tttw1003 / trainer_state.json
Kaipengm2's picture
Upload folder using huggingface_hub
63c6921 verified
Raw
History Blame Contribute Delete
189 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 5069,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0009864851533984414,
"grad_norm": 3.2353720664978027,
"learning_rate": 2.6143790849673204e-06,
"loss": 2.966313934326172,
"step": 5
},
{
"epoch": 0.0019729703067968827,
"grad_norm": 2.61134934425354,
"learning_rate": 5.882352941176471e-06,
"loss": 2.8623239517211916,
"step": 10
},
{
"epoch": 0.002959455460195324,
"grad_norm": 1.9227769374847412,
"learning_rate": 9.150326797385621e-06,
"loss": 2.8137367248535154,
"step": 15
},
{
"epoch": 0.003945940613593765,
"grad_norm": 1.2290284633636475,
"learning_rate": 1.2418300653594772e-05,
"loss": 2.7035795211791993,
"step": 20
},
{
"epoch": 0.0049324257669922066,
"grad_norm": 0.9648247957229614,
"learning_rate": 1.568627450980392e-05,
"loss": 2.6373886108398437,
"step": 25
},
{
"epoch": 0.005918910920390648,
"grad_norm": 0.6019311547279358,
"learning_rate": 1.895424836601307e-05,
"loss": 2.6448528289794924,
"step": 30
},
{
"epoch": 0.00690539607378909,
"grad_norm": 0.8435561656951904,
"learning_rate": 2.2222222222222223e-05,
"loss": 2.5108722686767577,
"step": 35
},
{
"epoch": 0.00789188122718753,
"grad_norm": 0.5805186629295349,
"learning_rate": 2.5490196078431373e-05,
"loss": 2.300589179992676,
"step": 40
},
{
"epoch": 0.008878366380585973,
"grad_norm": 0.5885493755340576,
"learning_rate": 2.8758169934640522e-05,
"loss": 2.332649040222168,
"step": 45
},
{
"epoch": 0.009864851533984413,
"grad_norm": 0.5239945650100708,
"learning_rate": 3.202614379084967e-05,
"loss": 2.283979797363281,
"step": 50
},
{
"epoch": 0.010851336687382855,
"grad_norm": 0.6694562435150146,
"learning_rate": 3.529411764705883e-05,
"loss": 2.2206966400146486,
"step": 55
},
{
"epoch": 0.011837821840781295,
"grad_norm": 0.49268701672554016,
"learning_rate": 3.8562091503267977e-05,
"loss": 2.1527887344360352,
"step": 60
},
{
"epoch": 0.012824306994179737,
"grad_norm": 0.4785483777523041,
"learning_rate": 4.1830065359477126e-05,
"loss": 2.1107677459716796,
"step": 65
},
{
"epoch": 0.01381079214757818,
"grad_norm": 0.5008707642555237,
"learning_rate": 4.5098039215686275e-05,
"loss": 2.095572280883789,
"step": 70
},
{
"epoch": 0.01479727730097662,
"grad_norm": 0.5966643691062927,
"learning_rate": 4.8366013071895424e-05,
"loss": 2.1681396484375,
"step": 75
},
{
"epoch": 0.01578376245437506,
"grad_norm": 0.5406578779220581,
"learning_rate": 5.163398692810458e-05,
"loss": 2.0038593292236326,
"step": 80
},
{
"epoch": 0.016770247607773504,
"grad_norm": 0.49416959285736084,
"learning_rate": 5.490196078431373e-05,
"loss": 1.9989547729492188,
"step": 85
},
{
"epoch": 0.017756732761171946,
"grad_norm": 0.818135678768158,
"learning_rate": 5.8169934640522886e-05,
"loss": 1.9833293914794923,
"step": 90
},
{
"epoch": 0.018743217914570384,
"grad_norm": 0.5396987795829773,
"learning_rate": 6.143790849673203e-05,
"loss": 1.9252159118652343,
"step": 95
},
{
"epoch": 0.019729703067968826,
"grad_norm": 0.6068782806396484,
"learning_rate": 6.470588235294118e-05,
"loss": 1.9406425476074218,
"step": 100
},
{
"epoch": 0.020716188221367268,
"grad_norm": 0.5170368552207947,
"learning_rate": 6.797385620915033e-05,
"loss": 1.8097198486328125,
"step": 105
},
{
"epoch": 0.02170267337476571,
"grad_norm": 0.5425059199333191,
"learning_rate": 7.124183006535948e-05,
"loss": 1.8881669998168946,
"step": 110
},
{
"epoch": 0.022689158528164152,
"grad_norm": 0.7135457396507263,
"learning_rate": 7.450980392156864e-05,
"loss": 1.8165441513061524,
"step": 115
},
{
"epoch": 0.02367564368156259,
"grad_norm": 0.677783727645874,
"learning_rate": 7.777777777777778e-05,
"loss": 1.8519512176513673,
"step": 120
},
{
"epoch": 0.024662128834961033,
"grad_norm": 0.5774139165878296,
"learning_rate": 8.104575163398694e-05,
"loss": 1.8139636993408204,
"step": 125
},
{
"epoch": 0.025648613988359475,
"grad_norm": 0.6886249780654907,
"learning_rate": 8.431372549019608e-05,
"loss": 1.7861183166503907,
"step": 130
},
{
"epoch": 0.026635099141757917,
"grad_norm": 0.6761642098426819,
"learning_rate": 8.758169934640524e-05,
"loss": 1.7952146530151367,
"step": 135
},
{
"epoch": 0.02762158429515636,
"grad_norm": 0.633804202079773,
"learning_rate": 9.084967320261438e-05,
"loss": 1.767338180541992,
"step": 140
},
{
"epoch": 0.0286080694485548,
"grad_norm": 0.777923047542572,
"learning_rate": 9.411764705882353e-05,
"loss": 1.7864761352539062,
"step": 145
},
{
"epoch": 0.02959455460195324,
"grad_norm": 0.6302430629730225,
"learning_rate": 9.738562091503268e-05,
"loss": 1.7012062072753906,
"step": 150
},
{
"epoch": 0.03058103975535168,
"grad_norm": 0.7225821018218994,
"learning_rate": 9.999998979022925e-05,
"loss": 1.7333024978637694,
"step": 155
},
{
"epoch": 0.03156752490875012,
"grad_norm": 0.6830950379371643,
"learning_rate": 9.999963244869056e-05,
"loss": 1.717333984375,
"step": 160
},
{
"epoch": 0.032554010062148565,
"grad_norm": 0.7561411261558533,
"learning_rate": 9.999876462278359e-05,
"loss": 1.7478538513183595,
"step": 165
},
{
"epoch": 0.03354049521554701,
"grad_norm": 0.5661100745201111,
"learning_rate": 9.999738632136865e-05,
"loss": 1.7332321166992188,
"step": 170
},
{
"epoch": 0.03452698036894545,
"grad_norm": 0.6723949313163757,
"learning_rate": 9.999549755851785e-05,
"loss": 1.667428207397461,
"step": 175
},
{
"epoch": 0.03551346552234389,
"grad_norm": 0.709575355052948,
"learning_rate": 9.999309835351502e-05,
"loss": 1.7241680145263671,
"step": 180
},
{
"epoch": 0.03649995067574233,
"grad_norm": 0.6242716312408447,
"learning_rate": 9.999018873085547e-05,
"loss": 1.7063037872314453,
"step": 185
},
{
"epoch": 0.03748643582914077,
"grad_norm": 0.6306405663490295,
"learning_rate": 9.998676872024577e-05,
"loss": 1.6588741302490235,
"step": 190
},
{
"epoch": 0.03847292098253921,
"grad_norm": 0.673144519329071,
"learning_rate": 9.998283835660338e-05,
"loss": 1.7226240158081054,
"step": 195
},
{
"epoch": 0.03945940613593765,
"grad_norm": 0.5926720499992371,
"learning_rate": 9.997839768005642e-05,
"loss": 1.6143610000610351,
"step": 200
},
{
"epoch": 0.040445891289336094,
"grad_norm": 0.6743038296699524,
"learning_rate": 9.997344673594313e-05,
"loss": 1.612361526489258,
"step": 205
},
{
"epoch": 0.041432376442734536,
"grad_norm": 0.6282744407653809,
"learning_rate": 9.996798557481146e-05,
"loss": 1.63587646484375,
"step": 210
},
{
"epoch": 0.04241886159613298,
"grad_norm": 0.7004051208496094,
"learning_rate": 9.996201425241858e-05,
"loss": 1.672429656982422,
"step": 215
},
{
"epoch": 0.04340534674953142,
"grad_norm": 0.9393495917320251,
"learning_rate": 9.995553282973025e-05,
"loss": 1.5298744201660157,
"step": 220
},
{
"epoch": 0.04439183190292986,
"grad_norm": 0.5777098536491394,
"learning_rate": 9.994854137292029e-05,
"loss": 1.6657585144042968,
"step": 225
},
{
"epoch": 0.045378317056328304,
"grad_norm": 0.6550586223602295,
"learning_rate": 9.994103995336979e-05,
"loss": 1.6891620635986329,
"step": 230
},
{
"epoch": 0.046364802209726746,
"grad_norm": 0.6067536473274231,
"learning_rate": 9.99330286476665e-05,
"loss": 1.6078147888183594,
"step": 235
},
{
"epoch": 0.04735128736312518,
"grad_norm": 0.8295519948005676,
"learning_rate": 9.992450753760388e-05,
"loss": 1.6302314758300782,
"step": 240
},
{
"epoch": 0.048337772516523624,
"grad_norm": 0.6381354928016663,
"learning_rate": 9.99154767101805e-05,
"loss": 1.6644575119018554,
"step": 245
},
{
"epoch": 0.049324257669922066,
"grad_norm": 0.7252200245857239,
"learning_rate": 9.990593625759891e-05,
"loss": 1.6170015335083008,
"step": 250
},
{
"epoch": 0.05031074282332051,
"grad_norm": 0.6090466976165771,
"learning_rate": 9.989588627726492e-05,
"loss": 1.6528291702270508,
"step": 255
},
{
"epoch": 0.05129722797671895,
"grad_norm": 0.8103018403053284,
"learning_rate": 9.98853268717864e-05,
"loss": 1.5820199012756349,
"step": 260
},
{
"epoch": 0.05228371313011739,
"grad_norm": 0.7244828343391418,
"learning_rate": 9.987425814897239e-05,
"loss": 1.5641636848449707,
"step": 265
},
{
"epoch": 0.053270198283515834,
"grad_norm": 0.7593345642089844,
"learning_rate": 9.986268022183192e-05,
"loss": 1.6236160278320313,
"step": 270
},
{
"epoch": 0.054256683436914276,
"grad_norm": 0.6060400605201721,
"learning_rate": 9.985059320857287e-05,
"loss": 1.5641530990600585,
"step": 275
},
{
"epoch": 0.05524316859031272,
"grad_norm": 0.8013601899147034,
"learning_rate": 9.983799723260076e-05,
"loss": 1.601710891723633,
"step": 280
},
{
"epoch": 0.05622965374371116,
"grad_norm": 0.832247793674469,
"learning_rate": 9.982489242251753e-05,
"loss": 1.6219011306762696,
"step": 285
},
{
"epoch": 0.0572161388971096,
"grad_norm": 0.5210625529289246,
"learning_rate": 9.981127891212019e-05,
"loss": 1.652933692932129,
"step": 290
},
{
"epoch": 0.05820262405050804,
"grad_norm": 0.6245460510253906,
"learning_rate": 9.979715684039943e-05,
"loss": 1.655886459350586,
"step": 295
},
{
"epoch": 0.05918910920390648,
"grad_norm": 0.7012748122215271,
"learning_rate": 9.978252635153824e-05,
"loss": 1.631332778930664,
"step": 300
},
{
"epoch": 0.06017559435730492,
"grad_norm": 0.6129679679870605,
"learning_rate": 9.976738759491044e-05,
"loss": 1.6186561584472656,
"step": 305
},
{
"epoch": 0.06116207951070336,
"grad_norm": 0.6349156498908997,
"learning_rate": 9.975174072507916e-05,
"loss": 1.5888310432434083,
"step": 310
},
{
"epoch": 0.062148564664101805,
"grad_norm": 0.544347882270813,
"learning_rate": 9.973558590179519e-05,
"loss": 1.5164371490478517,
"step": 315
},
{
"epoch": 0.06313504981750025,
"grad_norm": 0.5962512493133545,
"learning_rate": 9.971892328999545e-05,
"loss": 1.5541616439819337,
"step": 320
},
{
"epoch": 0.06412153497089869,
"grad_norm": 0.7568216919898987,
"learning_rate": 9.970175305980126e-05,
"loss": 1.6129310607910157,
"step": 325
},
{
"epoch": 0.06510802012429713,
"grad_norm": 0.6227603554725647,
"learning_rate": 9.968407538651659e-05,
"loss": 1.6116630554199218,
"step": 330
},
{
"epoch": 0.06609450527769557,
"grad_norm": 0.5311099886894226,
"learning_rate": 9.966589045062625e-05,
"loss": 1.5331324577331542,
"step": 335
},
{
"epoch": 0.06708099043109401,
"grad_norm": 0.8698157668113708,
"learning_rate": 9.964719843779417e-05,
"loss": 1.5205979347229004,
"step": 340
},
{
"epoch": 0.06806747558449246,
"grad_norm": 0.7399280667304993,
"learning_rate": 9.962799953886132e-05,
"loss": 1.5424941062927247,
"step": 345
},
{
"epoch": 0.0690539607378909,
"grad_norm": 0.5627014636993408,
"learning_rate": 9.96082939498439e-05,
"loss": 1.5411433219909667,
"step": 350
},
{
"epoch": 0.07004044589128934,
"grad_norm": 0.5627856850624084,
"learning_rate": 9.958808187193132e-05,
"loss": 1.4988202095031737,
"step": 355
},
{
"epoch": 0.07102693104468778,
"grad_norm": 0.7122822403907776,
"learning_rate": 9.956736351148406e-05,
"loss": 1.5739829063415527,
"step": 360
},
{
"epoch": 0.07201341619808622,
"grad_norm": 0.7875910997390747,
"learning_rate": 9.954613908003167e-05,
"loss": 1.5903581619262694,
"step": 365
},
{
"epoch": 0.07299990135148467,
"grad_norm": 0.582419216632843,
"learning_rate": 9.952440879427055e-05,
"loss": 1.557723045349121,
"step": 370
},
{
"epoch": 0.0739863865048831,
"grad_norm": 0.7579030990600586,
"learning_rate": 9.950217287606177e-05,
"loss": 1.5146641731262207,
"step": 375
},
{
"epoch": 0.07497287165828154,
"grad_norm": 0.7362158298492432,
"learning_rate": 9.947943155242876e-05,
"loss": 1.516170597076416,
"step": 380
},
{
"epoch": 0.07595935681167998,
"grad_norm": 0.5797917246818542,
"learning_rate": 9.945618505555503e-05,
"loss": 1.5603564262390137,
"step": 385
},
{
"epoch": 0.07694584196507842,
"grad_norm": 0.5989718437194824,
"learning_rate": 9.94324336227818e-05,
"loss": 1.5771269798278809,
"step": 390
},
{
"epoch": 0.07793232711847686,
"grad_norm": 0.6076745986938477,
"learning_rate": 9.940817749660554e-05,
"loss": 1.5456730842590332,
"step": 395
},
{
"epoch": 0.0789188122718753,
"grad_norm": 0.6797211766242981,
"learning_rate": 9.938341692467555e-05,
"loss": 1.4972328186035155,
"step": 400
},
{
"epoch": 0.07990529742527375,
"grad_norm": 0.6801923513412476,
"learning_rate": 9.935815215979138e-05,
"loss": 1.553426170349121,
"step": 405
},
{
"epoch": 0.08089178257867219,
"grad_norm": 0.7642124891281128,
"learning_rate": 9.933238345990027e-05,
"loss": 1.478465747833252,
"step": 410
},
{
"epoch": 0.08187826773207063,
"grad_norm": 0.5797330737113953,
"learning_rate": 9.930611108809453e-05,
"loss": 1.5628379821777343,
"step": 415
},
{
"epoch": 0.08286475288546907,
"grad_norm": 0.7465011477470398,
"learning_rate": 9.927933531260885e-05,
"loss": 1.5611876487731933,
"step": 420
},
{
"epoch": 0.08385123803886751,
"grad_norm": 0.6690396070480347,
"learning_rate": 9.925205640681751e-05,
"loss": 1.5046161651611327,
"step": 425
},
{
"epoch": 0.08483772319226596,
"grad_norm": 0.5959020853042603,
"learning_rate": 9.922427464923167e-05,
"loss": 1.4976409912109374,
"step": 430
},
{
"epoch": 0.0858242083456644,
"grad_norm": 0.6543482542037964,
"learning_rate": 9.919599032349648e-05,
"loss": 1.5222838401794434,
"step": 435
},
{
"epoch": 0.08681069349906284,
"grad_norm": 0.7326239347457886,
"learning_rate": 9.916720371838817e-05,
"loss": 1.48922700881958,
"step": 440
},
{
"epoch": 0.08779717865246128,
"grad_norm": 0.5853524804115295,
"learning_rate": 9.913791512781116e-05,
"loss": 1.5864839553833008,
"step": 445
},
{
"epoch": 0.08878366380585972,
"grad_norm": 0.6233240365982056,
"learning_rate": 9.910812485079499e-05,
"loss": 1.546849536895752,
"step": 450
},
{
"epoch": 0.08977014895925817,
"grad_norm": 0.5473343133926392,
"learning_rate": 9.907783319149129e-05,
"loss": 1.4771411895751954,
"step": 455
},
{
"epoch": 0.09075663411265661,
"grad_norm": 0.6460210680961609,
"learning_rate": 9.904704045917075e-05,
"loss": 1.5375116348266602,
"step": 460
},
{
"epoch": 0.09174311926605505,
"grad_norm": 0.7441430687904358,
"learning_rate": 9.901574696821981e-05,
"loss": 1.4876806259155273,
"step": 465
},
{
"epoch": 0.09272960441945349,
"grad_norm": 0.5090604424476624,
"learning_rate": 9.898395303813761e-05,
"loss": 1.5206018447875977,
"step": 470
},
{
"epoch": 0.09371608957285193,
"grad_norm": 0.5529798269271851,
"learning_rate": 9.895165899353259e-05,
"loss": 1.4785223007202148,
"step": 475
},
{
"epoch": 0.09470257472625036,
"grad_norm": 0.6639275550842285,
"learning_rate": 9.89188651641193e-05,
"loss": 1.4929648399353028,
"step": 480
},
{
"epoch": 0.0956890598796488,
"grad_norm": 1.0868035554885864,
"learning_rate": 9.888557188471492e-05,
"loss": 1.564094638824463,
"step": 485
},
{
"epoch": 0.09667554503304725,
"grad_norm": 0.4676032066345215,
"learning_rate": 9.885177949523597e-05,
"loss": 1.4715606689453125,
"step": 490
},
{
"epoch": 0.09766203018644569,
"grad_norm": 0.7386883497238159,
"learning_rate": 9.881748834069466e-05,
"loss": 1.5144580841064452,
"step": 495
},
{
"epoch": 0.09864851533984413,
"grad_norm": 0.5823488235473633,
"learning_rate": 9.878269877119557e-05,
"loss": 1.4615087509155273,
"step": 500
},
{
"epoch": 0.09963500049324257,
"grad_norm": 0.6547715663909912,
"learning_rate": 9.874741114193194e-05,
"loss": 1.5034366607666017,
"step": 505
},
{
"epoch": 0.10062148564664102,
"grad_norm": 0.5729165077209473,
"learning_rate": 9.871162581318205e-05,
"loss": 1.5307487487792968,
"step": 510
},
{
"epoch": 0.10160797080003946,
"grad_norm": 0.5744174122810364,
"learning_rate": 9.867534315030561e-05,
"loss": 1.554220962524414,
"step": 515
},
{
"epoch": 0.1025944559534379,
"grad_norm": 0.6196578741073608,
"learning_rate": 9.863856352374004e-05,
"loss": 1.5153225898742675,
"step": 520
},
{
"epoch": 0.10358094110683634,
"grad_norm": 0.7502561211585999,
"learning_rate": 9.860128730899653e-05,
"loss": 1.552648162841797,
"step": 525
},
{
"epoch": 0.10456742626023478,
"grad_norm": 0.7757261395454407,
"learning_rate": 9.85635148866564e-05,
"loss": 1.437141513824463,
"step": 530
},
{
"epoch": 0.10555391141363323,
"grad_norm": 0.5391474962234497,
"learning_rate": 9.852524664236709e-05,
"loss": 1.5158101081848145,
"step": 535
},
{
"epoch": 0.10654039656703167,
"grad_norm": 0.6243263483047485,
"learning_rate": 9.848648296683831e-05,
"loss": 1.5096162796020507,
"step": 540
},
{
"epoch": 0.10752688172043011,
"grad_norm": 0.5132644176483154,
"learning_rate": 9.844722425583797e-05,
"loss": 1.5241951942443848,
"step": 545
},
{
"epoch": 0.10851336687382855,
"grad_norm": 0.6474339365959167,
"learning_rate": 9.840747091018816e-05,
"loss": 1.5104016304016112,
"step": 550
},
{
"epoch": 0.109499852027227,
"grad_norm": 0.6036747097969055,
"learning_rate": 9.836722333576113e-05,
"loss": 1.4814903259277343,
"step": 555
},
{
"epoch": 0.11048633718062544,
"grad_norm": 0.6872631907463074,
"learning_rate": 9.832648194347503e-05,
"loss": 1.4320192337036133,
"step": 560
},
{
"epoch": 0.11147282233402388,
"grad_norm": 0.5338987112045288,
"learning_rate": 9.828524714928979e-05,
"loss": 1.4736339569091796,
"step": 565
},
{
"epoch": 0.11245930748742232,
"grad_norm": 0.5474720597267151,
"learning_rate": 9.824351937420288e-05,
"loss": 1.4333370208740235,
"step": 570
},
{
"epoch": 0.11344579264082076,
"grad_norm": 0.6396050453186035,
"learning_rate": 9.820129904424495e-05,
"loss": 1.5243054389953614,
"step": 575
},
{
"epoch": 0.1144322777942192,
"grad_norm": 0.6535886526107788,
"learning_rate": 9.815858659047557e-05,
"loss": 1.4550440788269043,
"step": 580
},
{
"epoch": 0.11541876294761765,
"grad_norm": 0.6298052668571472,
"learning_rate": 9.811538244897871e-05,
"loss": 1.4656028747558594,
"step": 585
},
{
"epoch": 0.11640524810101607,
"grad_norm": 0.75309157371521,
"learning_rate": 9.80716870608584e-05,
"loss": 1.4870981216430663,
"step": 590
},
{
"epoch": 0.11739173325441452,
"grad_norm": 0.6380122303962708,
"learning_rate": 9.802750087223418e-05,
"loss": 1.5880391120910644,
"step": 595
},
{
"epoch": 0.11837821840781296,
"grad_norm": 0.6400323510169983,
"learning_rate": 9.798282433423654e-05,
"loss": 1.4787846565246583,
"step": 600
},
{
"epoch": 0.1193647035612114,
"grad_norm": 0.5606880187988281,
"learning_rate": 9.793765790300231e-05,
"loss": 1.4515664100646972,
"step": 605
},
{
"epoch": 0.12035118871460984,
"grad_norm": 0.8095257878303528,
"learning_rate": 9.789200203967005e-05,
"loss": 1.560379409790039,
"step": 610
},
{
"epoch": 0.12133767386800828,
"grad_norm": 0.7955862879753113,
"learning_rate": 9.784585721037522e-05,
"loss": 1.3642581939697265,
"step": 615
},
{
"epoch": 0.12232415902140673,
"grad_norm": 0.6629276871681213,
"learning_rate": 9.77992238862456e-05,
"loss": 1.4662901878356933,
"step": 620
},
{
"epoch": 0.12331064417480517,
"grad_norm": 0.7091234922409058,
"learning_rate": 9.775210254339635e-05,
"loss": 1.4662946701049804,
"step": 625
},
{
"epoch": 0.12429712932820361,
"grad_norm": 0.6879212260246277,
"learning_rate": 9.770449366292521e-05,
"loss": 1.4349470138549805,
"step": 630
},
{
"epoch": 0.12528361448160205,
"grad_norm": 0.528315544128418,
"learning_rate": 9.765639773090747e-05,
"loss": 1.470627212524414,
"step": 635
},
{
"epoch": 0.1262700996350005,
"grad_norm": 0.7543030381202698,
"learning_rate": 9.760781523839121e-05,
"loss": 1.450231170654297,
"step": 640
},
{
"epoch": 0.12725658478839894,
"grad_norm": 0.5807738304138184,
"learning_rate": 9.755874668139218e-05,
"loss": 1.5048587799072266,
"step": 645
},
{
"epoch": 0.12824306994179738,
"grad_norm": 0.5668618679046631,
"learning_rate": 9.750919256088861e-05,
"loss": 1.4813220977783204,
"step": 650
},
{
"epoch": 0.12922955509519582,
"grad_norm": 0.6080701351165771,
"learning_rate": 9.745915338281635e-05,
"loss": 1.4730741500854492,
"step": 655
},
{
"epoch": 0.13021604024859426,
"grad_norm": 0.7769299149513245,
"learning_rate": 9.740862965806348e-05,
"loss": 1.4579566955566405,
"step": 660
},
{
"epoch": 0.1312025254019927,
"grad_norm": 0.5565757751464844,
"learning_rate": 9.735762190246527e-05,
"loss": 1.4760931015014649,
"step": 665
},
{
"epoch": 0.13218901055539115,
"grad_norm": 0.6028071045875549,
"learning_rate": 9.730613063679875e-05,
"loss": 1.5084136962890624,
"step": 670
},
{
"epoch": 0.1331754957087896,
"grad_norm": 0.5760068893432617,
"learning_rate": 9.725415638677753e-05,
"loss": 1.4650484085083009,
"step": 675
},
{
"epoch": 0.13416198086218803,
"grad_norm": 0.7157391905784607,
"learning_rate": 9.720169968304636e-05,
"loss": 1.5059303283691405,
"step": 680
},
{
"epoch": 0.13514846601558647,
"grad_norm": 0.49478286504745483,
"learning_rate": 9.71487610611757e-05,
"loss": 1.475875473022461,
"step": 685
},
{
"epoch": 0.1361349511689849,
"grad_norm": 0.6508141756057739,
"learning_rate": 9.70953410616563e-05,
"loss": 1.3789533615112304,
"step": 690
},
{
"epoch": 0.13712143632238336,
"grad_norm": 0.6664313077926636,
"learning_rate": 9.704144022989368e-05,
"loss": 1.467292594909668,
"step": 695
},
{
"epoch": 0.1381079214757818,
"grad_norm": 0.581700325012207,
"learning_rate": 9.698705911620255e-05,
"loss": 1.4467798233032227,
"step": 700
},
{
"epoch": 0.13909440662918024,
"grad_norm": 0.562256395816803,
"learning_rate": 9.69321982758011e-05,
"loss": 1.4940157890319825,
"step": 705
},
{
"epoch": 0.14008089178257868,
"grad_norm": 0.5404791235923767,
"learning_rate": 9.687685826880553e-05,
"loss": 1.472784423828125,
"step": 710
},
{
"epoch": 0.14106737693597712,
"grad_norm": 0.7314941883087158,
"learning_rate": 9.682103966022413e-05,
"loss": 1.5199737548828125,
"step": 715
},
{
"epoch": 0.14205386208937557,
"grad_norm": 0.6578221917152405,
"learning_rate": 9.676474301995165e-05,
"loss": 1.513593864440918,
"step": 720
},
{
"epoch": 0.143040347242774,
"grad_norm": 0.6490010619163513,
"learning_rate": 9.670796892276341e-05,
"loss": 1.3429681777954101,
"step": 725
},
{
"epoch": 0.14402683239617245,
"grad_norm": 0.7254768013954163,
"learning_rate": 9.665071794830945e-05,
"loss": 1.4317720413208008,
"step": 730
},
{
"epoch": 0.1450133175495709,
"grad_norm": 0.6505016684532166,
"learning_rate": 9.659299068110861e-05,
"loss": 1.4439896583557128,
"step": 735
},
{
"epoch": 0.14599980270296933,
"grad_norm": 0.5306701064109802,
"learning_rate": 9.65347877105426e-05,
"loss": 1.4680278778076172,
"step": 740
},
{
"epoch": 0.14698628785636778,
"grad_norm": 0.571385383605957,
"learning_rate": 9.647610963084989e-05,
"loss": 1.53929443359375,
"step": 745
},
{
"epoch": 0.1479727730097662,
"grad_norm": 0.5654305219650269,
"learning_rate": 9.641695704111975e-05,
"loss": 1.4372037887573241,
"step": 750
},
{
"epoch": 0.14895925816316463,
"grad_norm": 0.6540042161941528,
"learning_rate": 9.635733054528606e-05,
"loss": 1.4520246505737304,
"step": 755
},
{
"epoch": 0.14994574331656307,
"grad_norm": 0.6543537378311157,
"learning_rate": 9.629723075212118e-05,
"loss": 1.420071792602539,
"step": 760
},
{
"epoch": 0.15093222846996152,
"grad_norm": 0.46961069107055664,
"learning_rate": 9.623665827522973e-05,
"loss": 1.4734342575073243,
"step": 765
},
{
"epoch": 0.15191871362335996,
"grad_norm": 0.4992317855358124,
"learning_rate": 9.617561373304229e-05,
"loss": 1.4179805755615233,
"step": 770
},
{
"epoch": 0.1529051987767584,
"grad_norm": 0.4780406951904297,
"learning_rate": 9.611409774880914e-05,
"loss": 1.4355188369750977,
"step": 775
},
{
"epoch": 0.15389168393015684,
"grad_norm": 0.5523859262466431,
"learning_rate": 9.605211095059387e-05,
"loss": 1.4337491989135742,
"step": 780
},
{
"epoch": 0.15487816908355528,
"grad_norm": 0.5934846997261047,
"learning_rate": 9.598965397126694e-05,
"loss": 1.458352279663086,
"step": 785
},
{
"epoch": 0.15586465423695373,
"grad_norm": 0.6989786028862,
"learning_rate": 9.592672744849931e-05,
"loss": 1.5500715255737305,
"step": 790
},
{
"epoch": 0.15685113939035217,
"grad_norm": 0.9604306817054749,
"learning_rate": 9.586333202475579e-05,
"loss": 1.4432080268859864,
"step": 795
},
{
"epoch": 0.1578376245437506,
"grad_norm": 0.6653774380683899,
"learning_rate": 9.579946834728861e-05,
"loss": 1.5257307052612306,
"step": 800
},
{
"epoch": 0.15882410969714905,
"grad_norm": 0.595376193523407,
"learning_rate": 9.573513706813075e-05,
"loss": 1.455667495727539,
"step": 805
},
{
"epoch": 0.1598105948505475,
"grad_norm": 0.5112119317054749,
"learning_rate": 9.567033884408927e-05,
"loss": 1.4324947357177735,
"step": 810
},
{
"epoch": 0.16079708000394594,
"grad_norm": 0.6024064421653748,
"learning_rate": 9.560507433673866e-05,
"loss": 1.4875964164733886,
"step": 815
},
{
"epoch": 0.16178356515734438,
"grad_norm": 0.7176112532615662,
"learning_rate": 9.553934421241403e-05,
"loss": 1.441457748413086,
"step": 820
},
{
"epoch": 0.16277005031074282,
"grad_norm": 0.5914203524589539,
"learning_rate": 9.547314914220432e-05,
"loss": 1.443356418609619,
"step": 825
},
{
"epoch": 0.16375653546414126,
"grad_norm": 0.6007061004638672,
"learning_rate": 9.540648980194549e-05,
"loss": 1.482406234741211,
"step": 830
},
{
"epoch": 0.1647430206175397,
"grad_norm": 0.6062957644462585,
"learning_rate": 9.533936687221355e-05,
"loss": 1.373542594909668,
"step": 835
},
{
"epoch": 0.16572950577093815,
"grad_norm": 0.645159900188446,
"learning_rate": 9.527178103831767e-05,
"loss": 1.4086808204650878,
"step": 840
},
{
"epoch": 0.1667159909243366,
"grad_norm": 0.5422107577323914,
"learning_rate": 9.520373299029315e-05,
"loss": 1.3926645278930665,
"step": 845
},
{
"epoch": 0.16770247607773503,
"grad_norm": 0.49786874651908875,
"learning_rate": 9.513522342289443e-05,
"loss": 1.463076400756836,
"step": 850
},
{
"epoch": 0.16868896123113347,
"grad_norm": 0.5858026146888733,
"learning_rate": 9.506625303558786e-05,
"loss": 1.4894671440124512,
"step": 855
},
{
"epoch": 0.1696754463845319,
"grad_norm": 0.511616587638855,
"learning_rate": 9.499682253254474e-05,
"loss": 1.4372885704040528,
"step": 860
},
{
"epoch": 0.17066193153793036,
"grad_norm": 0.675003170967102,
"learning_rate": 9.492693262263402e-05,
"loss": 1.3997676849365235,
"step": 865
},
{
"epoch": 0.1716484166913288,
"grad_norm": 0.5352548956871033,
"learning_rate": 9.485658401941506e-05,
"loss": 1.4354936599731445,
"step": 870
},
{
"epoch": 0.17263490184472724,
"grad_norm": 0.6783424615859985,
"learning_rate": 9.47857774411304e-05,
"loss": 1.412605094909668,
"step": 875
},
{
"epoch": 0.17362138699812568,
"grad_norm": 0.730926513671875,
"learning_rate": 9.471451361069834e-05,
"loss": 1.4564047813415528,
"step": 880
},
{
"epoch": 0.17460787215152412,
"grad_norm": 0.6598925590515137,
"learning_rate": 9.46427932557057e-05,
"loss": 1.4466002464294434,
"step": 885
},
{
"epoch": 0.17559435730492257,
"grad_norm": 0.5154841542243958,
"learning_rate": 9.457061710840027e-05,
"loss": 1.462276554107666,
"step": 890
},
{
"epoch": 0.176580842458321,
"grad_norm": 0.5678347945213318,
"learning_rate": 9.449798590568335e-05,
"loss": 1.4446035385131837,
"step": 895
},
{
"epoch": 0.17756732761171945,
"grad_norm": 0.7407668828964233,
"learning_rate": 9.442490038910223e-05,
"loss": 1.429649829864502,
"step": 900
},
{
"epoch": 0.1785538127651179,
"grad_norm": 0.5583857297897339,
"learning_rate": 9.435136130484274e-05,
"loss": 1.4259547233581542,
"step": 905
},
{
"epoch": 0.17954029791851633,
"grad_norm": 0.5931476354598999,
"learning_rate": 9.42773694037214e-05,
"loss": 1.4426918983459474,
"step": 910
},
{
"epoch": 0.18052678307191478,
"grad_norm": 0.5862674713134766,
"learning_rate": 9.420292544117797e-05,
"loss": 1.4132636070251465,
"step": 915
},
{
"epoch": 0.18151326822531322,
"grad_norm": 0.6072817444801331,
"learning_rate": 9.412803017726762e-05,
"loss": 1.3843505859375,
"step": 920
},
{
"epoch": 0.18249975337871166,
"grad_norm": 0.7021059393882751,
"learning_rate": 9.405268437665318e-05,
"loss": 1.4372998237609864,
"step": 925
},
{
"epoch": 0.1834862385321101,
"grad_norm": 0.767218828201294,
"learning_rate": 9.397688880859741e-05,
"loss": 1.436506462097168,
"step": 930
},
{
"epoch": 0.18447272368550854,
"grad_norm": 0.5974704623222351,
"learning_rate": 9.390064424695501e-05,
"loss": 1.4508882522583009,
"step": 935
},
{
"epoch": 0.18545920883890699,
"grad_norm": 0.9095055460929871,
"learning_rate": 9.382395147016488e-05,
"loss": 1.5006205558776855,
"step": 940
},
{
"epoch": 0.18644569399230543,
"grad_norm": 0.5792879462242126,
"learning_rate": 9.3746811261242e-05,
"loss": 1.4897793769836425,
"step": 945
},
{
"epoch": 0.18743217914570387,
"grad_norm": 0.5870247483253479,
"learning_rate": 9.366922440776964e-05,
"loss": 1.466838550567627,
"step": 950
},
{
"epoch": 0.1884186642991023,
"grad_norm": 0.5905681848526001,
"learning_rate": 9.35911917018911e-05,
"loss": 1.3940668106079102,
"step": 955
},
{
"epoch": 0.18940514945250073,
"grad_norm": 0.5917418599128723,
"learning_rate": 9.351271394030177e-05,
"loss": 1.4312256813049316,
"step": 960
},
{
"epoch": 0.19039163460589917,
"grad_norm": 0.6832874417304993,
"learning_rate": 9.343379192424096e-05,
"loss": 1.410904312133789,
"step": 965
},
{
"epoch": 0.1913781197592976,
"grad_norm": 0.5834468603134155,
"learning_rate": 9.33544264594837e-05,
"loss": 1.4238100051879883,
"step": 970
},
{
"epoch": 0.19236460491269605,
"grad_norm": 0.6182828545570374,
"learning_rate": 9.327461835633252e-05,
"loss": 1.4452640533447265,
"step": 975
},
{
"epoch": 0.1933510900660945,
"grad_norm": 0.7788094878196716,
"learning_rate": 9.31943684296092e-05,
"loss": 1.3418388366699219,
"step": 980
},
{
"epoch": 0.19433757521949294,
"grad_norm": 0.7311539649963379,
"learning_rate": 9.311367749864645e-05,
"loss": 1.395836639404297,
"step": 985
},
{
"epoch": 0.19532406037289138,
"grad_norm": 0.5241447687149048,
"learning_rate": 9.303254638727947e-05,
"loss": 1.3089460372924804,
"step": 990
},
{
"epoch": 0.19631054552628982,
"grad_norm": 0.6210646033287048,
"learning_rate": 9.295097592383763e-05,
"loss": 1.3933066368103026,
"step": 995
},
{
"epoch": 0.19729703067968826,
"grad_norm": 0.5961835980415344,
"learning_rate": 9.286896694113601e-05,
"loss": 1.4005526542663573,
"step": 1000
},
{
"epoch": 0.1982835158330867,
"grad_norm": 0.5508049726486206,
"learning_rate": 9.278652027646682e-05,
"loss": 1.40365629196167,
"step": 1005
},
{
"epoch": 0.19927000098648515,
"grad_norm": 0.599801242351532,
"learning_rate": 9.270363677159092e-05,
"loss": 1.5600028038024902,
"step": 1010
},
{
"epoch": 0.2002564861398836,
"grad_norm": 0.869111955165863,
"learning_rate": 9.262031727272921e-05,
"loss": 1.3317771911621095,
"step": 1015
},
{
"epoch": 0.20124297129328203,
"grad_norm": 0.5815553665161133,
"learning_rate": 9.253656263055399e-05,
"loss": 1.4065570831298828,
"step": 1020
},
{
"epoch": 0.20222945644668047,
"grad_norm": 0.5509616732597351,
"learning_rate": 9.245237370018023e-05,
"loss": 1.4275922775268555,
"step": 1025
},
{
"epoch": 0.20321594160007891,
"grad_norm": 0.5519242286682129,
"learning_rate": 9.236775134115691e-05,
"loss": 1.3943086624145509,
"step": 1030
},
{
"epoch": 0.20420242675347736,
"grad_norm": 0.6026679873466492,
"learning_rate": 9.228269641745824e-05,
"loss": 1.3939952850341797,
"step": 1035
},
{
"epoch": 0.2051889119068758,
"grad_norm": 0.7139347195625305,
"learning_rate": 9.219720979747476e-05,
"loss": 1.370023536682129,
"step": 1040
},
{
"epoch": 0.20617539706027424,
"grad_norm": 0.6579439640045166,
"learning_rate": 9.211129235400456e-05,
"loss": 1.4684707641601562,
"step": 1045
},
{
"epoch": 0.20716188221367268,
"grad_norm": 0.6863147020339966,
"learning_rate": 9.20249449642443e-05,
"loss": 1.3825822830200196,
"step": 1050
},
{
"epoch": 0.20814836736707112,
"grad_norm": 0.5673623085021973,
"learning_rate": 9.193816850978035e-05,
"loss": 1.3995849609375,
"step": 1055
},
{
"epoch": 0.20913485252046957,
"grad_norm": 0.5348008275032043,
"learning_rate": 9.185096387657968e-05,
"loss": 1.4474531173706056,
"step": 1060
},
{
"epoch": 0.210121337673868,
"grad_norm": 0.6630058884620667,
"learning_rate": 9.176333195498086e-05,
"loss": 1.4350825309753419,
"step": 1065
},
{
"epoch": 0.21110782282726645,
"grad_norm": 0.46311473846435547,
"learning_rate": 9.167527363968501e-05,
"loss": 1.3441024780273438,
"step": 1070
},
{
"epoch": 0.2120943079806649,
"grad_norm": 0.6591059565544128,
"learning_rate": 9.15867898297466e-05,
"loss": 1.489658260345459,
"step": 1075
},
{
"epoch": 0.21308079313406333,
"grad_norm": 0.5710258483886719,
"learning_rate": 9.149788142856433e-05,
"loss": 1.4401814460754394,
"step": 1080
},
{
"epoch": 0.21406727828746178,
"grad_norm": 0.4761538803577423,
"learning_rate": 9.140854934387181e-05,
"loss": 1.3942955017089844,
"step": 1085
},
{
"epoch": 0.21505376344086022,
"grad_norm": 0.6157023906707764,
"learning_rate": 9.131879448772842e-05,
"loss": 1.3198326110839844,
"step": 1090
},
{
"epoch": 0.21604024859425866,
"grad_norm": 0.5419490933418274,
"learning_rate": 9.122861777650994e-05,
"loss": 1.4623618125915527,
"step": 1095
},
{
"epoch": 0.2170267337476571,
"grad_norm": 0.5312824249267578,
"learning_rate": 9.113802013089914e-05,
"loss": 1.4284454345703126,
"step": 1100
},
{
"epoch": 0.21801321890105554,
"grad_norm": 0.575139045715332,
"learning_rate": 9.104700247587648e-05,
"loss": 1.3328134536743164,
"step": 1105
},
{
"epoch": 0.218999704054454,
"grad_norm": 0.6362415552139282,
"learning_rate": 9.095556574071056e-05,
"loss": 1.445528793334961,
"step": 1110
},
{
"epoch": 0.21998618920785243,
"grad_norm": 0.5265237092971802,
"learning_rate": 9.086371085894873e-05,
"loss": 1.3516952514648437,
"step": 1115
},
{
"epoch": 0.22097267436125087,
"grad_norm": 0.5764641761779785,
"learning_rate": 9.077143876840754e-05,
"loss": 1.4117063522338866,
"step": 1120
},
{
"epoch": 0.2219591595146493,
"grad_norm": 0.5066831707954407,
"learning_rate": 9.067875041116309e-05,
"loss": 1.450549030303955,
"step": 1125
},
{
"epoch": 0.22294564466804775,
"grad_norm": 0.4971221387386322,
"learning_rate": 9.058564673354149e-05,
"loss": 1.4364774703979493,
"step": 1130
},
{
"epoch": 0.2239321298214462,
"grad_norm": 0.7196635603904724,
"learning_rate": 9.049212868610917e-05,
"loss": 1.3751482963562012,
"step": 1135
},
{
"epoch": 0.22491861497484464,
"grad_norm": 0.820759654045105,
"learning_rate": 9.039819722366318e-05,
"loss": 1.4462835311889648,
"step": 1140
},
{
"epoch": 0.22590510012824308,
"grad_norm": 0.5343965888023376,
"learning_rate": 9.030385330522142e-05,
"loss": 1.469766902923584,
"step": 1145
},
{
"epoch": 0.22689158528164152,
"grad_norm": 0.5673815608024597,
"learning_rate": 9.02090978940129e-05,
"loss": 1.4521542549133302,
"step": 1150
},
{
"epoch": 0.22787807043503996,
"grad_norm": 0.6005237102508545,
"learning_rate": 9.011393195746785e-05,
"loss": 1.4021113395690918,
"step": 1155
},
{
"epoch": 0.2288645555884384,
"grad_norm": 0.5934346914291382,
"learning_rate": 9.001835646720786e-05,
"loss": 1.3267120361328124,
"step": 1160
},
{
"epoch": 0.22985104074183685,
"grad_norm": 1.021802544593811,
"learning_rate": 8.9922372399036e-05,
"loss": 1.3490143775939942,
"step": 1165
},
{
"epoch": 0.2308375258952353,
"grad_norm": 0.49048376083374023,
"learning_rate": 8.982598073292678e-05,
"loss": 1.3833210945129395,
"step": 1170
},
{
"epoch": 0.2318240110486337,
"grad_norm": 0.6739670634269714,
"learning_rate": 8.972918245301623e-05,
"loss": 1.3661418914794923,
"step": 1175
},
{
"epoch": 0.23281049620203215,
"grad_norm": 0.7055224776268005,
"learning_rate": 8.963197854759178e-05,
"loss": 1.3576847076416017,
"step": 1180
},
{
"epoch": 0.2337969813554306,
"grad_norm": 0.6569607257843018,
"learning_rate": 8.953437000908221e-05,
"loss": 1.409212017059326,
"step": 1185
},
{
"epoch": 0.23478346650882903,
"grad_norm": 0.536245584487915,
"learning_rate": 8.943635783404751e-05,
"loss": 1.3718043327331544,
"step": 1190
},
{
"epoch": 0.23576995166222747,
"grad_norm": 0.5637156367301941,
"learning_rate": 8.933794302316871e-05,
"loss": 1.3906036376953126,
"step": 1195
},
{
"epoch": 0.23675643681562591,
"grad_norm": 0.5972879528999329,
"learning_rate": 8.923912658123763e-05,
"loss": 1.4119252204895019,
"step": 1200
},
{
"epoch": 0.23774292196902436,
"grad_norm": 0.6240541934967041,
"learning_rate": 8.913990951714668e-05,
"loss": 1.3591681480407716,
"step": 1205
},
{
"epoch": 0.2387294071224228,
"grad_norm": 0.46316784620285034,
"learning_rate": 8.90402928438785e-05,
"loss": 1.418950080871582,
"step": 1210
},
{
"epoch": 0.23971589227582124,
"grad_norm": 0.6697015762329102,
"learning_rate": 8.894027757849567e-05,
"loss": 1.3783536911010743,
"step": 1215
},
{
"epoch": 0.24070237742921968,
"grad_norm": 0.49618998169898987,
"learning_rate": 8.883986474213029e-05,
"loss": 1.4549713134765625,
"step": 1220
},
{
"epoch": 0.24168886258261812,
"grad_norm": 0.6125264763832092,
"learning_rate": 8.873905535997353e-05,
"loss": 1.3716602325439453,
"step": 1225
},
{
"epoch": 0.24267534773601657,
"grad_norm": 0.6195933818817139,
"learning_rate": 8.863785046126527e-05,
"loss": 1.4679327011108398,
"step": 1230
},
{
"epoch": 0.243661832889415,
"grad_norm": 0.772383451461792,
"learning_rate": 8.853625107928346e-05,
"loss": 1.3695890426635742,
"step": 1235
},
{
"epoch": 0.24464831804281345,
"grad_norm": 0.680544912815094,
"learning_rate": 8.843425825133366e-05,
"loss": 1.415188980102539,
"step": 1240
},
{
"epoch": 0.2456348031962119,
"grad_norm": 0.5220586061477661,
"learning_rate": 8.833187301873842e-05,
"loss": 1.3541433334350585,
"step": 1245
},
{
"epoch": 0.24662128834961033,
"grad_norm": 0.5799451470375061,
"learning_rate": 8.822909642682662e-05,
"loss": 1.271614170074463,
"step": 1250
},
{
"epoch": 0.24760777350300878,
"grad_norm": 0.7626891136169434,
"learning_rate": 8.812592952492285e-05,
"loss": 1.441151237487793,
"step": 1255
},
{
"epoch": 0.24859425865640722,
"grad_norm": 0.5133401155471802,
"learning_rate": 8.80223733663367e-05,
"loss": 1.4119951248168945,
"step": 1260
},
{
"epoch": 0.24958074380980566,
"grad_norm": 0.626316249370575,
"learning_rate": 8.791842900835188e-05,
"loss": 1.4289576530456543,
"step": 1265
},
{
"epoch": 0.2505672289632041,
"grad_norm": 0.6249300241470337,
"learning_rate": 8.781409751221565e-05,
"loss": 1.3933980941772461,
"step": 1270
},
{
"epoch": 0.25155371411660254,
"grad_norm": 0.8304049968719482,
"learning_rate": 8.770937994312778e-05,
"loss": 1.4216447830200196,
"step": 1275
},
{
"epoch": 0.252540199270001,
"grad_norm": 0.7824209332466125,
"learning_rate": 8.760427737022977e-05,
"loss": 1.4057207107543945,
"step": 1280
},
{
"epoch": 0.25352668442339943,
"grad_norm": 0.5530431866645813,
"learning_rate": 8.749879086659392e-05,
"loss": 1.4289255142211914,
"step": 1285
},
{
"epoch": 0.25451316957679787,
"grad_norm": 0.5862359404563904,
"learning_rate": 8.739292150921235e-05,
"loss": 1.3567360877990722,
"step": 1290
},
{
"epoch": 0.2554996547301963,
"grad_norm": 0.5090101361274719,
"learning_rate": 8.728667037898608e-05,
"loss": 1.406256866455078,
"step": 1295
},
{
"epoch": 0.25648613988359475,
"grad_norm": 0.47887495160102844,
"learning_rate": 8.718003856071388e-05,
"loss": 1.4069179534912108,
"step": 1300
},
{
"epoch": 0.2574726250369932,
"grad_norm": 0.45626330375671387,
"learning_rate": 8.707302714308129e-05,
"loss": 1.4302420616149902,
"step": 1305
},
{
"epoch": 0.25845911019039164,
"grad_norm": 0.673858642578125,
"learning_rate": 8.696563721864947e-05,
"loss": 1.3959161758422851,
"step": 1310
},
{
"epoch": 0.2594455953437901,
"grad_norm": 0.6544889211654663,
"learning_rate": 8.685786988384403e-05,
"loss": 1.3681744575500487,
"step": 1315
},
{
"epoch": 0.2604320804971885,
"grad_norm": 0.5582412481307983,
"learning_rate": 8.674972623894384e-05,
"loss": 1.3111011505126953,
"step": 1320
},
{
"epoch": 0.26141856565058696,
"grad_norm": 0.6188735365867615,
"learning_rate": 8.664120738806985e-05,
"loss": 1.3318187713623046,
"step": 1325
},
{
"epoch": 0.2624050508039854,
"grad_norm": 0.5725945830345154,
"learning_rate": 8.653231443917372e-05,
"loss": 1.4363153457641602,
"step": 1330
},
{
"epoch": 0.26339153595738385,
"grad_norm": 0.5722491145133972,
"learning_rate": 8.642304850402656e-05,
"loss": 1.3282783508300782,
"step": 1335
},
{
"epoch": 0.2643780211107823,
"grad_norm": 0.6460428833961487,
"learning_rate": 8.631341069820769e-05,
"loss": 1.3972877502441405,
"step": 1340
},
{
"epoch": 0.26536450626418073,
"grad_norm": 0.9819481372833252,
"learning_rate": 8.620340214109298e-05,
"loss": 1.330644702911377,
"step": 1345
},
{
"epoch": 0.2663509914175792,
"grad_norm": 0.5137161016464233,
"learning_rate": 8.609302395584371e-05,
"loss": 1.4094857215881347,
"step": 1350
},
{
"epoch": 0.2673374765709776,
"grad_norm": 0.5577951073646545,
"learning_rate": 8.59822772693949e-05,
"loss": 1.3833609580993653,
"step": 1355
},
{
"epoch": 0.26832396172437606,
"grad_norm": 0.49975770711898804,
"learning_rate": 8.587116321244393e-05,
"loss": 1.2978023529052733,
"step": 1360
},
{
"epoch": 0.2693104468777745,
"grad_norm": 0.5902668237686157,
"learning_rate": 8.57596829194389e-05,
"loss": 1.3106930732727051,
"step": 1365
},
{
"epoch": 0.27029693203117294,
"grad_norm": 0.5143404006958008,
"learning_rate": 8.564783752856713e-05,
"loss": 1.3758807182312012,
"step": 1370
},
{
"epoch": 0.2712834171845714,
"grad_norm": 0.726798951625824,
"learning_rate": 8.55356281817435e-05,
"loss": 1.3343554496765138,
"step": 1375
},
{
"epoch": 0.2722699023379698,
"grad_norm": 0.7748175859451294,
"learning_rate": 8.542305602459875e-05,
"loss": 1.3325206756591796,
"step": 1380
},
{
"epoch": 0.27325638749136827,
"grad_norm": 0.790576696395874,
"learning_rate": 8.531012220646788e-05,
"loss": 1.2693490982055664,
"step": 1385
},
{
"epoch": 0.2742428726447667,
"grad_norm": 0.6062843203544617,
"learning_rate": 8.519682788037835e-05,
"loss": 1.3619890213012695,
"step": 1390
},
{
"epoch": 0.27522935779816515,
"grad_norm": 0.610841691493988,
"learning_rate": 8.508317420303829e-05,
"loss": 1.376974868774414,
"step": 1395
},
{
"epoch": 0.2762158429515636,
"grad_norm": 0.6185135245323181,
"learning_rate": 8.496916233482476e-05,
"loss": 1.4190552711486817,
"step": 1400
},
{
"epoch": 0.27720232810496204,
"grad_norm": 0.5076117515563965,
"learning_rate": 8.485479343977184e-05,
"loss": 1.3702287673950195,
"step": 1405
},
{
"epoch": 0.2781888132583605,
"grad_norm": 0.7308570742607117,
"learning_rate": 8.474006868555878e-05,
"loss": 1.3855958938598634,
"step": 1410
},
{
"epoch": 0.2791752984117589,
"grad_norm": 0.7988835573196411,
"learning_rate": 8.462498924349807e-05,
"loss": 1.4211103439331054,
"step": 1415
},
{
"epoch": 0.28016178356515736,
"grad_norm": 0.5518186688423157,
"learning_rate": 8.450955628852345e-05,
"loss": 1.351612663269043,
"step": 1420
},
{
"epoch": 0.2811482687185558,
"grad_norm": 0.5241552591323853,
"learning_rate": 8.439377099917798e-05,
"loss": 1.3563777923583984,
"step": 1425
},
{
"epoch": 0.28213475387195425,
"grad_norm": 0.6125285029411316,
"learning_rate": 8.427763455760199e-05,
"loss": 1.3562525749206542,
"step": 1430
},
{
"epoch": 0.2831212390253527,
"grad_norm": 0.5519683361053467,
"learning_rate": 8.41611481495209e-05,
"loss": 1.3564645767211914,
"step": 1435
},
{
"epoch": 0.28410772417875113,
"grad_norm": 0.5959796905517578,
"learning_rate": 8.404431296423327e-05,
"loss": 1.3283239364624024,
"step": 1440
},
{
"epoch": 0.2850942093321496,
"grad_norm": 0.5448134541511536,
"learning_rate": 8.392713019459862e-05,
"loss": 1.3758052825927733,
"step": 1445
},
{
"epoch": 0.286080694485548,
"grad_norm": 0.8359006643295288,
"learning_rate": 8.380960103702516e-05,
"loss": 1.3918051719665527,
"step": 1450
},
{
"epoch": 0.28706717963894646,
"grad_norm": 0.7521634101867676,
"learning_rate": 8.369172669145766e-05,
"loss": 1.3350682258605957,
"step": 1455
},
{
"epoch": 0.2880536647923449,
"grad_norm": 0.6283717751502991,
"learning_rate": 8.357350836136522e-05,
"loss": 1.326828670501709,
"step": 1460
},
{
"epoch": 0.28904014994574334,
"grad_norm": 0.6766531467437744,
"learning_rate": 8.345494725372886e-05,
"loss": 1.3961981773376464,
"step": 1465
},
{
"epoch": 0.2900266350991418,
"grad_norm": 0.6832543611526489,
"learning_rate": 8.333604457902935e-05,
"loss": 1.4674219131469726,
"step": 1470
},
{
"epoch": 0.2910131202525402,
"grad_norm": 0.5512411594390869,
"learning_rate": 8.321680155123473e-05,
"loss": 1.3835647583007813,
"step": 1475
},
{
"epoch": 0.29199960540593867,
"grad_norm": 0.5400985479354858,
"learning_rate": 8.3097219387788e-05,
"loss": 1.4086020469665528,
"step": 1480
},
{
"epoch": 0.2929860905593371,
"grad_norm": 0.5161585211753845,
"learning_rate": 8.297729930959462e-05,
"loss": 1.2685392379760743,
"step": 1485
},
{
"epoch": 0.29397257571273555,
"grad_norm": 0.6080145239830017,
"learning_rate": 8.285704254101013e-05,
"loss": 1.3606173515319824,
"step": 1490
},
{
"epoch": 0.29495906086613394,
"grad_norm": 0.5382046699523926,
"learning_rate": 8.273645030982753e-05,
"loss": 1.4041426658630372,
"step": 1495
},
{
"epoch": 0.2959455460195324,
"grad_norm": 0.5880080461502075,
"learning_rate": 8.261552384726486e-05,
"loss": 1.3820586204528809,
"step": 1500
},
{
"epoch": 0.2969320311729308,
"grad_norm": 0.6372694373130798,
"learning_rate": 8.24942643879526e-05,
"loss": 1.3317127227783203,
"step": 1505
},
{
"epoch": 0.29791851632632926,
"grad_norm": 0.5431744456291199,
"learning_rate": 8.237267316992097e-05,
"loss": 1.3452259063720704,
"step": 1510
},
{
"epoch": 0.2989050014797277,
"grad_norm": 0.627873420715332,
"learning_rate": 8.225075143458747e-05,
"loss": 1.3627433776855469,
"step": 1515
},
{
"epoch": 0.29989148663312615,
"grad_norm": 0.6476534605026245,
"learning_rate": 8.212850042674401e-05,
"loss": 1.3387353897094727,
"step": 1520
},
{
"epoch": 0.3008779717865246,
"grad_norm": 0.6051098108291626,
"learning_rate": 8.200592139454437e-05,
"loss": 1.3802270889282227,
"step": 1525
},
{
"epoch": 0.30186445693992303,
"grad_norm": 0.7490224242210388,
"learning_rate": 8.188301558949132e-05,
"loss": 1.394545078277588,
"step": 1530
},
{
"epoch": 0.3028509420933215,
"grad_norm": 0.6829474568367004,
"learning_rate": 8.175978426642394e-05,
"loss": 1.3322031021118164,
"step": 1535
},
{
"epoch": 0.3038374272467199,
"grad_norm": 0.6685187220573425,
"learning_rate": 8.163622868350475e-05,
"loss": 1.3866594314575196,
"step": 1540
},
{
"epoch": 0.30482391240011836,
"grad_norm": 0.5692480206489563,
"learning_rate": 8.151235010220692e-05,
"loss": 1.380581760406494,
"step": 1545
},
{
"epoch": 0.3058103975535168,
"grad_norm": 0.5384566187858582,
"learning_rate": 8.138814978730132e-05,
"loss": 1.317976188659668,
"step": 1550
},
{
"epoch": 0.30679688270691524,
"grad_norm": 0.5023966431617737,
"learning_rate": 8.126362900684365e-05,
"loss": 1.43321533203125,
"step": 1555
},
{
"epoch": 0.3077833678603137,
"grad_norm": 0.5382370352745056,
"learning_rate": 8.113878903216152e-05,
"loss": 1.3258728981018066,
"step": 1560
},
{
"epoch": 0.3087698530137121,
"grad_norm": 0.6715255975723267,
"learning_rate": 8.101363113784139e-05,
"loss": 1.4534453392028808,
"step": 1565
},
{
"epoch": 0.30975633816711057,
"grad_norm": 0.6423724889755249,
"learning_rate": 8.08881566017156e-05,
"loss": 1.3444867134094238,
"step": 1570
},
{
"epoch": 0.310742823320509,
"grad_norm": 0.5984092950820923,
"learning_rate": 8.076236670484941e-05,
"loss": 1.3855459213256835,
"step": 1575
},
{
"epoch": 0.31172930847390745,
"grad_norm": 0.5952145457267761,
"learning_rate": 8.063626273152776e-05,
"loss": 1.421107292175293,
"step": 1580
},
{
"epoch": 0.3127157936273059,
"grad_norm": 0.5436971783638,
"learning_rate": 8.050984596924223e-05,
"loss": 1.3396149635314942,
"step": 1585
},
{
"epoch": 0.31370227878070434,
"grad_norm": 0.7394317984580994,
"learning_rate": 8.038311770867798e-05,
"loss": 1.375946044921875,
"step": 1590
},
{
"epoch": 0.3146887639341028,
"grad_norm": 0.6647710204124451,
"learning_rate": 8.025607924370037e-05,
"loss": 1.286642837524414,
"step": 1595
},
{
"epoch": 0.3156752490875012,
"grad_norm": 0.5024760961532593,
"learning_rate": 8.012873187134202e-05,
"loss": 1.2993341445922852,
"step": 1600
},
{
"epoch": 0.31666173424089966,
"grad_norm": 0.8035495281219482,
"learning_rate": 8.00010768917893e-05,
"loss": 1.3648031234741211,
"step": 1605
},
{
"epoch": 0.3176482193942981,
"grad_norm": 0.5939044952392578,
"learning_rate": 7.987311560836926e-05,
"loss": 1.3844295501708985,
"step": 1610
},
{
"epoch": 0.31863470454769655,
"grad_norm": 0.6447381973266602,
"learning_rate": 7.974484932753616e-05,
"loss": 1.3692025184631347,
"step": 1615
},
{
"epoch": 0.319621189701095,
"grad_norm": 0.6331056356430054,
"learning_rate": 7.961627935885828e-05,
"loss": 1.4030118942260743,
"step": 1620
},
{
"epoch": 0.32060767485449343,
"grad_norm": 0.7672907710075378,
"learning_rate": 7.948740701500444e-05,
"loss": 1.349720573425293,
"step": 1625
},
{
"epoch": 0.32159416000789187,
"grad_norm": 1.0480629205703735,
"learning_rate": 7.935823361173068e-05,
"loss": 1.35137996673584,
"step": 1630
},
{
"epoch": 0.3225806451612903,
"grad_norm": 0.5524029731750488,
"learning_rate": 7.922876046786672e-05,
"loss": 1.3474660873413087,
"step": 1635
},
{
"epoch": 0.32356713031468876,
"grad_norm": 0.8406732082366943,
"learning_rate": 7.909898890530261e-05,
"loss": 1.3530692100524901,
"step": 1640
},
{
"epoch": 0.3245536154680872,
"grad_norm": 0.6200135946273804,
"learning_rate": 7.89689202489752e-05,
"loss": 1.388188648223877,
"step": 1645
},
{
"epoch": 0.32554010062148564,
"grad_norm": 0.49200841784477234,
"learning_rate": 7.883855582685453e-05,
"loss": 1.2923940658569335,
"step": 1650
},
{
"epoch": 0.3265265857748841,
"grad_norm": 0.6728061437606812,
"learning_rate": 7.87078969699304e-05,
"loss": 1.3266336441040039,
"step": 1655
},
{
"epoch": 0.3275130709282825,
"grad_norm": 0.5868862271308899,
"learning_rate": 7.85769450121987e-05,
"loss": 1.3768096923828126,
"step": 1660
},
{
"epoch": 0.32849955608168097,
"grad_norm": 0.6203965544700623,
"learning_rate": 7.844570129064778e-05,
"loss": 1.3372143745422362,
"step": 1665
},
{
"epoch": 0.3294860412350794,
"grad_norm": 0.6243035793304443,
"learning_rate": 7.831416714524489e-05,
"loss": 1.351456356048584,
"step": 1670
},
{
"epoch": 0.33047252638847785,
"grad_norm": 0.52872633934021,
"learning_rate": 7.818234391892241e-05,
"loss": 1.3590909957885742,
"step": 1675
},
{
"epoch": 0.3314590115418763,
"grad_norm": 0.5473071932792664,
"learning_rate": 7.805023295756411e-05,
"loss": 1.2850920677185058,
"step": 1680
},
{
"epoch": 0.33244549669527473,
"grad_norm": 0.5982444882392883,
"learning_rate": 7.791783560999157e-05,
"loss": 1.3571319580078125,
"step": 1685
},
{
"epoch": 0.3334319818486732,
"grad_norm": 0.6122526526451111,
"learning_rate": 7.778515322795021e-05,
"loss": 1.4054675102233887,
"step": 1690
},
{
"epoch": 0.3344184670020716,
"grad_norm": 1.0291657447814941,
"learning_rate": 7.765218716609567e-05,
"loss": 1.2710465431213378,
"step": 1695
},
{
"epoch": 0.33540495215547006,
"grad_norm": 0.6338679194450378,
"learning_rate": 7.751893878197983e-05,
"loss": 1.3642587661743164,
"step": 1700
},
{
"epoch": 0.3363914373088685,
"grad_norm": 0.5614007115364075,
"learning_rate": 7.738540943603704e-05,
"loss": 1.3497204780578613,
"step": 1705
},
{
"epoch": 0.33737792246226694,
"grad_norm": 0.6538517475128174,
"learning_rate": 7.725160049157019e-05,
"loss": 1.3064023017883302,
"step": 1710
},
{
"epoch": 0.3383644076156654,
"grad_norm": 0.5769523978233337,
"learning_rate": 7.71175133147368e-05,
"loss": 1.368824291229248,
"step": 1715
},
{
"epoch": 0.3393508927690638,
"grad_norm": 0.5274118185043335,
"learning_rate": 7.698314927453512e-05,
"loss": 1.3131978988647461,
"step": 1720
},
{
"epoch": 0.34033737792246227,
"grad_norm": 0.5269600749015808,
"learning_rate": 7.684850974279007e-05,
"loss": 1.3798871994018556,
"step": 1725
},
{
"epoch": 0.3413238630758607,
"grad_norm": 0.588138997554779,
"learning_rate": 7.671359609413927e-05,
"loss": 1.294365692138672,
"step": 1730
},
{
"epoch": 0.34231034822925915,
"grad_norm": 0.6350558996200562,
"learning_rate": 7.657840970601901e-05,
"loss": 1.3693756103515624,
"step": 1735
},
{
"epoch": 0.3432968333826576,
"grad_norm": 0.5530595779418945,
"learning_rate": 7.644295195865024e-05,
"loss": 1.33154239654541,
"step": 1740
},
{
"epoch": 0.34428331853605604,
"grad_norm": 0.590160071849823,
"learning_rate": 7.630722423502432e-05,
"loss": 1.3611008644104003,
"step": 1745
},
{
"epoch": 0.3452698036894545,
"grad_norm": 0.5361330509185791,
"learning_rate": 7.617122792088908e-05,
"loss": 1.2934256553649903,
"step": 1750
},
{
"epoch": 0.3462562888428529,
"grad_norm": 0.6077249646186829,
"learning_rate": 7.603496440473462e-05,
"loss": 1.3284368515014648,
"step": 1755
},
{
"epoch": 0.34724277399625136,
"grad_norm": 0.662836492061615,
"learning_rate": 7.589843507777903e-05,
"loss": 1.3701604843139648,
"step": 1760
},
{
"epoch": 0.3482292591496498,
"grad_norm": 0.5313825011253357,
"learning_rate": 7.576164133395427e-05,
"loss": 1.3600330352783203,
"step": 1765
},
{
"epoch": 0.34921574430304825,
"grad_norm": 0.6143139600753784,
"learning_rate": 7.562458456989202e-05,
"loss": 1.3262263298034669,
"step": 1770
},
{
"epoch": 0.3502022294564467,
"grad_norm": 0.6019529700279236,
"learning_rate": 7.548726618490926e-05,
"loss": 1.3658750534057618,
"step": 1775
},
{
"epoch": 0.35118871460984513,
"grad_norm": 0.49150845408439636,
"learning_rate": 7.534968758099408e-05,
"loss": 1.2767610549926758,
"step": 1780
},
{
"epoch": 0.3521751997632436,
"grad_norm": 0.7461856603622437,
"learning_rate": 7.52118501627913e-05,
"loss": 1.2895452499389648,
"step": 1785
},
{
"epoch": 0.353161684916642,
"grad_norm": 0.5603261590003967,
"learning_rate": 7.507375533758827e-05,
"loss": 1.3417886734008788,
"step": 1790
},
{
"epoch": 0.35414817007004046,
"grad_norm": 0.6377080678939819,
"learning_rate": 7.493540451530033e-05,
"loss": 1.3576799392700196,
"step": 1795
},
{
"epoch": 0.3551346552234389,
"grad_norm": 0.6406121850013733,
"learning_rate": 7.479679910845647e-05,
"loss": 1.4028952598571778,
"step": 1800
},
{
"epoch": 0.35612114037683734,
"grad_norm": 0.5557231903076172,
"learning_rate": 7.465794053218501e-05,
"loss": 1.3010631561279298,
"step": 1805
},
{
"epoch": 0.3571076255302358,
"grad_norm": 0.5916813015937805,
"learning_rate": 7.4518830204199e-05,
"loss": 1.3707881927490235,
"step": 1810
},
{
"epoch": 0.3580941106836342,
"grad_norm": 0.6194364428520203,
"learning_rate": 7.437946954478184e-05,
"loss": 1.416722297668457,
"step": 1815
},
{
"epoch": 0.35908059583703267,
"grad_norm": 0.7916405200958252,
"learning_rate": 7.423985997677277e-05,
"loss": 1.3532354354858398,
"step": 1820
},
{
"epoch": 0.3600670809904311,
"grad_norm": 0.6148778796195984,
"learning_rate": 7.410000292555228e-05,
"loss": 1.3100946426391602,
"step": 1825
},
{
"epoch": 0.36105356614382955,
"grad_norm": 0.4952429234981537,
"learning_rate": 7.395989981902765e-05,
"loss": 1.247380828857422,
"step": 1830
},
{
"epoch": 0.362040051297228,
"grad_norm": 0.7478722929954529,
"learning_rate": 7.381955208761833e-05,
"loss": 1.2795291900634767,
"step": 1835
},
{
"epoch": 0.36302653645062644,
"grad_norm": 0.6232354044914246,
"learning_rate": 7.367896116424129e-05,
"loss": 1.3837846755981444,
"step": 1840
},
{
"epoch": 0.3640130216040249,
"grad_norm": 0.7537280917167664,
"learning_rate": 7.353812848429647e-05,
"loss": 1.2414480209350587,
"step": 1845
},
{
"epoch": 0.3649995067574233,
"grad_norm": 0.6165774464607239,
"learning_rate": 7.339705548565206e-05,
"loss": 1.3609989166259766,
"step": 1850
},
{
"epoch": 0.36598599191082176,
"grad_norm": 0.6720853447914124,
"learning_rate": 7.325574360862986e-05,
"loss": 1.268237590789795,
"step": 1855
},
{
"epoch": 0.3669724770642202,
"grad_norm": 0.5806273818016052,
"learning_rate": 7.311419429599057e-05,
"loss": 1.3741284370422364,
"step": 1860
},
{
"epoch": 0.36795896221761865,
"grad_norm": 0.698898434638977,
"learning_rate": 7.297240899291903e-05,
"loss": 1.3633082389831543,
"step": 1865
},
{
"epoch": 0.3689454473710171,
"grad_norm": 0.7180983424186707,
"learning_rate": 7.283038914700953e-05,
"loss": 1.2845206260681152,
"step": 1870
},
{
"epoch": 0.36993193252441553,
"grad_norm": 0.6578208804130554,
"learning_rate": 7.268813620825093e-05,
"loss": 1.2857697486877442,
"step": 1875
},
{
"epoch": 0.37091841767781397,
"grad_norm": 0.6658138632774353,
"learning_rate": 7.254565162901193e-05,
"loss": 1.3151821136474608,
"step": 1880
},
{
"epoch": 0.3719049028312124,
"grad_norm": 0.9386398792266846,
"learning_rate": 7.240293686402623e-05,
"loss": 1.316750431060791,
"step": 1885
},
{
"epoch": 0.37289138798461086,
"grad_norm": 0.5925989151000977,
"learning_rate": 7.225999337037771e-05,
"loss": 1.3902604103088378,
"step": 1890
},
{
"epoch": 0.3738778731380093,
"grad_norm": 0.5934818387031555,
"learning_rate": 7.211682260748546e-05,
"loss": 1.329233741760254,
"step": 1895
},
{
"epoch": 0.37486435829140774,
"grad_norm": 0.6553406715393066,
"learning_rate": 7.197342603708895e-05,
"loss": 1.2523588180541991,
"step": 1900
},
{
"epoch": 0.3758508434448062,
"grad_norm": 0.5263217687606812,
"learning_rate": 7.182980512323308e-05,
"loss": 1.294236183166504,
"step": 1905
},
{
"epoch": 0.3768373285982046,
"grad_norm": 0.576176106929779,
"learning_rate": 7.168596133225328e-05,
"loss": 1.386712646484375,
"step": 1910
},
{
"epoch": 0.37782381375160307,
"grad_norm": 0.6409969925880432,
"learning_rate": 7.154189613276048e-05,
"loss": 1.3919392585754395,
"step": 1915
},
{
"epoch": 0.37881029890500145,
"grad_norm": 0.5837907791137695,
"learning_rate": 7.139761099562616e-05,
"loss": 1.3163067817687988,
"step": 1920
},
{
"epoch": 0.3797967840583999,
"grad_norm": 0.6095336675643921,
"learning_rate": 7.125310739396723e-05,
"loss": 1.238929557800293,
"step": 1925
},
{
"epoch": 0.38078326921179834,
"grad_norm": 0.6900033354759216,
"learning_rate": 7.110838680313119e-05,
"loss": 1.3376704216003419,
"step": 1930
},
{
"epoch": 0.3817697543651968,
"grad_norm": 0.5560447573661804,
"learning_rate": 7.096345070068086e-05,
"loss": 1.2411578178405762,
"step": 1935
},
{
"epoch": 0.3827562395185952,
"grad_norm": 0.8668294548988342,
"learning_rate": 7.081830056637941e-05,
"loss": 1.2523038864135743,
"step": 1940
},
{
"epoch": 0.38374272467199366,
"grad_norm": 0.6804683208465576,
"learning_rate": 7.067293788217523e-05,
"loss": 1.2759183883666991,
"step": 1945
},
{
"epoch": 0.3847292098253921,
"grad_norm": 0.6282105445861816,
"learning_rate": 7.052736413218683e-05,
"loss": 1.3934737205505372,
"step": 1950
},
{
"epoch": 0.38571569497879055,
"grad_norm": 0.4964868724346161,
"learning_rate": 7.038158080268755e-05,
"loss": 1.3289667129516602,
"step": 1955
},
{
"epoch": 0.386702180132189,
"grad_norm": 0.9534741640090942,
"learning_rate": 7.023558938209058e-05,
"loss": 1.329804039001465,
"step": 1960
},
{
"epoch": 0.38768866528558743,
"grad_norm": 0.5716214179992676,
"learning_rate": 7.008939136093364e-05,
"loss": 1.3150169372558593,
"step": 1965
},
{
"epoch": 0.3886751504389859,
"grad_norm": 0.6827837228775024,
"learning_rate": 6.994298823186379e-05,
"loss": 1.340550994873047,
"step": 1970
},
{
"epoch": 0.3896616355923843,
"grad_norm": 0.6290521621704102,
"learning_rate": 6.979638148962218e-05,
"loss": 1.3575560569763183,
"step": 1975
},
{
"epoch": 0.39064812074578276,
"grad_norm": 0.7522562742233276,
"learning_rate": 6.964957263102883e-05,
"loss": 1.3603950500488282,
"step": 1980
},
{
"epoch": 0.3916346058991812,
"grad_norm": 0.6337642669677734,
"learning_rate": 6.95025631549673e-05,
"loss": 1.2885261535644532,
"step": 1985
},
{
"epoch": 0.39262109105257964,
"grad_norm": 0.5845955014228821,
"learning_rate": 6.935535456236946e-05,
"loss": 1.344526481628418,
"step": 1990
},
{
"epoch": 0.3936075762059781,
"grad_norm": 0.6726008057594299,
"learning_rate": 6.92079483562e-05,
"loss": 1.3362887382507325,
"step": 1995
},
{
"epoch": 0.3945940613593765,
"grad_norm": 0.5784204006195068,
"learning_rate": 6.906034604144128e-05,
"loss": 1.298848533630371,
"step": 2000
},
{
"epoch": 0.39558054651277497,
"grad_norm": 0.6113646626472473,
"learning_rate": 6.891254912507787e-05,
"loss": 1.315582847595215,
"step": 2005
},
{
"epoch": 0.3965670316661734,
"grad_norm": 0.5523650050163269,
"learning_rate": 6.87645591160812e-05,
"loss": 1.2081064224243163,
"step": 2010
},
{
"epoch": 0.39755351681957185,
"grad_norm": 0.6356167793273926,
"learning_rate": 6.861637752539407e-05,
"loss": 1.3052804946899415,
"step": 2015
},
{
"epoch": 0.3985400019729703,
"grad_norm": 0.5855145454406738,
"learning_rate": 6.846800586591533e-05,
"loss": 1.2717106819152832,
"step": 2020
},
{
"epoch": 0.39952648712636873,
"grad_norm": 0.6065287590026855,
"learning_rate": 6.831944565248436e-05,
"loss": 1.2940223693847657,
"step": 2025
},
{
"epoch": 0.4005129722797672,
"grad_norm": 0.8592427968978882,
"learning_rate": 6.817069840186566e-05,
"loss": 1.2965956687927247,
"step": 2030
},
{
"epoch": 0.4014994574331656,
"grad_norm": 0.6346101760864258,
"learning_rate": 6.802176563273331e-05,
"loss": 1.3930267333984374,
"step": 2035
},
{
"epoch": 0.40248594258656406,
"grad_norm": 0.5965458154678345,
"learning_rate": 6.78726488656555e-05,
"loss": 1.2906299591064454,
"step": 2040
},
{
"epoch": 0.4034724277399625,
"grad_norm": 0.730190634727478,
"learning_rate": 6.7723349623079e-05,
"loss": 1.442465591430664,
"step": 2045
},
{
"epoch": 0.40445891289336094,
"grad_norm": 0.6532384753227234,
"learning_rate": 6.75738694293136e-05,
"loss": 1.3926507949829101,
"step": 2050
},
{
"epoch": 0.4054453980467594,
"grad_norm": 0.6533120274543762,
"learning_rate": 6.742420981051656e-05,
"loss": 1.2911331176757812,
"step": 2055
},
{
"epoch": 0.40643188320015783,
"grad_norm": 0.6141111850738525,
"learning_rate": 6.727437229467704e-05,
"loss": 1.3386173248291016,
"step": 2060
},
{
"epoch": 0.40741836835355627,
"grad_norm": 0.7462441921234131,
"learning_rate": 6.712435841160048e-05,
"loss": 1.2048350334167481,
"step": 2065
},
{
"epoch": 0.4084048535069547,
"grad_norm": 0.631230354309082,
"learning_rate": 6.697416969289297e-05,
"loss": 1.2942930221557618,
"step": 2070
},
{
"epoch": 0.40939133866035315,
"grad_norm": 0.6422381401062012,
"learning_rate": 6.682380767194568e-05,
"loss": 1.4227178573608399,
"step": 2075
},
{
"epoch": 0.4103778238137516,
"grad_norm": 0.527824342250824,
"learning_rate": 6.667327388391906e-05,
"loss": 1.345162582397461,
"step": 2080
},
{
"epoch": 0.41136430896715004,
"grad_norm": 0.6163994073867798,
"learning_rate": 6.652256986572739e-05,
"loss": 1.3222397804260253,
"step": 2085
},
{
"epoch": 0.4123507941205485,
"grad_norm": 0.8280588984489441,
"learning_rate": 6.637169715602284e-05,
"loss": 1.2958997726440429,
"step": 2090
},
{
"epoch": 0.4133372792739469,
"grad_norm": 0.5802834630012512,
"learning_rate": 6.622065729517995e-05,
"loss": 1.399118137359619,
"step": 2095
},
{
"epoch": 0.41432376442734536,
"grad_norm": 0.8017279505729675,
"learning_rate": 6.606945182527983e-05,
"loss": 1.2858858108520508,
"step": 2100
},
{
"epoch": 0.4153102495807438,
"grad_norm": 0.6299742460250854,
"learning_rate": 6.591808229009439e-05,
"loss": 1.3155744552612305,
"step": 2105
},
{
"epoch": 0.41629673473414225,
"grad_norm": 0.6828439235687256,
"learning_rate": 6.576655023507062e-05,
"loss": 1.3274757385253906,
"step": 2110
},
{
"epoch": 0.4172832198875407,
"grad_norm": 0.698814868927002,
"learning_rate": 6.56148572073148e-05,
"loss": 1.2712442398071289,
"step": 2115
},
{
"epoch": 0.41826970504093913,
"grad_norm": 0.6453902721405029,
"learning_rate": 6.546300475557669e-05,
"loss": 1.3289005279541015,
"step": 2120
},
{
"epoch": 0.4192561901943376,
"grad_norm": 0.8425061702728271,
"learning_rate": 6.531099443023378e-05,
"loss": 1.3160091400146485,
"step": 2125
},
{
"epoch": 0.420242675347736,
"grad_norm": 0.6055358648300171,
"learning_rate": 6.515882778327536e-05,
"loss": 1.2763602256774902,
"step": 2130
},
{
"epoch": 0.42122916050113446,
"grad_norm": 0.5148352980613708,
"learning_rate": 6.500650636828673e-05,
"loss": 1.3284460067749024,
"step": 2135
},
{
"epoch": 0.4222156456545329,
"grad_norm": 0.6500594019889832,
"learning_rate": 6.485403174043336e-05,
"loss": 1.3336772918701172,
"step": 2140
},
{
"epoch": 0.42320213080793134,
"grad_norm": 0.47751089930534363,
"learning_rate": 6.470140545644498e-05,
"loss": 1.262111759185791,
"step": 2145
},
{
"epoch": 0.4241886159613298,
"grad_norm": 0.6590535640716553,
"learning_rate": 6.454862907459967e-05,
"loss": 1.3530981063842773,
"step": 2150
},
{
"epoch": 0.4251751011147282,
"grad_norm": 0.6921405792236328,
"learning_rate": 6.439570415470801e-05,
"loss": 1.3042779922485352,
"step": 2155
},
{
"epoch": 0.42616158626812667,
"grad_norm": 0.6713026165962219,
"learning_rate": 6.424263225809709e-05,
"loss": 1.3571932792663575,
"step": 2160
},
{
"epoch": 0.4271480714215251,
"grad_norm": 0.6041602492332458,
"learning_rate": 6.408941494759461e-05,
"loss": 1.316493606567383,
"step": 2165
},
{
"epoch": 0.42813455657492355,
"grad_norm": 0.5474830269813538,
"learning_rate": 6.393605378751293e-05,
"loss": 1.337926483154297,
"step": 2170
},
{
"epoch": 0.429121041728322,
"grad_norm": 0.6670401096343994,
"learning_rate": 6.378255034363302e-05,
"loss": 1.3425260543823243,
"step": 2175
},
{
"epoch": 0.43010752688172044,
"grad_norm": 0.6240463852882385,
"learning_rate": 6.362890618318858e-05,
"loss": 1.2913217544555664,
"step": 2180
},
{
"epoch": 0.4310940120351189,
"grad_norm": 0.7299990653991699,
"learning_rate": 6.347512287485001e-05,
"loss": 1.4104684829711913,
"step": 2185
},
{
"epoch": 0.4320804971885173,
"grad_norm": 0.5494000315666199,
"learning_rate": 6.332120198870835e-05,
"loss": 1.3143083572387695,
"step": 2190
},
{
"epoch": 0.43306698234191576,
"grad_norm": 0.5461328625679016,
"learning_rate": 6.316714509625924e-05,
"loss": 1.2770273208618164,
"step": 2195
},
{
"epoch": 0.4340534674953142,
"grad_norm": 0.8294140696525574,
"learning_rate": 6.301295377038698e-05,
"loss": 1.3323987007141114,
"step": 2200
},
{
"epoch": 0.43503995264871265,
"grad_norm": 0.5591831207275391,
"learning_rate": 6.28586295853484e-05,
"loss": 1.2820966720581055,
"step": 2205
},
{
"epoch": 0.4360264378021111,
"grad_norm": 0.7715386152267456,
"learning_rate": 6.270417411675671e-05,
"loss": 1.3211211204528808,
"step": 2210
},
{
"epoch": 0.43701292295550953,
"grad_norm": 0.6770358085632324,
"learning_rate": 6.254958894156557e-05,
"loss": 1.2514583587646484,
"step": 2215
},
{
"epoch": 0.437999408108908,
"grad_norm": 0.7189682126045227,
"learning_rate": 6.239487563805292e-05,
"loss": 1.2929292678833009,
"step": 2220
},
{
"epoch": 0.4389858932623064,
"grad_norm": 0.7726011872291565,
"learning_rate": 6.224003578580478e-05,
"loss": 1.3258827209472657,
"step": 2225
},
{
"epoch": 0.43997237841570486,
"grad_norm": 0.5027691125869751,
"learning_rate": 6.20850709656993e-05,
"loss": 1.299870491027832,
"step": 2230
},
{
"epoch": 0.4409588635691033,
"grad_norm": 0.5709672570228577,
"learning_rate": 6.192998275989043e-05,
"loss": 1.274820899963379,
"step": 2235
},
{
"epoch": 0.44194534872250174,
"grad_norm": 0.5759121775627136,
"learning_rate": 6.177477275179198e-05,
"loss": 1.3079337120056151,
"step": 2240
},
{
"epoch": 0.4429318338759002,
"grad_norm": 0.49961501359939575,
"learning_rate": 6.161944252606117e-05,
"loss": 1.288625144958496,
"step": 2245
},
{
"epoch": 0.4439183190292986,
"grad_norm": 0.6045638918876648,
"learning_rate": 6.146399366858276e-05,
"loss": 1.2426865577697754,
"step": 2250
},
{
"epoch": 0.44490480418269707,
"grad_norm": 0.8719197511672974,
"learning_rate": 6.130842776645263e-05,
"loss": 1.3069183349609375,
"step": 2255
},
{
"epoch": 0.4458912893360955,
"grad_norm": 0.622288167476654,
"learning_rate": 6.11527464079617e-05,
"loss": 1.300102996826172,
"step": 2260
},
{
"epoch": 0.44687777448949395,
"grad_norm": 0.6024958491325378,
"learning_rate": 6.099695118257962e-05,
"loss": 1.3699260711669923,
"step": 2265
},
{
"epoch": 0.4478642596428924,
"grad_norm": 0.6537339687347412,
"learning_rate": 6.0841043680938656e-05,
"loss": 1.262373924255371,
"step": 2270
},
{
"epoch": 0.44885074479629083,
"grad_norm": 0.5247714519500732,
"learning_rate": 6.0685025494817336e-05,
"loss": 1.3550918579101563,
"step": 2275
},
{
"epoch": 0.4498372299496893,
"grad_norm": 0.54839026927948,
"learning_rate": 6.0528898217124296e-05,
"loss": 1.334242057800293,
"step": 2280
},
{
"epoch": 0.4508237151030877,
"grad_norm": 0.5588928461074829,
"learning_rate": 6.037266344188194e-05,
"loss": 1.2948463439941407,
"step": 2285
},
{
"epoch": 0.45181020025648616,
"grad_norm": 0.5581854581832886,
"learning_rate": 6.0216322764210196e-05,
"loss": 1.2794687271118164,
"step": 2290
},
{
"epoch": 0.4527966854098846,
"grad_norm": 0.9220561981201172,
"learning_rate": 6.005987778031026e-05,
"loss": 1.3728206634521485,
"step": 2295
},
{
"epoch": 0.45378317056328304,
"grad_norm": 0.6274521946907043,
"learning_rate": 5.9903330087448225e-05,
"loss": 1.3434026718139649,
"step": 2300
},
{
"epoch": 0.4547696557166815,
"grad_norm": 0.6987320780754089,
"learning_rate": 5.9746681283938844e-05,
"loss": 1.3444314002990723,
"step": 2305
},
{
"epoch": 0.45575614087007993,
"grad_norm": 0.5950861573219299,
"learning_rate": 5.95899329691292e-05,
"loss": 1.318413543701172,
"step": 2310
},
{
"epoch": 0.45674262602347837,
"grad_norm": 0.8476980924606323,
"learning_rate": 5.9433086743382346e-05,
"loss": 1.2979576110839843,
"step": 2315
},
{
"epoch": 0.4577291111768768,
"grad_norm": 0.6287381649017334,
"learning_rate": 5.927614420806098e-05,
"loss": 1.3213831901550293,
"step": 2320
},
{
"epoch": 0.45871559633027525,
"grad_norm": 0.6041512489318848,
"learning_rate": 5.911910696551108e-05,
"loss": 1.3663217544555664,
"step": 2325
},
{
"epoch": 0.4597020814836737,
"grad_norm": 0.586740255355835,
"learning_rate": 5.89619766190456e-05,
"loss": 1.2764649391174316,
"step": 2330
},
{
"epoch": 0.46068856663707214,
"grad_norm": 0.6066878437995911,
"learning_rate": 5.880475477292805e-05,
"loss": 1.260331630706787,
"step": 2335
},
{
"epoch": 0.4616750517904706,
"grad_norm": 0.6039069890975952,
"learning_rate": 5.864744303235613e-05,
"loss": 1.3089725494384765,
"step": 2340
},
{
"epoch": 0.46266153694386897,
"grad_norm": 0.5299080014228821,
"learning_rate": 5.849004300344532e-05,
"loss": 1.327096176147461,
"step": 2345
},
{
"epoch": 0.4636480220972674,
"grad_norm": 0.5433689951896667,
"learning_rate": 5.833255629321253e-05,
"loss": 1.3444944381713868,
"step": 2350
},
{
"epoch": 0.46463450725066585,
"grad_norm": 0.8155282735824585,
"learning_rate": 5.8174984509559646e-05,
"loss": 1.332880973815918,
"step": 2355
},
{
"epoch": 0.4656209924040643,
"grad_norm": 0.5263012051582336,
"learning_rate": 5.801732926125716e-05,
"loss": 1.3378890991210937,
"step": 2360
},
{
"epoch": 0.46660747755746274,
"grad_norm": 0.6341373324394226,
"learning_rate": 5.785959215792769e-05,
"loss": 1.282596969604492,
"step": 2365
},
{
"epoch": 0.4675939627108612,
"grad_norm": 0.5949289202690125,
"learning_rate": 5.770177481002959e-05,
"loss": 1.3460208892822265,
"step": 2370
},
{
"epoch": 0.4685804478642596,
"grad_norm": 0.5771950483322144,
"learning_rate": 5.754387882884047e-05,
"loss": 1.2595620155334473,
"step": 2375
},
{
"epoch": 0.46956693301765806,
"grad_norm": 0.5534161329269409,
"learning_rate": 5.738590582644081e-05,
"loss": 1.2509942054748535,
"step": 2380
},
{
"epoch": 0.4705534181710565,
"grad_norm": 0.6139821410179138,
"learning_rate": 5.722785741569743e-05,
"loss": 1.2658857345581054,
"step": 2385
},
{
"epoch": 0.47153990332445495,
"grad_norm": 0.6152280569076538,
"learning_rate": 5.706973521024703e-05,
"loss": 1.2428781509399414,
"step": 2390
},
{
"epoch": 0.4725263884778534,
"grad_norm": 0.7122021317481995,
"learning_rate": 5.6911540824479795e-05,
"loss": 1.2491264343261719,
"step": 2395
},
{
"epoch": 0.47351287363125183,
"grad_norm": 0.5463324189186096,
"learning_rate": 5.6753275873522774e-05,
"loss": 1.3475863456726074,
"step": 2400
},
{
"epoch": 0.47449935878465027,
"grad_norm": 0.6332739591598511,
"learning_rate": 5.659494197322356e-05,
"loss": 1.2418846130371093,
"step": 2405
},
{
"epoch": 0.4754858439380487,
"grad_norm": 0.5899940729141235,
"learning_rate": 5.643654074013364e-05,
"loss": 1.2868310928344726,
"step": 2410
},
{
"epoch": 0.47647232909144716,
"grad_norm": 0.8350638151168823,
"learning_rate": 5.627807379149198e-05,
"loss": 1.2695045471191406,
"step": 2415
},
{
"epoch": 0.4774588142448456,
"grad_norm": 0.7122925519943237,
"learning_rate": 5.611954274520845e-05,
"loss": 1.156500244140625,
"step": 2420
},
{
"epoch": 0.47844529939824404,
"grad_norm": 0.8544057607650757,
"learning_rate": 5.596094921984738e-05,
"loss": 1.3231846809387207,
"step": 2425
},
{
"epoch": 0.4794317845516425,
"grad_norm": 0.5936956405639648,
"learning_rate": 5.5802294834611016e-05,
"loss": 1.3284900665283204,
"step": 2430
},
{
"epoch": 0.4804182697050409,
"grad_norm": 0.6106973886489868,
"learning_rate": 5.564358120932291e-05,
"loss": 1.2668222427368163,
"step": 2435
},
{
"epoch": 0.48140475485843937,
"grad_norm": 0.5120734572410583,
"learning_rate": 5.548480996441145e-05,
"loss": 1.2936574935913085,
"step": 2440
},
{
"epoch": 0.4823912400118378,
"grad_norm": 0.5015867352485657,
"learning_rate": 5.5325982720893366e-05,
"loss": 1.3047141075134276,
"step": 2445
},
{
"epoch": 0.48337772516523625,
"grad_norm": 0.6193450689315796,
"learning_rate": 5.516710110035703e-05,
"loss": 1.3068492889404297,
"step": 2450
},
{
"epoch": 0.4843642103186347,
"grad_norm": 0.7502596974372864,
"learning_rate": 5.500816672494611e-05,
"loss": 1.2675558090209962,
"step": 2455
},
{
"epoch": 0.48535069547203313,
"grad_norm": 0.6633110642433167,
"learning_rate": 5.484918121734276e-05,
"loss": 1.2609601974487306,
"step": 2460
},
{
"epoch": 0.4863371806254316,
"grad_norm": 0.6235228776931763,
"learning_rate": 5.4690146200751266e-05,
"loss": 1.3100595474243164,
"step": 2465
},
{
"epoch": 0.48732366577883,
"grad_norm": 0.5731393098831177,
"learning_rate": 5.453106329888136e-05,
"loss": 1.2778688430786134,
"step": 2470
},
{
"epoch": 0.48831015093222846,
"grad_norm": 0.8861362934112549,
"learning_rate": 5.437193413593167e-05,
"loss": 1.3638214111328124,
"step": 2475
},
{
"epoch": 0.4892966360856269,
"grad_norm": 0.7066556811332703,
"learning_rate": 5.421276033657313e-05,
"loss": 1.3259263038635254,
"step": 2480
},
{
"epoch": 0.49028312123902534,
"grad_norm": 0.6080769300460815,
"learning_rate": 5.405354352593245e-05,
"loss": 1.3311138153076172,
"step": 2485
},
{
"epoch": 0.4912696063924238,
"grad_norm": 0.5582625865936279,
"learning_rate": 5.389428532957541e-05,
"loss": 1.3339624404907227,
"step": 2490
},
{
"epoch": 0.4922560915458222,
"grad_norm": 0.722672700881958,
"learning_rate": 5.373498737349037e-05,
"loss": 1.2038657188415527,
"step": 2495
},
{
"epoch": 0.49324257669922067,
"grad_norm": 0.8127685189247131,
"learning_rate": 5.357565128407162e-05,
"loss": 1.3207258224487304,
"step": 2500
},
{
"epoch": 0.4942290618526191,
"grad_norm": 0.5719538927078247,
"learning_rate": 5.341627868810276e-05,
"loss": 1.360446834564209,
"step": 2505
},
{
"epoch": 0.49521554700601755,
"grad_norm": 0.622006356716156,
"learning_rate": 5.325687121274017e-05,
"loss": 1.2719401359558105,
"step": 2510
},
{
"epoch": 0.496202032159416,
"grad_norm": 0.7372507452964783,
"learning_rate": 5.3097430485496225e-05,
"loss": 1.2535398483276368,
"step": 2515
},
{
"epoch": 0.49718851731281444,
"grad_norm": 0.8012752532958984,
"learning_rate": 5.293795813422294e-05,
"loss": 1.252589225769043,
"step": 2520
},
{
"epoch": 0.4981750024662129,
"grad_norm": 0.8236263990402222,
"learning_rate": 5.2778455787095114e-05,
"loss": 1.253206729888916,
"step": 2525
},
{
"epoch": 0.4991614876196113,
"grad_norm": 0.7806019186973572,
"learning_rate": 5.261892507259383e-05,
"loss": 1.3268577575683593,
"step": 2530
},
{
"epoch": 0.5001479727730098,
"grad_norm": 0.6204727292060852,
"learning_rate": 5.2459367619489754e-05,
"loss": 1.3216068267822265,
"step": 2535
},
{
"epoch": 0.5011344579264082,
"grad_norm": 0.7056735754013062,
"learning_rate": 5.2299785056826555e-05,
"loss": 1.3040446281433105,
"step": 2540
},
{
"epoch": 0.5021209430798067,
"grad_norm": 0.8224450349807739,
"learning_rate": 5.2140179013904346e-05,
"loss": 1.263269805908203,
"step": 2545
},
{
"epoch": 0.5031074282332051,
"grad_norm": 0.5905019640922546,
"learning_rate": 5.198055112026288e-05,
"loss": 1.2911508560180665,
"step": 2550
},
{
"epoch": 0.5040939133866036,
"grad_norm": 0.9770287871360779,
"learning_rate": 5.1820903005665e-05,
"loss": 1.2736207008361817,
"step": 2555
},
{
"epoch": 0.505080398540002,
"grad_norm": 0.6376775503158569,
"learning_rate": 5.1661236300080054e-05,
"loss": 1.2824560165405274,
"step": 2560
},
{
"epoch": 0.5060668836934004,
"grad_norm": 0.7155457139015198,
"learning_rate": 5.150155263366715e-05,
"loss": 1.2564644813537598,
"step": 2565
},
{
"epoch": 0.5070533688467989,
"grad_norm": 0.5818341970443726,
"learning_rate": 5.134185363675862e-05,
"loss": 1.3089483261108399,
"step": 2570
},
{
"epoch": 0.5080398540001972,
"grad_norm": 0.4942404627799988,
"learning_rate": 5.118214093984325e-05,
"loss": 1.3135147094726562,
"step": 2575
},
{
"epoch": 0.5090263391535957,
"grad_norm": 0.5972030758857727,
"learning_rate": 5.102241617354974e-05,
"loss": 1.2330535888671874,
"step": 2580
},
{
"epoch": 0.5100128243069941,
"grad_norm": 0.8606343865394592,
"learning_rate": 5.086268096863e-05,
"loss": 1.2743648529052733,
"step": 2585
},
{
"epoch": 0.5109993094603926,
"grad_norm": 0.619056224822998,
"learning_rate": 5.0702936955942537e-05,
"loss": 1.2597833633422852,
"step": 2590
},
{
"epoch": 0.511985794613791,
"grad_norm": 0.637209951877594,
"learning_rate": 5.054318576643575e-05,
"loss": 1.3092596054077148,
"step": 2595
},
{
"epoch": 0.5129722797671895,
"grad_norm": 0.7981250286102295,
"learning_rate": 5.038342903113134e-05,
"loss": 1.2517135620117188,
"step": 2600
},
{
"epoch": 0.5139587649205879,
"grad_norm": 0.6200886368751526,
"learning_rate": 5.0223668381107625e-05,
"loss": 1.2324432373046874,
"step": 2605
},
{
"epoch": 0.5149452500739864,
"grad_norm": 0.7681851387023926,
"learning_rate": 5.006390544748287e-05,
"loss": 1.3301679611206054,
"step": 2610
},
{
"epoch": 0.5159317352273848,
"grad_norm": 0.7167775630950928,
"learning_rate": 4.990414186139868e-05,
"loss": 1.306093120574951,
"step": 2615
},
{
"epoch": 0.5169182203807833,
"grad_norm": 0.6668635010719299,
"learning_rate": 4.9744379254003334e-05,
"loss": 1.373420810699463,
"step": 2620
},
{
"epoch": 0.5179047055341817,
"grad_norm": 0.6026375889778137,
"learning_rate": 4.958461925643506e-05,
"loss": 1.3586023330688477,
"step": 2625
},
{
"epoch": 0.5188911906875802,
"grad_norm": 0.6150393486022949,
"learning_rate": 4.9424863499805486e-05,
"loss": 1.4242569923400878,
"step": 2630
},
{
"epoch": 0.5198776758409785,
"grad_norm": 0.6556458473205566,
"learning_rate": 4.9265113615182926e-05,
"loss": 1.2936998367309571,
"step": 2635
},
{
"epoch": 0.520864160994377,
"grad_norm": 0.8551809191703796,
"learning_rate": 4.910537123357577e-05,
"loss": 1.2494179725646972,
"step": 2640
},
{
"epoch": 0.5218506461477754,
"grad_norm": 0.540070652961731,
"learning_rate": 4.894563798591577e-05,
"loss": 1.30374755859375,
"step": 2645
},
{
"epoch": 0.5228371313011739,
"grad_norm": 0.523993730545044,
"learning_rate": 4.878591550304143e-05,
"loss": 1.2896368026733398,
"step": 2650
},
{
"epoch": 0.5238236164545723,
"grad_norm": 0.7759273052215576,
"learning_rate": 4.8626205415681384e-05,
"loss": 1.2786067008972168,
"step": 2655
},
{
"epoch": 0.5248101016079708,
"grad_norm": 0.5434350967407227,
"learning_rate": 4.846650935443764e-05,
"loss": 1.2534055709838867,
"step": 2660
},
{
"epoch": 0.5257965867613692,
"grad_norm": 0.7286601066589355,
"learning_rate": 4.830682894976905e-05,
"loss": 1.319723892211914,
"step": 2665
},
{
"epoch": 0.5267830719147677,
"grad_norm": 0.7513492703437805,
"learning_rate": 4.814716583197462e-05,
"loss": 1.2972120285034179,
"step": 2670
},
{
"epoch": 0.5277695570681661,
"grad_norm": 0.9861288070678711,
"learning_rate": 4.7987521631176845e-05,
"loss": 1.2634204864501952,
"step": 2675
},
{
"epoch": 0.5287560422215646,
"grad_norm": 0.7686546444892883,
"learning_rate": 4.7827897977305085e-05,
"loss": 1.297569751739502,
"step": 2680
},
{
"epoch": 0.529742527374963,
"grad_norm": 0.651456356048584,
"learning_rate": 4.7668296500078916e-05,
"loss": 1.2089714050292968,
"step": 2685
},
{
"epoch": 0.5307290125283615,
"grad_norm": 0.5578058362007141,
"learning_rate": 4.750871882899152e-05,
"loss": 1.269384002685547,
"step": 2690
},
{
"epoch": 0.5317154976817599,
"grad_norm": 0.6947818994522095,
"learning_rate": 4.734916659329302e-05,
"loss": 1.2427091598510742,
"step": 2695
},
{
"epoch": 0.5327019828351583,
"grad_norm": 0.5807385444641113,
"learning_rate": 4.718964142197379e-05,
"loss": 1.3292981147766114,
"step": 2700
},
{
"epoch": 0.5336884679885567,
"grad_norm": 0.6037646532058716,
"learning_rate": 4.703014494374793e-05,
"loss": 1.3633432388305664,
"step": 2705
},
{
"epoch": 0.5346749531419552,
"grad_norm": 0.5325913429260254,
"learning_rate": 4.6870678787036626e-05,
"loss": 1.3083860397338867,
"step": 2710
},
{
"epoch": 0.5356614382953536,
"grad_norm": 0.9301533699035645,
"learning_rate": 4.671124457995143e-05,
"loss": 1.2943355560302734,
"step": 2715
},
{
"epoch": 0.5366479234487521,
"grad_norm": 0.5465457439422607,
"learning_rate": 4.6551843950277734e-05,
"loss": 1.3357701301574707,
"step": 2720
},
{
"epoch": 0.5376344086021505,
"grad_norm": 0.5911902189254761,
"learning_rate": 4.639247852545808e-05,
"loss": 1.276066780090332,
"step": 2725
},
{
"epoch": 0.538620893755549,
"grad_norm": 0.9351704120635986,
"learning_rate": 4.62331499325756e-05,
"loss": 1.1888238906860351,
"step": 2730
},
{
"epoch": 0.5396073789089474,
"grad_norm": 0.714290976524353,
"learning_rate": 4.607385979833738e-05,
"loss": 1.348914337158203,
"step": 2735
},
{
"epoch": 0.5405938640623459,
"grad_norm": 0.691771388053894,
"learning_rate": 4.591460974905783e-05,
"loss": 1.2422422409057616,
"step": 2740
},
{
"epoch": 0.5415803492157443,
"grad_norm": 0.6136305928230286,
"learning_rate": 4.5755401410642124e-05,
"loss": 1.3646180152893066,
"step": 2745
},
{
"epoch": 0.5425668343691428,
"grad_norm": 0.6610255837440491,
"learning_rate": 4.559623640856957e-05,
"loss": 1.253035545349121,
"step": 2750
},
{
"epoch": 0.5435533195225412,
"grad_norm": 0.5677903294563293,
"learning_rate": 4.5437116367877023e-05,
"loss": 1.2843030929565429,
"step": 2755
},
{
"epoch": 0.5445398046759397,
"grad_norm": 0.6745204925537109,
"learning_rate": 4.527804291314231e-05,
"loss": 1.2579179763793946,
"step": 2760
},
{
"epoch": 0.545526289829338,
"grad_norm": 0.5611047148704529,
"learning_rate": 4.51190176684676e-05,
"loss": 1.2543134689331055,
"step": 2765
},
{
"epoch": 0.5465127749827365,
"grad_norm": 0.628821849822998,
"learning_rate": 4.4960042257462866e-05,
"loss": 1.282552146911621,
"step": 2770
},
{
"epoch": 0.5474992601361349,
"grad_norm": 0.6254104375839233,
"learning_rate": 4.480111830322928e-05,
"loss": 1.3153570175170899,
"step": 2775
},
{
"epoch": 0.5484857452895334,
"grad_norm": 0.7786182165145874,
"learning_rate": 4.464224742834265e-05,
"loss": 1.2361795425415039,
"step": 2780
},
{
"epoch": 0.5494722304429318,
"grad_norm": 0.6818842887878418,
"learning_rate": 4.448343125483687e-05,
"loss": 1.2627603530883789,
"step": 2785
},
{
"epoch": 0.5504587155963303,
"grad_norm": 0.8019605875015259,
"learning_rate": 4.432467140418734e-05,
"loss": 1.2016630172729492,
"step": 2790
},
{
"epoch": 0.5514452007497287,
"grad_norm": 0.6679314970970154,
"learning_rate": 4.416596949729441e-05,
"loss": 1.3287381172180175,
"step": 2795
},
{
"epoch": 0.5524316859031272,
"grad_norm": 0.534041702747345,
"learning_rate": 4.400732715446686e-05,
"loss": 1.2804701805114747,
"step": 2800
},
{
"epoch": 0.5534181710565256,
"grad_norm": 0.6065512299537659,
"learning_rate": 4.384874599540529e-05,
"loss": 1.2817874908447267,
"step": 2805
},
{
"epoch": 0.5544046562099241,
"grad_norm": 0.8300914168357849,
"learning_rate": 4.3690227639185724e-05,
"loss": 1.2281585693359376,
"step": 2810
},
{
"epoch": 0.5553911413633225,
"grad_norm": 0.7092686891555786,
"learning_rate": 4.353177370424284e-05,
"loss": 1.192383098602295,
"step": 2815
},
{
"epoch": 0.556377626516721,
"grad_norm": 0.7450085282325745,
"learning_rate": 4.3373385808353695e-05,
"loss": 1.244245719909668,
"step": 2820
},
{
"epoch": 0.5573641116701193,
"grad_norm": 0.8198437690734863,
"learning_rate": 4.321506556862109e-05,
"loss": 1.2955668449401856,
"step": 2825
},
{
"epoch": 0.5583505968235178,
"grad_norm": 0.5612547397613525,
"learning_rate": 4.305681460145705e-05,
"loss": 1.289536190032959,
"step": 2830
},
{
"epoch": 0.5593370819769162,
"grad_norm": 0.8305298686027527,
"learning_rate": 4.289863452256635e-05,
"loss": 1.1628555297851562,
"step": 2835
},
{
"epoch": 0.5603235671303147,
"grad_norm": 0.7152137756347656,
"learning_rate": 4.2740526946930014e-05,
"loss": 1.3210289001464843,
"step": 2840
},
{
"epoch": 0.5613100522837131,
"grad_norm": 0.7453590035438538,
"learning_rate": 4.2582493488788834e-05,
"loss": 1.2790961265563965,
"step": 2845
},
{
"epoch": 0.5622965374371116,
"grad_norm": 0.6487513780593872,
"learning_rate": 4.242453576162687e-05,
"loss": 1.2659735679626465,
"step": 2850
},
{
"epoch": 0.56328302259051,
"grad_norm": 0.6280403733253479,
"learning_rate": 4.226665537815498e-05,
"loss": 1.3162508010864258,
"step": 2855
},
{
"epoch": 0.5642695077439085,
"grad_norm": 0.666420578956604,
"learning_rate": 4.210885395029437e-05,
"loss": 1.3228586196899415,
"step": 2860
},
{
"epoch": 0.5652559928973069,
"grad_norm": 0.693835973739624,
"learning_rate": 4.1951133089160134e-05,
"loss": 1.270298957824707,
"step": 2865
},
{
"epoch": 0.5662424780507054,
"grad_norm": 0.7162954807281494,
"learning_rate": 4.179349440504478e-05,
"loss": 1.3217622756958007,
"step": 2870
},
{
"epoch": 0.5672289632041038,
"grad_norm": 0.5747545957565308,
"learning_rate": 4.163593950740184e-05,
"loss": 1.2470691680908204,
"step": 2875
},
{
"epoch": 0.5682154483575023,
"grad_norm": 0.6960130333900452,
"learning_rate": 4.147847000482937e-05,
"loss": 1.3104389190673829,
"step": 2880
},
{
"epoch": 0.5692019335109006,
"grad_norm": 0.6319047808647156,
"learning_rate": 4.1321087505053604e-05,
"loss": 1.3021565437316895,
"step": 2885
},
{
"epoch": 0.5701884186642991,
"grad_norm": 0.5702677369117737,
"learning_rate": 4.1163793614912446e-05,
"loss": 1.2373793601989747,
"step": 2890
},
{
"epoch": 0.5711749038176975,
"grad_norm": 0.8115822672843933,
"learning_rate": 4.100658994033916e-05,
"loss": 1.2513921737670899,
"step": 2895
},
{
"epoch": 0.572161388971096,
"grad_norm": 0.6213353872299194,
"learning_rate": 4.0849478086345904e-05,
"loss": 1.3069664001464845,
"step": 2900
},
{
"epoch": 0.5731478741244944,
"grad_norm": 0.6548802256584167,
"learning_rate": 4.0692459657007385e-05,
"loss": 1.194764995574951,
"step": 2905
},
{
"epoch": 0.5741343592778929,
"grad_norm": 0.6145977973937988,
"learning_rate": 4.053553625544446e-05,
"loss": 1.301835823059082,
"step": 2910
},
{
"epoch": 0.5751208444312913,
"grad_norm": 0.7434483170509338,
"learning_rate": 4.037870948380778e-05,
"loss": 1.280484962463379,
"step": 2915
},
{
"epoch": 0.5761073295846898,
"grad_norm": 0.8130502104759216,
"learning_rate": 4.0221980943261413e-05,
"loss": 1.154463005065918,
"step": 2920
},
{
"epoch": 0.5770938147380882,
"grad_norm": 0.6294492483139038,
"learning_rate": 4.006535223396654e-05,
"loss": 1.3500300407409669,
"step": 2925
},
{
"epoch": 0.5780802998914867,
"grad_norm": 0.567590057849884,
"learning_rate": 3.990882495506508e-05,
"loss": 1.2109825134277343,
"step": 2930
},
{
"epoch": 0.5790667850448851,
"grad_norm": 0.7338889837265015,
"learning_rate": 3.9752400704663333e-05,
"loss": 1.2079388618469238,
"step": 2935
},
{
"epoch": 0.5800532701982836,
"grad_norm": 0.7760624885559082,
"learning_rate": 3.9596081079815725e-05,
"loss": 1.2644323348999023,
"step": 2940
},
{
"epoch": 0.581039755351682,
"grad_norm": 0.6629675030708313,
"learning_rate": 3.943986767650852e-05,
"loss": 1.246302604675293,
"step": 2945
},
{
"epoch": 0.5820262405050804,
"grad_norm": 0.628167986869812,
"learning_rate": 3.928376208964344e-05,
"loss": 1.2584522247314454,
"step": 2950
},
{
"epoch": 0.5830127256584788,
"grad_norm": 0.6644889116287231,
"learning_rate": 3.912776591302144e-05,
"loss": 1.2831511497497559,
"step": 2955
},
{
"epoch": 0.5839992108118773,
"grad_norm": 0.7527005076408386,
"learning_rate": 3.897188073932642e-05,
"loss": 1.4331621170043944,
"step": 2960
},
{
"epoch": 0.5849856959652757,
"grad_norm": 0.6411287784576416,
"learning_rate": 3.881610816010897e-05,
"loss": 1.2337224960327149,
"step": 2965
},
{
"epoch": 0.5859721811186742,
"grad_norm": 0.6102463006973267,
"learning_rate": 3.8660449765770105e-05,
"loss": 1.334969425201416,
"step": 2970
},
{
"epoch": 0.5869586662720726,
"grad_norm": 0.6090612411499023,
"learning_rate": 3.850490714554505e-05,
"loss": 1.2560989379882812,
"step": 2975
},
{
"epoch": 0.5879451514254711,
"grad_norm": 0.5106144547462463,
"learning_rate": 3.8349481887487015e-05,
"loss": 1.2918489456176758,
"step": 2980
},
{
"epoch": 0.5889316365788695,
"grad_norm": 0.670214056968689,
"learning_rate": 3.8194175578450956e-05,
"loss": 1.2171284675598144,
"step": 2985
},
{
"epoch": 0.5899181217322679,
"grad_norm": 0.7328224778175354,
"learning_rate": 3.803898980407737e-05,
"loss": 1.2274720191955566,
"step": 2990
},
{
"epoch": 0.5909046068856664,
"grad_norm": 0.6471003293991089,
"learning_rate": 3.788392614877617e-05,
"loss": 1.2533182144165038,
"step": 2995
},
{
"epoch": 0.5918910920390648,
"grad_norm": 0.701018214225769,
"learning_rate": 3.772898619571043e-05,
"loss": 1.313788890838623,
"step": 3000
},
{
"epoch": 0.5928775771924633,
"grad_norm": 0.6119874715805054,
"learning_rate": 3.757417152678028e-05,
"loss": 1.2741676330566407,
"step": 3005
},
{
"epoch": 0.5938640623458616,
"grad_norm": 0.7143070101737976,
"learning_rate": 3.741948372260665e-05,
"loss": 1.2097087860107423,
"step": 3010
},
{
"epoch": 0.5948505474992601,
"grad_norm": 0.680537760257721,
"learning_rate": 3.726492436251532e-05,
"loss": 1.221632957458496,
"step": 3015
},
{
"epoch": 0.5958370326526585,
"grad_norm": 1.0982909202575684,
"learning_rate": 3.711049502452062e-05,
"loss": 1.270390510559082,
"step": 3020
},
{
"epoch": 0.596823517806057,
"grad_norm": 0.6590185761451721,
"learning_rate": 3.69561972853094e-05,
"loss": 1.2234973907470703,
"step": 3025
},
{
"epoch": 0.5978100029594554,
"grad_norm": 0.875187873840332,
"learning_rate": 3.680203272022493e-05,
"loss": 1.2828168869018555,
"step": 3030
},
{
"epoch": 0.5987964881128539,
"grad_norm": 0.5678836107254028,
"learning_rate": 3.664800290325078e-05,
"loss": 1.2891668319702148,
"step": 3035
},
{
"epoch": 0.5997829732662523,
"grad_norm": 0.5648377537727356,
"learning_rate": 3.6494109406994776e-05,
"loss": 1.272667407989502,
"step": 3040
},
{
"epoch": 0.6007694584196508,
"grad_norm": 0.7052152156829834,
"learning_rate": 3.6340353802673e-05,
"loss": 1.21891450881958,
"step": 3045
},
{
"epoch": 0.6017559435730492,
"grad_norm": 0.9625502824783325,
"learning_rate": 3.6186737660093594e-05,
"loss": 1.2393878936767577,
"step": 3050
},
{
"epoch": 0.6027424287264477,
"grad_norm": 0.6187904477119446,
"learning_rate": 3.603326254764089e-05,
"loss": 1.2499343872070312,
"step": 3055
},
{
"epoch": 0.6037289138798461,
"grad_norm": 0.5407034158706665,
"learning_rate": 3.5879930032259315e-05,
"loss": 1.2921064376831055,
"step": 3060
},
{
"epoch": 0.6047153990332446,
"grad_norm": 0.652245044708252,
"learning_rate": 3.572674167943744e-05,
"loss": 1.288829231262207,
"step": 3065
},
{
"epoch": 0.605701884186643,
"grad_norm": 0.7839142680168152,
"learning_rate": 3.5573699053191946e-05,
"loss": 1.184808349609375,
"step": 3070
},
{
"epoch": 0.6066883693400414,
"grad_norm": 0.6601583957672119,
"learning_rate": 3.542080371605169e-05,
"loss": 1.321641731262207,
"step": 3075
},
{
"epoch": 0.6076748544934398,
"grad_norm": 0.6391304731369019,
"learning_rate": 3.5268057229041745e-05,
"loss": 1.1616025924682618,
"step": 3080
},
{
"epoch": 0.6086613396468383,
"grad_norm": 0.596361517906189,
"learning_rate": 3.5115461151667435e-05,
"loss": 1.2123571395874024,
"step": 3085
},
{
"epoch": 0.6096478248002367,
"grad_norm": 0.5540140867233276,
"learning_rate": 3.496301704189846e-05,
"loss": 1.2987688064575196,
"step": 3090
},
{
"epoch": 0.6106343099536352,
"grad_norm": 0.5955787897109985,
"learning_rate": 3.481072645615298e-05,
"loss": 1.280453586578369,
"step": 3095
},
{
"epoch": 0.6116207951070336,
"grad_norm": 0.7479218244552612,
"learning_rate": 3.465859094928167e-05,
"loss": 1.3428421020507812,
"step": 3100
},
{
"epoch": 0.6126072802604321,
"grad_norm": 0.7732844948768616,
"learning_rate": 3.450661207455192e-05,
"loss": 1.2743149757385255,
"step": 3105
},
{
"epoch": 0.6135937654138305,
"grad_norm": 0.6871821880340576,
"learning_rate": 3.435479138363193e-05,
"loss": 1.2874011993408203,
"step": 3110
},
{
"epoch": 0.614580250567229,
"grad_norm": 0.8527916073799133,
"learning_rate": 3.42031304265749e-05,
"loss": 1.257390594482422,
"step": 3115
},
{
"epoch": 0.6155667357206274,
"grad_norm": 0.5993496775627136,
"learning_rate": 3.4051630751803174e-05,
"loss": 1.2911321640014648,
"step": 3120
},
{
"epoch": 0.6165532208740259,
"grad_norm": 0.6292076110839844,
"learning_rate": 3.39002939060924e-05,
"loss": 1.2540863037109375,
"step": 3125
},
{
"epoch": 0.6175397060274243,
"grad_norm": 0.6794880032539368,
"learning_rate": 3.374912143455585e-05,
"loss": 1.2051777839660645,
"step": 3130
},
{
"epoch": 0.6185261911808227,
"grad_norm": 0.6501321792602539,
"learning_rate": 3.359811488062853e-05,
"loss": 1.2653312683105469,
"step": 3135
},
{
"epoch": 0.6195126763342211,
"grad_norm": 0.823517918586731,
"learning_rate": 3.344727578605149e-05,
"loss": 1.3225866317749024,
"step": 3140
},
{
"epoch": 0.6204991614876196,
"grad_norm": 0.7574878931045532,
"learning_rate": 3.329660569085603e-05,
"loss": 1.2353111267089845,
"step": 3145
},
{
"epoch": 0.621485646641018,
"grad_norm": 0.6372307538986206,
"learning_rate": 3.3146106133348045e-05,
"loss": 1.1551859855651856,
"step": 3150
},
{
"epoch": 0.6224721317944165,
"grad_norm": 0.6747055649757385,
"learning_rate": 3.299577865009225e-05,
"loss": 1.2953763008117676,
"step": 3155
},
{
"epoch": 0.6234586169478149,
"grad_norm": 0.9490332007408142,
"learning_rate": 3.2845624775896554e-05,
"loss": 1.1598724365234374,
"step": 3160
},
{
"epoch": 0.6244451021012134,
"grad_norm": 0.9150590896606445,
"learning_rate": 3.2695646043796304e-05,
"loss": 1.316457176208496,
"step": 3165
},
{
"epoch": 0.6254315872546118,
"grad_norm": 0.9105712175369263,
"learning_rate": 3.254584398503874e-05,
"loss": 1.2646181106567382,
"step": 3170
},
{
"epoch": 0.6264180724080103,
"grad_norm": 0.7176617980003357,
"learning_rate": 3.2396220129067284e-05,
"loss": 1.273106575012207,
"step": 3175
},
{
"epoch": 0.6274045575614087,
"grad_norm": 0.6291666030883789,
"learning_rate": 3.2246776003505976e-05,
"loss": 1.219668483734131,
"step": 3180
},
{
"epoch": 0.6283910427148072,
"grad_norm": 0.6315911412239075,
"learning_rate": 3.209751313414382e-05,
"loss": 1.2776854515075684,
"step": 3185
},
{
"epoch": 0.6293775278682056,
"grad_norm": 0.6760464906692505,
"learning_rate": 3.194843304491926e-05,
"loss": 1.2513291358947753,
"step": 3190
},
{
"epoch": 0.630364013021604,
"grad_norm": 0.7457524538040161,
"learning_rate": 3.1799537257904565e-05,
"loss": 1.220335102081299,
"step": 3195
},
{
"epoch": 0.6313504981750024,
"grad_norm": 0.7505302429199219,
"learning_rate": 3.165082729329038e-05,
"loss": 1.2514754295349122,
"step": 3200
},
{
"epoch": 0.6323369833284009,
"grad_norm": 1.0568127632141113,
"learning_rate": 3.1502304669370086e-05,
"loss": 1.2489755630493165,
"step": 3205
},
{
"epoch": 0.6333234684817993,
"grad_norm": 0.8083130717277527,
"learning_rate": 3.135397090252439e-05,
"loss": 1.271303653717041,
"step": 3210
},
{
"epoch": 0.6343099536351978,
"grad_norm": 0.7235426306724548,
"learning_rate": 3.120582750720581e-05,
"loss": 1.2905401229858398,
"step": 3215
},
{
"epoch": 0.6352964387885962,
"grad_norm": 0.6699484586715698,
"learning_rate": 3.1057875995923224e-05,
"loss": 1.248713493347168,
"step": 3220
},
{
"epoch": 0.6362829239419947,
"grad_norm": 0.48013836145401,
"learning_rate": 3.09101178792264e-05,
"loss": 1.2338943481445312,
"step": 3225
},
{
"epoch": 0.6372694090953931,
"grad_norm": 0.7004823684692383,
"learning_rate": 3.0762554665690623e-05,
"loss": 1.2746992111206055,
"step": 3230
},
{
"epoch": 0.6382558942487916,
"grad_norm": 0.6915579438209534,
"learning_rate": 3.061518786190125e-05,
"loss": 1.2131980895996093,
"step": 3235
},
{
"epoch": 0.63924237940219,
"grad_norm": 0.6655653119087219,
"learning_rate": 3.0468018972438306e-05,
"loss": 1.2047163009643556,
"step": 3240
},
{
"epoch": 0.6402288645555885,
"grad_norm": 0.6790153384208679,
"learning_rate": 3.0321049499861187e-05,
"loss": 1.307823371887207,
"step": 3245
},
{
"epoch": 0.6412153497089869,
"grad_norm": 0.7771720886230469,
"learning_rate": 3.0174280944693317e-05,
"loss": 1.3104080200195312,
"step": 3250
},
{
"epoch": 0.6422018348623854,
"grad_norm": 0.8188883662223816,
"learning_rate": 3.0027714805406765e-05,
"loss": 1.1877461433410645,
"step": 3255
},
{
"epoch": 0.6431883200157837,
"grad_norm": 0.5972724556922913,
"learning_rate": 2.988135257840699e-05,
"loss": 1.2573711395263671,
"step": 3260
},
{
"epoch": 0.6441748051691822,
"grad_norm": 0.6972706317901611,
"learning_rate": 2.9735195758017553e-05,
"loss": 1.3350233078002929,
"step": 3265
},
{
"epoch": 0.6451612903225806,
"grad_norm": 0.6569691300392151,
"learning_rate": 2.9589245836464874e-05,
"loss": 1.1341728210449218,
"step": 3270
},
{
"epoch": 0.6461477754759791,
"grad_norm": 0.7379050850868225,
"learning_rate": 2.9443504303862985e-05,
"loss": 1.2505168914794922,
"step": 3275
},
{
"epoch": 0.6471342606293775,
"grad_norm": 0.8827411532402039,
"learning_rate": 2.9297972648198278e-05,
"loss": 1.3005850791931153,
"step": 3280
},
{
"epoch": 0.648120745782776,
"grad_norm": 0.5547851920127869,
"learning_rate": 2.9152652355314392e-05,
"loss": 1.2680097579956056,
"step": 3285
},
{
"epoch": 0.6491072309361744,
"grad_norm": 0.6803496479988098,
"learning_rate": 2.9007544908897022e-05,
"loss": 1.2209806442260742,
"step": 3290
},
{
"epoch": 0.6500937160895729,
"grad_norm": 0.5865746140480042,
"learning_rate": 2.8862651790458678e-05,
"loss": 1.2020627975463867,
"step": 3295
},
{
"epoch": 0.6510802012429713,
"grad_norm": 0.6626771688461304,
"learning_rate": 2.8717974479323718e-05,
"loss": 1.2622586250305177,
"step": 3300
},
{
"epoch": 0.6520666863963698,
"grad_norm": 0.701485276222229,
"learning_rate": 2.8573514452613063e-05,
"loss": 1.2559017181396483,
"step": 3305
},
{
"epoch": 0.6530531715497682,
"grad_norm": 0.6936823725700378,
"learning_rate": 2.8429273185229306e-05,
"loss": 1.3547216415405274,
"step": 3310
},
{
"epoch": 0.6540396567031667,
"grad_norm": 0.5725021362304688,
"learning_rate": 2.828525214984154e-05,
"loss": 1.291282844543457,
"step": 3315
},
{
"epoch": 0.655026141856565,
"grad_norm": 0.6768864989280701,
"learning_rate": 2.814145281687025e-05,
"loss": 1.2900625228881837,
"step": 3320
},
{
"epoch": 0.6560126270099635,
"grad_norm": 0.5812232494354248,
"learning_rate": 2.7997876654472515e-05,
"loss": 1.2744749069213868,
"step": 3325
},
{
"epoch": 0.6569991121633619,
"grad_norm": 0.5782356858253479,
"learning_rate": 2.7854525128526805e-05,
"loss": 1.1774396896362305,
"step": 3330
},
{
"epoch": 0.6579855973167604,
"grad_norm": 0.6956771016120911,
"learning_rate": 2.7711399702618164e-05,
"loss": 1.2629971504211426,
"step": 3335
},
{
"epoch": 0.6589720824701588,
"grad_norm": 0.6856631636619568,
"learning_rate": 2.7568501838023198e-05,
"loss": 1.2138921737670898,
"step": 3340
},
{
"epoch": 0.6599585676235573,
"grad_norm": 0.6038599014282227,
"learning_rate": 2.742583299369513e-05,
"loss": 1.1959566116333007,
"step": 3345
},
{
"epoch": 0.6609450527769557,
"grad_norm": 0.6338747143745422,
"learning_rate": 2.7283394626248992e-05,
"loss": 1.2070375442504884,
"step": 3350
},
{
"epoch": 0.6619315379303542,
"grad_norm": 0.592533528804779,
"learning_rate": 2.7141188189946655e-05,
"loss": 1.216831874847412,
"step": 3355
},
{
"epoch": 0.6629180230837526,
"grad_norm": 0.7979716658592224,
"learning_rate": 2.6999215136682032e-05,
"loss": 1.2753192901611328,
"step": 3360
},
{
"epoch": 0.6639045082371511,
"grad_norm": 0.5777809023857117,
"learning_rate": 2.685747691596631e-05,
"loss": 1.2421433448791503,
"step": 3365
},
{
"epoch": 0.6648909933905495,
"grad_norm": 0.5968480110168457,
"learning_rate": 2.6715974974913006e-05,
"loss": 1.2744085311889648,
"step": 3370
},
{
"epoch": 0.665877478543948,
"grad_norm": 0.8302050828933716,
"learning_rate": 2.6574710758223354e-05,
"loss": 1.1875201225280763,
"step": 3375
},
{
"epoch": 0.6668639636973464,
"grad_norm": 0.7582641243934631,
"learning_rate": 2.6433685708171406e-05,
"loss": 1.2816165924072265,
"step": 3380
},
{
"epoch": 0.6678504488507448,
"grad_norm": 0.6326640248298645,
"learning_rate": 2.6292901264589443e-05,
"loss": 1.2224931716918945,
"step": 3385
},
{
"epoch": 0.6688369340041432,
"grad_norm": 0.6629419922828674,
"learning_rate": 2.6152358864853222e-05,
"loss": 1.274005126953125,
"step": 3390
},
{
"epoch": 0.6698234191575417,
"grad_norm": 0.5562908053398132,
"learning_rate": 2.601205994386718e-05,
"loss": 1.1907829284667968,
"step": 3395
},
{
"epoch": 0.6708099043109401,
"grad_norm": 0.8455808758735657,
"learning_rate": 2.5872005934050003e-05,
"loss": 1.1719115257263184,
"step": 3400
},
{
"epoch": 0.6717963894643386,
"grad_norm": 0.561410129070282,
"learning_rate": 2.5732198265319873e-05,
"loss": 1.2491145133972168,
"step": 3405
},
{
"epoch": 0.672782874617737,
"grad_norm": 0.7136441469192505,
"learning_rate": 2.5592638365079837e-05,
"loss": 1.2955787658691407,
"step": 3410
},
{
"epoch": 0.6737693597711354,
"grad_norm": 0.5923577547073364,
"learning_rate": 2.5453327658203357e-05,
"loss": 1.2561634063720704,
"step": 3415
},
{
"epoch": 0.6747558449245339,
"grad_norm": 0.6192749738693237,
"learning_rate": 2.5314267567019612e-05,
"loss": 1.1964563369750976,
"step": 3420
},
{
"epoch": 0.6757423300779323,
"grad_norm": 0.6341261267662048,
"learning_rate": 2.5175459511299144e-05,
"loss": 1.218198013305664,
"step": 3425
},
{
"epoch": 0.6767288152313308,
"grad_norm": 0.744046151638031,
"learning_rate": 2.503690490823919e-05,
"loss": 1.174126148223877,
"step": 3430
},
{
"epoch": 0.6777153003847292,
"grad_norm": 0.6525372266769409,
"learning_rate": 2.4898605172449318e-05,
"loss": 1.2664989471435546,
"step": 3435
},
{
"epoch": 0.6787017855381277,
"grad_norm": 0.616605281829834,
"learning_rate": 2.4760561715937016e-05,
"loss": 1.1964638710021973,
"step": 3440
},
{
"epoch": 0.679688270691526,
"grad_norm": 0.6096294522285461,
"learning_rate": 2.4622775948093125e-05,
"loss": 1.2232364654541015,
"step": 3445
},
{
"epoch": 0.6806747558449245,
"grad_norm": 0.8499689698219299,
"learning_rate": 2.448524927567763e-05,
"loss": 1.388045883178711,
"step": 3450
},
{
"epoch": 0.6816612409983229,
"grad_norm": 0.6912193894386292,
"learning_rate": 2.43479831028052e-05,
"loss": 1.281625461578369,
"step": 3455
},
{
"epoch": 0.6826477261517214,
"grad_norm": 0.6856531500816345,
"learning_rate": 2.4210978830930802e-05,
"loss": 1.2008772850036622,
"step": 3460
},
{
"epoch": 0.6836342113051198,
"grad_norm": 1.069383978843689,
"learning_rate": 2.4074237858835554e-05,
"loss": 1.2502653121948242,
"step": 3465
},
{
"epoch": 0.6846206964585183,
"grad_norm": 0.614947497844696,
"learning_rate": 2.3937761582612263e-05,
"loss": 1.1920949935913085,
"step": 3470
},
{
"epoch": 0.6856071816119167,
"grad_norm": 0.6955264806747437,
"learning_rate": 2.3801551395651274e-05,
"loss": 1.212801170349121,
"step": 3475
},
{
"epoch": 0.6865936667653152,
"grad_norm": 0.8256301879882812,
"learning_rate": 2.366560868862626e-05,
"loss": 1.229395294189453,
"step": 3480
},
{
"epoch": 0.6875801519187136,
"grad_norm": 0.7527047395706177,
"learning_rate": 2.352993484947994e-05,
"loss": 1.1920668601989746,
"step": 3485
},
{
"epoch": 0.6885666370721121,
"grad_norm": 0.7431856393814087,
"learning_rate": 2.3394531263410007e-05,
"loss": 1.2293588638305664,
"step": 3490
},
{
"epoch": 0.6895531222255105,
"grad_norm": 0.7737157344818115,
"learning_rate": 2.3259399312854875e-05,
"loss": 1.2111127853393555,
"step": 3495
},
{
"epoch": 0.690539607378909,
"grad_norm": 0.5195772647857666,
"learning_rate": 2.3124540377479665e-05,
"loss": 1.2467874526977538,
"step": 3500
},
{
"epoch": 0.6915260925323073,
"grad_norm": 0.6953533887863159,
"learning_rate": 2.2989955834162104e-05,
"loss": 1.1929851531982423,
"step": 3505
},
{
"epoch": 0.6925125776857058,
"grad_norm": 0.6353641748428345,
"learning_rate": 2.2855647056978335e-05,
"loss": 1.320963478088379,
"step": 3510
},
{
"epoch": 0.6934990628391042,
"grad_norm": 0.7210065722465515,
"learning_rate": 2.2721615417189097e-05,
"loss": 1.1938302993774415,
"step": 3515
},
{
"epoch": 0.6944855479925027,
"grad_norm": 0.5775903463363647,
"learning_rate": 2.258786228322561e-05,
"loss": 1.2433890342712401,
"step": 3520
},
{
"epoch": 0.6954720331459011,
"grad_norm": 0.587211012840271,
"learning_rate": 2.2454389020675552e-05,
"loss": 1.2664244651794434,
"step": 3525
},
{
"epoch": 0.6964585182992996,
"grad_norm": 0.5521672368049622,
"learning_rate": 2.2321196992269273e-05,
"loss": 1.1534212112426758,
"step": 3530
},
{
"epoch": 0.697445003452698,
"grad_norm": 0.7396832704544067,
"learning_rate": 2.2188287557865687e-05,
"loss": 1.2141143798828125,
"step": 3535
},
{
"epoch": 0.6984314886060965,
"grad_norm": 0.6359992027282715,
"learning_rate": 2.205566207443859e-05,
"loss": 1.1850906372070313,
"step": 3540
},
{
"epoch": 0.6994179737594949,
"grad_norm": 0.6060147285461426,
"learning_rate": 2.192332189606263e-05,
"loss": 1.1797555923461913,
"step": 3545
},
{
"epoch": 0.7004044589128934,
"grad_norm": 0.7586352229118347,
"learning_rate": 2.1791268373899576e-05,
"loss": 1.233713722229004,
"step": 3550
},
{
"epoch": 0.7013909440662918,
"grad_norm": 0.584236204624176,
"learning_rate": 2.165950285618455e-05,
"loss": 1.2108146667480468,
"step": 3555
},
{
"epoch": 0.7023774292196903,
"grad_norm": 0.6863627433776855,
"learning_rate": 2.1528026688212144e-05,
"loss": 1.2762624740600585,
"step": 3560
},
{
"epoch": 0.7033639143730887,
"grad_norm": 0.6688559055328369,
"learning_rate": 2.1396841212322822e-05,
"loss": 1.1927000045776368,
"step": 3565
},
{
"epoch": 0.7043503995264871,
"grad_norm": 0.8499382138252258,
"learning_rate": 2.1265947767889134e-05,
"loss": 1.2211936950683593,
"step": 3570
},
{
"epoch": 0.7053368846798855,
"grad_norm": 0.6978778839111328,
"learning_rate": 2.1135347691302016e-05,
"loss": 1.2047572135925293,
"step": 3575
},
{
"epoch": 0.706323369833284,
"grad_norm": 0.6998097896575928,
"learning_rate": 2.1005042315957263e-05,
"loss": 1.1659759521484374,
"step": 3580
},
{
"epoch": 0.7073098549866824,
"grad_norm": 0.6424363255500793,
"learning_rate": 2.087503297224178e-05,
"loss": 1.2477108001708985,
"step": 3585
},
{
"epoch": 0.7082963401400809,
"grad_norm": 0.7973096966743469,
"learning_rate": 2.0745320987520057e-05,
"loss": 1.3090511322021485,
"step": 3590
},
{
"epoch": 0.7092828252934793,
"grad_norm": 0.8021285533905029,
"learning_rate": 2.061590768612067e-05,
"loss": 1.2245662689208985,
"step": 3595
},
{
"epoch": 0.7102693104468778,
"grad_norm": 0.6695612668991089,
"learning_rate": 2.048679438932266e-05,
"loss": 1.199762725830078,
"step": 3600
},
{
"epoch": 0.7112557956002762,
"grad_norm": 0.5868820548057556,
"learning_rate": 2.0357982415342143e-05,
"loss": 1.1535480499267579,
"step": 3605
},
{
"epoch": 0.7122422807536747,
"grad_norm": 0.6480816006660461,
"learning_rate": 2.022947307931873e-05,
"loss": 1.222834300994873,
"step": 3610
},
{
"epoch": 0.7132287659070731,
"grad_norm": 0.7549391984939575,
"learning_rate": 2.0101267693302224e-05,
"loss": 1.1557947158813477,
"step": 3615
},
{
"epoch": 0.7142152510604716,
"grad_norm": 0.7174381017684937,
"learning_rate": 1.997336756623918e-05,
"loss": 1.263605499267578,
"step": 3620
},
{
"epoch": 0.71520173621387,
"grad_norm": 0.7525988817214966,
"learning_rate": 1.9845774003959494e-05,
"loss": 1.2489173889160157,
"step": 3625
},
{
"epoch": 0.7161882213672685,
"grad_norm": 0.6560421586036682,
"learning_rate": 1.9718488309163117e-05,
"loss": 1.268783950805664,
"step": 3630
},
{
"epoch": 0.7171747065206668,
"grad_norm": 0.8618171215057373,
"learning_rate": 1.9591511781406746e-05,
"loss": 1.2355231285095214,
"step": 3635
},
{
"epoch": 0.7181611916740653,
"grad_norm": 0.5976682901382446,
"learning_rate": 1.946484571709056e-05,
"loss": 1.2503744125366212,
"step": 3640
},
{
"epoch": 0.7191476768274637,
"grad_norm": 0.7242641448974609,
"learning_rate": 1.933849140944502e-05,
"loss": 1.1406134605407714,
"step": 3645
},
{
"epoch": 0.7201341619808622,
"grad_norm": 0.6839653253555298,
"learning_rate": 1.9212450148517537e-05,
"loss": 1.1752308845520019,
"step": 3650
},
{
"epoch": 0.7211206471342606,
"grad_norm": 0.6410506367683411,
"learning_rate": 1.908672322115948e-05,
"loss": 1.280893039703369,
"step": 3655
},
{
"epoch": 0.7221071322876591,
"grad_norm": 1.1221973896026611,
"learning_rate": 1.896131191101289e-05,
"loss": 1.2831155776977539,
"step": 3660
},
{
"epoch": 0.7230936174410575,
"grad_norm": 0.6471509337425232,
"learning_rate": 1.8836217498497434e-05,
"loss": 1.2341423988342286,
"step": 3665
},
{
"epoch": 0.724080102594456,
"grad_norm": 0.6761668920516968,
"learning_rate": 1.8711441260797376e-05,
"loss": 1.146875,
"step": 3670
},
{
"epoch": 0.7250665877478544,
"grad_norm": 0.579704999923706,
"learning_rate": 1.8586984471848417e-05,
"loss": 1.228449249267578,
"step": 3675
},
{
"epoch": 0.7260530729012529,
"grad_norm": 0.7871549129486084,
"learning_rate": 1.846284840232485e-05,
"loss": 1.266524600982666,
"step": 3680
},
{
"epoch": 0.7270395580546513,
"grad_norm": 0.7466410994529724,
"learning_rate": 1.8339034319626407e-05,
"loss": 1.2731217384338378,
"step": 3685
},
{
"epoch": 0.7280260432080498,
"grad_norm": 0.6137645244598389,
"learning_rate": 1.821554348786548e-05,
"loss": 1.2103815078735352,
"step": 3690
},
{
"epoch": 0.7290125283614481,
"grad_norm": 0.6029869318008423,
"learning_rate": 1.809237716785414e-05,
"loss": 1.1383550643920899,
"step": 3695
},
{
"epoch": 0.7299990135148466,
"grad_norm": 0.8937312364578247,
"learning_rate": 1.796953661709123e-05,
"loss": 1.3221405029296875,
"step": 3700
},
{
"epoch": 0.730985498668245,
"grad_norm": 0.8550584316253662,
"learning_rate": 1.7847023089749575e-05,
"loss": 1.1921648025512694,
"step": 3705
},
{
"epoch": 0.7319719838216435,
"grad_norm": 0.6273738741874695,
"learning_rate": 1.7724837836663216e-05,
"loss": 1.2872068405151367,
"step": 3710
},
{
"epoch": 0.7329584689750419,
"grad_norm": 0.7247760891914368,
"learning_rate": 1.7602982105314518e-05,
"loss": 1.2260135650634765,
"step": 3715
},
{
"epoch": 0.7339449541284404,
"grad_norm": 0.7579748034477234,
"learning_rate": 1.7481457139821593e-05,
"loss": 1.2338243484497071,
"step": 3720
},
{
"epoch": 0.7349314392818388,
"grad_norm": 0.6567061543464661,
"learning_rate": 1.7360264180925423e-05,
"loss": 1.2360413551330567,
"step": 3725
},
{
"epoch": 0.7359179244352373,
"grad_norm": 0.8668411374092102,
"learning_rate": 1.7239404465977365e-05,
"loss": 1.2689210891723632,
"step": 3730
},
{
"epoch": 0.7369044095886357,
"grad_norm": 0.7307335138320923,
"learning_rate": 1.7118879228926353e-05,
"loss": 1.2083996772766112,
"step": 3735
},
{
"epoch": 0.7378908947420342,
"grad_norm": 0.7293202877044678,
"learning_rate": 1.6998689700306465e-05,
"loss": 1.2733915328979493,
"step": 3740
},
{
"epoch": 0.7388773798954326,
"grad_norm": 0.8396404385566711,
"learning_rate": 1.687883710722421e-05,
"loss": 1.0877705574035645,
"step": 3745
},
{
"epoch": 0.7398638650488311,
"grad_norm": 0.7708191871643066,
"learning_rate": 1.6759322673346074e-05,
"loss": 1.241110610961914,
"step": 3750
},
{
"epoch": 0.7408503502022294,
"grad_norm": 0.6098671555519104,
"learning_rate": 1.6640147618886048e-05,
"loss": 1.1960975646972656,
"step": 3755
},
{
"epoch": 0.7418368353556279,
"grad_norm": 0.6884947419166565,
"learning_rate": 1.652131316059315e-05,
"loss": 1.3021397590637207,
"step": 3760
},
{
"epoch": 0.7428233205090263,
"grad_norm": 0.6636525988578796,
"learning_rate": 1.6402820511738915e-05,
"loss": 1.2300701141357422,
"step": 3765
},
{
"epoch": 0.7438098056624248,
"grad_norm": 0.5936552882194519,
"learning_rate": 1.6284670882105185e-05,
"loss": 1.2187934875488282,
"step": 3770
},
{
"epoch": 0.7447962908158232,
"grad_norm": 0.7512790560722351,
"learning_rate": 1.6166865477971565e-05,
"loss": 1.275744342803955,
"step": 3775
},
{
"epoch": 0.7457827759692217,
"grad_norm": 0.8274773359298706,
"learning_rate": 1.604940550210329e-05,
"loss": 1.2067558288574218,
"step": 3780
},
{
"epoch": 0.7467692611226201,
"grad_norm": 1.1417531967163086,
"learning_rate": 1.593229215373877e-05,
"loss": 1.2055864334106445,
"step": 3785
},
{
"epoch": 0.7477557462760186,
"grad_norm": 0.6661544442176819,
"learning_rate": 1.581552662857747e-05,
"loss": 1.3098382949829102,
"step": 3790
},
{
"epoch": 0.748742231429417,
"grad_norm": 0.615811824798584,
"learning_rate": 1.5699110118767678e-05,
"loss": 1.2139877319335937,
"step": 3795
},
{
"epoch": 0.7497287165828155,
"grad_norm": 0.7433391213417053,
"learning_rate": 1.558304381289428e-05,
"loss": 1.2888544082641602,
"step": 3800
},
{
"epoch": 0.7507152017362139,
"grad_norm": 0.8470045328140259,
"learning_rate": 1.5467328895966698e-05,
"loss": 1.2692300796508789,
"step": 3805
},
{
"epoch": 0.7517016868896124,
"grad_norm": 0.6241767406463623,
"learning_rate": 1.535196654940676e-05,
"loss": 1.2418964385986329,
"step": 3810
},
{
"epoch": 0.7526881720430108,
"grad_norm": 0.7404466867446899,
"learning_rate": 1.5236957951036585e-05,
"loss": 1.2656991958618165,
"step": 3815
},
{
"epoch": 0.7536746571964092,
"grad_norm": 0.6952432990074158,
"learning_rate": 1.512230427506664e-05,
"loss": 1.2338417053222657,
"step": 3820
},
{
"epoch": 0.7546611423498076,
"grad_norm": 0.7083795070648193,
"learning_rate": 1.5008006692083742e-05,
"loss": 1.2619205474853517,
"step": 3825
},
{
"epoch": 0.7556476275032061,
"grad_norm": 0.9311999082565308,
"learning_rate": 1.4894066369039023e-05,
"loss": 1.3122014999389648,
"step": 3830
},
{
"epoch": 0.7566341126566045,
"grad_norm": 0.940156877040863,
"learning_rate": 1.478048446923615e-05,
"loss": 1.1117132186889649,
"step": 3835
},
{
"epoch": 0.7576205978100029,
"grad_norm": 0.7090609073638916,
"learning_rate": 1.46672621523193e-05,
"loss": 1.1701818466186524,
"step": 3840
},
{
"epoch": 0.7586070829634014,
"grad_norm": 0.6446618437767029,
"learning_rate": 1.4554400574261456e-05,
"loss": 1.2477872848510743,
"step": 3845
},
{
"epoch": 0.7595935681167998,
"grad_norm": 0.7125352621078491,
"learning_rate": 1.4441900887352495e-05,
"loss": 1.220113754272461,
"step": 3850
},
{
"epoch": 0.7605800532701983,
"grad_norm": 0.6272805333137512,
"learning_rate": 1.4329764240187515e-05,
"loss": 1.2100269317626953,
"step": 3855
},
{
"epoch": 0.7615665384235967,
"grad_norm": 0.6574905514717102,
"learning_rate": 1.4217991777655032e-05,
"loss": 1.1784603118896484,
"step": 3860
},
{
"epoch": 0.7625530235769952,
"grad_norm": 0.6214209198951721,
"learning_rate": 1.4106584640925307e-05,
"loss": 1.3334485054016114,
"step": 3865
},
{
"epoch": 0.7635395087303936,
"grad_norm": 0.6638909578323364,
"learning_rate": 1.3995543967438768e-05,
"loss": 1.213265609741211,
"step": 3870
},
{
"epoch": 0.764525993883792,
"grad_norm": 0.5960767269134521,
"learning_rate": 1.3884870890894318e-05,
"loss": 1.265770149230957,
"step": 3875
},
{
"epoch": 0.7655124790371904,
"grad_norm": 0.64935702085495,
"learning_rate": 1.3774566541237743e-05,
"loss": 1.1788627624511718,
"step": 3880
},
{
"epoch": 0.7664989641905889,
"grad_norm": 0.8633273839950562,
"learning_rate": 1.3664632044650277e-05,
"loss": 1.235267162322998,
"step": 3885
},
{
"epoch": 0.7674854493439873,
"grad_norm": 0.6872686743736267,
"learning_rate": 1.3555068523536984e-05,
"loss": 1.2033307075500488,
"step": 3890
},
{
"epoch": 0.7684719344973858,
"grad_norm": 0.7337464094161987,
"learning_rate": 1.3445877096515413e-05,
"loss": 1.1980605125427246,
"step": 3895
},
{
"epoch": 0.7694584196507842,
"grad_norm": 0.7682844400405884,
"learning_rate": 1.333705887840408e-05,
"loss": 1.2687780380249023,
"step": 3900
},
{
"epoch": 0.7704449048041827,
"grad_norm": 0.5740364193916321,
"learning_rate": 1.3228614980211113e-05,
"loss": 1.1739182472229004,
"step": 3905
},
{
"epoch": 0.7714313899575811,
"grad_norm": 0.5720030665397644,
"learning_rate": 1.3120546509122993e-05,
"loss": 1.2076943397521973,
"step": 3910
},
{
"epoch": 0.7724178751109796,
"grad_norm": 0.7429443597793579,
"learning_rate": 1.3012854568493094e-05,
"loss": 1.2207366943359375,
"step": 3915
},
{
"epoch": 0.773404360264378,
"grad_norm": 0.7344704270362854,
"learning_rate": 1.2905540257830568e-05,
"loss": 1.2863577842712401,
"step": 3920
},
{
"epoch": 0.7743908454177765,
"grad_norm": 0.6011790037155151,
"learning_rate": 1.2798604672789033e-05,
"loss": 1.1608015060424806,
"step": 3925
},
{
"epoch": 0.7753773305711749,
"grad_norm": 0.7186335921287537,
"learning_rate": 1.2692048905155402e-05,
"loss": 1.136014747619629,
"step": 3930
},
{
"epoch": 0.7763638157245734,
"grad_norm": 0.5801212787628174,
"learning_rate": 1.2585874042838747e-05,
"loss": 1.232620620727539,
"step": 3935
},
{
"epoch": 0.7773503008779717,
"grad_norm": 1.0780433416366577,
"learning_rate": 1.2480081169859165e-05,
"loss": 1.1188444137573241,
"step": 3940
},
{
"epoch": 0.7783367860313702,
"grad_norm": 0.6121573448181152,
"learning_rate": 1.2374671366336777e-05,
"loss": 1.1317131042480468,
"step": 3945
},
{
"epoch": 0.7793232711847686,
"grad_norm": 0.7263827323913574,
"learning_rate": 1.2269645708480643e-05,
"loss": 1.1772459030151368,
"step": 3950
},
{
"epoch": 0.7803097563381671,
"grad_norm": 0.6388430595397949,
"learning_rate": 1.2165005268577746e-05,
"loss": 1.2816486358642578,
"step": 3955
},
{
"epoch": 0.7812962414915655,
"grad_norm": 0.5850164294242859,
"learning_rate": 1.206075111498215e-05,
"loss": 1.3245516777038575,
"step": 3960
},
{
"epoch": 0.782282726644964,
"grad_norm": 0.6729406714439392,
"learning_rate": 1.1956884312103972e-05,
"loss": 1.2367045402526855,
"step": 3965
},
{
"epoch": 0.7832692117983624,
"grad_norm": 0.7395087480545044,
"learning_rate": 1.1853405920398609e-05,
"loss": 1.2816341400146485,
"step": 3970
},
{
"epoch": 0.7842556969517609,
"grad_norm": 0.7512933015823364,
"learning_rate": 1.1750316996355842e-05,
"loss": 1.2577844619750977,
"step": 3975
},
{
"epoch": 0.7852421821051593,
"grad_norm": 0.832805335521698,
"learning_rate": 1.164761859248909e-05,
"loss": 1.268585205078125,
"step": 3980
},
{
"epoch": 0.7862286672585578,
"grad_norm": 0.6334481239318848,
"learning_rate": 1.1545311757324678e-05,
"loss": 1.3201239585876465,
"step": 3985
},
{
"epoch": 0.7872151524119562,
"grad_norm": 0.7020235657691956,
"learning_rate": 1.1443397535391059e-05,
"loss": 1.199781608581543,
"step": 3990
},
{
"epoch": 0.7882016375653547,
"grad_norm": 0.8673114776611328,
"learning_rate": 1.1341876967208237e-05,
"loss": 1.2266029357910155,
"step": 3995
},
{
"epoch": 0.789188122718753,
"grad_norm": 0.5635994076728821,
"learning_rate": 1.1240751089277119e-05,
"loss": 1.2839754104614258,
"step": 4000
},
{
"epoch": 0.7901746078721515,
"grad_norm": 0.5986664891242981,
"learning_rate": 1.114002093406885e-05,
"loss": 1.175130844116211,
"step": 4005
},
{
"epoch": 0.7911610930255499,
"grad_norm": 0.655409574508667,
"learning_rate": 1.103968753001442e-05,
"loss": 1.2655525207519531,
"step": 4010
},
{
"epoch": 0.7921475781789484,
"grad_norm": 0.6396498680114746,
"learning_rate": 1.0939751901494033e-05,
"loss": 1.2779668807983398,
"step": 4015
},
{
"epoch": 0.7931340633323468,
"grad_norm": 0.6236947178840637,
"learning_rate": 1.0840215068826698e-05,
"loss": 1.249730110168457,
"step": 4020
},
{
"epoch": 0.7941205484857453,
"grad_norm": 0.6840082406997681,
"learning_rate": 1.0741078048259856e-05,
"loss": 1.2666340827941895,
"step": 4025
},
{
"epoch": 0.7951070336391437,
"grad_norm": 0.8755819797515869,
"learning_rate": 1.0642341851958903e-05,
"loss": 1.207362461090088,
"step": 4030
},
{
"epoch": 0.7960935187925422,
"grad_norm": 0.6677442789077759,
"learning_rate": 1.0544007487996977e-05,
"loss": 1.1208080291748046,
"step": 4035
},
{
"epoch": 0.7970800039459406,
"grad_norm": 0.7978340983390808,
"learning_rate": 1.0446075960344542e-05,
"loss": 1.2827581405639648,
"step": 4040
},
{
"epoch": 0.7980664890993391,
"grad_norm": 0.6629666686058044,
"learning_rate": 1.0348548268859249e-05,
"loss": 1.329923725128174,
"step": 4045
},
{
"epoch": 0.7990529742527375,
"grad_norm": 0.6499217748641968,
"learning_rate": 1.0251425409275673e-05,
"loss": 1.1210155487060547,
"step": 4050
},
{
"epoch": 0.800039459406136,
"grad_norm": 0.8563830852508545,
"learning_rate": 1.0154708373195083e-05,
"loss": 1.1945762634277344,
"step": 4055
},
{
"epoch": 0.8010259445595344,
"grad_norm": 0.6098916530609131,
"learning_rate": 1.0058398148075475e-05,
"loss": 1.1705703735351562,
"step": 4060
},
{
"epoch": 0.8020124297129329,
"grad_norm": 0.6772810816764832,
"learning_rate": 9.962495717221377e-06,
"loss": 1.154684066772461,
"step": 4065
},
{
"epoch": 0.8029989148663312,
"grad_norm": 0.932282567024231,
"learning_rate": 9.86700205977379e-06,
"loss": 1.2413950920104981,
"step": 4070
},
{
"epoch": 0.8039854000197297,
"grad_norm": 0.6044415831565857,
"learning_rate": 9.771918150700298e-06,
"loss": 1.2596137046813964,
"step": 4075
},
{
"epoch": 0.8049718851731281,
"grad_norm": 0.7235296368598938,
"learning_rate": 9.677244960784998e-06,
"loss": 1.2018991470336915,
"step": 4080
},
{
"epoch": 0.8059583703265266,
"grad_norm": 0.6422549486160278,
"learning_rate": 9.582983456618694e-06,
"loss": 1.1565820693969726,
"step": 4085
},
{
"epoch": 0.806944855479925,
"grad_norm": 0.6846360564231873,
"learning_rate": 9.489134600588933e-06,
"loss": 1.2013694763183593,
"step": 4090
},
{
"epoch": 0.8079313406333235,
"grad_norm": 0.6198418140411377,
"learning_rate": 9.395699350870225e-06,
"loss": 1.188382911682129,
"step": 4095
},
{
"epoch": 0.8089178257867219,
"grad_norm": 0.9237287044525146,
"learning_rate": 9.3026786614143e-06,
"loss": 1.2073503494262696,
"step": 4100
},
{
"epoch": 0.8099043109401204,
"grad_norm": 0.7590090036392212,
"learning_rate": 9.210073481940268e-06,
"loss": 1.0931904792785645,
"step": 4105
},
{
"epoch": 0.8108907960935188,
"grad_norm": 0.8027364015579224,
"learning_rate": 9.117884757925021e-06,
"loss": 1.2096965789794922,
"step": 4110
},
{
"epoch": 0.8118772812469173,
"grad_norm": 0.8245887160301208,
"learning_rate": 9.026113430593546e-06,
"loss": 1.206573486328125,
"step": 4115
},
{
"epoch": 0.8128637664003157,
"grad_norm": 0.7403396368026733,
"learning_rate": 8.934760436909273e-06,
"loss": 1.2779735565185546,
"step": 4120
},
{
"epoch": 0.8138502515537142,
"grad_norm": 1.0411396026611328,
"learning_rate": 8.843826709564573e-06,
"loss": 1.307113265991211,
"step": 4125
},
{
"epoch": 0.8148367367071125,
"grad_norm": 0.908251941204071,
"learning_rate": 8.753313176971195e-06,
"loss": 1.3073105812072754,
"step": 4130
},
{
"epoch": 0.815823221860511,
"grad_norm": 0.7219868898391724,
"learning_rate": 8.66322076325079e-06,
"loss": 1.265813159942627,
"step": 4135
},
{
"epoch": 0.8168097070139094,
"grad_norm": 0.7738665342330933,
"learning_rate": 8.57355038822552e-06,
"loss": 1.2150358200073241,
"step": 4140
},
{
"epoch": 0.8177961921673079,
"grad_norm": 0.9167988300323486,
"learning_rate": 8.484302967408585e-06,
"loss": 1.25528564453125,
"step": 4145
},
{
"epoch": 0.8187826773207063,
"grad_norm": 0.9312618374824524,
"learning_rate": 8.395479411994978e-06,
"loss": 1.2169861793518066,
"step": 4150
},
{
"epoch": 0.8197691624741048,
"grad_norm": 0.5853263735771179,
"learning_rate": 8.307080628852065e-06,
"loss": 1.1567497253417969,
"step": 4155
},
{
"epoch": 0.8207556476275032,
"grad_norm": 0.566801130771637,
"learning_rate": 8.219107520510433e-06,
"loss": 1.2267602920532226,
"step": 4160
},
{
"epoch": 0.8217421327809017,
"grad_norm": 0.877739667892456,
"learning_rate": 8.131560985154652e-06,
"loss": 1.2166537284851073,
"step": 4165
},
{
"epoch": 0.8227286179343001,
"grad_norm": 0.7193045020103455,
"learning_rate": 8.044441916613982e-06,
"loss": 1.1995410919189453,
"step": 4170
},
{
"epoch": 0.8237151030876986,
"grad_norm": 0.6103017330169678,
"learning_rate": 7.957751204353452e-06,
"loss": 1.1429649353027345,
"step": 4175
},
{
"epoch": 0.824701588241097,
"grad_norm": 0.5330570340156555,
"learning_rate": 7.871489733464649e-06,
"loss": 1.184005069732666,
"step": 4180
},
{
"epoch": 0.8256880733944955,
"grad_norm": 0.6953468322753906,
"learning_rate": 7.785658384656675e-06,
"loss": 1.2297355651855468,
"step": 4185
},
{
"epoch": 0.8266745585478938,
"grad_norm": 0.5787708759307861,
"learning_rate": 7.70025803424722e-06,
"loss": 1.1368911743164063,
"step": 4190
},
{
"epoch": 0.8276610437012923,
"grad_norm": 0.6307753920555115,
"learning_rate": 7.615289554153559e-06,
"loss": 1.308220386505127,
"step": 4195
},
{
"epoch": 0.8286475288546907,
"grad_norm": 0.7535484433174133,
"learning_rate": 7.530753811883711e-06,
"loss": 1.32164306640625,
"step": 4200
},
{
"epoch": 0.8296340140080892,
"grad_norm": 0.6035266518592834,
"learning_rate": 7.4466516705275105e-06,
"loss": 1.1468348503112793,
"step": 4205
},
{
"epoch": 0.8306204991614876,
"grad_norm": 0.5805249214172363,
"learning_rate": 7.362983988747818e-06,
"loss": 1.245039176940918,
"step": 4210
},
{
"epoch": 0.8316069843148861,
"grad_norm": 0.6887459754943848,
"learning_rate": 7.2797516207718135e-06,
"loss": 1.1813220977783203,
"step": 4215
},
{
"epoch": 0.8325934694682845,
"grad_norm": 0.6108956336975098,
"learning_rate": 7.196955416382179e-06,
"loss": 1.1665533065795899,
"step": 4220
},
{
"epoch": 0.833579954621683,
"grad_norm": 0.766027569770813,
"learning_rate": 7.114596220908504e-06,
"loss": 1.1568939208984375,
"step": 4225
},
{
"epoch": 0.8345664397750814,
"grad_norm": 0.7858899235725403,
"learning_rate": 7.032674875218614e-06,
"loss": 1.2345382690429687,
"step": 4230
},
{
"epoch": 0.8355529249284799,
"grad_norm": 0.6940763592720032,
"learning_rate": 6.951192215709962e-06,
"loss": 1.1764236450195313,
"step": 4235
},
{
"epoch": 0.8365394100818783,
"grad_norm": 0.6433379054069519,
"learning_rate": 6.870149074301185e-06,
"loss": 1.1978044509887695,
"step": 4240
},
{
"epoch": 0.8375258952352768,
"grad_norm": 0.7813035249710083,
"learning_rate": 6.789546278423447e-06,
"loss": 1.2522754669189453,
"step": 4245
},
{
"epoch": 0.8385123803886751,
"grad_norm": 0.9612628221511841,
"learning_rate": 6.7093846510121736e-06,
"loss": 1.1740789413452148,
"step": 4250
},
{
"epoch": 0.8394988655420736,
"grad_norm": 0.7509130835533142,
"learning_rate": 6.6296650104985426e-06,
"loss": 1.234006690979004,
"step": 4255
},
{
"epoch": 0.840485350695472,
"grad_norm": 0.6952577829360962,
"learning_rate": 6.55038817080112e-06,
"loss": 1.1954164505004883,
"step": 4260
},
{
"epoch": 0.8414718358488704,
"grad_norm": 0.6885148882865906,
"learning_rate": 6.471554941317631e-06,
"loss": 1.222324752807617,
"step": 4265
},
{
"epoch": 0.8424583210022689,
"grad_norm": 0.602550208568573,
"learning_rate": 6.3931661269166e-06,
"loss": 1.2029653549194337,
"step": 4270
},
{
"epoch": 0.8434448061556673,
"grad_norm": 0.8633717894554138,
"learning_rate": 6.315222527929205e-06,
"loss": 1.372523307800293,
"step": 4275
},
{
"epoch": 0.8444312913090658,
"grad_norm": 0.995661735534668,
"learning_rate": 6.237724940141088e-06,
"loss": 1.2466371536254883,
"step": 4280
},
{
"epoch": 0.8454177764624642,
"grad_norm": 0.6375277042388916,
"learning_rate": 6.16067415478418e-06,
"loss": 1.2364391326904296,
"step": 4285
},
{
"epoch": 0.8464042616158627,
"grad_norm": 0.5509450435638428,
"learning_rate": 6.084070958528714e-06,
"loss": 1.2184797286987306,
"step": 4290
},
{
"epoch": 0.8473907467692611,
"grad_norm": 0.6702736020088196,
"learning_rate": 6.0079161334751114e-06,
"loss": 1.2568435668945312,
"step": 4295
},
{
"epoch": 0.8483772319226596,
"grad_norm": 0.6265482902526855,
"learning_rate": 5.932210457146043e-06,
"loss": 1.2029899597167968,
"step": 4300
},
{
"epoch": 0.849363717076058,
"grad_norm": 0.6685672998428345,
"learning_rate": 5.856954702478495e-06,
"loss": 1.1792423248291015,
"step": 4305
},
{
"epoch": 0.8503502022294565,
"grad_norm": 0.5622189044952393,
"learning_rate": 5.782149637815825e-06,
"loss": 1.2113577842712402,
"step": 4310
},
{
"epoch": 0.8513366873828548,
"grad_norm": 0.6713023781776428,
"learning_rate": 5.7077960268999916e-06,
"loss": 1.2203788757324219,
"step": 4315
},
{
"epoch": 0.8523231725362533,
"grad_norm": 0.5652981996536255,
"learning_rate": 5.633894628863679e-06,
"loss": 1.2360023498535155,
"step": 4320
},
{
"epoch": 0.8533096576896517,
"grad_norm": 0.5809028744697571,
"learning_rate": 5.560446198222596e-06,
"loss": 1.2158353805541993,
"step": 4325
},
{
"epoch": 0.8542961428430502,
"grad_norm": 0.903544545173645,
"learning_rate": 5.487451484867795e-06,
"loss": 1.1978225708007812,
"step": 4330
},
{
"epoch": 0.8552826279964486,
"grad_norm": 0.6434166431427002,
"learning_rate": 5.414911234057935e-06,
"loss": 1.227414321899414,
"step": 4335
},
{
"epoch": 0.8562691131498471,
"grad_norm": 0.9673121571540833,
"learning_rate": 5.342826186411759e-06,
"loss": 1.2020007133483888,
"step": 4340
},
{
"epoch": 0.8572555983032455,
"grad_norm": 0.6131551861763,
"learning_rate": 5.271197077900459e-06,
"loss": 1.1618553161621095,
"step": 4345
},
{
"epoch": 0.858242083456644,
"grad_norm": 0.7144567370414734,
"learning_rate": 5.200024639840228e-06,
"loss": 1.1534433364868164,
"step": 4350
},
{
"epoch": 0.8592285686100424,
"grad_norm": 0.5641322731971741,
"learning_rate": 5.129309598884757e-06,
"loss": 1.158175563812256,
"step": 4355
},
{
"epoch": 0.8602150537634409,
"grad_norm": 0.7800830006599426,
"learning_rate": 5.0590526770178e-06,
"loss": 1.1411558151245118,
"step": 4360
},
{
"epoch": 0.8612015389168393,
"grad_norm": 0.6084363460540771,
"learning_rate": 4.989254591545833e-06,
"loss": 1.2227473258972168,
"step": 4365
},
{
"epoch": 0.8621880240702378,
"grad_norm": 0.6389532089233398,
"learning_rate": 4.91991605509074e-06,
"loss": 1.263646411895752,
"step": 4370
},
{
"epoch": 0.8631745092236361,
"grad_norm": 0.8883340954780579,
"learning_rate": 4.851037775582495e-06,
"loss": 1.2848438262939452,
"step": 4375
},
{
"epoch": 0.8641609943770346,
"grad_norm": 0.8165562748908997,
"learning_rate": 4.782620456251974e-06,
"loss": 1.2032787322998046,
"step": 4380
},
{
"epoch": 0.865147479530433,
"grad_norm": 0.7483526468276978,
"learning_rate": 4.714664795623741e-06,
"loss": 1.168528938293457,
"step": 4385
},
{
"epoch": 0.8661339646838315,
"grad_norm": 0.7091222405433655,
"learning_rate": 4.647171487508967e-06,
"loss": 1.235909366607666,
"step": 4390
},
{
"epoch": 0.8671204498372299,
"grad_norm": 0.8040610551834106,
"learning_rate": 4.58014122099828e-06,
"loss": 1.1748947143554687,
"step": 4395
},
{
"epoch": 0.8681069349906284,
"grad_norm": 0.6368282437324524,
"learning_rate": 4.513574680454763e-06,
"loss": 1.2476981163024903,
"step": 4400
},
{
"epoch": 0.8690934201440268,
"grad_norm": 0.6664937138557434,
"learning_rate": 4.4474725455070085e-06,
"loss": 1.1268077850341798,
"step": 4405
},
{
"epoch": 0.8700799052974253,
"grad_norm": 0.6298813223838806,
"learning_rate": 4.38183549104208e-06,
"loss": 1.2173730850219726,
"step": 4410
},
{
"epoch": 0.8710663904508237,
"grad_norm": 0.8542167544364929,
"learning_rate": 4.316664187198721e-06,
"loss": 1.1978675842285156,
"step": 4415
},
{
"epoch": 0.8720528756042222,
"grad_norm": 0.7301235795021057,
"learning_rate": 4.251959299360475e-06,
"loss": 1.2227022171020507,
"step": 4420
},
{
"epoch": 0.8730393607576206,
"grad_norm": 0.6046241521835327,
"learning_rate": 4.187721488148855e-06,
"loss": 1.213413619995117,
"step": 4425
},
{
"epoch": 0.8740258459110191,
"grad_norm": 0.6890552043914795,
"learning_rate": 4.123951409416666e-06,
"loss": 1.1941488265991211,
"step": 4430
},
{
"epoch": 0.8750123310644174,
"grad_norm": 0.6779873371124268,
"learning_rate": 4.060649714241249e-06,
"loss": 1.1976250648498534,
"step": 4435
},
{
"epoch": 0.875998816217816,
"grad_norm": 0.7285904288291931,
"learning_rate": 3.997817048917874e-06,
"loss": 1.3076556205749512,
"step": 4440
},
{
"epoch": 0.8769853013712143,
"grad_norm": 0.5105634927749634,
"learning_rate": 3.935454054953136e-06,
"loss": 1.243826675415039,
"step": 4445
},
{
"epoch": 0.8779717865246128,
"grad_norm": 0.7852115035057068,
"learning_rate": 3.87356136905837e-06,
"loss": 1.1994947433471679,
"step": 4450
},
{
"epoch": 0.8789582716780112,
"grad_norm": 0.8086354732513428,
"learning_rate": 3.8121396231432127e-06,
"loss": 1.2185371398925782,
"step": 4455
},
{
"epoch": 0.8799447568314097,
"grad_norm": 0.8336437344551086,
"learning_rate": 3.751189444309083e-06,
"loss": 1.2617269515991212,
"step": 4460
},
{
"epoch": 0.8809312419848081,
"grad_norm": 0.8013225197792053,
"learning_rate": 3.6907114548428324e-06,
"loss": 1.2356754302978517,
"step": 4465
},
{
"epoch": 0.8819177271382066,
"grad_norm": 0.6422016024589539,
"learning_rate": 3.6307062722103713e-06,
"loss": 1.1031051635742188,
"step": 4470
},
{
"epoch": 0.882904212291605,
"grad_norm": 0.9382597208023071,
"learning_rate": 3.5711745090503514e-06,
"loss": 1.2399019241333007,
"step": 4475
},
{
"epoch": 0.8838906974450035,
"grad_norm": 0.6045771241188049,
"learning_rate": 3.5121167731679226e-06,
"loss": 1.2082738876342773,
"step": 4480
},
{
"epoch": 0.8848771825984019,
"grad_norm": 0.7701643109321594,
"learning_rate": 3.4535336675285523e-06,
"loss": 1.175678825378418,
"step": 4485
},
{
"epoch": 0.8858636677518004,
"grad_norm": 0.6739333271980286,
"learning_rate": 3.395425790251816e-06,
"loss": 1.307490634918213,
"step": 4490
},
{
"epoch": 0.8868501529051988,
"grad_norm": 0.6314908862113953,
"learning_rate": 3.3377937346053545e-06,
"loss": 1.1802315711975098,
"step": 4495
},
{
"epoch": 0.8878366380585972,
"grad_norm": 0.6531069874763489,
"learning_rate": 3.2806380889987387e-06,
"loss": 1.205459213256836,
"step": 4500
},
{
"epoch": 0.8888231232119956,
"grad_norm": 0.609950065612793,
"learning_rate": 3.2239594369775584e-06,
"loss": 1.190005397796631,
"step": 4505
},
{
"epoch": 0.8898096083653941,
"grad_norm": 0.7288662791252136,
"learning_rate": 3.1677583572173707e-06,
"loss": 1.1768484115600586,
"step": 4510
},
{
"epoch": 0.8907960935187925,
"grad_norm": 0.6579195857048035,
"learning_rate": 3.1120354235178394e-06,
"loss": 1.192807388305664,
"step": 4515
},
{
"epoch": 0.891782578672191,
"grad_norm": 0.9012724161148071,
"learning_rate": 3.0567912047968936e-06,
"loss": 1.2069808959960937,
"step": 4520
},
{
"epoch": 0.8927690638255894,
"grad_norm": 0.5845980048179626,
"learning_rate": 3.0020262650848673e-06,
"loss": 1.261768913269043,
"step": 4525
},
{
"epoch": 0.8937555489789879,
"grad_norm": 0.5686200857162476,
"learning_rate": 2.9477411635187913e-06,
"loss": 1.1503599166870118,
"step": 4530
},
{
"epoch": 0.8947420341323863,
"grad_norm": 0.8169445991516113,
"learning_rate": 2.8939364543366654e-06,
"loss": 1.276069164276123,
"step": 4535
},
{
"epoch": 0.8957285192857848,
"grad_norm": 0.7000468373298645,
"learning_rate": 2.8406126868717744e-06,
"loss": 1.2005956649780274,
"step": 4540
},
{
"epoch": 0.8967150044391832,
"grad_norm": 0.7924646735191345,
"learning_rate": 2.787770405547119e-06,
"loss": 1.244744873046875,
"step": 4545
},
{
"epoch": 0.8977014895925817,
"grad_norm": 0.6633027195930481,
"learning_rate": 2.735410149869838e-06,
"loss": 1.2772028923034668,
"step": 4550
},
{
"epoch": 0.8986879747459801,
"grad_norm": 0.6178504824638367,
"learning_rate": 2.6835324544256967e-06,
"loss": 1.287399959564209,
"step": 4555
},
{
"epoch": 0.8996744598993786,
"grad_norm": 0.8301539421081543,
"learning_rate": 2.6321378488736504e-06,
"loss": 1.2425086975097657,
"step": 4560
},
{
"epoch": 0.9006609450527769,
"grad_norm": 0.6057320833206177,
"learning_rate": 2.5812268579403996e-06,
"loss": 1.2323110580444336,
"step": 4565
},
{
"epoch": 0.9016474302061754,
"grad_norm": 0.5955671072006226,
"learning_rate": 2.530800001415079e-06,
"loss": 1.3139602661132812,
"step": 4570
},
{
"epoch": 0.9026339153595738,
"grad_norm": 1.0247668027877808,
"learning_rate": 2.480857794143898e-06,
"loss": 1.3500228881835938,
"step": 4575
},
{
"epoch": 0.9036204005129723,
"grad_norm": 0.5576156973838806,
"learning_rate": 2.4314007460249312e-06,
"loss": 1.1277143478393554,
"step": 4580
},
{
"epoch": 0.9046068856663707,
"grad_norm": 0.8178988099098206,
"learning_rate": 2.382429362002908e-06,
"loss": 1.2583866119384766,
"step": 4585
},
{
"epoch": 0.9055933708197692,
"grad_norm": 0.7907400727272034,
"learning_rate": 2.3339441420640095e-06,
"loss": 1.264080810546875,
"step": 4590
},
{
"epoch": 0.9065798559731676,
"grad_norm": 0.6387272477149963,
"learning_rate": 2.285945581230814e-06,
"loss": 1.2724645614624024,
"step": 4595
},
{
"epoch": 0.9075663411265661,
"grad_norm": 0.7053713202476501,
"learning_rate": 2.238434169557213e-06,
"loss": 1.1759965896606446,
"step": 4600
},
{
"epoch": 0.9085528262799645,
"grad_norm": 0.7945536375045776,
"learning_rate": 2.191410392123444e-06,
"loss": 1.243608570098877,
"step": 4605
},
{
"epoch": 0.909539311433363,
"grad_norm": 0.9395650625228882,
"learning_rate": 2.14487472903111e-06,
"loss": 1.2648731231689454,
"step": 4610
},
{
"epoch": 0.9105257965867614,
"grad_norm": 0.5654191374778748,
"learning_rate": 2.0988276553982556e-06,
"loss": 1.2391172409057618,
"step": 4615
},
{
"epoch": 0.9115122817401599,
"grad_norm": 0.9005200266838074,
"learning_rate": 2.0532696413545826e-06,
"loss": 1.256478214263916,
"step": 4620
},
{
"epoch": 0.9124987668935582,
"grad_norm": 0.6227542757987976,
"learning_rate": 2.0082011520365707e-06,
"loss": 1.1761054992675781,
"step": 4625
},
{
"epoch": 0.9134852520469567,
"grad_norm": 0.5829877853393555,
"learning_rate": 1.963622647582808e-06,
"loss": 1.156816291809082,
"step": 4630
},
{
"epoch": 0.9144717372003551,
"grad_norm": 0.6080260872840881,
"learning_rate": 1.919534583129223e-06,
"loss": 1.3070613861083984,
"step": 4635
},
{
"epoch": 0.9154582223537536,
"grad_norm": 1.1772451400756836,
"learning_rate": 1.8759374088044734e-06,
"loss": 1.1401609420776366,
"step": 4640
},
{
"epoch": 0.916444707507152,
"grad_norm": 0.5841192007064819,
"learning_rate": 1.8328315697253684e-06,
"loss": 1.1951543807983398,
"step": 4645
},
{
"epoch": 0.9174311926605505,
"grad_norm": 0.731962263584137,
"learning_rate": 1.7902175059922643e-06,
"loss": 1.2066067695617675,
"step": 4650
},
{
"epoch": 0.9184176778139489,
"grad_norm": 0.7481285333633423,
"learning_rate": 1.7480956526846393e-06,
"loss": 1.2765756607055665,
"step": 4655
},
{
"epoch": 0.9194041629673474,
"grad_norm": 0.8645247220993042,
"learning_rate": 1.7064664398566133e-06,
"loss": 1.1822599411010741,
"step": 4660
},
{
"epoch": 0.9203906481207458,
"grad_norm": 0.6532875299453735,
"learning_rate": 1.665330292532552e-06,
"loss": 1.1630514144897461,
"step": 4665
},
{
"epoch": 0.9213771332741443,
"grad_norm": 0.6010720133781433,
"learning_rate": 1.624687630702737e-06,
"loss": 1.1973966598510741,
"step": 4670
},
{
"epoch": 0.9223636184275427,
"grad_norm": 0.6148731708526611,
"learning_rate": 1.5845388693191022e-06,
"loss": 1.2460204124450684,
"step": 4675
},
{
"epoch": 0.9233501035809412,
"grad_norm": 0.896836519241333,
"learning_rate": 1.5448844182909595e-06,
"loss": 1.207472801208496,
"step": 4680
},
{
"epoch": 0.9243365887343395,
"grad_norm": 0.6151934862136841,
"learning_rate": 1.5057246824808358e-06,
"loss": 1.1990650177001954,
"step": 4685
},
{
"epoch": 0.9253230738877379,
"grad_norm": 0.7115896344184875,
"learning_rate": 1.4670600617003205e-06,
"loss": 1.1894463539123534,
"step": 4690
},
{
"epoch": 0.9263095590411364,
"grad_norm": 0.5992312431335449,
"learning_rate": 1.4288909507060134e-06,
"loss": 1.11309757232666,
"step": 4695
},
{
"epoch": 0.9272960441945348,
"grad_norm": 0.8634612560272217,
"learning_rate": 1.3912177391954717e-06,
"loss": 1.198754119873047,
"step": 4700
},
{
"epoch": 0.9282825293479333,
"grad_norm": 1.1952672004699707,
"learning_rate": 1.3540408118032255e-06,
"loss": 1.089672565460205,
"step": 4705
},
{
"epoch": 0.9292690145013317,
"grad_norm": 0.6822053790092468,
"learning_rate": 1.3173605480968854e-06,
"loss": 1.2104829788208007,
"step": 4710
},
{
"epoch": 0.9302554996547302,
"grad_norm": 0.6624516248703003,
"learning_rate": 1.2811773225732071e-06,
"loss": 1.3110162734985351,
"step": 4715
},
{
"epoch": 0.9312419848081286,
"grad_norm": 0.6583454012870789,
"learning_rate": 1.2454915046543447e-06,
"loss": 1.1744534492492675,
"step": 4720
},
{
"epoch": 0.9322284699615271,
"grad_norm": 0.675237238407135,
"learning_rate": 1.2103034586840201e-06,
"loss": 1.2287273406982422,
"step": 4725
},
{
"epoch": 0.9332149551149255,
"grad_norm": 0.684691309928894,
"learning_rate": 1.1756135439238092e-06,
"loss": 1.222042179107666,
"step": 4730
},
{
"epoch": 0.934201440268324,
"grad_norm": 0.6616474986076355,
"learning_rate": 1.1414221145495174e-06,
"loss": 1.2868342399597168,
"step": 4735
},
{
"epoch": 0.9351879254217224,
"grad_norm": 0.893924355506897,
"learning_rate": 1.1077295196475046e-06,
"loss": 1.27929744720459,
"step": 4740
},
{
"epoch": 0.9361744105751209,
"grad_norm": 0.9433838725090027,
"learning_rate": 1.07453610321116e-06,
"loss": 1.1580618858337401,
"step": 4745
},
{
"epoch": 0.9371608957285192,
"grad_norm": 0.7204527258872986,
"learning_rate": 1.0418422041373887e-06,
"loss": 1.2097747802734375,
"step": 4750
},
{
"epoch": 0.9381473808819177,
"grad_norm": 0.59828120470047,
"learning_rate": 1.0096481562231142e-06,
"loss": 1.1047725677490234,
"step": 4755
},
{
"epoch": 0.9391338660353161,
"grad_norm": 0.555549144744873,
"learning_rate": 9.779542881619364e-07,
"loss": 1.1881049156188965,
"step": 4760
},
{
"epoch": 0.9401203511887146,
"grad_norm": 0.7596853375434875,
"learning_rate": 9.467609235407071e-07,
"loss": 1.1742735862731934,
"step": 4765
},
{
"epoch": 0.941106836342113,
"grad_norm": 0.7403998970985413,
"learning_rate": 9.160683808362714e-07,
"loss": 1.253631591796875,
"step": 4770
},
{
"epoch": 0.9420933214955115,
"grad_norm": 0.7962865233421326,
"learning_rate": 8.85876973412203e-07,
"loss": 1.2446669578552245,
"step": 4775
},
{
"epoch": 0.9430798066489099,
"grad_norm": 0.5877761244773865,
"learning_rate": 8.561870095155855e-07,
"loss": 1.1364049911499023,
"step": 4780
},
{
"epoch": 0.9440662918023084,
"grad_norm": 0.610772967338562,
"learning_rate": 8.269987922739087e-07,
"loss": 1.2133082389831542,
"step": 4785
},
{
"epoch": 0.9450527769557068,
"grad_norm": 0.6112104654312134,
"learning_rate": 7.983126196919377e-07,
"loss": 1.2320796966552734,
"step": 4790
},
{
"epoch": 0.9460392621091053,
"grad_norm": 0.5719399452209473,
"learning_rate": 7.70128784648666e-07,
"loss": 1.2707351684570312,
"step": 4795
},
{
"epoch": 0.9470257472625037,
"grad_norm": 0.7434406876564026,
"learning_rate": 7.424475748943671e-07,
"loss": 1.2662650108337403,
"step": 4800
},
{
"epoch": 0.9480122324159022,
"grad_norm": 1.0301364660263062,
"learning_rate": 7.152692730476029e-07,
"loss": 1.1646380424499512,
"step": 4805
},
{
"epoch": 0.9489987175693005,
"grad_norm": 0.6889775991439819,
"learning_rate": 6.885941565923925e-07,
"loss": 1.1487475395202638,
"step": 4810
},
{
"epoch": 0.949985202722699,
"grad_norm": 0.666320264339447,
"learning_rate": 6.624224978753257e-07,
"loss": 1.2685041427612305,
"step": 4815
},
{
"epoch": 0.9509716878760974,
"grad_norm": 0.7730403542518616,
"learning_rate": 6.367545641028261e-07,
"loss": 1.2028122901916505,
"step": 4820
},
{
"epoch": 0.9519581730294959,
"grad_norm": 0.7139598727226257,
"learning_rate": 6.115906173384034e-07,
"loss": 1.2060637474060059,
"step": 4825
},
{
"epoch": 0.9529446581828943,
"grad_norm": 1.3035459518432617,
"learning_rate": 5.869309144999613e-07,
"loss": 1.2906496047973632,
"step": 4830
},
{
"epoch": 0.9539311433362928,
"grad_norm": 0.6317528486251831,
"learning_rate": 5.627757073572159e-07,
"loss": 1.1726470947265626,
"step": 4835
},
{
"epoch": 0.9549176284896912,
"grad_norm": 0.9931450486183167,
"learning_rate": 5.391252425290982e-07,
"loss": 1.2547016143798828,
"step": 4840
},
{
"epoch": 0.9559041136430897,
"grad_norm": 0.64718097448349,
"learning_rate": 5.159797614812279e-07,
"loss": 1.2418943405151368,
"step": 4845
},
{
"epoch": 0.9568905987964881,
"grad_norm": 0.6341720819473267,
"learning_rate": 4.933395005234708e-07,
"loss": 1.1842002868652344,
"step": 4850
},
{
"epoch": 0.9578770839498866,
"grad_norm": 1.0610328912734985,
"learning_rate": 4.712046908075085e-07,
"loss": 1.2033771514892577,
"step": 4855
},
{
"epoch": 0.958863569103285,
"grad_norm": 0.7860101461410522,
"learning_rate": 4.4957555832449426e-07,
"loss": 1.2387046813964844,
"step": 4860
},
{
"epoch": 0.9598500542566835,
"grad_norm": 0.8293290734291077,
"learning_rate": 4.2845232390273384e-07,
"loss": 1.202822780609131,
"step": 4865
},
{
"epoch": 0.9608365394100818,
"grad_norm": 0.9116617441177368,
"learning_rate": 4.0783520320542e-07,
"loss": 1.151958179473877,
"step": 4870
},
{
"epoch": 0.9618230245634803,
"grad_norm": 0.6855465769767761,
"learning_rate": 3.8772440672847356e-07,
"loss": 1.244114875793457,
"step": 4875
},
{
"epoch": 0.9628095097168787,
"grad_norm": 0.7231633067131042,
"learning_rate": 3.681201397983336e-07,
"loss": 1.2055521965026856,
"step": 4880
},
{
"epoch": 0.9637959948702772,
"grad_norm": 0.6800587773323059,
"learning_rate": 3.4902260256991505e-07,
"loss": 1.281911277770996,
"step": 4885
},
{
"epoch": 0.9647824800236756,
"grad_norm": 0.7391085028648376,
"learning_rate": 3.3043199002453785e-07,
"loss": 1.3071533203125,
"step": 4890
},
{
"epoch": 0.9657689651770741,
"grad_norm": 0.6460412740707397,
"learning_rate": 3.1234849196793424e-07,
"loss": 1.2184791564941406,
"step": 4895
},
{
"epoch": 0.9667554503304725,
"grad_norm": 0.9166705012321472,
"learning_rate": 2.947722930283281e-07,
"loss": 1.2178396224975585,
"step": 4900
},
{
"epoch": 0.967741935483871,
"grad_norm": 0.8080833554267883,
"learning_rate": 2.777035726545307e-07,
"loss": 1.1572997093200683,
"step": 4905
},
{
"epoch": 0.9687284206372694,
"grad_norm": 0.7032297253608704,
"learning_rate": 2.6114250511412033e-07,
"loss": 1.1382969856262206,
"step": 4910
},
{
"epoch": 0.9697149057906679,
"grad_norm": 0.7416732907295227,
"learning_rate": 2.450892594916654e-07,
"loss": 1.250673484802246,
"step": 4915
},
{
"epoch": 0.9707013909440663,
"grad_norm": 0.7424551844596863,
"learning_rate": 2.2954399968699304e-07,
"loss": 1.1140318870544434,
"step": 4920
},
{
"epoch": 0.9716878760974648,
"grad_norm": 0.7498700618743896,
"learning_rate": 2.1450688441350675e-07,
"loss": 1.1579265594482422,
"step": 4925
},
{
"epoch": 0.9726743612508632,
"grad_norm": 0.655081570148468,
"learning_rate": 1.999780671965823e-07,
"loss": 1.2060396194458007,
"step": 4930
},
{
"epoch": 0.9736608464042616,
"grad_norm": 0.7647750973701477,
"learning_rate": 1.8595769637199113e-07,
"loss": 1.2389888763427734,
"step": 4935
},
{
"epoch": 0.97464733155766,
"grad_norm": 0.9867916703224182,
"learning_rate": 1.7244591508439047e-07,
"loss": 1.2211369514465331,
"step": 4940
},
{
"epoch": 0.9756338167110585,
"grad_norm": 0.6398246884346008,
"learning_rate": 1.5944286128585785e-07,
"loss": 1.190547561645508,
"step": 4945
},
{
"epoch": 0.9766203018644569,
"grad_norm": 0.6152133345603943,
"learning_rate": 1.4694866773448113e-07,
"loss": 1.1919227600097657,
"step": 4950
},
{
"epoch": 0.9776067870178554,
"grad_norm": 0.6231676936149597,
"learning_rate": 1.3496346199300403e-07,
"loss": 1.2857358932495118,
"step": 4955
},
{
"epoch": 0.9785932721712538,
"grad_norm": 0.6220934987068176,
"learning_rate": 1.2348736642753822e-07,
"loss": 1.204495620727539,
"step": 4960
},
{
"epoch": 0.9795797573246523,
"grad_norm": 0.6293838024139404,
"learning_rate": 1.1252049820629219e-07,
"loss": 1.2589128494262696,
"step": 4965
},
{
"epoch": 0.9805662424780507,
"grad_norm": 0.6287444233894348,
"learning_rate": 1.0206296929837212e-07,
"loss": 1.1795320510864258,
"step": 4970
},
{
"epoch": 0.9815527276314492,
"grad_norm": 0.7871685028076172,
"learning_rate": 9.211488647268284e-08,
"loss": 1.1962526321411133,
"step": 4975
},
{
"epoch": 0.9825392127848476,
"grad_norm": 0.7378242015838623,
"learning_rate": 8.267635129677321e-08,
"loss": 1.1561376571655273,
"step": 4980
},
{
"epoch": 0.9835256979382461,
"grad_norm": 0.7698005437850952,
"learning_rate": 7.374746013584788e-08,
"loss": 1.2471741676330566,
"step": 4985
},
{
"epoch": 0.9845121830916445,
"grad_norm": 0.6808744668960571,
"learning_rate": 6.532830415176273e-08,
"loss": 1.218988037109375,
"step": 4990
},
{
"epoch": 0.985498668245043,
"grad_norm": 0.898345410823822,
"learning_rate": 5.741896930210322e-08,
"loss": 1.131629467010498,
"step": 4995
},
{
"epoch": 0.9864851533984413,
"grad_norm": 0.6738277077674866,
"learning_rate": 5.0019536339301895e-08,
"loss": 1.2708320617675781,
"step": 5000
},
{
"epoch": 0.9874716385518398,
"grad_norm": 0.6729507446289062,
"learning_rate": 4.3130080809805626e-08,
"loss": 1.1585707664489746,
"step": 5005
},
{
"epoch": 0.9884581237052382,
"grad_norm": 0.6317814588546753,
"learning_rate": 3.675067305332069e-08,
"loss": 1.1529539108276368,
"step": 5010
},
{
"epoch": 0.9894446088586367,
"grad_norm": 1.0456939935684204,
"learning_rate": 3.088137820208004e-08,
"loss": 1.2157886505126954,
"step": 5015
},
{
"epoch": 0.9904310940120351,
"grad_norm": 0.6216485500335693,
"learning_rate": 2.5522256180199322e-08,
"loss": 1.3058615684509278,
"step": 5020
},
{
"epoch": 0.9914175791654336,
"grad_norm": 0.6462249159812927,
"learning_rate": 2.0673361703033024e-08,
"loss": 1.2741954803466797,
"step": 5025
},
{
"epoch": 0.992404064318832,
"grad_norm": 0.7657465934753418,
"learning_rate": 1.633474427664705e-08,
"loss": 1.2003997802734374,
"step": 5030
},
{
"epoch": 0.9933905494722305,
"grad_norm": 0.7415722012519836,
"learning_rate": 1.2506448197291409e-08,
"loss": 1.1379487991333008,
"step": 5035
},
{
"epoch": 0.9943770346256289,
"grad_norm": 0.8890054821968079,
"learning_rate": 9.188512550956096e-09,
"loss": 1.2447555541992188,
"step": 5040
},
{
"epoch": 0.9953635197790274,
"grad_norm": 0.8467966318130493,
"learning_rate": 6.380971212988085e-09,
"loss": 1.2432153701782227,
"step": 5045
},
{
"epoch": 0.9963500049324258,
"grad_norm": 0.6697431206703186,
"learning_rate": 4.0838528477027406e-09,
"loss": 1.252425193786621,
"step": 5050
},
{
"epoch": 0.9973364900858243,
"grad_norm": 0.7170816659927368,
"learning_rate": 2.2971809081395734e-09,
"loss": 1.2265609741210937,
"step": 5055
},
{
"epoch": 0.9983229752392226,
"grad_norm": 0.5926438570022583,
"learning_rate": 1.0209736357957856e-09,
"loss": 1.1267782211303712,
"step": 5060
},
{
"epoch": 0.9993094603926211,
"grad_norm": 0.7526215314865112,
"learning_rate": 2.552440604430828e-10,
"loss": 1.1800277709960938,
"step": 5065
}
],
"logging_steps": 5,
"max_steps": 5069,
"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": 3.0115375490688614e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}