PEFT
Safetensors
deu05232's picture
Upload folder using huggingface_hub
67636c5 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": 6.107433442294529,
"learning_rate": 4.9999999999999996e-05,
"loss": 6.4,
"step": 10
},
{
"epoch": 0.006041383476816191,
"grad_norm": 4.660591824324107,
"learning_rate": 6.505149978319905e-05,
"loss": 3.4203,
"step": 20
},
{
"epoch": 0.009062075215224286,
"grad_norm": 2.5455195791703775,
"learning_rate": 7.385606273598311e-05,
"loss": 1.7074,
"step": 30
},
{
"epoch": 0.012082766953632381,
"grad_norm": 2.371425231871035,
"learning_rate": 8.01029995663981e-05,
"loss": 1.3441,
"step": 40
},
{
"epoch": 0.015103458692040477,
"grad_norm": 2.9507148039059388,
"learning_rate": 8.494850021680092e-05,
"loss": 1.2033,
"step": 50
},
{
"epoch": 0.01812415043044857,
"grad_norm": 1.6469059513571374,
"learning_rate": 8.890756251918216e-05,
"loss": 1.0295,
"step": 60
},
{
"epoch": 0.02114484216885667,
"grad_norm": 2.0804468309368254,
"learning_rate": 9.225490200071284e-05,
"loss": 0.9693,
"step": 70
},
{
"epoch": 0.024165533907264763,
"grad_norm": 1.7068736962733648,
"learning_rate": 9.515449934959716e-05,
"loss": 0.9797,
"step": 80
},
{
"epoch": 0.02718622564567286,
"grad_norm": 1.4319371435452608,
"learning_rate": 9.771212547196623e-05,
"loss": 0.9646,
"step": 90
},
{
"epoch": 0.030206917384080954,
"grad_norm": 1.6355409473785059,
"learning_rate": 9.999999999999999e-05,
"loss": 0.9736,
"step": 100
},
{
"epoch": 0.03322760912248905,
"grad_norm": 1.5338040764461724,
"learning_rate": 9.97196261682243e-05,
"loss": 0.9596,
"step": 110
},
{
"epoch": 0.03624830086089714,
"grad_norm": 1.5781044711706278,
"learning_rate": 9.940809968847352e-05,
"loss": 0.9795,
"step": 120
},
{
"epoch": 0.03926899259930524,
"grad_norm": 1.3121521980332769,
"learning_rate": 9.909657320872274e-05,
"loss": 0.8775,
"step": 130
},
{
"epoch": 0.04228968433771334,
"grad_norm": 1.324861163031561,
"learning_rate": 9.878504672897197e-05,
"loss": 0.8459,
"step": 140
},
{
"epoch": 0.045310376076121435,
"grad_norm": 1.4277897482579234,
"learning_rate": 9.847352024922119e-05,
"loss": 0.8387,
"step": 150
},
{
"epoch": 0.048331067814529526,
"grad_norm": 0.9846724584363737,
"learning_rate": 9.816199376947041e-05,
"loss": 0.8412,
"step": 160
},
{
"epoch": 0.05135175955293762,
"grad_norm": 1.1379005965994144,
"learning_rate": 9.785046728971962e-05,
"loss": 0.858,
"step": 170
},
{
"epoch": 0.05437245129134572,
"grad_norm": 1.1113195573564463,
"learning_rate": 9.753894080996884e-05,
"loss": 0.8396,
"step": 180
},
{
"epoch": 0.05739314302975381,
"grad_norm": 1.1747287584404909,
"learning_rate": 9.722741433021807e-05,
"loss": 0.8455,
"step": 190
},
{
"epoch": 0.06041383476816191,
"grad_norm": 1.14829796924353,
"learning_rate": 9.691588785046729e-05,
"loss": 0.8162,
"step": 200
},
{
"epoch": 0.06343452650657,
"grad_norm": 1.6865130640347978,
"learning_rate": 9.660436137071651e-05,
"loss": 0.8049,
"step": 210
},
{
"epoch": 0.0664552182449781,
"grad_norm": 1.1210795530836306,
"learning_rate": 9.629283489096573e-05,
"loss": 0.8898,
"step": 220
},
{
"epoch": 0.0694759099833862,
"grad_norm": 1.3104253794887102,
"learning_rate": 9.598130841121496e-05,
"loss": 0.893,
"step": 230
},
{
"epoch": 0.07249660172179428,
"grad_norm": 0.9458837772298425,
"learning_rate": 9.566978193146418e-05,
"loss": 0.7352,
"step": 240
},
{
"epoch": 0.07551729346020239,
"grad_norm": 0.9994092468295118,
"learning_rate": 9.53582554517134e-05,
"loss": 0.8492,
"step": 250
},
{
"epoch": 0.07853798519861048,
"grad_norm": 1.2445159149963765,
"learning_rate": 9.504672897196261e-05,
"loss": 0.8211,
"step": 260
},
{
"epoch": 0.08155867693701857,
"grad_norm": 1.1031604885510584,
"learning_rate": 9.473520249221183e-05,
"loss": 0.8156,
"step": 270
},
{
"epoch": 0.08457936867542667,
"grad_norm": 1.08292480451744,
"learning_rate": 9.442367601246106e-05,
"loss": 0.6854,
"step": 280
},
{
"epoch": 0.08760006041383477,
"grad_norm": 0.8919075945445597,
"learning_rate": 9.411214953271028e-05,
"loss": 0.756,
"step": 290
},
{
"epoch": 0.09062075215224287,
"grad_norm": 1.2189933071330217,
"learning_rate": 9.38006230529595e-05,
"loss": 0.7294,
"step": 300
},
{
"epoch": 0.09364144389065096,
"grad_norm": 1.0428293335082384,
"learning_rate": 9.348909657320873e-05,
"loss": 0.7722,
"step": 310
},
{
"epoch": 0.09666213562905905,
"grad_norm": 1.1523183980366285,
"learning_rate": 9.317757009345795e-05,
"loss": 0.7602,
"step": 320
},
{
"epoch": 0.09968282736746716,
"grad_norm": 1.6661125066872844,
"learning_rate": 9.286604361370717e-05,
"loss": 0.793,
"step": 330
},
{
"epoch": 0.10270351910587525,
"grad_norm": 1.6388560265645526,
"learning_rate": 9.255451713395639e-05,
"loss": 0.7498,
"step": 340
},
{
"epoch": 0.10572421084428334,
"grad_norm": 1.1151288526789225,
"learning_rate": 9.22429906542056e-05,
"loss": 0.7476,
"step": 350
},
{
"epoch": 0.10874490258269144,
"grad_norm": 1.1716715572761998,
"learning_rate": 9.193146417445482e-05,
"loss": 0.7764,
"step": 360
},
{
"epoch": 0.11176559432109953,
"grad_norm": 1.1269421780352205,
"learning_rate": 9.161993769470405e-05,
"loss": 0.7246,
"step": 370
},
{
"epoch": 0.11478628605950762,
"grad_norm": 1.1869965443210357,
"learning_rate": 9.130841121495327e-05,
"loss": 0.7896,
"step": 380
},
{
"epoch": 0.11780697779791573,
"grad_norm": 1.0769160388337087,
"learning_rate": 9.099688473520249e-05,
"loss": 0.7696,
"step": 390
},
{
"epoch": 0.12082766953632382,
"grad_norm": 0.8590033803431932,
"learning_rate": 9.068535825545172e-05,
"loss": 0.7507,
"step": 400
},
{
"epoch": 0.12384836127473191,
"grad_norm": 1.0096874758539724,
"learning_rate": 9.037383177570094e-05,
"loss": 0.7424,
"step": 410
},
{
"epoch": 0.12686905301314,
"grad_norm": 1.1424291236899078,
"learning_rate": 9.006230529595016e-05,
"loss": 0.7352,
"step": 420
},
{
"epoch": 0.1298897447515481,
"grad_norm": 0.9309920132899709,
"learning_rate": 8.975077881619937e-05,
"loss": 0.8025,
"step": 430
},
{
"epoch": 0.1329104364899562,
"grad_norm": 0.9330207603118266,
"learning_rate": 8.943925233644859e-05,
"loss": 0.7444,
"step": 440
},
{
"epoch": 0.13593112822836428,
"grad_norm": 1.0426305811020373,
"learning_rate": 8.912772585669782e-05,
"loss": 0.728,
"step": 450
},
{
"epoch": 0.1389518199667724,
"grad_norm": 0.8256864364949085,
"learning_rate": 8.881619937694704e-05,
"loss": 0.7508,
"step": 460
},
{
"epoch": 0.1419725117051805,
"grad_norm": 0.9907438519838615,
"learning_rate": 8.850467289719626e-05,
"loss": 0.7418,
"step": 470
},
{
"epoch": 0.14499320344358857,
"grad_norm": 1.068101794840399,
"learning_rate": 8.819314641744548e-05,
"loss": 0.7414,
"step": 480
},
{
"epoch": 0.14801389518199667,
"grad_norm": 0.83782462293504,
"learning_rate": 8.78816199376947e-05,
"loss": 0.7424,
"step": 490
},
{
"epoch": 0.15103458692040478,
"grad_norm": 0.9497051551763392,
"learning_rate": 8.757009345794393e-05,
"loss": 0.7223,
"step": 500
},
{
"epoch": 0.15405527865881286,
"grad_norm": 0.9204069350053211,
"learning_rate": 8.725856697819315e-05,
"loss": 0.7381,
"step": 510
},
{
"epoch": 0.15707597039722096,
"grad_norm": 1.603736475383884,
"learning_rate": 8.694704049844236e-05,
"loss": 0.766,
"step": 520
},
{
"epoch": 0.16009666213562906,
"grad_norm": 0.9536306786605435,
"learning_rate": 8.663551401869158e-05,
"loss": 0.7324,
"step": 530
},
{
"epoch": 0.16311735387403714,
"grad_norm": 0.9251048779966528,
"learning_rate": 8.63239875389408e-05,
"loss": 0.7063,
"step": 540
},
{
"epoch": 0.16613804561244525,
"grad_norm": 0.9724970055024864,
"learning_rate": 8.601246105919003e-05,
"loss": 0.757,
"step": 550
},
{
"epoch": 0.16915873735085335,
"grad_norm": 1.0545809160156818,
"learning_rate": 8.570093457943925e-05,
"loss": 0.7216,
"step": 560
},
{
"epoch": 0.17217942908926145,
"grad_norm": 0.8476675542721974,
"learning_rate": 8.538940809968847e-05,
"loss": 0.7102,
"step": 570
},
{
"epoch": 0.17520012082766953,
"grad_norm": 1.1317207976980554,
"learning_rate": 8.50778816199377e-05,
"loss": 0.718,
"step": 580
},
{
"epoch": 0.17822081256607764,
"grad_norm": 0.9916049362507178,
"learning_rate": 8.476635514018692e-05,
"loss": 0.6864,
"step": 590
},
{
"epoch": 0.18124150430448574,
"grad_norm": 0.8652373721578226,
"learning_rate": 8.445482866043614e-05,
"loss": 0.7105,
"step": 600
},
{
"epoch": 0.18426219604289382,
"grad_norm": 1.1699410353273392,
"learning_rate": 8.414330218068535e-05,
"loss": 0.6818,
"step": 610
},
{
"epoch": 0.18728288778130192,
"grad_norm": 1.0789588004399187,
"learning_rate": 8.383177570093457e-05,
"loss": 0.7251,
"step": 620
},
{
"epoch": 0.19030357951971003,
"grad_norm": 1.0180526710834652,
"learning_rate": 8.35202492211838e-05,
"loss": 0.7223,
"step": 630
},
{
"epoch": 0.1933242712581181,
"grad_norm": 0.9048202099896194,
"learning_rate": 8.320872274143302e-05,
"loss": 0.7091,
"step": 640
},
{
"epoch": 0.1963449629965262,
"grad_norm": 0.7162765502274693,
"learning_rate": 8.289719626168224e-05,
"loss": 0.7302,
"step": 650
},
{
"epoch": 0.1993656547349343,
"grad_norm": 1.2538311853119164,
"learning_rate": 8.258566978193147e-05,
"loss": 0.7145,
"step": 660
},
{
"epoch": 0.2023863464733424,
"grad_norm": 1.0512084253999578,
"learning_rate": 8.227414330218069e-05,
"loss": 0.6951,
"step": 670
},
{
"epoch": 0.2054070382117505,
"grad_norm": 1.037367944272262,
"learning_rate": 8.196261682242991e-05,
"loss": 0.7394,
"step": 680
},
{
"epoch": 0.2084277299501586,
"grad_norm": 1.0297102626539856,
"learning_rate": 8.165109034267912e-05,
"loss": 0.6932,
"step": 690
},
{
"epoch": 0.21144842168856667,
"grad_norm": 1.1399417021007323,
"learning_rate": 8.133956386292834e-05,
"loss": 0.7127,
"step": 700
},
{
"epoch": 0.21446911342697478,
"grad_norm": 1.0467690678062298,
"learning_rate": 8.102803738317756e-05,
"loss": 0.7462,
"step": 710
},
{
"epoch": 0.21748980516538288,
"grad_norm": 0.9316716918640906,
"learning_rate": 8.071651090342679e-05,
"loss": 0.6787,
"step": 720
},
{
"epoch": 0.22051049690379096,
"grad_norm": 0.8657217241031085,
"learning_rate": 8.040498442367601e-05,
"loss": 0.7074,
"step": 730
},
{
"epoch": 0.22353118864219906,
"grad_norm": 1.0892372439549651,
"learning_rate": 8.009345794392523e-05,
"loss": 0.6419,
"step": 740
},
{
"epoch": 0.22655188038060717,
"grad_norm": 0.8388422615605526,
"learning_rate": 7.978193146417447e-05,
"loss": 0.7213,
"step": 750
},
{
"epoch": 0.22957257211901524,
"grad_norm": 0.8227586339047505,
"learning_rate": 7.947040498442369e-05,
"loss": 0.7275,
"step": 760
},
{
"epoch": 0.23259326385742335,
"grad_norm": 0.8604643672009238,
"learning_rate": 7.91588785046729e-05,
"loss": 0.684,
"step": 770
},
{
"epoch": 0.23561395559583145,
"grad_norm": 1.0386835816352997,
"learning_rate": 7.884735202492212e-05,
"loss": 0.687,
"step": 780
},
{
"epoch": 0.23863464733423953,
"grad_norm": 0.9089777843386366,
"learning_rate": 7.853582554517135e-05,
"loss": 0.7343,
"step": 790
},
{
"epoch": 0.24165533907264763,
"grad_norm": 0.9368976703259354,
"learning_rate": 7.822429906542057e-05,
"loss": 0.6764,
"step": 800
},
{
"epoch": 0.24467603081105574,
"grad_norm": 1.0484901190720786,
"learning_rate": 7.791277258566979e-05,
"loss": 0.7044,
"step": 810
},
{
"epoch": 0.24769672254946382,
"grad_norm": 0.9120908558656108,
"learning_rate": 7.760124610591901e-05,
"loss": 0.6773,
"step": 820
},
{
"epoch": 0.2507174142878719,
"grad_norm": 0.9933341742549251,
"learning_rate": 7.728971962616824e-05,
"loss": 0.7414,
"step": 830
},
{
"epoch": 0.25373810602628,
"grad_norm": 0.8906670834455529,
"learning_rate": 7.697819314641746e-05,
"loss": 0.699,
"step": 840
},
{
"epoch": 0.25675879776468813,
"grad_norm": 0.8754559347982087,
"learning_rate": 7.666666666666667e-05,
"loss": 0.7741,
"step": 850
},
{
"epoch": 0.2597794895030962,
"grad_norm": 1.1337062688099944,
"learning_rate": 7.635514018691589e-05,
"loss": 0.6972,
"step": 860
},
{
"epoch": 0.2628001812415043,
"grad_norm": 0.8921512134528322,
"learning_rate": 7.604361370716511e-05,
"loss": 0.7081,
"step": 870
},
{
"epoch": 0.2658208729799124,
"grad_norm": 1.139923437332287,
"learning_rate": 7.573208722741434e-05,
"loss": 0.7029,
"step": 880
},
{
"epoch": 0.2688415647183205,
"grad_norm": 0.8832629319764249,
"learning_rate": 7.542056074766356e-05,
"loss": 0.6877,
"step": 890
},
{
"epoch": 0.27186225645672857,
"grad_norm": 0.9566674159353702,
"learning_rate": 7.510903426791278e-05,
"loss": 0.7289,
"step": 900
},
{
"epoch": 0.2748829481951367,
"grad_norm": 0.8944322091168808,
"learning_rate": 7.4797507788162e-05,
"loss": 0.6798,
"step": 910
},
{
"epoch": 0.2779036399335448,
"grad_norm": 0.8741663428923058,
"learning_rate": 7.448598130841123e-05,
"loss": 0.6665,
"step": 920
},
{
"epoch": 0.28092433167195285,
"grad_norm": 1.2528199367337114,
"learning_rate": 7.417445482866045e-05,
"loss": 0.6814,
"step": 930
},
{
"epoch": 0.283945023410361,
"grad_norm": 0.8109577329944124,
"learning_rate": 7.386292834890966e-05,
"loss": 0.7122,
"step": 940
},
{
"epoch": 0.28696571514876906,
"grad_norm": 0.9453952285607337,
"learning_rate": 7.355140186915888e-05,
"loss": 0.6697,
"step": 950
},
{
"epoch": 0.28998640688717714,
"grad_norm": 1.1044067806495586,
"learning_rate": 7.32398753894081e-05,
"loss": 0.6553,
"step": 960
},
{
"epoch": 0.29300709862558527,
"grad_norm": 1.0302434791108652,
"learning_rate": 7.292834890965733e-05,
"loss": 0.724,
"step": 970
},
{
"epoch": 0.29602779036399335,
"grad_norm": 0.8626366458018235,
"learning_rate": 7.261682242990655e-05,
"loss": 0.6743,
"step": 980
},
{
"epoch": 0.2990484821024014,
"grad_norm": 0.9966319044864294,
"learning_rate": 7.230529595015577e-05,
"loss": 0.7529,
"step": 990
},
{
"epoch": 0.30206917384080956,
"grad_norm": 1.083570860936549,
"learning_rate": 7.1993769470405e-05,
"loss": 0.7124,
"step": 1000
},
{
"epoch": 0.30508986557921763,
"grad_norm": 0.8349298190301959,
"learning_rate": 7.168224299065422e-05,
"loss": 0.7182,
"step": 1010
},
{
"epoch": 0.3081105573176257,
"grad_norm": 1.2700422022124451,
"learning_rate": 7.137071651090344e-05,
"loss": 0.6926,
"step": 1020
},
{
"epoch": 0.31113124905603384,
"grad_norm": 1.1942224057675908,
"learning_rate": 7.105919003115265e-05,
"loss": 0.6666,
"step": 1030
},
{
"epoch": 0.3141519407944419,
"grad_norm": 0.905534922755039,
"learning_rate": 7.074766355140187e-05,
"loss": 0.7035,
"step": 1040
},
{
"epoch": 0.31717263253285,
"grad_norm": 0.8523161299606125,
"learning_rate": 7.04361370716511e-05,
"loss": 0.671,
"step": 1050
},
{
"epoch": 0.32019332427125813,
"grad_norm": 0.9817425166105982,
"learning_rate": 7.012461059190032e-05,
"loss": 0.7019,
"step": 1060
},
{
"epoch": 0.3232140160096662,
"grad_norm": 0.9551195155002241,
"learning_rate": 6.981308411214954e-05,
"loss": 0.6604,
"step": 1070
},
{
"epoch": 0.3262347077480743,
"grad_norm": 1.1486770353212066,
"learning_rate": 6.950155763239876e-05,
"loss": 0.6723,
"step": 1080
},
{
"epoch": 0.3292553994864824,
"grad_norm": 0.8661522940998408,
"learning_rate": 6.919003115264799e-05,
"loss": 0.6584,
"step": 1090
},
{
"epoch": 0.3322760912248905,
"grad_norm": 0.843165483336464,
"learning_rate": 6.887850467289721e-05,
"loss": 0.7084,
"step": 1100
},
{
"epoch": 0.3352967829632986,
"grad_norm": 1.0537988556287068,
"learning_rate": 6.856697819314643e-05,
"loss": 0.7206,
"step": 1110
},
{
"epoch": 0.3383174747017067,
"grad_norm": 1.083070190670757,
"learning_rate": 6.825545171339564e-05,
"loss": 0.6401,
"step": 1120
},
{
"epoch": 0.3413381664401148,
"grad_norm": 0.9463301153915464,
"learning_rate": 6.794392523364486e-05,
"loss": 0.6485,
"step": 1130
},
{
"epoch": 0.3443588581785229,
"grad_norm": 1.162687888943258,
"learning_rate": 6.763239875389409e-05,
"loss": 0.6798,
"step": 1140
},
{
"epoch": 0.347379549916931,
"grad_norm": 1.2355931335474732,
"learning_rate": 6.732087227414331e-05,
"loss": 0.6977,
"step": 1150
},
{
"epoch": 0.35040024165533906,
"grad_norm": 0.970203148116871,
"learning_rate": 6.700934579439253e-05,
"loss": 0.6578,
"step": 1160
},
{
"epoch": 0.3534209333937472,
"grad_norm": 0.6957094239818695,
"learning_rate": 6.669781931464175e-05,
"loss": 0.6967,
"step": 1170
},
{
"epoch": 0.35644162513215527,
"grad_norm": 1.0847258063198464,
"learning_rate": 6.638629283489098e-05,
"loss": 0.7067,
"step": 1180
},
{
"epoch": 0.35946231687056335,
"grad_norm": 1.016492287492506,
"learning_rate": 6.60747663551402e-05,
"loss": 0.6696,
"step": 1190
},
{
"epoch": 0.3624830086089715,
"grad_norm": 0.9585856007952032,
"learning_rate": 6.576323987538941e-05,
"loss": 0.6657,
"step": 1200
},
{
"epoch": 0.36550370034737956,
"grad_norm": 0.8864078227573705,
"learning_rate": 6.545171339563863e-05,
"loss": 0.6328,
"step": 1210
},
{
"epoch": 0.36852439208578763,
"grad_norm": 0.8943274613477806,
"learning_rate": 6.514018691588785e-05,
"loss": 0.6506,
"step": 1220
},
{
"epoch": 0.37154508382419577,
"grad_norm": 0.8332256298730885,
"learning_rate": 6.482866043613708e-05,
"loss": 0.6971,
"step": 1230
},
{
"epoch": 0.37456577556260384,
"grad_norm": 1.17330822337103,
"learning_rate": 6.45171339563863e-05,
"loss": 0.6422,
"step": 1240
},
{
"epoch": 0.3775864673010119,
"grad_norm": 0.9809759700466929,
"learning_rate": 6.420560747663552e-05,
"loss": 0.6917,
"step": 1250
},
{
"epoch": 0.38060715903942005,
"grad_norm": 0.832719810674294,
"learning_rate": 6.389408099688475e-05,
"loss": 0.6782,
"step": 1260
},
{
"epoch": 0.3836278507778281,
"grad_norm": 1.0543252158269494,
"learning_rate": 6.358255451713397e-05,
"loss": 0.7219,
"step": 1270
},
{
"epoch": 0.3866485425162362,
"grad_norm": 1.0467694974631614,
"learning_rate": 6.327102803738319e-05,
"loss": 0.6566,
"step": 1280
},
{
"epoch": 0.38966923425464434,
"grad_norm": 0.9218363850620362,
"learning_rate": 6.29595015576324e-05,
"loss": 0.69,
"step": 1290
},
{
"epoch": 0.3926899259930524,
"grad_norm": 1.1113741348986488,
"learning_rate": 6.264797507788162e-05,
"loss": 0.6374,
"step": 1300
},
{
"epoch": 0.3957106177314605,
"grad_norm": 1.006508842200885,
"learning_rate": 6.233644859813084e-05,
"loss": 0.6842,
"step": 1310
},
{
"epoch": 0.3987313094698686,
"grad_norm": 0.9247754231554504,
"learning_rate": 6.202492211838007e-05,
"loss": 0.6898,
"step": 1320
},
{
"epoch": 0.4017520012082767,
"grad_norm": 0.9397717237919477,
"learning_rate": 6.171339563862929e-05,
"loss": 0.6354,
"step": 1330
},
{
"epoch": 0.4047726929466848,
"grad_norm": 0.937179353987926,
"learning_rate": 6.140186915887851e-05,
"loss": 0.6588,
"step": 1340
},
{
"epoch": 0.4077933846850929,
"grad_norm": 1.0139109509887487,
"learning_rate": 6.109034267912774e-05,
"loss": 0.7113,
"step": 1350
},
{
"epoch": 0.410814076423501,
"grad_norm": 1.0157094514933067,
"learning_rate": 6.077881619937695e-05,
"loss": 0.7123,
"step": 1360
},
{
"epoch": 0.41383476816190906,
"grad_norm": 1.0322875686173636,
"learning_rate": 6.0467289719626174e-05,
"loss": 0.6943,
"step": 1370
},
{
"epoch": 0.4168554599003172,
"grad_norm": 1.0442005662047973,
"learning_rate": 6.01557632398754e-05,
"loss": 0.6535,
"step": 1380
},
{
"epoch": 0.41987615163872527,
"grad_norm": 0.9694302300151602,
"learning_rate": 5.984423676012462e-05,
"loss": 0.6777,
"step": 1390
},
{
"epoch": 0.42289684337713335,
"grad_norm": 0.856557806727891,
"learning_rate": 5.9532710280373836e-05,
"loss": 0.61,
"step": 1400
},
{
"epoch": 0.4259175351155415,
"grad_norm": 0.9901693752629647,
"learning_rate": 5.922118380062306e-05,
"loss": 0.6854,
"step": 1410
},
{
"epoch": 0.42893822685394956,
"grad_norm": 0.783217290208151,
"learning_rate": 5.890965732087228e-05,
"loss": 0.6706,
"step": 1420
},
{
"epoch": 0.43195891859235763,
"grad_norm": 0.8900001933069731,
"learning_rate": 5.8598130841121504e-05,
"loss": 0.6652,
"step": 1430
},
{
"epoch": 0.43497961033076576,
"grad_norm": 0.9412258235027727,
"learning_rate": 5.828660436137072e-05,
"loss": 0.637,
"step": 1440
},
{
"epoch": 0.43800030206917384,
"grad_norm": 0.8461686729968316,
"learning_rate": 5.797507788161994e-05,
"loss": 0.6709,
"step": 1450
},
{
"epoch": 0.4410209938075819,
"grad_norm": 0.8280644274865492,
"learning_rate": 5.7663551401869165e-05,
"loss": 0.6834,
"step": 1460
},
{
"epoch": 0.44404168554599005,
"grad_norm": 1.0036203683480829,
"learning_rate": 5.735202492211839e-05,
"loss": 0.6628,
"step": 1470
},
{
"epoch": 0.4470623772843981,
"grad_norm": 0.90528861522367,
"learning_rate": 5.704049844236761e-05,
"loss": 0.6304,
"step": 1480
},
{
"epoch": 0.4500830690228062,
"grad_norm": 0.9895764065229908,
"learning_rate": 5.6728971962616826e-05,
"loss": 0.6256,
"step": 1490
},
{
"epoch": 0.45310376076121434,
"grad_norm": 1.0865721318694312,
"learning_rate": 5.641744548286605e-05,
"loss": 0.6482,
"step": 1500
},
{
"epoch": 0.4561244524996224,
"grad_norm": 0.9252537781033792,
"learning_rate": 5.610591900311527e-05,
"loss": 0.664,
"step": 1510
},
{
"epoch": 0.4591451442380305,
"grad_norm": 0.9427528029529579,
"learning_rate": 5.5794392523364494e-05,
"loss": 0.6186,
"step": 1520
},
{
"epoch": 0.4621658359764386,
"grad_norm": 1.1501401377919116,
"learning_rate": 5.548286604361371e-05,
"loss": 0.6666,
"step": 1530
},
{
"epoch": 0.4651865277148467,
"grad_norm": 1.0127997306879968,
"learning_rate": 5.517133956386293e-05,
"loss": 0.633,
"step": 1540
},
{
"epoch": 0.4682072194532548,
"grad_norm": 0.9889020576837366,
"learning_rate": 5.4859813084112156e-05,
"loss": 0.6648,
"step": 1550
},
{
"epoch": 0.4712279111916629,
"grad_norm": 0.9133242828971371,
"learning_rate": 5.454828660436138e-05,
"loss": 0.6658,
"step": 1560
},
{
"epoch": 0.474248602930071,
"grad_norm": 0.8519147169548464,
"learning_rate": 5.4236760124610594e-05,
"loss": 0.6655,
"step": 1570
},
{
"epoch": 0.47726929466847906,
"grad_norm": 1.1425413437967364,
"learning_rate": 5.392523364485982e-05,
"loss": 0.6234,
"step": 1580
},
{
"epoch": 0.4802899864068872,
"grad_norm": 0.9407632886538477,
"learning_rate": 5.361370716510904e-05,
"loss": 0.6655,
"step": 1590
},
{
"epoch": 0.48331067814529527,
"grad_norm": 0.8778463761915637,
"learning_rate": 5.330218068535826e-05,
"loss": 0.6193,
"step": 1600
},
{
"epoch": 0.48633136988370335,
"grad_norm": 0.9088581291491995,
"learning_rate": 5.2990654205607485e-05,
"loss": 0.6959,
"step": 1610
},
{
"epoch": 0.4893520616221115,
"grad_norm": 0.8880069698564996,
"learning_rate": 5.26791277258567e-05,
"loss": 0.6302,
"step": 1620
},
{
"epoch": 0.49237275336051955,
"grad_norm": 1.1000634805590273,
"learning_rate": 5.2367601246105924e-05,
"loss": 0.6609,
"step": 1630
},
{
"epoch": 0.49539344509892763,
"grad_norm": 1.0376440595852419,
"learning_rate": 5.2056074766355146e-05,
"loss": 0.6493,
"step": 1640
},
{
"epoch": 0.49841413683733576,
"grad_norm": 1.154723883593218,
"learning_rate": 5.174454828660437e-05,
"loss": 0.623,
"step": 1650
},
{
"epoch": 0.5014348285757438,
"grad_norm": 1.0225782536084573,
"learning_rate": 5.1433021806853585e-05,
"loss": 0.6192,
"step": 1660
},
{
"epoch": 0.504455520314152,
"grad_norm": 0.8258028147484838,
"learning_rate": 5.112149532710281e-05,
"loss": 0.6226,
"step": 1670
},
{
"epoch": 0.50747621205256,
"grad_norm": 0.9881694368960537,
"learning_rate": 5.080996884735203e-05,
"loss": 0.6371,
"step": 1680
},
{
"epoch": 0.5104969037909681,
"grad_norm": 0.9387290792703848,
"learning_rate": 5.049844236760125e-05,
"loss": 0.6556,
"step": 1690
},
{
"epoch": 0.5135175955293763,
"grad_norm": 0.8124959833429422,
"learning_rate": 5.018691588785047e-05,
"loss": 0.6399,
"step": 1700
},
{
"epoch": 0.5165382872677843,
"grad_norm": 1.1013354676495737,
"learning_rate": 4.987538940809969e-05,
"loss": 0.6292,
"step": 1710
},
{
"epoch": 0.5195589790061924,
"grad_norm": 1.007870741398264,
"learning_rate": 4.9563862928348914e-05,
"loss": 0.6142,
"step": 1720
},
{
"epoch": 0.5225796707446005,
"grad_norm": 1.3444538643125503,
"learning_rate": 4.925233644859814e-05,
"loss": 0.6328,
"step": 1730
},
{
"epoch": 0.5256003624830086,
"grad_norm": 0.9385062120973191,
"learning_rate": 4.894080996884736e-05,
"loss": 0.64,
"step": 1740
},
{
"epoch": 0.5286210542214167,
"grad_norm": 1.0285328435343573,
"learning_rate": 4.8629283489096575e-05,
"loss": 0.6573,
"step": 1750
},
{
"epoch": 0.5316417459598248,
"grad_norm": 1.0469608587540762,
"learning_rate": 4.83177570093458e-05,
"loss": 0.647,
"step": 1760
},
{
"epoch": 0.5346624376982329,
"grad_norm": 1.357967431771865,
"learning_rate": 4.800623052959502e-05,
"loss": 0.6804,
"step": 1770
},
{
"epoch": 0.537683129436641,
"grad_norm": 1.0950853932379374,
"learning_rate": 4.7694704049844244e-05,
"loss": 0.6149,
"step": 1780
},
{
"epoch": 0.5407038211750491,
"grad_norm": 1.0159260957400553,
"learning_rate": 4.738317757009346e-05,
"loss": 0.6401,
"step": 1790
},
{
"epoch": 0.5437245129134571,
"grad_norm": 0.9884514226310773,
"learning_rate": 4.707165109034268e-05,
"loss": 0.6498,
"step": 1800
},
{
"epoch": 0.5467452046518653,
"grad_norm": 1.0160752710804692,
"learning_rate": 4.6760124610591905e-05,
"loss": 0.6315,
"step": 1810
},
{
"epoch": 0.5497658963902734,
"grad_norm": 1.147711843904364,
"learning_rate": 4.644859813084113e-05,
"loss": 0.6904,
"step": 1820
},
{
"epoch": 0.5527865881286814,
"grad_norm": 0.7714381685647183,
"learning_rate": 4.6137071651090343e-05,
"loss": 0.7019,
"step": 1830
},
{
"epoch": 0.5558072798670896,
"grad_norm": 0.8473230671521857,
"learning_rate": 4.5825545171339566e-05,
"loss": 0.6265,
"step": 1840
},
{
"epoch": 0.5588279716054977,
"grad_norm": 1.1527301512867654,
"learning_rate": 4.551401869158879e-05,
"loss": 0.6497,
"step": 1850
},
{
"epoch": 0.5618486633439057,
"grad_norm": 0.8079351530349161,
"learning_rate": 4.520249221183801e-05,
"loss": 0.6327,
"step": 1860
},
{
"epoch": 0.5648693550823138,
"grad_norm": 1.1580409714700377,
"learning_rate": 4.4890965732087234e-05,
"loss": 0.6779,
"step": 1870
},
{
"epoch": 0.567890046820722,
"grad_norm": 0.9514427291600485,
"learning_rate": 4.457943925233645e-05,
"loss": 0.6332,
"step": 1880
},
{
"epoch": 0.57091073855913,
"grad_norm": 0.7918488230351379,
"learning_rate": 4.426791277258567e-05,
"loss": 0.6119,
"step": 1890
},
{
"epoch": 0.5739314302975381,
"grad_norm": 1.0441374885870536,
"learning_rate": 4.3956386292834895e-05,
"loss": 0.6426,
"step": 1900
},
{
"epoch": 0.5769521220359463,
"grad_norm": 0.8939930818182591,
"learning_rate": 4.364485981308412e-05,
"loss": 0.5839,
"step": 1910
},
{
"epoch": 0.5799728137743543,
"grad_norm": 0.9746564212859139,
"learning_rate": 4.3333333333333334e-05,
"loss": 0.6926,
"step": 1920
},
{
"epoch": 0.5829935055127624,
"grad_norm": 1.0149681504761499,
"learning_rate": 4.302180685358256e-05,
"loss": 0.6386,
"step": 1930
},
{
"epoch": 0.5860141972511705,
"grad_norm": 0.875920209522558,
"learning_rate": 4.271028037383178e-05,
"loss": 0.7005,
"step": 1940
},
{
"epoch": 0.5890348889895786,
"grad_norm": 0.8668952964577089,
"learning_rate": 4.2398753894081e-05,
"loss": 0.6182,
"step": 1950
},
{
"epoch": 0.5920555807279867,
"grad_norm": 0.9953573736179951,
"learning_rate": 4.2087227414330225e-05,
"loss": 0.6162,
"step": 1960
},
{
"epoch": 0.5950762724663948,
"grad_norm": 0.8773254581979226,
"learning_rate": 4.177570093457944e-05,
"loss": 0.6612,
"step": 1970
},
{
"epoch": 0.5980969642048029,
"grad_norm": 1.0498588342027244,
"learning_rate": 4.1464174454828663e-05,
"loss": 0.5872,
"step": 1980
},
{
"epoch": 0.601117655943211,
"grad_norm": 0.8583607111635473,
"learning_rate": 4.1152647975077886e-05,
"loss": 0.6189,
"step": 1990
},
{
"epoch": 0.6041383476816191,
"grad_norm": 1.017062912575273,
"learning_rate": 4.084112149532711e-05,
"loss": 0.6654,
"step": 2000
},
{
"epoch": 0.6071590394200271,
"grad_norm": 0.924120469009652,
"learning_rate": 4.0529595015576325e-05,
"loss": 0.7291,
"step": 2010
},
{
"epoch": 0.6101797311584353,
"grad_norm": 0.8614010120137098,
"learning_rate": 4.021806853582555e-05,
"loss": 0.6339,
"step": 2020
},
{
"epoch": 0.6132004228968434,
"grad_norm": 1.0257955634224127,
"learning_rate": 3.990654205607477e-05,
"loss": 0.5762,
"step": 2030
},
{
"epoch": 0.6162211146352514,
"grad_norm": 0.8427293683655419,
"learning_rate": 3.959501557632399e-05,
"loss": 0.6203,
"step": 2040
},
{
"epoch": 0.6192418063736596,
"grad_norm": 0.9956600298614382,
"learning_rate": 3.928348909657321e-05,
"loss": 0.6715,
"step": 2050
},
{
"epoch": 0.6222624981120677,
"grad_norm": 0.9674162254082204,
"learning_rate": 3.897196261682243e-05,
"loss": 0.6394,
"step": 2060
},
{
"epoch": 0.6252831898504757,
"grad_norm": 0.9881381241452825,
"learning_rate": 3.8660436137071654e-05,
"loss": 0.6906,
"step": 2070
},
{
"epoch": 0.6283038815888838,
"grad_norm": 0.8420571914925713,
"learning_rate": 3.834890965732088e-05,
"loss": 0.6291,
"step": 2080
},
{
"epoch": 0.631324573327292,
"grad_norm": 0.7187740198739039,
"learning_rate": 3.80373831775701e-05,
"loss": 0.6202,
"step": 2090
},
{
"epoch": 0.6343452650657,
"grad_norm": 0.9804316800663279,
"learning_rate": 3.7725856697819315e-05,
"loss": 0.6561,
"step": 2100
},
{
"epoch": 0.6373659568041081,
"grad_norm": 0.8515569206563014,
"learning_rate": 3.741433021806854e-05,
"loss": 0.6293,
"step": 2110
},
{
"epoch": 0.6403866485425163,
"grad_norm": 0.9477963231654395,
"learning_rate": 3.710280373831776e-05,
"loss": 0.6406,
"step": 2120
},
{
"epoch": 0.6434073402809243,
"grad_norm": 0.8863714974409956,
"learning_rate": 3.6791277258566983e-05,
"loss": 0.6294,
"step": 2130
},
{
"epoch": 0.6464280320193324,
"grad_norm": 0.9053537817894837,
"learning_rate": 3.64797507788162e-05,
"loss": 0.6176,
"step": 2140
},
{
"epoch": 0.6494487237577405,
"grad_norm": 0.8847178095895625,
"learning_rate": 3.616822429906542e-05,
"loss": 0.6301,
"step": 2150
},
{
"epoch": 0.6524694154961486,
"grad_norm": 0.9228701990870082,
"learning_rate": 3.5856697819314645e-05,
"loss": 0.6339,
"step": 2160
},
{
"epoch": 0.6554901072345567,
"grad_norm": 0.9041368572529298,
"learning_rate": 3.554517133956387e-05,
"loss": 0.6957,
"step": 2170
},
{
"epoch": 0.6585107989729648,
"grad_norm": 1.0018714020772372,
"learning_rate": 3.523364485981308e-05,
"loss": 0.6757,
"step": 2180
},
{
"epoch": 0.6615314907113728,
"grad_norm": 0.7216294711027135,
"learning_rate": 3.4922118380062306e-05,
"loss": 0.6252,
"step": 2190
},
{
"epoch": 0.664552182449781,
"grad_norm": 1.0148006779880119,
"learning_rate": 3.461059190031153e-05,
"loss": 0.6176,
"step": 2200
},
{
"epoch": 0.6675728741881891,
"grad_norm": 1.041094582346374,
"learning_rate": 3.429906542056075e-05,
"loss": 0.632,
"step": 2210
},
{
"epoch": 0.6705935659265972,
"grad_norm": 1.001260036277429,
"learning_rate": 3.3987538940809974e-05,
"loss": 0.6531,
"step": 2220
},
{
"epoch": 0.6736142576650053,
"grad_norm": 1.0151906248677418,
"learning_rate": 3.367601246105919e-05,
"loss": 0.706,
"step": 2230
},
{
"epoch": 0.6766349494034134,
"grad_norm": 1.456634530516443,
"learning_rate": 3.336448598130841e-05,
"loss": 0.6607,
"step": 2240
},
{
"epoch": 0.6796556411418215,
"grad_norm": 0.9959148544905448,
"learning_rate": 3.3052959501557635e-05,
"loss": 0.6151,
"step": 2250
},
{
"epoch": 0.6826763328802296,
"grad_norm": 0.939441994632688,
"learning_rate": 3.274143302180686e-05,
"loss": 0.6225,
"step": 2260
},
{
"epoch": 0.6856970246186377,
"grad_norm": 1.1017419382315832,
"learning_rate": 3.2429906542056074e-05,
"loss": 0.5902,
"step": 2270
},
{
"epoch": 0.6887177163570458,
"grad_norm": 0.8743226098309229,
"learning_rate": 3.21183800623053e-05,
"loss": 0.629,
"step": 2280
},
{
"epoch": 0.6917384080954538,
"grad_norm": 0.8065027246881831,
"learning_rate": 3.180685358255452e-05,
"loss": 0.6421,
"step": 2290
},
{
"epoch": 0.694759099833862,
"grad_norm": 0.8404972374604114,
"learning_rate": 3.149532710280374e-05,
"loss": 0.6569,
"step": 2300
},
{
"epoch": 0.6977797915722701,
"grad_norm": 0.91033782819202,
"learning_rate": 3.118380062305296e-05,
"loss": 0.6354,
"step": 2310
},
{
"epoch": 0.7008004833106781,
"grad_norm": 0.8921637340751595,
"learning_rate": 3.087227414330218e-05,
"loss": 0.558,
"step": 2320
},
{
"epoch": 0.7038211750490863,
"grad_norm": 1.0088629034165304,
"learning_rate": 3.05607476635514e-05,
"loss": 0.5927,
"step": 2330
},
{
"epoch": 0.7068418667874944,
"grad_norm": 1.0472637263941902,
"learning_rate": 3.0249221183800626e-05,
"loss": 0.6821,
"step": 2340
},
{
"epoch": 0.7098625585259024,
"grad_norm": 1.0052257653394303,
"learning_rate": 2.9937694704049845e-05,
"loss": 0.6623,
"step": 2350
},
{
"epoch": 0.7128832502643105,
"grad_norm": 0.9928128899145213,
"learning_rate": 2.9626168224299068e-05,
"loss": 0.6671,
"step": 2360
},
{
"epoch": 0.7159039420027187,
"grad_norm": 0.9953651700320172,
"learning_rate": 2.9314641744548287e-05,
"loss": 0.6367,
"step": 2370
},
{
"epoch": 0.7189246337411267,
"grad_norm": 0.8954168445059429,
"learning_rate": 2.900311526479751e-05,
"loss": 0.6109,
"step": 2380
},
{
"epoch": 0.7219453254795348,
"grad_norm": 1.217556665995295,
"learning_rate": 2.869158878504673e-05,
"loss": 0.6311,
"step": 2390
},
{
"epoch": 0.724966017217943,
"grad_norm": 0.9874465787267548,
"learning_rate": 2.8380062305295952e-05,
"loss": 0.6042,
"step": 2400
},
{
"epoch": 0.727986708956351,
"grad_norm": 1.1136870123871143,
"learning_rate": 2.806853582554517e-05,
"loss": 0.5993,
"step": 2410
},
{
"epoch": 0.7310074006947591,
"grad_norm": 0.7783417690202461,
"learning_rate": 2.7757009345794394e-05,
"loss": 0.5967,
"step": 2420
},
{
"epoch": 0.7340280924331672,
"grad_norm": 0.951773138059301,
"learning_rate": 2.7445482866043613e-05,
"loss": 0.6521,
"step": 2430
},
{
"epoch": 0.7370487841715753,
"grad_norm": 0.9078553991976845,
"learning_rate": 2.7133956386292836e-05,
"loss": 0.6416,
"step": 2440
},
{
"epoch": 0.7400694759099834,
"grad_norm": 0.9242016113932383,
"learning_rate": 2.682242990654206e-05,
"loss": 0.6719,
"step": 2450
},
{
"epoch": 0.7430901676483915,
"grad_norm": 1.0353473116863448,
"learning_rate": 2.6510903426791278e-05,
"loss": 0.65,
"step": 2460
},
{
"epoch": 0.7461108593867996,
"grad_norm": 0.8426951444641244,
"learning_rate": 2.61993769470405e-05,
"loss": 0.6339,
"step": 2470
},
{
"epoch": 0.7491315511252077,
"grad_norm": 0.8383725124970911,
"learning_rate": 2.588785046728972e-05,
"loss": 0.675,
"step": 2480
},
{
"epoch": 0.7521522428636158,
"grad_norm": 0.9277064869500664,
"learning_rate": 2.5576323987538943e-05,
"loss": 0.6472,
"step": 2490
},
{
"epoch": 0.7551729346020238,
"grad_norm": 0.9557054194305636,
"learning_rate": 2.5264797507788162e-05,
"loss": 0.6151,
"step": 2500
},
{
"epoch": 0.758193626340432,
"grad_norm": 1.0139041673735976,
"learning_rate": 2.4953271028037385e-05,
"loss": 0.6385,
"step": 2510
},
{
"epoch": 0.7612143180788401,
"grad_norm": 1.0041711442441041,
"learning_rate": 2.4641744548286604e-05,
"loss": 0.6376,
"step": 2520
},
{
"epoch": 0.7642350098172481,
"grad_norm": 0.8875244791821518,
"learning_rate": 2.4330218068535827e-05,
"loss": 0.6062,
"step": 2530
},
{
"epoch": 0.7672557015556563,
"grad_norm": 1.130151408893583,
"learning_rate": 2.4018691588785046e-05,
"loss": 0.6085,
"step": 2540
},
{
"epoch": 0.7702763932940644,
"grad_norm": 1.0398637115998273,
"learning_rate": 2.370716510903427e-05,
"loss": 0.6562,
"step": 2550
},
{
"epoch": 0.7732970850324724,
"grad_norm": 0.841375986278148,
"learning_rate": 2.3395638629283488e-05,
"loss": 0.6214,
"step": 2560
},
{
"epoch": 0.7763177767708805,
"grad_norm": 0.8993507112928878,
"learning_rate": 2.308411214953271e-05,
"loss": 0.5868,
"step": 2570
},
{
"epoch": 0.7793384685092887,
"grad_norm": 1.075075042360366,
"learning_rate": 2.2772585669781933e-05,
"loss": 0.6384,
"step": 2580
},
{
"epoch": 0.7823591602476967,
"grad_norm": 0.9237623320118347,
"learning_rate": 2.2461059190031153e-05,
"loss": 0.6729,
"step": 2590
},
{
"epoch": 0.7853798519861048,
"grad_norm": 0.984394577152236,
"learning_rate": 2.2149532710280375e-05,
"loss": 0.6119,
"step": 2600
},
{
"epoch": 0.788400543724513,
"grad_norm": 0.888270911636166,
"learning_rate": 2.1838006230529595e-05,
"loss": 0.6118,
"step": 2610
},
{
"epoch": 0.791421235462921,
"grad_norm": 0.8732881800528585,
"learning_rate": 2.1526479750778817e-05,
"loss": 0.6432,
"step": 2620
},
{
"epoch": 0.7944419272013291,
"grad_norm": 0.9279191040641048,
"learning_rate": 2.1214953271028037e-05,
"loss": 0.6522,
"step": 2630
},
{
"epoch": 0.7974626189397372,
"grad_norm": 0.9382587852562286,
"learning_rate": 2.090342679127726e-05,
"loss": 0.6409,
"step": 2640
},
{
"epoch": 0.8004833106781453,
"grad_norm": 0.989385068087765,
"learning_rate": 2.059190031152648e-05,
"loss": 0.6885,
"step": 2650
},
{
"epoch": 0.8035040024165534,
"grad_norm": 0.9604084382069871,
"learning_rate": 2.02803738317757e-05,
"loss": 0.6805,
"step": 2660
},
{
"epoch": 0.8065246941549615,
"grad_norm": 0.9654367538592937,
"learning_rate": 1.996884735202492e-05,
"loss": 0.6632,
"step": 2670
},
{
"epoch": 0.8095453858933696,
"grad_norm": 0.9920470968943375,
"learning_rate": 1.9657320872274143e-05,
"loss": 0.6824,
"step": 2680
},
{
"epoch": 0.8125660776317777,
"grad_norm": 0.7281653572711693,
"learning_rate": 1.9345794392523366e-05,
"loss": 0.6115,
"step": 2690
},
{
"epoch": 0.8155867693701858,
"grad_norm": 0.8504737220071418,
"learning_rate": 1.9034267912772585e-05,
"loss": 0.6609,
"step": 2700
},
{
"epoch": 0.8186074611085938,
"grad_norm": 0.9110986375915231,
"learning_rate": 1.8722741433021808e-05,
"loss": 0.6534,
"step": 2710
},
{
"epoch": 0.821628152847002,
"grad_norm": 0.9177181751033111,
"learning_rate": 1.8411214953271027e-05,
"loss": 0.6276,
"step": 2720
},
{
"epoch": 0.8246488445854101,
"grad_norm": 1.086494084811123,
"learning_rate": 1.809968847352025e-05,
"loss": 0.6593,
"step": 2730
},
{
"epoch": 0.8276695363238181,
"grad_norm": 1.0609001562665614,
"learning_rate": 1.778816199376947e-05,
"loss": 0.6123,
"step": 2740
},
{
"epoch": 0.8306902280622263,
"grad_norm": 0.8745174453272212,
"learning_rate": 1.7476635514018695e-05,
"loss": 0.5829,
"step": 2750
},
{
"epoch": 0.8337109198006344,
"grad_norm": 0.9525366144528878,
"learning_rate": 1.7165109034267915e-05,
"loss": 0.6271,
"step": 2760
},
{
"epoch": 0.8367316115390424,
"grad_norm": 0.8989045432013612,
"learning_rate": 1.6853582554517137e-05,
"loss": 0.6768,
"step": 2770
},
{
"epoch": 0.8397523032774505,
"grad_norm": 0.840003478358324,
"learning_rate": 1.6542056074766357e-05,
"loss": 0.6186,
"step": 2780
},
{
"epoch": 0.8427729950158587,
"grad_norm": 0.9379207795270732,
"learning_rate": 1.623052959501558e-05,
"loss": 0.5944,
"step": 2790
},
{
"epoch": 0.8457936867542667,
"grad_norm": 1.080258331678067,
"learning_rate": 1.59190031152648e-05,
"loss": 0.6389,
"step": 2800
},
{
"epoch": 0.8488143784926748,
"grad_norm": 0.9069933666502163,
"learning_rate": 1.560747663551402e-05,
"loss": 0.6444,
"step": 2810
},
{
"epoch": 0.851835070231083,
"grad_norm": 0.8513737933030265,
"learning_rate": 1.529595015576324e-05,
"loss": 0.6139,
"step": 2820
},
{
"epoch": 0.854855761969491,
"grad_norm": 1.0436096847604104,
"learning_rate": 1.4984423676012463e-05,
"loss": 0.6757,
"step": 2830
},
{
"epoch": 0.8578764537078991,
"grad_norm": 1.036008741711385,
"learning_rate": 1.4672897196261684e-05,
"loss": 0.6221,
"step": 2840
},
{
"epoch": 0.8608971454463072,
"grad_norm": 0.9143395192776882,
"learning_rate": 1.4361370716510905e-05,
"loss": 0.6034,
"step": 2850
},
{
"epoch": 0.8639178371847153,
"grad_norm": 1.0024441386017884,
"learning_rate": 1.4049844236760126e-05,
"loss": 0.6449,
"step": 2860
},
{
"epoch": 0.8669385289231234,
"grad_norm": 0.9238811734598441,
"learning_rate": 1.3738317757009347e-05,
"loss": 0.6376,
"step": 2870
},
{
"epoch": 0.8699592206615315,
"grad_norm": 0.9865135031755546,
"learning_rate": 1.3426791277258568e-05,
"loss": 0.6552,
"step": 2880
},
{
"epoch": 0.8729799123999396,
"grad_norm": 0.9354858745761598,
"learning_rate": 1.311526479750779e-05,
"loss": 0.5966,
"step": 2890
},
{
"epoch": 0.8760006041383477,
"grad_norm": 0.8219054966280139,
"learning_rate": 1.280373831775701e-05,
"loss": 0.6313,
"step": 2900
},
{
"epoch": 0.8790212958767558,
"grad_norm": 1.1222331415622748,
"learning_rate": 1.2492211838006231e-05,
"loss": 0.6372,
"step": 2910
},
{
"epoch": 0.8820419876151638,
"grad_norm": 1.0297626108339775,
"learning_rate": 1.2180685358255452e-05,
"loss": 0.5962,
"step": 2920
},
{
"epoch": 0.885062679353572,
"grad_norm": 1.0309391754326147,
"learning_rate": 1.1869158878504673e-05,
"loss": 0.579,
"step": 2930
},
{
"epoch": 0.8880833710919801,
"grad_norm": 0.9200328981623718,
"learning_rate": 1.1557632398753894e-05,
"loss": 0.6271,
"step": 2940
},
{
"epoch": 0.8911040628303881,
"grad_norm": 1.124884533654081,
"learning_rate": 1.1246105919003117e-05,
"loss": 0.6542,
"step": 2950
},
{
"epoch": 0.8941247545687963,
"grad_norm": 0.9833353574823354,
"learning_rate": 1.0934579439252338e-05,
"loss": 0.6235,
"step": 2960
},
{
"epoch": 0.8971454463072044,
"grad_norm": 0.9665933597367096,
"learning_rate": 1.0623052959501559e-05,
"loss": 0.5958,
"step": 2970
},
{
"epoch": 0.9001661380456124,
"grad_norm": 0.8837829200974788,
"learning_rate": 1.031152647975078e-05,
"loss": 0.6561,
"step": 2980
},
{
"epoch": 0.9031868297840205,
"grad_norm": 1.0024721414278113,
"learning_rate": 1e-05,
"loss": 0.5681,
"step": 2990
},
{
"epoch": 0.9062075215224287,
"grad_norm": 1.0234363190903126,
"learning_rate": 9.688473520249222e-06,
"loss": 0.658,
"step": 3000
},
{
"epoch": 0.9092282132608367,
"grad_norm": 0.9838276704004294,
"learning_rate": 9.376947040498443e-06,
"loss": 0.593,
"step": 3010
},
{
"epoch": 0.9122489049992448,
"grad_norm": 0.9928052558627496,
"learning_rate": 9.065420560747664e-06,
"loss": 0.6912,
"step": 3020
},
{
"epoch": 0.915269596737653,
"grad_norm": 0.9696263922235319,
"learning_rate": 8.753894080996885e-06,
"loss": 0.6025,
"step": 3030
},
{
"epoch": 0.918290288476061,
"grad_norm": 0.9475128150610571,
"learning_rate": 8.442367601246107e-06,
"loss": 0.5789,
"step": 3040
},
{
"epoch": 0.9213109802144691,
"grad_norm": 1.0108090645384262,
"learning_rate": 8.130841121495328e-06,
"loss": 0.6225,
"step": 3050
},
{
"epoch": 0.9243316719528772,
"grad_norm": 0.736126172716179,
"learning_rate": 7.81931464174455e-06,
"loss": 0.6154,
"step": 3060
},
{
"epoch": 0.9273523636912853,
"grad_norm": 1.109978326602373,
"learning_rate": 7.5077881619937705e-06,
"loss": 0.6248,
"step": 3070
},
{
"epoch": 0.9303730554296934,
"grad_norm": 0.9959434394352312,
"learning_rate": 7.1962616822429915e-06,
"loss": 0.6562,
"step": 3080
},
{
"epoch": 0.9333937471681015,
"grad_norm": 1.0494254480922158,
"learning_rate": 6.8847352024922125e-06,
"loss": 0.6705,
"step": 3090
},
{
"epoch": 0.9364144389065095,
"grad_norm": 1.067674565055919,
"learning_rate": 6.5732087227414335e-06,
"loss": 0.6389,
"step": 3100
},
{
"epoch": 0.9394351306449177,
"grad_norm": 0.9838274913424749,
"learning_rate": 6.2616822429906544e-06,
"loss": 0.5787,
"step": 3110
},
{
"epoch": 0.9424558223833258,
"grad_norm": 0.7939176738520116,
"learning_rate": 5.9501557632398754e-06,
"loss": 0.6622,
"step": 3120
},
{
"epoch": 0.9454765141217338,
"grad_norm": 0.9618550363845632,
"learning_rate": 5.638629283489097e-06,
"loss": 0.6009,
"step": 3130
},
{
"epoch": 0.948497205860142,
"grad_norm": 0.8840790862007406,
"learning_rate": 5.327102803738318e-06,
"loss": 0.5571,
"step": 3140
},
{
"epoch": 0.9515178975985501,
"grad_norm": 1.0010773882850883,
"learning_rate": 5.015576323987539e-06,
"loss": 0.6224,
"step": 3150
},
{
"epoch": 0.9545385893369581,
"grad_norm": 1.0192208568405081,
"learning_rate": 4.70404984423676e-06,
"loss": 0.6137,
"step": 3160
},
{
"epoch": 0.9575592810753663,
"grad_norm": 0.9396020025901988,
"learning_rate": 4.392523364485981e-06,
"loss": 0.672,
"step": 3170
},
{
"epoch": 0.9605799728137744,
"grad_norm": 1.012741559195394,
"learning_rate": 4.080996884735202e-06,
"loss": 0.6187,
"step": 3180
},
{
"epoch": 0.9636006645521824,
"grad_norm": 0.9330143064677612,
"learning_rate": 3.7694704049844237e-06,
"loss": 0.5604,
"step": 3190
},
{
"epoch": 0.9666213562905905,
"grad_norm": 0.7854847787959368,
"learning_rate": 3.4579439252336447e-06,
"loss": 0.5812,
"step": 3200
},
{
"epoch": 0.9696420480289987,
"grad_norm": 0.9706724255216592,
"learning_rate": 3.146417445482866e-06,
"loss": 0.5765,
"step": 3210
},
{
"epoch": 0.9726627397674067,
"grad_norm": 1.0578377615563888,
"learning_rate": 2.8348909657320875e-06,
"loss": 0.6107,
"step": 3220
},
{
"epoch": 0.9756834315058148,
"grad_norm": 1.0768624831372862,
"learning_rate": 2.5233644859813085e-06,
"loss": 0.6097,
"step": 3230
},
{
"epoch": 0.978704123244223,
"grad_norm": 1.0432636589549298,
"learning_rate": 2.21183800623053e-06,
"loss": 0.594,
"step": 3240
},
{
"epoch": 0.981724814982631,
"grad_norm": 0.8847000237311595,
"learning_rate": 1.900311526479751e-06,
"loss": 0.6035,
"step": 3250
},
{
"epoch": 0.9847455067210391,
"grad_norm": 0.8730660744496067,
"learning_rate": 1.588785046728972e-06,
"loss": 0.6033,
"step": 3260
},
{
"epoch": 0.9877661984594472,
"grad_norm": 0.9491022107794316,
"learning_rate": 1.2772585669781931e-06,
"loss": 0.6328,
"step": 3270
},
{
"epoch": 0.9907868901978553,
"grad_norm": 0.9392025377265674,
"learning_rate": 9.657320872274145e-07,
"loss": 0.5992,
"step": 3280
},
{
"epoch": 0.9938075819362634,
"grad_norm": 0.8318935354284523,
"learning_rate": 6.542056074766356e-07,
"loss": 0.5914,
"step": 3290
},
{
"epoch": 0.9968282736746715,
"grad_norm": 1.0319593075212372,
"learning_rate": 3.4267912772585674e-07,
"loss": 0.62,
"step": 3300
},
{
"epoch": 0.9998489654130795,
"grad_norm": 0.9838443353358589,
"learning_rate": 3.1152647975077883e-08,
"loss": 0.6305,
"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
}