sparky-decoder-5 / checkpoint-1500 /trainer_state.json
Dohahemdann's picture
Checkpoint checkpoint-1500
2134d91 verified
Raw
History Blame Contribute Delete
259 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.262571007784557,
"eval_steps": 500,
"global_step": 1500,
"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
},
{
"epoch": 0.42162844519250997,
"grad_norm": 112.13333892822266,
"learning_rate": 4.800733413162206e-06,
"loss": 0.8266,
"step": 501
},
{
"epoch": 0.4224700189354092,
"grad_norm": 105.02662658691406,
"learning_rate": 4.799858823139532e-06,
"loss": 0.7877,
"step": 502
},
{
"epoch": 0.4233115926783084,
"grad_norm": 100.8410415649414,
"learning_rate": 4.798982398031976e-06,
"loss": 0.8558,
"step": 503
},
{
"epoch": 0.42415316642120765,
"grad_norm": 141.06207275390625,
"learning_rate": 4.798104138538848e-06,
"loss": 1.1098,
"step": 504
},
{
"epoch": 0.4249947401641069,
"grad_norm": 631.0702514648438,
"learning_rate": 4.79722404536092e-06,
"loss": 1.371,
"step": 505
},
{
"epoch": 0.4258363139070061,
"grad_norm": 133.77012634277344,
"learning_rate": 4.79634211920043e-06,
"loss": 1.0773,
"step": 506
},
{
"epoch": 0.42667788764990533,
"grad_norm": 104.05528259277344,
"learning_rate": 4.795458360761078e-06,
"loss": 0.8034,
"step": 507
},
{
"epoch": 0.42751946139280456,
"grad_norm": 135.33978271484375,
"learning_rate": 4.794572770748025e-06,
"loss": 0.9243,
"step": 508
},
{
"epoch": 0.4283610351357038,
"grad_norm": 143.45217895507812,
"learning_rate": 4.793685349867893e-06,
"loss": 1.0566,
"step": 509
},
{
"epoch": 0.429202608878603,
"grad_norm": 136.93051147460938,
"learning_rate": 4.792796098828768e-06,
"loss": 0.8205,
"step": 510
},
{
"epoch": 0.43004418262150224,
"grad_norm": 124.12754821777344,
"learning_rate": 4.791905018340191e-06,
"loss": 0.7642,
"step": 511
},
{
"epoch": 0.4308857563644014,
"grad_norm": 119.36278533935547,
"learning_rate": 4.791012109113168e-06,
"loss": 0.8458,
"step": 512
},
{
"epoch": 0.43172733010730063,
"grad_norm": 123.41871643066406,
"learning_rate": 4.790117371860162e-06,
"loss": 0.9155,
"step": 513
},
{
"epoch": 0.43256890385019986,
"grad_norm": 99.3111572265625,
"learning_rate": 4.789220807295092e-06,
"loss": 0.8699,
"step": 514
},
{
"epoch": 0.4334104775930991,
"grad_norm": 115.54022979736328,
"learning_rate": 4.788322416133341e-06,
"loss": 0.8808,
"step": 515
},
{
"epoch": 0.4342520513359983,
"grad_norm": 108.34188079833984,
"learning_rate": 4.7874221990917445e-06,
"loss": 0.8901,
"step": 516
},
{
"epoch": 0.43509362507889754,
"grad_norm": 105.6187973022461,
"learning_rate": 4.786520156888596e-06,
"loss": 0.8886,
"step": 517
},
{
"epoch": 0.43593519882179677,
"grad_norm": 108.33097076416016,
"learning_rate": 4.7856162902436465e-06,
"loss": 0.9418,
"step": 518
},
{
"epoch": 0.436776772564696,
"grad_norm": 121.62516784667969,
"learning_rate": 4.784710599878102e-06,
"loss": 0.8255,
"step": 519
},
{
"epoch": 0.4376183463075952,
"grad_norm": 151.26263427734375,
"learning_rate": 4.7838030865146236e-06,
"loss": 1.1101,
"step": 520
},
{
"epoch": 0.43845992005049444,
"grad_norm": 125.10527801513672,
"learning_rate": 4.782893750877326e-06,
"loss": 1.101,
"step": 521
},
{
"epoch": 0.43930149379339367,
"grad_norm": 122.728271484375,
"learning_rate": 4.781982593691781e-06,
"loss": 0.933,
"step": 522
},
{
"epoch": 0.44014306753629284,
"grad_norm": 133.12278747558594,
"learning_rate": 4.78106961568501e-06,
"loss": 0.9782,
"step": 523
},
{
"epoch": 0.44098464127919207,
"grad_norm": 138.73318481445312,
"learning_rate": 4.78015481758549e-06,
"loss": 0.834,
"step": 524
},
{
"epoch": 0.4418262150220913,
"grad_norm": 128.3526153564453,
"learning_rate": 4.77923820012315e-06,
"loss": 0.9034,
"step": 525
},
{
"epoch": 0.4426677887649905,
"grad_norm": 108.43571472167969,
"learning_rate": 4.778319764029368e-06,
"loss": 0.8541,
"step": 526
},
{
"epoch": 0.44350936250788975,
"grad_norm": 116.94001007080078,
"learning_rate": 4.777399510036977e-06,
"loss": 0.958,
"step": 527
},
{
"epoch": 0.444350936250789,
"grad_norm": 120.42687225341797,
"learning_rate": 4.776477438880258e-06,
"loss": 1.0345,
"step": 528
},
{
"epoch": 0.4451925099936882,
"grad_norm": 111.29932403564453,
"learning_rate": 4.775553551294942e-06,
"loss": 0.8609,
"step": 529
},
{
"epoch": 0.44603408373658743,
"grad_norm": 131.2679901123047,
"learning_rate": 4.774627848018211e-06,
"loss": 1.1318,
"step": 530
},
{
"epoch": 0.44687565747948665,
"grad_norm": 127.53345489501953,
"learning_rate": 4.773700329788696e-06,
"loss": 0.9082,
"step": 531
},
{
"epoch": 0.4477172312223859,
"grad_norm": 124.14596557617188,
"learning_rate": 4.772770997346471e-06,
"loss": 1.0633,
"step": 532
},
{
"epoch": 0.4485588049652851,
"grad_norm": 97.10338592529297,
"learning_rate": 4.7718398514330655e-06,
"loss": 0.7285,
"step": 533
},
{
"epoch": 0.4494003787081843,
"grad_norm": 86.35016632080078,
"learning_rate": 4.770906892791449e-06,
"loss": 0.5934,
"step": 534
},
{
"epoch": 0.4502419524510835,
"grad_norm": 112.21371459960938,
"learning_rate": 4.7699721221660425e-06,
"loss": 1.0579,
"step": 535
},
{
"epoch": 0.45108352619398273,
"grad_norm": 106.98262023925781,
"learning_rate": 4.7690355403027086e-06,
"loss": 0.8364,
"step": 536
},
{
"epoch": 0.45192509993688196,
"grad_norm": 122.70458984375,
"learning_rate": 4.768097147948759e-06,
"loss": 0.8282,
"step": 537
},
{
"epoch": 0.4527666736797812,
"grad_norm": 115.28579711914062,
"learning_rate": 4.767156945852948e-06,
"loss": 0.7753,
"step": 538
},
{
"epoch": 0.4536082474226804,
"grad_norm": 131.470458984375,
"learning_rate": 4.7662149347654725e-06,
"loss": 1.0653,
"step": 539
},
{
"epoch": 0.45444982116557964,
"grad_norm": 119.49797821044922,
"learning_rate": 4.7652711154379765e-06,
"loss": 0.9991,
"step": 540
},
{
"epoch": 0.45529139490847886,
"grad_norm": 126.73249053955078,
"learning_rate": 4.764325488623544e-06,
"loss": 1.1871,
"step": 541
},
{
"epoch": 0.4561329686513781,
"grad_norm": 95.57186889648438,
"learning_rate": 4.7633780550767e-06,
"loss": 0.9152,
"step": 542
},
{
"epoch": 0.4569745423942773,
"grad_norm": 119.88944244384766,
"learning_rate": 4.762428815553416e-06,
"loss": 0.818,
"step": 543
},
{
"epoch": 0.45781611613717654,
"grad_norm": 105.79959106445312,
"learning_rate": 4.761477770811101e-06,
"loss": 0.9361,
"step": 544
},
{
"epoch": 0.4586576898800757,
"grad_norm": 119.57502746582031,
"learning_rate": 4.760524921608604e-06,
"loss": 0.9895,
"step": 545
},
{
"epoch": 0.45949926362297494,
"grad_norm": 114.09158325195312,
"learning_rate": 4.759570268706214e-06,
"loss": 1.1529,
"step": 546
},
{
"epoch": 0.46034083736587417,
"grad_norm": 100.0584487915039,
"learning_rate": 4.758613812865662e-06,
"loss": 0.7656,
"step": 547
},
{
"epoch": 0.4611824111087734,
"grad_norm": 100.03468322753906,
"learning_rate": 4.757655554850115e-06,
"loss": 0.7772,
"step": 548
},
{
"epoch": 0.4620239848516726,
"grad_norm": 114.32467651367188,
"learning_rate": 4.756695495424177e-06,
"loss": 0.8262,
"step": 549
},
{
"epoch": 0.46286555859457185,
"grad_norm": 93.30988311767578,
"learning_rate": 4.755733635353894e-06,
"loss": 0.8206,
"step": 550
},
{
"epoch": 0.4637071323374711,
"grad_norm": 90.66636657714844,
"learning_rate": 4.7547699754067425e-06,
"loss": 0.7546,
"step": 551
},
{
"epoch": 0.4645487060803703,
"grad_norm": 84.4310302734375,
"learning_rate": 4.75380451635164e-06,
"loss": 0.7591,
"step": 552
},
{
"epoch": 0.4653902798232695,
"grad_norm": 155.35433959960938,
"learning_rate": 4.752837258958937e-06,
"loss": 1.2789,
"step": 553
},
{
"epoch": 0.46623185356616875,
"grad_norm": 82.25304412841797,
"learning_rate": 4.751868204000421e-06,
"loss": 0.686,
"step": 554
},
{
"epoch": 0.467073427309068,
"grad_norm": 110.43043518066406,
"learning_rate": 4.750897352249311e-06,
"loss": 0.8744,
"step": 555
},
{
"epoch": 0.4679150010519672,
"grad_norm": 104.57023620605469,
"learning_rate": 4.7499247044802625e-06,
"loss": 0.784,
"step": 556
},
{
"epoch": 0.4687565747948664,
"grad_norm": 82.92748260498047,
"learning_rate": 4.748950261469362e-06,
"loss": 0.7746,
"step": 557
},
{
"epoch": 0.4695981485377656,
"grad_norm": 118.4659194946289,
"learning_rate": 4.747974023994129e-06,
"loss": 0.7537,
"step": 558
},
{
"epoch": 0.47043972228066483,
"grad_norm": 108.45428466796875,
"learning_rate": 4.746995992833516e-06,
"loss": 0.9114,
"step": 559
},
{
"epoch": 0.47128129602356406,
"grad_norm": 92.51221466064453,
"learning_rate": 4.746016168767906e-06,
"loss": 0.819,
"step": 560
},
{
"epoch": 0.4721228697664633,
"grad_norm": 89.94612884521484,
"learning_rate": 4.745034552579111e-06,
"loss": 0.8209,
"step": 561
},
{
"epoch": 0.4729644435093625,
"grad_norm": 96.07715606689453,
"learning_rate": 4.744051145050375e-06,
"loss": 0.7277,
"step": 562
},
{
"epoch": 0.47380601725226174,
"grad_norm": 115.94178771972656,
"learning_rate": 4.743065946966371e-06,
"loss": 0.8407,
"step": 563
},
{
"epoch": 0.47464759099516096,
"grad_norm": 90.5213394165039,
"learning_rate": 4.7420789591132e-06,
"loss": 0.7891,
"step": 564
},
{
"epoch": 0.4754891647380602,
"grad_norm": 113.68289947509766,
"learning_rate": 4.741090182278391e-06,
"loss": 1.0005,
"step": 565
},
{
"epoch": 0.4763307384809594,
"grad_norm": 136.67337036132812,
"learning_rate": 4.740099617250902e-06,
"loss": 1.1607,
"step": 566
},
{
"epoch": 0.47717231222385864,
"grad_norm": 91.25704193115234,
"learning_rate": 4.739107264821115e-06,
"loss": 0.6981,
"step": 567
},
{
"epoch": 0.4780138859667578,
"grad_norm": 106.67892456054688,
"learning_rate": 4.738113125780841e-06,
"loss": 0.7727,
"step": 568
},
{
"epoch": 0.47885545970965704,
"grad_norm": 121.09901428222656,
"learning_rate": 4.7371172009233155e-06,
"loss": 0.8397,
"step": 569
},
{
"epoch": 0.47969703345255627,
"grad_norm": 80.89665222167969,
"learning_rate": 4.736119491043199e-06,
"loss": 0.7499,
"step": 570
},
{
"epoch": 0.4805386071954555,
"grad_norm": 130.6654815673828,
"learning_rate": 4.735119996936576e-06,
"loss": 0.992,
"step": 571
},
{
"epoch": 0.4813801809383547,
"grad_norm": 101.22091674804688,
"learning_rate": 4.734118719400954e-06,
"loss": 0.9348,
"step": 572
},
{
"epoch": 0.48222175468125394,
"grad_norm": 93.3582534790039,
"learning_rate": 4.733115659235265e-06,
"loss": 0.8983,
"step": 573
},
{
"epoch": 0.48306332842415317,
"grad_norm": 100.23347473144531,
"learning_rate": 4.7321108172398625e-06,
"loss": 0.8489,
"step": 574
},
{
"epoch": 0.4839049021670524,
"grad_norm": 83.09318542480469,
"learning_rate": 4.731104194216523e-06,
"loss": 0.709,
"step": 575
},
{
"epoch": 0.4847464759099516,
"grad_norm": 76.45307159423828,
"learning_rate": 4.730095790968442e-06,
"loss": 0.7334,
"step": 576
},
{
"epoch": 0.48558804965285085,
"grad_norm": 117.21244812011719,
"learning_rate": 4.7290856083002355e-06,
"loss": 1.0757,
"step": 577
},
{
"epoch": 0.4864296233957501,
"grad_norm": 90.2735595703125,
"learning_rate": 4.728073647017942e-06,
"loss": 0.8145,
"step": 578
},
{
"epoch": 0.48727119713864925,
"grad_norm": 104.04305267333984,
"learning_rate": 4.727059907929018e-06,
"loss": 1.0265,
"step": 579
},
{
"epoch": 0.4881127708815485,
"grad_norm": 103.80632019042969,
"learning_rate": 4.726044391842337e-06,
"loss": 0.9707,
"step": 580
},
{
"epoch": 0.4889543446244477,
"grad_norm": 91.39692687988281,
"learning_rate": 4.72502709956819e-06,
"loss": 0.8324,
"step": 581
},
{
"epoch": 0.4897959183673469,
"grad_norm": 103.67784881591797,
"learning_rate": 4.7240080319182905e-06,
"loss": 0.9852,
"step": 582
},
{
"epoch": 0.49063749211024615,
"grad_norm": 92.56071472167969,
"learning_rate": 4.722987189705761e-06,
"loss": 0.7628,
"step": 583
},
{
"epoch": 0.4914790658531454,
"grad_norm": 92.0321273803711,
"learning_rate": 4.721964573745145e-06,
"loss": 0.9345,
"step": 584
},
{
"epoch": 0.4923206395960446,
"grad_norm": 109.95680236816406,
"learning_rate": 4.7209401848524015e-06,
"loss": 0.9331,
"step": 585
},
{
"epoch": 0.49316221333894383,
"grad_norm": 109.92249298095703,
"learning_rate": 4.7199140238449e-06,
"loss": 1.0453,
"step": 586
},
{
"epoch": 0.49400378708184306,
"grad_norm": 109.58256530761719,
"learning_rate": 4.718886091541429e-06,
"loss": 1.1754,
"step": 587
},
{
"epoch": 0.4948453608247423,
"grad_norm": 109.40169525146484,
"learning_rate": 4.717856388762187e-06,
"loss": 1.0512,
"step": 588
},
{
"epoch": 0.4956869345676415,
"grad_norm": 113.2249755859375,
"learning_rate": 4.716824916328786e-06,
"loss": 0.8632,
"step": 589
},
{
"epoch": 0.4965285083105407,
"grad_norm": 110.68663787841797,
"learning_rate": 4.715791675064251e-06,
"loss": 1.1118,
"step": 590
},
{
"epoch": 0.4973700820534399,
"grad_norm": 101.6324691772461,
"learning_rate": 4.714756665793015e-06,
"loss": 0.9311,
"step": 591
},
{
"epoch": 0.49821165579633914,
"grad_norm": 156.46829223632812,
"learning_rate": 4.713719889340929e-06,
"loss": 0.943,
"step": 592
},
{
"epoch": 0.49905322953923836,
"grad_norm": 106.6951904296875,
"learning_rate": 4.712681346535245e-06,
"loss": 0.5679,
"step": 593
},
{
"epoch": 0.4998948032821376,
"grad_norm": 114.27716827392578,
"learning_rate": 4.71164103820463e-06,
"loss": 0.8942,
"step": 594
},
{
"epoch": 0.5007363770250368,
"grad_norm": 88.36841583251953,
"learning_rate": 4.71059896517916e-06,
"loss": 0.8059,
"step": 595
},
{
"epoch": 0.501577950767936,
"grad_norm": 77.92396545410156,
"learning_rate": 4.709555128290316e-06,
"loss": 0.789,
"step": 596
},
{
"epoch": 0.5024195245108353,
"grad_norm": 92.30797576904297,
"learning_rate": 4.708509528370988e-06,
"loss": 0.794,
"step": 597
},
{
"epoch": 0.5032610982537344,
"grad_norm": 114.52457427978516,
"learning_rate": 4.707462166255473e-06,
"loss": 0.9309,
"step": 598
},
{
"epoch": 0.5041026719966337,
"grad_norm": 206.55442810058594,
"learning_rate": 4.7064130427794745e-06,
"loss": 1.8481,
"step": 599
},
{
"epoch": 0.5049442457395329,
"grad_norm": 82.25666809082031,
"learning_rate": 4.705362158780099e-06,
"loss": 0.7768,
"step": 600
},
{
"epoch": 0.5057858194824322,
"grad_norm": 93.18348693847656,
"learning_rate": 4.70430951509586e-06,
"loss": 0.8917,
"step": 601
},
{
"epoch": 0.5066273932253313,
"grad_norm": 116.30068969726562,
"learning_rate": 4.703255112566676e-06,
"loss": 1.0472,
"step": 602
},
{
"epoch": 0.5074689669682306,
"grad_norm": 99.05022430419922,
"learning_rate": 4.702198952033866e-06,
"loss": 0.9099,
"step": 603
},
{
"epoch": 0.5083105407111298,
"grad_norm": 99.7860336303711,
"learning_rate": 4.701141034340153e-06,
"loss": 0.9021,
"step": 604
},
{
"epoch": 0.5091521144540291,
"grad_norm": 122.43360137939453,
"learning_rate": 4.700081360329664e-06,
"loss": 0.8791,
"step": 605
},
{
"epoch": 0.5099936881969283,
"grad_norm": 110.67709350585938,
"learning_rate": 4.699019930847923e-06,
"loss": 0.9684,
"step": 606
},
{
"epoch": 0.5108352619398274,
"grad_norm": 88.45003509521484,
"learning_rate": 4.697956746741858e-06,
"loss": 0.5716,
"step": 607
},
{
"epoch": 0.5116768356827267,
"grad_norm": 91.00718688964844,
"learning_rate": 4.696891808859797e-06,
"loss": 0.8279,
"step": 608
},
{
"epoch": 0.5125184094256259,
"grad_norm": 103.737548828125,
"learning_rate": 4.695825118051467e-06,
"loss": 0.8202,
"step": 609
},
{
"epoch": 0.5133599831685252,
"grad_norm": 94.69662475585938,
"learning_rate": 4.694756675167992e-06,
"loss": 0.9734,
"step": 610
},
{
"epoch": 0.5142015569114243,
"grad_norm": 114.9798355102539,
"learning_rate": 4.693686481061898e-06,
"loss": 1.0855,
"step": 611
},
{
"epoch": 0.5150431306543236,
"grad_norm": 95.23029327392578,
"learning_rate": 4.692614536587101e-06,
"loss": 0.7172,
"step": 612
},
{
"epoch": 0.5158847043972228,
"grad_norm": 99.6467514038086,
"learning_rate": 4.691540842598923e-06,
"loss": 0.7589,
"step": 613
},
{
"epoch": 0.5167262781401221,
"grad_norm": 103.03988647460938,
"learning_rate": 4.690465399954076e-06,
"loss": 0.8081,
"step": 614
},
{
"epoch": 0.5175678518830212,
"grad_norm": 90.93890380859375,
"learning_rate": 4.689388209510667e-06,
"loss": 0.7581,
"step": 615
},
{
"epoch": 0.5184094256259205,
"grad_norm": 105.35801696777344,
"learning_rate": 4.688309272128201e-06,
"loss": 0.8253,
"step": 616
},
{
"epoch": 0.5192509993688197,
"grad_norm": 133.68971252441406,
"learning_rate": 4.687228588667573e-06,
"loss": 0.8134,
"step": 617
},
{
"epoch": 0.520092573111719,
"grad_norm": 107.72319793701172,
"learning_rate": 4.686146159991075e-06,
"loss": 0.9942,
"step": 618
},
{
"epoch": 0.5209341468546181,
"grad_norm": 110.94938659667969,
"learning_rate": 4.685061986962391e-06,
"loss": 0.8772,
"step": 619
},
{
"epoch": 0.5217757205975173,
"grad_norm": 102.12600708007812,
"learning_rate": 4.6839760704465915e-06,
"loss": 0.7658,
"step": 620
},
{
"epoch": 0.5226172943404166,
"grad_norm": 96.44245910644531,
"learning_rate": 4.6828884113101454e-06,
"loss": 0.9078,
"step": 621
},
{
"epoch": 0.5234588680833158,
"grad_norm": 118.14047241210938,
"learning_rate": 4.681799010420909e-06,
"loss": 1.0003,
"step": 622
},
{
"epoch": 0.524300441826215,
"grad_norm": 85.10506439208984,
"learning_rate": 4.680707868648127e-06,
"loss": 0.8619,
"step": 623
},
{
"epoch": 0.5251420155691142,
"grad_norm": 86.90888214111328,
"learning_rate": 4.679614986862435e-06,
"loss": 0.9455,
"step": 624
},
{
"epoch": 0.5259835893120135,
"grad_norm": 86.18508911132812,
"learning_rate": 4.678520365935857e-06,
"loss": 0.8459,
"step": 625
},
{
"epoch": 0.5268251630549127,
"grad_norm": 82.22334289550781,
"learning_rate": 4.677424006741803e-06,
"loss": 0.8193,
"step": 626
},
{
"epoch": 0.527666736797812,
"grad_norm": 80.63818359375,
"learning_rate": 4.676325910155072e-06,
"loss": 0.9017,
"step": 627
},
{
"epoch": 0.5285083105407111,
"grad_norm": 95.19749450683594,
"learning_rate": 4.675226077051849e-06,
"loss": 0.9112,
"step": 628
},
{
"epoch": 0.5293498842836104,
"grad_norm": 103.03102111816406,
"learning_rate": 4.674124508309702e-06,
"loss": 0.7195,
"step": 629
},
{
"epoch": 0.5301914580265096,
"grad_norm": 100.74110412597656,
"learning_rate": 4.673021204807588e-06,
"loss": 0.8151,
"step": 630
},
{
"epoch": 0.5310330317694087,
"grad_norm": 148.33775329589844,
"learning_rate": 4.671916167425844e-06,
"loss": 1.073,
"step": 631
},
{
"epoch": 0.531874605512308,
"grad_norm": 131.19943237304688,
"learning_rate": 4.670809397046195e-06,
"loss": 1.0741,
"step": 632
},
{
"epoch": 0.5327161792552072,
"grad_norm": 109.5487289428711,
"learning_rate": 4.669700894551744e-06,
"loss": 0.715,
"step": 633
},
{
"epoch": 0.5335577529981065,
"grad_norm": 90.64634704589844,
"learning_rate": 4.6685906608269795e-06,
"loss": 0.8414,
"step": 634
},
{
"epoch": 0.5343993267410057,
"grad_norm": 144.5821075439453,
"learning_rate": 4.667478696757771e-06,
"loss": 1.2275,
"step": 635
},
{
"epoch": 0.5352409004839049,
"grad_norm": 94.56639099121094,
"learning_rate": 4.666365003231367e-06,
"loss": 0.8112,
"step": 636
},
{
"epoch": 0.5360824742268041,
"grad_norm": 88.34557342529297,
"learning_rate": 4.665249581136396e-06,
"loss": 0.7514,
"step": 637
},
{
"epoch": 0.5369240479697034,
"grad_norm": 112.50604248046875,
"learning_rate": 4.6641324313628685e-06,
"loss": 0.8959,
"step": 638
},
{
"epoch": 0.5377656217126026,
"grad_norm": 93.03011322021484,
"learning_rate": 4.6630135548021714e-06,
"loss": 0.7076,
"step": 639
},
{
"epoch": 0.5386071954555018,
"grad_norm": 97.32340240478516,
"learning_rate": 4.661892952347069e-06,
"loss": 0.9593,
"step": 640
},
{
"epoch": 0.539448769198401,
"grad_norm": 98.1068115234375,
"learning_rate": 4.660770624891703e-06,
"loss": 0.9363,
"step": 641
},
{
"epoch": 0.5402903429413002,
"grad_norm": 84.54730987548828,
"learning_rate": 4.659646573331593e-06,
"loss": 0.7832,
"step": 642
},
{
"epoch": 0.5411319166841995,
"grad_norm": 98.7477035522461,
"learning_rate": 4.658520798563632e-06,
"loss": 0.9713,
"step": 643
},
{
"epoch": 0.5419734904270986,
"grad_norm": 86.43669128417969,
"learning_rate": 4.65739330148609e-06,
"loss": 0.9527,
"step": 644
},
{
"epoch": 0.5428150641699979,
"grad_norm": 86.94834899902344,
"learning_rate": 4.656264082998611e-06,
"loss": 0.7682,
"step": 645
},
{
"epoch": 0.5436566379128971,
"grad_norm": 82.33644104003906,
"learning_rate": 4.65513314400221e-06,
"loss": 0.7847,
"step": 646
},
{
"epoch": 0.5444982116557964,
"grad_norm": 77.75247192382812,
"learning_rate": 4.654000485399278e-06,
"loss": 0.8332,
"step": 647
},
{
"epoch": 0.5453397853986955,
"grad_norm": 87.43701171875,
"learning_rate": 4.652866108093578e-06,
"loss": 0.7374,
"step": 648
},
{
"epoch": 0.5461813591415948,
"grad_norm": 104.72925567626953,
"learning_rate": 4.6517300129902415e-06,
"loss": 1.07,
"step": 649
},
{
"epoch": 0.547022932884494,
"grad_norm": 83.67887878417969,
"learning_rate": 4.650592200995773e-06,
"loss": 0.893,
"step": 650
},
{
"epoch": 0.5478645066273933,
"grad_norm": 74.65118408203125,
"learning_rate": 4.649452673018047e-06,
"loss": 0.7125,
"step": 651
},
{
"epoch": 0.5487060803702924,
"grad_norm": 97.34910583496094,
"learning_rate": 4.648311429966304e-06,
"loss": 0.8686,
"step": 652
},
{
"epoch": 0.5495476541131916,
"grad_norm": 91.98160552978516,
"learning_rate": 4.647168472751158e-06,
"loss": 0.8258,
"step": 653
},
{
"epoch": 0.5503892278560909,
"grad_norm": 86.38584899902344,
"learning_rate": 4.6460238022845885e-06,
"loss": 0.6989,
"step": 654
},
{
"epoch": 0.5512308015989901,
"grad_norm": 98.5851058959961,
"learning_rate": 4.64487741947994e-06,
"loss": 1.05,
"step": 655
},
{
"epoch": 0.5520723753418894,
"grad_norm": 74.66211700439453,
"learning_rate": 4.643729325251926e-06,
"loss": 0.8472,
"step": 656
},
{
"epoch": 0.5529139490847885,
"grad_norm": 100.06787872314453,
"learning_rate": 4.642579520516624e-06,
"loss": 0.8999,
"step": 657
},
{
"epoch": 0.5537555228276878,
"grad_norm": 111.21283721923828,
"learning_rate": 4.641428006191477e-06,
"loss": 0.8044,
"step": 658
},
{
"epoch": 0.554597096570587,
"grad_norm": 102.51060485839844,
"learning_rate": 4.6402747831952936e-06,
"loss": 0.9319,
"step": 659
},
{
"epoch": 0.5554386703134863,
"grad_norm": 86.2106704711914,
"learning_rate": 4.639119852448242e-06,
"loss": 0.9393,
"step": 660
},
{
"epoch": 0.5562802440563854,
"grad_norm": 100.43547821044922,
"learning_rate": 4.637963214871857e-06,
"loss": 0.9186,
"step": 661
},
{
"epoch": 0.5571218177992847,
"grad_norm": 87.44974517822266,
"learning_rate": 4.636804871389033e-06,
"loss": 0.775,
"step": 662
},
{
"epoch": 0.5579633915421839,
"grad_norm": 86.83062744140625,
"learning_rate": 4.635644822924026e-06,
"loss": 0.8735,
"step": 663
},
{
"epoch": 0.558804965285083,
"grad_norm": 93.20692443847656,
"learning_rate": 4.6344830704024525e-06,
"loss": 0.9683,
"step": 664
},
{
"epoch": 0.5596465390279823,
"grad_norm": 103.13855743408203,
"learning_rate": 4.633319614751288e-06,
"loss": 0.789,
"step": 665
},
{
"epoch": 0.5604881127708815,
"grad_norm": 85.04438781738281,
"learning_rate": 4.63215445689887e-06,
"loss": 0.8745,
"step": 666
},
{
"epoch": 0.5613296865137808,
"grad_norm": 92.2657699584961,
"learning_rate": 4.6309875977748895e-06,
"loss": 0.6973,
"step": 667
},
{
"epoch": 0.56217126025668,
"grad_norm": 107.98641967773438,
"learning_rate": 4.629819038310399e-06,
"loss": 0.9694,
"step": 668
},
{
"epoch": 0.5630128339995792,
"grad_norm": 105.74818420410156,
"learning_rate": 4.628648779437805e-06,
"loss": 1.0651,
"step": 669
},
{
"epoch": 0.5638544077424784,
"grad_norm": 93.05474853515625,
"learning_rate": 4.6274768220908715e-06,
"loss": 0.8287,
"step": 670
},
{
"epoch": 0.5646959814853777,
"grad_norm": 101.81683349609375,
"learning_rate": 4.626303167204717e-06,
"loss": 1.0467,
"step": 671
},
{
"epoch": 0.5655375552282769,
"grad_norm": 108.88080596923828,
"learning_rate": 4.6251278157158155e-06,
"loss": 1.0961,
"step": 672
},
{
"epoch": 0.5663791289711761,
"grad_norm": 99.9627456665039,
"learning_rate": 4.623950768561993e-06,
"loss": 0.9827,
"step": 673
},
{
"epoch": 0.5672207027140753,
"grad_norm": 74.0265884399414,
"learning_rate": 4.622772026682431e-06,
"loss": 0.7854,
"step": 674
},
{
"epoch": 0.5680622764569745,
"grad_norm": 97.81053161621094,
"learning_rate": 4.621591591017659e-06,
"loss": 0.8921,
"step": 675
},
{
"epoch": 0.5689038501998738,
"grad_norm": 98.89409637451172,
"learning_rate": 4.6204094625095645e-06,
"loss": 1.0033,
"step": 676
},
{
"epoch": 0.5697454239427729,
"grad_norm": 260.55377197265625,
"learning_rate": 4.619225642101379e-06,
"loss": 0.6616,
"step": 677
},
{
"epoch": 0.5705869976856722,
"grad_norm": 88.79230499267578,
"learning_rate": 4.618040130737689e-06,
"loss": 0.8913,
"step": 678
},
{
"epoch": 0.5714285714285714,
"grad_norm": 88.83299255371094,
"learning_rate": 4.616852929364426e-06,
"loss": 0.9479,
"step": 679
},
{
"epoch": 0.5722701451714707,
"grad_norm": 96.11365509033203,
"learning_rate": 4.615664038928874e-06,
"loss": 0.9281,
"step": 680
},
{
"epoch": 0.5731117189143698,
"grad_norm": 85.7844009399414,
"learning_rate": 4.6144734603796625e-06,
"loss": 0.6493,
"step": 681
},
{
"epoch": 0.5739532926572691,
"grad_norm": 92.83222961425781,
"learning_rate": 4.613281194666768e-06,
"loss": 0.8968,
"step": 682
},
{
"epoch": 0.5747948664001683,
"grad_norm": 99.30359649658203,
"learning_rate": 4.612087242741513e-06,
"loss": 0.841,
"step": 683
},
{
"epoch": 0.5756364401430676,
"grad_norm": 82.88372802734375,
"learning_rate": 4.610891605556568e-06,
"loss": 0.7645,
"step": 684
},
{
"epoch": 0.5764780138859668,
"grad_norm": 101.14331817626953,
"learning_rate": 4.609694284065945e-06,
"loss": 0.9215,
"step": 685
},
{
"epoch": 0.5773195876288659,
"grad_norm": 94.408203125,
"learning_rate": 4.608495279225e-06,
"loss": 0.7959,
"step": 686
},
{
"epoch": 0.5781611613717652,
"grad_norm": 105.04963684082031,
"learning_rate": 4.607294591990435e-06,
"loss": 1.0119,
"step": 687
},
{
"epoch": 0.5790027351146644,
"grad_norm": 110.18367767333984,
"learning_rate": 4.606092223320293e-06,
"loss": 1.0655,
"step": 688
},
{
"epoch": 0.5798443088575637,
"grad_norm": 97.40960693359375,
"learning_rate": 4.604888174173957e-06,
"loss": 0.9047,
"step": 689
},
{
"epoch": 0.5806858826004628,
"grad_norm": 82.741943359375,
"learning_rate": 4.603682445512153e-06,
"loss": 0.6259,
"step": 690
},
{
"epoch": 0.5815274563433621,
"grad_norm": 75.7352523803711,
"learning_rate": 4.602475038296946e-06,
"loss": 0.6802,
"step": 691
},
{
"epoch": 0.5823690300862613,
"grad_norm": 92.63795471191406,
"learning_rate": 4.601265953491741e-06,
"loss": 0.7462,
"step": 692
},
{
"epoch": 0.5832106038291606,
"grad_norm": 107.34371948242188,
"learning_rate": 4.60005519206128e-06,
"loss": 0.9377,
"step": 693
},
{
"epoch": 0.5840521775720597,
"grad_norm": 89.72711181640625,
"learning_rate": 4.5988427549716465e-06,
"loss": 0.8217,
"step": 694
},
{
"epoch": 0.584893751314959,
"grad_norm": 98.20314025878906,
"learning_rate": 4.597628643190256e-06,
"loss": 0.8939,
"step": 695
},
{
"epoch": 0.5857353250578582,
"grad_norm": 113.12525177001953,
"learning_rate": 4.596412857685865e-06,
"loss": 0.9188,
"step": 696
},
{
"epoch": 0.5865768988007574,
"grad_norm": 103.85059356689453,
"learning_rate": 4.595195399428562e-06,
"loss": 1.0561,
"step": 697
},
{
"epoch": 0.5874184725436566,
"grad_norm": 108.72787475585938,
"learning_rate": 4.5939762693897704e-06,
"loss": 0.8937,
"step": 698
},
{
"epoch": 0.5882600462865558,
"grad_norm": 99.8670425415039,
"learning_rate": 4.592755468542251e-06,
"loss": 0.8651,
"step": 699
},
{
"epoch": 0.5891016200294551,
"grad_norm": 91.36952209472656,
"learning_rate": 4.591532997860094e-06,
"loss": 0.8302,
"step": 700
},
{
"epoch": 0.5899431937723543,
"grad_norm": 99.10494995117188,
"learning_rate": 4.590308858318725e-06,
"loss": 0.8918,
"step": 701
},
{
"epoch": 0.5907847675152536,
"grad_norm": 88.3750228881836,
"learning_rate": 4.589083050894898e-06,
"loss": 0.8718,
"step": 702
},
{
"epoch": 0.5916263412581527,
"grad_norm": 82.0086669921875,
"learning_rate": 4.587855576566699e-06,
"loss": 0.7089,
"step": 703
},
{
"epoch": 0.592467915001052,
"grad_norm": 75.7967529296875,
"learning_rate": 4.586626436313546e-06,
"loss": 0.7745,
"step": 704
},
{
"epoch": 0.5933094887439512,
"grad_norm": 95.32615661621094,
"learning_rate": 4.585395631116184e-06,
"loss": 0.925,
"step": 705
},
{
"epoch": 0.5941510624868505,
"grad_norm": 91.64199829101562,
"learning_rate": 4.584163161956687e-06,
"loss": 0.7124,
"step": 706
},
{
"epoch": 0.5949926362297496,
"grad_norm": 96.98888397216797,
"learning_rate": 4.582929029818457e-06,
"loss": 0.8887,
"step": 707
},
{
"epoch": 0.5958342099726488,
"grad_norm": 130.055908203125,
"learning_rate": 4.5816932356862236e-06,
"loss": 1.0215,
"step": 708
},
{
"epoch": 0.5966757837155481,
"grad_norm": 91.83544158935547,
"learning_rate": 4.58045578054604e-06,
"loss": 0.9852,
"step": 709
},
{
"epoch": 0.5975173574584473,
"grad_norm": 121.96644592285156,
"learning_rate": 4.579216665385288e-06,
"loss": 0.7534,
"step": 710
},
{
"epoch": 0.5983589312013465,
"grad_norm": 95.6441650390625,
"learning_rate": 4.577975891192672e-06,
"loss": 0.848,
"step": 711
},
{
"epoch": 0.5992005049442457,
"grad_norm": 90.83454895019531,
"learning_rate": 4.57673345895822e-06,
"loss": 0.8351,
"step": 712
},
{
"epoch": 0.600042078687145,
"grad_norm": 83.06048583984375,
"learning_rate": 4.575489369673285e-06,
"loss": 0.9874,
"step": 713
},
{
"epoch": 0.6008836524300442,
"grad_norm": 81.80094146728516,
"learning_rate": 4.5742436243305385e-06,
"loss": 0.9504,
"step": 714
},
{
"epoch": 0.6017252261729434,
"grad_norm": 79.86914825439453,
"learning_rate": 4.572996223923979e-06,
"loss": 0.8618,
"step": 715
},
{
"epoch": 0.6025667999158426,
"grad_norm": 90.9134750366211,
"learning_rate": 4.571747169448918e-06,
"loss": 0.6886,
"step": 716
},
{
"epoch": 0.6034083736587419,
"grad_norm": 76.10773468017578,
"learning_rate": 4.570496461901994e-06,
"loss": 0.7606,
"step": 717
},
{
"epoch": 0.6042499474016411,
"grad_norm": 84.14070129394531,
"learning_rate": 4.569244102281161e-06,
"loss": 0.8042,
"step": 718
},
{
"epoch": 0.6050915211445403,
"grad_norm": 90.08390045166016,
"learning_rate": 4.567990091585691e-06,
"loss": 0.8287,
"step": 719
},
{
"epoch": 0.6059330948874395,
"grad_norm": 81.404296875,
"learning_rate": 4.566734430816174e-06,
"loss": 0.7245,
"step": 720
},
{
"epoch": 0.6067746686303387,
"grad_norm": 91.73657989501953,
"learning_rate": 4.565477120974517e-06,
"loss": 0.7274,
"step": 721
},
{
"epoch": 0.607616242373238,
"grad_norm": 99.26171112060547,
"learning_rate": 4.564218163063944e-06,
"loss": 0.8198,
"step": 722
},
{
"epoch": 0.6084578161161371,
"grad_norm": 85.09310913085938,
"learning_rate": 4.562957558088992e-06,
"loss": 0.6707,
"step": 723
},
{
"epoch": 0.6092993898590364,
"grad_norm": 63.053466796875,
"learning_rate": 4.56169530705551e-06,
"loss": 0.5859,
"step": 724
},
{
"epoch": 0.6101409636019356,
"grad_norm": 93.56147003173828,
"learning_rate": 4.560431410970667e-06,
"loss": 0.815,
"step": 725
},
{
"epoch": 0.6109825373448349,
"grad_norm": 83.997314453125,
"learning_rate": 4.55916587084294e-06,
"loss": 0.7257,
"step": 726
},
{
"epoch": 0.611824111087734,
"grad_norm": 81.07512664794922,
"learning_rate": 4.5578986876821165e-06,
"loss": 0.7207,
"step": 727
},
{
"epoch": 0.6126656848306333,
"grad_norm": 89.04486846923828,
"learning_rate": 4.556629862499299e-06,
"loss": 0.7028,
"step": 728
},
{
"epoch": 0.6135072585735325,
"grad_norm": 115.55461120605469,
"learning_rate": 4.555359396306898e-06,
"loss": 0.9597,
"step": 729
},
{
"epoch": 0.6143488323164318,
"grad_norm": 78.45919799804688,
"learning_rate": 4.5540872901186325e-06,
"loss": 0.737,
"step": 730
},
{
"epoch": 0.615190406059331,
"grad_norm": 105.65302276611328,
"learning_rate": 4.552813544949533e-06,
"loss": 0.8943,
"step": 731
},
{
"epoch": 0.6160319798022301,
"grad_norm": 86.27177429199219,
"learning_rate": 4.551538161815935e-06,
"loss": 0.6789,
"step": 732
},
{
"epoch": 0.6168735535451294,
"grad_norm": 111.21141052246094,
"learning_rate": 4.550261141735483e-06,
"loss": 1.0777,
"step": 733
},
{
"epoch": 0.6177151272880286,
"grad_norm": 75.09874725341797,
"learning_rate": 4.548982485727125e-06,
"loss": 0.6761,
"step": 734
},
{
"epoch": 0.6185567010309279,
"grad_norm": 108.20618438720703,
"learning_rate": 4.547702194811117e-06,
"loss": 1.0618,
"step": 735
},
{
"epoch": 0.619398274773827,
"grad_norm": 84.72933959960938,
"learning_rate": 4.546420270009018e-06,
"loss": 0.7433,
"step": 736
},
{
"epoch": 0.6202398485167263,
"grad_norm": 89.56011962890625,
"learning_rate": 4.5451367123436905e-06,
"loss": 0.8188,
"step": 737
},
{
"epoch": 0.6210814222596255,
"grad_norm": 91.85026550292969,
"learning_rate": 4.543851522839303e-06,
"loss": 0.9414,
"step": 738
},
{
"epoch": 0.6219229960025248,
"grad_norm": 84.19593048095703,
"learning_rate": 4.542564702521321e-06,
"loss": 0.8061,
"step": 739
},
{
"epoch": 0.6227645697454239,
"grad_norm": 71.855224609375,
"learning_rate": 4.541276252416516e-06,
"loss": 0.7419,
"step": 740
},
{
"epoch": 0.6236061434883232,
"grad_norm": 120.99076080322266,
"learning_rate": 4.539986173552955e-06,
"loss": 0.8866,
"step": 741
},
{
"epoch": 0.6244477172312224,
"grad_norm": 98.39362335205078,
"learning_rate": 4.53869446696001e-06,
"loss": 0.8933,
"step": 742
},
{
"epoch": 0.6252892909741216,
"grad_norm": 89.82457733154297,
"learning_rate": 4.537401133668347e-06,
"loss": 0.8136,
"step": 743
},
{
"epoch": 0.6261308647170208,
"grad_norm": 93.52442932128906,
"learning_rate": 4.536106174709933e-06,
"loss": 0.8235,
"step": 744
},
{
"epoch": 0.62697243845992,
"grad_norm": 118.60028076171875,
"learning_rate": 4.534809591118033e-06,
"loss": 1.1256,
"step": 745
},
{
"epoch": 0.6278140122028193,
"grad_norm": 72.14741516113281,
"learning_rate": 4.533511383927204e-06,
"loss": 0.7877,
"step": 746
},
{
"epoch": 0.6286555859457185,
"grad_norm": 73.36578369140625,
"learning_rate": 4.532211554173303e-06,
"loss": 0.7578,
"step": 747
},
{
"epoch": 0.6294971596886177,
"grad_norm": 120.38871002197266,
"learning_rate": 4.530910102893478e-06,
"loss": 1.3195,
"step": 748
},
{
"epoch": 0.6303387334315169,
"grad_norm": 97.33341979980469,
"learning_rate": 4.529607031126173e-06,
"loss": 1.1318,
"step": 749
},
{
"epoch": 0.6311803071744162,
"grad_norm": 81.80436706542969,
"learning_rate": 4.528302339911125e-06,
"loss": 0.7518,
"step": 750
},
{
"epoch": 0.6320218809173154,
"grad_norm": 86.9666519165039,
"learning_rate": 4.526996030289363e-06,
"loss": 0.8304,
"step": 751
},
{
"epoch": 0.6328634546602147,
"grad_norm": 77.8279800415039,
"learning_rate": 4.525688103303206e-06,
"loss": 0.7867,
"step": 752
},
{
"epoch": 0.6337050284031138,
"grad_norm": 67.7917251586914,
"learning_rate": 4.524378559996267e-06,
"loss": 0.7116,
"step": 753
},
{
"epoch": 0.634546602146013,
"grad_norm": 82.60803985595703,
"learning_rate": 4.5230674014134425e-06,
"loss": 0.7545,
"step": 754
},
{
"epoch": 0.6353881758889123,
"grad_norm": 84.61490631103516,
"learning_rate": 4.521754628600925e-06,
"loss": 0.7934,
"step": 755
},
{
"epoch": 0.6362297496318114,
"grad_norm": 87.46534729003906,
"learning_rate": 4.52044024260619e-06,
"loss": 0.8427,
"step": 756
},
{
"epoch": 0.6370713233747107,
"grad_norm": 165.8030548095703,
"learning_rate": 4.519124244478002e-06,
"loss": 0.7902,
"step": 757
},
{
"epoch": 0.6379128971176099,
"grad_norm": 72.46358489990234,
"learning_rate": 4.517806635266412e-06,
"loss": 0.6943,
"step": 758
},
{
"epoch": 0.6387544708605092,
"grad_norm": 85.27467346191406,
"learning_rate": 4.5164874160227555e-06,
"loss": 1.0195,
"step": 759
},
{
"epoch": 0.6395960446034084,
"grad_norm": 166.4009552001953,
"learning_rate": 4.5151665877996554e-06,
"loss": 0.8665,
"step": 760
},
{
"epoch": 0.6404376183463076,
"grad_norm": 120.591552734375,
"learning_rate": 4.513844151651013e-06,
"loss": 0.9978,
"step": 761
},
{
"epoch": 0.6412791920892068,
"grad_norm": 84.2991714477539,
"learning_rate": 4.512520108632019e-06,
"loss": 1.0485,
"step": 762
},
{
"epoch": 0.6421207658321061,
"grad_norm": 81.13017272949219,
"learning_rate": 4.5111944597991405e-06,
"loss": 0.6065,
"step": 763
},
{
"epoch": 0.6429623395750053,
"grad_norm": 83.88569641113281,
"learning_rate": 4.5098672062101316e-06,
"loss": 0.9585,
"step": 764
},
{
"epoch": 0.6438039133179044,
"grad_norm": 87.67027282714844,
"learning_rate": 4.5085383489240205e-06,
"loss": 0.5934,
"step": 765
},
{
"epoch": 0.6446454870608037,
"grad_norm": 144.13389587402344,
"learning_rate": 4.5072078890011206e-06,
"loss": 1.012,
"step": 766
},
{
"epoch": 0.6454870608037029,
"grad_norm": 110.44563293457031,
"learning_rate": 4.5058758275030215e-06,
"loss": 0.96,
"step": 767
},
{
"epoch": 0.6463286345466022,
"grad_norm": 94.65059661865234,
"learning_rate": 4.5045421654925905e-06,
"loss": 0.7704,
"step": 768
},
{
"epoch": 0.6471702082895013,
"grad_norm": 80.62268829345703,
"learning_rate": 4.503206904033973e-06,
"loss": 0.88,
"step": 769
},
{
"epoch": 0.6480117820324006,
"grad_norm": 80.2649154663086,
"learning_rate": 4.501870044192589e-06,
"loss": 0.7423,
"step": 770
},
{
"epoch": 0.6488533557752998,
"grad_norm": 182.56639099121094,
"learning_rate": 4.5005315870351365e-06,
"loss": 0.7048,
"step": 771
},
{
"epoch": 0.6496949295181991,
"grad_norm": 143.60955810546875,
"learning_rate": 4.499191533629586e-06,
"loss": 0.7246,
"step": 772
},
{
"epoch": 0.6505365032610982,
"grad_norm": 70.18280792236328,
"learning_rate": 4.4978498850451815e-06,
"loss": 0.7484,
"step": 773
},
{
"epoch": 0.6513780770039975,
"grad_norm": 71.79647064208984,
"learning_rate": 4.4965066423524405e-06,
"loss": 0.8044,
"step": 774
},
{
"epoch": 0.6522196507468967,
"grad_norm": 88.14927673339844,
"learning_rate": 4.495161806623153e-06,
"loss": 0.9869,
"step": 775
},
{
"epoch": 0.6530612244897959,
"grad_norm": 96.6579818725586,
"learning_rate": 4.493815378930379e-06,
"loss": 1.0004,
"step": 776
},
{
"epoch": 0.6539027982326951,
"grad_norm": 75.86454772949219,
"learning_rate": 4.492467360348451e-06,
"loss": 0.8189,
"step": 777
},
{
"epoch": 0.6547443719755943,
"grad_norm": 167.09136962890625,
"learning_rate": 4.491117751952966e-06,
"loss": 0.8028,
"step": 778
},
{
"epoch": 0.6555859457184936,
"grad_norm": 82.6733627319336,
"learning_rate": 4.489766554820795e-06,
"loss": 0.8775,
"step": 779
},
{
"epoch": 0.6564275194613928,
"grad_norm": 84.00127410888672,
"learning_rate": 4.488413770030074e-06,
"loss": 0.866,
"step": 780
},
{
"epoch": 0.657269093204292,
"grad_norm": 95.56571197509766,
"learning_rate": 4.4870593986602055e-06,
"loss": 0.7944,
"step": 781
},
{
"epoch": 0.6581106669471912,
"grad_norm": 76.46224975585938,
"learning_rate": 4.48570344179186e-06,
"loss": 0.7765,
"step": 782
},
{
"epoch": 0.6589522406900905,
"grad_norm": 95.7029800415039,
"learning_rate": 4.48434590050697e-06,
"loss": 0.9271,
"step": 783
},
{
"epoch": 0.6597938144329897,
"grad_norm": 74.58334350585938,
"learning_rate": 4.482986775888735e-06,
"loss": 0.8118,
"step": 784
},
{
"epoch": 0.660635388175889,
"grad_norm": 87.86770629882812,
"learning_rate": 4.481626069021616e-06,
"loss": 0.9888,
"step": 785
},
{
"epoch": 0.6614769619187881,
"grad_norm": 86.0281753540039,
"learning_rate": 4.480263780991338e-06,
"loss": 0.7254,
"step": 786
},
{
"epoch": 0.6623185356616873,
"grad_norm": 105.17807006835938,
"learning_rate": 4.478899912884888e-06,
"loss": 0.8356,
"step": 787
},
{
"epoch": 0.6631601094045866,
"grad_norm": 86.4527816772461,
"learning_rate": 4.47753446579051e-06,
"loss": 0.7711,
"step": 788
},
{
"epoch": 0.6640016831474858,
"grad_norm": 87.57376861572266,
"learning_rate": 4.476167440797713e-06,
"loss": 0.9198,
"step": 789
},
{
"epoch": 0.664843256890385,
"grad_norm": 75.603759765625,
"learning_rate": 4.474798838997262e-06,
"loss": 0.7014,
"step": 790
},
{
"epoch": 0.6656848306332842,
"grad_norm": 74.6491470336914,
"learning_rate": 4.47342866148118e-06,
"loss": 0.7734,
"step": 791
},
{
"epoch": 0.6665264043761835,
"grad_norm": 78.29629516601562,
"learning_rate": 4.472056909342748e-06,
"loss": 0.7462,
"step": 792
},
{
"epoch": 0.6673679781190827,
"grad_norm": 84.12577819824219,
"learning_rate": 4.470683583676504e-06,
"loss": 0.8609,
"step": 793
},
{
"epoch": 0.6682095518619819,
"grad_norm": 84.50099182128906,
"learning_rate": 4.469308685578241e-06,
"loss": 0.9761,
"step": 794
},
{
"epoch": 0.6690511256048811,
"grad_norm": 78.68242645263672,
"learning_rate": 4.467932216145006e-06,
"loss": 0.7648,
"step": 795
},
{
"epoch": 0.6698926993477804,
"grad_norm": 78.8578872680664,
"learning_rate": 4.466554176475102e-06,
"loss": 0.8425,
"step": 796
},
{
"epoch": 0.6707342730906796,
"grad_norm": 99.00289154052734,
"learning_rate": 4.465174567668081e-06,
"loss": 0.9641,
"step": 797
},
{
"epoch": 0.6715758468335787,
"grad_norm": 70.02457427978516,
"learning_rate": 4.463793390824752e-06,
"loss": 0.7527,
"step": 798
},
{
"epoch": 0.672417420576478,
"grad_norm": 95.30015563964844,
"learning_rate": 4.4624106470471704e-06,
"loss": 0.9996,
"step": 799
},
{
"epoch": 0.6732589943193772,
"grad_norm": 72.76326751708984,
"learning_rate": 4.461026337438646e-06,
"loss": 0.7481,
"step": 800
},
{
"epoch": 0.6741005680622765,
"grad_norm": 112.60525512695312,
"learning_rate": 4.459640463103733e-06,
"loss": 1.0518,
"step": 801
},
{
"epoch": 0.6749421418051756,
"grad_norm": 74.76424407958984,
"learning_rate": 4.45825302514824e-06,
"loss": 0.7562,
"step": 802
},
{
"epoch": 0.6757837155480749,
"grad_norm": 76.3163070678711,
"learning_rate": 4.456864024679219e-06,
"loss": 0.8997,
"step": 803
},
{
"epoch": 0.6766252892909741,
"grad_norm": 81.21278381347656,
"learning_rate": 4.4554734628049714e-06,
"loss": 0.7025,
"step": 804
},
{
"epoch": 0.6774668630338734,
"grad_norm": 97.0674819946289,
"learning_rate": 4.454081340635042e-06,
"loss": 0.8344,
"step": 805
},
{
"epoch": 0.6783084367767725,
"grad_norm": 74.96123504638672,
"learning_rate": 4.4526876592802205e-06,
"loss": 0.7259,
"step": 806
},
{
"epoch": 0.6791500105196718,
"grad_norm": 85.4431381225586,
"learning_rate": 4.451292419852544e-06,
"loss": 0.8788,
"step": 807
},
{
"epoch": 0.679991584262571,
"grad_norm": 90.90916442871094,
"learning_rate": 4.449895623465291e-06,
"loss": 1.0065,
"step": 808
},
{
"epoch": 0.6808331580054703,
"grad_norm": 79.66510009765625,
"learning_rate": 4.44849727123298e-06,
"loss": 0.7831,
"step": 809
},
{
"epoch": 0.6816747317483695,
"grad_norm": 87.62796783447266,
"learning_rate": 4.4470973642713745e-06,
"loss": 0.8283,
"step": 810
},
{
"epoch": 0.6825163054912686,
"grad_norm": 106.29439544677734,
"learning_rate": 4.445695903697477e-06,
"loss": 0.8776,
"step": 811
},
{
"epoch": 0.6833578792341679,
"grad_norm": 86.14681243896484,
"learning_rate": 4.444292890629529e-06,
"loss": 0.7907,
"step": 812
},
{
"epoch": 0.6841994529770671,
"grad_norm": 111.23360443115234,
"learning_rate": 4.442888326187012e-06,
"loss": 0.9141,
"step": 813
},
{
"epoch": 0.6850410267199664,
"grad_norm": 80.4526138305664,
"learning_rate": 4.441482211490644e-06,
"loss": 0.7239,
"step": 814
},
{
"epoch": 0.6858826004628655,
"grad_norm": 78.80682373046875,
"learning_rate": 4.440074547662382e-06,
"loss": 0.7941,
"step": 815
},
{
"epoch": 0.6867241742057648,
"grad_norm": 79.59489440917969,
"learning_rate": 4.438665335825416e-06,
"loss": 0.8866,
"step": 816
},
{
"epoch": 0.687565747948664,
"grad_norm": 72.70225524902344,
"learning_rate": 4.4372545771041755e-06,
"loss": 0.6917,
"step": 817
},
{
"epoch": 0.6884073216915633,
"grad_norm": 89.61310577392578,
"learning_rate": 4.43584227262432e-06,
"loss": 1.0724,
"step": 818
},
{
"epoch": 0.6892488954344624,
"grad_norm": 108.31492614746094,
"learning_rate": 4.4344284235127435e-06,
"loss": 1.0526,
"step": 819
},
{
"epoch": 0.6900904691773617,
"grad_norm": 64.2813491821289,
"learning_rate": 4.433013030897575e-06,
"loss": 0.5364,
"step": 820
},
{
"epoch": 0.6909320429202609,
"grad_norm": 83.37664794921875,
"learning_rate": 4.431596095908173e-06,
"loss": 0.7599,
"step": 821
},
{
"epoch": 0.6917736166631601,
"grad_norm": 82.26396942138672,
"learning_rate": 4.430177619675125e-06,
"loss": 0.8282,
"step": 822
},
{
"epoch": 0.6926151904060593,
"grad_norm": 90.45413970947266,
"learning_rate": 4.428757603330251e-06,
"loss": 0.853,
"step": 823
},
{
"epoch": 0.6934567641489585,
"grad_norm": 106.56082916259766,
"learning_rate": 4.4273360480066e-06,
"loss": 1.0193,
"step": 824
},
{
"epoch": 0.6942983378918578,
"grad_norm": 94.9800033569336,
"learning_rate": 4.425912954838448e-06,
"loss": 0.8258,
"step": 825
},
{
"epoch": 0.695139911634757,
"grad_norm": 90.53028106689453,
"learning_rate": 4.424488324961297e-06,
"loss": 0.8751,
"step": 826
},
{
"epoch": 0.6959814853776563,
"grad_norm": 101.14625549316406,
"learning_rate": 4.423062159511877e-06,
"loss": 0.9244,
"step": 827
},
{
"epoch": 0.6968230591205554,
"grad_norm": 79.58170318603516,
"learning_rate": 4.421634459628141e-06,
"loss": 0.7385,
"step": 828
},
{
"epoch": 0.6976646328634547,
"grad_norm": 86.39154052734375,
"learning_rate": 4.420205226449271e-06,
"loss": 0.8993,
"step": 829
},
{
"epoch": 0.6985062066063539,
"grad_norm": 129.21112060546875,
"learning_rate": 4.418774461115666e-06,
"loss": 0.8467,
"step": 830
},
{
"epoch": 0.6993477803492532,
"grad_norm": 74.53318786621094,
"learning_rate": 4.417342164768953e-06,
"loss": 0.7344,
"step": 831
},
{
"epoch": 0.7001893540921523,
"grad_norm": 86.27154541015625,
"learning_rate": 4.4159083385519765e-06,
"loss": 0.7898,
"step": 832
},
{
"epoch": 0.7010309278350515,
"grad_norm": 70.78800964355469,
"learning_rate": 4.4144729836088055e-06,
"loss": 0.7605,
"step": 833
},
{
"epoch": 0.7018725015779508,
"grad_norm": 89.29914855957031,
"learning_rate": 4.413036101084725e-06,
"loss": 0.9406,
"step": 834
},
{
"epoch": 0.70271407532085,
"grad_norm": 92.68386840820312,
"learning_rate": 4.411597692126242e-06,
"loss": 0.9553,
"step": 835
},
{
"epoch": 0.7035556490637492,
"grad_norm": 71.86311340332031,
"learning_rate": 4.410157757881081e-06,
"loss": 0.7552,
"step": 836
},
{
"epoch": 0.7043972228066484,
"grad_norm": 71.1396713256836,
"learning_rate": 4.4087162994981805e-06,
"loss": 0.718,
"step": 837
},
{
"epoch": 0.7052387965495477,
"grad_norm": 82.484619140625,
"learning_rate": 4.407273318127699e-06,
"loss": 0.9462,
"step": 838
},
{
"epoch": 0.7060803702924469,
"grad_norm": 84.81983184814453,
"learning_rate": 4.4058288149210085e-06,
"loss": 0.9518,
"step": 839
},
{
"epoch": 0.7069219440353461,
"grad_norm": 81.63459014892578,
"learning_rate": 4.4043827910306944e-06,
"loss": 0.9702,
"step": 840
},
{
"epoch": 0.7077635177782453,
"grad_norm": 91.86296081542969,
"learning_rate": 4.402935247610558e-06,
"loss": 0.9159,
"step": 841
},
{
"epoch": 0.7086050915211446,
"grad_norm": 70.82006072998047,
"learning_rate": 4.401486185815611e-06,
"loss": 0.6971,
"step": 842
},
{
"epoch": 0.7094466652640438,
"grad_norm": 93.87294006347656,
"learning_rate": 4.400035606802076e-06,
"loss": 0.859,
"step": 843
},
{
"epoch": 0.7102882390069429,
"grad_norm": 109.06575775146484,
"learning_rate": 4.398583511727388e-06,
"loss": 0.9812,
"step": 844
},
{
"epoch": 0.7111298127498422,
"grad_norm": 116.77679443359375,
"learning_rate": 4.397129901750192e-06,
"loss": 0.9749,
"step": 845
},
{
"epoch": 0.7119713864927414,
"grad_norm": 69.3492431640625,
"learning_rate": 4.39567477803034e-06,
"loss": 0.7911,
"step": 846
},
{
"epoch": 0.7128129602356407,
"grad_norm": 78.79708099365234,
"learning_rate": 4.394218141728893e-06,
"loss": 0.9871,
"step": 847
},
{
"epoch": 0.7136545339785398,
"grad_norm": 58.36460876464844,
"learning_rate": 4.392759994008118e-06,
"loss": 0.7013,
"step": 848
},
{
"epoch": 0.7144961077214391,
"grad_norm": 77.56861114501953,
"learning_rate": 4.391300336031489e-06,
"loss": 0.8301,
"step": 849
},
{
"epoch": 0.7153376814643383,
"grad_norm": 64.96390533447266,
"learning_rate": 4.3898391689636834e-06,
"loss": 0.7755,
"step": 850
},
{
"epoch": 0.7161792552072376,
"grad_norm": 75.965087890625,
"learning_rate": 4.388376493970585e-06,
"loss": 0.7274,
"step": 851
},
{
"epoch": 0.7170208289501367,
"grad_norm": 76.60365295410156,
"learning_rate": 4.38691231221928e-06,
"loss": 1.0479,
"step": 852
},
{
"epoch": 0.717862402693036,
"grad_norm": 97.6857681274414,
"learning_rate": 4.385446624878057e-06,
"loss": 1.1624,
"step": 853
},
{
"epoch": 0.7187039764359352,
"grad_norm": 83.55355072021484,
"learning_rate": 4.383979433116403e-06,
"loss": 0.8381,
"step": 854
},
{
"epoch": 0.7195455501788344,
"grad_norm": 99.44181823730469,
"learning_rate": 4.382510738105011e-06,
"loss": 1.0486,
"step": 855
},
{
"epoch": 0.7203871239217337,
"grad_norm": 71.99688720703125,
"learning_rate": 4.381040541015766e-06,
"loss": 0.8731,
"step": 856
},
{
"epoch": 0.7212286976646328,
"grad_norm": 86.90930938720703,
"learning_rate": 4.379568843021761e-06,
"loss": 0.9857,
"step": 857
},
{
"epoch": 0.7220702714075321,
"grad_norm": 81.49408721923828,
"learning_rate": 4.378095645297279e-06,
"loss": 0.8581,
"step": 858
},
{
"epoch": 0.7229118451504313,
"grad_norm": 74.4553451538086,
"learning_rate": 4.376620949017801e-06,
"loss": 0.851,
"step": 859
},
{
"epoch": 0.7237534188933306,
"grad_norm": 92.5252685546875,
"learning_rate": 4.3751447553600065e-06,
"loss": 0.7895,
"step": 860
},
{
"epoch": 0.7245949926362297,
"grad_norm": 70.13682556152344,
"learning_rate": 4.3736670655017675e-06,
"loss": 0.8247,
"step": 861
},
{
"epoch": 0.725436566379129,
"grad_norm": 87.14791870117188,
"learning_rate": 4.372187880622149e-06,
"loss": 0.8072,
"step": 862
},
{
"epoch": 0.7262781401220282,
"grad_norm": 214.95753479003906,
"learning_rate": 4.370707201901414e-06,
"loss": 0.8754,
"step": 863
},
{
"epoch": 0.7271197138649275,
"grad_norm": 85.0881118774414,
"learning_rate": 4.369225030521011e-06,
"loss": 0.9258,
"step": 864
},
{
"epoch": 0.7279612876078266,
"grad_norm": 79.23942565917969,
"learning_rate": 4.367741367663582e-06,
"loss": 0.9405,
"step": 865
},
{
"epoch": 0.7288028613507258,
"grad_norm": 77.9674301147461,
"learning_rate": 4.36625621451296e-06,
"loss": 0.8144,
"step": 866
},
{
"epoch": 0.7296444350936251,
"grad_norm": 58.98210906982422,
"learning_rate": 4.364769572254169e-06,
"loss": 0.5694,
"step": 867
},
{
"epoch": 0.7304860088365243,
"grad_norm": 103.85771179199219,
"learning_rate": 4.363281442073414e-06,
"loss": 0.812,
"step": 868
},
{
"epoch": 0.7313275825794235,
"grad_norm": 96.6537094116211,
"learning_rate": 4.361791825158095e-06,
"loss": 0.8771,
"step": 869
},
{
"epoch": 0.7321691563223227,
"grad_norm": 71.61028289794922,
"learning_rate": 4.360300722696794e-06,
"loss": 0.7049,
"step": 870
},
{
"epoch": 0.733010730065222,
"grad_norm": 67.1308364868164,
"learning_rate": 4.3588081358792796e-06,
"loss": 0.7327,
"step": 871
},
{
"epoch": 0.7338523038081212,
"grad_norm": 87.99250030517578,
"learning_rate": 4.357314065896506e-06,
"loss": 0.8441,
"step": 872
},
{
"epoch": 0.7346938775510204,
"grad_norm": 91.286376953125,
"learning_rate": 4.355818513940608e-06,
"loss": 0.8906,
"step": 873
},
{
"epoch": 0.7355354512939196,
"grad_norm": 77.45460510253906,
"learning_rate": 4.354321481204904e-06,
"loss": 0.8006,
"step": 874
},
{
"epoch": 0.7363770250368189,
"grad_norm": 83.27759552001953,
"learning_rate": 4.352822968883896e-06,
"loss": 0.6706,
"step": 875
},
{
"epoch": 0.7372185987797181,
"grad_norm": 131.8611297607422,
"learning_rate": 4.351322978173264e-06,
"loss": 0.7503,
"step": 876
},
{
"epoch": 0.7380601725226172,
"grad_norm": 66.10797119140625,
"learning_rate": 4.3498215102698695e-06,
"loss": 0.6593,
"step": 877
},
{
"epoch": 0.7389017462655165,
"grad_norm": 81.42675018310547,
"learning_rate": 4.348318566371749e-06,
"loss": 0.8397,
"step": 878
},
{
"epoch": 0.7397433200084157,
"grad_norm": 76.12470245361328,
"learning_rate": 4.346814147678124e-06,
"loss": 0.7452,
"step": 879
},
{
"epoch": 0.740584893751315,
"grad_norm": 88.55060577392578,
"learning_rate": 4.345308255389383e-06,
"loss": 1.0045,
"step": 880
},
{
"epoch": 0.7414264674942141,
"grad_norm": 69.48717498779297,
"learning_rate": 4.343800890707099e-06,
"loss": 0.6494,
"step": 881
},
{
"epoch": 0.7422680412371134,
"grad_norm": 73.43086242675781,
"learning_rate": 4.342292054834016e-06,
"loss": 0.7483,
"step": 882
},
{
"epoch": 0.7431096149800126,
"grad_norm": 102.50550079345703,
"learning_rate": 4.340781748974051e-06,
"loss": 0.7571,
"step": 883
},
{
"epoch": 0.7439511887229119,
"grad_norm": 74.73184204101562,
"learning_rate": 4.339269974332295e-06,
"loss": 0.7497,
"step": 884
},
{
"epoch": 0.744792762465811,
"grad_norm": 105.30278778076172,
"learning_rate": 4.337756732115011e-06,
"loss": 0.8387,
"step": 885
},
{
"epoch": 0.7456343362087103,
"grad_norm": 95.76217651367188,
"learning_rate": 4.336242023529635e-06,
"loss": 0.913,
"step": 886
},
{
"epoch": 0.7464759099516095,
"grad_norm": 81.93257904052734,
"learning_rate": 4.334725849784771e-06,
"loss": 0.8567,
"step": 887
},
{
"epoch": 0.7473174836945087,
"grad_norm": 86.54633331298828,
"learning_rate": 4.3332082120901895e-06,
"loss": 0.9386,
"step": 888
},
{
"epoch": 0.748159057437408,
"grad_norm": 77.6705093383789,
"learning_rate": 4.331689111656834e-06,
"loss": 0.8797,
"step": 889
},
{
"epoch": 0.7490006311803071,
"grad_norm": 63.00545120239258,
"learning_rate": 4.3301685496968114e-06,
"loss": 0.7039,
"step": 890
},
{
"epoch": 0.7498422049232064,
"grad_norm": 92.33924102783203,
"learning_rate": 4.328646527423398e-06,
"loss": 0.8826,
"step": 891
},
{
"epoch": 0.7506837786661056,
"grad_norm": 74.36640167236328,
"learning_rate": 4.3271230460510325e-06,
"loss": 0.7637,
"step": 892
},
{
"epoch": 0.7515253524090049,
"grad_norm": 428.94378662109375,
"learning_rate": 4.325598106795319e-06,
"loss": 0.8922,
"step": 893
},
{
"epoch": 0.752366926151904,
"grad_norm": 82.86460876464844,
"learning_rate": 4.324071710873025e-06,
"loss": 0.8002,
"step": 894
},
{
"epoch": 0.7532084998948033,
"grad_norm": 115.68354034423828,
"learning_rate": 4.322543859502081e-06,
"loss": 1.0063,
"step": 895
},
{
"epoch": 0.7540500736377025,
"grad_norm": 83.29459381103516,
"learning_rate": 4.3210145539015755e-06,
"loss": 0.8483,
"step": 896
},
{
"epoch": 0.7548916473806018,
"grad_norm": 73.96419525146484,
"learning_rate": 4.319483795291762e-06,
"loss": 0.8241,
"step": 897
},
{
"epoch": 0.7557332211235009,
"grad_norm": 77.9550552368164,
"learning_rate": 4.31795158489405e-06,
"loss": 0.7826,
"step": 898
},
{
"epoch": 0.7565747948664001,
"grad_norm": 103.94831085205078,
"learning_rate": 4.316417923931008e-06,
"loss": 0.8422,
"step": 899
},
{
"epoch": 0.7574163686092994,
"grad_norm": 107.88838195800781,
"learning_rate": 4.314882813626364e-06,
"loss": 1.0209,
"step": 900
},
{
"epoch": 0.7582579423521986,
"grad_norm": 76.37074279785156,
"learning_rate": 4.313346255205001e-06,
"loss": 0.8941,
"step": 901
},
{
"epoch": 0.7590995160950978,
"grad_norm": 61.950862884521484,
"learning_rate": 4.311808249892955e-06,
"loss": 0.624,
"step": 902
},
{
"epoch": 0.759941089837997,
"grad_norm": 89.68816375732422,
"learning_rate": 4.310268798917422e-06,
"loss": 1.0453,
"step": 903
},
{
"epoch": 0.7607826635808963,
"grad_norm": 74.37384033203125,
"learning_rate": 4.308727903506745e-06,
"loss": 0.933,
"step": 904
},
{
"epoch": 0.7616242373237955,
"grad_norm": 72.46370697021484,
"learning_rate": 4.307185564890426e-06,
"loss": 0.7905,
"step": 905
},
{
"epoch": 0.7624658110666948,
"grad_norm": 66.4675064086914,
"learning_rate": 4.305641784299113e-06,
"loss": 0.7179,
"step": 906
},
{
"epoch": 0.7633073848095939,
"grad_norm": 75.27861022949219,
"learning_rate": 4.304096562964608e-06,
"loss": 0.8811,
"step": 907
},
{
"epoch": 0.7641489585524932,
"grad_norm": 61.52951431274414,
"learning_rate": 4.302549902119863e-06,
"loss": 0.6958,
"step": 908
},
{
"epoch": 0.7649905322953924,
"grad_norm": 67.17115020751953,
"learning_rate": 4.301001802998974e-06,
"loss": 0.7002,
"step": 909
},
{
"epoch": 0.7658321060382917,
"grad_norm": 98.0196304321289,
"learning_rate": 4.299452266837191e-06,
"loss": 0.9624,
"step": 910
},
{
"epoch": 0.7666736797811908,
"grad_norm": 83.28306579589844,
"learning_rate": 4.2979012948709055e-06,
"loss": 0.9621,
"step": 911
},
{
"epoch": 0.76751525352409,
"grad_norm": 76.3790283203125,
"learning_rate": 4.296348888337657e-06,
"loss": 0.6902,
"step": 912
},
{
"epoch": 0.7683568272669893,
"grad_norm": 87.94258880615234,
"learning_rate": 4.29479504847613e-06,
"loss": 0.8355,
"step": 913
},
{
"epoch": 0.7691984010098885,
"grad_norm": 92.80834197998047,
"learning_rate": 4.2932397765261516e-06,
"loss": 1.0073,
"step": 914
},
{
"epoch": 0.7700399747527877,
"grad_norm": 71.12057495117188,
"learning_rate": 4.29168307372869e-06,
"loss": 0.7336,
"step": 915
},
{
"epoch": 0.7708815484956869,
"grad_norm": 67.44596862792969,
"learning_rate": 4.290124941325861e-06,
"loss": 0.7769,
"step": 916
},
{
"epoch": 0.7717231222385862,
"grad_norm": 71.54752349853516,
"learning_rate": 4.288565380560913e-06,
"loss": 0.726,
"step": 917
},
{
"epoch": 0.7725646959814854,
"grad_norm": 148.19482421875,
"learning_rate": 4.287004392678241e-06,
"loss": 0.9742,
"step": 918
},
{
"epoch": 0.7734062697243846,
"grad_norm": 61.75517272949219,
"learning_rate": 4.2854419789233744e-06,
"loss": 0.6872,
"step": 919
},
{
"epoch": 0.7742478434672838,
"grad_norm": 74.51992797851562,
"learning_rate": 4.2838781405429825e-06,
"loss": 0.6292,
"step": 920
},
{
"epoch": 0.7750894172101831,
"grad_norm": 87.5741958618164,
"learning_rate": 4.28231287878487e-06,
"loss": 0.8581,
"step": 921
},
{
"epoch": 0.7759309909530823,
"grad_norm": 105.05841064453125,
"learning_rate": 4.280746194897979e-06,
"loss": 0.8532,
"step": 922
},
{
"epoch": 0.7767725646959814,
"grad_norm": 80.18090057373047,
"learning_rate": 4.279178090132386e-06,
"loss": 0.7673,
"step": 923
},
{
"epoch": 0.7776141384388807,
"grad_norm": 117.09395599365234,
"learning_rate": 4.277608565739297e-06,
"loss": 0.9147,
"step": 924
},
{
"epoch": 0.7784557121817799,
"grad_norm": 98.15302276611328,
"learning_rate": 4.276037622971059e-06,
"loss": 0.8127,
"step": 925
},
{
"epoch": 0.7792972859246792,
"grad_norm": 81.57997131347656,
"learning_rate": 4.274465263081142e-06,
"loss": 0.7557,
"step": 926
},
{
"epoch": 0.7801388596675783,
"grad_norm": 110.0119400024414,
"learning_rate": 4.272891487324152e-06,
"loss": 0.7321,
"step": 927
},
{
"epoch": 0.7809804334104776,
"grad_norm": 126.8502197265625,
"learning_rate": 4.271316296955825e-06,
"loss": 0.8107,
"step": 928
},
{
"epoch": 0.7818220071533768,
"grad_norm": 72.49397277832031,
"learning_rate": 4.269739693233024e-06,
"loss": 0.6845,
"step": 929
},
{
"epoch": 0.7826635808962761,
"grad_norm": 74.62886810302734,
"learning_rate": 4.268161677413738e-06,
"loss": 0.714,
"step": 930
},
{
"epoch": 0.7835051546391752,
"grad_norm": 78.56126403808594,
"learning_rate": 4.266582250757086e-06,
"loss": 0.7087,
"step": 931
},
{
"epoch": 0.7843467283820745,
"grad_norm": 107.03934478759766,
"learning_rate": 4.265001414523312e-06,
"loss": 0.8672,
"step": 932
},
{
"epoch": 0.7851883021249737,
"grad_norm": 88.3293685913086,
"learning_rate": 4.263419169973784e-06,
"loss": 0.9362,
"step": 933
},
{
"epoch": 0.7860298758678729,
"grad_norm": 87.77210235595703,
"learning_rate": 4.261835518370992e-06,
"loss": 0.7801,
"step": 934
},
{
"epoch": 0.7868714496107722,
"grad_norm": 86.31080627441406,
"learning_rate": 4.2602504609785526e-06,
"loss": 0.9737,
"step": 935
},
{
"epoch": 0.7877130233536713,
"grad_norm": 84.96116638183594,
"learning_rate": 4.2586639990612015e-06,
"loss": 0.7804,
"step": 936
},
{
"epoch": 0.7885545970965706,
"grad_norm": 74.81805419921875,
"learning_rate": 4.257076133884795e-06,
"loss": 0.8313,
"step": 937
},
{
"epoch": 0.7893961708394698,
"grad_norm": 98.99002838134766,
"learning_rate": 4.255486866716311e-06,
"loss": 0.8527,
"step": 938
},
{
"epoch": 0.7902377445823691,
"grad_norm": 92.40936279296875,
"learning_rate": 4.253896198823844e-06,
"loss": 0.9288,
"step": 939
},
{
"epoch": 0.7910793183252682,
"grad_norm": 99.4084243774414,
"learning_rate": 4.252304131476607e-06,
"loss": 0.9823,
"step": 940
},
{
"epoch": 0.7919208920681675,
"grad_norm": 72.1796646118164,
"learning_rate": 4.2507106659449295e-06,
"loss": 0.7174,
"step": 941
},
{
"epoch": 0.7927624658110667,
"grad_norm": 71.70301055908203,
"learning_rate": 4.249115803500257e-06,
"loss": 0.8033,
"step": 942
},
{
"epoch": 0.793604039553966,
"grad_norm": 68.06544494628906,
"learning_rate": 4.247519545415149e-06,
"loss": 0.7506,
"step": 943
},
{
"epoch": 0.7944456132968651,
"grad_norm": 62.31113052368164,
"learning_rate": 4.245921892963279e-06,
"loss": 0.6848,
"step": 944
},
{
"epoch": 0.7952871870397643,
"grad_norm": 84.2646255493164,
"learning_rate": 4.2443228474194335e-06,
"loss": 0.9001,
"step": 945
},
{
"epoch": 0.7961287607826636,
"grad_norm": 87.14163208007812,
"learning_rate": 4.24272241005951e-06,
"loss": 1.0412,
"step": 946
},
{
"epoch": 0.7969703345255628,
"grad_norm": 89.12403869628906,
"learning_rate": 4.241120582160516e-06,
"loss": 0.8883,
"step": 947
},
{
"epoch": 0.797811908268462,
"grad_norm": 79.68348693847656,
"learning_rate": 4.239517365000569e-06,
"loss": 0.9308,
"step": 948
},
{
"epoch": 0.7986534820113612,
"grad_norm": 79.94398498535156,
"learning_rate": 4.237912759858896e-06,
"loss": 0.8429,
"step": 949
},
{
"epoch": 0.7994950557542605,
"grad_norm": 73.92453002929688,
"learning_rate": 4.236306768015831e-06,
"loss": 0.8694,
"step": 950
},
{
"epoch": 0.8003366294971597,
"grad_norm": 76.87979888916016,
"learning_rate": 4.234699390752812e-06,
"loss": 0.8229,
"step": 951
},
{
"epoch": 0.801178203240059,
"grad_norm": 66.13666534423828,
"learning_rate": 4.233090629352388e-06,
"loss": 0.7658,
"step": 952
},
{
"epoch": 0.8020197769829581,
"grad_norm": 77.92546844482422,
"learning_rate": 4.231480485098205e-06,
"loss": 0.7735,
"step": 953
},
{
"epoch": 0.8028613507258574,
"grad_norm": 91.31938934326172,
"learning_rate": 4.229868959275019e-06,
"loss": 1.0128,
"step": 954
},
{
"epoch": 0.8037029244687566,
"grad_norm": 103.90299224853516,
"learning_rate": 4.228256053168685e-06,
"loss": 1.1743,
"step": 955
},
{
"epoch": 0.8045444982116557,
"grad_norm": 80.80589294433594,
"learning_rate": 4.226641768066161e-06,
"loss": 0.8783,
"step": 956
},
{
"epoch": 0.805386071954555,
"grad_norm": 93.89417266845703,
"learning_rate": 4.225026105255503e-06,
"loss": 1.0368,
"step": 957
},
{
"epoch": 0.8062276456974542,
"grad_norm": 72.91015625,
"learning_rate": 4.223409066025867e-06,
"loss": 0.8157,
"step": 958
},
{
"epoch": 0.8070692194403535,
"grad_norm": 67.31309509277344,
"learning_rate": 4.221790651667511e-06,
"loss": 0.7382,
"step": 959
},
{
"epoch": 0.8079107931832527,
"grad_norm": 80.1012191772461,
"learning_rate": 4.220170863471785e-06,
"loss": 0.9073,
"step": 960
},
{
"epoch": 0.8087523669261519,
"grad_norm": 82.77555847167969,
"learning_rate": 4.218549702731138e-06,
"loss": 0.9337,
"step": 961
},
{
"epoch": 0.8095939406690511,
"grad_norm": 84.62960815429688,
"learning_rate": 4.216927170739113e-06,
"loss": 0.8536,
"step": 962
},
{
"epoch": 0.8104355144119504,
"grad_norm": 72.1750259399414,
"learning_rate": 4.2153032687903475e-06,
"loss": 0.8958,
"step": 963
},
{
"epoch": 0.8112770881548496,
"grad_norm": 73.40558624267578,
"learning_rate": 4.213677998180574e-06,
"loss": 0.8339,
"step": 964
},
{
"epoch": 0.8121186618977488,
"grad_norm": 99.02385711669922,
"learning_rate": 4.212051360206614e-06,
"loss": 0.7137,
"step": 965
},
{
"epoch": 0.812960235640648,
"grad_norm": 116.62301635742188,
"learning_rate": 4.21042335616638e-06,
"loss": 1.3858,
"step": 966
},
{
"epoch": 0.8138018093835472,
"grad_norm": 72.70298767089844,
"learning_rate": 4.208793987358879e-06,
"loss": 0.6905,
"step": 967
},
{
"epoch": 0.8146433831264465,
"grad_norm": 78.67459106445312,
"learning_rate": 4.207163255084202e-06,
"loss": 0.9802,
"step": 968
},
{
"epoch": 0.8154849568693456,
"grad_norm": 110.66291809082031,
"learning_rate": 4.205531160643531e-06,
"loss": 0.6545,
"step": 969
},
{
"epoch": 0.8163265306122449,
"grad_norm": 71.75804138183594,
"learning_rate": 4.203897705339132e-06,
"loss": 0.9127,
"step": 970
},
{
"epoch": 0.8171681043551441,
"grad_norm": 80.5187759399414,
"learning_rate": 4.20226289047436e-06,
"loss": 0.7607,
"step": 971
},
{
"epoch": 0.8180096780980434,
"grad_norm": 75.0357666015625,
"learning_rate": 4.200626717353653e-06,
"loss": 0.7147,
"step": 972
},
{
"epoch": 0.8188512518409425,
"grad_norm": 73.03073120117188,
"learning_rate": 4.198989187282532e-06,
"loss": 0.5297,
"step": 973
},
{
"epoch": 0.8196928255838418,
"grad_norm": 84.6255111694336,
"learning_rate": 4.197350301567603e-06,
"loss": 0.8926,
"step": 974
},
{
"epoch": 0.820534399326741,
"grad_norm": 66.8243637084961,
"learning_rate": 4.195710061516552e-06,
"loss": 0.8253,
"step": 975
},
{
"epoch": 0.8213759730696403,
"grad_norm": 57.61470031738281,
"learning_rate": 4.194068468438147e-06,
"loss": 0.6581,
"step": 976
},
{
"epoch": 0.8222175468125394,
"grad_norm": 66.89012908935547,
"learning_rate": 4.192425523642234e-06,
"loss": 0.81,
"step": 977
},
{
"epoch": 0.8230591205554386,
"grad_norm": 76.0096435546875,
"learning_rate": 4.1907812284397395e-06,
"loss": 0.8723,
"step": 978
},
{
"epoch": 0.8239006942983379,
"grad_norm": 80.26139831542969,
"learning_rate": 4.189135584142664e-06,
"loss": 0.8389,
"step": 979
},
{
"epoch": 0.8247422680412371,
"grad_norm": 71.46955871582031,
"learning_rate": 4.187488592064089e-06,
"loss": 0.7342,
"step": 980
},
{
"epoch": 0.8255838417841364,
"grad_norm": 68.54132843017578,
"learning_rate": 4.185840253518168e-06,
"loss": 0.9466,
"step": 981
},
{
"epoch": 0.8264254155270355,
"grad_norm": 74.63121032714844,
"learning_rate": 4.18419056982013e-06,
"loss": 0.9002,
"step": 982
},
{
"epoch": 0.8272669892699348,
"grad_norm": 68.22480010986328,
"learning_rate": 4.182539542286279e-06,
"loss": 0.7849,
"step": 983
},
{
"epoch": 0.828108563012834,
"grad_norm": 67.76121520996094,
"learning_rate": 4.1808871722339875e-06,
"loss": 0.4417,
"step": 984
},
{
"epoch": 0.8289501367557333,
"grad_norm": 80.49524688720703,
"learning_rate": 4.179233460981702e-06,
"loss": 0.9594,
"step": 985
},
{
"epoch": 0.8297917104986324,
"grad_norm": 82.17792510986328,
"learning_rate": 4.177578409848939e-06,
"loss": 0.842,
"step": 986
},
{
"epoch": 0.8306332842415317,
"grad_norm": 81.26715850830078,
"learning_rate": 4.1759220201562835e-06,
"loss": 0.9353,
"step": 987
},
{
"epoch": 0.8314748579844309,
"grad_norm": 77.98991394042969,
"learning_rate": 4.174264293225389e-06,
"loss": 0.753,
"step": 988
},
{
"epoch": 0.83231643172733,
"grad_norm": 83.93816375732422,
"learning_rate": 4.172605230378975e-06,
"loss": 0.9018,
"step": 989
},
{
"epoch": 0.8331580054702293,
"grad_norm": 74.68817138671875,
"learning_rate": 4.170944832940827e-06,
"loss": 0.6308,
"step": 990
},
{
"epoch": 0.8339995792131285,
"grad_norm": 78.59561157226562,
"learning_rate": 4.169283102235797e-06,
"loss": 0.867,
"step": 991
},
{
"epoch": 0.8348411529560278,
"grad_norm": 65.0259780883789,
"learning_rate": 4.167620039589799e-06,
"loss": 0.8615,
"step": 992
},
{
"epoch": 0.835682726698927,
"grad_norm": 79.45830535888672,
"learning_rate": 4.1659556463298124e-06,
"loss": 0.9266,
"step": 993
},
{
"epoch": 0.8365243004418262,
"grad_norm": 81.86888122558594,
"learning_rate": 4.164289923783874e-06,
"loss": 0.8939,
"step": 994
},
{
"epoch": 0.8373658741847254,
"grad_norm": 81.57584381103516,
"learning_rate": 4.162622873281086e-06,
"loss": 0.8272,
"step": 995
},
{
"epoch": 0.8382074479276247,
"grad_norm": 68.12931823730469,
"learning_rate": 4.160954496151607e-06,
"loss": 0.778,
"step": 996
},
{
"epoch": 0.8390490216705239,
"grad_norm": 77.3565673828125,
"learning_rate": 4.159284793726655e-06,
"loss": 0.695,
"step": 997
},
{
"epoch": 0.8398905954134231,
"grad_norm": 72.69764709472656,
"learning_rate": 4.157613767338505e-06,
"loss": 0.8958,
"step": 998
},
{
"epoch": 0.8407321691563223,
"grad_norm": 72.09540557861328,
"learning_rate": 4.15594141832049e-06,
"loss": 0.6611,
"step": 999
},
{
"epoch": 0.8415737428992216,
"grad_norm": 66.83956909179688,
"learning_rate": 4.154267748006996e-06,
"loss": 0.7814,
"step": 1000
},
{
"epoch": 0.8424153166421208,
"grad_norm": 69.9083251953125,
"learning_rate": 4.152592757733467e-06,
"loss": 0.7695,
"step": 1001
},
{
"epoch": 0.8432568903850199,
"grad_norm": 81.96198272705078,
"learning_rate": 4.150916448836395e-06,
"loss": 0.8362,
"step": 1002
},
{
"epoch": 0.8440984641279192,
"grad_norm": 60.79277420043945,
"learning_rate": 4.149238822653329e-06,
"loss": 0.6429,
"step": 1003
},
{
"epoch": 0.8449400378708184,
"grad_norm": 73.13953399658203,
"learning_rate": 4.147559880522866e-06,
"loss": 0.8873,
"step": 1004
},
{
"epoch": 0.8457816116137177,
"grad_norm": 75.53364562988281,
"learning_rate": 4.145879623784653e-06,
"loss": 0.7998,
"step": 1005
},
{
"epoch": 0.8466231853566168,
"grad_norm": 65.25211334228516,
"learning_rate": 4.144198053779389e-06,
"loss": 0.839,
"step": 1006
},
{
"epoch": 0.8474647590995161,
"grad_norm": 74.82688903808594,
"learning_rate": 4.142515171848819e-06,
"loss": 1.0005,
"step": 1007
},
{
"epoch": 0.8483063328424153,
"grad_norm": 65.3867416381836,
"learning_rate": 4.140830979335733e-06,
"loss": 0.5801,
"step": 1008
},
{
"epoch": 0.8491479065853146,
"grad_norm": 67.73045349121094,
"learning_rate": 4.139145477583968e-06,
"loss": 0.8043,
"step": 1009
},
{
"epoch": 0.8499894803282138,
"grad_norm": 92.07035064697266,
"learning_rate": 4.137458667938407e-06,
"loss": 0.888,
"step": 1010
},
{
"epoch": 0.850831054071113,
"grad_norm": 73.86541748046875,
"learning_rate": 4.135770551744977e-06,
"loss": 0.8436,
"step": 1011
},
{
"epoch": 0.8516726278140122,
"grad_norm": 83.31813049316406,
"learning_rate": 4.134081130350644e-06,
"loss": 1.0081,
"step": 1012
},
{
"epoch": 0.8525142015569114,
"grad_norm": 76.76337432861328,
"learning_rate": 4.132390405103419e-06,
"loss": 0.7978,
"step": 1013
},
{
"epoch": 0.8533557752998107,
"grad_norm": 96.06390380859375,
"learning_rate": 4.13069837735235e-06,
"loss": 0.9046,
"step": 1014
},
{
"epoch": 0.8541973490427098,
"grad_norm": 74.8203353881836,
"learning_rate": 4.1290050484475285e-06,
"loss": 0.9066,
"step": 1015
},
{
"epoch": 0.8550389227856091,
"grad_norm": 76.6991958618164,
"learning_rate": 4.127310419740081e-06,
"loss": 0.8842,
"step": 1016
},
{
"epoch": 0.8558804965285083,
"grad_norm": 59.34732437133789,
"learning_rate": 4.12561449258217e-06,
"loss": 0.7031,
"step": 1017
},
{
"epoch": 0.8567220702714076,
"grad_norm": 71.81965637207031,
"learning_rate": 4.123917268326999e-06,
"loss": 0.8189,
"step": 1018
},
{
"epoch": 0.8575636440143067,
"grad_norm": 58.81732940673828,
"learning_rate": 4.122218748328802e-06,
"loss": 0.8027,
"step": 1019
},
{
"epoch": 0.858405217757206,
"grad_norm": 75.25838470458984,
"learning_rate": 4.12051893394285e-06,
"loss": 0.8217,
"step": 1020
},
{
"epoch": 0.8592467915001052,
"grad_norm": 71.95877075195312,
"learning_rate": 4.118817826525443e-06,
"loss": 0.6275,
"step": 1021
},
{
"epoch": 0.8600883652430045,
"grad_norm": 82.2073974609375,
"learning_rate": 4.1171154274339164e-06,
"loss": 0.9913,
"step": 1022
},
{
"epoch": 0.8609299389859036,
"grad_norm": 80.4325942993164,
"learning_rate": 4.1154117380266346e-06,
"loss": 0.6675,
"step": 1023
},
{
"epoch": 0.8617715127288028,
"grad_norm": 79.63078308105469,
"learning_rate": 4.113706759662992e-06,
"loss": 0.7476,
"step": 1024
},
{
"epoch": 0.8626130864717021,
"grad_norm": 124.51466369628906,
"learning_rate": 4.112000493703411e-06,
"loss": 0.7556,
"step": 1025
},
{
"epoch": 0.8634546602146013,
"grad_norm": 69.8863525390625,
"learning_rate": 4.11029294150934e-06,
"loss": 0.8307,
"step": 1026
},
{
"epoch": 0.8642962339575005,
"grad_norm": 151.2084197998047,
"learning_rate": 4.108584104443259e-06,
"loss": 0.7547,
"step": 1027
},
{
"epoch": 0.8651378077003997,
"grad_norm": 58.940773010253906,
"learning_rate": 4.106873983868667e-06,
"loss": 0.8035,
"step": 1028
},
{
"epoch": 0.865979381443299,
"grad_norm": 68.62104034423828,
"learning_rate": 4.10516258115009e-06,
"loss": 0.7035,
"step": 1029
},
{
"epoch": 0.8668209551861982,
"grad_norm": 99.0032730102539,
"learning_rate": 4.1034498976530755e-06,
"loss": 1.0287,
"step": 1030
},
{
"epoch": 0.8676625289290975,
"grad_norm": 89.32034301757812,
"learning_rate": 4.101735934744197e-06,
"loss": 0.9512,
"step": 1031
},
{
"epoch": 0.8685041026719966,
"grad_norm": 59.35560989379883,
"learning_rate": 4.100020693791043e-06,
"loss": 0.7646,
"step": 1032
},
{
"epoch": 0.8693456764148959,
"grad_norm": 59.45350646972656,
"learning_rate": 4.098304176162228e-06,
"loss": 0.697,
"step": 1033
},
{
"epoch": 0.8701872501577951,
"grad_norm": 86.22278594970703,
"learning_rate": 4.09658638322738e-06,
"loss": 1.0046,
"step": 1034
},
{
"epoch": 0.8710288239006942,
"grad_norm": 69.21746063232422,
"learning_rate": 4.094867316357146e-06,
"loss": 0.6598,
"step": 1035
},
{
"epoch": 0.8718703976435935,
"grad_norm": 71.82403564453125,
"learning_rate": 4.093146976923191e-06,
"loss": 0.7445,
"step": 1036
},
{
"epoch": 0.8727119713864927,
"grad_norm": 72.35151672363281,
"learning_rate": 4.091425366298195e-06,
"loss": 0.82,
"step": 1037
},
{
"epoch": 0.873553545129392,
"grad_norm": 79.09820556640625,
"learning_rate": 4.08970248585585e-06,
"loss": 0.7896,
"step": 1038
},
{
"epoch": 0.8743951188722912,
"grad_norm": 70.99044036865234,
"learning_rate": 4.087978336970864e-06,
"loss": 0.7507,
"step": 1039
},
{
"epoch": 0.8752366926151904,
"grad_norm": 64.83211517333984,
"learning_rate": 4.0862529210189566e-06,
"loss": 0.8278,
"step": 1040
},
{
"epoch": 0.8760782663580896,
"grad_norm": 88.5916519165039,
"learning_rate": 4.084526239376857e-06,
"loss": 0.8102,
"step": 1041
},
{
"epoch": 0.8769198401009889,
"grad_norm": 100.17524719238281,
"learning_rate": 4.082798293422305e-06,
"loss": 0.9703,
"step": 1042
},
{
"epoch": 0.8777614138438881,
"grad_norm": 80.35630798339844,
"learning_rate": 4.08106908453405e-06,
"loss": 0.8869,
"step": 1043
},
{
"epoch": 0.8786029875867873,
"grad_norm": 72.74226379394531,
"learning_rate": 4.079338614091848e-06,
"loss": 0.73,
"step": 1044
},
{
"epoch": 0.8794445613296865,
"grad_norm": 87.65748596191406,
"learning_rate": 4.077606883476462e-06,
"loss": 0.9816,
"step": 1045
},
{
"epoch": 0.8802861350725857,
"grad_norm": 64.14987182617188,
"learning_rate": 4.075873894069661e-06,
"loss": 0.6598,
"step": 1046
},
{
"epoch": 0.881127708815485,
"grad_norm": 62.39348220825195,
"learning_rate": 4.074139647254218e-06,
"loss": 0.6845,
"step": 1047
},
{
"epoch": 0.8819692825583841,
"grad_norm": 57.6712532043457,
"learning_rate": 4.072404144413909e-06,
"loss": 0.6839,
"step": 1048
},
{
"epoch": 0.8828108563012834,
"grad_norm": 65.15045928955078,
"learning_rate": 4.070667386933512e-06,
"loss": 0.874,
"step": 1049
},
{
"epoch": 0.8836524300441826,
"grad_norm": 84.79566192626953,
"learning_rate": 4.068929376198808e-06,
"loss": 0.8565,
"step": 1050
},
{
"epoch": 0.8844940037870819,
"grad_norm": 85.54796600341797,
"learning_rate": 4.067190113596576e-06,
"loss": 1.0899,
"step": 1051
},
{
"epoch": 0.885335577529981,
"grad_norm": 66.54473876953125,
"learning_rate": 4.065449600514594e-06,
"loss": 0.7514,
"step": 1052
},
{
"epoch": 0.8861771512728803,
"grad_norm": 70.7523422241211,
"learning_rate": 4.0637078383416385e-06,
"loss": 0.7151,
"step": 1053
},
{
"epoch": 0.8870187250157795,
"grad_norm": 78.33436584472656,
"learning_rate": 4.061964828467482e-06,
"loss": 0.8094,
"step": 1054
},
{
"epoch": 0.8878602987586788,
"grad_norm": 63.319393157958984,
"learning_rate": 4.060220572282892e-06,
"loss": 0.7593,
"step": 1055
},
{
"epoch": 0.888701872501578,
"grad_norm": 77.5903549194336,
"learning_rate": 4.058475071179635e-06,
"loss": 0.7143,
"step": 1056
},
{
"epoch": 0.8895434462444771,
"grad_norm": 77.97570037841797,
"learning_rate": 4.056728326550463e-06,
"loss": 0.811,
"step": 1057
},
{
"epoch": 0.8903850199873764,
"grad_norm": 80.57714080810547,
"learning_rate": 4.0549803397891265e-06,
"loss": 0.8381,
"step": 1058
},
{
"epoch": 0.8912265937302756,
"grad_norm": 62.823917388916016,
"learning_rate": 4.053231112290365e-06,
"loss": 0.7561,
"step": 1059
},
{
"epoch": 0.8920681674731749,
"grad_norm": 68.00716400146484,
"learning_rate": 4.051480645449909e-06,
"loss": 0.7462,
"step": 1060
},
{
"epoch": 0.892909741216074,
"grad_norm": 60.848243713378906,
"learning_rate": 4.049728940664474e-06,
"loss": 0.6313,
"step": 1061
},
{
"epoch": 0.8937513149589733,
"grad_norm": 69.16812896728516,
"learning_rate": 4.04797599933177e-06,
"loss": 0.8213,
"step": 1062
},
{
"epoch": 0.8945928887018725,
"grad_norm": 88.9859848022461,
"learning_rate": 4.046221822850487e-06,
"loss": 0.9066,
"step": 1063
},
{
"epoch": 0.8954344624447718,
"grad_norm": 135.78160095214844,
"learning_rate": 4.044466412620305e-06,
"loss": 0.7938,
"step": 1064
},
{
"epoch": 0.8962760361876709,
"grad_norm": 89.73754119873047,
"learning_rate": 4.042709770041885e-06,
"loss": 0.926,
"step": 1065
},
{
"epoch": 0.8971176099305702,
"grad_norm": 68.3105239868164,
"learning_rate": 4.040951896516875e-06,
"loss": 0.671,
"step": 1066
},
{
"epoch": 0.8979591836734694,
"grad_norm": 57.36161422729492,
"learning_rate": 4.039192793447902e-06,
"loss": 0.6333,
"step": 1067
},
{
"epoch": 0.8988007574163686,
"grad_norm": 91.2534408569336,
"learning_rate": 4.037432462238578e-06,
"loss": 0.9629,
"step": 1068
},
{
"epoch": 0.8996423311592678,
"grad_norm": 97.27592468261719,
"learning_rate": 4.035670904293488e-06,
"loss": 1.0401,
"step": 1069
},
{
"epoch": 0.900483904902167,
"grad_norm": 96.43426513671875,
"learning_rate": 4.033908121018204e-06,
"loss": 0.9545,
"step": 1070
},
{
"epoch": 0.9013254786450663,
"grad_norm": 66.45618438720703,
"learning_rate": 4.03214411381927e-06,
"loss": 0.7543,
"step": 1071
},
{
"epoch": 0.9021670523879655,
"grad_norm": 107.90736389160156,
"learning_rate": 4.030378884104209e-06,
"loss": 0.7016,
"step": 1072
},
{
"epoch": 0.9030086261308647,
"grad_norm": 77.58516693115234,
"learning_rate": 4.0286124332815186e-06,
"loss": 0.8714,
"step": 1073
},
{
"epoch": 0.9038501998737639,
"grad_norm": 65.78289794921875,
"learning_rate": 4.0268447627606725e-06,
"loss": 0.5679,
"step": 1074
},
{
"epoch": 0.9046917736166632,
"grad_norm": 71.64932250976562,
"learning_rate": 4.025075873952113e-06,
"loss": 0.8684,
"step": 1075
},
{
"epoch": 0.9055333473595624,
"grad_norm": 69.41444396972656,
"learning_rate": 4.023305768267261e-06,
"loss": 0.7746,
"step": 1076
},
{
"epoch": 0.9063749211024617,
"grad_norm": 80.02897644042969,
"learning_rate": 4.021534447118502e-06,
"loss": 0.8513,
"step": 1077
},
{
"epoch": 0.9072164948453608,
"grad_norm": 69.74525451660156,
"learning_rate": 4.019761911919198e-06,
"loss": 0.823,
"step": 1078
},
{
"epoch": 0.90805806858826,
"grad_norm": 57.38936996459961,
"learning_rate": 4.017988164083673e-06,
"loss": 0.6093,
"step": 1079
},
{
"epoch": 0.9088996423311593,
"grad_norm": 68.37493133544922,
"learning_rate": 4.016213205027222e-06,
"loss": 0.7365,
"step": 1080
},
{
"epoch": 0.9097412160740584,
"grad_norm": 61.99602127075195,
"learning_rate": 4.0144370361661075e-06,
"loss": 0.7312,
"step": 1081
},
{
"epoch": 0.9105827898169577,
"grad_norm": 73.24932861328125,
"learning_rate": 4.012659658917555e-06,
"loss": 0.7137,
"step": 1082
},
{
"epoch": 0.9114243635598569,
"grad_norm": 73.64488983154297,
"learning_rate": 4.010881074699755e-06,
"loss": 0.8696,
"step": 1083
},
{
"epoch": 0.9122659373027562,
"grad_norm": 82.7219467163086,
"learning_rate": 4.009101284931861e-06,
"loss": 0.8684,
"step": 1084
},
{
"epoch": 0.9131075110456554,
"grad_norm": 86.55023193359375,
"learning_rate": 4.0073202910339895e-06,
"loss": 0.9902,
"step": 1085
},
{
"epoch": 0.9139490847885546,
"grad_norm": 61.76776123046875,
"learning_rate": 4.005538094427216e-06,
"loss": 0.767,
"step": 1086
},
{
"epoch": 0.9147906585314538,
"grad_norm": 70.59973907470703,
"learning_rate": 4.003754696533576e-06,
"loss": 0.8766,
"step": 1087
},
{
"epoch": 0.9156322322743531,
"grad_norm": 69.00643920898438,
"learning_rate": 4.001970098776066e-06,
"loss": 0.9194,
"step": 1088
},
{
"epoch": 0.9164738060172523,
"grad_norm": 88.84870147705078,
"learning_rate": 4.000184302578636e-06,
"loss": 1.0529,
"step": 1089
},
{
"epoch": 0.9173153797601514,
"grad_norm": 88.0756607055664,
"learning_rate": 3.998397309366194e-06,
"loss": 0.891,
"step": 1090
},
{
"epoch": 0.9181569535030507,
"grad_norm": 79.1492691040039,
"learning_rate": 3.996609120564605e-06,
"loss": 0.7358,
"step": 1091
},
{
"epoch": 0.9189985272459499,
"grad_norm": 76.13194274902344,
"learning_rate": 3.9948197376006855e-06,
"loss": 0.8981,
"step": 1092
},
{
"epoch": 0.9198401009888492,
"grad_norm": 69.00740051269531,
"learning_rate": 3.993029161902205e-06,
"loss": 0.8097,
"step": 1093
},
{
"epoch": 0.9206816747317483,
"grad_norm": 75.37550354003906,
"learning_rate": 3.991237394897886e-06,
"loss": 0.9167,
"step": 1094
},
{
"epoch": 0.9215232484746476,
"grad_norm": 86.60401916503906,
"learning_rate": 3.9894444380174005e-06,
"loss": 1.0947,
"step": 1095
},
{
"epoch": 0.9223648222175468,
"grad_norm": 68.166015625,
"learning_rate": 3.987650292691371e-06,
"loss": 0.7399,
"step": 1096
},
{
"epoch": 0.9232063959604461,
"grad_norm": 88.67655181884766,
"learning_rate": 3.985854960351365e-06,
"loss": 0.9703,
"step": 1097
},
{
"epoch": 0.9240479697033452,
"grad_norm": 67.01685333251953,
"learning_rate": 3.9840584424299034e-06,
"loss": 0.7215,
"step": 1098
},
{
"epoch": 0.9248895434462445,
"grad_norm": 79.09455871582031,
"learning_rate": 3.982260740360447e-06,
"loss": 0.8123,
"step": 1099
},
{
"epoch": 0.9257311171891437,
"grad_norm": 74.57783508300781,
"learning_rate": 3.980461855577403e-06,
"loss": 0.9284,
"step": 1100
},
{
"epoch": 0.926572690932043,
"grad_norm": 78.28739929199219,
"learning_rate": 3.978661789516126e-06,
"loss": 0.9738,
"step": 1101
},
{
"epoch": 0.9274142646749421,
"grad_norm": 69.92387390136719,
"learning_rate": 3.976860543612907e-06,
"loss": 0.7838,
"step": 1102
},
{
"epoch": 0.9282558384178413,
"grad_norm": 69.45232391357422,
"learning_rate": 3.975058119304984e-06,
"loss": 0.8191,
"step": 1103
},
{
"epoch": 0.9290974121607406,
"grad_norm": 85.00932312011719,
"learning_rate": 3.973254518030532e-06,
"loss": 0.8866,
"step": 1104
},
{
"epoch": 0.9299389859036398,
"grad_norm": 68.20960235595703,
"learning_rate": 3.971449741228666e-06,
"loss": 0.8838,
"step": 1105
},
{
"epoch": 0.930780559646539,
"grad_norm": 68.52924346923828,
"learning_rate": 3.969643790339441e-06,
"loss": 0.7283,
"step": 1106
},
{
"epoch": 0.9316221333894382,
"grad_norm": 85.05404663085938,
"learning_rate": 3.967836666803844e-06,
"loss": 0.8443,
"step": 1107
},
{
"epoch": 0.9324637071323375,
"grad_norm": 70.1175765991211,
"learning_rate": 3.966028372063801e-06,
"loss": 0.7288,
"step": 1108
},
{
"epoch": 0.9333052808752367,
"grad_norm": 72.65882873535156,
"learning_rate": 3.9642189075621726e-06,
"loss": 0.8062,
"step": 1109
},
{
"epoch": 0.934146854618136,
"grad_norm": 71.01957702636719,
"learning_rate": 3.962408274742753e-06,
"loss": 0.9103,
"step": 1110
},
{
"epoch": 0.9349884283610351,
"grad_norm": 86.82347106933594,
"learning_rate": 3.960596475050266e-06,
"loss": 0.6355,
"step": 1111
},
{
"epoch": 0.9358300021039344,
"grad_norm": 93.42919158935547,
"learning_rate": 3.95878350993037e-06,
"loss": 0.7063,
"step": 1112
},
{
"epoch": 0.9366715758468336,
"grad_norm": 92.69883728027344,
"learning_rate": 3.956969380829651e-06,
"loss": 0.8995,
"step": 1113
},
{
"epoch": 0.9375131495897328,
"grad_norm": 63.439788818359375,
"learning_rate": 3.955154089195624e-06,
"loss": 0.7768,
"step": 1114
},
{
"epoch": 0.938354723332632,
"grad_norm": 79.63054656982422,
"learning_rate": 3.953337636476731e-06,
"loss": 0.8086,
"step": 1115
},
{
"epoch": 0.9391962970755312,
"grad_norm": 79.53690338134766,
"learning_rate": 3.951520024122344e-06,
"loss": 0.8836,
"step": 1116
},
{
"epoch": 0.9400378708184305,
"grad_norm": 63.9859619140625,
"learning_rate": 3.949701253582755e-06,
"loss": 0.7912,
"step": 1117
},
{
"epoch": 0.9408794445613297,
"grad_norm": 66.5926742553711,
"learning_rate": 3.947881326309184e-06,
"loss": 0.8147,
"step": 1118
},
{
"epoch": 0.9417210183042289,
"grad_norm": 79.54521179199219,
"learning_rate": 3.946060243753773e-06,
"loss": 0.9673,
"step": 1119
},
{
"epoch": 0.9425625920471281,
"grad_norm": 86.67349243164062,
"learning_rate": 3.944238007369585e-06,
"loss": 0.8635,
"step": 1120
},
{
"epoch": 0.9434041657900274,
"grad_norm": 57.8299674987793,
"learning_rate": 3.942414618610604e-06,
"loss": 0.5855,
"step": 1121
},
{
"epoch": 0.9442457395329266,
"grad_norm": 81.97209167480469,
"learning_rate": 3.940590078931734e-06,
"loss": 0.7143,
"step": 1122
},
{
"epoch": 0.9450873132758258,
"grad_norm": 53.280616760253906,
"learning_rate": 3.938764389788798e-06,
"loss": 0.6815,
"step": 1123
},
{
"epoch": 0.945928887018725,
"grad_norm": 90.65757751464844,
"learning_rate": 3.936937552638534e-06,
"loss": 1.0972,
"step": 1124
},
{
"epoch": 0.9467704607616242,
"grad_norm": 97.28402709960938,
"learning_rate": 3.935109568938598e-06,
"loss": 0.7802,
"step": 1125
},
{
"epoch": 0.9476120345045235,
"grad_norm": 62.99158477783203,
"learning_rate": 3.9332804401475585e-06,
"loss": 0.6851,
"step": 1126
},
{
"epoch": 0.9484536082474226,
"grad_norm": 75.19110870361328,
"learning_rate": 3.9314501677249015e-06,
"loss": 0.7416,
"step": 1127
},
{
"epoch": 0.9492951819903219,
"grad_norm": 86.4506607055664,
"learning_rate": 3.929618753131023e-06,
"loss": 0.9711,
"step": 1128
},
{
"epoch": 0.9501367557332211,
"grad_norm": 68.20625305175781,
"learning_rate": 3.927786197827228e-06,
"loss": 0.7296,
"step": 1129
},
{
"epoch": 0.9509783294761204,
"grad_norm": 79.4578857421875,
"learning_rate": 3.9259525032757375e-06,
"loss": 0.9495,
"step": 1130
},
{
"epoch": 0.9518199032190195,
"grad_norm": 59.97657775878906,
"learning_rate": 3.924117670939676e-06,
"loss": 0.5803,
"step": 1131
},
{
"epoch": 0.9526614769619188,
"grad_norm": 69.94535064697266,
"learning_rate": 3.92228170228308e-06,
"loss": 0.8632,
"step": 1132
},
{
"epoch": 0.953503050704818,
"grad_norm": 70.57887268066406,
"learning_rate": 3.920444598770889e-06,
"loss": 0.8572,
"step": 1133
},
{
"epoch": 0.9543446244477173,
"grad_norm": 72.97533416748047,
"learning_rate": 3.918606361868952e-06,
"loss": 0.7592,
"step": 1134
},
{
"epoch": 0.9551861981906165,
"grad_norm": 76.23296356201172,
"learning_rate": 3.916766993044018e-06,
"loss": 0.774,
"step": 1135
},
{
"epoch": 0.9560277719335156,
"grad_norm": 69.50141143798828,
"learning_rate": 3.914926493763743e-06,
"loss": 0.8404,
"step": 1136
},
{
"epoch": 0.9568693456764149,
"grad_norm": 61.13656997680664,
"learning_rate": 3.9130848654966835e-06,
"loss": 0.8523,
"step": 1137
},
{
"epoch": 0.9577109194193141,
"grad_norm": 76.27356719970703,
"learning_rate": 3.911242109712296e-06,
"loss": 0.6067,
"step": 1138
},
{
"epoch": 0.9585524931622134,
"grad_norm": 70.06895446777344,
"learning_rate": 3.909398227880937e-06,
"loss": 0.5791,
"step": 1139
},
{
"epoch": 0.9593940669051125,
"grad_norm": 58.804359436035156,
"learning_rate": 3.907553221473863e-06,
"loss": 0.7047,
"step": 1140
},
{
"epoch": 0.9602356406480118,
"grad_norm": 91.75293731689453,
"learning_rate": 3.9057070919632265e-06,
"loss": 1.0091,
"step": 1141
},
{
"epoch": 0.961077214390911,
"grad_norm": 69.82617950439453,
"learning_rate": 3.903859840822076e-06,
"loss": 0.8492,
"step": 1142
},
{
"epoch": 0.9619187881338103,
"grad_norm": 50.22442626953125,
"learning_rate": 3.902011469524355e-06,
"loss": 0.6491,
"step": 1143
},
{
"epoch": 0.9627603618767094,
"grad_norm": 74.7395248413086,
"learning_rate": 3.900161979544901e-06,
"loss": 1.0323,
"step": 1144
},
{
"epoch": 0.9636019356196087,
"grad_norm": 70.21725463867188,
"learning_rate": 3.898311372359445e-06,
"loss": 0.6371,
"step": 1145
},
{
"epoch": 0.9644435093625079,
"grad_norm": 79.02413177490234,
"learning_rate": 3.896459649444607e-06,
"loss": 0.8239,
"step": 1146
},
{
"epoch": 0.9652850831054071,
"grad_norm": 83.71379852294922,
"learning_rate": 3.894606812277901e-06,
"loss": 0.9468,
"step": 1147
},
{
"epoch": 0.9661266568483063,
"grad_norm": 64.89826202392578,
"learning_rate": 3.892752862337728e-06,
"loss": 0.7002,
"step": 1148
},
{
"epoch": 0.9669682305912055,
"grad_norm": 66.54814147949219,
"learning_rate": 3.890897801103375e-06,
"loss": 0.649,
"step": 1149
},
{
"epoch": 0.9678098043341048,
"grad_norm": 101.90099334716797,
"learning_rate": 3.889041630055017e-06,
"loss": 0.7801,
"step": 1150
},
{
"epoch": 0.968651378077004,
"grad_norm": 63.44615936279297,
"learning_rate": 3.887184350673718e-06,
"loss": 0.6731,
"step": 1151
},
{
"epoch": 0.9694929518199032,
"grad_norm": 58.71804428100586,
"learning_rate": 3.885325964441422e-06,
"loss": 0.7604,
"step": 1152
},
{
"epoch": 0.9703345255628024,
"grad_norm": 61.444766998291016,
"learning_rate": 3.883466472840958e-06,
"loss": 0.7737,
"step": 1153
},
{
"epoch": 0.9711760993057017,
"grad_norm": 74.5656967163086,
"learning_rate": 3.881605877356035e-06,
"loss": 0.8615,
"step": 1154
},
{
"epoch": 0.9720176730486009,
"grad_norm": 102.34452056884766,
"learning_rate": 3.8797441794712465e-06,
"loss": 1.0315,
"step": 1155
},
{
"epoch": 0.9728592467915002,
"grad_norm": 80.07135772705078,
"learning_rate": 3.877881380672063e-06,
"loss": 0.8032,
"step": 1156
},
{
"epoch": 0.9737008205343993,
"grad_norm": 71.28023529052734,
"learning_rate": 3.876017482444832e-06,
"loss": 0.8414,
"step": 1157
},
{
"epoch": 0.9745423942772985,
"grad_norm": 69.85945129394531,
"learning_rate": 3.874152486276783e-06,
"loss": 0.8079,
"step": 1158
},
{
"epoch": 0.9753839680201978,
"grad_norm": 66.5711441040039,
"learning_rate": 3.872286393656017e-06,
"loss": 0.722,
"step": 1159
},
{
"epoch": 0.976225541763097,
"grad_norm": 177.06283569335938,
"learning_rate": 3.870419206071513e-06,
"loss": 0.766,
"step": 1160
},
{
"epoch": 0.9770671155059962,
"grad_norm": 92.01974487304688,
"learning_rate": 3.86855092501312e-06,
"loss": 1.0476,
"step": 1161
},
{
"epoch": 0.9779086892488954,
"grad_norm": 74.33403015136719,
"learning_rate": 3.866681551971565e-06,
"loss": 0.9104,
"step": 1162
},
{
"epoch": 0.9787502629917947,
"grad_norm": 66.5916748046875,
"learning_rate": 3.864811088438441e-06,
"loss": 0.7797,
"step": 1163
},
{
"epoch": 0.9795918367346939,
"grad_norm": 54.91474533081055,
"learning_rate": 3.862939535906213e-06,
"loss": 0.7155,
"step": 1164
},
{
"epoch": 0.9804334104775931,
"grad_norm": 61.410614013671875,
"learning_rate": 3.8610668958682165e-06,
"loss": 0.7638,
"step": 1165
},
{
"epoch": 0.9812749842204923,
"grad_norm": 71.29328155517578,
"learning_rate": 3.859193169818653e-06,
"loss": 0.7566,
"step": 1166
},
{
"epoch": 0.9821165579633916,
"grad_norm": 68.58023834228516,
"learning_rate": 3.85731835925259e-06,
"loss": 0.9146,
"step": 1167
},
{
"epoch": 0.9829581317062908,
"grad_norm": 63.331214904785156,
"learning_rate": 3.8554424656659625e-06,
"loss": 0.8314,
"step": 1168
},
{
"epoch": 0.9837997054491899,
"grad_norm": 78.91651916503906,
"learning_rate": 3.853565490555568e-06,
"loss": 0.8704,
"step": 1169
},
{
"epoch": 0.9846412791920892,
"grad_norm": 67.55413818359375,
"learning_rate": 3.851687435419067e-06,
"loss": 0.7655,
"step": 1170
},
{
"epoch": 0.9854828529349884,
"grad_norm": 69.61270904541016,
"learning_rate": 3.849808301754983e-06,
"loss": 0.8157,
"step": 1171
},
{
"epoch": 0.9863244266778877,
"grad_norm": 62.39061737060547,
"learning_rate": 3.847928091062698e-06,
"loss": 0.8609,
"step": 1172
},
{
"epoch": 0.9871660004207868,
"grad_norm": 73.21965026855469,
"learning_rate": 3.846046804842456e-06,
"loss": 0.6772,
"step": 1173
},
{
"epoch": 0.9880075741636861,
"grad_norm": 52.76114273071289,
"learning_rate": 3.844164444595356e-06,
"loss": 0.6729,
"step": 1174
},
{
"epoch": 0.9888491479065853,
"grad_norm": 65.8831558227539,
"learning_rate": 3.842281011823359e-06,
"loss": 0.6839,
"step": 1175
},
{
"epoch": 0.9896907216494846,
"grad_norm": 63.77772903442383,
"learning_rate": 3.840396508029275e-06,
"loss": 0.8853,
"step": 1176
},
{
"epoch": 0.9905322953923837,
"grad_norm": 70.06664276123047,
"learning_rate": 3.8385109347167735e-06,
"loss": 0.6873,
"step": 1177
},
{
"epoch": 0.991373869135283,
"grad_norm": 81.78180694580078,
"learning_rate": 3.836624293390378e-06,
"loss": 0.9209,
"step": 1178
},
{
"epoch": 0.9922154428781822,
"grad_norm": 66.87544250488281,
"learning_rate": 3.834736585555458e-06,
"loss": 0.6199,
"step": 1179
},
{
"epoch": 0.9930570166210814,
"grad_norm": 64.75494384765625,
"learning_rate": 3.832847812718242e-06,
"loss": 0.9293,
"step": 1180
},
{
"epoch": 0.9938985903639806,
"grad_norm": 67.62046813964844,
"learning_rate": 3.830957976385803e-06,
"loss": 0.8471,
"step": 1181
},
{
"epoch": 0.9947401641068798,
"grad_norm": 72.29881286621094,
"learning_rate": 3.829067078066063e-06,
"loss": 0.8786,
"step": 1182
},
{
"epoch": 0.9955817378497791,
"grad_norm": 57.97904968261719,
"learning_rate": 3.827175119267794e-06,
"loss": 0.716,
"step": 1183
},
{
"epoch": 0.9964233115926783,
"grad_norm": 57.827091217041016,
"learning_rate": 3.825282101500612e-06,
"loss": 0.7305,
"step": 1184
},
{
"epoch": 0.9972648853355776,
"grad_norm": 70.0528335571289,
"learning_rate": 3.823388026274979e-06,
"loss": 0.7974,
"step": 1185
},
{
"epoch": 0.9981064590784767,
"grad_norm": 73.72193908691406,
"learning_rate": 3.821492895102199e-06,
"loss": 0.8964,
"step": 1186
},
{
"epoch": 0.998948032821376,
"grad_norm": 67.32428741455078,
"learning_rate": 3.819596709494421e-06,
"loss": 0.7493,
"step": 1187
},
{
"epoch": 0.9997896065642752,
"grad_norm": 64.63594818115234,
"learning_rate": 3.817699470964634e-06,
"loss": 0.8268,
"step": 1188
},
{
"epoch": 1.0008415737428993,
"grad_norm": 118.80487823486328,
"learning_rate": 3.815801181026666e-06,
"loss": 1.8068,
"step": 1189
},
{
"epoch": 1.0016831474857983,
"grad_norm": 62.071449279785156,
"learning_rate": 3.813901841195188e-06,
"loss": 0.7766,
"step": 1190
},
{
"epoch": 1.0025247212286976,
"grad_norm": 71.35749053955078,
"learning_rate": 3.812001452985704e-06,
"loss": 0.7491,
"step": 1191
},
{
"epoch": 1.003366294971597,
"grad_norm": 65.00326538085938,
"learning_rate": 3.8101000179145564e-06,
"loss": 0.7828,
"step": 1192
},
{
"epoch": 1.0042078687144962,
"grad_norm": 65.48424530029297,
"learning_rate": 3.8081975374989237e-06,
"loss": 0.7553,
"step": 1193
},
{
"epoch": 1.0050494424573952,
"grad_norm": 71.35240936279297,
"learning_rate": 3.8062940132568183e-06,
"loss": 0.8126,
"step": 1194
},
{
"epoch": 1.0058910162002945,
"grad_norm": 75.91024780273438,
"learning_rate": 3.804389446707083e-06,
"loss": 0.9366,
"step": 1195
},
{
"epoch": 1.0067325899431938,
"grad_norm": 56.03441619873047,
"learning_rate": 3.8024838393693965e-06,
"loss": 0.6938,
"step": 1196
},
{
"epoch": 1.007574163686093,
"grad_norm": 75.76158142089844,
"learning_rate": 3.8005771927642655e-06,
"loss": 0.7218,
"step": 1197
},
{
"epoch": 1.0084157374289922,
"grad_norm": 72.80091857910156,
"learning_rate": 3.7986695084130244e-06,
"loss": 0.7922,
"step": 1198
},
{
"epoch": 1.0092573111718914,
"grad_norm": 67.57899475097656,
"learning_rate": 3.7967607878378375e-06,
"loss": 0.8067,
"step": 1199
},
{
"epoch": 1.0100988849147907,
"grad_norm": 64.86554718017578,
"learning_rate": 3.794851032561698e-06,
"loss": 0.7498,
"step": 1200
},
{
"epoch": 1.0109404586576898,
"grad_norm": 63.07143783569336,
"learning_rate": 3.7929402441084207e-06,
"loss": 0.8218,
"step": 1201
},
{
"epoch": 1.011782032400589,
"grad_norm": 70.4422378540039,
"learning_rate": 3.7910284240026474e-06,
"loss": 0.8098,
"step": 1202
},
{
"epoch": 1.0126236061434883,
"grad_norm": 73.65013885498047,
"learning_rate": 3.7891155737698415e-06,
"loss": 0.8102,
"step": 1203
},
{
"epoch": 1.0134651798863876,
"grad_norm": 61.945465087890625,
"learning_rate": 3.787201694936289e-06,
"loss": 0.8075,
"step": 1204
},
{
"epoch": 1.0143067536292867,
"grad_norm": 67.951904296875,
"learning_rate": 3.785286789029098e-06,
"loss": 0.6416,
"step": 1205
},
{
"epoch": 1.015148327372186,
"grad_norm": 68.7704086303711,
"learning_rate": 3.783370857576194e-06,
"loss": 0.6408,
"step": 1206
},
{
"epoch": 1.0159899011150852,
"grad_norm": 69.95020294189453,
"learning_rate": 3.7814539021063213e-06,
"loss": 0.731,
"step": 1207
},
{
"epoch": 1.0168314748579845,
"grad_norm": 75.72080993652344,
"learning_rate": 3.7795359241490425e-06,
"loss": 0.6204,
"step": 1208
},
{
"epoch": 1.0176730486008836,
"grad_norm": 69.6406478881836,
"learning_rate": 3.7776169252347344e-06,
"loss": 0.7158,
"step": 1209
},
{
"epoch": 1.0185146223437829,
"grad_norm": 74.77608489990234,
"learning_rate": 3.7756969068945893e-06,
"loss": 0.7458,
"step": 1210
},
{
"epoch": 1.0193561960866822,
"grad_norm": 97.5343017578125,
"learning_rate": 3.773775870660613e-06,
"loss": 0.9934,
"step": 1211
},
{
"epoch": 1.0201977698295812,
"grad_norm": 62.32524490356445,
"learning_rate": 3.7718538180656237e-06,
"loss": 0.6901,
"step": 1212
},
{
"epoch": 1.0210393435724805,
"grad_norm": 90.14856719970703,
"learning_rate": 3.7699307506432496e-06,
"loss": 0.9022,
"step": 1213
},
{
"epoch": 1.0218809173153798,
"grad_norm": 86.78469848632812,
"learning_rate": 3.7680066699279294e-06,
"loss": 0.8197,
"step": 1214
},
{
"epoch": 1.022722491058279,
"grad_norm": 67.51494598388672,
"learning_rate": 3.76608157745491e-06,
"loss": 0.7087,
"step": 1215
},
{
"epoch": 1.0235640648011781,
"grad_norm": 69.369873046875,
"learning_rate": 3.764155474760246e-06,
"loss": 0.774,
"step": 1216
},
{
"epoch": 1.0244056385440774,
"grad_norm": 65.6505355834961,
"learning_rate": 3.762228363380797e-06,
"loss": 0.6347,
"step": 1217
},
{
"epoch": 1.0252472122869767,
"grad_norm": 55.137664794921875,
"learning_rate": 3.760300244854229e-06,
"loss": 0.4952,
"step": 1218
},
{
"epoch": 1.026088786029876,
"grad_norm": 67.7928466796875,
"learning_rate": 3.758371120719011e-06,
"loss": 0.6969,
"step": 1219
},
{
"epoch": 1.026930359772775,
"grad_norm": 76.3191909790039,
"learning_rate": 3.756440992514414e-06,
"loss": 0.7473,
"step": 1220
},
{
"epoch": 1.0277719335156743,
"grad_norm": 81.80817413330078,
"learning_rate": 3.754509861780511e-06,
"loss": 0.7316,
"step": 1221
},
{
"epoch": 1.0286135072585736,
"grad_norm": 65.12232208251953,
"learning_rate": 3.7525777300581723e-06,
"loss": 0.7829,
"step": 1222
},
{
"epoch": 1.0294550810014726,
"grad_norm": 136.61734008789062,
"learning_rate": 3.7506445988890715e-06,
"loss": 0.9183,
"step": 1223
},
{
"epoch": 1.030296654744372,
"grad_norm": 76.92049407958984,
"learning_rate": 3.748710469815675e-06,
"loss": 0.815,
"step": 1224
},
{
"epoch": 1.0311382284872712,
"grad_norm": 76.00663757324219,
"learning_rate": 3.7467753443812493e-06,
"loss": 0.6683,
"step": 1225
},
{
"epoch": 1.0319798022301705,
"grad_norm": 82.0116195678711,
"learning_rate": 3.7448392241298532e-06,
"loss": 0.8531,
"step": 1226
},
{
"epoch": 1.0328213759730696,
"grad_norm": 68.03218841552734,
"learning_rate": 3.74290211060634e-06,
"loss": 0.7932,
"step": 1227
},
{
"epoch": 1.0336629497159688,
"grad_norm": 85.45331573486328,
"learning_rate": 3.7409640053563565e-06,
"loss": 0.8355,
"step": 1228
},
{
"epoch": 1.0345045234588681,
"grad_norm": 111.25006103515625,
"learning_rate": 3.739024909926339e-06,
"loss": 0.94,
"step": 1229
},
{
"epoch": 1.0353460972017674,
"grad_norm": 80.6521224975586,
"learning_rate": 3.7370848258635163e-06,
"loss": 0.6197,
"step": 1230
},
{
"epoch": 1.0361876709446665,
"grad_norm": 80.95207977294922,
"learning_rate": 3.735143754715903e-06,
"loss": 0.8114,
"step": 1231
},
{
"epoch": 1.0370292446875657,
"grad_norm": 73.94544219970703,
"learning_rate": 3.733201698032305e-06,
"loss": 0.692,
"step": 1232
},
{
"epoch": 1.037870818430465,
"grad_norm": 67.47168731689453,
"learning_rate": 3.7312586573623103e-06,
"loss": 0.7783,
"step": 1233
},
{
"epoch": 1.038712392173364,
"grad_norm": 64.5782241821289,
"learning_rate": 3.729314634256297e-06,
"loss": 0.8575,
"step": 1234
},
{
"epoch": 1.0395539659162634,
"grad_norm": 80.02053833007812,
"learning_rate": 3.7273696302654223e-06,
"loss": 0.6189,
"step": 1235
},
{
"epoch": 1.0403955396591626,
"grad_norm": 69.40652465820312,
"learning_rate": 3.7254236469416287e-06,
"loss": 0.8405,
"step": 1236
},
{
"epoch": 1.041237113402062,
"grad_norm": 77.50457763671875,
"learning_rate": 3.72347668583764e-06,
"loss": 0.7768,
"step": 1237
},
{
"epoch": 1.042078687144961,
"grad_norm": 58.28728485107422,
"learning_rate": 3.7215287485069598e-06,
"loss": 0.6804,
"step": 1238
},
{
"epoch": 1.0429202608878603,
"grad_norm": 81.27555084228516,
"learning_rate": 3.7195798365038695e-06,
"loss": 0.887,
"step": 1239
},
{
"epoch": 1.0437618346307596,
"grad_norm": 107.18325805664062,
"learning_rate": 3.7176299513834308e-06,
"loss": 1.1082,
"step": 1240
},
{
"epoch": 1.0446034083736588,
"grad_norm": 71.24678802490234,
"learning_rate": 3.7156790947014797e-06,
"loss": 0.8752,
"step": 1241
},
{
"epoch": 1.045444982116558,
"grad_norm": 64.26017761230469,
"learning_rate": 3.713727268014628e-06,
"loss": 0.5882,
"step": 1242
},
{
"epoch": 1.0462865558594572,
"grad_norm": 82.4547119140625,
"learning_rate": 3.711774472880262e-06,
"loss": 0.7684,
"step": 1243
},
{
"epoch": 1.0471281296023565,
"grad_norm": 73.11531066894531,
"learning_rate": 3.709820710856539e-06,
"loss": 0.7802,
"step": 1244
},
{
"epoch": 1.0479697033452555,
"grad_norm": 83.8969497680664,
"learning_rate": 3.707865983502391e-06,
"loss": 0.8114,
"step": 1245
},
{
"epoch": 1.0488112770881548,
"grad_norm": 69.56085205078125,
"learning_rate": 3.7059102923775175e-06,
"loss": 0.6799,
"step": 1246
},
{
"epoch": 1.049652850831054,
"grad_norm": 63.41591262817383,
"learning_rate": 3.7039536390423873e-06,
"loss": 0.7341,
"step": 1247
},
{
"epoch": 1.0504944245739534,
"grad_norm": 56.824581146240234,
"learning_rate": 3.7019960250582383e-06,
"loss": 0.6494,
"step": 1248
},
{
"epoch": 1.0513359983168524,
"grad_norm": 74.19931030273438,
"learning_rate": 3.700037451987074e-06,
"loss": 0.6798,
"step": 1249
},
{
"epoch": 1.0521775720597517,
"grad_norm": 65.28782653808594,
"learning_rate": 3.698077921391663e-06,
"loss": 0.6992,
"step": 1250
},
{
"epoch": 1.053019145802651,
"grad_norm": 119.53057861328125,
"learning_rate": 3.6961174348355383e-06,
"loss": 1.2016,
"step": 1251
},
{
"epoch": 1.0538607195455503,
"grad_norm": 59.280277252197266,
"learning_rate": 3.694155993882996e-06,
"loss": 0.5885,
"step": 1252
},
{
"epoch": 1.0547022932884493,
"grad_norm": 84.87373352050781,
"learning_rate": 3.692193600099092e-06,
"loss": 0.7715,
"step": 1253
},
{
"epoch": 1.0555438670313486,
"grad_norm": 71.60598754882812,
"learning_rate": 3.690230255049646e-06,
"loss": 0.8461,
"step": 1254
},
{
"epoch": 1.056385440774248,
"grad_norm": 81.19401550292969,
"learning_rate": 3.6882659603012325e-06,
"loss": 0.7127,
"step": 1255
},
{
"epoch": 1.0572270145171472,
"grad_norm": 90.88418579101562,
"learning_rate": 3.6863007174211873e-06,
"loss": 0.8569,
"step": 1256
},
{
"epoch": 1.0580685882600462,
"grad_norm": 60.42268753051758,
"learning_rate": 3.6843345279776e-06,
"loss": 0.6123,
"step": 1257
},
{
"epoch": 1.0589101620029455,
"grad_norm": 68.3895263671875,
"learning_rate": 3.6823673935393168e-06,
"loss": 0.6432,
"step": 1258
},
{
"epoch": 1.0597517357458448,
"grad_norm": 69.6331558227539,
"learning_rate": 3.6803993156759388e-06,
"loss": 0.7601,
"step": 1259
},
{
"epoch": 1.0605933094887439,
"grad_norm": 67.74185180664062,
"learning_rate": 3.6784302959578178e-06,
"loss": 0.9226,
"step": 1260
},
{
"epoch": 1.0614348832316431,
"grad_norm": 77.12291717529297,
"learning_rate": 3.6764603359560585e-06,
"loss": 0.862,
"step": 1261
},
{
"epoch": 1.0622764569745424,
"grad_norm": 71.94026947021484,
"learning_rate": 3.674489437242516e-06,
"loss": 0.7163,
"step": 1262
},
{
"epoch": 1.0631180307174417,
"grad_norm": 84.82723999023438,
"learning_rate": 3.6725176013897938e-06,
"loss": 0.654,
"step": 1263
},
{
"epoch": 1.0639596044603408,
"grad_norm": 55.34287643432617,
"learning_rate": 3.6705448299712433e-06,
"loss": 0.4786,
"step": 1264
},
{
"epoch": 1.06480117820324,
"grad_norm": 78.00069427490234,
"learning_rate": 3.668571124560962e-06,
"loss": 0.8155,
"step": 1265
},
{
"epoch": 1.0656427519461393,
"grad_norm": 68.24934387207031,
"learning_rate": 3.666596486733794e-06,
"loss": 0.8296,
"step": 1266
},
{
"epoch": 1.0664843256890384,
"grad_norm": 62.619388580322266,
"learning_rate": 3.664620918065325e-06,
"loss": 0.8412,
"step": 1267
},
{
"epoch": 1.0673258994319377,
"grad_norm": 62.24843978881836,
"learning_rate": 3.6626444201318865e-06,
"loss": 0.6693,
"step": 1268
},
{
"epoch": 1.068167473174837,
"grad_norm": 72.28678894042969,
"learning_rate": 3.6606669945105494e-06,
"loss": 0.8794,
"step": 1269
},
{
"epoch": 1.0690090469177362,
"grad_norm": 63.4146728515625,
"learning_rate": 3.6586886427791245e-06,
"loss": 0.7898,
"step": 1270
},
{
"epoch": 1.0698506206606353,
"grad_norm": 82.39759826660156,
"learning_rate": 3.6567093665161636e-06,
"loss": 0.811,
"step": 1271
},
{
"epoch": 1.0706921944035346,
"grad_norm": 76.58573150634766,
"learning_rate": 3.6547291673009533e-06,
"loss": 0.889,
"step": 1272
},
{
"epoch": 1.0715337681464339,
"grad_norm": 68.4346694946289,
"learning_rate": 3.6527480467135196e-06,
"loss": 0.7371,
"step": 1273
},
{
"epoch": 1.0723753418893331,
"grad_norm": 69.64159393310547,
"learning_rate": 3.6507660063346216e-06,
"loss": 0.7767,
"step": 1274
},
{
"epoch": 1.0732169156322322,
"grad_norm": 72.70745086669922,
"learning_rate": 3.648783047745753e-06,
"loss": 0.7388,
"step": 1275
},
{
"epoch": 1.0740584893751315,
"grad_norm": 61.45391845703125,
"learning_rate": 3.6467991725291406e-06,
"loss": 0.7133,
"step": 1276
},
{
"epoch": 1.0749000631180308,
"grad_norm": 59.521297454833984,
"learning_rate": 3.644814382267742e-06,
"loss": 0.6792,
"step": 1277
},
{
"epoch": 1.07574163686093,
"grad_norm": 72.21150207519531,
"learning_rate": 3.6428286785452454e-06,
"loss": 0.7792,
"step": 1278
},
{
"epoch": 1.076583210603829,
"grad_norm": 95.27154541015625,
"learning_rate": 3.640842062946067e-06,
"loss": 0.9082,
"step": 1279
},
{
"epoch": 1.0774247843467284,
"grad_norm": 62.20063781738281,
"learning_rate": 3.638854537055352e-06,
"loss": 0.645,
"step": 1280
},
{
"epoch": 1.0782663580896277,
"grad_norm": 61.36935043334961,
"learning_rate": 3.6368661024589702e-06,
"loss": 0.7807,
"step": 1281
},
{
"epoch": 1.0791079318325267,
"grad_norm": 68.19465637207031,
"learning_rate": 3.634876760743518e-06,
"loss": 0.8405,
"step": 1282
},
{
"epoch": 1.079949505575426,
"grad_norm": 99.6678466796875,
"learning_rate": 3.6328865134963144e-06,
"loss": 1.0025,
"step": 1283
},
{
"epoch": 1.0807910793183253,
"grad_norm": 67.33273315429688,
"learning_rate": 3.6308953623054025e-06,
"loss": 0.693,
"step": 1284
},
{
"epoch": 1.0816326530612246,
"grad_norm": 78.76142120361328,
"learning_rate": 3.628903308759545e-06,
"loss": 0.855,
"step": 1285
},
{
"epoch": 1.0824742268041236,
"grad_norm": 68.64391326904297,
"learning_rate": 3.6269103544482245e-06,
"loss": 0.8117,
"step": 1286
},
{
"epoch": 1.083315800547023,
"grad_norm": 79.25052642822266,
"learning_rate": 3.6249165009616436e-06,
"loss": 0.6921,
"step": 1287
},
{
"epoch": 1.0841573742899222,
"grad_norm": 82.95649719238281,
"learning_rate": 3.622921749890723e-06,
"loss": 0.7743,
"step": 1288
},
{
"epoch": 1.0849989480328213,
"grad_norm": 73.58319854736328,
"learning_rate": 3.620926102827097e-06,
"loss": 0.8296,
"step": 1289
},
{
"epoch": 1.0858405217757205,
"grad_norm": 71.20716857910156,
"learning_rate": 3.618929561363117e-06,
"loss": 0.7582,
"step": 1290
},
{
"epoch": 1.0866820955186198,
"grad_norm": 79.03595733642578,
"learning_rate": 3.616932127091848e-06,
"loss": 0.8906,
"step": 1291
},
{
"epoch": 1.087523669261519,
"grad_norm": 94.94636535644531,
"learning_rate": 3.6149338016070647e-06,
"loss": 0.9492,
"step": 1292
},
{
"epoch": 1.0883652430044182,
"grad_norm": 72.05924987792969,
"learning_rate": 3.612934586503257e-06,
"loss": 0.5987,
"step": 1293
},
{
"epoch": 1.0892068167473175,
"grad_norm": 80.31555938720703,
"learning_rate": 3.610934483375621e-06,
"loss": 0.9135,
"step": 1294
},
{
"epoch": 1.0900483904902167,
"grad_norm": 80.71312713623047,
"learning_rate": 3.6089334938200644e-06,
"loss": 0.8973,
"step": 1295
},
{
"epoch": 1.090889964233116,
"grad_norm": 69.00127410888672,
"learning_rate": 3.6069316194331995e-06,
"loss": 0.7587,
"step": 1296
},
{
"epoch": 1.091731537976015,
"grad_norm": 63.67856979370117,
"learning_rate": 3.6049288618123456e-06,
"loss": 0.7781,
"step": 1297
},
{
"epoch": 1.0925731117189144,
"grad_norm": 78.16109466552734,
"learning_rate": 3.6029252225555283e-06,
"loss": 0.8975,
"step": 1298
},
{
"epoch": 1.0934146854618136,
"grad_norm": 82.5450668334961,
"learning_rate": 3.6009207032614748e-06,
"loss": 0.7018,
"step": 1299
},
{
"epoch": 1.094256259204713,
"grad_norm": 67.94577026367188,
"learning_rate": 3.598915305529615e-06,
"loss": 0.7293,
"step": 1300
},
{
"epoch": 1.095097832947612,
"grad_norm": 76.59442138671875,
"learning_rate": 3.596909030960079e-06,
"loss": 0.8041,
"step": 1301
},
{
"epoch": 1.0959394066905113,
"grad_norm": 59.258644104003906,
"learning_rate": 3.5949018811536984e-06,
"loss": 0.8281,
"step": 1302
},
{
"epoch": 1.0967809804334105,
"grad_norm": 81.13563537597656,
"learning_rate": 3.5928938577120015e-06,
"loss": 0.7705,
"step": 1303
},
{
"epoch": 1.0976225541763096,
"grad_norm": 84.97859954833984,
"learning_rate": 3.590884962237214e-06,
"loss": 0.7768,
"step": 1304
},
{
"epoch": 1.0984641279192089,
"grad_norm": 75.2193832397461,
"learning_rate": 3.5888751963322587e-06,
"loss": 0.7057,
"step": 1305
},
{
"epoch": 1.0993057016621082,
"grad_norm": 62.515525817871094,
"learning_rate": 3.586864561600751e-06,
"loss": 0.607,
"step": 1306
},
{
"epoch": 1.1001472754050075,
"grad_norm": 85.55615234375,
"learning_rate": 3.5848530596470005e-06,
"loss": 0.9196,
"step": 1307
},
{
"epoch": 1.1009888491479065,
"grad_norm": 87.32113647460938,
"learning_rate": 3.5828406920760085e-06,
"loss": 1.0084,
"step": 1308
},
{
"epoch": 1.1018304228908058,
"grad_norm": 72.80986785888672,
"learning_rate": 3.580827460493468e-06,
"loss": 0.6663,
"step": 1309
},
{
"epoch": 1.102671996633705,
"grad_norm": 82.09026336669922,
"learning_rate": 3.5788133665057595e-06,
"loss": 0.8127,
"step": 1310
},
{
"epoch": 1.1035135703766041,
"grad_norm": 89.75031280517578,
"learning_rate": 3.5767984117199535e-06,
"loss": 0.8213,
"step": 1311
},
{
"epoch": 1.1043551441195034,
"grad_norm": 81.2878646850586,
"learning_rate": 3.5747825977438062e-06,
"loss": 0.8556,
"step": 1312
},
{
"epoch": 1.1051967178624027,
"grad_norm": 73.67637634277344,
"learning_rate": 3.57276592618576e-06,
"loss": 0.7873,
"step": 1313
},
{
"epoch": 1.106038291605302,
"grad_norm": 93.03740692138672,
"learning_rate": 3.5707483986549407e-06,
"loss": 0.7661,
"step": 1314
},
{
"epoch": 1.106879865348201,
"grad_norm": 73.1768569946289,
"learning_rate": 3.568730016761159e-06,
"loss": 0.8781,
"step": 1315
},
{
"epoch": 1.1077214390911003,
"grad_norm": 52.64601516723633,
"learning_rate": 3.566710782114904e-06,
"loss": 0.5982,
"step": 1316
},
{
"epoch": 1.1085630128339996,
"grad_norm": 67.40423583984375,
"learning_rate": 3.5646906963273482e-06,
"loss": 0.6887,
"step": 1317
},
{
"epoch": 1.1094045865768989,
"grad_norm": 76.27008819580078,
"learning_rate": 3.5626697610103424e-06,
"loss": 0.8059,
"step": 1318
},
{
"epoch": 1.110246160319798,
"grad_norm": 68.23004150390625,
"learning_rate": 3.5606479777764142e-06,
"loss": 0.7669,
"step": 1319
},
{
"epoch": 1.1110877340626972,
"grad_norm": 68.53443145751953,
"learning_rate": 3.55862534823877e-06,
"loss": 0.669,
"step": 1320
},
{
"epoch": 1.1119293078055965,
"grad_norm": 60.3872184753418,
"learning_rate": 3.556601874011289e-06,
"loss": 0.6156,
"step": 1321
},
{
"epoch": 1.1127708815484958,
"grad_norm": 57.09178161621094,
"learning_rate": 3.5545775567085263e-06,
"loss": 0.8307,
"step": 1322
},
{
"epoch": 1.1136124552913949,
"grad_norm": 77.64226531982422,
"learning_rate": 3.552552397945708e-06,
"loss": 0.963,
"step": 1323
},
{
"epoch": 1.1144540290342941,
"grad_norm": 67.55712127685547,
"learning_rate": 3.550526399338734e-06,
"loss": 0.7474,
"step": 1324
},
{
"epoch": 1.1152956027771934,
"grad_norm": 79.86772155761719,
"learning_rate": 3.5484995625041713e-06,
"loss": 0.7521,
"step": 1325
},
{
"epoch": 1.1161371765200925,
"grad_norm": 132.9801483154297,
"learning_rate": 3.5464718890592583e-06,
"loss": 0.6854,
"step": 1326
},
{
"epoch": 1.1169787502629918,
"grad_norm": 79.50981903076172,
"learning_rate": 3.5444433806218993e-06,
"loss": 0.9638,
"step": 1327
},
{
"epoch": 1.117820324005891,
"grad_norm": 84.78490447998047,
"learning_rate": 3.5424140388106652e-06,
"loss": 0.7145,
"step": 1328
},
{
"epoch": 1.1186618977487903,
"grad_norm": 68.53125,
"learning_rate": 3.540383865244793e-06,
"loss": 0.5723,
"step": 1329
},
{
"epoch": 1.1195034714916894,
"grad_norm": 65.09300231933594,
"learning_rate": 3.538352861544182e-06,
"loss": 0.6238,
"step": 1330
},
{
"epoch": 1.1203450452345887,
"grad_norm": 148.02352905273438,
"learning_rate": 3.536321029329395e-06,
"loss": 0.6512,
"step": 1331
},
{
"epoch": 1.121186618977488,
"grad_norm": 67.34825897216797,
"learning_rate": 3.5342883702216534e-06,
"loss": 0.7211,
"step": 1332
},
{
"epoch": 1.122028192720387,
"grad_norm": 83.49595642089844,
"learning_rate": 3.5322548858428424e-06,
"loss": 0.8661,
"step": 1333
},
{
"epoch": 1.1228697664632863,
"grad_norm": 66.2895736694336,
"learning_rate": 3.530220577815502e-06,
"loss": 0.7105,
"step": 1334
},
{
"epoch": 1.1237113402061856,
"grad_norm": 114.4986801147461,
"learning_rate": 3.5281854477628317e-06,
"loss": 0.7341,
"step": 1335
},
{
"epoch": 1.1245529139490849,
"grad_norm": 75.82929992675781,
"learning_rate": 3.526149497308686e-06,
"loss": 0.7586,
"step": 1336
},
{
"epoch": 1.125394487691984,
"grad_norm": 84.63530731201172,
"learning_rate": 3.524112728077575e-06,
"loss": 0.9024,
"step": 1337
},
{
"epoch": 1.1262360614348832,
"grad_norm": 81.52574920654297,
"learning_rate": 3.52207514169466e-06,
"loss": 0.8914,
"step": 1338
},
{
"epoch": 1.1270776351777825,
"grad_norm": 97.83477020263672,
"learning_rate": 3.5200367397857553e-06,
"loss": 1.0349,
"step": 1339
},
{
"epoch": 1.1279192089206818,
"grad_norm": 90.05326080322266,
"learning_rate": 3.5179975239773266e-06,
"loss": 0.952,
"step": 1340
},
{
"epoch": 1.1287607826635808,
"grad_norm": 73.44329071044922,
"learning_rate": 3.5159574958964895e-06,
"loss": 0.7403,
"step": 1341
},
{
"epoch": 1.12960235640648,
"grad_norm": 56.738311767578125,
"learning_rate": 3.513916657171006e-06,
"loss": 0.6308,
"step": 1342
},
{
"epoch": 1.1304439301493794,
"grad_norm": 60.14042282104492,
"learning_rate": 3.5118750094292853e-06,
"loss": 0.6327,
"step": 1343
},
{
"epoch": 1.1312855038922787,
"grad_norm": 73.36322021484375,
"learning_rate": 3.509832554300384e-06,
"loss": 0.7424,
"step": 1344
},
{
"epoch": 1.1321270776351777,
"grad_norm": 96.14129638671875,
"learning_rate": 3.5077892934140006e-06,
"loss": 0.9552,
"step": 1345
},
{
"epoch": 1.132968651378077,
"grad_norm": 85.27511596679688,
"learning_rate": 3.5057452284004767e-06,
"loss": 0.9572,
"step": 1346
},
{
"epoch": 1.1338102251209763,
"grad_norm": 65.14854431152344,
"learning_rate": 3.5037003608907975e-06,
"loss": 0.7357,
"step": 1347
},
{
"epoch": 1.1346517988638753,
"grad_norm": 85.38391876220703,
"learning_rate": 3.5016546925165866e-06,
"loss": 0.8456,
"step": 1348
},
{
"epoch": 1.1354933726067746,
"grad_norm": 83.07268524169922,
"learning_rate": 3.499608224910107e-06,
"loss": 1.0145,
"step": 1349
},
{
"epoch": 1.136334946349674,
"grad_norm": 82.41991424560547,
"learning_rate": 3.497560959704261e-06,
"loss": 0.9077,
"step": 1350
},
{
"epoch": 1.1371765200925732,
"grad_norm": 87.86802673339844,
"learning_rate": 3.495512898532585e-06,
"loss": 1.0205,
"step": 1351
},
{
"epoch": 1.1380180938354723,
"grad_norm": 67.33047485351562,
"learning_rate": 3.4934640430292512e-06,
"loss": 0.6918,
"step": 1352
},
{
"epoch": 1.1388596675783715,
"grad_norm": 58.26289367675781,
"learning_rate": 3.491414394829066e-06,
"loss": 0.6877,
"step": 1353
},
{
"epoch": 1.1397012413212708,
"grad_norm": 73.31893157958984,
"learning_rate": 3.4893639555674695e-06,
"loss": 0.8029,
"step": 1354
},
{
"epoch": 1.1405428150641699,
"grad_norm": 68.03288269042969,
"learning_rate": 3.487312726880531e-06,
"loss": 0.7889,
"step": 1355
},
{
"epoch": 1.1413843888070692,
"grad_norm": 66.80435180664062,
"learning_rate": 3.48526071040495e-06,
"loss": 0.7886,
"step": 1356
},
{
"epoch": 1.1422259625499684,
"grad_norm": 70.782958984375,
"learning_rate": 3.483207907778055e-06,
"loss": 0.8967,
"step": 1357
},
{
"epoch": 1.1430675362928677,
"grad_norm": 70.4962158203125,
"learning_rate": 3.481154320637802e-06,
"loss": 0.6492,
"step": 1358
},
{
"epoch": 1.143909110035767,
"grad_norm": 63.329246520996094,
"learning_rate": 3.4790999506227737e-06,
"loss": 0.7749,
"step": 1359
},
{
"epoch": 1.144750683778666,
"grad_norm": 75.54255676269531,
"learning_rate": 3.4770447993721747e-06,
"loss": 0.8227,
"step": 1360
},
{
"epoch": 1.1455922575215653,
"grad_norm": 62.19999694824219,
"learning_rate": 3.474988868525836e-06,
"loss": 0.7421,
"step": 1361
},
{
"epoch": 1.1464338312644646,
"grad_norm": 84.12921142578125,
"learning_rate": 3.4729321597242095e-06,
"loss": 0.8211,
"step": 1362
},
{
"epoch": 1.1472754050073637,
"grad_norm": 66.9234390258789,
"learning_rate": 3.470874674608367e-06,
"loss": 0.6031,
"step": 1363
},
{
"epoch": 1.148116978750263,
"grad_norm": 58.342872619628906,
"learning_rate": 3.468816414820001e-06,
"loss": 0.6996,
"step": 1364
},
{
"epoch": 1.1489585524931623,
"grad_norm": 69.41954040527344,
"learning_rate": 3.4667573820014215e-06,
"loss": 0.7793,
"step": 1365
},
{
"epoch": 1.1498001262360615,
"grad_norm": 93.1528549194336,
"learning_rate": 3.464697577795556e-06,
"loss": 0.8255,
"step": 1366
},
{
"epoch": 1.1506416999789606,
"grad_norm": 60.05856704711914,
"learning_rate": 3.4626370038459455e-06,
"loss": 0.5099,
"step": 1367
},
{
"epoch": 1.1514832737218599,
"grad_norm": 82.09307861328125,
"learning_rate": 3.4605756617967485e-06,
"loss": 0.7799,
"step": 1368
},
{
"epoch": 1.1523248474647592,
"grad_norm": 132.58616638183594,
"learning_rate": 3.458513553292733e-06,
"loss": 0.7991,
"step": 1369
},
{
"epoch": 1.1531664212076582,
"grad_norm": 62.63915252685547,
"learning_rate": 3.456450679979281e-06,
"loss": 0.6916,
"step": 1370
},
{
"epoch": 1.1540079949505575,
"grad_norm": 74.68429565429688,
"learning_rate": 3.454387043502383e-06,
"loss": 0.8426,
"step": 1371
},
{
"epoch": 1.1548495686934568,
"grad_norm": 64.6526107788086,
"learning_rate": 3.45232264550864e-06,
"loss": 0.7868,
"step": 1372
},
{
"epoch": 1.155691142436356,
"grad_norm": 71.72866821289062,
"learning_rate": 3.4502574876452594e-06,
"loss": 0.9175,
"step": 1373
},
{
"epoch": 1.1565327161792551,
"grad_norm": 65.70030975341797,
"learning_rate": 3.448191571560055e-06,
"loss": 0.6926,
"step": 1374
},
{
"epoch": 1.1573742899221544,
"grad_norm": 90.4015121459961,
"learning_rate": 3.4461248989014466e-06,
"loss": 0.9557,
"step": 1375
},
{
"epoch": 1.1582158636650537,
"grad_norm": 110.88024139404297,
"learning_rate": 3.444057471318457e-06,
"loss": 0.748,
"step": 1376
},
{
"epoch": 1.1590574374079528,
"grad_norm": 66.52738952636719,
"learning_rate": 3.4419892904607124e-06,
"loss": 0.7728,
"step": 1377
},
{
"epoch": 1.159899011150852,
"grad_norm": 81.06018829345703,
"learning_rate": 3.439920357978437e-06,
"loss": 0.8179,
"step": 1378
},
{
"epoch": 1.1607405848937513,
"grad_norm": 67.53968048095703,
"learning_rate": 3.437850675522457e-06,
"loss": 0.779,
"step": 1379
},
{
"epoch": 1.1615821586366506,
"grad_norm": 64.2708511352539,
"learning_rate": 3.435780244744199e-06,
"loss": 0.8699,
"step": 1380
},
{
"epoch": 1.1624237323795499,
"grad_norm": 65.01960754394531,
"learning_rate": 3.433709067295683e-06,
"loss": 0.6326,
"step": 1381
},
{
"epoch": 1.163265306122449,
"grad_norm": 61.47157669067383,
"learning_rate": 3.4316371448295256e-06,
"loss": 0.7747,
"step": 1382
},
{
"epoch": 1.1641068798653482,
"grad_norm": 74.15666198730469,
"learning_rate": 3.4295644789989397e-06,
"loss": 0.7089,
"step": 1383
},
{
"epoch": 1.1649484536082475,
"grad_norm": 59.576751708984375,
"learning_rate": 3.427491071457729e-06,
"loss": 0.7315,
"step": 1384
},
{
"epoch": 1.1657900273511466,
"grad_norm": 57.37242126464844,
"learning_rate": 3.425416923860292e-06,
"loss": 0.7324,
"step": 1385
},
{
"epoch": 1.1666316010940458,
"grad_norm": 67.24441528320312,
"learning_rate": 3.4233420378616144e-06,
"loss": 0.8165,
"step": 1386
},
{
"epoch": 1.1674731748369451,
"grad_norm": 72.99162292480469,
"learning_rate": 3.421266415117274e-06,
"loss": 0.8523,
"step": 1387
},
{
"epoch": 1.1683147485798444,
"grad_norm": 83.09717559814453,
"learning_rate": 3.4191900572834325e-06,
"loss": 0.6511,
"step": 1388
},
{
"epoch": 1.1691563223227435,
"grad_norm": 73.0621337890625,
"learning_rate": 3.4171129660168433e-06,
"loss": 0.7932,
"step": 1389
},
{
"epoch": 1.1699978960656428,
"grad_norm": 63.88787841796875,
"learning_rate": 3.4150351429748406e-06,
"loss": 0.7402,
"step": 1390
},
{
"epoch": 1.170839469808542,
"grad_norm": 68.88009643554688,
"learning_rate": 3.4129565898153448e-06,
"loss": 0.6272,
"step": 1391
},
{
"epoch": 1.171681043551441,
"grad_norm": 75.77027893066406,
"learning_rate": 3.4108773081968594e-06,
"loss": 0.8729,
"step": 1392
},
{
"epoch": 1.1725226172943404,
"grad_norm": 63.99619674682617,
"learning_rate": 3.408797299778467e-06,
"loss": 0.7848,
"step": 1393
},
{
"epoch": 1.1733641910372397,
"grad_norm": 64.0927505493164,
"learning_rate": 3.406716566219831e-06,
"loss": 0.8276,
"step": 1394
},
{
"epoch": 1.174205764780139,
"grad_norm": 224.97900390625,
"learning_rate": 3.4046351091811945e-06,
"loss": 0.8918,
"step": 1395
},
{
"epoch": 1.175047338523038,
"grad_norm": 79.71916198730469,
"learning_rate": 3.4025529303233764e-06,
"loss": 0.8166,
"step": 1396
},
{
"epoch": 1.1758889122659373,
"grad_norm": 72.94778442382812,
"learning_rate": 3.400470031307772e-06,
"loss": 0.6815,
"step": 1397
},
{
"epoch": 1.1767304860088366,
"grad_norm": 71.34203338623047,
"learning_rate": 3.3983864137963517e-06,
"loss": 0.6691,
"step": 1398
},
{
"epoch": 1.1775720597517356,
"grad_norm": 75.5274658203125,
"learning_rate": 3.396302079451659e-06,
"loss": 0.8337,
"step": 1399
},
{
"epoch": 1.178413633494635,
"grad_norm": 81.76123046875,
"learning_rate": 3.3942170299368094e-06,
"loss": 0.9503,
"step": 1400
},
{
"epoch": 1.1792552072375342,
"grad_norm": 198.34930419921875,
"learning_rate": 3.3921312669154883e-06,
"loss": 0.6294,
"step": 1401
},
{
"epoch": 1.1800967809804335,
"grad_norm": 73.18561553955078,
"learning_rate": 3.3900447920519507e-06,
"loss": 0.8083,
"step": 1402
},
{
"epoch": 1.1809383547233328,
"grad_norm": 77.73960876464844,
"learning_rate": 3.387957607011021e-06,
"loss": 0.7997,
"step": 1403
},
{
"epoch": 1.1817799284662318,
"grad_norm": 69.83214569091797,
"learning_rate": 3.3858697134580875e-06,
"loss": 0.5873,
"step": 1404
},
{
"epoch": 1.182621502209131,
"grad_norm": 51.2693977355957,
"learning_rate": 3.383781113059107e-06,
"loss": 0.7206,
"step": 1405
},
{
"epoch": 1.1834630759520304,
"grad_norm": 70.56368255615234,
"learning_rate": 3.3816918074805982e-06,
"loss": 0.7665,
"step": 1406
},
{
"epoch": 1.1843046496949294,
"grad_norm": 59.05124282836914,
"learning_rate": 3.379601798389642e-06,
"loss": 0.6055,
"step": 1407
},
{
"epoch": 1.1851462234378287,
"grad_norm": 61.26966094970703,
"learning_rate": 3.3775110874538834e-06,
"loss": 0.6029,
"step": 1408
},
{
"epoch": 1.185987797180728,
"grad_norm": 65.18826293945312,
"learning_rate": 3.3754196763415255e-06,
"loss": 0.8414,
"step": 1409
},
{
"epoch": 1.1868293709236273,
"grad_norm": 74.95602416992188,
"learning_rate": 3.373327566721329e-06,
"loss": 0.9045,
"step": 1410
},
{
"epoch": 1.1876709446665263,
"grad_norm": 54.38325500488281,
"learning_rate": 3.371234760262614e-06,
"loss": 0.6686,
"step": 1411
},
{
"epoch": 1.1885125184094256,
"grad_norm": 78.43930053710938,
"learning_rate": 3.3691412586352552e-06,
"loss": 0.7651,
"step": 1412
},
{
"epoch": 1.189354092152325,
"grad_norm": 72.94366455078125,
"learning_rate": 3.3670470635096837e-06,
"loss": 0.9488,
"step": 1413
},
{
"epoch": 1.190195665895224,
"grad_norm": 64.21903991699219,
"learning_rate": 3.364952176556882e-06,
"loss": 0.5698,
"step": 1414
},
{
"epoch": 1.1910372396381232,
"grad_norm": 79.49015045166016,
"learning_rate": 3.362856599448386e-06,
"loss": 0.9068,
"step": 1415
},
{
"epoch": 1.1918788133810225,
"grad_norm": 70.82846069335938,
"learning_rate": 3.360760333856282e-06,
"loss": 0.7983,
"step": 1416
},
{
"epoch": 1.1927203871239218,
"grad_norm": 85.83604431152344,
"learning_rate": 3.3586633814532043e-06,
"loss": 0.76,
"step": 1417
},
{
"epoch": 1.1935619608668209,
"grad_norm": 70.05342102050781,
"learning_rate": 3.3565657439123368e-06,
"loss": 0.7129,
"step": 1418
},
{
"epoch": 1.1944035346097202,
"grad_norm": 85.30274963378906,
"learning_rate": 3.3544674229074103e-06,
"loss": 0.9235,
"step": 1419
},
{
"epoch": 1.1952451083526194,
"grad_norm": 126.79804229736328,
"learning_rate": 3.3523684201126987e-06,
"loss": 0.5791,
"step": 1420
},
{
"epoch": 1.1960866820955187,
"grad_norm": 67.60092163085938,
"learning_rate": 3.3502687372030235e-06,
"loss": 0.758,
"step": 1421
},
{
"epoch": 1.1969282558384178,
"grad_norm": 98.4145278930664,
"learning_rate": 3.3481683758537454e-06,
"loss": 0.9113,
"step": 1422
},
{
"epoch": 1.197769829581317,
"grad_norm": 63.639102935791016,
"learning_rate": 3.3460673377407676e-06,
"loss": 0.7124,
"step": 1423
},
{
"epoch": 1.1986114033242163,
"grad_norm": 60.17888259887695,
"learning_rate": 3.343965624540535e-06,
"loss": 0.6899,
"step": 1424
},
{
"epoch": 1.1994529770671156,
"grad_norm": 86.66293334960938,
"learning_rate": 3.341863237930029e-06,
"loss": 1.0092,
"step": 1425
},
{
"epoch": 1.2002945508100147,
"grad_norm": 76.53636169433594,
"learning_rate": 3.3397601795867684e-06,
"loss": 0.7176,
"step": 1426
},
{
"epoch": 1.201136124552914,
"grad_norm": 72.173583984375,
"learning_rate": 3.3376564511888095e-06,
"loss": 0.8557,
"step": 1427
},
{
"epoch": 1.2019776982958132,
"grad_norm": 58.85466003417969,
"learning_rate": 3.335552054414742e-06,
"loss": 0.6287,
"step": 1428
},
{
"epoch": 1.2028192720387123,
"grad_norm": 58.9666748046875,
"learning_rate": 3.333446990943689e-06,
"loss": 0.6785,
"step": 1429
},
{
"epoch": 1.2036608457816116,
"grad_norm": 57.108009338378906,
"learning_rate": 3.3313412624553066e-06,
"loss": 0.7176,
"step": 1430
},
{
"epoch": 1.2045024195245109,
"grad_norm": 68.65019226074219,
"learning_rate": 3.3292348706297807e-06,
"loss": 0.8955,
"step": 1431
},
{
"epoch": 1.2053439932674102,
"grad_norm": 82.8580322265625,
"learning_rate": 3.327127817147825e-06,
"loss": 0.933,
"step": 1432
},
{
"epoch": 1.2061855670103092,
"grad_norm": 74.54534149169922,
"learning_rate": 3.3250201036906843e-06,
"loss": 0.6931,
"step": 1433
},
{
"epoch": 1.2070271407532085,
"grad_norm": 56.394012451171875,
"learning_rate": 3.322911731940128e-06,
"loss": 0.5127,
"step": 1434
},
{
"epoch": 1.2078687144961078,
"grad_norm": 71.54861450195312,
"learning_rate": 3.3208027035784513e-06,
"loss": 0.84,
"step": 1435
},
{
"epoch": 1.2087102882390068,
"grad_norm": 71.24790954589844,
"learning_rate": 3.3186930202884725e-06,
"loss": 0.7949,
"step": 1436
},
{
"epoch": 1.2095518619819061,
"grad_norm": 62.81068801879883,
"learning_rate": 3.316582683753534e-06,
"loss": 0.7365,
"step": 1437
},
{
"epoch": 1.2103934357248054,
"grad_norm": 68.5630111694336,
"learning_rate": 3.314471695657498e-06,
"loss": 0.5798,
"step": 1438
},
{
"epoch": 1.2112350094677047,
"grad_norm": 66.36724853515625,
"learning_rate": 3.3123600576847475e-06,
"loss": 0.8426,
"step": 1439
},
{
"epoch": 1.2120765832106037,
"grad_norm": 77.36634063720703,
"learning_rate": 3.3102477715201835e-06,
"loss": 0.9764,
"step": 1440
},
{
"epoch": 1.212918156953503,
"grad_norm": 77.31050872802734,
"learning_rate": 3.3081348388492247e-06,
"loss": 0.9308,
"step": 1441
},
{
"epoch": 1.2137597306964023,
"grad_norm": 74.0624771118164,
"learning_rate": 3.306021261357805e-06,
"loss": 0.8019,
"step": 1442
},
{
"epoch": 1.2146013044393016,
"grad_norm": 63.09843063354492,
"learning_rate": 3.303907040732374e-06,
"loss": 0.7373,
"step": 1443
},
{
"epoch": 1.2154428781822006,
"grad_norm": 69.69409942626953,
"learning_rate": 3.301792178659893e-06,
"loss": 0.8323,
"step": 1444
},
{
"epoch": 1.2162844519251,
"grad_norm": 48.24660110473633,
"learning_rate": 3.299676676827834e-06,
"loss": 0.6096,
"step": 1445
},
{
"epoch": 1.2171260256679992,
"grad_norm": 71.8796615600586,
"learning_rate": 3.297560536924185e-06,
"loss": 0.757,
"step": 1446
},
{
"epoch": 1.2179675994108985,
"grad_norm": 74.03809356689453,
"learning_rate": 3.295443760637437e-06,
"loss": 0.7347,
"step": 1447
},
{
"epoch": 1.2188091731537976,
"grad_norm": 75.21710968017578,
"learning_rate": 3.2933263496565908e-06,
"loss": 0.8716,
"step": 1448
},
{
"epoch": 1.2196507468966968,
"grad_norm": 77.39705657958984,
"learning_rate": 3.2912083056711545e-06,
"loss": 0.7371,
"step": 1449
},
{
"epoch": 1.2204923206395961,
"grad_norm": 82.25308990478516,
"learning_rate": 3.2890896303711417e-06,
"loss": 0.6782,
"step": 1450
},
{
"epoch": 1.2213338943824952,
"grad_norm": 57.6148681640625,
"learning_rate": 3.2869703254470677e-06,
"loss": 0.8079,
"step": 1451
},
{
"epoch": 1.2221754681253945,
"grad_norm": 62.759891510009766,
"learning_rate": 3.284850392589951e-06,
"loss": 0.7079,
"step": 1452
},
{
"epoch": 1.2230170418682937,
"grad_norm": 61.7097053527832,
"learning_rate": 3.282729833491312e-06,
"loss": 0.8856,
"step": 1453
},
{
"epoch": 1.223858615611193,
"grad_norm": 63.70922088623047,
"learning_rate": 3.2806086498431704e-06,
"loss": 0.8202,
"step": 1454
},
{
"epoch": 1.224700189354092,
"grad_norm": 59.12576675415039,
"learning_rate": 3.278486843338044e-06,
"loss": 0.6588,
"step": 1455
},
{
"epoch": 1.2255417630969914,
"grad_norm": 69.03202056884766,
"learning_rate": 3.276364415668947e-06,
"loss": 0.9184,
"step": 1456
},
{
"epoch": 1.2263833368398906,
"grad_norm": 63.162418365478516,
"learning_rate": 3.2742413685293906e-06,
"loss": 0.6063,
"step": 1457
},
{
"epoch": 1.2272249105827897,
"grad_norm": 47.80107879638672,
"learning_rate": 3.272117703613379e-06,
"loss": 0.6087,
"step": 1458
},
{
"epoch": 1.228066484325689,
"grad_norm": 69.43952178955078,
"learning_rate": 3.2699934226154093e-06,
"loss": 0.7593,
"step": 1459
},
{
"epoch": 1.2289080580685883,
"grad_norm": 80.62454986572266,
"learning_rate": 3.267868527230473e-06,
"loss": 0.7964,
"step": 1460
},
{
"epoch": 1.2297496318114876,
"grad_norm": 85.75826263427734,
"learning_rate": 3.2657430191540467e-06,
"loss": 0.7348,
"step": 1461
},
{
"epoch": 1.2305912055543866,
"grad_norm": 67.71492004394531,
"learning_rate": 3.2636169000821007e-06,
"loss": 0.8163,
"step": 1462
},
{
"epoch": 1.231432779297286,
"grad_norm": 59.86122131347656,
"learning_rate": 3.2614901717110893e-06,
"loss": 0.6843,
"step": 1463
},
{
"epoch": 1.2322743530401852,
"grad_norm": 61.91895294189453,
"learning_rate": 3.259362835737956e-06,
"loss": 0.6392,
"step": 1464
},
{
"epoch": 1.2331159267830845,
"grad_norm": 62.762943267822266,
"learning_rate": 3.257234893860126e-06,
"loss": 0.641,
"step": 1465
},
{
"epoch": 1.2339575005259835,
"grad_norm": 56.892093658447266,
"learning_rate": 3.255106347775512e-06,
"loss": 0.6627,
"step": 1466
},
{
"epoch": 1.2347990742688828,
"grad_norm": 64.63336944580078,
"learning_rate": 3.2529771991825043e-06,
"loss": 0.7472,
"step": 1467
},
{
"epoch": 1.235640648011782,
"grad_norm": 81.38465881347656,
"learning_rate": 3.250847449779977e-06,
"loss": 0.7303,
"step": 1468
},
{
"epoch": 1.2364822217546814,
"grad_norm": 66.22145080566406,
"learning_rate": 3.2487171012672825e-06,
"loss": 0.9322,
"step": 1469
},
{
"epoch": 1.2373237954975804,
"grad_norm": 77.17510986328125,
"learning_rate": 3.2465861553442514e-06,
"loss": 0.9179,
"step": 1470
},
{
"epoch": 1.2381653692404797,
"grad_norm": 84.92408752441406,
"learning_rate": 3.2444546137111905e-06,
"loss": 0.8133,
"step": 1471
},
{
"epoch": 1.239006942983379,
"grad_norm": 82.94153594970703,
"learning_rate": 3.2423224780688827e-06,
"loss": 0.7668,
"step": 1472
},
{
"epoch": 1.239848516726278,
"grad_norm": 89.27391815185547,
"learning_rate": 3.240189750118584e-06,
"loss": 0.8618,
"step": 1473
},
{
"epoch": 1.2406900904691773,
"grad_norm": 82.90675354003906,
"learning_rate": 3.2380564315620254e-06,
"loss": 0.9421,
"step": 1474
},
{
"epoch": 1.2415316642120766,
"grad_norm": 80.6243667602539,
"learning_rate": 3.2359225241014057e-06,
"loss": 0.6881,
"step": 1475
},
{
"epoch": 1.242373237954976,
"grad_norm": 79.01913452148438,
"learning_rate": 3.2337880294393953e-06,
"loss": 0.7986,
"step": 1476
},
{
"epoch": 1.243214811697875,
"grad_norm": 63.30573272705078,
"learning_rate": 3.231652949279133e-06,
"loss": 0.7715,
"step": 1477
},
{
"epoch": 1.2440563854407742,
"grad_norm": 67.03791809082031,
"learning_rate": 3.2295172853242252e-06,
"loss": 0.7908,
"step": 1478
},
{
"epoch": 1.2448979591836735,
"grad_norm": 68.64743041992188,
"learning_rate": 3.2273810392787443e-06,
"loss": 0.8484,
"step": 1479
},
{
"epoch": 1.2457395329265726,
"grad_norm": 67.99822998046875,
"learning_rate": 3.225244212847225e-06,
"loss": 0.7503,
"step": 1480
},
{
"epoch": 1.2465811066694719,
"grad_norm": 58.34498596191406,
"learning_rate": 3.2231068077346676e-06,
"loss": 0.6159,
"step": 1481
},
{
"epoch": 1.2474226804123711,
"grad_norm": 53.544925689697266,
"learning_rate": 3.220968825646534e-06,
"loss": 0.6952,
"step": 1482
},
{
"epoch": 1.2482642541552704,
"grad_norm": 66.74227905273438,
"learning_rate": 3.218830268288744e-06,
"loss": 0.6318,
"step": 1483
},
{
"epoch": 1.2491058278981695,
"grad_norm": 69.69791412353516,
"learning_rate": 3.21669113736768e-06,
"loss": 0.606,
"step": 1484
},
{
"epoch": 1.2499474016410688,
"grad_norm": 84.73765563964844,
"learning_rate": 3.2145514345901786e-06,
"loss": 0.791,
"step": 1485
},
{
"epoch": 1.250788975383968,
"grad_norm": 75.47090148925781,
"learning_rate": 3.212411161663535e-06,
"loss": 0.831,
"step": 1486
},
{
"epoch": 1.251630549126867,
"grad_norm": 88.83540344238281,
"learning_rate": 3.2102703202954984e-06,
"loss": 0.8213,
"step": 1487
},
{
"epoch": 1.2524721228697664,
"grad_norm": 64.17558288574219,
"learning_rate": 3.208128912194272e-06,
"loss": 0.763,
"step": 1488
},
{
"epoch": 1.2533136966126657,
"grad_norm": 66.27608489990234,
"learning_rate": 3.2059869390685106e-06,
"loss": 0.7587,
"step": 1489
},
{
"epoch": 1.254155270355565,
"grad_norm": 69.03297424316406,
"learning_rate": 3.2038444026273207e-06,
"loss": 0.7366,
"step": 1490
},
{
"epoch": 1.2549968440984642,
"grad_norm": 60.40120315551758,
"learning_rate": 3.2017013045802576e-06,
"loss": 0.6389,
"step": 1491
},
{
"epoch": 1.2558384178413633,
"grad_norm": 66.19833374023438,
"learning_rate": 3.1995576466373242e-06,
"loss": 0.7644,
"step": 1492
},
{
"epoch": 1.2566799915842626,
"grad_norm": 68.6891098022461,
"learning_rate": 3.197413430508971e-06,
"loss": 0.5621,
"step": 1493
},
{
"epoch": 1.2575215653271619,
"grad_norm": 88.04609680175781,
"learning_rate": 3.1952686579060943e-06,
"loss": 0.9415,
"step": 1494
},
{
"epoch": 1.258363139070061,
"grad_norm": 77.23043823242188,
"learning_rate": 3.193123330540034e-06,
"loss": 0.6554,
"step": 1495
},
{
"epoch": 1.2592047128129602,
"grad_norm": 62.28450012207031,
"learning_rate": 3.190977450122571e-06,
"loss": 0.6336,
"step": 1496
},
{
"epoch": 1.2600462865558595,
"grad_norm": 84.38449096679688,
"learning_rate": 3.188831018365931e-06,
"loss": 0.7267,
"step": 1497
},
{
"epoch": 1.2608878602987588,
"grad_norm": 55.069732666015625,
"learning_rate": 3.1866840369827764e-06,
"loss": 0.668,
"step": 1498
},
{
"epoch": 1.2617294340416578,
"grad_norm": 51.937782287597656,
"learning_rate": 3.1845365076862095e-06,
"loss": 0.6898,
"step": 1499
},
{
"epoch": 1.262571007784557,
"grad_norm": 73.25384521484375,
"learning_rate": 3.18238843218977e-06,
"loss": 0.6591,
"step": 1500
}
],
"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": 8.223871421972966e+16,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}