ttt1002 / trainer_state.json
Kaipengm2's picture
Upload folder using huggingface_hub
73b1a82 verified
Raw
History Blame Contribute Delete
92 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 2454,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.002037489812550937,
"grad_norm": 1.571320652961731,
"learning_rate": 5.405405405405406e-06,
"loss": 2.937255096435547,
"step": 5
},
{
"epoch": 0.004074979625101874,
"grad_norm": 1.180213212966919,
"learning_rate": 1.2162162162162164e-05,
"loss": 2.8778331756591795,
"step": 10
},
{
"epoch": 0.006112469437652812,
"grad_norm": 0.6333388686180115,
"learning_rate": 1.891891891891892e-05,
"loss": 2.7612342834472656,
"step": 15
},
{
"epoch": 0.008149959250203748,
"grad_norm": 0.29147869348526,
"learning_rate": 2.5675675675675675e-05,
"loss": 2.590922546386719,
"step": 20
},
{
"epoch": 0.010187449062754686,
"grad_norm": 0.27221187949180603,
"learning_rate": 3.2432432432432436e-05,
"loss": 2.4679227828979493,
"step": 25
},
{
"epoch": 0.012224938875305624,
"grad_norm": 0.35275858640670776,
"learning_rate": 3.918918918918919e-05,
"loss": 2.4889301300048827,
"step": 30
},
{
"epoch": 0.01426242868785656,
"grad_norm": 0.2352866381406784,
"learning_rate": 4.594594594594595e-05,
"loss": 2.3145923614501953,
"step": 35
},
{
"epoch": 0.016299918500407497,
"grad_norm": 0.18694627285003662,
"learning_rate": 5.27027027027027e-05,
"loss": 2.265145111083984,
"step": 40
},
{
"epoch": 0.018337408312958436,
"grad_norm": 0.20410102605819702,
"learning_rate": 5.9459459459459466e-05,
"loss": 2.1495094299316406,
"step": 45
},
{
"epoch": 0.020374898125509373,
"grad_norm": 0.1883523017168045,
"learning_rate": 6.621621621621621e-05,
"loss": 2.1063323974609376,
"step": 50
},
{
"epoch": 0.02241238793806031,
"grad_norm": 0.22693860530853271,
"learning_rate": 7.297297297297297e-05,
"loss": 2.0485776901245116,
"step": 55
},
{
"epoch": 0.02444987775061125,
"grad_norm": 0.22750967741012573,
"learning_rate": 7.972972972972974e-05,
"loss": 1.980290412902832,
"step": 60
},
{
"epoch": 0.026487367563162185,
"grad_norm": 0.1929962933063507,
"learning_rate": 8.64864864864865e-05,
"loss": 1.9449413299560547,
"step": 65
},
{
"epoch": 0.02852485737571312,
"grad_norm": 0.19021861255168915,
"learning_rate": 9.324324324324324e-05,
"loss": 1.8876354217529296,
"step": 70
},
{
"epoch": 0.030562347188264057,
"grad_norm": 0.19133411347866058,
"learning_rate": 0.0001,
"loss": 1.8819442749023438,
"step": 75
},
{
"epoch": 0.032599837000814993,
"grad_norm": 0.18844370543956757,
"learning_rate": 9.999891100904613e-05,
"loss": 1.8457191467285157,
"step": 80
},
{
"epoch": 0.03463732681336593,
"grad_norm": 0.17675529420375824,
"learning_rate": 9.999564408362054e-05,
"loss": 1.8305767059326172,
"step": 85
},
{
"epoch": 0.03667481662591687,
"grad_norm": 0.21320666372776031,
"learning_rate": 9.999019936602934e-05,
"loss": 1.7988300323486328,
"step": 90
},
{
"epoch": 0.038712306438467806,
"grad_norm": 0.23267436027526855,
"learning_rate": 9.998257709344246e-05,
"loss": 1.7630163192749024,
"step": 95
},
{
"epoch": 0.040749796251018745,
"grad_norm": 0.21504434943199158,
"learning_rate": 9.997277759788331e-05,
"loss": 1.753378677368164,
"step": 100
},
{
"epoch": 0.042787286063569685,
"grad_norm": 0.22522081434726715,
"learning_rate": 9.996080130621439e-05,
"loss": 1.738396453857422,
"step": 105
},
{
"epoch": 0.04482477587612062,
"grad_norm": 0.21060502529144287,
"learning_rate": 9.994664874011863e-05,
"loss": 1.6924251556396483,
"step": 110
},
{
"epoch": 0.04686226568867156,
"grad_norm": 0.22877274453639984,
"learning_rate": 9.993032051607669e-05,
"loss": 1.705193328857422,
"step": 115
},
{
"epoch": 0.0488997555012225,
"grad_norm": 0.22202537953853607,
"learning_rate": 9.99118173453401e-05,
"loss": 1.6783943176269531,
"step": 120
},
{
"epoch": 0.05093724531377343,
"grad_norm": 0.2630864679813385,
"learning_rate": 9.989114003390027e-05,
"loss": 1.661654281616211,
"step": 125
},
{
"epoch": 0.05297473512632437,
"grad_norm": 0.24131977558135986,
"learning_rate": 9.986828948245343e-05,
"loss": 1.6814453125,
"step": 130
},
{
"epoch": 0.0550122249388753,
"grad_norm": 0.2544144093990326,
"learning_rate": 9.984326668636131e-05,
"loss": 1.6966455459594727,
"step": 135
},
{
"epoch": 0.05704971475142624,
"grad_norm": 0.254102498292923,
"learning_rate": 9.981607273560785e-05,
"loss": 1.6166696548461914,
"step": 140
},
{
"epoch": 0.05908720456397718,
"grad_norm": 0.24406401813030243,
"learning_rate": 9.978670881475172e-05,
"loss": 1.6438676834106445,
"step": 145
},
{
"epoch": 0.061124694376528114,
"grad_norm": 0.23421962559223175,
"learning_rate": 9.97551762028747e-05,
"loss": 1.6600814819335938,
"step": 150
},
{
"epoch": 0.06316218418907905,
"grad_norm": 0.26790329813957214,
"learning_rate": 9.972147627352592e-05,
"loss": 1.632400894165039,
"step": 155
},
{
"epoch": 0.06519967400162999,
"grad_norm": 0.2244740128517151,
"learning_rate": 9.968561049466214e-05,
"loss": 1.619165802001953,
"step": 160
},
{
"epoch": 0.06723716381418093,
"grad_norm": 0.2931612730026245,
"learning_rate": 9.964758042858367e-05,
"loss": 1.5788841247558594,
"step": 165
},
{
"epoch": 0.06927465362673187,
"grad_norm": 0.22810588777065277,
"learning_rate": 9.960738773186648e-05,
"loss": 1.6591445922851562,
"step": 170
},
{
"epoch": 0.0713121434392828,
"grad_norm": 0.21488972008228302,
"learning_rate": 9.956503415528984e-05,
"loss": 1.6036659240722657,
"step": 175
},
{
"epoch": 0.07334963325183375,
"grad_norm": 0.21304139494895935,
"learning_rate": 9.952052154376026e-05,
"loss": 1.6186206817626954,
"step": 180
},
{
"epoch": 0.07538712306438468,
"grad_norm": 0.23776377737522125,
"learning_rate": 9.947385183623098e-05,
"loss": 1.6009157180786133,
"step": 185
},
{
"epoch": 0.07742461287693561,
"grad_norm": 0.19920478761196136,
"learning_rate": 9.942502706561757e-05,
"loss": 1.6162376403808594,
"step": 190
},
{
"epoch": 0.07946210268948656,
"grad_norm": 0.239252507686615,
"learning_rate": 9.937404935870938e-05,
"loss": 1.5773283958435058,
"step": 195
},
{
"epoch": 0.08149959250203749,
"grad_norm": 0.22820070385932922,
"learning_rate": 9.932092093607686e-05,
"loss": 1.5832679748535157,
"step": 200
},
{
"epoch": 0.08353708231458842,
"grad_norm": 0.25571751594543457,
"learning_rate": 9.926564411197488e-05,
"loss": 1.5769929885864258,
"step": 205
},
{
"epoch": 0.08557457212713937,
"grad_norm": 0.20927836000919342,
"learning_rate": 9.92082212942419e-05,
"loss": 1.5967325210571288,
"step": 210
},
{
"epoch": 0.0876120619396903,
"grad_norm": 0.22983776032924652,
"learning_rate": 9.91486549841951e-05,
"loss": 1.5639275550842284,
"step": 215
},
{
"epoch": 0.08964955175224124,
"grad_norm": 0.24595195055007935,
"learning_rate": 9.908694777652136e-05,
"loss": 1.5760770797729493,
"step": 220
},
{
"epoch": 0.09168704156479218,
"grad_norm": 0.26048213243484497,
"learning_rate": 9.902310235916435e-05,
"loss": 1.5929807662963866,
"step": 225
},
{
"epoch": 0.09372453137734311,
"grad_norm": 0.2384990155696869,
"learning_rate": 9.89571215132073e-05,
"loss": 1.5840772628784179,
"step": 230
},
{
"epoch": 0.09576202118989405,
"grad_norm": 0.2705344557762146,
"learning_rate": 9.888900811275204e-05,
"loss": 1.549574851989746,
"step": 235
},
{
"epoch": 0.097799511002445,
"grad_norm": 0.28348860144615173,
"learning_rate": 9.881876512479362e-05,
"loss": 1.5107032775878906,
"step": 240
},
{
"epoch": 0.09983700081499593,
"grad_norm": 0.2642553448677063,
"learning_rate": 9.874639560909117e-05,
"loss": 1.5189391136169434,
"step": 245
},
{
"epoch": 0.10187449062754686,
"grad_norm": 0.24302583932876587,
"learning_rate": 9.867190271803465e-05,
"loss": 1.5314508438110352,
"step": 250
},
{
"epoch": 0.1039119804400978,
"grad_norm": 0.24924422800540924,
"learning_rate": 9.859528969650738e-05,
"loss": 1.5717763900756836,
"step": 255
},
{
"epoch": 0.10594947025264874,
"grad_norm": 0.23828646540641785,
"learning_rate": 9.85165598817449e-05,
"loss": 1.545862865447998,
"step": 260
},
{
"epoch": 0.10798696006519967,
"grad_norm": 0.2610020935535431,
"learning_rate": 9.843571670318943e-05,
"loss": 1.5171467781066894,
"step": 265
},
{
"epoch": 0.1100244498777506,
"grad_norm": 0.25594744086265564,
"learning_rate": 9.835276368234059e-05,
"loss": 1.5452391624450683,
"step": 270
},
{
"epoch": 0.11206193969030155,
"grad_norm": 0.29088741540908813,
"learning_rate": 9.826770443260193e-05,
"loss": 1.4986861228942872,
"step": 275
},
{
"epoch": 0.11409942950285248,
"grad_norm": 0.234932079911232,
"learning_rate": 9.818054265912362e-05,
"loss": 1.544328498840332,
"step": 280
},
{
"epoch": 0.11613691931540342,
"grad_norm": 0.23521235585212708,
"learning_rate": 9.809128215864097e-05,
"loss": 1.5823867797851563,
"step": 285
},
{
"epoch": 0.11817440912795436,
"grad_norm": 0.24037106335163116,
"learning_rate": 9.799992681930905e-05,
"loss": 1.5799882888793946,
"step": 290
},
{
"epoch": 0.1202118989405053,
"grad_norm": 0.25879165530204773,
"learning_rate": 9.79064806205334e-05,
"loss": 1.5719776153564453,
"step": 295
},
{
"epoch": 0.12224938875305623,
"grad_norm": 0.24546164274215698,
"learning_rate": 9.781094763279666e-05,
"loss": 1.5144448280334473,
"step": 300
},
{
"epoch": 0.12428687856560718,
"grad_norm": 0.2443576604127884,
"learning_rate": 9.771333201748117e-05,
"loss": 1.5095788955688476,
"step": 305
},
{
"epoch": 0.1263243683781581,
"grad_norm": 0.2535657286643982,
"learning_rate": 9.76136380266878e-05,
"loss": 1.5216328620910644,
"step": 310
},
{
"epoch": 0.12836185819070906,
"grad_norm": 0.24777120351791382,
"learning_rate": 9.751187000305076e-05,
"loss": 1.5501005172729492,
"step": 315
},
{
"epoch": 0.13039934800325997,
"grad_norm": 0.2576065957546234,
"learning_rate": 9.74080323795483e-05,
"loss": 1.5214624404907227,
"step": 320
},
{
"epoch": 0.13243683781581092,
"grad_norm": 0.2413678616285324,
"learning_rate": 9.730212967930974e-05,
"loss": 1.5110502243041992,
"step": 325
},
{
"epoch": 0.13447432762836187,
"grad_norm": 0.30750271677970886,
"learning_rate": 9.719416651541839e-05,
"loss": 1.5003343582153321,
"step": 330
},
{
"epoch": 0.13651181744091279,
"grad_norm": 0.24753299355506897,
"learning_rate": 9.708414759071059e-05,
"loss": 1.5050297737121583,
"step": 335
},
{
"epoch": 0.13854930725346373,
"grad_norm": 0.24859397113323212,
"learning_rate": 9.697207769757089e-05,
"loss": 1.5112569808959961,
"step": 340
},
{
"epoch": 0.14058679706601468,
"grad_norm": 0.24517805874347687,
"learning_rate": 9.685796171772328e-05,
"loss": 1.5051301002502442,
"step": 345
},
{
"epoch": 0.1426242868785656,
"grad_norm": 0.23050017654895782,
"learning_rate": 9.674180462201856e-05,
"loss": 1.5167075157165528,
"step": 350
},
{
"epoch": 0.14466177669111654,
"grad_norm": 0.261944442987442,
"learning_rate": 9.662361147021779e-05,
"loss": 1.476831817626953,
"step": 355
},
{
"epoch": 0.1466992665036675,
"grad_norm": 0.28404876589775085,
"learning_rate": 9.65033874107719e-05,
"loss": 1.5023907661437987,
"step": 360
},
{
"epoch": 0.1487367563162184,
"grad_norm": 0.27228549122810364,
"learning_rate": 9.63811376805974e-05,
"loss": 1.51568660736084,
"step": 365
},
{
"epoch": 0.15077424612876936,
"grad_norm": 0.2329675555229187,
"learning_rate": 9.62568676048483e-05,
"loss": 1.4784177780151366,
"step": 370
},
{
"epoch": 0.1528117359413203,
"grad_norm": 0.2935916483402252,
"learning_rate": 9.613058259668416e-05,
"loss": 1.5028112411499024,
"step": 375
},
{
"epoch": 0.15484922575387122,
"grad_norm": 0.24432611465454102,
"learning_rate": 9.60022881570342e-05,
"loss": 1.5086502075195312,
"step": 380
},
{
"epoch": 0.15688671556642217,
"grad_norm": 0.26738467812538147,
"learning_rate": 9.587198987435782e-05,
"loss": 1.543716812133789,
"step": 385
},
{
"epoch": 0.15892420537897312,
"grad_norm": 0.2604532241821289,
"learning_rate": 9.573969342440106e-05,
"loss": 1.4996532440185546,
"step": 390
},
{
"epoch": 0.16096169519152403,
"grad_norm": 0.2646086513996124,
"learning_rate": 9.56054045699494e-05,
"loss": 1.5080904960632324,
"step": 395
},
{
"epoch": 0.16299918500407498,
"grad_norm": 0.2522474527359009,
"learning_rate": 9.546912916057676e-05,
"loss": 1.4308866500854491,
"step": 400
},
{
"epoch": 0.16503667481662593,
"grad_norm": 0.26946988701820374,
"learning_rate": 9.533087313239064e-05,
"loss": 1.4798666954040527,
"step": 405
},
{
"epoch": 0.16707416462917685,
"grad_norm": 0.2491266131401062,
"learning_rate": 9.51906425077736e-05,
"loss": 1.515095043182373,
"step": 410
},
{
"epoch": 0.1691116544417278,
"grad_norm": 0.25281429290771484,
"learning_rate": 9.504844339512095e-05,
"loss": 1.463273048400879,
"step": 415
},
{
"epoch": 0.17114914425427874,
"grad_norm": 0.32062724232673645,
"learning_rate": 9.490428198857457e-05,
"loss": 1.4602070808410645,
"step": 420
},
{
"epoch": 0.17318663406682966,
"grad_norm": 0.2846399247646332,
"learning_rate": 9.475816456775313e-05,
"loss": 1.4642086029052734,
"step": 425
},
{
"epoch": 0.1752241238793806,
"grad_norm": 0.2705329656600952,
"learning_rate": 9.461009749747863e-05,
"loss": 1.475681686401367,
"step": 430
},
{
"epoch": 0.17726161369193155,
"grad_norm": 0.2681976854801178,
"learning_rate": 9.446008722749905e-05,
"loss": 1.4666956901550292,
"step": 435
},
{
"epoch": 0.17929910350448247,
"grad_norm": 0.2682907283306122,
"learning_rate": 9.430814029220753e-05,
"loss": 1.468404769897461,
"step": 440
},
{
"epoch": 0.18133659331703342,
"grad_norm": 0.28875669836997986,
"learning_rate": 9.415426331035753e-05,
"loss": 1.4713603973388671,
"step": 445
},
{
"epoch": 0.18337408312958436,
"grad_norm": 0.2772611975669861,
"learning_rate": 9.399846298477474e-05,
"loss": 1.4709205627441406,
"step": 450
},
{
"epoch": 0.18541157294213528,
"grad_norm": 0.2741967439651489,
"learning_rate": 9.384074610206495e-05,
"loss": 1.4992780685424805,
"step": 455
},
{
"epoch": 0.18744906275468623,
"grad_norm": 0.23307016491889954,
"learning_rate": 9.368111953231848e-05,
"loss": 1.4394451141357423,
"step": 460
},
{
"epoch": 0.18948655256723718,
"grad_norm": 0.2436610758304596,
"learning_rate": 9.351959022881099e-05,
"loss": 1.4031631469726562,
"step": 465
},
{
"epoch": 0.1915240423797881,
"grad_norm": 0.2648111879825592,
"learning_rate": 9.335616522770045e-05,
"loss": 1.43841552734375,
"step": 470
},
{
"epoch": 0.19356153219233904,
"grad_norm": 0.2542089521884918,
"learning_rate": 9.319085164772082e-05,
"loss": 1.4984153747558593,
"step": 475
},
{
"epoch": 0.19559902200489,
"grad_norm": 0.27864205837249756,
"learning_rate": 9.302365668987179e-05,
"loss": 1.4818967819213866,
"step": 480
},
{
"epoch": 0.1976365118174409,
"grad_norm": 0.2655271887779236,
"learning_rate": 9.285458763710524e-05,
"loss": 1.4170777320861816,
"step": 485
},
{
"epoch": 0.19967400162999185,
"grad_norm": 0.2362072765827179,
"learning_rate": 9.268365185400793e-05,
"loss": 1.4463081359863281,
"step": 490
},
{
"epoch": 0.2017114914425428,
"grad_norm": 0.24712727963924408,
"learning_rate": 9.251085678648072e-05,
"loss": 1.389777946472168,
"step": 495
},
{
"epoch": 0.20374898125509372,
"grad_norm": 0.24617959558963776,
"learning_rate": 9.233620996141421e-05,
"loss": 1.4290151596069336,
"step": 500
},
{
"epoch": 0.20578647106764467,
"grad_norm": 0.31161096692085266,
"learning_rate": 9.215971898636094e-05,
"loss": 1.3988615036010743,
"step": 505
},
{
"epoch": 0.2078239608801956,
"grad_norm": 0.25244295597076416,
"learning_rate": 9.198139154920389e-05,
"loss": 1.455789279937744,
"step": 510
},
{
"epoch": 0.20986145069274653,
"grad_norm": 0.2520352602005005,
"learning_rate": 9.180123541782171e-05,
"loss": 1.4641700744628907,
"step": 515
},
{
"epoch": 0.21189894050529748,
"grad_norm": 0.2115061730146408,
"learning_rate": 9.161925843975028e-05,
"loss": 1.44570894241333,
"step": 520
},
{
"epoch": 0.2139364303178484,
"grad_norm": 0.2903922200202942,
"learning_rate": 9.143546854184094e-05,
"loss": 1.4460723876953125,
"step": 525
},
{
"epoch": 0.21597392013039934,
"grad_norm": 0.24372270703315735,
"learning_rate": 9.124987372991511e-05,
"loss": 1.4678077697753906,
"step": 530
},
{
"epoch": 0.2180114099429503,
"grad_norm": 0.26831796765327454,
"learning_rate": 9.106248208841569e-05,
"loss": 1.4859763145446778,
"step": 535
},
{
"epoch": 0.2200488997555012,
"grad_norm": 0.2345837950706482,
"learning_rate": 9.087330178005472e-05,
"loss": 1.4548344612121582,
"step": 540
},
{
"epoch": 0.22208638956805216,
"grad_norm": 0.25100767612457275,
"learning_rate": 9.068234104545801e-05,
"loss": 1.414707851409912,
"step": 545
},
{
"epoch": 0.2241238793806031,
"grad_norm": 0.2393769770860672,
"learning_rate": 9.048960820280605e-05,
"loss": 1.4750711441040039,
"step": 550
},
{
"epoch": 0.22616136919315402,
"grad_norm": 0.22969071567058563,
"learning_rate": 9.029511164747175e-05,
"loss": 1.4421704292297364,
"step": 555
},
{
"epoch": 0.22819885900570497,
"grad_norm": 0.2723912298679352,
"learning_rate": 9.009885985165464e-05,
"loss": 1.4354835510253907,
"step": 560
},
{
"epoch": 0.23023634881825591,
"grad_norm": 0.23489715158939362,
"learning_rate": 8.9900861364012e-05,
"loss": 1.445290470123291,
"step": 565
},
{
"epoch": 0.23227383863080683,
"grad_norm": 0.2519027888774872,
"learning_rate": 8.970112480928621e-05,
"loss": 1.4208389282226563,
"step": 570
},
{
"epoch": 0.23431132844335778,
"grad_norm": 0.2978993356227875,
"learning_rate": 8.949965888792941e-05,
"loss": 1.3885160446166993,
"step": 575
},
{
"epoch": 0.23634881825590873,
"grad_norm": 0.22895459830760956,
"learning_rate": 8.929647237572418e-05,
"loss": 1.4030443191528321,
"step": 580
},
{
"epoch": 0.23838630806845965,
"grad_norm": 0.2571552097797394,
"learning_rate": 8.90915741234015e-05,
"loss": 1.4485620498657226,
"step": 585
},
{
"epoch": 0.2404237978810106,
"grad_norm": 0.2530490458011627,
"learning_rate": 8.888497305625508e-05,
"loss": 1.397987461090088,
"step": 590
},
{
"epoch": 0.24246128769356154,
"grad_norm": 0.25649017095565796,
"learning_rate": 8.867667817375266e-05,
"loss": 1.4218891143798829,
"step": 595
},
{
"epoch": 0.24449877750611246,
"grad_norm": 0.23270413279533386,
"learning_rate": 8.846669854914396e-05,
"loss": 1.4246532440185546,
"step": 600
},
{
"epoch": 0.2465362673186634,
"grad_norm": 0.2551707625389099,
"learning_rate": 8.825504332906541e-05,
"loss": 1.4406282424926757,
"step": 605
},
{
"epoch": 0.24857375713121435,
"grad_norm": 0.2525927424430847,
"learning_rate": 8.804172173314184e-05,
"loss": 1.3947580337524415,
"step": 610
},
{
"epoch": 0.2506112469437653,
"grad_norm": 0.2448631376028061,
"learning_rate": 8.78267430535848e-05,
"loss": 1.432701301574707,
"step": 615
},
{
"epoch": 0.2526487367563162,
"grad_norm": 0.23594658076763153,
"learning_rate": 8.761011665478774e-05,
"loss": 1.4567941665649413,
"step": 620
},
{
"epoch": 0.25468622656886714,
"grad_norm": 0.25044554471969604,
"learning_rate": 8.739185197291823e-05,
"loss": 1.3989928245544434,
"step": 625
},
{
"epoch": 0.2567237163814181,
"grad_norm": 0.23957082629203796,
"learning_rate": 8.717195851550682e-05,
"loss": 1.4815309524536133,
"step": 630
},
{
"epoch": 0.25876120619396903,
"grad_norm": 0.29432135820388794,
"learning_rate": 8.695044586103296e-05,
"loss": 1.4535370826721192,
"step": 635
},
{
"epoch": 0.26079869600651995,
"grad_norm": 0.2810392379760742,
"learning_rate": 8.672732365850774e-05,
"loss": 1.4214276313781737,
"step": 640
},
{
"epoch": 0.2628361858190709,
"grad_norm": 0.27764493227005005,
"learning_rate": 8.650260162705351e-05,
"loss": 1.3979911804199219,
"step": 645
},
{
"epoch": 0.26487367563162184,
"grad_norm": 0.28854900598526,
"learning_rate": 8.62762895554807e-05,
"loss": 1.3852861404418946,
"step": 650
},
{
"epoch": 0.26691116544417276,
"grad_norm": 0.2634202241897583,
"learning_rate": 8.604839730186125e-05,
"loss": 1.3901723861694335,
"step": 655
},
{
"epoch": 0.26894865525672373,
"grad_norm": 0.24707822501659393,
"learning_rate": 8.581893479309925e-05,
"loss": 1.4080962181091308,
"step": 660
},
{
"epoch": 0.27098614506927465,
"grad_norm": 0.267046719789505,
"learning_rate": 8.558791202449856e-05,
"loss": 1.420369529724121,
"step": 665
},
{
"epoch": 0.27302363488182557,
"grad_norm": 0.24895347654819489,
"learning_rate": 8.535533905932738e-05,
"loss": 1.4124841690063477,
"step": 670
},
{
"epoch": 0.27506112469437655,
"grad_norm": 0.2543014883995056,
"learning_rate": 8.512122602837993e-05,
"loss": 1.3869916915893554,
"step": 675
},
{
"epoch": 0.27709861450692747,
"grad_norm": 0.2594490647315979,
"learning_rate": 8.488558312953512e-05,
"loss": 1.4187768936157226,
"step": 680
},
{
"epoch": 0.2791361043194784,
"grad_norm": 0.2667585015296936,
"learning_rate": 8.464842062731235e-05,
"loss": 1.360099697113037,
"step": 685
},
{
"epoch": 0.28117359413202936,
"grad_norm": 0.266170859336853,
"learning_rate": 8.44097488524244e-05,
"loss": 1.4031505584716797,
"step": 690
},
{
"epoch": 0.2832110839445803,
"grad_norm": 0.2641969323158264,
"learning_rate": 8.416957820132742e-05,
"loss": 1.4196790695190429,
"step": 695
},
{
"epoch": 0.2852485737571312,
"grad_norm": 0.23064430058002472,
"learning_rate": 8.392791913576809e-05,
"loss": 1.3981475830078125,
"step": 700
},
{
"epoch": 0.28728606356968217,
"grad_norm": 0.24963510036468506,
"learning_rate": 8.368478218232787e-05,
"loss": 1.3790153503417968,
"step": 705
},
{
"epoch": 0.2893235533822331,
"grad_norm": 0.26175132393836975,
"learning_rate": 8.344017793196442e-05,
"loss": 1.3812166213989259,
"step": 710
},
{
"epoch": 0.291361043194784,
"grad_norm": 0.2446986883878708,
"learning_rate": 8.319411703955042e-05,
"loss": 1.427156639099121,
"step": 715
},
{
"epoch": 0.293398533007335,
"grad_norm": 0.24531200528144836,
"learning_rate": 8.294661022340931e-05,
"loss": 1.3718086242675782,
"step": 720
},
{
"epoch": 0.2954360228198859,
"grad_norm": 0.2547568678855896,
"learning_rate": 8.269766826484842e-05,
"loss": 1.3975444793701173,
"step": 725
},
{
"epoch": 0.2974735126324368,
"grad_norm": 0.24539630115032196,
"learning_rate": 8.244730200768937e-05,
"loss": 1.4228199005126954,
"step": 730
},
{
"epoch": 0.2995110024449878,
"grad_norm": 0.2306804060935974,
"learning_rate": 8.219552235779578e-05,
"loss": 1.3604663848876952,
"step": 735
},
{
"epoch": 0.3015484922575387,
"grad_norm": 0.26924288272857666,
"learning_rate": 8.194234028259806e-05,
"loss": 1.3662386894226075,
"step": 740
},
{
"epoch": 0.30358598207008963,
"grad_norm": 0.3148330748081207,
"learning_rate": 8.168776681061582e-05,
"loss": 1.3941230773925781,
"step": 745
},
{
"epoch": 0.3056234718826406,
"grad_norm": 0.3298656642436981,
"learning_rate": 8.143181303097734e-05,
"loss": 1.34095458984375,
"step": 750
},
{
"epoch": 0.3076609616951915,
"grad_norm": 0.26874008774757385,
"learning_rate": 8.117449009293668e-05,
"loss": 1.3636792182922364,
"step": 755
},
{
"epoch": 0.30969845150774244,
"grad_norm": 0.2534666061401367,
"learning_rate": 8.09158092053879e-05,
"loss": 1.427321434020996,
"step": 760
},
{
"epoch": 0.3117359413202934,
"grad_norm": 0.25248315930366516,
"learning_rate": 8.065578163637686e-05,
"loss": 1.4066177368164063,
"step": 765
},
{
"epoch": 0.31377343113284434,
"grad_norm": 0.2795719504356384,
"learning_rate": 8.039441871261038e-05,
"loss": 1.4073759078979493,
"step": 770
},
{
"epoch": 0.31581092094539526,
"grad_norm": 0.25225991010665894,
"learning_rate": 8.013173181896283e-05,
"loss": 1.4331002235412598,
"step": 775
},
{
"epoch": 0.31784841075794623,
"grad_norm": 0.29578742384910583,
"learning_rate": 7.986773239798026e-05,
"loss": 1.3891535758972169,
"step": 780
},
{
"epoch": 0.31988590057049715,
"grad_norm": 0.24367083609104156,
"learning_rate": 7.960243194938192e-05,
"loss": 1.3595949172973634,
"step": 785
},
{
"epoch": 0.32192339038304807,
"grad_norm": 0.2300739288330078,
"learning_rate": 7.933584202955934e-05,
"loss": 1.3890039443969726,
"step": 790
},
{
"epoch": 0.32396088019559904,
"grad_norm": 0.29280030727386475,
"learning_rate": 7.9067974251073e-05,
"loss": 1.4000889778137207,
"step": 795
},
{
"epoch": 0.32599837000814996,
"grad_norm": 0.2583927810192108,
"learning_rate": 7.879884028214637e-05,
"loss": 1.3617567062377929,
"step": 800
},
{
"epoch": 0.3280358598207009,
"grad_norm": 0.2506610155105591,
"learning_rate": 7.852845184615776e-05,
"loss": 1.3741163253784179,
"step": 805
},
{
"epoch": 0.33007334963325186,
"grad_norm": 0.28492823243141174,
"learning_rate": 7.82568207211296e-05,
"loss": 1.3677806854248047,
"step": 810
},
{
"epoch": 0.3321108394458028,
"grad_norm": 0.28276413679122925,
"learning_rate": 7.79839587392154e-05,
"loss": 1.3788800239562988,
"step": 815
},
{
"epoch": 0.3341483292583537,
"grad_norm": 0.2577153146266937,
"learning_rate": 7.770987778618441e-05,
"loss": 1.4028077125549316,
"step": 820
},
{
"epoch": 0.33618581907090467,
"grad_norm": 0.260003924369812,
"learning_rate": 7.74345898009037e-05,
"loss": 1.3641690254211425,
"step": 825
},
{
"epoch": 0.3382233088834556,
"grad_norm": 0.28759846091270447,
"learning_rate": 7.715810677481834e-05,
"loss": 1.3745591163635253,
"step": 830
},
{
"epoch": 0.3402607986960065,
"grad_norm": 0.2516504228115082,
"learning_rate": 7.688044075142887e-05,
"loss": 1.3658802032470703,
"step": 835
},
{
"epoch": 0.3422982885085575,
"grad_norm": 0.2602832615375519,
"learning_rate": 7.660160382576683e-05,
"loss": 1.3842742919921875,
"step": 840
},
{
"epoch": 0.3443357783211084,
"grad_norm": 0.2655702829360962,
"learning_rate": 7.63216081438678e-05,
"loss": 1.3709933280944824,
"step": 845
},
{
"epoch": 0.3463732681336593,
"grad_norm": 0.29763323068618774,
"learning_rate": 7.604046590224235e-05,
"loss": 1.3394402503967284,
"step": 850
},
{
"epoch": 0.3484107579462103,
"grad_norm": 0.2478332668542862,
"learning_rate": 7.57581893473448e-05,
"loss": 1.356826400756836,
"step": 855
},
{
"epoch": 0.3504482477587612,
"grad_norm": 0.27973031997680664,
"learning_rate": 7.547479077503976e-05,
"loss": 1.3851838111877441,
"step": 860
},
{
"epoch": 0.35248573757131213,
"grad_norm": 0.25085529685020447,
"learning_rate": 7.519028253006648e-05,
"loss": 1.3550750732421875,
"step": 865
},
{
"epoch": 0.3545232273838631,
"grad_norm": 0.24575817584991455,
"learning_rate": 7.490467700550117e-05,
"loss": 1.3255870819091797,
"step": 870
},
{
"epoch": 0.356560717196414,
"grad_norm": 0.28239473700523376,
"learning_rate": 7.461798664221711e-05,
"loss": 1.3608011245727538,
"step": 875
},
{
"epoch": 0.35859820700896494,
"grad_norm": 0.24126704037189484,
"learning_rate": 7.433022392834282e-05,
"loss": 1.3341035842895508,
"step": 880
},
{
"epoch": 0.3606356968215159,
"grad_norm": 0.28654253482818604,
"learning_rate": 7.404140139871797e-05,
"loss": 1.3410005569458008,
"step": 885
},
{
"epoch": 0.36267318663406684,
"grad_norm": 0.2522304952144623,
"learning_rate": 7.375153163434746e-05,
"loss": 1.3798066139221192,
"step": 890
},
{
"epoch": 0.36471067644661775,
"grad_norm": 0.2636043429374695,
"learning_rate": 7.346062726185332e-05,
"loss": 1.370442771911621,
"step": 895
},
{
"epoch": 0.36674816625916873,
"grad_norm": 0.27951088547706604,
"learning_rate": 7.316870095292476e-05,
"loss": 1.4128787040710449,
"step": 900
},
{
"epoch": 0.36878565607171965,
"grad_norm": 0.22582846879959106,
"learning_rate": 7.287576542376617e-05,
"loss": 1.36837158203125,
"step": 905
},
{
"epoch": 0.37082314588427057,
"grad_norm": 0.2610475718975067,
"learning_rate": 7.258183343454318e-05,
"loss": 1.3420948028564452,
"step": 910
},
{
"epoch": 0.37286063569682154,
"grad_norm": 0.30103057622909546,
"learning_rate": 7.228691778882693e-05,
"loss": 1.3916706085205077,
"step": 915
},
{
"epoch": 0.37489812550937246,
"grad_norm": 0.24715907871723175,
"learning_rate": 7.199103133303618e-05,
"loss": 1.3638740539550782,
"step": 920
},
{
"epoch": 0.3769356153219234,
"grad_norm": 0.24132998287677765,
"learning_rate": 7.169418695587791e-05,
"loss": 1.370579719543457,
"step": 925
},
{
"epoch": 0.37897310513447435,
"grad_norm": 0.2682572603225708,
"learning_rate": 7.139639758778578e-05,
"loss": 1.3701016426086425,
"step": 930
},
{
"epoch": 0.38101059494702527,
"grad_norm": 0.22371429204940796,
"learning_rate": 7.109767620035689e-05,
"loss": 1.3316537857055664,
"step": 935
},
{
"epoch": 0.3830480847595762,
"grad_norm": 0.2753354012966156,
"learning_rate": 7.079803580578681e-05,
"loss": 1.3701014518737793,
"step": 940
},
{
"epoch": 0.38508557457212717,
"grad_norm": 0.20844975113868713,
"learning_rate": 7.049748945630269e-05,
"loss": 1.3398615837097168,
"step": 945
},
{
"epoch": 0.3871230643846781,
"grad_norm": 0.2711009979248047,
"learning_rate": 7.019605024359474e-05,
"loss": 1.3674811363220214,
"step": 950
},
{
"epoch": 0.389160554197229,
"grad_norm": 0.304538369178772,
"learning_rate": 6.989373129824604e-05,
"loss": 1.3223135948181153,
"step": 955
},
{
"epoch": 0.39119804400978,
"grad_norm": 0.2700227200984955,
"learning_rate": 6.959054578916044e-05,
"loss": 1.3195343017578125,
"step": 960
},
{
"epoch": 0.3932355338223309,
"grad_norm": 0.2265908122062683,
"learning_rate": 6.928650692298898e-05,
"loss": 1.3556050300598144,
"step": 965
},
{
"epoch": 0.3952730236348818,
"grad_norm": 0.2939392924308777,
"learning_rate": 6.89816279435547e-05,
"loss": 1.328925323486328,
"step": 970
},
{
"epoch": 0.3973105134474328,
"grad_norm": 0.2593058943748474,
"learning_rate": 6.867592213127558e-05,
"loss": 1.3408660888671875,
"step": 975
},
{
"epoch": 0.3993480032599837,
"grad_norm": 0.23454953730106354,
"learning_rate": 6.836940280258623e-05,
"loss": 1.3515377044677734,
"step": 980
},
{
"epoch": 0.4013854930725346,
"grad_norm": 0.2662452161312103,
"learning_rate": 6.806208330935766e-05,
"loss": 1.3709604263305664,
"step": 985
},
{
"epoch": 0.4034229828850856,
"grad_norm": 0.26926830410957336,
"learning_rate": 6.775397703831579e-05,
"loss": 1.3455933570861816,
"step": 990
},
{
"epoch": 0.4054604726976365,
"grad_norm": 0.22434349358081818,
"learning_rate": 6.744509741045835e-05,
"loss": 1.3225156784057617,
"step": 995
},
{
"epoch": 0.40749796251018744,
"grad_norm": 0.23916694521903992,
"learning_rate": 6.713545788047011e-05,
"loss": 1.3902326583862306,
"step": 1000
},
{
"epoch": 0.4095354523227384,
"grad_norm": 0.25313499569892883,
"learning_rate": 6.682507193613696e-05,
"loss": 1.3995287895202637,
"step": 1005
},
{
"epoch": 0.41157294213528933,
"grad_norm": 0.2778032124042511,
"learning_rate": 6.651395309775837e-05,
"loss": 1.3518735885620117,
"step": 1010
},
{
"epoch": 0.41361043194784025,
"grad_norm": 0.280178964138031,
"learning_rate": 6.62021149175583e-05,
"loss": 1.327726173400879,
"step": 1015
},
{
"epoch": 0.4156479217603912,
"grad_norm": 0.24925434589385986,
"learning_rate": 6.588957097909508e-05,
"loss": 1.317696762084961,
"step": 1020
},
{
"epoch": 0.41768541157294214,
"grad_norm": 0.26824769377708435,
"learning_rate": 6.557633489666958e-05,
"loss": 1.3888073921203614,
"step": 1025
},
{
"epoch": 0.41972290138549306,
"grad_norm": 0.24167698621749878,
"learning_rate": 6.52624203147322e-05,
"loss": 1.329982376098633,
"step": 1030
},
{
"epoch": 0.421760391198044,
"grad_norm": 0.3196124732494354,
"learning_rate": 6.494784090728852e-05,
"loss": 1.329521656036377,
"step": 1035
},
{
"epoch": 0.42379788101059496,
"grad_norm": 0.25458118319511414,
"learning_rate": 6.463261037730373e-05,
"loss": 1.3205717086791993,
"step": 1040
},
{
"epoch": 0.4258353708231459,
"grad_norm": 0.23451414704322815,
"learning_rate": 6.431674245610564e-05,
"loss": 1.3117390632629395,
"step": 1045
},
{
"epoch": 0.4278728606356968,
"grad_norm": 0.25196176767349243,
"learning_rate": 6.400025090278661e-05,
"loss": 1.31165771484375,
"step": 1050
},
{
"epoch": 0.42991035044824777,
"grad_norm": 0.29432404041290283,
"learning_rate": 6.368314950360415e-05,
"loss": 1.336604118347168,
"step": 1055
},
{
"epoch": 0.4319478402607987,
"grad_norm": 0.27889886498451233,
"learning_rate": 6.336545207138051e-05,
"loss": 1.279319953918457,
"step": 1060
},
{
"epoch": 0.4339853300733496,
"grad_norm": 0.2505008280277252,
"learning_rate": 6.304717244490084e-05,
"loss": 1.3398141860961914,
"step": 1065
},
{
"epoch": 0.4360228198859006,
"grad_norm": 0.30236688256263733,
"learning_rate": 6.272832448831054e-05,
"loss": 1.3090019226074219,
"step": 1070
},
{
"epoch": 0.4380603096984515,
"grad_norm": 0.2615617513656616,
"learning_rate": 6.240892209051121e-05,
"loss": 1.3184801101684571,
"step": 1075
},
{
"epoch": 0.4400977995110024,
"grad_norm": 0.28324833512306213,
"learning_rate": 6.20889791645557e-05,
"loss": 1.3101490974426269,
"step": 1080
},
{
"epoch": 0.4421352893235534,
"grad_norm": 0.2482040673494339,
"learning_rate": 6.176850964704213e-05,
"loss": 1.317474937438965,
"step": 1085
},
{
"epoch": 0.4441727791361043,
"grad_norm": 0.28971773386001587,
"learning_rate": 6.14475274975067e-05,
"loss": 1.3366613388061523,
"step": 1090
},
{
"epoch": 0.44621026894865523,
"grad_norm": 0.26008617877960205,
"learning_rate": 6.112604669781572e-05,
"loss": 1.3874540328979492,
"step": 1095
},
{
"epoch": 0.4482477587612062,
"grad_norm": 0.22775253653526306,
"learning_rate": 6.080408125155649e-05,
"loss": 1.3435938835144043,
"step": 1100
},
{
"epoch": 0.4502852485737571,
"grad_norm": 0.3380416929721832,
"learning_rate": 6.048164518342734e-05,
"loss": 1.2965017318725587,
"step": 1105
},
{
"epoch": 0.45232273838630804,
"grad_norm": 0.2921828329563141,
"learning_rate": 6.015875253862672e-05,
"loss": 1.3140098571777343,
"step": 1110
},
{
"epoch": 0.454360228198859,
"grad_norm": 0.2526835501194,
"learning_rate": 5.983541738224141e-05,
"loss": 1.3064233779907226,
"step": 1115
},
{
"epoch": 0.45639771801140994,
"grad_norm": 0.27323615550994873,
"learning_rate": 5.951165379863382e-05,
"loss": 1.3322813034057617,
"step": 1120
},
{
"epoch": 0.45843520782396086,
"grad_norm": 0.2531745731830597,
"learning_rate": 5.918747589082853e-05,
"loss": 1.3812932968139648,
"step": 1125
},
{
"epoch": 0.46047269763651183,
"grad_norm": 0.2473030537366867,
"learning_rate": 5.886289777989786e-05,
"loss": 1.2970403671264648,
"step": 1130
},
{
"epoch": 0.46251018744906275,
"grad_norm": 0.24724827706813812,
"learning_rate": 5.853793360434687e-05,
"loss": 1.2707000732421876,
"step": 1135
},
{
"epoch": 0.46454767726161367,
"grad_norm": 0.2586457133293152,
"learning_rate": 5.82125975194975e-05,
"loss": 1.3314603805541991,
"step": 1140
},
{
"epoch": 0.46658516707416464,
"grad_norm": 0.29006895422935486,
"learning_rate": 5.788690369687187e-05,
"loss": 1.3309256553649902,
"step": 1145
},
{
"epoch": 0.46862265688671556,
"grad_norm": 0.2463911920785904,
"learning_rate": 5.7560866323575056e-05,
"loss": 1.3553298950195312,
"step": 1150
},
{
"epoch": 0.4706601466992665,
"grad_norm": 0.25358062982559204,
"learning_rate": 5.7234499601677026e-05,
"loss": 1.3103775024414062,
"step": 1155
},
{
"epoch": 0.47269763651181745,
"grad_norm": 0.32736262679100037,
"learning_rate": 5.6907817747594116e-05,
"loss": 1.3686951637268066,
"step": 1160
},
{
"epoch": 0.4747351263243684,
"grad_norm": 0.27510881423950195,
"learning_rate": 5.6580834991469675e-05,
"loss": 1.3144912719726562,
"step": 1165
},
{
"epoch": 0.4767726161369193,
"grad_norm": 0.27266234159469604,
"learning_rate": 5.625356557655426e-05,
"loss": 1.3223678588867187,
"step": 1170
},
{
"epoch": 0.47881010594947027,
"grad_norm": 0.2711781859397888,
"learning_rate": 5.5926023758585146e-05,
"loss": 1.3262757301330566,
"step": 1175
},
{
"epoch": 0.4808475957620212,
"grad_norm": 0.22145110368728638,
"learning_rate": 5.559822380516539e-05,
"loss": 1.3080440521240235,
"step": 1180
},
{
"epoch": 0.4828850855745721,
"grad_norm": 0.2851805090904236,
"learning_rate": 5.527017999514239e-05,
"loss": 1.3531083106994628,
"step": 1185
},
{
"epoch": 0.4849225753871231,
"grad_norm": 0.26763734221458435,
"learning_rate": 5.4941906617985785e-05,
"loss": 1.342519760131836,
"step": 1190
},
{
"epoch": 0.486960065199674,
"grad_norm": 0.35442203283309937,
"learning_rate": 5.4613417973165106e-05,
"loss": 1.292084312438965,
"step": 1195
},
{
"epoch": 0.4889975550122249,
"grad_norm": 0.31827348470687866,
"learning_rate": 5.4284728369526846e-05,
"loss": 1.320939064025879,
"step": 1200
},
{
"epoch": 0.4910350448247759,
"grad_norm": 0.3328326642513275,
"learning_rate": 5.3955852124671234e-05,
"loss": 1.2905807495117188,
"step": 1205
},
{
"epoch": 0.4930725346373268,
"grad_norm": 0.2383362352848053,
"learning_rate": 5.3626803564328465e-05,
"loss": 1.3461034774780274,
"step": 1210
},
{
"epoch": 0.49511002444987773,
"grad_norm": 0.2907769978046417,
"learning_rate": 5.329759702173477e-05,
"loss": 1.360824203491211,
"step": 1215
},
{
"epoch": 0.4971475142624287,
"grad_norm": 0.2850855886936188,
"learning_rate": 5.2968246837008026e-05,
"loss": 1.3173855781555175,
"step": 1220
},
{
"epoch": 0.4991850040749796,
"grad_norm": 0.31968051195144653,
"learning_rate": 5.2638767356523125e-05,
"loss": 1.3406582832336427,
"step": 1225
},
{
"epoch": 0.5012224938875306,
"grad_norm": 0.24894443154335022,
"learning_rate": 5.230917293228699e-05,
"loss": 1.3107667922973634,
"step": 1230
},
{
"epoch": 0.5032599837000815,
"grad_norm": 0.2712059020996094,
"learning_rate": 5.197947792131348e-05,
"loss": 1.349766159057617,
"step": 1235
},
{
"epoch": 0.5052974735126324,
"grad_norm": 0.27703768014907837,
"learning_rate": 5.164969668499798e-05,
"loss": 1.332075309753418,
"step": 1240
},
{
"epoch": 0.5073349633251834,
"grad_norm": 0.27693548798561096,
"learning_rate": 5.131984358849182e-05,
"loss": 1.3212966918945312,
"step": 1245
},
{
"epoch": 0.5093724531377343,
"grad_norm": 0.24810165166854858,
"learning_rate": 5.0989933000076526e-05,
"loss": 1.3448861122131348,
"step": 1250
},
{
"epoch": 0.5114099429502853,
"grad_norm": 0.2883126735687256,
"learning_rate": 5.0659979290537954e-05,
"loss": 1.3556083679199218,
"step": 1255
},
{
"epoch": 0.5134474327628362,
"grad_norm": 0.2394496649503708,
"learning_rate": 5.0329996832540285e-05,
"loss": 1.3140595436096192,
"step": 1260
},
{
"epoch": 0.5154849225753871,
"grad_norm": 0.237855464220047,
"learning_rate": 5e-05,
"loss": 1.3213227272033692,
"step": 1265
},
{
"epoch": 0.5175224123879381,
"grad_norm": 0.25947558879852295,
"learning_rate": 4.967000316745972e-05,
"loss": 1.3036725997924805,
"step": 1270
},
{
"epoch": 0.519559902200489,
"grad_norm": 0.2595300078392029,
"learning_rate": 4.934002070946206e-05,
"loss": 1.3671839714050293,
"step": 1275
},
{
"epoch": 0.5215973920130399,
"grad_norm": 0.2309832125902176,
"learning_rate": 4.901006699992348e-05,
"loss": 1.2795485496520995,
"step": 1280
},
{
"epoch": 0.5236348818255909,
"grad_norm": 0.2816735506057739,
"learning_rate": 4.86801564115082e-05,
"loss": 1.2879011154174804,
"step": 1285
},
{
"epoch": 0.5256723716381418,
"grad_norm": 0.2892402410507202,
"learning_rate": 4.835030331500204e-05,
"loss": 1.2979843139648437,
"step": 1290
},
{
"epoch": 0.5277098614506928,
"grad_norm": 0.2680181860923767,
"learning_rate": 4.802052207868654e-05,
"loss": 1.3553829193115234,
"step": 1295
},
{
"epoch": 0.5297473512632437,
"grad_norm": 0.2874012887477875,
"learning_rate": 4.7690827067713035e-05,
"loss": 1.2482240676879883,
"step": 1300
},
{
"epoch": 0.5317848410757946,
"grad_norm": 0.24542248249053955,
"learning_rate": 4.736123264347688e-05,
"loss": 1.3517400741577148,
"step": 1305
},
{
"epoch": 0.5338223308883455,
"grad_norm": 0.2526733875274658,
"learning_rate": 4.7031753162991966e-05,
"loss": 1.3122305870056152,
"step": 1310
},
{
"epoch": 0.5358598207008965,
"grad_norm": 0.2901546359062195,
"learning_rate": 4.670240297826523e-05,
"loss": 1.332789707183838,
"step": 1315
},
{
"epoch": 0.5378973105134475,
"grad_norm": 0.23953253030776978,
"learning_rate": 4.637319643567154e-05,
"loss": 1.355057716369629,
"step": 1320
},
{
"epoch": 0.5399348003259984,
"grad_norm": 0.29978007078170776,
"learning_rate": 4.604414787532877e-05,
"loss": 1.3209955215454101,
"step": 1325
},
{
"epoch": 0.5419722901385493,
"grad_norm": 0.23924697935581207,
"learning_rate": 4.571527163047316e-05,
"loss": 1.3392011642456054,
"step": 1330
},
{
"epoch": 0.5440097799511002,
"grad_norm": 0.2822034955024719,
"learning_rate": 4.5386582026834906e-05,
"loss": 1.3059310913085938,
"step": 1335
},
{
"epoch": 0.5460472697636511,
"grad_norm": 0.28926926851272583,
"learning_rate": 4.5058093382014226e-05,
"loss": 1.250102424621582,
"step": 1340
},
{
"epoch": 0.5480847595762022,
"grad_norm": 0.33292192220687866,
"learning_rate": 4.472982000485762e-05,
"loss": 1.3017122268676757,
"step": 1345
},
{
"epoch": 0.5501222493887531,
"grad_norm": 0.29768046736717224,
"learning_rate": 4.4401776194834613e-05,
"loss": 1.3372932434082032,
"step": 1350
},
{
"epoch": 0.552159739201304,
"grad_norm": 0.3008410632610321,
"learning_rate": 4.407397624141487e-05,
"loss": 1.3267680168151856,
"step": 1355
},
{
"epoch": 0.5541972290138549,
"grad_norm": 0.22114711999893188,
"learning_rate": 4.374643442344576e-05,
"loss": 1.346128273010254,
"step": 1360
},
{
"epoch": 0.5562347188264058,
"grad_norm": 0.28926345705986023,
"learning_rate": 4.341916500853034e-05,
"loss": 1.3007698059082031,
"step": 1365
},
{
"epoch": 0.5582722086389568,
"grad_norm": 0.2524040937423706,
"learning_rate": 4.30921822524059e-05,
"loss": 1.3036186218261718,
"step": 1370
},
{
"epoch": 0.5603096984515077,
"grad_norm": 0.2812578082084656,
"learning_rate": 4.276550039832299e-05,
"loss": 1.2964239120483398,
"step": 1375
},
{
"epoch": 0.5623471882640587,
"grad_norm": 0.35602179169654846,
"learning_rate": 4.2439133676424955e-05,
"loss": 1.2641495704650878,
"step": 1380
},
{
"epoch": 0.5643846780766096,
"grad_norm": 0.2539024353027344,
"learning_rate": 4.2113096303128125e-05,
"loss": 1.3166223526000977,
"step": 1385
},
{
"epoch": 0.5664221678891606,
"grad_norm": 0.3124350905418396,
"learning_rate": 4.178740248050249e-05,
"loss": 1.3306081771850586,
"step": 1390
},
{
"epoch": 0.5684596577017115,
"grad_norm": 0.2921796441078186,
"learning_rate": 4.146206639565312e-05,
"loss": 1.3311893463134765,
"step": 1395
},
{
"epoch": 0.5704971475142624,
"grad_norm": 0.25485843420028687,
"learning_rate": 4.113710222010215e-05,
"loss": 1.2562400817871093,
"step": 1400
},
{
"epoch": 0.5725346373268133,
"grad_norm": 0.2926226854324341,
"learning_rate": 4.0812524109171476e-05,
"loss": 1.3223559379577636,
"step": 1405
},
{
"epoch": 0.5745721271393643,
"grad_norm": 0.23452463746070862,
"learning_rate": 4.048834620136618e-05,
"loss": 1.31746826171875,
"step": 1410
},
{
"epoch": 0.5766096169519153,
"grad_norm": 0.2829977869987488,
"learning_rate": 4.01645826177586e-05,
"loss": 1.3249542236328125,
"step": 1415
},
{
"epoch": 0.5786471067644662,
"grad_norm": 0.27074241638183594,
"learning_rate": 3.98412474613733e-05,
"loss": 1.235785961151123,
"step": 1420
},
{
"epoch": 0.5806845965770171,
"grad_norm": 0.2690928876399994,
"learning_rate": 3.951835481657268e-05,
"loss": 1.2997233390808105,
"step": 1425
},
{
"epoch": 0.582722086389568,
"grad_norm": 0.28461456298828125,
"learning_rate": 3.919591874844352e-05,
"loss": 1.3361207962036132,
"step": 1430
},
{
"epoch": 0.5847595762021189,
"grad_norm": 0.2862246632575989,
"learning_rate": 3.887395330218429e-05,
"loss": 1.301889991760254,
"step": 1435
},
{
"epoch": 0.58679706601467,
"grad_norm": 0.28518208861351013,
"learning_rate": 3.855247250249331e-05,
"loss": 1.3241774559020996,
"step": 1440
},
{
"epoch": 0.5888345558272209,
"grad_norm": 0.2596955895423889,
"learning_rate": 3.823149035295789e-05,
"loss": 1.3260196685791015,
"step": 1445
},
{
"epoch": 0.5908720456397718,
"grad_norm": 0.2542703449726105,
"learning_rate": 3.791102083544432e-05,
"loss": 1.2976004600524902,
"step": 1450
},
{
"epoch": 0.5929095354523227,
"grad_norm": 0.22579146921634674,
"learning_rate": 3.759107790948882e-05,
"loss": 1.3024371147155762,
"step": 1455
},
{
"epoch": 0.5949470252648736,
"grad_norm": 0.25704026222229004,
"learning_rate": 3.7271675511689475e-05,
"loss": 1.3003257751464843,
"step": 1460
},
{
"epoch": 0.5969845150774246,
"grad_norm": 0.2658834755420685,
"learning_rate": 3.695282755509917e-05,
"loss": 1.32208309173584,
"step": 1465
},
{
"epoch": 0.5990220048899756,
"grad_norm": 0.301856130361557,
"learning_rate": 3.6634547928619505e-05,
"loss": 1.2568475723266601,
"step": 1470
},
{
"epoch": 0.6010594947025265,
"grad_norm": 0.3496600091457367,
"learning_rate": 3.631685049639586e-05,
"loss": 1.28335599899292,
"step": 1475
},
{
"epoch": 0.6030969845150774,
"grad_norm": 0.30457738041877747,
"learning_rate": 3.599974909721341e-05,
"loss": 1.271221923828125,
"step": 1480
},
{
"epoch": 0.6051344743276283,
"grad_norm": 0.29012930393218994,
"learning_rate": 3.568325754389438e-05,
"loss": 1.2871403694152832,
"step": 1485
},
{
"epoch": 0.6071719641401793,
"grad_norm": 0.30023711919784546,
"learning_rate": 3.536738962269628e-05,
"loss": 1.2935770988464355,
"step": 1490
},
{
"epoch": 0.6092094539527302,
"grad_norm": 0.2748446762561798,
"learning_rate": 3.505215909271149e-05,
"loss": 1.2088833808898927,
"step": 1495
},
{
"epoch": 0.6112469437652812,
"grad_norm": 0.28603363037109375,
"learning_rate": 3.4737579685267815e-05,
"loss": 1.3596332550048829,
"step": 1500
},
{
"epoch": 0.6132844335778321,
"grad_norm": 0.2897932529449463,
"learning_rate": 3.442366510333043e-05,
"loss": 1.2609786987304688,
"step": 1505
},
{
"epoch": 0.615321923390383,
"grad_norm": 0.3109928369522095,
"learning_rate": 3.411042902090492e-05,
"loss": 1.2860849380493165,
"step": 1510
},
{
"epoch": 0.617359413202934,
"grad_norm": 0.24203228950500488,
"learning_rate": 3.379788508244171e-05,
"loss": 1.2845419883728026,
"step": 1515
},
{
"epoch": 0.6193969030154849,
"grad_norm": 0.27746057510375977,
"learning_rate": 3.3486046902241664e-05,
"loss": 1.3250798225402831,
"step": 1520
},
{
"epoch": 0.6214343928280358,
"grad_norm": 0.2685863971710205,
"learning_rate": 3.3174928063863055e-05,
"loss": 1.3168251037597656,
"step": 1525
},
{
"epoch": 0.6234718826405868,
"grad_norm": 0.30500712990760803,
"learning_rate": 3.286454211952989e-05,
"loss": 1.275892734527588,
"step": 1530
},
{
"epoch": 0.6255093724531378,
"grad_norm": 0.27914416790008545,
"learning_rate": 3.255490258954167e-05,
"loss": 1.2840620040893556,
"step": 1535
},
{
"epoch": 0.6275468622656887,
"grad_norm": 0.363704115152359,
"learning_rate": 3.2246022961684206e-05,
"loss": 1.2786981582641601,
"step": 1540
},
{
"epoch": 0.6295843520782396,
"grad_norm": 0.24931791424751282,
"learning_rate": 3.1937916690642356e-05,
"loss": 1.276559829711914,
"step": 1545
},
{
"epoch": 0.6316218418907905,
"grad_norm": 0.3054625988006592,
"learning_rate": 3.163059719741378e-05,
"loss": 1.2450448989868164,
"step": 1550
},
{
"epoch": 0.6336593317033414,
"grad_norm": 0.2961163818836212,
"learning_rate": 3.132407786872442e-05,
"loss": 1.2576936721801757,
"step": 1555
},
{
"epoch": 0.6356968215158925,
"grad_norm": 0.24456286430358887,
"learning_rate": 3.101837205644531e-05,
"loss": 1.3349946975708007,
"step": 1560
},
{
"epoch": 0.6377343113284434,
"grad_norm": 0.2532815933227539,
"learning_rate": 3.071349307701103e-05,
"loss": 1.3183099746704101,
"step": 1565
},
{
"epoch": 0.6397718011409943,
"grad_norm": 0.25841763615608215,
"learning_rate": 3.0409454210839577e-05,
"loss": 1.2757043838500977,
"step": 1570
},
{
"epoch": 0.6418092909535452,
"grad_norm": 0.2988349199295044,
"learning_rate": 3.0106268701753965e-05,
"loss": 1.2936763763427734,
"step": 1575
},
{
"epoch": 0.6438467807660961,
"grad_norm": 0.30920735001564026,
"learning_rate": 2.980394975640526e-05,
"loss": 1.2118300437927245,
"step": 1580
},
{
"epoch": 0.6458842705786471,
"grad_norm": 0.2675487995147705,
"learning_rate": 2.9502510543697325e-05,
"loss": 1.315651798248291,
"step": 1585
},
{
"epoch": 0.6479217603911981,
"grad_norm": 0.2591685652732849,
"learning_rate": 2.9201964194213204e-05,
"loss": 1.2554664611816406,
"step": 1590
},
{
"epoch": 0.649959250203749,
"grad_norm": 0.2772226333618164,
"learning_rate": 2.8902323799643116e-05,
"loss": 1.2888887405395508,
"step": 1595
},
{
"epoch": 0.6519967400162999,
"grad_norm": 0.2594970464706421,
"learning_rate": 2.8603602412214226e-05,
"loss": 1.245866298675537,
"step": 1600
},
{
"epoch": 0.6540342298288508,
"grad_norm": 0.2654174566268921,
"learning_rate": 2.8305813044122097e-05,
"loss": 1.2802406311035157,
"step": 1605
},
{
"epoch": 0.6560717196414018,
"grad_norm": 0.30692341923713684,
"learning_rate": 2.800896866696382e-05,
"loss": 1.2570459365844726,
"step": 1610
},
{
"epoch": 0.6581092094539527,
"grad_norm": 0.2689622938632965,
"learning_rate": 2.771308221117309e-05,
"loss": 1.2683204650878905,
"step": 1615
},
{
"epoch": 0.6601466992665037,
"grad_norm": 0.2888513207435608,
"learning_rate": 2.7418166565456815e-05,
"loss": 1.323173713684082,
"step": 1620
},
{
"epoch": 0.6621841890790546,
"grad_norm": 0.2877139151096344,
"learning_rate": 2.7124234576233847e-05,
"loss": 1.3259809494018555,
"step": 1625
},
{
"epoch": 0.6642216788916055,
"grad_norm": 0.24738578498363495,
"learning_rate": 2.6831299047075265e-05,
"loss": 1.2701099395751954,
"step": 1630
},
{
"epoch": 0.6662591687041565,
"grad_norm": 0.23475511372089386,
"learning_rate": 2.6539372738146695e-05,
"loss": 1.2258795738220214,
"step": 1635
},
{
"epoch": 0.6682966585167074,
"grad_norm": 0.2642034888267517,
"learning_rate": 2.6248468365652557e-05,
"loss": 1.2938674926757812,
"step": 1640
},
{
"epoch": 0.6703341483292583,
"grad_norm": 0.25303909182548523,
"learning_rate": 2.5958598601282035e-05,
"loss": 1.2401073455810547,
"step": 1645
},
{
"epoch": 0.6723716381418093,
"grad_norm": 0.2774398624897003,
"learning_rate": 2.5669776071657192e-05,
"loss": 1.2878467559814453,
"step": 1650
},
{
"epoch": 0.6744091279543603,
"grad_norm": 0.2558142840862274,
"learning_rate": 2.5382013357782893e-05,
"loss": 1.2604849815368653,
"step": 1655
},
{
"epoch": 0.6764466177669112,
"grad_norm": 0.2904720604419708,
"learning_rate": 2.509532299449885e-05,
"loss": 1.317274284362793,
"step": 1660
},
{
"epoch": 0.6784841075794621,
"grad_norm": 0.28084811568260193,
"learning_rate": 2.480971746993354e-05,
"loss": 1.272391128540039,
"step": 1665
},
{
"epoch": 0.680521597392013,
"grad_norm": 0.23356030881404877,
"learning_rate": 2.452520922496025e-05,
"loss": 1.2504362106323241,
"step": 1670
},
{
"epoch": 0.6825590872045639,
"grad_norm": 0.3247373700141907,
"learning_rate": 2.4241810652655196e-05,
"loss": 1.2986472129821778,
"step": 1675
},
{
"epoch": 0.684596577017115,
"grad_norm": 0.2745431065559387,
"learning_rate": 2.3959534097757664e-05,
"loss": 1.2283918380737304,
"step": 1680
},
{
"epoch": 0.6866340668296659,
"grad_norm": 0.26617926359176636,
"learning_rate": 2.3678391856132204e-05,
"loss": 1.2940399169921875,
"step": 1685
},
{
"epoch": 0.6886715566422168,
"grad_norm": 0.2801912724971771,
"learning_rate": 2.3398396174233178e-05,
"loss": 1.25750789642334,
"step": 1690
},
{
"epoch": 0.6907090464547677,
"grad_norm": 0.23978249728679657,
"learning_rate": 2.3119559248571128e-05,
"loss": 1.3003442764282227,
"step": 1695
},
{
"epoch": 0.6927465362673186,
"grad_norm": 0.31367555260658264,
"learning_rate": 2.2841893225181676e-05,
"loss": 1.2993706703186034,
"step": 1700
},
{
"epoch": 0.6947840260798696,
"grad_norm": 0.35199978947639465,
"learning_rate": 2.256541019909632e-05,
"loss": 1.264107894897461,
"step": 1705
},
{
"epoch": 0.6968215158924206,
"grad_norm": 0.2840383052825928,
"learning_rate": 2.2290122213815607e-05,
"loss": 1.27420654296875,
"step": 1710
},
{
"epoch": 0.6988590057049715,
"grad_norm": 0.2828579545021057,
"learning_rate": 2.2016041260784605e-05,
"loss": 1.3064224243164062,
"step": 1715
},
{
"epoch": 0.7008964955175224,
"grad_norm": 0.32567906379699707,
"learning_rate": 2.1743179278870407e-05,
"loss": 1.2433448791503907,
"step": 1720
},
{
"epoch": 0.7029339853300733,
"grad_norm": 0.2845408320426941,
"learning_rate": 2.1471548153842258e-05,
"loss": 1.3065531730651856,
"step": 1725
},
{
"epoch": 0.7049714751426243,
"grad_norm": 0.2809593677520752,
"learning_rate": 2.1201159717853637e-05,
"loss": 1.25469913482666,
"step": 1730
},
{
"epoch": 0.7070089649551752,
"grad_norm": 0.29171547293663025,
"learning_rate": 2.0932025748927013e-05,
"loss": 1.271986961364746,
"step": 1735
},
{
"epoch": 0.7090464547677262,
"grad_norm": 0.26354479789733887,
"learning_rate": 2.0664157970440674e-05,
"loss": 1.3261978149414062,
"step": 1740
},
{
"epoch": 0.7110839445802771,
"grad_norm": 0.25119608640670776,
"learning_rate": 2.0397568050618094e-05,
"loss": 1.2475339889526367,
"step": 1745
},
{
"epoch": 0.713121434392828,
"grad_norm": 0.30375808477401733,
"learning_rate": 2.0132267602019743e-05,
"loss": 1.2534145355224608,
"step": 1750
},
{
"epoch": 0.715158924205379,
"grad_norm": 0.3172304034233093,
"learning_rate": 1.9868268181037185e-05,
"loss": 1.2916114807128907,
"step": 1755
},
{
"epoch": 0.7171964140179299,
"grad_norm": 0.2723129689693451,
"learning_rate": 1.9605581287389635e-05,
"loss": 1.323664665222168,
"step": 1760
},
{
"epoch": 0.7192339038304808,
"grad_norm": 0.2418317049741745,
"learning_rate": 1.934421836362315e-05,
"loss": 1.2897408485412598,
"step": 1765
},
{
"epoch": 0.7212713936430318,
"grad_norm": 0.3190702497959137,
"learning_rate": 1.90841907946121e-05,
"loss": 1.2397636413574218,
"step": 1770
},
{
"epoch": 0.7233088834555828,
"grad_norm": 0.2643372416496277,
"learning_rate": 1.8825509907063327e-05,
"loss": 1.2604900360107423,
"step": 1775
},
{
"epoch": 0.7253463732681337,
"grad_norm": 0.27984854578971863,
"learning_rate": 1.8568186969022654e-05,
"loss": 1.24794340133667,
"step": 1780
},
{
"epoch": 0.7273838630806846,
"grad_norm": 0.26192235946655273,
"learning_rate": 1.831223318938419e-05,
"loss": 1.2820892333984375,
"step": 1785
},
{
"epoch": 0.7294213528932355,
"grad_norm": 0.34071704745292664,
"learning_rate": 1.8057659717401947e-05,
"loss": 1.3241848945617676,
"step": 1790
},
{
"epoch": 0.7314588427057864,
"grad_norm": 0.2529299557209015,
"learning_rate": 1.7804477642204222e-05,
"loss": 1.2486754417419434,
"step": 1795
},
{
"epoch": 0.7334963325183375,
"grad_norm": 0.3277510106563568,
"learning_rate": 1.7552697992310635e-05,
"loss": 1.2531171798706056,
"step": 1800
},
{
"epoch": 0.7355338223308884,
"grad_norm": 0.28675028681755066,
"learning_rate": 1.7302331735151594e-05,
"loss": 1.3075594902038574,
"step": 1805
},
{
"epoch": 0.7375713121434393,
"grad_norm": 0.25726768374443054,
"learning_rate": 1.705338977659071e-05,
"loss": 1.2466166496276856,
"step": 1810
},
{
"epoch": 0.7396088019559902,
"grad_norm": 0.2569176256656647,
"learning_rate": 1.6805882960449594e-05,
"loss": 1.327796459197998,
"step": 1815
},
{
"epoch": 0.7416462917685411,
"grad_norm": 0.26824307441711426,
"learning_rate": 1.6559822068035584e-05,
"loss": 1.2975038528442382,
"step": 1820
},
{
"epoch": 0.743683781581092,
"grad_norm": 0.26288747787475586,
"learning_rate": 1.631521781767214e-05,
"loss": 1.2798389434814452,
"step": 1825
},
{
"epoch": 0.7457212713936431,
"grad_norm": 0.27184709906578064,
"learning_rate": 1.6072080864231904e-05,
"loss": 1.263510036468506,
"step": 1830
},
{
"epoch": 0.747758761206194,
"grad_norm": 0.2663553059101105,
"learning_rate": 1.5830421798672568e-05,
"loss": 1.322945785522461,
"step": 1835
},
{
"epoch": 0.7497962510187449,
"grad_norm": 0.25886112451553345,
"learning_rate": 1.5590251147575608e-05,
"loss": 1.2786243438720704,
"step": 1840
},
{
"epoch": 0.7518337408312958,
"grad_norm": 0.27063053846359253,
"learning_rate": 1.535157937268766e-05,
"loss": 1.263995361328125,
"step": 1845
},
{
"epoch": 0.7538712306438468,
"grad_norm": 0.25247734785079956,
"learning_rate": 1.5114416870464898e-05,
"loss": 1.2966327667236328,
"step": 1850
},
{
"epoch": 0.7559087204563977,
"grad_norm": 0.2872999310493469,
"learning_rate": 1.4878773971620074e-05,
"loss": 1.3052764892578126,
"step": 1855
},
{
"epoch": 0.7579462102689487,
"grad_norm": 0.30310413241386414,
"learning_rate": 1.4644660940672627e-05,
"loss": 1.2672527313232422,
"step": 1860
},
{
"epoch": 0.7599837000814996,
"grad_norm": 0.2785632610321045,
"learning_rate": 1.4412087975501458e-05,
"loss": 1.2160486221313476,
"step": 1865
},
{
"epoch": 0.7620211898940505,
"grad_norm": 0.2611008882522583,
"learning_rate": 1.4181065206900762e-05,
"loss": 1.2843852996826173,
"step": 1870
},
{
"epoch": 0.7640586797066015,
"grad_norm": 0.2928135395050049,
"learning_rate": 1.3951602698138771e-05,
"loss": 1.3050471305847169,
"step": 1875
},
{
"epoch": 0.7660961695191524,
"grad_norm": 0.2900320887565613,
"learning_rate": 1.3723710444519306e-05,
"loss": 1.2945517539978026,
"step": 1880
},
{
"epoch": 0.7681336593317033,
"grad_norm": 0.25037550926208496,
"learning_rate": 1.3497398372946501e-05,
"loss": 1.2342204093933105,
"step": 1885
},
{
"epoch": 0.7701711491442543,
"grad_norm": 0.256669819355011,
"learning_rate": 1.3272676341492285e-05,
"loss": 1.284397792816162,
"step": 1890
},
{
"epoch": 0.7722086389568052,
"grad_norm": 0.3233170211315155,
"learning_rate": 1.3049554138967051e-05,
"loss": 1.2703520774841308,
"step": 1895
},
{
"epoch": 0.7742461287693562,
"grad_norm": 0.24304580688476562,
"learning_rate": 1.2828041484493187e-05,
"loss": 1.318847942352295,
"step": 1900
},
{
"epoch": 0.7762836185819071,
"grad_norm": 0.2866745591163635,
"learning_rate": 1.2608148027081773e-05,
"loss": 1.2475675582885741,
"step": 1905
},
{
"epoch": 0.778321108394458,
"grad_norm": 0.2676462233066559,
"learning_rate": 1.238988334521226e-05,
"loss": 1.3076879501342773,
"step": 1910
},
{
"epoch": 0.7803585982070089,
"grad_norm": 0.28914886713027954,
"learning_rate": 1.2173256946415212e-05,
"loss": 1.2790748596191406,
"step": 1915
},
{
"epoch": 0.78239608801956,
"grad_norm": 0.26682984828948975,
"learning_rate": 1.1958278266858147e-05,
"loss": 1.2656630516052245,
"step": 1920
},
{
"epoch": 0.7844335778321109,
"grad_norm": 0.3239537179470062,
"learning_rate": 1.1744956670934597e-05,
"loss": 1.344438362121582,
"step": 1925
},
{
"epoch": 0.7864710676446618,
"grad_norm": 0.25067269802093506,
"learning_rate": 1.1533301450856054e-05,
"loss": 1.235262107849121,
"step": 1930
},
{
"epoch": 0.7885085574572127,
"grad_norm": 0.2815246880054474,
"learning_rate": 1.1323321826247346e-05,
"loss": 1.252107524871826,
"step": 1935
},
{
"epoch": 0.7905460472697636,
"grad_norm": 0.34853169322013855,
"learning_rate": 1.1115026943744934e-05,
"loss": 1.3379150390625,
"step": 1940
},
{
"epoch": 0.7925835370823145,
"grad_norm": 0.27891600131988525,
"learning_rate": 1.090842587659851e-05,
"loss": 1.207740592956543,
"step": 1945
},
{
"epoch": 0.7946210268948656,
"grad_norm": 0.2608671188354492,
"learning_rate": 1.0703527624275832e-05,
"loss": 1.2680755615234376,
"step": 1950
},
{
"epoch": 0.7966585167074165,
"grad_norm": 0.29876336455345154,
"learning_rate": 1.0500341112070605e-05,
"loss": 1.3032147407531738,
"step": 1955
},
{
"epoch": 0.7986960065199674,
"grad_norm": 0.26807108521461487,
"learning_rate": 1.0298875190713802e-05,
"loss": 1.27540922164917,
"step": 1960
},
{
"epoch": 0.8007334963325183,
"grad_norm": 0.26686757802963257,
"learning_rate": 1.0099138635988026e-05,
"loss": 1.2521123886108398,
"step": 1965
},
{
"epoch": 0.8027709861450693,
"grad_norm": 0.2923862934112549,
"learning_rate": 9.901140148345361e-06,
"loss": 1.266144371032715,
"step": 1970
},
{
"epoch": 0.8048084759576202,
"grad_norm": 0.26123881340026855,
"learning_rate": 9.704888352528258e-06,
"loss": 1.290677833557129,
"step": 1975
},
{
"epoch": 0.8068459657701712,
"grad_norm": 0.28828051686286926,
"learning_rate": 9.51039179719394e-06,
"loss": 1.2674976348876954,
"step": 1980
},
{
"epoch": 0.8088834555827221,
"grad_norm": 0.28441664576530457,
"learning_rate": 9.317658954541992e-06,
"loss": 1.3033327102661132,
"step": 1985
},
{
"epoch": 0.810920945395273,
"grad_norm": 0.2592161297798157,
"learning_rate": 9.12669821994529e-06,
"loss": 1.214752197265625,
"step": 1990
},
{
"epoch": 0.812958435207824,
"grad_norm": 0.24296078085899353,
"learning_rate": 8.93751791158432e-06,
"loss": 1.305156135559082,
"step": 1995
},
{
"epoch": 0.8149959250203749,
"grad_norm": 0.26017627120018005,
"learning_rate": 8.75012627008489e-06,
"loss": 1.3143917083740235,
"step": 2000
},
{
"epoch": 0.8170334148329258,
"grad_norm": 0.28585124015808105,
"learning_rate": 8.56453145815907e-06,
"loss": 1.2376840591430665,
"step": 2005
},
{
"epoch": 0.8190709046454768,
"grad_norm": 0.2853546440601349,
"learning_rate": 8.380741560249727e-06,
"loss": 1.267874813079834,
"step": 2010
},
{
"epoch": 0.8211083944580277,
"grad_norm": 0.2658486068248749,
"learning_rate": 8.198764582178303e-06,
"loss": 1.2449701309204102,
"step": 2015
},
{
"epoch": 0.8231458842705787,
"grad_norm": 0.27855250239372253,
"learning_rate": 8.018608450796122e-06,
"loss": 1.2501028060913086,
"step": 2020
},
{
"epoch": 0.8251833740831296,
"grad_norm": 0.29566341638565063,
"learning_rate": 7.840281013639078e-06,
"loss": 1.2870569229125977,
"step": 2025
},
{
"epoch": 0.8272208638956805,
"grad_norm": 0.3037174642086029,
"learning_rate": 7.663790038585793e-06,
"loss": 1.255519199371338,
"step": 2030
},
{
"epoch": 0.8292583537082314,
"grad_norm": 0.2620980739593506,
"learning_rate": 7.489143213519301e-06,
"loss": 1.2635300636291504,
"step": 2035
},
{
"epoch": 0.8312958435207825,
"grad_norm": 0.2716981768608093,
"learning_rate": 7.31634814599208e-06,
"loss": 1.256894302368164,
"step": 2040
},
{
"epoch": 0.8333333333333334,
"grad_norm": 0.27713361382484436,
"learning_rate": 7.145412362894771e-06,
"loss": 1.2706600189208985,
"step": 2045
},
{
"epoch": 0.8353708231458843,
"grad_norm": 0.3421715795993805,
"learning_rate": 6.976343310128214e-06,
"loss": 1.2327119827270507,
"step": 2050
},
{
"epoch": 0.8374083129584352,
"grad_norm": 0.2683045566082001,
"learning_rate": 6.809148352279182e-06,
"loss": 1.2554466247558593,
"step": 2055
},
{
"epoch": 0.8394458027709861,
"grad_norm": 0.27507442235946655,
"learning_rate": 6.643834772299545e-06,
"loss": 1.2914720535278321,
"step": 2060
},
{
"epoch": 0.841483292583537,
"grad_norm": 0.2618330121040344,
"learning_rate": 6.480409771189028e-06,
"loss": 1.3056722640991212,
"step": 2065
},
{
"epoch": 0.843520782396088,
"grad_norm": 0.307759553194046,
"learning_rate": 6.318880467681526e-06,
"loss": 1.3092619895935058,
"step": 2070
},
{
"epoch": 0.845558272208639,
"grad_norm": 0.24145160615444183,
"learning_rate": 6.159253897935069e-06,
"loss": 1.2743852615356446,
"step": 2075
},
{
"epoch": 0.8475957620211899,
"grad_norm": 0.26536521315574646,
"learning_rate": 6.001537015225267e-06,
"loss": 1.2622230529785157,
"step": 2080
},
{
"epoch": 0.8496332518337408,
"grad_norm": 0.3143734335899353,
"learning_rate": 5.845736689642472e-06,
"loss": 1.2458198547363282,
"step": 2085
},
{
"epoch": 0.8516707416462918,
"grad_norm": 0.2654123306274414,
"learning_rate": 5.691859707792474e-06,
"loss": 1.2306365966796875,
"step": 2090
},
{
"epoch": 0.8537082314588427,
"grad_norm": 0.27503931522369385,
"learning_rate": 5.539912772500944e-06,
"loss": 1.2018237113952637,
"step": 2095
},
{
"epoch": 0.8557457212713936,
"grad_norm": 0.3079373240470886,
"learning_rate": 5.3899025025213935e-06,
"loss": 1.2758003234863282,
"step": 2100
},
{
"epoch": 0.8577832110839446,
"grad_norm": 0.26040902733802795,
"learning_rate": 5.241835432246889e-06,
"loss": 1.234812355041504,
"step": 2105
},
{
"epoch": 0.8598207008964955,
"grad_norm": 0.31597888469696045,
"learning_rate": 5.095718011425454e-06,
"loss": 1.2464513778686523,
"step": 2110
},
{
"epoch": 0.8618581907090465,
"grad_norm": 0.27159395813941956,
"learning_rate": 4.951556604879048e-06,
"loss": 1.2601951599121093,
"step": 2115
},
{
"epoch": 0.8638956805215974,
"grad_norm": 0.2514100968837738,
"learning_rate": 4.809357492226396e-06,
"loss": 1.3039180755615234,
"step": 2120
},
{
"epoch": 0.8659331703341483,
"grad_norm": 0.2648881673812866,
"learning_rate": 4.669126867609375e-06,
"loss": 1.2510721206665039,
"step": 2125
},
{
"epoch": 0.8679706601466992,
"grad_norm": 0.267498642206192,
"learning_rate": 4.530870839423257e-06,
"loss": 1.2672598838806153,
"step": 2130
},
{
"epoch": 0.8700081499592502,
"grad_norm": 0.2557595372200012,
"learning_rate": 4.394595430050613e-06,
"loss": 1.2957945823669434,
"step": 2135
},
{
"epoch": 0.8720456397718012,
"grad_norm": 0.30976709723472595,
"learning_rate": 4.260306575598949e-06,
"loss": 1.2747732162475587,
"step": 2140
},
{
"epoch": 0.8740831295843521,
"grad_norm": 0.28435832262039185,
"learning_rate": 4.12801012564219e-06,
"loss": 1.2773303985595703,
"step": 2145
},
{
"epoch": 0.876120619396903,
"grad_norm": 0.2486259937286377,
"learning_rate": 3.997711842965813e-06,
"loss": 1.3023883819580078,
"step": 2150
},
{
"epoch": 0.8781581092094539,
"grad_norm": 0.29785382747650146,
"learning_rate": 3.8694174033158555e-06,
"loss": 1.2375498771667481,
"step": 2155
},
{
"epoch": 0.8801955990220048,
"grad_norm": 0.27944067120552063,
"learning_rate": 3.743132395151705e-06,
"loss": 1.2415226936340331,
"step": 2160
},
{
"epoch": 0.8822330888345559,
"grad_norm": 0.27645689249038696,
"learning_rate": 3.6188623194026104e-06,
"loss": 1.248374843597412,
"step": 2165
},
{
"epoch": 0.8842705786471068,
"grad_norm": 0.23298542201519012,
"learning_rate": 3.4966125892281097e-06,
"loss": 1.2680888175964355,
"step": 2170
},
{
"epoch": 0.8863080684596577,
"grad_norm": 0.26010146737098694,
"learning_rate": 3.376388529782215e-06,
"loss": 1.3126177787780762,
"step": 2175
},
{
"epoch": 0.8883455582722086,
"grad_norm": 0.298277884721756,
"learning_rate": 3.258195377981443e-06,
"loss": 1.2544168472290038,
"step": 2180
},
{
"epoch": 0.8903830480847595,
"grad_norm": 0.29919010400772095,
"learning_rate": 3.1420382822767323e-06,
"loss": 1.295281982421875,
"step": 2185
},
{
"epoch": 0.8924205378973105,
"grad_norm": 0.3453187942504883,
"learning_rate": 3.027922302429126e-06,
"loss": 1.2434377670288086,
"step": 2190
},
{
"epoch": 0.8944580277098615,
"grad_norm": 0.30271825194358826,
"learning_rate": 2.9158524092894214e-06,
"loss": 1.271092414855957,
"step": 2195
},
{
"epoch": 0.8964955175224124,
"grad_norm": 0.3616655468940735,
"learning_rate": 2.8058334845816213e-06,
"loss": 1.2621625900268554,
"step": 2200
},
{
"epoch": 0.8985330073349633,
"grad_norm": 0.3653607964515686,
"learning_rate": 2.6978703206902657e-06,
"loss": 1.2719340324401855,
"step": 2205
},
{
"epoch": 0.9005704971475142,
"grad_norm": 0.2653542459011078,
"learning_rate": 2.591967620451707e-06,
"loss": 1.2712973594665526,
"step": 2210
},
{
"epoch": 0.9026079869600652,
"grad_norm": 0.30116185545921326,
"learning_rate": 2.488129996949251e-06,
"loss": 1.2984991073608398,
"step": 2215
},
{
"epoch": 0.9046454767726161,
"grad_norm": 0.2726503908634186,
"learning_rate": 2.3863619733122068e-06,
"loss": 1.2352977752685548,
"step": 2220
},
{
"epoch": 0.9066829665851671,
"grad_norm": 0.26609891653060913,
"learning_rate": 2.286667982518853e-06,
"loss": 1.305104637145996,
"step": 2225
},
{
"epoch": 0.908720456397718,
"grad_norm": 0.326118141412735,
"learning_rate": 2.1890523672033536e-06,
"loss": 1.2743569374084474,
"step": 2230
},
{
"epoch": 0.910757946210269,
"grad_norm": 0.32020553946495056,
"learning_rate": 2.093519379466602e-06,
"loss": 1.285715675354004,
"step": 2235
},
{
"epoch": 0.9127954360228199,
"grad_norm": 0.2822478413581848,
"learning_rate": 2.000073180690959e-06,
"loss": 1.3384413719177246,
"step": 2240
},
{
"epoch": 0.9148329258353708,
"grad_norm": 0.2956802546977997,
"learning_rate": 1.908717841359048e-06,
"loss": 1.2809405326843262,
"step": 2245
},
{
"epoch": 0.9168704156479217,
"grad_norm": 0.3025357127189636,
"learning_rate": 1.8194573408763794e-06,
"loss": 1.313020896911621,
"step": 2250
},
{
"epoch": 0.9189079054604727,
"grad_norm": 0.2663908302783966,
"learning_rate": 1.7322955673980678e-06,
"loss": 1.24508056640625,
"step": 2255
},
{
"epoch": 0.9209453952730237,
"grad_norm": 0.29095685482025146,
"learning_rate": 1.647236317659423e-06,
"loss": 1.2820388793945312,
"step": 2260
},
{
"epoch": 0.9229828850855746,
"grad_norm": 0.3127174377441406,
"learning_rate": 1.5642832968105758e-06,
"loss": 1.2095099449157716,
"step": 2265
},
{
"epoch": 0.9250203748981255,
"grad_norm": 0.2576475739479065,
"learning_rate": 1.4834401182551027e-06,
"loss": 1.217179775238037,
"step": 2270
},
{
"epoch": 0.9270578647106764,
"grad_norm": 0.2530013918876648,
"learning_rate": 1.404710303492618e-06,
"loss": 1.2629472732543945,
"step": 2275
},
{
"epoch": 0.9290953545232273,
"grad_norm": 0.3033037781715393,
"learning_rate": 1.328097281965357e-06,
"loss": 1.2430845260620118,
"step": 2280
},
{
"epoch": 0.9311328443357784,
"grad_norm": 0.3296666741371155,
"learning_rate": 1.2536043909088191e-06,
"loss": 1.292903518676758,
"step": 2285
},
{
"epoch": 0.9331703341483293,
"grad_norm": 0.24970361590385437,
"learning_rate": 1.1812348752063772e-06,
"loss": 1.2903746604919433,
"step": 2290
},
{
"epoch": 0.9352078239608802,
"grad_norm": 0.27012351155281067,
"learning_rate": 1.1109918872479642e-06,
"loss": 1.2502481460571289,
"step": 2295
},
{
"epoch": 0.9372453137734311,
"grad_norm": 0.30377525091171265,
"learning_rate": 1.0428784867926978e-06,
"loss": 1.2305086135864258,
"step": 2300
},
{
"epoch": 0.939282803585982,
"grad_norm": 0.2782386243343353,
"learning_rate": 9.768976408356666e-07,
"loss": 1.2385313987731934,
"step": 2305
},
{
"epoch": 0.941320293398533,
"grad_norm": 0.3249647617340088,
"learning_rate": 9.130522234786499e-07,
"loss": 1.2792120933532716,
"step": 2310
},
{
"epoch": 0.943357783211084,
"grad_norm": 0.291220486164093,
"learning_rate": 8.513450158049108e-07,
"loss": 1.2129091262817382,
"step": 2315
},
{
"epoch": 0.9453952730236349,
"grad_norm": 0.27387040853500366,
"learning_rate": 7.917787057581061e-07,
"loss": 1.2347321510314941,
"step": 2320
},
{
"epoch": 0.9474327628361858,
"grad_norm": 0.24083305895328522,
"learning_rate": 7.343558880251289e-07,
"loss": 1.2338086128234864,
"step": 2325
},
{
"epoch": 0.9494702526487367,
"grad_norm": 0.2876109480857849,
"learning_rate": 6.790790639231548e-07,
"loss": 1.2726865768432618,
"step": 2330
},
{
"epoch": 0.9515077424612877,
"grad_norm": 0.26769185066223145,
"learning_rate": 6.259506412906402e-07,
"loss": 1.2764947891235352,
"step": 2335
},
{
"epoch": 0.9535452322738386,
"grad_norm": 0.32385265827178955,
"learning_rate": 5.749729343824395e-07,
"loss": 1.2744218826293945,
"step": 2340
},
{
"epoch": 0.9555827220863896,
"grad_norm": 0.270440012216568,
"learning_rate": 5.261481637690246e-07,
"loss": 1.2826515197753907,
"step": 2345
},
{
"epoch": 0.9576202118989405,
"grad_norm": 0.2797473967075348,
"learning_rate": 4.794784562397458e-07,
"loss": 1.2687211990356446,
"step": 2350
},
{
"epoch": 0.9596577017114915,
"grad_norm": 0.32515427470207214,
"learning_rate": 4.349658447101612e-07,
"loss": 1.2687038421630858,
"step": 2355
},
{
"epoch": 0.9616951915240424,
"grad_norm": 0.34697550535202026,
"learning_rate": 3.9261226813353537e-07,
"loss": 1.2769744873046875,
"step": 2360
},
{
"epoch": 0.9637326813365933,
"grad_norm": 0.27114152908325195,
"learning_rate": 3.524195714163292e-07,
"loss": 1.3136171340942382,
"step": 2365
},
{
"epoch": 0.9657701711491442,
"grad_norm": 0.3094554841518402,
"learning_rate": 3.143895053378698e-07,
"loss": 1.2633909225463866,
"step": 2370
},
{
"epoch": 0.9678076609616952,
"grad_norm": 0.3390296697616577,
"learning_rate": 2.785237264740781e-07,
"loss": 1.2642793655395508,
"step": 2375
},
{
"epoch": 0.9698451507742462,
"grad_norm": 0.27997589111328125,
"learning_rate": 2.4482379712530423e-07,
"loss": 1.249699020385742,
"step": 2380
},
{
"epoch": 0.9718826405867971,
"grad_norm": 0.26257532835006714,
"learning_rate": 2.1329118524827662e-07,
"loss": 1.2510973930358886,
"step": 2385
},
{
"epoch": 0.973920130399348,
"grad_norm": 0.2797265350818634,
"learning_rate": 1.8392726439214746e-07,
"loss": 1.2273929595947266,
"step": 2390
},
{
"epoch": 0.9759576202118989,
"grad_norm": 0.26638010144233704,
"learning_rate": 1.567333136387017e-07,
"loss": 1.2225419998168945,
"step": 2395
},
{
"epoch": 0.9779951100244498,
"grad_norm": 0.2471892237663269,
"learning_rate": 1.3171051754657938e-07,
"loss": 1.2640929222106934,
"step": 2400
},
{
"epoch": 0.9800325998370009,
"grad_norm": 0.2521549463272095,
"learning_rate": 1.088599660997336e-07,
"loss": 1.2598012924194335,
"step": 2405
},
{
"epoch": 0.9820700896495518,
"grad_norm": 0.26020875573158264,
"learning_rate": 8.818265465991293e-08,
"loss": 1.2850156784057618,
"step": 2410
},
{
"epoch": 0.9841075794621027,
"grad_norm": 0.24907110631465912,
"learning_rate": 6.967948392331835e-08,
"loss": 1.2513031005859374,
"step": 2415
},
{
"epoch": 0.9861450692746536,
"grad_norm": 0.30028823018074036,
"learning_rate": 5.3351259881379014e-08,
"loss": 1.2602060317993165,
"step": 2420
},
{
"epoch": 0.9881825590872045,
"grad_norm": 0.23473893105983734,
"learning_rate": 3.9198693785619244e-08,
"loss": 1.255910873413086,
"step": 2425
},
{
"epoch": 0.9902200488997555,
"grad_norm": 0.2633054256439209,
"learning_rate": 2.722240211670557e-08,
"loss": 1.270261287689209,
"step": 2430
},
{
"epoch": 0.9922575387123065,
"grad_norm": 0.3252529799938202,
"learning_rate": 1.742290655755707e-08,
"loss": 1.2474273681640624,
"step": 2435
},
{
"epoch": 0.9942950285248574,
"grad_norm": 0.2814902365207672,
"learning_rate": 9.800633970669104e-09,
"loss": 1.2509822845458984,
"step": 2440
},
{
"epoch": 0.9963325183374083,
"grad_norm": 0.29194626212120056,
"learning_rate": 4.355916379467084e-09,
"loss": 1.2346750259399415,
"step": 2445
},
{
"epoch": 0.9983700081499592,
"grad_norm": 0.2590709924697876,
"learning_rate": 1.0889909538791543e-09,
"loss": 1.2552786827087403,
"step": 2450
}
],
"logging_steps": 5,
"max_steps": 2454,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 3.157508424539832e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}