pick_place_go1_air_4 / trainer_state.json
MartinB7's picture
Upload folder using huggingface_hub
13c5550 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 26.455026455026456,
"eval_steps": 500,
"global_step": 5000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.05291005291005291,
"grad_norm": 3.041571855545044,
"learning_rate": 4.000000000000001e-06,
"loss": 0.9501,
"step": 10
},
{
"epoch": 0.10582010582010581,
"grad_norm": 2.1699185371398926,
"learning_rate": 8.000000000000001e-06,
"loss": 0.8764,
"step": 20
},
{
"epoch": 0.15873015873015872,
"grad_norm": 1.5091040134429932,
"learning_rate": 1.2e-05,
"loss": 0.6398,
"step": 30
},
{
"epoch": 0.21164021164021163,
"grad_norm": 0.9642059206962585,
"learning_rate": 1.6000000000000003e-05,
"loss": 0.446,
"step": 40
},
{
"epoch": 0.26455026455026454,
"grad_norm": 0.5779881477355957,
"learning_rate": 2e-05,
"loss": 0.3417,
"step": 50
},
{
"epoch": 0.31746031746031744,
"grad_norm": 0.5219764709472656,
"learning_rate": 2.4e-05,
"loss": 0.2568,
"step": 60
},
{
"epoch": 0.37037037037037035,
"grad_norm": 0.3546525239944458,
"learning_rate": 2.8000000000000003e-05,
"loss": 0.1985,
"step": 70
},
{
"epoch": 0.42328042328042326,
"grad_norm": 0.3198654353618622,
"learning_rate": 3.2000000000000005e-05,
"loss": 0.1648,
"step": 80
},
{
"epoch": 0.47619047619047616,
"grad_norm": 0.3067784011363983,
"learning_rate": 3.6e-05,
"loss": 0.1332,
"step": 90
},
{
"epoch": 0.5291005291005291,
"grad_norm": 0.2967132031917572,
"learning_rate": 4e-05,
"loss": 0.1113,
"step": 100
},
{
"epoch": 0.582010582010582,
"grad_norm": 0.233621284365654,
"learning_rate": 4.4000000000000006e-05,
"loss": 0.0984,
"step": 110
},
{
"epoch": 0.6349206349206349,
"grad_norm": 0.28300589323043823,
"learning_rate": 4.8e-05,
"loss": 0.0869,
"step": 120
},
{
"epoch": 0.6878306878306878,
"grad_norm": 0.23296041786670685,
"learning_rate": 5.2000000000000004e-05,
"loss": 0.0757,
"step": 130
},
{
"epoch": 0.7407407407407407,
"grad_norm": 0.34818753600120544,
"learning_rate": 5.6000000000000006e-05,
"loss": 0.0695,
"step": 140
},
{
"epoch": 0.7936507936507936,
"grad_norm": 0.2807776629924774,
"learning_rate": 6e-05,
"loss": 0.0592,
"step": 150
},
{
"epoch": 0.8465608465608465,
"grad_norm": 0.2950199544429779,
"learning_rate": 6.400000000000001e-05,
"loss": 0.0557,
"step": 160
},
{
"epoch": 0.8994708994708994,
"grad_norm": 0.22396177053451538,
"learning_rate": 6.800000000000001e-05,
"loss": 0.0474,
"step": 170
},
{
"epoch": 0.9523809523809523,
"grad_norm": 0.20290444791316986,
"learning_rate": 7.2e-05,
"loss": 0.0449,
"step": 180
},
{
"epoch": 1.0052910052910053,
"grad_norm": 0.2324942648410797,
"learning_rate": 7.6e-05,
"loss": 0.0415,
"step": 190
},
{
"epoch": 1.0582010582010581,
"grad_norm": 0.2845875024795532,
"learning_rate": 8e-05,
"loss": 0.0387,
"step": 200
},
{
"epoch": 1.1111111111111112,
"grad_norm": 0.2593020498752594,
"learning_rate": 8.4e-05,
"loss": 0.0353,
"step": 210
},
{
"epoch": 1.164021164021164,
"grad_norm": 0.2940255105495453,
"learning_rate": 8.800000000000001e-05,
"loss": 0.038,
"step": 220
},
{
"epoch": 1.216931216931217,
"grad_norm": 0.25001412630081177,
"learning_rate": 9.200000000000001e-05,
"loss": 0.0364,
"step": 230
},
{
"epoch": 1.2698412698412698,
"grad_norm": 0.25423663854599,
"learning_rate": 9.6e-05,
"loss": 0.0323,
"step": 240
},
{
"epoch": 1.3227513227513228,
"grad_norm": 0.20486362278461456,
"learning_rate": 0.0001,
"loss": 0.0301,
"step": 250
},
{
"epoch": 1.3756613756613756,
"grad_norm": 0.21555016934871674,
"learning_rate": 0.00010400000000000001,
"loss": 0.0279,
"step": 260
},
{
"epoch": 1.4285714285714286,
"grad_norm": 0.2603663504123688,
"learning_rate": 0.00010800000000000001,
"loss": 0.0278,
"step": 270
},
{
"epoch": 1.4814814814814814,
"grad_norm": 0.2120589166879654,
"learning_rate": 0.00011200000000000001,
"loss": 0.0262,
"step": 280
},
{
"epoch": 1.5343915343915344,
"grad_norm": 0.19613762199878693,
"learning_rate": 0.000116,
"loss": 0.0246,
"step": 290
},
{
"epoch": 1.5873015873015874,
"grad_norm": 0.2326143980026245,
"learning_rate": 0.00012,
"loss": 0.0238,
"step": 300
},
{
"epoch": 1.6402116402116402,
"grad_norm": 0.22464260458946228,
"learning_rate": 0.000124,
"loss": 0.0233,
"step": 310
},
{
"epoch": 1.693121693121693,
"grad_norm": 0.21060794591903687,
"learning_rate": 0.00012800000000000002,
"loss": 0.0216,
"step": 320
},
{
"epoch": 1.746031746031746,
"grad_norm": 0.16642867028713226,
"learning_rate": 0.000132,
"loss": 0.02,
"step": 330
},
{
"epoch": 1.798941798941799,
"grad_norm": 0.2334732860326767,
"learning_rate": 0.00013600000000000003,
"loss": 0.0199,
"step": 340
},
{
"epoch": 1.8518518518518519,
"grad_norm": 0.2562713623046875,
"learning_rate": 0.00014,
"loss": 0.0198,
"step": 350
},
{
"epoch": 1.9047619047619047,
"grad_norm": 0.21125763654708862,
"learning_rate": 0.000144,
"loss": 0.0229,
"step": 360
},
{
"epoch": 1.9576719576719577,
"grad_norm": 0.1970241814851761,
"learning_rate": 0.000148,
"loss": 0.0218,
"step": 370
},
{
"epoch": 2.0105820105820107,
"grad_norm": 0.20859766006469727,
"learning_rate": 0.000152,
"loss": 0.0209,
"step": 380
},
{
"epoch": 2.0634920634920633,
"grad_norm": 0.24649283289909363,
"learning_rate": 0.00015600000000000002,
"loss": 0.02,
"step": 390
},
{
"epoch": 2.1164021164021163,
"grad_norm": 0.2227829247713089,
"learning_rate": 0.00016,
"loss": 0.0193,
"step": 400
},
{
"epoch": 2.1693121693121693,
"grad_norm": 0.18638582527637482,
"learning_rate": 0.000164,
"loss": 0.019,
"step": 410
},
{
"epoch": 2.2222222222222223,
"grad_norm": 0.17806550860404968,
"learning_rate": 0.000168,
"loss": 0.0172,
"step": 420
},
{
"epoch": 2.2751322751322753,
"grad_norm": 0.18050020933151245,
"learning_rate": 0.000172,
"loss": 0.0174,
"step": 430
},
{
"epoch": 2.328042328042328,
"grad_norm": 0.19839580357074738,
"learning_rate": 0.00017600000000000002,
"loss": 0.017,
"step": 440
},
{
"epoch": 2.380952380952381,
"grad_norm": 0.21962286531925201,
"learning_rate": 0.00018,
"loss": 0.0156,
"step": 450
},
{
"epoch": 2.433862433862434,
"grad_norm": 0.1837696135044098,
"learning_rate": 0.00018400000000000003,
"loss": 0.0156,
"step": 460
},
{
"epoch": 2.4867724867724865,
"grad_norm": 0.1841411590576172,
"learning_rate": 0.000188,
"loss": 0.0168,
"step": 470
},
{
"epoch": 2.5396825396825395,
"grad_norm": 0.1979677677154541,
"learning_rate": 0.000192,
"loss": 0.0204,
"step": 480
},
{
"epoch": 2.5925925925925926,
"grad_norm": 0.21689844131469727,
"learning_rate": 0.000196,
"loss": 0.0241,
"step": 490
},
{
"epoch": 2.6455026455026456,
"grad_norm": 0.1965128779411316,
"learning_rate": 0.0002,
"loss": 0.0194,
"step": 500
},
{
"epoch": 2.6984126984126986,
"grad_norm": 0.2420044094324112,
"learning_rate": 0.00020400000000000003,
"loss": 0.018,
"step": 510
},
{
"epoch": 2.751322751322751,
"grad_norm": 0.21214275062084198,
"learning_rate": 0.00020800000000000001,
"loss": 0.0163,
"step": 520
},
{
"epoch": 2.804232804232804,
"grad_norm": 0.19208742678165436,
"learning_rate": 0.00021200000000000003,
"loss": 0.0175,
"step": 530
},
{
"epoch": 2.857142857142857,
"grad_norm": 0.1883249282836914,
"learning_rate": 0.00021600000000000002,
"loss": 0.0161,
"step": 540
},
{
"epoch": 2.91005291005291,
"grad_norm": 0.18176694214344025,
"learning_rate": 0.00022000000000000003,
"loss": 0.0142,
"step": 550
},
{
"epoch": 2.962962962962963,
"grad_norm": 0.2530503273010254,
"learning_rate": 0.00022400000000000002,
"loss": 0.0132,
"step": 560
},
{
"epoch": 3.015873015873016,
"grad_norm": 0.23503687977790833,
"learning_rate": 0.00022799999999999999,
"loss": 0.0142,
"step": 570
},
{
"epoch": 3.068783068783069,
"grad_norm": 0.20202770829200745,
"learning_rate": 0.000232,
"loss": 0.0142,
"step": 580
},
{
"epoch": 3.121693121693122,
"grad_norm": 0.22023585438728333,
"learning_rate": 0.000236,
"loss": 0.0138,
"step": 590
},
{
"epoch": 3.1746031746031744,
"grad_norm": 0.17895381152629852,
"learning_rate": 0.00024,
"loss": 0.0123,
"step": 600
},
{
"epoch": 3.2275132275132274,
"grad_norm": 0.19829729199409485,
"learning_rate": 0.000244,
"loss": 0.0117,
"step": 610
},
{
"epoch": 3.2804232804232805,
"grad_norm": 0.27360883355140686,
"learning_rate": 0.000248,
"loss": 0.0118,
"step": 620
},
{
"epoch": 3.3333333333333335,
"grad_norm": 0.29356101155281067,
"learning_rate": 0.000252,
"loss": 0.0147,
"step": 630
},
{
"epoch": 3.386243386243386,
"grad_norm": 0.23610763251781464,
"learning_rate": 0.00025600000000000004,
"loss": 0.0137,
"step": 640
},
{
"epoch": 3.439153439153439,
"grad_norm": 0.18409186601638794,
"learning_rate": 0.00026000000000000003,
"loss": 0.0133,
"step": 650
},
{
"epoch": 3.492063492063492,
"grad_norm": 0.20087911188602448,
"learning_rate": 0.000264,
"loss": 0.0126,
"step": 660
},
{
"epoch": 3.544973544973545,
"grad_norm": 0.2380540668964386,
"learning_rate": 0.000268,
"loss": 0.0133,
"step": 670
},
{
"epoch": 3.597883597883598,
"grad_norm": 0.19549822807312012,
"learning_rate": 0.00027200000000000005,
"loss": 0.014,
"step": 680
},
{
"epoch": 3.6507936507936507,
"grad_norm": 0.17787310481071472,
"learning_rate": 0.000276,
"loss": 0.0128,
"step": 690
},
{
"epoch": 3.7037037037037037,
"grad_norm": 0.16858863830566406,
"learning_rate": 0.00028,
"loss": 0.012,
"step": 700
},
{
"epoch": 3.7566137566137567,
"grad_norm": 0.20191702246665955,
"learning_rate": 0.000284,
"loss": 0.012,
"step": 710
},
{
"epoch": 3.8095238095238093,
"grad_norm": 0.25565165281295776,
"learning_rate": 0.000288,
"loss": 0.0126,
"step": 720
},
{
"epoch": 3.8624338624338623,
"grad_norm": 0.16845475137233734,
"learning_rate": 0.000292,
"loss": 0.0124,
"step": 730
},
{
"epoch": 3.9153439153439153,
"grad_norm": 0.19665707647800446,
"learning_rate": 0.000296,
"loss": 0.0122,
"step": 740
},
{
"epoch": 3.9682539682539684,
"grad_norm": 0.2073742151260376,
"learning_rate": 0.00030000000000000003,
"loss": 0.0111,
"step": 750
},
{
"epoch": 4.021164021164021,
"grad_norm": 0.25036758184432983,
"learning_rate": 0.000304,
"loss": 0.0122,
"step": 760
},
{
"epoch": 4.074074074074074,
"grad_norm": 0.23370663821697235,
"learning_rate": 0.000308,
"loss": 0.0121,
"step": 770
},
{
"epoch": 4.1269841269841265,
"grad_norm": 0.219792902469635,
"learning_rate": 0.00031200000000000005,
"loss": 0.0113,
"step": 780
},
{
"epoch": 4.1798941798941796,
"grad_norm": 0.13535428047180176,
"learning_rate": 0.00031600000000000004,
"loss": 0.0099,
"step": 790
},
{
"epoch": 4.232804232804233,
"grad_norm": 0.1930338442325592,
"learning_rate": 0.00032,
"loss": 0.0099,
"step": 800
},
{
"epoch": 4.285714285714286,
"grad_norm": 0.1936197578907013,
"learning_rate": 0.000324,
"loss": 0.0103,
"step": 810
},
{
"epoch": 4.338624338624339,
"grad_norm": 0.16250453889369965,
"learning_rate": 0.000328,
"loss": 0.0102,
"step": 820
},
{
"epoch": 4.391534391534392,
"grad_norm": 0.19020062685012817,
"learning_rate": 0.000332,
"loss": 0.0105,
"step": 830
},
{
"epoch": 4.444444444444445,
"grad_norm": 0.1647062748670578,
"learning_rate": 0.000336,
"loss": 0.011,
"step": 840
},
{
"epoch": 4.497354497354498,
"grad_norm": 0.20398204028606415,
"learning_rate": 0.00034,
"loss": 0.0132,
"step": 850
},
{
"epoch": 4.550264550264551,
"grad_norm": 0.2011038213968277,
"learning_rate": 0.000344,
"loss": 0.0125,
"step": 860
},
{
"epoch": 4.603174603174603,
"grad_norm": 0.20113958418369293,
"learning_rate": 0.000348,
"loss": 0.013,
"step": 870
},
{
"epoch": 4.656084656084656,
"grad_norm": 0.21507562696933746,
"learning_rate": 0.00035200000000000005,
"loss": 0.0119,
"step": 880
},
{
"epoch": 4.708994708994709,
"grad_norm": 0.26027268171310425,
"learning_rate": 0.00035600000000000003,
"loss": 0.0153,
"step": 890
},
{
"epoch": 4.761904761904762,
"grad_norm": 0.24658744037151337,
"learning_rate": 0.00036,
"loss": 0.0156,
"step": 900
},
{
"epoch": 4.814814814814815,
"grad_norm": 0.24339796602725983,
"learning_rate": 0.000364,
"loss": 0.0151,
"step": 910
},
{
"epoch": 4.867724867724868,
"grad_norm": 0.7407196164131165,
"learning_rate": 0.00036800000000000005,
"loss": 0.0246,
"step": 920
},
{
"epoch": 4.920634920634921,
"grad_norm": 0.5323154330253601,
"learning_rate": 0.00037200000000000004,
"loss": 0.0428,
"step": 930
},
{
"epoch": 4.973544973544973,
"grad_norm": 0.3756681978702545,
"learning_rate": 0.000376,
"loss": 0.0337,
"step": 940
},
{
"epoch": 5.026455026455026,
"grad_norm": 0.2930818498134613,
"learning_rate": 0.00038,
"loss": 0.0262,
"step": 950
},
{
"epoch": 5.079365079365079,
"grad_norm": 0.2086581140756607,
"learning_rate": 0.000384,
"loss": 0.0214,
"step": 960
},
{
"epoch": 5.132275132275132,
"grad_norm": 0.221287801861763,
"learning_rate": 0.000388,
"loss": 0.0232,
"step": 970
},
{
"epoch": 5.185185185185185,
"grad_norm": 0.21039186418056488,
"learning_rate": 0.000392,
"loss": 0.0169,
"step": 980
},
{
"epoch": 5.238095238095238,
"grad_norm": 0.19493348896503448,
"learning_rate": 0.00039600000000000003,
"loss": 0.0137,
"step": 990
},
{
"epoch": 5.291005291005291,
"grad_norm": 0.17548353970050812,
"learning_rate": 0.0004,
"loss": 0.0123,
"step": 1000
},
{
"epoch": 5.343915343915344,
"grad_norm": 0.1651264727115631,
"learning_rate": 0.0003999996919696056,
"loss": 0.0122,
"step": 1010
},
{
"epoch": 5.396825396825397,
"grad_norm": 0.15470008552074432,
"learning_rate": 0.0003999987678793712,
"loss": 0.0112,
"step": 1020
},
{
"epoch": 5.449735449735449,
"grad_norm": 0.1465303599834442,
"learning_rate": 0.0003999972277321432,
"loss": 0.0122,
"step": 1030
},
{
"epoch": 5.502645502645502,
"grad_norm": 0.15165336430072784,
"learning_rate": 0.0003999950715326658,
"loss": 0.0101,
"step": 1040
},
{
"epoch": 5.555555555555555,
"grad_norm": 0.14673031866550446,
"learning_rate": 0.0003999922992875807,
"loss": 0.0096,
"step": 1050
},
{
"epoch": 5.608465608465608,
"grad_norm": 0.15030068159103394,
"learning_rate": 0.0003999889110054274,
"loss": 0.0102,
"step": 1060
},
{
"epoch": 5.661375661375661,
"grad_norm": 0.1340291053056717,
"learning_rate": 0.00039998490669664266,
"loss": 0.0099,
"step": 1070
},
{
"epoch": 5.714285714285714,
"grad_norm": 0.23277540504932404,
"learning_rate": 0.00039998028637356107,
"loss": 0.0105,
"step": 1080
},
{
"epoch": 5.767195767195767,
"grad_norm": 0.15930365025997162,
"learning_rate": 0.00039997505005041456,
"loss": 0.0115,
"step": 1090
},
{
"epoch": 5.8201058201058204,
"grad_norm": 0.22390137612819672,
"learning_rate": 0.0003999691977433327,
"loss": 0.0112,
"step": 1100
},
{
"epoch": 5.8730158730158735,
"grad_norm": 0.17377761006355286,
"learning_rate": 0.0003999627294703423,
"loss": 0.0108,
"step": 1110
},
{
"epoch": 5.925925925925926,
"grad_norm": 0.20633409917354584,
"learning_rate": 0.0003999556452513676,
"loss": 0.0111,
"step": 1120
},
{
"epoch": 5.978835978835979,
"grad_norm": 0.15734145045280457,
"learning_rate": 0.0003999479451082301,
"loss": 0.0098,
"step": 1130
},
{
"epoch": 6.031746031746032,
"grad_norm": 0.1325804442167282,
"learning_rate": 0.0003999396290646487,
"loss": 0.0095,
"step": 1140
},
{
"epoch": 6.084656084656085,
"grad_norm": 0.14096157252788544,
"learning_rate": 0.00039993069714623934,
"loss": 0.0088,
"step": 1150
},
{
"epoch": 6.137566137566138,
"grad_norm": 0.1900232434272766,
"learning_rate": 0.00039992114938051487,
"loss": 0.0091,
"step": 1160
},
{
"epoch": 6.190476190476191,
"grad_norm": 0.13816455006599426,
"learning_rate": 0.0003999109857968855,
"loss": 0.0081,
"step": 1170
},
{
"epoch": 6.243386243386244,
"grad_norm": 0.17951875925064087,
"learning_rate": 0.0003999002064266581,
"loss": 0.0086,
"step": 1180
},
{
"epoch": 6.296296296296296,
"grad_norm": 0.14678612351417542,
"learning_rate": 0.0003998888113030364,
"loss": 0.0086,
"step": 1190
},
{
"epoch": 6.349206349206349,
"grad_norm": 0.12591250240802765,
"learning_rate": 0.00039987680046112085,
"loss": 0.0081,
"step": 1200
},
{
"epoch": 6.402116402116402,
"grad_norm": 0.11895497143268585,
"learning_rate": 0.00039986417393790845,
"loss": 0.0073,
"step": 1210
},
{
"epoch": 6.455026455026455,
"grad_norm": 0.14046961069107056,
"learning_rate": 0.0003998509317722928,
"loss": 0.0073,
"step": 1220
},
{
"epoch": 6.507936507936508,
"grad_norm": 0.11593084037303925,
"learning_rate": 0.00039983707400506374,
"loss": 0.0077,
"step": 1230
},
{
"epoch": 6.560846560846561,
"grad_norm": 0.1327681690454483,
"learning_rate": 0.00039982260067890737,
"loss": 0.0071,
"step": 1240
},
{
"epoch": 6.613756613756614,
"grad_norm": 0.1821785569190979,
"learning_rate": 0.00039980751183840604,
"loss": 0.0084,
"step": 1250
},
{
"epoch": 6.666666666666667,
"grad_norm": 0.14916472136974335,
"learning_rate": 0.0003997918075300379,
"loss": 0.0078,
"step": 1260
},
{
"epoch": 6.71957671957672,
"grad_norm": 0.1373327672481537,
"learning_rate": 0.000399775487802177,
"loss": 0.0084,
"step": 1270
},
{
"epoch": 6.772486772486772,
"grad_norm": 0.1542566865682602,
"learning_rate": 0.0003997585527050931,
"loss": 0.0079,
"step": 1280
},
{
"epoch": 6.825396825396825,
"grad_norm": 0.1910487413406372,
"learning_rate": 0.0003997410022909514,
"loss": 0.009,
"step": 1290
},
{
"epoch": 6.878306878306878,
"grad_norm": 0.12972815334796906,
"learning_rate": 0.0003997228366138125,
"loss": 0.0079,
"step": 1300
},
{
"epoch": 6.931216931216931,
"grad_norm": 0.09012899547815323,
"learning_rate": 0.00039970405572963225,
"loss": 0.0064,
"step": 1310
},
{
"epoch": 6.984126984126984,
"grad_norm": 0.12821277976036072,
"learning_rate": 0.0003996846596962614,
"loss": 0.0069,
"step": 1320
},
{
"epoch": 7.037037037037037,
"grad_norm": 0.1257280707359314,
"learning_rate": 0.0003996646485734458,
"loss": 0.0063,
"step": 1330
},
{
"epoch": 7.08994708994709,
"grad_norm": 0.1407238394021988,
"learning_rate": 0.00039964402242282564,
"loss": 0.0067,
"step": 1340
},
{
"epoch": 7.142857142857143,
"grad_norm": 0.16583344340324402,
"learning_rate": 0.0003996227813079357,
"loss": 0.0063,
"step": 1350
},
{
"epoch": 7.195767195767195,
"grad_norm": 0.1333070546388626,
"learning_rate": 0.00039960092529420525,
"loss": 0.0069,
"step": 1360
},
{
"epoch": 7.248677248677248,
"grad_norm": 0.11976024508476257,
"learning_rate": 0.0003995784544489573,
"loss": 0.0066,
"step": 1370
},
{
"epoch": 7.301587301587301,
"grad_norm": 0.14129309356212616,
"learning_rate": 0.00039955536884140886,
"loss": 0.0069,
"step": 1380
},
{
"epoch": 7.354497354497354,
"grad_norm": 0.11559101939201355,
"learning_rate": 0.0003995316685426708,
"loss": 0.0068,
"step": 1390
},
{
"epoch": 7.407407407407407,
"grad_norm": 0.1609719693660736,
"learning_rate": 0.00039950735362574713,
"loss": 0.0068,
"step": 1400
},
{
"epoch": 7.4603174603174605,
"grad_norm": 0.14449729025363922,
"learning_rate": 0.00039948242416553516,
"loss": 0.0063,
"step": 1410
},
{
"epoch": 7.5132275132275135,
"grad_norm": 0.1491028368473053,
"learning_rate": 0.00039945688023882515,
"loss": 0.0067,
"step": 1420
},
{
"epoch": 7.5661375661375665,
"grad_norm": 0.14751970767974854,
"learning_rate": 0.00039943072192430033,
"loss": 0.0064,
"step": 1430
},
{
"epoch": 7.619047619047619,
"grad_norm": 0.15719734132289886,
"learning_rate": 0.00039940394930253615,
"loss": 0.0063,
"step": 1440
},
{
"epoch": 7.671957671957672,
"grad_norm": 0.13127799332141876,
"learning_rate": 0.00039937656245600045,
"loss": 0.0062,
"step": 1450
},
{
"epoch": 7.724867724867725,
"grad_norm": 0.11200472712516785,
"learning_rate": 0.00039934856146905305,
"loss": 0.006,
"step": 1460
},
{
"epoch": 7.777777777777778,
"grad_norm": 0.12920986115932465,
"learning_rate": 0.0003993199464279455,
"loss": 0.0057,
"step": 1470
},
{
"epoch": 7.830687830687831,
"grad_norm": 0.11783723533153534,
"learning_rate": 0.0003992907174208207,
"loss": 0.006,
"step": 1480
},
{
"epoch": 7.883597883597884,
"grad_norm": 0.13370218873023987,
"learning_rate": 0.00039926087453771305,
"loss": 0.0065,
"step": 1490
},
{
"epoch": 7.936507936507937,
"grad_norm": 0.13977813720703125,
"learning_rate": 0.00039923041787054766,
"loss": 0.0065,
"step": 1500
},
{
"epoch": 7.98941798941799,
"grad_norm": 0.0971393808722496,
"learning_rate": 0.00039919934751314026,
"loss": 0.0061,
"step": 1510
},
{
"epoch": 8.042328042328043,
"grad_norm": 0.1217234656214714,
"learning_rate": 0.00039916766356119706,
"loss": 0.0063,
"step": 1520
},
{
"epoch": 8.095238095238095,
"grad_norm": 0.10790567845106125,
"learning_rate": 0.00039913536611231425,
"loss": 0.0061,
"step": 1530
},
{
"epoch": 8.148148148148149,
"grad_norm": 0.13306866586208344,
"learning_rate": 0.0003991024552659777,
"loss": 0.0059,
"step": 1540
},
{
"epoch": 8.201058201058201,
"grad_norm": 0.11544477194547653,
"learning_rate": 0.000399068931123563,
"loss": 0.0059,
"step": 1550
},
{
"epoch": 8.253968253968253,
"grad_norm": 0.10638356953859329,
"learning_rate": 0.00039903479378833453,
"loss": 0.006,
"step": 1560
},
{
"epoch": 8.306878306878307,
"grad_norm": 0.0943189263343811,
"learning_rate": 0.00039900004336544567,
"loss": 0.0056,
"step": 1570
},
{
"epoch": 8.359788359788359,
"grad_norm": 0.14040158689022064,
"learning_rate": 0.0003989646799619384,
"loss": 0.0056,
"step": 1580
},
{
"epoch": 8.412698412698413,
"grad_norm": 0.12987872958183289,
"learning_rate": 0.00039892870368674266,
"loss": 0.0064,
"step": 1590
},
{
"epoch": 8.465608465608465,
"grad_norm": 0.1415356993675232,
"learning_rate": 0.00039889211465067635,
"loss": 0.0066,
"step": 1600
},
{
"epoch": 8.518518518518519,
"grad_norm": 0.10001295059919357,
"learning_rate": 0.0003988549129664448,
"loss": 0.0063,
"step": 1610
},
{
"epoch": 8.571428571428571,
"grad_norm": 0.1405894011259079,
"learning_rate": 0.0003988170987486405,
"loss": 0.0055,
"step": 1620
},
{
"epoch": 8.624338624338625,
"grad_norm": 0.1409602016210556,
"learning_rate": 0.0003987786721137428,
"loss": 0.0058,
"step": 1630
},
{
"epoch": 8.677248677248677,
"grad_norm": 0.16097617149353027,
"learning_rate": 0.00039873963318011734,
"loss": 0.0095,
"step": 1640
},
{
"epoch": 8.73015873015873,
"grad_norm": 0.16123740375041962,
"learning_rate": 0.0003986999820680159,
"loss": 0.0089,
"step": 1650
},
{
"epoch": 8.783068783068783,
"grad_norm": 0.12723775207996368,
"learning_rate": 0.00039865971889957604,
"loss": 0.0072,
"step": 1660
},
{
"epoch": 8.835978835978835,
"grad_norm": 0.14602605998516083,
"learning_rate": 0.0003986188437988204,
"loss": 0.0077,
"step": 1670
},
{
"epoch": 8.88888888888889,
"grad_norm": 0.11515416204929352,
"learning_rate": 0.00039857735689165685,
"loss": 0.0072,
"step": 1680
},
{
"epoch": 8.941798941798941,
"grad_norm": 0.1307208091020584,
"learning_rate": 0.00039853525830587766,
"loss": 0.0066,
"step": 1690
},
{
"epoch": 8.994708994708995,
"grad_norm": 0.11783187836408615,
"learning_rate": 0.00039849254817115925,
"loss": 0.0063,
"step": 1700
},
{
"epoch": 9.047619047619047,
"grad_norm": 0.12625868618488312,
"learning_rate": 0.00039844922661906183,
"loss": 0.0057,
"step": 1710
},
{
"epoch": 9.100529100529101,
"grad_norm": 0.14339777827262878,
"learning_rate": 0.0003984052937830289,
"loss": 0.0082,
"step": 1720
},
{
"epoch": 9.153439153439153,
"grad_norm": 0.14355403184890747,
"learning_rate": 0.00039836074979838695,
"loss": 0.0077,
"step": 1730
},
{
"epoch": 9.206349206349206,
"grad_norm": 0.1556662619113922,
"learning_rate": 0.00039831559480234507,
"loss": 0.0085,
"step": 1740
},
{
"epoch": 9.25925925925926,
"grad_norm": 0.1498294323682785,
"learning_rate": 0.0003982698289339943,
"loss": 0.0076,
"step": 1750
},
{
"epoch": 9.312169312169312,
"grad_norm": 0.28491827845573425,
"learning_rate": 0.0003982234523343074,
"loss": 0.0144,
"step": 1760
},
{
"epoch": 9.365079365079366,
"grad_norm": 0.2578721046447754,
"learning_rate": 0.0003981764651461385,
"loss": 0.0181,
"step": 1770
},
{
"epoch": 9.417989417989418,
"grad_norm": 0.21556247770786285,
"learning_rate": 0.00039812886751422233,
"loss": 0.0141,
"step": 1780
},
{
"epoch": 9.470899470899472,
"grad_norm": 0.15084554255008698,
"learning_rate": 0.00039808065958517417,
"loss": 0.0116,
"step": 1790
},
{
"epoch": 9.523809523809524,
"grad_norm": 0.179042786359787,
"learning_rate": 0.00039803184150748893,
"loss": 0.0097,
"step": 1800
},
{
"epoch": 9.576719576719576,
"grad_norm": 0.22134196758270264,
"learning_rate": 0.00039798241343154124,
"loss": 0.0102,
"step": 1810
},
{
"epoch": 9.62962962962963,
"grad_norm": 0.20731066167354584,
"learning_rate": 0.0003979323755095846,
"loss": 0.0132,
"step": 1820
},
{
"epoch": 9.682539682539682,
"grad_norm": 0.17891205847263336,
"learning_rate": 0.000397881727895751,
"loss": 0.0129,
"step": 1830
},
{
"epoch": 9.735449735449736,
"grad_norm": 0.18633806705474854,
"learning_rate": 0.00039783047074605043,
"loss": 0.0126,
"step": 1840
},
{
"epoch": 9.788359788359788,
"grad_norm": 0.13140429556369781,
"learning_rate": 0.00039777860421837054,
"loss": 0.0093,
"step": 1850
},
{
"epoch": 9.841269841269842,
"grad_norm": 0.10666497051715851,
"learning_rate": 0.000397726128472476,
"loss": 0.0076,
"step": 1860
},
{
"epoch": 9.894179894179894,
"grad_norm": 0.11088919639587402,
"learning_rate": 0.00039767304367000807,
"loss": 0.0067,
"step": 1870
},
{
"epoch": 9.947089947089948,
"grad_norm": 0.12122470140457153,
"learning_rate": 0.00039761934997448406,
"loss": 0.0066,
"step": 1880
},
{
"epoch": 10.0,
"grad_norm": 0.11925218254327774,
"learning_rate": 0.0003975650475512969,
"loss": 0.0065,
"step": 1890
},
{
"epoch": 10.052910052910052,
"grad_norm": 0.09949792176485062,
"learning_rate": 0.00039751013656771445,
"loss": 0.0054,
"step": 1900
},
{
"epoch": 10.105820105820106,
"grad_norm": 0.09856072068214417,
"learning_rate": 0.0003974546171928793,
"loss": 0.0056,
"step": 1910
},
{
"epoch": 10.158730158730158,
"grad_norm": 0.13713252544403076,
"learning_rate": 0.0003973984895978081,
"loss": 0.0057,
"step": 1920
},
{
"epoch": 10.211640211640212,
"grad_norm": 0.15290550887584686,
"learning_rate": 0.0003973417539553908,
"loss": 0.0061,
"step": 1930
},
{
"epoch": 10.264550264550264,
"grad_norm": 0.11667478084564209,
"learning_rate": 0.00039728441044039036,
"loss": 0.0059,
"step": 1940
},
{
"epoch": 10.317460317460318,
"grad_norm": 0.12303528189659119,
"learning_rate": 0.0003972264592294423,
"loss": 0.0065,
"step": 1950
},
{
"epoch": 10.37037037037037,
"grad_norm": 0.10757434368133545,
"learning_rate": 0.0003971679005010541,
"loss": 0.0065,
"step": 1960
},
{
"epoch": 10.423280423280424,
"grad_norm": 0.11240673065185547,
"learning_rate": 0.00039710873443560426,
"loss": 0.0057,
"step": 1970
},
{
"epoch": 10.476190476190476,
"grad_norm": 0.10644736886024475,
"learning_rate": 0.00039704896121534224,
"loss": 0.0065,
"step": 1980
},
{
"epoch": 10.529100529100528,
"grad_norm": 0.10123038291931152,
"learning_rate": 0.00039698858102438786,
"loss": 0.0059,
"step": 1990
},
{
"epoch": 10.582010582010582,
"grad_norm": 0.12077871710062027,
"learning_rate": 0.00039692759404873034,
"loss": 0.0061,
"step": 2000
},
{
"epoch": 10.634920634920634,
"grad_norm": 0.13381993770599365,
"learning_rate": 0.00039686600047622817,
"loss": 0.0058,
"step": 2010
},
{
"epoch": 10.687830687830688,
"grad_norm": 0.09834510833024979,
"learning_rate": 0.0003968038004966082,
"loss": 0.0052,
"step": 2020
},
{
"epoch": 10.74074074074074,
"grad_norm": 0.11062684655189514,
"learning_rate": 0.0003967409943014654,
"loss": 0.0047,
"step": 2030
},
{
"epoch": 10.793650793650794,
"grad_norm": 0.10785150527954102,
"learning_rate": 0.0003966775820842618,
"loss": 0.0049,
"step": 2040
},
{
"epoch": 10.846560846560847,
"grad_norm": 0.08249185979366302,
"learning_rate": 0.0003966135640403264,
"loss": 0.0047,
"step": 2050
},
{
"epoch": 10.899470899470899,
"grad_norm": 0.1601790189743042,
"learning_rate": 0.00039654894036685425,
"loss": 0.0068,
"step": 2060
},
{
"epoch": 10.952380952380953,
"grad_norm": 0.1896616369485855,
"learning_rate": 0.0003964837112629058,
"loss": 0.0113,
"step": 2070
},
{
"epoch": 11.005291005291005,
"grad_norm": 0.19728443026542664,
"learning_rate": 0.0003964178769294066,
"loss": 0.0111,
"step": 2080
},
{
"epoch": 11.058201058201059,
"grad_norm": 0.18131254613399506,
"learning_rate": 0.00039635143756914636,
"loss": 0.0104,
"step": 2090
},
{
"epoch": 11.11111111111111,
"grad_norm": 0.1904458999633789,
"learning_rate": 0.00039628439338677856,
"loss": 0.0086,
"step": 2100
},
{
"epoch": 11.164021164021165,
"grad_norm": 0.13467629253864288,
"learning_rate": 0.0003962167445888196,
"loss": 0.0076,
"step": 2110
},
{
"epoch": 11.216931216931217,
"grad_norm": 0.11170188337564468,
"learning_rate": 0.0003961484913836484,
"loss": 0.0074,
"step": 2120
},
{
"epoch": 11.26984126984127,
"grad_norm": 0.14301618933677673,
"learning_rate": 0.00039607963398150545,
"loss": 0.0086,
"step": 2130
},
{
"epoch": 11.322751322751323,
"grad_norm": 0.11310645937919617,
"learning_rate": 0.00039601017259449265,
"loss": 0.007,
"step": 2140
},
{
"epoch": 11.375661375661375,
"grad_norm": 0.15265315771102905,
"learning_rate": 0.00039594010743657207,
"loss": 0.0069,
"step": 2150
},
{
"epoch": 11.428571428571429,
"grad_norm": 0.1819579005241394,
"learning_rate": 0.0003958694387235657,
"loss": 0.0064,
"step": 2160
},
{
"epoch": 11.481481481481481,
"grad_norm": 0.12576521933078766,
"learning_rate": 0.00039579816667315466,
"loss": 0.0062,
"step": 2170
},
{
"epoch": 11.534391534391535,
"grad_norm": 0.13003787398338318,
"learning_rate": 0.0003957262915048786,
"loss": 0.0062,
"step": 2180
},
{
"epoch": 11.587301587301587,
"grad_norm": 0.13775634765625,
"learning_rate": 0.00039565381344013485,
"loss": 0.0057,
"step": 2190
},
{
"epoch": 11.640211640211641,
"grad_norm": 0.16895855963230133,
"learning_rate": 0.0003955807327021778,
"loss": 0.0057,
"step": 2200
},
{
"epoch": 11.693121693121693,
"grad_norm": 0.1317339539527893,
"learning_rate": 0.0003955070495161185,
"loss": 0.0055,
"step": 2210
},
{
"epoch": 11.746031746031747,
"grad_norm": 0.12059590220451355,
"learning_rate": 0.00039543276410892336,
"loss": 0.0074,
"step": 2220
},
{
"epoch": 11.798941798941799,
"grad_norm": 0.1815410554409027,
"learning_rate": 0.00039535787670941415,
"loss": 0.0068,
"step": 2230
},
{
"epoch": 11.851851851851851,
"grad_norm": 1.0670920610427856,
"learning_rate": 0.00039528238754826675,
"loss": 0.0335,
"step": 2240
},
{
"epoch": 11.904761904761905,
"grad_norm": 1.9259690046310425,
"learning_rate": 0.00039520629685801077,
"loss": 0.2453,
"step": 2250
},
{
"epoch": 11.957671957671957,
"grad_norm": 1.7839056253433228,
"learning_rate": 0.00039512960487302865,
"loss": 0.4239,
"step": 2260
},
{
"epoch": 12.010582010582011,
"grad_norm": 1.0937215089797974,
"learning_rate": 0.000395052311829555,
"loss": 0.2597,
"step": 2270
},
{
"epoch": 12.063492063492063,
"grad_norm": 0.693492591381073,
"learning_rate": 0.0003949744179656759,
"loss": 0.1695,
"step": 2280
},
{
"epoch": 12.116402116402117,
"grad_norm": 0.7162590026855469,
"learning_rate": 0.0003948959235213281,
"loss": 0.1095,
"step": 2290
},
{
"epoch": 12.16931216931217,
"grad_norm": 0.6560825109481812,
"learning_rate": 0.00039481682873829835,
"loss": 0.092,
"step": 2300
},
{
"epoch": 12.222222222222221,
"grad_norm": 0.49403268098831177,
"learning_rate": 0.0003947371338602227,
"loss": 0.0659,
"step": 2310
},
{
"epoch": 12.275132275132275,
"grad_norm": 0.4173797070980072,
"learning_rate": 0.0003946568391325855,
"loss": 0.0482,
"step": 2320
},
{
"epoch": 12.328042328042327,
"grad_norm": 0.3571629226207733,
"learning_rate": 0.00039457594480271895,
"loss": 0.0364,
"step": 2330
},
{
"epoch": 12.380952380952381,
"grad_norm": 0.25554782152175903,
"learning_rate": 0.00039449445111980217,
"loss": 0.027,
"step": 2340
},
{
"epoch": 12.433862433862434,
"grad_norm": 0.20498105883598328,
"learning_rate": 0.00039441235833486045,
"loss": 0.0205,
"step": 2350
},
{
"epoch": 12.486772486772487,
"grad_norm": 0.29463911056518555,
"learning_rate": 0.0003943296667007646,
"loss": 0.019,
"step": 2360
},
{
"epoch": 12.53968253968254,
"grad_norm": 0.16020773351192474,
"learning_rate": 0.0003942463764722299,
"loss": 0.0209,
"step": 2370
},
{
"epoch": 12.592592592592592,
"grad_norm": 0.17252512276172638,
"learning_rate": 0.00039416248790581567,
"loss": 0.0169,
"step": 2380
},
{
"epoch": 12.645502645502646,
"grad_norm": 0.18067197501659393,
"learning_rate": 0.0003940780012599241,
"loss": 0.0163,
"step": 2390
},
{
"epoch": 12.698412698412698,
"grad_norm": 0.12952257692813873,
"learning_rate": 0.0003939929167947997,
"loss": 0.0135,
"step": 2400
},
{
"epoch": 12.751322751322752,
"grad_norm": 0.14395594596862793,
"learning_rate": 0.00039390723477252866,
"loss": 0.0122,
"step": 2410
},
{
"epoch": 12.804232804232804,
"grad_norm": 0.11114363372325897,
"learning_rate": 0.00039382095545703746,
"loss": 0.0113,
"step": 2420
},
{
"epoch": 12.857142857142858,
"grad_norm": 0.10381834954023361,
"learning_rate": 0.0003937340791140927,
"loss": 0.0105,
"step": 2430
},
{
"epoch": 12.91005291005291,
"grad_norm": 0.1080816239118576,
"learning_rate": 0.00039364660601129994,
"loss": 0.0103,
"step": 2440
},
{
"epoch": 12.962962962962964,
"grad_norm": 0.09894987940788269,
"learning_rate": 0.0003935585364181029,
"loss": 0.0102,
"step": 2450
},
{
"epoch": 13.015873015873016,
"grad_norm": 0.10151786357164383,
"learning_rate": 0.00039346987060578264,
"loss": 0.0093,
"step": 2460
},
{
"epoch": 13.068783068783068,
"grad_norm": 0.09461592882871628,
"learning_rate": 0.0003933806088474569,
"loss": 0.0091,
"step": 2470
},
{
"epoch": 13.121693121693122,
"grad_norm": 0.10612160712480545,
"learning_rate": 0.00039329075141807904,
"loss": 0.0089,
"step": 2480
},
{
"epoch": 13.174603174603174,
"grad_norm": 0.11442465335130692,
"learning_rate": 0.00039320029859443717,
"loss": 0.01,
"step": 2490
},
{
"epoch": 13.227513227513228,
"grad_norm": 0.16065755486488342,
"learning_rate": 0.00039310925065515353,
"loss": 0.009,
"step": 2500
},
{
"epoch": 13.28042328042328,
"grad_norm": 0.11742605268955231,
"learning_rate": 0.00039301760788068346,
"loss": 0.009,
"step": 2510
},
{
"epoch": 13.333333333333334,
"grad_norm": 0.1333356350660324,
"learning_rate": 0.0003929253705533144,
"loss": 0.0095,
"step": 2520
},
{
"epoch": 13.386243386243386,
"grad_norm": 0.16482898592948914,
"learning_rate": 0.0003928325389571656,
"loss": 0.0118,
"step": 2530
},
{
"epoch": 13.43915343915344,
"grad_norm": 0.11537864804267883,
"learning_rate": 0.0003927391133781865,
"loss": 0.0105,
"step": 2540
},
{
"epoch": 13.492063492063492,
"grad_norm": 0.19208720326423645,
"learning_rate": 0.0003926450941041562,
"loss": 0.0097,
"step": 2550
},
{
"epoch": 13.544973544973544,
"grad_norm": 0.15091383457183838,
"learning_rate": 0.00039255048142468275,
"loss": 0.011,
"step": 2560
},
{
"epoch": 13.597883597883598,
"grad_norm": 0.11797965317964554,
"learning_rate": 0.0003924552756312019,
"loss": 0.0086,
"step": 2570
},
{
"epoch": 13.65079365079365,
"grad_norm": 0.09077858924865723,
"learning_rate": 0.00039235947701697643,
"loss": 0.0081,
"step": 2580
},
{
"epoch": 13.703703703703704,
"grad_norm": 0.09163253009319305,
"learning_rate": 0.0003922630858770951,
"loss": 0.0071,
"step": 2590
},
{
"epoch": 13.756613756613756,
"grad_norm": 0.09899520128965378,
"learning_rate": 0.00039216610250847214,
"loss": 0.0064,
"step": 2600
},
{
"epoch": 13.80952380952381,
"grad_norm": 0.13137751817703247,
"learning_rate": 0.0003920685272098457,
"loss": 0.0073,
"step": 2610
},
{
"epoch": 13.862433862433862,
"grad_norm": 0.1063733696937561,
"learning_rate": 0.0003919703602817772,
"loss": 0.007,
"step": 2620
},
{
"epoch": 13.915343915343914,
"grad_norm": 0.09142506867647171,
"learning_rate": 0.0003918716020266509,
"loss": 0.0068,
"step": 2630
},
{
"epoch": 13.968253968253968,
"grad_norm": 0.1315004676580429,
"learning_rate": 0.00039177225274867196,
"loss": 0.0065,
"step": 2640
},
{
"epoch": 14.02116402116402,
"grad_norm": 0.09566756337881088,
"learning_rate": 0.00039167231275386656,
"loss": 0.0061,
"step": 2650
},
{
"epoch": 14.074074074074074,
"grad_norm": 0.10887089371681213,
"learning_rate": 0.0003915717823500802,
"loss": 0.0055,
"step": 2660
},
{
"epoch": 14.126984126984127,
"grad_norm": 0.09512364864349365,
"learning_rate": 0.00039147066184697706,
"loss": 0.0057,
"step": 2670
},
{
"epoch": 14.17989417989418,
"grad_norm": 0.09984894841909409,
"learning_rate": 0.00039136895155603905,
"loss": 0.0058,
"step": 2680
},
{
"epoch": 14.232804232804233,
"grad_norm": 0.0963057205080986,
"learning_rate": 0.0003912666517905647,
"loss": 0.0055,
"step": 2690
},
{
"epoch": 14.285714285714286,
"grad_norm": 0.09049777686595917,
"learning_rate": 0.0003911637628656685,
"loss": 0.0062,
"step": 2700
},
{
"epoch": 14.338624338624339,
"grad_norm": 0.10670194029808044,
"learning_rate": 0.00039106028509827955,
"loss": 0.006,
"step": 2710
},
{
"epoch": 14.39153439153439,
"grad_norm": 0.0816589891910553,
"learning_rate": 0.0003909562188071408,
"loss": 0.0056,
"step": 2720
},
{
"epoch": 14.444444444444445,
"grad_norm": 0.10106207430362701,
"learning_rate": 0.00039085156431280813,
"loss": 0.0057,
"step": 2730
},
{
"epoch": 14.497354497354497,
"grad_norm": 0.08731462806463242,
"learning_rate": 0.0003907463219376491,
"loss": 0.0054,
"step": 2740
},
{
"epoch": 14.55026455026455,
"grad_norm": 0.09356683492660522,
"learning_rate": 0.00039064049200584225,
"loss": 0.0053,
"step": 2750
},
{
"epoch": 14.603174603174603,
"grad_norm": 0.10447552055120468,
"learning_rate": 0.000390534074843376,
"loss": 0.0054,
"step": 2760
},
{
"epoch": 14.656084656084657,
"grad_norm": 0.11416878551244736,
"learning_rate": 0.0003904270707780475,
"loss": 0.0055,
"step": 2770
},
{
"epoch": 14.708994708994709,
"grad_norm": 0.0949578583240509,
"learning_rate": 0.00039031948013946175,
"loss": 0.0053,
"step": 2780
},
{
"epoch": 14.761904761904763,
"grad_norm": 0.08923380821943283,
"learning_rate": 0.00039021130325903074,
"loss": 0.0052,
"step": 2790
},
{
"epoch": 14.814814814814815,
"grad_norm": 0.09841927140951157,
"learning_rate": 0.000390102540469972,
"loss": 0.005,
"step": 2800
},
{
"epoch": 14.867724867724867,
"grad_norm": 0.07885875552892685,
"learning_rate": 0.0003899931921073081,
"loss": 0.0049,
"step": 2810
},
{
"epoch": 14.920634920634921,
"grad_norm": 0.0910676047205925,
"learning_rate": 0.0003898832585078652,
"loss": 0.0048,
"step": 2820
},
{
"epoch": 14.973544973544973,
"grad_norm": 0.08391734957695007,
"learning_rate": 0.00038977274001027206,
"loss": 0.0049,
"step": 2830
},
{
"epoch": 15.026455026455027,
"grad_norm": 0.0831340104341507,
"learning_rate": 0.00038966163695495945,
"loss": 0.0049,
"step": 2840
},
{
"epoch": 15.079365079365079,
"grad_norm": 0.10346338152885437,
"learning_rate": 0.00038954994968415846,
"loss": 0.0055,
"step": 2850
},
{
"epoch": 15.132275132275133,
"grad_norm": 0.09366438537836075,
"learning_rate": 0.0003894376785418998,
"loss": 0.0054,
"step": 2860
},
{
"epoch": 15.185185185185185,
"grad_norm": 0.09279809892177582,
"learning_rate": 0.0003893248238740128,
"loss": 0.0056,
"step": 2870
},
{
"epoch": 15.238095238095237,
"grad_norm": 0.08680757135152817,
"learning_rate": 0.00038921138602812406,
"loss": 0.005,
"step": 2880
},
{
"epoch": 15.291005291005291,
"grad_norm": 0.08765140175819397,
"learning_rate": 0.0003890973653536566,
"loss": 0.0048,
"step": 2890
},
{
"epoch": 15.343915343915343,
"grad_norm": 0.08962959051132202,
"learning_rate": 0.0003889827622018289,
"loss": 0.0046,
"step": 2900
},
{
"epoch": 15.396825396825397,
"grad_norm": 0.0812198743224144,
"learning_rate": 0.00038886757692565334,
"loss": 0.005,
"step": 2910
},
{
"epoch": 15.44973544973545,
"grad_norm": 0.10678814351558685,
"learning_rate": 0.00038875180987993564,
"loss": 0.005,
"step": 2920
},
{
"epoch": 15.502645502645503,
"grad_norm": 0.06816115230321884,
"learning_rate": 0.0003886354614212735,
"loss": 0.0048,
"step": 2930
},
{
"epoch": 15.555555555555555,
"grad_norm": 0.09805359691381454,
"learning_rate": 0.0003885185319080555,
"loss": 0.0047,
"step": 2940
},
{
"epoch": 15.60846560846561,
"grad_norm": 0.082832932472229,
"learning_rate": 0.0003884010217004601,
"loss": 0.0046,
"step": 2950
},
{
"epoch": 15.661375661375661,
"grad_norm": 0.06807373464107513,
"learning_rate": 0.0003882829311604545,
"loss": 0.0047,
"step": 2960
},
{
"epoch": 15.714285714285714,
"grad_norm": 0.0932752788066864,
"learning_rate": 0.0003881642606517934,
"loss": 0.0051,
"step": 2970
},
{
"epoch": 15.767195767195767,
"grad_norm": 0.0888075903058052,
"learning_rate": 0.0003880450105400181,
"loss": 0.0046,
"step": 2980
},
{
"epoch": 15.82010582010582,
"grad_norm": 0.09350543469190598,
"learning_rate": 0.0003879251811924551,
"loss": 0.0044,
"step": 2990
},
{
"epoch": 15.873015873015873,
"grad_norm": 0.06647849828004837,
"learning_rate": 0.0003878047729782153,
"loss": 0.0041,
"step": 3000
},
{
"epoch": 15.925925925925926,
"grad_norm": 0.06874548643827438,
"learning_rate": 0.00038768378626819253,
"loss": 0.004,
"step": 3010
},
{
"epoch": 15.97883597883598,
"grad_norm": 0.08275860548019409,
"learning_rate": 0.0003875622214350626,
"loss": 0.0043,
"step": 3020
},
{
"epoch": 16.03174603174603,
"grad_norm": 0.07945531606674194,
"learning_rate": 0.00038744007885328227,
"loss": 0.0041,
"step": 3030
},
{
"epoch": 16.084656084656086,
"grad_norm": 0.06742502003908157,
"learning_rate": 0.00038731735889908773,
"loss": 0.0038,
"step": 3040
},
{
"epoch": 16.137566137566136,
"grad_norm": 0.08099709451198578,
"learning_rate": 0.0003871940619504938,
"loss": 0.004,
"step": 3050
},
{
"epoch": 16.19047619047619,
"grad_norm": 0.07479023188352585,
"learning_rate": 0.0003870701883872924,
"loss": 0.0041,
"step": 3060
},
{
"epoch": 16.243386243386244,
"grad_norm": 0.07875571399927139,
"learning_rate": 0.00038694573859105196,
"loss": 0.0041,
"step": 3070
},
{
"epoch": 16.296296296296298,
"grad_norm": 0.0765179693698883,
"learning_rate": 0.0003868207129451155,
"loss": 0.004,
"step": 3080
},
{
"epoch": 16.349206349206348,
"grad_norm": 0.07733309268951416,
"learning_rate": 0.0003866951118346001,
"loss": 0.0041,
"step": 3090
},
{
"epoch": 16.402116402116402,
"grad_norm": 0.11193585395812988,
"learning_rate": 0.00038656893564639543,
"loss": 0.0041,
"step": 3100
},
{
"epoch": 16.455026455026456,
"grad_norm": 0.0782993957400322,
"learning_rate": 0.0003864421847691624,
"loss": 0.0048,
"step": 3110
},
{
"epoch": 16.507936507936506,
"grad_norm": 0.11347425729036331,
"learning_rate": 0.00038631485959333224,
"loss": 0.0042,
"step": 3120
},
{
"epoch": 16.56084656084656,
"grad_norm": 0.12052897363901138,
"learning_rate": 0.0003861869605111053,
"loss": 0.0042,
"step": 3130
},
{
"epoch": 16.613756613756614,
"grad_norm": 0.1033359244465828,
"learning_rate": 0.00038605848791644937,
"loss": 0.0045,
"step": 3140
},
{
"epoch": 16.666666666666668,
"grad_norm": 0.083700992166996,
"learning_rate": 0.0003859294422050994,
"loss": 0.0045,
"step": 3150
},
{
"epoch": 16.719576719576718,
"grad_norm": 0.10319680720567703,
"learning_rate": 0.00038579982377455517,
"loss": 0.0039,
"step": 3160
},
{
"epoch": 16.772486772486772,
"grad_norm": 0.09331599622964859,
"learning_rate": 0.00038566963302408096,
"loss": 0.0039,
"step": 3170
},
{
"epoch": 16.825396825396826,
"grad_norm": 0.08213219791650772,
"learning_rate": 0.00038553887035470383,
"loss": 0.0038,
"step": 3180
},
{
"epoch": 16.87830687830688,
"grad_norm": 0.10324909538030624,
"learning_rate": 0.00038540753616921254,
"loss": 0.0045,
"step": 3190
},
{
"epoch": 16.93121693121693,
"grad_norm": 0.10563361644744873,
"learning_rate": 0.0003852756308721563,
"loss": 0.0044,
"step": 3200
},
{
"epoch": 16.984126984126984,
"grad_norm": 0.08582045137882233,
"learning_rate": 0.0003851431548698435,
"loss": 0.0038,
"step": 3210
},
{
"epoch": 17.037037037037038,
"grad_norm": 0.07344099879264832,
"learning_rate": 0.0003850101085703405,
"loss": 0.0038,
"step": 3220
},
{
"epoch": 17.08994708994709,
"grad_norm": 0.1084500402212143,
"learning_rate": 0.0003848764923834704,
"loss": 0.004,
"step": 3230
},
{
"epoch": 17.142857142857142,
"grad_norm": 0.0705375075340271,
"learning_rate": 0.00038474230672081166,
"loss": 0.0036,
"step": 3240
},
{
"epoch": 17.195767195767196,
"grad_norm": 0.0940157026052475,
"learning_rate": 0.00038460755199569684,
"loss": 0.0035,
"step": 3250
},
{
"epoch": 17.24867724867725,
"grad_norm": 0.062344059348106384,
"learning_rate": 0.0003844722286232114,
"loss": 0.0034,
"step": 3260
},
{
"epoch": 17.3015873015873,
"grad_norm": 0.06925034523010254,
"learning_rate": 0.00038433633702019256,
"loss": 0.0034,
"step": 3270
},
{
"epoch": 17.354497354497354,
"grad_norm": 0.08498704433441162,
"learning_rate": 0.0003841998776052278,
"loss": 0.0037,
"step": 3280
},
{
"epoch": 17.40740740740741,
"grad_norm": 0.0821915790438652,
"learning_rate": 0.00038406285079865344,
"loss": 0.0038,
"step": 3290
},
{
"epoch": 17.46031746031746,
"grad_norm": 0.08294922858476639,
"learning_rate": 0.00038392525702255387,
"loss": 0.0036,
"step": 3300
},
{
"epoch": 17.513227513227513,
"grad_norm": 0.08474984765052795,
"learning_rate": 0.0003837870967007596,
"loss": 0.0035,
"step": 3310
},
{
"epoch": 17.566137566137566,
"grad_norm": 0.06738153845071793,
"learning_rate": 0.00038364837025884653,
"loss": 0.0038,
"step": 3320
},
{
"epoch": 17.61904761904762,
"grad_norm": 0.08157813549041748,
"learning_rate": 0.00038350907812413415,
"loss": 0.0043,
"step": 3330
},
{
"epoch": 17.67195767195767,
"grad_norm": 0.11258726567029953,
"learning_rate": 0.00038336922072568463,
"loss": 0.0041,
"step": 3340
},
{
"epoch": 17.724867724867725,
"grad_norm": 0.08987373113632202,
"learning_rate": 0.0003832287984943013,
"loss": 0.0036,
"step": 3350
},
{
"epoch": 17.77777777777778,
"grad_norm": 0.07632698118686676,
"learning_rate": 0.0003830878118625272,
"loss": 0.0033,
"step": 3360
},
{
"epoch": 17.83068783068783,
"grad_norm": 0.07706711441278458,
"learning_rate": 0.0003829462612646441,
"loss": 0.0035,
"step": 3370
},
{
"epoch": 17.883597883597883,
"grad_norm": 0.06717482209205627,
"learning_rate": 0.0003828041471366709,
"loss": 0.0034,
"step": 3380
},
{
"epoch": 17.936507936507937,
"grad_norm": 0.08264566957950592,
"learning_rate": 0.00038266146991636224,
"loss": 0.0037,
"step": 3390
},
{
"epoch": 17.98941798941799,
"grad_norm": 0.08034229278564453,
"learning_rate": 0.0003825182300432073,
"loss": 0.0033,
"step": 3400
},
{
"epoch": 18.04232804232804,
"grad_norm": 0.08124092221260071,
"learning_rate": 0.0003823744279584285,
"loss": 0.0034,
"step": 3410
},
{
"epoch": 18.095238095238095,
"grad_norm": 0.06899471580982208,
"learning_rate": 0.00038223006410497996,
"loss": 0.0033,
"step": 3420
},
{
"epoch": 18.14814814814815,
"grad_norm": 0.12756113708019257,
"learning_rate": 0.00038208513892754617,
"loss": 0.0037,
"step": 3430
},
{
"epoch": 18.201058201058203,
"grad_norm": 0.08263306319713593,
"learning_rate": 0.0003819396528725408,
"loss": 0.0036,
"step": 3440
},
{
"epoch": 18.253968253968253,
"grad_norm": 0.07276030629873276,
"learning_rate": 0.00038179360638810505,
"loss": 0.0038,
"step": 3450
},
{
"epoch": 18.306878306878307,
"grad_norm": 0.060400381684303284,
"learning_rate": 0.0003816469999241065,
"loss": 0.0037,
"step": 3460
},
{
"epoch": 18.35978835978836,
"grad_norm": 0.09411425143480301,
"learning_rate": 0.00038149983393213763,
"loss": 0.0037,
"step": 3470
},
{
"epoch": 18.41269841269841,
"grad_norm": 0.07314950972795486,
"learning_rate": 0.0003813521088655144,
"loss": 0.0039,
"step": 3480
},
{
"epoch": 18.465608465608465,
"grad_norm": 0.0763072744011879,
"learning_rate": 0.000381203825179275,
"loss": 0.0033,
"step": 3490
},
{
"epoch": 18.51851851851852,
"grad_norm": 0.0973953828215599,
"learning_rate": 0.00038105498333017816,
"loss": 0.0033,
"step": 3500
},
{
"epoch": 18.571428571428573,
"grad_norm": 0.10556458681821823,
"learning_rate": 0.000380905583776702,
"loss": 0.0035,
"step": 3510
},
{
"epoch": 18.624338624338623,
"grad_norm": 0.11148570477962494,
"learning_rate": 0.0003807556269790427,
"loss": 0.0037,
"step": 3520
},
{
"epoch": 18.677248677248677,
"grad_norm": 0.09374136477708817,
"learning_rate": 0.0003806051133991127,
"loss": 0.0036,
"step": 3530
},
{
"epoch": 18.73015873015873,
"grad_norm": 0.08377573639154434,
"learning_rate": 0.0003804540435005395,
"loss": 0.0035,
"step": 3540
},
{
"epoch": 18.78306878306878,
"grad_norm": 0.09433193504810333,
"learning_rate": 0.00038030241774866433,
"loss": 0.0034,
"step": 3550
},
{
"epoch": 18.835978835978835,
"grad_norm": 0.07314567267894745,
"learning_rate": 0.00038015023661054057,
"loss": 0.0035,
"step": 3560
},
{
"epoch": 18.88888888888889,
"grad_norm": 0.10299256443977356,
"learning_rate": 0.00037999750055493255,
"loss": 0.0035,
"step": 3570
},
{
"epoch": 18.941798941798943,
"grad_norm": 0.06799434870481491,
"learning_rate": 0.00037984421005231355,
"loss": 0.0032,
"step": 3580
},
{
"epoch": 18.994708994708994,
"grad_norm": 0.058005448430776596,
"learning_rate": 0.000379690365574865,
"loss": 0.0029,
"step": 3590
},
{
"epoch": 19.047619047619047,
"grad_norm": 0.06635001301765442,
"learning_rate": 0.0003795359675964746,
"loss": 0.0029,
"step": 3600
},
{
"epoch": 19.1005291005291,
"grad_norm": 0.11281086504459381,
"learning_rate": 0.0003793810165927352,
"loss": 0.0033,
"step": 3610
},
{
"epoch": 19.15343915343915,
"grad_norm": 0.08719826489686966,
"learning_rate": 0.00037922551304094275,
"loss": 0.0034,
"step": 3620
},
{
"epoch": 19.206349206349206,
"grad_norm": 0.1060280054807663,
"learning_rate": 0.00037906945742009567,
"loss": 0.0031,
"step": 3630
},
{
"epoch": 19.25925925925926,
"grad_norm": 0.07241532951593399,
"learning_rate": 0.0003789128502108925,
"loss": 0.0031,
"step": 3640
},
{
"epoch": 19.312169312169313,
"grad_norm": 0.08218563348054886,
"learning_rate": 0.00037875569189573123,
"loss": 0.0033,
"step": 3650
},
{
"epoch": 19.365079365079364,
"grad_norm": 0.1528523564338684,
"learning_rate": 0.00037859798295870715,
"loss": 0.0036,
"step": 3660
},
{
"epoch": 19.417989417989418,
"grad_norm": 0.11239644885063171,
"learning_rate": 0.00037843972388561177,
"loss": 0.0039,
"step": 3670
},
{
"epoch": 19.47089947089947,
"grad_norm": 0.10018712282180786,
"learning_rate": 0.00037828091516393106,
"loss": 0.0036,
"step": 3680
},
{
"epoch": 19.523809523809526,
"grad_norm": 0.07450844347476959,
"learning_rate": 0.0003781215572828442,
"loss": 0.0033,
"step": 3690
},
{
"epoch": 19.576719576719576,
"grad_norm": 0.0734860748052597,
"learning_rate": 0.0003779616507332219,
"loss": 0.0032,
"step": 3700
},
{
"epoch": 19.62962962962963,
"grad_norm": 0.07196395099163055,
"learning_rate": 0.00037780119600762494,
"loss": 0.0031,
"step": 3710
},
{
"epoch": 19.682539682539684,
"grad_norm": 0.09709396213293076,
"learning_rate": 0.00037764019360030256,
"loss": 0.0032,
"step": 3720
},
{
"epoch": 19.735449735449734,
"grad_norm": 0.09168734401464462,
"learning_rate": 0.00037747864400719126,
"loss": 0.0033,
"step": 3730
},
{
"epoch": 19.788359788359788,
"grad_norm": 0.07125968486070633,
"learning_rate": 0.0003773165477259128,
"loss": 0.0034,
"step": 3740
},
{
"epoch": 19.841269841269842,
"grad_norm": 0.07378100603818893,
"learning_rate": 0.000377153905255773,
"loss": 0.0033,
"step": 3750
},
{
"epoch": 19.894179894179896,
"grad_norm": 0.07083296030759811,
"learning_rate": 0.0003769907170977601,
"loss": 0.0032,
"step": 3760
},
{
"epoch": 19.947089947089946,
"grad_norm": 0.06470996886491776,
"learning_rate": 0.00037682698375454324,
"loss": 0.003,
"step": 3770
},
{
"epoch": 20.0,
"grad_norm": 0.09688153117895126,
"learning_rate": 0.00037666270573047085,
"loss": 0.0033,
"step": 3780
},
{
"epoch": 20.052910052910054,
"grad_norm": 0.11122123897075653,
"learning_rate": 0.00037649788353156925,
"loss": 0.0036,
"step": 3790
},
{
"epoch": 20.105820105820104,
"grad_norm": 0.08080142736434937,
"learning_rate": 0.00037633251766554086,
"loss": 0.0033,
"step": 3800
},
{
"epoch": 20.158730158730158,
"grad_norm": 0.08291842043399811,
"learning_rate": 0.0003761666086417628,
"loss": 0.0034,
"step": 3810
},
{
"epoch": 20.211640211640212,
"grad_norm": 0.11718545109033585,
"learning_rate": 0.0003760001569712853,
"loss": 0.0037,
"step": 3820
},
{
"epoch": 20.264550264550266,
"grad_norm": 0.09750944375991821,
"learning_rate": 0.0003758331631668301,
"loss": 0.0037,
"step": 3830
},
{
"epoch": 20.317460317460316,
"grad_norm": 0.11106361448764801,
"learning_rate": 0.00037566562774278886,
"loss": 0.0032,
"step": 3840
},
{
"epoch": 20.37037037037037,
"grad_norm": 0.0689949169754982,
"learning_rate": 0.0003754975512152216,
"loss": 0.0035,
"step": 3850
},
{
"epoch": 20.423280423280424,
"grad_norm": 0.08527049422264099,
"learning_rate": 0.0003753289341018552,
"loss": 0.003,
"step": 3860
},
{
"epoch": 20.476190476190474,
"grad_norm": 0.0683789923787117,
"learning_rate": 0.00037515977692208154,
"loss": 0.0027,
"step": 3870
},
{
"epoch": 20.52910052910053,
"grad_norm": 0.06511671096086502,
"learning_rate": 0.00037499008019695613,
"loss": 0.0031,
"step": 3880
},
{
"epoch": 20.582010582010582,
"grad_norm": 0.09722796827554703,
"learning_rate": 0.00037481984444919647,
"loss": 0.0032,
"step": 3890
},
{
"epoch": 20.634920634920636,
"grad_norm": 0.0859525129199028,
"learning_rate": 0.0003746490702031805,
"loss": 0.0034,
"step": 3900
},
{
"epoch": 20.687830687830687,
"grad_norm": 0.07728194445371628,
"learning_rate": 0.0003744777579849447,
"loss": 0.003,
"step": 3910
},
{
"epoch": 20.74074074074074,
"grad_norm": 0.05846023187041283,
"learning_rate": 0.00037430590832218273,
"loss": 0.003,
"step": 3920
},
{
"epoch": 20.793650793650794,
"grad_norm": 0.06172306835651398,
"learning_rate": 0.00037413352174424396,
"loss": 0.0028,
"step": 3930
},
{
"epoch": 20.84656084656085,
"grad_norm": 0.08405909687280655,
"learning_rate": 0.0003739605987821313,
"loss": 0.0028,
"step": 3940
},
{
"epoch": 20.8994708994709,
"grad_norm": 0.07122877985239029,
"learning_rate": 0.0003737871399685001,
"loss": 0.0028,
"step": 3950
},
{
"epoch": 20.952380952380953,
"grad_norm": 0.06499795615673065,
"learning_rate": 0.00037361314583765615,
"loss": 0.0029,
"step": 3960
},
{
"epoch": 21.005291005291006,
"grad_norm": 0.09252621978521347,
"learning_rate": 0.0003734386169255544,
"loss": 0.0033,
"step": 3970
},
{
"epoch": 21.058201058201057,
"grad_norm": 0.0944569781422615,
"learning_rate": 0.00037326355376979676,
"loss": 0.003,
"step": 3980
},
{
"epoch": 21.11111111111111,
"grad_norm": 0.09000955522060394,
"learning_rate": 0.00037308795690963104,
"loss": 0.0031,
"step": 3990
},
{
"epoch": 21.164021164021165,
"grad_norm": 0.06754779070615768,
"learning_rate": 0.000372911826885949,
"loss": 0.0027,
"step": 4000
},
{
"epoch": 21.21693121693122,
"grad_norm": 0.08070409297943115,
"learning_rate": 0.00037273516424128465,
"loss": 0.0027,
"step": 4010
},
{
"epoch": 21.26984126984127,
"grad_norm": 0.09665092825889587,
"learning_rate": 0.00037255796951981255,
"loss": 0.003,
"step": 4020
},
{
"epoch": 21.322751322751323,
"grad_norm": 0.09719184041023254,
"learning_rate": 0.0003723802432673463,
"loss": 0.0031,
"step": 4030
},
{
"epoch": 21.375661375661377,
"grad_norm": 0.08830776065587997,
"learning_rate": 0.0003722019860313369,
"loss": 0.0031,
"step": 4040
},
{
"epoch": 21.428571428571427,
"grad_norm": 0.09745042771100998,
"learning_rate": 0.0003720231983608706,
"loss": 0.0034,
"step": 4050
},
{
"epoch": 21.48148148148148,
"grad_norm": 0.07377637177705765,
"learning_rate": 0.00037184388080666796,
"loss": 0.0027,
"step": 4060
},
{
"epoch": 21.534391534391535,
"grad_norm": 0.06053460016846657,
"learning_rate": 0.0003716640339210815,
"loss": 0.0028,
"step": 4070
},
{
"epoch": 21.58730158730159,
"grad_norm": 0.0902014896273613,
"learning_rate": 0.0003714836582580942,
"loss": 0.0031,
"step": 4080
},
{
"epoch": 21.64021164021164,
"grad_norm": 0.10117074847221375,
"learning_rate": 0.00037130275437331805,
"loss": 0.003,
"step": 4090
},
{
"epoch": 21.693121693121693,
"grad_norm": 0.0773254930973053,
"learning_rate": 0.000371121322823992,
"loss": 0.0029,
"step": 4100
},
{
"epoch": 21.746031746031747,
"grad_norm": 0.09695654362440109,
"learning_rate": 0.00037093936416898027,
"loss": 0.0032,
"step": 4110
},
{
"epoch": 21.798941798941797,
"grad_norm": 0.08522791415452957,
"learning_rate": 0.00037075687896877084,
"loss": 0.003,
"step": 4120
},
{
"epoch": 21.85185185185185,
"grad_norm": 0.06715506315231323,
"learning_rate": 0.0003705738677854737,
"loss": 0.0031,
"step": 4130
},
{
"epoch": 21.904761904761905,
"grad_norm": 0.07648234814405441,
"learning_rate": 0.0003703903311828188,
"loss": 0.0029,
"step": 4140
},
{
"epoch": 21.95767195767196,
"grad_norm": 0.07449754327535629,
"learning_rate": 0.00037020626972615465,
"loss": 0.0032,
"step": 4150
},
{
"epoch": 22.01058201058201,
"grad_norm": 0.09906895458698273,
"learning_rate": 0.00037002168398244664,
"loss": 0.0033,
"step": 4160
},
{
"epoch": 22.063492063492063,
"grad_norm": 0.0835467204451561,
"learning_rate": 0.00036983657452027474,
"loss": 0.0031,
"step": 4170
},
{
"epoch": 22.116402116402117,
"grad_norm": 0.09476753324270248,
"learning_rate": 0.00036965094190983253,
"loss": 0.003,
"step": 4180
},
{
"epoch": 22.16931216931217,
"grad_norm": 0.08117684721946716,
"learning_rate": 0.00036946478672292483,
"loss": 0.003,
"step": 4190
},
{
"epoch": 22.22222222222222,
"grad_norm": 0.11887771636247635,
"learning_rate": 0.00036927810953296615,
"loss": 0.0038,
"step": 4200
},
{
"epoch": 22.275132275132275,
"grad_norm": 0.23875179886817932,
"learning_rate": 0.000369090910914979,
"loss": 0.0056,
"step": 4210
},
{
"epoch": 22.32804232804233,
"grad_norm": 0.10867864638566971,
"learning_rate": 0.0003689031914455921,
"loss": 0.0053,
"step": 4220
},
{
"epoch": 22.38095238095238,
"grad_norm": 0.11437585949897766,
"learning_rate": 0.0003687149517030384,
"loss": 0.0043,
"step": 4230
},
{
"epoch": 22.433862433862434,
"grad_norm": 0.10560745000839233,
"learning_rate": 0.00036852619226715347,
"loss": 0.0041,
"step": 4240
},
{
"epoch": 22.486772486772487,
"grad_norm": 0.09894239902496338,
"learning_rate": 0.0003683369137193738,
"loss": 0.004,
"step": 4250
},
{
"epoch": 22.53968253968254,
"grad_norm": 0.10818574577569962,
"learning_rate": 0.00036814711664273484,
"loss": 0.0038,
"step": 4260
},
{
"epoch": 22.59259259259259,
"grad_norm": 0.0938066691160202,
"learning_rate": 0.0003679568016218693,
"loss": 0.0037,
"step": 4270
},
{
"epoch": 22.645502645502646,
"grad_norm": 0.08017753809690475,
"learning_rate": 0.00036776596924300523,
"loss": 0.0037,
"step": 4280
},
{
"epoch": 22.6984126984127,
"grad_norm": 0.09411531686782837,
"learning_rate": 0.0003675746200939644,
"loss": 0.0035,
"step": 4290
},
{
"epoch": 22.75132275132275,
"grad_norm": 0.07573402673006058,
"learning_rate": 0.0003673827547641604,
"loss": 0.0032,
"step": 4300
},
{
"epoch": 22.804232804232804,
"grad_norm": 0.07840324193239212,
"learning_rate": 0.00036719037384459665,
"loss": 0.0032,
"step": 4310
},
{
"epoch": 22.857142857142858,
"grad_norm": 0.08790623396635056,
"learning_rate": 0.00036699747792786496,
"loss": 0.0036,
"step": 4320
},
{
"epoch": 22.91005291005291,
"grad_norm": 0.17838507890701294,
"learning_rate": 0.0003668040676081433,
"loss": 0.0039,
"step": 4330
},
{
"epoch": 22.962962962962962,
"grad_norm": 0.4083213806152344,
"learning_rate": 0.0003666101434811943,
"loss": 0.0103,
"step": 4340
},
{
"epoch": 23.015873015873016,
"grad_norm": 0.42006683349609375,
"learning_rate": 0.0003664157061443632,
"loss": 0.0286,
"step": 4350
},
{
"epoch": 23.06878306878307,
"grad_norm": 0.44540899991989136,
"learning_rate": 0.00036622075619657606,
"loss": 0.0331,
"step": 4360
},
{
"epoch": 23.12169312169312,
"grad_norm": 0.5338546633720398,
"learning_rate": 0.00036602529423833803,
"loss": 0.0403,
"step": 4370
},
{
"epoch": 23.174603174603174,
"grad_norm": 0.4317573606967926,
"learning_rate": 0.00036582932087173123,
"loss": 0.0399,
"step": 4380
},
{
"epoch": 23.227513227513228,
"grad_norm": 0.36144527792930603,
"learning_rate": 0.0003656328367004134,
"loss": 0.0281,
"step": 4390
},
{
"epoch": 23.280423280423282,
"grad_norm": 0.2512885332107544,
"learning_rate": 0.0003654358423296153,
"loss": 0.0192,
"step": 4400
},
{
"epoch": 23.333333333333332,
"grad_norm": 0.176627978682518,
"learning_rate": 0.0003652383383661396,
"loss": 0.0135,
"step": 4410
},
{
"epoch": 23.386243386243386,
"grad_norm": 0.13672363758087158,
"learning_rate": 0.0003650403254183585,
"loss": 0.0101,
"step": 4420
},
{
"epoch": 23.43915343915344,
"grad_norm": 0.14289478957653046,
"learning_rate": 0.00036484180409621206,
"loss": 0.0078,
"step": 4430
},
{
"epoch": 23.49206349206349,
"grad_norm": 0.10004527121782303,
"learning_rate": 0.00036464277501120624,
"loss": 0.0061,
"step": 4440
},
{
"epoch": 23.544973544973544,
"grad_norm": 0.10518872737884521,
"learning_rate": 0.0003644432387764113,
"loss": 0.0057,
"step": 4450
},
{
"epoch": 23.597883597883598,
"grad_norm": 0.09654070436954498,
"learning_rate": 0.0003642431960064592,
"loss": 0.0049,
"step": 4460
},
{
"epoch": 23.650793650793652,
"grad_norm": 0.10660598427057266,
"learning_rate": 0.00036404264731754274,
"loss": 0.0049,
"step": 4470
},
{
"epoch": 23.703703703703702,
"grad_norm": 0.07862317562103271,
"learning_rate": 0.0003638415933274127,
"loss": 0.0045,
"step": 4480
},
{
"epoch": 23.756613756613756,
"grad_norm": 0.07714495062828064,
"learning_rate": 0.0003636400346553765,
"loss": 0.0042,
"step": 4490
},
{
"epoch": 23.80952380952381,
"grad_norm": 0.08995921164751053,
"learning_rate": 0.0003634379719222961,
"loss": 0.0042,
"step": 4500
},
{
"epoch": 23.862433862433864,
"grad_norm": 0.07015353441238403,
"learning_rate": 0.0003632354057505862,
"loss": 0.004,
"step": 4510
},
{
"epoch": 23.915343915343914,
"grad_norm": 0.07248852401971817,
"learning_rate": 0.00036303233676421206,
"loss": 0.0041,
"step": 4520
},
{
"epoch": 23.96825396825397,
"grad_norm": 0.07491960376501083,
"learning_rate": 0.000362828765588688,
"loss": 0.0037,
"step": 4530
},
{
"epoch": 24.021164021164022,
"grad_norm": 0.1126113086938858,
"learning_rate": 0.00036262469285107505,
"loss": 0.0043,
"step": 4540
},
{
"epoch": 24.074074074074073,
"grad_norm": 0.0801805779337883,
"learning_rate": 0.0003624201191799793,
"loss": 0.0036,
"step": 4550
},
{
"epoch": 24.126984126984127,
"grad_norm": 0.07318644970655441,
"learning_rate": 0.0003622150452055498,
"loss": 0.0034,
"step": 4560
},
{
"epoch": 24.17989417989418,
"grad_norm": 0.07393806427717209,
"learning_rate": 0.00036200947155947674,
"loss": 0.0035,
"step": 4570
},
{
"epoch": 24.232804232804234,
"grad_norm": 0.09606146067380905,
"learning_rate": 0.0003618033988749895,
"loss": 0.0036,
"step": 4580
},
{
"epoch": 24.285714285714285,
"grad_norm": 0.069748654961586,
"learning_rate": 0.00036159682778685447,
"loss": 0.0034,
"step": 4590
},
{
"epoch": 24.33862433862434,
"grad_norm": 0.11135067790746689,
"learning_rate": 0.00036138975893137347,
"loss": 0.0037,
"step": 4600
},
{
"epoch": 24.391534391534393,
"grad_norm": 0.07549436390399933,
"learning_rate": 0.00036118219294638146,
"loss": 0.0035,
"step": 4610
},
{
"epoch": 24.444444444444443,
"grad_norm": 0.07316253334283829,
"learning_rate": 0.0003609741304712448,
"loss": 0.0034,
"step": 4620
},
{
"epoch": 24.497354497354497,
"grad_norm": 0.0705738440155983,
"learning_rate": 0.000360765572146859,
"loss": 0.0037,
"step": 4630
},
{
"epoch": 24.55026455026455,
"grad_norm": 0.08228278160095215,
"learning_rate": 0.0003605565186156474,
"loss": 0.0034,
"step": 4640
},
{
"epoch": 24.603174603174605,
"grad_norm": 0.07518155127763748,
"learning_rate": 0.0003603469705215582,
"loss": 0.0032,
"step": 4650
},
{
"epoch": 24.656084656084655,
"grad_norm": 0.07805877178907394,
"learning_rate": 0.00036013692851006316,
"loss": 0.0037,
"step": 4660
},
{
"epoch": 24.70899470899471,
"grad_norm": 0.09085424989461899,
"learning_rate": 0.0003599263932281557,
"loss": 0.0043,
"step": 4670
},
{
"epoch": 24.761904761904763,
"grad_norm": 0.07668601721525192,
"learning_rate": 0.0003597153653243484,
"loss": 0.0045,
"step": 4680
},
{
"epoch": 24.814814814814813,
"grad_norm": 0.07025279849767685,
"learning_rate": 0.0003595038454486713,
"loss": 0.0042,
"step": 4690
},
{
"epoch": 24.867724867724867,
"grad_norm": 0.11927028745412827,
"learning_rate": 0.00035929183425266996,
"loss": 0.0075,
"step": 4700
},
{
"epoch": 24.92063492063492,
"grad_norm": 0.13428369164466858,
"learning_rate": 0.0003590793323894033,
"loss": 0.0104,
"step": 4710
},
{
"epoch": 24.973544973544975,
"grad_norm": 0.2348325103521347,
"learning_rate": 0.00035886634051344166,
"loss": 0.0096,
"step": 4720
},
{
"epoch": 25.026455026455025,
"grad_norm": 0.1860775500535965,
"learning_rate": 0.0003586528592808647,
"loss": 0.01,
"step": 4730
},
{
"epoch": 25.07936507936508,
"grad_norm": 0.12654832005500793,
"learning_rate": 0.00035843888934925964,
"loss": 0.0079,
"step": 4740
},
{
"epoch": 25.132275132275133,
"grad_norm": 0.13935686647891998,
"learning_rate": 0.0003582244313777187,
"loss": 0.0067,
"step": 4750
},
{
"epoch": 25.185185185185187,
"grad_norm": 0.1062978133559227,
"learning_rate": 0.0003580094860268377,
"loss": 0.0061,
"step": 4760
},
{
"epoch": 25.238095238095237,
"grad_norm": 0.13322557508945465,
"learning_rate": 0.0003577940539587137,
"loss": 0.0053,
"step": 4770
},
{
"epoch": 25.29100529100529,
"grad_norm": 0.16509148478507996,
"learning_rate": 0.00035757813583694293,
"loss": 0.0062,
"step": 4780
},
{
"epoch": 25.343915343915345,
"grad_norm": 0.07526130974292755,
"learning_rate": 0.0003573617323266188,
"loss": 0.0054,
"step": 4790
},
{
"epoch": 25.396825396825395,
"grad_norm": 0.09176628291606903,
"learning_rate": 0.0003571448440943299,
"loss": 0.0056,
"step": 4800
},
{
"epoch": 25.44973544973545,
"grad_norm": 0.09155473113059998,
"learning_rate": 0.00035692747180815797,
"loss": 0.005,
"step": 4810
},
{
"epoch": 25.502645502645503,
"grad_norm": 0.0963868498802185,
"learning_rate": 0.0003567096161376757,
"loss": 0.0047,
"step": 4820
},
{
"epoch": 25.555555555555557,
"grad_norm": 0.08302408456802368,
"learning_rate": 0.0003564912777539447,
"loss": 0.004,
"step": 4830
},
{
"epoch": 25.608465608465607,
"grad_norm": 0.06412070989608765,
"learning_rate": 0.0003562724573295136,
"loss": 0.0036,
"step": 4840
},
{
"epoch": 25.66137566137566,
"grad_norm": 0.07806012034416199,
"learning_rate": 0.00035605315553841575,
"loss": 0.0035,
"step": 4850
},
{
"epoch": 25.714285714285715,
"grad_norm": 0.08711991459131241,
"learning_rate": 0.0003558333730561675,
"loss": 0.0033,
"step": 4860
},
{
"epoch": 25.767195767195766,
"grad_norm": 0.0889301747083664,
"learning_rate": 0.00035561311055976556,
"loss": 0.0037,
"step": 4870
},
{
"epoch": 25.82010582010582,
"grad_norm": 0.11517832428216934,
"learning_rate": 0.0003553923687276854,
"loss": 0.0038,
"step": 4880
},
{
"epoch": 25.873015873015873,
"grad_norm": 0.07500556111335754,
"learning_rate": 0.00035517114823987886,
"loss": 0.0038,
"step": 4890
},
{
"epoch": 25.925925925925927,
"grad_norm": 0.06777866929769516,
"learning_rate": 0.00035494944977777245,
"loss": 0.0038,
"step": 4900
},
{
"epoch": 25.978835978835978,
"grad_norm": 0.0699387863278389,
"learning_rate": 0.00035472727402426475,
"loss": 0.0036,
"step": 4910
},
{
"epoch": 26.03174603174603,
"grad_norm": 0.06708484888076782,
"learning_rate": 0.0003545046216637245,
"loss": 0.0034,
"step": 4920
},
{
"epoch": 26.084656084656086,
"grad_norm": 0.08221206814050674,
"learning_rate": 0.0003542814933819888,
"loss": 0.0036,
"step": 4930
},
{
"epoch": 26.137566137566136,
"grad_norm": 0.08084139227867126,
"learning_rate": 0.00035405788986636057,
"loss": 0.0033,
"step": 4940
},
{
"epoch": 26.19047619047619,
"grad_norm": 0.05680591240525246,
"learning_rate": 0.00035383381180560653,
"loss": 0.0029,
"step": 4950
},
{
"epoch": 26.243386243386244,
"grad_norm": 0.06472446769475937,
"learning_rate": 0.00035360925988995516,
"loss": 0.0036,
"step": 4960
},
{
"epoch": 26.296296296296298,
"grad_norm": 0.07075701653957367,
"learning_rate": 0.0003533842348110947,
"loss": 0.003,
"step": 4970
},
{
"epoch": 26.349206349206348,
"grad_norm": 0.06456038355827332,
"learning_rate": 0.0003531587372621708,
"loss": 0.0027,
"step": 4980
},
{
"epoch": 26.402116402116402,
"grad_norm": 0.06482533365488052,
"learning_rate": 0.0003529327679377844,
"loss": 0.0031,
"step": 4990
},
{
"epoch": 26.455026455026456,
"grad_norm": 0.06165693700313568,
"learning_rate": 0.0003527063275339898,
"loss": 0.003,
"step": 5000
}
],
"logging_steps": 10,
"max_steps": 18900,
"num_input_tokens_seen": 0,
"num_train_epochs": 100,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 2.7759626173505107e+22,
"train_batch_size": 64,
"trial_name": null,
"trial_params": null
}