rqadri's picture
Upload folder using huggingface_hub
10b1cbf verified
Raw History Blame Contribute Delete
97.2 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.42533627175317273,
"eval_steps": 500,
"global_step": 5597,
"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": 1.9063805360822343,
"learning_rate": 2e-05,
"loss": 0.6063728332519531,
"step": 2550
},
{
"epoch": 0.19454365833270004,
"grad_norm": 1.339954525004715,
"learning_rate": 2e-05,
"loss": 0.6316645622253418,
"step": 2560
},
{
"epoch": 0.19530359449806217,
"grad_norm": 1.075456792572934,
"learning_rate": 2e-05,
"loss": 0.6289247035980224,
"step": 2570
},
{
"epoch": 0.19606353066342427,
"grad_norm": 0.8780966177223058,
"learning_rate": 2e-05,
"loss": 0.6316461086273193,
"step": 2580
},
{
"epoch": 0.19682346682878638,
"grad_norm": 0.8172704754373125,
"learning_rate": 2e-05,
"loss": 0.6350976467132569,
"step": 2590
},
{
"epoch": 0.1975834029941485,
"grad_norm": 1.314600777113447,
"learning_rate": 2e-05,
"loss": 0.6386783599853516,
"step": 2600
},
{
"epoch": 0.1983433391595106,
"grad_norm": 1.1204615377337357,
"learning_rate": 2e-05,
"loss": 0.6429242134094239,
"step": 2610
},
{
"epoch": 0.1991032753248727,
"grad_norm": 0.9207404992707692,
"learning_rate": 2e-05,
"loss": 0.6100504875183106,
"step": 2620
},
{
"epoch": 0.1998632114902348,
"grad_norm": 1.4479828471747938,
"learning_rate": 2e-05,
"loss": 0.6189369201660156,
"step": 2630
},
{
"epoch": 0.20062314765559694,
"grad_norm": 1.6538730364392782,
"learning_rate": 2e-05,
"loss": 0.6408646583557129,
"step": 2640
},
{
"epoch": 0.20138308382095904,
"grad_norm": 1.400833721181578,
"learning_rate": 2e-05,
"loss": 0.6331706523895264,
"step": 2650
},
{
"epoch": 0.20214301998632114,
"grad_norm": 2.4651840710400217,
"learning_rate": 2e-05,
"loss": 0.6359334945678711,
"step": 2660
},
{
"epoch": 0.20290295615168327,
"grad_norm": 2.8902390382766168,
"learning_rate": 2e-05,
"loss": 0.6367801666259766,
"step": 2670
},
{
"epoch": 0.20366289231704537,
"grad_norm": 1.0725705407050916,
"learning_rate": 2e-05,
"loss": 0.6252540588378906,
"step": 2680
},
{
"epoch": 0.20442282848240748,
"grad_norm": 1.453680714722172,
"learning_rate": 2e-05,
"loss": 0.6370233535766602,
"step": 2690
},
{
"epoch": 0.20518276464776958,
"grad_norm": 1.1494128351878075,
"learning_rate": 2e-05,
"loss": 0.6241844177246094,
"step": 2700
},
{
"epoch": 0.2059427008131317,
"grad_norm": 1.2561059487170374,
"learning_rate": 2e-05,
"loss": 0.6379550933837891,
"step": 2710
},
{
"epoch": 0.2067026369784938,
"grad_norm": 1.1763018989953011,
"learning_rate": 2e-05,
"loss": 0.617769193649292,
"step": 2720
},
{
"epoch": 0.2074625731438559,
"grad_norm": 0.8301947300077038,
"learning_rate": 2e-05,
"loss": 0.6172198295593262,
"step": 2730
},
{
"epoch": 0.208222509309218,
"grad_norm": 1.2654029656743424,
"learning_rate": 2e-05,
"loss": 0.6206151008605957,
"step": 2740
},
{
"epoch": 0.20898244547458014,
"grad_norm": 1.2906344649474475,
"learning_rate": 2e-05,
"loss": 0.6183318138122559,
"step": 2750
},
{
"epoch": 0.20974238163994224,
"grad_norm": 19.702502388770796,
"learning_rate": 2e-05,
"loss": 0.6330374717712403,
"step": 2760
},
{
"epoch": 0.21050231780530435,
"grad_norm": 2.896777724084538,
"learning_rate": 2e-05,
"loss": 0.6482915878295898,
"step": 2770
},
{
"epoch": 0.21126225397066647,
"grad_norm": 1.0591159342855259,
"learning_rate": 2e-05,
"loss": 0.6418767452239991,
"step": 2780
},
{
"epoch": 0.21202219013602858,
"grad_norm": 0.9814179371810936,
"learning_rate": 2e-05,
"loss": 0.6314129829406738,
"step": 2790
},
{
"epoch": 0.21278212630139068,
"grad_norm": 0.978536412787398,
"learning_rate": 2e-05,
"loss": 0.6173480033874512,
"step": 2800
},
{
"epoch": 0.21354206246675278,
"grad_norm": 1.0392932503872843,
"learning_rate": 2e-05,
"loss": 0.6070091247558593,
"step": 2810
},
{
"epoch": 0.2143019986321149,
"grad_norm": 0.7630311281085816,
"learning_rate": 2e-05,
"loss": 0.603149127960205,
"step": 2820
},
{
"epoch": 0.215061934797477,
"grad_norm": 0.7796521616886448,
"learning_rate": 2e-05,
"loss": 0.6206292629241943,
"step": 2830
},
{
"epoch": 0.2158218709628391,
"grad_norm": 1.135805965040796,
"learning_rate": 2e-05,
"loss": 0.6249284267425537,
"step": 2840
},
{
"epoch": 0.21658180712820124,
"grad_norm": 1.0034800711429899,
"learning_rate": 2e-05,
"loss": 0.6145776271820068,
"step": 2850
},
{
"epoch": 0.21734174329356334,
"grad_norm": 1.038072409401356,
"learning_rate": 2e-05,
"loss": 0.6203277587890625,
"step": 2860
},
{
"epoch": 0.21810167945892545,
"grad_norm": 1.586011294473222,
"learning_rate": 2e-05,
"loss": 0.6286195278167724,
"step": 2870
},
{
"epoch": 0.21886161562428755,
"grad_norm": 0.8925062852923543,
"learning_rate": 2e-05,
"loss": 0.6153251647949218,
"step": 2880
},
{
"epoch": 0.21962155178964968,
"grad_norm": 1.9996238470642247,
"learning_rate": 2e-05,
"loss": 0.6150319576263428,
"step": 2890
},
{
"epoch": 0.22038148795501178,
"grad_norm": 1.851709788656795,
"learning_rate": 2e-05,
"loss": 0.6348749160766601,
"step": 2900
},
{
"epoch": 0.22114142412037388,
"grad_norm": 1.5503903928626321,
"learning_rate": 2e-05,
"loss": 0.6155122280120849,
"step": 2910
},
{
"epoch": 0.221901360285736,
"grad_norm": 1.6014915313981002,
"learning_rate": 2e-05,
"loss": 0.6487886428833007,
"step": 2920
},
{
"epoch": 0.2226612964510981,
"grad_norm": 3.0888999458086306,
"learning_rate": 2e-05,
"loss": 0.6402599334716796,
"step": 2930
},
{
"epoch": 0.2234212326164602,
"grad_norm": 2.256985072645217,
"learning_rate": 2e-05,
"loss": 0.639389181137085,
"step": 2940
},
{
"epoch": 0.22418116878182232,
"grad_norm": 1.1953816080056705,
"learning_rate": 2e-05,
"loss": 0.6524765014648437,
"step": 2950
},
{
"epoch": 0.22494110494718444,
"grad_norm": 1.4837984558035446,
"learning_rate": 2e-05,
"loss": 0.6412689208984375,
"step": 2960
},
{
"epoch": 0.22570104111254655,
"grad_norm": 1.298620783185863,
"learning_rate": 2e-05,
"loss": 0.6097001552581787,
"step": 2970
},
{
"epoch": 0.22646097727790865,
"grad_norm": 0.9186623742719742,
"learning_rate": 2e-05,
"loss": 0.6124137878417969,
"step": 2980
},
{
"epoch": 0.22722091344327078,
"grad_norm": 1.2086327273935487,
"learning_rate": 2e-05,
"loss": 0.6018218994140625,
"step": 2990
},
{
"epoch": 0.22798084960863288,
"grad_norm": 1.1774896919937214,
"learning_rate": 2e-05,
"loss": 0.6359123229980469,
"step": 3000
},
{
"epoch": 0.22874078577399498,
"grad_norm": 0.8376140766835578,
"learning_rate": 2e-05,
"loss": 0.6134356498718262,
"step": 3010
},
{
"epoch": 0.22950072193935708,
"grad_norm": 1.3695480635864736,
"learning_rate": 2e-05,
"loss": 0.6094162940979004,
"step": 3020
},
{
"epoch": 0.2302606581047192,
"grad_norm": 8.759741635179273,
"learning_rate": 2e-05,
"loss": 0.6244755744934082,
"step": 3030
},
{
"epoch": 0.23102059427008131,
"grad_norm": 4.401741469652162,
"learning_rate": 2e-05,
"loss": 0.6653237819671631,
"step": 3040
},
{
"epoch": 0.23178053043544342,
"grad_norm": 0.9651591430969964,
"learning_rate": 2e-05,
"loss": 0.6163477897644043,
"step": 3050
},
{
"epoch": 0.23254046660080555,
"grad_norm": 1.2540306441172315,
"learning_rate": 2e-05,
"loss": 0.6076150417327881,
"step": 3060
},
{
"epoch": 0.23330040276616765,
"grad_norm": 1.512197160849794,
"learning_rate": 2e-05,
"loss": 0.6530462265014648,
"step": 3070
},
{
"epoch": 0.23406033893152975,
"grad_norm": 1.0873268561701026,
"learning_rate": 2e-05,
"loss": 0.6455690383911132,
"step": 3080
},
{
"epoch": 0.23482027509689185,
"grad_norm": 1.5428987528547071,
"learning_rate": 2e-05,
"loss": 0.6147666931152344,
"step": 3090
},
{
"epoch": 0.23558021126225398,
"grad_norm": 2.8908752896198657,
"learning_rate": 2e-05,
"loss": 0.6317712306976319,
"step": 3100
},
{
"epoch": 0.23634014742761608,
"grad_norm": 1.089006444137297,
"learning_rate": 2e-05,
"loss": 0.6073556423187256,
"step": 3110
},
{
"epoch": 0.23710008359297818,
"grad_norm": 1.062047237532833,
"learning_rate": 2e-05,
"loss": 0.629581356048584,
"step": 3120
},
{
"epoch": 0.2378600197583403,
"grad_norm": 1.0177972120401673,
"learning_rate": 2e-05,
"loss": 0.6323951721191406,
"step": 3130
},
{
"epoch": 0.23861995592370241,
"grad_norm": 1.0040279876940803,
"learning_rate": 2e-05,
"loss": 0.611222219467163,
"step": 3140
},
{
"epoch": 0.23937989208906452,
"grad_norm": 1.3386060294466204,
"learning_rate": 2e-05,
"loss": 0.6076837062835694,
"step": 3150
},
{
"epoch": 0.24013982825442662,
"grad_norm": 1.003288225197739,
"learning_rate": 2e-05,
"loss": 0.6263534545898437,
"step": 3160
},
{
"epoch": 0.24089976441978875,
"grad_norm": 0.9295245153210507,
"learning_rate": 2e-05,
"loss": 0.629744291305542,
"step": 3170
},
{
"epoch": 0.24165970058515085,
"grad_norm": 1.3354273823166982,
"learning_rate": 2e-05,
"loss": 0.6224825859069825,
"step": 3180
},
{
"epoch": 0.24241963675051295,
"grad_norm": 1.0760003107828335,
"learning_rate": 2e-05,
"loss": 0.6299269676208497,
"step": 3190
},
{
"epoch": 0.24317957291587505,
"grad_norm": 0.941133634643217,
"learning_rate": 2e-05,
"loss": 0.6484769821166992,
"step": 3200
},
{
"epoch": 0.24393950908123718,
"grad_norm": 1.3185461630351192,
"learning_rate": 2e-05,
"loss": 0.6362860679626465,
"step": 3210
},
{
"epoch": 0.24469944524659928,
"grad_norm": 1.183339782363713,
"learning_rate": 2e-05,
"loss": 0.6190390586853027,
"step": 3220
},
{
"epoch": 0.24545938141196139,
"grad_norm": 0.9387291080116852,
"learning_rate": 2e-05,
"loss": 0.6350534439086915,
"step": 3230
},
{
"epoch": 0.24621931757732352,
"grad_norm": 2.310301884693289,
"learning_rate": 2e-05,
"loss": 0.6099423885345459,
"step": 3240
},
{
"epoch": 0.24697925374268562,
"grad_norm": 4.51689150363215,
"learning_rate": 2e-05,
"loss": 0.6252852439880371,
"step": 3250
},
{
"epoch": 0.24773918990804772,
"grad_norm": 0.9038004061763136,
"learning_rate": 2e-05,
"loss": 0.6356831550598144,
"step": 3260
},
{
"epoch": 0.24849912607340982,
"grad_norm": 9.22641306881663,
"learning_rate": 2e-05,
"loss": 0.639389181137085,
"step": 3270
},
{
"epoch": 0.24925906223877195,
"grad_norm": 1.4908438703959253,
"learning_rate": 2e-05,
"loss": 0.628369140625,
"step": 3280
},
{
"epoch": 0.25001899840413405,
"grad_norm": 7.419626390823035,
"learning_rate": 2e-05,
"loss": 0.6123821258544921,
"step": 3290
},
{
"epoch": 0.2507789345694962,
"grad_norm": 1.2786580240513115,
"learning_rate": 2e-05,
"loss": 0.6278958320617676,
"step": 3300
},
{
"epoch": 0.25153887073485826,
"grad_norm": 1.0502472828946412,
"learning_rate": 2e-05,
"loss": 0.6296170711517334,
"step": 3310
},
{
"epoch": 0.2522988069002204,
"grad_norm": 1.0918861973419551,
"learning_rate": 2e-05,
"loss": 0.6308129787445068,
"step": 3320
},
{
"epoch": 0.2530587430655825,
"grad_norm": 1.1226136226715475,
"learning_rate": 2e-05,
"loss": 0.6414403438568115,
"step": 3330
},
{
"epoch": 0.2538186792309446,
"grad_norm": 1.3250141910916213,
"learning_rate": 2e-05,
"loss": 0.6283860206604004,
"step": 3340
},
{
"epoch": 0.2545786153963067,
"grad_norm": 1.3565252029111057,
"learning_rate": 2e-05,
"loss": 0.632968521118164,
"step": 3350
},
{
"epoch": 0.25533855156166885,
"grad_norm": 1.128898434457114,
"learning_rate": 2e-05,
"loss": 0.6104358673095703,
"step": 3360
},
{
"epoch": 0.2560984877270309,
"grad_norm": 1.6612843957044359,
"learning_rate": 2e-05,
"loss": 0.6286696434020996,
"step": 3370
},
{
"epoch": 0.25685842389239305,
"grad_norm": 1.156000264796932,
"learning_rate": 2e-05,
"loss": 0.6091503143310547,
"step": 3380
},
{
"epoch": 0.2576183600577551,
"grad_norm": 0.7528740182554841,
"learning_rate": 2e-05,
"loss": 0.6218111038208007,
"step": 3390
},
{
"epoch": 0.25837829622311725,
"grad_norm": 0.9275869426296971,
"learning_rate": 2e-05,
"loss": 0.621043586730957,
"step": 3400
},
{
"epoch": 0.2591382323884794,
"grad_norm": 1.0552086309608597,
"learning_rate": 2e-05,
"loss": 0.6073514461517334,
"step": 3410
},
{
"epoch": 0.25989816855384146,
"grad_norm": 1.1460119713030243,
"learning_rate": 2e-05,
"loss": 0.6165019512176514,
"step": 3420
},
{
"epoch": 0.2606581047192036,
"grad_norm": 1.0155774831457327,
"learning_rate": 2e-05,
"loss": 0.6184118270874024,
"step": 3430
},
{
"epoch": 0.2614180408845657,
"grad_norm": 1.3610565099669534,
"learning_rate": 2e-05,
"loss": 0.5858828544616699,
"step": 3440
},
{
"epoch": 0.2621779770499278,
"grad_norm": 1.6942539291614553,
"learning_rate": 2e-05,
"loss": 0.6389949798583985,
"step": 3450
},
{
"epoch": 0.2629379132152899,
"grad_norm": 0.9432722353200557,
"learning_rate": 2e-05,
"loss": 0.634234619140625,
"step": 3460
},
{
"epoch": 0.26369784938065205,
"grad_norm": 9.327485800969509,
"learning_rate": 2e-05,
"loss": 0.589467716217041,
"step": 3470
},
{
"epoch": 0.2644577855460141,
"grad_norm": 3.071442209769249,
"learning_rate": 2e-05,
"loss": 0.6044052124023438,
"step": 3480
},
{
"epoch": 0.26521772171137625,
"grad_norm": 1.2925607716528897,
"learning_rate": 2e-05,
"loss": 0.6095686912536621,
"step": 3490
},
{
"epoch": 0.2659776578767383,
"grad_norm": 16.06253345369598,
"learning_rate": 2e-05,
"loss": 0.6121344566345215,
"step": 3500
},
{
"epoch": 0.26673759404210046,
"grad_norm": 1.436188332832336,
"learning_rate": 2e-05,
"loss": 0.6014012336730957,
"step": 3510
},
{
"epoch": 0.2674975302074626,
"grad_norm": 1.7102705577739652,
"learning_rate": 2e-05,
"loss": 0.6195626258850098,
"step": 3520
},
{
"epoch": 0.26825746637282466,
"grad_norm": 1.4558527764040676,
"learning_rate": 2e-05,
"loss": 0.6225949764251709,
"step": 3530
},
{
"epoch": 0.2690174025381868,
"grad_norm": 1.3080733031627572,
"learning_rate": 2e-05,
"loss": 0.6161701202392578,
"step": 3540
},
{
"epoch": 0.2697773387035489,
"grad_norm": 1.1258102284127145,
"learning_rate": 2e-05,
"loss": 0.6220259666442871,
"step": 3550
},
{
"epoch": 0.270537274868911,
"grad_norm": 1.2422284824681804,
"learning_rate": 2e-05,
"loss": 0.645357084274292,
"step": 3560
},
{
"epoch": 0.2712972110342731,
"grad_norm": 1.2292013513428475,
"learning_rate": 2e-05,
"loss": 0.6096301078796387,
"step": 3570
},
{
"epoch": 0.27205714719963525,
"grad_norm": 1.0205303407560335,
"learning_rate": 2e-05,
"loss": 0.6461244106292725,
"step": 3580
},
{
"epoch": 0.2728170833649973,
"grad_norm": 2.6872881400099273,
"learning_rate": 2e-05,
"loss": 0.6358933925628663,
"step": 3590
},
{
"epoch": 0.27357701953035946,
"grad_norm": 1.1556669884783608,
"learning_rate": 2e-05,
"loss": 0.6323175430297852,
"step": 3600
},
{
"epoch": 0.2743369556957216,
"grad_norm": 3.314994024182842,
"learning_rate": 2e-05,
"loss": 0.5981500625610352,
"step": 3610
},
{
"epoch": 0.27509689186108366,
"grad_norm": 1.5758389088455522,
"learning_rate": 2e-05,
"loss": 0.6199132442474365,
"step": 3620
},
{
"epoch": 0.2758568280264458,
"grad_norm": 1.1523052929094744,
"learning_rate": 2e-05,
"loss": 0.6311004161834717,
"step": 3630
},
{
"epoch": 0.27661676419180786,
"grad_norm": 4.209511583399702,
"learning_rate": 2e-05,
"loss": 0.6306180477142334,
"step": 3640
},
{
"epoch": 0.27737670035717,
"grad_norm": 5.9265607841494905,
"learning_rate": 2e-05,
"loss": 0.614080810546875,
"step": 3650
},
{
"epoch": 0.2781366365225321,
"grad_norm": 1.339276521039795,
"learning_rate": 2e-05,
"loss": 0.6019871711730957,
"step": 3660
},
{
"epoch": 0.2788965726878942,
"grad_norm": 1.4009250479535327,
"learning_rate": 2e-05,
"loss": 0.6156785011291503,
"step": 3670
},
{
"epoch": 0.2796565088532563,
"grad_norm": 1.8701775141946877,
"learning_rate": 2e-05,
"loss": 0.6056217193603516,
"step": 3680
},
{
"epoch": 0.28041644501861845,
"grad_norm": 1.90893727770188,
"learning_rate": 2e-05,
"loss": 0.6284603118896485,
"step": 3690
},
{
"epoch": 0.28117638118398053,
"grad_norm": 2.744963869747775,
"learning_rate": 2e-05,
"loss": 0.6229351043701172,
"step": 3700
},
{
"epoch": 0.28193631734934266,
"grad_norm": 1.18153979601221,
"learning_rate": 2e-05,
"loss": 0.6146137237548828,
"step": 3710
},
{
"epoch": 0.2826962535147048,
"grad_norm": 1.2996533180738103,
"learning_rate": 2e-05,
"loss": 0.6364758491516114,
"step": 3720
},
{
"epoch": 0.28345618968006686,
"grad_norm": 1.158844048871961,
"learning_rate": 2e-05,
"loss": 0.5962699890136719,
"step": 3730
},
{
"epoch": 0.284216125845429,
"grad_norm": 1.2216472402171064,
"learning_rate": 2e-05,
"loss": 0.6206924438476562,
"step": 3740
},
{
"epoch": 0.2849760620107911,
"grad_norm": 0.890679821509559,
"learning_rate": 2e-05,
"loss": 0.6205419540405274,
"step": 3750
},
{
"epoch": 0.2857359981761532,
"grad_norm": 1.2912051550907495,
"learning_rate": 2e-05,
"loss": 0.624941635131836,
"step": 3760
},
{
"epoch": 0.2864959343415153,
"grad_norm": 1.0646251600564485,
"learning_rate": 2e-05,
"loss": 0.6251903533935547,
"step": 3770
},
{
"epoch": 0.2872558705068774,
"grad_norm": 1.4325260373232669,
"learning_rate": 2e-05,
"loss": 0.6093158721923828,
"step": 3780
},
{
"epoch": 0.2880158066722395,
"grad_norm": 1.8943205464020347,
"learning_rate": 2e-05,
"loss": 0.6362890243530274,
"step": 3790
},
{
"epoch": 0.28877574283760166,
"grad_norm": 2.3208592609691543,
"learning_rate": 2e-05,
"loss": 0.6275027751922607,
"step": 3800
},
{
"epoch": 0.28953567900296373,
"grad_norm": 2.1699449912440745,
"learning_rate": 2e-05,
"loss": 0.6083850860595703,
"step": 3810
},
{
"epoch": 0.29029561516832586,
"grad_norm": 0.9942819199698443,
"learning_rate": 2e-05,
"loss": 0.6335760354995728,
"step": 3820
},
{
"epoch": 0.291055551333688,
"grad_norm": 2.1188962468497134,
"learning_rate": 2e-05,
"loss": 0.6074498653411865,
"step": 3830
},
{
"epoch": 0.29181548749905006,
"grad_norm": 1.1946221873993108,
"learning_rate": 2e-05,
"loss": 0.6301320552825928,
"step": 3840
},
{
"epoch": 0.2925754236644122,
"grad_norm": 1.4654668135262237,
"learning_rate": 2e-05,
"loss": 0.615028190612793,
"step": 3850
},
{
"epoch": 0.2933353598297743,
"grad_norm": 0.9704782828686308,
"learning_rate": 2e-05,
"loss": 0.6268731594085694,
"step": 3860
},
{
"epoch": 0.2940952959951364,
"grad_norm": 0.973881954733979,
"learning_rate": 2e-05,
"loss": 0.622544240951538,
"step": 3870
},
{
"epoch": 0.2948552321604985,
"grad_norm": 0.9862174972117107,
"learning_rate": 2e-05,
"loss": 0.6262887477874756,
"step": 3880
},
{
"epoch": 0.2956151683258606,
"grad_norm": 1.8185012915464214,
"learning_rate": 2e-05,
"loss": 0.6259152889251709,
"step": 3890
},
{
"epoch": 0.29637510449122273,
"grad_norm": 0.9624840612294071,
"learning_rate": 2e-05,
"loss": 0.6382728099822998,
"step": 3900
},
{
"epoch": 0.29713504065658486,
"grad_norm": 1.27087177885983,
"learning_rate": 2e-05,
"loss": 0.6182097911834716,
"step": 3910
},
{
"epoch": 0.29789497682194693,
"grad_norm": 2.3890028433012644,
"learning_rate": 2e-05,
"loss": 0.6197179317474365,
"step": 3920
},
{
"epoch": 0.29865491298730906,
"grad_norm": 1.3648910447241072,
"learning_rate": 2e-05,
"loss": 0.6085414409637451,
"step": 3930
},
{
"epoch": 0.2994148491526712,
"grad_norm": 1.3264500482112846,
"learning_rate": 2e-05,
"loss": 0.6149806022644043,
"step": 3940
},
{
"epoch": 0.30017478531803327,
"grad_norm": 1.3200962604276947,
"learning_rate": 2e-05,
"loss": 0.6308249473571778,
"step": 3950
},
{
"epoch": 0.3009347214833954,
"grad_norm": 0.8595307476045679,
"learning_rate": 2e-05,
"loss": 0.6181689262390136,
"step": 3960
},
{
"epoch": 0.3016946576487575,
"grad_norm": 14.901845615980008,
"learning_rate": 2e-05,
"loss": 0.6143277168273926,
"step": 3970
},
{
"epoch": 0.3024545938141196,
"grad_norm": 4.449595695317806,
"learning_rate": 2e-05,
"loss": 0.6383512020111084,
"step": 3980
},
{
"epoch": 0.30321452997948173,
"grad_norm": 1.127083082452341,
"learning_rate": 2e-05,
"loss": 0.6203843116760254,
"step": 3990
},
{
"epoch": 0.30397446614484386,
"grad_norm": 0.8860818290392859,
"learning_rate": 2e-05,
"loss": 0.6098330974578857,
"step": 4000
},
{
"epoch": 0.30473440231020593,
"grad_norm": 1.0437370685884855,
"learning_rate": 2e-05,
"loss": 0.6101393699645996,
"step": 4010
},
{
"epoch": 0.30549433847556806,
"grad_norm": 1.278469365345407,
"learning_rate": 2e-05,
"loss": 0.6249421119689942,
"step": 4020
},
{
"epoch": 0.30625427464093014,
"grad_norm": 1.4953459408550605,
"learning_rate": 2e-05,
"loss": 0.6222182273864746,
"step": 4030
},
{
"epoch": 0.30701421080629226,
"grad_norm": 3.144699685900144,
"learning_rate": 2e-05,
"loss": 0.6221820831298828,
"step": 4040
},
{
"epoch": 0.3077741469716544,
"grad_norm": 1.6854712024923644,
"learning_rate": 2e-05,
"loss": 0.6161329269409179,
"step": 4050
},
{
"epoch": 0.30853408313701647,
"grad_norm": 2.3151959150874766,
"learning_rate": 2e-05,
"loss": 0.6204587936401367,
"step": 4060
},
{
"epoch": 0.3092940193023786,
"grad_norm": 1.4811534982681576,
"learning_rate": 2e-05,
"loss": 0.6056510925292968,
"step": 4070
},
{
"epoch": 0.3100539554677407,
"grad_norm": 1.0339459517187002,
"learning_rate": 2e-05,
"loss": 0.6259673595428467,
"step": 4080
},
{
"epoch": 0.3108138916331028,
"grad_norm": 1.2235117780386027,
"learning_rate": 2e-05,
"loss": 0.6084205627441406,
"step": 4090
},
{
"epoch": 0.31157382779846493,
"grad_norm": 1.7393673646949717,
"learning_rate": 2e-05,
"loss": 0.6320530891418457,
"step": 4100
},
{
"epoch": 0.31233376396382706,
"grad_norm": 0.9030435760410976,
"learning_rate": 2e-05,
"loss": 0.6125435829162598,
"step": 4110
},
{
"epoch": 0.31309370012918913,
"grad_norm": 0.9770233863461716,
"learning_rate": 2e-05,
"loss": 0.6324932098388671,
"step": 4120
},
{
"epoch": 0.31385363629455126,
"grad_norm": 11.164786625545675,
"learning_rate": 2e-05,
"loss": 0.6129942893981933,
"step": 4130
},
{
"epoch": 0.3146135724599134,
"grad_norm": 1.484858734640339,
"learning_rate": 2e-05,
"loss": 0.6236273765563964,
"step": 4140
},
{
"epoch": 0.31537350862527547,
"grad_norm": 1.0150188901770478,
"learning_rate": 2e-05,
"loss": 0.6191964626312256,
"step": 4150
},
{
"epoch": 0.3161334447906376,
"grad_norm": 0.734050746516317,
"learning_rate": 2e-05,
"loss": 0.6276535987854004,
"step": 4160
},
{
"epoch": 0.31689338095599967,
"grad_norm": 1.4167390319947213,
"learning_rate": 2e-05,
"loss": 0.6143552780151367,
"step": 4170
},
{
"epoch": 0.3176533171213618,
"grad_norm": 6.988334362027056,
"learning_rate": 2e-05,
"loss": 0.595700740814209,
"step": 4180
},
{
"epoch": 0.31841325328672393,
"grad_norm": 1.044747728571304,
"learning_rate": 2e-05,
"loss": 0.6026046752929688,
"step": 4190
},
{
"epoch": 0.319173189452086,
"grad_norm": 1.5761166174685894,
"learning_rate": 2e-05,
"loss": 0.6220703125,
"step": 4200
},
{
"epoch": 0.31993312561744813,
"grad_norm": 1.368582543235248,
"learning_rate": 2e-05,
"loss": 0.6257802009582519,
"step": 4210
},
{
"epoch": 0.32069306178281026,
"grad_norm": 0.8593932041129811,
"learning_rate": 2e-05,
"loss": 0.6099929809570312,
"step": 4220
},
{
"epoch": 0.32145299794817234,
"grad_norm": 2.866418221741655,
"learning_rate": 2e-05,
"loss": 0.6190606117248535,
"step": 4230
},
{
"epoch": 0.32221293411353447,
"grad_norm": 1.7731877645586673,
"learning_rate": 2e-05,
"loss": 0.6337327480316162,
"step": 4240
},
{
"epoch": 0.3229728702788966,
"grad_norm": 0.8654444995328492,
"learning_rate": 2e-05,
"loss": 0.6130261898040772,
"step": 4250
},
{
"epoch": 0.32373280644425867,
"grad_norm": 2.6189925006836585,
"learning_rate": 2e-05,
"loss": 0.6288496971130371,
"step": 4260
},
{
"epoch": 0.3244927426096208,
"grad_norm": 1.3466638193489775,
"learning_rate": 2e-05,
"loss": 0.6257670402526856,
"step": 4270
},
{
"epoch": 0.32525267877498293,
"grad_norm": 1.013486446763607,
"learning_rate": 2e-05,
"loss": 0.6344550609588623,
"step": 4280
},
{
"epoch": 0.326012614940345,
"grad_norm": 1.1856650844101666,
"learning_rate": 2e-05,
"loss": 0.6028676509857178,
"step": 4290
},
{
"epoch": 0.32677255110570713,
"grad_norm": 1.021941635755477,
"learning_rate": 2e-05,
"loss": 0.6205586433410645,
"step": 4300
},
{
"epoch": 0.3275324872710692,
"grad_norm": 1.0933995287761475,
"learning_rate": 2e-05,
"loss": 0.6085476875305176,
"step": 4310
},
{
"epoch": 0.32829242343643134,
"grad_norm": 1.9004377903753171,
"learning_rate": 2e-05,
"loss": 0.6211337089538574,
"step": 4320
},
{
"epoch": 0.32905235960179346,
"grad_norm": 2.530273155001998,
"learning_rate": 2e-05,
"loss": 0.6182102203369141,
"step": 4330
},
{
"epoch": 0.32981229576715554,
"grad_norm": 1.7906530704493195,
"learning_rate": 2e-05,
"loss": 0.6185105800628662,
"step": 4340
},
{
"epoch": 0.33057223193251767,
"grad_norm": 1.670584019028632,
"learning_rate": 2e-05,
"loss": 0.6119475364685059,
"step": 4350
},
{
"epoch": 0.3313321680978798,
"grad_norm": 1.5455256174452385,
"learning_rate": 2e-05,
"loss": 0.6008148193359375,
"step": 4360
},
{
"epoch": 0.33209210426324187,
"grad_norm": 3.208457109843326,
"learning_rate": 2e-05,
"loss": 0.6103523254394532,
"step": 4370
},
{
"epoch": 0.332852040428604,
"grad_norm": 4.734555401179033,
"learning_rate": 2e-05,
"loss": 0.6435708045959473,
"step": 4380
},
{
"epoch": 0.33361197659396613,
"grad_norm": 0.9834564438968257,
"learning_rate": 2e-05,
"loss": 0.613013744354248,
"step": 4390
},
{
"epoch": 0.3343719127593282,
"grad_norm": 1.5899940571927773,
"learning_rate": 2e-05,
"loss": 0.6370656490325928,
"step": 4400
},
{
"epoch": 0.33513184892469033,
"grad_norm": 0.9237244721295171,
"learning_rate": 2e-05,
"loss": 0.6245400905609131,
"step": 4410
},
{
"epoch": 0.3358917850900524,
"grad_norm": 0.9276064643002765,
"learning_rate": 2e-05,
"loss": 0.6174593925476074,
"step": 4420
},
{
"epoch": 0.33665172125541454,
"grad_norm": 2.8272846058313927,
"learning_rate": 2e-05,
"loss": 0.6339360237121582,
"step": 4430
},
{
"epoch": 0.33741165742077667,
"grad_norm": 0.8986064383731139,
"learning_rate": 2e-05,
"loss": 0.597989559173584,
"step": 4440
},
{
"epoch": 0.33817159358613874,
"grad_norm": 0.8793590754097262,
"learning_rate": 2e-05,
"loss": 0.6045150756835938,
"step": 4450
},
{
"epoch": 0.33893152975150087,
"grad_norm": 1.1779563215018218,
"learning_rate": 2e-05,
"loss": 0.6187587261199952,
"step": 4460
},
{
"epoch": 0.339691465916863,
"grad_norm": 1.3468609346078841,
"learning_rate": 2e-05,
"loss": 0.6251970291137695,
"step": 4470
},
{
"epoch": 0.3404514020822251,
"grad_norm": 0.8988935942814856,
"learning_rate": 2e-05,
"loss": 0.597871971130371,
"step": 4480
},
{
"epoch": 0.3412113382475872,
"grad_norm": 0.8607023127414716,
"learning_rate": 2e-05,
"loss": 0.6100861549377441,
"step": 4490
},
{
"epoch": 0.34197127441294933,
"grad_norm": 0.8479817886620996,
"learning_rate": 2e-05,
"loss": 0.6026260375976562,
"step": 4500
},
{
"epoch": 0.3427312105783114,
"grad_norm": 0.907490203208679,
"learning_rate": 2e-05,
"loss": 0.6259716033935547,
"step": 4510
},
{
"epoch": 0.34349114674367354,
"grad_norm": 0.9683122114373157,
"learning_rate": 2e-05,
"loss": 0.6468429088592529,
"step": 4520
},
{
"epoch": 0.34425108290903567,
"grad_norm": 0.9267267803730356,
"learning_rate": 2e-05,
"loss": 0.625063419342041,
"step": 4530
},
{
"epoch": 0.34501101907439774,
"grad_norm": 1.8581451268995983,
"learning_rate": 2e-05,
"loss": 0.6267961502075196,
"step": 4540
},
{
"epoch": 0.34577095523975987,
"grad_norm": 1.1135865038363975,
"learning_rate": 2e-05,
"loss": 0.6432443618774414,
"step": 4550
},
{
"epoch": 0.34653089140512194,
"grad_norm": 1.1244838971476872,
"learning_rate": 2e-05,
"loss": 0.6133568763732911,
"step": 4560
},
{
"epoch": 0.3472908275704841,
"grad_norm": 3.4229698437830103,
"learning_rate": 2e-05,
"loss": 0.602759838104248,
"step": 4570
},
{
"epoch": 0.3480507637358462,
"grad_norm": 1.7669657715748632,
"learning_rate": 2e-05,
"loss": 0.6133023262023926,
"step": 4580
},
{
"epoch": 0.3488106999012083,
"grad_norm": 0.9218128006407286,
"learning_rate": 2e-05,
"loss": 0.6523171424865722,
"step": 4590
},
{
"epoch": 0.3495706360665704,
"grad_norm": 0.8974549965072942,
"learning_rate": 2e-05,
"loss": 0.6212499141693115,
"step": 4600
},
{
"epoch": 0.35033057223193254,
"grad_norm": 1.1645952891980642,
"learning_rate": 2e-05,
"loss": 0.6346889495849609,
"step": 4610
},
{
"epoch": 0.3510905083972946,
"grad_norm": 1.4040413653568118,
"learning_rate": 2e-05,
"loss": 0.6068562507629395,
"step": 4620
},
{
"epoch": 0.35185044456265674,
"grad_norm": 1.4596732898533262,
"learning_rate": 2e-05,
"loss": 0.6284614562988281,
"step": 4630
},
{
"epoch": 0.35261038072801887,
"grad_norm": 1.0155555804241654,
"learning_rate": 2e-05,
"loss": 0.6058677673339844,
"step": 4640
},
{
"epoch": 0.35337031689338094,
"grad_norm": 1.2090689072922334,
"learning_rate": 2e-05,
"loss": 0.6285273551940918,
"step": 4650
},
{
"epoch": 0.35413025305874307,
"grad_norm": 1.2787108770004478,
"learning_rate": 2e-05,
"loss": 0.6266864776611328,
"step": 4660
},
{
"epoch": 0.3548901892241052,
"grad_norm": 0.9197107300964413,
"learning_rate": 2e-05,
"loss": 0.6152817726135253,
"step": 4670
},
{
"epoch": 0.3556501253894673,
"grad_norm": 18.919166842611308,
"learning_rate": 2e-05,
"loss": 0.6119517326354981,
"step": 4680
},
{
"epoch": 0.3564100615548294,
"grad_norm": 1.4058008566476412,
"learning_rate": 2e-05,
"loss": 0.6174506187438965,
"step": 4690
},
{
"epoch": 0.3571699977201915,
"grad_norm": 0.9988437823137366,
"learning_rate": 2e-05,
"loss": 0.6167577266693115,
"step": 4700
},
{
"epoch": 0.3579299338855536,
"grad_norm": 1.1556806181787362,
"learning_rate": 2e-05,
"loss": 0.6333663940429688,
"step": 4710
},
{
"epoch": 0.35868987005091574,
"grad_norm": 1.1098292654559276,
"learning_rate": 2e-05,
"loss": 0.617949104309082,
"step": 4720
},
{
"epoch": 0.3594498062162778,
"grad_norm": 1.9299990273309398,
"learning_rate": 2e-05,
"loss": 0.6180627346038818,
"step": 4730
},
{
"epoch": 0.36020974238163994,
"grad_norm": 1.554199460054541,
"learning_rate": 2e-05,
"loss": 0.6386888504028321,
"step": 4740
},
{
"epoch": 0.36096967854700207,
"grad_norm": 1.082013002346371,
"learning_rate": 2e-05,
"loss": 0.6285783767700195,
"step": 4750
},
{
"epoch": 0.36172961471236414,
"grad_norm": 1.0738717257172536,
"learning_rate": 2e-05,
"loss": 0.626548957824707,
"step": 4760
},
{
"epoch": 0.3624895508777263,
"grad_norm": 1.3833397306073538,
"learning_rate": 2e-05,
"loss": 0.6229054450988769,
"step": 4770
},
{
"epoch": 0.3632494870430884,
"grad_norm": 0.8969551611024454,
"learning_rate": 2e-05,
"loss": 0.6218652248382568,
"step": 4780
},
{
"epoch": 0.3640094232084505,
"grad_norm": 0.8497565266191878,
"learning_rate": 2e-05,
"loss": 0.6254983425140381,
"step": 4790
},
{
"epoch": 0.3647693593738126,
"grad_norm": 0.8030680111207589,
"learning_rate": 2e-05,
"loss": 0.5967194557189941,
"step": 4800
},
{
"epoch": 0.36552929553917474,
"grad_norm": 2.3725041063513617,
"learning_rate": 2e-05,
"loss": 0.6141952514648438,
"step": 4810
},
{
"epoch": 0.3662892317045368,
"grad_norm": 1.079090175426383,
"learning_rate": 2e-05,
"loss": 0.6035738945007324,
"step": 4820
},
{
"epoch": 0.36704916786989894,
"grad_norm": 0.7350181579763702,
"learning_rate": 2e-05,
"loss": 0.5995299339294433,
"step": 4830
},
{
"epoch": 0.367809104035261,
"grad_norm": 0.6998723708900272,
"learning_rate": 2e-05,
"loss": 0.6088802337646484,
"step": 4840
},
{
"epoch": 0.36856904020062314,
"grad_norm": 4.656688077966762,
"learning_rate": 2e-05,
"loss": 0.6377917766571045,
"step": 4850
},
{
"epoch": 0.3693289763659853,
"grad_norm": 0.9362102615883362,
"learning_rate": 2e-05,
"loss": 0.5998881816864013,
"step": 4860
},
{
"epoch": 0.37008891253134735,
"grad_norm": 0.7777138582558922,
"learning_rate": 2e-05,
"loss": 0.6036483764648437,
"step": 4870
},
{
"epoch": 0.3708488486967095,
"grad_norm": 0.7566663050177238,
"learning_rate": 2e-05,
"loss": 0.5968652725219726,
"step": 4880
},
{
"epoch": 0.3716087848620716,
"grad_norm": 0.814818990477887,
"learning_rate": 2e-05,
"loss": 0.6251842498779296,
"step": 4890
},
{
"epoch": 0.3723687210274337,
"grad_norm": 0.7635616353510719,
"learning_rate": 2e-05,
"loss": 0.619797658920288,
"step": 4900
},
{
"epoch": 0.3731286571927958,
"grad_norm": 0.7536889302602852,
"learning_rate": 2e-05,
"loss": 0.6170392513275147,
"step": 4910
},
{
"epoch": 0.37388859335815794,
"grad_norm": 0.9099848828220006,
"learning_rate": 2e-05,
"loss": 0.6249567985534668,
"step": 4920
},
{
"epoch": 0.37464852952352,
"grad_norm": 15.041580861087684,
"learning_rate": 2e-05,
"loss": 0.6162414073944091,
"step": 4930
},
{
"epoch": 0.37540846568888214,
"grad_norm": 0.8261042579927552,
"learning_rate": 2e-05,
"loss": 0.6113101005554199,
"step": 4940
},
{
"epoch": 0.3761684018542442,
"grad_norm": 1.0699305038057592,
"learning_rate": 2e-05,
"loss": 0.6028437614440918,
"step": 4950
},
{
"epoch": 0.37692833801960635,
"grad_norm": 0.7785165124192742,
"learning_rate": 2e-05,
"loss": 0.614793348312378,
"step": 4960
},
{
"epoch": 0.3776882741849685,
"grad_norm": 0.8881799045536933,
"learning_rate": 2e-05,
"loss": 0.6225779533386231,
"step": 4970
},
{
"epoch": 0.37844821035033055,
"grad_norm": 1.4278565114637936,
"learning_rate": 2e-05,
"loss": 0.5902219295501709,
"step": 4980
},
{
"epoch": 0.3792081465156927,
"grad_norm": 0.8827177845448362,
"learning_rate": 2e-05,
"loss": 0.5980781078338623,
"step": 4990
},
{
"epoch": 0.3799680826810548,
"grad_norm": 1.0285813124515226,
"learning_rate": 2e-05,
"loss": 0.6059083461761474,
"step": 5000
},
{
"epoch": 0.3807280188464169,
"grad_norm": 0.820793577407267,
"learning_rate": 2e-05,
"loss": 0.6184280395507813,
"step": 5010
},
{
"epoch": 0.381487955011779,
"grad_norm": 10.840810148333725,
"learning_rate": 2e-05,
"loss": 0.6353204727172852,
"step": 5020
},
{
"epoch": 0.38224789117714114,
"grad_norm": 0.8371736556282744,
"learning_rate": 2e-05,
"loss": 0.6085888862609863,
"step": 5030
},
{
"epoch": 0.3830078273425032,
"grad_norm": 2.557319897319922,
"learning_rate": 2e-05,
"loss": 0.6260895729064941,
"step": 5040
},
{
"epoch": 0.38376776350786534,
"grad_norm": 0.8642101569233044,
"learning_rate": 2e-05,
"loss": 0.6153550148010254,
"step": 5050
},
{
"epoch": 0.3845276996732275,
"grad_norm": 1.1032686543382764,
"learning_rate": 2e-05,
"loss": 0.6046747207641602,
"step": 5060
},
{
"epoch": 0.38528763583858955,
"grad_norm": 0.9365073025939038,
"learning_rate": 2e-05,
"loss": 0.5926566123962402,
"step": 5070
},
{
"epoch": 0.3860475720039517,
"grad_norm": 1.0246404884418376,
"learning_rate": 2e-05,
"loss": 0.6223001480102539,
"step": 5080
},
{
"epoch": 0.38680750816931375,
"grad_norm": 0.9300482280491423,
"learning_rate": 1.9999810479633703e-05,
"loss": 0.5798623561859131,
"step": 5090
},
{
"epoch": 0.3875674443346759,
"grad_norm": 1.5052808714934596,
"learning_rate": 1.997707677020786e-05,
"loss": 0.6030573844909668,
"step": 5100
},
{
"epoch": 0.388327380500038,
"grad_norm": 2.2568145779027002,
"learning_rate": 1.9916538196124637e-05,
"loss": 0.5952090740203857,
"step": 5110
},
{
"epoch": 0.3890873166654001,
"grad_norm": 2.466891095299882,
"learning_rate": 1.9818425303863402e-05,
"loss": 0.5759177684783936,
"step": 5120
},
{
"epoch": 0.3898472528307622,
"grad_norm": 2.45062727148961,
"learning_rate": 1.9683111732582887e-05,
"loss": 0.618531608581543,
"step": 5130
},
{
"epoch": 0.39060718899612434,
"grad_norm": 1.366275883059058,
"learning_rate": 1.951111279120706e-05,
"loss": 0.6189343929290771,
"step": 5140
},
{
"epoch": 0.3913671251614864,
"grad_norm": 0.7714894127012956,
"learning_rate": 1.9303083495996028e-05,
"loss": 0.6146251201629639,
"step": 5150
},
{
"epoch": 0.39212706132684855,
"grad_norm": 1.0375220086419477,
"learning_rate": 1.9059816076075482e-05,
"loss": 0.5992267608642579,
"step": 5160
},
{
"epoch": 0.3928869974922107,
"grad_norm": 0.8086435805016753,
"learning_rate": 1.8782236956424142e-05,
"loss": 0.614351463317871,
"step": 5170
},
{
"epoch": 0.39364693365757275,
"grad_norm": 0.66636069495425,
"learning_rate": 1.8471403229808897e-05,
"loss": 0.6243559837341308,
"step": 5180
},
{
"epoch": 0.3944068698229349,
"grad_norm": 0.7366348245062455,
"learning_rate": 1.8128498631103333e-05,
"loss": 0.5951040267944336,
"step": 5190
},
{
"epoch": 0.395166805988297,
"grad_norm": 0.9281118404991883,
"learning_rate": 1.7754829029320507e-05,
"loss": 0.6028473854064942,
"step": 5200
},
{
"epoch": 0.3959267421536591,
"grad_norm": 0.8017124304344392,
"learning_rate": 1.7351817454527473e-05,
"loss": 0.6082664012908936,
"step": 5210
},
{
"epoch": 0.3966866783190212,
"grad_norm": 0.7933405963496856,
"learning_rate": 1.6920998678580345e-05,
"loss": 0.5924901008605957,
"step": 5220
},
{
"epoch": 0.3974466144843833,
"grad_norm": 1.2633738997360422,
"learning_rate": 1.6464013370317854e-05,
"loss": 0.5934760093688964,
"step": 5230
},
{
"epoch": 0.3982065506497454,
"grad_norm": 1.134973888005499,
"learning_rate": 1.5982601847471915e-05,
"loss": 0.5857083320617675,
"step": 5240
},
{
"epoch": 0.39896648681510755,
"grad_norm": 0.9031779091890493,
"learning_rate": 1.5478597449089553e-05,
"loss": 0.6224403381347656,
"step": 5250
},
{
"epoch": 0.3997264229804696,
"grad_norm": 0.7756815211507662,
"learning_rate": 1.4953919553705725e-05,
"loss": 0.5830935955047607,
"step": 5260
},
{
"epoch": 0.40048635914583175,
"grad_norm": 0.7202722096336279,
"learning_rate": 1.441056626985559e-05,
"loss": 0.5759371757507324,
"step": 5270
},
{
"epoch": 0.4012462953111939,
"grad_norm": 1.3219063697681208,
"learning_rate": 1.3850606826762668e-05,
"loss": 0.6292336940765381,
"step": 5280
},
{
"epoch": 0.40200623147655595,
"grad_norm": 3.585669110053963,
"learning_rate": 1.3276173694181128e-05,
"loss": 0.6056513786315918,
"step": 5290
},
{
"epoch": 0.4027661676419181,
"grad_norm": 0.7746634353474207,
"learning_rate": 1.2689454461401822e-05,
"loss": 0.5831715583801269,
"step": 5300
},
{
"epoch": 0.4035261038072802,
"grad_norm": 0.7436771804000073,
"learning_rate": 1.2092683506348939e-05,
"loss": 0.5856090068817139,
"step": 5310
},
{
"epoch": 0.4042860399726423,
"grad_norm": 0.8243939991462585,
"learning_rate": 1.14881334864935e-05,
"loss": 0.6044112205505371,
"step": 5320
},
{
"epoch": 0.4050459761380044,
"grad_norm": 1.6562803048647476,
"learning_rate": 1.0878106683988481e-05,
"loss": 0.5971261024475097,
"step": 5330
},
{
"epoch": 0.40580591230336654,
"grad_norm": 4.910974949865658,
"learning_rate": 1.026492623798551e-05,
"loss": 0.5943801879882813,
"step": 5340
},
{
"epoch": 0.4065658484687286,
"grad_norm": 1.0253363768549901,
"learning_rate": 9.650927297522743e-06,
"loss": 0.6059672355651855,
"step": 5350
},
{
"epoch": 0.40732578463409075,
"grad_norm": 0.8284259223613327,
"learning_rate": 9.038448128676012e-06,
"loss": 0.5987912654876709,
"step": 5360
},
{
"epoch": 0.4080857207994528,
"grad_norm": 0.6774104652800589,
"learning_rate": 8.42982120983949e-06,
"loss": 0.6103313446044922,
"step": 5370
},
{
"epoch": 0.40884565696481495,
"grad_norm": 0.7602205308095877,
"learning_rate": 7.827364349047303e-06,
"loss": 0.597540283203125,
"step": 5380
},
{
"epoch": 0.4096055931301771,
"grad_norm": 1.2856730064413302,
"learning_rate": 7.233371857163706e-06,
"loss": 0.5856067657470703,
"step": 5390
},
{
"epoch": 0.41036552929553916,
"grad_norm": 0.8011635114293878,
"learning_rate": 6.650105810556492e-06,
"loss": 0.6171666145324707,
"step": 5400
},
{
"epoch": 0.4111254654609013,
"grad_norm": 0.7272457998770312,
"learning_rate": 6.079787436527681e-06,
"loss": 0.5883962631225585,
"step": 5410
},
{
"epoch": 0.4118854016262634,
"grad_norm": 1.0058818569289092,
"learning_rate": 5.524588654307961e-06,
"loss": 0.5919206619262696,
"step": 5420
},
{
"epoch": 0.4126453377916255,
"grad_norm": 0.9128994074260873,
"learning_rate": 4.986623803828985e-06,
"loss": 0.5826586723327637,
"step": 5430
},
{
"epoch": 0.4134052739569876,
"grad_norm": 0.6882162093119,
"learning_rate": 4.467941593772423e-06,
"loss": 0.5986693382263184,
"step": 5440
},
{
"epoch": 0.41416521012234975,
"grad_norm": 0.9872712818548395,
"learning_rate": 3.9705172995597606e-06,
"loss": 0.6001573085784913,
"step": 5450
},
{
"epoch": 0.4149251462877118,
"grad_norm": 0.7690240775482445,
"learning_rate": 3.4962452409948346e-06,
"loss": 0.5999698638916016,
"step": 5460
},
{
"epoch": 0.41568508245307395,
"grad_norm": 2.2856959583617873,
"learning_rate": 3.046931568206211e-06,
"loss": 0.5832877159118652,
"step": 5470
},
{
"epoch": 0.416445018618436,
"grad_norm": 1.072972601301357,
"learning_rate": 2.624287383362404e-06,
"loss": 0.5786107063293457,
"step": 5480
},
{
"epoch": 0.41720495478379815,
"grad_norm": 0.9233043740645318,
"learning_rate": 2.2299222243541845e-06,
"loss": 0.5964154243469239,
"step": 5490
},
{
"epoch": 0.4179648909491603,
"grad_norm": 1.0619589159137681,
"learning_rate": 1.8653379352598246e-06,
"loss": 0.6109145641326904,
"step": 5500
},
{
"epoch": 0.41872482711452236,
"grad_norm": 0.9280177062989923,
"learning_rate": 1.5319229469361075e-06,
"loss": 0.5728150844573975,
"step": 5510
},
{
"epoch": 0.4194847632798845,
"grad_norm": 0.7101698507077455,
"learning_rate": 1.2309469895160917e-06,
"loss": 0.583997106552124,
"step": 5520
},
{
"epoch": 0.4202446994452466,
"grad_norm": 0.7342117836819672,
"learning_rate": 9.635562569497974e-07,
"loss": 0.5807370662689209,
"step": 5530
},
{
"epoch": 0.4210046356106087,
"grad_norm": 0.7563653779635469,
"learning_rate": 7.307690420024867e-07,
"loss": 0.5899463653564453,
"step": 5540
},
{
"epoch": 0.4217645717759708,
"grad_norm": 0.6755406316412828,
"learning_rate": 5.334718583335942e-07,
"loss": 0.5911076545715332,
"step": 5550
},
{
"epoch": 0.42252450794133295,
"grad_norm": 0.9663205487939921,
"learning_rate": 3.7241606442442763e-07,
"loss": 0.5639224529266358,
"step": 5560
},
{
"epoch": 0.423284444106695,
"grad_norm": 0.8109774205713243,
"learning_rate": 2.482150022115879e-07,
"loss": 0.5898496627807617,
"step": 5570
},
{
"epoch": 0.42404438027205715,
"grad_norm": 0.9210917079827727,
"learning_rate": 1.6134166132293174e-07,
"loss": 0.5857495784759521,
"step": 5580
},
{
"epoch": 0.4248043164374193,
"grad_norm": 0.9024389102285274,
"learning_rate": 1.1212687781127357e-07,
"loss": 0.5854604721069336,
"step": 5590
},
{
"epoch": 0.42533627175317273,
"step": 5597,
"total_flos": 2650292248117248.0,
"train_loss": 0.05427270457854074,
"train_runtime": 7399.4304,
"train_samples_per_second": 9.077,
"train_steps_per_second": 0.756
}
],
"logging_steps": 10,
"max_steps": 5597,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 636,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 2650292248117248.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}