Files
OmniGPT-355M/last-checkpoint/trainer_state.json
ModelHub XC e0d4e3e57e 初始化项目,由ModelHub XC社区提供模型
Model: OnurDemircioglu/OmniGPT-355M
Source: Original Platform
2026-09-06 16:33:31 +08:00

1722 lines
44 KiB
JSON

{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 4818,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.00020757115798759762,
"grad_norm": 1.9747060537338257,
"learning_rate": 5e-05,
"loss": 3.7501306533813477,
"step": 1
},
{
"epoch": 0.0041514231597519526,
"grad_norm": 1.4219956398010254,
"learning_rate": 4.980282274802823e-05,
"loss": 4.4793243408203125,
"step": 20
},
{
"epoch": 0.008302846319503905,
"grad_norm": 2.1015241146087646,
"learning_rate": 4.9595267745952676e-05,
"loss": 4.568800354003907,
"step": 40
},
{
"epoch": 0.012454269479255858,
"grad_norm": 1.8942627906799316,
"learning_rate": 4.938771274387713e-05,
"loss": 4.616829299926758,
"step": 60
},
{
"epoch": 0.01660569263900781,
"grad_norm": 1.8736380338668823,
"learning_rate": 4.918015774180158e-05,
"loss": 4.5886375427246096,
"step": 80
},
{
"epoch": 0.020757115798759763,
"grad_norm": 2.0199742317199707,
"learning_rate": 4.8972602739726034e-05,
"loss": 4.620360565185547,
"step": 100
},
{
"epoch": 0.024908538958511715,
"grad_norm": 2.1432764530181885,
"learning_rate": 4.876504773765048e-05,
"loss": 4.6068885803222654,
"step": 120
},
{
"epoch": 0.029059962118263668,
"grad_norm": 1.762105107307434,
"learning_rate": 4.8557492735574926e-05,
"loss": 4.584332275390625,
"step": 140
},
{
"epoch": 0.03321138527801562,
"grad_norm": 1.89723539352417,
"learning_rate": 4.834993773349938e-05,
"loss": 4.604080581665039,
"step": 160
},
{
"epoch": 0.03736280843776757,
"grad_norm": 2.2306056022644043,
"learning_rate": 4.814238273142383e-05,
"loss": 4.588064575195313,
"step": 180
},
{
"epoch": 0.041514231597519526,
"grad_norm": 1.7941607236862183,
"learning_rate": 4.793482772934828e-05,
"loss": 4.564043045043945,
"step": 200
},
{
"epoch": 0.045665654757271475,
"grad_norm": 2.362287759780884,
"learning_rate": 4.772727272727273e-05,
"loss": 4.653321075439453,
"step": 220
},
{
"epoch": 0.04981707791702343,
"grad_norm": 1.7642165422439575,
"learning_rate": 4.7519717725197176e-05,
"loss": 4.614533233642578,
"step": 240
},
{
"epoch": 0.05396850107677538,
"grad_norm": 1.7014025449752808,
"learning_rate": 4.731216272312163e-05,
"loss": 4.562130355834961,
"step": 260
},
{
"epoch": 0.058119924236527336,
"grad_norm": 1.8154466152191162,
"learning_rate": 4.710460772104608e-05,
"loss": 4.517855453491211,
"step": 280
},
{
"epoch": 0.062271347396279285,
"grad_norm": 1.67521071434021,
"learning_rate": 4.689705271897053e-05,
"loss": 4.498160171508789,
"step": 300
},
{
"epoch": 0.06642277055603124,
"grad_norm": 2.6233015060424805,
"learning_rate": 4.668949771689498e-05,
"loss": 4.5643047332763675,
"step": 320
},
{
"epoch": 0.07057419371578319,
"grad_norm": 1.7439541816711426,
"learning_rate": 4.6481942714819433e-05,
"loss": 4.517733764648438,
"step": 340
},
{
"epoch": 0.07472561687553514,
"grad_norm": 1.8184090852737427,
"learning_rate": 4.627438771274388e-05,
"loss": 4.517937469482422,
"step": 360
},
{
"epoch": 0.0788770400352871,
"grad_norm": 2.135274887084961,
"learning_rate": 4.6066832710668325e-05,
"loss": 4.54853630065918,
"step": 380
},
{
"epoch": 0.08302846319503905,
"grad_norm": 2.0009491443634033,
"learning_rate": 4.585927770859278e-05,
"loss": 4.498529434204102,
"step": 400
},
{
"epoch": 0.087179886354791,
"grad_norm": 1.803911566734314,
"learning_rate": 4.565172270651723e-05,
"loss": 4.513103485107422,
"step": 420
},
{
"epoch": 0.09133130951454295,
"grad_norm": 1.7861387729644775,
"learning_rate": 4.5444167704441684e-05,
"loss": 4.456661605834961,
"step": 440
},
{
"epoch": 0.09548273267429491,
"grad_norm": 2.1625704765319824,
"learning_rate": 4.523661270236613e-05,
"loss": 4.533945846557617,
"step": 460
},
{
"epoch": 0.09963415583404686,
"grad_norm": 2.4035468101501465,
"learning_rate": 4.5029057700290576e-05,
"loss": 4.522591018676758,
"step": 480
},
{
"epoch": 0.10378557899379881,
"grad_norm": 1.579237699508667,
"learning_rate": 4.482150269821503e-05,
"loss": 4.502796173095703,
"step": 500
},
{
"epoch": 0.10793700215355076,
"grad_norm": 1.5584685802459717,
"learning_rate": 4.461394769613948e-05,
"loss": 4.617362213134766,
"step": 520
},
{
"epoch": 0.11208842531330272,
"grad_norm": 1.9391822814941406,
"learning_rate": 4.4406392694063934e-05,
"loss": 4.469199371337891,
"step": 540
},
{
"epoch": 0.11623984847305467,
"grad_norm": 1.661686897277832,
"learning_rate": 4.419883769198838e-05,
"loss": 4.474732971191406,
"step": 560
},
{
"epoch": 0.12039127163280662,
"grad_norm": 1.7620030641555786,
"learning_rate": 4.3991282689912826e-05,
"loss": 4.481924819946289,
"step": 580
},
{
"epoch": 0.12454269479255857,
"grad_norm": 1.7622147798538208,
"learning_rate": 4.378372768783728e-05,
"loss": 4.397687911987305,
"step": 600
},
{
"epoch": 0.12869411795231053,
"grad_norm": 1.8880603313446045,
"learning_rate": 4.357617268576173e-05,
"loss": 4.373026657104492,
"step": 620
},
{
"epoch": 0.13284554111206248,
"grad_norm": 1.8983092308044434,
"learning_rate": 4.336861768368618e-05,
"loss": 4.504365921020508,
"step": 640
},
{
"epoch": 0.13699696427181443,
"grad_norm": 1.7019997835159302,
"learning_rate": 4.316106268161063e-05,
"loss": 4.426007080078125,
"step": 660
},
{
"epoch": 0.14114838743156638,
"grad_norm": 1.8891173601150513,
"learning_rate": 4.2953507679535076e-05,
"loss": 4.482881927490235,
"step": 680
},
{
"epoch": 0.14529981059131833,
"grad_norm": 1.498939037322998,
"learning_rate": 4.274595267745953e-05,
"loss": 4.356070709228516,
"step": 700
},
{
"epoch": 0.14945123375107028,
"grad_norm": 2.1021225452423096,
"learning_rate": 4.253839767538398e-05,
"loss": 4.455725479125976,
"step": 720
},
{
"epoch": 0.15360265691082226,
"grad_norm": 1.6581143140792847,
"learning_rate": 4.233084267330843e-05,
"loss": 4.460345077514648,
"step": 740
},
{
"epoch": 0.1577540800705742,
"grad_norm": 2.000790596008301,
"learning_rate": 4.212328767123288e-05,
"loss": 4.41396484375,
"step": 760
},
{
"epoch": 0.16190550323032615,
"grad_norm": 2.255634069442749,
"learning_rate": 4.1915732669157326e-05,
"loss": 4.430901718139649,
"step": 780
},
{
"epoch": 0.1660569263900781,
"grad_norm": 1.521227240562439,
"learning_rate": 4.170817766708178e-05,
"loss": 4.428688049316406,
"step": 800
},
{
"epoch": 0.17020834954983005,
"grad_norm": 1.8831535577774048,
"learning_rate": 4.1500622665006225e-05,
"loss": 4.446892929077149,
"step": 820
},
{
"epoch": 0.174359772709582,
"grad_norm": 1.6413975954055786,
"learning_rate": 4.129306766293068e-05,
"loss": 4.484665298461914,
"step": 840
},
{
"epoch": 0.17851119586933395,
"grad_norm": 1.6965999603271484,
"learning_rate": 4.108551266085513e-05,
"loss": 4.374794387817383,
"step": 860
},
{
"epoch": 0.1826626190290859,
"grad_norm": 1.6068661212921143,
"learning_rate": 4.087795765877958e-05,
"loss": 4.386708831787109,
"step": 880
},
{
"epoch": 0.18681404218883785,
"grad_norm": 1.6991870403289795,
"learning_rate": 4.067040265670403e-05,
"loss": 4.362631607055664,
"step": 900
},
{
"epoch": 0.19096546534858982,
"grad_norm": 1.648937463760376,
"learning_rate": 4.0462847654628475e-05,
"loss": 4.352211380004883,
"step": 920
},
{
"epoch": 0.19511688850834177,
"grad_norm": 2.1894638538360596,
"learning_rate": 4.025529265255293e-05,
"loss": 4.468412017822265,
"step": 940
},
{
"epoch": 0.19926831166809372,
"grad_norm": 1.689712405204773,
"learning_rate": 4.004773765047738e-05,
"loss": 4.373916244506836,
"step": 960
},
{
"epoch": 0.20341973482784567,
"grad_norm": 1.724420189857483,
"learning_rate": 3.984018264840183e-05,
"loss": 4.3584449768066404,
"step": 980
},
{
"epoch": 0.20757115798759762,
"grad_norm": 1.4883451461791992,
"learning_rate": 3.963262764632628e-05,
"loss": 4.456719589233399,
"step": 1000
},
{
"epoch": 0.21172258114734957,
"grad_norm": 3.2922024726867676,
"learning_rate": 3.9425072644250725e-05,
"loss": 4.417997360229492,
"step": 1020
},
{
"epoch": 0.21587400430710152,
"grad_norm": 1.7016091346740723,
"learning_rate": 3.921751764217518e-05,
"loss": 4.3499900817871096,
"step": 1040
},
{
"epoch": 0.22002542746685347,
"grad_norm": 1.5597193241119385,
"learning_rate": 3.900996264009963e-05,
"loss": 4.464202880859375,
"step": 1060
},
{
"epoch": 0.22417685062660545,
"grad_norm": 1.7693027257919312,
"learning_rate": 3.8802407638024077e-05,
"loss": 4.411302947998047,
"step": 1080
},
{
"epoch": 0.2283282737863574,
"grad_norm": 1.5366003513336182,
"learning_rate": 3.859485263594853e-05,
"loss": 4.354082489013672,
"step": 1100
},
{
"epoch": 0.23247969694610934,
"grad_norm": 1.7182965278625488,
"learning_rate": 3.8387297633872975e-05,
"loss": 4.330610275268555,
"step": 1120
},
{
"epoch": 0.2366311201058613,
"grad_norm": 1.6730296611785889,
"learning_rate": 3.817974263179743e-05,
"loss": 4.348351669311524,
"step": 1140
},
{
"epoch": 0.24078254326561324,
"grad_norm": 1.642297387123108,
"learning_rate": 3.797218762972188e-05,
"loss": 4.296623992919922,
"step": 1160
},
{
"epoch": 0.2449339664253652,
"grad_norm": 1.4549849033355713,
"learning_rate": 3.776463262764633e-05,
"loss": 4.38079833984375,
"step": 1180
},
{
"epoch": 0.24908538958511714,
"grad_norm": 1.8313651084899902,
"learning_rate": 3.755707762557078e-05,
"loss": 4.365675735473633,
"step": 1200
},
{
"epoch": 0.2532368127448691,
"grad_norm": 1.7877867221832275,
"learning_rate": 3.7349522623495225e-05,
"loss": 4.326084899902344,
"step": 1220
},
{
"epoch": 0.25738823590462107,
"grad_norm": 1.765218734741211,
"learning_rate": 3.714196762141968e-05,
"loss": 4.333837890625,
"step": 1240
},
{
"epoch": 0.261539659064373,
"grad_norm": 1.752349853515625,
"learning_rate": 3.6934412619344124e-05,
"loss": 4.4339031219482425,
"step": 1260
},
{
"epoch": 0.26569108222412496,
"grad_norm": 1.6944761276245117,
"learning_rate": 3.672685761726858e-05,
"loss": 4.333243942260742,
"step": 1280
},
{
"epoch": 0.2698425053838769,
"grad_norm": 1.6764178276062012,
"learning_rate": 3.651930261519303e-05,
"loss": 4.3393302917480465,
"step": 1300
},
{
"epoch": 0.27399392854362886,
"grad_norm": 1.558286190032959,
"learning_rate": 3.631174761311748e-05,
"loss": 4.29670295715332,
"step": 1320
},
{
"epoch": 0.27814535170338084,
"grad_norm": 1.552565097808838,
"learning_rate": 3.610419261104193e-05,
"loss": 4.305424499511719,
"step": 1340
},
{
"epoch": 0.28229677486313276,
"grad_norm": 1.7239187955856323,
"learning_rate": 3.5896637608966374e-05,
"loss": 4.302777862548828,
"step": 1360
},
{
"epoch": 0.28644819802288474,
"grad_norm": 1.8057512044906616,
"learning_rate": 3.568908260689083e-05,
"loss": 4.289065170288086,
"step": 1380
},
{
"epoch": 0.29059962118263666,
"grad_norm": 1.9495364427566528,
"learning_rate": 3.548152760481528e-05,
"loss": 4.281164169311523,
"step": 1400
},
{
"epoch": 0.29475104434238864,
"grad_norm": 2.0150704383850098,
"learning_rate": 3.527397260273973e-05,
"loss": 4.300543975830078,
"step": 1420
},
{
"epoch": 0.29890246750214056,
"grad_norm": 1.7684186697006226,
"learning_rate": 3.506641760066418e-05,
"loss": 4.193937683105469,
"step": 1440
},
{
"epoch": 0.30305389066189253,
"grad_norm": 1.7381044626235962,
"learning_rate": 3.4858862598588625e-05,
"loss": 4.379112243652344,
"step": 1460
},
{
"epoch": 0.3072053138216445,
"grad_norm": 1.6002411842346191,
"learning_rate": 3.465130759651308e-05,
"loss": 4.297174072265625,
"step": 1480
},
{
"epoch": 0.31135673698139643,
"grad_norm": 1.5293762683868408,
"learning_rate": 3.444375259443753e-05,
"loss": 4.3323722839355465,
"step": 1500
},
{
"epoch": 0.3155081601411484,
"grad_norm": 1.5183730125427246,
"learning_rate": 3.4236197592361976e-05,
"loss": 4.206635284423828,
"step": 1520
},
{
"epoch": 0.31965958330090033,
"grad_norm": 1.6699546575546265,
"learning_rate": 3.402864259028642e-05,
"loss": 4.25066032409668,
"step": 1540
},
{
"epoch": 0.3238110064606523,
"grad_norm": 1.612926721572876,
"learning_rate": 3.3821087588210875e-05,
"loss": 4.241170120239258,
"step": 1560
},
{
"epoch": 0.3279624296204042,
"grad_norm": 1.836079478263855,
"learning_rate": 3.361353258613533e-05,
"loss": 4.270659255981445,
"step": 1580
},
{
"epoch": 0.3321138527801562,
"grad_norm": 1.9245415925979614,
"learning_rate": 3.340597758405978e-05,
"loss": 4.2110248565673825,
"step": 1600
},
{
"epoch": 0.3362652759399081,
"grad_norm": 2.754229784011841,
"learning_rate": 3.3198422581984226e-05,
"loss": 4.291688537597656,
"step": 1620
},
{
"epoch": 0.3404166990996601,
"grad_norm": 2.0228278636932373,
"learning_rate": 3.299086757990868e-05,
"loss": 4.2379310607910154,
"step": 1640
},
{
"epoch": 0.3445681222594121,
"grad_norm": 1.6231938600540161,
"learning_rate": 3.2783312577833125e-05,
"loss": 4.206072616577148,
"step": 1660
},
{
"epoch": 0.348719545419164,
"grad_norm": 1.7419497966766357,
"learning_rate": 3.257575757575758e-05,
"loss": 4.125591278076172,
"step": 1680
},
{
"epoch": 0.352870968578916,
"grad_norm": 1.8795361518859863,
"learning_rate": 3.236820257368203e-05,
"loss": 4.194551849365235,
"step": 1700
},
{
"epoch": 0.3570223917386679,
"grad_norm": 1.7395036220550537,
"learning_rate": 3.2160647571606476e-05,
"loss": 4.3172355651855465,
"step": 1720
},
{
"epoch": 0.3611738148984199,
"grad_norm": 2.3381433486938477,
"learning_rate": 3.195309256953093e-05,
"loss": 4.3080402374267575,
"step": 1740
},
{
"epoch": 0.3653252380581718,
"grad_norm": 1.4747214317321777,
"learning_rate": 3.1745537567455375e-05,
"loss": 4.1682899475097654,
"step": 1760
},
{
"epoch": 0.3694766612179238,
"grad_norm": 1.8914560079574585,
"learning_rate": 3.153798256537983e-05,
"loss": 4.221502304077148,
"step": 1780
},
{
"epoch": 0.3736280843776757,
"grad_norm": 1.6433378458023071,
"learning_rate": 3.1330427563304274e-05,
"loss": 4.272403717041016,
"step": 1800
},
{
"epoch": 0.3777795075374277,
"grad_norm": 1.6937401294708252,
"learning_rate": 3.1122872561228727e-05,
"loss": 4.236754989624023,
"step": 1820
},
{
"epoch": 0.38193093069717965,
"grad_norm": 1.7504234313964844,
"learning_rate": 3.091531755915318e-05,
"loss": 4.28508415222168,
"step": 1840
},
{
"epoch": 0.38608235385693157,
"grad_norm": 1.9228798151016235,
"learning_rate": 3.070776255707763e-05,
"loss": 4.103425598144531,
"step": 1860
},
{
"epoch": 0.39023377701668355,
"grad_norm": 1.5827641487121582,
"learning_rate": 3.0500207555002075e-05,
"loss": 4.163749694824219,
"step": 1880
},
{
"epoch": 0.39438520017643547,
"grad_norm": 1.950251817703247,
"learning_rate": 3.0292652552926527e-05,
"loss": 4.227020263671875,
"step": 1900
},
{
"epoch": 0.39853662333618745,
"grad_norm": 1.749630331993103,
"learning_rate": 3.0085097550850977e-05,
"loss": 4.166823196411133,
"step": 1920
},
{
"epoch": 0.40268804649593937,
"grad_norm": 1.5769996643066406,
"learning_rate": 2.987754254877543e-05,
"loss": 4.163990020751953,
"step": 1940
},
{
"epoch": 0.40683946965569134,
"grad_norm": 1.7626852989196777,
"learning_rate": 2.966998754669988e-05,
"loss": 4.200388336181641,
"step": 1960
},
{
"epoch": 0.4109908928154433,
"grad_norm": 1.8192287683486938,
"learning_rate": 2.9462432544624325e-05,
"loss": 4.156772232055664,
"step": 1980
},
{
"epoch": 0.41514231597519524,
"grad_norm": 1.9362469911575317,
"learning_rate": 2.9254877542548774e-05,
"loss": 4.3077339172363285,
"step": 2000
},
{
"epoch": 0.4192937391349472,
"grad_norm": 2.0180394649505615,
"learning_rate": 2.9047322540473227e-05,
"loss": 4.215380859375,
"step": 2020
},
{
"epoch": 0.42344516229469914,
"grad_norm": 1.5576096773147583,
"learning_rate": 2.8839767538397676e-05,
"loss": 4.081830215454102,
"step": 2040
},
{
"epoch": 0.4275965854544511,
"grad_norm": 1.7119743824005127,
"learning_rate": 2.863221253632213e-05,
"loss": 4.075210952758789,
"step": 2060
},
{
"epoch": 0.43174800861420304,
"grad_norm": 1.9975166320800781,
"learning_rate": 2.842465753424658e-05,
"loss": 4.085906600952148,
"step": 2080
},
{
"epoch": 0.435899431773955,
"grad_norm": 1.7588188648223877,
"learning_rate": 2.8217102532171024e-05,
"loss": 4.248412322998047,
"step": 2100
},
{
"epoch": 0.44005085493370694,
"grad_norm": 1.7669799327850342,
"learning_rate": 2.8009547530095477e-05,
"loss": 4.175002670288086,
"step": 2120
},
{
"epoch": 0.4442022780934589,
"grad_norm": 1.8094285726547241,
"learning_rate": 2.7801992528019926e-05,
"loss": 4.089089584350586,
"step": 2140
},
{
"epoch": 0.4483537012532109,
"grad_norm": 1.6908018589019775,
"learning_rate": 2.759443752594438e-05,
"loss": 4.142142105102539,
"step": 2160
},
{
"epoch": 0.4525051244129628,
"grad_norm": 1.474813461303711,
"learning_rate": 2.738688252386883e-05,
"loss": 4.158762741088867,
"step": 2180
},
{
"epoch": 0.4566565475727148,
"grad_norm": 1.6043535470962524,
"learning_rate": 2.7179327521793274e-05,
"loss": 4.0978271484375,
"step": 2200
},
{
"epoch": 0.4608079707324667,
"grad_norm": 2.129974126815796,
"learning_rate": 2.6971772519717724e-05,
"loss": 4.187951278686524,
"step": 2220
},
{
"epoch": 0.4649593938922187,
"grad_norm": 1.8921549320220947,
"learning_rate": 2.6764217517642177e-05,
"loss": 4.139292526245117,
"step": 2240
},
{
"epoch": 0.4691108170519706,
"grad_norm": 1.7402410507202148,
"learning_rate": 2.6556662515566626e-05,
"loss": 4.124836349487305,
"step": 2260
},
{
"epoch": 0.4732622402117226,
"grad_norm": 1.734351634979248,
"learning_rate": 2.634910751349108e-05,
"loss": 4.1282188415527346,
"step": 2280
},
{
"epoch": 0.4774136633714745,
"grad_norm": 1.5689741373062134,
"learning_rate": 2.6141552511415528e-05,
"loss": 4.073775100708008,
"step": 2300
},
{
"epoch": 0.4815650865312265,
"grad_norm": 1.8727039098739624,
"learning_rate": 2.5933997509339974e-05,
"loss": 4.214794158935547,
"step": 2320
},
{
"epoch": 0.48571650969097846,
"grad_norm": 1.715066909790039,
"learning_rate": 2.5726442507264427e-05,
"loss": 4.0818939208984375,
"step": 2340
},
{
"epoch": 0.4898679328507304,
"grad_norm": 1.7405266761779785,
"learning_rate": 2.5518887505188876e-05,
"loss": 4.194419097900391,
"step": 2360
},
{
"epoch": 0.49401935601048236,
"grad_norm": 1.8849917650222778,
"learning_rate": 2.531133250311333e-05,
"loss": 4.092476272583008,
"step": 2380
},
{
"epoch": 0.4981707791702343,
"grad_norm": 1.784807801246643,
"learning_rate": 2.5103777501037778e-05,
"loss": 4.206594085693359,
"step": 2400
},
{
"epoch": 0.5023222023299863,
"grad_norm": 1.6309819221496582,
"learning_rate": 2.4896222498962228e-05,
"loss": 4.067199325561523,
"step": 2420
},
{
"epoch": 0.5064736254897382,
"grad_norm": 1.8992376327514648,
"learning_rate": 2.4688667496886674e-05,
"loss": 4.026485824584961,
"step": 2440
},
{
"epoch": 0.5106250486494901,
"grad_norm": 2.6366453170776367,
"learning_rate": 2.4481112494811126e-05,
"loss": 4.160321426391602,
"step": 2460
},
{
"epoch": 0.5147764718092421,
"grad_norm": 1.9022001028060913,
"learning_rate": 2.4273557492735576e-05,
"loss": 4.082183456420898,
"step": 2480
},
{
"epoch": 0.518927894968994,
"grad_norm": 1.8230382204055786,
"learning_rate": 2.4066002490660025e-05,
"loss": 4.191316986083985,
"step": 2500
},
{
"epoch": 0.523079318128746,
"grad_norm": 2.047727346420288,
"learning_rate": 2.3858447488584478e-05,
"loss": 4.14111557006836,
"step": 2520
},
{
"epoch": 0.527230741288498,
"grad_norm": 1.5861501693725586,
"learning_rate": 2.3650892486508927e-05,
"loss": 4.128021240234375,
"step": 2540
},
{
"epoch": 0.5313821644482499,
"grad_norm": 2.292236566543579,
"learning_rate": 2.3443337484433376e-05,
"loss": 4.091572189331055,
"step": 2560
},
{
"epoch": 0.5355335876080018,
"grad_norm": 1.804754614830017,
"learning_rate": 2.3235782482357826e-05,
"loss": 4.194594192504883,
"step": 2580
},
{
"epoch": 0.5396850107677538,
"grad_norm": 1.6975075006484985,
"learning_rate": 2.302822748028228e-05,
"loss": 4.1291240692138675,
"step": 2600
},
{
"epoch": 0.5438364339275058,
"grad_norm": 1.6792091131210327,
"learning_rate": 2.2820672478206725e-05,
"loss": 4.039973449707031,
"step": 2620
},
{
"epoch": 0.5479878570872577,
"grad_norm": 1.8002510070800781,
"learning_rate": 2.2613117476131177e-05,
"loss": 4.076683807373047,
"step": 2640
},
{
"epoch": 0.5521392802470096,
"grad_norm": 1.6224623918533325,
"learning_rate": 2.2405562474055623e-05,
"loss": 4.095446014404297,
"step": 2660
},
{
"epoch": 0.5562907034067617,
"grad_norm": 1.974475383758545,
"learning_rate": 2.2198007471980076e-05,
"loss": 4.099486923217773,
"step": 2680
},
{
"epoch": 0.5604421265665136,
"grad_norm": 2.1149117946624756,
"learning_rate": 2.1990452469904525e-05,
"loss": 4.086590194702149,
"step": 2700
},
{
"epoch": 0.5645935497262655,
"grad_norm": 1.7754842042922974,
"learning_rate": 2.1782897467828975e-05,
"loss": 4.002002334594726,
"step": 2720
},
{
"epoch": 0.5687449728860174,
"grad_norm": 1.7504503726959229,
"learning_rate": 2.1575342465753427e-05,
"loss": 4.057027435302734,
"step": 2740
},
{
"epoch": 0.5728963960457695,
"grad_norm": 2.5411620140075684,
"learning_rate": 2.1367787463677877e-05,
"loss": 4.031216812133789,
"step": 2760
},
{
"epoch": 0.5770478192055214,
"grad_norm": 1.8988471031188965,
"learning_rate": 2.1160232461602326e-05,
"loss": 4.08874626159668,
"step": 2780
},
{
"epoch": 0.5811992423652733,
"grad_norm": 1.8526592254638672,
"learning_rate": 2.0952677459526776e-05,
"loss": 4.086692047119141,
"step": 2800
},
{
"epoch": 0.5853506655250253,
"grad_norm": 1.820813775062561,
"learning_rate": 2.0745122457451225e-05,
"loss": 4.064103698730468,
"step": 2820
},
{
"epoch": 0.5895020886847773,
"grad_norm": 1.9091445207595825,
"learning_rate": 2.0537567455375674e-05,
"loss": 4.169759750366211,
"step": 2840
},
{
"epoch": 0.5936535118445292,
"grad_norm": 1.9170401096343994,
"learning_rate": 2.0330012453300127e-05,
"loss": 4.0568359375,
"step": 2860
},
{
"epoch": 0.5978049350042811,
"grad_norm": 2.160008192062378,
"learning_rate": 2.0122457451224573e-05,
"loss": 4.070849609375,
"step": 2880
},
{
"epoch": 0.6019563581640331,
"grad_norm": 1.9741166830062866,
"learning_rate": 1.9914902449149026e-05,
"loss": 4.026998901367188,
"step": 2900
},
{
"epoch": 0.6061077813237851,
"grad_norm": 1.9789372682571411,
"learning_rate": 1.9707347447073475e-05,
"loss": 4.116487884521485,
"step": 2920
},
{
"epoch": 0.610259204483537,
"grad_norm": 1.7258014678955078,
"learning_rate": 1.9499792444997924e-05,
"loss": 4.0306743621826175,
"step": 2940
},
{
"epoch": 0.614410627643289,
"grad_norm": 1.7329745292663574,
"learning_rate": 1.9292237442922377e-05,
"loss": 4.183980941772461,
"step": 2960
},
{
"epoch": 0.6185620508030409,
"grad_norm": 1.7421926259994507,
"learning_rate": 1.9084682440846827e-05,
"loss": 4.154231262207031,
"step": 2980
},
{
"epoch": 0.6227134739627929,
"grad_norm": 1.8388780355453491,
"learning_rate": 1.8877127438771276e-05,
"loss": 3.9981304168701173,
"step": 3000
},
{
"epoch": 0.6268648971225448,
"grad_norm": 1.7260050773620605,
"learning_rate": 1.8669572436695725e-05,
"loss": 4.121049880981445,
"step": 3020
},
{
"epoch": 0.6310163202822968,
"grad_norm": 2.0676257610321045,
"learning_rate": 1.8462017434620175e-05,
"loss": 4.020652389526367,
"step": 3040
},
{
"epoch": 0.6351677434420487,
"grad_norm": 1.684615969657898,
"learning_rate": 1.8254462432544624e-05,
"loss": 4.100597763061524,
"step": 3060
},
{
"epoch": 0.6393191666018007,
"grad_norm": 1.8125957250595093,
"learning_rate": 1.8046907430469077e-05,
"loss": 3.976701354980469,
"step": 3080
},
{
"epoch": 0.6434705897615526,
"grad_norm": 1.6545263528823853,
"learning_rate": 1.7839352428393523e-05,
"loss": 4.052294921875,
"step": 3100
},
{
"epoch": 0.6476220129213046,
"grad_norm": 1.919820785522461,
"learning_rate": 1.7631797426317975e-05,
"loss": 4.060013961791992,
"step": 3120
},
{
"epoch": 0.6517734360810565,
"grad_norm": 1.9719918966293335,
"learning_rate": 1.7424242424242425e-05,
"loss": 4.023235321044922,
"step": 3140
},
{
"epoch": 0.6559248592408085,
"grad_norm": 1.6140228509902954,
"learning_rate": 1.7216687422166874e-05,
"loss": 4.122441101074219,
"step": 3160
},
{
"epoch": 0.6600762824005605,
"grad_norm": 1.955754280090332,
"learning_rate": 1.7009132420091327e-05,
"loss": 4.060406494140625,
"step": 3180
},
{
"epoch": 0.6642277055603124,
"grad_norm": 1.9840017557144165,
"learning_rate": 1.6801577418015773e-05,
"loss": 4.073451232910156,
"step": 3200
},
{
"epoch": 0.6683791287200643,
"grad_norm": 1.8778005838394165,
"learning_rate": 1.6594022415940226e-05,
"loss": 4.049271011352539,
"step": 3220
},
{
"epoch": 0.6725305518798163,
"grad_norm": 1.5047231912612915,
"learning_rate": 1.6386467413864675e-05,
"loss": 4.047217941284179,
"step": 3240
},
{
"epoch": 0.6766819750395683,
"grad_norm": 1.6387065649032593,
"learning_rate": 1.6178912411789124e-05,
"loss": 4.080060958862305,
"step": 3260
},
{
"epoch": 0.6808333981993202,
"grad_norm": 1.9958332777023315,
"learning_rate": 1.5971357409713574e-05,
"loss": 4.042473983764649,
"step": 3280
},
{
"epoch": 0.6849848213590721,
"grad_norm": 2.604090690612793,
"learning_rate": 1.5763802407638026e-05,
"loss": 4.063697052001953,
"step": 3300
},
{
"epoch": 0.6891362445188242,
"grad_norm": 2.2081997394561768,
"learning_rate": 1.5556247405562472e-05,
"loss": 4.024546432495117,
"step": 3320
},
{
"epoch": 0.6932876676785761,
"grad_norm": 1.7056355476379395,
"learning_rate": 1.5348692403486925e-05,
"loss": 3.986798095703125,
"step": 3340
},
{
"epoch": 0.697439090838328,
"grad_norm": 1.8961299657821655,
"learning_rate": 1.5141137401411376e-05,
"loss": 4.012807083129883,
"step": 3360
},
{
"epoch": 0.7015905139980799,
"grad_norm": 2.0114150047302246,
"learning_rate": 1.4933582399335824e-05,
"loss": 4.035384750366211,
"step": 3380
},
{
"epoch": 0.705741937157832,
"grad_norm": 2.143017292022705,
"learning_rate": 1.4726027397260275e-05,
"loss": 4.043708419799804,
"step": 3400
},
{
"epoch": 0.7098933603175839,
"grad_norm": 1.9830955266952515,
"learning_rate": 1.4518472395184724e-05,
"loss": 4.090269470214844,
"step": 3420
},
{
"epoch": 0.7140447834773358,
"grad_norm": 2.0425097942352295,
"learning_rate": 1.4310917393109175e-05,
"loss": 3.9867622375488283,
"step": 3440
},
{
"epoch": 0.7181962066370878,
"grad_norm": 1.9336873292922974,
"learning_rate": 1.4103362391033625e-05,
"loss": 3.9401180267333986,
"step": 3460
},
{
"epoch": 0.7223476297968398,
"grad_norm": 1.9117387533187866,
"learning_rate": 1.3895807388958074e-05,
"loss": 3.933444595336914,
"step": 3480
},
{
"epoch": 0.7264990529565917,
"grad_norm": 1.7063461542129517,
"learning_rate": 1.3688252386882525e-05,
"loss": 3.9604202270507813,
"step": 3500
},
{
"epoch": 0.7306504761163436,
"grad_norm": 1.6735029220581055,
"learning_rate": 1.3480697384806976e-05,
"loss": 3.9346424102783204,
"step": 3520
},
{
"epoch": 0.7348018992760956,
"grad_norm": 2.1304147243499756,
"learning_rate": 1.3273142382731424e-05,
"loss": 4.046992492675781,
"step": 3540
},
{
"epoch": 0.7389533224358475,
"grad_norm": 1.7260125875473022,
"learning_rate": 1.3065587380655875e-05,
"loss": 3.9742919921875,
"step": 3560
},
{
"epoch": 0.7431047455955995,
"grad_norm": 1.8843480348587036,
"learning_rate": 1.2858032378580322e-05,
"loss": 3.9342079162597656,
"step": 3580
},
{
"epoch": 0.7472561687553514,
"grad_norm": 1.9145067930221558,
"learning_rate": 1.2650477376504774e-05,
"loss": 3.945034408569336,
"step": 3600
},
{
"epoch": 0.7514075919151034,
"grad_norm": 1.7768443822860718,
"learning_rate": 1.2442922374429223e-05,
"loss": 4.006913375854492,
"step": 3620
},
{
"epoch": 0.7555590150748553,
"grad_norm": 1.6193681955337524,
"learning_rate": 1.2235367372353674e-05,
"loss": 3.9851921081542967,
"step": 3640
},
{
"epoch": 0.7597104382346073,
"grad_norm": 1.8195797204971313,
"learning_rate": 1.2027812370278125e-05,
"loss": 4.047183609008789,
"step": 3660
},
{
"epoch": 0.7638618613943593,
"grad_norm": 1.7672561407089233,
"learning_rate": 1.1820257368202574e-05,
"loss": 3.8762928009033204,
"step": 3680
},
{
"epoch": 0.7680132845541112,
"grad_norm": 1.561675786972046,
"learning_rate": 1.1612702366127025e-05,
"loss": 3.942879486083984,
"step": 3700
},
{
"epoch": 0.7721647077138631,
"grad_norm": 2.8441998958587646,
"learning_rate": 1.1405147364051475e-05,
"loss": 3.908550262451172,
"step": 3720
},
{
"epoch": 0.7763161308736151,
"grad_norm": 2.1282033920288086,
"learning_rate": 1.1197592361975924e-05,
"loss": 4.002864837646484,
"step": 3740
},
{
"epoch": 0.7804675540333671,
"grad_norm": 2.2484629154205322,
"learning_rate": 1.0990037359900373e-05,
"loss": 3.901498794555664,
"step": 3760
},
{
"epoch": 0.784618977193119,
"grad_norm": 1.747456431388855,
"learning_rate": 1.0782482357824825e-05,
"loss": 4.018042373657226,
"step": 3780
},
{
"epoch": 0.7887704003528709,
"grad_norm": 1.6550432443618774,
"learning_rate": 1.0574927355749274e-05,
"loss": 4.029907608032227,
"step": 3800
},
{
"epoch": 0.792921823512623,
"grad_norm": 1.901953935623169,
"learning_rate": 1.0367372353673723e-05,
"loss": 3.949459457397461,
"step": 3820
},
{
"epoch": 0.7970732466723749,
"grad_norm": 1.8096224069595337,
"learning_rate": 1.0159817351598173e-05,
"loss": 3.950168991088867,
"step": 3840
},
{
"epoch": 0.8012246698321268,
"grad_norm": 1.841707468032837,
"learning_rate": 9.952262349522624e-06,
"loss": 3.970652389526367,
"step": 3860
},
{
"epoch": 0.8053760929918787,
"grad_norm": 1.5807123184204102,
"learning_rate": 9.744707347447075e-06,
"loss": 3.9712162017822266,
"step": 3880
},
{
"epoch": 0.8095275161516308,
"grad_norm": 2.035214424133301,
"learning_rate": 9.537152345371524e-06,
"loss": 4.0423839569091795,
"step": 3900
},
{
"epoch": 0.8136789393113827,
"grad_norm": 1.813348650932312,
"learning_rate": 9.329597343295973e-06,
"loss": 3.961563491821289,
"step": 3920
},
{
"epoch": 0.8178303624711346,
"grad_norm": 1.9857704639434814,
"learning_rate": 9.122042341220424e-06,
"loss": 3.9110523223876954,
"step": 3940
},
{
"epoch": 0.8219817856308866,
"grad_norm": 1.9536502361297607,
"learning_rate": 8.914487339144874e-06,
"loss": 3.9576736450195313,
"step": 3960
},
{
"epoch": 0.8261332087906386,
"grad_norm": 1.9735718965530396,
"learning_rate": 8.706932337069323e-06,
"loss": 3.96722412109375,
"step": 3980
},
{
"epoch": 0.8302846319503905,
"grad_norm": 1.7947627305984497,
"learning_rate": 8.499377334993774e-06,
"loss": 3.940797805786133,
"step": 4000
},
{
"epoch": 0.8344360551101424,
"grad_norm": 1.9073888063430786,
"learning_rate": 8.291822332918224e-06,
"loss": 3.93328857421875,
"step": 4020
},
{
"epoch": 0.8385874782698944,
"grad_norm": 2.2240006923675537,
"learning_rate": 8.084267330842673e-06,
"loss": 3.9660671234130858,
"step": 4040
},
{
"epoch": 0.8427389014296464,
"grad_norm": 1.8106498718261719,
"learning_rate": 7.876712328767124e-06,
"loss": 3.9450511932373047,
"step": 4060
},
{
"epoch": 0.8468903245893983,
"grad_norm": 2.2481167316436768,
"learning_rate": 7.669157326691575e-06,
"loss": 3.9099327087402345,
"step": 4080
},
{
"epoch": 0.8510417477491502,
"grad_norm": 1.6156245470046997,
"learning_rate": 7.4616023246160236e-06,
"loss": 3.900218200683594,
"step": 4100
},
{
"epoch": 0.8551931709089022,
"grad_norm": 2.0962748527526855,
"learning_rate": 7.254047322540474e-06,
"loss": 3.9331405639648436,
"step": 4120
},
{
"epoch": 0.8593445940686542,
"grad_norm": 1.6167515516281128,
"learning_rate": 7.046492320464923e-06,
"loss": 3.880914306640625,
"step": 4140
},
{
"epoch": 0.8634960172284061,
"grad_norm": 2.121077299118042,
"learning_rate": 6.838937318389374e-06,
"loss": 3.917226028442383,
"step": 4160
},
{
"epoch": 0.8676474403881581,
"grad_norm": 1.990082025527954,
"learning_rate": 6.6313823163138235e-06,
"loss": 3.851776123046875,
"step": 4180
},
{
"epoch": 0.87179886354791,
"grad_norm": 1.6866909265518188,
"learning_rate": 6.423827314238273e-06,
"loss": 3.930047607421875,
"step": 4200
},
{
"epoch": 0.875950286707662,
"grad_norm": 1.5321828126907349,
"learning_rate": 6.216272312162723e-06,
"loss": 3.982738494873047,
"step": 4220
},
{
"epoch": 0.8801017098674139,
"grad_norm": 1.831310510635376,
"learning_rate": 6.008717310087173e-06,
"loss": 3.8260471343994142,
"step": 4240
},
{
"epoch": 0.8842531330271659,
"grad_norm": 1.8113958835601807,
"learning_rate": 5.8011623080116235e-06,
"loss": 3.843553924560547,
"step": 4260
},
{
"epoch": 0.8884045561869178,
"grad_norm": 2.2291505336761475,
"learning_rate": 5.593607305936073e-06,
"loss": 3.920602798461914,
"step": 4280
},
{
"epoch": 0.8925559793466697,
"grad_norm": 1.9720070362091064,
"learning_rate": 5.386052303860523e-06,
"loss": 3.9493858337402346,
"step": 4300
},
{
"epoch": 0.8967074025064218,
"grad_norm": 1.892719030380249,
"learning_rate": 5.178497301784973e-06,
"loss": 3.8462478637695314,
"step": 4320
},
{
"epoch": 0.9008588256661737,
"grad_norm": 1.7223536968231201,
"learning_rate": 4.9709422997094235e-06,
"loss": 3.9579647064208983,
"step": 4340
},
{
"epoch": 0.9050102488259256,
"grad_norm": 1.9189648628234863,
"learning_rate": 4.763387297633874e-06,
"loss": 3.856624221801758,
"step": 4360
},
{
"epoch": 0.9091616719856775,
"grad_norm": 1.683077335357666,
"learning_rate": 4.555832295558323e-06,
"loss": 3.9630027770996095,
"step": 4380
},
{
"epoch": 0.9133130951454296,
"grad_norm": 1.4068090915679932,
"learning_rate": 4.348277293482773e-06,
"loss": 3.9053329467773437,
"step": 4400
},
{
"epoch": 0.9174645183051815,
"grad_norm": 1.6004208326339722,
"learning_rate": 4.140722291407223e-06,
"loss": 3.970157241821289,
"step": 4420
},
{
"epoch": 0.9216159414649334,
"grad_norm": 1.7510566711425781,
"learning_rate": 3.933167289331674e-06,
"loss": 3.8958450317382813,
"step": 4440
},
{
"epoch": 0.9257673646246855,
"grad_norm": 1.9590911865234375,
"learning_rate": 3.725612287256123e-06,
"loss": 3.9771854400634767,
"step": 4460
},
{
"epoch": 0.9299187877844374,
"grad_norm": 2.2574198246002197,
"learning_rate": 3.5180572851805732e-06,
"loss": 3.971847915649414,
"step": 4480
},
{
"epoch": 0.9340702109441893,
"grad_norm": 1.7533721923828125,
"learning_rate": 3.310502283105023e-06,
"loss": 3.8168708801269533,
"step": 4500
},
{
"epoch": 0.9382216341039412,
"grad_norm": 1.9453550577163696,
"learning_rate": 3.102947281029473e-06,
"loss": 3.8188804626464843,
"step": 4520
},
{
"epoch": 0.9423730572636932,
"grad_norm": 1.9086698293685913,
"learning_rate": 2.895392278953923e-06,
"loss": 3.977469635009766,
"step": 4540
},
{
"epoch": 0.9465244804234452,
"grad_norm": 1.7968412637710571,
"learning_rate": 2.687837276878373e-06,
"loss": 3.9220932006835936,
"step": 4560
},
{
"epoch": 0.9506759035831971,
"grad_norm": 1.6008837223052979,
"learning_rate": 2.480282274802823e-06,
"loss": 3.927560806274414,
"step": 4580
},
{
"epoch": 0.954827326742949,
"grad_norm": 2.0438432693481445,
"learning_rate": 2.2727272727272728e-06,
"loss": 3.892019271850586,
"step": 4600
},
{
"epoch": 0.958978749902701,
"grad_norm": 1.7777832746505737,
"learning_rate": 2.065172270651723e-06,
"loss": 3.966938018798828,
"step": 4620
},
{
"epoch": 0.963130173062453,
"grad_norm": 1.823589563369751,
"learning_rate": 1.8576172685761727e-06,
"loss": 3.950274658203125,
"step": 4640
},
{
"epoch": 0.9672815962222049,
"grad_norm": 1.8060129880905151,
"learning_rate": 1.6500622665006227e-06,
"loss": 3.8800033569335937,
"step": 4660
},
{
"epoch": 0.9714330193819569,
"grad_norm": 1.9142494201660156,
"learning_rate": 1.4425072644250727e-06,
"loss": 3.871175003051758,
"step": 4680
},
{
"epoch": 0.9755844425417088,
"grad_norm": 1.6857857704162598,
"learning_rate": 1.2349522623495227e-06,
"loss": 3.895126724243164,
"step": 4700
},
{
"epoch": 0.9797358657014608,
"grad_norm": 1.9794493913650513,
"learning_rate": 1.0273972602739725e-06,
"loss": 3.9186450958251955,
"step": 4720
},
{
"epoch": 0.9838872888612127,
"grad_norm": 1.5562703609466553,
"learning_rate": 8.198422581984225e-07,
"loss": 3.9541133880615233,
"step": 4740
},
{
"epoch": 0.9880387120209647,
"grad_norm": 1.8461521863937378,
"learning_rate": 6.122872561228726e-07,
"loss": 3.854688262939453,
"step": 4760
},
{
"epoch": 0.9921901351807166,
"grad_norm": 2.31740665435791,
"learning_rate": 4.0473225404732254e-07,
"loss": 3.991476821899414,
"step": 4780
},
{
"epoch": 0.9963415583404686,
"grad_norm": 1.7164037227630615,
"learning_rate": 1.9717725197177253e-07,
"loss": 3.9244304656982423,
"step": 4800
}
],
"logging_steps": 20,
"max_steps": 4818,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 300,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 7.15847138845655e+16,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}