tttw1002 / trainer_state.json
Kaipengm2's picture
Upload folder using huggingface_hub
eefedd9 verified
Raw
History Blame Contribute Delete
196 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 5246,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0009531979792202841,
"grad_norm": 2.603611946105957,
"learning_rate": 2.531645569620253e-06,
"loss": 2.94761962890625,
"step": 5
},
{
"epoch": 0.0019063959584405682,
"grad_norm": 2.1450037956237793,
"learning_rate": 5.69620253164557e-06,
"loss": 3.0334014892578125,
"step": 10
},
{
"epoch": 0.0028595939376608523,
"grad_norm": 2.423077344894409,
"learning_rate": 8.860759493670886e-06,
"loss": 2.886791229248047,
"step": 15
},
{
"epoch": 0.0038127919168811364,
"grad_norm": 1.6740561723709106,
"learning_rate": 1.2025316455696203e-05,
"loss": 2.835204315185547,
"step": 20
},
{
"epoch": 0.0047659898961014205,
"grad_norm": 0.9845642447471619,
"learning_rate": 1.5189873417721521e-05,
"loss": 2.668177604675293,
"step": 25
},
{
"epoch": 0.005719187875321705,
"grad_norm": 0.7101410627365112,
"learning_rate": 1.8354430379746836e-05,
"loss": 2.4937652587890624,
"step": 30
},
{
"epoch": 0.006672385854541989,
"grad_norm": 0.6729114055633545,
"learning_rate": 2.1518987341772153e-05,
"loss": 2.4470664978027346,
"step": 35
},
{
"epoch": 0.007625583833762273,
"grad_norm": 0.665655255317688,
"learning_rate": 2.468354430379747e-05,
"loss": 2.420810317993164,
"step": 40
},
{
"epoch": 0.008578781812982557,
"grad_norm": 0.6502403020858765,
"learning_rate": 2.7848101265822786e-05,
"loss": 2.335397148132324,
"step": 45
},
{
"epoch": 0.009531979792202841,
"grad_norm": 0.5307614803314209,
"learning_rate": 3.10126582278481e-05,
"loss": 2.315672492980957,
"step": 50
},
{
"epoch": 0.010485177771423125,
"grad_norm": 0.5500478744506836,
"learning_rate": 3.4177215189873416e-05,
"loss": 2.2122943878173826,
"step": 55
},
{
"epoch": 0.01143837575064341,
"grad_norm": 0.5895763635635376,
"learning_rate": 3.7341772151898736e-05,
"loss": 2.283915138244629,
"step": 60
},
{
"epoch": 0.012391573729863693,
"grad_norm": 0.4846538007259369,
"learning_rate": 4.050632911392405e-05,
"loss": 2.166631317138672,
"step": 65
},
{
"epoch": 0.013344771709083977,
"grad_norm": 0.6095337867736816,
"learning_rate": 4.367088607594937e-05,
"loss": 2.100119781494141,
"step": 70
},
{
"epoch": 0.014297969688304261,
"grad_norm": 0.6169545650482178,
"learning_rate": 4.683544303797468e-05,
"loss": 2.0974822998046876,
"step": 75
},
{
"epoch": 0.015251167667524546,
"grad_norm": 0.6472786068916321,
"learning_rate": 5e-05,
"loss": 2.020016098022461,
"step": 80
},
{
"epoch": 0.01620436564674483,
"grad_norm": 0.5326976776123047,
"learning_rate": 5.3164556962025316e-05,
"loss": 1.9512775421142579,
"step": 85
},
{
"epoch": 0.017157563625965114,
"grad_norm": 0.5936091542243958,
"learning_rate": 5.6329113924050636e-05,
"loss": 1.9886482238769532,
"step": 90
},
{
"epoch": 0.018110761605185398,
"grad_norm": 0.5242489576339722,
"learning_rate": 5.949367088607595e-05,
"loss": 1.8959827423095703,
"step": 95
},
{
"epoch": 0.019063959584405682,
"grad_norm": 0.7063551545143127,
"learning_rate": 6.265822784810128e-05,
"loss": 1.901272964477539,
"step": 100
},
{
"epoch": 0.020017157563625966,
"grad_norm": 0.7186295986175537,
"learning_rate": 6.582278481012658e-05,
"loss": 1.8788005828857421,
"step": 105
},
{
"epoch": 0.02097035554284625,
"grad_norm": 0.48944440484046936,
"learning_rate": 6.89873417721519e-05,
"loss": 1.8759675979614259,
"step": 110
},
{
"epoch": 0.021923553522066534,
"grad_norm": 0.5304045081138611,
"learning_rate": 7.215189873417722e-05,
"loss": 1.7807559967041016,
"step": 115
},
{
"epoch": 0.02287675150128682,
"grad_norm": 0.6360881328582764,
"learning_rate": 7.531645569620254e-05,
"loss": 1.7849002838134767,
"step": 120
},
{
"epoch": 0.023829949480507102,
"grad_norm": 0.6153790354728699,
"learning_rate": 7.848101265822784e-05,
"loss": 1.8074321746826172,
"step": 125
},
{
"epoch": 0.024783147459727387,
"grad_norm": 0.7403739094734192,
"learning_rate": 8.164556962025317e-05,
"loss": 1.77244873046875,
"step": 130
},
{
"epoch": 0.02573634543894767,
"grad_norm": 0.6022510528564453,
"learning_rate": 8.481012658227848e-05,
"loss": 1.7288890838623048,
"step": 135
},
{
"epoch": 0.026689543418167955,
"grad_norm": 0.5608066320419312,
"learning_rate": 8.797468354430381e-05,
"loss": 1.770965576171875,
"step": 140
},
{
"epoch": 0.02764274139738824,
"grad_norm": 0.6921910643577576,
"learning_rate": 9.113924050632912e-05,
"loss": 1.7317293167114258,
"step": 145
},
{
"epoch": 0.028595939376608523,
"grad_norm": 0.64326012134552,
"learning_rate": 9.430379746835444e-05,
"loss": 1.7647701263427735,
"step": 150
},
{
"epoch": 0.029549137355828807,
"grad_norm": 0.8091812133789062,
"learning_rate": 9.746835443037975e-05,
"loss": 1.7537090301513671,
"step": 155
},
{
"epoch": 0.03050233533504909,
"grad_norm": 0.8823623657226562,
"learning_rate": 9.999999046884494e-05,
"loss": 1.7446464538574218,
"step": 160
},
{
"epoch": 0.031455533314269375,
"grad_norm": 0.7236696481704712,
"learning_rate": 9.999965687879942e-05,
"loss": 1.7527923583984375,
"step": 165
},
{
"epoch": 0.03240873129348966,
"grad_norm": 0.7431514859199524,
"learning_rate": 9.999884673463466e-05,
"loss": 1.712713623046875,
"step": 170
},
{
"epoch": 0.03336192927270994,
"grad_norm": 0.8012411594390869,
"learning_rate": 9.999756004407229e-05,
"loss": 1.7108882904052733,
"step": 175
},
{
"epoch": 0.03431512725193023,
"grad_norm": 0.791260838508606,
"learning_rate": 9.999579681937592e-05,
"loss": 1.7204954147338867,
"step": 180
},
{
"epoch": 0.03526832523115051,
"grad_norm": 0.6157879829406738,
"learning_rate": 9.99935570773511e-05,
"loss": 1.6766773223876954,
"step": 185
},
{
"epoch": 0.036221523210370796,
"grad_norm": 0.5641170144081116,
"learning_rate": 9.999084083934518e-05,
"loss": 1.60144100189209,
"step": 190
},
{
"epoch": 0.03717472118959108,
"grad_norm": 0.8239608407020569,
"learning_rate": 9.998764813124699e-05,
"loss": 1.5777477264404296,
"step": 195
},
{
"epoch": 0.038127919168811364,
"grad_norm": 0.9198218584060669,
"learning_rate": 9.99839789834867e-05,
"loss": 1.6275455474853515,
"step": 200
},
{
"epoch": 0.03908111714803165,
"grad_norm": 0.8572738170623779,
"learning_rate": 9.997983343103554e-05,
"loss": 1.6616302490234376,
"step": 205
},
{
"epoch": 0.04003431512725193,
"grad_norm": 1.1088039875030518,
"learning_rate": 9.997521151340535e-05,
"loss": 1.6321582794189453,
"step": 210
},
{
"epoch": 0.040987513106472216,
"grad_norm": 0.7507736086845398,
"learning_rate": 9.997011327464831e-05,
"loss": 1.6804428100585938,
"step": 215
},
{
"epoch": 0.0419407110856925,
"grad_norm": 0.7173669338226318,
"learning_rate": 9.99645387633565e-05,
"loss": 1.6310005187988281,
"step": 220
},
{
"epoch": 0.042893909064912784,
"grad_norm": 0.7343104481697083,
"learning_rate": 9.995848803266141e-05,
"loss": 1.648260498046875,
"step": 225
},
{
"epoch": 0.04384710704413307,
"grad_norm": 0.6716263890266418,
"learning_rate": 9.995196114023345e-05,
"loss": 1.682364273071289,
"step": 230
},
{
"epoch": 0.04480030502335335,
"grad_norm": 0.8079553842544556,
"learning_rate": 9.994495814828137e-05,
"loss": 1.653547477722168,
"step": 235
},
{
"epoch": 0.04575350300257364,
"grad_norm": 0.655113697052002,
"learning_rate": 9.993747912355175e-05,
"loss": 1.7656341552734376,
"step": 240
},
{
"epoch": 0.04670670098179392,
"grad_norm": 0.5484510064125061,
"learning_rate": 9.992952413732826e-05,
"loss": 1.7137901306152343,
"step": 245
},
{
"epoch": 0.047659898961014205,
"grad_norm": 0.9827032089233398,
"learning_rate": 9.992109326543108e-05,
"loss": 1.650299644470215,
"step": 250
},
{
"epoch": 0.04861309694023449,
"grad_norm": 0.6224775314331055,
"learning_rate": 9.991218658821608e-05,
"loss": 1.59774169921875,
"step": 255
},
{
"epoch": 0.04956629491945477,
"grad_norm": 0.6683592796325684,
"learning_rate": 9.99028041905741e-05,
"loss": 1.658073616027832,
"step": 260
},
{
"epoch": 0.05051949289867506,
"grad_norm": 0.5990311503410339,
"learning_rate": 9.989294616193017e-05,
"loss": 1.6002532958984375,
"step": 265
},
{
"epoch": 0.05147269087789534,
"grad_norm": 0.9066379070281982,
"learning_rate": 9.988261259624264e-05,
"loss": 1.6362133026123047,
"step": 270
},
{
"epoch": 0.052425888857115625,
"grad_norm": 0.6945815086364746,
"learning_rate": 9.987180359200222e-05,
"loss": 1.6218528747558594,
"step": 275
},
{
"epoch": 0.05337908683633591,
"grad_norm": 0.6819268465042114,
"learning_rate": 9.986051925223115e-05,
"loss": 1.5690045356750488,
"step": 280
},
{
"epoch": 0.054332284815556194,
"grad_norm": 0.46659737825393677,
"learning_rate": 9.98487596844821e-05,
"loss": 1.642005729675293,
"step": 285
},
{
"epoch": 0.05528548279477648,
"grad_norm": 0.7103817462921143,
"learning_rate": 9.983652500083732e-05,
"loss": 1.6343769073486327,
"step": 290
},
{
"epoch": 0.05623868077399676,
"grad_norm": 0.7017027139663696,
"learning_rate": 9.982381531790733e-05,
"loss": 1.6178388595581055,
"step": 295
},
{
"epoch": 0.057191878753217046,
"grad_norm": 1.0940369367599487,
"learning_rate": 9.981063075683e-05,
"loss": 1.6205331802368164,
"step": 300
},
{
"epoch": 0.05814507673243733,
"grad_norm": 0.6648183465003967,
"learning_rate": 9.979697144326934e-05,
"loss": 1.5720266342163085,
"step": 305
},
{
"epoch": 0.059098274711657614,
"grad_norm": 0.7526854872703552,
"learning_rate": 9.978283750741429e-05,
"loss": 1.574450969696045,
"step": 310
},
{
"epoch": 0.0600514726908779,
"grad_norm": 0.7190886735916138,
"learning_rate": 9.976822908397749e-05,
"loss": 1.6405771255493165,
"step": 315
},
{
"epoch": 0.06100467067009818,
"grad_norm": 0.6349034309387207,
"learning_rate": 9.975314631219397e-05,
"loss": 1.5865010261535644,
"step": 320
},
{
"epoch": 0.061957868649318466,
"grad_norm": 0.5627608299255371,
"learning_rate": 9.973758933581986e-05,
"loss": 1.62694091796875,
"step": 325
},
{
"epoch": 0.06291106662853875,
"grad_norm": 0.6480703353881836,
"learning_rate": 9.9721558303131e-05,
"loss": 1.6081056594848633,
"step": 330
},
{
"epoch": 0.06386426460775903,
"grad_norm": 0.6254872679710388,
"learning_rate": 9.970505336692153e-05,
"loss": 1.5399663925170899,
"step": 335
},
{
"epoch": 0.06481746258697932,
"grad_norm": 0.7078723311424255,
"learning_rate": 9.968807468450244e-05,
"loss": 1.5698416709899903,
"step": 340
},
{
"epoch": 0.0657706605661996,
"grad_norm": 0.687049388885498,
"learning_rate": 9.967062241770006e-05,
"loss": 1.5318476676940918,
"step": 345
},
{
"epoch": 0.06672385854541989,
"grad_norm": 0.6676340699195862,
"learning_rate": 9.965269673285453e-05,
"loss": 1.5753061294555664,
"step": 350
},
{
"epoch": 0.06767705652464016,
"grad_norm": 0.6658551692962646,
"learning_rate": 9.963429780081819e-05,
"loss": 1.5348722457885742,
"step": 355
},
{
"epoch": 0.06863025450386046,
"grad_norm": 0.6602298617362976,
"learning_rate": 9.961542579695398e-05,
"loss": 1.5966236114501953,
"step": 360
},
{
"epoch": 0.06958345248308073,
"grad_norm": 0.6053290963172913,
"learning_rate": 9.959608090113379e-05,
"loss": 1.6215503692626954,
"step": 365
},
{
"epoch": 0.07053665046230102,
"grad_norm": 0.6316519379615784,
"learning_rate": 9.957626329773662e-05,
"loss": 1.5686593055725098,
"step": 370
},
{
"epoch": 0.0714898484415213,
"grad_norm": 0.6437824368476868,
"learning_rate": 9.955597317564705e-05,
"loss": 1.538528060913086,
"step": 375
},
{
"epoch": 0.07244304642074159,
"grad_norm": 0.556100070476532,
"learning_rate": 9.953521072825315e-05,
"loss": 1.516727352142334,
"step": 380
},
{
"epoch": 0.07339624439996187,
"grad_norm": 0.586941123008728,
"learning_rate": 9.951397615344495e-05,
"loss": 1.458935260772705,
"step": 385
},
{
"epoch": 0.07434944237918216,
"grad_norm": 0.6599296927452087,
"learning_rate": 9.94922696536123e-05,
"loss": 1.5930103302001952,
"step": 390
},
{
"epoch": 0.07530264035840244,
"grad_norm": 0.6119373440742493,
"learning_rate": 9.947009143564303e-05,
"loss": 1.5218027114868165,
"step": 395
},
{
"epoch": 0.07625583833762273,
"grad_norm": 0.7252292633056641,
"learning_rate": 9.944744171092103e-05,
"loss": 1.4561975479125977,
"step": 400
},
{
"epoch": 0.077209036316843,
"grad_norm": 0.6261231899261475,
"learning_rate": 9.942432069532417e-05,
"loss": 1.574506664276123,
"step": 405
},
{
"epoch": 0.0781622342960633,
"grad_norm": 0.8522781133651733,
"learning_rate": 9.940072860922229e-05,
"loss": 1.5758988380432128,
"step": 410
},
{
"epoch": 0.07911543227528357,
"grad_norm": 0.7650408744812012,
"learning_rate": 9.937666567747501e-05,
"loss": 1.5137447357177733,
"step": 415
},
{
"epoch": 0.08006863025450386,
"grad_norm": 0.7556065320968628,
"learning_rate": 9.935213212942973e-05,
"loss": 1.593979263305664,
"step": 420
},
{
"epoch": 0.08102182823372414,
"grad_norm": 0.7380269169807434,
"learning_rate": 9.93271281989193e-05,
"loss": 1.5264683723449708,
"step": 425
},
{
"epoch": 0.08197502621294443,
"grad_norm": 0.5459885001182556,
"learning_rate": 9.930165412425987e-05,
"loss": 1.6289447784423827,
"step": 430
},
{
"epoch": 0.08292822419216471,
"grad_norm": 0.6383824348449707,
"learning_rate": 9.927571014824861e-05,
"loss": 1.5772947311401366,
"step": 435
},
{
"epoch": 0.083881422171385,
"grad_norm": 0.6348546743392944,
"learning_rate": 9.924929651816142e-05,
"loss": 1.5777009010314942,
"step": 440
},
{
"epoch": 0.08483462015060528,
"grad_norm": 1.0627590417861938,
"learning_rate": 9.922241348575047e-05,
"loss": 1.5587986946105956,
"step": 445
},
{
"epoch": 0.08578781812982557,
"grad_norm": 0.6058380007743835,
"learning_rate": 9.919506130724196e-05,
"loss": 1.5736598014831542,
"step": 450
},
{
"epoch": 0.08674101610904585,
"grad_norm": 0.714316189289093,
"learning_rate": 9.916724024333351e-05,
"loss": 1.3844944953918457,
"step": 455
},
{
"epoch": 0.08769421408826614,
"grad_norm": 0.6118850708007812,
"learning_rate": 9.91389505591918e-05,
"loss": 1.54150333404541,
"step": 460
},
{
"epoch": 0.08864741206748641,
"grad_norm": 0.9378138780593872,
"learning_rate": 9.911019252445e-05,
"loss": 1.5315143585205078,
"step": 465
},
{
"epoch": 0.0896006100467067,
"grad_norm": 0.6369871497154236,
"learning_rate": 9.908096641320519e-05,
"loss": 1.5936030387878417,
"step": 470
},
{
"epoch": 0.09055380802592698,
"grad_norm": 0.645124077796936,
"learning_rate": 9.905127250401573e-05,
"loss": 1.4397594451904296,
"step": 475
},
{
"epoch": 0.09150700600514727,
"grad_norm": 0.7292397022247314,
"learning_rate": 9.902111107989868e-05,
"loss": 1.5745635986328126,
"step": 480
},
{
"epoch": 0.09246020398436755,
"grad_norm": 0.5931456685066223,
"learning_rate": 9.899048242832705e-05,
"loss": 1.4752291679382323,
"step": 485
},
{
"epoch": 0.09341340196358784,
"grad_norm": 0.6176527142524719,
"learning_rate": 9.8959386841227e-05,
"loss": 1.5611001014709474,
"step": 490
},
{
"epoch": 0.09436659994280812,
"grad_norm": 0.7982703447341919,
"learning_rate": 9.89278246149752e-05,
"loss": 1.5105000495910645,
"step": 495
},
{
"epoch": 0.09531979792202841,
"grad_norm": 0.6238499283790588,
"learning_rate": 9.889579605039588e-05,
"loss": 1.507917881011963,
"step": 500
},
{
"epoch": 0.09627299590124869,
"grad_norm": 0.7363895773887634,
"learning_rate": 9.886330145275801e-05,
"loss": 1.5819868087768554,
"step": 505
},
{
"epoch": 0.09722619388046898,
"grad_norm": 0.5625659823417664,
"learning_rate": 9.883034113177243e-05,
"loss": 1.5337334632873536,
"step": 510
},
{
"epoch": 0.09817939185968926,
"grad_norm": 0.6615775227546692,
"learning_rate": 9.879691540158883e-05,
"loss": 1.5258471488952636,
"step": 515
},
{
"epoch": 0.09913258983890955,
"grad_norm": 0.5451887249946594,
"learning_rate": 9.876302458079275e-05,
"loss": 1.463639259338379,
"step": 520
},
{
"epoch": 0.10008578781812982,
"grad_norm": 0.5581617951393127,
"learning_rate": 9.872866899240265e-05,
"loss": 1.5490670204162598,
"step": 525
},
{
"epoch": 0.10103898579735011,
"grad_norm": 0.6998193264007568,
"learning_rate": 9.869384896386668e-05,
"loss": 1.5936518669128419,
"step": 530
},
{
"epoch": 0.10199218377657039,
"grad_norm": 0.5768121480941772,
"learning_rate": 9.865856482705973e-05,
"loss": 1.5422281265258788,
"step": 535
},
{
"epoch": 0.10294538175579068,
"grad_norm": 0.6640679836273193,
"learning_rate": 9.862281691828007e-05,
"loss": 1.5552820205688476,
"step": 540
},
{
"epoch": 0.10389857973501096,
"grad_norm": 0.5004339814186096,
"learning_rate": 9.858660557824634e-05,
"loss": 1.4664517402648927,
"step": 545
},
{
"epoch": 0.10485177771423125,
"grad_norm": 0.6273754239082336,
"learning_rate": 9.854993115209416e-05,
"loss": 1.53157377243042,
"step": 550
},
{
"epoch": 0.10580497569345153,
"grad_norm": 0.603412926197052,
"learning_rate": 9.85127939893729e-05,
"loss": 1.414932918548584,
"step": 555
},
{
"epoch": 0.10675817367267182,
"grad_norm": 0.7552788257598877,
"learning_rate": 9.847519444404234e-05,
"loss": 1.4488500595092773,
"step": 560
},
{
"epoch": 0.1077113716518921,
"grad_norm": 0.6311041712760925,
"learning_rate": 9.843713287446934e-05,
"loss": 1.5790462493896484,
"step": 565
},
{
"epoch": 0.10866456963111239,
"grad_norm": 0.6677698493003845,
"learning_rate": 9.83986096434243e-05,
"loss": 1.5158218383789062,
"step": 570
},
{
"epoch": 0.10961776761033266,
"grad_norm": 0.7698470950126648,
"learning_rate": 9.835962511807786e-05,
"loss": 1.5150428771972657,
"step": 575
},
{
"epoch": 0.11057096558955296,
"grad_norm": 0.8735120892524719,
"learning_rate": 9.832017966999727e-05,
"loss": 1.5016836166381835,
"step": 580
},
{
"epoch": 0.11152416356877323,
"grad_norm": 0.6346865296363831,
"learning_rate": 9.828027367514296e-05,
"loss": 1.5396503448486327,
"step": 585
},
{
"epoch": 0.11247736154799352,
"grad_norm": 0.7268425822257996,
"learning_rate": 9.823990751386483e-05,
"loss": 1.460188865661621,
"step": 590
},
{
"epoch": 0.1134305595272138,
"grad_norm": 0.9180682301521301,
"learning_rate": 9.819908157089872e-05,
"loss": 1.4552988052368163,
"step": 595
},
{
"epoch": 0.11438375750643409,
"grad_norm": 0.5978934168815613,
"learning_rate": 9.815779623536276e-05,
"loss": 1.5380775451660156,
"step": 600
},
{
"epoch": 0.11533695548565437,
"grad_norm": 0.5722628235816956,
"learning_rate": 9.811605190075359e-05,
"loss": 1.4910564422607422,
"step": 605
},
{
"epoch": 0.11629015346487466,
"grad_norm": 0.5002546906471252,
"learning_rate": 9.807384896494258e-05,
"loss": 1.4244789123535155,
"step": 610
},
{
"epoch": 0.11724335144409494,
"grad_norm": 0.8268494009971619,
"learning_rate": 9.803118783017222e-05,
"loss": 1.4637470245361328,
"step": 615
},
{
"epoch": 0.11819654942331523,
"grad_norm": 0.6143020391464233,
"learning_rate": 9.798806890305202e-05,
"loss": 1.4786890029907227,
"step": 620
},
{
"epoch": 0.1191497474025355,
"grad_norm": 0.6437013745307922,
"learning_rate": 9.794449259455489e-05,
"loss": 1.5790328979492188,
"step": 625
},
{
"epoch": 0.1201029453817558,
"grad_norm": 0.6218923926353455,
"learning_rate": 9.790045932001306e-05,
"loss": 1.5234998703002929,
"step": 630
},
{
"epoch": 0.12105614336097607,
"grad_norm": 0.6830564737319946,
"learning_rate": 9.785596949911417e-05,
"loss": 1.5065134048461915,
"step": 635
},
{
"epoch": 0.12200934134019636,
"grad_norm": 0.5134453773498535,
"learning_rate": 9.781102355589728e-05,
"loss": 1.4854091644287108,
"step": 640
},
{
"epoch": 0.12296253931941664,
"grad_norm": 0.7742736339569092,
"learning_rate": 9.776562191874882e-05,
"loss": 1.493219757080078,
"step": 645
},
{
"epoch": 0.12391573729863693,
"grad_norm": 0.6689919233322144,
"learning_rate": 9.771976502039853e-05,
"loss": 1.4934368133544922,
"step": 650
},
{
"epoch": 0.12486893527785721,
"grad_norm": 0.7283974885940552,
"learning_rate": 9.767345329791522e-05,
"loss": 1.4888264656066894,
"step": 655
},
{
"epoch": 0.1258221332570775,
"grad_norm": 0.5391749143600464,
"learning_rate": 9.762668719270281e-05,
"loss": 1.509228515625,
"step": 660
},
{
"epoch": 0.12677533123629778,
"grad_norm": 0.5379536151885986,
"learning_rate": 9.757946715049595e-05,
"loss": 1.4341736793518067,
"step": 665
},
{
"epoch": 0.12772852921551806,
"grad_norm": 0.7166261076927185,
"learning_rate": 9.753179362135585e-05,
"loss": 1.3572914123535156,
"step": 670
},
{
"epoch": 0.12868172719473836,
"grad_norm": 0.5180948972702026,
"learning_rate": 9.748366705966594e-05,
"loss": 1.4610020637512207,
"step": 675
},
{
"epoch": 0.12963492517395864,
"grad_norm": 0.6936877965927124,
"learning_rate": 9.74350879241276e-05,
"loss": 1.4436023712158204,
"step": 680
},
{
"epoch": 0.13058812315317891,
"grad_norm": 0.6196861863136292,
"learning_rate": 9.738605667775578e-05,
"loss": 1.4700481414794921,
"step": 685
},
{
"epoch": 0.1315413211323992,
"grad_norm": 0.5676122903823853,
"learning_rate": 9.73365737878745e-05,
"loss": 1.5323741912841797,
"step": 690
},
{
"epoch": 0.1324945191116195,
"grad_norm": 0.5992201566696167,
"learning_rate": 9.728663972611251e-05,
"loss": 1.4702093124389648,
"step": 695
},
{
"epoch": 0.13344771709083977,
"grad_norm": 0.7591404914855957,
"learning_rate": 9.723625496839872e-05,
"loss": 1.352162742614746,
"step": 700
},
{
"epoch": 0.13440091507006005,
"grad_norm": 0.6924651861190796,
"learning_rate": 9.718541999495773e-05,
"loss": 1.4715144157409668,
"step": 705
},
{
"epoch": 0.13535411304928033,
"grad_norm": 0.6223724484443665,
"learning_rate": 9.713413529030514e-05,
"loss": 1.4769960403442384,
"step": 710
},
{
"epoch": 0.13630731102850063,
"grad_norm": 0.5391125679016113,
"learning_rate": 9.708240134324311e-05,
"loss": 1.45025691986084,
"step": 715
},
{
"epoch": 0.1372605090077209,
"grad_norm": 0.637563169002533,
"learning_rate": 9.703021864685549e-05,
"loss": 1.5019891738891602,
"step": 720
},
{
"epoch": 0.1382137069869412,
"grad_norm": 0.5522620677947998,
"learning_rate": 9.697758769850327e-05,
"loss": 1.43154935836792,
"step": 725
},
{
"epoch": 0.13916690496616146,
"grad_norm": 0.616644024848938,
"learning_rate": 9.692450899981981e-05,
"loss": 1.347580909729004,
"step": 730
},
{
"epoch": 0.14012010294538177,
"grad_norm": 0.6153479218482971,
"learning_rate": 9.687098305670605e-05,
"loss": 1.4593775749206543,
"step": 735
},
{
"epoch": 0.14107330092460205,
"grad_norm": 0.7453744411468506,
"learning_rate": 9.681701037932565e-05,
"loss": 1.4191917419433593,
"step": 740
},
{
"epoch": 0.14202649890382232,
"grad_norm": 0.5880297422409058,
"learning_rate": 9.676259148210013e-05,
"loss": 1.4744148254394531,
"step": 745
},
{
"epoch": 0.1429796968830426,
"grad_norm": 0.6558246612548828,
"learning_rate": 9.670772688370413e-05,
"loss": 1.378889560699463,
"step": 750
},
{
"epoch": 0.1439328948622629,
"grad_norm": 0.503994345664978,
"learning_rate": 9.665241710706019e-05,
"loss": 1.3663111686706544,
"step": 755
},
{
"epoch": 0.14488609284148318,
"grad_norm": 0.5992655158042908,
"learning_rate": 9.659666267933395e-05,
"loss": 1.4828500747680664,
"step": 760
},
{
"epoch": 0.14583929082070346,
"grad_norm": 0.5429707169532776,
"learning_rate": 9.654046413192916e-05,
"loss": 1.498521327972412,
"step": 765
},
{
"epoch": 0.14679248879992374,
"grad_norm": 0.5560146570205688,
"learning_rate": 9.648382200048243e-05,
"loss": 1.4589380264282226,
"step": 770
},
{
"epoch": 0.14774568677914404,
"grad_norm": 0.644885241985321,
"learning_rate": 9.64267368248583e-05,
"loss": 1.4834991455078126,
"step": 775
},
{
"epoch": 0.14869888475836432,
"grad_norm": 0.5343101620674133,
"learning_rate": 9.636920914914402e-05,
"loss": 1.4405736923217773,
"step": 780
},
{
"epoch": 0.1496520827375846,
"grad_norm": 0.6613175868988037,
"learning_rate": 9.631123952164437e-05,
"loss": 1.4065616607666016,
"step": 785
},
{
"epoch": 0.15060528071680487,
"grad_norm": 0.6537813544273376,
"learning_rate": 9.625282849487644e-05,
"loss": 1.475318431854248,
"step": 790
},
{
"epoch": 0.15155847869602515,
"grad_norm": 0.5663841962814331,
"learning_rate": 9.619397662556435e-05,
"loss": 1.4459699630737304,
"step": 795
},
{
"epoch": 0.15251167667524546,
"grad_norm": 0.5424874424934387,
"learning_rate": 9.613468447463397e-05,
"loss": 1.4708211898803711,
"step": 800
},
{
"epoch": 0.15346487465446573,
"grad_norm": 0.5572993159294128,
"learning_rate": 9.607495260720754e-05,
"loss": 1.4485332489013671,
"step": 805
},
{
"epoch": 0.154418072633686,
"grad_norm": 0.6531555652618408,
"learning_rate": 9.601478159259833e-05,
"loss": 1.5056473731994628,
"step": 810
},
{
"epoch": 0.1553712706129063,
"grad_norm": 0.5149410963058472,
"learning_rate": 9.595417200430516e-05,
"loss": 1.457845401763916,
"step": 815
},
{
"epoch": 0.1563244685921266,
"grad_norm": 0.7332213521003723,
"learning_rate": 9.589312442000702e-05,
"loss": 1.4969018936157226,
"step": 820
},
{
"epoch": 0.15727766657134687,
"grad_norm": 0.5197286009788513,
"learning_rate": 9.58316394215574e-05,
"loss": 1.430741310119629,
"step": 825
},
{
"epoch": 0.15823086455056715,
"grad_norm": 0.5908288359642029,
"learning_rate": 9.576971759497892e-05,
"loss": 1.416817855834961,
"step": 830
},
{
"epoch": 0.15918406252978742,
"grad_norm": 0.6981506943702698,
"learning_rate": 9.570735953045769e-05,
"loss": 1.432063102722168,
"step": 835
},
{
"epoch": 0.16013726050900773,
"grad_norm": 0.5579051375389099,
"learning_rate": 9.56445658223376e-05,
"loss": 1.4493148803710938,
"step": 840
},
{
"epoch": 0.161090458488228,
"grad_norm": 1.0297809839248657,
"learning_rate": 9.55813370691148e-05,
"loss": 1.4447443962097168,
"step": 845
},
{
"epoch": 0.16204365646744828,
"grad_norm": 0.5966977477073669,
"learning_rate": 9.551767387343184e-05,
"loss": 1.4523069381713867,
"step": 850
},
{
"epoch": 0.16299685444666856,
"grad_norm": 0.5328999757766724,
"learning_rate": 9.54535768420721e-05,
"loss": 1.4199484825134276,
"step": 855
},
{
"epoch": 0.16395005242588886,
"grad_norm": 0.5872285962104797,
"learning_rate": 9.53890465859538e-05,
"loss": 1.4303491592407227,
"step": 860
},
{
"epoch": 0.16490325040510914,
"grad_norm": 0.566717267036438,
"learning_rate": 9.53240837201244e-05,
"loss": 1.4515490531921387,
"step": 865
},
{
"epoch": 0.16585644838432942,
"grad_norm": 0.7435690760612488,
"learning_rate": 9.525868886375455e-05,
"loss": 1.419486904144287,
"step": 870
},
{
"epoch": 0.1668096463635497,
"grad_norm": 0.6318017244338989,
"learning_rate": 9.519286264013228e-05,
"loss": 1.4467649459838867,
"step": 875
},
{
"epoch": 0.16776284434277,
"grad_norm": 0.5930330157279968,
"learning_rate": 9.512660567665707e-05,
"loss": 1.4508071899414063,
"step": 880
},
{
"epoch": 0.16871604232199028,
"grad_norm": 0.585934579372406,
"learning_rate": 9.505991860483383e-05,
"loss": 1.419721794128418,
"step": 885
},
{
"epoch": 0.16966924030121056,
"grad_norm": 0.5505164265632629,
"learning_rate": 9.49928020602669e-05,
"loss": 1.4339097023010254,
"step": 890
},
{
"epoch": 0.17062243828043083,
"grad_norm": 0.5828437209129333,
"learning_rate": 9.492525668265399e-05,
"loss": 1.3837672233581544,
"step": 895
},
{
"epoch": 0.17157563625965114,
"grad_norm": 0.6347310543060303,
"learning_rate": 9.485728311578006e-05,
"loss": 1.4076103210449218,
"step": 900
},
{
"epoch": 0.17252883423887141,
"grad_norm": 0.579509437084198,
"learning_rate": 9.478888200751123e-05,
"loss": 1.4327677726745605,
"step": 905
},
{
"epoch": 0.1734820322180917,
"grad_norm": 0.4839983582496643,
"learning_rate": 9.47200540097886e-05,
"loss": 1.4768704414367675,
"step": 910
},
{
"epoch": 0.17443523019731197,
"grad_norm": 0.6111943125724792,
"learning_rate": 9.465079977862193e-05,
"loss": 1.5077296257019044,
"step": 915
},
{
"epoch": 0.17538842817653227,
"grad_norm": 0.5450233817100525,
"learning_rate": 9.458111997408359e-05,
"loss": 1.3619820594787597,
"step": 920
},
{
"epoch": 0.17634162615575255,
"grad_norm": 0.5677049160003662,
"learning_rate": 9.451101526030203e-05,
"loss": 1.4970988273620605,
"step": 925
},
{
"epoch": 0.17729482413497283,
"grad_norm": 0.5915372967720032,
"learning_rate": 9.444048630545573e-05,
"loss": 1.4513702392578125,
"step": 930
},
{
"epoch": 0.1782480221141931,
"grad_norm": 0.6052073240280151,
"learning_rate": 9.43695337817665e-05,
"loss": 1.4654881477355957,
"step": 935
},
{
"epoch": 0.1792012200934134,
"grad_norm": 0.5419918298721313,
"learning_rate": 9.429815836549335e-05,
"loss": 1.331681251525879,
"step": 940
},
{
"epoch": 0.1801544180726337,
"grad_norm": 0.9381572604179382,
"learning_rate": 9.422636073692594e-05,
"loss": 1.527089500427246,
"step": 945
},
{
"epoch": 0.18110761605185396,
"grad_norm": 0.5322099328041077,
"learning_rate": 9.415414158037807e-05,
"loss": 1.4385725021362306,
"step": 950
},
{
"epoch": 0.18206081403107424,
"grad_norm": 0.6048387289047241,
"learning_rate": 9.40815015841812e-05,
"loss": 1.4098733901977538,
"step": 955
},
{
"epoch": 0.18301401201029455,
"grad_norm": 0.5957016348838806,
"learning_rate": 9.400844144067785e-05,
"loss": 1.3700874328613282,
"step": 960
},
{
"epoch": 0.18396720998951482,
"grad_norm": 0.5258905291557312,
"learning_rate": 9.393496184621506e-05,
"loss": 1.5229506492614746,
"step": 965
},
{
"epoch": 0.1849204079687351,
"grad_norm": 0.5643404126167297,
"learning_rate": 9.386106350113773e-05,
"loss": 1.4227136611938476,
"step": 970
},
{
"epoch": 0.18587360594795538,
"grad_norm": 0.5638192296028137,
"learning_rate": 9.378674710978185e-05,
"loss": 1.326858901977539,
"step": 975
},
{
"epoch": 0.18682680392717568,
"grad_norm": 0.638519287109375,
"learning_rate": 9.371201338046797e-05,
"loss": 1.4697874069213868,
"step": 980
},
{
"epoch": 0.18778000190639596,
"grad_norm": 0.8361124992370605,
"learning_rate": 9.363686302549429e-05,
"loss": 1.4617888450622558,
"step": 985
},
{
"epoch": 0.18873319988561624,
"grad_norm": 0.6748219728469849,
"learning_rate": 9.356129676112997e-05,
"loss": 1.4383960723876954,
"step": 990
},
{
"epoch": 0.18968639786483651,
"grad_norm": 0.5921084880828857,
"learning_rate": 9.348531530760823e-05,
"loss": 1.4091617584228515,
"step": 995
},
{
"epoch": 0.19063959584405682,
"grad_norm": 0.572279155254364,
"learning_rate": 9.340891938911955e-05,
"loss": 1.2878050804138184,
"step": 1000
},
{
"epoch": 0.1915927938232771,
"grad_norm": 0.5492361187934875,
"learning_rate": 9.333210973380468e-05,
"loss": 1.4598516464233398,
"step": 1005
},
{
"epoch": 0.19254599180249737,
"grad_norm": 0.6517593264579773,
"learning_rate": 9.325488707374786e-05,
"loss": 1.392833137512207,
"step": 1010
},
{
"epoch": 0.19349918978171765,
"grad_norm": 0.5690333843231201,
"learning_rate": 9.31772521449696e-05,
"loss": 1.3799049377441406,
"step": 1015
},
{
"epoch": 0.19445238776093796,
"grad_norm": 0.5540326237678528,
"learning_rate": 9.309920568741993e-05,
"loss": 1.4240671157836915,
"step": 1020
},
{
"epoch": 0.19540558574015823,
"grad_norm": 0.4939487874507904,
"learning_rate": 9.302074844497117e-05,
"loss": 1.5023568153381348,
"step": 1025
},
{
"epoch": 0.1963587837193785,
"grad_norm": 0.5948994159698486,
"learning_rate": 9.294188116541088e-05,
"loss": 1.4755451202392578,
"step": 1030
},
{
"epoch": 0.1973119816985988,
"grad_norm": 0.6184943318367004,
"learning_rate": 9.286260460043475e-05,
"loss": 1.413205337524414,
"step": 1035
},
{
"epoch": 0.1982651796778191,
"grad_norm": 0.6401903033256531,
"learning_rate": 9.278291950563946e-05,
"loss": 1.4191042900085449,
"step": 1040
},
{
"epoch": 0.19921837765703937,
"grad_norm": 0.7051407098770142,
"learning_rate": 9.27028266405154e-05,
"loss": 1.385235595703125,
"step": 1045
},
{
"epoch": 0.20017157563625965,
"grad_norm": 0.6343866586685181,
"learning_rate": 9.262232676843953e-05,
"loss": 1.4120969772338867,
"step": 1050
},
{
"epoch": 0.20112477361547992,
"grad_norm": 0.5595751404762268,
"learning_rate": 9.254142065666801e-05,
"loss": 1.3800504684448243,
"step": 1055
},
{
"epoch": 0.20207797159470023,
"grad_norm": 1.0234235525131226,
"learning_rate": 9.246010907632895e-05,
"loss": 1.3642923355102539,
"step": 1060
},
{
"epoch": 0.2030311695739205,
"grad_norm": 0.6397507786750793,
"learning_rate": 9.237839280241505e-05,
"loss": 1.4000006675720216,
"step": 1065
},
{
"epoch": 0.20398436755314078,
"grad_norm": 0.6452104449272156,
"learning_rate": 9.229627261377618e-05,
"loss": 1.4264166831970215,
"step": 1070
},
{
"epoch": 0.20493756553236106,
"grad_norm": 0.6453445553779602,
"learning_rate": 9.2213749293112e-05,
"loss": 1.4670862197875976,
"step": 1075
},
{
"epoch": 0.20589076351158137,
"grad_norm": 0.9865917563438416,
"learning_rate": 9.213082362696448e-05,
"loss": 1.3475669860839843,
"step": 1080
},
{
"epoch": 0.20684396149080164,
"grad_norm": 0.5215176939964294,
"learning_rate": 9.20474964057104e-05,
"loss": 1.4212247848510742,
"step": 1085
},
{
"epoch": 0.20779715947002192,
"grad_norm": 0.8144325613975525,
"learning_rate": 9.196376842355382e-05,
"loss": 1.4531272888183593,
"step": 1090
},
{
"epoch": 0.2087503574492422,
"grad_norm": 0.6696116328239441,
"learning_rate": 9.187964047851852e-05,
"loss": 1.4349409103393556,
"step": 1095
},
{
"epoch": 0.2097035554284625,
"grad_norm": 0.5102435946464539,
"learning_rate": 9.179511337244034e-05,
"loss": 1.4025270462036132,
"step": 1100
},
{
"epoch": 0.21065675340768278,
"grad_norm": 0.561460018157959,
"learning_rate": 9.171018791095966e-05,
"loss": 1.3480627059936523,
"step": 1105
},
{
"epoch": 0.21160995138690306,
"grad_norm": 0.569500207901001,
"learning_rate": 9.162486490351358e-05,
"loss": 1.369408416748047,
"step": 1110
},
{
"epoch": 0.21256314936612333,
"grad_norm": 0.5028200149536133,
"learning_rate": 9.153914516332833e-05,
"loss": 1.3912203788757325,
"step": 1115
},
{
"epoch": 0.21351634734534364,
"grad_norm": 0.6325492858886719,
"learning_rate": 9.145302950741139e-05,
"loss": 1.3620031356811524,
"step": 1120
},
{
"epoch": 0.21446954532456391,
"grad_norm": 0.6650065779685974,
"learning_rate": 9.13665187565438e-05,
"loss": 1.4505338668823242,
"step": 1125
},
{
"epoch": 0.2154227433037842,
"grad_norm": 0.6167837977409363,
"learning_rate": 9.127961373527234e-05,
"loss": 1.431904125213623,
"step": 1130
},
{
"epoch": 0.21637594128300447,
"grad_norm": 0.7874214053153992,
"learning_rate": 9.119231527190158e-05,
"loss": 1.3958396911621094,
"step": 1135
},
{
"epoch": 0.21732913926222477,
"grad_norm": 0.5030773878097534,
"learning_rate": 9.110462419848611e-05,
"loss": 1.4253235816955567,
"step": 1140
},
{
"epoch": 0.21828233724144505,
"grad_norm": 0.639376699924469,
"learning_rate": 9.101654135082248e-05,
"loss": 1.4379282951354981,
"step": 1145
},
{
"epoch": 0.21923553522066533,
"grad_norm": 0.5737013220787048,
"learning_rate": 9.092806756844134e-05,
"loss": 1.3839303970336914,
"step": 1150
},
{
"epoch": 0.2201887331998856,
"grad_norm": 0.654533863067627,
"learning_rate": 9.083920369459941e-05,
"loss": 1.3634964942932128,
"step": 1155
},
{
"epoch": 0.2211419311791059,
"grad_norm": 0.5694367289543152,
"learning_rate": 9.074995057627136e-05,
"loss": 1.3617011070251466,
"step": 1160
},
{
"epoch": 0.2220951291583262,
"grad_norm": 0.7163636088371277,
"learning_rate": 9.066030906414189e-05,
"loss": 1.4148844718933105,
"step": 1165
},
{
"epoch": 0.22304832713754646,
"grad_norm": 0.7204518914222717,
"learning_rate": 9.057028001259747e-05,
"loss": 1.4406082153320312,
"step": 1170
},
{
"epoch": 0.22400152511676674,
"grad_norm": 0.6264455318450928,
"learning_rate": 9.047986427971831e-05,
"loss": 1.2642692565917968,
"step": 1175
},
{
"epoch": 0.22495472309598705,
"grad_norm": 0.6186636090278625,
"learning_rate": 9.038906272727012e-05,
"loss": 1.4535820007324218,
"step": 1180
},
{
"epoch": 0.22590792107520732,
"grad_norm": 0.75081467628479,
"learning_rate": 9.029787622069594e-05,
"loss": 1.3068763732910156,
"step": 1185
},
{
"epoch": 0.2268611190544276,
"grad_norm": 0.6259666681289673,
"learning_rate": 9.02063056291078e-05,
"loss": 1.4070331573486328,
"step": 1190
},
{
"epoch": 0.22781431703364788,
"grad_norm": 0.5153995156288147,
"learning_rate": 9.011435182527857e-05,
"loss": 1.4499921798706055,
"step": 1195
},
{
"epoch": 0.22876751501286818,
"grad_norm": 0.5387507081031799,
"learning_rate": 9.002201568563354e-05,
"loss": 1.425567626953125,
"step": 1200
},
{
"epoch": 0.22972071299208846,
"grad_norm": 0.5215945243835449,
"learning_rate": 8.992929809024209e-05,
"loss": 1.4283903121948243,
"step": 1205
},
{
"epoch": 0.23067391097130874,
"grad_norm": 0.5825516581535339,
"learning_rate": 8.983619992280934e-05,
"loss": 1.3858247756958009,
"step": 1210
},
{
"epoch": 0.23162710895052901,
"grad_norm": 0.7696310877799988,
"learning_rate": 8.974272207066767e-05,
"loss": 1.3599931716918945,
"step": 1215
},
{
"epoch": 0.23258030692974932,
"grad_norm": 0.5581028461456299,
"learning_rate": 8.964886542476831e-05,
"loss": 1.4406028747558595,
"step": 1220
},
{
"epoch": 0.2335335049089696,
"grad_norm": 0.6212132573127747,
"learning_rate": 8.955463087967283e-05,
"loss": 1.3897615432739259,
"step": 1225
},
{
"epoch": 0.23448670288818987,
"grad_norm": 0.7411231398582458,
"learning_rate": 8.94600193335446e-05,
"loss": 1.361074161529541,
"step": 1230
},
{
"epoch": 0.23543990086741015,
"grad_norm": 0.5625526905059814,
"learning_rate": 8.936503168814024e-05,
"loss": 1.4169339179992675,
"step": 1235
},
{
"epoch": 0.23639309884663046,
"grad_norm": 0.7929979562759399,
"learning_rate": 8.926966884880105e-05,
"loss": 1.2988113403320312,
"step": 1240
},
{
"epoch": 0.23734629682585073,
"grad_norm": 0.8460327386856079,
"learning_rate": 8.917393172444435e-05,
"loss": 1.4165689468383789,
"step": 1245
},
{
"epoch": 0.238299494805071,
"grad_norm": 0.6199972033500671,
"learning_rate": 8.907782122755482e-05,
"loss": 1.3829384803771974,
"step": 1250
},
{
"epoch": 0.2392526927842913,
"grad_norm": 0.7463774085044861,
"learning_rate": 8.898133827417578e-05,
"loss": 1.346090316772461,
"step": 1255
},
{
"epoch": 0.2402058907635116,
"grad_norm": 0.5334345698356628,
"learning_rate": 8.888448378390055e-05,
"loss": 1.3929250717163086,
"step": 1260
},
{
"epoch": 0.24115908874273187,
"grad_norm": 0.5729868412017822,
"learning_rate": 8.878725867986358e-05,
"loss": 1.4423667907714843,
"step": 1265
},
{
"epoch": 0.24211228672195215,
"grad_norm": 0.624984622001648,
"learning_rate": 8.868966388873173e-05,
"loss": 1.3437192916870118,
"step": 1270
},
{
"epoch": 0.24306548470117242,
"grad_norm": 0.7201520800590515,
"learning_rate": 8.859170034069533e-05,
"loss": 1.4097161293029785,
"step": 1275
},
{
"epoch": 0.24401868268039273,
"grad_norm": 0.4854135513305664,
"learning_rate": 8.849336896945946e-05,
"loss": 1.4177908897399902,
"step": 1280
},
{
"epoch": 0.244971880659613,
"grad_norm": 0.5305654406547546,
"learning_rate": 8.839467071223496e-05,
"loss": 1.5288206100463868,
"step": 1285
},
{
"epoch": 0.24592507863883328,
"grad_norm": 0.5070072412490845,
"learning_rate": 8.82956065097295e-05,
"loss": 1.376577377319336,
"step": 1290
},
{
"epoch": 0.24687827661805356,
"grad_norm": 0.4628710448741913,
"learning_rate": 8.819617730613862e-05,
"loss": 1.347555923461914,
"step": 1295
},
{
"epoch": 0.24783147459727387,
"grad_norm": 0.5474252104759216,
"learning_rate": 8.809638404913678e-05,
"loss": 1.3857503890991212,
"step": 1300
},
{
"epoch": 0.24878467257649414,
"grad_norm": 0.490693062543869,
"learning_rate": 8.799622768986826e-05,
"loss": 1.3779126167297364,
"step": 1305
},
{
"epoch": 0.24973787055571442,
"grad_norm": 0.7508081197738647,
"learning_rate": 8.78957091829381e-05,
"loss": 1.340792465209961,
"step": 1310
},
{
"epoch": 0.2506910685349347,
"grad_norm": 0.5256382822990417,
"learning_rate": 8.779482948640305e-05,
"loss": 1.3755708694458009,
"step": 1315
},
{
"epoch": 0.251644266514155,
"grad_norm": 0.8270961046218872,
"learning_rate": 8.769358956176241e-05,
"loss": 1.3478235244750976,
"step": 1320
},
{
"epoch": 0.2525974644933753,
"grad_norm": 0.5768178105354309,
"learning_rate": 8.759199037394887e-05,
"loss": 1.3513019561767579,
"step": 1325
},
{
"epoch": 0.25355066247259556,
"grad_norm": 0.736600399017334,
"learning_rate": 8.749003289131931e-05,
"loss": 1.456697654724121,
"step": 1330
},
{
"epoch": 0.25450386045181583,
"grad_norm": 0.5873891711235046,
"learning_rate": 8.738771808564555e-05,
"loss": 1.378398609161377,
"step": 1335
},
{
"epoch": 0.2554570584310361,
"grad_norm": 0.7747564315795898,
"learning_rate": 8.728504693210514e-05,
"loss": 1.3838523864746093,
"step": 1340
},
{
"epoch": 0.2564102564102564,
"grad_norm": 0.6339495778083801,
"learning_rate": 8.7182020409272e-05,
"loss": 1.3945033073425293,
"step": 1345
},
{
"epoch": 0.2573634543894767,
"grad_norm": 0.5631842017173767,
"learning_rate": 8.707863949910716e-05,
"loss": 1.378108024597168,
"step": 1350
},
{
"epoch": 0.258316652368697,
"grad_norm": 0.6121353507041931,
"learning_rate": 8.697490518694935e-05,
"loss": 1.4062651634216308,
"step": 1355
},
{
"epoch": 0.2592698503479173,
"grad_norm": 0.6150690317153931,
"learning_rate": 8.687081846150561e-05,
"loss": 1.3605571746826173,
"step": 1360
},
{
"epoch": 0.26022304832713755,
"grad_norm": 0.5333664417266846,
"learning_rate": 8.676638031484196e-05,
"loss": 1.4339089393615723,
"step": 1365
},
{
"epoch": 0.26117624630635783,
"grad_norm": 0.625784158706665,
"learning_rate": 8.666159174237373e-05,
"loss": 1.385106086730957,
"step": 1370
},
{
"epoch": 0.2621294442855781,
"grad_norm": 0.625914454460144,
"learning_rate": 8.655645374285637e-05,
"loss": 1.36302490234375,
"step": 1375
},
{
"epoch": 0.2630826422647984,
"grad_norm": 0.5408570170402527,
"learning_rate": 8.645096731837566e-05,
"loss": 1.3285486221313476,
"step": 1380
},
{
"epoch": 0.26403584024401866,
"grad_norm": 0.5265281796455383,
"learning_rate": 8.634513347433827e-05,
"loss": 1.3816985130310058,
"step": 1385
},
{
"epoch": 0.264989038223239,
"grad_norm": 0.6113120913505554,
"learning_rate": 8.623895321946225e-05,
"loss": 1.3977375984191895,
"step": 1390
},
{
"epoch": 0.26594223620245927,
"grad_norm": 0.5564927458763123,
"learning_rate": 8.613242756576728e-05,
"loss": 1.3723347663879395,
"step": 1395
},
{
"epoch": 0.26689543418167955,
"grad_norm": 0.494344562292099,
"learning_rate": 8.602555752856512e-05,
"loss": 1.4047321319580077,
"step": 1400
},
{
"epoch": 0.2678486321608998,
"grad_norm": 0.6288126111030579,
"learning_rate": 8.59183441264499e-05,
"loss": 1.4278359413146973,
"step": 1405
},
{
"epoch": 0.2688018301401201,
"grad_norm": 0.7214698791503906,
"learning_rate": 8.581078838128837e-05,
"loss": 1.347008228302002,
"step": 1410
},
{
"epoch": 0.2697550281193404,
"grad_norm": 0.481616735458374,
"learning_rate": 8.570289131821025e-05,
"loss": 1.3708770751953125,
"step": 1415
},
{
"epoch": 0.27070822609856066,
"grad_norm": 0.5352104306221008,
"learning_rate": 8.55946539655984e-05,
"loss": 1.4136503219604493,
"step": 1420
},
{
"epoch": 0.27166142407778093,
"grad_norm": 0.7826181054115295,
"learning_rate": 8.548607735507904e-05,
"loss": 1.3423415184020997,
"step": 1425
},
{
"epoch": 0.27261462205700127,
"grad_norm": 0.5781493782997131,
"learning_rate": 8.537716252151184e-05,
"loss": 1.3649326324462892,
"step": 1430
},
{
"epoch": 0.27356782003622154,
"grad_norm": 0.693625271320343,
"learning_rate": 8.526791050298022e-05,
"loss": 1.3296145439147948,
"step": 1435
},
{
"epoch": 0.2745210180154418,
"grad_norm": 0.673933744430542,
"learning_rate": 8.515832234078129e-05,
"loss": 1.350517463684082,
"step": 1440
},
{
"epoch": 0.2754742159946621,
"grad_norm": 0.54261314868927,
"learning_rate": 8.504839907941604e-05,
"loss": 1.4441131591796874,
"step": 1445
},
{
"epoch": 0.2764274139738824,
"grad_norm": 0.7024149298667908,
"learning_rate": 8.493814176657933e-05,
"loss": 1.3249446868896484,
"step": 1450
},
{
"epoch": 0.27738061195310265,
"grad_norm": 0.6024031043052673,
"learning_rate": 8.482755145314986e-05,
"loss": 1.380367374420166,
"step": 1455
},
{
"epoch": 0.27833380993232293,
"grad_norm": 0.588101327419281,
"learning_rate": 8.471662919318027e-05,
"loss": 1.3683594703674316,
"step": 1460
},
{
"epoch": 0.2792870079115432,
"grad_norm": 0.6452014446258545,
"learning_rate": 8.460537604388704e-05,
"loss": 1.3404038429260254,
"step": 1465
},
{
"epoch": 0.28024020589076354,
"grad_norm": 0.5389564633369446,
"learning_rate": 8.449379306564035e-05,
"loss": 1.3373002052307128,
"step": 1470
},
{
"epoch": 0.2811934038699838,
"grad_norm": 0.5641931891441345,
"learning_rate": 8.438188132195406e-05,
"loss": 1.4321128845214843,
"step": 1475
},
{
"epoch": 0.2821466018492041,
"grad_norm": 0.5689829587936401,
"learning_rate": 8.426964187947554e-05,
"loss": 1.4158197402954102,
"step": 1480
},
{
"epoch": 0.28309979982842437,
"grad_norm": 0.6428666710853577,
"learning_rate": 8.415707580797551e-05,
"loss": 1.3772966384887695,
"step": 1485
},
{
"epoch": 0.28405299780764465,
"grad_norm": 0.5933529734611511,
"learning_rate": 8.404418418033785e-05,
"loss": 1.3336894989013672,
"step": 1490
},
{
"epoch": 0.2850061957868649,
"grad_norm": 0.5442245006561279,
"learning_rate": 8.39309680725493e-05,
"loss": 1.3275571823120118,
"step": 1495
},
{
"epoch": 0.2859593937660852,
"grad_norm": 0.6149893999099731,
"learning_rate": 8.381742856368937e-05,
"loss": 1.3826486587524414,
"step": 1500
},
{
"epoch": 0.2869125917453055,
"grad_norm": 0.5523366928100586,
"learning_rate": 8.370356673591985e-05,
"loss": 1.3565295219421387,
"step": 1505
},
{
"epoch": 0.2878657897245258,
"grad_norm": 0.5619015097618103,
"learning_rate": 8.358938367447468e-05,
"loss": 1.4171290397644043,
"step": 1510
},
{
"epoch": 0.2888189877037461,
"grad_norm": 0.5124524235725403,
"learning_rate": 8.347488046764948e-05,
"loss": 1.343014621734619,
"step": 1515
},
{
"epoch": 0.28977218568296637,
"grad_norm": 0.514398992061615,
"learning_rate": 8.336005820679126e-05,
"loss": 1.3443027496337892,
"step": 1520
},
{
"epoch": 0.29072538366218664,
"grad_norm": 0.4980037212371826,
"learning_rate": 8.324491798628793e-05,
"loss": 1.3882343292236328,
"step": 1525
},
{
"epoch": 0.2916785816414069,
"grad_norm": 0.707893967628479,
"learning_rate": 8.312946090355796e-05,
"loss": 1.3469395637512207,
"step": 1530
},
{
"epoch": 0.2926317796206272,
"grad_norm": 0.5966976284980774,
"learning_rate": 8.301368805903988e-05,
"loss": 1.3640374183654784,
"step": 1535
},
{
"epoch": 0.2935849775998475,
"grad_norm": 0.6900899410247803,
"learning_rate": 8.289760055618175e-05,
"loss": 1.3795743942260743,
"step": 1540
},
{
"epoch": 0.29453817557906775,
"grad_norm": 0.7353556156158447,
"learning_rate": 8.278119950143074e-05,
"loss": 1.3588271141052246,
"step": 1545
},
{
"epoch": 0.2954913735582881,
"grad_norm": 0.6737638115882874,
"learning_rate": 8.26644860042225e-05,
"loss": 1.3469388008117675,
"step": 1550
},
{
"epoch": 0.29644457153750836,
"grad_norm": 0.49913325905799866,
"learning_rate": 8.254746117697062e-05,
"loss": 1.3773557662963867,
"step": 1555
},
{
"epoch": 0.29739776951672864,
"grad_norm": 0.5710513591766357,
"learning_rate": 8.243012613505604e-05,
"loss": 1.3305269241333009,
"step": 1560
},
{
"epoch": 0.2983509674959489,
"grad_norm": 0.6924476027488708,
"learning_rate": 8.231248199681635e-05,
"loss": 1.3393621444702148,
"step": 1565
},
{
"epoch": 0.2993041654751692,
"grad_norm": 0.6819959878921509,
"learning_rate": 8.219452988353526e-05,
"loss": 1.439009189605713,
"step": 1570
},
{
"epoch": 0.30025736345438947,
"grad_norm": 0.5717564225196838,
"learning_rate": 8.207627091943178e-05,
"loss": 1.3435260772705078,
"step": 1575
},
{
"epoch": 0.30121056143360975,
"grad_norm": 0.5868562459945679,
"learning_rate": 8.195770623164955e-05,
"loss": 1.3461615562438964,
"step": 1580
},
{
"epoch": 0.30216375941283,
"grad_norm": 0.6372678875923157,
"learning_rate": 8.183883695024615e-05,
"loss": 1.2999251365661622,
"step": 1585
},
{
"epoch": 0.3031169573920503,
"grad_norm": 0.5561612248420715,
"learning_rate": 8.171966420818228e-05,
"loss": 1.4569865226745606,
"step": 1590
},
{
"epoch": 0.30407015537127063,
"grad_norm": 0.5445943474769592,
"learning_rate": 8.160018914131094e-05,
"loss": 1.3226649284362793,
"step": 1595
},
{
"epoch": 0.3050233533504909,
"grad_norm": 0.5619344115257263,
"learning_rate": 8.148041288836667e-05,
"loss": 1.2719925880432128,
"step": 1600
},
{
"epoch": 0.3059765513297112,
"grad_norm": 0.6340746879577637,
"learning_rate": 8.13603365909546e-05,
"loss": 1.3441352844238281,
"step": 1605
},
{
"epoch": 0.30692974930893147,
"grad_norm": 0.678502082824707,
"learning_rate": 8.123996139353971e-05,
"loss": 1.306410312652588,
"step": 1610
},
{
"epoch": 0.30788294728815174,
"grad_norm": 0.6480040550231934,
"learning_rate": 8.11192884434358e-05,
"loss": 1.342271900177002,
"step": 1615
},
{
"epoch": 0.308836145267372,
"grad_norm": 0.7765945792198181,
"learning_rate": 8.099831889079455e-05,
"loss": 1.3604805946350098,
"step": 1620
},
{
"epoch": 0.3097893432465923,
"grad_norm": 0.7199779748916626,
"learning_rate": 8.087705388859467e-05,
"loss": 1.477169132232666,
"step": 1625
},
{
"epoch": 0.3107425412258126,
"grad_norm": 0.640831470489502,
"learning_rate": 8.07554945926308e-05,
"loss": 1.3985925674438477,
"step": 1630
},
{
"epoch": 0.3116957392050329,
"grad_norm": 0.5893611311912537,
"learning_rate": 8.063364216150257e-05,
"loss": 1.3131418228149414,
"step": 1635
},
{
"epoch": 0.3126489371842532,
"grad_norm": 0.8980715274810791,
"learning_rate": 8.051149775660354e-05,
"loss": 1.427363967895508,
"step": 1640
},
{
"epoch": 0.31360213516347346,
"grad_norm": 0.5138471126556396,
"learning_rate": 8.038906254211002e-05,
"loss": 1.3387574195861816,
"step": 1645
},
{
"epoch": 0.31455533314269374,
"grad_norm": 0.5718988180160522,
"learning_rate": 8.026633768497019e-05,
"loss": 1.4060593605041505,
"step": 1650
},
{
"epoch": 0.315508531121914,
"grad_norm": 0.6414536237716675,
"learning_rate": 8.014332435489275e-05,
"loss": 1.3734132766723632,
"step": 1655
},
{
"epoch": 0.3164617291011343,
"grad_norm": 0.6421763896942139,
"learning_rate": 8.0020023724336e-05,
"loss": 1.4304315567016601,
"step": 1660
},
{
"epoch": 0.31741492708035457,
"grad_norm": 0.6136577725410461,
"learning_rate": 7.98964369684964e-05,
"loss": 1.3949074745178223,
"step": 1665
},
{
"epoch": 0.31836812505957485,
"grad_norm": 0.5601505041122437,
"learning_rate": 7.977256526529762e-05,
"loss": 1.334375,
"step": 1670
},
{
"epoch": 0.3193213230387952,
"grad_norm": 0.4467509984970093,
"learning_rate": 7.964840979537917e-05,
"loss": 1.3619421005249024,
"step": 1675
},
{
"epoch": 0.32027452101801546,
"grad_norm": 0.5007718801498413,
"learning_rate": 7.952397174208519e-05,
"loss": 1.331326675415039,
"step": 1680
},
{
"epoch": 0.32122771899723573,
"grad_norm": 0.641846239566803,
"learning_rate": 7.939925229145313e-05,
"loss": 1.3683222770690917,
"step": 1685
},
{
"epoch": 0.322180916976456,
"grad_norm": 0.8255288600921631,
"learning_rate": 7.927425263220252e-05,
"loss": 1.4838945388793945,
"step": 1690
},
{
"epoch": 0.3231341149556763,
"grad_norm": 0.6215581297874451,
"learning_rate": 7.91489739557236e-05,
"loss": 1.385828971862793,
"step": 1695
},
{
"epoch": 0.32408731293489657,
"grad_norm": 0.5082218647003174,
"learning_rate": 7.902341745606594e-05,
"loss": 1.348279094696045,
"step": 1700
},
{
"epoch": 0.32504051091411684,
"grad_norm": 0.4819370210170746,
"learning_rate": 7.889758432992711e-05,
"loss": 1.3172736167907715,
"step": 1705
},
{
"epoch": 0.3259937088933371,
"grad_norm": 0.7424606084823608,
"learning_rate": 7.87714757766412e-05,
"loss": 1.4057170867919921,
"step": 1710
},
{
"epoch": 0.32694690687255745,
"grad_norm": 0.893409252166748,
"learning_rate": 7.864509299816747e-05,
"loss": 1.3335143089294434,
"step": 1715
},
{
"epoch": 0.32790010485177773,
"grad_norm": 1.3331201076507568,
"learning_rate": 7.851843719907888e-05,
"loss": 1.3815743446350097,
"step": 1720
},
{
"epoch": 0.328853302830998,
"grad_norm": 0.6320205926895142,
"learning_rate": 7.839150958655058e-05,
"loss": 1.3576692581176757,
"step": 1725
},
{
"epoch": 0.3298065008102183,
"grad_norm": 0.582838237285614,
"learning_rate": 7.826431137034838e-05,
"loss": 1.4382132530212401,
"step": 1730
},
{
"epoch": 0.33075969878943856,
"grad_norm": 0.561447024345398,
"learning_rate": 7.813684376281729e-05,
"loss": 1.4237019538879394,
"step": 1735
},
{
"epoch": 0.33171289676865884,
"grad_norm": 0.5873743891716003,
"learning_rate": 7.800910797886991e-05,
"loss": 1.3614691734313964,
"step": 1740
},
{
"epoch": 0.3326660947478791,
"grad_norm": 0.6277682781219482,
"learning_rate": 7.788110523597488e-05,
"loss": 1.3796740531921388,
"step": 1745
},
{
"epoch": 0.3336192927270994,
"grad_norm": 0.6571743488311768,
"learning_rate": 7.775283675414524e-05,
"loss": 1.275069808959961,
"step": 1750
},
{
"epoch": 0.3345724907063197,
"grad_norm": 0.6006145477294922,
"learning_rate": 7.762430375592688e-05,
"loss": 1.289985752105713,
"step": 1755
},
{
"epoch": 0.33552568868554,
"grad_norm": 0.7694780230522156,
"learning_rate": 7.749550746638679e-05,
"loss": 1.3432666778564453,
"step": 1760
},
{
"epoch": 0.3364788866647603,
"grad_norm": 0.5260142087936401,
"learning_rate": 7.736644911310144e-05,
"loss": 1.3697632789611816,
"step": 1765
},
{
"epoch": 0.33743208464398056,
"grad_norm": 0.5768413543701172,
"learning_rate": 7.723712992614504e-05,
"loss": 1.3696261405944825,
"step": 1770
},
{
"epoch": 0.33838528262320083,
"grad_norm": 0.4571114182472229,
"learning_rate": 7.710755113807794e-05,
"loss": 1.3056211471557617,
"step": 1775
},
{
"epoch": 0.3393384806024211,
"grad_norm": 0.49580568075180054,
"learning_rate": 7.697771398393466e-05,
"loss": 1.276703453063965,
"step": 1780
},
{
"epoch": 0.3402916785816414,
"grad_norm": 0.56279456615448,
"learning_rate": 7.68476197012123e-05,
"loss": 1.3663080215454102,
"step": 1785
},
{
"epoch": 0.34124487656086167,
"grad_norm": 0.6360243558883667,
"learning_rate": 7.671726952985876e-05,
"loss": 1.3344181060791016,
"step": 1790
},
{
"epoch": 0.342198074540082,
"grad_norm": 0.49206170439720154,
"learning_rate": 7.658666471226073e-05,
"loss": 1.4072901725769043,
"step": 1795
},
{
"epoch": 0.3431512725193023,
"grad_norm": 0.44681113958358765,
"learning_rate": 7.645580649323205e-05,
"loss": 1.37835693359375,
"step": 1800
},
{
"epoch": 0.34410447049852255,
"grad_norm": 0.7092109322547913,
"learning_rate": 7.632469612000173e-05,
"loss": 1.3130971908569335,
"step": 1805
},
{
"epoch": 0.34505766847774283,
"grad_norm": 0.5687137842178345,
"learning_rate": 7.619333484220212e-05,
"loss": 1.2733522415161134,
"step": 1810
},
{
"epoch": 0.3460108664569631,
"grad_norm": 0.5895804762840271,
"learning_rate": 7.6061723911857e-05,
"loss": 1.2574093818664551,
"step": 1815
},
{
"epoch": 0.3469640644361834,
"grad_norm": 0.7050639390945435,
"learning_rate": 7.592986458336958e-05,
"loss": 1.3524909973144532,
"step": 1820
},
{
"epoch": 0.34791726241540366,
"grad_norm": 0.7112770676612854,
"learning_rate": 7.579775811351057e-05,
"loss": 1.3599508285522461,
"step": 1825
},
{
"epoch": 0.34887046039462394,
"grad_norm": 0.7098567485809326,
"learning_rate": 7.566540576140633e-05,
"loss": 1.3201171875,
"step": 1830
},
{
"epoch": 0.34982365837384427,
"grad_norm": 0.5906610488891602,
"learning_rate": 7.553280878852664e-05,
"loss": 1.3277406692504883,
"step": 1835
},
{
"epoch": 0.35077685635306455,
"grad_norm": 0.600531280040741,
"learning_rate": 7.539996845867285e-05,
"loss": 1.3474562644958497,
"step": 1840
},
{
"epoch": 0.3517300543322848,
"grad_norm": 0.598839282989502,
"learning_rate": 7.526688603796578e-05,
"loss": 1.3831668853759767,
"step": 1845
},
{
"epoch": 0.3526832523115051,
"grad_norm": 0.7975032329559326,
"learning_rate": 7.513356279483366e-05,
"loss": 1.3465482711791992,
"step": 1850
},
{
"epoch": 0.3536364502907254,
"grad_norm": 0.6491721868515015,
"learning_rate": 7.500000000000001e-05,
"loss": 1.3161048889160156,
"step": 1855
},
{
"epoch": 0.35458964826994566,
"grad_norm": 0.4972705841064453,
"learning_rate": 7.486619892647158e-05,
"loss": 1.401839828491211,
"step": 1860
},
{
"epoch": 0.35554284624916593,
"grad_norm": 0.5851622223854065,
"learning_rate": 7.473216084952615e-05,
"loss": 1.4101516723632812,
"step": 1865
},
{
"epoch": 0.3564960442283862,
"grad_norm": 0.6072251796722412,
"learning_rate": 7.45978870467005e-05,
"loss": 1.416058349609375,
"step": 1870
},
{
"epoch": 0.35744924220760654,
"grad_norm": 0.5288138389587402,
"learning_rate": 7.446337879777802e-05,
"loss": 1.3266029357910156,
"step": 1875
},
{
"epoch": 0.3584024401868268,
"grad_norm": 0.5310401916503906,
"learning_rate": 7.432863738477675e-05,
"loss": 1.400155258178711,
"step": 1880
},
{
"epoch": 0.3593556381660471,
"grad_norm": 0.48231241106987,
"learning_rate": 7.419366409193701e-05,
"loss": 1.30098876953125,
"step": 1885
},
{
"epoch": 0.3603088361452674,
"grad_norm": 0.6846333742141724,
"learning_rate": 7.40584602057092e-05,
"loss": 1.3059898376464845,
"step": 1890
},
{
"epoch": 0.36126203412448765,
"grad_norm": 0.5746827125549316,
"learning_rate": 7.392302701474151e-05,
"loss": 1.3219364166259766,
"step": 1895
},
{
"epoch": 0.36221523210370793,
"grad_norm": 0.5340244770050049,
"learning_rate": 7.378736580986774e-05,
"loss": 1.415522003173828,
"step": 1900
},
{
"epoch": 0.3631684300829282,
"grad_norm": 0.6440497636795044,
"learning_rate": 7.365147788409485e-05,
"loss": 1.3212701797485351,
"step": 1905
},
{
"epoch": 0.3641216280621485,
"grad_norm": 0.7144160270690918,
"learning_rate": 7.351536453259079e-05,
"loss": 1.2544239997863769,
"step": 1910
},
{
"epoch": 0.3650748260413688,
"grad_norm": 0.8565750122070312,
"learning_rate": 7.337902705267198e-05,
"loss": 1.4466636657714844,
"step": 1915
},
{
"epoch": 0.3660280240205891,
"grad_norm": 0.6030147075653076,
"learning_rate": 7.324246674379112e-05,
"loss": 1.2723915100097656,
"step": 1920
},
{
"epoch": 0.36698122199980937,
"grad_norm": 0.6239649057388306,
"learning_rate": 7.310568490752467e-05,
"loss": 1.417379093170166,
"step": 1925
},
{
"epoch": 0.36793441997902965,
"grad_norm": 0.8152230381965637,
"learning_rate": 7.296868284756057e-05,
"loss": 1.3639314651489258,
"step": 1930
},
{
"epoch": 0.3688876179582499,
"grad_norm": 0.7496471405029297,
"learning_rate": 7.283146186968565e-05,
"loss": 1.333154296875,
"step": 1935
},
{
"epoch": 0.3698408159374702,
"grad_norm": 0.5171647667884827,
"learning_rate": 7.269402328177337e-05,
"loss": 1.3112310409545898,
"step": 1940
},
{
"epoch": 0.3707940139166905,
"grad_norm": 0.6781322956085205,
"learning_rate": 7.255636839377122e-05,
"loss": 1.364420509338379,
"step": 1945
},
{
"epoch": 0.37174721189591076,
"grad_norm": 0.774675190448761,
"learning_rate": 7.241849851768825e-05,
"loss": 1.4390934944152831,
"step": 1950
},
{
"epoch": 0.3727004098751311,
"grad_norm": 0.5805970430374146,
"learning_rate": 7.228041496758265e-05,
"loss": 1.3490175247192382,
"step": 1955
},
{
"epoch": 0.37365360785435137,
"grad_norm": 0.5821093916893005,
"learning_rate": 7.214211905954914e-05,
"loss": 1.3155420303344727,
"step": 1960
},
{
"epoch": 0.37460680583357164,
"grad_norm": 0.6285253763198853,
"learning_rate": 7.200361211170648e-05,
"loss": 1.3470318794250489,
"step": 1965
},
{
"epoch": 0.3755600038127919,
"grad_norm": 0.6625807881355286,
"learning_rate": 7.186489544418483e-05,
"loss": 1.2834866523742676,
"step": 1970
},
{
"epoch": 0.3765132017920122,
"grad_norm": 0.5142903327941895,
"learning_rate": 7.172597037911323e-05,
"loss": 1.3514581680297852,
"step": 1975
},
{
"epoch": 0.3774663997712325,
"grad_norm": 0.5874062180519104,
"learning_rate": 7.158683824060706e-05,
"loss": 1.380577850341797,
"step": 1980
},
{
"epoch": 0.37841959775045275,
"grad_norm": 0.6100233197212219,
"learning_rate": 7.144750035475525e-05,
"loss": 1.400775146484375,
"step": 1985
},
{
"epoch": 0.37937279572967303,
"grad_norm": 0.6496831178665161,
"learning_rate": 7.130795804960783e-05,
"loss": 1.4105688095092774,
"step": 1990
},
{
"epoch": 0.38032599370889336,
"grad_norm": 0.5364809632301331,
"learning_rate": 7.116821265516307e-05,
"loss": 1.362969970703125,
"step": 1995
},
{
"epoch": 0.38127919168811364,
"grad_norm": 0.5413024425506592,
"learning_rate": 7.102826550335503e-05,
"loss": 1.3479754447937011,
"step": 2000
},
{
"epoch": 0.3822323896673339,
"grad_norm": 0.6250216960906982,
"learning_rate": 7.088811792804068e-05,
"loss": 1.3507553100585938,
"step": 2005
},
{
"epoch": 0.3831855876465542,
"grad_norm": 0.565956175327301,
"learning_rate": 7.074777126498728e-05,
"loss": 1.3483230590820312,
"step": 2010
},
{
"epoch": 0.38413878562577447,
"grad_norm": 0.8557022213935852,
"learning_rate": 7.060722685185961e-05,
"loss": 1.3211019515991211,
"step": 2015
},
{
"epoch": 0.38509198360499475,
"grad_norm": 0.6648387908935547,
"learning_rate": 7.046648602820724e-05,
"loss": 1.274226474761963,
"step": 2020
},
{
"epoch": 0.386045181584215,
"grad_norm": 0.6867877840995789,
"learning_rate": 7.032555013545176e-05,
"loss": 1.3185123443603515,
"step": 2025
},
{
"epoch": 0.3869983795634353,
"grad_norm": 0.5501325726509094,
"learning_rate": 7.0184420516874e-05,
"loss": 1.2785281181335448,
"step": 2030
},
{
"epoch": 0.38795157754265563,
"grad_norm": 0.6135473251342773,
"learning_rate": 7.00430985176012e-05,
"loss": 1.2404916763305665,
"step": 2035
},
{
"epoch": 0.3889047755218759,
"grad_norm": 0.8233889937400818,
"learning_rate": 6.990158548459424e-05,
"loss": 1.383447265625,
"step": 2040
},
{
"epoch": 0.3898579735010962,
"grad_norm": 0.8463984131813049,
"learning_rate": 6.975988276663474e-05,
"loss": 1.3010026931762695,
"step": 2045
},
{
"epoch": 0.39081117148031647,
"grad_norm": 0.6072444915771484,
"learning_rate": 6.961799171431225e-05,
"loss": 1.3599861145019532,
"step": 2050
},
{
"epoch": 0.39176436945953674,
"grad_norm": 0.6704472899436951,
"learning_rate": 6.947591368001138e-05,
"loss": 1.3024486541748046,
"step": 2055
},
{
"epoch": 0.392717567438757,
"grad_norm": 0.5563452243804932,
"learning_rate": 6.933365001789882e-05,
"loss": 1.2490873336791992,
"step": 2060
},
{
"epoch": 0.3936707654179773,
"grad_norm": 0.5691594481468201,
"learning_rate": 6.919120208391061e-05,
"loss": 1.4150970458984375,
"step": 2065
},
{
"epoch": 0.3946239633971976,
"grad_norm": 0.6240338087081909,
"learning_rate": 6.904857123573904e-05,
"loss": 1.3098100662231444,
"step": 2070
},
{
"epoch": 0.3955771613764179,
"grad_norm": 0.5713334083557129,
"learning_rate": 6.89057588328198e-05,
"loss": 1.3023897171020509,
"step": 2075
},
{
"epoch": 0.3965303593556382,
"grad_norm": 0.9345459342002869,
"learning_rate": 6.876276623631901e-05,
"loss": 1.369023609161377,
"step": 2080
},
{
"epoch": 0.39748355733485846,
"grad_norm": 0.6146746873855591,
"learning_rate": 6.861959480912027e-05,
"loss": 1.2818717002868651,
"step": 2085
},
{
"epoch": 0.39843675531407874,
"grad_norm": 0.9866006374359131,
"learning_rate": 6.847624591581157e-05,
"loss": 1.3727745056152343,
"step": 2090
},
{
"epoch": 0.399389953293299,
"grad_norm": 0.7147544622421265,
"learning_rate": 6.833272092267241e-05,
"loss": 1.3544178009033203,
"step": 2095
},
{
"epoch": 0.4003431512725193,
"grad_norm": 0.7019825577735901,
"learning_rate": 6.818902119766073e-05,
"loss": 1.2577683448791503,
"step": 2100
},
{
"epoch": 0.40129634925173957,
"grad_norm": 0.7285730242729187,
"learning_rate": 6.804514811039983e-05,
"loss": 1.4660480499267579,
"step": 2105
},
{
"epoch": 0.40224954723095985,
"grad_norm": 0.5752015709877014,
"learning_rate": 6.790110303216537e-05,
"loss": 1.2868889808654784,
"step": 2110
},
{
"epoch": 0.4032027452101802,
"grad_norm": 0.7527331113815308,
"learning_rate": 6.775688733587228e-05,
"loss": 1.3297164916992188,
"step": 2115
},
{
"epoch": 0.40415594318940046,
"grad_norm": 0.5745465755462646,
"learning_rate": 6.761250239606169e-05,
"loss": 1.2335416793823242,
"step": 2120
},
{
"epoch": 0.40510914116862073,
"grad_norm": 0.5933486223220825,
"learning_rate": 6.746794958888777e-05,
"loss": 1.3321504592895508,
"step": 2125
},
{
"epoch": 0.406062339147841,
"grad_norm": 0.6915920972824097,
"learning_rate": 6.73232302921047e-05,
"loss": 1.3351354598999023,
"step": 2130
},
{
"epoch": 0.4070155371270613,
"grad_norm": 0.5966864228248596,
"learning_rate": 6.71783458850535e-05,
"loss": 1.34144287109375,
"step": 2135
},
{
"epoch": 0.40796873510628157,
"grad_norm": 0.5685322284698486,
"learning_rate": 6.703329774864886e-05,
"loss": 1.3440410614013671,
"step": 2140
},
{
"epoch": 0.40892193308550184,
"grad_norm": 0.6365430355072021,
"learning_rate": 6.6888087265366e-05,
"loss": 1.3618284225463868,
"step": 2145
},
{
"epoch": 0.4098751310647221,
"grad_norm": 0.6044729351997375,
"learning_rate": 6.67427158192275e-05,
"loss": 1.302448558807373,
"step": 2150
},
{
"epoch": 0.41082832904394245,
"grad_norm": 0.5664344429969788,
"learning_rate": 6.659718479579009e-05,
"loss": 1.4417905807495117,
"step": 2155
},
{
"epoch": 0.41178152702316273,
"grad_norm": 0.6466103196144104,
"learning_rate": 6.645149558213148e-05,
"loss": 1.3929347038269042,
"step": 2160
},
{
"epoch": 0.412734725002383,
"grad_norm": 0.6547885537147522,
"learning_rate": 6.630564956683708e-05,
"loss": 1.3017690658569336,
"step": 2165
},
{
"epoch": 0.4136879229816033,
"grad_norm": 0.5610964298248291,
"learning_rate": 6.615964813998682e-05,
"loss": 1.3484243392944335,
"step": 2170
},
{
"epoch": 0.41464112096082356,
"grad_norm": 0.5627769231796265,
"learning_rate": 6.601349269314188e-05,
"loss": 1.352993106842041,
"step": 2175
},
{
"epoch": 0.41559431894004384,
"grad_norm": 0.4858628511428833,
"learning_rate": 6.586718461933143e-05,
"loss": 1.3508516311645509,
"step": 2180
},
{
"epoch": 0.4165475169192641,
"grad_norm": 0.6281948089599609,
"learning_rate": 6.572072531303934e-05,
"loss": 1.331397247314453,
"step": 2185
},
{
"epoch": 0.4175007148984844,
"grad_norm": 0.6066363453865051,
"learning_rate": 6.557411617019092e-05,
"loss": 1.2853527069091797,
"step": 2190
},
{
"epoch": 0.4184539128777047,
"grad_norm": 0.6890729665756226,
"learning_rate": 6.542735858813954e-05,
"loss": 1.3202073097229003,
"step": 2195
},
{
"epoch": 0.419407110856925,
"grad_norm": 0.8165856599807739,
"learning_rate": 6.528045396565342e-05,
"loss": 1.353419303894043,
"step": 2200
},
{
"epoch": 0.4203603088361453,
"grad_norm": 0.5850958228111267,
"learning_rate": 6.513340370290224e-05,
"loss": 1.3310669898986816,
"step": 2205
},
{
"epoch": 0.42131350681536556,
"grad_norm": 0.5966853499412537,
"learning_rate": 6.498620920144378e-05,
"loss": 1.372230052947998,
"step": 2210
},
{
"epoch": 0.42226670479458583,
"grad_norm": 0.5578535199165344,
"learning_rate": 6.483887186421058e-05,
"loss": 1.2029244422912597,
"step": 2215
},
{
"epoch": 0.4232199027738061,
"grad_norm": 0.588769793510437,
"learning_rate": 6.46913930954966e-05,
"loss": 1.2575939178466797,
"step": 2220
},
{
"epoch": 0.4241731007530264,
"grad_norm": 0.5605327486991882,
"learning_rate": 6.454377430094376e-05,
"loss": 1.2515182495117188,
"step": 2225
},
{
"epoch": 0.42512629873224667,
"grad_norm": 0.6388108730316162,
"learning_rate": 6.43960168875286e-05,
"loss": 1.2630409240722655,
"step": 2230
},
{
"epoch": 0.426079496711467,
"grad_norm": 0.6354728937149048,
"learning_rate": 6.42481222635489e-05,
"loss": 1.29013090133667,
"step": 2235
},
{
"epoch": 0.4270326946906873,
"grad_norm": 0.6923354268074036,
"learning_rate": 6.410009183861013e-05,
"loss": 1.2888442993164062,
"step": 2240
},
{
"epoch": 0.42798589266990755,
"grad_norm": 0.5017269849777222,
"learning_rate": 6.39519270236122e-05,
"loss": 1.2488239288330079,
"step": 2245
},
{
"epoch": 0.42893909064912783,
"grad_norm": 0.6971740126609802,
"learning_rate": 6.38036292307358e-05,
"loss": 1.3083078384399414,
"step": 2250
},
{
"epoch": 0.4298922886283481,
"grad_norm": 0.7388419508934021,
"learning_rate": 6.365519987342917e-05,
"loss": 1.326636505126953,
"step": 2255
},
{
"epoch": 0.4308454866075684,
"grad_norm": 0.5238728523254395,
"learning_rate": 6.350664036639442e-05,
"loss": 1.2866081237792968,
"step": 2260
},
{
"epoch": 0.43179868458678866,
"grad_norm": 0.7681960463523865,
"learning_rate": 6.335795212557417e-05,
"loss": 1.3370766639709473,
"step": 2265
},
{
"epoch": 0.43275188256600894,
"grad_norm": 0.4777337610721588,
"learning_rate": 6.320913656813802e-05,
"loss": 1.3494007110595703,
"step": 2270
},
{
"epoch": 0.43370508054522927,
"grad_norm": 0.5424591898918152,
"learning_rate": 6.306019511246905e-05,
"loss": 1.2188060760498047,
"step": 2275
},
{
"epoch": 0.43465827852444955,
"grad_norm": 0.7049203515052795,
"learning_rate": 6.291112917815027e-05,
"loss": 1.226613426208496,
"step": 2280
},
{
"epoch": 0.4356114765036698,
"grad_norm": 0.7803995609283447,
"learning_rate": 6.276194018595114e-05,
"loss": 1.2571246147155761,
"step": 2285
},
{
"epoch": 0.4365646744828901,
"grad_norm": 0.7541532516479492,
"learning_rate": 6.261262955781398e-05,
"loss": 1.2014997482299805,
"step": 2290
},
{
"epoch": 0.4375178724621104,
"grad_norm": 0.6911846399307251,
"learning_rate": 6.246319871684048e-05,
"loss": 1.3261570930480957,
"step": 2295
},
{
"epoch": 0.43847107044133066,
"grad_norm": 0.5578048229217529,
"learning_rate": 6.231364908727806e-05,
"loss": 1.3646205902099608,
"step": 2300
},
{
"epoch": 0.43942426842055093,
"grad_norm": 0.5710150003433228,
"learning_rate": 6.216398209450633e-05,
"loss": 1.2794734954833984,
"step": 2305
},
{
"epoch": 0.4403774663997712,
"grad_norm": 0.6587945818901062,
"learning_rate": 6.201419916502351e-05,
"loss": 1.3150693893432617,
"step": 2310
},
{
"epoch": 0.44133066437899154,
"grad_norm": 0.7044800519943237,
"learning_rate": 6.186430172643288e-05,
"loss": 1.3477896690368651,
"step": 2315
},
{
"epoch": 0.4422838623582118,
"grad_norm": 0.5416322350502014,
"learning_rate": 6.171429120742903e-05,
"loss": 1.3167153358459474,
"step": 2320
},
{
"epoch": 0.4432370603374321,
"grad_norm": 0.8006619811058044,
"learning_rate": 6.156416903778443e-05,
"loss": 1.2461322784423827,
"step": 2325
},
{
"epoch": 0.4441902583166524,
"grad_norm": 0.7887236475944519,
"learning_rate": 6.141393664833562e-05,
"loss": 1.322500228881836,
"step": 2330
},
{
"epoch": 0.44514345629587265,
"grad_norm": 0.6436631083488464,
"learning_rate": 6.126359547096975e-05,
"loss": 1.2609729766845703,
"step": 2335
},
{
"epoch": 0.44609665427509293,
"grad_norm": 0.5567827820777893,
"learning_rate": 6.111314693861078e-05,
"loss": 1.3153230667114257,
"step": 2340
},
{
"epoch": 0.4470498522543132,
"grad_norm": 0.6213957071304321,
"learning_rate": 6.096259248520591e-05,
"loss": 1.2853681564331054,
"step": 2345
},
{
"epoch": 0.4480030502335335,
"grad_norm": 0.7204575538635254,
"learning_rate": 6.081193354571191e-05,
"loss": 1.3712921142578125,
"step": 2350
},
{
"epoch": 0.44895624821275376,
"grad_norm": 0.5651723742485046,
"learning_rate": 6.066117155608135e-05,
"loss": 1.2652931213378906,
"step": 2355
},
{
"epoch": 0.4499094461919741,
"grad_norm": 0.48197153210639954,
"learning_rate": 6.051030795324906e-05,
"loss": 1.3778664588928222,
"step": 2360
},
{
"epoch": 0.45086264417119437,
"grad_norm": 0.5710497498512268,
"learning_rate": 6.035934417511834e-05,
"loss": 1.2601179122924804,
"step": 2365
},
{
"epoch": 0.45181584215041465,
"grad_norm": 0.6569958925247192,
"learning_rate": 6.020828166054726e-05,
"loss": 1.2517206192016601,
"step": 2370
},
{
"epoch": 0.4527690401296349,
"grad_norm": 0.8211113810539246,
"learning_rate": 6.0057121849334975e-05,
"loss": 1.3188791275024414,
"step": 2375
},
{
"epoch": 0.4537222381088552,
"grad_norm": 0.8274943828582764,
"learning_rate": 5.9905866182207994e-05,
"loss": 1.2817185401916504,
"step": 2380
},
{
"epoch": 0.4546754360880755,
"grad_norm": 0.9456292986869812,
"learning_rate": 5.9754516100806423e-05,
"loss": 1.3292264938354492,
"step": 2385
},
{
"epoch": 0.45562863406729576,
"grad_norm": 0.552264928817749,
"learning_rate": 5.9603073047670253e-05,
"loss": 1.3574498176574707,
"step": 2390
},
{
"epoch": 0.45658183204651603,
"grad_norm": 0.8571290969848633,
"learning_rate": 5.945153846622563e-05,
"loss": 1.3702621459960938,
"step": 2395
},
{
"epoch": 0.45753503002573637,
"grad_norm": 0.593041181564331,
"learning_rate": 5.929991380077102e-05,
"loss": 1.2561967849731446,
"step": 2400
},
{
"epoch": 0.45848822800495664,
"grad_norm": 0.7197123169898987,
"learning_rate": 5.9148200496463524e-05,
"loss": 1.2986156463623046,
"step": 2405
},
{
"epoch": 0.4594414259841769,
"grad_norm": 0.7309663891792297,
"learning_rate": 5.8996399999305084e-05,
"loss": 1.2270036697387696,
"step": 2410
},
{
"epoch": 0.4603946239633972,
"grad_norm": 0.5782007575035095,
"learning_rate": 5.884451375612865e-05,
"loss": 1.2581504821777343,
"step": 2415
},
{
"epoch": 0.4613478219426175,
"grad_norm": 0.7056921720504761,
"learning_rate": 5.869254321458447e-05,
"loss": 1.310012435913086,
"step": 2420
},
{
"epoch": 0.46230101992183775,
"grad_norm": 0.5803052186965942,
"learning_rate": 5.854048982312623e-05,
"loss": 1.3429342269897462,
"step": 2425
},
{
"epoch": 0.46325421790105803,
"grad_norm": 0.9722602367401123,
"learning_rate": 5.838835503099727e-05,
"loss": 1.293086624145508,
"step": 2430
},
{
"epoch": 0.4642074158802783,
"grad_norm": 0.5890440940856934,
"learning_rate": 5.8236140288216776e-05,
"loss": 1.3354226112365724,
"step": 2435
},
{
"epoch": 0.46516061385949864,
"grad_norm": 0.5847529768943787,
"learning_rate": 5.8083847045565964e-05,
"loss": 1.2724483489990235,
"step": 2440
},
{
"epoch": 0.4661138118387189,
"grad_norm": 0.5705927014350891,
"learning_rate": 5.7931476754574246e-05,
"loss": 1.246432113647461,
"step": 2445
},
{
"epoch": 0.4670670098179392,
"grad_norm": 0.6440657377243042,
"learning_rate": 5.777903086750537e-05,
"loss": 1.3789167404174805,
"step": 2450
},
{
"epoch": 0.46802020779715947,
"grad_norm": 0.6465033888816833,
"learning_rate": 5.762651083734363e-05,
"loss": 1.3318554878234863,
"step": 2455
},
{
"epoch": 0.46897340577637975,
"grad_norm": 0.53423672914505,
"learning_rate": 5.747391811777996e-05,
"loss": 1.3092180252075196,
"step": 2460
},
{
"epoch": 0.4699266037556,
"grad_norm": 0.5637331008911133,
"learning_rate": 5.7321254163198126e-05,
"loss": 1.2770545959472657,
"step": 2465
},
{
"epoch": 0.4708798017348203,
"grad_norm": 0.769513726234436,
"learning_rate": 5.716852042866083e-05,
"loss": 1.3958868026733398,
"step": 2470
},
{
"epoch": 0.4718329997140406,
"grad_norm": 0.5739859342575073,
"learning_rate": 5.701571836989591e-05,
"loss": 1.338501262664795,
"step": 2475
},
{
"epoch": 0.4727861976932609,
"grad_norm": 0.6028321981430054,
"learning_rate": 5.686284944328233e-05,
"loss": 1.277390480041504,
"step": 2480
},
{
"epoch": 0.4737393956724812,
"grad_norm": 0.7873733639717102,
"learning_rate": 5.670991510583643e-05,
"loss": 1.3972469329833985,
"step": 2485
},
{
"epoch": 0.47469259365170147,
"grad_norm": 0.5674543976783752,
"learning_rate": 5.6556916815198004e-05,
"loss": 1.2728700637817383,
"step": 2490
},
{
"epoch": 0.47564579163092174,
"grad_norm": 0.5882276892662048,
"learning_rate": 5.640385602961634e-05,
"loss": 1.2748459815979003,
"step": 2495
},
{
"epoch": 0.476598989610142,
"grad_norm": 0.6623244285583496,
"learning_rate": 5.625073420793644e-05,
"loss": 1.3471666336059571,
"step": 2500
},
{
"epoch": 0.4775521875893623,
"grad_norm": 0.549636721611023,
"learning_rate": 5.6097552809584996e-05,
"loss": 1.268274974822998,
"step": 2505
},
{
"epoch": 0.4785053855685826,
"grad_norm": 0.5313766598701477,
"learning_rate": 5.594431329455655e-05,
"loss": 1.2008712768554688,
"step": 2510
},
{
"epoch": 0.47945858354780285,
"grad_norm": 0.6205539107322693,
"learning_rate": 5.579101712339958e-05,
"loss": 1.2783683776855468,
"step": 2515
},
{
"epoch": 0.4804117815270232,
"grad_norm": 1.1922192573547363,
"learning_rate": 5.563766575720254e-05,
"loss": 1.3177761077880858,
"step": 2520
},
{
"epoch": 0.48136497950624346,
"grad_norm": 0.5861597657203674,
"learning_rate": 5.5484260657579965e-05,
"loss": 1.1803934097290039,
"step": 2525
},
{
"epoch": 0.48231817748546374,
"grad_norm": 1.1449154615402222,
"learning_rate": 5.533080328665855e-05,
"loss": 1.2816280364990233,
"step": 2530
},
{
"epoch": 0.483271375464684,
"grad_norm": 0.6446986198425293,
"learning_rate": 5.517729510706315e-05,
"loss": 1.3410715103149413,
"step": 2535
},
{
"epoch": 0.4842245734439043,
"grad_norm": 0.7727839946746826,
"learning_rate": 5.5023737581902935e-05,
"loss": 1.2542783737182617,
"step": 2540
},
{
"epoch": 0.48517777142312457,
"grad_norm": 0.6695873141288757,
"learning_rate": 5.487013217475735e-05,
"loss": 1.2204108238220215,
"step": 2545
},
{
"epoch": 0.48613096940234485,
"grad_norm": 0.8227330446243286,
"learning_rate": 5.471648034966225e-05,
"loss": 1.3047113418579102,
"step": 2550
},
{
"epoch": 0.4870841673815651,
"grad_norm": 0.63616544008255,
"learning_rate": 5.456278357109589e-05,
"loss": 1.2772872924804688,
"step": 2555
},
{
"epoch": 0.48803736536078546,
"grad_norm": 0.6261322498321533,
"learning_rate": 5.4409043303964954e-05,
"loss": 1.3274253845214843,
"step": 2560
},
{
"epoch": 0.48899056334000573,
"grad_norm": 0.6103684306144714,
"learning_rate": 5.425526101359067e-05,
"loss": 1.3406309127807616,
"step": 2565
},
{
"epoch": 0.489943761319226,
"grad_norm": 0.6387801170349121,
"learning_rate": 5.41014381656948e-05,
"loss": 1.3044708251953125,
"step": 2570
},
{
"epoch": 0.4908969592984463,
"grad_norm": 0.559608519077301,
"learning_rate": 5.39475762263856e-05,
"loss": 1.2652557373046875,
"step": 2575
},
{
"epoch": 0.49185015727766657,
"grad_norm": 0.5396852493286133,
"learning_rate": 5.3793676662143965e-05,
"loss": 1.3324545860290526,
"step": 2580
},
{
"epoch": 0.49280335525688684,
"grad_norm": 0.672410786151886,
"learning_rate": 5.3639740939809405e-05,
"loss": 1.2873205184936523,
"step": 2585
},
{
"epoch": 0.4937565532361071,
"grad_norm": 0.5460557341575623,
"learning_rate": 5.3485770526566e-05,
"loss": 1.238821792602539,
"step": 2590
},
{
"epoch": 0.4947097512153274,
"grad_norm": 0.48951706290245056,
"learning_rate": 5.3331766889928556e-05,
"loss": 1.2931809425354004,
"step": 2595
},
{
"epoch": 0.49566294919454773,
"grad_norm": 0.6401025652885437,
"learning_rate": 5.3177731497728475e-05,
"loss": 1.2673799514770507,
"step": 2600
},
{
"epoch": 0.496616147173768,
"grad_norm": 0.5934286117553711,
"learning_rate": 5.302366581809984e-05,
"loss": 1.2936947822570801,
"step": 2605
},
{
"epoch": 0.4975693451529883,
"grad_norm": 0.6554251313209534,
"learning_rate": 5.2869571319465435e-05,
"loss": 1.30528564453125,
"step": 2610
},
{
"epoch": 0.49852254313220856,
"grad_norm": 0.5471514463424683,
"learning_rate": 5.271544947052267e-05,
"loss": 1.1838308334350587,
"step": 2615
},
{
"epoch": 0.49947574111142884,
"grad_norm": 0.7644234895706177,
"learning_rate": 5.2561301740229686e-05,
"loss": 1.283384418487549,
"step": 2620
},
{
"epoch": 0.5004289390906491,
"grad_norm": 0.5817055106163025,
"learning_rate": 5.2407129597791295e-05,
"loss": 1.2347601890563964,
"step": 2625
},
{
"epoch": 0.5013821370698694,
"grad_norm": 0.6453908085823059,
"learning_rate": 5.2252934512644935e-05,
"loss": 1.3328070640563965,
"step": 2630
},
{
"epoch": 0.5023353350490897,
"grad_norm": 0.532061755657196,
"learning_rate": 5.2098717954446784e-05,
"loss": 1.333824920654297,
"step": 2635
},
{
"epoch": 0.50328853302831,
"grad_norm": 0.6687843203544617,
"learning_rate": 5.194448139305762e-05,
"loss": 1.3482172012329101,
"step": 2640
},
{
"epoch": 0.5042417310075302,
"grad_norm": 0.6556944847106934,
"learning_rate": 5.1790226298528924e-05,
"loss": 1.257643985748291,
"step": 2645
},
{
"epoch": 0.5051949289867506,
"grad_norm": 0.6609522104263306,
"learning_rate": 5.1635954141088813e-05,
"loss": 1.2046517372131347,
"step": 2650
},
{
"epoch": 0.5061481269659708,
"grad_norm": 0.6581369042396545,
"learning_rate": 5.148166639112799e-05,
"loss": 1.256479263305664,
"step": 2655
},
{
"epoch": 0.5071013249451911,
"grad_norm": 0.5723243951797485,
"learning_rate": 5.132736451918582e-05,
"loss": 1.3054783821105957,
"step": 2660
},
{
"epoch": 0.5080545229244114,
"grad_norm": 0.6024762392044067,
"learning_rate": 5.117304999593625e-05,
"loss": 1.2113540649414063,
"step": 2665
},
{
"epoch": 0.5090077209036317,
"grad_norm": 0.6620156168937683,
"learning_rate": 5.101872429217379e-05,
"loss": 1.2767284393310547,
"step": 2670
},
{
"epoch": 0.509960918882852,
"grad_norm": 0.5391150712966919,
"learning_rate": 5.086438887879955e-05,
"loss": 1.3419584274291991,
"step": 2675
},
{
"epoch": 0.5109141168620722,
"grad_norm": 0.6278124451637268,
"learning_rate": 5.0710045226807157e-05,
"loss": 1.2554492950439453,
"step": 2680
},
{
"epoch": 0.5118673148412926,
"grad_norm": 0.677617073059082,
"learning_rate": 5.055569480726876e-05,
"loss": 1.2829187393188477,
"step": 2685
},
{
"epoch": 0.5128205128205128,
"grad_norm": 0.6371073126792908,
"learning_rate": 5.040133909132104e-05,
"loss": 1.2125523567199707,
"step": 2690
},
{
"epoch": 0.5137737107997331,
"grad_norm": 0.7256470918655396,
"learning_rate": 5.0246979550151116e-05,
"loss": 1.2278175354003906,
"step": 2695
},
{
"epoch": 0.5147269087789534,
"grad_norm": 0.6447271704673767,
"learning_rate": 5.00926176549826e-05,
"loss": 1.3057592391967774,
"step": 2700
},
{
"epoch": 0.5156801067581737,
"grad_norm": 0.6097456216812134,
"learning_rate": 4.993825487706153e-05,
"loss": 1.2595368385314942,
"step": 2705
},
{
"epoch": 0.516633304737394,
"grad_norm": 0.5032343864440918,
"learning_rate": 4.978389268764236e-05,
"loss": 1.2212257385253906,
"step": 2710
},
{
"epoch": 0.5175865027166142,
"grad_norm": 0.5995299816131592,
"learning_rate": 4.96295325579739e-05,
"loss": 1.3145025253295899,
"step": 2715
},
{
"epoch": 0.5185397006958345,
"grad_norm": 0.6602983474731445,
"learning_rate": 4.9475175959285394e-05,
"loss": 1.326936626434326,
"step": 2720
},
{
"epoch": 0.5194928986750548,
"grad_norm": 0.6034419536590576,
"learning_rate": 4.9320824362772396e-05,
"loss": 1.3041457176208495,
"step": 2725
},
{
"epoch": 0.5204460966542751,
"grad_norm": 0.8866453170776367,
"learning_rate": 4.916647923958276e-05,
"loss": 1.2453497886657714,
"step": 2730
},
{
"epoch": 0.5213992946334953,
"grad_norm": 0.5730390548706055,
"learning_rate": 4.901214206080269e-05,
"loss": 1.1771400451660157,
"step": 2735
},
{
"epoch": 0.5223524926127157,
"grad_norm": 0.7555006146430969,
"learning_rate": 4.885781429744263e-05,
"loss": 1.2875048637390136,
"step": 2740
},
{
"epoch": 0.523305690591936,
"grad_norm": 0.8437433838844299,
"learning_rate": 4.870349742042328e-05,
"loss": 1.300255584716797,
"step": 2745
},
{
"epoch": 0.5242588885711562,
"grad_norm": 0.7320507168769836,
"learning_rate": 4.8549192900561666e-05,
"loss": 1.2990969657897948,
"step": 2750
},
{
"epoch": 0.5252120865503765,
"grad_norm": 0.714146077632904,
"learning_rate": 4.839490220855692e-05,
"loss": 1.2314547538757323,
"step": 2755
},
{
"epoch": 0.5261652845295968,
"grad_norm": 0.6077867746353149,
"learning_rate": 4.824062681497643e-05,
"loss": 1.3071678161621094,
"step": 2760
},
{
"epoch": 0.5271184825088171,
"grad_norm": 0.615460216999054,
"learning_rate": 4.8086368190241804e-05,
"loss": 1.2812042236328125,
"step": 2765
},
{
"epoch": 0.5280716804880373,
"grad_norm": 0.572293221950531,
"learning_rate": 4.793212780461478e-05,
"loss": 1.2879579544067383,
"step": 2770
},
{
"epoch": 0.5290248784672577,
"grad_norm": 0.5470404028892517,
"learning_rate": 4.777790712818324e-05,
"loss": 1.2061164855957032,
"step": 2775
},
{
"epoch": 0.529978076446478,
"grad_norm": 0.7726988792419434,
"learning_rate": 4.7623707630847285e-05,
"loss": 1.2886631965637207,
"step": 2780
},
{
"epoch": 0.5309312744256982,
"grad_norm": 0.5666267275810242,
"learning_rate": 4.74695307823051e-05,
"loss": 1.2591033935546876,
"step": 2785
},
{
"epoch": 0.5318844724049185,
"grad_norm": 0.799761176109314,
"learning_rate": 4.731537805203901e-05,
"loss": 1.2433523178100585,
"step": 2790
},
{
"epoch": 0.5328376703841388,
"grad_norm": 0.5671012997627258,
"learning_rate": 4.716125090930146e-05,
"loss": 1.302900505065918,
"step": 2795
},
{
"epoch": 0.5337908683633591,
"grad_norm": 0.6657522320747375,
"learning_rate": 4.700715082310103e-05,
"loss": 1.296823215484619,
"step": 2800
},
{
"epoch": 0.5347440663425793,
"grad_norm": 0.7177772521972656,
"learning_rate": 4.685307926218844e-05,
"loss": 1.2669149398803712,
"step": 2805
},
{
"epoch": 0.5356972643217996,
"grad_norm": 0.6294072866439819,
"learning_rate": 4.669903769504248e-05,
"loss": 1.2358671188354493,
"step": 2810
},
{
"epoch": 0.5366504623010199,
"grad_norm": 0.5263190269470215,
"learning_rate": 4.654502758985611e-05,
"loss": 1.2559940338134765,
"step": 2815
},
{
"epoch": 0.5376036602802402,
"grad_norm": 0.7350649833679199,
"learning_rate": 4.63910504145224e-05,
"loss": 1.2670574188232422,
"step": 2820
},
{
"epoch": 0.5385568582594605,
"grad_norm": 0.6216763257980347,
"learning_rate": 4.6237107636620535e-05,
"loss": 1.2593679428100586,
"step": 2825
},
{
"epoch": 0.5395100562386808,
"grad_norm": 0.647221028804779,
"learning_rate": 4.6083200723401935e-05,
"loss": 1.2642337799072265,
"step": 2830
},
{
"epoch": 0.5404632542179011,
"grad_norm": 0.6106157302856445,
"learning_rate": 4.59293311417761e-05,
"loss": 1.3094030380249024,
"step": 2835
},
{
"epoch": 0.5414164521971213,
"grad_norm": 0.6650776267051697,
"learning_rate": 4.5775500358296755e-05,
"loss": 1.3401853561401367,
"step": 2840
},
{
"epoch": 0.5423696501763416,
"grad_norm": 0.7189481854438782,
"learning_rate": 4.562170983914785e-05,
"loss": 1.3148399353027345,
"step": 2845
},
{
"epoch": 0.5433228481555619,
"grad_norm": 0.7908103466033936,
"learning_rate": 4.546796105012955e-05,
"loss": 1.185401153564453,
"step": 2850
},
{
"epoch": 0.5442760461347822,
"grad_norm": 0.5463522672653198,
"learning_rate": 4.531425545664425e-05,
"loss": 1.2353577613830566,
"step": 2855
},
{
"epoch": 0.5452292441140025,
"grad_norm": 0.7208181023597717,
"learning_rate": 4.5160594523682734e-05,
"loss": 1.1626194953918456,
"step": 2860
},
{
"epoch": 0.5461824420932228,
"grad_norm": 0.6463826298713684,
"learning_rate": 4.5006979715810034e-05,
"loss": 1.2271671295166016,
"step": 2865
},
{
"epoch": 0.5471356400724431,
"grad_norm": 0.5353156328201294,
"learning_rate": 4.485341249715157e-05,
"loss": 1.2597745895385741,
"step": 2870
},
{
"epoch": 0.5480888380516633,
"grad_norm": 0.5640683174133301,
"learning_rate": 4.469989433137923e-05,
"loss": 1.2277194023132325,
"step": 2875
},
{
"epoch": 0.5490420360308836,
"grad_norm": 0.6965321898460388,
"learning_rate": 4.454642668169732e-05,
"loss": 1.3282282829284668,
"step": 2880
},
{
"epoch": 0.5499952340101039,
"grad_norm": 0.5201863050460815,
"learning_rate": 4.439301101082871e-05,
"loss": 1.1863132476806642,
"step": 2885
},
{
"epoch": 0.5509484319893242,
"grad_norm": 0.7312695384025574,
"learning_rate": 4.423964878100081e-05,
"loss": 1.3062297821044921,
"step": 2890
},
{
"epoch": 0.5519016299685444,
"grad_norm": 0.6597520112991333,
"learning_rate": 4.4086341453931716e-05,
"loss": 1.3157255172729492,
"step": 2895
},
{
"epoch": 0.5528548279477647,
"grad_norm": 0.8693311810493469,
"learning_rate": 4.3933090490816217e-05,
"loss": 1.2623749732971192,
"step": 2900
},
{
"epoch": 0.5538080259269851,
"grad_norm": 0.8025595545768738,
"learning_rate": 4.3779897352311864e-05,
"loss": 1.2473636627197267,
"step": 2905
},
{
"epoch": 0.5547612239062053,
"grad_norm": 0.8665235638618469,
"learning_rate": 4.362676349852515e-05,
"loss": 1.2450160026550292,
"step": 2910
},
{
"epoch": 0.5557144218854256,
"grad_norm": 0.5059013962745667,
"learning_rate": 4.347369038899744e-05,
"loss": 1.3125247955322266,
"step": 2915
},
{
"epoch": 0.5566676198646459,
"grad_norm": 0.5233684182167053,
"learning_rate": 4.332067948269113e-05,
"loss": 1.282141876220703,
"step": 2920
},
{
"epoch": 0.5576208178438662,
"grad_norm": 0.6547322273254395,
"learning_rate": 4.316773223797585e-05,
"loss": 1.325181007385254,
"step": 2925
},
{
"epoch": 0.5585740158230864,
"grad_norm": 0.6160426139831543,
"learning_rate": 4.3014850112614355e-05,
"loss": 1.227367115020752,
"step": 2930
},
{
"epoch": 0.5595272138023067,
"grad_norm": 0.6052421927452087,
"learning_rate": 4.286203456374877e-05,
"loss": 1.2693682670593263,
"step": 2935
},
{
"epoch": 0.5604804117815271,
"grad_norm": 0.9350481033325195,
"learning_rate": 4.27092870478867e-05,
"loss": 1.2604055404663086,
"step": 2940
},
{
"epoch": 0.5614336097607473,
"grad_norm": 0.6442693471908569,
"learning_rate": 4.255660902088729e-05,
"loss": 1.2549070358276366,
"step": 2945
},
{
"epoch": 0.5623868077399676,
"grad_norm": 0.8175579309463501,
"learning_rate": 4.240400193794736e-05,
"loss": 1.3217493057250977,
"step": 2950
},
{
"epoch": 0.5633400057191879,
"grad_norm": 0.6251521706581116,
"learning_rate": 4.225146725358759e-05,
"loss": 1.159871482849121,
"step": 2955
},
{
"epoch": 0.5642932036984082,
"grad_norm": 0.8438305854797363,
"learning_rate": 4.209900642163859e-05,
"loss": 1.3238512039184571,
"step": 2960
},
{
"epoch": 0.5652464016776284,
"grad_norm": 0.7203558683395386,
"learning_rate": 4.194662089522707e-05,
"loss": 1.2954622268676759,
"step": 2965
},
{
"epoch": 0.5661995996568487,
"grad_norm": 0.5530749559402466,
"learning_rate": 4.1794312126762025e-05,
"loss": 1.296596145629883,
"step": 2970
},
{
"epoch": 0.567152797636069,
"grad_norm": 0.5645231008529663,
"learning_rate": 4.1642081567920846e-05,
"loss": 1.2336214065551758,
"step": 2975
},
{
"epoch": 0.5681059956152893,
"grad_norm": 0.5967411994934082,
"learning_rate": 4.148993066963546e-05,
"loss": 1.2412754058837892,
"step": 2980
},
{
"epoch": 0.5690591935945096,
"grad_norm": 0.5813344120979309,
"learning_rate": 4.133786088207857e-05,
"loss": 1.3021140098571777,
"step": 2985
},
{
"epoch": 0.5700123915737298,
"grad_norm": 1.3175607919692993,
"learning_rate": 4.118587365464981e-05,
"loss": 1.3055715560913086,
"step": 2990
},
{
"epoch": 0.5709655895529502,
"grad_norm": 0.616934061050415,
"learning_rate": 4.103397043596191e-05,
"loss": 1.2695520401000977,
"step": 2995
},
{
"epoch": 0.5719187875321704,
"grad_norm": 0.6335171461105347,
"learning_rate": 4.0882152673826854e-05,
"loss": 1.225665283203125,
"step": 3000
},
{
"epoch": 0.5728719855113907,
"grad_norm": 0.8340386748313904,
"learning_rate": 4.073042181524224e-05,
"loss": 1.301685619354248,
"step": 3005
},
{
"epoch": 0.573825183490611,
"grad_norm": 0.7248563766479492,
"learning_rate": 4.057877930637726e-05,
"loss": 1.2224946975708009,
"step": 3010
},
{
"epoch": 0.5747783814698313,
"grad_norm": 0.940913736820221,
"learning_rate": 4.0427226592559065e-05,
"loss": 1.244685173034668,
"step": 3015
},
{
"epoch": 0.5757315794490516,
"grad_norm": 0.8180844783782959,
"learning_rate": 4.0275765118259016e-05,
"loss": 1.199371337890625,
"step": 3020
},
{
"epoch": 0.5766847774282718,
"grad_norm": 0.7802680730819702,
"learning_rate": 4.012439632707878e-05,
"loss": 1.2432506561279297,
"step": 3025
},
{
"epoch": 0.5776379754074922,
"grad_norm": 0.6335556507110596,
"learning_rate": 3.9973121661736666e-05,
"loss": 1.2290749549865723,
"step": 3030
},
{
"epoch": 0.5785911733867124,
"grad_norm": 0.6272526383399963,
"learning_rate": 3.98219425640539e-05,
"loss": 1.2388141632080079,
"step": 3035
},
{
"epoch": 0.5795443713659327,
"grad_norm": 0.8064735531806946,
"learning_rate": 3.967086047494078e-05,
"loss": 1.2962579727172852,
"step": 3040
},
{
"epoch": 0.580497569345153,
"grad_norm": 0.8494065403938293,
"learning_rate": 3.9519876834383014e-05,
"loss": 1.2536334991455078,
"step": 3045
},
{
"epoch": 0.5814507673243733,
"grad_norm": 0.9386687278747559,
"learning_rate": 3.936899308142803e-05,
"loss": 1.302174949645996,
"step": 3050
},
{
"epoch": 0.5824039653035935,
"grad_norm": 0.9092950820922852,
"learning_rate": 3.921821065417116e-05,
"loss": 1.293523406982422,
"step": 3055
},
{
"epoch": 0.5833571632828138,
"grad_norm": 1.1398800611495972,
"learning_rate": 3.9067530989742e-05,
"loss": 1.2533843994140625,
"step": 3060
},
{
"epoch": 0.5843103612620342,
"grad_norm": 0.6835266351699829,
"learning_rate": 3.8916955524290684e-05,
"loss": 1.2517004013061523,
"step": 3065
},
{
"epoch": 0.5852635592412544,
"grad_norm": 0.6582618355751038,
"learning_rate": 3.8766485692974264e-05,
"loss": 1.1177049636840821,
"step": 3070
},
{
"epoch": 0.5862167572204747,
"grad_norm": 0.5887477397918701,
"learning_rate": 3.861612292994292e-05,
"loss": 1.2374506950378419,
"step": 3075
},
{
"epoch": 0.587169955199695,
"grad_norm": 0.7007836103439331,
"learning_rate": 3.846586866832635e-05,
"loss": 1.3616958618164063,
"step": 3080
},
{
"epoch": 0.5881231531789153,
"grad_norm": 0.76991868019104,
"learning_rate": 3.831572434022017e-05,
"loss": 1.1976142883300782,
"step": 3085
},
{
"epoch": 0.5890763511581355,
"grad_norm": 0.8526211977005005,
"learning_rate": 3.816569137667214e-05,
"loss": 1.3073230743408204,
"step": 3090
},
{
"epoch": 0.5900295491373558,
"grad_norm": 0.7521684765815735,
"learning_rate": 3.8015771207668596e-05,
"loss": 1.3025712966918945,
"step": 3095
},
{
"epoch": 0.5909827471165762,
"grad_norm": 0.7034772634506226,
"learning_rate": 3.786596526212084e-05,
"loss": 1.2449336051940918,
"step": 3100
},
{
"epoch": 0.5919359450957964,
"grad_norm": 0.8000289797782898,
"learning_rate": 3.771627496785148e-05,
"loss": 1.2548463821411133,
"step": 3105
},
{
"epoch": 0.5928891430750167,
"grad_norm": 0.6360304951667786,
"learning_rate": 3.756670175158083e-05,
"loss": 1.3065912246704101,
"step": 3110
},
{
"epoch": 0.5938423410542369,
"grad_norm": 0.6742876172065735,
"learning_rate": 3.741724703891333e-05,
"loss": 1.2310423851013184,
"step": 3115
},
{
"epoch": 0.5947955390334573,
"grad_norm": 0.5950530767440796,
"learning_rate": 3.726791225432392e-05,
"loss": 1.220978355407715,
"step": 3120
},
{
"epoch": 0.5957487370126775,
"grad_norm": 0.7483922243118286,
"learning_rate": 3.711869882114449e-05,
"loss": 1.2474722862243652,
"step": 3125
},
{
"epoch": 0.5967019349918978,
"grad_norm": 0.6180773973464966,
"learning_rate": 3.696960816155036e-05,
"loss": 1.2909536361694336,
"step": 3130
},
{
"epoch": 0.597655132971118,
"grad_norm": 0.8564450740814209,
"learning_rate": 3.682064169654663e-05,
"loss": 1.2528623580932616,
"step": 3135
},
{
"epoch": 0.5986083309503384,
"grad_norm": 0.6046072840690613,
"learning_rate": 3.667180084595468e-05,
"loss": 1.261697769165039,
"step": 3140
},
{
"epoch": 0.5995615289295587,
"grad_norm": 0.7372415661811829,
"learning_rate": 3.652308702839864e-05,
"loss": 1.2479530334472657,
"step": 3145
},
{
"epoch": 0.6005147269087789,
"grad_norm": 0.6007838249206543,
"learning_rate": 3.637450166129193e-05,
"loss": 1.251002025604248,
"step": 3150
},
{
"epoch": 0.6014679248879993,
"grad_norm": 0.9622913002967834,
"learning_rate": 3.622604616082361e-05,
"loss": 1.2388505935668945,
"step": 3155
},
{
"epoch": 0.6024211228672195,
"grad_norm": 0.8016634583473206,
"learning_rate": 3.607772194194499e-05,
"loss": 1.2663208961486816,
"step": 3160
},
{
"epoch": 0.6033743208464398,
"grad_norm": 0.5737648010253906,
"learning_rate": 3.592953041835615e-05,
"loss": 1.3093027114868163,
"step": 3165
},
{
"epoch": 0.60432751882566,
"grad_norm": 0.7579228281974792,
"learning_rate": 3.5781473002492393e-05,
"loss": 1.2580357551574708,
"step": 3170
},
{
"epoch": 0.6052807168048804,
"grad_norm": 0.7541826963424683,
"learning_rate": 3.563355110551081e-05,
"loss": 1.3639699935913085,
"step": 3175
},
{
"epoch": 0.6062339147841006,
"grad_norm": 0.614702582359314,
"learning_rate": 3.5485766137276894e-05,
"loss": 1.3074114799499512,
"step": 3180
},
{
"epoch": 0.6071871127633209,
"grad_norm": 0.5839883089065552,
"learning_rate": 3.533811950635099e-05,
"loss": 1.2523385047912599,
"step": 3185
},
{
"epoch": 0.6081403107425413,
"grad_norm": 0.5816488265991211,
"learning_rate": 3.519061261997497e-05,
"loss": 1.2530695915222168,
"step": 3190
},
{
"epoch": 0.6090935087217615,
"grad_norm": 0.6914941072463989,
"learning_rate": 3.504324688405878e-05,
"loss": 1.2479660034179687,
"step": 3195
},
{
"epoch": 0.6100467067009818,
"grad_norm": 0.6355175375938416,
"learning_rate": 3.4896023703167026e-05,
"loss": 1.221681785583496,
"step": 3200
},
{
"epoch": 0.610999904680202,
"grad_norm": 0.9013419151306152,
"learning_rate": 3.4748944480505564e-05,
"loss": 1.1884735107421875,
"step": 3205
},
{
"epoch": 0.6119531026594224,
"grad_norm": 0.5984360575675964,
"learning_rate": 3.460201061790825e-05,
"loss": 1.265510654449463,
"step": 3210
},
{
"epoch": 0.6129063006386426,
"grad_norm": 0.9662471413612366,
"learning_rate": 3.4455223515823446e-05,
"loss": 1.2142179489135743,
"step": 3215
},
{
"epoch": 0.6138594986178629,
"grad_norm": 0.5978898406028748,
"learning_rate": 3.430858457330066e-05,
"loss": 1.3043466567993165,
"step": 3220
},
{
"epoch": 0.6148126965970833,
"grad_norm": 0.8540678024291992,
"learning_rate": 3.4162095187977396e-05,
"loss": 1.2650224685668945,
"step": 3225
},
{
"epoch": 0.6157658945763035,
"grad_norm": 0.672986626625061,
"learning_rate": 3.40157567560656e-05,
"loss": 1.1735028266906737,
"step": 3230
},
{
"epoch": 0.6167190925555238,
"grad_norm": 0.7440739274024963,
"learning_rate": 3.386957067233848e-05,
"loss": 1.2253644943237305,
"step": 3235
},
{
"epoch": 0.617672290534744,
"grad_norm": 0.8787075877189636,
"learning_rate": 3.372353833011722e-05,
"loss": 1.119975471496582,
"step": 3240
},
{
"epoch": 0.6186254885139644,
"grad_norm": 0.7388913631439209,
"learning_rate": 3.357766112125767e-05,
"loss": 1.2302642822265626,
"step": 3245
},
{
"epoch": 0.6195786864931846,
"grad_norm": 0.5929700136184692,
"learning_rate": 3.343194043613705e-05,
"loss": 1.2942583084106445,
"step": 3250
},
{
"epoch": 0.6205318844724049,
"grad_norm": 0.5497426390647888,
"learning_rate": 3.328637766364075e-05,
"loss": 1.2651148796081544,
"step": 3255
},
{
"epoch": 0.6214850824516251,
"grad_norm": 0.7928446531295776,
"learning_rate": 3.3140974191149074e-05,
"loss": 1.231013298034668,
"step": 3260
},
{
"epoch": 0.6224382804308455,
"grad_norm": 0.6611085534095764,
"learning_rate": 3.299573140452399e-05,
"loss": 1.1703400611877441,
"step": 3265
},
{
"epoch": 0.6233914784100658,
"grad_norm": 0.5699756145477295,
"learning_rate": 3.285065068809597e-05,
"loss": 1.2961910247802735,
"step": 3270
},
{
"epoch": 0.624344676389286,
"grad_norm": 0.6091726422309875,
"learning_rate": 3.270573342465079e-05,
"loss": 1.3418394088745118,
"step": 3275
},
{
"epoch": 0.6252978743685064,
"grad_norm": 0.8048937916755676,
"learning_rate": 3.256098099541628e-05,
"loss": 1.2975360870361328,
"step": 3280
},
{
"epoch": 0.6262510723477266,
"grad_norm": 0.5461477637290955,
"learning_rate": 3.2416394780049216e-05,
"loss": 1.2592443466186523,
"step": 3285
},
{
"epoch": 0.6272042703269469,
"grad_norm": 0.6404699683189392,
"learning_rate": 3.227197615662223e-05,
"loss": 1.323004913330078,
"step": 3290
},
{
"epoch": 0.6281574683061671,
"grad_norm": 0.7028511762619019,
"learning_rate": 3.212772650161056e-05,
"loss": 1.221904182434082,
"step": 3295
},
{
"epoch": 0.6291106662853875,
"grad_norm": 0.7548397779464722,
"learning_rate": 3.198364718987893e-05,
"loss": 1.308700942993164,
"step": 3300
},
{
"epoch": 0.6300638642646078,
"grad_norm": 0.7729750275611877,
"learning_rate": 3.183973959466863e-05,
"loss": 1.2036958694458009,
"step": 3305
},
{
"epoch": 0.631017062243828,
"grad_norm": 0.6055728793144226,
"learning_rate": 3.169600508758418e-05,
"loss": 1.2661842346191405,
"step": 3310
},
{
"epoch": 0.6319702602230484,
"grad_norm": 0.5860031843185425,
"learning_rate": 3.1552445038580416e-05,
"loss": 1.2778533935546874,
"step": 3315
},
{
"epoch": 0.6329234582022686,
"grad_norm": 0.6527698636054993,
"learning_rate": 3.140906081594935e-05,
"loss": 1.2230241775512696,
"step": 3320
},
{
"epoch": 0.6338766561814889,
"grad_norm": 0.5826326608657837,
"learning_rate": 3.126585378630726e-05,
"loss": 1.2406131744384765,
"step": 3325
},
{
"epoch": 0.6348298541607091,
"grad_norm": 0.6854681372642517,
"learning_rate": 3.112282531458147e-05,
"loss": 1.2818424224853515,
"step": 3330
},
{
"epoch": 0.6357830521399295,
"grad_norm": 0.5114747285842896,
"learning_rate": 3.097997676399749e-05,
"loss": 1.1926551818847657,
"step": 3335
},
{
"epoch": 0.6367362501191497,
"grad_norm": 0.5886870622634888,
"learning_rate": 3.083730949606597e-05,
"loss": 1.3072022438049316,
"step": 3340
},
{
"epoch": 0.63768944809837,
"grad_norm": 0.9239357709884644,
"learning_rate": 3.069482487056974e-05,
"loss": 1.2589837074279786,
"step": 3345
},
{
"epoch": 0.6386426460775904,
"grad_norm": 0.5014736652374268,
"learning_rate": 3.0552524245550774e-05,
"loss": 1.250617504119873,
"step": 3350
},
{
"epoch": 0.6395958440568106,
"grad_norm": 0.616330623626709,
"learning_rate": 3.0410408977297417e-05,
"loss": 1.1623845100402832,
"step": 3355
},
{
"epoch": 0.6405490420360309,
"grad_norm": 0.5648524165153503,
"learning_rate": 3.026848042033127e-05,
"loss": 1.2155937194824218,
"step": 3360
},
{
"epoch": 0.6415022400152511,
"grad_norm": 0.6168292164802551,
"learning_rate": 3.0126739927394372e-05,
"loss": 1.2222782135009767,
"step": 3365
},
{
"epoch": 0.6424554379944715,
"grad_norm": 0.543231189250946,
"learning_rate": 2.998518884943634e-05,
"loss": 1.1905147552490234,
"step": 3370
},
{
"epoch": 0.6434086359736917,
"grad_norm": 0.6179722547531128,
"learning_rate": 2.98438285356014e-05,
"loss": 1.180801773071289,
"step": 3375
},
{
"epoch": 0.644361833952912,
"grad_norm": 0.6018236875534058,
"learning_rate": 2.9702660333215575e-05,
"loss": 1.2459759712219238,
"step": 3380
},
{
"epoch": 0.6453150319321324,
"grad_norm": 0.946191668510437,
"learning_rate": 2.9561685587773913e-05,
"loss": 1.2836873054504394,
"step": 3385
},
{
"epoch": 0.6462682299113526,
"grad_norm": 0.5618352293968201,
"learning_rate": 2.942090564292752e-05,
"loss": 1.244155502319336,
"step": 3390
},
{
"epoch": 0.6472214278905729,
"grad_norm": 0.6654608845710754,
"learning_rate": 2.9280321840470846e-05,
"loss": 1.2071935653686523,
"step": 3395
},
{
"epoch": 0.6481746258697931,
"grad_norm": 0.6032199859619141,
"learning_rate": 2.9139935520328898e-05,
"loss": 1.317270278930664,
"step": 3400
},
{
"epoch": 0.6491278238490135,
"grad_norm": 0.6821097731590271,
"learning_rate": 2.899974802054446e-05,
"loss": 1.2333221435546875,
"step": 3405
},
{
"epoch": 0.6500810218282337,
"grad_norm": 0.5638805627822876,
"learning_rate": 2.8859760677265296e-05,
"loss": 1.2344874382019042,
"step": 3410
},
{
"epoch": 0.651034219807454,
"grad_norm": 0.7227931618690491,
"learning_rate": 2.8719974824731442e-05,
"loss": 1.2080360412597657,
"step": 3415
},
{
"epoch": 0.6519874177866742,
"grad_norm": 0.6364921927452087,
"learning_rate": 2.858039179526256e-05,
"loss": 1.2339427947998047,
"step": 3420
},
{
"epoch": 0.6529406157658946,
"grad_norm": 0.5939121246337891,
"learning_rate": 2.844101291924511e-05,
"loss": 1.2578203201293945,
"step": 3425
},
{
"epoch": 0.6538938137451149,
"grad_norm": 0.7106569409370422,
"learning_rate": 2.830183952511974e-05,
"loss": 1.3066545486450196,
"step": 3430
},
{
"epoch": 0.6548470117243351,
"grad_norm": 0.8327311873435974,
"learning_rate": 2.8162872939368677e-05,
"loss": 1.305269145965576,
"step": 3435
},
{
"epoch": 0.6558002097035555,
"grad_norm": 0.8016763925552368,
"learning_rate": 2.8024114486502972e-05,
"loss": 1.1706339836120605,
"step": 3440
},
{
"epoch": 0.6567534076827757,
"grad_norm": 0.958949089050293,
"learning_rate": 2.7885565489049946e-05,
"loss": 1.219830322265625,
"step": 3445
},
{
"epoch": 0.657706605661996,
"grad_norm": 0.650306224822998,
"learning_rate": 2.7747227267540565e-05,
"loss": 1.190234088897705,
"step": 3450
},
{
"epoch": 0.6586598036412162,
"grad_norm": 0.8644406795501709,
"learning_rate": 2.7609101140496863e-05,
"loss": 1.2023894309997558,
"step": 3455
},
{
"epoch": 0.6596130016204366,
"grad_norm": 0.7740634083747864,
"learning_rate": 2.747118842441936e-05,
"loss": 1.1613030433654785,
"step": 3460
},
{
"epoch": 0.6605661995996569,
"grad_norm": 0.7851150035858154,
"learning_rate": 2.7333490433774566e-05,
"loss": 1.2760469436645507,
"step": 3465
},
{
"epoch": 0.6615193975788771,
"grad_norm": 0.7144232392311096,
"learning_rate": 2.719600848098236e-05,
"loss": 1.2046222686767578,
"step": 3470
},
{
"epoch": 0.6624725955580975,
"grad_norm": 0.7233515381813049,
"learning_rate": 2.705874387640354e-05,
"loss": 1.2585116386413575,
"step": 3475
},
{
"epoch": 0.6634257935373177,
"grad_norm": 0.8118272423744202,
"learning_rate": 2.6921697928327377e-05,
"loss": 1.2871943473815919,
"step": 3480
},
{
"epoch": 0.664378991516538,
"grad_norm": 0.7319923639297485,
"learning_rate": 2.6784871942959034e-05,
"loss": 1.1926034927368163,
"step": 3485
},
{
"epoch": 0.6653321894957582,
"grad_norm": 0.6281217336654663,
"learning_rate": 2.6648267224407197e-05,
"loss": 1.2875242233276367,
"step": 3490
},
{
"epoch": 0.6662853874749786,
"grad_norm": 0.580879271030426,
"learning_rate": 2.651188507467161e-05,
"loss": 1.2922750473022462,
"step": 3495
},
{
"epoch": 0.6672385854541988,
"grad_norm": 0.6035364866256714,
"learning_rate": 2.637572679363074e-05,
"loss": 1.2198362350463867,
"step": 3500
},
{
"epoch": 0.6681917834334191,
"grad_norm": 0.5943630933761597,
"learning_rate": 2.6239793679029268e-05,
"loss": 1.1869538307189942,
"step": 3505
},
{
"epoch": 0.6691449814126395,
"grad_norm": 0.9928993582725525,
"learning_rate": 2.6104087026465772e-05,
"loss": 1.3231023788452148,
"step": 3510
},
{
"epoch": 0.6700981793918597,
"grad_norm": 0.701328456401825,
"learning_rate": 2.5968608129380457e-05,
"loss": 1.2900423049926757,
"step": 3515
},
{
"epoch": 0.67105137737108,
"grad_norm": 0.9241816401481628,
"learning_rate": 2.5833358279042696e-05,
"loss": 1.1883378982543946,
"step": 3520
},
{
"epoch": 0.6720045753503002,
"grad_norm": 0.6395987272262573,
"learning_rate": 2.5698338764538787e-05,
"loss": 1.299351406097412,
"step": 3525
},
{
"epoch": 0.6729577733295206,
"grad_norm": 0.6330081820487976,
"learning_rate": 2.55635508727597e-05,
"loss": 1.2223155975341797,
"step": 3530
},
{
"epoch": 0.6739109713087408,
"grad_norm": 0.6576858162879944,
"learning_rate": 2.542899588838875e-05,
"loss": 1.2195650100708009,
"step": 3535
},
{
"epoch": 0.6748641692879611,
"grad_norm": 0.640916645526886,
"learning_rate": 2.529467509388935e-05,
"loss": 1.1820967674255372,
"step": 3540
},
{
"epoch": 0.6758173672671814,
"grad_norm": 0.7734649181365967,
"learning_rate": 2.5160589769492882e-05,
"loss": 1.2204526901245116,
"step": 3545
},
{
"epoch": 0.6767705652464017,
"grad_norm": 0.688849151134491,
"learning_rate": 2.5026741193186376e-05,
"loss": 1.3522113800048827,
"step": 3550
},
{
"epoch": 0.677723763225622,
"grad_norm": 0.6019042730331421,
"learning_rate": 2.489313064070037e-05,
"loss": 1.235138511657715,
"step": 3555
},
{
"epoch": 0.6786769612048422,
"grad_norm": 0.5748273134231567,
"learning_rate": 2.4759759385496828e-05,
"loss": 1.241206169128418,
"step": 3560
},
{
"epoch": 0.6796301591840626,
"grad_norm": 0.8615065217018127,
"learning_rate": 2.462662869875687e-05,
"loss": 1.2812816619873046,
"step": 3565
},
{
"epoch": 0.6805833571632828,
"grad_norm": 0.5118948221206665,
"learning_rate": 2.4493739849368734e-05,
"loss": 1.2726932525634767,
"step": 3570
},
{
"epoch": 0.6815365551425031,
"grad_norm": 0.5440770983695984,
"learning_rate": 2.4361094103915726e-05,
"loss": 1.2532058715820313,
"step": 3575
},
{
"epoch": 0.6824897531217233,
"grad_norm": 0.6594852209091187,
"learning_rate": 2.422869272666402e-05,
"loss": 1.2583551406860352,
"step": 3580
},
{
"epoch": 0.6834429511009437,
"grad_norm": 0.5646781325340271,
"learning_rate": 2.4096536979550717e-05,
"loss": 1.2261823654174804,
"step": 3585
},
{
"epoch": 0.684396149080164,
"grad_norm": 0.6829425692558289,
"learning_rate": 2.3964628122171767e-05,
"loss": 1.3109420776367187,
"step": 3590
},
{
"epoch": 0.6853493470593842,
"grad_norm": 0.5610309839248657,
"learning_rate": 2.383296741176997e-05,
"loss": 1.1917716026306153,
"step": 3595
},
{
"epoch": 0.6863025450386046,
"grad_norm": 0.5548216700553894,
"learning_rate": 2.3701556103223053e-05,
"loss": 1.259854793548584,
"step": 3600
},
{
"epoch": 0.6872557430178248,
"grad_norm": 0.5742631554603577,
"learning_rate": 2.3570395449031607e-05,
"loss": 1.2464586257934571,
"step": 3605
},
{
"epoch": 0.6882089409970451,
"grad_norm": 0.6340937614440918,
"learning_rate": 2.3439486699307206e-05,
"loss": 1.1482309341430663,
"step": 3610
},
{
"epoch": 0.6891621389762653,
"grad_norm": 0.6905898451805115,
"learning_rate": 2.3308831101760486e-05,
"loss": 1.2200018882751464,
"step": 3615
},
{
"epoch": 0.6901153369554857,
"grad_norm": 0.6480242013931274,
"learning_rate": 2.317842990168924e-05,
"loss": 1.2016407012939454,
"step": 3620
},
{
"epoch": 0.691068534934706,
"grad_norm": 0.9651988744735718,
"learning_rate": 2.3048284341966625e-05,
"loss": 1.2657808303833007,
"step": 3625
},
{
"epoch": 0.6920217329139262,
"grad_norm": 0.900599479675293,
"learning_rate": 2.2918395663029175e-05,
"loss": 1.1350242614746093,
"step": 3630
},
{
"epoch": 0.6929749308931465,
"grad_norm": 0.6493949890136719,
"learning_rate": 2.278876510286507e-05,
"loss": 1.218827247619629,
"step": 3635
},
{
"epoch": 0.6939281288723668,
"grad_norm": 0.6841947436332703,
"learning_rate": 2.2659393897002378e-05,
"loss": 1.2333038330078125,
"step": 3640
},
{
"epoch": 0.6948813268515871,
"grad_norm": 0.7985919117927551,
"learning_rate": 2.2530283278497162e-05,
"loss": 1.3237688064575195,
"step": 3645
},
{
"epoch": 0.6958345248308073,
"grad_norm": 0.7954857349395752,
"learning_rate": 2.2401434477921784e-05,
"loss": 1.2194201469421386,
"step": 3650
},
{
"epoch": 0.6967877228100277,
"grad_norm": 0.54095059633255,
"learning_rate": 2.227284872335325e-05,
"loss": 1.2793880462646485,
"step": 3655
},
{
"epoch": 0.6977409207892479,
"grad_norm": 0.8957579731941223,
"learning_rate": 2.214452724036137e-05,
"loss": 1.2481760025024413,
"step": 3660
},
{
"epoch": 0.6986941187684682,
"grad_norm": 0.5143315196037292,
"learning_rate": 2.201647125199714e-05,
"loss": 1.199508285522461,
"step": 3665
},
{
"epoch": 0.6996473167476885,
"grad_norm": 0.7957437634468079,
"learning_rate": 2.1888681978781157e-05,
"loss": 1.2408424377441407,
"step": 3670
},
{
"epoch": 0.7006005147269088,
"grad_norm": 0.5331310629844666,
"learning_rate": 2.176116063869184e-05,
"loss": 1.2055547714233399,
"step": 3675
},
{
"epoch": 0.7015537127061291,
"grad_norm": 0.9300455451011658,
"learning_rate": 2.163390844715394e-05,
"loss": 1.2586519241333007,
"step": 3680
},
{
"epoch": 0.7025069106853493,
"grad_norm": 0.7345501184463501,
"learning_rate": 2.1506926617026873e-05,
"loss": 1.260438823699951,
"step": 3685
},
{
"epoch": 0.7034601086645696,
"grad_norm": 0.7026432156562805,
"learning_rate": 2.1380216358593275e-05,
"loss": 1.204383087158203,
"step": 3690
},
{
"epoch": 0.7044133066437899,
"grad_norm": 0.6597341299057007,
"learning_rate": 2.125377887954732e-05,
"loss": 1.351447296142578,
"step": 3695
},
{
"epoch": 0.7053665046230102,
"grad_norm": 0.6570689678192139,
"learning_rate": 2.112761538498329e-05,
"loss": 1.2551791191101074,
"step": 3700
},
{
"epoch": 0.7063197026022305,
"grad_norm": 0.6351315379142761,
"learning_rate": 2.100172707738413e-05,
"loss": 1.2678667068481446,
"step": 3705
},
{
"epoch": 0.7072729005814508,
"grad_norm": 0.9237719178199768,
"learning_rate": 2.08761151566099e-05,
"loss": 1.1760089874267579,
"step": 3710
},
{
"epoch": 0.7082260985606711,
"grad_norm": 0.5916315317153931,
"learning_rate": 2.075078081988635e-05,
"loss": 1.2464874267578125,
"step": 3715
},
{
"epoch": 0.7091792965398913,
"grad_norm": 0.8199937343597412,
"learning_rate": 2.062572526179361e-05,
"loss": 1.2140378952026367,
"step": 3720
},
{
"epoch": 0.7101324945191116,
"grad_norm": 0.6741486191749573,
"learning_rate": 2.0500949674254665e-05,
"loss": 1.1921680450439454,
"step": 3725
},
{
"epoch": 0.7110856924983319,
"grad_norm": 0.8311759829521179,
"learning_rate": 2.037645524652406e-05,
"loss": 1.2360509872436523,
"step": 3730
},
{
"epoch": 0.7120388904775522,
"grad_norm": 0.8071461319923401,
"learning_rate": 2.0252243165176632e-05,
"loss": 1.205699634552002,
"step": 3735
},
{
"epoch": 0.7129920884567724,
"grad_norm": 0.7058308124542236,
"learning_rate": 2.0128314614096054e-05,
"loss": 1.2220178604125977,
"step": 3740
},
{
"epoch": 0.7139452864359928,
"grad_norm": 0.7297564148902893,
"learning_rate": 2.000467077446367e-05,
"loss": 1.1973310470581056,
"step": 3745
},
{
"epoch": 0.7148984844152131,
"grad_norm": 0.6562042832374573,
"learning_rate": 1.988131282474719e-05,
"loss": 1.2018284797668457,
"step": 3750
},
{
"epoch": 0.7158516823944333,
"grad_norm": 1.0852255821228027,
"learning_rate": 1.975824194068946e-05,
"loss": 1.3041958808898926,
"step": 3755
},
{
"epoch": 0.7168048803736536,
"grad_norm": 0.657896101474762,
"learning_rate": 1.9635459295297244e-05,
"loss": 1.1873765945434571,
"step": 3760
},
{
"epoch": 0.7177580783528739,
"grad_norm": 0.595609724521637,
"learning_rate": 1.9512966058830122e-05,
"loss": 1.1639683723449707,
"step": 3765
},
{
"epoch": 0.7187112763320942,
"grad_norm": 0.7022997140884399,
"learning_rate": 1.9390763398789223e-05,
"loss": 1.1171220779418944,
"step": 3770
},
{
"epoch": 0.7196644743113144,
"grad_norm": 0.7257121205329895,
"learning_rate": 1.9268852479906147e-05,
"loss": 1.1456874847412108,
"step": 3775
},
{
"epoch": 0.7206176722905347,
"grad_norm": 0.7017017602920532,
"learning_rate": 1.9147234464131868e-05,
"loss": 1.1874860763549804,
"step": 3780
},
{
"epoch": 0.7215708702697551,
"grad_norm": 0.6350660920143127,
"learning_rate": 1.9025910510625705e-05,
"loss": 1.2646387100219727,
"step": 3785
},
{
"epoch": 0.7225240682489753,
"grad_norm": 0.5961688756942749,
"learning_rate": 1.890488177574417e-05,
"loss": 1.1608770370483399,
"step": 3790
},
{
"epoch": 0.7234772662281956,
"grad_norm": 0.6065945625305176,
"learning_rate": 1.8784149413030005e-05,
"loss": 1.2965071678161622,
"step": 3795
},
{
"epoch": 0.7244304642074159,
"grad_norm": 0.7318583726882935,
"learning_rate": 1.8663714573201236e-05,
"loss": 1.172421646118164,
"step": 3800
},
{
"epoch": 0.7253836621866362,
"grad_norm": 0.6596642732620239,
"learning_rate": 1.8543578404140107e-05,
"loss": 1.227908992767334,
"step": 3805
},
{
"epoch": 0.7263368601658564,
"grad_norm": 0.6395574808120728,
"learning_rate": 1.8423742050882177e-05,
"loss": 1.2192790985107422,
"step": 3810
},
{
"epoch": 0.7272900581450767,
"grad_norm": 0.6519439816474915,
"learning_rate": 1.8304206655605476e-05,
"loss": 1.3065692901611328,
"step": 3815
},
{
"epoch": 0.728243256124297,
"grad_norm": 0.612903356552124,
"learning_rate": 1.818497335761951e-05,
"loss": 1.2053973197937011,
"step": 3820
},
{
"epoch": 0.7291964541035173,
"grad_norm": 0.6368285417556763,
"learning_rate": 1.8066043293354446e-05,
"loss": 1.2521958351135254,
"step": 3825
},
{
"epoch": 0.7301496520827376,
"grad_norm": 0.5930647850036621,
"learning_rate": 1.7947417596350335e-05,
"loss": 1.2521787643432618,
"step": 3830
},
{
"epoch": 0.7311028500619579,
"grad_norm": 1.255863070487976,
"learning_rate": 1.7829097397246208e-05,
"loss": 1.238549041748047,
"step": 3835
},
{
"epoch": 0.7320560480411782,
"grad_norm": 0.6461191773414612,
"learning_rate": 1.7711083823769354e-05,
"loss": 1.1778066635131836,
"step": 3840
},
{
"epoch": 0.7330092460203984,
"grad_norm": 0.6845282316207886,
"learning_rate": 1.759337800072462e-05,
"loss": 1.2988571166992187,
"step": 3845
},
{
"epoch": 0.7339624439996187,
"grad_norm": 0.6393148899078369,
"learning_rate": 1.747598104998358e-05,
"loss": 1.2095186233520507,
"step": 3850
},
{
"epoch": 0.734915641978839,
"grad_norm": 0.575240433216095,
"learning_rate": 1.7358894090473925e-05,
"loss": 1.1965845108032227,
"step": 3855
},
{
"epoch": 0.7358688399580593,
"grad_norm": 0.644023060798645,
"learning_rate": 1.7242118238168753e-05,
"loss": 1.2200517654418945,
"step": 3860
},
{
"epoch": 0.7368220379372796,
"grad_norm": 0.8921502828598022,
"learning_rate": 1.712565460607601e-05,
"loss": 1.2336673736572266,
"step": 3865
},
{
"epoch": 0.7377752359164998,
"grad_norm": 0.6045573949813843,
"learning_rate": 1.700950430422775e-05,
"loss": 1.1552182197570802,
"step": 3870
},
{
"epoch": 0.7387284338957202,
"grad_norm": 0.6928856372833252,
"learning_rate": 1.6893668439669673e-05,
"loss": 1.3072986602783203,
"step": 3875
},
{
"epoch": 0.7396816318749404,
"grad_norm": 1.4058945178985596,
"learning_rate": 1.6778148116450543e-05,
"loss": 1.13934326171875,
"step": 3880
},
{
"epoch": 0.7406348298541607,
"grad_norm": 0.6174265742301941,
"learning_rate": 1.6662944435611623e-05,
"loss": 1.2171598434448243,
"step": 3885
},
{
"epoch": 0.741588027833381,
"grad_norm": 0.554641842842102,
"learning_rate": 1.654805849517621e-05,
"loss": 1.2246965408325194,
"step": 3890
},
{
"epoch": 0.7425412258126013,
"grad_norm": 0.8113451600074768,
"learning_rate": 1.6433491390139177e-05,
"loss": 1.3295331954956056,
"step": 3895
},
{
"epoch": 0.7434944237918215,
"grad_norm": 0.8403462767601013,
"learning_rate": 1.631924421245657e-05,
"loss": 1.2600353240966797,
"step": 3900
},
{
"epoch": 0.7444476217710418,
"grad_norm": 0.7383481860160828,
"learning_rate": 1.6205318051035095e-05,
"loss": 1.2012214660644531,
"step": 3905
},
{
"epoch": 0.7454008197502622,
"grad_norm": 0.5221326351165771,
"learning_rate": 1.609171399172184e-05,
"loss": 1.2928386688232423,
"step": 3910
},
{
"epoch": 0.7463540177294824,
"grad_norm": 0.8107960820198059,
"learning_rate": 1.5978433117293883e-05,
"loss": 1.2678289413452148,
"step": 3915
},
{
"epoch": 0.7473072157087027,
"grad_norm": 0.5199189186096191,
"learning_rate": 1.586547650744797e-05,
"loss": 1.2676198959350586,
"step": 3920
},
{
"epoch": 0.748260413687923,
"grad_norm": 0.7182909250259399,
"learning_rate": 1.575284523879027e-05,
"loss": 1.2382710456848145,
"step": 3925
},
{
"epoch": 0.7492136116671433,
"grad_norm": 0.693976104259491,
"learning_rate": 1.564054038482603e-05,
"loss": 1.2486501693725587,
"step": 3930
},
{
"epoch": 0.7501668096463635,
"grad_norm": 0.6236864328384399,
"learning_rate": 1.552856301594942e-05,
"loss": 1.1453128814697267,
"step": 3935
},
{
"epoch": 0.7511200076255838,
"grad_norm": 0.8047870993614197,
"learning_rate": 1.541691419943327e-05,
"loss": 1.235875701904297,
"step": 3940
},
{
"epoch": 0.7520732056048041,
"grad_norm": 0.7706174850463867,
"learning_rate": 1.5305594999418987e-05,
"loss": 1.1984367370605469,
"step": 3945
},
{
"epoch": 0.7530264035840244,
"grad_norm": 0.524729311466217,
"learning_rate": 1.5194606476906299e-05,
"loss": 1.2197100639343261,
"step": 3950
},
{
"epoch": 0.7539796015632447,
"grad_norm": 0.6807271242141724,
"learning_rate": 1.5083949689743203e-05,
"loss": 1.1406570434570313,
"step": 3955
},
{
"epoch": 0.754932799542465,
"grad_norm": 0.7323083281517029,
"learning_rate": 1.4973625692615922e-05,
"loss": 1.1842288017272948,
"step": 3960
},
{
"epoch": 0.7558859975216853,
"grad_norm": 0.7220993638038635,
"learning_rate": 1.4863635537038767e-05,
"loss": 1.2957634925842285,
"step": 3965
},
{
"epoch": 0.7568391955009055,
"grad_norm": 0.6333289742469788,
"learning_rate": 1.4753980271344142e-05,
"loss": 1.1837800979614257,
"step": 3970
},
{
"epoch": 0.7577923934801258,
"grad_norm": 0.6561208963394165,
"learning_rate": 1.4644660940672627e-05,
"loss": 1.2387582778930664,
"step": 3975
},
{
"epoch": 0.7587455914593461,
"grad_norm": 0.7606480717658997,
"learning_rate": 1.4535678586962902e-05,
"loss": 1.17692289352417,
"step": 3980
},
{
"epoch": 0.7596987894385664,
"grad_norm": 0.5796291828155518,
"learning_rate": 1.4427034248941873e-05,
"loss": 1.2010177612304687,
"step": 3985
},
{
"epoch": 0.7606519874177867,
"grad_norm": 0.6456603407859802,
"learning_rate": 1.4318728962114813e-05,
"loss": 1.1805098533630372,
"step": 3990
},
{
"epoch": 0.761605185397007,
"grad_norm": 0.5619194507598877,
"learning_rate": 1.4210763758755402e-05,
"loss": 1.1843601226806642,
"step": 3995
},
{
"epoch": 0.7625583833762273,
"grad_norm": 0.6663962006568909,
"learning_rate": 1.4103139667895932e-05,
"loss": 1.302177906036377,
"step": 4000
},
{
"epoch": 0.7635115813554475,
"grad_norm": 0.7648465037345886,
"learning_rate": 1.399585771531755e-05,
"loss": 1.3074260711669923,
"step": 4005
},
{
"epoch": 0.7644647793346678,
"grad_norm": 0.6551898121833801,
"learning_rate": 1.388891892354039e-05,
"loss": 1.301952075958252,
"step": 4010
},
{
"epoch": 0.765417977313888,
"grad_norm": 0.830543041229248,
"learning_rate": 1.378232431181386e-05,
"loss": 1.2780150413513183,
"step": 4015
},
{
"epoch": 0.7663711752931084,
"grad_norm": 0.6601079106330872,
"learning_rate": 1.3676074896106994e-05,
"loss": 1.2808897018432617,
"step": 4020
},
{
"epoch": 0.7673243732723286,
"grad_norm": 0.5462204813957214,
"learning_rate": 1.357017168909866e-05,
"loss": 1.2064345359802247,
"step": 4025
},
{
"epoch": 0.7682775712515489,
"grad_norm": 0.5006428360939026,
"learning_rate": 1.3464615700167965e-05,
"loss": 1.2465780258178711,
"step": 4030
},
{
"epoch": 0.7692307692307693,
"grad_norm": 0.5989637970924377,
"learning_rate": 1.3359407935384644e-05,
"loss": 1.2092927932739257,
"step": 4035
},
{
"epoch": 0.7701839672099895,
"grad_norm": 0.907219409942627,
"learning_rate": 1.3254549397499466e-05,
"loss": 1.2018006324768067,
"step": 4040
},
{
"epoch": 0.7711371651892098,
"grad_norm": 0.568661093711853,
"learning_rate": 1.3150041085934638e-05,
"loss": 1.240280532836914,
"step": 4045
},
{
"epoch": 0.77209036316843,
"grad_norm": 0.6743282675743103,
"learning_rate": 1.3045883996774339e-05,
"loss": 1.2279913902282715,
"step": 4050
},
{
"epoch": 0.7730435611476504,
"grad_norm": 0.5384716391563416,
"learning_rate": 1.2942079122755162e-05,
"loss": 1.2751991271972656,
"step": 4055
},
{
"epoch": 0.7739967591268706,
"grad_norm": 0.5935498476028442,
"learning_rate": 1.2838627453256696e-05,
"loss": 1.1956563949584962,
"step": 4060
},
{
"epoch": 0.7749499571060909,
"grad_norm": 0.7832478880882263,
"learning_rate": 1.273552997429212e-05,
"loss": 1.2472781181335448,
"step": 4065
},
{
"epoch": 0.7759031550853113,
"grad_norm": 0.5618842840194702,
"learning_rate": 1.263278766849873e-05,
"loss": 1.2372567176818847,
"step": 4070
},
{
"epoch": 0.7768563530645315,
"grad_norm": 0.6840872764587402,
"learning_rate": 1.253040151512862e-05,
"loss": 1.2406797409057617,
"step": 4075
},
{
"epoch": 0.7778095510437518,
"grad_norm": 0.6183333992958069,
"learning_rate": 1.2428372490039347e-05,
"loss": 1.1819196701049806,
"step": 4080
},
{
"epoch": 0.778762749022972,
"grad_norm": 0.744036853313446,
"learning_rate": 1.2326701565684645e-05,
"loss": 1.2907965660095215,
"step": 4085
},
{
"epoch": 0.7797159470021924,
"grad_norm": 0.5243722796440125,
"learning_rate": 1.2225389711105111e-05,
"loss": 1.3109117507934571,
"step": 4090
},
{
"epoch": 0.7806691449814126,
"grad_norm": 0.6731049418449402,
"learning_rate": 1.2124437891918993e-05,
"loss": 1.2946283340454101,
"step": 4095
},
{
"epoch": 0.7816223429606329,
"grad_norm": 0.7544031143188477,
"learning_rate": 1.2023847070313027e-05,
"loss": 1.1947206497192382,
"step": 4100
},
{
"epoch": 0.7825755409398532,
"grad_norm": 0.847855269908905,
"learning_rate": 1.1923618205033183e-05,
"loss": 1.1523775100708007,
"step": 4105
},
{
"epoch": 0.7835287389190735,
"grad_norm": 0.6558092832565308,
"learning_rate": 1.1823752251375586e-05,
"loss": 1.2275083541870118,
"step": 4110
},
{
"epoch": 0.7844819368982938,
"grad_norm": 0.773508608341217,
"learning_rate": 1.1724250161177392e-05,
"loss": 1.324901008605957,
"step": 4115
},
{
"epoch": 0.785435134877514,
"grad_norm": 0.555661678314209,
"learning_rate": 1.1625112882807754e-05,
"loss": 1.22003231048584,
"step": 4120
},
{
"epoch": 0.7863883328567344,
"grad_norm": 0.6107597351074219,
"learning_rate": 1.152634136115871e-05,
"loss": 1.2048500061035157,
"step": 4125
},
{
"epoch": 0.7873415308359546,
"grad_norm": 0.8006135821342468,
"learning_rate": 1.1427936537636214e-05,
"loss": 1.265638828277588,
"step": 4130
},
{
"epoch": 0.7882947288151749,
"grad_norm": 0.6387047171592712,
"learning_rate": 1.1329899350151213e-05,
"loss": 1.1716151237487793,
"step": 4135
},
{
"epoch": 0.7892479267943951,
"grad_norm": 0.6204472184181213,
"learning_rate": 1.1232230733110615e-05,
"loss": 1.0868881225585938,
"step": 4140
},
{
"epoch": 0.7902011247736155,
"grad_norm": 0.7070595622062683,
"learning_rate": 1.1134931617408435e-05,
"loss": 1.2030377388000488,
"step": 4145
},
{
"epoch": 0.7911543227528358,
"grad_norm": 0.5742433667182922,
"learning_rate": 1.1038002930416941e-05,
"loss": 1.1542322158813476,
"step": 4150
},
{
"epoch": 0.792107520732056,
"grad_norm": 0.6153649687767029,
"learning_rate": 1.0941445595977767e-05,
"loss": 1.2294605255126954,
"step": 4155
},
{
"epoch": 0.7930607187112764,
"grad_norm": 0.6542089581489563,
"learning_rate": 1.0845260534393125e-05,
"loss": 1.2671607971191405,
"step": 4160
},
{
"epoch": 0.7940139166904966,
"grad_norm": 0.5247233510017395,
"learning_rate": 1.0749448662417084e-05,
"loss": 1.1785138130187989,
"step": 4165
},
{
"epoch": 0.7949671146697169,
"grad_norm": 0.6570800542831421,
"learning_rate": 1.065401089324673e-05,
"loss": 1.2902163505554198,
"step": 4170
},
{
"epoch": 0.7959203126489371,
"grad_norm": 0.8899843096733093,
"learning_rate": 1.0558948136513535e-05,
"loss": 1.238455581665039,
"step": 4175
},
{
"epoch": 0.7968735106281575,
"grad_norm": 0.8128093481063843,
"learning_rate": 1.0464261298274725e-05,
"loss": 1.1051462173461915,
"step": 4180
},
{
"epoch": 0.7978267086073777,
"grad_norm": 0.6244000196456909,
"learning_rate": 1.0369951281004513e-05,
"loss": 1.2674825668334961,
"step": 4185
},
{
"epoch": 0.798779906586598,
"grad_norm": 0.820381760597229,
"learning_rate": 1.0276018983585633e-05,
"loss": 1.2259267807006835,
"step": 4190
},
{
"epoch": 0.7997331045658184,
"grad_norm": 0.6989737749099731,
"learning_rate": 1.018246530130067e-05,
"loss": 1.180887794494629,
"step": 4195
},
{
"epoch": 0.8006863025450386,
"grad_norm": 0.6637066006660461,
"learning_rate": 1.008929112582362e-05,
"loss": 1.225269889831543,
"step": 4200
},
{
"epoch": 0.8016395005242589,
"grad_norm": 0.6584336161613464,
"learning_rate": 9.996497345211313e-06,
"loss": 1.207047462463379,
"step": 4205
},
{
"epoch": 0.8025926985034791,
"grad_norm": 0.8894166350364685,
"learning_rate": 9.904084843894983e-06,
"loss": 1.30599946975708,
"step": 4210
},
{
"epoch": 0.8035458964826995,
"grad_norm": 0.8106604218482971,
"learning_rate": 9.812054502671836e-06,
"loss": 1.1555444717407226,
"step": 4215
},
{
"epoch": 0.8044990944619197,
"grad_norm": 0.7093089818954468,
"learning_rate": 9.720407198696662e-06,
"loss": 1.2678872108459474,
"step": 4220
},
{
"epoch": 0.80545229244114,
"grad_norm": 0.7287440896034241,
"learning_rate": 9.62914380547344e-06,
"loss": 1.2558979034423827,
"step": 4225
},
{
"epoch": 0.8064054904203604,
"grad_norm": 0.5888250470161438,
"learning_rate": 9.53826519284709e-06,
"loss": 1.266513729095459,
"step": 4230
},
{
"epoch": 0.8073586883995806,
"grad_norm": 0.6939835548400879,
"learning_rate": 9.447772226995084e-06,
"loss": 1.2332074165344238,
"step": 4235
},
{
"epoch": 0.8083118863788009,
"grad_norm": 0.8536332845687866,
"learning_rate": 9.357665770419244e-06,
"loss": 1.2781600952148438,
"step": 4240
},
{
"epoch": 0.8092650843580211,
"grad_norm": 0.6815188527107239,
"learning_rate": 9.267946681937544e-06,
"loss": 1.213609218597412,
"step": 4245
},
{
"epoch": 0.8102182823372415,
"grad_norm": 0.8423677682876587,
"learning_rate": 9.178615816675867e-06,
"loss": 1.2570457458496094,
"step": 4250
},
{
"epoch": 0.8111714803164617,
"grad_norm": 0.7196339964866638,
"learning_rate": 9.08967402605988e-06,
"loss": 1.269762134552002,
"step": 4255
},
{
"epoch": 0.812124678295682,
"grad_norm": 0.725196361541748,
"learning_rate": 9.001122157806957e-06,
"loss": 1.257029914855957,
"step": 4260
},
{
"epoch": 0.8130778762749022,
"grad_norm": 0.6335158944129944,
"learning_rate": 8.912961055918035e-06,
"loss": 1.247712993621826,
"step": 4265
},
{
"epoch": 0.8140310742541226,
"grad_norm": 0.6774193644523621,
"learning_rate": 8.825191560669589e-06,
"loss": 1.26491117477417,
"step": 4270
},
{
"epoch": 0.8149842722333429,
"grad_norm": 0.805121898651123,
"learning_rate": 8.737814508605674e-06,
"loss": 1.2180566787719727,
"step": 4275
},
{
"epoch": 0.8159374702125631,
"grad_norm": 0.663702130317688,
"learning_rate": 8.650830732529869e-06,
"loss": 1.220667266845703,
"step": 4280
},
{
"epoch": 0.8168906681917835,
"grad_norm": 0.6967896819114685,
"learning_rate": 8.564241061497413e-06,
"loss": 1.1456944465637207,
"step": 4285
},
{
"epoch": 0.8178438661710037,
"grad_norm": 0.7834393382072449,
"learning_rate": 8.478046320807232e-06,
"loss": 1.2307169914245606,
"step": 4290
},
{
"epoch": 0.818797064150224,
"grad_norm": 0.836487889289856,
"learning_rate": 8.392247331994173e-06,
"loss": 1.163895606994629,
"step": 4295
},
{
"epoch": 0.8197502621294442,
"grad_norm": 0.6468165516853333,
"learning_rate": 8.30684491282106e-06,
"loss": 1.1135669708251954,
"step": 4300
},
{
"epoch": 0.8207034601086646,
"grad_norm": 0.7467713952064514,
"learning_rate": 8.221839877270964e-06,
"loss": 1.2942085266113281,
"step": 4305
},
{
"epoch": 0.8216566580878849,
"grad_norm": 0.9583889842033386,
"learning_rate": 8.13723303553946e-06,
"loss": 1.2471393585205077,
"step": 4310
},
{
"epoch": 0.8226098560671051,
"grad_norm": 0.8676281571388245,
"learning_rate": 8.053025194026858e-06,
"loss": 1.2271574020385743,
"step": 4315
},
{
"epoch": 0.8235630540463255,
"grad_norm": 0.8718356490135193,
"learning_rate": 7.969217155330516e-06,
"loss": 1.214697265625,
"step": 4320
},
{
"epoch": 0.8245162520255457,
"grad_norm": 0.9373974800109863,
"learning_rate": 7.885809718237274e-06,
"loss": 1.173631477355957,
"step": 4325
},
{
"epoch": 0.825469450004766,
"grad_norm": 0.5150532126426697,
"learning_rate": 7.802803677715726e-06,
"loss": 1.23596134185791,
"step": 4330
},
{
"epoch": 0.8264226479839862,
"grad_norm": 0.6421557068824768,
"learning_rate": 7.720199824908692e-06,
"loss": 1.2475356101989745,
"step": 4335
},
{
"epoch": 0.8273758459632066,
"grad_norm": 0.5453834533691406,
"learning_rate": 7.63799894712573e-06,
"loss": 1.2999723434448243,
"step": 4340
},
{
"epoch": 0.8283290439424268,
"grad_norm": 0.69022136926651,
"learning_rate": 7.556201827835552e-06,
"loss": 1.2229093551635741,
"step": 4345
},
{
"epoch": 0.8292822419216471,
"grad_norm": 0.8374666571617126,
"learning_rate": 7.4748092466585764e-06,
"loss": 1.237013816833496,
"step": 4350
},
{
"epoch": 0.8302354399008675,
"grad_norm": 0.8628894686698914,
"learning_rate": 7.393821979359528e-06,
"loss": 1.2680168151855469,
"step": 4355
},
{
"epoch": 0.8311886378800877,
"grad_norm": 0.6808069348335266,
"learning_rate": 7.3132407978400155e-06,
"loss": 1.2018478393554688,
"step": 4360
},
{
"epoch": 0.832141835859308,
"grad_norm": 0.8288878202438354,
"learning_rate": 7.233066470131205e-06,
"loss": 1.1832545280456543,
"step": 4365
},
{
"epoch": 0.8330950338385282,
"grad_norm": 0.6376352310180664,
"learning_rate": 7.153299760386456e-06,
"loss": 1.2254414558410645,
"step": 4370
},
{
"epoch": 0.8340482318177486,
"grad_norm": 0.6286559700965881,
"learning_rate": 7.073941428874064e-06,
"loss": 1.1868556022644043,
"step": 4375
},
{
"epoch": 0.8350014297969688,
"grad_norm": 0.5295135974884033,
"learning_rate": 6.994992231970016e-06,
"loss": 1.1714859008789062,
"step": 4380
},
{
"epoch": 0.8359546277761891,
"grad_norm": 0.7237422466278076,
"learning_rate": 6.9164529221507675e-06,
"loss": 1.213966178894043,
"step": 4385
},
{
"epoch": 0.8369078257554095,
"grad_norm": 0.8221988677978516,
"learning_rate": 6.838324247986111e-06,
"loss": 1.2056529998779297,
"step": 4390
},
{
"epoch": 0.8378610237346297,
"grad_norm": 0.7373778820037842,
"learning_rate": 6.760606954131965e-06,
"loss": 1.2847192764282227,
"step": 4395
},
{
"epoch": 0.83881422171385,
"grad_norm": 0.8068478107452393,
"learning_rate": 6.683301781323348e-06,
"loss": 1.224747657775879,
"step": 4400
},
{
"epoch": 0.8397674196930702,
"grad_norm": 0.6083196997642517,
"learning_rate": 6.606409466367297e-06,
"loss": 1.2385419845581054,
"step": 4405
},
{
"epoch": 0.8407206176722906,
"grad_norm": 0.682094156742096,
"learning_rate": 6.529930742135826e-06,
"loss": 1.213953971862793,
"step": 4410
},
{
"epoch": 0.8416738156515108,
"grad_norm": 0.5927014946937561,
"learning_rate": 6.45386633755894e-06,
"loss": 1.2639514923095703,
"step": 4415
},
{
"epoch": 0.8426270136307311,
"grad_norm": 0.8485562801361084,
"learning_rate": 6.3782169776177516e-06,
"loss": 1.2114716529846192,
"step": 4420
},
{
"epoch": 0.8435802116099513,
"grad_norm": 0.6819233298301697,
"learning_rate": 6.302983383337474e-06,
"loss": 1.2224900245666503,
"step": 4425
},
{
"epoch": 0.8445334095891717,
"grad_norm": 0.6738782525062561,
"learning_rate": 6.2281662717806e-06,
"loss": 1.2218244552612305,
"step": 4430
},
{
"epoch": 0.845486607568392,
"grad_norm": 0.5365089774131775,
"learning_rate": 6.1537663560401085e-06,
"loss": 1.2705909729003906,
"step": 4435
},
{
"epoch": 0.8464398055476122,
"grad_norm": 0.8223872780799866,
"learning_rate": 6.07978434523257e-06,
"loss": 1.2167354583740235,
"step": 4440
},
{
"epoch": 0.8473930035268326,
"grad_norm": 0.7721960544586182,
"learning_rate": 6.006220944491459e-06,
"loss": 1.166141891479492,
"step": 4445
},
{
"epoch": 0.8483462015060528,
"grad_norm": 0.5896003842353821,
"learning_rate": 5.93307685496044e-06,
"loss": 1.2694342613220215,
"step": 4450
},
{
"epoch": 0.8492993994852731,
"grad_norm": 0.75387042760849,
"learning_rate": 5.860352773786631e-06,
"loss": 1.2268791198730469,
"step": 4455
},
{
"epoch": 0.8502525974644933,
"grad_norm": 0.7291093468666077,
"learning_rate": 5.7880493941139926e-06,
"loss": 1.2128201484680177,
"step": 4460
},
{
"epoch": 0.8512057954437137,
"grad_norm": 0.7240249514579773,
"learning_rate": 5.7161674050767275e-06,
"loss": 1.3292516708374023,
"step": 4465
},
{
"epoch": 0.852158993422934,
"grad_norm": 0.6796298623085022,
"learning_rate": 5.6447074917927e-06,
"loss": 1.2751702308654784,
"step": 4470
},
{
"epoch": 0.8531121914021542,
"grad_norm": 0.6727436780929565,
"learning_rate": 5.573670335356901e-06,
"loss": 1.2269893646240235,
"step": 4475
},
{
"epoch": 0.8540653893813746,
"grad_norm": 0.6166629791259766,
"learning_rate": 5.5030566128349556e-06,
"loss": 1.3168445587158204,
"step": 4480
},
{
"epoch": 0.8550185873605948,
"grad_norm": 0.7454910278320312,
"learning_rate": 5.432866997256708e-06,
"loss": 1.1884767532348632,
"step": 4485
},
{
"epoch": 0.8559717853398151,
"grad_norm": 0.7811619639396667,
"learning_rate": 5.363102157609751e-06,
"loss": 1.1711655616760255,
"step": 4490
},
{
"epoch": 0.8569249833190353,
"grad_norm": 0.6651840806007385,
"learning_rate": 5.293762758833071e-06,
"loss": 1.1538570404052735,
"step": 4495
},
{
"epoch": 0.8578781812982557,
"grad_norm": 0.8838904500007629,
"learning_rate": 5.224849461810749e-06,
"loss": 1.2227632522583007,
"step": 4500
},
{
"epoch": 0.8588313792774759,
"grad_norm": 0.5597884654998779,
"learning_rate": 5.156362923365588e-06,
"loss": 1.2299514770507813,
"step": 4505
},
{
"epoch": 0.8597845772566962,
"grad_norm": 0.7257062196731567,
"learning_rate": 5.088303796252908e-06,
"loss": 1.2300275802612304,
"step": 4510
},
{
"epoch": 0.8607377752359165,
"grad_norm": 0.8183436989784241,
"learning_rate": 5.020672729154308e-06,
"loss": 1.1926952362060548,
"step": 4515
},
{
"epoch": 0.8616909732151368,
"grad_norm": 0.7929868102073669,
"learning_rate": 4.9534703666714925e-06,
"loss": 1.2333523750305175,
"step": 4520
},
{
"epoch": 0.8626441711943571,
"grad_norm": 0.7421267032623291,
"learning_rate": 4.886697349320085e-06,
"loss": 1.2376693725585937,
"step": 4525
},
{
"epoch": 0.8635973691735773,
"grad_norm": 0.6524899005889893,
"learning_rate": 4.820354313523617e-06,
"loss": 1.1557720184326172,
"step": 4530
},
{
"epoch": 0.8645505671527977,
"grad_norm": 0.6860274076461792,
"learning_rate": 4.754441891607347e-06,
"loss": 1.2184629440307617,
"step": 4535
},
{
"epoch": 0.8655037651320179,
"grad_norm": 0.6499245762825012,
"learning_rate": 4.688960711792306e-06,
"loss": 1.1541982650756837,
"step": 4540
},
{
"epoch": 0.8664569631112382,
"grad_norm": 0.6724962592124939,
"learning_rate": 4.623911398189318e-06,
"loss": 1.234123420715332,
"step": 4545
},
{
"epoch": 0.8674101610904585,
"grad_norm": 0.6820011138916016,
"learning_rate": 4.559294570792988e-06,
"loss": 1.2424484252929688,
"step": 4550
},
{
"epoch": 0.8683633590696788,
"grad_norm": 0.9211460947990417,
"learning_rate": 4.495110845475848e-06,
"loss": 1.199906826019287,
"step": 4555
},
{
"epoch": 0.8693165570488991,
"grad_norm": 0.7585262060165405,
"learning_rate": 4.431360833982468e-06,
"loss": 1.1647123336791991,
"step": 4560
},
{
"epoch": 0.8702697550281193,
"grad_norm": 0.6126020550727844,
"learning_rate": 4.368045143923638e-06,
"loss": 1.241098403930664,
"step": 4565
},
{
"epoch": 0.8712229530073397,
"grad_norm": 0.5670948028564453,
"learning_rate": 4.305164378770554e-06,
"loss": 1.1584237098693848,
"step": 4570
},
{
"epoch": 0.8721761509865599,
"grad_norm": 0.5973881483078003,
"learning_rate": 4.242719137849077e-06,
"loss": 1.1756168365478517,
"step": 4575
},
{
"epoch": 0.8731293489657802,
"grad_norm": 0.6933620572090149,
"learning_rate": 4.180710016334039e-06,
"loss": 1.1758663177490234,
"step": 4580
},
{
"epoch": 0.8740825469450004,
"grad_norm": 0.7207223176956177,
"learning_rate": 4.119137605243528e-06,
"loss": 1.1413729667663575,
"step": 4585
},
{
"epoch": 0.8750357449242208,
"grad_norm": 0.9575616121292114,
"learning_rate": 4.0580024914332995e-06,
"loss": 1.199721336364746,
"step": 4590
},
{
"epoch": 0.8759889429034411,
"grad_norm": 0.6890310049057007,
"learning_rate": 3.997305257591161e-06,
"loss": 1.2197802543640137,
"step": 4595
},
{
"epoch": 0.8769421408826613,
"grad_norm": 0.8953237533569336,
"learning_rate": 3.937046482231416e-06,
"loss": 1.1360723495483398,
"step": 4600
},
{
"epoch": 0.8778953388618816,
"grad_norm": 0.6974352598190308,
"learning_rate": 3.877226739689355e-06,
"loss": 1.258785343170166,
"step": 4605
},
{
"epoch": 0.8788485368411019,
"grad_norm": 0.5826160907745361,
"learning_rate": 3.817846600115793e-06,
"loss": 1.2580406188964843,
"step": 4610
},
{
"epoch": 0.8798017348203222,
"grad_norm": 1.011110544204712,
"learning_rate": 3.7589066294716147e-06,
"loss": 1.248322105407715,
"step": 4615
},
{
"epoch": 0.8807549327995424,
"grad_norm": 0.6812794804573059,
"learning_rate": 3.700407389522381e-06,
"loss": 1.091628074645996,
"step": 4620
},
{
"epoch": 0.8817081307787628,
"grad_norm": 0.555317223072052,
"learning_rate": 3.6423494378330035e-06,
"loss": 1.1832502365112305,
"step": 4625
},
{
"epoch": 0.8826613287579831,
"grad_norm": 0.7968437075614929,
"learning_rate": 3.5847333277624017e-06,
"loss": 1.1362716674804687,
"step": 4630
},
{
"epoch": 0.8836145267372033,
"grad_norm": 0.7184310555458069,
"learning_rate": 3.5275596084582253e-06,
"loss": 1.2132604598999024,
"step": 4635
},
{
"epoch": 0.8845677247164236,
"grad_norm": 0.7875999212265015,
"learning_rate": 3.4708288248516516e-06,
"loss": 1.1182334899902344,
"step": 4640
},
{
"epoch": 0.8855209226956439,
"grad_norm": 0.9749584197998047,
"learning_rate": 3.41454151765217e-06,
"loss": 1.1107658386230468,
"step": 4645
},
{
"epoch": 0.8864741206748642,
"grad_norm": 0.721259593963623,
"learning_rate": 3.3586982233424102e-06,
"loss": 1.1024761199951172,
"step": 4650
},
{
"epoch": 0.8874273186540844,
"grad_norm": 0.6477571725845337,
"learning_rate": 3.303299474173066e-06,
"loss": 1.1984753608703613,
"step": 4655
},
{
"epoch": 0.8883805166333048,
"grad_norm": 0.5940926671028137,
"learning_rate": 3.2483457981578092e-06,
"loss": 1.1844069480895996,
"step": 4660
},
{
"epoch": 0.889333714612525,
"grad_norm": 0.7926877737045288,
"learning_rate": 3.1938377190682323e-06,
"loss": 1.282862091064453,
"step": 4665
},
{
"epoch": 0.8902869125917453,
"grad_norm": 0.5801262259483337,
"learning_rate": 3.1397757564289143e-06,
"loss": 1.2212105751037599,
"step": 4670
},
{
"epoch": 0.8912401105709656,
"grad_norm": 0.6794660091400146,
"learning_rate": 3.0861604255124076e-06,
"loss": 1.1792485237121582,
"step": 4675
},
{
"epoch": 0.8921933085501859,
"grad_norm": 0.8272831439971924,
"learning_rate": 3.0329922373343422e-06,
"loss": 1.2231868743896483,
"step": 4680
},
{
"epoch": 0.8931465065294062,
"grad_norm": 0.8906814455986023,
"learning_rate": 2.980271698648579e-06,
"loss": 1.1177072525024414,
"step": 4685
},
{
"epoch": 0.8940997045086264,
"grad_norm": 0.881855309009552,
"learning_rate": 2.9279993119423753e-06,
"loss": 1.3396747589111329,
"step": 4690
},
{
"epoch": 0.8950529024878467,
"grad_norm": 0.6932026147842407,
"learning_rate": 2.8761755754315668e-06,
"loss": 1.1601354598999023,
"step": 4695
},
{
"epoch": 0.896006100467067,
"grad_norm": 0.583479106426239,
"learning_rate": 2.824800983055842e-06,
"loss": 1.1855619430541993,
"step": 4700
},
{
"epoch": 0.8969592984462873,
"grad_norm": 0.7233911156654358,
"learning_rate": 2.7738760244740425e-06,
"loss": 1.2095130920410155,
"step": 4705
},
{
"epoch": 0.8979124964255075,
"grad_norm": 0.8217453360557556,
"learning_rate": 2.723401185059471e-06,
"loss": 1.2268701553344727,
"step": 4710
},
{
"epoch": 0.8988656944047279,
"grad_norm": 0.700840950012207,
"learning_rate": 2.673376945895273e-06,
"loss": 1.3245098114013671,
"step": 4715
},
{
"epoch": 0.8998188923839482,
"grad_norm": 0.6691460013389587,
"learning_rate": 2.6238037837698793e-06,
"loss": 1.2844621658325195,
"step": 4720
},
{
"epoch": 0.9007720903631684,
"grad_norm": 0.9386864900588989,
"learning_rate": 2.574682171172421e-06,
"loss": 1.281606674194336,
"step": 4725
},
{
"epoch": 0.9017252883423887,
"grad_norm": 0.7843084335327148,
"learning_rate": 2.526012576288245e-06,
"loss": 1.126380157470703,
"step": 4730
},
{
"epoch": 0.902678486321609,
"grad_norm": 0.7220968008041382,
"learning_rate": 2.4777954629944477e-06,
"loss": 1.107554054260254,
"step": 4735
},
{
"epoch": 0.9036316843008293,
"grad_norm": 0.7478969693183899,
"learning_rate": 2.430031290855472e-06,
"loss": 1.1156665802001953,
"step": 4740
},
{
"epoch": 0.9045848822800495,
"grad_norm": 0.6611006855964661,
"learning_rate": 2.382720515118697e-06,
"loss": 1.1280939102172851,
"step": 4745
},
{
"epoch": 0.9055380802592699,
"grad_norm": 0.7239766120910645,
"learning_rate": 2.3358635867101132e-06,
"loss": 1.1275172233581543,
"step": 4750
},
{
"epoch": 0.9064912782384902,
"grad_norm": 0.9137570261955261,
"learning_rate": 2.289460952230038e-06,
"loss": 1.2814844131469727,
"step": 4755
},
{
"epoch": 0.9074444762177104,
"grad_norm": 0.8071094751358032,
"learning_rate": 2.243513053948837e-06,
"loss": 1.1880189895629882,
"step": 4760
},
{
"epoch": 0.9083976741969307,
"grad_norm": 0.7546805739402771,
"learning_rate": 2.1980203298027135e-06,
"loss": 1.1844099044799805,
"step": 4765
},
{
"epoch": 0.909350872176151,
"grad_norm": 0.8106234073638916,
"learning_rate": 2.152983213389559e-06,
"loss": 1.2425337791442872,
"step": 4770
},
{
"epoch": 0.9103040701553713,
"grad_norm": 0.9149326086044312,
"learning_rate": 2.108402133964771e-06,
"loss": 1.1953652381896973,
"step": 4775
},
{
"epoch": 0.9112572681345915,
"grad_norm": 0.96515953540802,
"learning_rate": 2.0642775164372075e-06,
"loss": 1.2789098739624023,
"step": 4780
},
{
"epoch": 0.9122104661138118,
"grad_norm": 0.720737874507904,
"learning_rate": 2.0206097813651347e-06,
"loss": 1.2279577255249023,
"step": 4785
},
{
"epoch": 0.9131636640930321,
"grad_norm": 0.6376658082008362,
"learning_rate": 1.9773993449521744e-06,
"loss": 1.1840669631958007,
"step": 4790
},
{
"epoch": 0.9141168620722524,
"grad_norm": 0.6828863024711609,
"learning_rate": 1.9346466190433845e-06,
"loss": 1.3013590812683105,
"step": 4795
},
{
"epoch": 0.9150700600514727,
"grad_norm": 0.6297360062599182,
"learning_rate": 1.8923520111213122e-06,
"loss": 1.2824861526489257,
"step": 4800
},
{
"epoch": 0.916023258030693,
"grad_norm": 0.5658792853355408,
"learning_rate": 1.8505159243021263e-06,
"loss": 1.3033199310302734,
"step": 4805
},
{
"epoch": 0.9169764560099133,
"grad_norm": 0.6617979407310486,
"learning_rate": 1.8091387573317464e-06,
"loss": 1.299100112915039,
"step": 4810
},
{
"epoch": 0.9179296539891335,
"grad_norm": 0.7780975103378296,
"learning_rate": 1.7682209045820686e-06,
"loss": 1.2176671981811524,
"step": 4815
},
{
"epoch": 0.9188828519683538,
"grad_norm": 0.6944255232810974,
"learning_rate": 1.727762756047191e-06,
"loss": 1.2845424652099608,
"step": 4820
},
{
"epoch": 0.9198360499475741,
"grad_norm": 0.6028406620025635,
"learning_rate": 1.687764697339711e-06,
"loss": 1.1866291999816894,
"step": 4825
},
{
"epoch": 0.9207892479267944,
"grad_norm": 1.0062178373336792,
"learning_rate": 1.6482271096870283e-06,
"loss": 1.3127197265625,
"step": 4830
},
{
"epoch": 0.9217424459060147,
"grad_norm": 0.8297625780105591,
"learning_rate": 1.6091503699277476e-06,
"loss": 1.2448455810546875,
"step": 4835
},
{
"epoch": 0.922695643885235,
"grad_norm": 0.7154349088668823,
"learning_rate": 1.570534850508043e-06,
"loss": 1.2162049293518067,
"step": 4840
},
{
"epoch": 0.9236488418644553,
"grad_norm": 0.6263326406478882,
"learning_rate": 1.5323809194781269e-06,
"loss": 1.1927183151245118,
"step": 4845
},
{
"epoch": 0.9246020398436755,
"grad_norm": 0.8444077968597412,
"learning_rate": 1.494688940488781e-06,
"loss": 1.1999497413635254,
"step": 4850
},
{
"epoch": 0.9255552378228958,
"grad_norm": 0.6820685863494873,
"learning_rate": 1.457459272787809e-06,
"loss": 1.2206403732299804,
"step": 4855
},
{
"epoch": 0.9265084358021161,
"grad_norm": 0.7888537645339966,
"learning_rate": 1.4206922712166725e-06,
"loss": 1.3106781005859376,
"step": 4860
},
{
"epoch": 0.9274616337813364,
"grad_norm": 0.7488760948181152,
"learning_rate": 1.384388286207111e-06,
"loss": 1.247926902770996,
"step": 4865
},
{
"epoch": 0.9284148317605566,
"grad_norm": 0.6490522027015686,
"learning_rate": 1.348547663777766e-06,
"loss": 1.2276580810546875,
"step": 4870
},
{
"epoch": 0.929368029739777,
"grad_norm": 0.8611010909080505,
"learning_rate": 1.3131707455309006e-06,
"loss": 1.1637331008911134,
"step": 4875
},
{
"epoch": 0.9303212277189973,
"grad_norm": 0.6082998514175415,
"learning_rate": 1.2782578686491575e-06,
"loss": 1.2539389610290528,
"step": 4880
},
{
"epoch": 0.9312744256982175,
"grad_norm": 0.7314949035644531,
"learning_rate": 1.2438093658923234e-06,
"loss": 1.2367414474487304,
"step": 4885
},
{
"epoch": 0.9322276236774378,
"grad_norm": 0.7305496335029602,
"learning_rate": 1.2098255655941692e-06,
"loss": 1.2181676864624023,
"step": 4890
},
{
"epoch": 0.933180821656658,
"grad_norm": 0.7598952651023865,
"learning_rate": 1.1763067916593262e-06,
"loss": 1.2679986000061034,
"step": 4895
},
{
"epoch": 0.9341340196358784,
"grad_norm": 0.6616954207420349,
"learning_rate": 1.1432533635601705e-06,
"loss": 1.1741644859313964,
"step": 4900
},
{
"epoch": 0.9350872176150986,
"grad_norm": 0.7101142406463623,
"learning_rate": 1.1106655963338152e-06,
"loss": 1.125762939453125,
"step": 4905
},
{
"epoch": 0.9360404155943189,
"grad_norm": 1.0871772766113281,
"learning_rate": 1.0785438005790849e-06,
"loss": 1.225977897644043,
"step": 4910
},
{
"epoch": 0.9369936135735393,
"grad_norm": 0.7330594062805176,
"learning_rate": 1.0468882824535674e-06,
"loss": 1.2087047576904297,
"step": 4915
},
{
"epoch": 0.9379468115527595,
"grad_norm": 0.5671722292900085,
"learning_rate": 1.0156993436706785e-06,
"loss": 1.2688034057617188,
"step": 4920
},
{
"epoch": 0.9389000095319798,
"grad_norm": 0.5870729088783264,
"learning_rate": 9.849772814968017e-07,
"loss": 1.2404970169067382,
"step": 4925
},
{
"epoch": 0.9398532075112,
"grad_norm": 0.8348343372344971,
"learning_rate": 9.54722388748458e-07,
"loss": 1.1828479766845703,
"step": 4930
},
{
"epoch": 0.9408064054904204,
"grad_norm": 0.6649323105812073,
"learning_rate": 9.249349537894969e-07,
"loss": 1.2183140754699706,
"step": 4935
},
{
"epoch": 0.9417596034696406,
"grad_norm": 0.8177406787872314,
"learning_rate": 8.956152605283652e-07,
"loss": 1.2287372589111327,
"step": 4940
},
{
"epoch": 0.9427128014488609,
"grad_norm": 0.741066038608551,
"learning_rate": 8.667635884153979e-07,
"loss": 1.1585744857788085,
"step": 4945
},
{
"epoch": 0.9436659994280812,
"grad_norm": 0.6713501811027527,
"learning_rate": 8.383802124401374e-07,
"loss": 1.2805383682250977,
"step": 4950
},
{
"epoch": 0.9446191974073015,
"grad_norm": 0.775189995765686,
"learning_rate": 8.104654031287406e-07,
"loss": 1.2360420227050781,
"step": 4955
},
{
"epoch": 0.9455723953865218,
"grad_norm": 0.7879636287689209,
"learning_rate": 7.830194265413759e-07,
"loss": 1.2383962631225587,
"step": 4960
},
{
"epoch": 0.946525593365742,
"grad_norm": 0.674618661403656,
"learning_rate": 7.560425442697028e-07,
"loss": 1.2733049392700195,
"step": 4965
},
{
"epoch": 0.9474787913449624,
"grad_norm": 0.652667760848999,
"learning_rate": 7.295350134343793e-07,
"loss": 1.1778074264526368,
"step": 4970
},
{
"epoch": 0.9484319893241826,
"grad_norm": 0.7130538821220398,
"learning_rate": 7.034970866825974e-07,
"loss": 1.154177761077881,
"step": 4975
},
{
"epoch": 0.9493851873034029,
"grad_norm": 0.6210470795631409,
"learning_rate": 6.779290121856796e-07,
"loss": 1.2111547470092774,
"step": 4980
},
{
"epoch": 0.9503383852826232,
"grad_norm": 0.6812198758125305,
"learning_rate": 6.52831033636725e-07,
"loss": 1.1659784317016602,
"step": 4985
},
{
"epoch": 0.9512915832618435,
"grad_norm": 0.9577103853225708,
"learning_rate": 6.282033902482831e-07,
"loss": 1.1558143615722656,
"step": 4990
},
{
"epoch": 0.9522447812410638,
"grad_norm": 0.6614917516708374,
"learning_rate": 6.04046316750051e-07,
"loss": 1.1943655014038086,
"step": 4995
},
{
"epoch": 0.953197979220284,
"grad_norm": 0.7641220688819885,
"learning_rate": 5.80360043386674e-07,
"loss": 1.282275390625,
"step": 5000
},
{
"epoch": 0.9541511771995044,
"grad_norm": 0.6209828853607178,
"learning_rate": 5.571447959155263e-07,
"loss": 1.2226948738098145,
"step": 5005
},
{
"epoch": 0.9551043751787246,
"grad_norm": 0.6895093321800232,
"learning_rate": 5.34400795604556e-07,
"loss": 1.2642387390136718,
"step": 5010
},
{
"epoch": 0.9560575731579449,
"grad_norm": 0.7569581270217896,
"learning_rate": 5.121282592301934e-07,
"loss": 1.1798085212707519,
"step": 5015
},
{
"epoch": 0.9570107711371652,
"grad_norm": 0.6797343492507935,
"learning_rate": 4.903273990752744e-07,
"loss": 1.2560643196105956,
"step": 5020
},
{
"epoch": 0.9579639691163855,
"grad_norm": 0.9098554849624634,
"learning_rate": 4.68998422927025e-07,
"loss": 1.1735345840454101,
"step": 5025
},
{
"epoch": 0.9589171670956057,
"grad_norm": 0.8553006649017334,
"learning_rate": 4.481415340750639e-07,
"loss": 1.2700979232788085,
"step": 5030
},
{
"epoch": 0.959870365074826,
"grad_norm": 0.8495411276817322,
"learning_rate": 4.277569313094809e-07,
"loss": 1.0952903747558593,
"step": 5035
},
{
"epoch": 0.9608235630540464,
"grad_norm": 0.7445310354232788,
"learning_rate": 4.0784480891893906e-07,
"loss": 1.2339123725891112,
"step": 5040
},
{
"epoch": 0.9617767610332666,
"grad_norm": 0.6860194802284241,
"learning_rate": 3.884053566888146e-07,
"loss": 1.2385534286499023,
"step": 5045
},
{
"epoch": 0.9627299590124869,
"grad_norm": 0.6623778343200684,
"learning_rate": 3.694387598994098e-07,
"loss": 1.2900373458862304,
"step": 5050
},
{
"epoch": 0.9636831569917071,
"grad_norm": 0.6038623452186584,
"learning_rate": 3.5094519932415417e-07,
"loss": 1.139063262939453,
"step": 5055
},
{
"epoch": 0.9646363549709275,
"grad_norm": 0.8033509254455566,
"learning_rate": 3.3292485122790593e-07,
"loss": 1.2035551071166992,
"step": 5060
},
{
"epoch": 0.9655895529501477,
"grad_norm": 0.7523816823959351,
"learning_rate": 3.1537788736527017e-07,
"loss": 1.1845973968505858,
"step": 5065
},
{
"epoch": 0.966542750929368,
"grad_norm": 0.7979245185852051,
"learning_rate": 2.983044749789554e-07,
"loss": 1.1925546646118164,
"step": 5070
},
{
"epoch": 0.9674959489085884,
"grad_norm": 0.8654818534851074,
"learning_rate": 2.817047767981695e-07,
"loss": 1.1745403289794922,
"step": 5075
},
{
"epoch": 0.9684491468878086,
"grad_norm": 0.639470100402832,
"learning_rate": 2.6557895103709875e-07,
"loss": 1.247914695739746,
"step": 5080
},
{
"epoch": 0.9694023448670289,
"grad_norm": 0.6247613430023193,
"learning_rate": 2.4992715139335877e-07,
"loss": 1.233420181274414,
"step": 5085
},
{
"epoch": 0.9703555428462491,
"grad_norm": 0.7666287422180176,
"learning_rate": 2.3474952704657939e-07,
"loss": 1.1596334457397461,
"step": 5090
},
{
"epoch": 0.9713087408254695,
"grad_norm": 0.6626922488212585,
"learning_rate": 2.2004622265693887e-07,
"loss": 1.2604512214660644,
"step": 5095
},
{
"epoch": 0.9722619388046897,
"grad_norm": 0.9594115614891052,
"learning_rate": 2.0581737836379289e-07,
"loss": 1.2246232986450196,
"step": 5100
},
{
"epoch": 0.97321513678391,
"grad_norm": 0.5604175925254822,
"learning_rate": 1.9206312978437557e-07,
"loss": 1.1933224678039551,
"step": 5105
},
{
"epoch": 0.9741683347631303,
"grad_norm": 0.9626731872558594,
"learning_rate": 1.7878360801245054e-07,
"loss": 1.154996395111084,
"step": 5110
},
{
"epoch": 0.9751215327423506,
"grad_norm": 0.7064977288246155,
"learning_rate": 1.659789396171063e-07,
"loss": 1.2570127487182616,
"step": 5115
},
{
"epoch": 0.9760747307215709,
"grad_norm": 0.5774402618408203,
"learning_rate": 1.5364924664152957e-07,
"loss": 1.2565927505493164,
"step": 5120
},
{
"epoch": 0.9770279287007911,
"grad_norm": 0.5336069464683533,
"learning_rate": 1.4179464660184493e-07,
"loss": 1.1792776107788085,
"step": 5125
},
{
"epoch": 0.9779811266800115,
"grad_norm": 0.7254840731620789,
"learning_rate": 1.3041525248599916e-07,
"loss": 1.2528850555419921,
"step": 5130
},
{
"epoch": 0.9789343246592317,
"grad_norm": 0.6289318203926086,
"learning_rate": 1.195111727526843e-07,
"loss": 1.2054006576538085,
"step": 5135
},
{
"epoch": 0.979887522638452,
"grad_norm": 0.6436653733253479,
"learning_rate": 1.0908251133028847e-07,
"loss": 1.134188175201416,
"step": 5140
},
{
"epoch": 0.9808407206176722,
"grad_norm": 0.6223757266998291,
"learning_rate": 9.91293676159244e-08,
"loss": 1.1787416458129882,
"step": 5145
},
{
"epoch": 0.9817939185968926,
"grad_norm": 0.933802604675293,
"learning_rate": 8.965183647448028e-08,
"loss": 1.2939325332641602,
"step": 5150
},
{
"epoch": 0.9827471165761129,
"grad_norm": 0.8259690999984741,
"learning_rate": 8.06500082377093e-08,
"loss": 1.0475672721862792,
"step": 5155
},
{
"epoch": 0.9837003145553331,
"grad_norm": 0.6043301820755005,
"learning_rate": 7.212396870335258e-08,
"loss": 1.1976526260375977,
"step": 5160
},
{
"epoch": 0.9846535125345535,
"grad_norm": 0.7538432478904724,
"learning_rate": 6.407379913436762e-08,
"loss": 1.1764219284057618,
"step": 5165
},
{
"epoch": 0.9856067105137737,
"grad_norm": 0.6621540188789368,
"learning_rate": 5.649957625810664e-08,
"loss": 1.2455005645751953,
"step": 5170
},
{
"epoch": 0.986559908492994,
"grad_norm": 0.8748796582221985,
"learning_rate": 4.940137226560615e-08,
"loss": 1.177157211303711,
"step": 5175
},
{
"epoch": 0.9875131064722142,
"grad_norm": 0.673653244972229,
"learning_rate": 4.2779254810892953e-08,
"loss": 1.1055315017700196,
"step": 5180
},
{
"epoch": 0.9884663044514346,
"grad_norm": 0.6079859733581543,
"learning_rate": 3.663328701035695e-08,
"loss": 1.2303492546081543,
"step": 5185
},
{
"epoch": 0.9894195024306548,
"grad_norm": 0.7685458660125732,
"learning_rate": 3.096352744211828e-08,
"loss": 1.2230051040649415,
"step": 5190
},
{
"epoch": 0.9903727004098751,
"grad_norm": 0.6509291529655457,
"learning_rate": 2.5770030145494394e-08,
"loss": 1.1461639404296875,
"step": 5195
},
{
"epoch": 0.9913258983890955,
"grad_norm": 0.7126690745353699,
"learning_rate": 2.1052844620472743e-08,
"loss": 1.2010046005249024,
"step": 5200
},
{
"epoch": 0.9922790963683157,
"grad_norm": 0.7618613243103027,
"learning_rate": 1.6812015827255555e-08,
"loss": 1.2073261260986328,
"step": 5205
},
{
"epoch": 0.993232294347536,
"grad_norm": 0.6913411021232605,
"learning_rate": 1.3047584185793549e-08,
"loss": 1.1698974609375,
"step": 5210
},
{
"epoch": 0.9941854923267562,
"grad_norm": 0.6507450938224792,
"learning_rate": 9.75958557545842e-09,
"loss": 1.1786861419677734,
"step": 5215
},
{
"epoch": 0.9951386903059766,
"grad_norm": 0.5883746147155762,
"learning_rate": 6.948051334637606e-09,
"loss": 1.1779633522033692,
"step": 5220
},
{
"epoch": 0.9960918882851968,
"grad_norm": 0.7004103064537048,
"learning_rate": 4.613008260490048e-09,
"loss": 1.2079018592834472,
"step": 5225
},
{
"epoch": 0.9970450862644171,
"grad_norm": 0.718757152557373,
"learning_rate": 2.7544786086464157e-09,
"loss": 1.1970481872558594,
"step": 5230
},
{
"epoch": 0.9979982842436375,
"grad_norm": 0.6223225593566895,
"learning_rate": 1.3724800930314807e-09,
"loss": 1.2528276443481445,
"step": 5235
},
{
"epoch": 0.9989514822228577,
"grad_norm": 0.7650284171104431,
"learning_rate": 4.670258856753762e-10,
"loss": 1.2802021026611328,
"step": 5240
},
{
"epoch": 0.999904680202078,
"grad_norm": 0.5030611753463745,
"learning_rate": 3.812461660257505e-11,
"loss": 1.2648646354675293,
"step": 5245
}
],
"logging_steps": 5,
"max_steps": 5246,
"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": 3.043371807492014e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}