss": 1.4991,
"step": 51
},
{
"epoch": 0.21052631578947367,
"grad_norm": 0.6388403898305072,
"learning_rate": 1.6774193548387098e-05,
"loss": 1.4816,
"step": 52
},
{
"epoch": 0.2145748987854251,
"grad_norm": 0.7853787833015233,
"learning_rate": 1.7096774193548388e-05,
"loss": 1.4578,
"step": 53
},
{
"epoch": 0.21862348178137653,
"grad_norm": 0.5855857407765253,
"learning_rate": 1.741935483870968e-05,
"loss": 1.5005,
"step": 54
},
{
"epoch": 0.22267206477732793,
"grad_norm": 0.5976034527834093,
"learning_rate": 1.774193548387097e-05,
"loss": 1.5234,
"step": 55
},
{
"epoch": 0.22672064777327935,
"grad_norm": 0.9145371573104308,
"learning_rate": 1.806451612903226e-05,
"loss": 1.4912,
"step": 56
},
{
"epoch": 0.23076923076923078,
"grad_norm": 0.7345993696865815,
"learning_rate": 1.838709677419355e-05,
"loss": 1.5002,
"step": 57
},
{
"epoch": 0.23481781376518218,
"grad_norm": 0.5974259167020819,
"learning_rate": 1.870967741935484e-05,
"loss": 1.4765,
"step": 58
},
{
"epoch": 0.2388663967611336,
"grad_norm": 0.7246925807524477,
"learning_rate": 1.903225806451613e-05,
"loss": 1.4631,
"step": 59
},
{
"epoch": 0.242914979757085,
"grad_norm": 0.6686845806856508,
"learning_rate": 1.935483870967742e-05,
"loss": 1.4905,
"step": 60
},
{
"epoch": 0.24696356275303644,
"grad_norm": 0.6464322678592415,
"learning_rate": 1.967741935483871e-05,
"loss": 1.4546,
"step": 61
},
{
"epoch": 0.25101214574898784,
"grad_norm": 0.8180276524871607,
"learning_rate": 2e-05,
"loss": 1.4036,
"step": 62
},
{
"epoch": 0.2550607287449393,
"grad_norm": 0.7184441287980796,
"learning_rate": 2.0322580645161292e-05,
"loss": 1.4988,
"step": 63
},
{
"epoch": 0.2591093117408907,
"grad_norm": 0.7105733541196354,
"learning_rate": 2.0645161290322582e-05,
"loss": 1.4565,
"step": 64
},
{
"epoch": 0.2631578947368421,
"grad_norm": 0.9131515923724282,
"learning_rate": 2.0967741935483873e-05,
"loss": 1.5024,
"step": 65
},
{
"epoch": 0.26720647773279355,
"grad_norm": 0.8748271662494419,
"learning_rate": 2.1290322580645163e-05,
"loss": 1.4858,
"step": 66
},
{
"epoch": 0.27125506072874495,
"grad_norm": 0.7267931450760191,
"learning_rate": 2.161290322580645e-05,
"loss": 1.4913,
"step": 67
},
{
"epoch": 0.27530364372469635,
"grad_norm": 0.7336137065705268,
"learning_rate": 2.193548387096774e-05,
"loss": 1.4424,
"step": 68
},
{
"epoch": 0.2793522267206478,
"grad_norm": 0.7835771045274686,
"learning_rate": 2.225806451612903e-05,
"loss": 1.4627,
"step": 69
},
{
"epoch": 0.2834008097165992,
"grad_norm": 0.7803487971770654,
"learning_rate": 2.2580645161290328e-05,
"loss": 1.5023,
"step": 70
},
{
"epoch": 0.2874493927125506,
"grad_norm": 0.5920262117156554,
"learning_rate": 2.2903225806451615e-05,
"loss": 1.5133,
"step": 71
},
{
"epoch": 0.291497975708502,
"grad_norm": 0.9809560785514037,
"learning_rate": 2.3225806451612906e-05,
"loss": 1.4899,
"step": 72
},
{
"epoch": 0.29554655870445345,
"grad_norm": 0.8091887730254227,
"learning_rate": 2.3548387096774196e-05,
"loss": 1.4575,
"step": 73
},
{
"epoch": 0.29959514170040485,
"grad_norm": 1.0207770291813947,
"learning_rate": 2.3870967741935486e-05,
"loss": 1.4258,
"step": 74
},
{
"epoch": 0.30364372469635625,
"grad_norm": 1.4808998151641,
"learning_rate": 2.4193548387096777e-05,
"loss": 1.5116,
"step": 75
},
{
"epoch": 0.3076923076923077,
"grad_norm": 0.6957045833595951,
"learning_rate": 2.4516129032258067e-05,
"loss": 1.5221,
"step": 76
},
{
"epoch": 0.3117408906882591,
"grad_norm": 1.645347680091355,
"learning_rate": 2.4838709677419358e-05,
"loss": 1.4471,
"step": 77
},
{
"epoch": 0.3157894736842105,
"grad_norm": 0.8080589483987229,
"learning_rate": 2.5161290322580648e-05,
"loss": 1.4862,
"step": 78
},
{
"epoch": 0.31983805668016196,
"grad_norm": 1.1086100017085756,
"learning_rate": 2.548387096774194e-05,
"loss": 1.5077,
"step": 79
},
{
"epoch": 0.32388663967611336,
"grad_norm": 1.0452483696341917,
"learning_rate": 2.5806451612903226e-05,
"loss": 1.4788,
"step": 80
},
{
"epoch": 0.32793522267206476,
"grad_norm": 1.2108991808912686,
"learning_rate": 2.6129032258064516e-05,
"loss": 1.4786,
"step": 81
},
{
"epoch": 0.3319838056680162,
"grad_norm": 1.1929498348455305,
"learning_rate": 2.6451612903225806e-05,
"loss": 1.4523,
"step": 82
},
{
"epoch": 0.3360323886639676,
"grad_norm": 0.8914691126569384,
"learning_rate": 2.6774193548387097e-05,
"loss": 1.4503,
"step": 83
},
{
"epoch": 0.340080971659919,
"grad_norm": 0.8416203010660562,
"learning_rate": 2.7096774193548387e-05,
"loss": 1.4835,
"step": 84
},
{
"epoch": 0.3441295546558704,
"grad_norm": 0.829776044167902,
"learning_rate": 2.7419354838709678e-05,
"loss": 1.4872,
"step": 85
},
{
"epoch": 0.3481781376518219,
"grad_norm": 1.0364987214831005,
"learning_rate": 2.774193548387097e-05,
"loss": 1.4591,
"step": 86
},
{
"epoch": 0.3522267206477733,
"grad_norm": 1.3192196114204893,
"learning_rate": 2.8064516129032262e-05,
"loss": 1.4454,
"step": 87
},
{
"epoch": 0.3562753036437247,
"grad_norm": 0.991535519473574,
"learning_rate": 2.8387096774193552e-05,
"loss": 1.4773,
"step": 88
},
{
"epoch": 0.3603238866396761,
"grad_norm": 0.772653878408877,
"learning_rate": 2.8709677419354843e-05,
"loss": 1.4868,
"step": 89
},
{
"epoch": 0.3643724696356275,
"grad_norm": 1.1198935107502297,
"learning_rate": 2.9032258064516133e-05,
"loss": 1.4774,
"step": 90
},
{
"epoch": 0.3684210526315789,
"grad_norm": 1.2800753626524828,
"learning_rate": 2.9354838709677423e-05,
"loss": 1.488,
"step": 91
},
{
"epoch": 0.3724696356275304,
"grad_norm": 0.9513763248283649,
"learning_rate": 2.9677419354838714e-05,
"loss": 1.4469,
"step": 92
},
{
"epoch": 0.3765182186234818,
"grad_norm": 1.0668434756361387,
"learning_rate": 3.0000000000000004e-05,
"loss": 1.4784,
"step": 93
},
{
"epoch": 0.3805668016194332,
"grad_norm": 0.7578471944363252,
"learning_rate": 3.032258064516129e-05,
"loss": 1.4928,
"step": 94
},
{
"epoch": 0.38461538461538464,
"grad_norm": 0.9047756200899297,
"learning_rate": 3.0645161290322585e-05,
"loss": 1.4368,
"step": 95
},
{
"epoch": 0.38866396761133604,
"grad_norm": 1.0455886278025504,
"learning_rate": 3.096774193548387e-05,
"loss": 1.4576,
"step": 96
},
{
"epoch": 0.39271255060728744,
"grad_norm": 1.3551218667911635,
"learning_rate": 3.1290322580645166e-05,
"loss": 1.4587,
"step": 97
},
{
"epoch": 0.3967611336032389,
"grad_norm": 0.8465232891785122,
"learning_rate": 3.161290322580645e-05,
"loss": 1.4508,
"step": 98
},
{
"epoch": 0.4008097165991903,
"grad_norm": 1.1742216069305886,
"learning_rate": 3.193548387096774e-05,
"loss": 1.4684,
"step": 99
},
{
"epoch": 0.4048582995951417,
"grad_norm": 1.0522366412258324,
"learning_rate": 3.2258064516129034e-05,
"loss": 1.4197,
"step": 100
},
{
"epoch": 0.4089068825910931,
"grad_norm": 1.0838626191410978,
"learning_rate": 3.258064516129033e-05,
"loss": 1.4837,
"step": 101
},
{
"epoch": 0.41295546558704455,
"grad_norm": 0.9484713751195007,
"learning_rate": 3.2903225806451614e-05,
"loss": 1.4705,
"step": 102
},
{
"epoch": 0.41700404858299595,
"grad_norm": 0.8381888254081903,
"learning_rate": 3.322580645161291e-05,
"loss": 1.4773,
"step": 103
},
{
"epoch": 0.42105263157894735,
"grad_norm": 1.0112096449509995,
"learning_rate": 3.3548387096774195e-05,
"loss": 1.4629,
"step": 104
},
{
"epoch": 0.4251012145748988,
"grad_norm": 1.3071798872461056,
"learning_rate": 3.387096774193549e-05,
"loss": 1.4976,
"step": 105
},
{
"epoch": 0.4291497975708502,
"grad_norm": 1.1904720862901332,
"learning_rate": 3.4193548387096776e-05,
"loss": 1.4616,
"step": 106
},
{
"epoch": 0.4331983805668016,
"grad_norm": 1.4096019170211587,
"learning_rate": 3.451612903225807e-05,
"loss": 1.4644,
"step": 107
},
{
"epoch": 0.43724696356275305,
"grad_norm": 0.8758089668498651,
"learning_rate": 3.483870967741936e-05,
"loss": 1.4662,
"step": 108
},
{
"epoch": 0.44129554655870445,
"grad_norm": 1.1709297553068376,
"learning_rate": 3.516129032258065e-05,
"loss": 1.4319,
"step": 109
},
{
"epoch": 0.44534412955465585,
"grad_norm": 1.4284545044629804,
"learning_rate": 3.548387096774194e-05,
"loss": 1.4796,
"step": 110
},
{
"epoch": 0.4493927125506073,
"grad_norm": 1.2669694197667367,
"learning_rate": 3.5806451612903225e-05,
"loss": 1.5083,
"step": 111
},
{
"epoch": 0.4534412955465587,
"grad_norm": 0.8780991243791991,
"learning_rate": 3.612903225806452e-05,
"loss": 1.4512,
"step": 112
},
{
"epoch": 0.4574898785425101,
"grad_norm": 1.164133167066795,
"learning_rate": 3.6451612903225805e-05,
"loss": 1.4447,
"step": 113
},
{
"epoch": 0.46153846153846156,
"grad_norm": 1.7759229538131707,
"learning_rate": 3.67741935483871e-05,
"loss": 1.4436,
"step": 114
},
{
"epoch": 0.46558704453441296,
"grad_norm": 1.0201777307655573,
"learning_rate": 3.7096774193548386e-05,
"loss": 1.4657,
"step": 115
},
{
"epoch": 0.46963562753036436,
"grad_norm": 0.9968353817673279,
"learning_rate": 3.741935483870968e-05,
"loss": 1.4904,
"step": 116
},
{
"epoch": 0.47368421052631576,
"grad_norm": 2.355954173997913,
"learning_rate": 3.7741935483870974e-05,
"loss": 1.466,
"step": 117
},
{
"epoch": 0.4777327935222672,
"grad_norm": 1.1707140012298214,
"learning_rate": 3.806451612903226e-05,
"loss": 1.4497,
"step": 118
},
{
"epoch": 0.4817813765182186,
"grad_norm": 3.141201224808401,
"learning_rate": 3.8387096774193555e-05,
"loss": 1.4969,
"step": 119
},
{
"epoch": 0.48582995951417,
"grad_norm": 2.6953547544063476,
"learning_rate": 3.870967741935484e-05,
"loss": 1.4171,
"step": 120
},
{
"epoch": 0.4898785425101215,
"grad_norm": 2.473471485981855,
"learning_rate": 3.9032258064516135e-05,
"loss": 1.451,
"step": 121
},
{
"epoch": 0.4939271255060729,
"grad_norm": 2.727116683859162,
"learning_rate": 3.935483870967742e-05,
"loss": 1.4757,
"step": 122
},
{
"epoch": 0.4979757085020243,
"grad_norm": 1.5078962353978118,
"learning_rate": 3.9677419354838716e-05,
"loss": 1.4457,
"step": 123
},
{
"epoch": 0.5020242914979757,
"grad_norm": 1.6128021551782532,
"learning_rate": 4e-05,
"loss": 1.444,
"step": 124
},
{
"epoch": 0.5060728744939271,
"grad_norm": 1.635985681468839,
"learning_rate": 3.9999920040266474e-05,
"loss": 1.4609,
"step": 125
},
{
"epoch": 0.5101214574898786,
"grad_norm": 1.105733797208387,
"learning_rate": 3.999968016170525e-05,
"loss": 1.4521,
"step": 126
},
{
"epoch": 0.5141700404858299,
"grad_norm": 1.6427875032025154,
"learning_rate": 3.9999280366234394e-05,
"loss": 1.4918,
"step": 127
},
{
"epoch": 0.5182186234817814,
"grad_norm": 1.3376707584260004,
"learning_rate": 3.999872065705065e-05,
"loss": 1.4715,
"step": 128
},
{
"epoch": 0.5222672064777328,
"grad_norm": 1.287963979204365,
"learning_rate": 3.999800103862945e-05,
"loss": 1.4675,
"step": 129
},
{
"epoch": 0.5263157894736842,
"grad_norm": 1.2453137016805433,
"learning_rate": 3.9997121516724834e-05,
"loss": 1.4487,
"step": 130
},
{
"epoch": 0.5303643724696356,
"grad_norm": 1.4596225927658026,
"learning_rate": 3.999608209836944e-05,
"loss": 1.4575,
"step": 131
},
{
"epoch": 0.5344129554655871,
"grad_norm": 1.6139274982965632,
"learning_rate": 3.999488279187443e-05,
"loss": 1.4384,
"step": 132
},
{
"epoch": 0.5384615384615384,
"grad_norm": 1.3759478324258247,
"learning_rate": 3.9993523606829425e-05,
"loss": 1.4491,
"step": 133
},
{
"epoch": 0.5425101214574899,
"grad_norm": 1.9642798966498003,
"learning_rate": 3.9992004554102435e-05,
"loss": 1.4772,
"step": 134
},
{
"epoch": 0.5465587044534413,
"grad_norm": 1.7182821150968743,
"learning_rate": 3.999032564583976e-05,
"loss": 1.4185,
"step": 135
},
{
"epoch": 0.5506072874493927,
"grad_norm": 1.4989843477840996,
"learning_rate": 3.998848689546591e-05,
"loss": 1.4238,
"step": 136
},
{
"epoch": 0.5546558704453441,
"grad_norm": 1.5474254486447436,
"learning_rate": 3.9986488317683496e-05,
"loss": 1.4668,
"step": 137
},
{
"epoch": 0.5587044534412956,
"grad_norm": 1.277819886941497,
"learning_rate": 3.9984329928473067e-05,
"loss": 1.4635,
"step": 138
},
{
"epoch": 0.562753036437247,
"grad_norm": 1.864170585995185,
"learning_rate": 3.998201174509306e-05,
"loss": 1.4632,
"step": 139
},
{
"epoch": 0.5668016194331984,
"grad_norm": 1.3297200740897912,
"learning_rate": 3.997953378607961e-05,
"loss": 1.4666,
"step": 140
},
{
"epoch": 0.5708502024291497,
"grad_norm": 1.9266001414345117,
"learning_rate": 3.9976896071246403e-05,
"loss": 1.4401,
"step": 141
},
{
"epoch": 0.5748987854251012,
"grad_norm": 1.5590251762930223,
"learning_rate": 3.997409862168454e-05,
"loss": 1.4667,
"step": 142
},
{
"epoch": 0.5789473684210527,
"grad_norm": 1.5715000073432852,
"learning_rate": 3.997114145976237e-05,
"loss": 1.4737,
"step": 143
},
{
"epoch": 0.582995951417004,
"grad_norm": 1.082552530728037,
"learning_rate": 3.9968024609125245e-05,
"loss": 1.4495,
"step": 144
},
{
"epoch": 0.5870445344129555,
"grad_norm": 0.9185682543244243,
"learning_rate": 3.996474809469545e-05,
"loss": 1.4048,
"step": 145
},
{
"epoch": 0.5910931174089069,
"grad_norm": 1.511641514887913,
"learning_rate": 3.996131194267189e-05,
"loss": 1.4617,
"step": 146
},
{
"epoch": 0.5951417004048583,
"grad_norm": 1.0672369838325564,
"learning_rate": 3.995771618052995e-05,
"loss": 1.4321,
"step": 147
},
{
"epoch": 0.5991902834008097,
"grad_norm": 1.1365561508500535,
"learning_rate": 3.9953960837021256e-05,
"loss": 1.469,
"step": 148
},
{
"epoch": 0.6032388663967612,
"grad_norm": 1.3292581388756313,
"learning_rate": 3.995004594217343e-05,
"loss": 1.4415,
"step": 149
},
{
"epoch": 0.6072874493927125,
"grad_norm": 0.9625869219593246,
"learning_rate": 3.994597152728987e-05,
"loss": 1.4565,
"step": 150
},
{
"epoch": 0.611336032388664,
"grad_norm": 1.6825190577747537,
"learning_rate": 3.994173762494947e-05,
"loss": 1.3995,
"step": 151
},
{
"epoch": 0.6153846153846154,
"grad_norm": 1.3999376900788516,
"learning_rate": 3.9937344269006424e-05,
"loss": 1.4371,
"step": 152
},
{
"epoch": 0.6194331983805668,
"grad_norm": 1.215819608646709,
"learning_rate": 3.993279149458988e-05,
"loss": 1.4572,
"step": 153
},
{
"epoch": 0.6234817813765182,
"grad_norm": 1.4785690531427118,
"learning_rate": 3.9928079338103695e-05,
"loss": 1.4428,
"step": 154
},
{
"epoch": 0.6275303643724697,
"grad_norm": 1.0327689338776804,
"learning_rate": 3.992320783722616e-05,
"loss": 1.4753,
"step": 155
},
{
"epoch": 0.631578947368421,
"grad_norm": 1.540245233660307,
"learning_rate": 3.991817703090965e-05,
"loss": 1.4415,
"step": 156
},
{
"epoch": 0.6356275303643725,
"grad_norm": 1.1698580076613396,
"learning_rate": 3.991298695938038e-05,
"loss": 1.419,
"step": 157
},
{
"epoch": 0.6396761133603239,
"grad_norm": 1.4728036343585944,
"learning_rate": 3.9907637664138e-05,
"loss": 1.416,
"step": 158
},
{
"epoch": 0.6437246963562753,
"grad_norm": 1.3036351502871044,
"learning_rate": 3.990212918795534e-05,
"loss": 1.4377,
"step": 159
},
{
"epoch": 0.6477732793522267,
"grad_norm": 1.3286430802687763,
"learning_rate": 3.9896461574878036e-05,
"loss": 1.469,
"step": 160
},
{
"epoch": 0.6518218623481782,
"grad_norm": 1.0866297387917692,
"learning_rate": 3.989063487022416e-05,
"loss": 1.451,
"step": 161
},
{
"epoch": 0.6558704453441295,
"grad_norm": 1.1881163354190234,
"learning_rate": 3.9884649120583903e-05,
"loss": 1.4719,
"step": 162
},
{
"epoch": 0.659919028340081,
"grad_norm": 1.1010695642644635,
"learning_rate": 3.987850437381915e-05,
"loss": 1.5041,
"step": 163
},
{
"epoch": 0.6639676113360324,
"grad_norm": 1.1066277036916576,
"learning_rate": 3.987220067906313e-05,
"loss": 1.4493,
"step": 164
},
{
"epoch": 0.6680161943319838,
"grad_norm": 0.9710132169341941,
"learning_rate": 3.986573808672002e-05,
"loss": 1.4117,
"step": 165
},
{
"epoch": 0.6720647773279352,
"grad_norm": 1.2876211891226204,
"learning_rate": 3.9859116648464534e-05,
"loss": 1.4797,
"step": 166
},
{
"epoch": 0.6761133603238867,
"grad_norm": 0.8904262972697528,
"learning_rate": 3.985233641724152e-05,
"loss": 1.4648,
"step": 167
},
{
"epoch": 0.680161943319838,
"grad_norm": 1.7502455226737832,
"learning_rate": 3.984539744726554e-05,
"loss": 1.4389,
"step": 168
},
{
"epoch": 0.6842105263157895,
"grad_norm": 1.1080471306015431,
"learning_rate": 3.983829979402038e-05,
"loss": 1.4482,
"step": 169
},
{
"epoch": 0.6882591093117408,
"grad_norm": 2.028149266000291,
"learning_rate": 3.983104351425871e-05,
"loss": 1.4615,
"step": 170
},
{
"epoch": 0.6923076923076923,
"grad_norm": 1.8217620933719565,
"learning_rate": 3.982362866600154e-05,
"loss": 1.4641,
"step": 171
},
{
"epoch": 0.6963562753036437,
"grad_norm": 1.3076284375387828,
"learning_rate": 3.9816055308537806e-05,
"loss": 1.4597,
"step": 172
},
{
"epoch": 0.7004048582995951,
"grad_norm": 1.4846123110035605,
"learning_rate": 3.980832350242388e-05,
"loss": 1.4464,
"step": 173
},
{
"epoch": 0.7044534412955465,
"grad_norm": 1.0952566977130556,
"learning_rate": 3.980043330948306e-05,
"loss": 1.4376,
"step": 174
},
{
"epoch": 0.708502024291498,
"grad_norm": 1.0736127861437472,
"learning_rate": 3.9792384792805124e-05,
"loss": 1.3983,
"step": 175
},
{
"epoch": 0.7125506072874493,
"grad_norm": 1.125815578019142,
"learning_rate": 3.978417801674579e-05,
"loss": 1.4353,
"step": 176
},
{
"epoch": 0.7165991902834008,
"grad_norm": 1.0847057387922805,
"learning_rate": 3.9775813046926244e-05,
"loss": 1.4151,
"step": 177
},
{
"epoch": 0.7206477732793523,
"grad_norm": 0.8286311106168794,
"learning_rate": 3.976728995023254e-05,
"loss": 1.4577,
"step": 178
},
{
"epoch": 0.7246963562753036,
"grad_norm": 0.9221071397996441,
"learning_rate": 3.975860879481514e-05,
"loss": 1.4641,
"step": 179
},
{
"epoch": 0.728744939271255,
"grad_norm": 1.0069377052419717,
"learning_rate": 3.974976965008833e-05,
"loss": 1.4391,
"step": 180
},
{
"epoch": 0.7327935222672065,
"grad_norm": 1.0716522533380493,
"learning_rate": 3.9740772586729674e-05,
"loss": 1.4418,
"step": 181
},
{
"epoch": 0.7368421052631579,
"grad_norm": 1.1130197490462124,
"learning_rate": 3.973161767667946e-05,
"loss": 1.4149,
"step": 182
},
{
"epoch": 0.7408906882591093,
"grad_norm": 1.0147293780131004,
"learning_rate": 3.972230499314009e-05,
"loss": 1.4744,
"step": 183
},
{
"epoch": 0.7449392712550608,
"grad_norm": 0.8636860550282619,
"learning_rate": 3.9712834610575545e-05,
"loss": 1.489,
"step": 184
},
{
"epoch": 0.7489878542510121,
"grad_norm": 0.8852249113105359,
"learning_rate": 3.9703206604710746e-05,
"loss": 1.4438,
"step": 185
},
{
"epoch": 0.7530364372469636,
"grad_norm": 1.3385513547183816,
"learning_rate": 3.969342105253098e-05,
"loss": 1.4423,
"step": 186
},
{
"epoch": 0.757085020242915,
"grad_norm": 0.9449323994057769,
"learning_rate": 3.9683478032281246e-05,
"loss": 1.4461,
"step": 187
},
{
"epoch": 0.7611336032388664,
"grad_norm": 1.5012058670106085,
"learning_rate": 3.967337762346568e-05,
"loss": 1.472,
"step": 188
},
{
"epoch": 0.7651821862348178,
"grad_norm": 1.3066581962480466,
"learning_rate": 3.9663119906846885e-05,
"loss": 1.4284,
"step": 189
},
{
"epoch": 0.7692307692307693,
"grad_norm": 0.9217438428080573,
"learning_rate": 3.965270496444528e-05,
"loss": 1.4528,
"step": 190
},
{
"epoch": 0.7732793522267206,
"grad_norm": 0.9568062776906358,
"learning_rate": 3.9642132879538465e-05,
"loss": 1.4589,
"step": 191
},
{
"epoch": 0.7773279352226721,
"grad_norm": 0.7770106339057873,
"learning_rate": 3.9631403736660566e-05,
"loss": 1.4115,
"step": 192
},
{
"epoch": 0.7813765182186235,
"grad_norm": 1.208448478313322,
"learning_rate": 3.962051762160151e-05,
"loss": 1.4188,
"step": 193
},
{
"epoch": 0.7854251012145749,
"grad_norm": 0.839613711417774,
"learning_rate": 3.960947462140638e-05,
"loss": 1.4427,
"step": 194
},
{
"epoch": 0.7894736842105263,
"grad_norm": 1.1555345472007346,
"learning_rate": 3.959827482437472e-05,
"loss": 1.3812,
"step": 195
},
{
"epoch": 0.7935222672064778,
"grad_norm": 0.7719899581282356,
"learning_rate": 3.958691832005981e-05,
"loss": 1.4182,
"step": 196
},
{
"epoch": 0.7975708502024291,
"grad_norm": 1.040696561867317,
"learning_rate": 3.9575405199267944e-05,
"loss": 1.471,
"step": 197
},
{
"epoch": 0.8016194331983806,
"grad_norm": 0.6843895432630372,
"learning_rate": 3.956373555405774e-05,
"loss": 1.4602,
"step": 198
},
{
"epoch": 0.805668016194332,
"grad_norm": 0.8881700699326408,
"learning_rate": 3.955190947773937e-05,
"loss": 1.4881,
"step": 199
},
{
"epoch": 0.8097165991902834,
"grad_norm": 0.7004429703708895,
"learning_rate": 3.9539927064873824e-05,
"loss": 1.4337,
"step": 200
},
{
"epoch": 0.8137651821862348,
"grad_norm": 0.8546698252141217,
"learning_rate": 3.952778841127214e-05,
"loss": 1.464,
"step": 201
},
{
"epoch": 0.8178137651821862,
"grad_norm": 0.6162215132351873,
"learning_rate": 3.95154936139947e-05,
"loss": 1.461,
"step": 202
},
{
"epoch": 0.8218623481781376,
"grad_norm": 0.7512824599542013,
"learning_rate": 3.950304277135034e-05,
"loss": 1.4285,
"step": 203
},
{
"epoch": 0.8259109311740891,
"grad_norm": 0.8547414364152552,
"learning_rate": 3.94904359828957e-05,
"loss": 1.4456,
"step": 204
},
{
"epoch": 0.8299595141700404,
"grad_norm": 1.0528341177612186,
"learning_rate": 3.94776733494343e-05,
"loss": 1.4375,
"step": 205
},
{
"epoch": 0.8340080971659919,
"grad_norm": 1.1225337026298936,
"learning_rate": 3.9464754973015826e-05,
"loss": 1.4275,
"step": 206
},
{
"epoch": 0.8380566801619433,
"grad_norm": 0.6667551437324072,
"learning_rate": 3.9451680956935275e-05,
"loss": 1.4141,
"step": 207
},
{
"epoch": 0.8421052631578947,
"grad_norm": 0.9712308189095885,
"learning_rate": 3.943845140573213e-05,
"loss": 1.4542,
"step": 208
},
{
"epoch": 0.8461538461538461,
"grad_norm": 1.28029724333711,
"learning_rate": 3.942506642518952e-05,
"loss": 1.4364,
"step": 209
},
{
"epoch": 0.8502024291497976,
"grad_norm": 0.9761983584128348,
"learning_rate": 3.9411526122333396e-05,
"loss": 1.4451,
"step": 210
},
{
"epoch": 0.854251012145749,
"grad_norm": 0.7419945662027494,
"learning_rate": 3.939783060543167e-05,
"loss": 1.4936,
"step": 211
},
{
"epoch": 0.8582995951417004,
"grad_norm": 1.0940045962345781,
"learning_rate": 3.938397998399333e-05,
"loss": 1.4132,
"step": 212
},
{
"epoch": 0.8623481781376519,
"grad_norm": 0.906874507349303,
"learning_rate": 3.936997436876756e-05,
"loss": 1.468,
"step": 213
},
{
"epoch": 0.8663967611336032,
"grad_norm": 0.6619462110875662,
"learning_rate": 3.935581387174289e-05,
"loss": 1.4622,
"step": 214
},
{
"epoch": 0.8704453441295547,
"grad_norm": 0.8850531743027442,
"learning_rate": 3.9341498606146295e-05,
"loss": 1.4377,
"step": 215
},
{
"epoch": 0.8744939271255061,
"grad_norm": 0.7858136802850174,
"learning_rate": 3.932702868644224e-05,
"loss": 1.4642,
"step": 216
},
{
"epoch": 0.8785425101214575,
"grad_norm": 1.0805048650631535,
"learning_rate": 3.931240422833181e-05,
"loss": 1.4674,
"step": 217
},
{
"epoch": 0.8825910931174089,
"grad_norm": 0.7400234005140444,
"learning_rate": 3.9297625348751806e-05,
"loss": 1.4061,
"step": 218
},
{
"epoch": 0.8866396761133604,
"grad_norm": 0.8156515434372876,
"learning_rate": 3.9282692165873744e-05,
"loss": 1.4865,
"step": 219
},
{
"epoch": 0.8906882591093117,
"grad_norm": 0.9085837075091509,
"learning_rate": 3.9267604799102945e-05,
"loss": 1.4557,
"step": 220
},
{
"epoch": 0.8947368421052632,
"grad_norm": 0.8200360712169551,
"learning_rate": 3.9252363369077615e-05,
"loss": 1.4495,
"step": 221
},
{
"epoch": 0.8987854251012146,
"grad_norm": 0.9504370041779868,
"learning_rate": 3.9236967997667796e-05,
"loss": 1.4176,
"step": 222
},
{
"epoch": 0.902834008097166,
"grad_norm": 0.9389531434091761,
"learning_rate": 3.9221418807974496e-05,
"loss": 1.4086,
"step": 223
},
{
"epoch": 0.9068825910931174,
"grad_norm": 0.8590813355255267,
"learning_rate": 3.92057159243286e-05,
"loss": 1.4329,
"step": 224
},
{
"epoch": 0.9109311740890689,
"grad_norm": 0.7138105328379782,
"learning_rate": 3.9189859472289956e-05,
"loss": 1.4228,
"step": 225
},
{
"epoch": 0.9149797570850202,
"grad_norm": 0.7885911558309601,
"learning_rate": 3.9173849578646324e-05,
"loss": 1.3926,
"step": 226
},
{
"epoch": 0.9190283400809717,
"grad_norm": 0.8819160527105082,
"learning_rate": 3.9157686371412395e-05,
"loss": 1.4309,
"step": 227
},
{
"epoch": 0.9230769230769231,
"grad_norm": 0.8117441284604736,
"learning_rate": 3.914136997982874e-05,
"loss": 1.4585,
"step": 228
},
{
"epoch": 0.9271255060728745,
"grad_norm": 0.7249176530587264,
"learning_rate": 3.912490053436079e-05,
"loss": 1.4618,
"step": 229
},
{
"epoch": 0.9311740890688259,
"grad_norm": 0.7431362849850806,
"learning_rate": 3.9108278166697805e-05,
"loss": 1.442,
"step": 230
},
{
"epoch": 0.9352226720647774,
"grad_norm": 0.8189611224322949,
"learning_rate": 3.909150300975178e-05,
"loss": 1.4142,
"step": 231
},
{
"epoch": 0.9392712550607287,
"grad_norm": 0.7951678719974796,
"learning_rate": 3.907457519765642e-05,
"loss": 1.4629,
"step": 232
},
{
"epoch": 0.9433198380566802,
"grad_norm": 0.7065837979195885,
"learning_rate": 3.905749486576607e-05,
"loss": 1.4602,
"step": 233
},
{
"epoch": 0.9473684210526315,
"grad_norm": 0.6981096690435673,
"learning_rate": 3.90402621506546e-05,
"loss": 1.4411,
"step": 234
},
{
"epoch": 0.951417004048583,
"grad_norm": 0.5822531700667379,
"learning_rate": 3.902287719011435e-05,
"loss": 1.432,
"step": 235
},
{
"epoch": 0.9554655870445344,
"grad_norm": 0.6161799423927152,
"learning_rate": 3.900534012315499e-05,
"loss": 1.4299,
"step": 236
},
{
"epoch": 0.9595141700404858,
"grad_norm": 0.6169361276301292,
"learning_rate": 3.8987651090002454e-05,
"loss": 1.4579,
"step": 237
},
{
"epoch": 0.9635627530364372,
"grad_norm": 0.7148454860141688,
"learning_rate": 3.896981023209777e-05,
"loss": 1.4184,
"step": 238
},
{
"epoch": 0.9676113360323887,
"grad_norm": 0.5680268734837902,
"learning_rate": 3.8951817692095966e-05,
"loss": 1.4347,
"step": 239
},
{
"epoch": 0.97165991902834,
"grad_norm": 0.6084029411128079,
"learning_rate": 3.893367361386491e-05,
"loss": 1.4543,
"step": 240
},
{
"epoch": 0.9757085020242915,
"grad_norm": 0.667326049445474,
"learning_rate": 3.891537814248417e-05,
"loss": 1.4172,
"step": 241
},
{
"epoch": 0.979757085020243,
"grad_norm": 0.6289625781052549,
"learning_rate": 3.889693142424384e-05,
"loss": 1.4144,
"step": 242
},
{
"epoch": 0.9838056680161943,
"grad_norm": 0.6397130256969309,
"learning_rate": 3.8878333606643405e-05,
"loss": 1.4388,
"step": 243
},
{
"epoch": 0.9878542510121457,
"grad_norm": 0.6293864957841452,
"learning_rate": 3.885958483839049e-05,
"loss": 1.4545,
"step": 244
},
{
"epoch": 0.9919028340080972,
"grad_norm": 0.6004779678918897,
"learning_rate": 3.8840685269399786e-05,
"loss": 1.4332,
"step": 245
},
{
"epoch": 0.9959514170040485,
"grad_norm": 0.5997395370617209,
"learning_rate": 3.882163505079171e-05,
"loss": 1.4658,
"step": 246
},
{
"epoch": 1.0,
"grad_norm": 0.6531213947369271,
"learning_rate": 3.880243433489132e-05,
"loss": 1.407,
"step": 247
},
{
"epoch": 1.0040485829959513,
"grad_norm": 0.6685558354103437,
"learning_rate": 3.878308327522702e-05,
"loss": 1.415,
"step": 248
},
{
"epoch": 1.008097165991903,
"grad_norm": 0.8583048569328354,
"learning_rate": 3.876358202652936e-05,
"loss": 1.4118,
"step": 249
},
{
"epoch": 1.0121457489878543,
"grad_norm": 0.697922190458052,
"learning_rate": 3.8743930744729835e-05,
"loss": 1.3932,
"step": 250
},
{
"epoch": 1.0161943319838056,
"grad_norm": 0.7868265781608158,
"learning_rate": 3.8724129586959535e-05,
"loss": 1.3646,
"step": 251
},
{
"epoch": 1.0202429149797572,
"grad_norm": 0.7231864996622832,
"learning_rate": 3.870417871154801e-05,
"loss": 1.4262,
"step": 252
},
{
"epoch": 1.0242914979757085,
"grad_norm": 0.7808848835444173,
"learning_rate": 3.868407827802192e-05,
"loss": 1.398,
"step": 253
},
{
"epoch": 1.0283400809716599,
"grad_norm": 0.842539149164592,
"learning_rate": 3.8663828447103804e-05,
"loss": 1.4206,
"step": 254
},
{
"epoch": 1.0323886639676114,
"grad_norm": 0.5982335594320104,
"learning_rate": 3.8643429380710765e-05,
"loss": 1.3812,
"step": 255
},
{
"epoch": 1.0364372469635628,
"grad_norm": 0.6002337143147122,
"learning_rate": 3.862288124195319e-05,
"loss": 1.3521,
"step": 256
},
{
"epoch": 1.040485829959514,
"grad_norm": 0.7095023566820967,
"learning_rate": 3.8602184195133464e-05,
"loss": 1.3703,
"step": 257
},
{
"epoch": 1.0445344129554657,
"grad_norm": 0.664173442496212,
"learning_rate": 3.8581338405744604e-05,
"loss": 1.3789,
"step": 258
},
{
"epoch": 1.048582995951417,
"grad_norm": 0.6333353278374255,
"learning_rate": 3.8560344040468996e-05,
"loss": 1.4028,
"step": 259
},
{
"epoch": 1.0526315789473684,
"grad_norm": 0.6144553252081912,
"learning_rate": 3.853920126717703e-05,
"loss": 1.3814,
"step": 260
},
{
"epoch": 1.05668016194332,
"grad_norm": 0.7198852099762024,
"learning_rate": 3.851791025492574e-05,
"loss": 1.3892,
"step": 261
},
{
"epoch": 1.0607287449392713,
"grad_norm": 0.5307831466979526,
"learning_rate": 3.8496471173957515e-05,
"loss": 1.3964,
"step": 262
},
{
"epoch": 1.0647773279352226,
"grad_norm": 0.7165237367154584,
"learning_rate": 3.847488419569866e-05,
"loss": 1.3772,
"step": 263
},
{
"epoch": 1.0688259109311742,
"grad_norm": 0.629165654421554,
"learning_rate": 3.8453149492758077e-05,
"loss": 1.3805,
"step": 264
},
{
"epoch": 1.0728744939271255,
"grad_norm": 0.6099476860648617,
"learning_rate": 3.8431267238925886e-05,
"loss": 1.3743,
"step": 265
},
{
"epoch": 1.0769230769230769,
"grad_norm": 0.6986055328085804,
"learning_rate": 3.840923760917199e-05,
"loss": 1.3865,
"step": 266
},
{
"epoch": 1.0809716599190284,
"grad_norm": 0.7605745665406622,
"learning_rate": 3.838706077964473e-05,
"loss": 1.3858,
"step": 267
},
{
"epoch": 1.0850202429149798,
"grad_norm": 0.8219001820348313,
"learning_rate": 3.836473692766944e-05,
"loss": 1.4024,
"step": 268
},
{
"epoch": 1.0890688259109311,
"grad_norm": 0.7979128106440478,
"learning_rate": 3.834226623174705e-05,
"loss": 1.4097,
"step": 269
},
{
"epoch": 1.0931174089068827,
"grad_norm": 0.5574421911446628,
"learning_rate": 3.831964887155264e-05,
"loss": 1.3715,
"step": 270
},
{
"epoch": 1.097165991902834,
"grad_norm": 0.5831160379589343,
"learning_rate": 3.829688502793403e-05,
"loss": 1.3623,
"step": 271
},
{
"epoch": 1.1012145748987854,
"grad_norm": 0.5895929468190468,
"learning_rate": 3.827397488291029e-05,
"loss": 1.3832,
"step": 272
},
{
"epoch": 1.1052631578947367,
"grad_norm": 0.6963075821845434,
"learning_rate": 3.8250918619670343e-05,
"loss": 1.3317,
"step": 273
},
{
"epoch": 1.1093117408906883,
"grad_norm": 0.9038087936649127,
"learning_rate": 3.822771642257145e-05,
"loss": 1.3904,
"step": 274
},
{
"epoch": 1.1133603238866396,
"grad_norm": 0.5637541753319327,
"learning_rate": 3.820436847713776e-05,
"loss": 1.3773,
"step": 275
},
{
"epoch": 1.117408906882591,
"grad_norm": 0.6989298180519041,
"learning_rate": 3.8180874970058826e-05,
"loss": 1.3793,
"step": 276
},
{
"epoch": 1.1214574898785425,
"grad_norm": 0.7141689984476253,
"learning_rate": 3.8157236089188103e-05,
"loss": 1.3842,
"step": 277
},
{
"epoch": 1.125506072874494,
"grad_norm": 0.687013392387765,
"learning_rate": 3.813345202354145e-05,
"loss": 1.3867,
"step": 278
},
{
"epoch": 1.1295546558704452,
"grad_norm": 0.5729705709772869,
"learning_rate": 3.810952296329563e-05,
"loss": 1.3877,
"step": 279
},
{
"epoch": 1.1336032388663968,
"grad_norm": 0.661241089389352,
"learning_rate": 3.808544909978676e-05,
"loss": 1.3803,
"step": 280
},
{
"epoch": 1.1376518218623481,
"grad_norm": 0.7277521635061901,
"learning_rate": 3.806123062550882e-05,
"loss": 1.4342,
"step": 281
},
{
"epoch": 1.1417004048582995,
"grad_norm": 0.7339349292544272,
"learning_rate": 3.803686773411208e-05,
"loss": 1.3863,
"step": 282
},
{
"epoch": 1.145748987854251,
"grad_norm": 0.5970321899802905,
"learning_rate": 3.801236062040158e-05,
"loss": 1.3382,
"step": 283
},
{
"epoch": 1.1497975708502024,
"grad_norm": 0.8296411155386302,
"learning_rate": 3.798770948033553e-05,
"loss": 1.3951,
"step": 284
},
{
"epoch": 1.1538461538461537,
"grad_norm": 0.6621834813359324,
"learning_rate": 3.796291451102381e-05,
"loss": 1.3615,
"step": 285
},
{
"epoch": 1.1578947368421053,
"grad_norm": 0.6307792064578708,
"learning_rate": 3.793797591072631e-05,
"loss": 1.3601,
"step": 286
},
{
"epoch": 1.1619433198380567,
"grad_norm": 0.6970411734324247,
"learning_rate": 3.791289387885144e-05,
"loss": 1.3963,
"step": 287
},
{
"epoch": 1.165991902834008,
"grad_norm": 0.7102312005586533,
"learning_rate": 3.788766861595443e-05,
"loss": 1.3721,
"step": 288
},
{
"epoch": 1.1700404858299596,
"grad_norm": 0.7629439156470526,
"learning_rate": 3.7862300323735835e-05,
"loss": 1.3911,
"step": 289
},
{
"epoch": 1.174089068825911,
"grad_norm": 0.600223866212318,
"learning_rate": 3.783678920503983e-05,
"loss": 1.415,
"step": 290
},
{
"epoch": 1.1781376518218623,
"grad_norm": 0.5063311784391415,
"learning_rate": 3.781113546385265e-05,
"loss": 1.3812,
"step": 291
},
{
"epoch": 1.1821862348178138,
"grad_norm": 0.5739612304498843,
"learning_rate": 3.7785339305300916e-05,
"loss": 1.3715,
"step": 292
},
{
"epoch": 1.1862348178137652,
"grad_norm": 0.6444131442034535,
"learning_rate": 3.7759400935650025e-05,
"loss": 1.3582,
"step": 293
},
{
"epoch": 1.1902834008097165,
"grad_norm": 0.5014727336755345,
"learning_rate": 3.77333205623025e-05,
"loss": 1.3994,
"step": 294
},
{
"epoch": 1.194331983805668,
"grad_norm": 0.6321555899905817,
"learning_rate": 3.7707098393796295e-05,
"loss": 1.3691,
"step": 295
},
{
"epoch": 1.1983805668016194,
"grad_norm": 0.5678203937991235,
"learning_rate": 3.7680734639803185e-05,
"loss": 1.3983,
"step": 296
},
{
"epoch": 1.2024291497975708,
"grad_norm": 0.5822510294278924,
"learning_rate": 3.765422951112704e-05,
"loss": 1.3869,
"step": 297
},
{
"epoch": 1.2064777327935223,
"grad_norm": 0.6501435021717704,
"learning_rate": 3.762758321970216e-05,
"loss": 1.3904,
"step": 298
},
{
"epoch": 1.2105263157894737,
"grad_norm": 0.5514454595430551,
"learning_rate": 3.7600795978591584e-05,
"loss": 1.3882,
"step": 299
},
{
"epoch": 1.214574898785425,
"grad_norm": 0.5582215644761732,
"learning_rate": 3.757386800198538e-05,
"loss": 1.3292,
"step": 300
},
{
"epoch": 1.2186234817813766,
"grad_norm": 0.49839973345570304,
"learning_rate": 3.7546799505198925e-05,
"loss": 1.3389,
"step": 301
},
{
"epoch": 1.222672064777328,
"grad_norm": 0.6272598695402243,
"learning_rate": 3.7519590704671197e-05,
"loss": 1.3756,
"step": 302
},
{
"epoch": 1.2267206477732793,
"grad_norm": 0.5335771355804929,
"learning_rate": 3.749224181796305e-05,
"loss": 1.4108,
"step": 303
},
{
"epoch": 1.2307692307692308,
"grad_norm": 0.5164589251620814,
"learning_rate": 3.746475306375545e-05,
"loss": 1.38,
"step": 304
},
{
"epoch": 1.2348178137651822,
"grad_norm": 0.5378874330565743,
"learning_rate": 3.743712466184774e-05,
"loss": 1.3951,
"step": 305
},
{
"epoch": 1.2388663967611335,
"grad_norm": 0.627492569954726,
"learning_rate": 3.7409356833155885e-05,
"loss": 1.379,
"step": 306
},
{
"epoch": 1.242914979757085,
"grad_norm": 0.7084947402750295,
"learning_rate": 3.73814497997107e-05,
"loss": 1.3788,
"step": 307
},
{
"epoch": 1.2469635627530364,
"grad_norm": 0.6900996447545404,
"learning_rate": 3.7353403784656095e-05,
"loss": 1.3705,
"step": 308
},
{
"epoch": 1.2510121457489878,
"grad_norm": 0.6265628663414077,
"learning_rate": 3.732521901224724e-05,
"loss": 1.3466,
"step": 309
},
{
"epoch": 1.2550607287449393,
"grad_norm": 0.5568519126849876,
"learning_rate": 3.7296895707848844e-05,
"loss": 1.3682,
"step": 310
},
{
"epoch": 1.2591093117408907,
"grad_norm": 0.7931681575240622,
"learning_rate": 3.7268434097933275e-05,
"loss": 1.3403,
"step": 311
},
{
"epoch": 1.263157894736842,
"grad_norm": 0.8285746837403117,
"learning_rate": 3.723983441007882e-05,
"loss": 1.4101,
"step": 312
},
{
"epoch": 1.2672064777327936,
"grad_norm": 0.8224840807311248,
"learning_rate": 3.721109687296781e-05,
"loss": 1.3623,
"step": 313
},
{
"epoch": 1.271255060728745,
"grad_norm": 0.7119494302394005,
"learning_rate": 3.718222171638484e-05,
"loss": 1.3426,
"step": 314
},
{
"epoch": 1.2753036437246963,
"grad_norm": 0.5369934691033277,
"learning_rate": 3.715320917121488e-05,
"loss": 1.3776,
"step": 315
},
{
"epoch": 1.2793522267206479,
"grad_norm": 0.8062519049953825,
"learning_rate": 3.712405946944149e-05,
"loss": 1.369,
"step": 316
},
{
"epoch": 1.2834008097165992,
"grad_norm": 0.9237275305496779,
"learning_rate": 3.709477284414488e-05,
"loss": 1.3836,
"step": 317
},
{
"epoch": 1.2874493927125505,
"grad_norm": 0.7846938835286075,
"learning_rate": 3.7065349529500136e-05,
"loss": 1.3768,
"step": 318
},
{
"epoch": 1.291497975708502,
"grad_norm": 0.5652170879774373,
"learning_rate": 3.7035789760775305e-05,
"loss": 1.4003,
"step": 319
},
{
"epoch": 1.2955465587044535,
"grad_norm": 0.9197214583786602,
"learning_rate": 3.70060937743295e-05,
"loss": 1.3735,
"step": 320
},
{
"epoch": 1.2995951417004048,
"grad_norm": 0.9613420195813668,
"learning_rate": 3.6976261807611045e-05,
"loss": 1.3575,
"step": 321
},
{
"epoch": 1.3036437246963564,
"grad_norm": 0.4916017545951097,
"learning_rate": 3.694629409915555e-05,
"loss": 1.4155,
"step": 322
},
{
"epoch": 1.3076923076923077,
"grad_norm": 0.7726893562582218,
"learning_rate": 3.691619088858401e-05,
"loss": 1.4194,
"step": 323
},
{
"epoch": 1.311740890688259,
"grad_norm": 1.0847646284564254,
"learning_rate": 3.68859524166009e-05,
"loss": 1.4258,
"step": 324
},
{
"epoch": 1.3157894736842106,
"grad_norm": 0.7623774055606887,
"learning_rate": 3.6855578924992226e-05,
"loss": 1.3577,
"step": 325
},
{
"epoch": 1.319838056680162,
"grad_norm": 0.611238863798574,
"learning_rate": 3.6825070656623626e-05,
"loss": 1.3448,
"step": 326
},
{
"epoch": 1.3238866396761133,
"grad_norm": 0.6876249934813652,
"learning_rate": 3.67944278554384e-05,
"loss": 1.4238,
"step": 327
},
{
"epoch": 1.3279352226720649,
"grad_norm": 0.6477149414721328,
"learning_rate": 3.676365076645557e-05,
"loss": 1.316,
"step": 328
},
{
"epoch": 1.3319838056680162,
"grad_norm": 0.5889414764932491,
"learning_rate": 3.673273963576791e-05,
"loss": 1.3493,
"step": 329
},
{
"epoch": 1.3360323886639676,
"grad_norm": 0.7035319037115692,
"learning_rate": 3.670169471054001e-05,
"loss": 1.3585,
"step": 330
},
{
"epoch": 1.3400809716599191,
"grad_norm": 0.766707725357067,
"learning_rate": 3.6670516239006254e-05,
"loss": 1.367,
"step": 331
},
{
"epoch": 1.3441295546558705,
"grad_norm": 0.5853844198689601,
"learning_rate": 3.663920447046888e-05,
"loss": 1.371,
"step": 332
},
{
"epoch": 1.3481781376518218,
"grad_norm": 0.581892631573,
"learning_rate": 3.6607759655295954e-05,
"loss": 1.3567,
"step": 333
},
{
"epoch": 1.3522267206477734,
"grad_norm": 0.6287337601385642,
"learning_rate": 3.657618204491937e-05,
"loss": 1.349,
"step": 334
},
{
"epoch": 1.3562753036437247,
"grad_norm": 0.5504578718000623,
"learning_rate": 3.6544471891832875e-05,
"loss": 1.416,
"step": 335
},
{
"epoch": 1.360323886639676,
"grad_norm": 0.5108653245388801,
"learning_rate": 3.6512629449589983e-05,
"loss": 1.3694,
"step": 336
},
{
"epoch": 1.3643724696356276,
"grad_norm": 0.5530371647784502,
"learning_rate": 3.6480654972802045e-05,
"loss": 1.3703,
"step": 337
},
{
"epoch": 1.368421052631579,
"grad_norm": 0.5670802991341252,
"learning_rate": 3.644854871713611e-05,
"loss": 1.3854,
"step": 338
},
{
"epoch": 1.3724696356275303,
"grad_norm": 0.5493082083946805,
"learning_rate": 3.6416310939312924e-05,
"loss": 1.3812,
"step": 339
},
{
"epoch": 1.376518218623482,
"grad_norm": 0.5307184305007718,
"learning_rate": 3.638394189710493e-05,
"loss": 1.3746,
"step": 340
},
{
"epoch": 1.3805668016194332,
"grad_norm": 0.5093074571774802,
"learning_rate": 3.635144184933412e-05,
"loss": 1.3777,
"step": 341
},
{
"epoch": 1.3846153846153846,
"grad_norm": 0.5900800573134718,
"learning_rate": 3.6318811055869994e-05,
"loss": 1.4063,
"step": 342
},
{
"epoch": 1.3886639676113361,
"grad_norm": 0.5310914676672811,
"learning_rate": 3.628604977762752e-05,
"loss": 1.3845,
"step": 343
},
{
"epoch": 1.3927125506072875,
"grad_norm": 0.6185292930618294,
"learning_rate": 3.6253158276565004e-05,
"loss": 1.3454,
"step": 344
},
{
"epoch": 1.3967611336032388,
"grad_norm": 0.5378676954658407,
"learning_rate": 3.6220136815682016e-05,
"loss": 1.4096,
"step": 345
},
{
"epoch": 1.4008097165991904,
"grad_norm": 0.6033847808474099,
"learning_rate": 3.618698565901727e-05,
"loss": 1.3671,
"step": 346
},
{
"epoch": 1.4048582995951417,
"grad_norm": 0.6147573513660067,
"learning_rate": 3.615370507164654e-05,
"loss": 1.3859,
"step": 347
},
{
"epoch": 1.408906882591093,
"grad_norm": 0.5103086719924701,
"learning_rate": 3.61202953196805e-05,
"loss": 1.4,
"step": 348
},
{
"epoch": 1.4129554655870447,
"grad_norm": 0.5243806866220297,
"learning_rate": 3.6086756670262654e-05,
"loss": 1.3855,
"step": 349
},
{
"epoch": 1.417004048582996,
"grad_norm": 0.5588336099097755,
"learning_rate": 3.6053089391567135e-05,
"loss": 1.3628,
"step": 350
},
{
"epoch": 1.4210526315789473,
"grad_norm": 0.5912189191907244,
"learning_rate": 3.601929375279662e-05,
"loss": 1.3857,
"step": 351
},
{
"epoch": 1.425101214574899,
"grad_norm": 0.5872851988093222,
"learning_rate": 3.598537002418012e-05,
"loss": 1.4177,
"step": 352
},
{
"epoch": 1.4291497975708503,
"grad_norm": 0.5547662524198659,
"learning_rate": 3.5951318476970885e-05,
"loss": 1.3572,
"step": 353
},
{
"epoch": 1.4331983805668016,
"grad_norm": 0.7334661739511349,
"learning_rate": 3.591713938344416e-05,
"loss": 1.3658,
"step": 354
},
{
"epoch": 1.4372469635627532,
"grad_norm": 0.5303222313164324,
"learning_rate": 3.588283301689507e-05,
"loss": 1.3727,
"step": 355
},
{
"epoch": 1.4412955465587045,
"grad_norm": 0.6286587074524215,
"learning_rate": 3.5848399651636424e-05,
"loss": 1.4096,
"step": 356
},
{
"epoch": 1.4453441295546559,
"grad_norm": 0.5021175256487025,
"learning_rate": 3.5813839562996473e-05,
"loss": 1.3624,
"step": 357
},
{
"epoch": 1.4493927125506074,
"grad_norm": 0.6308369325223174,
"learning_rate": 3.5779153027316775e-05,
"loss": 1.3744,
"step": 358
},
{
"epoch": 1.4534412955465588,
"grad_norm": 0.5750096146348971,
"learning_rate": 3.574434032194994e-05,
"loss": 1.393,
"step": 359
},
{
"epoch": 1.45748987854251,
"grad_norm": 0.5142664830270662,
"learning_rate": 3.5709401725257436e-05,
"loss": 1.4,
"step": 360
},
{
"epoch": 1.4615384615384617,
"grad_norm": 0.5002157429613521,
"learning_rate": 3.5674337516607346e-05,
"loss": 1.3795,
"step": 361
},
{
"epoch": 1.465587044534413,
"grad_norm": 0.5686826544345225,
"learning_rate": 3.563914797637216e-05,
"loss": 1.3985,
"step": 362
},
{
"epoch": 1.4696356275303644,
"grad_norm": 0.6087211850191244,
"learning_rate": 3.560383338592649e-05,
"loss": 1.3723,
"step": 363
},
{
"epoch": 1.4736842105263157,
"grad_norm": 0.501824770929623,
"learning_rate": 3.556839402764487e-05,
"loss": 1.3632,
"step": 364
},
{
"epoch": 1.4777327935222673,
"grad_norm": 0.5423363628827459,
"learning_rate": 3.553283018489946e-05,
"loss": 1.3704,
"step": 365
},
{
"epoch": 1.4817813765182186,
"grad_norm": 0.5341763158157213,
"learning_rate": 3.54971421420578e-05,
"loss": 1.3291,
"step": 366
},
{
"epoch": 1.48582995951417,
"grad_norm": 0.5827195165994913,
"learning_rate": 3.5461330184480524e-05,
"loss": 1.3841,
"step": 367
},
{
"epoch": 1.4898785425101215,
"grad_norm": 0.5190320821242866,
"learning_rate": 3.54253945985191e-05,
"loss": 1.3407,
"step": 368
},
{
"epoch": 1.4939271255060729,
"grad_norm": 0.5970221096177575,
"learning_rate": 3.5389335671513524e-05,
"loss": 1.389,
"step": 369
},
{
"epoch": 1.4979757085020242,
"grad_norm": 0.6213617214159624,
"learning_rate": 3.5353153691789986e-05,
"loss": 1.3405,
"step": 370
},
{
"epoch": 1.5020242914979756,
"grad_norm": 0.47863468999095776,
"learning_rate": 3.531684894865866e-05,
"loss": 1.3571,
"step": 371
},
{
"epoch": 1.5060728744939271,
"grad_norm": 0.622340547024952,
"learning_rate": 3.528042173241128e-05,
"loss": 1.3961,
"step": 372
},
{
"epoch": 1.5101214574898787,
"grad_norm": 0.7204356098493967,
"learning_rate": 3.524387233431892e-05,
"loss": 1.3638,
"step": 373
},
{
"epoch": 1.5141700404858298,
"grad_norm": 0.5850418285048404,
"learning_rate": 3.520720104662958e-05,
"loss": 1.3462,
"step": 374
},
{
"epoch": 1.5182186234817814,
"grad_norm": 0.6223443716624953,
"learning_rate": 3.5170408162565904e-05,
"loss": 1.3579,
"step": 375
},
{
"epoch": 1.522267206477733,
"grad_norm": 0.704082210186702,
"learning_rate": 3.513349397632281e-05,
"loss": 1.3717,
"step": 376
},
{
"epoch": 1.526315789473684,
"grad_norm": 0.8107530513032872,
"learning_rate": 3.5096458783065145e-05,
"loss": 1.3785,
"step": 377
},
{
"epoch": 1.5303643724696356,
"grad_norm": 0.5413519440434359,
"learning_rate": 3.505930287892533e-05,
"loss": 1.3621,
"step": 378
},
{
"epoch": 1.5344129554655872,
"grad_norm": 0.5363239737840794,
"learning_rate": 3.502202656100099e-05,
"loss": 1.3485,
"step": 379
},
{
"epoch": 1.5384615384615383,
"grad_norm": 0.5464741975070547,
"learning_rate": 3.498463012735256e-05,
"loss": 1.4154,
"step": 380
},
{
"epoch": 1.54251012145749,
"grad_norm": 0.5114675472494571,
"learning_rate": 3.494711387700094e-05,
"loss": 1.3429,
"step": 381
},
{
"epoch": 1.5465587044534415,
"grad_norm": 0.44361704325480816,
"learning_rate": 3.490947810992505e-05,
"loss": 1.3724,
"step": 382
},
{
"epoch": 1.5506072874493926,
"grad_norm": 0.5598354308607173,
"learning_rate": 3.48717231270595e-05,
"loss": 1.3708,
"step": 383
},
{
"epoch": 1.5546558704453441,
"grad_norm": 0.4513814826318252,
"learning_rate": 3.483384923029211e-05,
"loss": 1.3439,
"step": 384
},
{
"epoch": 1.5587044534412957,
"grad_norm": 0.6637441803549535,
"learning_rate": 3.479585672246156e-05,
"loss": 1.3988,
"step": 385
},
{
"epoch": 1.5627530364372468,
"grad_norm": 0.5873164779052186,
"learning_rate": 3.4757745907354924e-05,
"loss": 1.3207,
"step": 386
},
{
"epoch": 1.5668016194331984,
"grad_norm": 0.6351572935021823,
"learning_rate": 3.471951708970527e-05,
"loss": 1.3774,
"step": 387
},
{
"epoch": 1.5708502024291497,
"grad_norm": 0.539899314563376,
"learning_rate": 3.468117057518921e-05,
"loss": 1.3833,
"step": 388
},
{
"epoch": 1.574898785425101,
"grad_norm": 0.6344658578628452,
"learning_rate": 3.4642706670424426e-05,
"loss": 1.3589,
"step": 389
},
{
"epoch": 1.5789473684210527,
"grad_norm": 0.47200478341546054,
"learning_rate": 3.460412568296731e-05,
"loss": 1.3393,
"step": 390
},
{
"epoch": 1.582995951417004,
"grad_norm": 0.7009336400167785,
"learning_rate": 3.456542792131039e-05,
"loss": 1.4019,
"step": 391
},
{
"epoch": 1.5870445344129553,
"grad_norm": 0.6850801306106404,
"learning_rate": 3.4526613694879936e-05,
"loss": 1.421,
"step": 392
},
{
"epoch": 1.591093117408907,
"grad_norm": 0.5820910666853931,
"learning_rate": 3.448768331403348e-05,
"loss": 1.4005,
"step": 393
},
{
"epoch": 1.5951417004048583,
"grad_norm": 0.5621083562536767,
"learning_rate": 3.44486370900573e-05,
"loss": 1.3585,
"step": 394
},
{
"epoch": 1.5991902834008096,
"grad_norm": 0.5545675465358257,
"learning_rate": 3.440947533516396e-05,
"loss": 1.3918,
"step": 395
},
{
"epoch": 1.6032388663967612,
"grad_norm": 0.6286978024323933,
"learning_rate": 3.4370198362489816e-05,
"loss": 1.4129,
"step": 396
},
{
"epoch": 1.6072874493927125,
"grad_norm": 0.5816349496003744,
"learning_rate": 3.43308064860925e-05,
"loss": 1.3652,
"step": 397
},
{
"epoch": 1.6113360323886639,
"grad_norm": 0.5790548627018058,
"learning_rate": 3.4291300020948393e-05,
"loss": 1.3559,
"step": 398
},
{
"epoch": 1.6153846153846154,
"grad_norm": 0.576723275475117,
"learning_rate": 3.425167928295015e-05,
"loss": 1.3561,
"step": 399
},
{
"epoch": 1.6194331983805668,
"grad_norm": 0.7639689269041396,
"learning_rate": 3.421194458890411e-05,
"loss": 1.3695,
"step": 400
},
{
"epoch": 1.623481781376518,
"grad_norm": 0.4856048459407262,
"learning_rate": 3.4172096256527865e-05,
"loss": 1.3431,
"step": 401
},
{
"epoch": 1.6275303643724697,
"grad_norm": 0.676103445472939,
"learning_rate": 3.4132134604447595e-05,
"loss": 1.4001,
"step": 402
},
{
"epoch": 1.631578947368421,
"grad_norm": 0.664491502080572,
"learning_rate": 3.409205995219561e-05,
"loss": 1.3656,
"step": 403
},
{
"epoch": 1.6356275303643724,
"grad_norm": 0.5401115119801213,
"learning_rate": 3.4051872620207765e-05,
"loss": 1.3821,
"step": 404
},
{
"epoch": 1.639676113360324,
"grad_norm": 0.5818741354517444,
"learning_rate": 3.40115729298209e-05,
"loss": 1.3866,
"step": 405
},
{
"epoch": 1.6437246963562753,
"grad_norm": 0.5632198773763333,
"learning_rate": 3.397116120327025e-05,
"loss": 1.4389,
"step": 406
},
{
"epoch": 1.6477732793522266,
"grad_norm": 0.5893395439146693,
"learning_rate": 3.3930637763686916e-05,
"loss": 1.401,
"step": 407
},
{
"epoch": 1.6518218623481782,
"grad_norm": 0.5381503546318057,
"learning_rate": 3.389000293509524e-05,
"loss": 1.3992,
"step": 408
},
{
"epoch": 1.6558704453441295,
"grad_norm": 0.5804024478265414,
"learning_rate": 3.3849257042410225e-05,
"loss": 1.3811,
"step": 409
},
{
"epoch": 1.6599190283400809,
"grad_norm": 0.49079752534749416,
"learning_rate": 3.380840041143494e-05,
"loss": 1.365,
"step": 410
},
{
"epoch": 1.6639676113360324,
"grad_norm": 0.48909931793202643,
"learning_rate": 3.376743336885792e-05,
"loss": 1.3941,
"step": 411
},
{
"epoch": 1.6680161943319838,
"grad_norm": 0.5370962688987705,
"learning_rate": 3.372635624225055e-05,
"loss": 1.3587,
"step": 412
},
{
"epoch": 1.6720647773279351,
"grad_norm": 0.588010764279834,
"learning_rate": 3.3685169360064445e-05,
"loss": 1.3951,
"step": 413
},
{
"epoch": 1.6761133603238867,
"grad_norm": 0.5829395860019619,
"learning_rate": 3.3643873051628804e-05,
"loss": 1.3947,
"step": 414
},
{
"epoch": 1.680161943319838,
"grad_norm": 0.5720045303392852,
"learning_rate": 3.3602467647147814e-05,
"loss": 1.3712,
"step": 415
},
{
"epoch": 1.6842105263157894,
"grad_norm": 0.5995456538782289,
"learning_rate": 3.356095347769799e-05,
"loss": 1.3663,
"step": 416
},
{
"epoch": 1.688259109311741,
"grad_norm": 0.5529000283676447,
"learning_rate": 3.351933087522552e-05,
"loss": 1.3863,
"step": 417
},
{
"epoch": 1.6923076923076923,
"grad_norm": 0.5649724501932367,
"learning_rate": 3.3477600172543626e-05,
"loss": 1.4228,
"step": 418
},
{
"epoch": 1.6963562753036436,
"grad_norm": 0.6665543341965108,
"learning_rate": 3.3435761703329894e-05,
"loss": 1.3586,
"step": 419
},
{
"epoch": 1.7004048582995952,
"grad_norm": 0.49939182882920896,
"learning_rate": 3.3393815802123606e-05,
"loss": 1.3822,
"step": 420
},
{
"epoch": 1.7044534412955465,
"grad_norm": 0.6139706011449371,
"learning_rate": 3.335176280432308e-05,
"loss": 1.3782,
"step": 421
},
{
"epoch": 1.708502024291498,
"grad_norm": 0.5096477858072345,
"learning_rate": 3.330960304618295e-05,
"loss": 1.3883,
"step": 422
},
{
"epoch": 1.7125506072874495,
"grad_norm": 0.46915374164003093,
"learning_rate": 3.326733686481153e-05,
"loss": 1.3322,
"step": 423
},
{
"epoch": 1.7165991902834008,
"grad_norm": 0.49305332452091677,
"learning_rate": 3.3224964598168084e-05,
"loss": 1.3664,
"step": 424
},
{
"epoch": 1.7206477732793521,
"grad_norm": 0.4876554493960633,
"learning_rate": 3.318248658506012e-05,
"loss": 1.3722,
"step": 425
},
{
"epoch": 1.7246963562753037,
"grad_norm": 0.5399962734743278,
"learning_rate": 3.313990316514069e-05,
"loss": 1.3964,
"step": 426
},
{
"epoch": 1.728744939271255,
"grad_norm": 0.49651098799150417,
"learning_rate": 3.309721467890571e-05,
"loss": 1.3865,
"step": 427
},
{
"epoch": 1.7327935222672064,
"grad_norm": 0.7139299941403744,
"learning_rate": 3.305442146769115e-05,
"loss": 1.3857,
"step": 428
},
{
"epoch": 1.736842105263158,
"grad_norm": 0.6975767193137785,
"learning_rate": 3.30115238736704e-05,
"loss": 1.3704,
"step": 429
},
{
"epoch": 1.7408906882591093,
"grad_norm": 0.5642142543900814,
"learning_rate": 3.296852223985148e-05,
"loss": 1.3467,
"step": 430
},
{
"epoch": 1.7449392712550607,
"grad_norm": 0.5862630587493917,
"learning_rate": 3.2925416910074315e-05,
"loss": 1.4016,
"step": 431
},
{
"epoch": 1.7489878542510122,
"grad_norm": 0.6806844471918478,
"learning_rate": 3.288220822900796e-05,
"loss": 1.3857,
"step": 432
},
{
"epoch": 1.7530364372469636,
"grad_norm": 0.5092900962329729,
"learning_rate": 3.283889654214788e-05,
"loss": 1.3835,
"step": 433
},
{
"epoch": 1.757085020242915,
"grad_norm": 0.566587524629417,
"learning_rate": 3.279548219581317e-05,
"loss": 1.3573,
"step": 434
},
{
"epoch": 1.7611336032388665,
"grad_norm": 0.5440507245432396,
"learning_rate": 3.275196553714379e-05,
"loss": 1.3957,
"step": 435
},
{
"epoch": 1.7651821862348178,
"grad_norm": 0.6121783858490832,
"learning_rate": 3.270834691409778e-05,
"loss": 1.353,
"step": 436
},
{
"epoch": 1.7692307692307692,
"grad_norm": 0.5528293652804136,
"learning_rate": 3.266462667544849e-05,
"loss": 1.3882,
"step": 437
},
{
"epoch": 1.7732793522267207,
"grad_norm": 0.5507759896413812,
"learning_rate": 3.262080517078178e-05,
"loss": 1.3913,
"step": 438
},
{
"epoch": 1.777327935222672,
"grad_norm": 0.5059704291153286,
"learning_rate": 3.257688275049323e-05,
"loss": 1.4079,
"step": 439
},
{
"epoch": 1.7813765182186234,
"grad_norm": 0.5725207945674248,
"learning_rate": 3.253285976578535e-05,
"loss": 1.3672,
"step": 440
},
{
"epoch": 1.785425101214575,
"grad_norm": 0.6497931345961506,
"learning_rate": 3.2488736568664756e-05,
"loss": 1.3598,
"step": 441
},
{
"epoch": 1.7894736842105263,
"grad_norm": 0.614917859348756,
"learning_rate": 3.244451351193935e-05,
"loss": 1.3777,
"step": 442
},
{
"epoch": 1.7935222672064777,
"grad_norm": 0.6521965147829103,
"learning_rate": 3.240019094921551e-05,
"loss": 1.4134,
"step": 443
},
{
"epoch": 1.7975708502024292,
"grad_norm": 0.6281959792605809,
"learning_rate": 3.235576923489527e-05,
"loss": 1.3601,
"step": 444
},
{
"epoch": 1.8016194331983806,
"grad_norm": 0.6418729945764865,
"learning_rate": 3.2311248724173476e-05,
"loss": 1.3905,
"step": 445
},
{
"epoch": 1.805668016194332,
"grad_norm": 0.8528523655881421,
"learning_rate": 3.226662977303494e-05,
"loss": 1.3864,
"step": 446
},
{
"epoch": 1.8097165991902835,
"grad_norm": 0.5068694871501774,
"learning_rate": 3.2221912738251626e-05,
"loss": 1.368,
"step": 447
},
{
"epoch": 1.8137651821862348,
"grad_norm": 0.6519140744549372,
"learning_rate": 3.2177097977379736e-05,
"loss": 1.3661,
"step": 448
},
{
"epoch": 1.8178137651821862,
"grad_norm": 0.5334847202626184,
"learning_rate": 3.21321858487569e-05,
"loss": 1.3306,
"step": 449
},
{
"epoch": 1.8218623481781377,
"grad_norm": 0.5337510202694541,
"learning_rate": 3.208717671149931e-05,
"loss": 1.3341,
"step": 450
},
{
"epoch": 1.825910931174089,
"grad_norm": 0.5072593699054928,
"learning_rate": 3.204207092549882e-05,
"loss": 1.3937,
"step": 451
},
{
"epoch": 1.8299595141700404,
"grad_norm": 0.5706084403515487,
"learning_rate": 3.19968688514201e-05,
"loss": 1.3897,
"step": 452
},
{
"epoch": 1.834008097165992,
"grad_norm": 0.5917245486560782,
"learning_rate": 3.195157085069774e-05,
"loss": 1.3845,
"step": 453
},
{
"epoch": 1.8380566801619433,
"grad_norm": 0.43756985952548216,
"learning_rate": 3.190617728553332e-05,
"loss": 1.3643,
"step": 454
},
{
"epoch": 1.8421052631578947,
"grad_norm": 0.5312525518873622,
"learning_rate": 3.1860688518892606e-05,
"loss": 1.3823,
"step": 455
},
{
"epoch": 1.8461538461538463,
"grad_norm": 0.5742824717929932,
"learning_rate": 3.181510491450254e-05,
"loss": 1.3075,
"step": 456
},
{
"epoch": 1.8502024291497976,
"grad_norm": 0.5412991830550173,
"learning_rate": 3.176942683684842e-05,
"loss": 1.3659,
"step": 457
},
{
"epoch": 1.854251012145749,
"grad_norm": 0.7508994825509778,
"learning_rate": 3.1723654651170934e-05,
"loss": 1.3713,
"step": 458
},
{
"epoch": 1.8582995951417005,
"grad_norm": 0.5150612492235142,
"learning_rate": 3.167778872346327e-05,
"loss": 1.3528,
"step": 459
},
{
"epoch": 1.8623481781376519,
"grad_norm": 0.5439880197437115,
"learning_rate": 3.163182942046815e-05,
"loss": 1.3791,
"step": 460
},
{
"epoch": 1.8663967611336032,
"grad_norm": 0.5127724051765892,
"learning_rate": 3.158577710967494e-05,
"loss": 1.4003,
"step": 461
},
{
"epoch": 1.8704453441295548,
"grad_norm": 0.5445738005355268,
"learning_rate": 3.153963215931669e-05,
"loss": 1.3932,
"step": 462
},
{
"epoch": 1.874493927125506,
"grad_norm": 0.592452111710725,
"learning_rate": 3.14933949383672e-05,
"loss": 1.3505,
"step": 463
},
{
"epoch": 1.8785425101214575,
"grad_norm": 0.5740087210668031,
"learning_rate": 3.144706581653805e-05,
"loss": 1.3381,
"step": 464
},
{
"epoch": 1.882591093117409,
"grad_norm": 0.4976233493388172,
"learning_rate": 3.140064516427566e-05,
"loss": 1.3805,
"step": 465
},
{
"epoch": 1.8866396761133604,
"grad_norm": 0.42727496560837147,
"learning_rate": 3.135413335275833e-05,
"loss": 1.3722,
"step": 466
},
{
"epoch": 1.8906882591093117,
"grad_norm": 0.5126871767818264,
"learning_rate": 3.130753075389327e-05,
"loss": 1.3509,
"step": 467
},
{
"epoch": 1.8947368421052633,
"grad_norm": 0.4010328076121951,
"learning_rate": 3.1260837740313625e-05,
"loss": 1.3827,
"step": 468
},
{
"epoch": 1.8987854251012146,
"grad_norm": 0.43835343945752325,
"learning_rate": 3.121405468537547e-05,
"loss": 1.387,
"step": 469
},
{
"epoch": 1.902834008097166,
"grad_norm": 0.5094169753313056,
"learning_rate": 3.116718196315488e-05,
"loss": 1.3608,
"step": 470
},
{
"epoch": 1.9068825910931175,
"grad_norm": 0.4935021256882946,
"learning_rate": 3.1120219948444894e-05,
"loss": 1.3834,
"step": 471
},
{
"epoch": 1.9109311740890689,
"grad_norm": 0.4277655373916513,
"learning_rate": 3.107316901675252e-05,
"loss": 1.3894,
"step": 472
},
{
"epoch": 1.9149797570850202,
"grad_norm": 0.5038651205590757,
"learning_rate": 3.1026029544295745e-05,
"loss": 1.3532,
"step": 473
},
{
"epoch": 1.9190283400809718,
"grad_norm": 0.4951535460138383,
"learning_rate": 3.097880190800057e-05,
"loss": 1.3834,
"step": 474
},
{
"epoch": 1.9230769230769231,
"grad_norm": 0.4511274934014721,
"learning_rate": 3.093148648549788e-05,
"loss": 1.3415,
"step": 475
},
{
"epoch": 1.9271255060728745,
"grad_norm": 0.4726880650766076,
"learning_rate": 3.088408365512055e-05,
"loss": 1.3787,
"step": 476
},
{
"epoch": 1.931174089068826,
"grad_norm": 0.5998625325350694,
"learning_rate": 3.083659379590034e-05,
"loss": 1.3814,
"step": 477
},
{
"epoch": 1.9352226720647774,
"grad_norm": 0.4938314705487526,
"learning_rate": 3.0789017287564913e-05,
"loss": 1.4029,
"step": 478
},
{
"epoch": 1.9392712550607287,
"grad_norm": 0.5512402883913451,
"learning_rate": 3.074135451053475e-05,
"loss": 1.3847,
"step": 479
},
{
"epoch": 1.9433198380566803,
"grad_norm": 0.5367033174141453,
"learning_rate": 3.069360584592015e-05,
"loss": 1.3712,
"step": 480
},
{
"epoch": 1.9473684210526314,
"grad_norm": 0.5512459746641305,
"learning_rate": 3.064577167551816e-05,
"loss": 1.3594,
"step": 481
},
{
"epoch": 1.951417004048583,
"grad_norm": 0.6546284697217643,
"learning_rate": 3.059785238180953e-05,
"loss": 1.3789,
"step": 482
},
{
"epoch": 1.9554655870445345,
"grad_norm": 0.4756525906604495,
"learning_rate": 3.054984834795566e-05,
"loss": 1.374,
"step": 483
},
{
"epoch": 1.9595141700404857,
"grad_norm": 0.4624876799697758,
"learning_rate": 3.0501759957795537e-05,
"loss": 1.3535,
"step": 484
},
{
"epoch": 1.9635627530364372,
"grad_norm": 0.5011308381983439,
"learning_rate": 3.0453587595842627e-05,
"loss": 1.3741,
"step": 485
},
{
"epoch": 1.9676113360323888,
"grad_norm": 0.44826949318433407,
"learning_rate": 3.0405331647281856e-05,
"loss": 1.3785,
"step": 486
},
{
"epoch": 1.97165991902834,
"grad_norm": 0.41357443542832406,
"learning_rate": 3.0356992497966508e-05,
"loss": 1.3618,
"step": 487
},
{
"epoch": 1.9757085020242915,
"grad_norm": 0.5460770061108906,
"learning_rate": 3.0308570534415137e-05,
"loss": 1.3844,
"step": 488
},
{
"epoch": 1.979757085020243,
"grad_norm": 0.5349874114220654,
"learning_rate": 3.0260066143808467e-05,
"loss": 1.3824,
"step": 489
},
{
"epoch": 1.9838056680161942,
"grad_norm": 0.689501406091512,
"learning_rate": 3.021147971398632e-05,
"loss": 1.372,
"step": 490
},
{
"epoch": 1.9878542510121457,
"grad_norm": 0.5350946808264438,
"learning_rate": 3.0162811633444478e-05,
"loss": 1.3467,
"step": 491
},
{
"epoch": 1.9919028340080973,
"grad_norm": 0.6837044313175963,
"learning_rate": 3.0114062291331635e-05,
"loss": 1.3995,
"step": 492
},
{
"epoch": 1.9959514170040484,
"grad_norm": 0.5101796645439566,
"learning_rate": 3.0065232077446217e-05,
"loss": 1.3756,
"step": 493
},
{
"epoch": 2.0,
"grad_norm": 0.736855814713231,
"learning_rate": 3.001632138223332e-05,
"loss": 1.3524,
"step": 494
},
{
"epoch": 2.0040485829959516,
"grad_norm": 0.6428081017986733,
"learning_rate": 2.9967330596781565e-05,
"loss": 1.2768,
"step": 495
},
{
"epoch": 2.0080971659919027,
"grad_norm": 0.6585603090970296,
"learning_rate": 2.9918260112819952e-05,
"loss": 1.2286,
"step": 496
},
{
"epoch": 2.0121457489878543,
"grad_norm": 0.7875977352495007,
"learning_rate": 2.9869110322714778e-05,
"loss": 1.3012,
"step": 497
},
{
"epoch": 2.016194331983806,
"grad_norm": 0.8256651249286182,
"learning_rate": 2.9819881619466447e-05,
"loss": 1.328,
"step": 498
},
{
"epoch": 2.020242914979757,
"grad_norm": 0.7252816201666361,
"learning_rate": 2.9770574396706364e-05,
"loss": 1.277,
"step": 499
},
{
"epoch": 2.0242914979757085,
"grad_norm": 0.7765763891176009,
"learning_rate": 2.972118904869376e-05,
"loss": 1.2714,
"step": 500
},
{
"epoch": 2.02834008097166,
"grad_norm": 0.7459120579178277,
"learning_rate": 2.9671725970312563e-05,
"loss": 1.3305,
"step": 501
},
{
"epoch": 2.032388663967611,
"grad_norm": 0.728984400167187,
"learning_rate": 2.9622185557068238e-05,
"loss": 1.3139,
"step": 502
},
{
"epoch": 2.0364372469635628,
"grad_norm": 0.9564634406766338,
"learning_rate": 2.95725682050846e-05,
"loss": 1.2955,
"step": 503
},
{
"epoch": 2.0404858299595143,
"grad_norm": 0.5789936492459943,
"learning_rate": 2.9522874311100684e-05,
"loss": 1.2918,
"step": 504
},
{
"epoch": 2.0445344129554655,
"grad_norm": 0.862518482640073,
"learning_rate": 2.9473104272467526e-05,
"loss": 1.2919,
"step": 505
},
{
"epoch": 2.048582995951417,
"grad_norm": 0.5334142822384859,
"learning_rate": 2.942325848714505e-05,
"loss": 1.2925,
"step": 506
},
{
"epoch": 2.0526315789473686,
"grad_norm": 0.6854570768019675,
"learning_rate": 2.9373337353698804e-05,
"loss": 1.2676,
"step": 507
},
{
"epoch": 2.0566801619433197,
"grad_norm": 0.6209937993674114,
"learning_rate": 2.932334127129686e-05,
"loss": 1.3086,
"step": 508
},
{
"epoch": 2.0607287449392713,
"grad_norm": 0.5327377474886907,
"learning_rate": 2.9273270639706547e-05,
"loss": 1.2645,
"step": 509
},
{
"epoch": 2.064777327935223,
"grad_norm": 0.7487658783717519,
"learning_rate": 2.922312585929131e-05,
"loss": 1.2741,
"step": 510
},
{
"epoch": 2.068825910931174,
"grad_norm": 0.5290664636560628,
"learning_rate": 2.9172907331007478e-05,
"loss": 1.2986,
"step": 511
},
{
"epoch": 2.0728744939271255,
"grad_norm": 0.6316474144468632,
"learning_rate": 2.9122615456401053e-05,
"loss": 1.3329,
"step": 512
},
{
"epoch": 2.076923076923077,
"grad_norm": 0.6477736725155299,
"learning_rate": 2.9072250637604535e-05,
"loss": 1.3237,
"step": 513
},
{
"epoch": 2.080971659919028,
"grad_norm": 0.526410086773937,
"learning_rate": 2.9021813277333667e-05,
"loss": 1.3008,
"step": 514
},
{
"epoch": 2.08502024291498,
"grad_norm": 0.5490986298796344,
"learning_rate": 2.8971303778884244e-05,
"loss": 1.2826,
"step": 515
},
{
"epoch": 2.0890688259109313,
"grad_norm": 0.6333216899424138,
"learning_rate": 2.8920722546128863e-05,
"loss": 1.2989,
"step": 516
},
{
"epoch": 2.0931174089068825,
"grad_norm": 0.5584213660417928,
"learning_rate": 2.887006998351372e-05,
"loss": 1.2903,
"step": 517
},
{
"epoch": 2.097165991902834,
"grad_norm": 0.5529201234068806,
"learning_rate": 2.881934649605535e-05,
"loss": 1.283,
"step": 518
},
{
"epoch": 2.1012145748987856,
"grad_norm": 0.6649614558795209,
"learning_rate": 2.8768552489337408e-05,
"loss": 1.3079,
"step": 519
},
{
"epoch": 2.1052631578947367,
"grad_norm": 0.4904932361015151,
"learning_rate": 2.8717688369507425e-05,
"loss": 1.2856,
"step": 520
},
{
"epoch": 2.1093117408906883,
"grad_norm": 0.7159761547084958,
"learning_rate": 2.8666754543273535e-05,
"loss": 1.2967,
"step": 521
},
{
"epoch": 2.11336032388664,
"grad_norm": 0.4943269821674619,
"learning_rate": 2.8615751417901273e-05,
"loss": 1.2854,
"step": 522
},
{
"epoch": 2.117408906882591,
"grad_norm": 0.5682359649121878,
"learning_rate": 2.856467940121025e-05,
"loss": 1.2812,
"step": 523
},
{
"epoch": 2.1214574898785425,
"grad_norm": 0.496094773249813,
"learning_rate": 2.8513538901570964e-05,
"loss": 1.2675,
"step": 524
},
{
"epoch": 2.125506072874494,
"grad_norm": 0.5699729044506303,
"learning_rate": 2.846233032790149e-05,
"loss": 1.2617,
"step": 525
},
{
"epoch": 2.1295546558704452,
"grad_norm": 0.5434434480993015,
"learning_rate": 2.84110540896642e-05,
"loss": 1.2442,
"step": 526
},
{
"epoch": 2.133603238866397,
"grad_norm": 0.7855839734797125,
"learning_rate": 2.8359710596862555e-05,
"loss": 1.2656,
"step": 527
},
{
"epoch": 2.1376518218623484,
"grad_norm": 0.45505617933943326,
"learning_rate": 2.8308300260037734e-05,
"loss": 1.2782,
"step": 528
},
{
"epoch": 2.1417004048582995,
"grad_norm": 0.5072059503758407,
"learning_rate": 2.8256823490265424e-05,
"loss": 1.2874,
"step": 529
},
{
"epoch": 2.145748987854251,
"grad_norm": 0.4517381694984218,
"learning_rate": 2.8205280699152515e-05,
"loss": 1.2642,
"step": 530
},
{
"epoch": 2.1497975708502026,
"grad_norm": 0.4635132952468857,
"learning_rate": 2.8153672298833777e-05,
"loss": 1.2593,
"step": 531
},
{
"epoch": 2.1538461538461537,
"grad_norm": 0.5337061127057077,
"learning_rate": 2.8101998701968615e-05,
"loss": 1.271,
"step": 532
},
{
"epoch": 2.1578947368421053,
"grad_norm": 0.5073116548451597,
"learning_rate": 2.805026032173773e-05,
"loss": 1.2607,
"step": 533
},
{
"epoch": 2.161943319838057,
"grad_norm": 0.5555526050618494,
"learning_rate": 2.7998457571839832e-05,
"loss": 1.2767,
"step": 534
},
{
"epoch": 2.165991902834008,
"grad_norm": 0.5992476668336323,
"learning_rate": 2.7946590866488322e-05,
"loss": 1.3162,
"step": 535
},
{
"epoch": 2.1700404858299596,
"grad_norm": 0.47880758251830824,
"learning_rate": 2.7894660620408e-05,
"loss": 1.3036,
"step": 536
},
{
"epoch": 2.174089068825911,
"grad_norm": 0.4849666437125043,
"learning_rate": 2.784266724883173e-05,
"loss": 1.2593,
"step": 537
},
{
"epoch": 2.1781376518218623,
"grad_norm": 0.4992339664794073,
"learning_rate": 2.7790611167497128e-05,
"loss": 1.3381,
"step": 538
},
{
"epoch": 2.182186234817814,
"grad_norm": 0.5462632094051942,
"learning_rate": 2.7738492792643224e-05,
"loss": 1.3082,
"step": 539
},
{
"epoch": 2.1862348178137654,
"grad_norm": 0.5009990189661042,
"learning_rate": 2.768631254100716e-05,
"loss": 1.2563,
"step": 540
},
{
"epoch": 2.1902834008097165,
"grad_norm": 0.513803741432287,
"learning_rate": 2.7634070829820836e-05,
"loss": 1.306,
"step": 541
},
{
"epoch": 2.194331983805668,
"grad_norm": 0.4935089650901549,
"learning_rate": 2.7581768076807592e-05,
"loss": 1.3011,
"step": 542
},
{
"epoch": 2.1983805668016196,
"grad_norm": 0.5087266098655004,
"learning_rate": 2.7529404700178832e-05,
"loss": 1.271,
"step": 543
},
{
"epoch": 2.2024291497975708,
"grad_norm": 0.515430351382149,
"learning_rate": 2.7476981118630735e-05,
"loss": 1.2938,
"step": 544
},
{
"epoch": 2.2064777327935223,
"grad_norm": 0.492124745059608,
"learning_rate": 2.742449775134085e-05,
"loss": 1.3152,
"step": 545
},
{
"epoch": 2.2105263157894735,
"grad_norm": 0.5051366704656183,
"learning_rate": 2.7371955017964788e-05,
"loss": 1.2769,
"step": 546
},
{
"epoch": 2.214574898785425,
"grad_norm": 0.4517465149281084,
"learning_rate": 2.7319353338632845e-05,
"loss": 1.2721,
"step": 547
},
{
"epoch": 2.2186234817813766,
"grad_norm": 0.41707172382128277,
"learning_rate": 2.7266693133946645e-05,
"loss": 1.2911,
"step": 548
},
{
"epoch": 2.2226720647773277,
"grad_norm": 0.48192864263872043,
"learning_rate": 2.7213974824975785e-05,
"loss": 1.2651,
"step": 549
},
{
"epoch": 2.2267206477732793,
"grad_norm": 0.4782312791720754,
"learning_rate": 2.7161198833254464e-05,
"loss": 1.3328,
"step": 550
},
{
"epoch": 2.230769230769231,
"grad_norm": 0.5103504177715468,
"learning_rate": 2.7108365580778095e-05,
"loss": 1.2699,
"step": 551
},
{
"epoch": 2.234817813765182,
"grad_norm": 0.473034557152573,
"learning_rate": 2.705547548999996e-05,
"loss": 1.2991,
"step": 552
},
{
"epoch": 2.2388663967611335,
"grad_norm": 0.5929169102083783,
"learning_rate": 2.7002528983827817e-05,
"loss": 1.3125,
"step": 553
},
{
"epoch": 2.242914979757085,
"grad_norm": 0.4676037755517163,
"learning_rate": 2.6949526485620517e-05,
"loss": 1.2843,
"step": 554
},
{
"epoch": 2.246963562753036,
"grad_norm": 0.5977011059889631,
"learning_rate": 2.6896468419184623e-05,
"loss": 1.3408,
"step": 555
},
{
"epoch": 2.251012145748988,
"grad_norm": 0.4765607039024156,
"learning_rate": 2.6843355208771024e-05,
"loss": 1.2641,
"step": 556
},
{
"epoch": 2.2550607287449393,
"grad_norm": 0.5439789246344054,
"learning_rate": 2.6790187279071536e-05,
"loss": 1.3067,
"step": 557
},
{
"epoch": 2.2591093117408905,
"grad_norm": 0.43995168715579025,
"learning_rate": 2.67369650552155e-05,
"loss": 1.2639,
"step": 558
},
{
"epoch": 2.263157894736842,
"grad_norm": 0.4826905299822355,
"learning_rate": 2.6683688962766407e-05,
"loss": 1.2607,
"step": 559
},
{
"epoch": 2.2672064777327936,
"grad_norm": 0.5502596778882005,
"learning_rate": 2.6630359427718468e-05,
"loss": 1.2645,
"step": 560
},
{
"epoch": 2.2712550607287447,
"grad_norm": 0.4424004369454904,
"learning_rate": 2.6576976876493225e-05,
"loss": 1.2186,
"step": 561
},
{
"epoch": 2.2753036437246963,
"grad_norm": 0.6420996670349446,
"learning_rate": 2.652354173593614e-05,
"loss": 1.2768,
"step": 562
},
{
"epoch": 2.279352226720648,
"grad_norm": 0.5164761324424306,
"learning_rate": 2.647005443331317e-05,
"loss": 1.2766,
"step": 563
},
{
"epoch": 2.283400809716599,
"grad_norm": 0.5374313844137613,
"learning_rate": 2.6416515396307354e-05,
"loss": 1.2746,
"step": 564
},
{
"epoch": 2.2874493927125505,
"grad_norm": 0.4478268793821207,
"learning_rate": 2.636292505301542e-05,
"loss": 1.283,
"step": 565
},
{
"epoch": 2.291497975708502,
"grad_norm": 0.48023810079066165,
"learning_rate": 2.630928383194431e-05,
"loss": 1.2655,
"step": 566
},
{
"epoch": 2.2955465587044532,
"grad_norm": 0.41729302581500705,
"learning_rate": 2.625559216200781e-05,
"loss": 1.2995,
"step": 567
},
{
"epoch": 2.299595141700405,
"grad_norm": 0.5407368009882402,
"learning_rate": 2.620185047252308e-05,
"loss": 1.2864,
"step": 568
},
{
"epoch": 2.3036437246963564,
"grad_norm": 0.51208515915542,
"learning_rate": 2.614805919320723e-05,
"loss": 1.282,
"step": 569
},
{
"epoch": 2.3076923076923075,
"grad_norm": 0.49879511103847346,
"learning_rate": 2.6094218754173895e-05,
"loss": 1.2685,
"step": 570
},
{
"epoch": 2.311740890688259,
"grad_norm": 0.6154043710312979,
"learning_rate": 2.60403295859298e-05,
"loss": 1.2794,
"step": 571
},
{
"epoch": 2.3157894736842106,
"grad_norm": 0.46557962659362445,
"learning_rate": 2.5986392119371297e-05,
"loss": 1.2888,
"step": 572
},
{
"epoch": 2.3198380566801617,
"grad_norm": 0.6238169599873594,
"learning_rate": 2.5932406785780928e-05,
"loss": 1.2845,
"step": 573
},
{
"epoch": 2.3238866396761133,
"grad_norm": 0.5483989816049093,
"learning_rate": 2.5878374016823987e-05,
"loss": 1.2974,
"step": 574
},
{
"epoch": 2.327935222672065,
"grad_norm": 0.5004978129429855,
"learning_rate": 2.5824294244545045e-05,
"loss": 1.3173,
"step": 575
},
{
"epoch": 2.331983805668016,
"grad_norm": 0.6125294583148055,
"learning_rate": 2.577016790136453e-05,
"loss": 1.2997,
"step": 576
},
{
"epoch": 2.3360323886639676,
"grad_norm": 0.49315737106486096,
"learning_rate": 2.5715995420075225e-05,
"loss": 1.2952,
"step": 577
},
{
"epoch": 2.340080971659919,
"grad_norm": 0.584718401213755,
"learning_rate": 2.5661777233838864e-05,
"loss": 1.2915,
"step": 578
},
{
"epoch": 2.3441295546558703,
"grad_norm": 0.5032542807212903,
"learning_rate": 2.5607513776182616e-05,
"loss": 1.2626,
"step": 579
},
{
"epoch": 2.348178137651822,
"grad_norm": 0.5534285747498516,
"learning_rate": 2.5553205480995633e-05,
"loss": 1.262,
"step": 580
},
{
"epoch": 2.3522267206477734,
"grad_norm": 0.673247412351994,
"learning_rate": 2.5498852782525604e-05,
"loss": 1.2939,
"step": 581
},
{
"epoch": 2.3562753036437245,
"grad_norm": 0.5294009232916105,
"learning_rate": 2.5444456115375252e-05,
"loss": 1.2768,
"step": 582
},
{
"epoch": 2.360323886639676,
"grad_norm": 0.7046360884973728,
"learning_rate": 2.539001591449889e-05,
"loss": 1.2948,
"step": 583
},
{
"epoch": 2.3643724696356276,
"grad_norm": 0.5731069988336823,
"learning_rate": 2.53355326151989e-05,
"loss": 1.2628,
"step": 584
},
{
"epoch": 2.3684210526315788,
"grad_norm": 0.6456572405823144,
"learning_rate": 2.5281006653122288e-05,
"loss": 1.2992,
"step": 585
},
{
"epoch": 2.3724696356275303,
"grad_norm": 0.49736492566874585,
"learning_rate": 2.5226438464257217e-05,
"loss": 1.2649,
"step": 586
},
{
"epoch": 2.376518218623482,
"grad_norm": 0.5312828223216624,
"learning_rate": 2.5171828484929447e-05,
"loss": 1.2774,
"step": 587
},
{
"epoch": 2.380566801619433,
"grad_norm": 0.5291882217232715,
"learning_rate": 2.5117177151798934e-05,
"loss": 1.2836,
"step": 588
},
{
"epoch": 2.3846153846153846,
"grad_norm": 0.510828929570924,
"learning_rate": 2.506248490185627e-05,
"loss": 1.3029,
"step": 589
},
{
"epoch": 2.388663967611336,
"grad_norm": 0.5483207269076205,
"learning_rate": 2.500775217241923e-05,
"loss": 1.304,
"step": 590
},
{
"epoch": 2.3927125506072873,
"grad_norm": 0.42263938872932444,
"learning_rate": 2.4952979401129266e-05,
"loss": 1.3156,
"step": 591
},
{
"epoch": 2.396761133603239,
"grad_norm": 0.5992275622214346,
"learning_rate": 2.4898167025948e-05,
"loss": 1.307,
"step": 592
},
{
"epoch": 2.4008097165991904,
"grad_norm": 0.39424753176450056,
"learning_rate": 2.4843315485153703e-05,
"loss": 1.3115,
"step": 593
},
{
"epoch": 2.4048582995951415,
"grad_norm": 0.5863624117791234,
"learning_rate": 2.4788425217337856e-05,
"loss": 1.3156,
"step": 594
},
{
"epoch": 2.408906882591093,
"grad_norm": 0.37479977464468234,
"learning_rate": 2.4733496661401572e-05,
"loss": 1.2892,
"step": 595
},
{
"epoch": 2.4129554655870447,
"grad_norm": 0.5111250580498063,
"learning_rate": 2.4678530256552114e-05,
"loss": 1.2886,
"step": 596
},
{
"epoch": 2.417004048582996,
"grad_norm": 0.37977334167941534,
"learning_rate": 2.46235264422994e-05,
"loss": 1.2714,
"step": 597
},
{
"epoch": 2.4210526315789473,
"grad_norm": 0.38510735063692003,
"learning_rate": 2.4568485658452452e-05,
"loss": 1.2881,
"step": 598
},
{
"epoch": 2.425101214574899,
"grad_norm": 0.43526541656156603,
"learning_rate": 2.451340834511592e-05,
"loss": 1.3067,
"step": 599
},
{
"epoch": 2.42914979757085,
"grad_norm": 0.4384059026249326,
"learning_rate": 2.4458294942686528e-05,
"loss": 1.2793,
"step": 600
},
{
"epoch": 2.4331983805668016,
"grad_norm": 0.4678762524158813,
"learning_rate": 2.440314589184958e-05,
"loss": 1.3167,
"step": 601
},
{
"epoch": 2.437246963562753,
"grad_norm": 0.5005589926223257,
"learning_rate": 2.434796163357541e-05,
"loss": 1.2916,
"step": 602
},
{
"epoch": 2.4412955465587043,
"grad_norm": 0.5087546102315266,
"learning_rate": 2.429274260911588e-05,
"loss": 1.324,
"step": 603
},
{
"epoch": 2.445344129554656,
"grad_norm": 0.47432023294648384,
"learning_rate": 2.423748926000084e-05,
"loss": 1.3099,
"step": 604
},
{
"epoch": 2.4493927125506074,
"grad_norm": 0.5512266709459933,
"learning_rate": 2.418220202803459e-05,
"loss": 1.3185,
"step": 605
},
{
"epoch": 2.4534412955465585,
"grad_norm": 0.41750425626537185,
"learning_rate": 2.412688135529237e-05,
"loss": 1.2772,
"step": 606
},
{
"epoch": 2.45748987854251,
"grad_norm": 0.5261850565972286,
"learning_rate": 2.4071527684116807e-05,
"loss": 1.263,
"step": 607
},
{
"epoch": 2.4615384615384617,
"grad_norm": 0.41491087954344746,
"learning_rate": 2.4016141457114372e-05,
"loss": 1.3188,
"step": 608
},
{
"epoch": 2.465587044534413,
"grad_norm": 0.578993211970881,
"learning_rate": 2.396072311715187e-05,
"loss": 1.2871,
"step": 609
},
{
"epoch": 2.4696356275303644,
"grad_norm": 0.4226575942698871,
"learning_rate": 2.3905273107352863e-05,
"loss": 1.2384,
"step": 610
},
{
"epoch": 2.473684210526316,
"grad_norm": 0.4569887115772879,
"learning_rate": 2.3849791871094157e-05,
"loss": 1.269,
"step": 611
},
{
"epoch": 2.477732793522267,
"grad_norm": 0.5048359359819701,
"learning_rate": 2.379427985200223e-05,
"loss": 1.3161,
"step": 612
},
{
"epoch": 2.4817813765182186,
"grad_norm": 0.45447090494294384,
"learning_rate": 2.373873749394972e-05,
"loss": 1.2951,
"step": 613
},
{
"epoch": 2.48582995951417,
"grad_norm": 0.4549382682757696,
"learning_rate": 2.3683165241051836e-05,
"loss": 1.283,
"step": 614
},
{
"epoch": 2.4898785425101213,
"grad_norm": 0.41476460235497226,
"learning_rate": 2.3627563537662823e-05,
"loss": 1.2833,
"step": 615
},
{
"epoch": 2.493927125506073,
"grad_norm": 0.4187428014834106,
"learning_rate": 2.3571932828372433e-05,
"loss": 1.3324,
"step": 616
},
{
"epoch": 2.4979757085020244,
"grad_norm": 0.40563568223222973,
"learning_rate": 2.351627355800233e-05,
"loss": 1.2959,
"step": 617
},
{
"epoch": 2.5020242914979756,
"grad_norm": 0.5414961601183284,
"learning_rate": 2.346058617160256e-05,
"loss": 1.2838,
"step": 618
},
{
"epoch": 2.506072874493927,
"grad_norm": 0.4457487272910913,
"learning_rate": 2.3404871114447976e-05,
"loss": 1.2634,
"step": 619
},
{
"epoch": 2.5101214574898787,
"grad_norm": 0.5653996825926697,
"learning_rate": 2.334912883203469e-05,
"loss": 1.2579,
"step": 620
},
{
"epoch": 2.51417004048583,
"grad_norm": 0.594505236742439,
"learning_rate": 2.3293359770076505e-05,
"loss": 1.3101,
"step": 621
},
{
"epoch": 2.5182186234817814,
"grad_norm": 0.42238304894959733,
"learning_rate": 2.323756437450136e-05,
"loss": 1.2768,
"step": 622
},
{
"epoch": 2.522267206477733,
"grad_norm": 0.4373105447810298,
"learning_rate": 2.3181743091447754e-05,
"loss": 1.3037,
"step": 623
},
{
"epoch": 2.526315789473684,
"grad_norm": 0.439265093077708,
"learning_rate": 2.3125896367261164e-05,
"loss": 1.2622,
"step": 624
},
{
"epoch": 2.5303643724696356,
"grad_norm": 0.5017990337721747,
"learning_rate": 2.307002464849052e-05,
"loss": 1.3019,
"step": 625
},
{
"epoch": 2.534412955465587,
"grad_norm": 0.4501992919587179,
"learning_rate": 2.30141283818846e-05,
"loss": 1.3094,
"step": 626
},
{
"epoch": 2.5384615384615383,
"grad_norm": 0.48304281584744735,
"learning_rate": 2.2958208014388462e-05,
"loss": 1.3002,
"step": 627
},
{
"epoch": 2.54251012145749,
"grad_norm": 0.6240091255397798,
"learning_rate": 2.2902263993139857e-05,
"loss": 1.284,
"step": 628
},
{
"epoch": 2.5465587044534415,
"grad_norm": 0.4001608216745486,
"learning_rate": 2.2846296765465708e-05,
"loss": 1.3118,
"step": 629
},
{
"epoch": 2.5506072874493926,
"grad_norm": 0.625175703949816,
"learning_rate": 2.2790306778878465e-05,
"loss": 1.2813,
"step": 630
},
{
"epoch": 2.554655870445344,
"grad_norm": 0.4837541737696685,
"learning_rate": 2.273429448107257e-05,
"loss": 1.3179,
"step": 631
},
{
"epoch": 2.5587044534412957,
"grad_norm": 0.5628089907105301,
"learning_rate": 2.2678260319920868e-05,
"loss": 1.2626,
"step": 632
},
{
"epoch": 2.562753036437247,
"grad_norm": 0.4833151472909849,
"learning_rate": 2.262220474347102e-05,
"loss": 1.323,
"step": 633
},
{
"epoch": 2.5668016194331984,
"grad_norm": 0.4363314142944152,
"learning_rate": 2.2566128199941918e-05,
"loss": 1.266,
"step": 634
},
{
"epoch": 2.57085020242915,
"grad_norm": 0.4552687079525855,
"learning_rate": 2.2510031137720104e-05,
"loss": 1.2838,
"step": 635
},
{
"epoch": 2.574898785425101,
"grad_norm": 0.4891344237116978,
"learning_rate": 2.2453914005356206e-05,
"loss": 1.2746,
"step": 636
},
{
"epoch": 2.5789473684210527,
"grad_norm": 0.45960292157100263,
"learning_rate": 2.2397777251561302e-05,
"loss": 1.2578,
"step": 637
},
{
"epoch": 2.582995951417004,
"grad_norm": 0.4621254980253756,
"learning_rate": 2.2341621325203393e-05,
"loss": 1.3022,
"step": 638
},
{
"epoch": 2.5870445344129553,
"grad_norm": 0.4546234699160453,
"learning_rate": 2.2285446675303764e-05,
"loss": 1.285,
"step": 639
},
{
"epoch": 2.591093117408907,
"grad_norm": 0.4250447472216337,
"learning_rate": 2.2229253751033423e-05,
"loss": 1.2997,
"step": 640
},
{
"epoch": 2.5951417004048585,
"grad_norm": 0.4564923281759718,
"learning_rate": 2.2173043001709492e-05,
"loss": 1.3275,
"step": 641
},
{
"epoch": 2.5991902834008096,
"grad_norm": 0.4361919203014018,
"learning_rate": 2.2116814876791627e-05,
"loss": 1.2828,
"step": 642
},
{
"epoch": 2.603238866396761,
"grad_norm": 0.43374990560527693,
"learning_rate": 2.2060569825878413e-05,
"loss": 1.2928,
"step": 643
},
{
"epoch": 2.6072874493927127,
"grad_norm": 0.38911428348122634,
"learning_rate": 2.2004308298703777e-05,
"loss": 1.2729,
"step": 644
},
{
"epoch": 2.611336032388664,
"grad_norm": 0.4394825205734053,
"learning_rate": 2.1948030745133396e-05,
"loss": 1.2593,
"step": 645
},
{
"epoch": 2.6153846153846154,
"grad_norm": 0.41199683972653856,
"learning_rate": 2.1891737615161083e-05,
"loss": 1.2803,
"step": 646
},
{
"epoch": 2.619433198380567,
"grad_norm": 0.4571831647916046,
"learning_rate": 2.1835429358905218e-05,
"loss": 1.3122,
"step": 647
},
{
"epoch": 2.623481781376518,
"grad_norm": 0.42837492455175225,
"learning_rate": 2.17791064266051e-05,
"loss": 1.2993,
"step": 648
},
{
"epoch": 2.6275303643724697,
"grad_norm": 0.41073685564986756,
"learning_rate": 2.172276926861741e-05,
"loss": 1.2703,
"step": 649
},
{
"epoch": 2.6315789473684212,
"grad_norm": 0.40947911477581495,
"learning_rate": 2.166641833541255e-05,
"loss": 1.2986,
"step": 650
},
{
"epoch": 2.6356275303643724,
"grad_norm": 0.43246175914572366,
"learning_rate": 2.161005407757109e-05,
"loss": 1.3095,
"step": 651
},
{
"epoch": 2.639676113360324,
"grad_norm": 0.4335042315640144,
"learning_rate": 2.155367694578013e-05,
"loss": 1.2986,
"step": 652
},
{
"epoch": 2.6437246963562755,
"grad_norm": 0.4137972990695582,
"learning_rate": 2.1497287390829712e-05,
"loss": 1.2739,
"step": 653
},
{
"epoch": 2.6477732793522266,
"grad_norm": 0.4598840619133685,
"learning_rate": 2.1440885863609208e-05,
"loss": 1.2988,
"step": 654
},
{
"epoch": 2.651821862348178,
"grad_norm": 0.440291447164743,
"learning_rate": 2.1384472815103737e-05,
"loss": 1.305,
"step": 655
},
{
"epoch": 2.6558704453441297,
"grad_norm": 0.4953821785191996,
"learning_rate": 2.1328048696390534e-05,
"loss": 1.2969,
"step": 656
},
{
"epoch": 2.659919028340081,
"grad_norm": 0.44342245437353,
"learning_rate": 2.1271613958635342e-05,
"loss": 1.256,
"step": 657
},
{
"epoch": 2.6639676113360324,
"grad_norm": 0.4998785209004264,
"learning_rate": 2.1215169053088825e-05,
"loss": 1.2874,
"step": 658
},
{
"epoch": 2.668016194331984,
"grad_norm": 0.5315464077539014,
"learning_rate": 2.1158714431082937e-05,
"loss": 1.2877,
"step": 659
},
{
"epoch": 2.672064777327935,
"grad_norm": 0.4225685829626049,
"learning_rate": 2.110225054402733e-05,
"loss": 1.3223,
"step": 660
},
{
"epoch": 2.6761133603238867,
"grad_norm": 0.46244434036397053,
"learning_rate": 2.1045777843405747e-05,
"loss": 1.2915,
"step": 661
},
{
"epoch": 2.6801619433198383,
"grad_norm": 0.4072584902907476,
"learning_rate": 2.09892967807724e-05,
"loss": 1.2993,
"step": 662
},
{
"epoch": 2.6842105263157894,
"grad_norm": 0.49686488483186225,
"learning_rate": 2.093280780774835e-05,
"loss": 1.3308,
"step": 663
},
{
"epoch": 2.688259109311741,
"grad_norm": 0.4072628264357235,
"learning_rate": 2.087631137601793e-05,
"loss": 1.2933,
"step": 664
},
{
"epoch": 2.6923076923076925,
"grad_norm": 0.50164280535948,
"learning_rate": 2.0819807937325104e-05,
"loss": 1.2945,
"step": 665
},
{
"epoch": 2.6963562753036436,
"grad_norm": 0.41814201680794066,
"learning_rate": 2.0763297943469847e-05,
"loss": 1.2481,
"step": 666
},
{
"epoch": 2.700404858299595,
"grad_norm": 0.530359481381438,
"learning_rate": 2.070678184630458e-05,
"loss": 1.2728,
"step": 667
},
{
"epoch": 2.7044534412955468,
"grad_norm": 0.4153000801503012,
"learning_rate": 2.06502600977305e-05,
"loss": 1.2636,
"step": 668
},
{
"epoch": 2.708502024291498,
"grad_norm": 0.4020950993530895,
"learning_rate": 2.0593733149694008e-05,
"loss": 1.2676,
"step": 669
},
{
"epoch": 2.7125506072874495,
"grad_norm": 0.45745505237514816,
"learning_rate": 2.0537201454183074e-05,
"loss": 1.2394,
"step": 670
},
{
"epoch": 2.716599190283401,
"grad_norm": 0.40952941250196034,
"learning_rate": 2.0480665463223626e-05,
"loss": 1.3068,
"step": 671
},
{
"epoch": 2.720647773279352,
"grad_norm": 0.4235751192737178,
"learning_rate": 2.042412562887594e-05,
"loss": 1.3138,
"step": 672
},
{
"epoch": 2.7246963562753037,
"grad_norm": 0.4687485618201297,
"learning_rate": 2.0367582403231033e-05,
"loss": 1.3115,
"step": 673
},
{
"epoch": 2.7287449392712553,
"grad_norm": 0.3996653664072773,
"learning_rate": 2.0311036238407023e-05,
"loss": 1.3098,
"step": 674
},
{
"epoch": 2.7327935222672064,
"grad_norm": 0.44428029817877485,
"learning_rate": 2.0254487586545534e-05,
"loss": 1.3067,
"step": 675
},
{
"epoch": 2.736842105263158,
"grad_norm": 0.39455167609357245,
"learning_rate": 2.0197936899808088e-05,
"loss": 1.2614,
"step": 676
},
{
"epoch": 2.7408906882591095,
"grad_norm": 0.4572106942420455,
"learning_rate": 2.0141384630372466e-05,
"loss": 1.291,
"step": 677
},
{
"epoch": 2.7449392712550607,
"grad_norm": 0.4568700481797311,
"learning_rate": 2.0084831230429106e-05,
"loss": 1.2849,
"step": 678
},
{
"epoch": 2.748987854251012,
"grad_norm": 0.41487214988721294,
"learning_rate": 2.0028277152177492e-05,
"loss": 1.2994,
"step": 679
},
{
"epoch": 2.753036437246964,
"grad_norm": 0.5463127737994212,
"learning_rate": 1.9971722847822518e-05,
"loss": 1.2918,
"step": 680
},
{
"epoch": 2.757085020242915,
"grad_norm": 0.38274698642759464,
"learning_rate": 1.99151687695709e-05,
"loss": 1.3094,
"step": 681
},
{
"epoch": 2.7611336032388665,
"grad_norm": 0.5043220348944546,
"learning_rate": 1.9858615369627537e-05,
"loss": 1.2985,
"step": 682
},
{
"epoch": 2.765182186234818,
"grad_norm": 0.4452662681265526,
"learning_rate": 1.9802063100191916e-05,
"loss": 1.3111,
"step": 683
},
{
"epoch": 2.769230769230769,
"grad_norm": 0.40699569128158536,
"learning_rate": 1.974551241345447e-05,
"loss": 1.2849,
"step": 684
},
{
"epoch": 2.7732793522267207,
"grad_norm": 0.41025092641057553,
"learning_rate": 1.9688963761592984e-05,
"loss": 1.2674,
"step": 685
},
{
"epoch": 2.7773279352226723,
"grad_norm": 0.4630547655554322,
"learning_rate": 1.963241759676897e-05,
"loss": 1.2532,
"step": 686
},
{
"epoch": 2.7813765182186234,
"grad_norm": 0.4066418826201579,
"learning_rate": 1.9575874371124066e-05,
"loss": 1.2886,
"step": 687
},
{
"epoch": 2.785425101214575,
"grad_norm": 0.39793930961024027,
"learning_rate": 1.951933453677638e-05,
"loss": 1.2551,
"step": 688
},
{
"epoch": 2.7894736842105265,
"grad_norm": 0.4516802700611041,
"learning_rate": 1.9462798545816932e-05,
"loss": 1.2978,
"step": 689
},
{
"epoch": 2.7935222672064777,
"grad_norm": 0.39873435621148495,
"learning_rate": 1.9406266850305996e-05,
"loss": 1.25,
"step": 690
},
{
"epoch": 2.7975708502024292,
"grad_norm": 0.44912110797925786,
"learning_rate": 1.9349739902269503e-05,
"loss": 1.2819,
"step": 691
},
{
"epoch": 2.801619433198381,
"grad_norm": 0.438761221295816,
"learning_rate": 1.9293218153695425e-05,
"loss": 1.295,
"step": 692
},
{
"epoch": 2.805668016194332,
"grad_norm": 0.4422978803954671,
"learning_rate": 1.9236702056530153e-05,
"loss": 1.2857,
"step": 693
},
{
"epoch": 2.8097165991902835,
"grad_norm": 0.4437092381436285,
"learning_rate": 1.9180192062674907e-05,
"loss": 1.2783,
"step": 694
},
{
"epoch": 2.813765182186235,
"grad_norm": 0.46507383624326953,
"learning_rate": 1.9123688623982076e-05,
"loss": 1.3042,
"step": 695
},
{
"epoch": 2.817813765182186,
"grad_norm": 0.40985505253941995,
"learning_rate": 1.906719219225165e-05,
"loss": 1.3203,
"step": 696
},
{
"epoch": 2.8218623481781377,
"grad_norm": 0.4125520792137272,
"learning_rate": 1.9010703219227603e-05,
"loss": 1.2848,
"step": 697
},
{
"epoch": 2.8259109311740893,
"grad_norm": 0.4141535418975292,
"learning_rate": 1.8954222156594256e-05,
"loss": 1.2933,
"step": 698
},
{
"epoch": 2.8299595141700404,
"grad_norm": 0.472010806024362,
"learning_rate": 1.8897749455972673e-05,
"loss": 1.271,
"step": 699
},
{
"epoch": 2.834008097165992,
"grad_norm": 0.3704148598335097,
"learning_rate": 1.884128556891707e-05,
"loss": 1.2736,
"step": 700
},
{
"epoch": 2.8380566801619436,
"grad_norm": 0.46957818091632014,
"learning_rate": 1.878483094691119e-05,
"loss": 1.2758,
"step": 701
},
{
"epoch": 2.8421052631578947,
"grad_norm": 0.3708305856578212,
"learning_rate": 1.8728386041364664e-05,
"loss": 1.3258,
"step": 702
},
{
"epoch": 2.8461538461538463,
"grad_norm": 0.4224951550074662,
"learning_rate": 1.867195130360947e-05,
"loss": 1.275,
"step": 703
},
{
"epoch": 2.850202429149798,
"grad_norm": 0.4045861767556225,
"learning_rate": 1.8615527184896263e-05,
"loss": 1.3331,
"step": 704
},
{
"epoch": 2.854251012145749,
"grad_norm": 0.4831987186453239,
"learning_rate": 1.8559114136390795e-05,
"loss": 1.2748,
"step": 705
},
{
"epoch": 2.8582995951417005,
"grad_norm": 0.3805292210283063,
"learning_rate": 1.8502712609170298e-05,
"loss": 1.2792,
"step": 706
},
{
"epoch": 2.862348178137652,
"grad_norm": 0.46777527395595736,
"learning_rate": 1.8446323054219876e-05,
"loss": 1.2669,
"step": 707
},
{
"epoch": 2.866396761133603,
"grad_norm": 0.4894543104271215,
"learning_rate": 1.838994592242891e-05,
"loss": 1.2977,
"step": 708
},
{
"epoch": 2.8704453441295548,
"grad_norm": 0.4061614067297631,
"learning_rate": 1.8333581664587453e-05,
"loss": 1.3176,
"step": 709
},
{
"epoch": 2.8744939271255063,
"grad_norm": 0.5187951654888106,
"learning_rate": 1.8277230731382593e-05,
"loss": 1.2783,
"step": 710
},
{
"epoch": 2.8785425101214575,
"grad_norm": 0.3666520574090162,
"learning_rate": 1.82208935733949e-05,
"loss": 1.2742,
"step": 711
},
{
"epoch": 2.882591093117409,
"grad_norm": 0.45731436086237276,
"learning_rate": 1.8164570641094793e-05,
"loss": 1.2616,
"step": 712
},
{
"epoch": 2.8866396761133606,
"grad_norm": 0.41026708652980015,
"learning_rate": 1.810826238483892e-05,
"loss": 1.3278,
"step": 713
},
{
"epoch": 2.8906882591093117,
"grad_norm": 0.44272043904090114,
"learning_rate": 1.8051969254866608e-05,
"loss": 1.2991,
"step": 714
},
{
"epoch": 2.8947368421052633,
"grad_norm": 0.3911433634910705,
"learning_rate": 1.7995691701296226e-05,
"loss": 1.303,
"step": 715
},
{
"epoch": 2.898785425101215,
"grad_norm": 0.4923327333289708,
"learning_rate": 1.7939430174121593e-05,
"loss": 1.3074,
"step": 716
},
{
"epoch": 2.902834008097166,
"grad_norm": 0.4164631830806745,
"learning_rate": 1.788318512320838e-05,
"loss": 1.295,
"step": 717
},
{
"epoch": 2.9068825910931175,
"grad_norm": 0.49168400071088597,
"learning_rate": 1.782695699829051e-05,
"loss": 1.2934,
"step": 718
},
{
"epoch": 2.910931174089069,
"grad_norm": 0.5133006500940642,
"learning_rate": 1.7770746248966587e-05,
"loss": 1.2818,
"step": 719
},
{
"epoch": 2.91497975708502,
"grad_norm": 0.5910859436785207,
"learning_rate": 1.7714553324696243e-05,
"loss": 1.2827,
"step": 720
},
{
"epoch": 2.919028340080972,
"grad_norm": 0.4127399475072077,
"learning_rate": 1.7658378674796614e-05,
"loss": 1.2896,
"step": 721
},
{
"epoch": 2.9230769230769234,
"grad_norm": 0.4598272764192443,
"learning_rate": 1.76022227484387e-05,
"loss": 1.284,
"step": 722
},
{
"epoch": 2.9271255060728745,
"grad_norm": 0.44373195242535174,
"learning_rate": 1.7546085994643807e-05,
"loss": 1.2792,
"step": 723
},
{
"epoch": 2.931174089068826,
"grad_norm": 0.5307293103536116,
"learning_rate": 1.7489968862279902e-05,
"loss": 1.292,
"step": 724
},
{
"epoch": 2.9352226720647776,
"grad_norm": 0.4057865128750227,
"learning_rate": 1.743387180005809e-05,
"loss": 1.2931,
"step": 725
},
{
"epoch": 2.9392712550607287,
"grad_norm": 0.4769868925022623,
"learning_rate": 1.737779525652899e-05,
"loss": 1.3502,
"step": 726
},
{
"epoch": 2.9433198380566803,
"grad_norm": 0.46654675044628174,
"learning_rate": 1.732173968007914e-05,
"loss": 1.2743,
"step": 727
},
{
"epoch": 2.9473684210526314,
"grad_norm": 0.5118736820125256,
"learning_rate": 1.7265705518927437e-05,
"loss": 1.2934,
"step": 728
},
{
"epoch": 2.951417004048583,
"grad_norm": 0.4131503539473802,
"learning_rate": 1.7209693221121542e-05,
"loss": 1.2732,
"step": 729
},
{
"epoch": 2.9554655870445345,
"grad_norm": 0.560058736836292,
"learning_rate": 1.7153703234534302e-05,
"loss": 1.2871,
"step": 730
},
{
"epoch": 2.9595141700404857,
"grad_norm": 0.3923426446853355,
"learning_rate": 1.709773600686015e-05,
"loss": 1.2714,
"step": 731
},
{
"epoch": 2.9635627530364372,
"grad_norm": 0.4950910697942294,
"learning_rate": 1.7041791985611548e-05,
"loss": 1.2978,
"step": 732
},
{
"epoch": 2.967611336032389,
"grad_norm": 0.42230607855658575,
"learning_rate": 1.69858716181154e-05,
"loss": 1.2554,
"step": 733
},
{
"epoch": 2.97165991902834,
"grad_norm": 0.5704428725517385,
"learning_rate": 1.692997535150948e-05,
"loss": 1.2863,
"step": 734
},
{
"epoch": 2.9757085020242915,
"grad_norm": 0.3901505173237882,
"learning_rate": 1.687410363273884e-05,
"loss": 1.2701,
"step": 735
},
{
"epoch": 2.979757085020243,
"grad_norm": 0.5221539635192982,
"learning_rate": 1.6818256908552257e-05,
"loss": 1.2822,
"step": 736
},
{
"epoch": 2.983805668016194,
"grad_norm": 0.4212396814583793,
"learning_rate": 1.6762435625498646e-05,
"loss": 1.3013,
"step": 737
},
{
"epoch": 2.9878542510121457,
"grad_norm": 0.4892037297247971,
"learning_rate": 1.67066402299235e-05,
"loss": 1.296,
"step": 738
},
{
"epoch": 2.9919028340080973,
"grad_norm": 0.42231718675273056,
"learning_rate": 1.6650871167965313e-05,
"loss": 1.2619,
"step": 739
},
{
"epoch": 2.9959514170040484,
"grad_norm": 0.42965605254988404,
"learning_rate": 1.6595128885552028e-05,
"loss": 1.2961,
"step": 740
},
{
"epoch": 3.0,
"grad_norm": 0.4374759985448403,
"learning_rate": 1.653941382839745e-05,
"loss": 1.2124,
"step": 741
},
{
"epoch": 3.0040485829959516,
"grad_norm": 0.6078394431011476,
"learning_rate": 1.6483726441997673e-05,
"loss": 1.1738,
"step": 742
},
{
"epoch": 3.0080971659919027,
"grad_norm": 0.5822759566734187,
"learning_rate": 1.642806717162757e-05,
"loss": 1.1621,
"step": 743
},
{
"epoch": 3.0121457489878543,
"grad_norm": 0.9178822121207427,
"learning_rate": 1.637243646233718e-05,
"loss": 1.2011,
"step": 744
},
{
"epoch": 3.016194331983806,
"grad_norm": 0.7746000759506798,
"learning_rate": 1.6316834758948174e-05,
"loss": 1.2139,
"step": 745
},
{
"epoch": 3.020242914979757,
"grad_norm": 0.7264824055237907,
"learning_rate": 1.6261262506050282e-05,
"loss": 1.1798,
"step": 746
},
{
"epoch": 3.0242914979757085,
"grad_norm": 0.6334659184397462,
"learning_rate": 1.620572014799777e-05,
"loss": 1.1732,
"step": 747
},
{
"epoch": 3.02834008097166,
"grad_norm": 0.8228333538317738,
"learning_rate": 1.6150208128905857e-05,
"loss": 1.1865,
"step": 748
},
{
"epoch": 3.032388663967611,
"grad_norm": 0.6769701262453409,
"learning_rate": 1.6094726892647147e-05,
"loss": 1.1679,
"step": 749
},
{
"epoch": 3.0364372469635628,
"grad_norm": 0.8228898396928098,
"learning_rate": 1.6039276882848135e-05,
"loss": 1.1575,
"step": 750
},
{
"epoch": 3.0404858299595143,
"grad_norm": 0.7846217960391885,
"learning_rate": 1.5983858542885635e-05,
"loss": 1.22,
"step": 751
},
{
"epoch": 3.0445344129554655,
"grad_norm": 0.7330166533389478,
"learning_rate": 1.5928472315883203e-05,
"loss": 1.1552,
"step": 752
},
{
"epoch": 3.048582995951417,
"grad_norm": 0.7768162421212228,
"learning_rate": 1.5873118644707633e-05,
"loss": 1.1871,
"step": 753
},
{
"epoch": 3.0526315789473686,
"grad_norm": 0.7860747854523958,
"learning_rate": 1.5817797971965413e-05,
"loss": 1.1786,
"step": 754
},
{
"epoch": 3.0566801619433197,
"grad_norm": 0.7242047045194603,
"learning_rate": 1.576251073999917e-05,
"loss": 1.1623,
"step": 755
},
{
"epoch": 3.0607287449392713,
"grad_norm": 0.7842993151144695,
"learning_rate": 1.5707257390884126e-05,
"loss": 1.1761,
"step": 756
},
{
"epoch": 3.064777327935223,
"grad_norm": 0.7634392744649442,
"learning_rate": 1.5652038366424595e-05,
"loss": 1.1979,
"step": 757
},
{
"epoch": 3.068825910931174,
"grad_norm": 0.6795220683579706,
"learning_rate": 1.5596854108150424e-05,
"loss": 1.2024,
"step": 758
},
{
"epoch": 3.0728744939271255,
"grad_norm": 0.7498388668597464,
"learning_rate": 1.5541705057313476e-05,
"loss": 1.1726,
"step": 759
},
{
"epoch": 3.076923076923077,
"grad_norm": 0.6241212839294635,
"learning_rate": 1.5486591654884086e-05,
"loss": 1.1631,
"step": 760
},
{
"epoch": 3.080971659919028,
"grad_norm": 0.6725423266097267,
"learning_rate": 1.543151434154755e-05,
"loss": 1.2001,
"step": 761
},
{
"epoch": 3.08502024291498,
"grad_norm": 0.6371954152604786,
"learning_rate": 1.5376473557700615e-05,
"loss": 1.1914,
"step": 762
},
{
"epoch": 3.0890688259109313,
"grad_norm": 0.6998180315836338,
"learning_rate": 1.5321469743447893e-05,
"loss": 1.158,
"step": 763
},
{
"epoch": 3.0931174089068825,
"grad_norm": 0.648838353535273,
"learning_rate": 1.5266503338598434e-05,
"loss": 1.1955,
"step": 764
},
{
"epoch": 3.097165991902834,
"grad_norm": 0.5217989295738109,
"learning_rate": 1.5211574782662149e-05,
"loss": 1.1839,
"step": 765
},
{
"epoch": 3.1012145748987856,
"grad_norm": 0.6443386099369397,
"learning_rate": 1.5156684514846307e-05,
"loss": 1.1832,
"step": 766
},
{
"epoch": 3.1052631578947367,
"grad_norm": 0.5782989855437146,
"learning_rate": 1.5101832974052012e-05,
"loss": 1.1611,
"step": 767
},
{
"epoch": 3.1093117408906883,
"grad_norm": 0.5393244576157568,
"learning_rate": 1.5047020598870737e-05,
"loss": 1.1905,
"step": 768
},
{
"epoch": 3.11336032388664,
"grad_norm": 0.5296294039969957,
"learning_rate": 1.4992247827580778e-05,
"loss": 1.2174,
"step": 769
},
{
"epoch": 3.117408906882591,
"grad_norm": 0.5121522702959266,
"learning_rate": 1.493751509814374e-05,
"loss": 1.2009,
"step": 770
},
{
"epoch": 3.1214574898785425,
"grad_norm": 0.49841636633739556,
"learning_rate": 1.4882822848201073e-05,
"loss": 1.1788,
"step": 771
},
{
"epoch": 3.125506072874494,
"grad_norm": 0.4279309784298134,
"learning_rate": 1.4828171515070553e-05,
"loss": 1.1661,
"step": 772
},
{
"epoch": 3.1295546558704452,
"grad_norm": 0.44536139014205345,
"learning_rate": 1.4773561535742793e-05,
"loss": 1.174,
"step": 773
},
{
"epoch": 3.133603238866397,
"grad_norm": 0.47982809459792897,
"learning_rate": 1.4718993346877715e-05,
"loss": 1.1889,
"step": 774
},
{
"epoch": 3.1376518218623484,
"grad_norm": 0.4733467085646248,
"learning_rate": 1.466446738480111e-05,
"loss": 1.1647,
"step": 775
},
{
"epoch": 3.1417004048582995,
"grad_norm": 0.5061967274384327,
"learning_rate": 1.460998408550112e-05,
"loss": 1.1657,
"step": 776
},
{
"epoch": 3.145748987854251,
"grad_norm": 0.5030226510703024,
"learning_rate": 1.4555543884624751e-05,
"loss": 1.2145,
"step": 777
},
{
"epoch": 3.1497975708502026,
"grad_norm": 0.4332162690108985,
"learning_rate": 1.4501147217474402e-05,
"loss": 1.1871,
"step": 778
},
{
"epoch": 3.1538461538461537,
"grad_norm": 0.5084048946163269,
"learning_rate": 1.444679451900437e-05,
"loss": 1.1793,
"step": 779
},
{
"epoch": 3.1578947368421053,
"grad_norm": 0.4450345405076333,
"learning_rate": 1.4392486223817397e-05,
"loss": 1.2125,
"step": 780
},
{
"epoch": 3.161943319838057,
"grad_norm": 0.3955220256422019,
"learning_rate": 1.4338222766161143e-05,
"loss": 1.196,
"step": 781
},
{
"epoch": 3.165991902834008,
"grad_norm": 0.4667495338254247,
"learning_rate": 1.4284004579924777e-05,
"loss": 1.1767,
"step": 782
},
{
"epoch": 3.1700404858299596,
"grad_norm": 0.4106821733175521,
"learning_rate": 1.4229832098635479e-05,
"loss": 1.1651,
"step": 783
},
{
"epoch": 3.174089068825911,
"grad_norm": 0.37939660842448664,
"learning_rate": 1.4175705755454963e-05,
"loss": 1.205,
"step": 784
},
{
"epoch": 3.1781376518218623,
"grad_norm": 0.37451174181919894,
"learning_rate": 1.412162598317602e-05,
"loss": 1.2172,
"step": 785
},
{
"epoch": 3.182186234817814,
"grad_norm": 0.4208002525867822,
"learning_rate": 1.406759321421907e-05,
"loss": 1.2053,
"step": 786
},
{
"epoch": 3.1862348178137654,
"grad_norm": 0.37173574753830463,
"learning_rate": 1.401360788062871e-05,
"loss": 1.2158,
"step": 787
},
{
"epoch": 3.1902834008097165,
"grad_norm": 0.3864093100956334,
"learning_rate": 1.3959670414070204e-05,
"loss": 1.1925,
"step": 788
},
{
"epoch": 3.194331983805668,
"grad_norm": 0.4174469706414744,
"learning_rate": 1.3905781245826108e-05,
"loss": 1.1968,
"step": 789
},
{
"epoch": 3.1983805668016196,
"grad_norm": 0.4436685421583235,
"learning_rate": 1.3851940806792778e-05,
"loss": 1.2045,
"step": 790
},
{
"epoch": 3.2024291497975708,
"grad_norm": 0.42524844404935724,
"learning_rate": 1.379814952747693e-05,
"loss": 1.1454,
"step": 791
},
{
"epoch": 3.2064777327935223,
"grad_norm": 0.37066368394062255,
"learning_rate": 1.3744407837992193e-05,
"loss": 1.2006,
"step": 792
},
{
"epoch": 3.2105263157894735,
"grad_norm": 0.4099134895132954,
"learning_rate": 1.3690716168055692e-05,
"loss": 1.1439,
"step": 793
},
{
"epoch": 3.214574898785425,
"grad_norm": 0.4185750519704096,
"learning_rate": 1.363707494698459e-05,
"loss": 1.1719,
"step": 794
},
{
"epoch": 3.2186234817813766,
"grad_norm": 0.41878941475280046,
"learning_rate": 1.358348460369265e-05,
"loss": 1.1926,
"step": 795
},
{
"epoch": 3.2226720647773277,
"grad_norm": 0.42290007887575165,
"learning_rate": 1.352994556668684e-05,
"loss": 1.1836,
"step": 796
},
{
"epoch": 3.2267206477732793,
"grad_norm": 0.3940765851974513,
"learning_rate": 1.3476458264063865e-05,
"loss": 1.1845,
"step": 797
},
{
"epoch": 3.230769230769231,
"grad_norm": 0.43987531888703546,
"learning_rate": 1.342302312350678e-05,
"loss": 1.1649,
"step": 798
},
{
"epoch": 3.234817813765182,
"grad_norm": 0.37629660182215874,
"learning_rate": 1.3369640572281537e-05,
"loss": 1.1688,
"step": 799
},
{
"epoch": 3.2388663967611335,
"grad_norm": 0.37958309389341127,
"learning_rate": 1.3316311037233596e-05,
"loss": 1.1995,
"step": 800
},
{
"epoch": 3.242914979757085,
"grad_norm": 0.395341607068679,
"learning_rate": 1.326303494478451e-05,
"loss": 1.1943,
"step": 801
},
{
"epoch": 3.246963562753036,
"grad_norm": 0.4220094418201543,
"learning_rate": 1.3209812720928472e-05,
"loss": 1.2223,
"step": 802
},
{
"epoch": 3.251012145748988,
"grad_norm": 0.4245943690539546,
"learning_rate": 1.315664479122898e-05,
"loss": 1.1532,
"step": 803
},
{
"epoch": 3.2550607287449393,
"grad_norm": 0.43289578272681,
"learning_rate": 1.3103531580815378e-05,
"loss": 1.1795,
"step": 804
},
{
"epoch": 3.2591093117408905,
"grad_norm": 0.3546788854586895,
"learning_rate": 1.3050473514379488e-05,
"loss": 1.1878,
"step": 805
},
{
"epoch": 3.263157894736842,
"grad_norm": 0.4155804931691732,
"learning_rate": 1.2997471016172188e-05,
"loss": 1.2272,
"step": 806
},
{
"epoch": 3.2672064777327936,
"grad_norm": 0.40394931181454574,
"learning_rate": 1.2944524510000042e-05,
"loss": 1.1946,
"step": 807
},
{
"epoch": 3.2712550607287447,
"grad_norm": 0.3524464838036561,
"learning_rate": 1.2891634419221908e-05,
"loss": 1.1654,
"step": 808
},
{
"epoch": 3.2753036437246963,
"grad_norm": 0.3890699258761701,
"learning_rate": 1.2838801166745545e-05,
"loss": 1.238,
"step": 809
},
{
"epoch": 3.279352226720648,
"grad_norm": 0.4029451707916382,
"learning_rate": 1.2786025175024216e-05,
"loss": 1.1566,
"step": 810
},
{
"epoch": 3.283400809716599,
"grad_norm": 0.3681091655035043,
"learning_rate": 1.2733306866053357e-05,
"loss": 1.2204,
"step": 811
},
{
"epoch": 3.2874493927125505,
"grad_norm": 0.4369200040202172,
"learning_rate": 1.2680646661367163e-05,
"loss": 1.2175,
"step": 812
},
{
"epoch": 3.291497975708502,
"grad_norm": 0.39616151431217594,
"learning_rate": 1.2628044982035219e-05,
"loss": 1.1843,
"step": 813
},
{
"epoch": 3.2955465587044532,
"grad_norm": 0.4177617926818041,
"learning_rate": 1.2575502248659155e-05,
"loss": 1.1745,
"step": 814
},
{
"epoch": 3.299595141700405,
"grad_norm": 0.3795242166581197,
"learning_rate": 1.2523018881369269e-05,
"loss": 1.1884,
"step": 815
},
{
"epoch": 3.3036437246963564,
"grad_norm": 0.3931594916887269,
"learning_rate": 1.2470595299821176e-05,
"loss": 1.1896,
"step": 816
},
{
"epoch": 3.3076923076923075,
"grad_norm": 0.40343135903581423,
"learning_rate": 1.2418231923192416e-05,
"loss": 1.1936,
"step": 817
},
{
"epoch": 3.311740890688259,
"grad_norm": 0.3913552844407808,
"learning_rate": 1.2365929170179165e-05,
"loss": 1.1724,
"step": 818
},
{
"epoch": 3.3157894736842106,
"grad_norm": 0.37835657472007966,
"learning_rate": 1.2313687458992848e-05,
"loss": 1.2224,
"step": 819
},
{
"epoch": 3.3198380566801617,
"grad_norm": 0.39371972485013584,
"learning_rate": 1.2261507207356785e-05,
"loss": 1.1861,
"step": 820
},
{
"epoch": 3.3238866396761133,
"grad_norm": 0.3627368225730571,
"learning_rate": 1.220938883250288e-05,
"loss": 1.2168,
"step": 821
},
{
"epoch": 3.327935222672065,
"grad_norm": 0.3823885662444042,
"learning_rate": 1.215733275116827e-05,
"loss": 1.1852,
"step": 822
},
{
"epoch": 3.331983805668016,
"grad_norm": 0.41552999441657085,
"learning_rate": 1.2105339379592005e-05,
"loss": 1.2022,
"step": 823
},
{
"epoch": 3.3360323886639676,
"grad_norm": 0.36744231752627954,
"learning_rate": 1.2053409133511681e-05,
"loss": 1.2349,
"step": 824
},
{
"epoch": 3.340080971659919,
"grad_norm": 0.35983368150993633,
"learning_rate": 1.2001542428160173e-05,
"loss": 1.1785,
"step": 825
},
{
"epoch": 3.3441295546558703,
"grad_norm": 0.4032779007406651,
"learning_rate": 1.1949739678262278e-05,
"loss": 1.2163,
"step": 826
},
{
"epoch": 3.348178137651822,
"grad_norm": 0.39350549208081237,
"learning_rate": 1.189800129803139e-05,
"loss": 1.1714,
"step": 827
},
{
"epoch": 3.3522267206477734,
"grad_norm": 0.37446650731852527,
"learning_rate": 1.1846327701166228e-05,
"loss": 1.1798,
"step": 828
},
{
"epoch": 3.3562753036437245,
"grad_norm": 0.5171095026927718,
"learning_rate": 1.1794719300847493e-05,
"loss": 1.1542,
"step": 829
},
{
"epoch": 3.360323886639676,
"grad_norm": 0.34340915080015577,
"learning_rate": 1.1743176509734582e-05,
"loss": 1.2004,
"step": 830
},
{
"epoch": 3.3643724696356276,
"grad_norm": 0.4407521271830069,
"learning_rate": 1.1691699739962275e-05,
"loss": 1.1687,
"step": 831
},
{
"epoch": 3.3684210526315788,
"grad_norm": 0.4318829437010958,
"learning_rate": 1.1640289403137456e-05,
"loss": 1.2025,
"step": 832
},
{
"epoch": 3.3724696356275303,
"grad_norm": 0.36405175495923553,
"learning_rate": 1.1588945910335802e-05,
"loss": 1.1857,
"step": 833
},
{
"epoch": 3.376518218623482,
"grad_norm": 0.4087973896584326,
"learning_rate": 1.1537669672098517e-05,
"loss": 1.1986,
"step": 834
},
{
"epoch": 3.380566801619433,
"grad_norm": 0.3899128909575809,
"learning_rate": 1.1486461098429036e-05,
"loss": 1.2087,
"step": 835
},
{
"epoch": 3.3846153846153846,
"grad_norm": 0.4476558330615122,
"learning_rate": 1.1435320598789747e-05,
"loss": 1.1517,
"step": 836
},
{
"epoch": 3.388663967611336,
"grad_norm": 0.38510015118051155,
"learning_rate": 1.1384248582098736e-05,
"loss": 1.2044,
"step": 837
},
{
"epoch": 3.3927125506072873,
"grad_norm": 0.47855404304332183,
"learning_rate": 1.1333245456726467e-05,
"loss": 1.1315,
"step": 838
},
{
"epoch": 3.396761133603239,
"grad_norm": 0.3826331338459264,
"learning_rate": 1.128231163049258e-05,
"loss": 1.1535,
"step": 839
},
{
"epoch": 3.4008097165991904,
"grad_norm": 0.42693357955539246,
"learning_rate": 1.123144751066259e-05,
"loss": 1.17,
"step": 840
},
{
"epoch": 3.4048582995951415,
"grad_norm": 0.4441396010968801,
"learning_rate": 1.1180653503944657e-05,
"loss": 1.2113,
"step": 841
},
{
"epoch": 3.408906882591093,
"grad_norm": 0.4197453089772406,
"learning_rate": 1.1129930016486287e-05,
"loss": 1.2182,
"step": 842
},
{
"epoch": 3.4129554655870447,
"grad_norm": 0.3624340744210912,
"learning_rate": 1.107927745387114e-05,
"loss": 1.216,
"step": 843
},
{
"epoch": 3.417004048582996,
"grad_norm": 0.40486219789759254,
"learning_rate": 1.1028696221115766e-05,
"loss": 1.1726,
"step": 844
},
{
"epoch": 3.4210526315789473,
"grad_norm": 0.40162665763458183,
"learning_rate": 1.0978186722666341e-05,
"loss": 1.2168,
"step": 845
},
{
"epoch": 3.425101214574899,
"grad_norm": 0.39222242405311797,
"learning_rate": 1.0927749362395475e-05,
"loss": 1.1804,
"step": 846
},
{
"epoch": 3.42914979757085,
"grad_norm": 0.3741854924383862,
"learning_rate": 1.0877384543598954e-05,
"loss": 1.1896,
"step": 847
},
{
"epoch": 3.4331983805668016,
"grad_norm": 0.34799552219482055,
"learning_rate": 1.0827092668992529e-05,
"loss": 1.2228,
"step": 848
},
{
"epoch": 3.437246963562753,
"grad_norm": 0.41028813581245877,
"learning_rate": 1.077687414070869e-05,
"loss": 1.1466,
"step": 849
},
{
"epoch": 3.4412955465587043,
"grad_norm": 0.3723063294465725,
"learning_rate": 1.0726729360293451e-05,
"loss": 1.1396,
"step": 850
},
{
"epoch": 3.445344129554656,
"grad_norm": 0.3860876189554463,
"learning_rate": 1.0676658728703147e-05,
"loss": 1.1953,
"step": 851
},
{
"epoch": 3.4493927125506074,
"grad_norm": 0.35289651832984387,
"learning_rate": 1.0626662646301198e-05,
"loss": 1.1725,
"step": 852
},
{
"epoch": 3.4534412955465585,
"grad_norm": 0.4214247923773692,
"learning_rate": 1.0576741512854958e-05,
"loss": 1.2063,
"step": 853
},
{
"epoch": 3.45748987854251,
"grad_norm": 0.3634899803778953,
"learning_rate": 1.052689572753247e-05,
"loss": 1.1881,
"step": 854
},
{
"epoch": 3.4615384615384617,
"grad_norm": 0.46526397882486076,
"learning_rate": 1.0477125688899328e-05,
"loss": 1.186,
"step": 855
},
{
"epoch": 3.465587044534413,
"grad_norm": 0.38977790742836915,
"learning_rate": 1.0427431794915406e-05,
"loss": 1.1696,
"step": 856
},
{
"epoch": 3.4696356275303644,
"grad_norm": 0.3865900057921488,
"learning_rate": 1.0377814442931769e-05,
"loss": 1.1683,
"step": 857
},
{
"epoch": 3.473684210526316,
"grad_norm": 0.37714310485893704,
"learning_rate": 1.0328274029687439e-05,
"loss": 1.1845,
"step": 858
},
{
"epoch": 3.477732793522267,
"grad_norm": 0.382504539379788,
"learning_rate": 1.0278810951306253e-05,
"loss": 1.1713,
"step": 859
},
{
"epoch": 3.4817813765182186,
"grad_norm": 0.4019084290023694,
"learning_rate": 1.0229425603293647e-05,
"loss": 1.2189,
"step": 860
},
{
"epoch": 3.48582995951417,
"grad_norm": 0.3616503085328899,
"learning_rate": 1.0180118380533558e-05,
"loss": 1.2034,
"step": 861
},
{
"epoch": 3.4898785425101213,
"grad_norm": 0.34071971843734794,
"learning_rate": 1.0130889677285225e-05,
"loss": 1.1733,
"step": 862
},
{
"epoch": 3.493927125506073,
"grad_norm": 0.43726328159413497,
"learning_rate": 1.008173988718005e-05,
"loss": 1.1958,
"step": 863
},
{
"epoch": 3.4979757085020244,
"grad_norm": 0.3584166616977485,
"learning_rate": 1.0032669403218439e-05,
"loss": 1.1837,
"step": 864
},
{
"epoch": 3.5020242914979756,
"grad_norm": 0.3665514236103231,
"learning_rate": 9.983678617766677e-06,
"loss": 1.1914,
"step": 865
},
{
"epoch": 3.506072874493927,
"grad_norm": 0.361644589120467,
"learning_rate": 9.934767922553788e-06,
"loss": 1.1708,
"step": 866
},
{
"epoch": 3.5101214574898787,
"grad_norm": 0.38802024852522377,
"learning_rate": 9.885937708668371e-06,
"loss": 1.1518,
"step": 867
},
{
"epoch": 3.51417004048583,
"grad_norm": 0.4311932255057576,
"learning_rate": 9.837188366555524e-06,
"loss": 1.2089,
"step": 868
},
{
"epoch": 3.5182186234817814,
"grad_norm": 0.3739415458334539,
"learning_rate": 9.788520286013691e-06,
"loss": 1.19,
"step": 869
},
{
"epoch": 3.522267206477733,
"grad_norm": 0.41289441591081855,
"learning_rate": 9.739933856191537e-06,
"loss": 1.2084,
"step": 870
},
{
"epoch": 3.526315789473684,
"grad_norm": 0.418956037728057,
"learning_rate": 9.691429465584866e-06,
"loss": 1.1802,
"step": 871
},
{
"epoch": 3.5303643724696356,
"grad_norm": 0.372171809588552,
"learning_rate": 9.643007502033492e-06,
"loss": 1.1736,
"step": 872
},
{
"epoch": 3.534412955465587,
"grad_norm": 0.3511782084608693,
"learning_rate": 9.594668352718152e-06,
"loss": 1.1749,
"step": 873
},
{
"epoch": 3.5384615384615383,
"grad_norm": 0.413278367145341,
"learning_rate": 9.546412404157383e-06,
"loss": 1.1945,
"step": 874
},
{
"epoch": 3.54251012145749,
"grad_norm": 0.3935949600418023,
"learning_rate": 9.498240042204471e-06,
"loss": 1.2052,
"step": 875
},
{
"epoch": 3.5465587044534415,
"grad_norm": 0.3786232696311741,
"learning_rate": 9.450151652044339e-06,
"loss": 1.1738,
"step": 876
},
{
"epoch": 3.5506072874493926,
"grad_norm": 0.42430455949214885,
"learning_rate": 9.402147618190472e-06,
"loss": 1.2058,
"step": 877
},
{
"epoch": 3.554655870445344,
"grad_norm": 0.43892912236116155,
"learning_rate": 9.354228324481841e-06,
"loss": 1.1916,
"step": 878
},
{
"epoch": 3.5587044534412957,
"grad_norm": 0.3947343893087211,
"learning_rate": 9.306394154079852e-06,
"loss": 1.1774,
"step": 879
},
{
"epoch": 3.562753036437247,
"grad_norm": 0.3764184027484958,
"learning_rate": 9.258645489465254e-06,
"loss": 1.1581,
"step": 880
},
{
"epoch": 3.5668016194331984,
"grad_norm": 0.4151100088589348,
"learning_rate": 9.21098271243509e-06,
"loss": 1.2095,
"step": 881
},
{
"epoch": 3.57085020242915,
"grad_norm": 0.353624391958425,
"learning_rate": 9.163406204099659e-06,
"loss": 1.2262,
"step": 882
},
{
"epoch": 3.574898785425101,
"grad_norm": 0.3695452019471619,
"learning_rate": 9.115916344879452e-06,
"loss": 1.1687,
"step": 883
},
{
"epoch": 3.5789473684210527,
"grad_norm": 0.3912917143315633,
"learning_rate": 9.068513514502126e-06,
"loss": 1.1815,
"step": 884
},
{
"epoch": 3.582995951417004,
"grad_norm": 0.41291805587940283,
"learning_rate": 9.021198091999439e-06,
"loss": 1.1995,
"step": 885
},
{
"epoch": 3.5870445344129553,
"grad_norm": 0.365977609866257,
"learning_rate": 8.973970455704251e-06,
"loss": 1.2344,
"step": 886
},
{
"epoch": 3.591093117408907,
"grad_norm": 0.38402209015510935,
"learning_rate": 8.926830983247495e-06,
"loss": 1.1487,
"step": 887
},
{
"epoch": 3.5951417004048585,
"grad_norm": 0.38081810944477634,
"learning_rate": 8.879780051555116e-06,
"loss": 1.1795,
"step": 888
},
{
"epoch": 3.5991902834008096,
"grad_norm": 0.33676451740929736,
"learning_rate": 8.83281803684512e-06,
"loss": 1.1773,
"step": 889
},
{
"epoch": 3.603238866396761,
"grad_norm": 0.3546446297064312,
"learning_rate": 8.785945314624532e-06,
"loss": 1.2276,
"step": 890
},
{
"epoch": 3.6072874493927127,
"grad_norm": 0.4461174481779982,
"learning_rate": 8.739162259686385e-06,
"loss": 1.1992,
"step": 891
},
{
"epoch": 3.611336032388664,
"grad_norm": 0.3517360859584748,
"learning_rate": 8.692469246106734e-06,
"loss": 1.2324,
"step": 892
},
{
"epoch": 3.6153846153846154,
"grad_norm": 0.40712843550165967,
"learning_rate": 8.645866647241677e-06,
"loss": 1.2195,
"step": 893
},
{
"epoch": 3.619433198380567,
"grad_norm": 0.3715039485107766,
"learning_rate": 8.599354835724348e-06,
"loss": 1.177,
"step": 894
},
{
"epoch": 3.623481781376518,
"grad_norm": 0.37713836913916765,
"learning_rate": 8.552934183461958e-06,
"loss": 1.2152,
"step": 895
},
{
"epoch": 3.6275303643724697,
"grad_norm": 0.3324657048263331,
"learning_rate": 8.506605061632805e-06,
"loss": 1.1865,
"step": 896
},
{
"epoch": 3.6315789473684212,
"grad_norm": 0.35670251667576497,
"learning_rate": 8.460367840683308e-06,
"loss": 1.1838,
"step": 897
},
{
"epoch": 3.6356275303643724,
"grad_norm": 0.35373339263458337,
"learning_rate": 8.414222890325065e-06,
"loss": 1.1792,
"step": 898
},
{
"epoch": 3.639676113360324,
"grad_norm": 0.368100635944373,
"learning_rate": 8.368170579531855e-06,
"loss": 1.1879,
"step": 899
},
{
"epoch": 3.6437246963562755,
"grad_norm": 0.3810672732844277,
"learning_rate": 8.322211276536733e-06,
"loss": 1.2026,
"step": 900
},
{
"epoch": 3.6477732793522266,
"grad_norm": 0.39588796839302065,
"learning_rate": 8.27634534882907e-06,
"loss": 1.1847,
"step": 901
},
{
"epoch": 3.651821862348178,
"grad_norm": 0.3529028007902776,
"learning_rate": 8.230573163151585e-06,
"loss": 1.2115,
"step": 902
},
{
"epoch": 3.6558704453441297,
"grad_norm": 0.35690398879490076,
"learning_rate": 8.184895085497466e-06,
"loss": 1.1847,
"step": 903
},
{
"epoch": 3.659919028340081,
"grad_norm": 0.38509718499657175,
"learning_rate": 8.139311481107401e-06,
"loss": 1.1938,
"step": 904
},
{
"epoch": 3.6639676113360324,
"grad_norm": 0.3763822241603648,
"learning_rate": 8.093822714466686e-06,
"loss": 1.1857,
"step": 905
},
{
"epoch": 3.668016194331984,
"grad_norm": 0.3577962510698488,
"learning_rate": 8.048429149302271e-06,
"loss": 1.1972,
"step": 906
},
{
"epoch": 3.672064777327935,
"grad_norm": 0.3697958670005469,
"learning_rate": 8.003131148579902e-06,
"loss": 1.1739,
"step": 907
},
{
"epoch": 3.6761133603238867,
"grad_norm": 0.3356782124913771,
"learning_rate": 7.957929074501182e-06,
"loss": 1.1848,
"step": 908
},
{
"epoch": 3.6801619433198383,
"grad_norm": 0.36664182976010773,
"learning_rate": 7.912823288500695e-06,
"loss": 1.179,
"step": 909
},
{
"epoch": 3.6842105263157894,
"grad_norm": 0.3893054403167499,
"learning_rate": 7.867814151243102e-06,
"loss": 1.1657,
"step": 910
},
{
"epoch": 3.688259109311741,
"grad_norm": 0.4004652771234261,
"learning_rate": 7.822902022620267e-06,
"loss": 1.1959,
"step": 911
},
{
"epoch": 3.6923076923076925,
"grad_norm": 0.360572459724826,
"learning_rate": 7.778087261748377e-06,
"loss": 1.1943,
"step": 912
},
{
"epoch": 3.6963562753036436,
"grad_norm": 0.4008356705409762,
"learning_rate": 7.733370226965056e-06,
"loss": 1.1648,
"step": 913
},
{
"epoch": 3.700404858299595,
"grad_norm": 0.3779328467334184,
"learning_rate": 7.688751275826527e-06,
"loss": 1.1914,
"step": 914
},
{
"epoch": 3.7044534412955468,
"grad_norm": 0.3655997052560777,
"learning_rate": 7.644230765104735e-06,
"loss": 1.1876,
"step": 915
},
{
"epoch": 3.708502024291498,
"grad_norm": 0.3609911563006185,
"learning_rate": 7.599809050784502e-06,
"loss": 1.1944,
"step": 916
},
{
"epoch": 3.7125506072874495,
"grad_norm": 0.4044526377036624,
"learning_rate": 7.555486488060661e-06,
"loss": 1.1715,
"step": 917
},
{
"epoch": 3.716599190283401,
"grad_norm": 0.43011986943496094,
"learning_rate": 7.511263431335249e-06,
"loss": 1.1764,
"step": 918
},
{
"epoch": 3.720647773279352,
"grad_norm": 0.3551575435855887,
"learning_rate": 7.467140234214656e-06,
"loss": 1.1865,
"step": 919
},
{
"epoch": 3.7246963562753037,
"grad_norm": 0.3578124297022588,
"learning_rate": 7.423117249506777e-06,
"loss": 1.1938,
"step": 920
},
{
"epoch": 3.7287449392712553,
"grad_norm": 0.3675218911708793,
"learning_rate": 7.379194829218231e-06,
"loss": 1.2072,
"step": 921
},
{
"epoch": 3.7327935222672064,
"grad_norm": 0.37410111312280553,
"learning_rate": 7.335373324551518e-06,
"loss": 1.168,
"step": 922
},
{
"epoch": 3.736842105263158,
"grad_norm": 0.34054495574785054,
"learning_rate": 7.291653085902224e-06,
"loss": 1.1873,
"step": 923
},
{
"epoch": 3.7408906882591095,
"grad_norm": 0.4141426882950444,
"learning_rate": 7.248034462856213e-06,
"loss": 1.2031,
"step": 924
},
{
"epoch": 3.7449392712550607,
"grad_norm": 0.3426471623608404,
"learning_rate": 7.20451780418683e-06,
"loss": 1.165,
"step": 925
},
{
"epoch": 3.748987854251012,
"grad_norm": 0.33399616804686605,
"learning_rate": 7.161103457852128e-06,
"loss": 1.216,
"step": 926
},
{
"epoch": 3.753036437246964,
"grad_norm": 0.44411143532532205,
"learning_rate": 7.117791770992046e-06,
"loss": 1.1795,
"step": 927
},
{
"epoch": 3.757085020242915,
"grad_norm": 0.36724685281926245,
"learning_rate": 7.074583089925688e-06,
"loss": 1.1861,
"step": 928
},
{
"epoch": 3.7611336032388665,
"grad_norm": 0.3556005549562317,
"learning_rate": 7.0314777601485155e-06,
"loss": 1.1965,
"step": 929
},
{
"epoch": 3.765182186234818,
"grad_norm": 0.37230803644622834,
"learning_rate": 6.988476126329604e-06,
"loss": 1.1979,
"step": 930
},
{
"epoch": 3.769230769230769,
"grad_norm": 0.40772516838423445,
"learning_rate": 6.945578532308857e-06,
"loss": 1.1926,
"step": 931
},
{
"epoch": 3.7732793522267207,
"grad_norm": 0.37363090510914077,
"learning_rate": 6.902785321094301e-06,
"loss": 1.235,
"step": 932
},
{
"epoch": 3.7773279352226723,
"grad_norm": 0.3382507527281336,
"learning_rate": 6.8600968348593115e-06,
"loss": 1.1806,
"step": 933
},
{
"epoch": 3.7813765182186234,
"grad_norm": 0.35315887091062337,
"learning_rate": 6.817513414939892e-06,
"loss": 1.1511,
"step": 934
},
{
"epoch": 3.785425101214575,
"grad_norm": 0.40421032487942793,
"learning_rate": 6.775035401831924e-06,
"loss": 1.243,
"step": 935
},
{
"epoch": 3.7894736842105265,
"grad_norm": 0.3469716818880439,
"learning_rate": 6.732663135188473e-06,
"loss": 1.1798,
"step": 936
},
{
"epoch": 3.7935222672064777,
"grad_norm": 0.33056812471318264,
"learning_rate": 6.690396953817055e-06,
"loss": 1.1761,
"step": 937
},
{
"epoch": 3.7975708502024292,
"grad_norm": 0.3554717301734223,
"learning_rate": 6.648237195676927e-06,
"loss": 1.1861,
"step": 938
},
{
"epoch": 3.801619433198381,
"grad_norm": 0.3370427600657742,
"learning_rate": 6.606184197876395e-06,
"loss": 1.1999,
"step": 939
},
{
"epoch": 3.805668016194332,
"grad_norm": 0.3679577991789021,
"learning_rate": 6.564238296670109e-06,
"loss": 1.1972,
"step": 940
},
{
"epoch": 3.8097165991902835,
"grad_norm": 0.3459974433716084,
"learning_rate": 6.52239982745638e-06,
"loss": 1.161,
"step": 941
},
{
"epoch": 3.813765182186235,
"grad_norm": 0.3625437705877451,
"learning_rate": 6.480669124774484e-06,
"loss": 1.1745,
"step": 942
},
{
"epoch": 3.817813765182186,
"grad_norm": 0.3285879478288195,
"learning_rate": 6.439046522302013e-06,
"loss": 1.1507,
"step": 943
},
{
"epoch": 3.8218623481781377,
"grad_norm": 0.35380917109551774,
"learning_rate": 6.397532352852191e-06,
"loss": 1.2031,
"step": 944
},
{
"epoch": 3.8259109311740893,
"grad_norm": 0.3136097163995498,
"learning_rate": 6.356126948371202e-06,
"loss": 1.1998,
"step": 945
},
{
"epoch": 3.8299595141700404,
"grad_norm": 0.35767822477790395,
"learning_rate": 6.314830639935559e-06,
"loss": 1.1904,
"step": 946
},
{
"epoch": 3.834008097165992,
"grad_norm": 0.34431918204840883,
"learning_rate": 6.273643757749448e-06,
"loss": 1.204,
"step": 947
},
{
"epoch": 3.8380566801619436,
"grad_norm": 0.3350671963105035,
"learning_rate": 6.2325666311420856e-06,
"loss": 1.1991,
"step": 948
},
{
"epoch": 3.8421052631578947,
"grad_norm": 0.30041137771293336,
"learning_rate": 6.1915995885650675e-06,
"loss": 1.1777,
"step": 949
},
{
"epoch": 3.8461538461538463,
"grad_norm": 0.34467730745369735,
"learning_rate": 6.150742957589784e-06,
"loss": 1.1784,
"step": 950
},
{
"epoch": 3.850202429149798,
"grad_norm": 0.34674916425557056,
"learning_rate": 6.109997064904774e-06,
"loss": 1.196,
"step": 951
},
{
"epoch": 3.854251012145749,
"grad_norm": 0.33913822426726753,
"learning_rate": 6.069362236313092e-06,
"loss": 1.1966,
"step": 952
},
{
"epoch": 3.8582995951417005,
"grad_norm": 0.3448769697848159,
"learning_rate": 6.0288387967297565e-06,
"loss": 1.1652,
"step": 953
},
{
"epoch": 3.862348178137652,
"grad_norm": 0.3357409859256295,
"learning_rate": 5.988427070179108e-06,
"loss": 1.1917,
"step": 954
},
{
"epoch": 3.866396761133603,
"grad_norm": 0.3894279212430365,
"learning_rate": 5.948127379792234e-06,
"loss": 1.1855,
"step": 955
},
{
"epoch": 3.8704453441295548,
"grad_norm": 0.33144469330712656,
"learning_rate": 5.90794004780439e-06,
"loss": 1.2123,
"step": 956
},
{
"epoch": 3.8744939271255063,
"grad_norm": 0.3670507489821827,
"learning_rate": 5.867865395552405e-06,
"loss": 1.1804,
"step": 957
},
{
"epoch": 3.8785425101214575,
"grad_norm": 0.32857629887609935,
"learning_rate": 5.8279037434721364e-06,
"loss": 1.1936,
"step": 958
},
{
"epoch": 3.882591093117409,
"grad_norm": 0.3345379835725316,
"learning_rate": 5.788055411095892e-06,
"loss": 1.1765,
"step": 959
},
{
"epoch": 3.8866396761133606,
"grad_norm": 0.3499140289748021,
"learning_rate": 5.7483207170498625e-06,
"loss": 1.1935,
"step": 960
},
{
"epoch": 3.8906882591093117,
"grad_norm": 0.330357340909801,
"learning_rate": 5.708699979051613e-06,
"loss": 1.1774,
"step": 961
},
{
"epoch": 3.8947368421052633,
"grad_norm": 0.3353680418531212,
"learning_rate": 5.66919351390751e-06,
"loss": 1.1714,
"step": 962
},
{
"epoch": 3.898785425101215,
"grad_norm": 0.32185027039953196,
"learning_rate": 5.629801637510188e-06,
"loss": 1.1895,
"step": 963
},
{
"epoch": 3.902834008097166,
"grad_norm": 0.35016106738338343,
"learning_rate": 5.590524664836043e-06,
"loss": 1.2041,
"step": 964
},
{
"epoch": 3.9068825910931175,
"grad_norm": 0.3261344734938588,
"learning_rate": 5.551362909942708e-06,
"loss": 1.1971,
"step": 965
},
{
"epoch": 3.910931174089069,
"grad_norm": 0.32894229637674754,
"learning_rate": 5.51231668596653e-06,
"loss": 1.2449,
"step": 966
},
{
"epoch": 3.91497975708502,
"grad_norm": 0.3214378751266099,
"learning_rate": 5.47338630512007e-06,
"loss": 1.1583,
"step": 967
},
{
"epoch": 3.919028340080972,
"grad_norm": 0.34196216850186273,
"learning_rate": 5.434572078689617e-06,
"loss": 1.1818,
"step": 968
},
{
"epoch": 3.9230769230769234,
"grad_norm": 0.33505632879500064,
"learning_rate": 5.395874317032694e-06,
"loss": 1.1982,
"step": 969
},
{
"epoch": 3.9271255060728745,
"grad_norm": 0.34145502943429246,
"learning_rate": 5.357293329575573e-06,
"loss": 1.1754,
"step": 970
},
{
"epoch": 3.931174089068826,
"grad_norm": 0.3192356558805668,
"learning_rate": 5.318829424810797e-06,
"loss": 1.1991,
"step": 971
},
{
"epoch": 3.9352226720647776,
"grad_norm": 0.3078208327130786,
"learning_rate": 5.280482910294728e-06,
"loss": 1.1956,
"step": 972
},
{
"epoch": 3.9392712550607287,
"grad_norm": 0.34199264962308223,
"learning_rate": 5.242254092645078e-06,
"loss": 1.1746,
"step": 973
},
{
"epoch": 3.9433198380566803,
"grad_norm": 0.34435119892429994,
"learning_rate": 5.2041432775384445e-06,
"loss": 1.1984,
"step": 974
},
{
"epoch": 3.9473684210526314,
"grad_norm": 0.32771564927290364,
"learning_rate": 5.166150769707892e-06,
"loss": 1.2471,
"step": 975
},
{
"epoch": 3.951417004048583,
"grad_norm": 0.35149712603491423,
"learning_rate": 5.128276872940503e-06,
"loss": 1.1782,
"step": 976
},
{
"epoch": 3.9554655870445345,
"grad_norm": 0.3327413460328318,
"learning_rate": 5.090521890074953e-06,
"loss": 1.1935,
"step": 977
},
{
"epoch": 3.9595141700404857,
"grad_norm": 0.3223038500869169,
"learning_rate": 5.0528861229990675e-06,
"loss": 1.167,
"step": 978
},
{
"epoch": 3.9635627530364372,
"grad_norm": 0.3130010979011477,
"learning_rate": 5.015369872647442e-06,
"loss": 1.2219,
"step": 979
},
{
"epoch": 3.967611336032389,
"grad_norm": 0.31329442396510154,
"learning_rate": 4.9779734389990195e-06,
"loss": 1.2072,
"step": 980
},
{
"epoch": 3.97165991902834,
"grad_norm": 0.3339213263782977,
"learning_rate": 4.940697121074675e-06,
"loss": 1.1892,
"step": 981
},
{
"epoch": 3.9757085020242915,
"grad_norm": 0.3273124066128923,
"learning_rate": 4.903541216934864e-06,
"loss": 1.1605,
"step": 982
},
{
"epoch": 3.979757085020243,
"grad_norm": 0.31548713460767885,
"learning_rate": 4.8665060236772e-06,
"loss": 1.212,
"step": 983
},
{
"epoch": 3.983805668016194,
"grad_norm": 0.3174526753039003,
"learning_rate": 4.829591837434102e-06,
"loss": 1.1936,
"step": 984
},
{
"epoch": 3.9878542510121457,
"grad_norm": 0.32054267797405245,
"learning_rate": 4.792798953370424e-06,
"loss": 1.1852,
"step": 985
},
{
"epoch": 3.9919028340080973,
"grad_norm": 0.32690347616937915,
"learning_rate": 4.756127665681083e-06,
"loss": 1.2051,
"step": 986
},
{
"epoch": 3.9959514170040484,
"grad_norm": 0.3276466681735962,
"learning_rate": 4.719578267588725e-06,
"loss": 1.2028,
"step": 987
},
{
"epoch": 4.0,
"grad_norm": 0.31911290978946194,
"learning_rate": 4.6831510513413525e-06,
"loss": 1.1703,
"step": 988
},
{
"epoch": 4.004048582995951,
"grad_norm": 0.7590917525184308,
"learning_rate": 4.6468463082100176e-06,
"loss": 1.0995,
"step": 989
},
{
"epoch": 4.008097165991903,
"grad_norm": 0.487019994698456,
"learning_rate": 4.610664328486485e-06,
"loss": 1.0923,
"step": 990
},
{
"epoch": 4.012145748987854,
"grad_norm": 0.465644202158436,
"learning_rate": 4.5746054014809e-06,
"loss": 1.1074,
"step": 991
},
{
"epoch": 4.016194331983805,
"grad_norm": 0.8127328309404142,
"learning_rate": 4.5386698155194766e-06,
"loss": 1.1735,
"step": 992
},
{
"epoch": 4.020242914979757,
"grad_norm": 0.6832728917724232,
"learning_rate": 4.502857857942204e-06,
"loss": 1.0754,
"step": 993
},
{
"epoch": 4.0242914979757085,
"grad_norm": 0.48667507177327207,
"learning_rate": 4.467169815100549e-06,
"loss": 1.0706,
"step": 994
},
{
"epoch": 4.02834008097166,
"grad_norm": 0.5289102784956959,
"learning_rate": 4.431605972355137e-06,
"loss": 1.0707,
"step": 995
},
{
"epoch": 4.032388663967612,
"grad_norm": 0.5550058747611406,
"learning_rate": 4.396166614073516e-06,
"loss": 1.1104,
"step": 996
},
{
"epoch": 4.036437246963563,
"grad_norm": 0.5333254683671137,
"learning_rate": 4.360852023627846e-06,
"loss": 1.0998,
"step": 997
},
{
"epoch": 4.040485829959514,
"grad_norm": 0.44463773806459905,
"learning_rate": 4.325662483392656e-06,
"loss": 1.0845,
"step": 998
},
{
"epoch": 4.044534412955466,
"grad_norm": 0.5344257099140275,
"learning_rate": 4.29059827474257e-06,
"loss": 1.0771,
"step": 999
},
{
"epoch": 4.048582995951417,
"grad_norm": 0.6068998638848839,
"learning_rate": 4.255659678050063e-06,
"loss": 1.0902,
"step": 1000
},
{
"epoch": 4.052631578947368,
"grad_norm": 0.48270816003173384,
"learning_rate": 4.22084697268323e-06,
"loss": 1.1122,
"step": 1001
},
{
"epoch": 4.05668016194332,
"grad_norm": 0.44864774989437056,
"learning_rate": 4.186160437003535e-06,
"loss": 1.1197,
"step": 1002
},
{
"epoch": 4.060728744939271,
"grad_norm": 0.5454334855573534,
"learning_rate": 4.151600348363582e-06,
"loss": 1.1332,
"step": 1003
},
{
"epoch": 4.064777327935222,
"grad_norm": 0.4696771056288324,
"learning_rate": 4.11716698310493e-06,
"loss": 1.1084,
"step": 1004
},
{
"epoch": 4.068825910931174,
"grad_norm": 0.4589531721249273,
"learning_rate": 4.082860616555848e-06,
"loss": 1.0964,
"step": 1005
},
{
"epoch": 4.0728744939271255,
"grad_norm": 0.40646199841250735,
"learning_rate": 4.048681523029125e-06,
"loss": 1.1668,
"step": 1006
},
{
"epoch": 4.076923076923077,
"grad_norm": 0.3989818344788565,
"learning_rate": 4.014629975819882e-06,
"loss": 1.1055,
"step": 1007
},
{
"epoch": 4.080971659919029,
"grad_norm": 0.4569220087382976,
"learning_rate": 3.9807062472033855e-06,
"loss": 1.0538,
"step": 1008
},
{
"epoch": 4.08502024291498,
"grad_norm": 0.45540656748366254,
"learning_rate": 3.94691060843287e-06,
"loss": 1.1163,
"step": 1009
},
{
"epoch": 4.089068825910931,
"grad_norm": 0.46822024073319535,
"learning_rate": 3.913243329737353e-06,
"loss": 1.115,
"step": 1010
},
{
"epoch": 4.093117408906883,
"grad_norm": 0.3702155417467374,
"learning_rate": 3.879704680319503e-06,
"loss": 1.0974,
"step": 1011
},
{
"epoch": 4.097165991902834,
"grad_norm": 0.3960695008428155,
"learning_rate": 3.846294928353471e-06,
"loss": 1.0902,
"step": 1012
},
{
"epoch": 4.101214574898785,
"grad_norm": 0.4432033393840935,
"learning_rate": 3.8130143409827345e-06,
"loss": 1.1418,
"step": 1013
},
{
"epoch": 4.105263157894737,
"grad_norm": 0.42603739326342305,
"learning_rate": 3.7798631843179887e-06,
"loss": 1.0941,
"step": 1014
},
{
"epoch": 4.109311740890688,
"grad_norm": 0.42947822066060726,
"learning_rate": 3.7468417234349973e-06,
"loss": 1.1213,
"step": 1015
},
{
"epoch": 4.113360323886639,
"grad_norm": 0.3471670286078278,
"learning_rate": 3.7139502223724823e-06,
"loss": 1.1127,
"step": 1016
},
{
"epoch": 4.117408906882591,
"grad_norm": 0.4032875911372414,
"learning_rate": 3.6811889441300095e-06,
"loss": 1.1126,
"step": 1017
},
{
"epoch": 4.1214574898785425,
"grad_norm": 0.41515231412814363,
"learning_rate": 3.648558150665886e-06,
"loss": 1.1038,
"step": 1018
},
{
"epoch": 4.125506072874494,
"grad_norm": 0.4260169683578617,
"learning_rate": 3.616058102895075e-06,
"loss": 1.1205,
"step": 1019
},
{
"epoch": 4.129554655870446,
"grad_norm": 0.359438913264633,
"learning_rate": 3.583689060687079e-06,
"loss": 1.1652,
"step": 1020
},
{
"epoch": 4.133603238866397,
"grad_norm": 0.34008107980225377,
"learning_rate": 3.5514512828639023e-06,
"loss": 1.0882,
"step": 1021
},
{
"epoch": 4.137651821862348,
"grad_norm": 0.4168420203443667,
"learning_rate": 3.51934502719796e-06,
"loss": 1.0767,
"step": 1022
},
{
"epoch": 4.1417004048583,
"grad_norm": 0.3936760294577784,
"learning_rate": 3.4873705504100185e-06,
"loss": 1.0979,
"step": 1023
},
{
"epoch": 4.145748987854251,
"grad_norm": 0.3704732748972756,
"learning_rate": 3.4555281081671345e-06,
"loss": 1.0753,
"step": 1024
},
{
"epoch": 4.149797570850202,
"grad_norm": 0.360575288238212,
"learning_rate": 3.4238179550806326e-06,
"loss": 1.0751,
"step": 1025
},
{
"epoch": 4.153846153846154,
"grad_norm": 0.3960968741586434,
"learning_rate": 3.3922403447040497e-06,
"loss": 1.0882,
"step": 1026
},
{
"epoch": 4.157894736842105,
"grad_norm": 0.39632007157959964,
"learning_rate": 3.3607955295311247e-06,
"loss": 1.0925,
"step": 1027
},
{
"epoch": 4.161943319838056,
"grad_norm": 0.41494929279627163,
"learning_rate": 3.32948376099375e-06,
"loss": 1.0993,
"step": 1028
},
{
"epoch": 4.165991902834008,
"grad_norm": 0.3629544845921003,
"learning_rate": 3.2983052894599976e-06,
"loss": 1.0695,
"step": 1029
},
{
"epoch": 4.17004048582996,
"grad_norm": 0.3511838454301867,
"learning_rate": 3.2672603642320945e-06,
"loss": 1.1021,
"step": 1030
},
{
"epoch": 4.174089068825911,
"grad_norm": 0.427680812504248,
"learning_rate": 3.236349233544438e-06,
"loss": 1.0912,
"step": 1031
},
{
"epoch": 4.178137651821863,
"grad_norm": 0.3802527789610484,
"learning_rate": 3.205572144561604e-06,
"loss": 1.095,
"step": 1032
},
{
"epoch": 4.182186234817814,
"grad_norm": 0.3334848468949266,
"learning_rate": 3.174929343376374e-06,
"loss": 1.0854,
"step": 1033
},
{
"epoch": 4.186234817813765,
"grad_norm": 0.3634774209425702,
"learning_rate": 3.1444210750077777e-06,
"loss": 1.071,
"step": 1034
},
{
"epoch": 4.190283400809717,
"grad_norm": 0.3375697145987058,
"learning_rate": 3.1140475833991046e-06,
"loss": 1.0887,
"step": 1035
},
{
"epoch": 4.194331983805668,
"grad_norm": 0.3396787089374625,
"learning_rate": 3.0838091114159895e-06,
"loss": 1.0857,
"step": 1036
},
{
"epoch": 4.198380566801619,
"grad_norm": 0.37362412759543395,
"learning_rate": 3.0537059008444524e-06,
"loss": 1.1092,
"step": 1037
},
{
"epoch": 4.202429149797571,
"grad_norm": 0.35169504965672865,
"learning_rate": 3.0237381923889563e-06,
"loss": 1.141,
"step": 1038
},
{
"epoch": 4.206477732793522,
"grad_norm": 0.3487053963149441,
"learning_rate": 2.9939062256704997e-06,
"loss": 1.1232,
"step": 1039
},
{
"epoch": 4.2105263157894735,
"grad_norm": 0.3326484487031238,
"learning_rate": 2.964210239224696e-06,
"loss": 1.0895,
"step": 1040
},
{
"epoch": 4.2145748987854255,
"grad_norm": 0.34270570284513185,
"learning_rate": 2.934650470499869e-06,
"loss": 1.1339,
"step": 1041
},
{
"epoch": 4.218623481781377,
"grad_norm": 0.33971110853111997,
"learning_rate": 2.905227155855128e-06,
"loss": 1.0875,
"step": 1042
},
{
"epoch": 4.222672064777328,
"grad_norm": 0.3454591501339845,
"learning_rate": 2.875940530558519e-06,
"loss": 1.0742,
"step": 1043
},
{
"epoch": 4.22672064777328,
"grad_norm": 0.3274886002542428,
"learning_rate": 2.8467908287851154e-06,
"loss": 1.0965,
"step": 1044
},
{
"epoch": 4.230769230769231,
"grad_norm": 0.3093319433548498,
"learning_rate": 2.8177782836151603e-06,
"loss": 1.1356,
"step": 1045
},
{
"epoch": 4.234817813765182,
"grad_norm": 0.359190115496107,
"learning_rate": 2.7889031270321874e-06,
"loss": 1.0681,
"step": 1046
},
{
"epoch": 4.238866396761134,
"grad_norm": 0.35441706603781276,
"learning_rate": 2.760165589921184e-06,
"loss": 1.1267,
"step": 1047
},
{
"epoch": 4.242914979757085,
"grad_norm": 0.3597364046230048,
"learning_rate": 2.7315659020667305e-06,
"loss": 1.1026,
"step": 1048
},
{
"epoch": 4.246963562753036,
"grad_norm": 0.35193464689563253,
"learning_rate": 2.703104292151164e-06,
"loss": 1.1207,
"step": 1049
},
{
"epoch": 4.251012145748988,
"grad_norm": 0.3831170030291526,
"learning_rate": 2.6747809877527607e-06,
"loss": 1.0982,
"step": 1050
},
{
"epoch": 4.255060728744939,
"grad_norm": 0.33582402159545593,
"learning_rate": 2.6465962153439082e-06,
"loss": 1.0868,
"step": 1051
},
{
"epoch": 4.2591093117408905,
"grad_norm": 0.3385373522504172,
"learning_rate": 2.618550200289303e-06,
"loss": 1.1084,
"step": 1052
},
{
"epoch": 4.2631578947368425,
"grad_norm": 0.36826798970047986,
"learning_rate": 2.59064316684412e-06,
"loss": 1.0862,
"step": 1053
},
{
"epoch": 4.267206477732794,
"grad_norm": 0.34530680154310567,
"learning_rate": 2.562875338152264e-06,
"loss": 1.1017,
"step": 1054
},
{
"epoch": 4.271255060728745,
"grad_norm": 0.3424278202814088,
"learning_rate": 2.5352469362445552e-06,
"loss": 1.0918,
"step": 1055
},
{
"epoch": 4.275303643724697,
"grad_norm": 0.3526144640080365,
"learning_rate": 2.5077581820369523e-06,
"loss": 1.0386,
"step": 1056
},
{
"epoch": 4.279352226720648,
"grad_norm": 0.33467932674695133,
"learning_rate": 2.480409295328803e-06,
"loss": 1.0903,
"step": 1057
},
{
"epoch": 4.283400809716599,
"grad_norm": 0.3515767272508767,
"learning_rate": 2.45320049480108e-06,
"loss": 1.0456,
"step": 1058
},
{
"epoch": 4.287449392712551,
"grad_norm": 0.34956743341005714,
"learning_rate": 2.4261319980146293e-06,
"loss": 1.1186,
"step": 1059
},
{
"epoch": 4.291497975708502,
"grad_norm": 0.3483956078994242,
"learning_rate": 2.399204021408421e-06,
"loss": 1.0949,
"step": 1060
},
{
"epoch": 4.295546558704453,
"grad_norm": 0.3118835986809752,
"learning_rate": 2.3724167802978395e-06,
"loss": 1.1452,
"step": 1061
},
{
"epoch": 4.299595141700405,
"grad_norm": 0.3474003279462389,
"learning_rate": 2.3457704888729625e-06,
"loss": 1.1017,
"step": 1062
},
{
"epoch": 4.303643724696356,
"grad_norm": 0.3317210997243994,
"learning_rate": 2.3192653601968162e-06,
"loss": 1.0831,
"step": 1063
},
{
"epoch": 4.3076923076923075,
"grad_norm": 0.34288424809091106,
"learning_rate": 2.2929016062037057e-06,
"loss": 1.0914,
"step": 1064
},
{
"epoch": 4.3117408906882595,
"grad_norm": 0.3494292918655118,
"learning_rate": 2.2666794376975055e-06,
"loss": 1.1079,
"step": 1065
},
{
"epoch": 4.315789473684211,
"grad_norm": 0.34044481180191805,
"learning_rate": 2.24059906434998e-06,
"loss": 1.1202,
"step": 1066
},
{
"epoch": 4.319838056680162,
"grad_norm": 0.363832377948552,
"learning_rate": 2.214660694699091e-06,
"loss": 1.1355,
"step": 1067
},
{
"epoch": 4.323886639676114,
"grad_norm": 0.37181364160398545,
"learning_rate": 2.188864536147355e-06,
"loss": 1.1262,
"step": 1068
},
{
"epoch": 4.327935222672065,
"grad_norm": 0.3369370767120824,
"learning_rate": 2.1632107949601757e-06,
"loss": 1.0881,
"step": 1069
},
{
"epoch": 4.331983805668016,
"grad_norm": 0.36203558906698097,
"learning_rate": 2.137699676264171e-06,
"loss": 1.0701,
"step": 1070
},
{
"epoch": 4.336032388663968,
"grad_norm": 0.3437150363104081,
"learning_rate": 2.1123313840455738e-06,
"loss": 1.1412,
"step": 1071
},
{
"epoch": 4.340080971659919,
"grad_norm": 0.37745637275584487,
"learning_rate": 2.08710612114857e-06,
"loss": 1.0877,
"step": 1072
},
{
"epoch": 4.34412955465587,
"grad_norm": 0.347470573212044,
"learning_rate": 2.0620240892736953e-06,
"loss": 1.1036,
"step": 1073
},
{
"epoch": 4.348178137651822,
"grad_norm": 0.3487370377488197,
"learning_rate": 2.0370854889762e-06,
"loss": 1.124,
"step": 1074
},
{
"epoch": 4.352226720647773,
"grad_norm": 0.3283652181345829,
"learning_rate": 2.0122905196644726e-06,
"loss": 1.1498,
"step": 1075
},
{
"epoch": 4.3562753036437245,
"grad_norm": 0.3442048398613211,
"learning_rate": 1.987639379598427e-06,
"loss": 1.0886,
"step": 1076
},
{
"epoch": 4.3603238866396765,
"grad_norm": 0.350126697306271,
"learning_rate": 1.963132265887919e-06,
"loss": 1.1195,
"step": 1077
},
{
"epoch": 4.364372469635628,
"grad_norm": 0.34062828513537974,
"learning_rate": 1.9387693744911784e-06,
"loss": 1.1287,
"step": 1078
},
{
"epoch": 4.368421052631579,
"grad_norm": 0.35895492311937194,
"learning_rate": 1.9145509002132368e-06,
"loss": 1.139,
"step": 1079
},
{
"epoch": 4.372469635627531,
"grad_norm": 0.32359315533802113,
"learning_rate": 1.8904770367043723e-06,
"loss": 1.1123,
"step": 1080
},
{
"epoch": 4.376518218623482,
"grad_norm": 0.33953779484575325,
"learning_rate": 1.8665479764585504e-06,
"loss": 1.1126,
"step": 1081
},
{
"epoch": 4.380566801619433,
"grad_norm": 0.3102013108706209,
"learning_rate": 1.8427639108118977e-06,
"loss": 1.1563,
"step": 1082
},
{
"epoch": 4.384615384615385,
"grad_norm": 0.3280308047075404,
"learning_rate": 1.8191250299411755e-06,
"loss": 1.0807,
"step": 1083
},
{
"epoch": 4.388663967611336,
"grad_norm": 0.3375322156045213,
"learning_rate": 1.7956315228622446e-06,
"loss": 1.0552,
"step": 1084
},
{
"epoch": 4.392712550607287,
"grad_norm": 0.34736636400730897,
"learning_rate": 1.7722835774285553e-06,
"loss": 1.1096,
"step": 1085
},
{
"epoch": 4.396761133603239,
"grad_norm": 0.3483909431625543,
"learning_rate": 1.7490813803296624e-06,
"loss": 1.1187,
"step": 1086
},
{
"epoch": 4.40080971659919,
"grad_norm": 0.35385865584901277,
"learning_rate": 1.726025117089718e-06,
"loss": 1.1124,
"step": 1087
},
{
"epoch": 4.4048582995951415,
"grad_norm": 0.31462604960341894,
"learning_rate": 1.70311497206598e-06,
"loss": 1.1186,
"step": 1088
},
{
"epoch": 4.4089068825910935,
"grad_norm": 0.3616378025423594,
"learning_rate": 1.6803511284473638e-06,
"loss": 1.1173,
"step": 1089
},
{
"epoch": 4.412955465587045,
"grad_norm": 0.321451331720842,
"learning_rate": 1.6577337682529538e-06,
"loss": 1.1026,
"step": 1090
},
{
"epoch": 4.417004048582996,
"grad_norm": 0.31735459709110597,
"learning_rate": 1.6352630723305617e-06,
"loss": 1.1042,
"step": 1091
},
{
"epoch": 4.421052631578947,
"grad_norm": 0.34180408738798274,
"learning_rate": 1.612939220355272e-06,
"loss": 1.1544,
"step": 1092
},
{
"epoch": 4.425101214574899,
"grad_norm": 0.3530346656218542,
"learning_rate": 1.5907623908280112e-06,
"loss": 1.0913,
"step": 1093
},
{
"epoch": 4.42914979757085,
"grad_norm": 0.3281451357155445,
"learning_rate": 1.568732761074121e-06,
"loss": 1.0866,
"step": 1094
},
{
"epoch": 4.433198380566802,
"grad_norm": 0.3365149589904803,
"learning_rate": 1.5468505072419237e-06,
"loss": 1.1256,
"step": 1095
},
{
"epoch": 4.437246963562753,
"grad_norm": 0.34112916910438557,
"learning_rate": 1.5251158043013448e-06,
"loss": 1.1216,
"step": 1096
},
{
"epoch": 4.441295546558704,
"grad_norm": 0.33468663455539954,
"learning_rate": 1.5035288260424885e-06,
"loss": 1.0992,
"step": 1097
},
{
"epoch": 4.445344129554655,
"grad_norm": 0.3497504372369661,
"learning_rate": 1.4820897450742623e-06,
"loss": 1.0953,
"step": 1098
},
{
"epoch": 4.449392712550607,
"grad_norm": 0.3503281799066174,
"learning_rate": 1.4607987328229767e-06,
"loss": 1.0686,
"step": 1099
},
{
"epoch": 4.4534412955465585,
"grad_norm": 0.32372167592895074,
"learning_rate": 1.4396559595310056e-06,
"loss": 1.1391,
"step": 1100
},
{
"epoch": 4.4574898785425106,
"grad_norm": 0.3283856989377848,
"learning_rate": 1.4186615942553973e-06,
"loss": 1.148,
"step": 1101
},
{
"epoch": 4.461538461538462,
"grad_norm": 0.3229723124129563,
"learning_rate": 1.3978158048665424e-06,
"loss": 1.1159,
"step": 1102
},
{
"epoch": 4.465587044534413,
"grad_norm": 0.35425157052445516,
"learning_rate": 1.3771187580468115e-06,
"loss": 1.1033,
"step": 1103
},
{
"epoch": 4.469635627530364,
"grad_norm": 0.32235367599459197,
"learning_rate": 1.3565706192892392e-06,
"loss": 1.106,
"step": 1104
},
{
"epoch": 4.473684210526316,
"grad_norm": 0.330846965222404,
"learning_rate": 1.3361715528961993e-06,
"loss": 1.0936,
"step": 1105
},
{
"epoch": 4.477732793522267,
"grad_norm": 0.3567418212368603,
"learning_rate": 1.3159217219780817e-06,
"loss": 1.1071,
"step": 1106
},
{
"epoch": 4.481781376518219,
"grad_norm": 0.32418054194750734,
"learning_rate": 1.2958212884519949e-06,
"loss": 1.1277,
"step": 1107
},
{
"epoch": 4.48582995951417,
"grad_norm": 0.32982497992628534,
"learning_rate": 1.2758704130404675e-06,
"loss": 1.1138,
"step": 1108
},
{
"epoch": 4.489878542510121,
"grad_norm": 0.3158871383728131,
"learning_rate": 1.2560692552701738e-06,
"loss": 1.0888,
"step": 1109
},
{
"epoch": 4.493927125506072,
"grad_norm": 0.3295819794840626,
"learning_rate": 1.2364179734706382e-06,
"loss": 1.1255,
"step": 1110
},
{
"epoch": 4.497975708502024,
"grad_norm": 0.3492560577428405,
"learning_rate": 1.2169167247729852e-06,
"loss": 1.0739,
"step": 1111
},
{
"epoch": 4.502024291497976,
"grad_norm": 0.3201168616575043,
"learning_rate": 1.1975656651086864e-06,
"loss": 1.0951,
"step": 1112
},
{
"epoch": 4.506072874493928,
"grad_norm": 0.33793219720820267,
"learning_rate": 1.1783649492082926e-06,
"loss": 1.1119,
"step": 1113
},
{
"epoch": 4.510121457489879,
"grad_norm": 0.31184204960061274,
"learning_rate": 1.1593147306002183e-06,
"loss": 1.1016,
"step": 1114
},
{
"epoch": 4.51417004048583,
"grad_norm": 0.3123076004552112,
"learning_rate": 1.140415161609505e-06,
"loss": 1.136,
"step": 1115
},
{
"epoch": 4.518218623481781,
"grad_norm": 0.3218789102764629,
"learning_rate": 1.121666393356604e-06,
"loss": 1.1253,
"step": 1116
},
{
"epoch": 4.522267206477733,
"grad_norm": 0.3114018660412308,
"learning_rate": 1.1030685757561632e-06,
"loss": 1.1577,
"step": 1117
},
{
"epoch": 4.526315789473684,
"grad_norm": 0.3195552058504605,
"learning_rate": 1.0846218575158373e-06,
"loss": 1.1081,
"step": 1118
},
{
"epoch": 4.530364372469636,
"grad_norm": 0.35312132507809363,
"learning_rate": 1.066326386135097e-06,
"loss": 1.0927,
"step": 1119
},
{
"epoch": 4.534412955465587,
"grad_norm": 0.3445542152368858,
"learning_rate": 1.0481823079040398e-06,
"loss": 1.0927,
"step": 1120
},
{
"epoch": 4.538461538461538,
"grad_norm": 0.31984395297708373,
"learning_rate": 1.0301897679022344e-06,
"loss": 1.0975,
"step": 1121
},
{
"epoch": 4.5425101214574894,
"grad_norm": 0.3174112599643119,
"learning_rate": 1.0123489099975491e-06,
"loss": 1.1483,
"step": 1122
},
{
"epoch": 4.5465587044534415,
"grad_norm": 0.32686602073941373,
"learning_rate": 9.9465987684501e-07,
"loss": 1.1293,
"step": 1123
},
{
"epoch": 4.550607287449393,
"grad_norm": 0.3329560766408977,
"learning_rate": 9.771228098856534e-07,
"loss": 1.0891,
"step": 1124
},
{
"epoch": 4.554655870445345,
"grad_norm": 0.3238124627167755,
"learning_rate": 9.597378493454013e-07,
"loss": 1.1174,
"step": 1125
},
{
"epoch": 4.558704453441296,
"grad_norm": 0.3232138328200124,
"learning_rate": 9.425051342339353e-07,
"loss": 1.1281,
"step": 1126
},
{
"epoch": 4.562753036437247,
"grad_norm": 0.34467180702511563,
"learning_rate": 9.254248023435864e-07,
"loss": 1.0713,
"step": 1127
},
{
"epoch": 4.566801619433198,
"grad_norm": 0.32616034045911346,
"learning_rate": 9.084969902482288e-07,
"loss": 1.1149,
"step": 1128
},
{
"epoch": 4.57085020242915,
"grad_norm": 0.336941908074731,
"learning_rate": 8.917218333021993e-07,
"loss": 1.0785,
"step": 1129
},
{
"epoch": 4.574898785425101,
"grad_norm": 0.3114233345897134,
"learning_rate": 8.750994656392109e-07,
"loss": 1.1103,
"step": 1130
},
{
"epoch": 4.578947368421053,
"grad_norm": 0.3254943695973851,
"learning_rate": 8.58630020171265e-07,
"loss": 1.1089,
"step": 1131
},
{
"epoch": 4.582995951417004,
"grad_norm": 0.322119550015437,
"learning_rate": 8.423136285876099e-07,
"loss": 1.1001,
"step": 1132
},
{
"epoch": 4.587044534412955,
"grad_norm": 0.3100350805898905,
"learning_rate": 8.26150421353682e-07,
"loss": 1.1164,
"step": 1133
},
{
"epoch": 4.5910931174089065,
"grad_norm": 0.33705187572525125,
"learning_rate": 8.101405277100549e-07,
"loss": 1.104,
"step": 1134
},
{
"epoch": 4.5951417004048585,
"grad_norm": 0.33620362780725616,
"learning_rate": 7.942840756714077e-07,
"loss": 1.1093,
"step": 1135
},
{
"epoch": 4.59919028340081,
"grad_norm": 0.33862885635115875,
"learning_rate": 7.785811920255093e-07,
"loss": 1.1019,
"step": 1136
},
{
"epoch": 4.603238866396762,
"grad_norm": 0.3412300048780369,
"learning_rate": 7.630320023322024e-07,
"loss": 1.0939,
"step": 1137
},
{
"epoch": 4.607287449392713,
"grad_norm": 0.33145714449249364,
"learning_rate": 7.476366309223903e-07,
"loss": 1.0879,
"step": 1138
},
{
"epoch": 4.611336032388664,
"grad_norm": 0.32248571446077096,
"learning_rate": 7.323952008970537e-07,
"loss": 1.0931,
"step": 1139
},
{
"epoch": 4.615384615384615,
"grad_norm": 0.3314447926284423,
"learning_rate": 7.173078341262618e-07,
"loss": 1.1197,
"step": 1140
},
{
"epoch": 4.619433198380567,
"grad_norm": 0.3269868495355912,
"learning_rate": 7.023746512481989e-07,
"loss": 1.1262,
"step": 1141
},
{
"epoch": 4.623481781376518,
"grad_norm": 0.3323484167722332,
"learning_rate": 6.875957716681902e-07,
"loss": 1.107,
"step": 1142
},
{
"epoch": 4.62753036437247,
"grad_norm": 0.3378714680195475,
"learning_rate": 6.729713135577687e-07,
"loss": 1.0992,
"step": 1143
},
{
"epoch": 4.631578947368421,
"grad_norm": 0.33837561927614346,
"learning_rate": 6.585013938537143e-07,
"loss": 1.1069,
"step": 1144
},
{
"epoch": 4.635627530364372,
"grad_norm": 0.31396840182358626,
"learning_rate": 6.441861282571115e-07,
"loss": 1.0881,
"step": 1145
},
{
"epoch": 4.6396761133603235,
"grad_norm": 0.30630102097494266,
"learning_rate": 6.300256312324471e-07,
"loss": 1.0955,
"step": 1146
},
{
"epoch": 4.6437246963562755,
"grad_norm": 0.32505724084352117,
"learning_rate": 6.160200160066776e-07,
"loss": 1.0985,
"step": 1147
},
{
"epoch": 4.647773279352227,
"grad_norm": 0.3154643633110579,
"learning_rate": 6.021693945683327e-07,
"loss": 1.102,
"step": 1148
},
{
"epoch": 4.651821862348179,
"grad_norm": 0.33329757438427976,
"learning_rate": 5.884738776666044e-07,
"loss": 1.0857,
"step": 1149
},
{
"epoch": 4.65587044534413,
"grad_norm": 0.32827136300143267,
"learning_rate": 5.749335748104878e-07,
"loss": 1.1092,
"step": 1150
},
{
"epoch": 4.659919028340081,
"grad_norm": 0.3210914611632018,
"learning_rate": 5.615485942678778e-07,
"loss": 1.1055,
"step": 1151
},
{
"epoch": 4.663967611336032,
"grad_norm": 0.35117002445886375,
"learning_rate": 5.483190430647289e-07,
"loss": 1.0784,
"step": 1152
},
{
"epoch": 4.668016194331984,
"grad_norm": 0.3293898156197792,
"learning_rate": 5.352450269841747e-07,
"loss": 1.1009,
"step": 1153
},
{
"epoch": 4.672064777327935,
"grad_norm": 0.32599808114439177,
"learning_rate": 5.22326650565701e-07,
"loss": 1.1023,
"step": 1154
},
{
"epoch": 4.676113360323887,
"grad_norm": 0.33323265346651926,
"learning_rate": 5.095640171043048e-07,
"loss": 1.0922,
"step": 1155
},
{
"epoch": 4.680161943319838,
"grad_norm": 0.30826048534258543,
"learning_rate": 4.969572286496594e-07,
"loss": 1.0802,
"step": 1156
},
{
"epoch": 4.684210526315789,
"grad_norm": 0.3199485047673001,
"learning_rate": 4.845063860053079e-07,
"loss": 1.1162,
"step": 1157
},
{
"epoch": 4.6882591093117405,
"grad_norm": 0.33820257298862144,
"learning_rate": 4.7221158872785997e-07,
"loss": 1.0994,
"step": 1158
},
{
"epoch": 4.6923076923076925,
"grad_norm": 0.32104416395488883,
"learning_rate": 4.600729351261857e-07,
"loss": 1.1171,
"step": 1159
},
{
"epoch": 4.696356275303644,
"grad_norm": 0.31831386955376134,
"learning_rate": 4.4809052226063355e-07,
"loss": 1.1043,
"step": 1160
},
{
"epoch": 4.700404858299595,
"grad_norm": 0.33968625510638956,
"learning_rate": 4.3626444594226046e-07,
"loss": 1.0442,
"step": 1161
},
{
"epoch": 4.704453441295547,
"grad_norm": 0.33347683754318136,
"learning_rate": 4.245948007320588e-07,
"loss": 1.0926,
"step": 1162
},
{
"epoch": 4.708502024291498,
"grad_norm": 0.3195554171737926,
"learning_rate": 4.130816799401971e-07,
"loss": 1.0953,
"step": 1163
},
{
"epoch": 4.712550607287449,
"grad_norm": 0.31725986750786617,
"learning_rate": 4.017251756252827e-07,
"loss": 1.0921,
"step": 1164
},
{
"epoch": 4.716599190283401,
"grad_norm": 0.2980969079296831,
"learning_rate": 3.9052537859362473e-07,
"loss": 1.1546,
"step": 1165
},
{
"epoch": 4.720647773279352,
"grad_norm": 0.3319642238619746,
"learning_rate": 3.7948237839849687e-07,
"loss": 1.0954,
"step": 1166
},
{
"epoch": 4.724696356275303,
"grad_norm": 0.3262412741040614,
"learning_rate": 3.6859626333943797e-07,
"loss": 1.1285,
"step": 1167
},
{
"epoch": 4.728744939271255,
"grad_norm": 0.3224285447194617,
"learning_rate": 3.578671204615347e-07,
"loss": 1.0828,
"step": 1168
},
{
"epoch": 4.732793522267206,
"grad_norm": 0.34612567781672804,
"learning_rate": 3.472950355547267e-07,
"loss": 1.1427,
"step": 1169
},
{
"epoch": 4.7368421052631575,
"grad_norm": 0.3491819874227095,
"learning_rate": 3.3688009315312244e-07,
"loss": 1.0825,
"step": 1170
},
{
"epoch": 4.7408906882591095,
"grad_norm": 0.31959889546218256,
"learning_rate": 3.266223765343224e-07,
"loss": 1.114,
"step": 1171
},
{
"epoch": 4.744939271255061,
"grad_norm": 0.30581769632656025,
"learning_rate": 3.16521967718757e-07,
"loss": 1.1265,
"step": 1172
},
{
"epoch": 4.748987854251012,
"grad_norm": 0.2957359934593013,
"learning_rate": 3.0657894746902726e-07,
"loss": 1.12,
"step": 1173
},
{
"epoch": 4.753036437246964,
"grad_norm": 0.3297836179304678,
"learning_rate": 2.9679339528925654e-07,
"loss": 1.0911,
"step": 1174
},
{
"epoch": 4.757085020242915,
"grad_norm": 0.333025290132935,
"learning_rate": 2.871653894244597e-07,
"loss": 1.123,
"step": 1175
},
{
"epoch": 4.761133603238866,
"grad_norm": 0.31390509501438285,
"learning_rate": 2.7769500685991266e-07,
"loss": 1.0929,
"step": 1176
},
{
"epoch": 4.765182186234818,
"grad_norm": 0.34165988809140974,
"learning_rate": 2.683823233205485e-07,
"loss": 1.0818,
"step": 1177
},
{
"epoch": 4.769230769230769,
"grad_norm": 0.2986572185619355,
"learning_rate": 2.592274132703287e-07,
"loss": 1.1215,
"step": 1178
},
{
"epoch": 4.77327935222672,
"grad_norm": 0.320647893795316,
"learning_rate": 2.502303499116754e-07,
"loss": 1.128,
"step": 1179
},
{
"epoch": 4.777327935222672,
"grad_norm": 0.32860404756212613,
"learning_rate": 2.413912051848666e-07,
"loss": 1.1165,
"step": 1180
},
{
"epoch": 4.781376518218623,
"grad_norm": 0.32729709195899415,
"learning_rate": 2.3271004976746613e-07,
"loss": 1.1353,
"step": 1181
},
{
"epoch": 4.7854251012145745,
"grad_norm": 0.335026458774907,
"learning_rate": 2.2418695307376393e-07,
"loss": 1.0974,
"step": 1182
},
{
"epoch": 4.7894736842105265,
"grad_norm": 0.2937242253026348,
"learning_rate": 2.1582198325420968e-07,
"loss": 1.1365,
"step": 1183
},
{
"epoch": 4.793522267206478,
"grad_norm": 0.3237127801645354,
"learning_rate": 2.0761520719488447e-07,
"loss": 1.0852,
"step": 1184
},
{
"epoch": 4.797570850202429,
"grad_norm": 0.30286866111007754,
"learning_rate": 1.9956669051694798e-07,
"loss": 1.1017,
"step": 1185
},
{
"epoch": 4.801619433198381,
"grad_norm": 0.3272705718865014,
"learning_rate": 1.916764975761254e-07,
"loss": 1.0834,
"step": 1186
},
{
"epoch": 4.805668016194332,
"grad_norm": 0.32688422498760905,
"learning_rate": 1.839446914621923e-07,
"loss": 1.0789,
"step": 1187
},
{
"epoch": 4.809716599190283,
"grad_norm": 0.29996609705813604,
"learning_rate": 1.7637133399845962e-07,
"loss": 1.1234,
"step": 1188
},
{
"epoch": 4.813765182186235,
"grad_norm": 0.3318368123370343,
"learning_rate": 1.6895648574129397e-07,
"loss": 1.1145,
"step": 1189
},
{
"epoch": 4.817813765182186,
"grad_norm": 0.3181550500235424,
"learning_rate": 1.6170020597962465e-07,
"loss": 1.1104,
"step": 1190
},
{
"epoch": 4.821862348178137,
"grad_norm": 0.31538994958712047,
"learning_rate": 1.5460255273447078e-07,
"loss": 1.1026,
"step": 1191
},
{
"epoch": 4.825910931174089,
"grad_norm": 0.31122931585494634,
"learning_rate": 1.4766358275847715e-07,
"loss": 1.131,
"step": 1192
},
{
"epoch": 4.82995951417004,
"grad_norm": 0.32161508208103845,
"learning_rate": 1.4088335153546573e-07,
"loss": 1.1204,
"step": 1193
},
{
"epoch": 4.834008097165992,
"grad_norm": 0.34138243312063615,
"learning_rate": 1.3426191327998496e-07,
"loss": 1.0583,
"step": 1194
},
{
"epoch": 4.838056680161944,
"grad_norm": 0.3334954746854171,
"learning_rate": 1.277993209368744e-07,
"loss": 1.0842,
"step": 1195
},
{
"epoch": 4.842105263157895,
"grad_norm": 0.3167706002013669,
"learning_rate": 1.214956261808542e-07,
"loss": 1.1328,
"step": 1196
},
{
"epoch": 4.846153846153846,
"grad_norm": 0.3176431680250027,
"learning_rate": 1.1535087941609846e-07,
"loss": 1.1164,
"step": 1197
},
{
"epoch": 4.850202429149798,
"grad_norm": 0.3045162006915953,
"learning_rate": 1.0936512977583802e-07,
"loss": 1.176,
"step": 1198
},
{
"epoch": 4.854251012145749,
"grad_norm": 0.32571339069409544,
"learning_rate": 1.0353842512196732e-07,
"loss": 1.1271,
"step": 1199
},
{
"epoch": 4.8582995951417,
"grad_norm": 0.31272784354580757,
"learning_rate": 9.787081204466253e-08,
"loss": 1.1239,
"step": 1200
},
{
"epoch": 4.862348178137652,
"grad_norm": 0.31565811455442405,
"learning_rate": 9.236233586200627e-08,
"loss": 1.1306,
"step": 1201
},
{
"epoch": 4.866396761133603,
"grad_norm": 0.3320867516986725,
"learning_rate": 8.70130406196279e-08,
"loss": 1.1315,
"step": 1202
},
{
"epoch": 4.870445344129554,
"grad_norm": 0.302759012431199,
"learning_rate": 8.182296909035047e-08,
"loss": 1.1123,
"step": 1203
},
{
"epoch": 4.874493927125506,
"grad_norm": 0.3256548347714428,
"learning_rate": 7.679216277384439e-08,
"loss": 1.1028,
"step": 1204
},
{
"epoch": 4.8785425101214575,
"grad_norm": 0.31686211586793095,
"learning_rate": 7.19206618963053e-08,
"loss": 1.1236,
"step": 1205
},
{
"epoch": 4.882591093117409,
"grad_norm": 0.3223781095030876,
"learning_rate": 6.720850541012347e-08,
"loss": 1.1032,
"step": 1206
},
{
"epoch": 4.886639676113361,
"grad_norm": 0.34333816817745866,
"learning_rate": 6.265573099357714e-08,
"loss": 1.0614,
"step": 1207
},
{
"epoch": 4.890688259109312,
"grad_norm": 0.3344677739089476,
"learning_rate": 5.826237505053067e-08,
"loss": 1.0867,
"step": 1208
},
{
"epoch": 4.894736842105263,
"grad_norm": 0.33966716848608597,
"learning_rate": 5.4028472710137e-08,
"loss": 1.0925,
"step": 1209
},
{
"epoch": 4.898785425101215,
"grad_norm": 0.32359224167224987,
"learning_rate": 4.9954057826571146e-08,
"loss": 1.0945,
"step": 1210
},
{
"epoch": 4.902834008097166,
"grad_norm": 0.3326142428072715,
"learning_rate": 4.6039162978743776e-08,
"loss": 1.0826,
"step": 1211
},
{
"epoch": 4.906882591093117,
"grad_norm": 0.32372406288216726,
"learning_rate": 4.228381947004812e-08,
"loss": 1.0837,
"step": 1212
},
{
"epoch": 4.910931174089069,
"grad_norm": 0.32721542348794536,
"learning_rate": 3.868805732811343e-08,
"loss": 1.0925,
"step": 1213
},
{
"epoch": 4.91497975708502,
"grad_norm": 0.3126042422892713,
"learning_rate": 3.525190530455636e-08,
"loss": 1.1549,
"step": 1214
},
{
"epoch": 4.919028340080971,
"grad_norm": 0.31704416033560123,
"learning_rate": 3.197539087475887e-08,
"loss": 1.1081,
"step": 1215
},
{
"epoch": 4.923076923076923,
"grad_norm": 0.327187607173195,
"learning_rate": 2.8858540237641786e-08,
"loss": 1.0484,
"step": 1216
},
{
"epoch": 4.9271255060728745,
"grad_norm": 0.3345304183870678,
"learning_rate": 2.5901378315458247e-08,
"loss": 1.1067,
"step": 1217
},
{
"epoch": 4.931174089068826,
"grad_norm": 0.32943185701495026,
"learning_rate": 2.3103928753598347e-08,
"loss": 1.1341,
"step": 1218
},
{
"epoch": 4.935222672064778,
"grad_norm": 0.30647521104792297,
"learning_rate": 2.0466213920393718e-08,
"loss": 1.1183,
"step": 1219
},
{
"epoch": 4.939271255060729,
"grad_norm": 0.34941237478483245,
"learning_rate": 1.7988254906942115e-08,
"loss": 1.0941,
"step": 1220
},
{
"epoch": 4.94331983805668,
"grad_norm": 0.31206713509905354,
"learning_rate": 1.5670071526936447e-08,
"loss": 1.1184,
"step": 1221
},
{
"epoch": 4.947368421052632,
"grad_norm": 0.320295107624868,
"learning_rate": 1.3511682316509344e-08,
"loss": 1.0875,
"step": 1222
},
{
"epoch": 4.951417004048583,
"grad_norm": 0.3351097341912554,
"learning_rate": 1.1513104534086605e-08,
"loss": 1.0496,
"step": 1223
},
{
"epoch": 4.955465587044534,
"grad_norm": 0.29273772833508593,
"learning_rate": 9.67435416023843e-09,
"loss": 1.115,
"step": 1224
},
{
"epoch": 4.959514170040486,
"grad_norm": 0.32387472524294286,
"learning_rate": 7.995445897566178e-09,
"loss": 1.0622,
"step": 1225
},
{
"epoch": 4.963562753036437,
"grad_norm": 0.3042866628176389,
"learning_rate": 6.476393170578021e-09,
"loss": 1.1029,
"step": 1226
},
{
"epoch": 4.967611336032388,
"grad_norm": 0.3226638356242307,
"learning_rate": 5.11720812557348e-09,
"loss": 1.0991,
"step": 1227
},
{
"epoch": 4.97165991902834,
"grad_norm": 0.3034870770868901,
"learning_rate": 3.917901630563492e-09,
"loss": 1.085,
"step": 1228
},
{
"epoch": 4.9757085020242915,
"grad_norm": 0.32130586231968705,
"learning_rate": 2.878483275170485e-09,
"loss": 1.1113,
"step": 1229
},
{
"epoch": 4.979757085020243,
"grad_norm": 0.32716811607291846,
"learning_rate": 1.998961370555108e-09,
"loss": 1.1312,
"step": 1230
},
{
"epoch": 4.983805668016195,
"grad_norm": 0.3137863401182357,
"learning_rate": 1.2793429493518361e-09,
"loss": 1.0935,
"step": 1231
},
{
"epoch": 4.987854251012146,
"grad_norm": 0.31246502936907594,
"learning_rate": 7.196337656112384e-10,
"loss": 1.1141,
"step": 1232
},
{
"epoch": 4.991902834008097,
"grad_norm": 0.3338806907053388,
"learning_rate": 3.198382947511292e-10,
"loss": 1.118,
"step": 1233
},
{
"epoch": 4.995951417004049,
"grad_norm": 0.3051288829844342,
"learning_rate": 7.995973352770137e-11,
"loss": 1.0554,
"step": 1234
},
{
"epoch": 5.0,
"grad_norm": 0.2930568939918333,
"learning_rate": 0.0,
"loss": 1.0946,
"step": 1235
},
{
"epoch": 5.0,
"step": 1235,
"total_flos": 3.995134388089127e+18,
"train_loss": 1.2891550317949612,
"train_runtime": 26946.1158,
"train_samples_per_second": 5.864,
"train_steps_per_second": 0.046
}
],
"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": 3.995134388089127e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}