rqadri's picture
Upload folder using huggingface_hub
a4c5268 verified
Raw History Blame Contribute Delete
49.2 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.21255414545178206,
"eval_steps": 500,
"global_step": 2797,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0007599361653621095,
"grad_norm": 191.63430175303478,
"learning_rate": 1.1061797752808989e-06,
"loss": 2.883259582519531,
"step": 10
},
{
"epoch": 0.001519872330724219,
"grad_norm": 124.58368761068755,
"learning_rate": 2.2241573033707866e-06,
"loss": 1.8781267166137696,
"step": 20
},
{
"epoch": 0.0022798084960863286,
"grad_norm": 87.1340034334807,
"learning_rate": 3.342134831460675e-06,
"loss": 1.1774589538574218,
"step": 30
},
{
"epoch": 0.003039744661448438,
"grad_norm": 844.0347817685458,
"learning_rate": 4.4601123595505624e-06,
"loss": 0.9150102615356446,
"step": 40
},
{
"epoch": 0.0037996808268105477,
"grad_norm": 26.071347281400854,
"learning_rate": 5.5780898876404495e-06,
"loss": 0.7994430065155029,
"step": 50
},
{
"epoch": 0.004559616992172657,
"grad_norm": 106.14143580914507,
"learning_rate": 6.6960674157303374e-06,
"loss": 0.7530351161956788,
"step": 60
},
{
"epoch": 0.005319553157534767,
"grad_norm": 150.7760094384226,
"learning_rate": 7.814044943820226e-06,
"loss": 0.7612151622772216,
"step": 70
},
{
"epoch": 0.006079489322896876,
"grad_norm": 17.649974068212234,
"learning_rate": 8.932022471910112e-06,
"loss": 0.7631851196289062,
"step": 80
},
{
"epoch": 0.006839425488258986,
"grad_norm": 74.83280873477604,
"learning_rate": 1.005e-05,
"loss": 0.7748248100280761,
"step": 90
},
{
"epoch": 0.007599361653621095,
"grad_norm": 923.6270136131341,
"learning_rate": 1.1167977528089888e-05,
"loss": 0.7622076988220214,
"step": 100
},
{
"epoch": 0.008359297818983206,
"grad_norm": 191.99807923455484,
"learning_rate": 1.2285955056179776e-05,
"loss": 0.7419852256774903,
"step": 110
},
{
"epoch": 0.009119233984345314,
"grad_norm": 8.180188792042815,
"learning_rate": 1.3403932584269664e-05,
"loss": 0.7526190757751465,
"step": 120
},
{
"epoch": 0.009879170149707425,
"grad_norm": 6.875791274720428,
"learning_rate": 1.4521910112359552e-05,
"loss": 0.7423999309539795,
"step": 130
},
{
"epoch": 0.010639106315069534,
"grad_norm": 4.139161233642084,
"learning_rate": 1.563988764044944e-05,
"loss": 0.736327075958252,
"step": 140
},
{
"epoch": 0.011399042480431644,
"grad_norm": 22.480724459482204,
"learning_rate": 1.6757865168539328e-05,
"loss": 0.7192594528198242,
"step": 150
},
{
"epoch": 0.012158978645793753,
"grad_norm": 42.20418057032992,
"learning_rate": 1.7875842696629214e-05,
"loss": 0.7199824333190918,
"step": 160
},
{
"epoch": 0.012918914811155863,
"grad_norm": 22.176716309419565,
"learning_rate": 1.89938202247191e-05,
"loss": 0.7676156520843506,
"step": 170
},
{
"epoch": 0.013678850976517972,
"grad_norm": 39.06503779613347,
"learning_rate": 2e-05,
"loss": 0.723706579208374,
"step": 180
},
{
"epoch": 0.014438787141880082,
"grad_norm": 72.77156957569211,
"learning_rate": 2e-05,
"loss": 0.7594737052917481,
"step": 190
},
{
"epoch": 0.01519872330724219,
"grad_norm": 121.81035558320448,
"learning_rate": 2e-05,
"loss": 0.7727550506591797,
"step": 200
},
{
"epoch": 0.015958659472604303,
"grad_norm": 17.261218474875754,
"learning_rate": 2e-05,
"loss": 0.7319492340087891,
"step": 210
},
{
"epoch": 0.01671859563796641,
"grad_norm": 13.6677607141103,
"learning_rate": 2e-05,
"loss": 0.7261901378631592,
"step": 220
},
{
"epoch": 0.01747853180332852,
"grad_norm": 19.87687501769399,
"learning_rate": 2e-05,
"loss": 0.7509442806243897,
"step": 230
},
{
"epoch": 0.01823846796869063,
"grad_norm": 71.02797556619042,
"learning_rate": 2e-05,
"loss": 0.7291659355163574,
"step": 240
},
{
"epoch": 0.01899840413405274,
"grad_norm": 7.361788981118384,
"learning_rate": 2e-05,
"loss": 0.7559044837951661,
"step": 250
},
{
"epoch": 0.01975834029941485,
"grad_norm": 13.834043992751518,
"learning_rate": 2e-05,
"loss": 0.7548636436462403,
"step": 260
},
{
"epoch": 0.02051827646477696,
"grad_norm": 18.40212932805268,
"learning_rate": 2e-05,
"loss": 0.7729722499847412,
"step": 270
},
{
"epoch": 0.021278212630139067,
"grad_norm": 3.7670360731030317,
"learning_rate": 2e-05,
"loss": 0.7427117824554443,
"step": 280
},
{
"epoch": 0.02203814879550118,
"grad_norm": 3.2551327080904002,
"learning_rate": 2e-05,
"loss": 0.696455717086792,
"step": 290
},
{
"epoch": 0.022798084960863288,
"grad_norm": 12.681218354218718,
"learning_rate": 2e-05,
"loss": 0.7120409965515136,
"step": 300
},
{
"epoch": 0.023558021126225397,
"grad_norm": 54.34344654534163,
"learning_rate": 2e-05,
"loss": 0.6876229286193848,
"step": 310
},
{
"epoch": 0.024317957291587505,
"grad_norm": 5.6674147357259415,
"learning_rate": 2e-05,
"loss": 0.7949447631835938,
"step": 320
},
{
"epoch": 0.025077893456949617,
"grad_norm": 4.542555847806229,
"learning_rate": 2e-05,
"loss": 0.7002223014831543,
"step": 330
},
{
"epoch": 0.025837829622311726,
"grad_norm": 6.420887665016644,
"learning_rate": 2e-05,
"loss": 0.74368314743042,
"step": 340
},
{
"epoch": 0.026597765787673835,
"grad_norm": 66.78918732414809,
"learning_rate": 2e-05,
"loss": 0.7450421810150146,
"step": 350
},
{
"epoch": 0.027357701953035943,
"grad_norm": 8.341774953333017,
"learning_rate": 2e-05,
"loss": 0.7264707565307618,
"step": 360
},
{
"epoch": 0.028117638118398056,
"grad_norm": 2.4168868991630457,
"learning_rate": 2e-05,
"loss": 0.7024238586425782,
"step": 370
},
{
"epoch": 0.028877574283760164,
"grad_norm": 4.431590851345387,
"learning_rate": 2e-05,
"loss": 0.7049625396728516,
"step": 380
},
{
"epoch": 0.029637510449122273,
"grad_norm": 36.43429145728405,
"learning_rate": 2e-05,
"loss": 0.7121419429779052,
"step": 390
},
{
"epoch": 0.03039744661448438,
"grad_norm": 8.501774499445727,
"learning_rate": 2e-05,
"loss": 0.7381654262542725,
"step": 400
},
{
"epoch": 0.031157382779846494,
"grad_norm": 2.1284121866050194,
"learning_rate": 2e-05,
"loss": 0.7352887153625488,
"step": 410
},
{
"epoch": 0.031917318945208606,
"grad_norm": 1.6959741724878892,
"learning_rate": 2e-05,
"loss": 0.709722900390625,
"step": 420
},
{
"epoch": 0.032677255110570715,
"grad_norm": 2.777848797318434,
"learning_rate": 2e-05,
"loss": 0.6909010887145997,
"step": 430
},
{
"epoch": 0.03343719127593282,
"grad_norm": 5.759466391991063,
"learning_rate": 2e-05,
"loss": 0.6841095924377442,
"step": 440
},
{
"epoch": 0.03419712744129493,
"grad_norm": 1.9894311227745336,
"learning_rate": 2e-05,
"loss": 0.6729912757873535,
"step": 450
},
{
"epoch": 0.03495706360665704,
"grad_norm": 2.313482340982159,
"learning_rate": 2e-05,
"loss": 0.6638118743896484,
"step": 460
},
{
"epoch": 0.03571699977201915,
"grad_norm": 7.013232779450187,
"learning_rate": 2e-05,
"loss": 0.6692584037780762,
"step": 470
},
{
"epoch": 0.03647693593738126,
"grad_norm": 2.574330643991075,
"learning_rate": 2e-05,
"loss": 0.70731201171875,
"step": 480
},
{
"epoch": 0.03723687210274337,
"grad_norm": 1.5052842559853645,
"learning_rate": 2e-05,
"loss": 0.6873155117034913,
"step": 490
},
{
"epoch": 0.03799680826810548,
"grad_norm": 1.116178924419754,
"learning_rate": 2e-05,
"loss": 0.669003677368164,
"step": 500
},
{
"epoch": 0.03875674443346759,
"grad_norm": 1.938914338338345,
"learning_rate": 2e-05,
"loss": 0.69815092086792,
"step": 510
},
{
"epoch": 0.0395166805988297,
"grad_norm": 1.3811669929651738,
"learning_rate": 2e-05,
"loss": 0.712796401977539,
"step": 520
},
{
"epoch": 0.04027661676419181,
"grad_norm": 1.091199227854418,
"learning_rate": 2e-05,
"loss": 0.6600155830383301,
"step": 530
},
{
"epoch": 0.04103655292955392,
"grad_norm": 1.105571901275732,
"learning_rate": 2e-05,
"loss": 0.6982758998870849,
"step": 540
},
{
"epoch": 0.041796489094916026,
"grad_norm": 0.9941727634113782,
"learning_rate": 2e-05,
"loss": 0.6900170803070068,
"step": 550
},
{
"epoch": 0.042556425260278134,
"grad_norm": 0.9835023836804377,
"learning_rate": 2e-05,
"loss": 0.6558903694152832,
"step": 560
},
{
"epoch": 0.04331636142564024,
"grad_norm": 1.1595810180524677,
"learning_rate": 2e-05,
"loss": 0.6865039825439453,
"step": 570
},
{
"epoch": 0.04407629759100236,
"grad_norm": 2.2186915798904154,
"learning_rate": 2e-05,
"loss": 0.6749402999877929,
"step": 580
},
{
"epoch": 0.04483623375636447,
"grad_norm": 1.3118057884159302,
"learning_rate": 2e-05,
"loss": 0.675851583480835,
"step": 590
},
{
"epoch": 0.045596169921726576,
"grad_norm": 1.692496987832561,
"learning_rate": 2e-05,
"loss": 0.6745028495788574,
"step": 600
},
{
"epoch": 0.046356106087088685,
"grad_norm": 1.9730825283501034,
"learning_rate": 2e-05,
"loss": 0.6651962280273438,
"step": 610
},
{
"epoch": 0.04711604225245079,
"grad_norm": 1.331976977887419,
"learning_rate": 2e-05,
"loss": 0.6501778602600098,
"step": 620
},
{
"epoch": 0.0478759784178129,
"grad_norm": 1.5637084886303765,
"learning_rate": 2e-05,
"loss": 0.7042662620544433,
"step": 630
},
{
"epoch": 0.04863591458317501,
"grad_norm": 1.4726774498399116,
"learning_rate": 2e-05,
"loss": 0.6678030967712403,
"step": 640
},
{
"epoch": 0.049395850748537126,
"grad_norm": 2.868807910273408,
"learning_rate": 2e-05,
"loss": 0.6779534339904785,
"step": 650
},
{
"epoch": 0.050155786913899235,
"grad_norm": 1.4317134087883574,
"learning_rate": 2e-05,
"loss": 0.6799858093261719,
"step": 660
},
{
"epoch": 0.050915723079261344,
"grad_norm": 1.7227333001916365,
"learning_rate": 2e-05,
"loss": 0.6573725700378418,
"step": 670
},
{
"epoch": 0.05167565924462345,
"grad_norm": 1.3271414274442734,
"learning_rate": 2e-05,
"loss": 0.6701514244079589,
"step": 680
},
{
"epoch": 0.05243559540998556,
"grad_norm": 1.5421124248057487,
"learning_rate": 2e-05,
"loss": 0.6800753593444824,
"step": 690
},
{
"epoch": 0.05319553157534767,
"grad_norm": 1.0972092785185035,
"learning_rate": 2e-05,
"loss": 0.6628331184387207,
"step": 700
},
{
"epoch": 0.05395546774070978,
"grad_norm": 1.8016898257371647,
"learning_rate": 2e-05,
"loss": 0.6686877250671387,
"step": 710
},
{
"epoch": 0.05471540390607189,
"grad_norm": 1.3817551649658868,
"learning_rate": 2e-05,
"loss": 0.691933012008667,
"step": 720
},
{
"epoch": 0.055475340071434,
"grad_norm": 4.390986381327533,
"learning_rate": 2e-05,
"loss": 0.6459454536437989,
"step": 730
},
{
"epoch": 0.05623527623679611,
"grad_norm": 1.049270882156613,
"learning_rate": 2e-05,
"loss": 0.7004701137542725,
"step": 740
},
{
"epoch": 0.05699521240215822,
"grad_norm": 2.5788017211289134,
"learning_rate": 2e-05,
"loss": 0.6631406784057617,
"step": 750
},
{
"epoch": 0.05775514856752033,
"grad_norm": 4.003076354285841,
"learning_rate": 2e-05,
"loss": 0.6525260925292968,
"step": 760
},
{
"epoch": 0.05851508473288244,
"grad_norm": 3.856641478939966,
"learning_rate": 2e-05,
"loss": 0.6481318950653077,
"step": 770
},
{
"epoch": 0.059275020898244546,
"grad_norm": 2.1050201546733516,
"learning_rate": 2e-05,
"loss": 0.6629796504974366,
"step": 780
},
{
"epoch": 0.060034957063606655,
"grad_norm": 2.4972482058867684,
"learning_rate": 2e-05,
"loss": 0.6793940544128418,
"step": 790
},
{
"epoch": 0.06079489322896876,
"grad_norm": 2.9023733973854338,
"learning_rate": 2e-05,
"loss": 0.6557552337646484,
"step": 800
},
{
"epoch": 0.06155482939433088,
"grad_norm": 1.753191005851473,
"learning_rate": 2e-05,
"loss": 0.6550214767456055,
"step": 810
},
{
"epoch": 0.06231476555969299,
"grad_norm": 3.6008366995268677,
"learning_rate": 2e-05,
"loss": 0.6532149314880371,
"step": 820
},
{
"epoch": 0.0630747017250551,
"grad_norm": 3.9574066156632264,
"learning_rate": 2e-05,
"loss": 0.664048147201538,
"step": 830
},
{
"epoch": 0.06383463789041721,
"grad_norm": 1.4500186208031258,
"learning_rate": 2e-05,
"loss": 0.668964147567749,
"step": 840
},
{
"epoch": 0.06459457405577931,
"grad_norm": 1.8650680203909065,
"learning_rate": 2e-05,
"loss": 0.6597201824188232,
"step": 850
},
{
"epoch": 0.06535451022114143,
"grad_norm": 1.286387371803429,
"learning_rate": 2e-05,
"loss": 0.6578952312469483,
"step": 860
},
{
"epoch": 0.06611444638650353,
"grad_norm": 44.161007706426496,
"learning_rate": 2e-05,
"loss": 0.6617650508880615,
"step": 870
},
{
"epoch": 0.06687438255186565,
"grad_norm": 8.70858945322814,
"learning_rate": 2e-05,
"loss": 0.6729881286621093,
"step": 880
},
{
"epoch": 0.06763431871722775,
"grad_norm": 2.0716744352445295,
"learning_rate": 2e-05,
"loss": 0.6804790496826172,
"step": 890
},
{
"epoch": 0.06839425488258986,
"grad_norm": 2.3995890552247063,
"learning_rate": 2e-05,
"loss": 0.6576640605926514,
"step": 900
},
{
"epoch": 0.06915419104795197,
"grad_norm": 7.020838251383462,
"learning_rate": 2e-05,
"loss": 0.6924066066741943,
"step": 910
},
{
"epoch": 0.06991412721331408,
"grad_norm": 1.2111035584957595,
"learning_rate": 2e-05,
"loss": 0.6660052299499511,
"step": 920
},
{
"epoch": 0.0706740633786762,
"grad_norm": 1.8673454485620073,
"learning_rate": 2e-05,
"loss": 0.6890796661376953,
"step": 930
},
{
"epoch": 0.0714339995440383,
"grad_norm": 2.462929162354763,
"learning_rate": 2e-05,
"loss": 0.6732182979583741,
"step": 940
},
{
"epoch": 0.07219393570940041,
"grad_norm": 1.8507788704761199,
"learning_rate": 2e-05,
"loss": 0.6905992031097412,
"step": 950
},
{
"epoch": 0.07295387187476252,
"grad_norm": 1.5233611761953871,
"learning_rate": 2e-05,
"loss": 0.6639741897583008,
"step": 960
},
{
"epoch": 0.07371380804012463,
"grad_norm": 1.0039057924990893,
"learning_rate": 2e-05,
"loss": 0.6618095397949219,
"step": 970
},
{
"epoch": 0.07447374420548673,
"grad_norm": 1.59146144447183,
"learning_rate": 2e-05,
"loss": 0.6541851043701172,
"step": 980
},
{
"epoch": 0.07523368037084885,
"grad_norm": 1.671042860219738,
"learning_rate": 2e-05,
"loss": 0.6654860496520996,
"step": 990
},
{
"epoch": 0.07599361653621096,
"grad_norm": 1.198120650365706,
"learning_rate": 2e-05,
"loss": 0.6552873134613038,
"step": 1000
},
{
"epoch": 0.07675355270157307,
"grad_norm": 8.77583610709108,
"learning_rate": 2e-05,
"loss": 0.6433052062988281,
"step": 1010
},
{
"epoch": 0.07751348886693518,
"grad_norm": 1.111330096476852,
"learning_rate": 2e-05,
"loss": 0.642775297164917,
"step": 1020
},
{
"epoch": 0.07827342503229728,
"grad_norm": 1.348192309325266,
"learning_rate": 2e-05,
"loss": 0.6692739009857178,
"step": 1030
},
{
"epoch": 0.0790333611976594,
"grad_norm": 1.0407807523199857,
"learning_rate": 2e-05,
"loss": 0.6457259178161621,
"step": 1040
},
{
"epoch": 0.0797932973630215,
"grad_norm": 1.0582735573079416,
"learning_rate": 2e-05,
"loss": 0.6610038757324219,
"step": 1050
},
{
"epoch": 0.08055323352838362,
"grad_norm": 2.1749317446696894,
"learning_rate": 2e-05,
"loss": 0.6602062225341797,
"step": 1060
},
{
"epoch": 0.08131316969374573,
"grad_norm": 1.2523925736905532,
"learning_rate": 2e-05,
"loss": 0.6483442306518554,
"step": 1070
},
{
"epoch": 0.08207310585910783,
"grad_norm": 1.2772297314001628,
"learning_rate": 2e-05,
"loss": 0.6683950424194336,
"step": 1080
},
{
"epoch": 0.08283304202446995,
"grad_norm": 3.3387060909226665,
"learning_rate": 2e-05,
"loss": 0.6565097808837891,
"step": 1090
},
{
"epoch": 0.08359297818983205,
"grad_norm": 1.0681126627729676,
"learning_rate": 2e-05,
"loss": 0.6453925609588623,
"step": 1100
},
{
"epoch": 0.08435291435519417,
"grad_norm": 1.0491354731716318,
"learning_rate": 2e-05,
"loss": 0.672363567352295,
"step": 1110
},
{
"epoch": 0.08511285052055627,
"grad_norm": 2.5229554034222126,
"learning_rate": 2e-05,
"loss": 0.6442315101623535,
"step": 1120
},
{
"epoch": 0.08587278668591838,
"grad_norm": 0.9549713568548055,
"learning_rate": 2e-05,
"loss": 0.6467645168304443,
"step": 1130
},
{
"epoch": 0.08663272285128049,
"grad_norm": 1.0994614474759026,
"learning_rate": 2e-05,
"loss": 0.6226651668548584,
"step": 1140
},
{
"epoch": 0.0873926590166426,
"grad_norm": 1.5931672722656118,
"learning_rate": 2e-05,
"loss": 0.6385224819183349,
"step": 1150
},
{
"epoch": 0.08815259518200472,
"grad_norm": 1.3058693072378955,
"learning_rate": 2e-05,
"loss": 0.6373834133148193,
"step": 1160
},
{
"epoch": 0.08891253134736682,
"grad_norm": 1.498709878738794,
"learning_rate": 2e-05,
"loss": 0.6416314601898193,
"step": 1170
},
{
"epoch": 0.08967246751272893,
"grad_norm": 1.6779770963133274,
"learning_rate": 2e-05,
"loss": 0.6468165874481201,
"step": 1180
},
{
"epoch": 0.09043240367809104,
"grad_norm": 1.454215379835705,
"learning_rate": 2e-05,
"loss": 0.6477084636688233,
"step": 1190
},
{
"epoch": 0.09119233984345315,
"grad_norm": 1.086597371932175,
"learning_rate": 2e-05,
"loss": 0.6345207214355468,
"step": 1200
},
{
"epoch": 0.09195227600881525,
"grad_norm": 1.1041391627395598,
"learning_rate": 2e-05,
"loss": 0.6753363609313965,
"step": 1210
},
{
"epoch": 0.09271221217417737,
"grad_norm": 1.0442979911159636,
"learning_rate": 2e-05,
"loss": 0.6602447509765625,
"step": 1220
},
{
"epoch": 0.09347214833953948,
"grad_norm": 1.1068350401318283,
"learning_rate": 2e-05,
"loss": 0.6334839820861816,
"step": 1230
},
{
"epoch": 0.09423208450490159,
"grad_norm": 3.9758019839117336,
"learning_rate": 2e-05,
"loss": 0.6504351615905761,
"step": 1240
},
{
"epoch": 0.0949920206702637,
"grad_norm": 0.9541829838291032,
"learning_rate": 2e-05,
"loss": 0.6485028743743897,
"step": 1250
},
{
"epoch": 0.0957519568356258,
"grad_norm": 1.164445477460499,
"learning_rate": 2e-05,
"loss": 0.6839859962463379,
"step": 1260
},
{
"epoch": 0.09651189300098792,
"grad_norm": 1.8779857955368715,
"learning_rate": 2e-05,
"loss": 0.6411514282226562,
"step": 1270
},
{
"epoch": 0.09727182916635002,
"grad_norm": 1.147603065647347,
"learning_rate": 2e-05,
"loss": 0.6255774974822998,
"step": 1280
},
{
"epoch": 0.09803176533171214,
"grad_norm": 1.6550437152636595,
"learning_rate": 2e-05,
"loss": 0.6291728019714355,
"step": 1290
},
{
"epoch": 0.09879170149707425,
"grad_norm": 1.7682464495720873,
"learning_rate": 2e-05,
"loss": 0.659593915939331,
"step": 1300
},
{
"epoch": 0.09955163766243635,
"grad_norm": 3.7022883686619323,
"learning_rate": 2e-05,
"loss": 0.650638484954834,
"step": 1310
},
{
"epoch": 0.10031157382779847,
"grad_norm": 1.0466634365071796,
"learning_rate": 2e-05,
"loss": 0.6600216865539551,
"step": 1320
},
{
"epoch": 0.10107150999316057,
"grad_norm": 6.303642093550111,
"learning_rate": 2e-05,
"loss": 0.6597589015960693,
"step": 1330
},
{
"epoch": 0.10183144615852269,
"grad_norm": 1.093978067554938,
"learning_rate": 2e-05,
"loss": 0.6573177337646484,
"step": 1340
},
{
"epoch": 0.10259138232388479,
"grad_norm": 2.0424198171963748,
"learning_rate": 2e-05,
"loss": 0.6553116798400879,
"step": 1350
},
{
"epoch": 0.1033513184892469,
"grad_norm": 1.5430270988592787,
"learning_rate": 2e-05,
"loss": 0.6318343162536622,
"step": 1360
},
{
"epoch": 0.104111254654609,
"grad_norm": 2.1316696565809528,
"learning_rate": 2e-05,
"loss": 0.6384878158569336,
"step": 1370
},
{
"epoch": 0.10487119081997112,
"grad_norm": 2.91741562357673,
"learning_rate": 2e-05,
"loss": 0.6429308891296387,
"step": 1380
},
{
"epoch": 0.10563112698533324,
"grad_norm": 1.0434912962088092,
"learning_rate": 2e-05,
"loss": 0.6408050537109375,
"step": 1390
},
{
"epoch": 0.10639106315069534,
"grad_norm": 1.224807798537247,
"learning_rate": 2e-05,
"loss": 0.6579841613769531,
"step": 1400
},
{
"epoch": 0.10715099931605745,
"grad_norm": 2.0847642989918778,
"learning_rate": 2e-05,
"loss": 0.6498456954956054,
"step": 1410
},
{
"epoch": 0.10791093548141956,
"grad_norm": 1.108872735327033,
"learning_rate": 2e-05,
"loss": 0.6492026329040528,
"step": 1420
},
{
"epoch": 0.10867087164678167,
"grad_norm": 0.8206739744796532,
"learning_rate": 2e-05,
"loss": 0.6415225028991699,
"step": 1430
},
{
"epoch": 0.10943080781214377,
"grad_norm": 1.5508982512432574,
"learning_rate": 2e-05,
"loss": 0.6333587169647217,
"step": 1440
},
{
"epoch": 0.11019074397750589,
"grad_norm": 1.215268801016134,
"learning_rate": 2e-05,
"loss": 0.6485206127166748,
"step": 1450
},
{
"epoch": 0.110950680142868,
"grad_norm": 1.270810180433496,
"learning_rate": 2e-05,
"loss": 0.6261978149414062,
"step": 1460
},
{
"epoch": 0.1117106163082301,
"grad_norm": 1.2131825076098512,
"learning_rate": 2e-05,
"loss": 0.6661453247070312,
"step": 1470
},
{
"epoch": 0.11247055247359222,
"grad_norm": 0.9731960092017137,
"learning_rate": 2e-05,
"loss": 0.659299087524414,
"step": 1480
},
{
"epoch": 0.11323048863895432,
"grad_norm": 1.433503129247427,
"learning_rate": 2e-05,
"loss": 0.6485359191894531,
"step": 1490
},
{
"epoch": 0.11399042480431644,
"grad_norm": 1.3500939927983868,
"learning_rate": 2e-05,
"loss": 0.6578350067138672,
"step": 1500
},
{
"epoch": 0.11475036096967854,
"grad_norm": 1.4918825355601473,
"learning_rate": 2e-05,
"loss": 0.6433755397796631,
"step": 1510
},
{
"epoch": 0.11551029713504066,
"grad_norm": 1.8872836071366936,
"learning_rate": 2e-05,
"loss": 0.6468602657318115,
"step": 1520
},
{
"epoch": 0.11627023330040277,
"grad_norm": 2.3590696751049545,
"learning_rate": 2e-05,
"loss": 0.6331050395965576,
"step": 1530
},
{
"epoch": 0.11703016946576487,
"grad_norm": 1.333851094236431,
"learning_rate": 2e-05,
"loss": 0.667790412902832,
"step": 1540
},
{
"epoch": 0.11779010563112699,
"grad_norm": 1.259990456534997,
"learning_rate": 2e-05,
"loss": 0.6469200134277344,
"step": 1550
},
{
"epoch": 0.11855004179648909,
"grad_norm": 1.7108862137007483,
"learning_rate": 2e-05,
"loss": 0.6639731884002685,
"step": 1560
},
{
"epoch": 0.11930997796185121,
"grad_norm": 1.4148747471318508,
"learning_rate": 2e-05,
"loss": 0.6579952239990234,
"step": 1570
},
{
"epoch": 0.12006991412721331,
"grad_norm": 2.8707717322914195,
"learning_rate": 2e-05,
"loss": 0.6224180221557617,
"step": 1580
},
{
"epoch": 0.12082985029257542,
"grad_norm": 2.6283336063490403,
"learning_rate": 2e-05,
"loss": 0.6329303741455078,
"step": 1590
},
{
"epoch": 0.12158978645793753,
"grad_norm": 3.150067996130026,
"learning_rate": 2e-05,
"loss": 0.6741673946380615,
"step": 1600
},
{
"epoch": 0.12234972262329964,
"grad_norm": 1.8279905616934546,
"learning_rate": 2e-05,
"loss": 0.659599781036377,
"step": 1610
},
{
"epoch": 0.12310965878866176,
"grad_norm": 4.748332853059669,
"learning_rate": 2e-05,
"loss": 0.6577554702758789,
"step": 1620
},
{
"epoch": 0.12386959495402386,
"grad_norm": 1.4163870558356635,
"learning_rate": 2e-05,
"loss": 0.6562213897705078,
"step": 1630
},
{
"epoch": 0.12462953111938598,
"grad_norm": 6.694979985306562,
"learning_rate": 2e-05,
"loss": 0.6411036491394043,
"step": 1640
},
{
"epoch": 0.1253894672847481,
"grad_norm": 6.209427745695381,
"learning_rate": 2e-05,
"loss": 0.6805436611175537,
"step": 1650
},
{
"epoch": 0.1261494034501102,
"grad_norm": 13.756950170865382,
"learning_rate": 2e-05,
"loss": 0.6436635017395019,
"step": 1660
},
{
"epoch": 0.1269093396154723,
"grad_norm": 3.997563874538566,
"learning_rate": 2e-05,
"loss": 0.6643521785736084,
"step": 1670
},
{
"epoch": 0.12766927578083442,
"grad_norm": 2.114093971367723,
"learning_rate": 2e-05,
"loss": 0.6300763607025146,
"step": 1680
},
{
"epoch": 0.12842921194619653,
"grad_norm": 7.777372686568945,
"learning_rate": 2e-05,
"loss": 0.6559924125671387,
"step": 1690
},
{
"epoch": 0.12918914811155863,
"grad_norm": 5.240897859578015,
"learning_rate": 2e-05,
"loss": 0.6357455253601074,
"step": 1700
},
{
"epoch": 0.12994908427692073,
"grad_norm": 6.9535155450420545,
"learning_rate": 2e-05,
"loss": 0.6478095531463623,
"step": 1710
},
{
"epoch": 0.13070902044228286,
"grad_norm": 2.358848628288785,
"learning_rate": 2e-05,
"loss": 0.6790260314941406,
"step": 1720
},
{
"epoch": 0.13146895660764496,
"grad_norm": 2.0138319749117786,
"learning_rate": 2e-05,
"loss": 0.6263014793395996,
"step": 1730
},
{
"epoch": 0.13222889277300706,
"grad_norm": 1.4025723929068667,
"learning_rate": 2e-05,
"loss": 0.6489272117614746,
"step": 1740
},
{
"epoch": 0.13298882893836916,
"grad_norm": 2.723181912456477,
"learning_rate": 2e-05,
"loss": 0.6445198059082031,
"step": 1750
},
{
"epoch": 0.1337487651037313,
"grad_norm": 2.9660220967238637,
"learning_rate": 2e-05,
"loss": 0.6173060417175293,
"step": 1760
},
{
"epoch": 0.1345087012690934,
"grad_norm": 1.548833654772482,
"learning_rate": 2e-05,
"loss": 0.6361001014709473,
"step": 1770
},
{
"epoch": 0.1352686374344555,
"grad_norm": 1.4209226521258322,
"learning_rate": 2e-05,
"loss": 0.6101439476013184,
"step": 1780
},
{
"epoch": 0.13602857359981763,
"grad_norm": 5.083546421164765,
"learning_rate": 2e-05,
"loss": 0.6381206512451172,
"step": 1790
},
{
"epoch": 0.13678850976517973,
"grad_norm": 2.496640617510042,
"learning_rate": 2e-05,
"loss": 0.6482817649841308,
"step": 1800
},
{
"epoch": 0.13754844593054183,
"grad_norm": 1.9610047412853615,
"learning_rate": 2e-05,
"loss": 0.6429681301116943,
"step": 1810
},
{
"epoch": 0.13830838209590393,
"grad_norm": 1.3428122362755035,
"learning_rate": 2e-05,
"loss": 0.6612567901611328,
"step": 1820
},
{
"epoch": 0.13906831826126606,
"grad_norm": 1.0480266633178372,
"learning_rate": 2e-05,
"loss": 0.634760570526123,
"step": 1830
},
{
"epoch": 0.13982825442662816,
"grad_norm": 5.767390275358731,
"learning_rate": 2e-05,
"loss": 0.6517911434173584,
"step": 1840
},
{
"epoch": 0.14058819059199026,
"grad_norm": 1.0716026596322161,
"learning_rate": 2e-05,
"loss": 0.6544319152832031,
"step": 1850
},
{
"epoch": 0.1413481267573524,
"grad_norm": 1.846290383372757,
"learning_rate": 2e-05,
"loss": 0.6552516460418701,
"step": 1860
},
{
"epoch": 0.1421080629227145,
"grad_norm": 1.5831829672964501,
"learning_rate": 2e-05,
"loss": 0.6423295497894287,
"step": 1870
},
{
"epoch": 0.1428679990880766,
"grad_norm": 5.440399705277345,
"learning_rate": 2e-05,
"loss": 0.6394572257995605,
"step": 1880
},
{
"epoch": 0.1436279352534387,
"grad_norm": 1.9378312669131104,
"learning_rate": 2e-05,
"loss": 0.6410587310791016,
"step": 1890
},
{
"epoch": 0.14438787141880083,
"grad_norm": 1.4028689103168956,
"learning_rate": 2e-05,
"loss": 0.6463975429534912,
"step": 1900
},
{
"epoch": 0.14514780758416293,
"grad_norm": 1.5006186537122213,
"learning_rate": 2e-05,
"loss": 0.6331391334533691,
"step": 1910
},
{
"epoch": 0.14590774374952503,
"grad_norm": 1.4896410731969767,
"learning_rate": 2e-05,
"loss": 0.6415359497070312,
"step": 1920
},
{
"epoch": 0.14666767991488716,
"grad_norm": 1.4531220265216627,
"learning_rate": 2e-05,
"loss": 0.6483102321624756,
"step": 1930
},
{
"epoch": 0.14742761608024926,
"grad_norm": 4.060292836092577,
"learning_rate": 2e-05,
"loss": 0.6400858879089355,
"step": 1940
},
{
"epoch": 0.14818755224561136,
"grad_norm": 2.01518756849028,
"learning_rate": 2e-05,
"loss": 0.6560895442962646,
"step": 1950
},
{
"epoch": 0.14894748841097347,
"grad_norm": 1.1943584364475957,
"learning_rate": 2e-05,
"loss": 0.6442777156829834,
"step": 1960
},
{
"epoch": 0.1497074245763356,
"grad_norm": 1.2365114760287763,
"learning_rate": 2e-05,
"loss": 0.615467643737793,
"step": 1970
},
{
"epoch": 0.1504673607416977,
"grad_norm": 1.4776860920715695,
"learning_rate": 2e-05,
"loss": 0.6682824611663818,
"step": 1980
},
{
"epoch": 0.1512272969070598,
"grad_norm": 1.4579971244761316,
"learning_rate": 2e-05,
"loss": 0.6257425308227539,
"step": 1990
},
{
"epoch": 0.15198723307242193,
"grad_norm": 10.148842969854169,
"learning_rate": 2e-05,
"loss": 0.6339943885803223,
"step": 2000
},
{
"epoch": 0.15274716923778403,
"grad_norm": 5.442411967582219,
"learning_rate": 2e-05,
"loss": 0.6389657497406006,
"step": 2010
},
{
"epoch": 0.15350710540314613,
"grad_norm": 16.30352466832802,
"learning_rate": 2e-05,
"loss": 0.6474606990814209,
"step": 2020
},
{
"epoch": 0.15426704156850823,
"grad_norm": 4.362878921974434,
"learning_rate": 2e-05,
"loss": 0.6363828659057618,
"step": 2030
},
{
"epoch": 0.15502697773387036,
"grad_norm": 1.2676245703999107,
"learning_rate": 2e-05,
"loss": 0.6665027618408204,
"step": 2040
},
{
"epoch": 0.15578691389923247,
"grad_norm": 1.0765397406434796,
"learning_rate": 2e-05,
"loss": 0.6307790279388428,
"step": 2050
},
{
"epoch": 0.15654685006459457,
"grad_norm": 2.6315823864120396,
"learning_rate": 2e-05,
"loss": 0.661501693725586,
"step": 2060
},
{
"epoch": 0.1573067862299567,
"grad_norm": 0.917152666671035,
"learning_rate": 2e-05,
"loss": 0.6306610107421875,
"step": 2070
},
{
"epoch": 0.1580667223953188,
"grad_norm": 1.0959801674463898,
"learning_rate": 2e-05,
"loss": 0.6308591365814209,
"step": 2080
},
{
"epoch": 0.1588266585606809,
"grad_norm": 2.2239738262841513,
"learning_rate": 2e-05,
"loss": 0.6311299800872803,
"step": 2090
},
{
"epoch": 0.159586594726043,
"grad_norm": 3.4139655597175462,
"learning_rate": 2e-05,
"loss": 0.6391436576843261,
"step": 2100
},
{
"epoch": 0.16034653089140513,
"grad_norm": 1.5249315901105078,
"learning_rate": 2e-05,
"loss": 0.6251719951629638,
"step": 2110
},
{
"epoch": 0.16110646705676723,
"grad_norm": 5.887416970765259,
"learning_rate": 2e-05,
"loss": 0.6304204940795899,
"step": 2120
},
{
"epoch": 0.16186640322212933,
"grad_norm": 1.3922439274444605,
"learning_rate": 2e-05,
"loss": 0.6613870143890381,
"step": 2130
},
{
"epoch": 0.16262633938749146,
"grad_norm": 1.9337982890171186,
"learning_rate": 2e-05,
"loss": 0.6491190910339355,
"step": 2140
},
{
"epoch": 0.16338627555285357,
"grad_norm": 1.3724422818318218,
"learning_rate": 2e-05,
"loss": 0.6484801769256592,
"step": 2150
},
{
"epoch": 0.16414621171821567,
"grad_norm": 2.591228095277901,
"learning_rate": 2e-05,
"loss": 0.651193380355835,
"step": 2160
},
{
"epoch": 0.16490614788357777,
"grad_norm": 1.6151307202072216,
"learning_rate": 2e-05,
"loss": 0.6365288734436035,
"step": 2170
},
{
"epoch": 0.1656660840489399,
"grad_norm": 1.546438395722935,
"learning_rate": 2e-05,
"loss": 0.6704031944274902,
"step": 2180
},
{
"epoch": 0.166426020214302,
"grad_norm": 2.407511950274536,
"learning_rate": 2e-05,
"loss": 0.6353227138519287,
"step": 2190
},
{
"epoch": 0.1671859563796641,
"grad_norm": 1.6797534787574688,
"learning_rate": 2e-05,
"loss": 0.6494597434997559,
"step": 2200
},
{
"epoch": 0.1679458925450262,
"grad_norm": 4.8613761266230835,
"learning_rate": 2e-05,
"loss": 0.6252596855163575,
"step": 2210
},
{
"epoch": 0.16870582871038833,
"grad_norm": 0.99515726828342,
"learning_rate": 2e-05,
"loss": 0.6384762763977051,
"step": 2220
},
{
"epoch": 0.16946576487575044,
"grad_norm": 1.1520248314117032,
"learning_rate": 2e-05,
"loss": 0.6748219966888428,
"step": 2230
},
{
"epoch": 0.17022570104111254,
"grad_norm": 1.861088340323805,
"learning_rate": 2e-05,
"loss": 0.6419329643249512,
"step": 2240
},
{
"epoch": 0.17098563720647467,
"grad_norm": 1.412748370534858,
"learning_rate": 2e-05,
"loss": 0.6323102474212646,
"step": 2250
},
{
"epoch": 0.17174557337183677,
"grad_norm": 0.8560502316583172,
"learning_rate": 2e-05,
"loss": 0.6213559150695801,
"step": 2260
},
{
"epoch": 0.17250550953719887,
"grad_norm": 1.2057655077981024,
"learning_rate": 2e-05,
"loss": 0.6556430339813233,
"step": 2270
},
{
"epoch": 0.17326544570256097,
"grad_norm": 0.9239999124842788,
"learning_rate": 2e-05,
"loss": 0.6087024211883545,
"step": 2280
},
{
"epoch": 0.1740253818679231,
"grad_norm": 0.9516737623691233,
"learning_rate": 2e-05,
"loss": 0.6084947109222412,
"step": 2290
},
{
"epoch": 0.1747853180332852,
"grad_norm": 2.815786294259301,
"learning_rate": 2e-05,
"loss": 0.6483430385589599,
"step": 2300
},
{
"epoch": 0.1755452541986473,
"grad_norm": 0.8652265177148013,
"learning_rate": 2e-05,
"loss": 0.6458633422851563,
"step": 2310
},
{
"epoch": 0.17630519036400943,
"grad_norm": 1.6069854127598282,
"learning_rate": 2e-05,
"loss": 0.639268445968628,
"step": 2320
},
{
"epoch": 0.17706512652937154,
"grad_norm": 1.2760797527559364,
"learning_rate": 2e-05,
"loss": 0.6464717864990235,
"step": 2330
},
{
"epoch": 0.17782506269473364,
"grad_norm": 1.1069755420870164,
"learning_rate": 2e-05,
"loss": 0.6435245513916016,
"step": 2340
},
{
"epoch": 0.17858499886009574,
"grad_norm": 3.3877427199201224,
"learning_rate": 2e-05,
"loss": 0.6325599193572998,
"step": 2350
},
{
"epoch": 0.17934493502545787,
"grad_norm": 1.1611639959868179,
"learning_rate": 2e-05,
"loss": 0.6352363109588623,
"step": 2360
},
{
"epoch": 0.18010487119081997,
"grad_norm": 1.530158351349937,
"learning_rate": 2e-05,
"loss": 0.6382848262786865,
"step": 2370
},
{
"epoch": 0.18086480735618207,
"grad_norm": 0.828975583603307,
"learning_rate": 2e-05,
"loss": 0.6329381465911865,
"step": 2380
},
{
"epoch": 0.1816247435215442,
"grad_norm": 1.9691779255291941,
"learning_rate": 2e-05,
"loss": 0.6218951225280762,
"step": 2390
},
{
"epoch": 0.1823846796869063,
"grad_norm": 8.233905278346574,
"learning_rate": 2e-05,
"loss": 0.6378360748291015,
"step": 2400
},
{
"epoch": 0.1831446158522684,
"grad_norm": 1.253360788923102,
"learning_rate": 2e-05,
"loss": 0.639088773727417,
"step": 2410
},
{
"epoch": 0.1839045520176305,
"grad_norm": 1.0639498166629777,
"learning_rate": 2e-05,
"loss": 0.648853063583374,
"step": 2420
},
{
"epoch": 0.18466448818299264,
"grad_norm": 1.224113932980045,
"learning_rate": 2e-05,
"loss": 0.6247108459472657,
"step": 2430
},
{
"epoch": 0.18542442434835474,
"grad_norm": 0.862916274591829,
"learning_rate": 2e-05,
"loss": 0.6458800315856934,
"step": 2440
},
{
"epoch": 0.18618436051371684,
"grad_norm": 1.0838844297963148,
"learning_rate": 2e-05,
"loss": 0.6351897239685058,
"step": 2450
},
{
"epoch": 0.18694429667907897,
"grad_norm": 0.9652999261625115,
"learning_rate": 2e-05,
"loss": 0.6378411293029785,
"step": 2460
},
{
"epoch": 0.18770423284444107,
"grad_norm": 0.7979754847702989,
"learning_rate": 2e-05,
"loss": 0.6047022819519043,
"step": 2470
},
{
"epoch": 0.18846416900980317,
"grad_norm": 2.021558850612358,
"learning_rate": 2e-05,
"loss": 0.6230844497680664,
"step": 2480
},
{
"epoch": 0.18922410517516527,
"grad_norm": 0.8946816777799294,
"learning_rate": 2e-05,
"loss": 0.6341133117675781,
"step": 2490
},
{
"epoch": 0.1899840413405274,
"grad_norm": 1.479766364730666,
"learning_rate": 2e-05,
"loss": 0.6266453742980957,
"step": 2500
},
{
"epoch": 0.1907439775058895,
"grad_norm": 1.9037844174773604,
"learning_rate": 2e-05,
"loss": 0.6139904975891113,
"step": 2510
},
{
"epoch": 0.1915039136712516,
"grad_norm": 0.8961866794407122,
"learning_rate": 2e-05,
"loss": 0.608923864364624,
"step": 2520
},
{
"epoch": 0.19226384983661374,
"grad_norm": 1.6522838694053719,
"learning_rate": 2e-05,
"loss": 0.6544345378875732,
"step": 2530
},
{
"epoch": 0.19302378600197584,
"grad_norm": 1.1182433232512783,
"learning_rate": 2e-05,
"loss": 0.6411478042602539,
"step": 2540
},
{
"epoch": 0.19378372216733794,
"grad_norm": 1.1192615245073028,
"learning_rate": 1.997261400182901e-05,
"loss": 0.6062673841203962,
"step": 2550
},
{
"epoch": 0.19454365833270004,
"grad_norm": 1.0632178721311252,
"learning_rate": 1.9805800746810423e-05,
"loss": 0.6316331386566162,
"step": 2560
},
{
"epoch": 0.19530359449806217,
"grad_norm": 1.4237346885812376,
"learning_rate": 1.9489934220110274e-05,
"loss": 0.6297570705413819,
"step": 2570
},
{
"epoch": 0.19606353066342427,
"grad_norm": 4.799706403442472,
"learning_rate": 1.9029840364820392e-05,
"loss": 0.633021879196167,
"step": 2580
},
{
"epoch": 0.19682346682878638,
"grad_norm": 5.308900534860684,
"learning_rate": 1.8432548690479908e-05,
"loss": 0.6384659767150879,
"step": 2590
},
{
"epoch": 0.1975834029941485,
"grad_norm": 1.630333387718921,
"learning_rate": 1.7707184873239772e-05,
"loss": 0.6404977321624756,
"step": 2600
},
{
"epoch": 0.1983433391595106,
"grad_norm": 1.123797533518854,
"learning_rate": 1.6864831329902565e-05,
"loss": 0.6424628257751465,
"step": 2610
},
{
"epoch": 0.1991032753248727,
"grad_norm": 1.198275545430331,
"learning_rate": 1.591835789604667e-05,
"loss": 0.6066534996032715,
"step": 2620
},
{
"epoch": 0.1998632114902348,
"grad_norm": 0.9776581648334995,
"learning_rate": 1.4882225195206459e-05,
"loss": 0.6132757186889648,
"step": 2630
},
{
"epoch": 0.20062314765559694,
"grad_norm": 0.9842453668936922,
"learning_rate": 1.3772263703317695e-05,
"loss": 0.6338860988616943,
"step": 2640
},
{
"epoch": 0.20138308382095904,
"grad_norm": 0.9184542263499781,
"learning_rate": 1.2605431883975547e-05,
"loss": 0.6243062496185303,
"step": 2650
},
{
"epoch": 0.20214301998632114,
"grad_norm": 0.83304000086466,
"learning_rate": 1.1399557089817574e-05,
"loss": 0.6261422157287597,
"step": 2660
},
{
"epoch": 0.20290295615168327,
"grad_norm": 0.9581824501892245,
"learning_rate": 1.0173063188650463e-05,
"loss": 0.622321605682373,
"step": 2670
},
{
"epoch": 0.20366289231704537,
"grad_norm": 0.7902384583807414,
"learning_rate": 8.944689075764396e-06,
"loss": 0.608195686340332,
"step": 2680
},
{
"epoch": 0.20442282848240748,
"grad_norm": 1.123981063630698,
"learning_rate": 7.73320237312362e-06,
"loss": 0.6179349422454834,
"step": 2690
},
{
"epoch": 0.20518276464776958,
"grad_norm": 1.0951027626268137,
"learning_rate": 6.557112689658629e-06,
"loss": 0.6065043449401856,
"step": 2700
},
{
"epoch": 0.2059427008131317,
"grad_norm": 0.9814994421962618,
"learning_rate": 5.434388823591213e-06,
"loss": 0.6199812889099121,
"step": 2710
},
{
"epoch": 0.2067026369784938,
"grad_norm": 1.0931236372713993,
"learning_rate": 4.382184227495424e-06,
"loss": 0.6003854274749756,
"step": 2720
},
{
"epoch": 0.2074625731438559,
"grad_norm": 2.155359385509788,
"learning_rate": 3.4165749305560203e-06,
"loss": 0.5984349727630616,
"step": 2730
},
{
"epoch": 0.208222509309218,
"grad_norm": 3.9420694279102046,
"learning_rate": 2.5523139221595556e-06,
"loss": 0.6005512237548828,
"step": 2740
},
{
"epoch": 0.20898244547458014,
"grad_norm": 1.3759060505447278,
"learning_rate": 1.8026057494499821e-06,
"loss": 0.5939900398254394,
"step": 2750
},
{
"epoch": 0.20974238163994224,
"grad_norm": 3.131703698922485,
"learning_rate": 1.178904772643223e-06,
"loss": 0.6074230194091796,
"step": 2760
},
{
"epoch": 0.21050231780530435,
"grad_norm": 1.1907942914472174,
"learning_rate": 6.90740160441754e-07,
"loss": 0.6195389747619628,
"step": 2770
},
{
"epoch": 0.21126225397066647,
"grad_norm": 1.2539651296893224,
"learning_rate": 3.45570299343632e-07,
"loss": 0.6137242794036866,
"step": 2780
},
{
"epoch": 0.21202219013602858,
"grad_norm": 1.2535870401839126,
"learning_rate": 1.486688412425152e-07,
"loss": 0.607922887802124,
"step": 2790
},
{
"epoch": 0.21255414545178206,
"step": 2797,
"total_flos": 1324435843842048.0,
"train_loss": 0.05602183613045794,
"train_runtime": 3663.6732,
"train_samples_per_second": 9.161,
"train_steps_per_second": 0.763
}
],
"logging_steps": 10,
"max_steps": 2797,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 2543,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1324435843842048.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}