Jinwu01's picture
Upload folder using huggingface_hub
c3dc34b verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 5000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.002,
"grad_norm": 0.0,
"learning_rate": 2e-08,
"loss": 0.1903,
"step": 10
},
{
"epoch": 0.004,
"grad_norm": 0.0,
"learning_rate": 4e-08,
"loss": 0.2403,
"step": 20
},
{
"epoch": 0.006,
"grad_norm": 0.0,
"learning_rate": 6e-08,
"loss": 0.3415,
"step": 30
},
{
"epoch": 0.008,
"grad_norm": 0.15908482670783997,
"learning_rate": 8e-08,
"loss": 0.2319,
"step": 40
},
{
"epoch": 0.01,
"grad_norm": 0.0,
"learning_rate": 1e-07,
"loss": 0.1534,
"step": 50
},
{
"epoch": 0.012,
"grad_norm": 0.0,
"learning_rate": 1.2e-07,
"loss": 0.2303,
"step": 60
},
{
"epoch": 0.014,
"grad_norm": 0.15090326964855194,
"learning_rate": 1.4e-07,
"loss": 0.3153,
"step": 70
},
{
"epoch": 0.016,
"grad_norm": 0.18264611065387726,
"learning_rate": 1.6e-07,
"loss": 0.2112,
"step": 80
},
{
"epoch": 0.018,
"grad_norm": 0.13211487233638763,
"learning_rate": 1.8e-07,
"loss": 0.204,
"step": 90
},
{
"epoch": 0.02,
"grad_norm": 0.0,
"learning_rate": 2e-07,
"loss": 0.1409,
"step": 100
},
{
"epoch": 0.022,
"grad_norm": 0.3466247022151947,
"learning_rate": 2.1999999999999998e-07,
"loss": 0.2463,
"step": 110
},
{
"epoch": 0.024,
"grad_norm": 0.19141152501106262,
"learning_rate": 2.4e-07,
"loss": 0.2355,
"step": 120
},
{
"epoch": 0.026,
"grad_norm": 0.18133391439914703,
"learning_rate": 2.6e-07,
"loss": 0.2277,
"step": 130
},
{
"epoch": 0.028,
"grad_norm": 0.0,
"learning_rate": 2.8e-07,
"loss": 0.1573,
"step": 140
},
{
"epoch": 0.03,
"grad_norm": 0.0,
"learning_rate": 3e-07,
"loss": 0.2451,
"step": 150
},
{
"epoch": 0.032,
"grad_norm": 0.0,
"learning_rate": 3.2e-07,
"loss": 0.1121,
"step": 160
},
{
"epoch": 0.034,
"grad_norm": 0.15121112763881683,
"learning_rate": 3.4000000000000003e-07,
"loss": 0.2163,
"step": 170
},
{
"epoch": 0.036,
"grad_norm": 0.0,
"learning_rate": 3.6e-07,
"loss": 0.2381,
"step": 180
},
{
"epoch": 0.038,
"grad_norm": 0.0,
"learning_rate": 3.7999999999999996e-07,
"loss": 0.2397,
"step": 190
},
{
"epoch": 0.04,
"grad_norm": 0.0,
"learning_rate": 4e-07,
"loss": 0.2632,
"step": 200
},
{
"epoch": 0.042,
"grad_norm": 0.1544131636619568,
"learning_rate": 4.1999999999999995e-07,
"loss": 0.216,
"step": 210
},
{
"epoch": 0.044,
"grad_norm": 0.0,
"learning_rate": 4.3999999999999997e-07,
"loss": 0.1347,
"step": 220
},
{
"epoch": 0.046,
"grad_norm": 0.0,
"learning_rate": 4.6e-07,
"loss": 0.1863,
"step": 230
},
{
"epoch": 0.048,
"grad_norm": 0.0,
"learning_rate": 4.8e-07,
"loss": 0.0353,
"step": 240
},
{
"epoch": 0.05,
"grad_norm": 0.1916782557964325,
"learning_rate": 5e-07,
"loss": 0.2479,
"step": 250
},
{
"epoch": 0.052,
"grad_norm": 0.0,
"learning_rate": 5.2e-07,
"loss": 0.2976,
"step": 260
},
{
"epoch": 0.054,
"grad_norm": 0.2502536475658417,
"learning_rate": 5.4e-07,
"loss": 0.218,
"step": 270
},
{
"epoch": 0.056,
"grad_norm": 0.1519690901041031,
"learning_rate": 5.6e-07,
"loss": 0.1472,
"step": 280
},
{
"epoch": 0.058,
"grad_norm": 0.0,
"learning_rate": 5.8e-07,
"loss": 0.1957,
"step": 290
},
{
"epoch": 0.06,
"grad_norm": 0.1677587926387787,
"learning_rate": 6e-07,
"loss": 0.2272,
"step": 300
},
{
"epoch": 0.062,
"grad_norm": 0.0,
"learning_rate": 6.2e-07,
"loss": 0.201,
"step": 310
},
{
"epoch": 0.064,
"grad_norm": 0.10945549607276917,
"learning_rate": 6.4e-07,
"loss": 0.188,
"step": 320
},
{
"epoch": 0.066,
"grad_norm": 0.1568213850259781,
"learning_rate": 6.6e-07,
"loss": 0.2443,
"step": 330
},
{
"epoch": 0.068,
"grad_norm": 0.2246701866388321,
"learning_rate": 6.800000000000001e-07,
"loss": 0.3293,
"step": 340
},
{
"epoch": 0.07,
"grad_norm": 0.13563767075538635,
"learning_rate": 7e-07,
"loss": 0.1562,
"step": 350
},
{
"epoch": 0.072,
"grad_norm": 0.16457843780517578,
"learning_rate": 7.2e-07,
"loss": 0.174,
"step": 360
},
{
"epoch": 0.074,
"grad_norm": 0.0,
"learning_rate": 7.4e-07,
"loss": 0.2139,
"step": 370
},
{
"epoch": 0.076,
"grad_norm": 0.0,
"learning_rate": 7.599999999999999e-07,
"loss": 0.1278,
"step": 380
},
{
"epoch": 0.078,
"grad_norm": 0.2623412609100342,
"learning_rate": 7.799999999999999e-07,
"loss": 0.2084,
"step": 390
},
{
"epoch": 0.08,
"grad_norm": 0.0,
"learning_rate": 8e-07,
"loss": 0.1944,
"step": 400
},
{
"epoch": 0.082,
"grad_norm": 0.0,
"learning_rate": 8.199999999999999e-07,
"loss": 0.2289,
"step": 410
},
{
"epoch": 0.084,
"grad_norm": 0.0,
"learning_rate": 8.399999999999999e-07,
"loss": 0.2451,
"step": 420
},
{
"epoch": 0.086,
"grad_norm": 0.3384997844696045,
"learning_rate": 8.599999999999999e-07,
"loss": 0.2605,
"step": 430
},
{
"epoch": 0.088,
"grad_norm": 0.0,
"learning_rate": 8.799999999999999e-07,
"loss": 0.1546,
"step": 440
},
{
"epoch": 0.09,
"grad_norm": 0.19793939590454102,
"learning_rate": 9e-07,
"loss": 0.201,
"step": 450
},
{
"epoch": 0.092,
"grad_norm": 0.0,
"learning_rate": 9.2e-07,
"loss": 0.1955,
"step": 460
},
{
"epoch": 0.094,
"grad_norm": 0.0,
"learning_rate": 9.399999999999999e-07,
"loss": 0.3065,
"step": 470
},
{
"epoch": 0.096,
"grad_norm": 0.2155238538980484,
"learning_rate": 9.6e-07,
"loss": 0.2489,
"step": 480
},
{
"epoch": 0.098,
"grad_norm": 0.24407915771007538,
"learning_rate": 9.8e-07,
"loss": 0.4369,
"step": 490
},
{
"epoch": 0.1,
"grad_norm": 0.23500211536884308,
"learning_rate": 1e-06,
"loss": 0.2276,
"step": 500
},
{
"epoch": 0.102,
"grad_norm": 0.13844609260559082,
"learning_rate": 9.999878153526972e-07,
"loss": 0.1132,
"step": 510
},
{
"epoch": 0.104,
"grad_norm": 0.13867129385471344,
"learning_rate": 9.99951262004652e-07,
"loss": 0.3207,
"step": 520
},
{
"epoch": 0.106,
"grad_norm": 0.16211336851119995,
"learning_rate": 9.998903417374226e-07,
"loss": 0.1522,
"step": 530
},
{
"epoch": 0.108,
"grad_norm": 0.0,
"learning_rate": 9.99805057520177e-07,
"loss": 0.1751,
"step": 540
},
{
"epoch": 0.11,
"grad_norm": 0.0,
"learning_rate": 9.996954135095478e-07,
"loss": 0.1791,
"step": 550
},
{
"epoch": 0.112,
"grad_norm": 0.15618827939033508,
"learning_rate": 9.99561415049429e-07,
"loss": 0.2804,
"step": 560
},
{
"epoch": 0.114,
"grad_norm": 0.17624777555465698,
"learning_rate": 9.99403068670717e-07,
"loss": 0.1606,
"step": 570
},
{
"epoch": 0.116,
"grad_norm": 0.17222286760807037,
"learning_rate": 9.992203820909905e-07,
"loss": 0.2415,
"step": 580
},
{
"epoch": 0.118,
"grad_norm": 0.15965154767036438,
"learning_rate": 9.990133642141357e-07,
"loss": 0.204,
"step": 590
},
{
"epoch": 0.12,
"grad_norm": 0.2776583731174469,
"learning_rate": 9.98782025129912e-07,
"loss": 0.2245,
"step": 600
},
{
"epoch": 0.122,
"grad_norm": 0.0,
"learning_rate": 9.9852637611346e-07,
"loss": 0.2463,
"step": 610
},
{
"epoch": 0.124,
"grad_norm": 0.0,
"learning_rate": 9.982464296247522e-07,
"loss": 0.212,
"step": 620
},
{
"epoch": 0.126,
"grad_norm": 0.0,
"learning_rate": 9.97942199307985e-07,
"loss": 0.2933,
"step": 630
},
{
"epoch": 0.128,
"grad_norm": 0.262826532125473,
"learning_rate": 9.976136999909155e-07,
"loss": 0.2437,
"step": 640
},
{
"epoch": 0.13,
"grad_norm": 0.14430199563503265,
"learning_rate": 9.972609476841365e-07,
"loss": 0.4017,
"step": 650
},
{
"epoch": 0.132,
"grad_norm": 0.0,
"learning_rate": 9.968839595802981e-07,
"loss": 0.2055,
"step": 660
},
{
"epoch": 0.134,
"grad_norm": 0.0,
"learning_rate": 9.964827540532684e-07,
"loss": 0.0685,
"step": 670
},
{
"epoch": 0.136,
"grad_norm": 0.0,
"learning_rate": 9.960573506572389e-07,
"loss": 0.2145,
"step": 680
},
{
"epoch": 0.138,
"grad_norm": 0.18930301070213318,
"learning_rate": 9.956077701257707e-07,
"loss": 0.2836,
"step": 690
},
{
"epoch": 0.14,
"grad_norm": 0.21128208935260773,
"learning_rate": 9.95134034370785e-07,
"loss": 0.3285,
"step": 700
},
{
"epoch": 0.142,
"grad_norm": 0.0,
"learning_rate": 9.946361664814943e-07,
"loss": 0.2677,
"step": 710
},
{
"epoch": 0.144,
"grad_norm": 0.14169941842556,
"learning_rate": 9.941141907232763e-07,
"loss": 0.1567,
"step": 720
},
{
"epoch": 0.146,
"grad_norm": 0.0,
"learning_rate": 9.93568132536494e-07,
"loss": 0.149,
"step": 730
},
{
"epoch": 0.148,
"grad_norm": 0.0,
"learning_rate": 9.929980185352525e-07,
"loss": 0.3243,
"step": 740
},
{
"epoch": 0.15,
"grad_norm": 0.4330027103424072,
"learning_rate": 9.92403876506104e-07,
"loss": 0.3521,
"step": 750
},
{
"epoch": 0.152,
"grad_norm": 0.0,
"learning_rate": 9.91785735406693e-07,
"loss": 0.2648,
"step": 760
},
{
"epoch": 0.154,
"grad_norm": 0.2584955096244812,
"learning_rate": 9.911436253643443e-07,
"loss": 0.2882,
"step": 770
},
{
"epoch": 0.156,
"grad_norm": 0.2793239653110504,
"learning_rate": 9.904775776745956e-07,
"loss": 0.1974,
"step": 780
},
{
"epoch": 0.158,
"grad_norm": 0.7728231549263,
"learning_rate": 9.89787624799672e-07,
"loss": 0.3992,
"step": 790
},
{
"epoch": 0.16,
"grad_norm": 0.0,
"learning_rate": 9.890738003669027e-07,
"loss": 0.1949,
"step": 800
},
{
"epoch": 0.162,
"grad_norm": 0.0,
"learning_rate": 9.883361391670839e-07,
"loss": 0.1667,
"step": 810
},
{
"epoch": 0.164,
"grad_norm": 0.0,
"learning_rate": 9.875746771527815e-07,
"loss": 0.151,
"step": 820
},
{
"epoch": 0.166,
"grad_norm": 0.0,
"learning_rate": 9.8678945143658e-07,
"loss": 0.2299,
"step": 830
},
{
"epoch": 0.168,
"grad_norm": 0.3418106436729431,
"learning_rate": 9.859805002892731e-07,
"loss": 0.2426,
"step": 840
},
{
"epoch": 0.17,
"grad_norm": 0.23645935952663422,
"learning_rate": 9.851478631379982e-07,
"loss": 0.3274,
"step": 850
},
{
"epoch": 0.172,
"grad_norm": 0.0,
"learning_rate": 9.842915805643156e-07,
"loss": 0.1971,
"step": 860
},
{
"epoch": 0.174,
"grad_norm": 0.0,
"learning_rate": 9.834116943022297e-07,
"loss": 0.1315,
"step": 870
},
{
"epoch": 0.176,
"grad_norm": 0.0,
"learning_rate": 9.825082472361556e-07,
"loss": 0.2402,
"step": 880
},
{
"epoch": 0.178,
"grad_norm": 0.17213283479213715,
"learning_rate": 9.81581283398829e-07,
"loss": 0.198,
"step": 890
},
{
"epoch": 0.18,
"grad_norm": 0.0,
"learning_rate": 9.806308479691594e-07,
"loss": 0.2376,
"step": 900
},
{
"epoch": 0.182,
"grad_norm": 0.0,
"learning_rate": 9.796569872700287e-07,
"loss": 0.2244,
"step": 910
},
{
"epoch": 0.184,
"grad_norm": 0.0,
"learning_rate": 9.786597487660335e-07,
"loss": 0.1149,
"step": 920
},
{
"epoch": 0.186,
"grad_norm": 0.2780471742153168,
"learning_rate": 9.776391810611718e-07,
"loss": 0.2305,
"step": 930
},
{
"epoch": 0.188,
"grad_norm": 0.1760900914669037,
"learning_rate": 9.765953338964734e-07,
"loss": 0.1016,
"step": 940
},
{
"epoch": 0.19,
"grad_norm": 0.0,
"learning_rate": 9.755282581475767e-07,
"loss": 0.1416,
"step": 950
},
{
"epoch": 0.192,
"grad_norm": 0.23833173513412476,
"learning_rate": 9.744380058222482e-07,
"loss": 0.3388,
"step": 960
},
{
"epoch": 0.194,
"grad_norm": 0.23290500044822693,
"learning_rate": 9.733246300578482e-07,
"loss": 0.2386,
"step": 970
},
{
"epoch": 0.196,
"grad_norm": 0.0,
"learning_rate": 9.721881851187405e-07,
"loss": 0.1735,
"step": 980
},
{
"epoch": 0.198,
"grad_norm": 0.31412091851234436,
"learning_rate": 9.710287263936483e-07,
"loss": 0.2637,
"step": 990
},
{
"epoch": 0.2,
"grad_norm": 0.27369225025177,
"learning_rate": 9.698463103929541e-07,
"loss": 0.2161,
"step": 1000
},
{
"epoch": 0.202,
"grad_norm": 0.0,
"learning_rate": 9.686409947459457e-07,
"loss": 0.2611,
"step": 1010
},
{
"epoch": 0.204,
"grad_norm": 0.18912163376808167,
"learning_rate": 9.674128381980071e-07,
"loss": 0.1744,
"step": 1020
},
{
"epoch": 0.206,
"grad_norm": 0.2728767395019531,
"learning_rate": 9.661619006077561e-07,
"loss": 0.2382,
"step": 1030
},
{
"epoch": 0.208,
"grad_norm": 0.17648717761039734,
"learning_rate": 9.648882429441256e-07,
"loss": 0.2746,
"step": 1040
},
{
"epoch": 0.21,
"grad_norm": 0.0,
"learning_rate": 9.635919272833937e-07,
"loss": 0.1473,
"step": 1050
},
{
"epoch": 0.212,
"grad_norm": 0.1783466935157776,
"learning_rate": 9.622730168061567e-07,
"loss": 0.2144,
"step": 1060
},
{
"epoch": 0.214,
"grad_norm": 0.18328428268432617,
"learning_rate": 9.609315757942502e-07,
"loss": 0.2516,
"step": 1070
},
{
"epoch": 0.216,
"grad_norm": 0.43628615140914917,
"learning_rate": 9.595676696276171e-07,
"loss": 0.3733,
"step": 1080
},
{
"epoch": 0.218,
"grad_norm": 0.0,
"learning_rate": 9.581813647811197e-07,
"loss": 0.1464,
"step": 1090
},
{
"epoch": 0.22,
"grad_norm": 0.3243227005004883,
"learning_rate": 9.567727288213004e-07,
"loss": 0.249,
"step": 1100
},
{
"epoch": 0.222,
"grad_norm": 0.2554561495780945,
"learning_rate": 9.553418304030885e-07,
"loss": 0.1656,
"step": 1110
},
{
"epoch": 0.224,
"grad_norm": 0.0,
"learning_rate": 9.538887392664543e-07,
"loss": 0.0984,
"step": 1120
},
{
"epoch": 0.226,
"grad_norm": 0.0,
"learning_rate": 9.524135262330098e-07,
"loss": 0.1974,
"step": 1130
},
{
"epoch": 0.228,
"grad_norm": 0.0,
"learning_rate": 9.509162632025569e-07,
"loss": 0.2129,
"step": 1140
},
{
"epoch": 0.23,
"grad_norm": 0.23683281242847443,
"learning_rate": 9.493970231495834e-07,
"loss": 0.3028,
"step": 1150
},
{
"epoch": 0.232,
"grad_norm": 0.18687139451503754,
"learning_rate": 9.478558801197064e-07,
"loss": 0.1714,
"step": 1160
},
{
"epoch": 0.234,
"grad_norm": 0.2143907994031906,
"learning_rate": 9.462929092260628e-07,
"loss": 0.1583,
"step": 1170
},
{
"epoch": 0.236,
"grad_norm": 0.0,
"learning_rate": 9.447081866456487e-07,
"loss": 0.237,
"step": 1180
},
{
"epoch": 0.238,
"grad_norm": 0.0,
"learning_rate": 9.431017896156073e-07,
"loss": 0.1604,
"step": 1190
},
{
"epoch": 0.24,
"grad_norm": 0.4592614769935608,
"learning_rate": 9.414737964294634e-07,
"loss": 0.2498,
"step": 1200
},
{
"epoch": 0.242,
"grad_norm": 0.17820818722248077,
"learning_rate": 9.398242864333083e-07,
"loss": 0.1914,
"step": 1210
},
{
"epoch": 0.244,
"grad_norm": 0.2759917080402374,
"learning_rate": 9.381533400219317e-07,
"loss": 0.2974,
"step": 1220
},
{
"epoch": 0.246,
"grad_norm": 0.25706782937049866,
"learning_rate": 9.364610386349047e-07,
"loss": 0.1905,
"step": 1230
},
{
"epoch": 0.248,
"grad_norm": 0.36785247921943665,
"learning_rate": 9.347474647526095e-07,
"loss": 0.3763,
"step": 1240
},
{
"epoch": 0.25,
"grad_norm": 0.22583401203155518,
"learning_rate": 9.330127018922193e-07,
"loss": 0.2157,
"step": 1250
},
{
"epoch": 0.252,
"grad_norm": 0.5489715337753296,
"learning_rate": 9.312568346036287e-07,
"loss": 0.2944,
"step": 1260
},
{
"epoch": 0.254,
"grad_norm": 0.0,
"learning_rate": 9.294799484653322e-07,
"loss": 0.3299,
"step": 1270
},
{
"epoch": 0.256,
"grad_norm": 0.28369149565696716,
"learning_rate": 9.276821300802533e-07,
"loss": 0.2198,
"step": 1280
},
{
"epoch": 0.258,
"grad_norm": 0.17388980090618134,
"learning_rate": 9.258634670715237e-07,
"loss": 0.1441,
"step": 1290
},
{
"epoch": 0.26,
"grad_norm": 0.0,
"learning_rate": 9.240240480782129e-07,
"loss": 0.2248,
"step": 1300
},
{
"epoch": 0.262,
"grad_norm": 0.0,
"learning_rate": 9.221639627510075e-07,
"loss": 0.1801,
"step": 1310
},
{
"epoch": 0.264,
"grad_norm": 0.0,
"learning_rate": 9.202833017478421e-07,
"loss": 0.1365,
"step": 1320
},
{
"epoch": 0.266,
"grad_norm": 0.0,
"learning_rate": 9.183821567294808e-07,
"loss": 0.1392,
"step": 1330
},
{
"epoch": 0.268,
"grad_norm": 0.0,
"learning_rate": 9.164606203550497e-07,
"loss": 0.2375,
"step": 1340
},
{
"epoch": 0.27,
"grad_norm": 0.0,
"learning_rate": 9.145187862775208e-07,
"loss": 0.2166,
"step": 1350
},
{
"epoch": 0.272,
"grad_norm": 0.6230157613754272,
"learning_rate": 9.125567491391475e-07,
"loss": 0.2524,
"step": 1360
},
{
"epoch": 0.274,
"grad_norm": 0.1825215071439743,
"learning_rate": 9.10574604566852e-07,
"loss": 0.116,
"step": 1370
},
{
"epoch": 0.276,
"grad_norm": 0.6285405158996582,
"learning_rate": 9.085724491675642e-07,
"loss": 0.1629,
"step": 1380
},
{
"epoch": 0.278,
"grad_norm": 0.19837217032909393,
"learning_rate": 9.065503805235137e-07,
"loss": 0.2715,
"step": 1390
},
{
"epoch": 0.28,
"grad_norm": 0.0,
"learning_rate": 9.045084971874737e-07,
"loss": 0.1993,
"step": 1400
},
{
"epoch": 0.282,
"grad_norm": 0.5574557185173035,
"learning_rate": 9.02446898677957e-07,
"loss": 0.227,
"step": 1410
},
{
"epoch": 0.284,
"grad_norm": 0.21934154629707336,
"learning_rate": 9.003656854743666e-07,
"loss": 0.2065,
"step": 1420
},
{
"epoch": 0.286,
"grad_norm": 0.19448724389076233,
"learning_rate": 8.982649590120981e-07,
"loss": 0.2645,
"step": 1430
},
{
"epoch": 0.288,
"grad_norm": 0.39887791872024536,
"learning_rate": 8.961448216775953e-07,
"loss": 0.2705,
"step": 1440
},
{
"epoch": 0.29,
"grad_norm": 0.32678404450416565,
"learning_rate": 8.940053768033608e-07,
"loss": 0.2481,
"step": 1450
},
{
"epoch": 0.292,
"grad_norm": 0.2931132912635803,
"learning_rate": 8.918467286629198e-07,
"loss": 0.1207,
"step": 1460
},
{
"epoch": 0.294,
"grad_norm": 0.0,
"learning_rate": 8.896689824657371e-07,
"loss": 0.1948,
"step": 1470
},
{
"epoch": 0.296,
"grad_norm": 0.0,
"learning_rate": 8.874722443520898e-07,
"loss": 0.1255,
"step": 1480
},
{
"epoch": 0.298,
"grad_norm": 0.0,
"learning_rate": 8.852566213878946e-07,
"loss": 0.0875,
"step": 1490
},
{
"epoch": 0.3,
"grad_norm": 0.0,
"learning_rate": 8.83022221559489e-07,
"loss": 0.4261,
"step": 1500
},
{
"epoch": 0.302,
"grad_norm": 0.0,
"learning_rate": 8.807691537683684e-07,
"loss": 0.2974,
"step": 1510
},
{
"epoch": 0.304,
"grad_norm": 0.0,
"learning_rate": 8.784975278258782e-07,
"loss": 0.2012,
"step": 1520
},
{
"epoch": 0.306,
"grad_norm": 0.37348565459251404,
"learning_rate": 8.762074544478621e-07,
"loss": 0.2722,
"step": 1530
},
{
"epoch": 0.308,
"grad_norm": 0.24141186475753784,
"learning_rate": 8.73899045249266e-07,
"loss": 0.1886,
"step": 1540
},
{
"epoch": 0.31,
"grad_norm": 0.33376821875572205,
"learning_rate": 8.71572412738697e-07,
"loss": 0.2372,
"step": 1550
},
{
"epoch": 0.312,
"grad_norm": 0.32474586367607117,
"learning_rate": 8.69227670312942e-07,
"loss": 0.274,
"step": 1560
},
{
"epoch": 0.314,
"grad_norm": 0.0,
"learning_rate": 8.668649322514381e-07,
"loss": 0.0607,
"step": 1570
},
{
"epoch": 0.316,
"grad_norm": 0.3872072696685791,
"learning_rate": 8.644843137107057e-07,
"loss": 0.2427,
"step": 1580
},
{
"epoch": 0.318,
"grad_norm": 0.2110668420791626,
"learning_rate": 8.620859307187338e-07,
"loss": 0.1347,
"step": 1590
},
{
"epoch": 0.32,
"grad_norm": 0.22994905710220337,
"learning_rate": 8.596699001693255e-07,
"loss": 0.1957,
"step": 1600
},
{
"epoch": 0.322,
"grad_norm": 0.22339865565299988,
"learning_rate": 8.572363398164016e-07,
"loss": 0.2206,
"step": 1610
},
{
"epoch": 0.324,
"grad_norm": 0.0,
"learning_rate": 8.547853682682604e-07,
"loss": 0.2685,
"step": 1620
},
{
"epoch": 0.326,
"grad_norm": 0.23870423436164856,
"learning_rate": 8.523171049817973e-07,
"loss": 0.2696,
"step": 1630
},
{
"epoch": 0.328,
"grad_norm": 0.2892666459083557,
"learning_rate": 8.498316702566826e-07,
"loss": 0.2392,
"step": 1640
},
{
"epoch": 0.33,
"grad_norm": 0.3370724618434906,
"learning_rate": 8.473291852294986e-07,
"loss": 0.2293,
"step": 1650
},
{
"epoch": 0.332,
"grad_norm": 0.26216867566108704,
"learning_rate": 8.448097718678348e-07,
"loss": 0.2758,
"step": 1660
},
{
"epoch": 0.334,
"grad_norm": 0.30991676449775696,
"learning_rate": 8.422735529643443e-07,
"loss": 0.1503,
"step": 1670
},
{
"epoch": 0.336,
"grad_norm": 0.0,
"learning_rate": 8.397206521307583e-07,
"loss": 0.0285,
"step": 1680
},
{
"epoch": 0.338,
"grad_norm": 0.0,
"learning_rate": 8.371511937918617e-07,
"loss": 0.135,
"step": 1690
},
{
"epoch": 0.34,
"grad_norm": 0.0,
"learning_rate": 8.34565303179429e-07,
"loss": 0.1756,
"step": 1700
},
{
"epoch": 0.342,
"grad_norm": 0.0,
"learning_rate": 8.319631063261207e-07,
"loss": 0.12,
"step": 1710
},
{
"epoch": 0.344,
"grad_norm": 0.0,
"learning_rate": 8.293447300593402e-07,
"loss": 0.2617,
"step": 1720
},
{
"epoch": 0.346,
"grad_norm": 0.31137949228286743,
"learning_rate": 8.267103019950528e-07,
"loss": 0.2053,
"step": 1730
},
{
"epoch": 0.348,
"grad_norm": 0.44446438550949097,
"learning_rate": 8.240599505315654e-07,
"loss": 0.2521,
"step": 1740
},
{
"epoch": 0.35,
"grad_norm": 0.0,
"learning_rate": 8.213938048432696e-07,
"loss": 0.1694,
"step": 1750
},
{
"epoch": 0.352,
"grad_norm": 0.3610154986381531,
"learning_rate": 8.187119948743449e-07,
"loss": 0.1555,
"step": 1760
},
{
"epoch": 0.354,
"grad_norm": 0.0,
"learning_rate": 8.160146513324254e-07,
"loss": 0.2196,
"step": 1770
},
{
"epoch": 0.356,
"grad_norm": 0.31506335735321045,
"learning_rate": 8.133019056822302e-07,
"loss": 0.2752,
"step": 1780
},
{
"epoch": 0.358,
"grad_norm": 0.24824751913547516,
"learning_rate": 8.105738901391551e-07,
"loss": 0.212,
"step": 1790
},
{
"epoch": 0.36,
"grad_norm": 0.0,
"learning_rate": 8.07830737662829e-07,
"loss": 0.2336,
"step": 1800
},
{
"epoch": 0.362,
"grad_norm": 0.21576835215091705,
"learning_rate": 8.050725819506339e-07,
"loss": 0.1185,
"step": 1810
},
{
"epoch": 0.364,
"grad_norm": 0.2118854969739914,
"learning_rate": 8.022995574311875e-07,
"loss": 0.2184,
"step": 1820
},
{
"epoch": 0.366,
"grad_norm": 0.3406828045845032,
"learning_rate": 7.995117992577928e-07,
"loss": 0.2807,
"step": 1830
},
{
"epoch": 0.368,
"grad_norm": 0.0,
"learning_rate": 7.967094433018508e-07,
"loss": 0.1165,
"step": 1840
},
{
"epoch": 0.37,
"grad_norm": 0.22280316054821014,
"learning_rate": 7.938926261462365e-07,
"loss": 0.1799,
"step": 1850
},
{
"epoch": 0.372,
"grad_norm": 0.20756544172763824,
"learning_rate": 7.910614850786447e-07,
"loss": 0.2926,
"step": 1860
},
{
"epoch": 0.374,
"grad_norm": 0.22667911648750305,
"learning_rate": 7.882161580848966e-07,
"loss": 0.2575,
"step": 1870
},
{
"epoch": 0.376,
"grad_norm": 0.0,
"learning_rate": 7.853567838422159e-07,
"loss": 0.1874,
"step": 1880
},
{
"epoch": 0.378,
"grad_norm": 0.16809621453285217,
"learning_rate": 7.82483501712469e-07,
"loss": 0.1535,
"step": 1890
},
{
"epoch": 0.38,
"grad_norm": 0.0,
"learning_rate": 7.795964517353733e-07,
"loss": 0.0704,
"step": 1900
},
{
"epoch": 0.382,
"grad_norm": 0.2382143884897232,
"learning_rate": 7.76695774621672e-07,
"loss": 0.2371,
"step": 1910
},
{
"epoch": 0.384,
"grad_norm": 0.21829308569431305,
"learning_rate": 7.737816117462751e-07,
"loss": 0.1903,
"step": 1920
},
{
"epoch": 0.386,
"grad_norm": 0.26749011874198914,
"learning_rate": 7.7085410514137e-07,
"loss": 0.1282,
"step": 1930
},
{
"epoch": 0.388,
"grad_norm": 0.0,
"learning_rate": 7.679133974894982e-07,
"loss": 0.2183,
"step": 1940
},
{
"epoch": 0.39,
"grad_norm": 0.0,
"learning_rate": 7.649596321166024e-07,
"loss": 0.1082,
"step": 1950
},
{
"epoch": 0.392,
"grad_norm": 0.0,
"learning_rate": 7.619929529850396e-07,
"loss": 0.1404,
"step": 1960
},
{
"epoch": 0.394,
"grad_norm": 0.3156719207763672,
"learning_rate": 7.590135046865651e-07,
"loss": 0.1441,
"step": 1970
},
{
"epoch": 0.396,
"grad_norm": 0.0,
"learning_rate": 7.560214324352858e-07,
"loss": 0.1038,
"step": 1980
},
{
"epoch": 0.398,
"grad_norm": 0.19355212152004242,
"learning_rate": 7.530168820605818e-07,
"loss": 0.2091,
"step": 1990
},
{
"epoch": 0.4,
"grad_norm": 0.3190907835960388,
"learning_rate": 7.5e-07,
"loss": 0.1659,
"step": 2000
},
{
"epoch": 0.402,
"grad_norm": 0.0,
"learning_rate": 7.469709332921154e-07,
"loss": 0.1804,
"step": 2010
},
{
"epoch": 0.404,
"grad_norm": 0.3254469633102417,
"learning_rate": 7.439298295693663e-07,
"loss": 0.2499,
"step": 2020
},
{
"epoch": 0.406,
"grad_norm": 0.21677596867084503,
"learning_rate": 7.408768370508576e-07,
"loss": 0.2162,
"step": 2030
},
{
"epoch": 0.408,
"grad_norm": 0.3386067748069763,
"learning_rate": 7.378121045351377e-07,
"loss": 0.1557,
"step": 2040
},
{
"epoch": 0.41,
"grad_norm": 0.0,
"learning_rate": 7.347357813929454e-07,
"loss": 0.2083,
"step": 2050
},
{
"epoch": 0.412,
"grad_norm": 0.0,
"learning_rate": 7.316480175599308e-07,
"loss": 0.1489,
"step": 2060
},
{
"epoch": 0.414,
"grad_norm": 0.0,
"learning_rate": 7.285489635293471e-07,
"loss": 0.1723,
"step": 2070
},
{
"epoch": 0.416,
"grad_norm": 0.23418666422367096,
"learning_rate": 7.254387703447153e-07,
"loss": 0.1862,
"step": 2080
},
{
"epoch": 0.418,
"grad_norm": 0.0,
"learning_rate": 7.223175895924637e-07,
"loss": 0.1223,
"step": 2090
},
{
"epoch": 0.42,
"grad_norm": 0.0,
"learning_rate": 7.191855733945386e-07,
"loss": 0.1192,
"step": 2100
},
{
"epoch": 0.422,
"grad_norm": 0.5469371676445007,
"learning_rate": 7.160428744009912e-07,
"loss": 0.2137,
"step": 2110
},
{
"epoch": 0.424,
"grad_norm": 0.2084398865699768,
"learning_rate": 7.128896457825363e-07,
"loss": 0.2785,
"step": 2120
},
{
"epoch": 0.426,
"grad_norm": 0.0,
"learning_rate": 7.097260412230885e-07,
"loss": 0.137,
"step": 2130
},
{
"epoch": 0.428,
"grad_norm": 0.0,
"learning_rate": 7.065522149122709e-07,
"loss": 0.164,
"step": 2140
},
{
"epoch": 0.43,
"grad_norm": 0.0,
"learning_rate": 7.033683215379002e-07,
"loss": 0.1038,
"step": 2150
},
{
"epoch": 0.432,
"grad_norm": 0.47716024518013,
"learning_rate": 7.001745162784475e-07,
"loss": 0.3109,
"step": 2160
},
{
"epoch": 0.434,
"grad_norm": 0.0,
"learning_rate": 6.969709547954755e-07,
"loss": 0.2861,
"step": 2170
},
{
"epoch": 0.436,
"grad_norm": 0.0,
"learning_rate": 6.937577932260514e-07,
"loss": 0.1104,
"step": 2180
},
{
"epoch": 0.438,
"grad_norm": 0.0,
"learning_rate": 6.905351881751371e-07,
"loss": 0.144,
"step": 2190
},
{
"epoch": 0.44,
"grad_norm": 0.0,
"learning_rate": 6.87303296707956e-07,
"loss": 0.0941,
"step": 2200
},
{
"epoch": 0.442,
"grad_norm": 0.30282729864120483,
"learning_rate": 6.840622763423391e-07,
"loss": 0.2653,
"step": 2210
},
{
"epoch": 0.444,
"grad_norm": 0.0,
"learning_rate": 6.80812285041046e-07,
"loss": 0.2248,
"step": 2220
},
{
"epoch": 0.446,
"grad_norm": 0.0,
"learning_rate": 6.775534812040686e-07,
"loss": 0.1519,
"step": 2230
},
{
"epoch": 0.448,
"grad_norm": 0.0,
"learning_rate": 6.742860236609076e-07,
"loss": 0.201,
"step": 2240
},
{
"epoch": 0.45,
"grad_norm": 0.17793890833854675,
"learning_rate": 6.710100716628344e-07,
"loss": 0.232,
"step": 2250
},
{
"epoch": 0.452,
"grad_norm": 0.2877851724624634,
"learning_rate": 6.677257848751276e-07,
"loss": 0.1519,
"step": 2260
},
{
"epoch": 0.454,
"grad_norm": 0.20634020864963531,
"learning_rate": 6.644333233692916e-07,
"loss": 0.135,
"step": 2270
},
{
"epoch": 0.456,
"grad_norm": 0.0,
"learning_rate": 6.611328476152556e-07,
"loss": 0.1017,
"step": 2280
},
{
"epoch": 0.458,
"grad_norm": 0.1838294267654419,
"learning_rate": 6.578245184735512e-07,
"loss": 0.2447,
"step": 2290
},
{
"epoch": 0.46,
"grad_norm": 0.2564513385295868,
"learning_rate": 6.545084971874736e-07,
"loss": 0.2115,
"step": 2300
},
{
"epoch": 0.462,
"grad_norm": 0.0,
"learning_rate": 6.511849453752223e-07,
"loss": 0.1666,
"step": 2310
},
{
"epoch": 0.464,
"grad_norm": 0.0,
"learning_rate": 6.478540250220233e-07,
"loss": 0.1613,
"step": 2320
},
{
"epoch": 0.466,
"grad_norm": 0.41829341650009155,
"learning_rate": 6.445158984722358e-07,
"loss": 0.1578,
"step": 2330
},
{
"epoch": 0.468,
"grad_norm": 0.20675064623355865,
"learning_rate": 6.411707284214383e-07,
"loss": 0.246,
"step": 2340
},
{
"epoch": 0.47,
"grad_norm": 0.0,
"learning_rate": 6.378186779084995e-07,
"loss": 0.2042,
"step": 2350
},
{
"epoch": 0.472,
"grad_norm": 0.4196428656578064,
"learning_rate": 6.344599103076328e-07,
"loss": 0.229,
"step": 2360
},
{
"epoch": 0.474,
"grad_norm": 0.3443352282047272,
"learning_rate": 6.310945893204324e-07,
"loss": 0.297,
"step": 2370
},
{
"epoch": 0.476,
"grad_norm": 0.0,
"learning_rate": 6.277228789678953e-07,
"loss": 0.1169,
"step": 2380
},
{
"epoch": 0.478,
"grad_norm": 0.17485401034355164,
"learning_rate": 6.243449435824276e-07,
"loss": 0.256,
"step": 2390
},
{
"epoch": 0.48,
"grad_norm": 0.2088983952999115,
"learning_rate": 6.209609477998338e-07,
"loss": 0.143,
"step": 2400
},
{
"epoch": 0.482,
"grad_norm": 0.33278387784957886,
"learning_rate": 6.17571056551295e-07,
"loss": 0.1413,
"step": 2410
},
{
"epoch": 0.484,
"grad_norm": 0.0,
"learning_rate": 6.141754350553279e-07,
"loss": 0.2189,
"step": 2420
},
{
"epoch": 0.486,
"grad_norm": 0.0,
"learning_rate": 6.107742488097338e-07,
"loss": 0.1325,
"step": 2430
},
{
"epoch": 0.488,
"grad_norm": 0.0,
"learning_rate": 6.073676635835316e-07,
"loss": 0.1428,
"step": 2440
},
{
"epoch": 0.49,
"grad_norm": 0.0,
"learning_rate": 6.039558454088795e-07,
"loss": 0.1964,
"step": 2450
},
{
"epoch": 0.492,
"grad_norm": 0.19888480007648468,
"learning_rate": 6.005389605729824e-07,
"loss": 0.1928,
"step": 2460
},
{
"epoch": 0.494,
"grad_norm": 0.0,
"learning_rate": 5.97117175609986e-07,
"loss": 0.1368,
"step": 2470
},
{
"epoch": 0.496,
"grad_norm": 0.0,
"learning_rate": 5.936906572928624e-07,
"loss": 0.2274,
"step": 2480
},
{
"epoch": 0.498,
"grad_norm": 0.18393106758594513,
"learning_rate": 5.9025957262528e-07,
"loss": 0.1892,
"step": 2490
},
{
"epoch": 0.5,
"grad_norm": 0.21718409657478333,
"learning_rate": 5.868240888334652e-07,
"loss": 0.214,
"step": 2500
},
{
"epoch": 0.502,
"grad_norm": 0.21103844046592712,
"learning_rate": 5.833843733580512e-07,
"loss": 0.1929,
"step": 2510
},
{
"epoch": 0.504,
"grad_norm": 0.3171471059322357,
"learning_rate": 5.799405938459174e-07,
"loss": 0.0885,
"step": 2520
},
{
"epoch": 0.506,
"grad_norm": 0.2388087958097458,
"learning_rate": 5.764929181420191e-07,
"loss": 0.2027,
"step": 2530
},
{
"epoch": 0.508,
"grad_norm": 0.0,
"learning_rate": 5.730415142812058e-07,
"loss": 0.2108,
"step": 2540
},
{
"epoch": 0.51,
"grad_norm": 0.4884037971496582,
"learning_rate": 5.695865504800327e-07,
"loss": 0.2387,
"step": 2550
},
{
"epoch": 0.512,
"grad_norm": 0.1944325864315033,
"learning_rate": 5.661281951285612e-07,
"loss": 0.2252,
"step": 2560
},
{
"epoch": 0.514,
"grad_norm": 0.42049652338027954,
"learning_rate": 5.626666167821521e-07,
"loss": 0.2047,
"step": 2570
},
{
"epoch": 0.516,
"grad_norm": 0.0,
"learning_rate": 5.592019841532506e-07,
"loss": 0.1942,
"step": 2580
},
{
"epoch": 0.518,
"grad_norm": 0.24343137443065643,
"learning_rate": 5.557344661031627e-07,
"loss": 0.2018,
"step": 2590
},
{
"epoch": 0.52,
"grad_norm": 0.27307412028312683,
"learning_rate": 5.522642316338268e-07,
"loss": 0.1997,
"step": 2600
},
{
"epoch": 0.522,
"grad_norm": 0.2629737854003906,
"learning_rate": 5.487914498795747e-07,
"loss": 0.1112,
"step": 2610
},
{
"epoch": 0.524,
"grad_norm": 0.22226367890834808,
"learning_rate": 5.453162900988901e-07,
"loss": 0.2175,
"step": 2620
},
{
"epoch": 0.526,
"grad_norm": 0.0,
"learning_rate": 5.418389216661578e-07,
"loss": 0.2222,
"step": 2630
},
{
"epoch": 0.528,
"grad_norm": 0.0,
"learning_rate": 5.383595140634093e-07,
"loss": 0.1489,
"step": 2640
},
{
"epoch": 0.53,
"grad_norm": 0.25698596239089966,
"learning_rate": 5.348782368720625e-07,
"loss": 0.2534,
"step": 2650
},
{
"epoch": 0.532,
"grad_norm": 0.23344826698303223,
"learning_rate": 5.313952597646567e-07,
"loss": 0.2596,
"step": 2660
},
{
"epoch": 0.534,
"grad_norm": 0.309211403131485,
"learning_rate": 5.27910752496582e-07,
"loss": 0.254,
"step": 2670
},
{
"epoch": 0.536,
"grad_norm": 0.0,
"learning_rate": 5.244248848978067e-07,
"loss": 0.1684,
"step": 2680
},
{
"epoch": 0.538,
"grad_norm": 0.29229891300201416,
"learning_rate": 5.209378268645997e-07,
"loss": 0.18,
"step": 2690
},
{
"epoch": 0.54,
"grad_norm": 0.3112972676753998,
"learning_rate": 5.174497483512505e-07,
"loss": 0.2821,
"step": 2700
},
{
"epoch": 0.542,
"grad_norm": 0.0,
"learning_rate": 5.139608193617844e-07,
"loss": 0.1157,
"step": 2710
},
{
"epoch": 0.544,
"grad_norm": 0.0,
"learning_rate": 5.104712099416785e-07,
"loss": 0.0984,
"step": 2720
},
{
"epoch": 0.546,
"grad_norm": 0.17804303765296936,
"learning_rate": 5.069810901695727e-07,
"loss": 0.2217,
"step": 2730
},
{
"epoch": 0.548,
"grad_norm": 0.2273794710636139,
"learning_rate": 5.034906301489807e-07,
"loss": 0.1595,
"step": 2740
},
{
"epoch": 0.55,
"grad_norm": 0.0,
"learning_rate": 5e-07,
"loss": 0.1954,
"step": 2750
},
{
"epoch": 0.552,
"grad_norm": 0.29391637444496155,
"learning_rate": 4.965093698510192e-07,
"loss": 0.2079,
"step": 2760
},
{
"epoch": 0.554,
"grad_norm": 0.18587817251682281,
"learning_rate": 4.930189098304274e-07,
"loss": 0.2526,
"step": 2770
},
{
"epoch": 0.556,
"grad_norm": 0.3287542164325714,
"learning_rate": 4.895287900583216e-07,
"loss": 0.1786,
"step": 2780
},
{
"epoch": 0.558,
"grad_norm": 0.0,
"learning_rate": 4.860391806382156e-07,
"loss": 0.1701,
"step": 2790
},
{
"epoch": 0.56,
"grad_norm": 0.5386554598808289,
"learning_rate": 4.825502516487496e-07,
"loss": 0.1811,
"step": 2800
},
{
"epoch": 0.562,
"grad_norm": 0.0,
"learning_rate": 4.790621731354002e-07,
"loss": 0.2274,
"step": 2810
},
{
"epoch": 0.564,
"grad_norm": 0.0,
"learning_rate": 4.7557511510219335e-07,
"loss": 0.0856,
"step": 2820
},
{
"epoch": 0.566,
"grad_norm": 0.21392494440078735,
"learning_rate": 4.7208924750341805e-07,
"loss": 0.149,
"step": 2830
},
{
"epoch": 0.568,
"grad_norm": 0.304354727268219,
"learning_rate": 4.686047402353433e-07,
"loss": 0.2554,
"step": 2840
},
{
"epoch": 0.57,
"grad_norm": 0.0,
"learning_rate": 4.6512176312793735e-07,
"loss": 0.059,
"step": 2850
},
{
"epoch": 0.572,
"grad_norm": 0.0,
"learning_rate": 4.6164048593659065e-07,
"loss": 0.1526,
"step": 2860
},
{
"epoch": 0.574,
"grad_norm": 0.0,
"learning_rate": 4.5816107833384233e-07,
"loss": 0.2571,
"step": 2870
},
{
"epoch": 0.576,
"grad_norm": 0.0,
"learning_rate": 4.5468370990110997e-07,
"loss": 0.1313,
"step": 2880
},
{
"epoch": 0.578,
"grad_norm": 0.24898090958595276,
"learning_rate": 4.512085501204253e-07,
"loss": 0.2444,
"step": 2890
},
{
"epoch": 0.58,
"grad_norm": 0.2554855942726135,
"learning_rate": 4.477357683661733e-07,
"loss": 0.1362,
"step": 2900
},
{
"epoch": 0.582,
"grad_norm": 0.0,
"learning_rate": 4.442655338968373e-07,
"loss": 0.1642,
"step": 2910
},
{
"epoch": 0.584,
"grad_norm": 0.0,
"learning_rate": 4.407980158467495e-07,
"loss": 0.2219,
"step": 2920
},
{
"epoch": 0.586,
"grad_norm": 0.0,
"learning_rate": 4.3733338321784777e-07,
"loss": 0.1358,
"step": 2930
},
{
"epoch": 0.588,
"grad_norm": 0.0,
"learning_rate": 4.338718048714387e-07,
"loss": 0.2484,
"step": 2940
},
{
"epoch": 0.59,
"grad_norm": 0.5872920751571655,
"learning_rate": 4.304134495199674e-07,
"loss": 0.2279,
"step": 2950
},
{
"epoch": 0.592,
"grad_norm": 0.0,
"learning_rate": 4.2695848571879424e-07,
"loss": 0.155,
"step": 2960
},
{
"epoch": 0.594,
"grad_norm": 0.38492071628570557,
"learning_rate": 4.23507081857981e-07,
"loss": 0.2144,
"step": 2970
},
{
"epoch": 0.596,
"grad_norm": 0.0,
"learning_rate": 4.200594061540826e-07,
"loss": 0.1034,
"step": 2980
},
{
"epoch": 0.598,
"grad_norm": 0.0,
"learning_rate": 4.166156266419489e-07,
"loss": 0.084,
"step": 2990
},
{
"epoch": 0.6,
"grad_norm": 0.22598810493946075,
"learning_rate": 4.131759111665348e-07,
"loss": 0.1944,
"step": 3000
},
{
"epoch": 0.602,
"grad_norm": 0.0,
"learning_rate": 4.0974042737472005e-07,
"loss": 0.1805,
"step": 3010
},
{
"epoch": 0.604,
"grad_norm": 0.30624961853027344,
"learning_rate": 4.0630934270713755e-07,
"loss": 0.121,
"step": 3020
},
{
"epoch": 0.606,
"grad_norm": 0.0,
"learning_rate": 4.028828243900141e-07,
"loss": 0.1631,
"step": 3030
},
{
"epoch": 0.608,
"grad_norm": 0.0,
"learning_rate": 3.9946103942701775e-07,
"loss": 0.1618,
"step": 3040
},
{
"epoch": 0.61,
"grad_norm": 0.35461926460266113,
"learning_rate": 3.960441545911204e-07,
"loss": 0.3025,
"step": 3050
},
{
"epoch": 0.612,
"grad_norm": 0.5769699215888977,
"learning_rate": 3.9263233641646836e-07,
"loss": 0.3172,
"step": 3060
},
{
"epoch": 0.614,
"grad_norm": 0.0,
"learning_rate": 3.8922575119026635e-07,
"loss": 0.1701,
"step": 3070
},
{
"epoch": 0.616,
"grad_norm": 0.0,
"learning_rate": 3.8582456494467206e-07,
"loss": 0.2202,
"step": 3080
},
{
"epoch": 0.618,
"grad_norm": 0.20652209222316742,
"learning_rate": 3.8242894344870495e-07,
"loss": 0.1956,
"step": 3090
},
{
"epoch": 0.62,
"grad_norm": 0.0,
"learning_rate": 3.790390522001662e-07,
"loss": 0.1548,
"step": 3100
},
{
"epoch": 0.622,
"grad_norm": 0.0,
"learning_rate": 3.7565505641757266e-07,
"loss": 0.1647,
"step": 3110
},
{
"epoch": 0.624,
"grad_norm": 0.0,
"learning_rate": 3.722771210321048e-07,
"loss": 0.1113,
"step": 3120
},
{
"epoch": 0.626,
"grad_norm": 0.0,
"learning_rate": 3.689054106795677e-07,
"loss": 0.2053,
"step": 3130
},
{
"epoch": 0.628,
"grad_norm": 0.20418575406074524,
"learning_rate": 3.6554008969236715e-07,
"loss": 0.153,
"step": 3140
},
{
"epoch": 0.63,
"grad_norm": 0.0,
"learning_rate": 3.621813220915004e-07,
"loss": 0.1937,
"step": 3150
},
{
"epoch": 0.632,
"grad_norm": 0.0,
"learning_rate": 3.5882927157856167e-07,
"loss": 0.0739,
"step": 3160
},
{
"epoch": 0.634,
"grad_norm": 0.23595549166202545,
"learning_rate": 3.554841015277641e-07,
"loss": 0.1403,
"step": 3170
},
{
"epoch": 0.636,
"grad_norm": 0.0,
"learning_rate": 3.521459749779768e-07,
"loss": 0.186,
"step": 3180
},
{
"epoch": 0.638,
"grad_norm": 0.0,
"learning_rate": 3.488150546247778e-07,
"loss": 0.2311,
"step": 3190
},
{
"epoch": 0.64,
"grad_norm": 0.2079978585243225,
"learning_rate": 3.454915028125263e-07,
"loss": 0.2173,
"step": 3200
},
{
"epoch": 0.642,
"grad_norm": 0.0,
"learning_rate": 3.421754815264488e-07,
"loss": 0.1359,
"step": 3210
},
{
"epoch": 0.644,
"grad_norm": 0.37681078910827637,
"learning_rate": 3.388671523847445e-07,
"loss": 0.2212,
"step": 3220
},
{
"epoch": 0.646,
"grad_norm": 0.2023986130952835,
"learning_rate": 3.3556667663070835e-07,
"loss": 0.2237,
"step": 3230
},
{
"epoch": 0.648,
"grad_norm": 0.0,
"learning_rate": 3.3227421512487255e-07,
"loss": 0.1364,
"step": 3240
},
{
"epoch": 0.65,
"grad_norm": 0.0,
"learning_rate": 3.2898992833716563e-07,
"loss": 0.1204,
"step": 3250
},
{
"epoch": 0.652,
"grad_norm": 0.0,
"learning_rate": 3.257139763390925e-07,
"loss": 0.1476,
"step": 3260
},
{
"epoch": 0.654,
"grad_norm": 0.2869522273540497,
"learning_rate": 3.2244651879593156e-07,
"loss": 0.2089,
"step": 3270
},
{
"epoch": 0.656,
"grad_norm": 0.0,
"learning_rate": 3.191877149589539e-07,
"loss": 0.1223,
"step": 3280
},
{
"epoch": 0.658,
"grad_norm": 0.6116024851799011,
"learning_rate": 3.15937723657661e-07,
"loss": 0.1637,
"step": 3290
},
{
"epoch": 0.66,
"grad_norm": 0.2069709151983261,
"learning_rate": 3.1269670329204393e-07,
"loss": 0.1503,
"step": 3300
},
{
"epoch": 0.662,
"grad_norm": 0.0,
"learning_rate": 3.0946481182486297e-07,
"loss": 0.1257,
"step": 3310
},
{
"epoch": 0.664,
"grad_norm": 0.18059736490249634,
"learning_rate": 3.0624220677394854e-07,
"loss": 0.214,
"step": 3320
},
{
"epoch": 0.666,
"grad_norm": 0.27141156792640686,
"learning_rate": 3.0302904520452443e-07,
"loss": 0.3025,
"step": 3330
},
{
"epoch": 0.668,
"grad_norm": 0.0,
"learning_rate": 2.9982548372155256e-07,
"loss": 0.2296,
"step": 3340
},
{
"epoch": 0.67,
"grad_norm": 0.0,
"learning_rate": 2.9663167846209996e-07,
"loss": 0.1282,
"step": 3350
},
{
"epoch": 0.672,
"grad_norm": 0.34350427985191345,
"learning_rate": 2.9344778508772914e-07,
"loss": 0.2626,
"step": 3360
},
{
"epoch": 0.674,
"grad_norm": 0.25415468215942383,
"learning_rate": 2.902739587769114e-07,
"loss": 0.1824,
"step": 3370
},
{
"epoch": 0.676,
"grad_norm": 0.39462947845458984,
"learning_rate": 2.8711035421746363e-07,
"loss": 0.0924,
"step": 3380
},
{
"epoch": 0.678,
"grad_norm": 0.0,
"learning_rate": 2.8395712559900874e-07,
"loss": 0.2372,
"step": 3390
},
{
"epoch": 0.68,
"grad_norm": 0.24075475335121155,
"learning_rate": 2.808144266054612e-07,
"loss": 0.3402,
"step": 3400
},
{
"epoch": 0.682,
"grad_norm": 0.240759015083313,
"learning_rate": 2.776824104075364e-07,
"loss": 0.1164,
"step": 3410
},
{
"epoch": 0.684,
"grad_norm": 0.0,
"learning_rate": 2.745612296552847e-07,
"loss": 0.1703,
"step": 3420
},
{
"epoch": 0.686,
"grad_norm": 0.3092927932739258,
"learning_rate": 2.71451036470653e-07,
"loss": 0.1506,
"step": 3430
},
{
"epoch": 0.688,
"grad_norm": 0.0,
"learning_rate": 2.683519824400692e-07,
"loss": 0.0584,
"step": 3440
},
{
"epoch": 0.69,
"grad_norm": 0.0,
"learning_rate": 2.6526421860705473e-07,
"loss": 0.191,
"step": 3450
},
{
"epoch": 0.692,
"grad_norm": 0.0,
"learning_rate": 2.621878954648623e-07,
"loss": 0.105,
"step": 3460
},
{
"epoch": 0.694,
"grad_norm": 0.0,
"learning_rate": 2.591231629491423e-07,
"loss": 0.1391,
"step": 3470
},
{
"epoch": 0.696,
"grad_norm": 0.0,
"learning_rate": 2.5607017043063353e-07,
"loss": 0.0587,
"step": 3480
},
{
"epoch": 0.698,
"grad_norm": 0.27812764048576355,
"learning_rate": 2.530290667078846e-07,
"loss": 0.1216,
"step": 3490
},
{
"epoch": 0.7,
"grad_norm": 0.0,
"learning_rate": 2.500000000000001e-07,
"loss": 0.135,
"step": 3500
},
{
"epoch": 0.702,
"grad_norm": 0.26827505230903625,
"learning_rate": 2.469831179394182e-07,
"loss": 0.1703,
"step": 3510
},
{
"epoch": 0.704,
"grad_norm": 0.0,
"learning_rate": 2.439785675647143e-07,
"loss": 0.1405,
"step": 3520
},
{
"epoch": 0.706,
"grad_norm": 0.0,
"learning_rate": 2.4098649531343494e-07,
"loss": 0.1304,
"step": 3530
},
{
"epoch": 0.708,
"grad_norm": 0.2695796489715576,
"learning_rate": 2.380070470149605e-07,
"loss": 0.1459,
"step": 3540
},
{
"epoch": 0.71,
"grad_norm": 0.3325313627719879,
"learning_rate": 2.350403678833976e-07,
"loss": 0.1686,
"step": 3550
},
{
"epoch": 0.712,
"grad_norm": 0.0,
"learning_rate": 2.3208660251050156e-07,
"loss": 0.0577,
"step": 3560
},
{
"epoch": 0.714,
"grad_norm": 0.2125588357448578,
"learning_rate": 2.2914589485863012e-07,
"loss": 0.1926,
"step": 3570
},
{
"epoch": 0.716,
"grad_norm": 0.20440517365932465,
"learning_rate": 2.262183882537249e-07,
"loss": 0.1804,
"step": 3580
},
{
"epoch": 0.718,
"grad_norm": 0.243080735206604,
"learning_rate": 2.23304225378328e-07,
"loss": 0.1719,
"step": 3590
},
{
"epoch": 0.72,
"grad_norm": 0.0,
"learning_rate": 2.2040354826462664e-07,
"loss": 0.1243,
"step": 3600
},
{
"epoch": 0.722,
"grad_norm": 0.3800593912601471,
"learning_rate": 2.1751649828753106e-07,
"loss": 0.2673,
"step": 3610
},
{
"epoch": 0.724,
"grad_norm": 0.26765140891075134,
"learning_rate": 2.146432161577842e-07,
"loss": 0.2706,
"step": 3620
},
{
"epoch": 0.726,
"grad_norm": 0.0,
"learning_rate": 2.117838419151034e-07,
"loss": 0.1071,
"step": 3630
},
{
"epoch": 0.728,
"grad_norm": 0.0,
"learning_rate": 2.0893851492135532e-07,
"loss": 0.1986,
"step": 3640
},
{
"epoch": 0.73,
"grad_norm": 0.0,
"learning_rate": 2.0610737385376348e-07,
"loss": 0.119,
"step": 3650
},
{
"epoch": 0.732,
"grad_norm": 0.3050713837146759,
"learning_rate": 2.0329055669814933e-07,
"loss": 0.2265,
"step": 3660
},
{
"epoch": 0.734,
"grad_norm": 0.4194519519805908,
"learning_rate": 2.0048820074220711e-07,
"loss": 0.2161,
"step": 3670
},
{
"epoch": 0.736,
"grad_norm": 0.31268754601478577,
"learning_rate": 1.9770044256881258e-07,
"loss": 0.16,
"step": 3680
},
{
"epoch": 0.738,
"grad_norm": 0.0,
"learning_rate": 1.9492741804936618e-07,
"loss": 0.1463,
"step": 3690
},
{
"epoch": 0.74,
"grad_norm": 0.22846342623233795,
"learning_rate": 1.9216926233717084e-07,
"loss": 0.1947,
"step": 3700
},
{
"epoch": 0.742,
"grad_norm": 0.31964659690856934,
"learning_rate": 1.8942610986084484e-07,
"loss": 0.1406,
"step": 3710
},
{
"epoch": 0.744,
"grad_norm": 0.0,
"learning_rate": 1.8669809431776988e-07,
"loss": 0.1202,
"step": 3720
},
{
"epoch": 0.746,
"grad_norm": 0.0,
"learning_rate": 1.8398534866757455e-07,
"loss": 0.2124,
"step": 3730
},
{
"epoch": 0.748,
"grad_norm": 0.0,
"learning_rate": 1.812880051256551e-07,
"loss": 0.1895,
"step": 3740
},
{
"epoch": 0.75,
"grad_norm": 0.24558274447917938,
"learning_rate": 1.7860619515673032e-07,
"loss": 0.2038,
"step": 3750
},
{
"epoch": 0.752,
"grad_norm": 0.2237321436405182,
"learning_rate": 1.7594004946843454e-07,
"loss": 0.196,
"step": 3760
},
{
"epoch": 0.754,
"grad_norm": 0.4649571180343628,
"learning_rate": 1.7328969800494726e-07,
"loss": 0.2282,
"step": 3770
},
{
"epoch": 0.756,
"grad_norm": 0.21489180624485016,
"learning_rate": 1.7065526994065972e-07,
"loss": 0.218,
"step": 3780
},
{
"epoch": 0.758,
"grad_norm": 0.0,
"learning_rate": 1.6803689367387918e-07,
"loss": 0.1441,
"step": 3790
},
{
"epoch": 0.76,
"grad_norm": 0.36996012926101685,
"learning_rate": 1.6543469682057104e-07,
"loss": 0.1192,
"step": 3800
},
{
"epoch": 0.762,
"grad_norm": 0.19936543703079224,
"learning_rate": 1.6284880620813846e-07,
"loss": 0.312,
"step": 3810
},
{
"epoch": 0.764,
"grad_norm": 0.28389972448349,
"learning_rate": 1.6027934786924185e-07,
"loss": 0.1183,
"step": 3820
},
{
"epoch": 0.766,
"grad_norm": 0.20603181421756744,
"learning_rate": 1.5772644703565564e-07,
"loss": 0.1443,
"step": 3830
},
{
"epoch": 0.768,
"grad_norm": 0.0,
"learning_rate": 1.551902281321651e-07,
"loss": 0.1898,
"step": 3840
},
{
"epoch": 0.77,
"grad_norm": 0.2706297039985657,
"learning_rate": 1.5267081477050131e-07,
"loss": 0.2005,
"step": 3850
},
{
"epoch": 0.772,
"grad_norm": 0.0,
"learning_rate": 1.5016832974331723e-07,
"loss": 0.233,
"step": 3860
},
{
"epoch": 0.774,
"grad_norm": 0.0,
"learning_rate": 1.4768289501820263e-07,
"loss": 0.116,
"step": 3870
},
{
"epoch": 0.776,
"grad_norm": 0.26223212480545044,
"learning_rate": 1.4521463173173965e-07,
"loss": 0.1664,
"step": 3880
},
{
"epoch": 0.778,
"grad_norm": 0.21312423050403595,
"learning_rate": 1.4276366018359842e-07,
"loss": 0.2515,
"step": 3890
},
{
"epoch": 0.78,
"grad_norm": 0.0,
"learning_rate": 1.4033009983067452e-07,
"loss": 0.1169,
"step": 3900
},
{
"epoch": 0.782,
"grad_norm": 0.0,
"learning_rate": 1.3791406928126635e-07,
"loss": 0.0598,
"step": 3910
},
{
"epoch": 0.784,
"grad_norm": 0.2988565266132355,
"learning_rate": 1.3551568628929432e-07,
"loss": 0.313,
"step": 3920
},
{
"epoch": 0.786,
"grad_norm": 0.2568652927875519,
"learning_rate": 1.3313506774856175e-07,
"loss": 0.1434,
"step": 3930
},
{
"epoch": 0.788,
"grad_norm": 0.36155828833580017,
"learning_rate": 1.3077232968705805e-07,
"loss": 0.1699,
"step": 3940
},
{
"epoch": 0.79,
"grad_norm": 0.0,
"learning_rate": 1.284275872613028e-07,
"loss": 0.0944,
"step": 3950
},
{
"epoch": 0.792,
"grad_norm": 0.33918094635009766,
"learning_rate": 1.2610095475073413e-07,
"loss": 0.1457,
"step": 3960
},
{
"epoch": 0.794,
"grad_norm": 0.2919306457042694,
"learning_rate": 1.2379254555213786e-07,
"loss": 0.1773,
"step": 3970
},
{
"epoch": 0.796,
"grad_norm": 0.5223463177680969,
"learning_rate": 1.2150247217412185e-07,
"loss": 0.1956,
"step": 3980
},
{
"epoch": 0.798,
"grad_norm": 0.26477882266044617,
"learning_rate": 1.192308462316317e-07,
"loss": 0.2178,
"step": 3990
},
{
"epoch": 0.8,
"grad_norm": 0.22203296422958374,
"learning_rate": 1.1697777844051104e-07,
"loss": 0.1097,
"step": 4000
},
{
"epoch": 0.802,
"grad_norm": 0.0,
"learning_rate": 1.1474337861210543e-07,
"loss": 0.1238,
"step": 4010
},
{
"epoch": 0.804,
"grad_norm": 0.26538366079330444,
"learning_rate": 1.1252775564791023e-07,
"loss": 0.1531,
"step": 4020
},
{
"epoch": 0.806,
"grad_norm": 0.2682723104953766,
"learning_rate": 1.1033101753426282e-07,
"loss": 0.1954,
"step": 4030
},
{
"epoch": 0.808,
"grad_norm": 0.22357577085494995,
"learning_rate": 1.0815327133708013e-07,
"loss": 0.1523,
"step": 4040
},
{
"epoch": 0.81,
"grad_norm": 0.0,
"learning_rate": 1.0599462319663904e-07,
"loss": 0.2394,
"step": 4050
},
{
"epoch": 0.812,
"grad_norm": 0.0,
"learning_rate": 1.038551783224047e-07,
"loss": 0.177,
"step": 4060
},
{
"epoch": 0.814,
"grad_norm": 0.425686776638031,
"learning_rate": 1.0173504098790186e-07,
"loss": 0.2096,
"step": 4070
},
{
"epoch": 0.816,
"grad_norm": 0.2931308448314667,
"learning_rate": 9.963431452563331e-08,
"loss": 0.1841,
"step": 4080
},
{
"epoch": 0.818,
"grad_norm": 0.5810412764549255,
"learning_rate": 9.755310132204297e-08,
"loss": 0.2342,
"step": 4090
},
{
"epoch": 0.82,
"grad_norm": 0.3016006648540497,
"learning_rate": 9.549150281252632e-08,
"loss": 0.1852,
"step": 4100
},
{
"epoch": 0.822,
"grad_norm": 0.0,
"learning_rate": 9.344961947648622e-08,
"loss": 0.1833,
"step": 4110
},
{
"epoch": 0.824,
"grad_norm": 0.0,
"learning_rate": 9.142755083243575e-08,
"loss": 0.1947,
"step": 4120
},
{
"epoch": 0.826,
"grad_norm": 0.0,
"learning_rate": 8.942539543314798e-08,
"loss": 0.076,
"step": 4130
},
{
"epoch": 0.828,
"grad_norm": 0.3835859000682831,
"learning_rate": 8.744325086085247e-08,
"loss": 0.2069,
"step": 4140
},
{
"epoch": 0.83,
"grad_norm": 0.0,
"learning_rate": 8.548121372247919e-08,
"loss": 0.1643,
"step": 4150
},
{
"epoch": 0.832,
"grad_norm": 0.5051850080490112,
"learning_rate": 8.353937964495028e-08,
"loss": 0.2224,
"step": 4160
},
{
"epoch": 0.834,
"grad_norm": 0.22442878782749176,
"learning_rate": 8.161784327051919e-08,
"loss": 0.1301,
"step": 4170
},
{
"epoch": 0.836,
"grad_norm": 0.0,
"learning_rate": 7.971669825215787e-08,
"loss": 0.1773,
"step": 4180
},
{
"epoch": 0.838,
"grad_norm": 0.0,
"learning_rate": 7.783603724899257e-08,
"loss": 0.1138,
"step": 4190
},
{
"epoch": 0.84,
"grad_norm": 0.2998628318309784,
"learning_rate": 7.597595192178702e-08,
"loss": 0.1488,
"step": 4200
},
{
"epoch": 0.842,
"grad_norm": 0.259321391582489,
"learning_rate": 7.413653292847616e-08,
"loss": 0.1796,
"step": 4210
},
{
"epoch": 0.844,
"grad_norm": 0.0,
"learning_rate": 7.23178699197467e-08,
"loss": 0.1397,
"step": 4220
},
{
"epoch": 0.846,
"grad_norm": 0.3644345700740814,
"learning_rate": 7.052005153466778e-08,
"loss": 0.2052,
"step": 4230
},
{
"epoch": 0.848,
"grad_norm": 0.0,
"learning_rate": 6.874316539637126e-08,
"loss": 0.1092,
"step": 4240
},
{
"epoch": 0.85,
"grad_norm": 0.0,
"learning_rate": 6.698729810778064e-08,
"loss": 0.1719,
"step": 4250
},
{
"epoch": 0.852,
"grad_norm": 0.270352303981781,
"learning_rate": 6.52525352473905e-08,
"loss": 0.1219,
"step": 4260
},
{
"epoch": 0.854,
"grad_norm": 0.26021769642829895,
"learning_rate": 6.353896136509524e-08,
"loss": 0.2,
"step": 4270
},
{
"epoch": 0.856,
"grad_norm": 0.245030015707016,
"learning_rate": 6.184665997806831e-08,
"loss": 0.1799,
"step": 4280
},
{
"epoch": 0.858,
"grad_norm": 0.0,
"learning_rate": 6.017571356669182e-08,
"loss": 0.1433,
"step": 4290
},
{
"epoch": 0.86,
"grad_norm": 0.0,
"learning_rate": 5.8526203570536504e-08,
"loss": 0.1186,
"step": 4300
},
{
"epoch": 0.862,
"grad_norm": 0.0,
"learning_rate": 5.689821038439263e-08,
"loss": 0.0735,
"step": 4310
},
{
"epoch": 0.864,
"grad_norm": 0.0,
"learning_rate": 5.529181335435124e-08,
"loss": 0.1492,
"step": 4320
},
{
"epoch": 0.866,
"grad_norm": 0.24768459796905518,
"learning_rate": 5.37070907739372e-08,
"loss": 0.1799,
"step": 4330
},
{
"epoch": 0.868,
"grad_norm": 0.25161197781562805,
"learning_rate": 5.2144119880293544e-08,
"loss": 0.1065,
"step": 4340
},
{
"epoch": 0.87,
"grad_norm": 0.0,
"learning_rate": 5.060297685041659e-08,
"loss": 0.099,
"step": 4350
},
{
"epoch": 0.872,
"grad_norm": 0.0,
"learning_rate": 4.908373679744315e-08,
"loss": 0.1629,
"step": 4360
},
{
"epoch": 0.874,
"grad_norm": 0.0,
"learning_rate": 4.758647376699032e-08,
"loss": 0.2036,
"step": 4370
},
{
"epoch": 0.876,
"grad_norm": 0.42183464765548706,
"learning_rate": 4.611126073354571e-08,
"loss": 0.2889,
"step": 4380
},
{
"epoch": 0.878,
"grad_norm": 0.27400001883506775,
"learning_rate": 4.465816959691149e-08,
"loss": 0.1851,
"step": 4390
},
{
"epoch": 0.88,
"grad_norm": 0.0,
"learning_rate": 4.322727117869951e-08,
"loss": 0.1456,
"step": 4400
},
{
"epoch": 0.882,
"grad_norm": 0.322198748588562,
"learning_rate": 4.181863521888018e-08,
"loss": 0.1768,
"step": 4410
},
{
"epoch": 0.884,
"grad_norm": 0.48080357909202576,
"learning_rate": 4.043233037238281e-08,
"loss": 0.1277,
"step": 4420
},
{
"epoch": 0.886,
"grad_norm": 0.0,
"learning_rate": 3.9068424205749794e-08,
"loss": 0.1703,
"step": 4430
},
{
"epoch": 0.888,
"grad_norm": 0.45539432764053345,
"learning_rate": 3.7726983193843485e-08,
"loss": 0.2544,
"step": 4440
},
{
"epoch": 0.89,
"grad_norm": 0.6837408542633057,
"learning_rate": 3.6408072716606345e-08,
"loss": 0.2151,
"step": 4450
},
{
"epoch": 0.892,
"grad_norm": 0.41328099370002747,
"learning_rate": 3.5111757055874326e-08,
"loss": 0.176,
"step": 4460
},
{
"epoch": 0.894,
"grad_norm": 0.0,
"learning_rate": 3.3838099392243915e-08,
"loss": 0.2373,
"step": 4470
},
{
"epoch": 0.896,
"grad_norm": 0.0,
"learning_rate": 3.258716180199278e-08,
"loss": 0.164,
"step": 4480
},
{
"epoch": 0.898,
"grad_norm": 0.0,
"learning_rate": 3.135900525405427e-08,
"loss": 0.1046,
"step": 4490
},
{
"epoch": 0.9,
"grad_norm": 0.31756848096847534,
"learning_rate": 3.015368960704584e-08,
"loss": 0.2611,
"step": 4500
},
{
"epoch": 0.902,
"grad_norm": 0.0,
"learning_rate": 2.8971273606351655e-08,
"loss": 0.1775,
"step": 4510
},
{
"epoch": 0.904,
"grad_norm": 0.0,
"learning_rate": 2.78118148812595e-08,
"loss": 0.1356,
"step": 4520
},
{
"epoch": 0.906,
"grad_norm": 0.0,
"learning_rate": 2.667536994215186e-08,
"loss": 0.1012,
"step": 4530
},
{
"epoch": 0.908,
"grad_norm": 0.3601909279823303,
"learning_rate": 2.5561994177751732e-08,
"loss": 0.161,
"step": 4540
},
{
"epoch": 0.91,
"grad_norm": 0.0,
"learning_rate": 2.4471741852423233e-08,
"loss": 0.1071,
"step": 4550
},
{
"epoch": 0.912,
"grad_norm": 0.0,
"learning_rate": 2.3404666103526537e-08,
"loss": 0.1885,
"step": 4560
},
{
"epoch": 0.914,
"grad_norm": 0.2765796482563019,
"learning_rate": 2.2360818938828187e-08,
"loss": 0.1625,
"step": 4570
},
{
"epoch": 0.916,
"grad_norm": 0.0,
"learning_rate": 2.1340251233966377e-08,
"loss": 0.1428,
"step": 4580
},
{
"epoch": 0.918,
"grad_norm": 0.4303652048110962,
"learning_rate": 2.0343012729971243e-08,
"loss": 0.0951,
"step": 4590
},
{
"epoch": 0.92,
"grad_norm": 0.0,
"learning_rate": 1.936915203084055e-08,
"loss": 0.1465,
"step": 4600
},
{
"epoch": 0.922,
"grad_norm": 0.0,
"learning_rate": 1.8418716601170947e-08,
"loss": 0.1271,
"step": 4610
},
{
"epoch": 0.924,
"grad_norm": 0.4692068099975586,
"learning_rate": 1.7491752763844292e-08,
"loss": 0.2774,
"step": 4620
},
{
"epoch": 0.926,
"grad_norm": 0.23148785531520844,
"learning_rate": 1.658830569777031e-08,
"loss": 0.0971,
"step": 4630
},
{
"epoch": 0.928,
"grad_norm": 0.0,
"learning_rate": 1.570841943568446e-08,
"loss": 0.1732,
"step": 4640
},
{
"epoch": 0.93,
"grad_norm": 0.0,
"learning_rate": 1.4852136862001763e-08,
"loss": 0.2104,
"step": 4650
},
{
"epoch": 0.932,
"grad_norm": 0.0,
"learning_rate": 1.4019499710726911e-08,
"loss": 0.1566,
"step": 4660
},
{
"epoch": 0.934,
"grad_norm": 0.0,
"learning_rate": 1.3210548563419855e-08,
"loss": 0.1474,
"step": 4670
},
{
"epoch": 0.936,
"grad_norm": 0.0,
"learning_rate": 1.2425322847218367e-08,
"loss": 0.188,
"step": 4680
},
{
"epoch": 0.938,
"grad_norm": 0.19969218969345093,
"learning_rate": 1.166386083291604e-08,
"loss": 0.267,
"step": 4690
},
{
"epoch": 0.94,
"grad_norm": 0.0,
"learning_rate": 1.0926199633097154e-08,
"loss": 0.1695,
"step": 4700
},
{
"epoch": 0.942,
"grad_norm": 0.0,
"learning_rate": 1.0212375200327972e-08,
"loss": 0.1638,
"step": 4710
},
{
"epoch": 0.944,
"grad_norm": 0.0,
"learning_rate": 9.522422325404233e-09,
"loss": 0.0975,
"step": 4720
},
{
"epoch": 0.946,
"grad_norm": 0.0,
"learning_rate": 8.856374635655695e-09,
"loss": 0.222,
"step": 4730
},
{
"epoch": 0.948,
"grad_norm": 0.0,
"learning_rate": 8.214264593307096e-09,
"loss": 0.231,
"step": 4740
},
{
"epoch": 0.95,
"grad_norm": 0.0,
"learning_rate": 7.59612349389599e-09,
"loss": 0.1663,
"step": 4750
},
{
"epoch": 0.952,
"grad_norm": 0.0,
"learning_rate": 7.0019814647475636e-09,
"loss": 0.0257,
"step": 4760
},
{
"epoch": 0.954,
"grad_norm": 0.3837907016277313,
"learning_rate": 6.431867463506046e-09,
"loss": 0.1889,
"step": 4770
},
{
"epoch": 0.956,
"grad_norm": 0.20743760466575623,
"learning_rate": 5.8858092767236076e-09,
"loss": 0.1868,
"step": 4780
},
{
"epoch": 0.958,
"grad_norm": 0.0,
"learning_rate": 5.3638335185058335e-09,
"loss": 0.1717,
"step": 4790
},
{
"epoch": 0.96,
"grad_norm": 0.37456387281417847,
"learning_rate": 4.865965629214819e-09,
"loss": 0.2297,
"step": 4800
},
{
"epoch": 0.962,
"grad_norm": 0.35872021317481995,
"learning_rate": 4.3922298742291585e-09,
"loss": 0.1361,
"step": 4810
},
{
"epoch": 0.964,
"grad_norm": 0.3316497504711151,
"learning_rate": 3.9426493427611175e-09,
"loss": 0.2255,
"step": 4820
},
{
"epoch": 0.966,
"grad_norm": 0.0,
"learning_rate": 3.5172459467315286e-09,
"loss": 0.0901,
"step": 4830
},
{
"epoch": 0.968,
"grad_norm": 0.3181476593017578,
"learning_rate": 3.116040419701815e-09,
"loss": 0.1494,
"step": 4840
},
{
"epoch": 0.97,
"grad_norm": 0.0,
"learning_rate": 2.739052315863355e-09,
"loss": 0.1942,
"step": 4850
},
{
"epoch": 0.972,
"grad_norm": 0.0,
"learning_rate": 2.3863000090844076e-09,
"loss": 0.2808,
"step": 4860
},
{
"epoch": 0.974,
"grad_norm": 0.2747316360473633,
"learning_rate": 2.057800692014833e-09,
"loss": 0.1568,
"step": 4870
},
{
"epoch": 0.976,
"grad_norm": 0.3130796253681183,
"learning_rate": 1.7535703752478147e-09,
"loss": 0.2903,
"step": 4880
},
{
"epoch": 0.978,
"grad_norm": 0.0,
"learning_rate": 1.4736238865398765e-09,
"loss": 0.1769,
"step": 4890
},
{
"epoch": 0.98,
"grad_norm": 0.0,
"learning_rate": 1.217974870087901e-09,
"loss": 0.2085,
"step": 4900
},
{
"epoch": 0.982,
"grad_norm": 0.0,
"learning_rate": 9.866357858642205e-10,
"loss": 0.2805,
"step": 4910
},
{
"epoch": 0.984,
"grad_norm": 0.0,
"learning_rate": 7.79617909009489e-10,
"loss": 0.1946,
"step": 4920
},
{
"epoch": 0.986,
"grad_norm": 0.0,
"learning_rate": 5.969313292830125e-10,
"loss": 0.1962,
"step": 4930
},
{
"epoch": 0.988,
"grad_norm": 0.0,
"learning_rate": 4.3858495057080836e-10,
"loss": 0.2275,
"step": 4940
},
{
"epoch": 0.99,
"grad_norm": 0.0,
"learning_rate": 3.0458649045211894e-10,
"loss": 0.1399,
"step": 4950
},
{
"epoch": 0.992,
"grad_norm": 0.0,
"learning_rate": 1.9494247982282387e-10,
"loss": 0.0858,
"step": 4960
},
{
"epoch": 0.994,
"grad_norm": 0.24740253388881683,
"learning_rate": 1.0965826257725019e-10,
"loss": 0.1723,
"step": 4970
},
{
"epoch": 0.996,
"grad_norm": 0.2049749344587326,
"learning_rate": 4.873799534788059e-11,
"loss": 0.1224,
"step": 4980
},
{
"epoch": 0.998,
"grad_norm": 0.3774297535419464,
"learning_rate": 1.2184647302626582e-11,
"loss": 0.2228,
"step": 4990
},
{
"epoch": 1.0,
"grad_norm": 0.33722221851348877,
"learning_rate": 0.0,
"loss": 0.2303,
"step": 5000
},
{
"epoch": 1.0,
"step": 5000,
"total_flos": 8414511274868736.0,
"train_loss": 0.19171363378167153,
"train_runtime": 843.4917,
"train_samples_per_second": 5.928,
"train_steps_per_second": 5.928
}
],
"logging_steps": 10,
"max_steps": 5000,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 4000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 8414511274868736.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}