PEFT
Safetensors
deu05232's picture
Upload folder using huggingface_hub
69e33d6 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.9998489654130795,
"eval_steps": 500,
"global_step": 3310,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0030206917384080953,
"grad_norm": 7.860902826143009,
"learning_rate": 4.9999999999999996e-05,
"loss": 5.2938,
"step": 10
},
{
"epoch": 0.006041383476816191,
"grad_norm": 4.511291722305653,
"learning_rate": 6.505149978319905e-05,
"loss": 2.3098,
"step": 20
},
{
"epoch": 0.009062075215224286,
"grad_norm": 2.7069657833195726,
"learning_rate": 7.385606273598311e-05,
"loss": 1.1035,
"step": 30
},
{
"epoch": 0.012082766953632381,
"grad_norm": 1.6517779530383951,
"learning_rate": 8.01029995663981e-05,
"loss": 0.9473,
"step": 40
},
{
"epoch": 0.015103458692040477,
"grad_norm": 1.5824900592761986,
"learning_rate": 8.494850021680092e-05,
"loss": 0.7474,
"step": 50
},
{
"epoch": 0.01812415043044857,
"grad_norm": 1.4927944897096168,
"learning_rate": 8.890756251918216e-05,
"loss": 0.8,
"step": 60
},
{
"epoch": 0.02114484216885667,
"grad_norm": 1.4512369318084755,
"learning_rate": 9.225490200071284e-05,
"loss": 0.7104,
"step": 70
},
{
"epoch": 0.024165533907264763,
"grad_norm": 1.2235263602555901,
"learning_rate": 9.515449934959716e-05,
"loss": 0.672,
"step": 80
},
{
"epoch": 0.02718622564567286,
"grad_norm": 1.0865733201242376,
"learning_rate": 9.771212547196623e-05,
"loss": 0.7005,
"step": 90
},
{
"epoch": 0.030206917384080954,
"grad_norm": 1.0916398917893568,
"learning_rate": 9.999999999999999e-05,
"loss": 0.6422,
"step": 100
},
{
"epoch": 0.03322760912248905,
"grad_norm": 0.9984336118107756,
"learning_rate": 9.97196261682243e-05,
"loss": 0.6322,
"step": 110
},
{
"epoch": 0.03624830086089714,
"grad_norm": 1.1464001917159417,
"learning_rate": 9.940809968847352e-05,
"loss": 0.6456,
"step": 120
},
{
"epoch": 0.03926899259930524,
"grad_norm": 1.0575258767552795,
"learning_rate": 9.909657320872274e-05,
"loss": 0.6428,
"step": 130
},
{
"epoch": 0.04228968433771334,
"grad_norm": 0.906321333672499,
"learning_rate": 9.878504672897197e-05,
"loss": 0.5997,
"step": 140
},
{
"epoch": 0.045310376076121435,
"grad_norm": 0.8757644190599452,
"learning_rate": 9.847352024922119e-05,
"loss": 0.6227,
"step": 150
},
{
"epoch": 0.048331067814529526,
"grad_norm": 1.0214855267308176,
"learning_rate": 9.816199376947041e-05,
"loss": 0.6208,
"step": 160
},
{
"epoch": 0.05135175955293762,
"grad_norm": 0.8601713217999793,
"learning_rate": 9.785046728971962e-05,
"loss": 0.6462,
"step": 170
},
{
"epoch": 0.05437245129134572,
"grad_norm": 1.050085429963326,
"learning_rate": 9.753894080996884e-05,
"loss": 0.5955,
"step": 180
},
{
"epoch": 0.05739314302975381,
"grad_norm": 0.8205900418247731,
"learning_rate": 9.722741433021807e-05,
"loss": 0.5634,
"step": 190
},
{
"epoch": 0.06041383476816191,
"grad_norm": 0.7968684735669824,
"learning_rate": 9.691588785046729e-05,
"loss": 0.5238,
"step": 200
},
{
"epoch": 0.06343452650657,
"grad_norm": 0.7595311038179066,
"learning_rate": 9.660436137071651e-05,
"loss": 0.5756,
"step": 210
},
{
"epoch": 0.0664552182449781,
"grad_norm": 0.783835223389923,
"learning_rate": 9.629283489096573e-05,
"loss": 0.6058,
"step": 220
},
{
"epoch": 0.0694759099833862,
"grad_norm": 0.8819755880813628,
"learning_rate": 9.598130841121496e-05,
"loss": 0.5973,
"step": 230
},
{
"epoch": 0.07249660172179428,
"grad_norm": 0.8208034762186932,
"learning_rate": 9.566978193146418e-05,
"loss": 0.5653,
"step": 240
},
{
"epoch": 0.07551729346020239,
"grad_norm": 0.7732868130876457,
"learning_rate": 9.53582554517134e-05,
"loss": 0.513,
"step": 250
},
{
"epoch": 0.07853798519861048,
"grad_norm": 0.9597274496832807,
"learning_rate": 9.504672897196261e-05,
"loss": 0.6034,
"step": 260
},
{
"epoch": 0.08155867693701857,
"grad_norm": 0.6985522547546542,
"learning_rate": 9.473520249221183e-05,
"loss": 0.6076,
"step": 270
},
{
"epoch": 0.08457936867542667,
"grad_norm": 0.7054831014230869,
"learning_rate": 9.442367601246106e-05,
"loss": 0.553,
"step": 280
},
{
"epoch": 0.08760006041383477,
"grad_norm": 0.5696603228373256,
"learning_rate": 9.411214953271028e-05,
"loss": 0.505,
"step": 290
},
{
"epoch": 0.09062075215224287,
"grad_norm": 0.8376021526765258,
"learning_rate": 9.38006230529595e-05,
"loss": 0.5247,
"step": 300
},
{
"epoch": 0.09364144389065096,
"grad_norm": 0.7886851489104535,
"learning_rate": 9.348909657320873e-05,
"loss": 0.5436,
"step": 310
},
{
"epoch": 0.09666213562905905,
"grad_norm": 0.6106377526973836,
"learning_rate": 9.317757009345795e-05,
"loss": 0.5165,
"step": 320
},
{
"epoch": 0.09968282736746716,
"grad_norm": 0.7131012981207684,
"learning_rate": 9.286604361370717e-05,
"loss": 0.5526,
"step": 330
},
{
"epoch": 0.10270351910587525,
"grad_norm": 0.6233943907609755,
"learning_rate": 9.255451713395639e-05,
"loss": 0.5225,
"step": 340
},
{
"epoch": 0.10572421084428334,
"grad_norm": 0.7249996194183139,
"learning_rate": 9.22429906542056e-05,
"loss": 0.5455,
"step": 350
},
{
"epoch": 0.10874490258269144,
"grad_norm": 0.6265859328133128,
"learning_rate": 9.193146417445482e-05,
"loss": 0.5757,
"step": 360
},
{
"epoch": 0.11176559432109953,
"grad_norm": 0.7289630589742836,
"learning_rate": 9.161993769470405e-05,
"loss": 0.5513,
"step": 370
},
{
"epoch": 0.11478628605950762,
"grad_norm": 0.732914371784668,
"learning_rate": 9.130841121495327e-05,
"loss": 0.59,
"step": 380
},
{
"epoch": 0.11780697779791573,
"grad_norm": 0.8864895281320011,
"learning_rate": 9.099688473520249e-05,
"loss": 0.5993,
"step": 390
},
{
"epoch": 0.12082766953632382,
"grad_norm": 0.8427509184519877,
"learning_rate": 9.068535825545172e-05,
"loss": 0.5375,
"step": 400
},
{
"epoch": 0.12384836127473191,
"grad_norm": 0.7962371354744809,
"learning_rate": 9.037383177570094e-05,
"loss": 0.5892,
"step": 410
},
{
"epoch": 0.12686905301314,
"grad_norm": 0.5844098721184025,
"learning_rate": 9.006230529595016e-05,
"loss": 0.5025,
"step": 420
},
{
"epoch": 0.1298897447515481,
"grad_norm": 0.6894143217833443,
"learning_rate": 8.975077881619937e-05,
"loss": 0.5127,
"step": 430
},
{
"epoch": 0.1329104364899562,
"grad_norm": 0.9651737660018649,
"learning_rate": 8.943925233644859e-05,
"loss": 0.5628,
"step": 440
},
{
"epoch": 0.13593112822836428,
"grad_norm": 0.5968288306984006,
"learning_rate": 8.912772585669782e-05,
"loss": 0.5762,
"step": 450
},
{
"epoch": 0.1389518199667724,
"grad_norm": 0.9277080892180378,
"learning_rate": 8.881619937694704e-05,
"loss": 0.5489,
"step": 460
},
{
"epoch": 0.1419725117051805,
"grad_norm": 0.7221871473312308,
"learning_rate": 8.850467289719626e-05,
"loss": 0.5279,
"step": 470
},
{
"epoch": 0.14499320344358857,
"grad_norm": 0.5469965514589646,
"learning_rate": 8.819314641744548e-05,
"loss": 0.5367,
"step": 480
},
{
"epoch": 0.14801389518199667,
"grad_norm": 0.7449772597908582,
"learning_rate": 8.78816199376947e-05,
"loss": 0.5254,
"step": 490
},
{
"epoch": 0.15103458692040478,
"grad_norm": 0.5698736361874787,
"learning_rate": 8.757009345794393e-05,
"loss": 0.5337,
"step": 500
},
{
"epoch": 0.15405527865881286,
"grad_norm": 0.8201534561520991,
"learning_rate": 8.725856697819315e-05,
"loss": 0.5286,
"step": 510
},
{
"epoch": 0.15707597039722096,
"grad_norm": 0.7093761430875013,
"learning_rate": 8.694704049844236e-05,
"loss": 0.5812,
"step": 520
},
{
"epoch": 0.16009666213562906,
"grad_norm": 0.6683695731458494,
"learning_rate": 8.663551401869158e-05,
"loss": 0.5362,
"step": 530
},
{
"epoch": 0.16311735387403714,
"grad_norm": 0.861928287205589,
"learning_rate": 8.63239875389408e-05,
"loss": 0.5214,
"step": 540
},
{
"epoch": 0.16613804561244525,
"grad_norm": 0.7126631544642709,
"learning_rate": 8.601246105919003e-05,
"loss": 0.565,
"step": 550
},
{
"epoch": 0.16915873735085335,
"grad_norm": 0.709173184156536,
"learning_rate": 8.570093457943925e-05,
"loss": 0.5749,
"step": 560
},
{
"epoch": 0.17217942908926145,
"grad_norm": 0.8311379874918872,
"learning_rate": 8.538940809968847e-05,
"loss": 0.5459,
"step": 570
},
{
"epoch": 0.17520012082766953,
"grad_norm": 0.7913562407787625,
"learning_rate": 8.50778816199377e-05,
"loss": 0.5264,
"step": 580
},
{
"epoch": 0.17822081256607764,
"grad_norm": 0.6011445816625751,
"learning_rate": 8.476635514018692e-05,
"loss": 0.4878,
"step": 590
},
{
"epoch": 0.18124150430448574,
"grad_norm": 1.0082749419806356,
"learning_rate": 8.445482866043614e-05,
"loss": 0.4913,
"step": 600
},
{
"epoch": 0.18426219604289382,
"grad_norm": 0.5497983725939031,
"learning_rate": 8.414330218068535e-05,
"loss": 0.5235,
"step": 610
},
{
"epoch": 0.18728288778130192,
"grad_norm": 0.8986078230720201,
"learning_rate": 8.383177570093457e-05,
"loss": 0.5854,
"step": 620
},
{
"epoch": 0.19030357951971003,
"grad_norm": 0.7284462570376612,
"learning_rate": 8.35202492211838e-05,
"loss": 0.5746,
"step": 630
},
{
"epoch": 0.1933242712581181,
"grad_norm": 0.724150442704867,
"learning_rate": 8.320872274143302e-05,
"loss": 0.485,
"step": 640
},
{
"epoch": 0.1963449629965262,
"grad_norm": 0.6535182731169173,
"learning_rate": 8.289719626168224e-05,
"loss": 0.5631,
"step": 650
},
{
"epoch": 0.1993656547349343,
"grad_norm": 0.63387702383207,
"learning_rate": 8.258566978193147e-05,
"loss": 0.4996,
"step": 660
},
{
"epoch": 0.2023863464733424,
"grad_norm": 0.6811034194798663,
"learning_rate": 8.227414330218069e-05,
"loss": 0.5504,
"step": 670
},
{
"epoch": 0.2054070382117505,
"grad_norm": 0.7799202419110538,
"learning_rate": 8.196261682242991e-05,
"loss": 0.556,
"step": 680
},
{
"epoch": 0.2084277299501586,
"grad_norm": 0.7386112508975338,
"learning_rate": 8.165109034267912e-05,
"loss": 0.4909,
"step": 690
},
{
"epoch": 0.21144842168856667,
"grad_norm": 0.6265425656178647,
"learning_rate": 8.133956386292834e-05,
"loss": 0.5035,
"step": 700
},
{
"epoch": 0.21446911342697478,
"grad_norm": 0.5527127947239024,
"learning_rate": 8.102803738317756e-05,
"loss": 0.4742,
"step": 710
},
{
"epoch": 0.21748980516538288,
"grad_norm": 0.7210817085101476,
"learning_rate": 8.071651090342679e-05,
"loss": 0.5064,
"step": 720
},
{
"epoch": 0.22051049690379096,
"grad_norm": 0.7748477898562701,
"learning_rate": 8.040498442367601e-05,
"loss": 0.5112,
"step": 730
},
{
"epoch": 0.22353118864219906,
"grad_norm": 0.5987724728648355,
"learning_rate": 8.009345794392523e-05,
"loss": 0.5583,
"step": 740
},
{
"epoch": 0.22655188038060717,
"grad_norm": 0.6315284050914015,
"learning_rate": 7.978193146417447e-05,
"loss": 0.4908,
"step": 750
},
{
"epoch": 0.22957257211901524,
"grad_norm": 0.7439147404236649,
"learning_rate": 7.947040498442369e-05,
"loss": 0.5044,
"step": 760
},
{
"epoch": 0.23259326385742335,
"grad_norm": 0.8232950848407529,
"learning_rate": 7.91588785046729e-05,
"loss": 0.5303,
"step": 770
},
{
"epoch": 0.23561395559583145,
"grad_norm": 0.6244019640292158,
"learning_rate": 7.884735202492212e-05,
"loss": 0.5016,
"step": 780
},
{
"epoch": 0.23863464733423953,
"grad_norm": 0.8429729495386874,
"learning_rate": 7.853582554517135e-05,
"loss": 0.5619,
"step": 790
},
{
"epoch": 0.24165533907264763,
"grad_norm": 0.961363118958505,
"learning_rate": 7.822429906542057e-05,
"loss": 0.491,
"step": 800
},
{
"epoch": 0.24467603081105574,
"grad_norm": 0.6386803465961769,
"learning_rate": 7.791277258566979e-05,
"loss": 0.5101,
"step": 810
},
{
"epoch": 0.24769672254946382,
"grad_norm": 0.6282019054793118,
"learning_rate": 7.760124610591901e-05,
"loss": 0.5059,
"step": 820
},
{
"epoch": 0.2507174142878719,
"grad_norm": 0.6033470702936573,
"learning_rate": 7.728971962616824e-05,
"loss": 0.4973,
"step": 830
},
{
"epoch": 0.25373810602628,
"grad_norm": 0.6569464828424888,
"learning_rate": 7.697819314641746e-05,
"loss": 0.5134,
"step": 840
},
{
"epoch": 0.25675879776468813,
"grad_norm": 0.7129931789969072,
"learning_rate": 7.666666666666667e-05,
"loss": 0.5178,
"step": 850
},
{
"epoch": 0.2597794895030962,
"grad_norm": 0.717582336051238,
"learning_rate": 7.635514018691589e-05,
"loss": 0.5421,
"step": 860
},
{
"epoch": 0.2628001812415043,
"grad_norm": 0.5589361089618974,
"learning_rate": 7.604361370716511e-05,
"loss": 0.5866,
"step": 870
},
{
"epoch": 0.2658208729799124,
"grad_norm": 0.7907041871549095,
"learning_rate": 7.573208722741434e-05,
"loss": 0.5116,
"step": 880
},
{
"epoch": 0.2688415647183205,
"grad_norm": 0.6066796049688076,
"learning_rate": 7.542056074766356e-05,
"loss": 0.5484,
"step": 890
},
{
"epoch": 0.27186225645672857,
"grad_norm": 0.8302061575819333,
"learning_rate": 7.510903426791278e-05,
"loss": 0.4718,
"step": 900
},
{
"epoch": 0.2748829481951367,
"grad_norm": 0.664952371989958,
"learning_rate": 7.4797507788162e-05,
"loss": 0.5073,
"step": 910
},
{
"epoch": 0.2779036399335448,
"grad_norm": 0.557055900394115,
"learning_rate": 7.448598130841123e-05,
"loss": 0.4833,
"step": 920
},
{
"epoch": 0.28092433167195285,
"grad_norm": 0.73864614315634,
"learning_rate": 7.417445482866045e-05,
"loss": 0.5188,
"step": 930
},
{
"epoch": 0.283945023410361,
"grad_norm": 0.7098086742823411,
"learning_rate": 7.386292834890966e-05,
"loss": 0.5025,
"step": 940
},
{
"epoch": 0.28696571514876906,
"grad_norm": 0.8126982976288293,
"learning_rate": 7.355140186915888e-05,
"loss": 0.4671,
"step": 950
},
{
"epoch": 0.28998640688717714,
"grad_norm": 0.8594101389961067,
"learning_rate": 7.32398753894081e-05,
"loss": 0.5115,
"step": 960
},
{
"epoch": 0.29300709862558527,
"grad_norm": 0.6318400724828283,
"learning_rate": 7.292834890965733e-05,
"loss": 0.5426,
"step": 970
},
{
"epoch": 0.29602779036399335,
"grad_norm": 0.607343193432092,
"learning_rate": 7.261682242990655e-05,
"loss": 0.5355,
"step": 980
},
{
"epoch": 0.2990484821024014,
"grad_norm": 0.5812424643125412,
"learning_rate": 7.230529595015577e-05,
"loss": 0.4685,
"step": 990
},
{
"epoch": 0.30206917384080956,
"grad_norm": 0.7976096140166644,
"learning_rate": 7.1993769470405e-05,
"loss": 0.5284,
"step": 1000
},
{
"epoch": 0.30508986557921763,
"grad_norm": 0.694432937558643,
"learning_rate": 7.168224299065422e-05,
"loss": 0.5584,
"step": 1010
},
{
"epoch": 0.3081105573176257,
"grad_norm": 0.869210767617519,
"learning_rate": 7.137071651090344e-05,
"loss": 0.4834,
"step": 1020
},
{
"epoch": 0.31113124905603384,
"grad_norm": 0.7733774645954711,
"learning_rate": 7.105919003115265e-05,
"loss": 0.4988,
"step": 1030
},
{
"epoch": 0.3141519407944419,
"grad_norm": 0.668133872609032,
"learning_rate": 7.074766355140187e-05,
"loss": 0.4852,
"step": 1040
},
{
"epoch": 0.31717263253285,
"grad_norm": 0.713889105372433,
"learning_rate": 7.04361370716511e-05,
"loss": 0.5226,
"step": 1050
},
{
"epoch": 0.32019332427125813,
"grad_norm": 0.609924063317524,
"learning_rate": 7.012461059190032e-05,
"loss": 0.5146,
"step": 1060
},
{
"epoch": 0.3232140160096662,
"grad_norm": 0.6156658182193047,
"learning_rate": 6.981308411214954e-05,
"loss": 0.4713,
"step": 1070
},
{
"epoch": 0.3262347077480743,
"grad_norm": 0.6861828495019596,
"learning_rate": 6.950155763239876e-05,
"loss": 0.5544,
"step": 1080
},
{
"epoch": 0.3292553994864824,
"grad_norm": 0.6231776449176035,
"learning_rate": 6.919003115264799e-05,
"loss": 0.5564,
"step": 1090
},
{
"epoch": 0.3322760912248905,
"grad_norm": 0.7844621173061321,
"learning_rate": 6.887850467289721e-05,
"loss": 0.5205,
"step": 1100
},
{
"epoch": 0.3352967829632986,
"grad_norm": 1.160293626946917,
"learning_rate": 6.856697819314643e-05,
"loss": 0.5101,
"step": 1110
},
{
"epoch": 0.3383174747017067,
"grad_norm": 0.6207523671493266,
"learning_rate": 6.825545171339564e-05,
"loss": 0.5085,
"step": 1120
},
{
"epoch": 0.3413381664401148,
"grad_norm": 0.5082261716094172,
"learning_rate": 6.794392523364486e-05,
"loss": 0.5101,
"step": 1130
},
{
"epoch": 0.3443588581785229,
"grad_norm": 0.7455308166848186,
"learning_rate": 6.763239875389409e-05,
"loss": 0.5342,
"step": 1140
},
{
"epoch": 0.347379549916931,
"grad_norm": 0.7622167905417062,
"learning_rate": 6.732087227414331e-05,
"loss": 0.4958,
"step": 1150
},
{
"epoch": 0.35040024165533906,
"grad_norm": 0.5921992970055514,
"learning_rate": 6.700934579439253e-05,
"loss": 0.4644,
"step": 1160
},
{
"epoch": 0.3534209333937472,
"grad_norm": 0.6617160160272711,
"learning_rate": 6.669781931464175e-05,
"loss": 0.4816,
"step": 1170
},
{
"epoch": 0.35644162513215527,
"grad_norm": 0.7047555631507049,
"learning_rate": 6.638629283489098e-05,
"loss": 0.551,
"step": 1180
},
{
"epoch": 0.35946231687056335,
"grad_norm": 0.7211612122230344,
"learning_rate": 6.60747663551402e-05,
"loss": 0.5147,
"step": 1190
},
{
"epoch": 0.3624830086089715,
"grad_norm": 0.5463976820460816,
"learning_rate": 6.576323987538941e-05,
"loss": 0.4945,
"step": 1200
},
{
"epoch": 0.36550370034737956,
"grad_norm": 0.6133595626698687,
"learning_rate": 6.545171339563863e-05,
"loss": 0.5363,
"step": 1210
},
{
"epoch": 0.36852439208578763,
"grad_norm": 0.4760803106984906,
"learning_rate": 6.514018691588785e-05,
"loss": 0.4918,
"step": 1220
},
{
"epoch": 0.37154508382419577,
"grad_norm": 0.6756140458486795,
"learning_rate": 6.482866043613708e-05,
"loss": 0.4963,
"step": 1230
},
{
"epoch": 0.37456577556260384,
"grad_norm": 0.8143804547667383,
"learning_rate": 6.45171339563863e-05,
"loss": 0.5301,
"step": 1240
},
{
"epoch": 0.3775864673010119,
"grad_norm": 0.5966112187371303,
"learning_rate": 6.420560747663552e-05,
"loss": 0.5172,
"step": 1250
},
{
"epoch": 0.38060715903942005,
"grad_norm": 0.6150062623844954,
"learning_rate": 6.389408099688475e-05,
"loss": 0.464,
"step": 1260
},
{
"epoch": 0.3836278507778281,
"grad_norm": 0.7194197777579191,
"learning_rate": 6.358255451713397e-05,
"loss": 0.4937,
"step": 1270
},
{
"epoch": 0.3866485425162362,
"grad_norm": 0.6696235801036751,
"learning_rate": 6.327102803738319e-05,
"loss": 0.533,
"step": 1280
},
{
"epoch": 0.38966923425464434,
"grad_norm": 0.7024944116635744,
"learning_rate": 6.29595015576324e-05,
"loss": 0.4763,
"step": 1290
},
{
"epoch": 0.3926899259930524,
"grad_norm": 0.7572346094078879,
"learning_rate": 6.264797507788162e-05,
"loss": 0.5063,
"step": 1300
},
{
"epoch": 0.3957106177314605,
"grad_norm": 0.7346206609994042,
"learning_rate": 6.233644859813084e-05,
"loss": 0.5276,
"step": 1310
},
{
"epoch": 0.3987313094698686,
"grad_norm": 0.6814367858074142,
"learning_rate": 6.202492211838007e-05,
"loss": 0.4662,
"step": 1320
},
{
"epoch": 0.4017520012082767,
"grad_norm": 0.6902775515516495,
"learning_rate": 6.171339563862929e-05,
"loss": 0.559,
"step": 1330
},
{
"epoch": 0.4047726929466848,
"grad_norm": 0.6120211975726814,
"learning_rate": 6.140186915887851e-05,
"loss": 0.4707,
"step": 1340
},
{
"epoch": 0.4077933846850929,
"grad_norm": 0.6329979353157215,
"learning_rate": 6.109034267912774e-05,
"loss": 0.5148,
"step": 1350
},
{
"epoch": 0.410814076423501,
"grad_norm": 0.6687805103119827,
"learning_rate": 6.077881619937695e-05,
"loss": 0.5416,
"step": 1360
},
{
"epoch": 0.41383476816190906,
"grad_norm": 0.6242235017669654,
"learning_rate": 6.0467289719626174e-05,
"loss": 0.4599,
"step": 1370
},
{
"epoch": 0.4168554599003172,
"grad_norm": 0.9184272721334844,
"learning_rate": 6.01557632398754e-05,
"loss": 0.4841,
"step": 1380
},
{
"epoch": 0.41987615163872527,
"grad_norm": 0.7213791381113127,
"learning_rate": 5.984423676012462e-05,
"loss": 0.5283,
"step": 1390
},
{
"epoch": 0.42289684337713335,
"grad_norm": 0.5862938152436937,
"learning_rate": 5.9532710280373836e-05,
"loss": 0.4839,
"step": 1400
},
{
"epoch": 0.4259175351155415,
"grad_norm": 0.5276429716038666,
"learning_rate": 5.922118380062306e-05,
"loss": 0.4928,
"step": 1410
},
{
"epoch": 0.42893822685394956,
"grad_norm": 0.5790780479504153,
"learning_rate": 5.890965732087228e-05,
"loss": 0.4831,
"step": 1420
},
{
"epoch": 0.43195891859235763,
"grad_norm": 0.5146029281623179,
"learning_rate": 5.8598130841121504e-05,
"loss": 0.4994,
"step": 1430
},
{
"epoch": 0.43497961033076576,
"grad_norm": 0.6505167034934826,
"learning_rate": 5.828660436137072e-05,
"loss": 0.4873,
"step": 1440
},
{
"epoch": 0.43800030206917384,
"grad_norm": 0.8179918379603428,
"learning_rate": 5.797507788161994e-05,
"loss": 0.5129,
"step": 1450
},
{
"epoch": 0.4410209938075819,
"grad_norm": 0.6273361780644717,
"learning_rate": 5.7663551401869165e-05,
"loss": 0.5137,
"step": 1460
},
{
"epoch": 0.44404168554599005,
"grad_norm": 0.7146608331163231,
"learning_rate": 5.735202492211839e-05,
"loss": 0.4884,
"step": 1470
},
{
"epoch": 0.4470623772843981,
"grad_norm": 0.6081112811434682,
"learning_rate": 5.704049844236761e-05,
"loss": 0.5333,
"step": 1480
},
{
"epoch": 0.4500830690228062,
"grad_norm": 0.731045030074276,
"learning_rate": 5.6728971962616826e-05,
"loss": 0.4508,
"step": 1490
},
{
"epoch": 0.45310376076121434,
"grad_norm": 0.6153816965846339,
"learning_rate": 5.641744548286605e-05,
"loss": 0.4827,
"step": 1500
},
{
"epoch": 0.4561244524996224,
"grad_norm": 0.45726750703880226,
"learning_rate": 5.610591900311527e-05,
"loss": 0.4875,
"step": 1510
},
{
"epoch": 0.4591451442380305,
"grad_norm": 0.663659630127711,
"learning_rate": 5.5794392523364494e-05,
"loss": 0.5083,
"step": 1520
},
{
"epoch": 0.4621658359764386,
"grad_norm": 0.8268100391510672,
"learning_rate": 5.548286604361371e-05,
"loss": 0.578,
"step": 1530
},
{
"epoch": 0.4651865277148467,
"grad_norm": 0.7547559996949195,
"learning_rate": 5.517133956386293e-05,
"loss": 0.519,
"step": 1540
},
{
"epoch": 0.4682072194532548,
"grad_norm": 0.41803236981554814,
"learning_rate": 5.4859813084112156e-05,
"loss": 0.4619,
"step": 1550
},
{
"epoch": 0.4712279111916629,
"grad_norm": 0.7150056601602596,
"learning_rate": 5.454828660436138e-05,
"loss": 0.4682,
"step": 1560
},
{
"epoch": 0.474248602930071,
"grad_norm": 0.6992223210785882,
"learning_rate": 5.4236760124610594e-05,
"loss": 0.4903,
"step": 1570
},
{
"epoch": 0.47726929466847906,
"grad_norm": 0.5592167440420206,
"learning_rate": 5.392523364485982e-05,
"loss": 0.4952,
"step": 1580
},
{
"epoch": 0.4802899864068872,
"grad_norm": 0.7785572985568321,
"learning_rate": 5.361370716510904e-05,
"loss": 0.5094,
"step": 1590
},
{
"epoch": 0.48331067814529527,
"grad_norm": 0.762116977455126,
"learning_rate": 5.330218068535826e-05,
"loss": 0.5107,
"step": 1600
},
{
"epoch": 0.48633136988370335,
"grad_norm": 0.7746813762837843,
"learning_rate": 5.2990654205607485e-05,
"loss": 0.5304,
"step": 1610
},
{
"epoch": 0.4893520616221115,
"grad_norm": 0.6024056219450453,
"learning_rate": 5.26791277258567e-05,
"loss": 0.4474,
"step": 1620
},
{
"epoch": 0.49237275336051955,
"grad_norm": 0.8132413189133607,
"learning_rate": 5.2367601246105924e-05,
"loss": 0.4889,
"step": 1630
},
{
"epoch": 0.49539344509892763,
"grad_norm": 0.5960780268330229,
"learning_rate": 5.2056074766355146e-05,
"loss": 0.4501,
"step": 1640
},
{
"epoch": 0.49841413683733576,
"grad_norm": 0.542967395272246,
"learning_rate": 5.174454828660437e-05,
"loss": 0.4714,
"step": 1650
},
{
"epoch": 0.5014348285757438,
"grad_norm": 0.8234624235000848,
"learning_rate": 5.1433021806853585e-05,
"loss": 0.5459,
"step": 1660
},
{
"epoch": 0.504455520314152,
"grad_norm": 0.652614480564709,
"learning_rate": 5.112149532710281e-05,
"loss": 0.4976,
"step": 1670
},
{
"epoch": 0.50747621205256,
"grad_norm": 0.5614317647789608,
"learning_rate": 5.080996884735203e-05,
"loss": 0.46,
"step": 1680
},
{
"epoch": 0.5104969037909681,
"grad_norm": 0.7627505052748401,
"learning_rate": 5.049844236760125e-05,
"loss": 0.4787,
"step": 1690
},
{
"epoch": 0.5135175955293763,
"grad_norm": 0.6318545707006004,
"learning_rate": 5.018691588785047e-05,
"loss": 0.4864,
"step": 1700
},
{
"epoch": 0.5165382872677843,
"grad_norm": 0.6898665688037914,
"learning_rate": 4.987538940809969e-05,
"loss": 0.5021,
"step": 1710
},
{
"epoch": 0.5195589790061924,
"grad_norm": 0.6182530903547616,
"learning_rate": 4.9563862928348914e-05,
"loss": 0.5367,
"step": 1720
},
{
"epoch": 0.5225796707446005,
"grad_norm": 0.6069638822221811,
"learning_rate": 4.925233644859814e-05,
"loss": 0.4947,
"step": 1730
},
{
"epoch": 0.5256003624830086,
"grad_norm": 0.5975176087573361,
"learning_rate": 4.894080996884736e-05,
"loss": 0.5071,
"step": 1740
},
{
"epoch": 0.5286210542214167,
"grad_norm": 0.5793526703493967,
"learning_rate": 4.8629283489096575e-05,
"loss": 0.5061,
"step": 1750
},
{
"epoch": 0.5316417459598248,
"grad_norm": 0.6140452049938298,
"learning_rate": 4.83177570093458e-05,
"loss": 0.5172,
"step": 1760
},
{
"epoch": 0.5346624376982329,
"grad_norm": 0.7474521417691947,
"learning_rate": 4.800623052959502e-05,
"loss": 0.532,
"step": 1770
},
{
"epoch": 0.537683129436641,
"grad_norm": 0.6281548556304867,
"learning_rate": 4.7694704049844244e-05,
"loss": 0.4434,
"step": 1780
},
{
"epoch": 0.5407038211750491,
"grad_norm": 0.8824533759891267,
"learning_rate": 4.738317757009346e-05,
"loss": 0.4346,
"step": 1790
},
{
"epoch": 0.5437245129134571,
"grad_norm": 0.6092344633414732,
"learning_rate": 4.707165109034268e-05,
"loss": 0.4707,
"step": 1800
},
{
"epoch": 0.5467452046518653,
"grad_norm": 0.6315726597136827,
"learning_rate": 4.6760124610591905e-05,
"loss": 0.4814,
"step": 1810
},
{
"epoch": 0.5497658963902734,
"grad_norm": 0.6211320346993994,
"learning_rate": 4.644859813084113e-05,
"loss": 0.4981,
"step": 1820
},
{
"epoch": 0.5527865881286814,
"grad_norm": 0.6595155780900703,
"learning_rate": 4.6137071651090343e-05,
"loss": 0.5091,
"step": 1830
},
{
"epoch": 0.5558072798670896,
"grad_norm": 0.6804930869394946,
"learning_rate": 4.5825545171339566e-05,
"loss": 0.4524,
"step": 1840
},
{
"epoch": 0.5588279716054977,
"grad_norm": 0.7386089771166297,
"learning_rate": 4.551401869158879e-05,
"loss": 0.424,
"step": 1850
},
{
"epoch": 0.5618486633439057,
"grad_norm": 0.6169986160154738,
"learning_rate": 4.520249221183801e-05,
"loss": 0.4565,
"step": 1860
},
{
"epoch": 0.5648693550823138,
"grad_norm": 0.6788139316166171,
"learning_rate": 4.4890965732087234e-05,
"loss": 0.4896,
"step": 1870
},
{
"epoch": 0.567890046820722,
"grad_norm": 0.6307731794916878,
"learning_rate": 4.457943925233645e-05,
"loss": 0.4282,
"step": 1880
},
{
"epoch": 0.57091073855913,
"grad_norm": 0.7340656415399682,
"learning_rate": 4.426791277258567e-05,
"loss": 0.5055,
"step": 1890
},
{
"epoch": 0.5739314302975381,
"grad_norm": 0.7589871803898481,
"learning_rate": 4.3956386292834895e-05,
"loss": 0.5192,
"step": 1900
},
{
"epoch": 0.5769521220359463,
"grad_norm": 0.5953092849878492,
"learning_rate": 4.364485981308412e-05,
"loss": 0.5085,
"step": 1910
},
{
"epoch": 0.5799728137743543,
"grad_norm": 0.5649150390088848,
"learning_rate": 4.3333333333333334e-05,
"loss": 0.4754,
"step": 1920
},
{
"epoch": 0.5829935055127624,
"grad_norm": 0.6719520987790143,
"learning_rate": 4.302180685358256e-05,
"loss": 0.4598,
"step": 1930
},
{
"epoch": 0.5860141972511705,
"grad_norm": 0.7413895967328512,
"learning_rate": 4.271028037383178e-05,
"loss": 0.5045,
"step": 1940
},
{
"epoch": 0.5890348889895786,
"grad_norm": 0.7436883322002826,
"learning_rate": 4.2398753894081e-05,
"loss": 0.4521,
"step": 1950
},
{
"epoch": 0.5920555807279867,
"grad_norm": 0.827127778770197,
"learning_rate": 4.2087227414330225e-05,
"loss": 0.4804,
"step": 1960
},
{
"epoch": 0.5950762724663948,
"grad_norm": 0.9203480291774605,
"learning_rate": 4.177570093457944e-05,
"loss": 0.4887,
"step": 1970
},
{
"epoch": 0.5980969642048029,
"grad_norm": 0.6907822772913312,
"learning_rate": 4.1464174454828663e-05,
"loss": 0.5302,
"step": 1980
},
{
"epoch": 0.601117655943211,
"grad_norm": 0.5296351234090546,
"learning_rate": 4.1152647975077886e-05,
"loss": 0.444,
"step": 1990
},
{
"epoch": 0.6041383476816191,
"grad_norm": 0.5641337119040449,
"learning_rate": 4.084112149532711e-05,
"loss": 0.4531,
"step": 2000
},
{
"epoch": 0.6071590394200271,
"grad_norm": 0.676312705325477,
"learning_rate": 4.0529595015576325e-05,
"loss": 0.508,
"step": 2010
},
{
"epoch": 0.6101797311584353,
"grad_norm": 0.6120325079612706,
"learning_rate": 4.021806853582555e-05,
"loss": 0.4748,
"step": 2020
},
{
"epoch": 0.6132004228968434,
"grad_norm": 0.9676013773079569,
"learning_rate": 3.990654205607477e-05,
"loss": 0.4858,
"step": 2030
},
{
"epoch": 0.6162211146352514,
"grad_norm": 0.6009545456697323,
"learning_rate": 3.959501557632399e-05,
"loss": 0.4209,
"step": 2040
},
{
"epoch": 0.6192418063736596,
"grad_norm": 0.6020534832439192,
"learning_rate": 3.928348909657321e-05,
"loss": 0.4868,
"step": 2050
},
{
"epoch": 0.6222624981120677,
"grad_norm": 0.6494496327682583,
"learning_rate": 3.897196261682243e-05,
"loss": 0.462,
"step": 2060
},
{
"epoch": 0.6252831898504757,
"grad_norm": 0.6759831682026574,
"learning_rate": 3.8660436137071654e-05,
"loss": 0.4546,
"step": 2070
},
{
"epoch": 0.6283038815888838,
"grad_norm": 0.6199001698834578,
"learning_rate": 3.834890965732088e-05,
"loss": 0.4463,
"step": 2080
},
{
"epoch": 0.631324573327292,
"grad_norm": 0.7262152575230303,
"learning_rate": 3.80373831775701e-05,
"loss": 0.466,
"step": 2090
},
{
"epoch": 0.6343452650657,
"grad_norm": 0.7076099327551308,
"learning_rate": 3.7725856697819315e-05,
"loss": 0.5031,
"step": 2100
},
{
"epoch": 0.6373659568041081,
"grad_norm": 0.9643779843336525,
"learning_rate": 3.741433021806854e-05,
"loss": 0.5146,
"step": 2110
},
{
"epoch": 0.6403866485425163,
"grad_norm": 0.6243649441821072,
"learning_rate": 3.710280373831776e-05,
"loss": 0.5105,
"step": 2120
},
{
"epoch": 0.6434073402809243,
"grad_norm": 0.5786120381695404,
"learning_rate": 3.6791277258566983e-05,
"loss": 0.4741,
"step": 2130
},
{
"epoch": 0.6464280320193324,
"grad_norm": 0.6492128763607345,
"learning_rate": 3.64797507788162e-05,
"loss": 0.4777,
"step": 2140
},
{
"epoch": 0.6494487237577405,
"grad_norm": 0.7701725108706596,
"learning_rate": 3.616822429906542e-05,
"loss": 0.4802,
"step": 2150
},
{
"epoch": 0.6524694154961486,
"grad_norm": 0.6157266282096414,
"learning_rate": 3.5856697819314645e-05,
"loss": 0.5053,
"step": 2160
},
{
"epoch": 0.6554901072345567,
"grad_norm": 0.6633463330124277,
"learning_rate": 3.554517133956387e-05,
"loss": 0.4847,
"step": 2170
},
{
"epoch": 0.6585107989729648,
"grad_norm": 0.6795119892813151,
"learning_rate": 3.523364485981308e-05,
"loss": 0.5012,
"step": 2180
},
{
"epoch": 0.6615314907113728,
"grad_norm": 0.6692112466043711,
"learning_rate": 3.4922118380062306e-05,
"loss": 0.5036,
"step": 2190
},
{
"epoch": 0.664552182449781,
"grad_norm": 0.6780839446658836,
"learning_rate": 3.461059190031153e-05,
"loss": 0.4893,
"step": 2200
},
{
"epoch": 0.6675728741881891,
"grad_norm": 0.6907983692263772,
"learning_rate": 3.429906542056075e-05,
"loss": 0.4635,
"step": 2210
},
{
"epoch": 0.6705935659265972,
"grad_norm": 0.6015605168384579,
"learning_rate": 3.3987538940809974e-05,
"loss": 0.5006,
"step": 2220
},
{
"epoch": 0.6736142576650053,
"grad_norm": 0.6296326044284368,
"learning_rate": 3.367601246105919e-05,
"loss": 0.5172,
"step": 2230
},
{
"epoch": 0.6766349494034134,
"grad_norm": 0.6643923767390163,
"learning_rate": 3.336448598130841e-05,
"loss": 0.4813,
"step": 2240
},
{
"epoch": 0.6796556411418215,
"grad_norm": 0.5757590558165718,
"learning_rate": 3.3052959501557635e-05,
"loss": 0.4934,
"step": 2250
},
{
"epoch": 0.6826763328802296,
"grad_norm": 0.5855872780472199,
"learning_rate": 3.274143302180686e-05,
"loss": 0.501,
"step": 2260
},
{
"epoch": 0.6856970246186377,
"grad_norm": 0.846989357267261,
"learning_rate": 3.2429906542056074e-05,
"loss": 0.5219,
"step": 2270
},
{
"epoch": 0.6887177163570458,
"grad_norm": 0.6996689031422824,
"learning_rate": 3.21183800623053e-05,
"loss": 0.4626,
"step": 2280
},
{
"epoch": 0.6917384080954538,
"grad_norm": 0.81509620990659,
"learning_rate": 3.180685358255452e-05,
"loss": 0.5085,
"step": 2290
},
{
"epoch": 0.694759099833862,
"grad_norm": 0.7719046228841456,
"learning_rate": 3.149532710280374e-05,
"loss": 0.4892,
"step": 2300
},
{
"epoch": 0.6977797915722701,
"grad_norm": 0.6730694043119466,
"learning_rate": 3.118380062305296e-05,
"loss": 0.4577,
"step": 2310
},
{
"epoch": 0.7008004833106781,
"grad_norm": 0.8856588670570701,
"learning_rate": 3.087227414330218e-05,
"loss": 0.5122,
"step": 2320
},
{
"epoch": 0.7038211750490863,
"grad_norm": 0.6367972726689247,
"learning_rate": 3.05607476635514e-05,
"loss": 0.475,
"step": 2330
},
{
"epoch": 0.7068418667874944,
"grad_norm": 0.5849716773219861,
"learning_rate": 3.0249221183800626e-05,
"loss": 0.4869,
"step": 2340
},
{
"epoch": 0.7098625585259024,
"grad_norm": 0.5667390432429509,
"learning_rate": 2.9937694704049845e-05,
"loss": 0.4821,
"step": 2350
},
{
"epoch": 0.7128832502643105,
"grad_norm": 0.8158007174195662,
"learning_rate": 2.9626168224299068e-05,
"loss": 0.4708,
"step": 2360
},
{
"epoch": 0.7159039420027187,
"grad_norm": 0.6954612915854044,
"learning_rate": 2.9314641744548287e-05,
"loss": 0.4809,
"step": 2370
},
{
"epoch": 0.7189246337411267,
"grad_norm": 0.738176885953241,
"learning_rate": 2.900311526479751e-05,
"loss": 0.475,
"step": 2380
},
{
"epoch": 0.7219453254795348,
"grad_norm": 0.6713964950175404,
"learning_rate": 2.869158878504673e-05,
"loss": 0.482,
"step": 2390
},
{
"epoch": 0.724966017217943,
"grad_norm": 0.5355810916081953,
"learning_rate": 2.8380062305295952e-05,
"loss": 0.4719,
"step": 2400
},
{
"epoch": 0.727986708956351,
"grad_norm": 0.6323075129817857,
"learning_rate": 2.806853582554517e-05,
"loss": 0.4877,
"step": 2410
},
{
"epoch": 0.7310074006947591,
"grad_norm": 0.6663982474306439,
"learning_rate": 2.7757009345794394e-05,
"loss": 0.4769,
"step": 2420
},
{
"epoch": 0.7340280924331672,
"grad_norm": 0.6129890965925909,
"learning_rate": 2.7445482866043613e-05,
"loss": 0.4329,
"step": 2430
},
{
"epoch": 0.7370487841715753,
"grad_norm": 0.7020430789719889,
"learning_rate": 2.7133956386292836e-05,
"loss": 0.5035,
"step": 2440
},
{
"epoch": 0.7400694759099834,
"grad_norm": 0.6548362671497355,
"learning_rate": 2.682242990654206e-05,
"loss": 0.481,
"step": 2450
},
{
"epoch": 0.7430901676483915,
"grad_norm": 0.5896793799165079,
"learning_rate": 2.6510903426791278e-05,
"loss": 0.4862,
"step": 2460
},
{
"epoch": 0.7461108593867996,
"grad_norm": 0.6068195954465485,
"learning_rate": 2.61993769470405e-05,
"loss": 0.4552,
"step": 2470
},
{
"epoch": 0.7491315511252077,
"grad_norm": 0.5230646373745321,
"learning_rate": 2.588785046728972e-05,
"loss": 0.4873,
"step": 2480
},
{
"epoch": 0.7521522428636158,
"grad_norm": 0.680283833221463,
"learning_rate": 2.5576323987538943e-05,
"loss": 0.4085,
"step": 2490
},
{
"epoch": 0.7551729346020238,
"grad_norm": 0.7532041269948455,
"learning_rate": 2.5264797507788162e-05,
"loss": 0.4896,
"step": 2500
},
{
"epoch": 0.758193626340432,
"grad_norm": 0.6689006902036698,
"learning_rate": 2.4953271028037385e-05,
"loss": 0.4913,
"step": 2510
},
{
"epoch": 0.7612143180788401,
"grad_norm": 0.7606354790524109,
"learning_rate": 2.4641744548286604e-05,
"loss": 0.4693,
"step": 2520
},
{
"epoch": 0.7642350098172481,
"grad_norm": 0.7655897760238763,
"learning_rate": 2.4330218068535827e-05,
"loss": 0.5203,
"step": 2530
},
{
"epoch": 0.7672557015556563,
"grad_norm": 0.7272836555361831,
"learning_rate": 2.4018691588785046e-05,
"loss": 0.4882,
"step": 2540
},
{
"epoch": 0.7702763932940644,
"grad_norm": 0.7115409752131671,
"learning_rate": 2.370716510903427e-05,
"loss": 0.4736,
"step": 2550
},
{
"epoch": 0.7732970850324724,
"grad_norm": 0.7165805473587474,
"learning_rate": 2.3395638629283488e-05,
"loss": 0.4672,
"step": 2560
},
{
"epoch": 0.7763177767708805,
"grad_norm": 0.7038069485178577,
"learning_rate": 2.308411214953271e-05,
"loss": 0.4636,
"step": 2570
},
{
"epoch": 0.7793384685092887,
"grad_norm": 0.6655589862457777,
"learning_rate": 2.2772585669781933e-05,
"loss": 0.4635,
"step": 2580
},
{
"epoch": 0.7823591602476967,
"grad_norm": 0.9868766672662372,
"learning_rate": 2.2461059190031153e-05,
"loss": 0.483,
"step": 2590
},
{
"epoch": 0.7853798519861048,
"grad_norm": 0.7128396439929989,
"learning_rate": 2.2149532710280375e-05,
"loss": 0.4509,
"step": 2600
},
{
"epoch": 0.788400543724513,
"grad_norm": 0.6977419571506234,
"learning_rate": 2.1838006230529595e-05,
"loss": 0.494,
"step": 2610
},
{
"epoch": 0.791421235462921,
"grad_norm": 0.5400014125415836,
"learning_rate": 2.1526479750778817e-05,
"loss": 0.4394,
"step": 2620
},
{
"epoch": 0.7944419272013291,
"grad_norm": 0.6763721220518477,
"learning_rate": 2.1214953271028037e-05,
"loss": 0.4484,
"step": 2630
},
{
"epoch": 0.7974626189397372,
"grad_norm": 0.7309543198827074,
"learning_rate": 2.090342679127726e-05,
"loss": 0.4605,
"step": 2640
},
{
"epoch": 0.8004833106781453,
"grad_norm": 0.6930081216590248,
"learning_rate": 2.059190031152648e-05,
"loss": 0.4179,
"step": 2650
},
{
"epoch": 0.8035040024165534,
"grad_norm": 0.8168610777468801,
"learning_rate": 2.02803738317757e-05,
"loss": 0.4575,
"step": 2660
},
{
"epoch": 0.8065246941549615,
"grad_norm": 0.6022566279246642,
"learning_rate": 1.996884735202492e-05,
"loss": 0.4,
"step": 2670
},
{
"epoch": 0.8095453858933696,
"grad_norm": 0.5949859271480968,
"learning_rate": 1.9657320872274143e-05,
"loss": 0.4264,
"step": 2680
},
{
"epoch": 0.8125660776317777,
"grad_norm": 0.5770913852760331,
"learning_rate": 1.9345794392523366e-05,
"loss": 0.4456,
"step": 2690
},
{
"epoch": 0.8155867693701858,
"grad_norm": 0.9360609848011735,
"learning_rate": 1.9034267912772585e-05,
"loss": 0.4633,
"step": 2700
},
{
"epoch": 0.8186074611085938,
"grad_norm": 0.5540802818065433,
"learning_rate": 1.8722741433021808e-05,
"loss": 0.4762,
"step": 2710
},
{
"epoch": 0.821628152847002,
"grad_norm": 0.62030309075617,
"learning_rate": 1.8411214953271027e-05,
"loss": 0.5352,
"step": 2720
},
{
"epoch": 0.8246488445854101,
"grad_norm": 0.5905823042658712,
"learning_rate": 1.809968847352025e-05,
"loss": 0.5097,
"step": 2730
},
{
"epoch": 0.8276695363238181,
"grad_norm": 0.6505033755133793,
"learning_rate": 1.778816199376947e-05,
"loss": 0.4432,
"step": 2740
},
{
"epoch": 0.8306902280622263,
"grad_norm": 0.6032430190153859,
"learning_rate": 1.7476635514018695e-05,
"loss": 0.4666,
"step": 2750
},
{
"epoch": 0.8337109198006344,
"grad_norm": 0.5832675483994434,
"learning_rate": 1.7165109034267915e-05,
"loss": 0.4883,
"step": 2760
},
{
"epoch": 0.8367316115390424,
"grad_norm": 1.0043922280478583,
"learning_rate": 1.6853582554517137e-05,
"loss": 0.4652,
"step": 2770
},
{
"epoch": 0.8397523032774505,
"grad_norm": 0.6737120621664583,
"learning_rate": 1.6542056074766357e-05,
"loss": 0.4704,
"step": 2780
},
{
"epoch": 0.8427729950158587,
"grad_norm": 0.7074156380724425,
"learning_rate": 1.623052959501558e-05,
"loss": 0.478,
"step": 2790
},
{
"epoch": 0.8457936867542667,
"grad_norm": 0.6969288011075542,
"learning_rate": 1.59190031152648e-05,
"loss": 0.4712,
"step": 2800
},
{
"epoch": 0.8488143784926748,
"grad_norm": 0.6539134454898755,
"learning_rate": 1.560747663551402e-05,
"loss": 0.4603,
"step": 2810
},
{
"epoch": 0.851835070231083,
"grad_norm": 0.8574345423647629,
"learning_rate": 1.529595015576324e-05,
"loss": 0.4764,
"step": 2820
},
{
"epoch": 0.854855761969491,
"grad_norm": 0.7681171318897773,
"learning_rate": 1.4984423676012463e-05,
"loss": 0.4653,
"step": 2830
},
{
"epoch": 0.8578764537078991,
"grad_norm": 0.6088873864466463,
"learning_rate": 1.4672897196261684e-05,
"loss": 0.4377,
"step": 2840
},
{
"epoch": 0.8608971454463072,
"grad_norm": 0.601123003489307,
"learning_rate": 1.4361370716510905e-05,
"loss": 0.4311,
"step": 2850
},
{
"epoch": 0.8639178371847153,
"grad_norm": 0.6809468400448008,
"learning_rate": 1.4049844236760126e-05,
"loss": 0.4892,
"step": 2860
},
{
"epoch": 0.8669385289231234,
"grad_norm": 0.5314529225728805,
"learning_rate": 1.3738317757009347e-05,
"loss": 0.4862,
"step": 2870
},
{
"epoch": 0.8699592206615315,
"grad_norm": 0.743623953204313,
"learning_rate": 1.3426791277258568e-05,
"loss": 0.4717,
"step": 2880
},
{
"epoch": 0.8729799123999396,
"grad_norm": 0.6911048797521679,
"learning_rate": 1.311526479750779e-05,
"loss": 0.4297,
"step": 2890
},
{
"epoch": 0.8760006041383477,
"grad_norm": 0.7443596626758798,
"learning_rate": 1.280373831775701e-05,
"loss": 0.489,
"step": 2900
},
{
"epoch": 0.8790212958767558,
"grad_norm": 0.7517966421871881,
"learning_rate": 1.2492211838006231e-05,
"loss": 0.4632,
"step": 2910
},
{
"epoch": 0.8820419876151638,
"grad_norm": 0.5808758369392492,
"learning_rate": 1.2180685358255452e-05,
"loss": 0.4099,
"step": 2920
},
{
"epoch": 0.885062679353572,
"grad_norm": 0.6369115713060849,
"learning_rate": 1.1869158878504673e-05,
"loss": 0.4728,
"step": 2930
},
{
"epoch": 0.8880833710919801,
"grad_norm": 0.9179975692680791,
"learning_rate": 1.1557632398753894e-05,
"loss": 0.4856,
"step": 2940
},
{
"epoch": 0.8911040628303881,
"grad_norm": 0.6773305310854225,
"learning_rate": 1.1246105919003117e-05,
"loss": 0.4557,
"step": 2950
},
{
"epoch": 0.8941247545687963,
"grad_norm": 0.722663791187088,
"learning_rate": 1.0934579439252338e-05,
"loss": 0.4732,
"step": 2960
},
{
"epoch": 0.8971454463072044,
"grad_norm": 0.6532478901527927,
"learning_rate": 1.0623052959501559e-05,
"loss": 0.4777,
"step": 2970
},
{
"epoch": 0.9001661380456124,
"grad_norm": 0.6331812888758941,
"learning_rate": 1.031152647975078e-05,
"loss": 0.5272,
"step": 2980
},
{
"epoch": 0.9031868297840205,
"grad_norm": 0.7749496461228453,
"learning_rate": 1e-05,
"loss": 0.513,
"step": 2990
},
{
"epoch": 0.9062075215224287,
"grad_norm": 0.7653109544518386,
"learning_rate": 9.688473520249222e-06,
"loss": 0.4854,
"step": 3000
},
{
"epoch": 0.9092282132608367,
"grad_norm": 0.6688150761391862,
"learning_rate": 9.376947040498443e-06,
"loss": 0.522,
"step": 3010
},
{
"epoch": 0.9122489049992448,
"grad_norm": 0.6862947854651005,
"learning_rate": 9.065420560747664e-06,
"loss": 0.4747,
"step": 3020
},
{
"epoch": 0.915269596737653,
"grad_norm": 0.5454396007926078,
"learning_rate": 8.753894080996885e-06,
"loss": 0.4653,
"step": 3030
},
{
"epoch": 0.918290288476061,
"grad_norm": 0.6921863003888915,
"learning_rate": 8.442367601246107e-06,
"loss": 0.4502,
"step": 3040
},
{
"epoch": 0.9213109802144691,
"grad_norm": 0.7052109555603467,
"learning_rate": 8.130841121495328e-06,
"loss": 0.4797,
"step": 3050
},
{
"epoch": 0.9243316719528772,
"grad_norm": 0.8178618265398909,
"learning_rate": 7.81931464174455e-06,
"loss": 0.4774,
"step": 3060
},
{
"epoch": 0.9273523636912853,
"grad_norm": 0.7616412002758305,
"learning_rate": 7.5077881619937705e-06,
"loss": 0.4944,
"step": 3070
},
{
"epoch": 0.9303730554296934,
"grad_norm": 0.5855415675806104,
"learning_rate": 7.1962616822429915e-06,
"loss": 0.4981,
"step": 3080
},
{
"epoch": 0.9333937471681015,
"grad_norm": 0.6051491728311711,
"learning_rate": 6.8847352024922125e-06,
"loss": 0.4907,
"step": 3090
},
{
"epoch": 0.9364144389065095,
"grad_norm": 0.623048815626897,
"learning_rate": 6.5732087227414335e-06,
"loss": 0.5104,
"step": 3100
},
{
"epoch": 0.9394351306449177,
"grad_norm": 0.6676510238617815,
"learning_rate": 6.2616822429906544e-06,
"loss": 0.4472,
"step": 3110
},
{
"epoch": 0.9424558223833258,
"grad_norm": 0.7021671692225062,
"learning_rate": 5.9501557632398754e-06,
"loss": 0.4495,
"step": 3120
},
{
"epoch": 0.9454765141217338,
"grad_norm": 0.5409778417569182,
"learning_rate": 5.638629283489097e-06,
"loss": 0.4534,
"step": 3130
},
{
"epoch": 0.948497205860142,
"grad_norm": 0.6533122214052549,
"learning_rate": 5.327102803738318e-06,
"loss": 0.503,
"step": 3140
},
{
"epoch": 0.9515178975985501,
"grad_norm": 0.6670668593007272,
"learning_rate": 5.015576323987539e-06,
"loss": 0.4685,
"step": 3150
},
{
"epoch": 0.9545385893369581,
"grad_norm": 0.6832342091874272,
"learning_rate": 4.70404984423676e-06,
"loss": 0.4956,
"step": 3160
},
{
"epoch": 0.9575592810753663,
"grad_norm": 0.7708609034197621,
"learning_rate": 4.392523364485981e-06,
"loss": 0.4686,
"step": 3170
},
{
"epoch": 0.9605799728137744,
"grad_norm": 0.6684641605237255,
"learning_rate": 4.080996884735202e-06,
"loss": 0.4902,
"step": 3180
},
{
"epoch": 0.9636006645521824,
"grad_norm": 0.7432259731703129,
"learning_rate": 3.7694704049844237e-06,
"loss": 0.4668,
"step": 3190
},
{
"epoch": 0.9666213562905905,
"grad_norm": 0.7582361471424034,
"learning_rate": 3.4579439252336447e-06,
"loss": 0.4613,
"step": 3200
},
{
"epoch": 0.9696420480289987,
"grad_norm": 0.6472728279580365,
"learning_rate": 3.146417445482866e-06,
"loss": 0.4424,
"step": 3210
},
{
"epoch": 0.9726627397674067,
"grad_norm": 0.704669184808671,
"learning_rate": 2.8348909657320875e-06,
"loss": 0.4475,
"step": 3220
},
{
"epoch": 0.9756834315058148,
"grad_norm": 0.716266089867387,
"learning_rate": 2.5233644859813085e-06,
"loss": 0.4623,
"step": 3230
},
{
"epoch": 0.978704123244223,
"grad_norm": 0.5942572536685375,
"learning_rate": 2.21183800623053e-06,
"loss": 0.3935,
"step": 3240
},
{
"epoch": 0.981724814982631,
"grad_norm": 0.6653648014910388,
"learning_rate": 1.900311526479751e-06,
"loss": 0.4209,
"step": 3250
},
{
"epoch": 0.9847455067210391,
"grad_norm": 0.6129065506951287,
"learning_rate": 1.588785046728972e-06,
"loss": 0.4299,
"step": 3260
},
{
"epoch": 0.9877661984594472,
"grad_norm": 0.684618245926937,
"learning_rate": 1.2772585669781931e-06,
"loss": 0.4901,
"step": 3270
},
{
"epoch": 0.9907868901978553,
"grad_norm": 0.6670737405214955,
"learning_rate": 9.657320872274145e-07,
"loss": 0.4904,
"step": 3280
},
{
"epoch": 0.9938075819362634,
"grad_norm": 0.593768140583266,
"learning_rate": 6.542056074766356e-07,
"loss": 0.4567,
"step": 3290
},
{
"epoch": 0.9968282736746715,
"grad_norm": 0.7284904929213815,
"learning_rate": 3.4267912772585674e-07,
"loss": 0.4917,
"step": 3300
},
{
"epoch": 0.9998489654130795,
"grad_norm": 0.5502017590847491,
"learning_rate": 3.1152647975077883e-08,
"loss": 0.4985,
"step": 3310
}
],
"logging_steps": 10,
"max_steps": 3310,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 0.0,
"train_batch_size": 32,
"trial_name": null,
"trial_params": null
}