PEFT
Safetensors
deu05232's picture
Upload folder using huggingface_hub
2f73625 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": 6.19309427213461,
"learning_rate": 4.9999999999999996e-05,
"loss": 6.2766,
"step": 10
},
{
"epoch": 0.006041383476816191,
"grad_norm": 4.662664520855649,
"learning_rate": 6.505149978319905e-05,
"loss": 3.3371,
"step": 20
},
{
"epoch": 0.009062075215224286,
"grad_norm": 3.4740694746564333,
"learning_rate": 7.385606273598311e-05,
"loss": 1.6125,
"step": 30
},
{
"epoch": 0.012082766953632381,
"grad_norm": 2.4504825462615445,
"learning_rate": 8.01029995663981e-05,
"loss": 1.2844,
"step": 40
},
{
"epoch": 0.015103458692040477,
"grad_norm": 2.979439099694636,
"learning_rate": 8.494850021680092e-05,
"loss": 1.1324,
"step": 50
},
{
"epoch": 0.01812415043044857,
"grad_norm": 2.0640372504873756,
"learning_rate": 8.890756251918216e-05,
"loss": 1.1203,
"step": 60
},
{
"epoch": 0.02114484216885667,
"grad_norm": 1.6439321309921382,
"learning_rate": 9.225490200071284e-05,
"loss": 1.0453,
"step": 70
},
{
"epoch": 0.024165533907264763,
"grad_norm": 1.6111224175364611,
"learning_rate": 9.515449934959716e-05,
"loss": 0.9713,
"step": 80
},
{
"epoch": 0.02718622564567286,
"grad_norm": 1.594476964360827,
"learning_rate": 9.771212547196623e-05,
"loss": 0.9342,
"step": 90
},
{
"epoch": 0.030206917384080954,
"grad_norm": 1.4499083844782663,
"learning_rate": 9.999999999999999e-05,
"loss": 0.8637,
"step": 100
},
{
"epoch": 0.03322760912248905,
"grad_norm": 1.3677640337703747,
"learning_rate": 9.97196261682243e-05,
"loss": 0.9406,
"step": 110
},
{
"epoch": 0.03624830086089714,
"grad_norm": 1.119404831347932,
"learning_rate": 9.940809968847352e-05,
"loss": 0.9506,
"step": 120
},
{
"epoch": 0.03926899259930524,
"grad_norm": 1.5618042154973153,
"learning_rate": 9.909657320872274e-05,
"loss": 0.8656,
"step": 130
},
{
"epoch": 0.04228968433771334,
"grad_norm": 1.5561772704529109,
"learning_rate": 9.878504672897197e-05,
"loss": 0.8889,
"step": 140
},
{
"epoch": 0.045310376076121435,
"grad_norm": 1.272567947440116,
"learning_rate": 9.847352024922119e-05,
"loss": 0.8756,
"step": 150
},
{
"epoch": 0.048331067814529526,
"grad_norm": 1.7068521135192898,
"learning_rate": 9.816199376947041e-05,
"loss": 0.882,
"step": 160
},
{
"epoch": 0.05135175955293762,
"grad_norm": 1.0894569556092386,
"learning_rate": 9.785046728971962e-05,
"loss": 0.8803,
"step": 170
},
{
"epoch": 0.05437245129134572,
"grad_norm": 1.3080888030272213,
"learning_rate": 9.753894080996884e-05,
"loss": 0.8285,
"step": 180
},
{
"epoch": 0.05739314302975381,
"grad_norm": 0.9550283589269429,
"learning_rate": 9.722741433021807e-05,
"loss": 0.8578,
"step": 190
},
{
"epoch": 0.06041383476816191,
"grad_norm": 1.3615437434812387,
"learning_rate": 9.691588785046729e-05,
"loss": 0.8359,
"step": 200
},
{
"epoch": 0.06343452650657,
"grad_norm": 0.9605970568605654,
"learning_rate": 9.660436137071651e-05,
"loss": 0.8076,
"step": 210
},
{
"epoch": 0.0664552182449781,
"grad_norm": 1.1329343940597405,
"learning_rate": 9.629283489096573e-05,
"loss": 0.8449,
"step": 220
},
{
"epoch": 0.0694759099833862,
"grad_norm": 1.1282886155957252,
"learning_rate": 9.598130841121496e-05,
"loss": 0.7922,
"step": 230
},
{
"epoch": 0.07249660172179428,
"grad_norm": 1.292056905093199,
"learning_rate": 9.566978193146418e-05,
"loss": 0.749,
"step": 240
},
{
"epoch": 0.07551729346020239,
"grad_norm": 1.2921850415396794,
"learning_rate": 9.53582554517134e-05,
"loss": 0.7672,
"step": 250
},
{
"epoch": 0.07853798519861048,
"grad_norm": 1.3022393245969244,
"learning_rate": 9.504672897196261e-05,
"loss": 0.8271,
"step": 260
},
{
"epoch": 0.08155867693701857,
"grad_norm": 1.1267601057630086,
"learning_rate": 9.473520249221183e-05,
"loss": 0.7986,
"step": 270
},
{
"epoch": 0.08457936867542667,
"grad_norm": 1.1205137168491301,
"learning_rate": 9.442367601246106e-05,
"loss": 0.7918,
"step": 280
},
{
"epoch": 0.08760006041383477,
"grad_norm": 1.1263421895412242,
"learning_rate": 9.411214953271028e-05,
"loss": 0.7566,
"step": 290
},
{
"epoch": 0.09062075215224287,
"grad_norm": 0.9340845037361807,
"learning_rate": 9.38006230529595e-05,
"loss": 0.744,
"step": 300
},
{
"epoch": 0.09364144389065096,
"grad_norm": 1.0623928380404488,
"learning_rate": 9.348909657320873e-05,
"loss": 0.8087,
"step": 310
},
{
"epoch": 0.09666213562905905,
"grad_norm": 0.9254895956168991,
"learning_rate": 9.317757009345795e-05,
"loss": 0.8035,
"step": 320
},
{
"epoch": 0.09968282736746716,
"grad_norm": 1.1100514511341473,
"learning_rate": 9.286604361370717e-05,
"loss": 0.81,
"step": 330
},
{
"epoch": 0.10270351910587525,
"grad_norm": 1.1642882203194822,
"learning_rate": 9.255451713395639e-05,
"loss": 0.7413,
"step": 340
},
{
"epoch": 0.10572421084428334,
"grad_norm": 1.0574631945981023,
"learning_rate": 9.22429906542056e-05,
"loss": 0.7035,
"step": 350
},
{
"epoch": 0.10874490258269144,
"grad_norm": 1.124067882434976,
"learning_rate": 9.193146417445482e-05,
"loss": 0.7619,
"step": 360
},
{
"epoch": 0.11176559432109953,
"grad_norm": 1.1787596255028467,
"learning_rate": 9.161993769470405e-05,
"loss": 0.7737,
"step": 370
},
{
"epoch": 0.11478628605950762,
"grad_norm": 0.9786890412818356,
"learning_rate": 9.130841121495327e-05,
"loss": 0.7513,
"step": 380
},
{
"epoch": 0.11780697779791573,
"grad_norm": 0.8877789393750584,
"learning_rate": 9.099688473520249e-05,
"loss": 0.8102,
"step": 390
},
{
"epoch": 0.12082766953632382,
"grad_norm": 1.461821958474022,
"learning_rate": 9.068535825545172e-05,
"loss": 0.7563,
"step": 400
},
{
"epoch": 0.12384836127473191,
"grad_norm": 1.078980625367024,
"learning_rate": 9.037383177570094e-05,
"loss": 0.7738,
"step": 410
},
{
"epoch": 0.12686905301314,
"grad_norm": 0.904976356792518,
"learning_rate": 9.006230529595016e-05,
"loss": 0.7701,
"step": 420
},
{
"epoch": 0.1298897447515481,
"grad_norm": 1.0192231852961207,
"learning_rate": 8.975077881619937e-05,
"loss": 0.7383,
"step": 430
},
{
"epoch": 0.1329104364899562,
"grad_norm": 1.0392965346479355,
"learning_rate": 8.943925233644859e-05,
"loss": 0.7619,
"step": 440
},
{
"epoch": 0.13593112822836428,
"grad_norm": 1.1003519530119685,
"learning_rate": 8.912772585669782e-05,
"loss": 0.7875,
"step": 450
},
{
"epoch": 0.1389518199667724,
"grad_norm": 1.0523930769134358,
"learning_rate": 8.881619937694704e-05,
"loss": 0.7482,
"step": 460
},
{
"epoch": 0.1419725117051805,
"grad_norm": 1.174632687542033,
"learning_rate": 8.850467289719626e-05,
"loss": 0.7738,
"step": 470
},
{
"epoch": 0.14499320344358857,
"grad_norm": 0.9776217932648514,
"learning_rate": 8.819314641744548e-05,
"loss": 0.7409,
"step": 480
},
{
"epoch": 0.14801389518199667,
"grad_norm": 1.0118904788844576,
"learning_rate": 8.78816199376947e-05,
"loss": 0.7258,
"step": 490
},
{
"epoch": 0.15103458692040478,
"grad_norm": 1.0568731288041235,
"learning_rate": 8.757009345794393e-05,
"loss": 0.716,
"step": 500
},
{
"epoch": 0.15405527865881286,
"grad_norm": 1.3079989179988274,
"learning_rate": 8.725856697819315e-05,
"loss": 0.7342,
"step": 510
},
{
"epoch": 0.15707597039722096,
"grad_norm": 1.1341303669370166,
"learning_rate": 8.694704049844236e-05,
"loss": 0.7221,
"step": 520
},
{
"epoch": 0.16009666213562906,
"grad_norm": 0.8552347998617839,
"learning_rate": 8.663551401869158e-05,
"loss": 0.7293,
"step": 530
},
{
"epoch": 0.16311735387403714,
"grad_norm": 0.9609818158633973,
"learning_rate": 8.63239875389408e-05,
"loss": 0.7479,
"step": 540
},
{
"epoch": 0.16613804561244525,
"grad_norm": 0.9432720426458855,
"learning_rate": 8.601246105919003e-05,
"loss": 0.7363,
"step": 550
},
{
"epoch": 0.16915873735085335,
"grad_norm": 1.1687881617165012,
"learning_rate": 8.570093457943925e-05,
"loss": 0.7099,
"step": 560
},
{
"epoch": 0.17217942908926145,
"grad_norm": 1.1176630402906178,
"learning_rate": 8.538940809968847e-05,
"loss": 0.743,
"step": 570
},
{
"epoch": 0.17520012082766953,
"grad_norm": 1.1412515199025173,
"learning_rate": 8.50778816199377e-05,
"loss": 0.7658,
"step": 580
},
{
"epoch": 0.17822081256607764,
"grad_norm": 1.02867994536102,
"learning_rate": 8.476635514018692e-05,
"loss": 0.6729,
"step": 590
},
{
"epoch": 0.18124150430448574,
"grad_norm": 1.082154820840065,
"learning_rate": 8.445482866043614e-05,
"loss": 0.7262,
"step": 600
},
{
"epoch": 0.18426219604289382,
"grad_norm": 1.2694919790994081,
"learning_rate": 8.414330218068535e-05,
"loss": 0.6877,
"step": 610
},
{
"epoch": 0.18728288778130192,
"grad_norm": 0.838053554184396,
"learning_rate": 8.383177570093457e-05,
"loss": 0.7487,
"step": 620
},
{
"epoch": 0.19030357951971003,
"grad_norm": 0.9221421412802141,
"learning_rate": 8.35202492211838e-05,
"loss": 0.7513,
"step": 630
},
{
"epoch": 0.1933242712581181,
"grad_norm": 1.392283589831917,
"learning_rate": 8.320872274143302e-05,
"loss": 0.7356,
"step": 640
},
{
"epoch": 0.1963449629965262,
"grad_norm": 0.9718812974756135,
"learning_rate": 8.289719626168224e-05,
"loss": 0.7398,
"step": 650
},
{
"epoch": 0.1993656547349343,
"grad_norm": 1.0116415820607176,
"learning_rate": 8.258566978193147e-05,
"loss": 0.7148,
"step": 660
},
{
"epoch": 0.2023863464733424,
"grad_norm": 1.0199465148237918,
"learning_rate": 8.227414330218069e-05,
"loss": 0.6546,
"step": 670
},
{
"epoch": 0.2054070382117505,
"grad_norm": 1.2125727840596712,
"learning_rate": 8.196261682242991e-05,
"loss": 0.6545,
"step": 680
},
{
"epoch": 0.2084277299501586,
"grad_norm": 0.9404678991103539,
"learning_rate": 8.165109034267912e-05,
"loss": 0.7408,
"step": 690
},
{
"epoch": 0.21144842168856667,
"grad_norm": 0.9716578982835177,
"learning_rate": 8.133956386292834e-05,
"loss": 0.6768,
"step": 700
},
{
"epoch": 0.21446911342697478,
"grad_norm": 0.7918631504914017,
"learning_rate": 8.102803738317756e-05,
"loss": 0.7063,
"step": 710
},
{
"epoch": 0.21748980516538288,
"grad_norm": 0.9263059529725423,
"learning_rate": 8.071651090342679e-05,
"loss": 0.7164,
"step": 720
},
{
"epoch": 0.22051049690379096,
"grad_norm": 1.0260520096805317,
"learning_rate": 8.040498442367601e-05,
"loss": 0.7434,
"step": 730
},
{
"epoch": 0.22353118864219906,
"grad_norm": 0.8723377162489638,
"learning_rate": 8.009345794392523e-05,
"loss": 0.7002,
"step": 740
},
{
"epoch": 0.22655188038060717,
"grad_norm": 1.0740728368307533,
"learning_rate": 7.978193146417447e-05,
"loss": 0.6652,
"step": 750
},
{
"epoch": 0.22957257211901524,
"grad_norm": 0.9456556214898678,
"learning_rate": 7.947040498442369e-05,
"loss": 0.7201,
"step": 760
},
{
"epoch": 0.23259326385742335,
"grad_norm": 0.8073940402881862,
"learning_rate": 7.91588785046729e-05,
"loss": 0.6856,
"step": 770
},
{
"epoch": 0.23561395559583145,
"grad_norm": 1.0593516460675978,
"learning_rate": 7.884735202492212e-05,
"loss": 0.6857,
"step": 780
},
{
"epoch": 0.23863464733423953,
"grad_norm": 0.8521007828434903,
"learning_rate": 7.853582554517135e-05,
"loss": 0.7057,
"step": 790
},
{
"epoch": 0.24165533907264763,
"grad_norm": 1.0583599191053652,
"learning_rate": 7.822429906542057e-05,
"loss": 0.7232,
"step": 800
},
{
"epoch": 0.24467603081105574,
"grad_norm": 0.9551767616126388,
"learning_rate": 7.791277258566979e-05,
"loss": 0.6908,
"step": 810
},
{
"epoch": 0.24769672254946382,
"grad_norm": 0.8615297590153658,
"learning_rate": 7.760124610591901e-05,
"loss": 0.7184,
"step": 820
},
{
"epoch": 0.2507174142878719,
"grad_norm": 1.3152099264067771,
"learning_rate": 7.728971962616824e-05,
"loss": 0.7364,
"step": 830
},
{
"epoch": 0.25373810602628,
"grad_norm": 1.0635463723018044,
"learning_rate": 7.697819314641746e-05,
"loss": 0.7004,
"step": 840
},
{
"epoch": 0.25675879776468813,
"grad_norm": 0.9572383370970278,
"learning_rate": 7.666666666666667e-05,
"loss": 0.7464,
"step": 850
},
{
"epoch": 0.2597794895030962,
"grad_norm": 1.064383252851316,
"learning_rate": 7.635514018691589e-05,
"loss": 0.7277,
"step": 860
},
{
"epoch": 0.2628001812415043,
"grad_norm": 0.9049317589458534,
"learning_rate": 7.604361370716511e-05,
"loss": 0.6551,
"step": 870
},
{
"epoch": 0.2658208729799124,
"grad_norm": 0.9799744579764643,
"learning_rate": 7.573208722741434e-05,
"loss": 0.6875,
"step": 880
},
{
"epoch": 0.2688415647183205,
"grad_norm": 0.9279170964443864,
"learning_rate": 7.542056074766356e-05,
"loss": 0.7053,
"step": 890
},
{
"epoch": 0.27186225645672857,
"grad_norm": 0.8393064524929871,
"learning_rate": 7.510903426791278e-05,
"loss": 0.7562,
"step": 900
},
{
"epoch": 0.2748829481951367,
"grad_norm": 0.8473010959065636,
"learning_rate": 7.4797507788162e-05,
"loss": 0.6984,
"step": 910
},
{
"epoch": 0.2779036399335448,
"grad_norm": 0.9692282290480679,
"learning_rate": 7.448598130841123e-05,
"loss": 0.6755,
"step": 920
},
{
"epoch": 0.28092433167195285,
"grad_norm": 1.0247211218005006,
"learning_rate": 7.417445482866045e-05,
"loss": 0.7045,
"step": 930
},
{
"epoch": 0.283945023410361,
"grad_norm": 0.8017304920286634,
"learning_rate": 7.386292834890966e-05,
"loss": 0.6949,
"step": 940
},
{
"epoch": 0.28696571514876906,
"grad_norm": 0.9954668010151982,
"learning_rate": 7.355140186915888e-05,
"loss": 0.6987,
"step": 950
},
{
"epoch": 0.28998640688717714,
"grad_norm": 0.7614508342259325,
"learning_rate": 7.32398753894081e-05,
"loss": 0.7365,
"step": 960
},
{
"epoch": 0.29300709862558527,
"grad_norm": 1.0122965704452649,
"learning_rate": 7.292834890965733e-05,
"loss": 0.6671,
"step": 970
},
{
"epoch": 0.29602779036399335,
"grad_norm": 0.9599792773793446,
"learning_rate": 7.261682242990655e-05,
"loss": 0.6937,
"step": 980
},
{
"epoch": 0.2990484821024014,
"grad_norm": 0.8680572994187495,
"learning_rate": 7.230529595015577e-05,
"loss": 0.7078,
"step": 990
},
{
"epoch": 0.30206917384080956,
"grad_norm": 0.8834934535143681,
"learning_rate": 7.1993769470405e-05,
"loss": 0.6958,
"step": 1000
},
{
"epoch": 0.30508986557921763,
"grad_norm": 0.9243495158552708,
"learning_rate": 7.168224299065422e-05,
"loss": 0.6791,
"step": 1010
},
{
"epoch": 0.3081105573176257,
"grad_norm": 0.8955219617632852,
"learning_rate": 7.137071651090344e-05,
"loss": 0.7097,
"step": 1020
},
{
"epoch": 0.31113124905603384,
"grad_norm": 0.9931149366193008,
"learning_rate": 7.105919003115265e-05,
"loss": 0.6227,
"step": 1030
},
{
"epoch": 0.3141519407944419,
"grad_norm": 0.7569849851606368,
"learning_rate": 7.074766355140187e-05,
"loss": 0.7429,
"step": 1040
},
{
"epoch": 0.31717263253285,
"grad_norm": 0.771881005125769,
"learning_rate": 7.04361370716511e-05,
"loss": 0.6765,
"step": 1050
},
{
"epoch": 0.32019332427125813,
"grad_norm": 1.143381387938975,
"learning_rate": 7.012461059190032e-05,
"loss": 0.6813,
"step": 1060
},
{
"epoch": 0.3232140160096662,
"grad_norm": 0.8238947611161652,
"learning_rate": 6.981308411214954e-05,
"loss": 0.6717,
"step": 1070
},
{
"epoch": 0.3262347077480743,
"grad_norm": 0.8835494463241778,
"learning_rate": 6.950155763239876e-05,
"loss": 0.6936,
"step": 1080
},
{
"epoch": 0.3292553994864824,
"grad_norm": 0.9211582290817895,
"learning_rate": 6.919003115264799e-05,
"loss": 0.6479,
"step": 1090
},
{
"epoch": 0.3322760912248905,
"grad_norm": 0.8975205838470279,
"learning_rate": 6.887850467289721e-05,
"loss": 0.7252,
"step": 1100
},
{
"epoch": 0.3352967829632986,
"grad_norm": 1.012640948447073,
"learning_rate": 6.856697819314643e-05,
"loss": 0.7231,
"step": 1110
},
{
"epoch": 0.3383174747017067,
"grad_norm": 1.0723892527532837,
"learning_rate": 6.825545171339564e-05,
"loss": 0.66,
"step": 1120
},
{
"epoch": 0.3413381664401148,
"grad_norm": 0.9730571481063612,
"learning_rate": 6.794392523364486e-05,
"loss": 0.6449,
"step": 1130
},
{
"epoch": 0.3443588581785229,
"grad_norm": 1.067984338089875,
"learning_rate": 6.763239875389409e-05,
"loss": 0.7296,
"step": 1140
},
{
"epoch": 0.347379549916931,
"grad_norm": 0.9490599476113804,
"learning_rate": 6.732087227414331e-05,
"loss": 0.6509,
"step": 1150
},
{
"epoch": 0.35040024165533906,
"grad_norm": 1.0467534497439532,
"learning_rate": 6.700934579439253e-05,
"loss": 0.6699,
"step": 1160
},
{
"epoch": 0.3534209333937472,
"grad_norm": 1.025561971018954,
"learning_rate": 6.669781931464175e-05,
"loss": 0.6731,
"step": 1170
},
{
"epoch": 0.35644162513215527,
"grad_norm": 0.8559227700219806,
"learning_rate": 6.638629283489098e-05,
"loss": 0.7385,
"step": 1180
},
{
"epoch": 0.35946231687056335,
"grad_norm": 0.9811074108069637,
"learning_rate": 6.60747663551402e-05,
"loss": 0.7004,
"step": 1190
},
{
"epoch": 0.3624830086089715,
"grad_norm": 1.2629840514118404,
"learning_rate": 6.576323987538941e-05,
"loss": 0.7,
"step": 1200
},
{
"epoch": 0.36550370034737956,
"grad_norm": 0.8368147777189291,
"learning_rate": 6.545171339563863e-05,
"loss": 0.6681,
"step": 1210
},
{
"epoch": 0.36852439208578763,
"grad_norm": 0.9065786530560539,
"learning_rate": 6.514018691588785e-05,
"loss": 0.7471,
"step": 1220
},
{
"epoch": 0.37154508382419577,
"grad_norm": 0.845163729233175,
"learning_rate": 6.482866043613708e-05,
"loss": 0.6894,
"step": 1230
},
{
"epoch": 0.37456577556260384,
"grad_norm": 1.0596543107543936,
"learning_rate": 6.45171339563863e-05,
"loss": 0.6566,
"step": 1240
},
{
"epoch": 0.3775864673010119,
"grad_norm": 0.9732698632672376,
"learning_rate": 6.420560747663552e-05,
"loss": 0.7103,
"step": 1250
},
{
"epoch": 0.38060715903942005,
"grad_norm": 0.8505649011210936,
"learning_rate": 6.389408099688475e-05,
"loss": 0.6642,
"step": 1260
},
{
"epoch": 0.3836278507778281,
"grad_norm": 0.9630092097486271,
"learning_rate": 6.358255451713397e-05,
"loss": 0.6745,
"step": 1270
},
{
"epoch": 0.3866485425162362,
"grad_norm": 0.9557213652111145,
"learning_rate": 6.327102803738319e-05,
"loss": 0.6952,
"step": 1280
},
{
"epoch": 0.38966923425464434,
"grad_norm": 1.1176168484612952,
"learning_rate": 6.29595015576324e-05,
"loss": 0.7395,
"step": 1290
},
{
"epoch": 0.3926899259930524,
"grad_norm": 0.8571565860376751,
"learning_rate": 6.264797507788162e-05,
"loss": 0.7111,
"step": 1300
},
{
"epoch": 0.3957106177314605,
"grad_norm": 0.9134972865790056,
"learning_rate": 6.233644859813084e-05,
"loss": 0.6683,
"step": 1310
},
{
"epoch": 0.3987313094698686,
"grad_norm": 1.0992598415620052,
"learning_rate": 6.202492211838007e-05,
"loss": 0.643,
"step": 1320
},
{
"epoch": 0.4017520012082767,
"grad_norm": 0.8394762760205097,
"learning_rate": 6.171339563862929e-05,
"loss": 0.6573,
"step": 1330
},
{
"epoch": 0.4047726929466848,
"grad_norm": 0.9597186024683255,
"learning_rate": 6.140186915887851e-05,
"loss": 0.6533,
"step": 1340
},
{
"epoch": 0.4077933846850929,
"grad_norm": 1.0291805836678007,
"learning_rate": 6.109034267912774e-05,
"loss": 0.6914,
"step": 1350
},
{
"epoch": 0.410814076423501,
"grad_norm": 0.9098749247300835,
"learning_rate": 6.077881619937695e-05,
"loss": 0.7174,
"step": 1360
},
{
"epoch": 0.41383476816190906,
"grad_norm": 0.9748564539894223,
"learning_rate": 6.0467289719626174e-05,
"loss": 0.7098,
"step": 1370
},
{
"epoch": 0.4168554599003172,
"grad_norm": 0.8332710644006146,
"learning_rate": 6.01557632398754e-05,
"loss": 0.6571,
"step": 1380
},
{
"epoch": 0.41987615163872527,
"grad_norm": 1.034002380247954,
"learning_rate": 5.984423676012462e-05,
"loss": 0.6515,
"step": 1390
},
{
"epoch": 0.42289684337713335,
"grad_norm": 0.8765270890141124,
"learning_rate": 5.9532710280373836e-05,
"loss": 0.6687,
"step": 1400
},
{
"epoch": 0.4259175351155415,
"grad_norm": 0.8972730862256827,
"learning_rate": 5.922118380062306e-05,
"loss": 0.6566,
"step": 1410
},
{
"epoch": 0.42893822685394956,
"grad_norm": 0.990782741867072,
"learning_rate": 5.890965732087228e-05,
"loss": 0.6968,
"step": 1420
},
{
"epoch": 0.43195891859235763,
"grad_norm": 0.9710394039056895,
"learning_rate": 5.8598130841121504e-05,
"loss": 0.6864,
"step": 1430
},
{
"epoch": 0.43497961033076576,
"grad_norm": 0.9070606227395045,
"learning_rate": 5.828660436137072e-05,
"loss": 0.6248,
"step": 1440
},
{
"epoch": 0.43800030206917384,
"grad_norm": 0.9414728155949651,
"learning_rate": 5.797507788161994e-05,
"loss": 0.6698,
"step": 1450
},
{
"epoch": 0.4410209938075819,
"grad_norm": 0.7858430906023061,
"learning_rate": 5.7663551401869165e-05,
"loss": 0.6714,
"step": 1460
},
{
"epoch": 0.44404168554599005,
"grad_norm": 0.9086044428924012,
"learning_rate": 5.735202492211839e-05,
"loss": 0.7047,
"step": 1470
},
{
"epoch": 0.4470623772843981,
"grad_norm": 1.483392296727074,
"learning_rate": 5.704049844236761e-05,
"loss": 0.6729,
"step": 1480
},
{
"epoch": 0.4500830690228062,
"grad_norm": 1.2221155973651154,
"learning_rate": 5.6728971962616826e-05,
"loss": 0.6549,
"step": 1490
},
{
"epoch": 0.45310376076121434,
"grad_norm": 0.9537155935809545,
"learning_rate": 5.641744548286605e-05,
"loss": 0.6159,
"step": 1500
},
{
"epoch": 0.4561244524996224,
"grad_norm": 0.9622095662501553,
"learning_rate": 5.610591900311527e-05,
"loss": 0.6979,
"step": 1510
},
{
"epoch": 0.4591451442380305,
"grad_norm": 0.9758624698389782,
"learning_rate": 5.5794392523364494e-05,
"loss": 0.646,
"step": 1520
},
{
"epoch": 0.4621658359764386,
"grad_norm": 0.7544959077134319,
"learning_rate": 5.548286604361371e-05,
"loss": 0.6247,
"step": 1530
},
{
"epoch": 0.4651865277148467,
"grad_norm": 1.1533746560653724,
"learning_rate": 5.517133956386293e-05,
"loss": 0.6904,
"step": 1540
},
{
"epoch": 0.4682072194532548,
"grad_norm": 1.0820947508876602,
"learning_rate": 5.4859813084112156e-05,
"loss": 0.6553,
"step": 1550
},
{
"epoch": 0.4712279111916629,
"grad_norm": 0.9732497051241004,
"learning_rate": 5.454828660436138e-05,
"loss": 0.641,
"step": 1560
},
{
"epoch": 0.474248602930071,
"grad_norm": 1.0002172762804926,
"learning_rate": 5.4236760124610594e-05,
"loss": 0.6021,
"step": 1570
},
{
"epoch": 0.47726929466847906,
"grad_norm": 1.0211904068150215,
"learning_rate": 5.392523364485982e-05,
"loss": 0.7062,
"step": 1580
},
{
"epoch": 0.4802899864068872,
"grad_norm": 0.9215201787115269,
"learning_rate": 5.361370716510904e-05,
"loss": 0.6423,
"step": 1590
},
{
"epoch": 0.48331067814529527,
"grad_norm": 1.0853776893432,
"learning_rate": 5.330218068535826e-05,
"loss": 0.7374,
"step": 1600
},
{
"epoch": 0.48633136988370335,
"grad_norm": 1.111343817288821,
"learning_rate": 5.2990654205607485e-05,
"loss": 0.6563,
"step": 1610
},
{
"epoch": 0.4893520616221115,
"grad_norm": 1.0514243279918507,
"learning_rate": 5.26791277258567e-05,
"loss": 0.6827,
"step": 1620
},
{
"epoch": 0.49237275336051955,
"grad_norm": 1.0619541347822374,
"learning_rate": 5.2367601246105924e-05,
"loss": 0.6242,
"step": 1630
},
{
"epoch": 0.49539344509892763,
"grad_norm": 0.9097722349251828,
"learning_rate": 5.2056074766355146e-05,
"loss": 0.615,
"step": 1640
},
{
"epoch": 0.49841413683733576,
"grad_norm": 0.9219734475697886,
"learning_rate": 5.174454828660437e-05,
"loss": 0.662,
"step": 1650
},
{
"epoch": 0.5014348285757438,
"grad_norm": 0.9690364090647586,
"learning_rate": 5.1433021806853585e-05,
"loss": 0.6632,
"step": 1660
},
{
"epoch": 0.504455520314152,
"grad_norm": 0.870971102845472,
"learning_rate": 5.112149532710281e-05,
"loss": 0.6061,
"step": 1670
},
{
"epoch": 0.50747621205256,
"grad_norm": 0.9160879376768967,
"learning_rate": 5.080996884735203e-05,
"loss": 0.6727,
"step": 1680
},
{
"epoch": 0.5104969037909681,
"grad_norm": 0.8236516373917104,
"learning_rate": 5.049844236760125e-05,
"loss": 0.679,
"step": 1690
},
{
"epoch": 0.5135175955293763,
"grad_norm": 1.1544029383623384,
"learning_rate": 5.018691588785047e-05,
"loss": 0.6687,
"step": 1700
},
{
"epoch": 0.5165382872677843,
"grad_norm": 0.9415259266877539,
"learning_rate": 4.987538940809969e-05,
"loss": 0.6988,
"step": 1710
},
{
"epoch": 0.5195589790061924,
"grad_norm": 1.005313152550806,
"learning_rate": 4.9563862928348914e-05,
"loss": 0.6871,
"step": 1720
},
{
"epoch": 0.5225796707446005,
"grad_norm": 0.8721485622772522,
"learning_rate": 4.925233644859814e-05,
"loss": 0.7242,
"step": 1730
},
{
"epoch": 0.5256003624830086,
"grad_norm": 0.9585413211074175,
"learning_rate": 4.894080996884736e-05,
"loss": 0.6469,
"step": 1740
},
{
"epoch": 0.5286210542214167,
"grad_norm": 1.19585806105442,
"learning_rate": 4.8629283489096575e-05,
"loss": 0.6231,
"step": 1750
},
{
"epoch": 0.5316417459598248,
"grad_norm": 1.1101720657226213,
"learning_rate": 4.83177570093458e-05,
"loss": 0.6153,
"step": 1760
},
{
"epoch": 0.5346624376982329,
"grad_norm": 1.01126944002228,
"learning_rate": 4.800623052959502e-05,
"loss": 0.6766,
"step": 1770
},
{
"epoch": 0.537683129436641,
"grad_norm": 0.9755471413606605,
"learning_rate": 4.7694704049844244e-05,
"loss": 0.5983,
"step": 1780
},
{
"epoch": 0.5407038211750491,
"grad_norm": 1.0374363204506591,
"learning_rate": 4.738317757009346e-05,
"loss": 0.582,
"step": 1790
},
{
"epoch": 0.5437245129134571,
"grad_norm": 1.0114337393595922,
"learning_rate": 4.707165109034268e-05,
"loss": 0.6469,
"step": 1800
},
{
"epoch": 0.5467452046518653,
"grad_norm": 0.7718847238435554,
"learning_rate": 4.6760124610591905e-05,
"loss": 0.6209,
"step": 1810
},
{
"epoch": 0.5497658963902734,
"grad_norm": 0.9625054212371464,
"learning_rate": 4.644859813084113e-05,
"loss": 0.6366,
"step": 1820
},
{
"epoch": 0.5527865881286814,
"grad_norm": 1.0349710792742046,
"learning_rate": 4.6137071651090343e-05,
"loss": 0.5803,
"step": 1830
},
{
"epoch": 0.5558072798670896,
"grad_norm": 1.1320954110896704,
"learning_rate": 4.5825545171339566e-05,
"loss": 0.6988,
"step": 1840
},
{
"epoch": 0.5588279716054977,
"grad_norm": 1.0220144920125607,
"learning_rate": 4.551401869158879e-05,
"loss": 0.6072,
"step": 1850
},
{
"epoch": 0.5618486633439057,
"grad_norm": 1.0791799785815506,
"learning_rate": 4.520249221183801e-05,
"loss": 0.6198,
"step": 1860
},
{
"epoch": 0.5648693550823138,
"grad_norm": 0.9841150881681497,
"learning_rate": 4.4890965732087234e-05,
"loss": 0.662,
"step": 1870
},
{
"epoch": 0.567890046820722,
"grad_norm": 1.0018821293811744,
"learning_rate": 4.457943925233645e-05,
"loss": 0.6349,
"step": 1880
},
{
"epoch": 0.57091073855913,
"grad_norm": 0.913754158406669,
"learning_rate": 4.426791277258567e-05,
"loss": 0.6809,
"step": 1890
},
{
"epoch": 0.5739314302975381,
"grad_norm": 0.7961432611640711,
"learning_rate": 4.3956386292834895e-05,
"loss": 0.6104,
"step": 1900
},
{
"epoch": 0.5769521220359463,
"grad_norm": 0.9894651687363102,
"learning_rate": 4.364485981308412e-05,
"loss": 0.5729,
"step": 1910
},
{
"epoch": 0.5799728137743543,
"grad_norm": 0.8147185677420008,
"learning_rate": 4.3333333333333334e-05,
"loss": 0.6087,
"step": 1920
},
{
"epoch": 0.5829935055127624,
"grad_norm": 0.9783085144604581,
"learning_rate": 4.302180685358256e-05,
"loss": 0.6262,
"step": 1930
},
{
"epoch": 0.5860141972511705,
"grad_norm": 0.7784275704986368,
"learning_rate": 4.271028037383178e-05,
"loss": 0.6636,
"step": 1940
},
{
"epoch": 0.5890348889895786,
"grad_norm": 0.9310267092458656,
"learning_rate": 4.2398753894081e-05,
"loss": 0.6632,
"step": 1950
},
{
"epoch": 0.5920555807279867,
"grad_norm": 1.0903553721195471,
"learning_rate": 4.2087227414330225e-05,
"loss": 0.6353,
"step": 1960
},
{
"epoch": 0.5950762724663948,
"grad_norm": 1.128799949747586,
"learning_rate": 4.177570093457944e-05,
"loss": 0.6573,
"step": 1970
},
{
"epoch": 0.5980969642048029,
"grad_norm": 0.8610297939057725,
"learning_rate": 4.1464174454828663e-05,
"loss": 0.6328,
"step": 1980
},
{
"epoch": 0.601117655943211,
"grad_norm": 0.9604742989982925,
"learning_rate": 4.1152647975077886e-05,
"loss": 0.6519,
"step": 1990
},
{
"epoch": 0.6041383476816191,
"grad_norm": 0.9402231206039308,
"learning_rate": 4.084112149532711e-05,
"loss": 0.6516,
"step": 2000
},
{
"epoch": 0.6071590394200271,
"grad_norm": 0.9545250774883017,
"learning_rate": 4.0529595015576325e-05,
"loss": 0.5866,
"step": 2010
},
{
"epoch": 0.6101797311584353,
"grad_norm": 1.0796127712688048,
"learning_rate": 4.021806853582555e-05,
"loss": 0.7043,
"step": 2020
},
{
"epoch": 0.6132004228968434,
"grad_norm": 1.1333119978393902,
"learning_rate": 3.990654205607477e-05,
"loss": 0.674,
"step": 2030
},
{
"epoch": 0.6162211146352514,
"grad_norm": 1.029611011718845,
"learning_rate": 3.959501557632399e-05,
"loss": 0.6823,
"step": 2040
},
{
"epoch": 0.6192418063736596,
"grad_norm": 0.8866452647064959,
"learning_rate": 3.928348909657321e-05,
"loss": 0.6755,
"step": 2050
},
{
"epoch": 0.6222624981120677,
"grad_norm": 0.9810289783702174,
"learning_rate": 3.897196261682243e-05,
"loss": 0.6846,
"step": 2060
},
{
"epoch": 0.6252831898504757,
"grad_norm": 0.9655687966209278,
"learning_rate": 3.8660436137071654e-05,
"loss": 0.642,
"step": 2070
},
{
"epoch": 0.6283038815888838,
"grad_norm": 0.859596529702297,
"learning_rate": 3.834890965732088e-05,
"loss": 0.5939,
"step": 2080
},
{
"epoch": 0.631324573327292,
"grad_norm": 0.921220108830949,
"learning_rate": 3.80373831775701e-05,
"loss": 0.6443,
"step": 2090
},
{
"epoch": 0.6343452650657,
"grad_norm": 0.9645040704285783,
"learning_rate": 3.7725856697819315e-05,
"loss": 0.6755,
"step": 2100
},
{
"epoch": 0.6373659568041081,
"grad_norm": 0.8113343875987217,
"learning_rate": 3.741433021806854e-05,
"loss": 0.6873,
"step": 2110
},
{
"epoch": 0.6403866485425163,
"grad_norm": 0.9875722670281138,
"learning_rate": 3.710280373831776e-05,
"loss": 0.6757,
"step": 2120
},
{
"epoch": 0.6434073402809243,
"grad_norm": 0.9212045663948729,
"learning_rate": 3.6791277258566983e-05,
"loss": 0.6152,
"step": 2130
},
{
"epoch": 0.6464280320193324,
"grad_norm": 1.1795807247471173,
"learning_rate": 3.64797507788162e-05,
"loss": 0.6206,
"step": 2140
},
{
"epoch": 0.6494487237577405,
"grad_norm": 1.02016484907723,
"learning_rate": 3.616822429906542e-05,
"loss": 0.6145,
"step": 2150
},
{
"epoch": 0.6524694154961486,
"grad_norm": 1.0910288266708823,
"learning_rate": 3.5856697819314645e-05,
"loss": 0.6827,
"step": 2160
},
{
"epoch": 0.6554901072345567,
"grad_norm": 0.9260898097849999,
"learning_rate": 3.554517133956387e-05,
"loss": 0.6265,
"step": 2170
},
{
"epoch": 0.6585107989729648,
"grad_norm": 0.8033545643142217,
"learning_rate": 3.523364485981308e-05,
"loss": 0.5972,
"step": 2180
},
{
"epoch": 0.6615314907113728,
"grad_norm": 1.0236521911766159,
"learning_rate": 3.4922118380062306e-05,
"loss": 0.6367,
"step": 2190
},
{
"epoch": 0.664552182449781,
"grad_norm": 0.9203341940281535,
"learning_rate": 3.461059190031153e-05,
"loss": 0.6529,
"step": 2200
},
{
"epoch": 0.6675728741881891,
"grad_norm": 0.8978799856543846,
"learning_rate": 3.429906542056075e-05,
"loss": 0.6297,
"step": 2210
},
{
"epoch": 0.6705935659265972,
"grad_norm": 0.7847116258760808,
"learning_rate": 3.3987538940809974e-05,
"loss": 0.6372,
"step": 2220
},
{
"epoch": 0.6736142576650053,
"grad_norm": 0.8284178797894528,
"learning_rate": 3.367601246105919e-05,
"loss": 0.6496,
"step": 2230
},
{
"epoch": 0.6766349494034134,
"grad_norm": 0.9028088991532255,
"learning_rate": 3.336448598130841e-05,
"loss": 0.612,
"step": 2240
},
{
"epoch": 0.6796556411418215,
"grad_norm": 0.9644339999971415,
"learning_rate": 3.3052959501557635e-05,
"loss": 0.6577,
"step": 2250
},
{
"epoch": 0.6826763328802296,
"grad_norm": 0.8435903838116136,
"learning_rate": 3.274143302180686e-05,
"loss": 0.6565,
"step": 2260
},
{
"epoch": 0.6856970246186377,
"grad_norm": 1.0057414649400669,
"learning_rate": 3.2429906542056074e-05,
"loss": 0.6863,
"step": 2270
},
{
"epoch": 0.6887177163570458,
"grad_norm": 1.0282432957302625,
"learning_rate": 3.21183800623053e-05,
"loss": 0.6397,
"step": 2280
},
{
"epoch": 0.6917384080954538,
"grad_norm": 0.9733238446682813,
"learning_rate": 3.180685358255452e-05,
"loss": 0.6157,
"step": 2290
},
{
"epoch": 0.694759099833862,
"grad_norm": 1.107461230355507,
"learning_rate": 3.149532710280374e-05,
"loss": 0.6401,
"step": 2300
},
{
"epoch": 0.6977797915722701,
"grad_norm": 1.1302701977875342,
"learning_rate": 3.118380062305296e-05,
"loss": 0.6478,
"step": 2310
},
{
"epoch": 0.7008004833106781,
"grad_norm": 0.8727697902704742,
"learning_rate": 3.087227414330218e-05,
"loss": 0.6183,
"step": 2320
},
{
"epoch": 0.7038211750490863,
"grad_norm": 1.0811023575925298,
"learning_rate": 3.05607476635514e-05,
"loss": 0.6678,
"step": 2330
},
{
"epoch": 0.7068418667874944,
"grad_norm": 0.910662174723694,
"learning_rate": 3.0249221183800626e-05,
"loss": 0.6134,
"step": 2340
},
{
"epoch": 0.7098625585259024,
"grad_norm": 0.8701473143660866,
"learning_rate": 2.9937694704049845e-05,
"loss": 0.684,
"step": 2350
},
{
"epoch": 0.7128832502643105,
"grad_norm": 0.9502715450958549,
"learning_rate": 2.9626168224299068e-05,
"loss": 0.6117,
"step": 2360
},
{
"epoch": 0.7159039420027187,
"grad_norm": 0.9326968943217765,
"learning_rate": 2.9314641744548287e-05,
"loss": 0.6553,
"step": 2370
},
{
"epoch": 0.7189246337411267,
"grad_norm": 1.085479042571671,
"learning_rate": 2.900311526479751e-05,
"loss": 0.6746,
"step": 2380
},
{
"epoch": 0.7219453254795348,
"grad_norm": 0.8618276796564664,
"learning_rate": 2.869158878504673e-05,
"loss": 0.6343,
"step": 2390
},
{
"epoch": 0.724966017217943,
"grad_norm": 1.0814238125070605,
"learning_rate": 2.8380062305295952e-05,
"loss": 0.6238,
"step": 2400
},
{
"epoch": 0.727986708956351,
"grad_norm": 0.9771905961761014,
"learning_rate": 2.806853582554517e-05,
"loss": 0.5864,
"step": 2410
},
{
"epoch": 0.7310074006947591,
"grad_norm": 0.9767500457929595,
"learning_rate": 2.7757009345794394e-05,
"loss": 0.6386,
"step": 2420
},
{
"epoch": 0.7340280924331672,
"grad_norm": 1.123445002483322,
"learning_rate": 2.7445482866043613e-05,
"loss": 0.6035,
"step": 2430
},
{
"epoch": 0.7370487841715753,
"grad_norm": 1.0442903344618233,
"learning_rate": 2.7133956386292836e-05,
"loss": 0.6466,
"step": 2440
},
{
"epoch": 0.7400694759099834,
"grad_norm": 1.1309919373809965,
"learning_rate": 2.682242990654206e-05,
"loss": 0.6004,
"step": 2450
},
{
"epoch": 0.7430901676483915,
"grad_norm": 1.0080726531395334,
"learning_rate": 2.6510903426791278e-05,
"loss": 0.6576,
"step": 2460
},
{
"epoch": 0.7461108593867996,
"grad_norm": 0.9924935740259064,
"learning_rate": 2.61993769470405e-05,
"loss": 0.6122,
"step": 2470
},
{
"epoch": 0.7491315511252077,
"grad_norm": 1.0768441001403686,
"learning_rate": 2.588785046728972e-05,
"loss": 0.6349,
"step": 2480
},
{
"epoch": 0.7521522428636158,
"grad_norm": 1.0720344843351275,
"learning_rate": 2.5576323987538943e-05,
"loss": 0.684,
"step": 2490
},
{
"epoch": 0.7551729346020238,
"grad_norm": 0.9812143360946806,
"learning_rate": 2.5264797507788162e-05,
"loss": 0.6326,
"step": 2500
},
{
"epoch": 0.758193626340432,
"grad_norm": 0.8459161365628051,
"learning_rate": 2.4953271028037385e-05,
"loss": 0.6498,
"step": 2510
},
{
"epoch": 0.7612143180788401,
"grad_norm": 1.062011499231393,
"learning_rate": 2.4641744548286604e-05,
"loss": 0.5903,
"step": 2520
},
{
"epoch": 0.7642350098172481,
"grad_norm": 1.2676346445894033,
"learning_rate": 2.4330218068535827e-05,
"loss": 0.6346,
"step": 2530
},
{
"epoch": 0.7672557015556563,
"grad_norm": 0.9982039436166976,
"learning_rate": 2.4018691588785046e-05,
"loss": 0.6086,
"step": 2540
},
{
"epoch": 0.7702763932940644,
"grad_norm": 1.1218149129480335,
"learning_rate": 2.370716510903427e-05,
"loss": 0.5995,
"step": 2550
},
{
"epoch": 0.7732970850324724,
"grad_norm": 1.119893392548989,
"learning_rate": 2.3395638629283488e-05,
"loss": 0.6021,
"step": 2560
},
{
"epoch": 0.7763177767708805,
"grad_norm": 0.98486579720627,
"learning_rate": 2.308411214953271e-05,
"loss": 0.6024,
"step": 2570
},
{
"epoch": 0.7793384685092887,
"grad_norm": 0.8364224095096793,
"learning_rate": 2.2772585669781933e-05,
"loss": 0.6449,
"step": 2580
},
{
"epoch": 0.7823591602476967,
"grad_norm": 0.9213982322355779,
"learning_rate": 2.2461059190031153e-05,
"loss": 0.6273,
"step": 2590
},
{
"epoch": 0.7853798519861048,
"grad_norm": 1.0159207681918687,
"learning_rate": 2.2149532710280375e-05,
"loss": 0.6399,
"step": 2600
},
{
"epoch": 0.788400543724513,
"grad_norm": 1.145425432922919,
"learning_rate": 2.1838006230529595e-05,
"loss": 0.6163,
"step": 2610
},
{
"epoch": 0.791421235462921,
"grad_norm": 0.9322134975256001,
"learning_rate": 2.1526479750778817e-05,
"loss": 0.6677,
"step": 2620
},
{
"epoch": 0.7944419272013291,
"grad_norm": 0.935489630914845,
"learning_rate": 2.1214953271028037e-05,
"loss": 0.6627,
"step": 2630
},
{
"epoch": 0.7974626189397372,
"grad_norm": 0.9480142924396203,
"learning_rate": 2.090342679127726e-05,
"loss": 0.614,
"step": 2640
},
{
"epoch": 0.8004833106781453,
"grad_norm": 1.0419224396210682,
"learning_rate": 2.059190031152648e-05,
"loss": 0.6425,
"step": 2650
},
{
"epoch": 0.8035040024165534,
"grad_norm": 1.1997744692346006,
"learning_rate": 2.02803738317757e-05,
"loss": 0.6496,
"step": 2660
},
{
"epoch": 0.8065246941549615,
"grad_norm": 1.0212167604964282,
"learning_rate": 1.996884735202492e-05,
"loss": 0.6233,
"step": 2670
},
{
"epoch": 0.8095453858933696,
"grad_norm": 0.8681772917726166,
"learning_rate": 1.9657320872274143e-05,
"loss": 0.6538,
"step": 2680
},
{
"epoch": 0.8125660776317777,
"grad_norm": 1.091843651657025,
"learning_rate": 1.9345794392523366e-05,
"loss": 0.6272,
"step": 2690
},
{
"epoch": 0.8155867693701858,
"grad_norm": 1.2768445724046573,
"learning_rate": 1.9034267912772585e-05,
"loss": 0.6181,
"step": 2700
},
{
"epoch": 0.8186074611085938,
"grad_norm": 0.9166319020452763,
"learning_rate": 1.8722741433021808e-05,
"loss": 0.6557,
"step": 2710
},
{
"epoch": 0.821628152847002,
"grad_norm": 0.8050333909093216,
"learning_rate": 1.8411214953271027e-05,
"loss": 0.594,
"step": 2720
},
{
"epoch": 0.8246488445854101,
"grad_norm": 1.0875099162903203,
"learning_rate": 1.809968847352025e-05,
"loss": 0.6467,
"step": 2730
},
{
"epoch": 0.8276695363238181,
"grad_norm": 1.084354704975054,
"learning_rate": 1.778816199376947e-05,
"loss": 0.6525,
"step": 2740
},
{
"epoch": 0.8306902280622263,
"grad_norm": 0.8160931529433489,
"learning_rate": 1.7476635514018695e-05,
"loss": 0.5978,
"step": 2750
},
{
"epoch": 0.8337109198006344,
"grad_norm": 0.838873047824145,
"learning_rate": 1.7165109034267915e-05,
"loss": 0.5907,
"step": 2760
},
{
"epoch": 0.8367316115390424,
"grad_norm": 1.0454181565883631,
"learning_rate": 1.6853582554517137e-05,
"loss": 0.5869,
"step": 2770
},
{
"epoch": 0.8397523032774505,
"grad_norm": 1.047934242393408,
"learning_rate": 1.6542056074766357e-05,
"loss": 0.6021,
"step": 2780
},
{
"epoch": 0.8427729950158587,
"grad_norm": 0.9172506322417012,
"learning_rate": 1.623052959501558e-05,
"loss": 0.6181,
"step": 2790
},
{
"epoch": 0.8457936867542667,
"grad_norm": 1.0342214473578908,
"learning_rate": 1.59190031152648e-05,
"loss": 0.6476,
"step": 2800
},
{
"epoch": 0.8488143784926748,
"grad_norm": 1.0576515362756844,
"learning_rate": 1.560747663551402e-05,
"loss": 0.6443,
"step": 2810
},
{
"epoch": 0.851835070231083,
"grad_norm": 0.6692869313084945,
"learning_rate": 1.529595015576324e-05,
"loss": 0.6229,
"step": 2820
},
{
"epoch": 0.854855761969491,
"grad_norm": 0.9210101682143316,
"learning_rate": 1.4984423676012463e-05,
"loss": 0.6246,
"step": 2830
},
{
"epoch": 0.8578764537078991,
"grad_norm": 1.156766881429858,
"learning_rate": 1.4672897196261684e-05,
"loss": 0.6044,
"step": 2840
},
{
"epoch": 0.8608971454463072,
"grad_norm": 0.9789512390026106,
"learning_rate": 1.4361370716510905e-05,
"loss": 0.6389,
"step": 2850
},
{
"epoch": 0.8639178371847153,
"grad_norm": 1.191949240447118,
"learning_rate": 1.4049844236760126e-05,
"loss": 0.6602,
"step": 2860
},
{
"epoch": 0.8669385289231234,
"grad_norm": 0.9708955037564521,
"learning_rate": 1.3738317757009347e-05,
"loss": 0.6651,
"step": 2870
},
{
"epoch": 0.8699592206615315,
"grad_norm": 0.8288865619618823,
"learning_rate": 1.3426791277258568e-05,
"loss": 0.599,
"step": 2880
},
{
"epoch": 0.8729799123999396,
"grad_norm": 0.9521891563981253,
"learning_rate": 1.311526479750779e-05,
"loss": 0.6146,
"step": 2890
},
{
"epoch": 0.8760006041383477,
"grad_norm": 0.8630474026750076,
"learning_rate": 1.280373831775701e-05,
"loss": 0.5642,
"step": 2900
},
{
"epoch": 0.8790212958767558,
"grad_norm": 1.049571278684741,
"learning_rate": 1.2492211838006231e-05,
"loss": 0.6267,
"step": 2910
},
{
"epoch": 0.8820419876151638,
"grad_norm": 1.0590797502360565,
"learning_rate": 1.2180685358255452e-05,
"loss": 0.5879,
"step": 2920
},
{
"epoch": 0.885062679353572,
"grad_norm": 1.0239937477648127,
"learning_rate": 1.1869158878504673e-05,
"loss": 0.6303,
"step": 2930
},
{
"epoch": 0.8880833710919801,
"grad_norm": 1.0807935195579166,
"learning_rate": 1.1557632398753894e-05,
"loss": 0.6165,
"step": 2940
},
{
"epoch": 0.8911040628303881,
"grad_norm": 1.2104250578230216,
"learning_rate": 1.1246105919003117e-05,
"loss": 0.6377,
"step": 2950
},
{
"epoch": 0.8941247545687963,
"grad_norm": 0.9833051516741547,
"learning_rate": 1.0934579439252338e-05,
"loss": 0.6356,
"step": 2960
},
{
"epoch": 0.8971454463072044,
"grad_norm": 0.9195910071930194,
"learning_rate": 1.0623052959501559e-05,
"loss": 0.5927,
"step": 2970
},
{
"epoch": 0.9001661380456124,
"grad_norm": 0.9508331555941772,
"learning_rate": 1.031152647975078e-05,
"loss": 0.6539,
"step": 2980
},
{
"epoch": 0.9031868297840205,
"grad_norm": 1.0172540761990956,
"learning_rate": 1e-05,
"loss": 0.6099,
"step": 2990
},
{
"epoch": 0.9062075215224287,
"grad_norm": 0.8454569627496504,
"learning_rate": 9.688473520249222e-06,
"loss": 0.6446,
"step": 3000
},
{
"epoch": 0.9092282132608367,
"grad_norm": 1.0229837834556614,
"learning_rate": 9.376947040498443e-06,
"loss": 0.5646,
"step": 3010
},
{
"epoch": 0.9122489049992448,
"grad_norm": 0.8108694312533385,
"learning_rate": 9.065420560747664e-06,
"loss": 0.6081,
"step": 3020
},
{
"epoch": 0.915269596737653,
"grad_norm": 1.122597148836237,
"learning_rate": 8.753894080996885e-06,
"loss": 0.5925,
"step": 3030
},
{
"epoch": 0.918290288476061,
"grad_norm": 0.8644525693487572,
"learning_rate": 8.442367601246107e-06,
"loss": 0.6021,
"step": 3040
},
{
"epoch": 0.9213109802144691,
"grad_norm": 0.9380249510498911,
"learning_rate": 8.130841121495328e-06,
"loss": 0.5896,
"step": 3050
},
{
"epoch": 0.9243316719528772,
"grad_norm": 0.9316208234480511,
"learning_rate": 7.81931464174455e-06,
"loss": 0.6168,
"step": 3060
},
{
"epoch": 0.9273523636912853,
"grad_norm": 0.8978678144978878,
"learning_rate": 7.5077881619937705e-06,
"loss": 0.5979,
"step": 3070
},
{
"epoch": 0.9303730554296934,
"grad_norm": 1.016592081776955,
"learning_rate": 7.1962616822429915e-06,
"loss": 0.6034,
"step": 3080
},
{
"epoch": 0.9333937471681015,
"grad_norm": 0.8642654535948303,
"learning_rate": 6.8847352024922125e-06,
"loss": 0.6082,
"step": 3090
},
{
"epoch": 0.9364144389065095,
"grad_norm": 1.02041087422202,
"learning_rate": 6.5732087227414335e-06,
"loss": 0.601,
"step": 3100
},
{
"epoch": 0.9394351306449177,
"grad_norm": 0.8116761149764674,
"learning_rate": 6.2616822429906544e-06,
"loss": 0.619,
"step": 3110
},
{
"epoch": 0.9424558223833258,
"grad_norm": 0.7974672131278376,
"learning_rate": 5.9501557632398754e-06,
"loss": 0.5918,
"step": 3120
},
{
"epoch": 0.9454765141217338,
"grad_norm": 1.0260273583236317,
"learning_rate": 5.638629283489097e-06,
"loss": 0.624,
"step": 3130
},
{
"epoch": 0.948497205860142,
"grad_norm": 1.0810678454481608,
"learning_rate": 5.327102803738318e-06,
"loss": 0.6249,
"step": 3140
},
{
"epoch": 0.9515178975985501,
"grad_norm": 0.9410480605089832,
"learning_rate": 5.015576323987539e-06,
"loss": 0.615,
"step": 3150
},
{
"epoch": 0.9545385893369581,
"grad_norm": 0.8663290980844977,
"learning_rate": 4.70404984423676e-06,
"loss": 0.6023,
"step": 3160
},
{
"epoch": 0.9575592810753663,
"grad_norm": 0.8857325255058867,
"learning_rate": 4.392523364485981e-06,
"loss": 0.6022,
"step": 3170
},
{
"epoch": 0.9605799728137744,
"grad_norm": 0.9687226928742629,
"learning_rate": 4.080996884735202e-06,
"loss": 0.6729,
"step": 3180
},
{
"epoch": 0.9636006645521824,
"grad_norm": 1.0927550182300916,
"learning_rate": 3.7694704049844237e-06,
"loss": 0.6365,
"step": 3190
},
{
"epoch": 0.9666213562905905,
"grad_norm": 0.8978647497382573,
"learning_rate": 3.4579439252336447e-06,
"loss": 0.601,
"step": 3200
},
{
"epoch": 0.9696420480289987,
"grad_norm": 0.9487851003875312,
"learning_rate": 3.146417445482866e-06,
"loss": 0.6064,
"step": 3210
},
{
"epoch": 0.9726627397674067,
"grad_norm": 0.9251513877632064,
"learning_rate": 2.8348909657320875e-06,
"loss": 0.5759,
"step": 3220
},
{
"epoch": 0.9756834315058148,
"grad_norm": 1.0597043780475088,
"learning_rate": 2.5233644859813085e-06,
"loss": 0.6105,
"step": 3230
},
{
"epoch": 0.978704123244223,
"grad_norm": 0.8426038200196306,
"learning_rate": 2.21183800623053e-06,
"loss": 0.5598,
"step": 3240
},
{
"epoch": 0.981724814982631,
"grad_norm": 0.9047413406086182,
"learning_rate": 1.900311526479751e-06,
"loss": 0.5793,
"step": 3250
},
{
"epoch": 0.9847455067210391,
"grad_norm": 1.0374829901076585,
"learning_rate": 1.588785046728972e-06,
"loss": 0.5269,
"step": 3260
},
{
"epoch": 0.9877661984594472,
"grad_norm": 0.8339586589651915,
"learning_rate": 1.2772585669781931e-06,
"loss": 0.6187,
"step": 3270
},
{
"epoch": 0.9907868901978553,
"grad_norm": 0.9187545091121259,
"learning_rate": 9.657320872274145e-07,
"loss": 0.6223,
"step": 3280
},
{
"epoch": 0.9938075819362634,
"grad_norm": 0.8684873265208356,
"learning_rate": 6.542056074766356e-07,
"loss": 0.5936,
"step": 3290
},
{
"epoch": 0.9968282736746715,
"grad_norm": 0.9952715787699572,
"learning_rate": 3.4267912772585674e-07,
"loss": 0.5693,
"step": 3300
},
{
"epoch": 0.9998489654130795,
"grad_norm": 0.9894526942742425,
"learning_rate": 3.1152647975077883e-08,
"loss": 0.655,
"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
}