sparky-decoder-5 / checkpoint-500 /trainer_state.json
Dohahemdann's picture
Checkpoint checkpoint-500
948d0a0 verified
Raw
History Blame Contribute Delete
86.8 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.4207868714496108,
"eval_steps": 500,
"global_step": 500,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0008415737428992215,
"grad_norm": 1220866.75,
"learning_rate": 0.0,
"loss": 6.7045,
"step": 1
},
{
"epoch": 0.001683147485798443,
"grad_norm": 1159900.75,
"learning_rate": 1.0000000000000001e-07,
"loss": 5.5719,
"step": 2
},
{
"epoch": 0.0025247212286976647,
"grad_norm": 179539.0,
"learning_rate": 2.0000000000000002e-07,
"loss": 2.9643,
"step": 3
},
{
"epoch": 0.003366294971596886,
"grad_norm": 60915.5234375,
"learning_rate": 3.0000000000000004e-07,
"loss": 2.3161,
"step": 4
},
{
"epoch": 0.004207868714496108,
"grad_norm": 41937.33984375,
"learning_rate": 4.0000000000000003e-07,
"loss": 2.3402,
"step": 5
},
{
"epoch": 0.005049442457395329,
"grad_norm": 25547.8046875,
"learning_rate": 5.000000000000001e-07,
"loss": 2.0138,
"step": 6
},
{
"epoch": 0.005891016200294551,
"grad_norm": 22664.328125,
"learning_rate": 6.000000000000001e-07,
"loss": 2.0857,
"step": 7
},
{
"epoch": 0.006732589943193772,
"grad_norm": 14687.037109375,
"learning_rate": 7.000000000000001e-07,
"loss": 2.0964,
"step": 8
},
{
"epoch": 0.007574163686092994,
"grad_norm": 13980.6904296875,
"learning_rate": 8.000000000000001e-07,
"loss": 2.4171,
"step": 9
},
{
"epoch": 0.008415737428992216,
"grad_norm": 9177.841796875,
"learning_rate": 9.000000000000001e-07,
"loss": 2.0959,
"step": 10
},
{
"epoch": 0.009257311171891438,
"grad_norm": 7599.8466796875,
"learning_rate": 1.0000000000000002e-06,
"loss": 1.7852,
"step": 11
},
{
"epoch": 0.010098884914790659,
"grad_norm": 6499.0732421875,
"learning_rate": 1.1e-06,
"loss": 1.6924,
"step": 12
},
{
"epoch": 0.01094045865768988,
"grad_norm": 6869.11865234375,
"learning_rate": 1.2000000000000002e-06,
"loss": 1.6957,
"step": 13
},
{
"epoch": 0.011782032400589101,
"grad_norm": 6527.53662109375,
"learning_rate": 1.3e-06,
"loss": 1.6633,
"step": 14
},
{
"epoch": 0.012623606143488323,
"grad_norm": 6013.7412109375,
"learning_rate": 1.4000000000000001e-06,
"loss": 1.8745,
"step": 15
},
{
"epoch": 0.013465179886387544,
"grad_norm": 7231.5712890625,
"learning_rate": 1.5e-06,
"loss": 1.5122,
"step": 16
},
{
"epoch": 0.014306753629286767,
"grad_norm": 5269.26318359375,
"learning_rate": 1.6000000000000001e-06,
"loss": 1.8746,
"step": 17
},
{
"epoch": 0.015148327372185988,
"grad_norm": 6138.6845703125,
"learning_rate": 1.7000000000000002e-06,
"loss": 1.9535,
"step": 18
},
{
"epoch": 0.01598990111508521,
"grad_norm": 5054.857421875,
"learning_rate": 1.8000000000000001e-06,
"loss": 1.8551,
"step": 19
},
{
"epoch": 0.016831474857984433,
"grad_norm": 5286.1708984375,
"learning_rate": 1.9000000000000002e-06,
"loss": 1.8266,
"step": 20
},
{
"epoch": 0.017673048600883652,
"grad_norm": 3078.6181640625,
"learning_rate": 2.0000000000000003e-06,
"loss": 1.5621,
"step": 21
},
{
"epoch": 0.018514622343782875,
"grad_norm": 4725.95654296875,
"learning_rate": 2.1000000000000002e-06,
"loss": 1.4557,
"step": 22
},
{
"epoch": 0.019356196086682095,
"grad_norm": 3049.34375,
"learning_rate": 2.2e-06,
"loss": 1.6599,
"step": 23
},
{
"epoch": 0.020197769829581318,
"grad_norm": 2984.544921875,
"learning_rate": 2.3000000000000004e-06,
"loss": 1.7747,
"step": 24
},
{
"epoch": 0.021039343572480537,
"grad_norm": 6122.8681640625,
"learning_rate": 2.4000000000000003e-06,
"loss": 2.2994,
"step": 25
},
{
"epoch": 0.02188091731537976,
"grad_norm": 2451.735107421875,
"learning_rate": 2.5e-06,
"loss": 1.6198,
"step": 26
},
{
"epoch": 0.022722491058278983,
"grad_norm": 2788.764404296875,
"learning_rate": 2.6e-06,
"loss": 1.6456,
"step": 27
},
{
"epoch": 0.023564064801178203,
"grad_norm": 2508.93408203125,
"learning_rate": 2.7000000000000004e-06,
"loss": 1.6448,
"step": 28
},
{
"epoch": 0.024405638544077426,
"grad_norm": 2768.5078125,
"learning_rate": 2.8000000000000003e-06,
"loss": 1.7862,
"step": 29
},
{
"epoch": 0.025247212286976645,
"grad_norm": 1970.8114013671875,
"learning_rate": 2.9e-06,
"loss": 1.6794,
"step": 30
},
{
"epoch": 0.02608878602987587,
"grad_norm": 2229.092041015625,
"learning_rate": 3e-06,
"loss": 1.5213,
"step": 31
},
{
"epoch": 0.026930359772775088,
"grad_norm": 2219.89990234375,
"learning_rate": 3.1000000000000004e-06,
"loss": 1.5728,
"step": 32
},
{
"epoch": 0.02777193351567431,
"grad_norm": 2461.462158203125,
"learning_rate": 3.2000000000000003e-06,
"loss": 1.8106,
"step": 33
},
{
"epoch": 0.028613507258573534,
"grad_norm": 1786.85693359375,
"learning_rate": 3.3000000000000006e-06,
"loss": 1.5036,
"step": 34
},
{
"epoch": 0.029455081001472753,
"grad_norm": 1430.862060546875,
"learning_rate": 3.4000000000000005e-06,
"loss": 1.2056,
"step": 35
},
{
"epoch": 0.030296654744371977,
"grad_norm": 1801.9127197265625,
"learning_rate": 3.5e-06,
"loss": 1.3607,
"step": 36
},
{
"epoch": 0.031138228487271196,
"grad_norm": 1663.1851806640625,
"learning_rate": 3.6000000000000003e-06,
"loss": 1.4841,
"step": 37
},
{
"epoch": 0.03197980223017042,
"grad_norm": 2743.8642578125,
"learning_rate": 3.7e-06,
"loss": 1.629,
"step": 38
},
{
"epoch": 0.03282137597306964,
"grad_norm": 1889.03662109375,
"learning_rate": 3.8000000000000005e-06,
"loss": 1.5254,
"step": 39
},
{
"epoch": 0.033662949715968865,
"grad_norm": 1672.36279296875,
"learning_rate": 3.900000000000001e-06,
"loss": 1.3192,
"step": 40
},
{
"epoch": 0.034504523458868085,
"grad_norm": 1701.8798828125,
"learning_rate": 4.000000000000001e-06,
"loss": 1.5442,
"step": 41
},
{
"epoch": 0.035346097201767304,
"grad_norm": 1488.308349609375,
"learning_rate": 4.1e-06,
"loss": 1.7156,
"step": 42
},
{
"epoch": 0.036187670944666524,
"grad_norm": 1526.427001953125,
"learning_rate": 4.2000000000000004e-06,
"loss": 1.8971,
"step": 43
},
{
"epoch": 0.03702924468756575,
"grad_norm": 1431.5985107421875,
"learning_rate": 4.3e-06,
"loss": 1.7197,
"step": 44
},
{
"epoch": 0.03787081843046497,
"grad_norm": 1157.3121337890625,
"learning_rate": 4.4e-06,
"loss": 1.1669,
"step": 45
},
{
"epoch": 0.03871239217336419,
"grad_norm": 1217.6405029296875,
"learning_rate": 4.5e-06,
"loss": 1.7516,
"step": 46
},
{
"epoch": 0.039553965916263416,
"grad_norm": 1782.4837646484375,
"learning_rate": 4.600000000000001e-06,
"loss": 1.5455,
"step": 47
},
{
"epoch": 0.040395539659162635,
"grad_norm": 1759.3140869140625,
"learning_rate": 4.7e-06,
"loss": 1.559,
"step": 48
},
{
"epoch": 0.041237113402061855,
"grad_norm": 1439.0577392578125,
"learning_rate": 4.800000000000001e-06,
"loss": 1.6599,
"step": 49
},
{
"epoch": 0.042078687144961074,
"grad_norm": 1564.3431396484375,
"learning_rate": 4.9000000000000005e-06,
"loss": 1.666,
"step": 50
},
{
"epoch": 0.0429202608878603,
"grad_norm": 1380.912109375,
"learning_rate": 5e-06,
"loss": 1.6822,
"step": 51
},
{
"epoch": 0.04376183463075952,
"grad_norm": 1470.3021240234375,
"learning_rate": 4.999999002610039e-06,
"loss": 1.5958,
"step": 52
},
{
"epoch": 0.04460340837365874,
"grad_norm": 2070.1259765625,
"learning_rate": 4.999996010440949e-06,
"loss": 1.599,
"step": 53
},
{
"epoch": 0.045444982116557966,
"grad_norm": 1041.359375,
"learning_rate": 4.999991023495121e-06,
"loss": 1.5113,
"step": 54
},
{
"epoch": 0.046286555859457186,
"grad_norm": 917.322998046875,
"learning_rate": 4.9999840417765314e-06,
"loss": 1.7166,
"step": 55
},
{
"epoch": 0.047128129602356406,
"grad_norm": 1014.3453979492188,
"learning_rate": 4.999975065290751e-06,
"loss": 1.4541,
"step": 56
},
{
"epoch": 0.047969703345255625,
"grad_norm": 1376.1556396484375,
"learning_rate": 4.999964094044944e-06,
"loss": 1.5334,
"step": 57
},
{
"epoch": 0.04881127708815485,
"grad_norm": 763.3536376953125,
"learning_rate": 4.999951128047863e-06,
"loss": 1.4616,
"step": 58
},
{
"epoch": 0.04965285083105407,
"grad_norm": 782.1888427734375,
"learning_rate": 4.999936167309855e-06,
"loss": 1.2268,
"step": 59
},
{
"epoch": 0.05049442457395329,
"grad_norm": 1264.4327392578125,
"learning_rate": 4.999919211842856e-06,
"loss": 1.8287,
"step": 60
},
{
"epoch": 0.05133599831685252,
"grad_norm": 792.3248901367188,
"learning_rate": 4.999900261660395e-06,
"loss": 1.4994,
"step": 61
},
{
"epoch": 0.05217757205975174,
"grad_norm": 707.390380859375,
"learning_rate": 4.999879316777594e-06,
"loss": 1.4387,
"step": 62
},
{
"epoch": 0.053019145802650956,
"grad_norm": 753.5997924804688,
"learning_rate": 4.999856377211164e-06,
"loss": 1.3524,
"step": 63
},
{
"epoch": 0.053860719545550176,
"grad_norm": 1315.026611328125,
"learning_rate": 4.99983144297941e-06,
"loss": 1.438,
"step": 64
},
{
"epoch": 0.0547022932884494,
"grad_norm": 999.6796875,
"learning_rate": 4.999804514102226e-06,
"loss": 1.6885,
"step": 65
},
{
"epoch": 0.05554386703134862,
"grad_norm": 1015.9467163085938,
"learning_rate": 4.999775590601098e-06,
"loss": 1.5048,
"step": 66
},
{
"epoch": 0.05638544077424784,
"grad_norm": 904.549560546875,
"learning_rate": 4.999744672499107e-06,
"loss": 1.3633,
"step": 67
},
{
"epoch": 0.05722701451714707,
"grad_norm": 802.7755737304688,
"learning_rate": 4.999711759820921e-06,
"loss": 1.3909,
"step": 68
},
{
"epoch": 0.05806858826004629,
"grad_norm": 988.9061279296875,
"learning_rate": 4.999676852592803e-06,
"loss": 1.7298,
"step": 69
},
{
"epoch": 0.05891016200294551,
"grad_norm": 804.6920776367188,
"learning_rate": 4.9996399508426045e-06,
"loss": 1.5774,
"step": 70
},
{
"epoch": 0.059751735745844727,
"grad_norm": 983.9691772460938,
"learning_rate": 4.99960105459977e-06,
"loss": 1.4138,
"step": 71
},
{
"epoch": 0.06059330948874395,
"grad_norm": 1331.3961181640625,
"learning_rate": 4.999560163895335e-06,
"loss": 1.5044,
"step": 72
},
{
"epoch": 0.06143488323164317,
"grad_norm": 623.4003295898438,
"learning_rate": 4.9995172787619285e-06,
"loss": 1.3104,
"step": 73
},
{
"epoch": 0.06227645697454239,
"grad_norm": 602.7695922851562,
"learning_rate": 4.999472399233767e-06,
"loss": 1.6953,
"step": 74
},
{
"epoch": 0.06311803071744161,
"grad_norm": 963.92724609375,
"learning_rate": 4.999425525346661e-06,
"loss": 1.4376,
"step": 75
},
{
"epoch": 0.06395960446034084,
"grad_norm": 625.63037109375,
"learning_rate": 4.9993766571380124e-06,
"loss": 1.6049,
"step": 76
},
{
"epoch": 0.06480117820324006,
"grad_norm": 686.1634521484375,
"learning_rate": 4.999325794646813e-06,
"loss": 1.5269,
"step": 77
},
{
"epoch": 0.06564275194613928,
"grad_norm": 644.1351928710938,
"learning_rate": 4.999272937913647e-06,
"loss": 1.3191,
"step": 78
},
{
"epoch": 0.0664843256890385,
"grad_norm": 672.7321166992188,
"learning_rate": 4.99921808698069e-06,
"loss": 1.1279,
"step": 79
},
{
"epoch": 0.06732589943193773,
"grad_norm": 993.0891723632812,
"learning_rate": 4.999161241891707e-06,
"loss": 1.6974,
"step": 80
},
{
"epoch": 0.06816747317483694,
"grad_norm": 778.0184936523438,
"learning_rate": 4.999102402692056e-06,
"loss": 1.6307,
"step": 81
},
{
"epoch": 0.06900904691773617,
"grad_norm": 601.3452758789062,
"learning_rate": 4.9990415694286845e-06,
"loss": 1.7501,
"step": 82
},
{
"epoch": 0.06985062066063538,
"grad_norm": 640.2106323242188,
"learning_rate": 4.998978742150133e-06,
"loss": 1.3542,
"step": 83
},
{
"epoch": 0.07069219440353461,
"grad_norm": 906.6204833984375,
"learning_rate": 4.998913920906534e-06,
"loss": 1.2639,
"step": 84
},
{
"epoch": 0.07153376814643383,
"grad_norm": 454.73944091796875,
"learning_rate": 4.998847105749606e-06,
"loss": 1.1869,
"step": 85
},
{
"epoch": 0.07237534188933305,
"grad_norm": 697.6354370117188,
"learning_rate": 4.998778296732663e-06,
"loss": 1.3469,
"step": 86
},
{
"epoch": 0.07321691563223227,
"grad_norm": 518.2359619140625,
"learning_rate": 4.998707493910607e-06,
"loss": 1.4724,
"step": 87
},
{
"epoch": 0.0740584893751315,
"grad_norm": 793.8117065429688,
"learning_rate": 4.998634697339936e-06,
"loss": 1.559,
"step": 88
},
{
"epoch": 0.07490006311803071,
"grad_norm": 452.31536865234375,
"learning_rate": 4.998559907078732e-06,
"loss": 1.1228,
"step": 89
},
{
"epoch": 0.07574163686092994,
"grad_norm": 671.4263305664062,
"learning_rate": 4.998483123186673e-06,
"loss": 1.3658,
"step": 90
},
{
"epoch": 0.07658321060382917,
"grad_norm": 494.4920959472656,
"learning_rate": 4.998404345725023e-06,
"loss": 1.4983,
"step": 91
},
{
"epoch": 0.07742478434672838,
"grad_norm": 564.9657592773438,
"learning_rate": 4.998323574756643e-06,
"loss": 1.4565,
"step": 92
},
{
"epoch": 0.0782663580896276,
"grad_norm": 588.97216796875,
"learning_rate": 4.998240810345979e-06,
"loss": 1.4131,
"step": 93
},
{
"epoch": 0.07910793183252683,
"grad_norm": 912.6625366210938,
"learning_rate": 4.998156052559071e-06,
"loss": 1.3753,
"step": 94
},
{
"epoch": 0.07994950557542604,
"grad_norm": 623.6527709960938,
"learning_rate": 4.998069301463546e-06,
"loss": 1.2489,
"step": 95
},
{
"epoch": 0.08079107931832527,
"grad_norm": 479.7945251464844,
"learning_rate": 4.997980557128627e-06,
"loss": 1.127,
"step": 96
},
{
"epoch": 0.08163265306122448,
"grad_norm": 445.030517578125,
"learning_rate": 4.997889819625121e-06,
"loss": 1.2625,
"step": 97
},
{
"epoch": 0.08247422680412371,
"grad_norm": 576.0341796875,
"learning_rate": 4.99779708902543e-06,
"loss": 1.5363,
"step": 98
},
{
"epoch": 0.08331580054702294,
"grad_norm": 646.4935302734375,
"learning_rate": 4.997702365403546e-06,
"loss": 1.9151,
"step": 99
},
{
"epoch": 0.08415737428992215,
"grad_norm": 688.9935302734375,
"learning_rate": 4.997605648835048e-06,
"loss": 1.3353,
"step": 100
},
{
"epoch": 0.08499894803282138,
"grad_norm": 547.1505737304688,
"learning_rate": 4.9975069393971076e-06,
"loss": 1.2941,
"step": 101
},
{
"epoch": 0.0858405217757206,
"grad_norm": 585.2498779296875,
"learning_rate": 4.997406237168488e-06,
"loss": 1.4254,
"step": 102
},
{
"epoch": 0.08668209551861981,
"grad_norm": 503.3772277832031,
"learning_rate": 4.997303542229539e-06,
"loss": 1.2362,
"step": 103
},
{
"epoch": 0.08752366926151904,
"grad_norm": 455.1918640136719,
"learning_rate": 4.997198854662203e-06,
"loss": 1.182,
"step": 104
},
{
"epoch": 0.08836524300441827,
"grad_norm": 571.8682861328125,
"learning_rate": 4.9970921745500104e-06,
"loss": 1.5275,
"step": 105
},
{
"epoch": 0.08920681674731748,
"grad_norm": 494.3360900878906,
"learning_rate": 4.996983501978084e-06,
"loss": 1.3781,
"step": 106
},
{
"epoch": 0.0900483904902167,
"grad_norm": 550.50732421875,
"learning_rate": 4.996872837033134e-06,
"loss": 1.4291,
"step": 107
},
{
"epoch": 0.09088996423311593,
"grad_norm": 601.7146606445312,
"learning_rate": 4.996760179803461e-06,
"loss": 1.4825,
"step": 108
},
{
"epoch": 0.09173153797601515,
"grad_norm": 441.38037109375,
"learning_rate": 4.996645530378957e-06,
"loss": 1.4476,
"step": 109
},
{
"epoch": 0.09257311171891437,
"grad_norm": 413.4273986816406,
"learning_rate": 4.996528888851101e-06,
"loss": 1.2163,
"step": 110
},
{
"epoch": 0.09341468546181358,
"grad_norm": 580.2811279296875,
"learning_rate": 4.996410255312962e-06,
"loss": 1.1997,
"step": 111
},
{
"epoch": 0.09425625920471281,
"grad_norm": 347.80322265625,
"learning_rate": 4.996289629859201e-06,
"loss": 1.3247,
"step": 112
},
{
"epoch": 0.09509783294761204,
"grad_norm": 521.5128173828125,
"learning_rate": 4.996167012586065e-06,
"loss": 1.3204,
"step": 113
},
{
"epoch": 0.09593940669051125,
"grad_norm": 384.0645446777344,
"learning_rate": 4.996042403591393e-06,
"loss": 1.3969,
"step": 114
},
{
"epoch": 0.09678098043341048,
"grad_norm": 413.4881591796875,
"learning_rate": 4.99591580297461e-06,
"loss": 1.4374,
"step": 115
},
{
"epoch": 0.0976225541763097,
"grad_norm": 389.8695983886719,
"learning_rate": 4.995787210836734e-06,
"loss": 1.1474,
"step": 116
},
{
"epoch": 0.09846412791920892,
"grad_norm": 382.87823486328125,
"learning_rate": 4.995656627280371e-06,
"loss": 1.3191,
"step": 117
},
{
"epoch": 0.09930570166210814,
"grad_norm": 471.7465515136719,
"learning_rate": 4.995524052409713e-06,
"loss": 1.5302,
"step": 118
},
{
"epoch": 0.10014727540500737,
"grad_norm": 347.9407958984375,
"learning_rate": 4.995389486330543e-06,
"loss": 1.1651,
"step": 119
},
{
"epoch": 0.10098884914790658,
"grad_norm": 490.6868896484375,
"learning_rate": 4.995252929150234e-06,
"loss": 1.2531,
"step": 120
},
{
"epoch": 0.10183042289080581,
"grad_norm": 452.0323486328125,
"learning_rate": 4.995114380977747e-06,
"loss": 1.3209,
"step": 121
},
{
"epoch": 0.10267199663370503,
"grad_norm": 406.12701416015625,
"learning_rate": 4.99497384192363e-06,
"loss": 1.5792,
"step": 122
},
{
"epoch": 0.10351357037660425,
"grad_norm": 394.3304138183594,
"learning_rate": 4.994831312100023e-06,
"loss": 1.2446,
"step": 123
},
{
"epoch": 0.10435514411950347,
"grad_norm": 377.8374938964844,
"learning_rate": 4.994686791620649e-06,
"loss": 0.9982,
"step": 124
},
{
"epoch": 0.1051967178624027,
"grad_norm": 365.93560791015625,
"learning_rate": 4.994540280600825e-06,
"loss": 1.1529,
"step": 125
},
{
"epoch": 0.10603829160530191,
"grad_norm": 368.8994140625,
"learning_rate": 4.994391779157452e-06,
"loss": 1.2085,
"step": 126
},
{
"epoch": 0.10687986534820114,
"grad_norm": 381.2408752441406,
"learning_rate": 4.994241287409024e-06,
"loss": 1.2553,
"step": 127
},
{
"epoch": 0.10772143909110035,
"grad_norm": 387.8236083984375,
"learning_rate": 4.994088805475618e-06,
"loss": 1.4945,
"step": 128
},
{
"epoch": 0.10856301283399958,
"grad_norm": 305.5185546875,
"learning_rate": 4.993934333478901e-06,
"loss": 1.0922,
"step": 129
},
{
"epoch": 0.1094045865768988,
"grad_norm": 408.90057373046875,
"learning_rate": 4.993777871542129e-06,
"loss": 1.2649,
"step": 130
},
{
"epoch": 0.11024616031979802,
"grad_norm": 420.7751159667969,
"learning_rate": 4.993619419790145e-06,
"loss": 1.1042,
"step": 131
},
{
"epoch": 0.11108773406269724,
"grad_norm": 331.85504150390625,
"learning_rate": 4.993458978349379e-06,
"loss": 1.1527,
"step": 132
},
{
"epoch": 0.11192930780559647,
"grad_norm": 558.6054077148438,
"learning_rate": 4.993296547347848e-06,
"loss": 1.4046,
"step": 133
},
{
"epoch": 0.11277088154849568,
"grad_norm": 365.9854736328125,
"learning_rate": 4.99313212691516e-06,
"loss": 1.3445,
"step": 134
},
{
"epoch": 0.11361245529139491,
"grad_norm": 312.7580261230469,
"learning_rate": 4.992965717182507e-06,
"loss": 1.3768,
"step": 135
},
{
"epoch": 0.11445402903429414,
"grad_norm": 460.6139831542969,
"learning_rate": 4.992797318282668e-06,
"loss": 1.4361,
"step": 136
},
{
"epoch": 0.11529560277719335,
"grad_norm": 378.0125732421875,
"learning_rate": 4.992626930350013e-06,
"loss": 1.0356,
"step": 137
},
{
"epoch": 0.11613717652009257,
"grad_norm": 435.4710388183594,
"learning_rate": 4.992454553520495e-06,
"loss": 1.3194,
"step": 138
},
{
"epoch": 0.1169787502629918,
"grad_norm": 347.2532043457031,
"learning_rate": 4.992280187931656e-06,
"loss": 1.5423,
"step": 139
},
{
"epoch": 0.11782032400589101,
"grad_norm": 314.25067138671875,
"learning_rate": 4.992103833722623e-06,
"loss": 0.9756,
"step": 140
},
{
"epoch": 0.11866189774879024,
"grad_norm": 304.3933410644531,
"learning_rate": 4.991925491034113e-06,
"loss": 1.3259,
"step": 141
},
{
"epoch": 0.11950347149168945,
"grad_norm": 301.6213073730469,
"learning_rate": 4.9917451600084275e-06,
"loss": 1.1353,
"step": 142
},
{
"epoch": 0.12034504523458868,
"grad_norm": 408.5636901855469,
"learning_rate": 4.991562840789455e-06,
"loss": 1.402,
"step": 143
},
{
"epoch": 0.1211866189774879,
"grad_norm": 340.44317626953125,
"learning_rate": 4.991378533522669e-06,
"loss": 1.353,
"step": 144
},
{
"epoch": 0.12202819272038712,
"grad_norm": 268.15936279296875,
"learning_rate": 4.99119223835513e-06,
"loss": 1.2545,
"step": 145
},
{
"epoch": 0.12286976646328635,
"grad_norm": 372.2171325683594,
"learning_rate": 4.991003955435487e-06,
"loss": 1.2183,
"step": 146
},
{
"epoch": 0.12371134020618557,
"grad_norm": 278.7896423339844,
"learning_rate": 4.990813684913972e-06,
"loss": 1.1768,
"step": 147
},
{
"epoch": 0.12455291394908478,
"grad_norm": 266.3719787597656,
"learning_rate": 4.990621426942404e-06,
"loss": 1.2064,
"step": 148
},
{
"epoch": 0.125394487691984,
"grad_norm": 308.0953369140625,
"learning_rate": 4.99042718167419e-06,
"loss": 1.3979,
"step": 149
},
{
"epoch": 0.12623606143488322,
"grad_norm": 414.5055236816406,
"learning_rate": 4.990230949264319e-06,
"loss": 1.1933,
"step": 150
},
{
"epoch": 0.12707763517778245,
"grad_norm": 328.179931640625,
"learning_rate": 4.990032729869366e-06,
"loss": 1.157,
"step": 151
},
{
"epoch": 0.12791920892068168,
"grad_norm": 468.81768798828125,
"learning_rate": 4.989832523647494e-06,
"loss": 1.15,
"step": 152
},
{
"epoch": 0.1287607826635809,
"grad_norm": 352.3136291503906,
"learning_rate": 4.989630330758451e-06,
"loss": 1.1188,
"step": 153
},
{
"epoch": 0.12960235640648013,
"grad_norm": 388.4230041503906,
"learning_rate": 4.9894261513635674e-06,
"loss": 0.9838,
"step": 154
},
{
"epoch": 0.13044393014937933,
"grad_norm": 251.35621643066406,
"learning_rate": 4.98921998562576e-06,
"loss": 1.0673,
"step": 155
},
{
"epoch": 0.13128550389227855,
"grad_norm": 438.343505859375,
"learning_rate": 4.989011833709534e-06,
"loss": 1.2557,
"step": 156
},
{
"epoch": 0.13212707763517778,
"grad_norm": 316.020751953125,
"learning_rate": 4.988801695780972e-06,
"loss": 1.0859,
"step": 157
},
{
"epoch": 0.132968651378077,
"grad_norm": 342.6540222167969,
"learning_rate": 4.988589572007749e-06,
"loss": 1.122,
"step": 158
},
{
"epoch": 0.13381022512097623,
"grad_norm": 396.69403076171875,
"learning_rate": 4.988375462559119e-06,
"loss": 1.2415,
"step": 159
},
{
"epoch": 0.13465179886387546,
"grad_norm": 480.0825500488281,
"learning_rate": 4.988159367605924e-06,
"loss": 1.5157,
"step": 160
},
{
"epoch": 0.13549337260677466,
"grad_norm": 414.4588928222656,
"learning_rate": 4.987941287320588e-06,
"loss": 1.0775,
"step": 161
},
{
"epoch": 0.13633494634967389,
"grad_norm": 345.3420715332031,
"learning_rate": 4.987721221877119e-06,
"loss": 1.3017,
"step": 162
},
{
"epoch": 0.1371765200925731,
"grad_norm": 395.992431640625,
"learning_rate": 4.987499171451112e-06,
"loss": 1.176,
"step": 163
},
{
"epoch": 0.13801809383547234,
"grad_norm": 241.21575927734375,
"learning_rate": 4.987275136219742e-06,
"loss": 1.1583,
"step": 164
},
{
"epoch": 0.13885966757837157,
"grad_norm": 459.3298034667969,
"learning_rate": 4.987049116361769e-06,
"loss": 1.2788,
"step": 165
},
{
"epoch": 0.13970124132127076,
"grad_norm": 408.43121337890625,
"learning_rate": 4.9868211120575396e-06,
"loss": 1.4012,
"step": 166
},
{
"epoch": 0.14054281506417,
"grad_norm": 457.90081787109375,
"learning_rate": 4.986591123488978e-06,
"loss": 1.445,
"step": 167
},
{
"epoch": 0.14138438880706922,
"grad_norm": 281.84515380859375,
"learning_rate": 4.9863591508395965e-06,
"loss": 1.3696,
"step": 168
},
{
"epoch": 0.14222596254996844,
"grad_norm": 332.0083923339844,
"learning_rate": 4.986125194294489e-06,
"loss": 1.0008,
"step": 169
},
{
"epoch": 0.14306753629286767,
"grad_norm": 339.5845031738281,
"learning_rate": 4.985889254040332e-06,
"loss": 1.3454,
"step": 170
},
{
"epoch": 0.1439091100357669,
"grad_norm": 841.4637451171875,
"learning_rate": 4.985651330265383e-06,
"loss": 1.2247,
"step": 171
},
{
"epoch": 0.1447506837786661,
"grad_norm": 336.28045654296875,
"learning_rate": 4.985411423159487e-06,
"loss": 1.3671,
"step": 172
},
{
"epoch": 0.14559225752156532,
"grad_norm": 392.8457946777344,
"learning_rate": 4.985169532914069e-06,
"loss": 1.1262,
"step": 173
},
{
"epoch": 0.14643383126446455,
"grad_norm": 301.209228515625,
"learning_rate": 4.984925659722134e-06,
"loss": 1.1485,
"step": 174
},
{
"epoch": 0.14727540500736377,
"grad_norm": 277.6703796386719,
"learning_rate": 4.984679803778273e-06,
"loss": 1.1578,
"step": 175
},
{
"epoch": 0.148116978750263,
"grad_norm": 302.1859436035156,
"learning_rate": 4.984431965278655e-06,
"loss": 1.2856,
"step": 176
},
{
"epoch": 0.1489585524931622,
"grad_norm": 300.76025390625,
"learning_rate": 4.984182144421037e-06,
"loss": 1.0614,
"step": 177
},
{
"epoch": 0.14980012623606143,
"grad_norm": 332.2203674316406,
"learning_rate": 4.983930341404751e-06,
"loss": 1.1719,
"step": 178
},
{
"epoch": 0.15064169997896065,
"grad_norm": 204.2970733642578,
"learning_rate": 4.983676556430715e-06,
"loss": 0.9967,
"step": 179
},
{
"epoch": 0.15148327372185988,
"grad_norm": 320.5247802734375,
"learning_rate": 4.983420789701427e-06,
"loss": 1.255,
"step": 180
},
{
"epoch": 0.1523248474647591,
"grad_norm": 249.64166259765625,
"learning_rate": 4.983163041420965e-06,
"loss": 0.8233,
"step": 181
},
{
"epoch": 0.15316642120765833,
"grad_norm": 246.04092407226562,
"learning_rate": 4.982903311794991e-06,
"loss": 1.3713,
"step": 182
},
{
"epoch": 0.15400799495055753,
"grad_norm": 302.2898864746094,
"learning_rate": 4.9826416010307464e-06,
"loss": 1.159,
"step": 183
},
{
"epoch": 0.15484956869345676,
"grad_norm": 252.8140411376953,
"learning_rate": 4.982377909337053e-06,
"loss": 1.0992,
"step": 184
},
{
"epoch": 0.15569114243635598,
"grad_norm": 262.8348693847656,
"learning_rate": 4.982112236924312e-06,
"loss": 0.9386,
"step": 185
},
{
"epoch": 0.1565327161792552,
"grad_norm": 262.5164489746094,
"learning_rate": 4.98184458400451e-06,
"loss": 1.2239,
"step": 186
},
{
"epoch": 0.15737428992215444,
"grad_norm": 231.84715270996094,
"learning_rate": 4.981574950791207e-06,
"loss": 0.9828,
"step": 187
},
{
"epoch": 0.15821586366505366,
"grad_norm": 240.302001953125,
"learning_rate": 4.981303337499549e-06,
"loss": 0.991,
"step": 188
},
{
"epoch": 0.15905743740795286,
"grad_norm": 246.06973266601562,
"learning_rate": 4.981029744346258e-06,
"loss": 1.1315,
"step": 189
},
{
"epoch": 0.1598990111508521,
"grad_norm": 302.11181640625,
"learning_rate": 4.980754171549638e-06,
"loss": 1.1767,
"step": 190
},
{
"epoch": 0.16074058489375131,
"grad_norm": 253.83128356933594,
"learning_rate": 4.980476619329571e-06,
"loss": 1.1659,
"step": 191
},
{
"epoch": 0.16158215863665054,
"grad_norm": 261.5288391113281,
"learning_rate": 4.980197087907521e-06,
"loss": 1.296,
"step": 192
},
{
"epoch": 0.16242373237954977,
"grad_norm": 240.8744659423828,
"learning_rate": 4.979915577506527e-06,
"loss": 1.2207,
"step": 193
},
{
"epoch": 0.16326530612244897,
"grad_norm": 247.27427673339844,
"learning_rate": 4.979632088351211e-06,
"loss": 1.1433,
"step": 194
},
{
"epoch": 0.1641068798653482,
"grad_norm": 225.68511962890625,
"learning_rate": 4.979346620667774e-06,
"loss": 1.1316,
"step": 195
},
{
"epoch": 0.16494845360824742,
"grad_norm": 235.21524047851562,
"learning_rate": 4.979059174683991e-06,
"loss": 1.0509,
"step": 196
},
{
"epoch": 0.16579002735114665,
"grad_norm": 211.19992065429688,
"learning_rate": 4.9787697506292195e-06,
"loss": 1.1184,
"step": 197
},
{
"epoch": 0.16663160109404587,
"grad_norm": 299.4044494628906,
"learning_rate": 4.978478348734396e-06,
"loss": 1.2405,
"step": 198
},
{
"epoch": 0.1674731748369451,
"grad_norm": 226.4676055908203,
"learning_rate": 4.978184969232032e-06,
"loss": 1.2609,
"step": 199
},
{
"epoch": 0.1683147485798443,
"grad_norm": 302.3034973144531,
"learning_rate": 4.977889612356218e-06,
"loss": 1.0461,
"step": 200
},
{
"epoch": 0.16915632232274352,
"grad_norm": 336.5556335449219,
"learning_rate": 4.9775922783426246e-06,
"loss": 1.1755,
"step": 201
},
{
"epoch": 0.16999789606564275,
"grad_norm": 159.33302307128906,
"learning_rate": 4.977292967428497e-06,
"loss": 0.8049,
"step": 202
},
{
"epoch": 0.17083946980854198,
"grad_norm": 273.533935546875,
"learning_rate": 4.97699167985266e-06,
"loss": 0.9883,
"step": 203
},
{
"epoch": 0.1716810435514412,
"grad_norm": 333.1833801269531,
"learning_rate": 4.976688415855513e-06,
"loss": 1.172,
"step": 204
},
{
"epoch": 0.17252261729434043,
"grad_norm": 317.8067321777344,
"learning_rate": 4.976383175679034e-06,
"loss": 1.1309,
"step": 205
},
{
"epoch": 0.17336419103723963,
"grad_norm": 275.27545166015625,
"learning_rate": 4.97607595956678e-06,
"loss": 1.0488,
"step": 206
},
{
"epoch": 0.17420576478013886,
"grad_norm": 225.5316162109375,
"learning_rate": 4.97576676776388e-06,
"loss": 1.0911,
"step": 207
},
{
"epoch": 0.17504733852303808,
"grad_norm": 236.5315399169922,
"learning_rate": 4.975455600517043e-06,
"loss": 0.9873,
"step": 208
},
{
"epoch": 0.1758889122659373,
"grad_norm": 200.12295532226562,
"learning_rate": 4.975142458074553e-06,
"loss": 1.3244,
"step": 209
},
{
"epoch": 0.17673048600883653,
"grad_norm": 836.041259765625,
"learning_rate": 4.97482734068627e-06,
"loss": 1.3085,
"step": 210
},
{
"epoch": 0.17757205975173573,
"grad_norm": 208.06675720214844,
"learning_rate": 4.97451024860363e-06,
"loss": 1.2389,
"step": 211
},
{
"epoch": 0.17841363349463496,
"grad_norm": 250.1989288330078,
"learning_rate": 4.974191182079644e-06,
"loss": 1.0525,
"step": 212
},
{
"epoch": 0.1792552072375342,
"grad_norm": 273.5354919433594,
"learning_rate": 4.9738701413689006e-06,
"loss": 1.2381,
"step": 213
},
{
"epoch": 0.1800967809804334,
"grad_norm": 646.2859497070312,
"learning_rate": 4.97354712672756e-06,
"loss": 0.9527,
"step": 214
},
{
"epoch": 0.18093835472333264,
"grad_norm": 351.3960266113281,
"learning_rate": 4.97322213841336e-06,
"loss": 1.1817,
"step": 215
},
{
"epoch": 0.18177992846623187,
"grad_norm": 270.67529296875,
"learning_rate": 4.972895176685614e-06,
"loss": 1.2518,
"step": 216
},
{
"epoch": 0.18262150220913106,
"grad_norm": 191.62271118164062,
"learning_rate": 4.972566241805208e-06,
"loss": 0.991,
"step": 217
},
{
"epoch": 0.1834630759520303,
"grad_norm": 233.8841094970703,
"learning_rate": 4.972235334034601e-06,
"loss": 1.171,
"step": 218
},
{
"epoch": 0.18430464969492952,
"grad_norm": 260.0747375488281,
"learning_rate": 4.971902453637831e-06,
"loss": 1.0585,
"step": 219
},
{
"epoch": 0.18514622343782874,
"grad_norm": 339.6416015625,
"learning_rate": 4.971567600880507e-06,
"loss": 0.9673,
"step": 220
},
{
"epoch": 0.18598779718072797,
"grad_norm": 227.68484497070312,
"learning_rate": 4.97123077602981e-06,
"loss": 1.1898,
"step": 221
},
{
"epoch": 0.18682937092362717,
"grad_norm": 221.54664611816406,
"learning_rate": 4.970891979354499e-06,
"loss": 1.0242,
"step": 222
},
{
"epoch": 0.1876709446665264,
"grad_norm": 354.9067077636719,
"learning_rate": 4.970551211124902e-06,
"loss": 1.3887,
"step": 223
},
{
"epoch": 0.18851251840942562,
"grad_norm": 351.784912109375,
"learning_rate": 4.970208471612924e-06,
"loss": 1.2994,
"step": 224
},
{
"epoch": 0.18935409215232485,
"grad_norm": 449.1482849121094,
"learning_rate": 4.969863761092038e-06,
"loss": 1.196,
"step": 225
},
{
"epoch": 0.19019566589522408,
"grad_norm": 306.9056396484375,
"learning_rate": 4.9695170798372966e-06,
"loss": 0.8384,
"step": 226
},
{
"epoch": 0.1910372396381233,
"grad_norm": 276.563720703125,
"learning_rate": 4.969168428125318e-06,
"loss": 1.1098,
"step": 227
},
{
"epoch": 0.1918788133810225,
"grad_norm": 269.2872009277344,
"learning_rate": 4.968817806234296e-06,
"loss": 1.1772,
"step": 228
},
{
"epoch": 0.19272038712392173,
"grad_norm": 231.6291046142578,
"learning_rate": 4.968465214443996e-06,
"loss": 0.9755,
"step": 229
},
{
"epoch": 0.19356196086682095,
"grad_norm": 256.9276123046875,
"learning_rate": 4.968110653035757e-06,
"loss": 1.1614,
"step": 230
},
{
"epoch": 0.19440353460972018,
"grad_norm": 243.2183074951172,
"learning_rate": 4.967754122292485e-06,
"loss": 1.1839,
"step": 231
},
{
"epoch": 0.1952451083526194,
"grad_norm": 250.42715454101562,
"learning_rate": 4.967395622498662e-06,
"loss": 1.1289,
"step": 232
},
{
"epoch": 0.19608668209551863,
"grad_norm": 220.76181030273438,
"learning_rate": 4.9670351539403395e-06,
"loss": 1.1068,
"step": 233
},
{
"epoch": 0.19692825583841783,
"grad_norm": 199.235107421875,
"learning_rate": 4.9666727169051385e-06,
"loss": 0.9976,
"step": 234
},
{
"epoch": 0.19776982958131706,
"grad_norm": 225.22341918945312,
"learning_rate": 4.9663083116822515e-06,
"loss": 0.9086,
"step": 235
},
{
"epoch": 0.19861140332421628,
"grad_norm": 203.07470703125,
"learning_rate": 4.965941938562444e-06,
"loss": 0.977,
"step": 236
},
{
"epoch": 0.1994529770671155,
"grad_norm": 204.74114990234375,
"learning_rate": 4.965573597838047e-06,
"loss": 0.8848,
"step": 237
},
{
"epoch": 0.20029455081001474,
"grad_norm": 196.8950653076172,
"learning_rate": 4.965203289802966e-06,
"loss": 1.0225,
"step": 238
},
{
"epoch": 0.20113612455291394,
"grad_norm": 200.91964721679688,
"learning_rate": 4.964831014752673e-06,
"loss": 0.9533,
"step": 239
},
{
"epoch": 0.20197769829581316,
"grad_norm": 214.7240753173828,
"learning_rate": 4.964456772984211e-06,
"loss": 1.1088,
"step": 240
},
{
"epoch": 0.2028192720387124,
"grad_norm": 324.22711181640625,
"learning_rate": 4.964080564796192e-06,
"loss": 1.0528,
"step": 241
},
{
"epoch": 0.20366084578161162,
"grad_norm": 363.13189697265625,
"learning_rate": 4.963702390488797e-06,
"loss": 1.1069,
"step": 242
},
{
"epoch": 0.20450241952451084,
"grad_norm": 176.63153076171875,
"learning_rate": 4.963322250363776e-06,
"loss": 0.9716,
"step": 243
},
{
"epoch": 0.20534399326741007,
"grad_norm": 209.2146759033203,
"learning_rate": 4.962940144724446e-06,
"loss": 1.2695,
"step": 244
},
{
"epoch": 0.20618556701030927,
"grad_norm": 230.91806030273438,
"learning_rate": 4.9625560738756965e-06,
"loss": 1.1252,
"step": 245
},
{
"epoch": 0.2070271407532085,
"grad_norm": 219.78872680664062,
"learning_rate": 4.96217003812398e-06,
"loss": 0.9209,
"step": 246
},
{
"epoch": 0.20786871449610772,
"grad_norm": 331.7774353027344,
"learning_rate": 4.961782037777319e-06,
"loss": 1.0271,
"step": 247
},
{
"epoch": 0.20871028823900695,
"grad_norm": 238.35067749023438,
"learning_rate": 4.961392073145305e-06,
"loss": 1.0553,
"step": 248
},
{
"epoch": 0.20955186198190617,
"grad_norm": 219.87989807128906,
"learning_rate": 4.961000144539094e-06,
"loss": 1.1054,
"step": 249
},
{
"epoch": 0.2103934357248054,
"grad_norm": 213.93609619140625,
"learning_rate": 4.960606252271413e-06,
"loss": 1.0722,
"step": 250
},
{
"epoch": 0.2112350094677046,
"grad_norm": 173.34837341308594,
"learning_rate": 4.960210396656549e-06,
"loss": 0.7469,
"step": 251
},
{
"epoch": 0.21207658321060383,
"grad_norm": 176.90809631347656,
"learning_rate": 4.959812578010364e-06,
"loss": 0.8142,
"step": 252
},
{
"epoch": 0.21291815695350305,
"grad_norm": 145.44351196289062,
"learning_rate": 4.959412796650281e-06,
"loss": 0.6995,
"step": 253
},
{
"epoch": 0.21375973069640228,
"grad_norm": 164.6792755126953,
"learning_rate": 4.959011052895289e-06,
"loss": 0.8751,
"step": 254
},
{
"epoch": 0.2146013044393015,
"grad_norm": 179.0094451904297,
"learning_rate": 4.9586073470659465e-06,
"loss": 0.8738,
"step": 255
},
{
"epoch": 0.2154428781822007,
"grad_norm": 275.41729736328125,
"learning_rate": 4.958201679484373e-06,
"loss": 1.1572,
"step": 256
},
{
"epoch": 0.21628445192509993,
"grad_norm": 211.57325744628906,
"learning_rate": 4.9577940504742576e-06,
"loss": 1.1856,
"step": 257
},
{
"epoch": 0.21712602566799916,
"grad_norm": 164.41294860839844,
"learning_rate": 4.95738446036085e-06,
"loss": 0.7646,
"step": 258
},
{
"epoch": 0.21796759941089838,
"grad_norm": 204.46368408203125,
"learning_rate": 4.95697290947097e-06,
"loss": 0.9968,
"step": 259
},
{
"epoch": 0.2188091731537976,
"grad_norm": 166.3022003173828,
"learning_rate": 4.956559398132995e-06,
"loss": 0.9949,
"step": 260
},
{
"epoch": 0.21965074689669684,
"grad_norm": 183.39434814453125,
"learning_rate": 4.956143926676874e-06,
"loss": 0.8959,
"step": 261
},
{
"epoch": 0.22049232063959603,
"grad_norm": 184.88546752929688,
"learning_rate": 4.955726495434116e-06,
"loss": 0.9911,
"step": 262
},
{
"epoch": 0.22133389438249526,
"grad_norm": 261.6977844238281,
"learning_rate": 4.955307104737794e-06,
"loss": 1.0551,
"step": 263
},
{
"epoch": 0.2221754681253945,
"grad_norm": 206.3694610595703,
"learning_rate": 4.954885754922544e-06,
"loss": 1.0895,
"step": 264
},
{
"epoch": 0.22301704186829371,
"grad_norm": 192.2148895263672,
"learning_rate": 4.954462446324567e-06,
"loss": 0.9871,
"step": 265
},
{
"epoch": 0.22385861561119294,
"grad_norm": 181.06787109375,
"learning_rate": 4.954037179281626e-06,
"loss": 1.0505,
"step": 266
},
{
"epoch": 0.22470018935409214,
"grad_norm": 211.8572540283203,
"learning_rate": 4.9536099541330475e-06,
"loss": 0.9762,
"step": 267
},
{
"epoch": 0.22554176309699137,
"grad_norm": 229.56036376953125,
"learning_rate": 4.953180771219718e-06,
"loss": 0.9214,
"step": 268
},
{
"epoch": 0.2263833368398906,
"grad_norm": 175.8040771484375,
"learning_rate": 4.952749630884088e-06,
"loss": 0.8592,
"step": 269
},
{
"epoch": 0.22722491058278982,
"grad_norm": 205.4430389404297,
"learning_rate": 4.952316533470169e-06,
"loss": 1.093,
"step": 270
},
{
"epoch": 0.22806648432568905,
"grad_norm": 208.4731903076172,
"learning_rate": 4.951881479323537e-06,
"loss": 1.0825,
"step": 271
},
{
"epoch": 0.22890805806858827,
"grad_norm": 257.52972412109375,
"learning_rate": 4.951444468791325e-06,
"loss": 1.0441,
"step": 272
},
{
"epoch": 0.22974963181148747,
"grad_norm": 155.39646911621094,
"learning_rate": 4.951005502222229e-06,
"loss": 0.7643,
"step": 273
},
{
"epoch": 0.2305912055543867,
"grad_norm": 187.8433380126953,
"learning_rate": 4.950564579966506e-06,
"loss": 0.8654,
"step": 274
},
{
"epoch": 0.23143277929728592,
"grad_norm": 185.39447021484375,
"learning_rate": 4.950121702375973e-06,
"loss": 1.096,
"step": 275
},
{
"epoch": 0.23227435304018515,
"grad_norm": 164.85166931152344,
"learning_rate": 4.949676869804008e-06,
"loss": 0.9997,
"step": 276
},
{
"epoch": 0.23311592678308438,
"grad_norm": 231.10238647460938,
"learning_rate": 4.949230082605546e-06,
"loss": 0.8651,
"step": 277
},
{
"epoch": 0.2339575005259836,
"grad_norm": 220.4220428466797,
"learning_rate": 4.948781341137087e-06,
"loss": 1.245,
"step": 278
},
{
"epoch": 0.2347990742688828,
"grad_norm": 193.20501708984375,
"learning_rate": 4.948330645756686e-06,
"loss": 1.1347,
"step": 279
},
{
"epoch": 0.23564064801178203,
"grad_norm": 138.90548706054688,
"learning_rate": 4.947877996823957e-06,
"loss": 0.9499,
"step": 280
},
{
"epoch": 0.23648222175468125,
"grad_norm": 217.0552520751953,
"learning_rate": 4.947423394700075e-06,
"loss": 1.3172,
"step": 281
},
{
"epoch": 0.23732379549758048,
"grad_norm": 214.4515380859375,
"learning_rate": 4.946966839747773e-06,
"loss": 1.0737,
"step": 282
},
{
"epoch": 0.2381653692404797,
"grad_norm": 222.87826538085938,
"learning_rate": 4.946508332331341e-06,
"loss": 1.0215,
"step": 283
},
{
"epoch": 0.2390069429833789,
"grad_norm": 242.82290649414062,
"learning_rate": 4.946047872816628e-06,
"loss": 1.0079,
"step": 284
},
{
"epoch": 0.23984851672627813,
"grad_norm": 190.513427734375,
"learning_rate": 4.9455854615710385e-06,
"loss": 0.9054,
"step": 285
},
{
"epoch": 0.24069009046917736,
"grad_norm": 180.5825958251953,
"learning_rate": 4.945121098963538e-06,
"loss": 0.8978,
"step": 286
},
{
"epoch": 0.24153166421207659,
"grad_norm": 222.07452392578125,
"learning_rate": 4.944654785364647e-06,
"loss": 0.7585,
"step": 287
},
{
"epoch": 0.2423732379549758,
"grad_norm": 323.4588928222656,
"learning_rate": 4.94418652114644e-06,
"loss": 1.0618,
"step": 288
},
{
"epoch": 0.24321481169787504,
"grad_norm": 178.20001220703125,
"learning_rate": 4.943716306682555e-06,
"loss": 0.9913,
"step": 289
},
{
"epoch": 0.24405638544077424,
"grad_norm": 211.6590576171875,
"learning_rate": 4.943244142348177e-06,
"loss": 0.872,
"step": 290
},
{
"epoch": 0.24489795918367346,
"grad_norm": 155.99195861816406,
"learning_rate": 4.942770028520055e-06,
"loss": 0.82,
"step": 291
},
{
"epoch": 0.2457395329265727,
"grad_norm": 186.73419189453125,
"learning_rate": 4.942293965576489e-06,
"loss": 1.0884,
"step": 292
},
{
"epoch": 0.24658110666947192,
"grad_norm": 246.95408630371094,
"learning_rate": 4.941815953897335e-06,
"loss": 1.0209,
"step": 293
},
{
"epoch": 0.24742268041237114,
"grad_norm": 255.74241638183594,
"learning_rate": 4.941335993864005e-06,
"loss": 1.0589,
"step": 294
},
{
"epoch": 0.24826425415527034,
"grad_norm": 240.6260528564453,
"learning_rate": 4.940854085859465e-06,
"loss": 1.2218,
"step": 295
},
{
"epoch": 0.24910582789816957,
"grad_norm": 248.5775146484375,
"learning_rate": 4.940370230268233e-06,
"loss": 0.9993,
"step": 296
},
{
"epoch": 0.2499474016410688,
"grad_norm": 217.07452392578125,
"learning_rate": 4.939884427476386e-06,
"loss": 1.1372,
"step": 297
},
{
"epoch": 0.250788975383968,
"grad_norm": 196.2444610595703,
"learning_rate": 4.9393966778715505e-06,
"loss": 0.7739,
"step": 298
},
{
"epoch": 0.2516305491268672,
"grad_norm": 138.59149169921875,
"learning_rate": 4.938906981842907e-06,
"loss": 0.8014,
"step": 299
},
{
"epoch": 0.25247212286976645,
"grad_norm": 152.51995849609375,
"learning_rate": 4.938415339781192e-06,
"loss": 1.0103,
"step": 300
},
{
"epoch": 0.2533136966126657,
"grad_norm": 136.28709411621094,
"learning_rate": 4.93792175207869e-06,
"loss": 0.7962,
"step": 301
},
{
"epoch": 0.2541552703555649,
"grad_norm": 171.7638397216797,
"learning_rate": 4.937426219129242e-06,
"loss": 1.0729,
"step": 302
},
{
"epoch": 0.2549968440984641,
"grad_norm": 307.9116516113281,
"learning_rate": 4.93692874132824e-06,
"loss": 1.0687,
"step": 303
},
{
"epoch": 0.25583841784136335,
"grad_norm": 135.71917724609375,
"learning_rate": 4.936429319072626e-06,
"loss": 0.7077,
"step": 304
},
{
"epoch": 0.2566799915842626,
"grad_norm": 220.42298889160156,
"learning_rate": 4.9359279527608975e-06,
"loss": 1.295,
"step": 305
},
{
"epoch": 0.2575215653271618,
"grad_norm": 143.9759063720703,
"learning_rate": 4.935424642793099e-06,
"loss": 0.9554,
"step": 306
},
{
"epoch": 0.25836313907006103,
"grad_norm": 155.3295440673828,
"learning_rate": 4.934919389570826e-06,
"loss": 0.8669,
"step": 307
},
{
"epoch": 0.25920471281296026,
"grad_norm": 173.58909606933594,
"learning_rate": 4.934412193497229e-06,
"loss": 0.9683,
"step": 308
},
{
"epoch": 0.2600462865558595,
"grad_norm": 164.60137939453125,
"learning_rate": 4.933903054977005e-06,
"loss": 1.0304,
"step": 309
},
{
"epoch": 0.26088786029875866,
"grad_norm": 148.02781677246094,
"learning_rate": 4.9333919744164005e-06,
"loss": 0.9517,
"step": 310
},
{
"epoch": 0.2617294340416579,
"grad_norm": 167.60630798339844,
"learning_rate": 4.932878952223213e-06,
"loss": 0.9157,
"step": 311
},
{
"epoch": 0.2625710077845571,
"grad_norm": 165.19789123535156,
"learning_rate": 4.932363988806791e-06,
"loss": 0.7926,
"step": 312
},
{
"epoch": 0.26341258152745634,
"grad_norm": 179.5117645263672,
"learning_rate": 4.931847084578028e-06,
"loss": 0.9573,
"step": 313
},
{
"epoch": 0.26425415527035556,
"grad_norm": 199.1255645751953,
"learning_rate": 4.931328239949369e-06,
"loss": 1.1345,
"step": 314
},
{
"epoch": 0.2650957290132548,
"grad_norm": 192.1922607421875,
"learning_rate": 4.930807455334807e-06,
"loss": 1.0578,
"step": 315
},
{
"epoch": 0.265937302756154,
"grad_norm": 252.30825805664062,
"learning_rate": 4.930284731149879e-06,
"loss": 1.1559,
"step": 316
},
{
"epoch": 0.26677887649905324,
"grad_norm": 189.8568572998047,
"learning_rate": 4.929760067811676e-06,
"loss": 1.211,
"step": 317
},
{
"epoch": 0.26762045024195247,
"grad_norm": 184.5948486328125,
"learning_rate": 4.929233465738833e-06,
"loss": 0.9276,
"step": 318
},
{
"epoch": 0.2684620239848517,
"grad_norm": 270.45855712890625,
"learning_rate": 4.928704925351532e-06,
"loss": 1.0734,
"step": 319
},
{
"epoch": 0.2693035977277509,
"grad_norm": 224.6494598388672,
"learning_rate": 4.9281744470715e-06,
"loss": 0.9032,
"step": 320
},
{
"epoch": 0.2701451714706501,
"grad_norm": 141.63299560546875,
"learning_rate": 4.927642031322013e-06,
"loss": 0.7141,
"step": 321
},
{
"epoch": 0.2709867452135493,
"grad_norm": 178.1015625,
"learning_rate": 4.927107678527893e-06,
"loss": 0.7973,
"step": 322
},
{
"epoch": 0.27182831895644854,
"grad_norm": 155.43727111816406,
"learning_rate": 4.9265713891155055e-06,
"loss": 0.9052,
"step": 323
},
{
"epoch": 0.27266989269934777,
"grad_norm": 140.11785888671875,
"learning_rate": 4.926033163512761e-06,
"loss": 0.7327,
"step": 324
},
{
"epoch": 0.273511466442247,
"grad_norm": 153.34718322753906,
"learning_rate": 4.925493002149119e-06,
"loss": 0.8394,
"step": 325
},
{
"epoch": 0.2743530401851462,
"grad_norm": 191.73004150390625,
"learning_rate": 4.924950905455578e-06,
"loss": 0.9868,
"step": 326
},
{
"epoch": 0.27519461392804545,
"grad_norm": 150.3389129638672,
"learning_rate": 4.924406873864685e-06,
"loss": 0.6314,
"step": 327
},
{
"epoch": 0.2760361876709447,
"grad_norm": 200.07156372070312,
"learning_rate": 4.9238609078105285e-06,
"loss": 0.889,
"step": 328
},
{
"epoch": 0.2768777614138439,
"grad_norm": 133.3046417236328,
"learning_rate": 4.9233130077287425e-06,
"loss": 0.7299,
"step": 329
},
{
"epoch": 0.27771933515674313,
"grad_norm": 159.61390686035156,
"learning_rate": 4.9227631740565015e-06,
"loss": 0.9479,
"step": 330
},
{
"epoch": 0.27856090889964236,
"grad_norm": 148.41209411621094,
"learning_rate": 4.922211407232525e-06,
"loss": 1.0882,
"step": 331
},
{
"epoch": 0.2794024826425415,
"grad_norm": 167.6956329345703,
"learning_rate": 4.921657707697077e-06,
"loss": 1.0162,
"step": 332
},
{
"epoch": 0.28024405638544075,
"grad_norm": 200.65501403808594,
"learning_rate": 4.9211020758919555e-06,
"loss": 1.0155,
"step": 333
},
{
"epoch": 0.28108563012834,
"grad_norm": 168.57797241210938,
"learning_rate": 4.9205445122605095e-06,
"loss": 0.9225,
"step": 334
},
{
"epoch": 0.2819272038712392,
"grad_norm": 148.2366943359375,
"learning_rate": 4.919985017247625e-06,
"loss": 0.8756,
"step": 335
},
{
"epoch": 0.28276877761413843,
"grad_norm": 215.92623901367188,
"learning_rate": 4.91942359129973e-06,
"loss": 1.1371,
"step": 336
},
{
"epoch": 0.28361035135703766,
"grad_norm": 183.8272247314453,
"learning_rate": 4.918860234864793e-06,
"loss": 0.9762,
"step": 337
},
{
"epoch": 0.2844519250999369,
"grad_norm": 213.24313354492188,
"learning_rate": 4.918294948392322e-06,
"loss": 1.0352,
"step": 338
},
{
"epoch": 0.2852934988428361,
"grad_norm": 162.19821166992188,
"learning_rate": 4.9177277323333665e-06,
"loss": 1.1227,
"step": 339
},
{
"epoch": 0.28613507258573534,
"grad_norm": 174.2341766357422,
"learning_rate": 4.917158587140515e-06,
"loss": 0.9655,
"step": 340
},
{
"epoch": 0.28697664632863457,
"grad_norm": 147.44154357910156,
"learning_rate": 4.916587513267895e-06,
"loss": 0.8849,
"step": 341
},
{
"epoch": 0.2878182200715338,
"grad_norm": 162.8009033203125,
"learning_rate": 4.916014511171173e-06,
"loss": 0.9787,
"step": 342
},
{
"epoch": 0.28865979381443296,
"grad_norm": 192.03756713867188,
"learning_rate": 4.915439581307555e-06,
"loss": 0.8299,
"step": 343
},
{
"epoch": 0.2895013675573322,
"grad_norm": 239.55003356933594,
"learning_rate": 4.914862724135784e-06,
"loss": 0.8902,
"step": 344
},
{
"epoch": 0.2903429413002314,
"grad_norm": 143.41920471191406,
"learning_rate": 4.91428394011614e-06,
"loss": 0.8735,
"step": 345
},
{
"epoch": 0.29118451504313064,
"grad_norm": 151.96649169921875,
"learning_rate": 4.9137032297104445e-06,
"loss": 1.0496,
"step": 346
},
{
"epoch": 0.29202608878602987,
"grad_norm": 143.7827606201172,
"learning_rate": 4.913120593382051e-06,
"loss": 0.9842,
"step": 347
},
{
"epoch": 0.2928676625289291,
"grad_norm": 214.41659545898438,
"learning_rate": 4.912536031595853e-06,
"loss": 1.0993,
"step": 348
},
{
"epoch": 0.2937092362718283,
"grad_norm": 192.0199737548828,
"learning_rate": 4.911949544818278e-06,
"loss": 0.9345,
"step": 349
},
{
"epoch": 0.29455081001472755,
"grad_norm": 140.22434997558594,
"learning_rate": 4.911361133517293e-06,
"loss": 0.7077,
"step": 350
},
{
"epoch": 0.2953923837576268,
"grad_norm": 149.80943298339844,
"learning_rate": 4.910770798162396e-06,
"loss": 1.0209,
"step": 351
},
{
"epoch": 0.296233957500526,
"grad_norm": 188.54193115234375,
"learning_rate": 4.910178539224626e-06,
"loss": 0.7282,
"step": 352
},
{
"epoch": 0.29707553124342523,
"grad_norm": 182.12388610839844,
"learning_rate": 4.90958435717655e-06,
"loss": 1.1027,
"step": 353
},
{
"epoch": 0.2979171049863244,
"grad_norm": 155.85667419433594,
"learning_rate": 4.908988252492274e-06,
"loss": 1.1697,
"step": 354
},
{
"epoch": 0.2987586787292236,
"grad_norm": 174.2142791748047,
"learning_rate": 4.908390225647438e-06,
"loss": 1.1215,
"step": 355
},
{
"epoch": 0.29960025247212285,
"grad_norm": 191.0850830078125,
"learning_rate": 4.907790277119214e-06,
"loss": 1.1317,
"step": 356
},
{
"epoch": 0.3004418262150221,
"grad_norm": 166.85194396972656,
"learning_rate": 4.907188407386308e-06,
"loss": 0.9083,
"step": 357
},
{
"epoch": 0.3012833999579213,
"grad_norm": 115.8208236694336,
"learning_rate": 4.90658461692896e-06,
"loss": 0.8165,
"step": 358
},
{
"epoch": 0.30212497370082053,
"grad_norm": 136.56874084472656,
"learning_rate": 4.905978906228942e-06,
"loss": 0.7891,
"step": 359
},
{
"epoch": 0.30296654744371976,
"grad_norm": 153.9842071533203,
"learning_rate": 4.905371275769555e-06,
"loss": 1.1678,
"step": 360
},
{
"epoch": 0.303808121186619,
"grad_norm": 183.4450225830078,
"learning_rate": 4.9047617260356375e-06,
"loss": 0.9229,
"step": 361
},
{
"epoch": 0.3046496949295182,
"grad_norm": 132.87869262695312,
"learning_rate": 4.904150257513555e-06,
"loss": 0.8403,
"step": 362
},
{
"epoch": 0.30549126867241744,
"grad_norm": 143.56069946289062,
"learning_rate": 4.903536870691206e-06,
"loss": 0.9998,
"step": 363
},
{
"epoch": 0.30633284241531666,
"grad_norm": 151.4114532470703,
"learning_rate": 4.90292156605802e-06,
"loss": 1.0212,
"step": 364
},
{
"epoch": 0.3071744161582159,
"grad_norm": 139.41395568847656,
"learning_rate": 4.902304344104954e-06,
"loss": 1.0311,
"step": 365
},
{
"epoch": 0.30801598990111506,
"grad_norm": 155.7967071533203,
"learning_rate": 4.901685205324499e-06,
"loss": 0.9561,
"step": 366
},
{
"epoch": 0.3088575636440143,
"grad_norm": 153.0625457763672,
"learning_rate": 4.901064150210672e-06,
"loss": 0.984,
"step": 367
},
{
"epoch": 0.3096991373869135,
"grad_norm": 204.0398406982422,
"learning_rate": 4.900441179259019e-06,
"loss": 1.152,
"step": 368
},
{
"epoch": 0.31054071112981274,
"grad_norm": 178.45762634277344,
"learning_rate": 4.899816292966619e-06,
"loss": 0.9165,
"step": 369
},
{
"epoch": 0.31138228487271197,
"grad_norm": 154.21519470214844,
"learning_rate": 4.899189491832075e-06,
"loss": 1.1589,
"step": 370
},
{
"epoch": 0.3122238586156112,
"grad_norm": 152.87892150878906,
"learning_rate": 4.8985607763555175e-06,
"loss": 0.9606,
"step": 371
},
{
"epoch": 0.3130654323585104,
"grad_norm": 125.60699462890625,
"learning_rate": 4.897930147038609e-06,
"loss": 0.9289,
"step": 372
},
{
"epoch": 0.31390700610140965,
"grad_norm": 141.3811492919922,
"learning_rate": 4.897297604384533e-06,
"loss": 0.7742,
"step": 373
},
{
"epoch": 0.3147485798443089,
"grad_norm": 123.92646789550781,
"learning_rate": 4.8966631488980055e-06,
"loss": 0.8978,
"step": 374
},
{
"epoch": 0.3155901535872081,
"grad_norm": 147.79969787597656,
"learning_rate": 4.896026781085265e-06,
"loss": 1.042,
"step": 375
},
{
"epoch": 0.3164317273301073,
"grad_norm": 274.53399658203125,
"learning_rate": 4.895388501454077e-06,
"loss": 1.0025,
"step": 376
},
{
"epoch": 0.3172733010730065,
"grad_norm": 154.58265686035156,
"learning_rate": 4.894748310513733e-06,
"loss": 0.99,
"step": 377
},
{
"epoch": 0.3181148748159057,
"grad_norm": 110.33751678466797,
"learning_rate": 4.894106208775048e-06,
"loss": 0.8535,
"step": 378
},
{
"epoch": 0.31895644855880495,
"grad_norm": 118.87710571289062,
"learning_rate": 4.893462196750362e-06,
"loss": 0.9556,
"step": 379
},
{
"epoch": 0.3197980223017042,
"grad_norm": 119.17797088623047,
"learning_rate": 4.8928162749535435e-06,
"loss": 0.8088,
"step": 380
},
{
"epoch": 0.3206395960446034,
"grad_norm": 96.09217834472656,
"learning_rate": 4.892168443899978e-06,
"loss": 0.7595,
"step": 381
},
{
"epoch": 0.32148116978750263,
"grad_norm": 127.31585693359375,
"learning_rate": 4.8915187041065785e-06,
"loss": 0.8242,
"step": 382
},
{
"epoch": 0.32232274353040186,
"grad_norm": 125.22929382324219,
"learning_rate": 4.89086705609178e-06,
"loss": 0.8291,
"step": 383
},
{
"epoch": 0.3231643172733011,
"grad_norm": 144.8851776123047,
"learning_rate": 4.89021350037554e-06,
"loss": 1.1254,
"step": 384
},
{
"epoch": 0.3240058910162003,
"grad_norm": 121.80813598632812,
"learning_rate": 4.88955803747934e-06,
"loss": 0.7422,
"step": 385
},
{
"epoch": 0.32484746475909954,
"grad_norm": 159.96646118164062,
"learning_rate": 4.88890066792618e-06,
"loss": 1.0759,
"step": 386
},
{
"epoch": 0.32568903850199876,
"grad_norm": 114.78094482421875,
"learning_rate": 4.888241392240584e-06,
"loss": 0.563,
"step": 387
},
{
"epoch": 0.32653061224489793,
"grad_norm": 140.37147521972656,
"learning_rate": 4.887580210948595e-06,
"loss": 0.7917,
"step": 388
},
{
"epoch": 0.32737218598779716,
"grad_norm": 143.88070678710938,
"learning_rate": 4.886917124577778e-06,
"loss": 1.0749,
"step": 389
},
{
"epoch": 0.3282137597306964,
"grad_norm": 142.34536743164062,
"learning_rate": 4.886252133657218e-06,
"loss": 0.9067,
"step": 390
},
{
"epoch": 0.3290553334735956,
"grad_norm": 157.0981903076172,
"learning_rate": 4.8855852387175185e-06,
"loss": 1.1051,
"step": 391
},
{
"epoch": 0.32989690721649484,
"grad_norm": 272.75946044921875,
"learning_rate": 4.8849164402908024e-06,
"loss": 0.8886,
"step": 392
},
{
"epoch": 0.33073848095939407,
"grad_norm": 132.60775756835938,
"learning_rate": 4.884245738910714e-06,
"loss": 0.8409,
"step": 393
},
{
"epoch": 0.3315800547022933,
"grad_norm": 127.93871307373047,
"learning_rate": 4.883573135112413e-06,
"loss": 0.7745,
"step": 394
},
{
"epoch": 0.3324216284451925,
"grad_norm": 119.92588806152344,
"learning_rate": 4.882898629432577e-06,
"loss": 0.9234,
"step": 395
},
{
"epoch": 0.33326320218809175,
"grad_norm": 149.8199005126953,
"learning_rate": 4.882222222409403e-06,
"loss": 1.017,
"step": 396
},
{
"epoch": 0.33410477593099097,
"grad_norm": 141.49534606933594,
"learning_rate": 4.881543914582604e-06,
"loss": 0.7122,
"step": 397
},
{
"epoch": 0.3349463496738902,
"grad_norm": 118.47855377197266,
"learning_rate": 4.880863706493411e-06,
"loss": 0.928,
"step": 398
},
{
"epoch": 0.33578792341678937,
"grad_norm": 123.57112884521484,
"learning_rate": 4.8801815986845685e-06,
"loss": 0.8434,
"step": 399
},
{
"epoch": 0.3366294971596886,
"grad_norm": 139.73097229003906,
"learning_rate": 4.879497591700339e-06,
"loss": 0.7714,
"step": 400
},
{
"epoch": 0.3374710709025878,
"grad_norm": 138.10569763183594,
"learning_rate": 4.878811686086502e-06,
"loss": 0.8897,
"step": 401
},
{
"epoch": 0.33831264464548705,
"grad_norm": 157.8883514404297,
"learning_rate": 4.878123882390347e-06,
"loss": 1.1479,
"step": 402
},
{
"epoch": 0.3391542183883863,
"grad_norm": 138.03404235839844,
"learning_rate": 4.877434181160681e-06,
"loss": 0.9142,
"step": 403
},
{
"epoch": 0.3399957921312855,
"grad_norm": 142.91648864746094,
"learning_rate": 4.876742582947826e-06,
"loss": 0.8765,
"step": 404
},
{
"epoch": 0.34083736587418473,
"grad_norm": 140.9148406982422,
"learning_rate": 4.876049088303616e-06,
"loss": 1.0419,
"step": 405
},
{
"epoch": 0.34167893961708395,
"grad_norm": 127.75152587890625,
"learning_rate": 4.8753536977814e-06,
"loss": 0.839,
"step": 406
},
{
"epoch": 0.3425205133599832,
"grad_norm": 137.12606811523438,
"learning_rate": 4.874656411936035e-06,
"loss": 0.8453,
"step": 407
},
{
"epoch": 0.3433620871028824,
"grad_norm": 158.32156372070312,
"learning_rate": 4.8739572313238984e-06,
"loss": 1.0524,
"step": 408
},
{
"epoch": 0.34420366084578163,
"grad_norm": 142.4107208251953,
"learning_rate": 4.873256156502871e-06,
"loss": 0.9642,
"step": 409
},
{
"epoch": 0.34504523458868086,
"grad_norm": 165.7876739501953,
"learning_rate": 4.87255318803235e-06,
"loss": 0.8204,
"step": 410
},
{
"epoch": 0.34588680833158003,
"grad_norm": 134.32568359375,
"learning_rate": 4.871848326473243e-06,
"loss": 0.8822,
"step": 411
},
{
"epoch": 0.34672838207447926,
"grad_norm": 148.0491180419922,
"learning_rate": 4.871141572387966e-06,
"loss": 0.8069,
"step": 412
},
{
"epoch": 0.3475699558173785,
"grad_norm": 159.5659637451172,
"learning_rate": 4.870432926340448e-06,
"loss": 1.1664,
"step": 413
},
{
"epoch": 0.3484115295602777,
"grad_norm": 170.69534301757812,
"learning_rate": 4.869722388896127e-06,
"loss": 1.2899,
"step": 414
},
{
"epoch": 0.34925310330317694,
"grad_norm": 156.73345947265625,
"learning_rate": 4.869009960621947e-06,
"loss": 0.9722,
"step": 415
},
{
"epoch": 0.35009467704607616,
"grad_norm": 113.5978775024414,
"learning_rate": 4.868295642086363e-06,
"loss": 0.9496,
"step": 416
},
{
"epoch": 0.3509362507889754,
"grad_norm": 97.61703491210938,
"learning_rate": 4.86757943385934e-06,
"loss": 0.6348,
"step": 417
},
{
"epoch": 0.3517778245318746,
"grad_norm": 170.4888916015625,
"learning_rate": 4.8668613365123485e-06,
"loss": 0.8473,
"step": 418
},
{
"epoch": 0.35261939827477384,
"grad_norm": 122.68419647216797,
"learning_rate": 4.866141350618367e-06,
"loss": 0.9248,
"step": 419
},
{
"epoch": 0.35346097201767307,
"grad_norm": 145.2256622314453,
"learning_rate": 4.8654194767518815e-06,
"loss": 0.9411,
"step": 420
},
{
"epoch": 0.3543025457605723,
"grad_norm": 123.21392059326172,
"learning_rate": 4.864695715488882e-06,
"loss": 0.9264,
"step": 421
},
{
"epoch": 0.35514411950347147,
"grad_norm": 127.80828857421875,
"learning_rate": 4.863970067406868e-06,
"loss": 0.9816,
"step": 422
},
{
"epoch": 0.3559856932463707,
"grad_norm": 108.34115600585938,
"learning_rate": 4.863242533084842e-06,
"loss": 0.9395,
"step": 423
},
{
"epoch": 0.3568272669892699,
"grad_norm": 106.89410400390625,
"learning_rate": 4.862513113103312e-06,
"loss": 0.8584,
"step": 424
},
{
"epoch": 0.35766884073216915,
"grad_norm": 110.82831573486328,
"learning_rate": 4.861781808044292e-06,
"loss": 0.7783,
"step": 425
},
{
"epoch": 0.3585104144750684,
"grad_norm": 170.09359741210938,
"learning_rate": 4.861048618491298e-06,
"loss": 0.9577,
"step": 426
},
{
"epoch": 0.3593519882179676,
"grad_norm": 126.96100616455078,
"learning_rate": 4.8603135450293516e-06,
"loss": 0.9036,
"step": 427
},
{
"epoch": 0.3601935619608668,
"grad_norm": 165.4973907470703,
"learning_rate": 4.8595765882449766e-06,
"loss": 1.0496,
"step": 428
},
{
"epoch": 0.36103513570376605,
"grad_norm": 126.7000503540039,
"learning_rate": 4.858837748726199e-06,
"loss": 0.9302,
"step": 429
},
{
"epoch": 0.3618767094466653,
"grad_norm": 154.56597900390625,
"learning_rate": 4.858097027062548e-06,
"loss": 0.9942,
"step": 430
},
{
"epoch": 0.3627182831895645,
"grad_norm": 106.47966766357422,
"learning_rate": 4.857354423845054e-06,
"loss": 0.7642,
"step": 431
},
{
"epoch": 0.36355985693246373,
"grad_norm": 111.3463363647461,
"learning_rate": 4.856609939666251e-06,
"loss": 0.8334,
"step": 432
},
{
"epoch": 0.3644014306753629,
"grad_norm": 113.79244995117188,
"learning_rate": 4.855863575120168e-06,
"loss": 0.9112,
"step": 433
},
{
"epoch": 0.36524300441826213,
"grad_norm": 123.11966705322266,
"learning_rate": 4.855115330802342e-06,
"loss": 1.0529,
"step": 434
},
{
"epoch": 0.36608457816116136,
"grad_norm": 108.89775085449219,
"learning_rate": 4.854365207309803e-06,
"loss": 0.7978,
"step": 435
},
{
"epoch": 0.3669261519040606,
"grad_norm": 164.04800415039062,
"learning_rate": 4.853613205241086e-06,
"loss": 0.9432,
"step": 436
},
{
"epoch": 0.3677677256469598,
"grad_norm": 136.918701171875,
"learning_rate": 4.8528593251962206e-06,
"loss": 0.982,
"step": 437
},
{
"epoch": 0.36860929938985904,
"grad_norm": 93.6369400024414,
"learning_rate": 4.852103567776738e-06,
"loss": 0.6342,
"step": 438
},
{
"epoch": 0.36945087313275826,
"grad_norm": 161.514892578125,
"learning_rate": 4.851345933585666e-06,
"loss": 1.0082,
"step": 439
},
{
"epoch": 0.3702924468756575,
"grad_norm": 120.8597183227539,
"learning_rate": 4.850586423227529e-06,
"loss": 1.0075,
"step": 440
},
{
"epoch": 0.3711340206185567,
"grad_norm": 146.28897094726562,
"learning_rate": 4.849825037308351e-06,
"loss": 0.8819,
"step": 441
},
{
"epoch": 0.37197559436145594,
"grad_norm": 133.81219482421875,
"learning_rate": 4.849061776435649e-06,
"loss": 0.7981,
"step": 442
},
{
"epoch": 0.37281716810435517,
"grad_norm": 126.19327545166016,
"learning_rate": 4.8482966412184394e-06,
"loss": 0.9471,
"step": 443
},
{
"epoch": 0.37365874184725434,
"grad_norm": 128.17840576171875,
"learning_rate": 4.847529632267233e-06,
"loss": 0.7527,
"step": 444
},
{
"epoch": 0.37450031559015357,
"grad_norm": 122.65428161621094,
"learning_rate": 4.846760750194033e-06,
"loss": 0.93,
"step": 445
},
{
"epoch": 0.3753418893330528,
"grad_norm": 127.0604019165039,
"learning_rate": 4.845989995612342e-06,
"loss": 0.8162,
"step": 446
},
{
"epoch": 0.376183463075952,
"grad_norm": 141.19271850585938,
"learning_rate": 4.8452173691371545e-06,
"loss": 0.9245,
"step": 447
},
{
"epoch": 0.37702503681885124,
"grad_norm": 145.30938720703125,
"learning_rate": 4.844442871384957e-06,
"loss": 1.1939,
"step": 448
},
{
"epoch": 0.37786661056175047,
"grad_norm": 95.39657592773438,
"learning_rate": 4.843666502973731e-06,
"loss": 0.8952,
"step": 449
},
{
"epoch": 0.3787081843046497,
"grad_norm": 143.34825134277344,
"learning_rate": 4.84288826452295e-06,
"loss": 1.1639,
"step": 450
},
{
"epoch": 0.3795497580475489,
"grad_norm": 110.71271514892578,
"learning_rate": 4.842108156653581e-06,
"loss": 0.8824,
"step": 451
},
{
"epoch": 0.38039133179044815,
"grad_norm": 100.15702819824219,
"learning_rate": 4.84132617998808e-06,
"loss": 0.7559,
"step": 452
},
{
"epoch": 0.3812329055333474,
"grad_norm": 108.70226287841797,
"learning_rate": 4.840542335150395e-06,
"loss": 0.9015,
"step": 453
},
{
"epoch": 0.3820744792762466,
"grad_norm": 127.7223892211914,
"learning_rate": 4.839756622765966e-06,
"loss": 0.8032,
"step": 454
},
{
"epoch": 0.38291605301914583,
"grad_norm": 183.36643981933594,
"learning_rate": 4.838969043461725e-06,
"loss": 1.3108,
"step": 455
},
{
"epoch": 0.383757626762045,
"grad_norm": 93.0749740600586,
"learning_rate": 4.838179597866085e-06,
"loss": 0.6992,
"step": 456
},
{
"epoch": 0.3845992005049442,
"grad_norm": 218.0874481201172,
"learning_rate": 4.83738828660896e-06,
"loss": 1.1059,
"step": 457
},
{
"epoch": 0.38544077424784345,
"grad_norm": 129.31700134277344,
"learning_rate": 4.836595110321744e-06,
"loss": 0.686,
"step": 458
},
{
"epoch": 0.3862823479907427,
"grad_norm": 113.3348388671875,
"learning_rate": 4.835800069637321e-06,
"loss": 0.7151,
"step": 459
},
{
"epoch": 0.3871239217336419,
"grad_norm": 242.91207885742188,
"learning_rate": 4.835003165190064e-06,
"loss": 1.1047,
"step": 460
},
{
"epoch": 0.38796549547654113,
"grad_norm": 107.70856475830078,
"learning_rate": 4.8342043976158344e-06,
"loss": 0.8914,
"step": 461
},
{
"epoch": 0.38880706921944036,
"grad_norm": 102.79827117919922,
"learning_rate": 4.8334037675519775e-06,
"loss": 0.8817,
"step": 462
},
{
"epoch": 0.3896486429623396,
"grad_norm": 87.40475463867188,
"learning_rate": 4.832601275637324e-06,
"loss": 0.606,
"step": 463
},
{
"epoch": 0.3904902167052388,
"grad_norm": 118.46971893310547,
"learning_rate": 4.831796922512194e-06,
"loss": 1.0383,
"step": 464
},
{
"epoch": 0.39133179044813804,
"grad_norm": 122.6368179321289,
"learning_rate": 4.8309907088183894e-06,
"loss": 0.9457,
"step": 465
},
{
"epoch": 0.39217336419103727,
"grad_norm": 123.12105560302734,
"learning_rate": 4.830182635199197e-06,
"loss": 0.9501,
"step": 466
},
{
"epoch": 0.39301493793393644,
"grad_norm": 168.8681640625,
"learning_rate": 4.829372702299391e-06,
"loss": 0.7714,
"step": 467
},
{
"epoch": 0.39385651167683566,
"grad_norm": 128.96510314941406,
"learning_rate": 4.828560910765223e-06,
"loss": 1.0639,
"step": 468
},
{
"epoch": 0.3946980854197349,
"grad_norm": 130.435302734375,
"learning_rate": 4.827747261244434e-06,
"loss": 0.9294,
"step": 469
},
{
"epoch": 0.3955396591626341,
"grad_norm": 116.36153411865234,
"learning_rate": 4.826931754386244e-06,
"loss": 0.9145,
"step": 470
},
{
"epoch": 0.39638123290553334,
"grad_norm": 170.28231811523438,
"learning_rate": 4.8261143908413554e-06,
"loss": 1.1118,
"step": 471
},
{
"epoch": 0.39722280664843257,
"grad_norm": 140.88491821289062,
"learning_rate": 4.825295171261952e-06,
"loss": 0.8239,
"step": 472
},
{
"epoch": 0.3980643803913318,
"grad_norm": 150.85214233398438,
"learning_rate": 4.8244740963017e-06,
"loss": 0.8503,
"step": 473
},
{
"epoch": 0.398905954134231,
"grad_norm": 108.1480712890625,
"learning_rate": 4.823651166615742e-06,
"loss": 1.0041,
"step": 474
},
{
"epoch": 0.39974752787713025,
"grad_norm": 319.3303527832031,
"learning_rate": 4.822826382860708e-06,
"loss": 1.003,
"step": 475
},
{
"epoch": 0.4005891016200295,
"grad_norm": 123.97578430175781,
"learning_rate": 4.8219997456947e-06,
"loss": 0.8949,
"step": 476
},
{
"epoch": 0.4014306753629287,
"grad_norm": 209.10208129882812,
"learning_rate": 4.821171255777302e-06,
"loss": 1.0194,
"step": 477
},
{
"epoch": 0.4022722491058279,
"grad_norm": 141.95541381835938,
"learning_rate": 4.820340913769577e-06,
"loss": 0.9419,
"step": 478
},
{
"epoch": 0.4031138228487271,
"grad_norm": 140.51358032226562,
"learning_rate": 4.819508720334063e-06,
"loss": 1.1065,
"step": 479
},
{
"epoch": 0.4039553965916263,
"grad_norm": 97.90631103515625,
"learning_rate": 4.818674676134778e-06,
"loss": 0.919,
"step": 480
},
{
"epoch": 0.40479697033452555,
"grad_norm": 151.83956909179688,
"learning_rate": 4.8178387818372165e-06,
"loss": 1.0386,
"step": 481
},
{
"epoch": 0.4056385440774248,
"grad_norm": 112.4679946899414,
"learning_rate": 4.817001038108349e-06,
"loss": 0.7155,
"step": 482
},
{
"epoch": 0.406480117820324,
"grad_norm": 110.14672088623047,
"learning_rate": 4.816161445616618e-06,
"loss": 0.8235,
"step": 483
},
{
"epoch": 0.40732169156322323,
"grad_norm": 166.98773193359375,
"learning_rate": 4.815320005031949e-06,
"loss": 1.3208,
"step": 484
},
{
"epoch": 0.40816326530612246,
"grad_norm": 116.70498657226562,
"learning_rate": 4.814476717025734e-06,
"loss": 0.9825,
"step": 485
},
{
"epoch": 0.4090048390490217,
"grad_norm": 114.12043762207031,
"learning_rate": 4.813631582270844e-06,
"loss": 0.9469,
"step": 486
},
{
"epoch": 0.4098464127919209,
"grad_norm": 118.60401916503906,
"learning_rate": 4.812784601441622e-06,
"loss": 0.8652,
"step": 487
},
{
"epoch": 0.41068798653482014,
"grad_norm": 142.69345092773438,
"learning_rate": 4.811935775213884e-06,
"loss": 0.7706,
"step": 488
},
{
"epoch": 0.4115295602777193,
"grad_norm": 129.11827087402344,
"learning_rate": 4.811085104264919e-06,
"loss": 0.986,
"step": 489
},
{
"epoch": 0.41237113402061853,
"grad_norm": 126.18577575683594,
"learning_rate": 4.810232589273487e-06,
"loss": 0.9475,
"step": 490
},
{
"epoch": 0.41321270776351776,
"grad_norm": 120.70259857177734,
"learning_rate": 4.80937823091982e-06,
"loss": 0.9647,
"step": 491
},
{
"epoch": 0.414054281506417,
"grad_norm": 113.89292907714844,
"learning_rate": 4.808522029885622e-06,
"loss": 0.889,
"step": 492
},
{
"epoch": 0.4148958552493162,
"grad_norm": 97.43743896484375,
"learning_rate": 4.8076639868540635e-06,
"loss": 0.7151,
"step": 493
},
{
"epoch": 0.41573742899221544,
"grad_norm": 105.58618927001953,
"learning_rate": 4.8068041025097895e-06,
"loss": 0.9715,
"step": 494
},
{
"epoch": 0.41657900273511467,
"grad_norm": 127.87680053710938,
"learning_rate": 4.805942377538911e-06,
"loss": 0.8187,
"step": 495
},
{
"epoch": 0.4174205764780139,
"grad_norm": 134.050048828125,
"learning_rate": 4.805078812629009e-06,
"loss": 0.8815,
"step": 496
},
{
"epoch": 0.4182621502209131,
"grad_norm": 108.08966064453125,
"learning_rate": 4.804213408469133e-06,
"loss": 0.8102,
"step": 497
},
{
"epoch": 0.41910372396381235,
"grad_norm": 129.20497131347656,
"learning_rate": 4.803346165749798e-06,
"loss": 0.6703,
"step": 498
},
{
"epoch": 0.4199452977067116,
"grad_norm": 146.5807647705078,
"learning_rate": 4.802477085162988e-06,
"loss": 0.857,
"step": 499
},
{
"epoch": 0.4207868714496108,
"grad_norm": 121.63455200195312,
"learning_rate": 4.801606167402152e-06,
"loss": 0.9737,
"step": 500
}
],
"logging_steps": 1,
"max_steps": 3567,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 2.7603496702462464e+16,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}