PEFT
Safetensors
python_tutor / trainer_state.json
Cherry4321's picture
Upload fine-tuned Qwen 2.5 Coder 7B
b09928a verified
Raw
History Blame Contribute Delete
177 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.022886075949367,
"eval_steps": 500,
"global_step": 5050,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0010126582278481013,
"grad_norm": 0.2890350818634033,
"learning_rate": 6.4e-07,
"loss": 0.4252,
"step": 5
},
{
"epoch": 0.0020253164556962027,
"grad_norm": 0.38236403465270996,
"learning_rate": 1.44e-06,
"loss": 0.5874,
"step": 10
},
{
"epoch": 0.0030379746835443038,
"grad_norm": 0.35676905512809753,
"learning_rate": 2.24e-06,
"loss": 0.6046,
"step": 15
},
{
"epoch": 0.004050632911392405,
"grad_norm": 0.24803373217582703,
"learning_rate": 3.0399999999999997e-06,
"loss": 0.5374,
"step": 20
},
{
"epoch": 0.005063291139240506,
"grad_norm": 0.24468094110488892,
"learning_rate": 3.84e-06,
"loss": 0.5436,
"step": 25
},
{
"epoch": 0.0060759493670886075,
"grad_norm": 0.23739802837371826,
"learning_rate": 4.64e-06,
"loss": 0.5003,
"step": 30
},
{
"epoch": 0.007088607594936709,
"grad_norm": 0.2143125683069229,
"learning_rate": 5.4400000000000004e-06,
"loss": 0.4294,
"step": 35
},
{
"epoch": 0.00810126582278481,
"grad_norm": 0.2776055335998535,
"learning_rate": 6.2399999999999995e-06,
"loss": 0.4609,
"step": 40
},
{
"epoch": 0.009113924050632912,
"grad_norm": 0.2547096312046051,
"learning_rate": 7.0399999999999995e-06,
"loss": 0.4772,
"step": 45
},
{
"epoch": 0.010126582278481013,
"grad_norm": 0.7673835158348083,
"learning_rate": 7.84e-06,
"loss": 0.5448,
"step": 50
},
{
"epoch": 0.011139240506329114,
"grad_norm": 0.10860913246870041,
"learning_rate": 7.874332644514525e-06,
"loss": 0.4309,
"step": 55
},
{
"epoch": 0.012151898734177215,
"grad_norm": 0.14334945380687714,
"learning_rate": 7.37731170200806e-06,
"loss": 0.5759,
"step": 60
},
{
"epoch": 0.013164556962025316,
"grad_norm": 0.15437713265419006,
"learning_rate": 6.549695958994759e-06,
"loss": 0.4921,
"step": 65
},
{
"epoch": 0.014177215189873417,
"grad_norm": 0.17790572345256805,
"learning_rate": 5.4724982107387124e-06,
"loss": 0.4864,
"step": 70
},
{
"epoch": 0.015189873417721518,
"grad_norm": 0.16902197897434235,
"learning_rate": 4.251162078117254e-06,
"loss": 0.4801,
"step": 75
},
{
"epoch": 0.01620253164556962,
"grad_norm": 0.16751091182231903,
"learning_rate": 3.0052404513405813e-06,
"loss": 0.4284,
"step": 80
},
{
"epoch": 0.017215189873417722,
"grad_norm": 0.19641244411468506,
"learning_rate": 1.8566928200840125e-06,
"loss": 0.4211,
"step": 85
},
{
"epoch": 0.018227848101265823,
"grad_norm": 0.2027222365140915,
"learning_rate": 9.179470288968434e-07,
"loss": 0.4226,
"step": 90
},
{
"epoch": 0.019240506329113925,
"grad_norm": 0.25514501333236694,
"learning_rate": 2.808940564469946e-07,
"loss": 0.4273,
"step": 95
},
{
"epoch": 0.020253164556962026,
"grad_norm": 0.5317699909210205,
"learning_rate": 7.893086286913764e-09,
"loss": 0.4665,
"step": 100
},
{
"epoch": 0.021265822784810127,
"grad_norm": 0.09557212144136429,
"learning_rate": 7.936391428253735e-06,
"loss": 0.4278,
"step": 105
},
{
"epoch": 0.022278481012658228,
"grad_norm": 0.13262175023555756,
"learning_rate": 7.924105832633914e-06,
"loss": 0.5694,
"step": 110
},
{
"epoch": 0.02329113924050633,
"grad_norm": 0.1363184005022049,
"learning_rate": 7.910747425661406e-06,
"loss": 0.5771,
"step": 115
},
{
"epoch": 0.02430379746835443,
"grad_norm": 0.15607652068138123,
"learning_rate": 7.896319859391291e-06,
"loss": 0.4534,
"step": 120
},
{
"epoch": 0.02531645569620253,
"grad_norm": 0.1568451225757599,
"learning_rate": 7.880827078176131e-06,
"loss": 0.4774,
"step": 125
},
{
"epoch": 0.026329113924050632,
"grad_norm": 0.1714460253715515,
"learning_rate": 7.864273317587615e-06,
"loss": 0.4163,
"step": 130
},
{
"epoch": 0.027341772151898733,
"grad_norm": 0.1982555091381073,
"learning_rate": 7.846663103258601e-06,
"loss": 0.424,
"step": 135
},
{
"epoch": 0.028354430379746835,
"grad_norm": 0.20806045830249786,
"learning_rate": 7.828001249645847e-06,
"loss": 0.4141,
"step": 140
},
{
"epoch": 0.029367088607594936,
"grad_norm": 0.29137369990348816,
"learning_rate": 7.808292858713795e-06,
"loss": 0.4502,
"step": 145
},
{
"epoch": 0.030379746835443037,
"grad_norm": 0.4427628517150879,
"learning_rate": 7.787543318539739e-06,
"loss": 0.4446,
"step": 150
},
{
"epoch": 0.03139240506329114,
"grad_norm": 0.10811079293489456,
"learning_rate": 7.765758301840778e-06,
"loss": 0.4315,
"step": 155
},
{
"epoch": 0.03240506329113924,
"grad_norm": 0.1357722282409668,
"learning_rate": 7.742943764422958e-06,
"loss": 0.5206,
"step": 160
},
{
"epoch": 0.033417721518987344,
"grad_norm": 0.1358698010444641,
"learning_rate": 7.719105943553005e-06,
"loss": 0.5026,
"step": 165
},
{
"epoch": 0.034430379746835445,
"grad_norm": 0.1749633401632309,
"learning_rate": 7.694251356253128e-06,
"loss": 0.514,
"step": 170
},
{
"epoch": 0.035443037974683546,
"grad_norm": 0.15725477039813995,
"learning_rate": 7.668386797519319e-06,
"loss": 0.4445,
"step": 175
},
{
"epoch": 0.03645569620253165,
"grad_norm": 0.1692255735397339,
"learning_rate": 7.641519338463685e-06,
"loss": 0.4387,
"step": 180
},
{
"epoch": 0.03746835443037975,
"grad_norm": 0.21342308819293976,
"learning_rate": 7.613656324381259e-06,
"loss": 0.4048,
"step": 185
},
{
"epoch": 0.03848101265822785,
"grad_norm": 0.2070913016796112,
"learning_rate": 7.584805372741881e-06,
"loss": 0.4175,
"step": 190
},
{
"epoch": 0.03949367088607595,
"grad_norm": 0.3016558885574341,
"learning_rate": 7.5549743711076545e-06,
"loss": 0.4523,
"step": 195
},
{
"epoch": 0.04050632911392405,
"grad_norm": 0.6172907948493958,
"learning_rate": 7.524171474976565e-06,
"loss": 0.4796,
"step": 200
},
{
"epoch": 0.04151898734177215,
"grad_norm": 0.13224560022354126,
"learning_rate": 7.492405105552852e-06,
"loss": 0.4505,
"step": 205
},
{
"epoch": 0.042531645569620254,
"grad_norm": 0.11821311712265015,
"learning_rate": 7.459683947444728e-06,
"loss": 0.4879,
"step": 210
},
{
"epoch": 0.043544303797468355,
"grad_norm": 0.14326579868793488,
"learning_rate": 7.426016946290103e-06,
"loss": 0.5523,
"step": 215
},
{
"epoch": 0.044556962025316456,
"grad_norm": 0.14366313815116882,
"learning_rate": 7.391413306310926e-06,
"loss": 0.4587,
"step": 220
},
{
"epoch": 0.04556962025316456,
"grad_norm": 0.174359992146492,
"learning_rate": 7.355882487796858e-06,
"loss": 0.4058,
"step": 225
},
{
"epoch": 0.04658227848101266,
"grad_norm": 0.18399442732334137,
"learning_rate": 7.319434204518904e-06,
"loss": 0.4233,
"step": 230
},
{
"epoch": 0.04759493670886076,
"grad_norm": 0.19887953996658325,
"learning_rate": 7.282078421073778e-06,
"loss": 0.4057,
"step": 235
},
{
"epoch": 0.04860759493670886,
"grad_norm": 0.2326246052980423,
"learning_rate": 7.24382535015968e-06,
"loss": 0.3484,
"step": 240
},
{
"epoch": 0.04962025316455696,
"grad_norm": 0.2785189747810364,
"learning_rate": 7.204685449784243e-06,
"loss": 0.4128,
"step": 245
},
{
"epoch": 0.05063291139240506,
"grad_norm": 0.6590746641159058,
"learning_rate": 7.16466942040542e-06,
"loss": 0.4627,
"step": 250
},
{
"epoch": 0.051645569620253164,
"grad_norm": 0.10315769165754318,
"learning_rate": 7.123788202006087e-06,
"loss": 0.3976,
"step": 255
},
{
"epoch": 0.052658227848101265,
"grad_norm": 0.12270621210336685,
"learning_rate": 7.082052971103157e-06,
"loss": 0.4965,
"step": 260
},
{
"epoch": 0.053670886075949366,
"grad_norm": 0.1367034763097763,
"learning_rate": 7.039475137692034e-06,
"loss": 0.5032,
"step": 265
},
{
"epoch": 0.05468354430379747,
"grad_norm": 0.15129856765270233,
"learning_rate": 6.996066342127237e-06,
"loss": 0.5059,
"step": 270
},
{
"epoch": 0.05569620253164557,
"grad_norm": 0.17595066130161285,
"learning_rate": 6.951838451940039e-06,
"loss": 0.4856,
"step": 275
},
{
"epoch": 0.05670886075949367,
"grad_norm": 0.18321171402931213,
"learning_rate": 6.906803558594003e-06,
"loss": 0.4277,
"step": 280
},
{
"epoch": 0.05772151898734177,
"grad_norm": 0.1897234469652176,
"learning_rate": 6.8609739741793015e-06,
"loss": 0.3998,
"step": 285
},
{
"epoch": 0.05873417721518987,
"grad_norm": 0.2459779679775238,
"learning_rate": 6.814362228046707e-06,
"loss": 0.398,
"step": 290
},
{
"epoch": 0.05974683544303797,
"grad_norm": 0.27000924944877625,
"learning_rate": 6.7669810633821976e-06,
"loss": 0.4316,
"step": 295
},
{
"epoch": 0.060759493670886074,
"grad_norm": 0.5964180827140808,
"learning_rate": 6.718843433723092e-06,
"loss": 0.3984,
"step": 300
},
{
"epoch": 0.061772151898734175,
"grad_norm": 0.10210335999727249,
"learning_rate": 6.6699624994166895e-06,
"loss": 0.4643,
"step": 305
},
{
"epoch": 0.06278481012658228,
"grad_norm": 0.12209000438451767,
"learning_rate": 6.6203516240223575e-06,
"loss": 0.51,
"step": 310
},
{
"epoch": 0.06379746835443038,
"grad_norm": 0.15346437692642212,
"learning_rate": 6.570024370658066e-06,
"loss": 0.4933,
"step": 315
},
{
"epoch": 0.06481012658227848,
"grad_norm": 0.20867754518985748,
"learning_rate": 6.51899449829237e-06,
"loss": 0.526,
"step": 320
},
{
"epoch": 0.06582278481012659,
"grad_norm": 0.19501474499702454,
"learning_rate": 6.4672759579828435e-06,
"loss": 0.41,
"step": 325
},
{
"epoch": 0.06683544303797469,
"grad_norm": 0.19708237051963806,
"learning_rate": 6.414882889062003e-06,
"loss": 0.46,
"step": 330
},
{
"epoch": 0.06784810126582279,
"grad_norm": 0.21802377700805664,
"learning_rate": 6.361829615271748e-06,
"loss": 0.3705,
"step": 335
},
{
"epoch": 0.06886075949367089,
"grad_norm": 0.263461709022522,
"learning_rate": 6.3081306408474e-06,
"loss": 0.4137,
"step": 340
},
{
"epoch": 0.06987341772151899,
"grad_norm": 0.290580689907074,
"learning_rate": 6.253800646552387e-06,
"loss": 0.4061,
"step": 345
},
{
"epoch": 0.07088607594936709,
"grad_norm": 0.4870627224445343,
"learning_rate": 6.198854485664664e-06,
"loss": 0.4233,
"step": 350
},
{
"epoch": 0.07189873417721519,
"grad_norm": 0.11330131441354752,
"learning_rate": 6.143307179915986e-06,
"loss": 0.4924,
"step": 355
},
{
"epoch": 0.0729113924050633,
"grad_norm": 0.12951083481311798,
"learning_rate": 6.087173915385109e-06,
"loss": 0.4488,
"step": 360
},
{
"epoch": 0.0739240506329114,
"grad_norm": 0.14505082368850708,
"learning_rate": 6.030470038346067e-06,
"loss": 0.526,
"step": 365
},
{
"epoch": 0.0749367088607595,
"grad_norm": 0.15361474454402924,
"learning_rate": 5.9732110510726595e-06,
"loss": 0.421,
"step": 370
},
{
"epoch": 0.0759493670886076,
"grad_norm": 0.1695505827665329,
"learning_rate": 5.915412607600278e-06,
"loss": 0.3937,
"step": 375
},
{
"epoch": 0.0769620253164557,
"grad_norm": 0.17293846607208252,
"learning_rate": 5.857090509446255e-06,
"loss": 0.3942,
"step": 380
},
{
"epoch": 0.0779746835443038,
"grad_norm": 0.1847795844078064,
"learning_rate": 5.7982607012898815e-06,
"loss": 0.3982,
"step": 385
},
{
"epoch": 0.0789873417721519,
"grad_norm": 0.22609223425388336,
"learning_rate": 5.738939266613299e-06,
"loss": 0.4267,
"step": 390
},
{
"epoch": 0.08,
"grad_norm": 0.2740250825881958,
"learning_rate": 5.679142423304423e-06,
"loss": 0.3658,
"step": 395
},
{
"epoch": 0.0810126582278481,
"grad_norm": 0.6746684908866882,
"learning_rate": 5.618886519223149e-06,
"loss": 0.4483,
"step": 400
},
{
"epoch": 0.0820253164556962,
"grad_norm": 0.10393477231264114,
"learning_rate": 5.558188027732001e-06,
"loss": 0.4223,
"step": 405
},
{
"epoch": 0.0830379746835443,
"grad_norm": 0.11308884620666504,
"learning_rate": 5.4970635431924805e-06,
"loss": 0.49,
"step": 410
},
{
"epoch": 0.0840506329113924,
"grad_norm": 0.15513402223587036,
"learning_rate": 5.43552977642834e-06,
"loss": 0.6736,
"step": 415
},
{
"epoch": 0.08506329113924051,
"grad_norm": 0.1626657396554947,
"learning_rate": 5.373603550156999e-06,
"loss": 0.4875,
"step": 420
},
{
"epoch": 0.08607594936708861,
"grad_norm": 0.1769031286239624,
"learning_rate": 5.31130179439039e-06,
"loss": 0.4655,
"step": 425
},
{
"epoch": 0.08708860759493671,
"grad_norm": 0.17216484248638153,
"learning_rate": 5.248641541806463e-06,
"loss": 0.3709,
"step": 430
},
{
"epoch": 0.08810126582278481,
"grad_norm": 0.19002921879291534,
"learning_rate": 5.185639923092615e-06,
"loss": 0.4099,
"step": 435
},
{
"epoch": 0.08911392405063291,
"grad_norm": 0.24008767306804657,
"learning_rate": 5.1223141622623324e-06,
"loss": 0.3837,
"step": 440
},
{
"epoch": 0.09012658227848101,
"grad_norm": 0.2640061378479004,
"learning_rate": 5.058681571946326e-06,
"loss": 0.417,
"step": 445
},
{
"epoch": 0.09113924050632911,
"grad_norm": 0.5950750708580017,
"learning_rate": 4.9947595486594205e-06,
"loss": 0.4172,
"step": 450
},
{
"epoch": 0.09215189873417722,
"grad_norm": 0.10812418907880783,
"learning_rate": 4.930565568044521e-06,
"loss": 0.4353,
"step": 455
},
{
"epoch": 0.09316455696202532,
"grad_norm": 0.11884841322898865,
"learning_rate": 4.866117180094959e-06,
"loss": 0.4917,
"step": 460
},
{
"epoch": 0.09417721518987342,
"grad_norm": 0.13334554433822632,
"learning_rate": 4.801432004356485e-06,
"loss": 0.546,
"step": 465
},
{
"epoch": 0.09518987341772152,
"grad_norm": 0.1469433307647705,
"learning_rate": 4.736527725110268e-06,
"loss": 0.5123,
"step": 470
},
{
"epoch": 0.09620253164556962,
"grad_norm": 0.15326330065727234,
"learning_rate": 4.67142208653819e-06,
"loss": 0.4421,
"step": 475
},
{
"epoch": 0.09721518987341772,
"grad_norm": 0.1922120749950409,
"learning_rate": 4.6061328878717614e-06,
"loss": 0.3653,
"step": 480
},
{
"epoch": 0.09822784810126582,
"grad_norm": 0.22665756940841675,
"learning_rate": 4.540677978525988e-06,
"loss": 0.4022,
"step": 485
},
{
"epoch": 0.09924050632911392,
"grad_norm": 0.24490761756896973,
"learning_rate": 4.47507525321952e-06,
"loss": 0.4495,
"step": 490
},
{
"epoch": 0.10025316455696202,
"grad_norm": 0.2760622203350067,
"learning_rate": 4.40934264708242e-06,
"loss": 0.3994,
"step": 495
},
{
"epoch": 0.10126582278481013,
"grad_norm": 0.4926779568195343,
"learning_rate": 4.343498130752868e-06,
"loss": 0.4233,
"step": 500
},
{
"epoch": 0.10227848101265823,
"grad_norm": 0.100213423371315,
"learning_rate": 4.277559705464176e-06,
"loss": 0.4346,
"step": 505
},
{
"epoch": 0.10329113924050633,
"grad_norm": 0.14627546072006226,
"learning_rate": 4.211545398123426e-06,
"loss": 0.4828,
"step": 510
},
{
"epoch": 0.10430379746835443,
"grad_norm": 0.1387825906276703,
"learning_rate": 4.145473256383098e-06,
"loss": 0.5354,
"step": 515
},
{
"epoch": 0.10531645569620253,
"grad_norm": 0.14282335340976715,
"learning_rate": 4.0793613437070145e-06,
"loss": 0.5289,
"step": 520
},
{
"epoch": 0.10632911392405063,
"grad_norm": 0.14829204976558685,
"learning_rate": 4.013227734431981e-06,
"loss": 0.4303,
"step": 525
},
{
"epoch": 0.10734177215189873,
"grad_norm": 0.19231268763542175,
"learning_rate": 3.947090508826431e-06,
"loss": 0.438,
"step": 530
},
{
"epoch": 0.10835443037974683,
"grad_norm": 0.20014595985412598,
"learning_rate": 3.88096774814747e-06,
"loss": 0.3829,
"step": 535
},
{
"epoch": 0.10936708860759493,
"grad_norm": 0.22177746891975403,
"learning_rate": 3.814877529697636e-06,
"loss": 0.4263,
"step": 540
},
{
"epoch": 0.11037974683544304,
"grad_norm": 0.2902078628540039,
"learning_rate": 3.748837921882746e-06,
"loss": 0.4259,
"step": 545
},
{
"epoch": 0.11139240506329114,
"grad_norm": 0.8105320334434509,
"learning_rate": 3.682866979272176e-06,
"loss": 0.4032,
"step": 550
},
{
"epoch": 0.11240506329113924,
"grad_norm": 0.1199047714471817,
"learning_rate": 3.616982737662919e-06,
"loss": 0.452,
"step": 555
},
{
"epoch": 0.11341772151898734,
"grad_norm": 0.12158358842134476,
"learning_rate": 3.551203209148783e-06,
"loss": 0.5097,
"step": 560
},
{
"epoch": 0.11443037974683544,
"grad_norm": 0.14767540991306305,
"learning_rate": 3.4855463771960563e-06,
"loss": 0.576,
"step": 565
},
{
"epoch": 0.11544303797468354,
"grad_norm": 0.14708371460437775,
"learning_rate": 3.4200301917270166e-06,
"loss": 0.4402,
"step": 570
},
{
"epoch": 0.11645569620253164,
"grad_norm": 0.16839522123336792,
"learning_rate": 3.354672564212596e-06,
"loss": 0.4255,
"step": 575
},
{
"epoch": 0.11746835443037974,
"grad_norm": 0.1900736689567566,
"learning_rate": 3.289491362775565e-06,
"loss": 0.3884,
"step": 580
},
{
"epoch": 0.11848101265822784,
"grad_norm": 0.21712316572666168,
"learning_rate": 3.2245044073055673e-06,
"loss": 0.3787,
"step": 585
},
{
"epoch": 0.11949367088607595,
"grad_norm": 0.23528164625167847,
"learning_rate": 3.159729464587337e-06,
"loss": 0.3975,
"step": 590
},
{
"epoch": 0.12050632911392405,
"grad_norm": 0.28719690442085266,
"learning_rate": 3.0951842434434397e-06,
"loss": 0.4288,
"step": 595
},
{
"epoch": 0.12151898734177215,
"grad_norm": 0.5678369402885437,
"learning_rate": 3.030886389892855e-06,
"loss": 0.4045,
"step": 600
},
{
"epoch": 0.12253164556962025,
"grad_norm": 0.10973752290010452,
"learning_rate": 2.9668534823267344e-06,
"loss": 0.4225,
"step": 605
},
{
"epoch": 0.12354430379746835,
"grad_norm": 0.11315113306045532,
"learning_rate": 2.9031030267026413e-06,
"loss": 0.5046,
"step": 610
},
{
"epoch": 0.12455696202531645,
"grad_norm": 0.13727732002735138,
"learning_rate": 2.8396524517586003e-06,
"loss": 0.5213,
"step": 615
},
{
"epoch": 0.12556962025316457,
"grad_norm": 0.15396752953529358,
"learning_rate": 2.7765191042482504e-06,
"loss": 0.4664,
"step": 620
},
{
"epoch": 0.12658227848101267,
"grad_norm": 0.18092885613441467,
"learning_rate": 2.713720244198419e-06,
"loss": 0.4453,
"step": 625
},
{
"epoch": 0.12759493670886077,
"grad_norm": 0.15994560718536377,
"learning_rate": 2.651273040190399e-06,
"loss": 0.3989,
"step": 630
},
{
"epoch": 0.12860759493670887,
"grad_norm": 0.19212107360363007,
"learning_rate": 2.5891945646662354e-06,
"loss": 0.3812,
"step": 635
},
{
"epoch": 0.12962025316455697,
"grad_norm": 0.21311832964420319,
"learning_rate": 2.527501789261288e-06,
"loss": 0.3816,
"step": 640
},
{
"epoch": 0.13063291139240507,
"grad_norm": 0.3330369293689728,
"learning_rate": 2.4662115801643535e-06,
"loss": 0.4601,
"step": 645
},
{
"epoch": 0.13164556962025317,
"grad_norm": 0.6984276175498962,
"learning_rate": 2.40534069350662e-06,
"loss": 0.4338,
"step": 650
},
{
"epoch": 0.13265822784810127,
"grad_norm": 0.11435388773679733,
"learning_rate": 2.344905770780707e-06,
"loss": 0.471,
"step": 655
},
{
"epoch": 0.13367088607594937,
"grad_norm": 0.12216036766767502,
"learning_rate": 2.284923334291051e-06,
"loss": 0.5131,
"step": 660
},
{
"epoch": 0.13468354430379748,
"grad_norm": 0.1551143079996109,
"learning_rate": 2.225409782636865e-06,
"loss": 0.5183,
"step": 665
},
{
"epoch": 0.13569620253164558,
"grad_norm": 0.1415749043226242,
"learning_rate": 2.166381386228933e-06,
"loss": 0.4561,
"step": 670
},
{
"epoch": 0.13670886075949368,
"grad_norm": 0.17465831339359283,
"learning_rate": 2.107854282841437e-06,
"loss": 0.3942,
"step": 675
},
{
"epoch": 0.13772151898734178,
"grad_norm": 0.189087375998497,
"learning_rate": 2.049844473200055e-06,
"loss": 0.4062,
"step": 680
},
{
"epoch": 0.13873417721518988,
"grad_norm": 0.18520382046699524,
"learning_rate": 1.992367816607518e-06,
"loss": 0.3883,
"step": 685
},
{
"epoch": 0.13974683544303798,
"grad_norm": 0.24047788977622986,
"learning_rate": 1.9354400266078383e-06,
"loss": 0.3774,
"step": 690
},
{
"epoch": 0.14075949367088608,
"grad_norm": 0.29270535707473755,
"learning_rate": 1.8790766666903818e-06,
"loss": 0.4195,
"step": 695
},
{
"epoch": 0.14177215189873418,
"grad_norm": 0.5180591940879822,
"learning_rate": 1.8232931460349637e-06,
"loss": 0.4163,
"step": 700
},
{
"epoch": 0.14278481012658228,
"grad_norm": 0.10784682631492615,
"learning_rate": 1.768104715299136e-06,
"loss": 0.4312,
"step": 705
},
{
"epoch": 0.14379746835443039,
"grad_norm": 0.1298905611038208,
"learning_rate": 1.7135264624488098e-06,
"loss": 0.4392,
"step": 710
},
{
"epoch": 0.1448101265822785,
"grad_norm": 0.12554921209812164,
"learning_rate": 1.659573308633354e-06,
"loss": 0.4694,
"step": 715
},
{
"epoch": 0.1458227848101266,
"grad_norm": 0.1577264368534088,
"learning_rate": 1.6062600041063026e-06,
"loss": 0.5188,
"step": 720
},
{
"epoch": 0.1468354430379747,
"grad_norm": 0.15761785209178925,
"learning_rate": 1.553601124192783e-06,
"loss": 0.5072,
"step": 725
},
{
"epoch": 0.1478481012658228,
"grad_norm": 0.20060843229293823,
"learning_rate": 1.5016110653047688e-06,
"loss": 0.4121,
"step": 730
},
{
"epoch": 0.1488607594936709,
"grad_norm": 0.18824118375778198,
"learning_rate": 1.4503040410052409e-06,
"loss": 0.3682,
"step": 735
},
{
"epoch": 0.149873417721519,
"grad_norm": 0.20293931663036346,
"learning_rate": 1.399694078122347e-06,
"loss": 0.3776,
"step": 740
},
{
"epoch": 0.1508860759493671,
"grad_norm": 0.25646811723709106,
"learning_rate": 1.3497950129146013e-06,
"loss": 0.4024,
"step": 745
},
{
"epoch": 0.1518987341772152,
"grad_norm": 0.45194464921951294,
"learning_rate": 1.300620487288192e-06,
"loss": 0.4111,
"step": 750
},
{
"epoch": 0.1529113924050633,
"grad_norm": 0.10275962203741074,
"learning_rate": 1.2521839450674216e-06,
"loss": 0.4164,
"step": 755
},
{
"epoch": 0.1539240506329114,
"grad_norm": 0.1149837076663971,
"learning_rate": 1.2044986283192967e-06,
"loss": 0.4825,
"step": 760
},
{
"epoch": 0.1549367088607595,
"grad_norm": 0.1332281082868576,
"learning_rate": 1.1575775737332834e-06,
"loss": 0.5074,
"step": 765
},
{
"epoch": 0.1559493670886076,
"grad_norm": 0.14507196843624115,
"learning_rate": 1.1114336090571983e-06,
"loss": 0.4963,
"step": 770
},
{
"epoch": 0.1569620253164557,
"grad_norm": 0.15169334411621094,
"learning_rate": 1.066079349590241e-06,
"loss": 0.4424,
"step": 775
},
{
"epoch": 0.1579746835443038,
"grad_norm": 0.1872536540031433,
"learning_rate": 1.02152719473409e-06,
"loss": 0.4428,
"step": 780
},
{
"epoch": 0.1589873417721519,
"grad_norm": 0.16057899594306946,
"learning_rate": 9.777893246030328e-07,
"loss": 0.3703,
"step": 785
},
{
"epoch": 0.16,
"grad_norm": 0.20160511136054993,
"learning_rate": 9.34877696694043e-07,
"loss": 0.3881,
"step": 790
},
{
"epoch": 0.1610126582278481,
"grad_norm": 0.24652443826198578,
"learning_rate": 8.928040426177222e-07,
"loss": 0.4117,
"step": 795
},
{
"epoch": 0.1620253164556962,
"grad_norm": 0.44853565096855164,
"learning_rate": 8.5157986489099e-07,
"loss": 0.3905,
"step": 800
},
{
"epoch": 0.1630379746835443,
"grad_norm": 0.10627669095993042,
"learning_rate": 8.11216433792417e-07,
"loss": 0.4148,
"step": 805
},
{
"epoch": 0.1640506329113924,
"grad_norm": 0.11713217943906784,
"learning_rate": 7.717247842810395e-07,
"loss": 0.5088,
"step": 810
},
{
"epoch": 0.1650632911392405,
"grad_norm": 0.11783024668693542,
"learning_rate": 7.331157129795116e-07,
"loss": 0.5066,
"step": 815
},
{
"epoch": 0.1660759493670886,
"grad_norm": 0.14592789113521576,
"learning_rate": 6.953997752224179e-07,
"loss": 0.465,
"step": 820
},
{
"epoch": 0.1670886075949367,
"grad_norm": 0.14927402138710022,
"learning_rate": 6.585872821705507e-07,
"loss": 0.4678,
"step": 825
},
{
"epoch": 0.1681012658227848,
"grad_norm": 0.19253399968147278,
"learning_rate": 6.226882979919406e-07,
"loss": 0.4072,
"step": 830
},
{
"epoch": 0.1691139240506329,
"grad_norm": 0.18113583326339722,
"learning_rate": 5.877126371104108e-07,
"loss": 0.3919,
"step": 835
},
{
"epoch": 0.17012658227848101,
"grad_norm": 0.22322966158390045,
"learning_rate": 5.536698615224176e-07,
"loss": 0.4318,
"step": 840
},
{
"epoch": 0.17113924050632912,
"grad_norm": 0.2774063050746918,
"learning_rate": 5.205692781828914e-07,
"loss": 0.3694,
"step": 845
},
{
"epoch": 0.17215189873417722,
"grad_norm": 0.45838215947151184,
"learning_rate": 4.884199364608101e-07,
"loss": 0.4058,
"step": 850
},
{
"epoch": 0.17316455696202532,
"grad_norm": 0.1019487977027893,
"learning_rate": 4.5723062566519077e-07,
"loss": 0.4625,
"step": 855
},
{
"epoch": 0.17417721518987342,
"grad_norm": 0.11792553216218948,
"learning_rate": 4.2700987264218335e-07,
"loss": 0.4959,
"step": 860
},
{
"epoch": 0.17518987341772152,
"grad_norm": 0.13125309348106384,
"learning_rate": 3.977659394439117e-07,
"loss": 0.5385,
"step": 865
},
{
"epoch": 0.17620253164556962,
"grad_norm": 0.16696500778198242,
"learning_rate": 3.6950682106971165e-07,
"loss": 0.489,
"step": 870
},
{
"epoch": 0.17721518987341772,
"grad_norm": 0.17606593668460846,
"learning_rate": 3.4224024328038237e-07,
"loss": 0.409,
"step": 875
},
{
"epoch": 0.17822784810126582,
"grad_norm": 0.17009328305721283,
"learning_rate": 3.159736604860357e-07,
"loss": 0.4325,
"step": 880
},
{
"epoch": 0.17924050632911392,
"grad_norm": 0.22070448100566864,
"learning_rate": 2.907142537081384e-07,
"loss": 0.4555,
"step": 885
},
{
"epoch": 0.18025316455696203,
"grad_norm": 0.22417625784873962,
"learning_rate": 2.664689286162889e-07,
"loss": 0.4207,
"step": 890
},
{
"epoch": 0.18126582278481013,
"grad_norm": 0.2567766010761261,
"learning_rate": 2.432443136402802e-07,
"loss": 0.3849,
"step": 895
},
{
"epoch": 0.18227848101265823,
"grad_norm": 0.46946048736572266,
"learning_rate": 2.2104675815795005e-07,
"loss": 0.4488,
"step": 900
},
{
"epoch": 0.18329113924050633,
"grad_norm": 0.11608965694904327,
"learning_rate": 1.9988233075932538e-07,
"loss": 0.4282,
"step": 905
},
{
"epoch": 0.18430379746835443,
"grad_norm": 0.1248716339468956,
"learning_rate": 1.7975681758753258e-07,
"loss": 0.5059,
"step": 910
},
{
"epoch": 0.18531645569620253,
"grad_norm": 0.13459211587905884,
"learning_rate": 1.6067572075691983e-07,
"loss": 0.4774,
"step": 915
},
{
"epoch": 0.18632911392405063,
"grad_norm": 0.14493702352046967,
"learning_rate": 1.4264425684883797e-07,
"loss": 0.4831,
"step": 920
},
{
"epoch": 0.18734177215189873,
"grad_norm": 0.14861345291137695,
"learning_rate": 1.256673554854757e-07,
"loss": 0.4305,
"step": 925
},
{
"epoch": 0.18835443037974683,
"grad_norm": 0.15066850185394287,
"learning_rate": 1.0974965798215307e-07,
"loss": 0.3985,
"step": 930
},
{
"epoch": 0.18936708860759494,
"grad_norm": 0.19650596380233765,
"learning_rate": 9.489551607842949e-08,
"loss": 0.3718,
"step": 935
},
{
"epoch": 0.19037974683544304,
"grad_norm": 0.21272413432598114,
"learning_rate": 8.110899074838462e-08,
"loss": 0.4131,
"step": 940
},
{
"epoch": 0.19139240506329114,
"grad_norm": 0.2408161163330078,
"learning_rate": 6.83938510903852e-08,
"loss": 0.3826,
"step": 945
},
{
"epoch": 0.19240506329113924,
"grad_norm": 0.43301400542259216,
"learning_rate": 5.675357329665464e-08,
"loss": 0.4119,
"step": 950
},
{
"epoch": 0.19341772151898734,
"grad_norm": 0.105130635201931,
"learning_rate": 4.6191339702914115e-08,
"loss": 0.4472,
"step": 955
},
{
"epoch": 0.19443037974683544,
"grad_norm": 0.11819428950548172,
"learning_rate": 3.6710037918364686e-08,
"loss": 0.5325,
"step": 960
},
{
"epoch": 0.19544303797468354,
"grad_norm": 0.13096201419830322,
"learning_rate": 2.831226003624376e-08,
"loss": 0.5285,
"step": 965
},
{
"epoch": 0.19645569620253164,
"grad_norm": 0.16127194464206696,
"learning_rate": 2.1000301925174102e-08,
"loss": 0.473,
"step": 970
},
{
"epoch": 0.19746835443037974,
"grad_norm": 0.17393045127391815,
"learning_rate": 1.4776162601492614e-08,
"loss": 0.4401,
"step": 975
},
{
"epoch": 0.19848101265822785,
"grad_norm": 0.16996300220489502,
"learning_rate": 9.641543682743058e-09,
"loss": 0.3624,
"step": 980
},
{
"epoch": 0.19949367088607595,
"grad_norm": 0.19939649105072021,
"learning_rate": 5.597848922469506e-09,
"loss": 0.3985,
"step": 985
},
{
"epoch": 0.20050632911392405,
"grad_norm": 0.2079152762889862,
"learning_rate": 2.646183826443327e-09,
"loss": 0.3647,
"step": 990
},
{
"epoch": 0.20151898734177215,
"grad_norm": 0.25647786259651184,
"learning_rate": 7.873553504289531e-10,
"loss": 0.3993,
"step": 995
},
{
"epoch": 0.20253164556962025,
"grad_norm": 0.5233524441719055,
"learning_rate": 2.1871679571461298e-11,
"loss": 0.3959,
"step": 1000
},
{
"epoch": 0.20354430379746835,
"grad_norm": 0.10694823414087296,
"learning_rate": 3.43461784275586e-06,
"loss": 0.439,
"step": 1005
},
{
"epoch": 0.20455696202531645,
"grad_norm": 0.12545199692249298,
"learning_rate": 3.3990976435169715e-06,
"loss": 0.5481,
"step": 1010
},
{
"epoch": 0.20556962025316455,
"grad_norm": 0.14048460125923157,
"learning_rate": 3.363625857573787e-06,
"loss": 0.477,
"step": 1015
},
{
"epoch": 0.20658227848101265,
"grad_norm": 0.15151582658290863,
"learning_rate": 3.3282053428050322e-06,
"loss": 0.4496,
"step": 1020
},
{
"epoch": 0.20759493670886076,
"grad_norm": 0.17850059270858765,
"learning_rate": 3.2928389529586387e-06,
"loss": 0.483,
"step": 1025
},
{
"epoch": 0.20860759493670886,
"grad_norm": 0.1780131757259369,
"learning_rate": 3.2575295374218165e-06,
"loss": 0.3974,
"step": 1030
},
{
"epoch": 0.20962025316455696,
"grad_norm": 0.3242931663990021,
"learning_rate": 3.2222799409914895e-06,
"loss": 0.4333,
"step": 1035
},
{
"epoch": 0.21063291139240506,
"grad_norm": 0.20353111624717712,
"learning_rate": 3.1870930036450966e-06,
"loss": 0.3864,
"step": 1040
},
{
"epoch": 0.21164556962025316,
"grad_norm": 0.2553807199001312,
"learning_rate": 3.151971560311782e-06,
"loss": 0.4259,
"step": 1045
},
{
"epoch": 0.21265822784810126,
"grad_norm": 0.4868514835834503,
"learning_rate": 3.1169184406439887e-06,
"loss": 0.3913,
"step": 1050
},
{
"epoch": 0.21367088607594936,
"grad_norm": 0.10679679363965988,
"learning_rate": 3.0819364687894824e-06,
"loss": 0.4395,
"step": 1055
},
{
"epoch": 0.21468354430379746,
"grad_norm": 0.1106216236948967,
"learning_rate": 3.0470284631638185e-06,
"loss": 0.458,
"step": 1060
},
{
"epoch": 0.21569620253164556,
"grad_norm": 0.1427021473646164,
"learning_rate": 3.0121972362232646e-06,
"loss": 0.5329,
"step": 1065
},
{
"epoch": 0.21670886075949367,
"grad_norm": 0.14735770225524902,
"learning_rate": 2.977445594238208e-06,
"loss": 0.4113,
"step": 1070
},
{
"epoch": 0.21772151898734177,
"grad_norm": 0.20452076196670532,
"learning_rate": 2.942776337067065e-06,
"loss": 0.4977,
"step": 1075
},
{
"epoch": 0.21873417721518987,
"grad_norm": 0.1985035240650177,
"learning_rate": 2.9081922579306996e-06,
"loss": 0.4163,
"step": 1080
},
{
"epoch": 0.21974683544303797,
"grad_norm": 0.22656890749931335,
"learning_rate": 2.87369614318738e-06,
"loss": 0.3908,
"step": 1085
},
{
"epoch": 0.22075949367088607,
"grad_norm": 0.2971227467060089,
"learning_rate": 2.8392907721082894e-06,
"loss": 0.4146,
"step": 1090
},
{
"epoch": 0.22177215189873417,
"grad_norm": 0.2702869474887848,
"learning_rate": 2.804978916653611e-06,
"loss": 0.4041,
"step": 1095
},
{
"epoch": 0.22278481012658227,
"grad_norm": 0.467525452375412,
"learning_rate": 2.7707633412491837e-06,
"loss": 0.4333,
"step": 1100
},
{
"epoch": 0.22379746835443037,
"grad_norm": 0.10149203985929489,
"learning_rate": 2.7366468025637973e-06,
"loss": 0.4055,
"step": 1105
},
{
"epoch": 0.22481012658227847,
"grad_norm": 0.1414402574300766,
"learning_rate": 2.702632049287082e-06,
"loss": 0.4703,
"step": 1110
},
{
"epoch": 0.22582278481012658,
"grad_norm": 0.14163967967033386,
"learning_rate": 2.6687218219080534e-06,
"loss": 0.5799,
"step": 1115
},
{
"epoch": 0.22683544303797468,
"grad_norm": 0.14414837956428528,
"learning_rate": 2.634918852494324e-06,
"loss": 0.5006,
"step": 1120
},
{
"epoch": 0.22784810126582278,
"grad_norm": 0.17992720007896423,
"learning_rate": 2.60122586447198e-06,
"loss": 0.4479,
"step": 1125
},
{
"epoch": 0.22886075949367088,
"grad_norm": 0.17456406354904175,
"learning_rate": 2.5676455724061683e-06,
"loss": 0.3772,
"step": 1130
},
{
"epoch": 0.22987341772151898,
"grad_norm": 0.20985274016857147,
"learning_rate": 2.5341806817823827e-06,
"loss": 0.3954,
"step": 1135
},
{
"epoch": 0.23088607594936708,
"grad_norm": 0.25371554493904114,
"learning_rate": 2.5008338887884936e-06,
"loss": 0.4388,
"step": 1140
},
{
"epoch": 0.23189873417721518,
"grad_norm": 0.3039631247520447,
"learning_rate": 2.467607880097526e-06,
"loss": 0.3854,
"step": 1145
},
{
"epoch": 0.23291139240506328,
"grad_norm": 0.6863613128662109,
"learning_rate": 2.43450533265119e-06,
"loss": 0.4592,
"step": 1150
},
{
"epoch": 0.23392405063291138,
"grad_norm": 0.12260797619819641,
"learning_rate": 2.4015289134442183e-06,
"loss": 0.4252,
"step": 1155
},
{
"epoch": 0.23493670886075949,
"grad_norm": 0.12175462394952774,
"learning_rate": 2.3686812793094836e-06,
"loss": 0.5593,
"step": 1160
},
{
"epoch": 0.2359493670886076,
"grad_norm": 0.1424095332622528,
"learning_rate": 2.3359650767039463e-06,
"loss": 0.5693,
"step": 1165
},
{
"epoch": 0.2369620253164557,
"grad_norm": 0.15477095544338226,
"learning_rate": 2.3033829414954386e-06,
"loss": 0.5274,
"step": 1170
},
{
"epoch": 0.2379746835443038,
"grad_norm": 0.15436388552188873,
"learning_rate": 2.270937498750295e-06,
"loss": 0.4195,
"step": 1175
},
{
"epoch": 0.2389873417721519,
"grad_norm": 0.16602319478988647,
"learning_rate": 2.2386313625218575e-06,
"loss": 0.3755,
"step": 1180
},
{
"epoch": 0.24,
"grad_norm": 0.17724846303462982,
"learning_rate": 2.206467135639872e-06,
"loss": 0.3982,
"step": 1185
},
{
"epoch": 0.2410126582278481,
"grad_norm": 0.25246885418891907,
"learning_rate": 2.1744474095007733e-06,
"loss": 0.4287,
"step": 1190
},
{
"epoch": 0.2420253164556962,
"grad_norm": 0.2653447091579437,
"learning_rate": 2.1425747638589207e-06,
"loss": 0.4071,
"step": 1195
},
{
"epoch": 0.2430379746835443,
"grad_norm": 0.5125321745872498,
"learning_rate": 2.1108517666187323e-06,
"loss": 0.4196,
"step": 1200
},
{
"epoch": 0.2440506329113924,
"grad_norm": 0.11791334301233292,
"learning_rate": 2.0792809736278077e-06,
"loss": 0.4958,
"step": 1205
},
{
"epoch": 0.2450632911392405,
"grad_norm": 0.12004737555980682,
"learning_rate": 2.0478649284710106e-06,
"loss": 0.4868,
"step": 1210
},
{
"epoch": 0.2460759493670886,
"grad_norm": 0.13654638826847076,
"learning_rate": 2.0166061622655283e-06,
"loss": 0.5159,
"step": 1215
},
{
"epoch": 0.2470886075949367,
"grad_norm": 0.15790481865406036,
"learning_rate": 1.9855071934569565e-06,
"loss": 0.4413,
"step": 1220
},
{
"epoch": 0.2481012658227848,
"grad_norm": 0.16185115277767181,
"learning_rate": 1.9545705276163857e-06,
"loss": 0.4031,
"step": 1225
},
{
"epoch": 0.2491139240506329,
"grad_norm": 0.19042210280895233,
"learning_rate": 1.923798657238534e-06,
"loss": 0.4108,
"step": 1230
},
{
"epoch": 0.250126582278481,
"grad_norm": 0.22220680117607117,
"learning_rate": 1.8931940615409418e-06,
"loss": 0.4323,
"step": 1235
},
{
"epoch": 0.25113924050632913,
"grad_norm": 0.22606489062309265,
"learning_rate": 1.862759206264215e-06,
"loss": 0.4315,
"step": 1240
},
{
"epoch": 0.2521518987341772,
"grad_norm": 0.40939441323280334,
"learning_rate": 1.832496543473371e-06,
"loss": 0.406,
"step": 1245
},
{
"epoch": 0.25316455696202533,
"grad_norm": 0.4478609561920166,
"learning_rate": 1.8024085113602867e-06,
"loss": 0.3977,
"step": 1250
},
{
"epoch": 0.2541772151898734,
"grad_norm": 0.09748850762844086,
"learning_rate": 1.7724975340472495e-06,
"loss": 0.4293,
"step": 1255
},
{
"epoch": 0.25518987341772154,
"grad_norm": 0.1282418966293335,
"learning_rate": 1.7427660213916555e-06,
"loss": 0.4978,
"step": 1260
},
{
"epoch": 0.2562025316455696,
"grad_norm": 0.1652185171842575,
"learning_rate": 1.7132163687918583e-06,
"loss": 0.586,
"step": 1265
},
{
"epoch": 0.25721518987341774,
"grad_norm": 0.18384326994419098,
"learning_rate": 1.683850956994171e-06,
"loss": 0.5046,
"step": 1270
},
{
"epoch": 0.2582278481012658,
"grad_norm": 0.18458569049835205,
"learning_rate": 1.654672151901055e-06,
"loss": 0.4794,
"step": 1275
},
{
"epoch": 0.25924050632911394,
"grad_norm": 0.16467557847499847,
"learning_rate": 1.6256823043805055e-06,
"loss": 0.439,
"step": 1280
},
{
"epoch": 0.260253164556962,
"grad_norm": 0.19161252677440643,
"learning_rate": 1.5968837500766536e-06,
"loss": 0.3848,
"step": 1285
},
{
"epoch": 0.26126582278481014,
"grad_norm": 0.21991680562496185,
"learning_rate": 1.5682788092215794e-06,
"loss": 0.4433,
"step": 1290
},
{
"epoch": 0.2622784810126582,
"grad_norm": 0.2523345351219177,
"learning_rate": 1.5398697864483804e-06,
"loss": 0.4174,
"step": 1295
},
{
"epoch": 0.26329113924050634,
"grad_norm": 0.5151788592338562,
"learning_rate": 1.5116589706054965e-06,
"loss": 0.3779,
"step": 1300
},
{
"epoch": 0.2643037974683544,
"grad_norm": 0.09810342639684677,
"learning_rate": 1.4836486345723009e-06,
"loss": 0.4675,
"step": 1305
},
{
"epoch": 0.26531645569620255,
"grad_norm": 0.11820008605718613,
"learning_rate": 1.4558410350759683e-06,
"loss": 0.5105,
"step": 1310
},
{
"epoch": 0.2663291139240506,
"grad_norm": 0.14217756688594818,
"learning_rate": 1.4282384125096755e-06,
"loss": 0.5005,
"step": 1315
},
{
"epoch": 0.26734177215189875,
"grad_norm": 0.16685616970062256,
"learning_rate": 1.4008429907520855e-06,
"loss": 0.4641,
"step": 1320
},
{
"epoch": 0.2683544303797468,
"grad_norm": 0.1626589149236679,
"learning_rate": 1.3736569769881748e-06,
"loss": 0.4566,
"step": 1325
},
{
"epoch": 0.26936708860759495,
"grad_norm": 0.17022362351417542,
"learning_rate": 1.346682561531409e-06,
"loss": 0.4517,
"step": 1330
},
{
"epoch": 0.270379746835443,
"grad_norm": 0.2058844119310379,
"learning_rate": 1.3199219176472757e-06,
"loss": 0.3797,
"step": 1335
},
{
"epoch": 0.27139240506329115,
"grad_norm": 0.2771223187446594,
"learning_rate": 1.293377201378186e-06,
"loss": 0.4179,
"step": 1340
},
{
"epoch": 0.2724050632911392,
"grad_norm": 0.29106470942497253,
"learning_rate": 1.267050551369767e-06,
"loss": 0.4006,
"step": 1345
},
{
"epoch": 0.27341772151898736,
"grad_norm": 0.5391093492507935,
"learning_rate": 1.240944088698561e-06,
"loss": 0.4131,
"step": 1350
},
{
"epoch": 0.27443037974683543,
"grad_norm": 0.10237446427345276,
"learning_rate": 1.2150599167011337e-06,
"loss": 0.4661,
"step": 1355
},
{
"epoch": 0.27544303797468356,
"grad_norm": 0.12482398748397827,
"learning_rate": 1.1894001208046036e-06,
"loss": 0.5128,
"step": 1360
},
{
"epoch": 0.27645569620253163,
"grad_norm": 0.12247582525014877,
"learning_rate": 1.1639667683586396e-06,
"loss": 0.4258,
"step": 1365
},
{
"epoch": 0.27746835443037976,
"grad_norm": 0.19288122653961182,
"learning_rate": 1.1387619084688917e-06,
"loss": 0.4464,
"step": 1370
},
{
"epoch": 0.27848101265822783,
"grad_norm": 0.17132723331451416,
"learning_rate": 1.1137875718318964e-06,
"loss": 0.4244,
"step": 1375
},
{
"epoch": 0.27949367088607596,
"grad_norm": 0.18715178966522217,
"learning_rate": 1.0890457705714714e-06,
"loss": 0.4194,
"step": 1380
},
{
"epoch": 0.28050632911392404,
"grad_norm": 0.2150091528892517,
"learning_rate": 1.064538498076609e-06,
"loss": 0.3551,
"step": 1385
},
{
"epoch": 0.28151898734177216,
"grad_norm": 0.21580003201961517,
"learning_rate": 1.0402677288408624e-06,
"loss": 0.4105,
"step": 1390
},
{
"epoch": 0.28253164556962024,
"grad_norm": 0.29213687777519226,
"learning_rate": 1.0162354183032715e-06,
"loss": 0.3604,
"step": 1395
},
{
"epoch": 0.28354430379746837,
"grad_norm": 0.7079854607582092,
"learning_rate": 9.924435026908207e-07,
"loss": 0.4572,
"step": 1400
},
{
"epoch": 0.28455696202531644,
"grad_norm": 0.0977628231048584,
"learning_rate": 9.688938988624373e-07,
"loss": 0.4338,
"step": 1405
},
{
"epoch": 0.28556962025316457,
"grad_norm": 0.11893545836210251,
"learning_rate": 9.455885041545535e-07,
"loss": 0.4922,
"step": 1410
},
{
"epoch": 0.28658227848101264,
"grad_norm": 0.13641275465488434,
"learning_rate": 9.225291962282442e-07,
"loss": 0.477,
"step": 1415
},
{
"epoch": 0.28759493670886077,
"grad_norm": 0.16472046077251434,
"learning_rate": 8.997178329179531e-07,
"loss": 0.4883,
"step": 1420
},
{
"epoch": 0.28860759493670884,
"grad_norm": 0.17243064939975739,
"learning_rate": 8.771562520818018e-07,
"loss": 0.3958,
"step": 1425
},
{
"epoch": 0.289620253164557,
"grad_norm": 0.1709485650062561,
"learning_rate": 8.548462714535243e-07,
"loss": 0.4144,
"step": 1430
},
{
"epoch": 0.29063291139240505,
"grad_norm": 0.20217914879322052,
"learning_rate": 8.327896884960175e-07,
"loss": 0.3837,
"step": 1435
},
{
"epoch": 0.2916455696202532,
"grad_norm": 0.19909925758838654,
"learning_rate": 8.109882802565188e-07,
"loss": 0.3884,
"step": 1440
},
{
"epoch": 0.29265822784810125,
"grad_norm": 0.25214800238609314,
"learning_rate": 7.894438032234352e-07,
"loss": 0.405,
"step": 1445
},
{
"epoch": 0.2936708860759494,
"grad_norm": 0.6173389554023743,
"learning_rate": 7.681579931848321e-07,
"loss": 0.4049,
"step": 1450
},
{
"epoch": 0.29468354430379745,
"grad_norm": 0.13174627721309662,
"learning_rate": 7.47132565088576e-07,
"loss": 0.4734,
"step": 1455
},
{
"epoch": 0.2956962025316456,
"grad_norm": 0.13449151813983917,
"learning_rate": 7.263692129041752e-07,
"loss": 0.5292,
"step": 1460
},
{
"epoch": 0.29670886075949365,
"grad_norm": 0.1471068412065506,
"learning_rate": 7.058696094862902e-07,
"loss": 0.5087,
"step": 1465
},
{
"epoch": 0.2977215189873418,
"grad_norm": 0.17206205427646637,
"learning_rate": 6.856354064399652e-07,
"loss": 0.5156,
"step": 1470
},
{
"epoch": 0.29873417721518986,
"grad_norm": 0.15566039085388184,
"learning_rate": 6.656682339875566e-07,
"loss": 0.3905,
"step": 1475
},
{
"epoch": 0.299746835443038,
"grad_norm": 0.21175558865070343,
"learning_rate": 6.459697008373886e-07,
"loss": 0.4312,
"step": 1480
},
{
"epoch": 0.30075949367088606,
"grad_norm": 0.22644248604774475,
"learning_rate": 6.265413940541511e-07,
"loss": 0.4038,
"step": 1485
},
{
"epoch": 0.3017721518987342,
"grad_norm": 0.24314475059509277,
"learning_rate": 6.073848789310232e-07,
"loss": 0.3796,
"step": 1490
},
{
"epoch": 0.30278481012658226,
"grad_norm": 0.3007188141345978,
"learning_rate": 5.885016988635709e-07,
"loss": 0.4242,
"step": 1495
},
{
"epoch": 0.3037974683544304,
"grad_norm": 0.750256359577179,
"learning_rate": 5.698933752253916e-07,
"loss": 0.4566,
"step": 1500
},
{
"epoch": 0.30481012658227846,
"grad_norm": 0.10441678762435913,
"learning_rate": 5.515614072455439e-07,
"loss": 0.4428,
"step": 1505
},
{
"epoch": 0.3058227848101266,
"grad_norm": 0.1318226307630539,
"learning_rate": 5.335072718877609e-07,
"loss": 0.538,
"step": 1510
},
{
"epoch": 0.30683544303797466,
"grad_norm": 0.1475210189819336,
"learning_rate": 5.157324237314498e-07,
"loss": 0.4972,
"step": 1515
},
{
"epoch": 0.3078481012658228,
"grad_norm": 0.13725890219211578,
"learning_rate": 4.982382948545014e-07,
"loss": 0.4501,
"step": 1520
},
{
"epoch": 0.30886075949367087,
"grad_norm": 0.16876070201396942,
"learning_rate": 4.81026294717914e-07,
"loss": 0.4272,
"step": 1525
},
{
"epoch": 0.309873417721519,
"grad_norm": 0.19023726880550385,
"learning_rate": 4.6409781005223126e-07,
"loss": 0.4076,
"step": 1530
},
{
"epoch": 0.31088607594936707,
"grad_norm": 0.1916356384754181,
"learning_rate": 4.474542047458221e-07,
"loss": 0.3823,
"step": 1535
},
{
"epoch": 0.3118987341772152,
"grad_norm": 0.23694798350334167,
"learning_rate": 4.310968197349893e-07,
"loss": 0.4093,
"step": 1540
},
{
"epoch": 0.31291139240506327,
"grad_norm": 0.26376596093177795,
"learning_rate": 4.150269728959399e-07,
"loss": 0.4143,
"step": 1545
},
{
"epoch": 0.3139240506329114,
"grad_norm": 0.5081406235694885,
"learning_rate": 3.992459589386019e-07,
"loss": 0.4261,
"step": 1550
},
{
"epoch": 0.3149367088607595,
"grad_norm": 0.10619200766086578,
"learning_rate": 3.837550493023136e-07,
"loss": 0.4679,
"step": 1555
},
{
"epoch": 0.3159493670886076,
"grad_norm": 0.1179724633693695,
"learning_rate": 3.685554920533902e-07,
"loss": 0.4763,
"step": 1560
},
{
"epoch": 0.3169620253164557,
"grad_norm": 0.13078778982162476,
"learning_rate": 3.536485117845651e-07,
"loss": 0.4345,
"step": 1565
},
{
"epoch": 0.3179746835443038,
"grad_norm": 0.17319700121879578,
"learning_rate": 3.3903530951632987e-07,
"loss": 0.4699,
"step": 1570
},
{
"epoch": 0.3189873417721519,
"grad_norm": 0.17910867929458618,
"learning_rate": 3.247170626001714e-07,
"loss": 0.4459,
"step": 1575
},
{
"epoch": 0.32,
"grad_norm": 0.18782804906368256,
"learning_rate": 3.1069492462371515e-07,
"loss": 0.3877,
"step": 1580
},
{
"epoch": 0.3210126582278481,
"grad_norm": 0.1973138451576233,
"learning_rate": 2.9697002531777937e-07,
"loss": 0.4405,
"step": 1585
},
{
"epoch": 0.3220253164556962,
"grad_norm": 0.25930511951446533,
"learning_rate": 2.8354347046536207e-07,
"loss": 0.398,
"step": 1590
},
{
"epoch": 0.3230379746835443,
"grad_norm": 0.2588368058204651,
"learning_rate": 2.7041634181254714e-07,
"loss": 0.4011,
"step": 1595
},
{
"epoch": 0.3240506329113924,
"grad_norm": 0.646274983882904,
"learning_rate": 2.5758969698134847e-07,
"loss": 0.3979,
"step": 1600
},
{
"epoch": 0.3250632911392405,
"grad_norm": 0.1041562557220459,
"learning_rate": 2.4506456938450373e-07,
"loss": 0.4247,
"step": 1605
},
{
"epoch": 0.3260759493670886,
"grad_norm": 0.12852942943572998,
"learning_rate": 2.3284196814221313e-07,
"loss": 0.5499,
"step": 1610
},
{
"epoch": 0.3270886075949367,
"grad_norm": 0.14556489884853363,
"learning_rate": 2.2092287800083632e-07,
"loss": 0.445,
"step": 1615
},
{
"epoch": 0.3281012658227848,
"grad_norm": 0.1491594910621643,
"learning_rate": 2.0930825925355423e-07,
"loss": 0.5151,
"step": 1620
},
{
"epoch": 0.3291139240506329,
"grad_norm": 0.15905530750751495,
"learning_rate": 1.9799904766300134e-07,
"loss": 0.3956,
"step": 1625
},
{
"epoch": 0.330126582278481,
"grad_norm": 0.19776026904582977,
"learning_rate": 1.8699615438587313e-07,
"loss": 0.4313,
"step": 1630
},
{
"epoch": 0.3311392405063291,
"grad_norm": 0.1957055926322937,
"learning_rate": 1.7630046589951265e-07,
"loss": 0.3696,
"step": 1635
},
{
"epoch": 0.3321518987341772,
"grad_norm": 0.24615226686000824,
"learning_rate": 1.659128439304962e-07,
"loss": 0.382,
"step": 1640
},
{
"epoch": 0.3331645569620253,
"grad_norm": 0.29114511609077454,
"learning_rate": 1.5583412538520136e-07,
"loss": 0.3869,
"step": 1645
},
{
"epoch": 0.3341772151898734,
"grad_norm": 0.455689936876297,
"learning_rate": 1.4606512228237855e-07,
"loss": 0.3963,
"step": 1650
},
{
"epoch": 0.3351898734177215,
"grad_norm": 0.09605590254068375,
"learning_rate": 1.3660662168773062e-07,
"loss": 0.4219,
"step": 1655
},
{
"epoch": 0.3362025316455696,
"grad_norm": 0.11081603169441223,
"learning_rate": 1.2745938565050264e-07,
"loss": 0.5219,
"step": 1660
},
{
"epoch": 0.3372151898734177,
"grad_norm": 0.13300614058971405,
"learning_rate": 1.1862415114208158e-07,
"loss": 0.504,
"step": 1665
},
{
"epoch": 0.3382278481012658,
"grad_norm": 0.160051167011261,
"learning_rate": 1.1010162999662043e-07,
"loss": 0.5673,
"step": 1670
},
{
"epoch": 0.3392405063291139,
"grad_norm": 0.1791842132806778,
"learning_rate": 1.0189250885369105e-07,
"loss": 0.4801,
"step": 1675
},
{
"epoch": 0.34025316455696203,
"grad_norm": 0.17864088714122772,
"learning_rate": 9.399744910296137e-08,
"loss": 0.4088,
"step": 1680
},
{
"epoch": 0.3412658227848101,
"grad_norm": 0.18893688917160034,
"learning_rate": 8.641708683090599e-08,
"loss": 0.4036,
"step": 1685
},
{
"epoch": 0.34227848101265823,
"grad_norm": 0.2292567938566208,
"learning_rate": 7.915203276956228e-08,
"loss": 0.3785,
"step": 1690
},
{
"epoch": 0.3432911392405063,
"grad_norm": 0.4482201933860779,
"learning_rate": 7.220287224732358e-08,
"loss": 0.4292,
"step": 1695
},
{
"epoch": 0.34430379746835443,
"grad_norm": 0.4612099528312683,
"learning_rate": 6.557016514177949e-08,
"loss": 0.3968,
"step": 1700
},
{
"epoch": 0.3453164556962025,
"grad_norm": 0.12291746586561203,
"learning_rate": 5.925444583460892e-08,
"loss": 0.4729,
"step": 1705
},
{
"epoch": 0.34632911392405064,
"grad_norm": 0.12686744332313538,
"learning_rate": 5.3256223168527845e-08,
"loss": 0.5565,
"step": 1710
},
{
"epoch": 0.3473417721518987,
"grad_norm": 0.12894991040229797,
"learning_rate": 4.7575980406289626e-08,
"loss": 0.513,
"step": 1715
},
{
"epoch": 0.34835443037974684,
"grad_norm": 0.16988959908485413,
"learning_rate": 4.22141751917513e-08,
"loss": 0.4667,
"step": 1720
},
{
"epoch": 0.3493670886075949,
"grad_norm": 0.15458881855010986,
"learning_rate": 3.717123951300438e-08,
"loss": 0.4129,
"step": 1725
},
{
"epoch": 0.35037974683544304,
"grad_norm": 0.2089594155550003,
"learning_rate": 3.244757966756584e-08,
"loss": 0.4471,
"step": 1730
},
{
"epoch": 0.3513924050632911,
"grad_norm": 0.1924118995666504,
"learning_rate": 2.8043576229649148e-08,
"loss": 0.3561,
"step": 1735
},
{
"epoch": 0.35240506329113924,
"grad_norm": 0.21856747567653656,
"learning_rate": 2.395958401949727e-08,
"loss": 0.3673,
"step": 1740
},
{
"epoch": 0.3534177215189873,
"grad_norm": 0.28095388412475586,
"learning_rate": 2.0195932074799305e-08,
"loss": 0.3738,
"step": 1745
},
{
"epoch": 0.35443037974683544,
"grad_norm": 0.4498268961906433,
"learning_rate": 1.6752923624179237e-08,
"loss": 0.3584,
"step": 1750
},
{
"epoch": 0.3554430379746835,
"grad_norm": 0.09736178815364838,
"learning_rate": 1.3630836062764384e-08,
"loss": 0.4111,
"step": 1755
},
{
"epoch": 0.35645569620253165,
"grad_norm": 0.13336490094661713,
"learning_rate": 1.0829920929839698e-08,
"loss": 0.5663,
"step": 1760
},
{
"epoch": 0.3574683544303797,
"grad_norm": 0.1261349618434906,
"learning_rate": 8.350403888577329e-09,
"loss": 0.4662,
"step": 1765
},
{
"epoch": 0.35848101265822785,
"grad_norm": 0.1632804125547409,
"learning_rate": 6.192484707859158e-09,
"loss": 0.4908,
"step": 1770
},
{
"epoch": 0.3594936708860759,
"grad_norm": 0.17994073033332825,
"learning_rate": 4.356337246180786e-09,
"loss": 0.4855,
"step": 1775
},
{
"epoch": 0.36050632911392405,
"grad_norm": 0.17971830070018768,
"learning_rate": 2.842109437643181e-09,
"loss": 0.4014,
"step": 1780
},
{
"epoch": 0.3615189873417721,
"grad_norm": 0.1954912692308426,
"learning_rate": 1.6499232800346597e-09,
"loss": 0.4128,
"step": 1785
},
{
"epoch": 0.36253164556962025,
"grad_norm": 0.23222939670085907,
"learning_rate": 7.798748250023024e-10,
"loss": 0.3697,
"step": 1790
},
{
"epoch": 0.3635443037974683,
"grad_norm": 0.24149201810359955,
"learning_rate": 2.3203417031236738e-10,
"loss": 0.4264,
"step": 1795
},
{
"epoch": 0.36455696202531646,
"grad_norm": 0.5369536280632019,
"learning_rate": 6.445454204584422e-12,
"loss": 0.3989,
"step": 1800
},
{
"epoch": 0.3655696202531646,
"grad_norm": 0.12446800619363785,
"learning_rate": 1.4900332043982445e-06,
"loss": 0.4249,
"step": 1805
},
{
"epoch": 0.36658227848101266,
"grad_norm": 0.1284579336643219,
"learning_rate": 1.4701166259034258e-06,
"loss": 0.4117,
"step": 1810
},
{
"epoch": 0.3675949367088608,
"grad_norm": 0.14047998189926147,
"learning_rate": 1.4503040410052409e-06,
"loss": 0.4614,
"step": 1815
},
{
"epoch": 0.36860759493670886,
"grad_norm": 0.1662559062242508,
"learning_rate": 1.4305962641214545e-06,
"loss": 0.4622,
"step": 1820
},
{
"epoch": 0.369620253164557,
"grad_norm": 0.1931740790605545,
"learning_rate": 1.410994105361584e-06,
"loss": 0.4318,
"step": 1825
},
{
"epoch": 0.37063291139240506,
"grad_norm": 0.19616465270519257,
"learning_rate": 1.3914983704935993e-06,
"loss": 0.4327,
"step": 1830
},
{
"epoch": 0.3716455696202532,
"grad_norm": 0.21658243238925934,
"learning_rate": 1.372109860910803e-06,
"loss": 0.3855,
"step": 1835
},
{
"epoch": 0.37265822784810126,
"grad_norm": 0.2893512547016144,
"learning_rate": 1.3528293735988814e-06,
"loss": 0.3777,
"step": 1840
},
{
"epoch": 0.3736708860759494,
"grad_norm": 0.3063053786754608,
"learning_rate": 1.3336577011031534e-06,
"loss": 0.4206,
"step": 1845
},
{
"epoch": 0.37468354430379747,
"grad_norm": 0.5033496022224426,
"learning_rate": 1.314595631495984e-06,
"loss": 0.3802,
"step": 1850
},
{
"epoch": 0.3756962025316456,
"grad_norm": 0.12718456983566284,
"learning_rate": 1.2956439483443959e-06,
"loss": 0.4771,
"step": 1855
},
{
"epoch": 0.37670886075949367,
"grad_norm": 0.11955064535140991,
"learning_rate": 1.276803430677852e-06,
"loss": 0.5227,
"step": 1860
},
{
"epoch": 0.3777215189873418,
"grad_norm": 0.13488246500492096,
"learning_rate": 1.2580748529562417e-06,
"loss": 0.5132,
"step": 1865
},
{
"epoch": 0.37873417721518987,
"grad_norm": 0.15127943456172943,
"learning_rate": 1.2394589850380395e-06,
"loss": 0.4784,
"step": 1870
},
{
"epoch": 0.379746835443038,
"grad_norm": 0.18862000107765198,
"learning_rate": 1.220956592148663e-06,
"loss": 0.42,
"step": 1875
},
{
"epoch": 0.3807594936708861,
"grad_norm": 0.19870607554912567,
"learning_rate": 1.2025684348490152e-06,
"loss": 0.4123,
"step": 1880
},
{
"epoch": 0.3817721518987342,
"grad_norm": 0.21817389130592346,
"learning_rate": 1.1842952690042203e-06,
"loss": 0.4247,
"step": 1885
},
{
"epoch": 0.3827848101265823,
"grad_norm": 0.20650418102741241,
"learning_rate": 1.1661378457525554e-06,
"loss": 0.3891,
"step": 1890
},
{
"epoch": 0.3837974683544304,
"grad_norm": 0.2906585931777954,
"learning_rate": 1.1480969114745728e-06,
"loss": 0.3704,
"step": 1895
},
{
"epoch": 0.3848101265822785,
"grad_norm": 0.5335307717323303,
"learning_rate": 1.1301732077624206e-06,
"loss": 0.4108,
"step": 1900
},
{
"epoch": 0.3858227848101266,
"grad_norm": 0.10487744212150574,
"learning_rate": 1.112367471389353e-06,
"loss": 0.4532,
"step": 1905
},
{
"epoch": 0.3868354430379747,
"grad_norm": 0.1331753134727478,
"learning_rate": 1.094680434279454e-06,
"loss": 0.5345,
"step": 1910
},
{
"epoch": 0.3878481012658228,
"grad_norm": 0.1402808427810669,
"learning_rate": 1.077112823477543e-06,
"loss": 0.4952,
"step": 1915
},
{
"epoch": 0.3888607594936709,
"grad_norm": 0.17083363234996796,
"learning_rate": 1.059665361119293e-06,
"loss": 0.4731,
"step": 1920
},
{
"epoch": 0.389873417721519,
"grad_norm": 0.1720193475484848,
"learning_rate": 1.042338764401545e-06,
"loss": 0.4466,
"step": 1925
},
{
"epoch": 0.3908860759493671,
"grad_norm": 0.20235848426818848,
"learning_rate": 1.025133745552825e-06,
"loss": 0.3934,
"step": 1930
},
{
"epoch": 0.3918987341772152,
"grad_norm": 0.18851284682750702,
"learning_rate": 1.008051011804071e-06,
"loss": 0.3986,
"step": 1935
},
{
"epoch": 0.3929113924050633,
"grad_norm": 0.2193611115217209,
"learning_rate": 9.91091265359559e-07,
"loss": 0.3907,
"step": 1940
},
{
"epoch": 0.3939240506329114,
"grad_norm": 0.43451058864593506,
"learning_rate": 9.742552033680392e-07,
"loss": 0.397,
"step": 1945
},
{
"epoch": 0.3949367088607595,
"grad_norm": 0.5794976353645325,
"learning_rate": 9.575435178940767e-07,
"loss": 0.409,
"step": 1950
},
{
"epoch": 0.3959493670886076,
"grad_norm": 0.11215135455131531,
"learning_rate": 9.409568958896064e-07,
"loss": 0.4264,
"step": 1955
},
{
"epoch": 0.3969620253164557,
"grad_norm": 0.12815797328948975,
"learning_rate": 9.244960191656939e-07,
"loss": 0.5694,
"step": 1960
},
{
"epoch": 0.3979746835443038,
"grad_norm": 0.12770868837833405,
"learning_rate": 9.081615643645086e-07,
"loss": 0.444,
"step": 1965
},
{
"epoch": 0.3989873417721519,
"grad_norm": 0.16232244670391083,
"learning_rate": 8.919542029315104e-07,
"loss": 0.4552,
"step": 1970
},
{
"epoch": 0.4,
"grad_norm": 0.17293521761894226,
"learning_rate": 8.758746010878457e-07,
"loss": 0.431,
"step": 1975
},
{
"epoch": 0.4010126582278481,
"grad_norm": 0.19090458750724792,
"learning_rate": 8.599234198029677e-07,
"loss": 0.4058,
"step": 1980
},
{
"epoch": 0.4020253164556962,
"grad_norm": 0.2036549597978592,
"learning_rate": 8.441013147674634e-07,
"loss": 0.4358,
"step": 1985
},
{
"epoch": 0.4030379746835443,
"grad_norm": 0.22251509130001068,
"learning_rate": 8.284089363660989e-07,
"loss": 0.3914,
"step": 1990
},
{
"epoch": 0.4040506329113924,
"grad_norm": 0.27179235219955444,
"learning_rate": 8.128469296510894e-07,
"loss": 0.4115,
"step": 1995
},
{
"epoch": 0.4050632911392405,
"grad_norm": 0.7619171142578125,
"learning_rate": 7.974159343155755e-07,
"loss": 0.4404,
"step": 2000
},
{
"epoch": 0.40607594936708863,
"grad_norm": 0.10011362284421921,
"learning_rate": 7.821165846673387e-07,
"loss": 0.4388,
"step": 2005
},
{
"epoch": 0.4070886075949367,
"grad_norm": 0.12640056014060974,
"learning_rate": 7.669495096027177e-07,
"loss": 0.4905,
"step": 2010
},
{
"epoch": 0.40810126582278483,
"grad_norm": 0.13997557759284973,
"learning_rate": 7.519153325807659e-07,
"loss": 0.524,
"step": 2015
},
{
"epoch": 0.4091139240506329,
"grad_norm": 0.14068852365016937,
"learning_rate": 7.370146715976138e-07,
"loss": 0.4809,
"step": 2020
},
{
"epoch": 0.41012658227848103,
"grad_norm": 0.18665578961372375,
"learning_rate": 7.222481391610724e-07,
"loss": 0.4055,
"step": 2025
},
{
"epoch": 0.4111392405063291,
"grad_norm": 0.16774073243141174,
"learning_rate": 7.076163422654544e-07,
"loss": 0.3714,
"step": 2030
},
{
"epoch": 0.41215189873417724,
"grad_norm": 0.2382858544588089,
"learning_rate": 6.931198823666214e-07,
"loss": 0.4396,
"step": 2035
},
{
"epoch": 0.4131645569620253,
"grad_norm": 0.21281184256076813,
"learning_rate": 6.787593553572613e-07,
"loss": 0.3661,
"step": 2040
},
{
"epoch": 0.41417721518987344,
"grad_norm": 0.26028022170066833,
"learning_rate": 6.645353515423915e-07,
"loss": 0.3836,
"step": 2045
},
{
"epoch": 0.4151898734177215,
"grad_norm": 0.8616901636123657,
"learning_rate": 6.504484556150989e-07,
"loss": 0.4813,
"step": 2050
},
{
"epoch": 0.41620253164556964,
"grad_norm": 0.1189771443605423,
"learning_rate": 6.364992466324994e-07,
"loss": 0.4214,
"step": 2055
},
{
"epoch": 0.4172151898734177,
"grad_norm": 0.11853820830583572,
"learning_rate": 6.226882979919406e-07,
"loss": 0.5015,
"step": 2060
},
{
"epoch": 0.41822784810126584,
"grad_norm": 0.13352267444133759,
"learning_rate": 6.090161774074252e-07,
"loss": 0.5043,
"step": 2065
},
{
"epoch": 0.4192405063291139,
"grad_norm": 0.1625378280878067,
"learning_rate": 5.954834468862806e-07,
"loss": 0.4998,
"step": 2070
},
{
"epoch": 0.42025316455696204,
"grad_norm": 0.182509183883667,
"learning_rate": 5.820906627060558e-07,
"loss": 0.3892,
"step": 2075
},
{
"epoch": 0.4212658227848101,
"grad_norm": 0.22160086035728455,
"learning_rate": 5.688383753916524e-07,
"loss": 0.421,
"step": 2080
},
{
"epoch": 0.42227848101265825,
"grad_norm": 0.21978890895843506,
"learning_rate": 5.557271296926962e-07,
"loss": 0.4145,
"step": 2085
},
{
"epoch": 0.4232911392405063,
"grad_norm": 0.23363438248634338,
"learning_rate": 5.427574645611436e-07,
"loss": 0.3904,
"step": 2090
},
{
"epoch": 0.42430379746835445,
"grad_norm": 0.3062392473220825,
"learning_rate": 5.299299131291297e-07,
"loss": 0.3926,
"step": 2095
},
{
"epoch": 0.4253164556962025,
"grad_norm": 0.5166776180267334,
"learning_rate": 5.172450026870528e-07,
"loss": 0.4133,
"step": 2100
},
{
"epoch": 0.42632911392405065,
"grad_norm": 0.10753131657838821,
"learning_rate": 5.047032546618974e-07,
"loss": 0.4676,
"step": 2105
},
{
"epoch": 0.4273417721518987,
"grad_norm": 0.13093893229961395,
"learning_rate": 4.923051845958013e-07,
"loss": 0.5038,
"step": 2110
},
{
"epoch": 0.42835443037974685,
"grad_norm": 0.153261199593544,
"learning_rate": 4.800513021248651e-07,
"loss": 0.5759,
"step": 2115
},
{
"epoch": 0.4293670886075949,
"grad_norm": 0.1721193641424179,
"learning_rate": 4.679421109582029e-07,
"loss": 0.4686,
"step": 2120
},
{
"epoch": 0.43037974683544306,
"grad_norm": 0.17097310721874237,
"learning_rate": 4.559781088572348e-07,
"loss": 0.418,
"step": 2125
},
{
"epoch": 0.43139240506329113,
"grad_norm": 0.2163458615541458,
"learning_rate": 4.4415978761522724e-07,
"loss": 0.432,
"step": 2130
},
{
"epoch": 0.43240506329113926,
"grad_norm": 0.20527294278144836,
"learning_rate": 4.3248763303707703e-07,
"loss": 0.3977,
"step": 2135
},
{
"epoch": 0.43341772151898733,
"grad_norm": 0.24217315018177032,
"learning_rate": 4.2096212491934225e-07,
"loss": 0.3806,
"step": 2140
},
{
"epoch": 0.43443037974683546,
"grad_norm": 0.2926272451877594,
"learning_rate": 4.0958373703051976e-07,
"loss": 0.4288,
"step": 2145
},
{
"epoch": 0.43544303797468353,
"grad_norm": 0.5068588852882385,
"learning_rate": 3.983529370915701e-07,
"loss": 0.3931,
"step": 2150
},
{
"epoch": 0.43645569620253166,
"grad_norm": 0.10863950848579407,
"learning_rate": 3.872701867566892e-07,
"loss": 0.4569,
"step": 2155
},
{
"epoch": 0.43746835443037974,
"grad_norm": 0.12185105681419373,
"learning_rate": 3.763359415943359e-07,
"loss": 0.4345,
"step": 2160
},
{
"epoch": 0.43848101265822786,
"grad_norm": 0.15586481988430023,
"learning_rate": 3.65550651068502e-07,
"loss": 0.4708,
"step": 2165
},
{
"epoch": 0.43949367088607594,
"grad_norm": 0.1565941870212555,
"learning_rate": 3.549147585202386e-07,
"loss": 0.4028,
"step": 2170
},
{
"epoch": 0.44050632911392407,
"grad_norm": 0.16850170493125916,
"learning_rate": 3.4442870114943066e-07,
"loss": 0.4342,
"step": 2175
},
{
"epoch": 0.44151898734177214,
"grad_norm": 0.2067377120256424,
"learning_rate": 3.3409290999682417e-07,
"loss": 0.4058,
"step": 2180
},
{
"epoch": 0.44253164556962027,
"grad_norm": 0.20416751503944397,
"learning_rate": 3.2390780992631194e-07,
"loss": 0.4218,
"step": 2185
},
{
"epoch": 0.44354430379746834,
"grad_norm": 0.2940517067909241,
"learning_rate": 3.1387381960746506e-07,
"loss": 0.4351,
"step": 2190
},
{
"epoch": 0.44455696202531647,
"grad_norm": 0.30264705419540405,
"learning_rate": 3.0399135149832546e-07,
"loss": 0.4144,
"step": 2195
},
{
"epoch": 0.44556962025316454,
"grad_norm": 0.49135756492614746,
"learning_rate": 2.942608118284502e-07,
"loss": 0.4439,
"step": 2200
},
{
"epoch": 0.4465822784810127,
"grad_norm": 0.09968866407871246,
"learning_rate": 2.8468260058221295e-07,
"loss": 0.4879,
"step": 2205
},
{
"epoch": 0.44759493670886075,
"grad_norm": 0.13854998350143433,
"learning_rate": 2.7525711148236276e-07,
"loss": 0.534,
"step": 2210
},
{
"epoch": 0.4486075949367089,
"grad_norm": 0.14843471348285675,
"learning_rate": 2.659847319738393e-07,
"loss": 0.455,
"step": 2215
},
{
"epoch": 0.44962025316455695,
"grad_norm": 0.15449483692646027,
"learning_rate": 2.5686584320784786e-07,
"loss": 0.4964,
"step": 2220
},
{
"epoch": 0.4506329113924051,
"grad_norm": 0.18671952188014984,
"learning_rate": 2.479008200261874e-07,
"loss": 0.4517,
"step": 2225
},
{
"epoch": 0.45164556962025315,
"grad_norm": 0.18353737890720367,
"learning_rate": 2.3909003094584856e-07,
"loss": 0.4155,
"step": 2230
},
{
"epoch": 0.4526582278481013,
"grad_norm": 0.21654197573661804,
"learning_rate": 2.304338381438611e-07,
"loss": 0.3943,
"step": 2235
},
{
"epoch": 0.45367088607594935,
"grad_norm": 0.23084385693073273,
"learning_rate": 2.21932597442406e-07,
"loss": 0.3996,
"step": 2240
},
{
"epoch": 0.4546835443037975,
"grad_norm": 0.2896442115306854,
"learning_rate": 2.135866582941932e-07,
"loss": 0.3828,
"step": 2245
},
{
"epoch": 0.45569620253164556,
"grad_norm": 0.5373207330703735,
"learning_rate": 2.0539636376809066e-07,
"loss": 0.3998,
"step": 2250
},
{
"epoch": 0.4567088607594937,
"grad_norm": 0.09597853571176529,
"learning_rate": 1.9736205053502774e-07,
"loss": 0.4169,
"step": 2255
},
{
"epoch": 0.45772151898734176,
"grad_norm": 0.1262124925851822,
"learning_rate": 1.8948404885415291e-07,
"loss": 0.5286,
"step": 2260
},
{
"epoch": 0.4587341772151899,
"grad_norm": 0.1337854415178299,
"learning_rate": 1.8176268255925975e-07,
"loss": 0.4516,
"step": 2265
},
{
"epoch": 0.45974683544303796,
"grad_norm": 0.16479073464870453,
"learning_rate": 1.7419826904547398e-07,
"loss": 0.5414,
"step": 2270
},
{
"epoch": 0.4607594936708861,
"grad_norm": 0.1927000880241394,
"learning_rate": 1.6679111925620704e-07,
"loss": 0.4988,
"step": 2275
},
{
"epoch": 0.46177215189873416,
"grad_norm": 0.17869296669960022,
"learning_rate": 1.5954153767037525e-07,
"loss": 0.3817,
"step": 2280
},
{
"epoch": 0.4627848101265823,
"grad_norm": 0.19914336502552032,
"learning_rate": 1.524498222898827e-07,
"loss": 0.3862,
"step": 2285
},
{
"epoch": 0.46379746835443036,
"grad_norm": 0.27144744992256165,
"learning_rate": 1.455162646273731e-07,
"loss": 0.4033,
"step": 2290
},
{
"epoch": 0.4648101265822785,
"grad_norm": 0.29997050762176514,
"learning_rate": 1.387411496942441e-07,
"loss": 0.3682,
"step": 2295
},
{
"epoch": 0.46582278481012657,
"grad_norm": 0.5696393847465515,
"learning_rate": 1.321247559889347e-07,
"loss": 0.3941,
"step": 2300
},
{
"epoch": 0.4668354430379747,
"grad_norm": 0.10038802772760391,
"learning_rate": 1.256673554854757e-07,
"loss": 0.4332,
"step": 2305
},
{
"epoch": 0.46784810126582277,
"grad_norm": 0.12659814953804016,
"learning_rate": 1.1936921362230946e-07,
"loss": 0.4953,
"step": 2310
},
{
"epoch": 0.4688607594936709,
"grad_norm": 0.14239275455474854,
"learning_rate": 1.1323058929137986e-07,
"loss": 0.5252,
"step": 2315
},
{
"epoch": 0.46987341772151897,
"grad_norm": 0.15920379757881165,
"learning_rate": 1.0725173482749016e-07,
"loss": 0.5316,
"step": 2320
},
{
"epoch": 0.4708860759493671,
"grad_norm": 0.16068552434444427,
"learning_rate": 1.0143289599792959e-07,
"loss": 0.4272,
"step": 2325
},
{
"epoch": 0.4718987341772152,
"grad_norm": 0.16718363761901855,
"learning_rate": 9.577431199237107e-08,
"loss": 0.3845,
"step": 2330
},
{
"epoch": 0.4729113924050633,
"grad_norm": 0.19578270614147186,
"learning_rate": 9.027621541304099e-08,
"loss": 0.3934,
"step": 2335
},
{
"epoch": 0.4739240506329114,
"grad_norm": 0.26058533787727356,
"learning_rate": 8.49388322651543e-08,
"loss": 0.3687,
"step": 2340
},
{
"epoch": 0.4749367088607595,
"grad_norm": 0.269424706697464,
"learning_rate": 7.976238194762785e-08,
"loss": 0.4036,
"step": 2345
},
{
"epoch": 0.4759493670886076,
"grad_norm": 0.4983403980731964,
"learning_rate": 7.47470772440586e-08,
"loss": 0.4376,
"step": 2350
},
{
"epoch": 0.4769620253164557,
"grad_norm": 0.10910938680171967,
"learning_rate": 6.989312431397998e-08,
"loss": 0.4566,
"step": 2355
},
{
"epoch": 0.4779746835443038,
"grad_norm": 0.1394750475883484,
"learning_rate": 6.520072268438515e-08,
"loss": 0.5027,
"step": 2360
},
{
"epoch": 0.4789873417721519,
"grad_norm": 0.14517003297805786,
"learning_rate": 6.067006524152551e-08,
"loss": 0.4066,
"step": 2365
},
{
"epoch": 0.48,
"grad_norm": 0.16347290575504303,
"learning_rate": 5.630133822298422e-08,
"loss": 0.4592,
"step": 2370
},
{
"epoch": 0.4810126582278481,
"grad_norm": 0.14922384917736053,
"learning_rate": 5.2094721210017344e-08,
"loss": 0.4147,
"step": 2375
},
{
"epoch": 0.4820253164556962,
"grad_norm": 0.267541766166687,
"learning_rate": 4.8050387120174504e-08,
"loss": 0.4218,
"step": 2380
},
{
"epoch": 0.4830379746835443,
"grad_norm": 0.20772917568683624,
"learning_rate": 4.416850220018853e-08,
"loss": 0.3547,
"step": 2385
},
{
"epoch": 0.4840506329113924,
"grad_norm": 0.2326534241437912,
"learning_rate": 4.044922601914535e-08,
"loss": 0.3728,
"step": 2390
},
{
"epoch": 0.4850632911392405,
"grad_norm": 0.30080971121788025,
"learning_rate": 3.689271146192086e-08,
"loss": 0.3819,
"step": 2395
},
{
"epoch": 0.4860759493670886,
"grad_norm": 0.6655465364456177,
"learning_rate": 3.34991047228983e-08,
"loss": 0.4111,
"step": 2400
},
{
"epoch": 0.4870886075949367,
"grad_norm": 0.10431922227144241,
"learning_rate": 3.026854529995937e-08,
"loss": 0.4185,
"step": 2405
},
{
"epoch": 0.4881012658227848,
"grad_norm": 0.11591701209545135,
"learning_rate": 2.7201165988748332e-08,
"loss": 0.5377,
"step": 2410
},
{
"epoch": 0.4891139240506329,
"grad_norm": 0.14209790527820587,
"learning_rate": 2.4297092877215043e-08,
"loss": 0.4822,
"step": 2415
},
{
"epoch": 0.490126582278481,
"grad_norm": 0.16605523228645325,
"learning_rate": 2.155644534043022e-08,
"loss": 0.4199,
"step": 2420
},
{
"epoch": 0.4911392405063291,
"grad_norm": 0.17735035717487335,
"learning_rate": 1.8979336035680027e-08,
"loss": 0.4481,
"step": 2425
},
{
"epoch": 0.4921518987341772,
"grad_norm": 0.20083750784397125,
"learning_rate": 1.656587089783379e-08,
"loss": 0.4077,
"step": 2430
},
{
"epoch": 0.4931645569620253,
"grad_norm": 0.20666271448135376,
"learning_rate": 1.4316149134991017e-08,
"loss": 0.4377,
"step": 2435
},
{
"epoch": 0.4941772151898734,
"grad_norm": 0.23543819785118103,
"learning_rate": 1.2230263224400684e-08,
"loss": 0.4154,
"step": 2440
},
{
"epoch": 0.4951898734177215,
"grad_norm": 0.2703807055950165,
"learning_rate": 1.0308298908663803e-08,
"loss": 0.3971,
"step": 2445
},
{
"epoch": 0.4962025316455696,
"grad_norm": 0.5880486965179443,
"learning_rate": 8.550335192205161e-09,
"loss": 0.4251,
"step": 2450
},
{
"epoch": 0.49721518987341773,
"grad_norm": 0.09348611533641815,
"learning_rate": 6.956444338027889e-09,
"loss": 0.4186,
"step": 2455
},
{
"epoch": 0.4982278481012658,
"grad_norm": 0.1480860859155655,
"learning_rate": 5.52669186474164e-09,
"loss": 0.509,
"step": 2460
},
{
"epoch": 0.49924050632911393,
"grad_norm": 0.15586763620376587,
"learning_rate": 4.2611365438713905e-09,
"loss": 0.486,
"step": 2465
},
{
"epoch": 0.500253164556962,
"grad_norm": 0.18108093738555908,
"learning_rate": 3.159830397438945e-09,
"loss": 0.4567,
"step": 2470
},
{
"epoch": 0.5012658227848101,
"grad_norm": 0.1902192085981369,
"learning_rate": 2.2228186958268556e-09,
"loss": 0.4181,
"step": 2475
},
{
"epoch": 0.5022784810126583,
"grad_norm": 0.16405874490737915,
"learning_rate": 1.4501399559154747e-09,
"loss": 0.389,
"step": 2480
},
{
"epoch": 0.5032911392405063,
"grad_norm": 0.21526110172271729,
"learning_rate": 8.418259395019944e-10,
"loss": 0.3984,
"step": 2485
},
{
"epoch": 0.5043037974683544,
"grad_norm": 0.3269748389720917,
"learning_rate": 3.97901651993493e-10,
"loss": 0.3936,
"step": 2490
},
{
"epoch": 0.5053164556962025,
"grad_norm": 0.27853924036026,
"learning_rate": 1.1838534137753597e-10,
"loss": 0.3835,
"step": 2495
},
{
"epoch": 0.5063291139240507,
"grad_norm": 0.7066273093223572,
"learning_rate": 3.2884974756619555e-12,
"loss": 0.4042,
"step": 2500
},
{
"epoch": 0.5073417721518987,
"grad_norm": 0.0975538045167923,
"learning_rate": 2.794380470215918e-08,
"loss": 0.4414,
"step": 2505
},
{
"epoch": 0.5083544303797468,
"grad_norm": 0.1161501482129097,
"learning_rate": 2.5111764414195823e-08,
"loss": 0.4881,
"step": 2510
},
{
"epoch": 0.5093670886075949,
"grad_norm": 0.1251804530620575,
"learning_rate": 2.243055246291936e-08,
"loss": 0.4934,
"step": 2515
},
{
"epoch": 0.5103797468354431,
"grad_norm": 0.15802332758903503,
"learning_rate": 1.9900270587728474e-08,
"loss": 0.5103,
"step": 2520
},
{
"epoch": 0.5113924050632911,
"grad_norm": 0.15753835439682007,
"learning_rate": 1.752101480093415e-08,
"loss": 0.4844,
"step": 2525
},
{
"epoch": 0.5124050632911392,
"grad_norm": 0.1545649617910385,
"learning_rate": 1.5292875384116832e-08,
"loss": 0.3984,
"step": 2530
},
{
"epoch": 0.5134177215189873,
"grad_norm": 0.2382134050130844,
"learning_rate": 1.321593688470113e-08,
"loss": 0.3778,
"step": 2535
},
{
"epoch": 0.5144303797468355,
"grad_norm": 0.24257385730743408,
"learning_rate": 1.1290278112747298e-08,
"loss": 0.3882,
"step": 2540
},
{
"epoch": 0.5154430379746835,
"grad_norm": 0.2859530448913574,
"learning_rate": 9.515972137958961e-09,
"loss": 0.4232,
"step": 2545
},
{
"epoch": 0.5164556962025316,
"grad_norm": 0.731451690196991,
"learning_rate": 7.893086286913764e-09,
"loss": 0.423,
"step": 2550
},
{
"epoch": 0.5174683544303798,
"grad_norm": 0.14217010140419006,
"learning_rate": 6.421682140506312e-09,
"loss": 0.4463,
"step": 2555
},
{
"epoch": 0.5184810126582279,
"grad_norm": 0.1456647515296936,
"learning_rate": 5.101815531612264e-09,
"loss": 0.6181,
"step": 2560
},
{
"epoch": 0.5194936708860759,
"grad_norm": 0.14817620813846588,
"learning_rate": 3.933536542968685e-09,
"loss": 0.4546,
"step": 2565
},
{
"epoch": 0.520506329113924,
"grad_norm": 0.18737301230430603,
"learning_rate": 2.916889505276021e-09,
"loss": 0.48,
"step": 2570
},
{
"epoch": 0.5215189873417722,
"grad_norm": 0.20854100584983826,
"learning_rate": 2.0519129955136606e-09,
"loss": 0.4711,
"step": 2575
},
{
"epoch": 0.5225316455696203,
"grad_norm": 0.17956523597240448,
"learning_rate": 1.3386398354762185e-09,
"loss": 0.3807,
"step": 2580
},
{
"epoch": 0.5235443037974683,
"grad_norm": 0.21315374970436096,
"learning_rate": 7.770970905305318e-10,
"loss": 0.4167,
"step": 2585
},
{
"epoch": 0.5245569620253164,
"grad_norm": 0.25169533491134644,
"learning_rate": 3.673060685849272e-10,
"loss": 0.4137,
"step": 2590
},
{
"epoch": 0.5255696202531646,
"grad_norm": 0.25980013608932495,
"learning_rate": 1.0928231928497566e-10,
"loss": 0.3961,
"step": 2595
},
{
"epoch": 0.5265822784810127,
"grad_norm": 0.6642146706581116,
"learning_rate": 3.035633418857486e-12,
"loss": 0.4063,
"step": 2600
},
{
"epoch": 0.5275949367088607,
"grad_norm": 0.11889899522066116,
"learning_rate": 1.455762992274968e-06,
"loss": 0.4235,
"step": 2605
},
{
"epoch": 0.5286075949367088,
"grad_norm": 0.1415431648492813,
"learning_rate": 1.4421305953258104e-06,
"loss": 0.4819,
"step": 2610
},
{
"epoch": 0.529620253164557,
"grad_norm": 0.14658379554748535,
"learning_rate": 1.4285482779599597e-06,
"loss": 0.544,
"step": 2615
},
{
"epoch": 0.5306329113924051,
"grad_norm": 0.1469714194536209,
"learning_rate": 1.4150163061006066e-06,
"loss": 0.5139,
"step": 2620
},
{
"epoch": 0.5316455696202531,
"grad_norm": 0.17124968767166138,
"learning_rate": 1.4015349446852462e-06,
"loss": 0.4299,
"step": 2625
},
{
"epoch": 0.5326582278481012,
"grad_norm": 0.1571967601776123,
"learning_rate": 1.3881044576604897e-06,
"loss": 0.4186,
"step": 2630
},
{
"epoch": 0.5336708860759494,
"grad_norm": 0.24192066490650177,
"learning_rate": 1.3747251079768974e-06,
"loss": 0.4007,
"step": 2635
},
{
"epoch": 0.5346835443037975,
"grad_norm": 0.22104869782924652,
"learning_rate": 1.3613971575838314e-06,
"loss": 0.3914,
"step": 2640
},
{
"epoch": 0.5356962025316455,
"grad_norm": 0.27553728222846985,
"learning_rate": 1.3481208674243246e-06,
"loss": 0.3683,
"step": 2645
},
{
"epoch": 0.5367088607594936,
"grad_norm": 0.4800131916999817,
"learning_rate": 1.3348964974299745e-06,
"loss": 0.4087,
"step": 2650
},
{
"epoch": 0.5377215189873418,
"grad_norm": 0.09944673627614975,
"learning_rate": 1.3217243065158518e-06,
"loss": 0.4186,
"step": 2655
},
{
"epoch": 0.5387341772151899,
"grad_norm": 0.11233413219451904,
"learning_rate": 1.3086045525754327e-06,
"loss": 0.5264,
"step": 2660
},
{
"epoch": 0.5397468354430379,
"grad_norm": 0.14689946174621582,
"learning_rate": 1.2955374924755495e-06,
"loss": 0.5095,
"step": 2665
},
{
"epoch": 0.540759493670886,
"grad_norm": 0.1830890029668808,
"learning_rate": 1.2825233820513602e-06,
"loss": 0.4974,
"step": 2670
},
{
"epoch": 0.5417721518987342,
"grad_norm": 0.15380920469760895,
"learning_rate": 1.269562476101342e-06,
"loss": 0.4354,
"step": 2675
},
{
"epoch": 0.5427848101265823,
"grad_norm": 0.17962035536766052,
"learning_rate": 1.2566550283822993e-06,
"loss": 0.4164,
"step": 2680
},
{
"epoch": 0.5437974683544303,
"grad_norm": 0.21246092021465302,
"learning_rate": 1.2438012916043997e-06,
"loss": 0.4162,
"step": 2685
},
{
"epoch": 0.5448101265822785,
"grad_norm": 0.22833871841430664,
"learning_rate": 1.2310015174262228e-06,
"loss": 0.3858,
"step": 2690
},
{
"epoch": 0.5458227848101266,
"grad_norm": 0.2646944224834442,
"learning_rate": 1.218255956449834e-06,
"loss": 0.4128,
"step": 2695
},
{
"epoch": 0.5468354430379747,
"grad_norm": 0.47620370984077454,
"learning_rate": 1.2055648582158794e-06,
"loss": 0.3937,
"step": 2700
},
{
"epoch": 0.5478481012658228,
"grad_norm": 0.1026020497083664,
"learning_rate": 1.1929284711986994e-06,
"loss": 0.4544,
"step": 2705
},
{
"epoch": 0.5488607594936709,
"grad_norm": 0.13041728734970093,
"learning_rate": 1.1803470428014612e-06,
"loss": 0.4827,
"step": 2710
},
{
"epoch": 0.549873417721519,
"grad_norm": 0.132302924990654,
"learning_rate": 1.1678208193513208e-06,
"loss": 0.508,
"step": 2715
},
{
"epoch": 0.5508860759493671,
"grad_norm": 0.1621437668800354,
"learning_rate": 1.1553500460945937e-06,
"loss": 0.477,
"step": 2720
},
{
"epoch": 0.5518987341772152,
"grad_norm": 0.1835368424654007,
"learning_rate": 1.1429349671919593e-06,
"loss": 0.433,
"step": 2725
},
{
"epoch": 0.5529113924050633,
"grad_norm": 0.16705666482448578,
"learning_rate": 1.1305758257136755e-06,
"loss": 0.4118,
"step": 2730
},
{
"epoch": 0.5539240506329114,
"grad_norm": 0.2378142923116684,
"learning_rate": 1.118272863634823e-06,
"loss": 0.383,
"step": 2735
},
{
"epoch": 0.5549367088607595,
"grad_norm": 0.2359638810157776,
"learning_rate": 1.106026321830567e-06,
"loss": 0.4056,
"step": 2740
},
{
"epoch": 0.5559493670886076,
"grad_norm": 0.2995934784412384,
"learning_rate": 1.0938364400714388e-06,
"loss": 0.4165,
"step": 2745
},
{
"epoch": 0.5569620253164557,
"grad_norm": 0.5750916004180908,
"learning_rate": 1.0817034570186477e-06,
"loss": 0.4282,
"step": 2750
},
{
"epoch": 0.5579746835443038,
"grad_norm": 0.07957121729850769,
"learning_rate": 1.069627610219401e-06,
"loss": 0.4181,
"step": 2755
},
{
"epoch": 0.5589873417721519,
"grad_norm": 0.13279347121715546,
"learning_rate": 1.0576091361022563e-06,
"loss": 0.5386,
"step": 2760
},
{
"epoch": 0.56,
"grad_norm": 0.15844392776489258,
"learning_rate": 1.045648269972494e-06,
"loss": 0.4838,
"step": 2765
},
{
"epoch": 0.5610126582278481,
"grad_norm": 0.15711435675621033,
"learning_rate": 1.0337452460075083e-06,
"loss": 0.4998,
"step": 2770
},
{
"epoch": 0.5620253164556962,
"grad_norm": 0.16822844743728638,
"learning_rate": 1.0219002972522219e-06,
"loss": 0.4461,
"step": 2775
},
{
"epoch": 0.5630379746835443,
"grad_norm": 0.2024250328540802,
"learning_rate": 1.0101136556145266e-06,
"loss": 0.4168,
"step": 2780
},
{
"epoch": 0.5640506329113925,
"grad_norm": 0.1919785588979721,
"learning_rate": 9.983855518607382e-07,
"loss": 0.3942,
"step": 2785
},
{
"epoch": 0.5650632911392405,
"grad_norm": 0.22147142887115479,
"learning_rate": 9.86716215611083e-07,
"loss": 0.3952,
"step": 2790
},
{
"epoch": 0.5660759493670886,
"grad_norm": 0.28781041502952576,
"learning_rate": 9.751058753351991e-07,
"loss": 0.3809,
"step": 2795
},
{
"epoch": 0.5670886075949367,
"grad_norm": 0.6933401226997375,
"learning_rate": 9.635547583476627e-07,
"loss": 0.3959,
"step": 2800
},
{
"epoch": 0.5681012658227849,
"grad_norm": 0.10298442840576172,
"learning_rate": 9.520630908035414e-07,
"loss": 0.4014,
"step": 2805
},
{
"epoch": 0.5691139240506329,
"grad_norm": 0.13080711662769318,
"learning_rate": 9.406310976939629e-07,
"loss": 0.5034,
"step": 2810
},
{
"epoch": 0.570126582278481,
"grad_norm": 0.1588020622730255,
"learning_rate": 9.292590028417109e-07,
"loss": 0.5223,
"step": 2815
},
{
"epoch": 0.5711392405063291,
"grad_norm": 0.15460827946662903,
"learning_rate": 9.179470288968434e-07,
"loss": 0.4344,
"step": 2820
},
{
"epoch": 0.5721518987341773,
"grad_norm": 0.2036442905664444,
"learning_rate": 9.066953973323341e-07,
"loss": 0.4414,
"step": 2825
},
{
"epoch": 0.5731645569620253,
"grad_norm": 0.213650643825531,
"learning_rate": 8.955043284397338e-07,
"loss": 0.4025,
"step": 2830
},
{
"epoch": 0.5741772151898734,
"grad_norm": 0.19585683941841125,
"learning_rate": 8.843740413248606e-07,
"loss": 0.4186,
"step": 2835
},
{
"epoch": 0.5751898734177215,
"grad_norm": 0.23203453421592712,
"learning_rate": 8.733047539035077e-07,
"loss": 0.4464,
"step": 2840
},
{
"epoch": 0.5762025316455697,
"grad_norm": 0.3242339491844177,
"learning_rate": 8.622966828971781e-07,
"loss": 0.3964,
"step": 2845
},
{
"epoch": 0.5772151898734177,
"grad_norm": 0.47551751136779785,
"learning_rate": 8.51350043828841e-07,
"loss": 0.4272,
"step": 2850
},
{
"epoch": 0.5782278481012658,
"grad_norm": 0.10742074251174927,
"learning_rate": 8.404650510187123e-07,
"loss": 0.4219,
"step": 2855
},
{
"epoch": 0.579240506329114,
"grad_norm": 0.11851336061954498,
"learning_rate": 8.296419175800573e-07,
"loss": 0.5157,
"step": 2860
},
{
"epoch": 0.5802531645569621,
"grad_norm": 0.1675945371389389,
"learning_rate": 8.188808554150229e-07,
"loss": 0.5129,
"step": 2865
},
{
"epoch": 0.5812658227848101,
"grad_norm": 0.16009923815727234,
"learning_rate": 8.081820752104826e-07,
"loss": 0.4889,
"step": 2870
},
{
"epoch": 0.5822784810126582,
"grad_norm": 0.18945632874965668,
"learning_rate": 7.975457864339144e-07,
"loss": 0.3923,
"step": 2875
},
{
"epoch": 0.5832911392405064,
"grad_norm": 0.1865212470293045,
"learning_rate": 7.869721973292999e-07,
"loss": 0.428,
"step": 2880
},
{
"epoch": 0.5843037974683545,
"grad_norm": 0.20608146488666534,
"learning_rate": 7.76461514913048e-07,
"loss": 0.407,
"step": 2885
},
{
"epoch": 0.5853164556962025,
"grad_norm": 0.29125773906707764,
"learning_rate": 7.660139449699387e-07,
"loss": 0.407,
"step": 2890
},
{
"epoch": 0.5863291139240506,
"grad_norm": 0.33646541833877563,
"learning_rate": 7.556296920490987e-07,
"loss": 0.4333,
"step": 2895
},
{
"epoch": 0.5873417721518988,
"grad_norm": 0.5913756489753723,
"learning_rate": 7.453089594599915e-07,
"loss": 0.4502,
"step": 2900
},
{
"epoch": 0.5883544303797469,
"grad_norm": 0.13217048346996307,
"learning_rate": 7.350519492684407e-07,
"loss": 0.4547,
"step": 2905
},
{
"epoch": 0.5893670886075949,
"grad_norm": 0.14116889238357544,
"learning_rate": 7.248588622926725e-07,
"loss": 0.5199,
"step": 2910
},
{
"epoch": 0.590379746835443,
"grad_norm": 0.15484601259231567,
"learning_rate": 7.147298980993843e-07,
"loss": 0.548,
"step": 2915
},
{
"epoch": 0.5913924050632912,
"grad_norm": 0.15964430570602417,
"learning_rate": 7.046652549998362e-07,
"loss": 0.4814,
"step": 2920
},
{
"epoch": 0.5924050632911393,
"grad_norm": 0.1844719797372818,
"learning_rate": 6.946651300459701e-07,
"loss": 0.4685,
"step": 2925
},
{
"epoch": 0.5934177215189873,
"grad_norm": 0.19145998358726501,
"learning_rate": 6.847297190265501e-07,
"loss": 0.4308,
"step": 2930
},
{
"epoch": 0.5944303797468354,
"grad_norm": 0.21867701411247253,
"learning_rate": 6.748592164633313e-07,
"loss": 0.4295,
"step": 2935
},
{
"epoch": 0.5954430379746836,
"grad_norm": 0.2490256279706955,
"learning_rate": 6.650538156072483e-07,
"loss": 0.3908,
"step": 2940
},
{
"epoch": 0.5964556962025317,
"grad_norm": 0.2814614474773407,
"learning_rate": 6.55313708434635e-07,
"loss": 0.4496,
"step": 2945
},
{
"epoch": 0.5974683544303797,
"grad_norm": 0.6335835456848145,
"learning_rate": 6.45639085643463e-07,
"loss": 0.4608,
"step": 2950
},
{
"epoch": 0.5984810126582278,
"grad_norm": 0.12605907022953033,
"learning_rate": 1.1337485346248307e-06,
"loss": 0.4293,
"step": 2955
},
{
"epoch": 0.599493670886076,
"grad_norm": 0.1293596625328064,
"learning_rate": 1.1223888816257035e-06,
"loss": 0.5123,
"step": 2960
},
{
"epoch": 0.6005063291139241,
"grad_norm": 0.1349225491285324,
"learning_rate": 1.1110771302018184e-06,
"loss": 0.4865,
"step": 2965
},
{
"epoch": 0.6015189873417721,
"grad_norm": 0.18334391713142395,
"learning_rate": 1.0998134686519725e-06,
"loss": 0.4882,
"step": 2970
},
{
"epoch": 0.6025316455696202,
"grad_norm": 0.1953776478767395,
"learning_rate": 1.088598084474444e-06,
"loss": 0.3952,
"step": 2975
},
{
"epoch": 0.6035443037974684,
"grad_norm": 0.17717419564723969,
"learning_rate": 1.0774311643638725e-06,
"loss": 0.4286,
"step": 2980
},
{
"epoch": 0.6045569620253165,
"grad_norm": 0.2070053219795227,
"learning_rate": 1.0663128942081482e-06,
"loss": 0.3948,
"step": 2985
},
{
"epoch": 0.6055696202531645,
"grad_norm": 0.27012762427330017,
"learning_rate": 1.0552434590853247e-06,
"loss": 0.4096,
"step": 2990
},
{
"epoch": 0.6065822784810126,
"grad_norm": 0.27783793210983276,
"learning_rate": 1.044223043260525e-06,
"loss": 0.3866,
"step": 2995
},
{
"epoch": 0.6075949367088608,
"grad_norm": 0.6585811972618103,
"learning_rate": 1.0332518301828908e-06,
"loss": 0.4278,
"step": 3000
},
{
"epoch": 0.6086075949367089,
"grad_norm": 0.09339703619480133,
"learning_rate": 1.022330002482513e-06,
"loss": 0.3972,
"step": 3005
},
{
"epoch": 0.6096202531645569,
"grad_norm": 0.14275415241718292,
"learning_rate": 1.0114577419674013e-06,
"loss": 0.5359,
"step": 3010
},
{
"epoch": 0.610632911392405,
"grad_norm": 0.13809597492218018,
"learning_rate": 1.0006352296204595e-06,
"loss": 0.5597,
"step": 3015
},
{
"epoch": 0.6116455696202532,
"grad_norm": 0.15734300017356873,
"learning_rate": 9.898626455964594e-07,
"loss": 0.4792,
"step": 3020
},
{
"epoch": 0.6126582278481013,
"grad_norm": 0.17018863558769226,
"learning_rate": 9.79140169219062e-07,
"loss": 0.4098,
"step": 3025
},
{
"epoch": 0.6136708860759493,
"grad_norm": 0.18591512739658356,
"learning_rate": 9.684679789778098e-07,
"loss": 0.4217,
"step": 3030
},
{
"epoch": 0.6146835443037975,
"grad_norm": 0.18555332720279694,
"learning_rate": 9.578462525251776e-07,
"loss": 0.3826,
"step": 3035
},
{
"epoch": 0.6156962025316456,
"grad_norm": 0.20953959226608276,
"learning_rate": 9.472751666735988e-07,
"loss": 0.3649,
"step": 3040
},
{
"epoch": 0.6167088607594937,
"grad_norm": 0.2972744405269623,
"learning_rate": 9.367548973925302e-07,
"loss": 0.4014,
"step": 3045
},
{
"epoch": 0.6177215189873417,
"grad_norm": 0.4772799015045166,
"learning_rate": 9.262856198055198e-07,
"loss": 0.432,
"step": 3050
},
{
"epoch": 0.6187341772151899,
"grad_norm": 0.1068698912858963,
"learning_rate": 9.158675081872936e-07,
"loss": 0.4168,
"step": 3055
},
{
"epoch": 0.619746835443038,
"grad_norm": 0.13200756907463074,
"learning_rate": 9.055007359608517e-07,
"loss": 0.5179,
"step": 3060
},
{
"epoch": 0.6207594936708861,
"grad_norm": 0.12744776904582977,
"learning_rate": 8.951854756945887e-07,
"loss": 0.5078,
"step": 3065
},
{
"epoch": 0.6217721518987341,
"grad_norm": 0.16431675851345062,
"learning_rate": 8.849218990994072e-07,
"loss": 0.4998,
"step": 3070
},
{
"epoch": 0.6227848101265823,
"grad_norm": 0.1980285942554474,
"learning_rate": 8.747101770258769e-07,
"loss": 0.4425,
"step": 3075
},
{
"epoch": 0.6237974683544304,
"grad_norm": 0.19649136066436768,
"learning_rate": 8.645504794613727e-07,
"loss": 0.4448,
"step": 3080
},
{
"epoch": 0.6248101265822785,
"grad_norm": 0.23119045794010162,
"learning_rate": 8.544429755272626e-07,
"loss": 0.4068,
"step": 3085
},
{
"epoch": 0.6258227848101265,
"grad_norm": 0.24636401236057281,
"learning_rate": 8.443878334760777e-07,
"loss": 0.4346,
"step": 3090
},
{
"epoch": 0.6268354430379747,
"grad_norm": 0.33371180295944214,
"learning_rate": 8.343852206887208e-07,
"loss": 0.374,
"step": 3095
},
{
"epoch": 0.6278481012658228,
"grad_norm": 0.4640437066555023,
"learning_rate": 8.244353036716742e-07,
"loss": 0.3928,
"step": 3100
},
{
"epoch": 0.6288607594936709,
"grad_norm": 0.10186227411031723,
"learning_rate": 8.145382480542338e-07,
"loss": 0.4351,
"step": 3105
},
{
"epoch": 0.629873417721519,
"grad_norm": 0.11600786447525024,
"learning_rate": 8.046942185857459e-07,
"loss": 0.5473,
"step": 3110
},
{
"epoch": 0.6308860759493671,
"grad_norm": 0.14960376918315887,
"learning_rate": 7.949033791328692e-07,
"loss": 0.5021,
"step": 3115
},
{
"epoch": 0.6318987341772152,
"grad_norm": 0.15927369892597198,
"learning_rate": 7.851658926768449e-07,
"loss": 0.4915,
"step": 3120
},
{
"epoch": 0.6329113924050633,
"grad_norm": 0.21519607305526733,
"learning_rate": 7.754819213107869e-07,
"loss": 0.4531,
"step": 3125
},
{
"epoch": 0.6339240506329114,
"grad_norm": 0.19581706821918488,
"learning_rate": 7.658516262369747e-07,
"loss": 0.4027,
"step": 3130
},
{
"epoch": 0.6349367088607595,
"grad_norm": 0.20079204440116882,
"learning_rate": 7.562751677641835e-07,
"loss": 0.4482,
"step": 3135
},
{
"epoch": 0.6359493670886076,
"grad_norm": 0.2671821117401123,
"learning_rate": 7.467527053050036e-07,
"loss": 0.4217,
"step": 3140
},
{
"epoch": 0.6369620253164557,
"grad_norm": 0.2694138288497925,
"learning_rate": 7.372843973731933e-07,
"loss": 0.4188,
"step": 3145
},
{
"epoch": 0.6379746835443038,
"grad_norm": 0.4967109262943268,
"learning_rate": 7.278704015810371e-07,
"loss": 0.4075,
"step": 3150
},
{
"epoch": 0.6389873417721519,
"grad_norm": 0.13129402697086334,
"learning_rate": 7.185108746367246e-07,
"loss": 0.4101,
"step": 3155
},
{
"epoch": 0.64,
"grad_norm": 0.13725103437900543,
"learning_rate": 7.092059723417407e-07,
"loss": 0.4936,
"step": 3160
},
{
"epoch": 0.6410126582278481,
"grad_norm": 0.15618817508220673,
"learning_rate": 6.999558495882718e-07,
"loss": 0.5314,
"step": 3165
},
{
"epoch": 0.6420253164556962,
"grad_norm": 0.162822425365448,
"learning_rate": 6.907606603566258e-07,
"loss": 0.4752,
"step": 3170
},
{
"epoch": 0.6430379746835443,
"grad_norm": 0.21904124319553375,
"learning_rate": 6.816205577126762e-07,
"loss": 0.4544,
"step": 3175
},
{
"epoch": 0.6440506329113924,
"grad_norm": 0.2016332447528839,
"learning_rate": 6.725356938053006e-07,
"loss": 0.3698,
"step": 3180
},
{
"epoch": 0.6450632911392405,
"grad_norm": 0.22290003299713135,
"learning_rate": 6.635062198638635e-07,
"loss": 0.3635,
"step": 3185
},
{
"epoch": 0.6460759493670886,
"grad_norm": 0.21098457276821136,
"learning_rate": 6.545322861956881e-07,
"loss": 0.3863,
"step": 3190
},
{
"epoch": 0.6470886075949367,
"grad_norm": 0.36327552795410156,
"learning_rate": 6.456140421835572e-07,
"loss": 0.4089,
"step": 3195
},
{
"epoch": 0.6481012658227848,
"grad_norm": 0.5956647992134094,
"learning_rate": 6.367516362832277e-07,
"loss": 0.3957,
"step": 3200
},
{
"epoch": 0.649113924050633,
"grad_norm": 0.08976675570011139,
"learning_rate": 6.279452160209588e-07,
"loss": 0.4714,
"step": 3205
},
{
"epoch": 0.650126582278481,
"grad_norm": 0.11679979413747787,
"learning_rate": 6.191949279910558e-07,
"loss": 0.4448,
"step": 3210
},
{
"epoch": 0.6511392405063291,
"grad_norm": 0.14985623955726624,
"learning_rate": 6.105009178534301e-07,
"loss": 0.5321,
"step": 3215
},
{
"epoch": 0.6521518987341772,
"grad_norm": 0.16550008952617645,
"learning_rate": 6.018633303311733e-07,
"loss": 0.5148,
"step": 3220
},
{
"epoch": 0.6531645569620254,
"grad_norm": 0.1526990532875061,
"learning_rate": 5.932823092081536e-07,
"loss": 0.3914,
"step": 3225
},
{
"epoch": 0.6541772151898734,
"grad_norm": 0.20120437443256378,
"learning_rate": 5.847579973266108e-07,
"loss": 0.4138,
"step": 3230
},
{
"epoch": 0.6551898734177215,
"grad_norm": 0.22462603449821472,
"learning_rate": 5.762905365847936e-07,
"loss": 0.4191,
"step": 3235
},
{
"epoch": 0.6562025316455696,
"grad_norm": 0.21973134577274323,
"learning_rate": 5.678800679345816e-07,
"loss": 0.351,
"step": 3240
},
{
"epoch": 0.6572151898734178,
"grad_norm": 0.2744171917438507,
"learning_rate": 5.595267313791545e-07,
"loss": 0.4064,
"step": 3245
},
{
"epoch": 0.6582278481012658,
"grad_norm": 0.4992412328720093,
"learning_rate": 5.512306659706496e-07,
"loss": 0.3707,
"step": 3250
},
{
"epoch": 0.6592405063291139,
"grad_norm": 0.10248786211013794,
"learning_rate": 5.429920098078526e-07,
"loss": 0.3939,
"step": 3255
},
{
"epoch": 0.660253164556962,
"grad_norm": 0.14512722194194794,
"learning_rate": 5.348109000338978e-07,
"loss": 0.4847,
"step": 3260
},
{
"epoch": 0.6612658227848102,
"grad_norm": 0.13799966871738434,
"learning_rate": 5.266874728339848e-07,
"loss": 0.5207,
"step": 3265
},
{
"epoch": 0.6622784810126582,
"grad_norm": 0.20723897218704224,
"learning_rate": 5.186218634331129e-07,
"loss": 0.5036,
"step": 3270
},
{
"epoch": 0.6632911392405063,
"grad_norm": 0.18485288321971893,
"learning_rate": 5.106142060938276e-07,
"loss": 0.4884,
"step": 3275
},
{
"epoch": 0.6643037974683544,
"grad_norm": 0.16878673434257507,
"learning_rate": 5.026646341139868e-07,
"loss": 0.3754,
"step": 3280
},
{
"epoch": 0.6653164556962026,
"grad_norm": 0.20705561339855194,
"learning_rate": 4.947732798245465e-07,
"loss": 0.39,
"step": 3285
},
{
"epoch": 0.6663291139240506,
"grad_norm": 0.22199900448322296,
"learning_rate": 4.869402745873454e-07,
"loss": 0.3827,
"step": 3290
},
{
"epoch": 0.6673417721518987,
"grad_norm": 0.3176887333393097,
"learning_rate": 4.791657487929348e-07,
"loss": 0.395,
"step": 3295
},
{
"epoch": 0.6683544303797468,
"grad_norm": 0.4770476520061493,
"learning_rate": 4.714498318583948e-07,
"loss": 0.3991,
"step": 3300
},
{
"epoch": 0.669367088607595,
"grad_norm": 0.10591458529233932,
"learning_rate": 4.637926522251865e-07,
"loss": 0.4178,
"step": 3305
},
{
"epoch": 0.670379746835443,
"grad_norm": 0.12099853157997131,
"learning_rate": 4.5619433735701165e-07,
"loss": 0.5278,
"step": 3310
},
{
"epoch": 0.6713924050632911,
"grad_norm": 0.1402314007282257,
"learning_rate": 4.4865501373769186e-07,
"loss": 0.5253,
"step": 3315
},
{
"epoch": 0.6724050632911392,
"grad_norm": 0.16952082514762878,
"learning_rate": 4.4117480686906283e-07,
"loss": 0.4734,
"step": 3320
},
{
"epoch": 0.6734177215189874,
"grad_norm": 0.16695508360862732,
"learning_rate": 4.3375384126888504e-07,
"loss": 0.4518,
"step": 3325
},
{
"epoch": 0.6744303797468354,
"grad_norm": 0.20437532663345337,
"learning_rate": 4.26392240468771e-07,
"loss": 0.4321,
"step": 3330
},
{
"epoch": 0.6754430379746835,
"grad_norm": 0.1856561303138733,
"learning_rate": 4.190901270121303e-07,
"loss": 0.3806,
"step": 3335
},
{
"epoch": 0.6764556962025317,
"grad_norm": 0.24451293051242828,
"learning_rate": 4.1184762245212525e-07,
"loss": 0.3897,
"step": 3340
},
{
"epoch": 0.6774683544303798,
"grad_norm": 0.2876019775867462,
"learning_rate": 4.0466484734965433e-07,
"loss": 0.3632,
"step": 3345
},
{
"epoch": 0.6784810126582278,
"grad_norm": 0.4984877407550812,
"learning_rate": 3.9754192127133957e-07,
"loss": 0.4028,
"step": 3350
},
{
"epoch": 0.6794936708860759,
"grad_norm": 0.09997934848070145,
"learning_rate": 3.9047896278753844e-07,
"loss": 0.4502,
"step": 3355
},
{
"epoch": 0.6805063291139241,
"grad_norm": 0.1204715296626091,
"learning_rate": 3.8347608947037014e-07,
"loss": 0.5394,
"step": 3360
},
{
"epoch": 0.6815189873417722,
"grad_norm": 0.14950430393218994,
"learning_rate": 3.765334178917583e-07,
"loss": 0.4733,
"step": 3365
},
{
"epoch": 0.6825316455696202,
"grad_norm": 0.1534213125705719,
"learning_rate": 3.6965106362149e-07,
"loss": 0.4995,
"step": 3370
},
{
"epoch": 0.6835443037974683,
"grad_norm": 0.1603599637746811,
"learning_rate": 3.628291412252924e-07,
"loss": 0.4559,
"step": 3375
},
{
"epoch": 0.6845569620253165,
"grad_norm": 0.1949257254600525,
"learning_rate": 3.560677642629257e-07,
"loss": 0.4296,
"step": 3380
},
{
"epoch": 0.6855696202531646,
"grad_norm": 0.2206084430217743,
"learning_rate": 3.493670452862947e-07,
"loss": 0.4297,
"step": 3385
},
{
"epoch": 0.6865822784810126,
"grad_norm": 0.23324523866176605,
"learning_rate": 3.427270958375685e-07,
"loss": 0.4457,
"step": 3390
},
{
"epoch": 0.6875949367088607,
"grad_norm": 0.30350419878959656,
"learning_rate": 3.361480264473342e-07,
"loss": 0.4041,
"step": 3395
},
{
"epoch": 0.6886075949367089,
"grad_norm": 0.5744687914848328,
"learning_rate": 3.2962994663274567e-07,
"loss": 0.4043,
"step": 3400
},
{
"epoch": 0.689620253164557,
"grad_norm": 0.12565582990646362,
"learning_rate": 3.231729648957109e-07,
"loss": 0.4517,
"step": 3405
},
{
"epoch": 0.690632911392405,
"grad_norm": 0.11841224879026413,
"learning_rate": 3.1677718872107796e-07,
"loss": 0.4729,
"step": 3410
},
{
"epoch": 0.6916455696202531,
"grad_norm": 0.14457449316978455,
"learning_rate": 3.1044272457485e-07,
"loss": 0.5286,
"step": 3415
},
{
"epoch": 0.6926582278481013,
"grad_norm": 0.19540651142597198,
"learning_rate": 3.04169677902411e-07,
"loss": 0.481,
"step": 3420
},
{
"epoch": 0.6936708860759494,
"grad_norm": 0.16388900578022003,
"learning_rate": 2.9795815312677254e-07,
"loss": 0.3597,
"step": 3425
},
{
"epoch": 0.6946835443037974,
"grad_norm": 0.1902589499950409,
"learning_rate": 2.9180825364683335e-07,
"loss": 0.4312,
"step": 3430
},
{
"epoch": 0.6956962025316455,
"grad_norm": 0.22043222188949585,
"learning_rate": 2.8572008183565954e-07,
"loss": 0.363,
"step": 3435
},
{
"epoch": 0.6967088607594937,
"grad_norm": 0.21098974347114563,
"learning_rate": 2.796937390387786e-07,
"loss": 0.3886,
"step": 3440
},
{
"epoch": 0.6977215189873418,
"grad_norm": 0.3231201767921448,
"learning_rate": 2.737293255724973e-07,
"loss": 0.3982,
"step": 3445
},
{
"epoch": 0.6987341772151898,
"grad_norm": 1.050829291343689,
"learning_rate": 2.678269407222231e-07,
"loss": 0.4543,
"step": 3450
},
{
"epoch": 0.699746835443038,
"grad_norm": 0.11217325180768967,
"learning_rate": 2.619866827408219e-07,
"loss": 0.4463,
"step": 3455
},
{
"epoch": 0.7007594936708861,
"grad_norm": 0.14886336028575897,
"learning_rate": 2.5620864884697303e-07,
"loss": 0.5608,
"step": 3460
},
{
"epoch": 0.7017721518987342,
"grad_norm": 0.15907342731952667,
"learning_rate": 2.504929352235572e-07,
"loss": 0.4828,
"step": 3465
},
{
"epoch": 0.7027848101265822,
"grad_norm": 0.1800810545682907,
"learning_rate": 2.4483963701605257e-07,
"loss": 0.4766,
"step": 3470
},
{
"epoch": 0.7037974683544304,
"grad_norm": 0.16972433030605316,
"learning_rate": 2.392488483309516e-07,
"loss": 0.3882,
"step": 3475
},
{
"epoch": 0.7048101265822785,
"grad_norm": 0.1915484517812729,
"learning_rate": 2.337206622341954e-07,
"loss": 0.389,
"step": 3480
},
{
"epoch": 0.7058227848101266,
"grad_norm": 0.2930021286010742,
"learning_rate": 2.2825517074962186e-07,
"loss": 0.4086,
"step": 3485
},
{
"epoch": 0.7068354430379746,
"grad_norm": 0.25938400626182556,
"learning_rate": 2.2285246485743703e-07,
"loss": 0.3592,
"step": 3490
},
{
"epoch": 0.7078481012658228,
"grad_norm": 0.266793429851532,
"learning_rate": 2.1751263449270074e-07,
"loss": 0.4103,
"step": 3495
},
{
"epoch": 0.7088607594936709,
"grad_norm": 0.6035900115966797,
"learning_rate": 2.1223576854382297e-07,
"loss": 0.3819,
"step": 3500
},
{
"epoch": 0.709873417721519,
"grad_norm": 0.08965824544429779,
"learning_rate": 9.487686635552891e-07,
"loss": 0.3868,
"step": 3505
},
{
"epoch": 0.710886075949367,
"grad_norm": 0.14689555764198303,
"learning_rate": 9.396576552413696e-07,
"loss": 0.5019,
"step": 3510
},
{
"epoch": 0.7118987341772152,
"grad_norm": 0.1451750546693802,
"learning_rate": 9.305847789018373e-07,
"loss": 0.5091,
"step": 3515
},
{
"epoch": 0.7129113924050633,
"grad_norm": 0.15510936081409454,
"learning_rate": 9.215501475850529e-07,
"loss": 0.4715,
"step": 3520
},
{
"epoch": 0.7139240506329114,
"grad_norm": 0.21404242515563965,
"learning_rate": 9.125538738628478e-07,
"loss": 0.4516,
"step": 3525
},
{
"epoch": 0.7149367088607594,
"grad_norm": 0.21132291853427887,
"learning_rate": 9.035960698291067e-07,
"loss": 0.3797,
"step": 3530
},
{
"epoch": 0.7159493670886076,
"grad_norm": 0.23155702650547028,
"learning_rate": 8.946768470983906e-07,
"loss": 0.3713,
"step": 3535
},
{
"epoch": 0.7169620253164557,
"grad_norm": 0.22461672127246857,
"learning_rate": 8.857963168045301e-07,
"loss": 0.3757,
"step": 3540
},
{
"epoch": 0.7179746835443038,
"grad_norm": 0.24911552667617798,
"learning_rate": 8.769545895992472e-07,
"loss": 0.4037,
"step": 3545
},
{
"epoch": 0.7189873417721518,
"grad_norm": 0.48576197028160095,
"learning_rate": 8.681517756507806e-07,
"loss": 0.4365,
"step": 3550
},
{
"epoch": 0.72,
"grad_norm": 0.093582384288311,
"learning_rate": 8.593879846425002e-07,
"loss": 0.4422,
"step": 3555
},
{
"epoch": 0.7210126582278481,
"grad_norm": 0.12010160088539124,
"learning_rate": 8.506633257715577e-07,
"loss": 0.4805,
"step": 3560
},
{
"epoch": 0.7220253164556962,
"grad_norm": 0.14595255255699158,
"learning_rate": 8.419779077475104e-07,
"loss": 0.4995,
"step": 3565
},
{
"epoch": 0.7230379746835442,
"grad_norm": 0.1681118756532669,
"learning_rate": 8.333318387909764e-07,
"loss": 0.482,
"step": 3570
},
{
"epoch": 0.7240506329113924,
"grad_norm": 0.1851746290922165,
"learning_rate": 8.247252266322822e-07,
"loss": 0.4098,
"step": 3575
},
{
"epoch": 0.7250632911392405,
"grad_norm": 0.19346100091934204,
"learning_rate": 8.161581785101197e-07,
"loss": 0.3622,
"step": 3580
},
{
"epoch": 0.7260759493670886,
"grad_norm": 0.2945711314678192,
"learning_rate": 8.076308011702155e-07,
"loss": 0.3625,
"step": 3585
},
{
"epoch": 0.7270886075949367,
"grad_norm": 0.24430105090141296,
"learning_rate": 7.991432008639929e-07,
"loss": 0.4068,
"step": 3590
},
{
"epoch": 0.7281012658227848,
"grad_norm": 0.2994179129600525,
"learning_rate": 7.906954833472537e-07,
"loss": 0.3901,
"step": 3595
},
{
"epoch": 0.7291139240506329,
"grad_norm": 0.531819224357605,
"learning_rate": 7.822877538788581e-07,
"loss": 0.4053,
"step": 3600
},
{
"epoch": 0.730126582278481,
"grad_norm": 0.11160524189472198,
"learning_rate": 7.739201172194132e-07,
"loss": 0.425,
"step": 3605
},
{
"epoch": 0.7311392405063292,
"grad_norm": 0.14233025908470154,
"learning_rate": 7.655926776299723e-07,
"loss": 0.514,
"step": 3610
},
{
"epoch": 0.7321518987341772,
"grad_norm": 0.1432143747806549,
"learning_rate": 7.57305538870725e-07,
"loss": 0.4969,
"step": 3615
},
{
"epoch": 0.7331645569620253,
"grad_norm": 0.17136462032794952,
"learning_rate": 7.490588041997173e-07,
"loss": 0.4799,
"step": 3620
},
{
"epoch": 0.7341772151898734,
"grad_norm": 0.17079944908618927,
"learning_rate": 7.408525763715552e-07,
"loss": 0.4471,
"step": 3625
},
{
"epoch": 0.7351898734177216,
"grad_norm": 0.20650941133499146,
"learning_rate": 7.326869576361297e-07,
"loss": 0.398,
"step": 3630
},
{
"epoch": 0.7362025316455696,
"grad_norm": 0.21803735196590424,
"learning_rate": 7.245620497373402e-07,
"loss": 0.373,
"step": 3635
},
{
"epoch": 0.7372151898734177,
"grad_norm": 0.27619433403015137,
"learning_rate": 7.164779539118262e-07,
"loss": 0.3923,
"step": 3640
},
{
"epoch": 0.7382278481012658,
"grad_norm": 0.42795369029045105,
"learning_rate": 7.084347708877123e-07,
"loss": 0.4059,
"step": 3645
},
{
"epoch": 0.739240506329114,
"grad_norm": 0.7586501836776733,
"learning_rate": 7.004326008833428e-07,
"loss": 0.3722,
"step": 3650
},
{
"epoch": 0.740253164556962,
"grad_norm": 0.10004276782274246,
"learning_rate": 6.924715436060409e-07,
"loss": 0.4142,
"step": 3655
},
{
"epoch": 0.7412658227848101,
"grad_norm": 0.12116418778896332,
"learning_rate": 6.845516982508628e-07,
"loss": 0.514,
"step": 3660
},
{
"epoch": 0.7422784810126583,
"grad_norm": 0.14440514147281647,
"learning_rate": 6.766731634993639e-07,
"loss": 0.5125,
"step": 3665
},
{
"epoch": 0.7432911392405064,
"grad_norm": 0.2194243222475052,
"learning_rate": 6.68836037518365e-07,
"loss": 0.4733,
"step": 3670
},
{
"epoch": 0.7443037974683544,
"grad_norm": 0.1936502903699875,
"learning_rate": 6.610404179587376e-07,
"loss": 0.4607,
"step": 3675
},
{
"epoch": 0.7453164556962025,
"grad_norm": 0.19898279011249542,
"learning_rate": 6.532864019541771e-07,
"loss": 0.4122,
"step": 3680
},
{
"epoch": 0.7463291139240507,
"grad_norm": 0.3182351291179657,
"learning_rate": 6.455740861199981e-07,
"loss": 0.3976,
"step": 3685
},
{
"epoch": 0.7473417721518988,
"grad_norm": 0.2338692545890808,
"learning_rate": 6.379035665519304e-07,
"loss": 0.3635,
"step": 3690
},
{
"epoch": 0.7483544303797468,
"grad_norm": 0.2808414399623871,
"learning_rate": 6.302749388249195e-07,
"loss": 0.3903,
"step": 3695
},
{
"epoch": 0.7493670886075949,
"grad_norm": 0.7633646130561829,
"learning_rate": 6.226882979919406e-07,
"loss": 0.4081,
"step": 3700
},
{
"epoch": 0.7503797468354431,
"grad_norm": 0.09589074552059174,
"learning_rate": 6.151437385828044e-07,
"loss": 0.4141,
"step": 3705
},
{
"epoch": 0.7513924050632912,
"grad_norm": 0.13365186750888824,
"learning_rate": 6.076413546029928e-07,
"loss": 0.517,
"step": 3710
},
{
"epoch": 0.7524050632911392,
"grad_norm": 0.16160084307193756,
"learning_rate": 6.001812395324748e-07,
"loss": 0.5141,
"step": 3715
},
{
"epoch": 0.7534177215189873,
"grad_norm": 0.18057221174240112,
"learning_rate": 5.927634863245492e-07,
"loss": 0.4528,
"step": 3720
},
{
"epoch": 0.7544303797468355,
"grad_norm": 0.15431033074855804,
"learning_rate": 5.853881874046869e-07,
"loss": 0.4152,
"step": 3725
},
{
"epoch": 0.7554430379746836,
"grad_norm": 0.19977331161499023,
"learning_rate": 5.780554346693711e-07,
"loss": 0.3954,
"step": 3730
},
{
"epoch": 0.7564556962025316,
"grad_norm": 0.21549344062805176,
"learning_rate": 5.707653194849645e-07,
"loss": 0.3993,
"step": 3735
},
{
"epoch": 0.7574683544303797,
"grad_norm": 0.2258233278989792,
"learning_rate": 5.6351793268656e-07,
"loss": 0.3773,
"step": 3740
},
{
"epoch": 0.7584810126582279,
"grad_norm": 0.40919986367225647,
"learning_rate": 5.563133645768543e-07,
"loss": 0.4361,
"step": 3745
},
{
"epoch": 0.759493670886076,
"grad_norm": 1.0033652782440186,
"learning_rate": 5.491517049250243e-07,
"loss": 0.419,
"step": 3750
},
{
"epoch": 0.760506329113924,
"grad_norm": 0.10725365579128265,
"learning_rate": 5.420330429656008e-07,
"loss": 0.4195,
"step": 3755
},
{
"epoch": 0.7615189873417721,
"grad_norm": 0.12712964415550232,
"learning_rate": 5.349574673973669e-07,
"loss": 0.5357,
"step": 3760
},
{
"epoch": 0.7625316455696203,
"grad_norm": 0.14178912341594696,
"learning_rate": 5.279250663822453e-07,
"loss": 0.4903,
"step": 3765
},
{
"epoch": 0.7635443037974684,
"grad_norm": 0.16502144932746887,
"learning_rate": 5.20935927544202e-07,
"loss": 0.4563,
"step": 3770
},
{
"epoch": 0.7645569620253164,
"grad_norm": 0.16947367787361145,
"learning_rate": 5.139901379681562e-07,
"loss": 0.3973,
"step": 3775
},
{
"epoch": 0.7655696202531646,
"grad_norm": 0.1997070163488388,
"learning_rate": 5.07087784198891e-07,
"loss": 0.4224,
"step": 3780
},
{
"epoch": 0.7665822784810127,
"grad_norm": 0.22137194871902466,
"learning_rate": 5.002289522399823e-07,
"loss": 0.4397,
"step": 3785
},
{
"epoch": 0.7675949367088608,
"grad_norm": 0.27175015211105347,
"learning_rate": 4.934137275527188e-07,
"loss": 0.4022,
"step": 3790
},
{
"epoch": 0.7686075949367088,
"grad_norm": 0.2860874533653259,
"learning_rate": 4.866421950550421e-07,
"loss": 0.407,
"step": 3795
},
{
"epoch": 0.769620253164557,
"grad_norm": 0.5431458950042725,
"learning_rate": 4.799144391204883e-07,
"loss": 0.4005,
"step": 3800
},
{
"epoch": 0.7706329113924051,
"grad_norm": 0.10014721751213074,
"learning_rate": 4.7323054357713487e-07,
"loss": 0.4579,
"step": 3805
},
{
"epoch": 0.7716455696202532,
"grad_norm": 0.11959011107683182,
"learning_rate": 4.6659059170655843e-07,
"loss": 0.4702,
"step": 3810
},
{
"epoch": 0.7726582278481012,
"grad_norm": 0.14652705192565918,
"learning_rate": 4.5999466624279513e-07,
"loss": 0.5059,
"step": 3815
},
{
"epoch": 0.7736708860759494,
"grad_norm": 0.17129305005073547,
"learning_rate": 4.5344284937131095e-07,
"loss": 0.5076,
"step": 3820
},
{
"epoch": 0.7746835443037975,
"grad_norm": 0.18978917598724365,
"learning_rate": 4.469352227279768e-07,
"loss": 0.4032,
"step": 3825
},
{
"epoch": 0.7756962025316456,
"grad_norm": 0.19936999678611755,
"learning_rate": 4.404718673980512e-07,
"loss": 0.4833,
"step": 3830
},
{
"epoch": 0.7767088607594936,
"grad_norm": 0.22007428109645844,
"learning_rate": 4.3405286391517173e-07,
"loss": 0.3645,
"step": 3835
},
{
"epoch": 0.7777215189873418,
"grad_norm": 0.25941571593284607,
"learning_rate": 4.2767829226035037e-07,
"loss": 0.4139,
"step": 3840
},
{
"epoch": 0.7787341772151899,
"grad_norm": 0.3206968605518341,
"learning_rate": 4.213482318609767e-07,
"loss": 0.4351,
"step": 3845
},
{
"epoch": 0.779746835443038,
"grad_norm": 0.849869430065155,
"learning_rate": 4.1506276158982744e-07,
"loss": 0.4489,
"step": 3850
},
{
"epoch": 0.780759493670886,
"grad_norm": 0.11732237786054611,
"learning_rate": 4.0882195976408607e-07,
"loss": 0.4208,
"step": 3855
},
{
"epoch": 0.7817721518987342,
"grad_norm": 0.14925657212734222,
"learning_rate": 4.0262590414436424e-07,
"loss": 0.5574,
"step": 3860
},
{
"epoch": 0.7827848101265823,
"grad_norm": 0.15243734419345856,
"learning_rate": 3.9647467193373795e-07,
"loss": 0.5599,
"step": 3865
},
{
"epoch": 0.7837974683544304,
"grad_norm": 0.1502576768398285,
"learning_rate": 3.903683397767752e-07,
"loss": 0.4451,
"step": 3870
},
{
"epoch": 0.7848101265822784,
"grad_norm": 0.191819429397583,
"learning_rate": 3.843069837585946e-07,
"loss": 0.4375,
"step": 3875
},
{
"epoch": 0.7858227848101266,
"grad_norm": 0.1775212585926056,
"learning_rate": 3.7829067940390666e-07,
"loss": 0.4219,
"step": 3880
},
{
"epoch": 0.7868354430379747,
"grad_norm": 0.20777887105941772,
"learning_rate": 3.7231950167607674e-07,
"loss": 0.4507,
"step": 3885
},
{
"epoch": 0.7878481012658228,
"grad_norm": 0.23604603111743927,
"learning_rate": 3.6639352497619405e-07,
"loss": 0.3698,
"step": 3890
},
{
"epoch": 0.7888607594936708,
"grad_norm": 0.257485955953598,
"learning_rate": 3.6051282314213527e-07,
"loss": 0.4041,
"step": 3895
},
{
"epoch": 0.789873417721519,
"grad_norm": 0.548670768737793,
"learning_rate": 3.546774694476569e-07,
"loss": 0.3936,
"step": 3900
},
{
"epoch": 0.7908860759493671,
"grad_norm": 0.13488371670246124,
"learning_rate": 3.4888753660147297e-07,
"loss": 0.4395,
"step": 3905
},
{
"epoch": 0.7918987341772152,
"grad_norm": 0.13122034072875977,
"learning_rate": 3.431430967463513e-07,
"loss": 0.5337,
"step": 3910
},
{
"epoch": 0.7929113924050633,
"grad_norm": 0.13643068075180054,
"learning_rate": 3.3744422145821893e-07,
"loss": 0.4381,
"step": 3915
},
{
"epoch": 0.7939240506329114,
"grad_norm": 0.15460450947284698,
"learning_rate": 3.3179098174526177e-07,
"loss": 0.3952,
"step": 3920
},
{
"epoch": 0.7949367088607595,
"grad_norm": 0.18093320727348328,
"learning_rate": 3.2618344804704954e-07,
"loss": 0.4627,
"step": 3925
},
{
"epoch": 0.7959493670886076,
"grad_norm": 0.2061995565891266,
"learning_rate": 3.206216902336507e-07,
"loss": 0.4491,
"step": 3930
},
{
"epoch": 0.7969620253164557,
"grad_norm": 0.2234310656785965,
"learning_rate": 3.1510577760476586e-07,
"loss": 0.3708,
"step": 3935
},
{
"epoch": 0.7979746835443038,
"grad_norm": 0.22730328142642975,
"learning_rate": 3.096357788888615e-07,
"loss": 0.4253,
"step": 3940
},
{
"epoch": 0.7989873417721519,
"grad_norm": 0.24523144960403442,
"learning_rate": 3.0421176224231594e-07,
"loss": 0.4152,
"step": 3945
},
{
"epoch": 0.8,
"grad_norm": 0.47707587480545044,
"learning_rate": 2.9883379524857064e-07,
"loss": 0.3822,
"step": 3950
},
{
"epoch": 0.8010126582278481,
"grad_norm": 0.10347505658864975,
"learning_rate": 2.9350194491728484e-07,
"loss": 0.4948,
"step": 3955
},
{
"epoch": 0.8020253164556962,
"grad_norm": 0.12368808686733246,
"learning_rate": 2.882162776835031e-07,
"loss": 0.479,
"step": 3960
},
{
"epoch": 0.8030379746835443,
"grad_norm": 0.16210949420928955,
"learning_rate": 2.829768594068267e-07,
"loss": 0.5109,
"step": 3965
},
{
"epoch": 0.8040506329113924,
"grad_norm": 0.15744373202323914,
"learning_rate": 2.777837553705935e-07,
"loss": 0.4726,
"step": 3970
},
{
"epoch": 0.8050632911392405,
"grad_norm": 0.1994476467370987,
"learning_rate": 2.7263703028106566e-07,
"loss": 0.4601,
"step": 3975
},
{
"epoch": 0.8060759493670886,
"grad_norm": 0.18476442992687225,
"learning_rate": 2.6753674826661954e-07,
"loss": 0.3917,
"step": 3980
},
{
"epoch": 0.8070886075949367,
"grad_norm": 0.21709558367729187,
"learning_rate": 2.624829728769509e-07,
"loss": 0.4254,
"step": 3985
},
{
"epoch": 0.8081012658227849,
"grad_norm": 0.21527519822120667,
"learning_rate": 2.5747576708228025e-07,
"loss": 0.3935,
"step": 3990
},
{
"epoch": 0.8091139240506329,
"grad_norm": 0.2815234363079071,
"learning_rate": 2.5251519327256977e-07,
"loss": 0.3794,
"step": 3995
},
{
"epoch": 0.810126582278481,
"grad_norm": 0.5821972489356995,
"learning_rate": 2.476013132567454e-07,
"loss": 0.3868,
"step": 4000
},
{
"epoch": 0.8111392405063291,
"grad_norm": 0.1082138866186142,
"learning_rate": 2.4273418826192736e-07,
"loss": 0.4239,
"step": 4005
},
{
"epoch": 0.8121518987341773,
"grad_norm": 0.1272335797548294,
"learning_rate": 2.3791387893266602e-07,
"loss": 0.5149,
"step": 4010
},
{
"epoch": 0.8131645569620253,
"grad_norm": 0.17124220728874207,
"learning_rate": 2.3314044533018618e-07,
"loss": 0.4957,
"step": 4015
},
{
"epoch": 0.8141772151898734,
"grad_norm": 0.17160958051681519,
"learning_rate": 2.2841394693164086e-07,
"loss": 0.5115,
"step": 4020
},
{
"epoch": 0.8151898734177215,
"grad_norm": 0.1839609593153,
"learning_rate": 2.2373444262936724e-07,
"loss": 0.4393,
"step": 4025
},
{
"epoch": 0.8162025316455697,
"grad_norm": 0.19161073863506317,
"learning_rate": 2.1910199073015724e-07,
"loss": 0.4096,
"step": 4030
},
{
"epoch": 0.8172151898734177,
"grad_norm": 0.19516105949878693,
"learning_rate": 2.145166489545236e-07,
"loss": 0.4265,
"step": 4035
},
{
"epoch": 0.8182278481012658,
"grad_norm": 0.2757415473461151,
"learning_rate": 2.099784744359896e-07,
"loss": 0.389,
"step": 4040
},
{
"epoch": 0.8192405063291139,
"grad_norm": 0.32892289757728577,
"learning_rate": 2.0548752372037083e-07,
"loss": 0.4007,
"step": 4045
},
{
"epoch": 0.8202531645569621,
"grad_norm": 0.5428944826126099,
"learning_rate": 2.010438527650713e-07,
"loss": 0.3928,
"step": 4050
},
{
"epoch": 0.8212658227848101,
"grad_norm": 0.12007920444011688,
"learning_rate": 1.9664751693839166e-07,
"loss": 0.4631,
"step": 4055
},
{
"epoch": 0.8222784810126582,
"grad_norm": 0.12291346490383148,
"learning_rate": 1.9229857101883007e-07,
"loss": 0.5238,
"step": 4060
},
{
"epoch": 0.8232911392405063,
"grad_norm": 0.17395402491092682,
"learning_rate": 1.8799706919440904e-07,
"loss": 0.4979,
"step": 4065
},
{
"epoch": 0.8243037974683545,
"grad_norm": 0.16262653470039368,
"learning_rate": 1.8374306506199334e-07,
"loss": 0.4664,
"step": 4070
},
{
"epoch": 0.8253164556962025,
"grad_norm": 0.1795371174812317,
"learning_rate": 1.7953661162662504e-07,
"loss": 0.4336,
"step": 4075
},
{
"epoch": 0.8263291139240506,
"grad_norm": 0.2007470577955246,
"learning_rate": 1.7537776130086558e-07,
"loss": 0.3647,
"step": 4080
},
{
"epoch": 0.8273417721518987,
"grad_norm": 0.2502759099006653,
"learning_rate": 1.7126656590413523e-07,
"loss": 0.3679,
"step": 4085
},
{
"epoch": 0.8283544303797469,
"grad_norm": 0.21397298574447632,
"learning_rate": 1.6720307666207645e-07,
"loss": 0.3645,
"step": 4090
},
{
"epoch": 0.8293670886075949,
"grad_norm": 0.2824733853340149,
"learning_rate": 1.6318734420590928e-07,
"loss": 0.3871,
"step": 4095
},
{
"epoch": 0.830379746835443,
"grad_norm": 0.7403417229652405,
"learning_rate": 1.5921941857180277e-07,
"loss": 0.4183,
"step": 4100
},
{
"epoch": 0.8313924050632912,
"grad_norm": 0.12091076374053955,
"learning_rate": 1.5529934920025123e-07,
"loss": 0.4705,
"step": 4105
},
{
"epoch": 0.8324050632911393,
"grad_norm": 0.15003369748592377,
"learning_rate": 1.5142718493545803e-07,
"loss": 0.508,
"step": 4110
},
{
"epoch": 0.8334177215189873,
"grad_norm": 0.16014327108860016,
"learning_rate": 1.4760297402472843e-07,
"loss": 0.5583,
"step": 4115
},
{
"epoch": 0.8344303797468354,
"grad_norm": 0.16555018723011017,
"learning_rate": 1.438267641178661e-07,
"loss": 0.4875,
"step": 4120
},
{
"epoch": 0.8354430379746836,
"grad_norm": 0.16284528374671936,
"learning_rate": 1.4009860226657977e-07,
"loss": 0.4613,
"step": 4125
},
{
"epoch": 0.8364556962025317,
"grad_norm": 0.17814618349075317,
"learning_rate": 1.3641853492389976e-07,
"loss": 0.3998,
"step": 4130
},
{
"epoch": 0.8374683544303797,
"grad_norm": 0.2173750251531601,
"learning_rate": 1.327866079435953e-07,
"loss": 0.4046,
"step": 4135
},
{
"epoch": 0.8384810126582278,
"grad_norm": 0.2331230491399765,
"learning_rate": 1.2920286657960523e-07,
"loss": 0.4004,
"step": 4140
},
{
"epoch": 0.839493670886076,
"grad_norm": 0.31456252932548523,
"learning_rate": 1.256673554854757e-07,
"loss": 0.416,
"step": 4145
},
{
"epoch": 0.8405063291139241,
"grad_norm": 0.4124484956264496,
"learning_rate": 1.221801187137994e-07,
"loss": 0.392,
"step": 4150
},
{
"epoch": 0.8415189873417721,
"grad_norm": 0.11082947254180908,
"learning_rate": 1.1874119971567109e-07,
"loss": 0.4219,
"step": 4155
},
{
"epoch": 0.8425316455696202,
"grad_norm": 0.137724369764328,
"learning_rate": 1.1535064134014349e-07,
"loss": 0.5606,
"step": 4160
},
{
"epoch": 0.8435443037974684,
"grad_norm": 0.148650124669075,
"learning_rate": 1.120084858336936e-07,
"loss": 0.5352,
"step": 4165
},
{
"epoch": 0.8445569620253165,
"grad_norm": 0.19314394891262054,
"learning_rate": 1.0871477483969949e-07,
"loss": 0.4619,
"step": 4170
},
{
"epoch": 0.8455696202531645,
"grad_norm": 0.17370128631591797,
"learning_rate": 1.054695493979163e-07,
"loss": 0.4683,
"step": 4175
},
{
"epoch": 0.8465822784810126,
"grad_norm": 0.1875048279762268,
"learning_rate": 1.0227284994396868e-07,
"loss": 0.4172,
"step": 4180
},
{
"epoch": 0.8475949367088608,
"grad_norm": 0.23327000439167023,
"learning_rate": 9.912471630884489e-08,
"loss": 0.44,
"step": 4185
},
{
"epoch": 0.8486075949367089,
"grad_norm": 0.24482522904872894,
"learning_rate": 9.602518771840218e-08,
"loss": 0.4094,
"step": 4190
},
{
"epoch": 0.8496202531645569,
"grad_norm": 0.40738484263420105,
"learning_rate": 9.297430279287777e-08,
"loss": 0.3816,
"step": 4195
},
{
"epoch": 0.850632911392405,
"grad_norm": 0.5299381613731384,
"learning_rate": 8.997209954640439e-08,
"loss": 0.3866,
"step": 4200
},
{
"epoch": 0.8516455696202532,
"grad_norm": 0.11590830236673355,
"learning_rate": 8.701861538654309e-08,
"loss": 0.4854,
"step": 4205
},
{
"epoch": 0.8526582278481013,
"grad_norm": 0.12266714870929718,
"learning_rate": 8.411388711381073e-08,
"loss": 0.4658,
"step": 4210
},
{
"epoch": 0.8536708860759493,
"grad_norm": 0.1612856090068817,
"learning_rate": 8.125795092122434e-08,
"loss": 0.4938,
"step": 4215
},
{
"epoch": 0.8546835443037974,
"grad_norm": 0.1906183362007141,
"learning_rate": 7.845084239385169e-08,
"loss": 0.4947,
"step": 4220
},
{
"epoch": 0.8556962025316456,
"grad_norm": 0.17465056478977203,
"learning_rate": 7.569259650836279e-08,
"loss": 0.4358,
"step": 4225
},
{
"epoch": 0.8567088607594937,
"grad_norm": 0.17438243329524994,
"learning_rate": 7.298324763260132e-08,
"loss": 0.4226,
"step": 4230
},
{
"epoch": 0.8577215189873417,
"grad_norm": 0.22130343317985535,
"learning_rate": 7.03228295251499e-08,
"loss": 0.3903,
"step": 4235
},
{
"epoch": 0.8587341772151899,
"grad_norm": 0.22619099915027618,
"learning_rate": 6.771137533491255e-08,
"loss": 0.3699,
"step": 4240
},
{
"epoch": 0.859746835443038,
"grad_norm": 0.29880502820014954,
"learning_rate": 6.51489176007014e-08,
"loss": 0.4389,
"step": 4245
},
{
"epoch": 0.8607594936708861,
"grad_norm": 0.5453495383262634,
"learning_rate": 6.263548825083065e-08,
"loss": 0.391,
"step": 4250
},
{
"epoch": 0.8617721518987341,
"grad_norm": 0.10160661488771439,
"learning_rate": 6.017111860272007e-08,
"loss": 0.454,
"step": 4255
},
{
"epoch": 0.8627848101265823,
"grad_norm": 0.14306581020355225,
"learning_rate": 5.775583936250239e-08,
"loss": 0.4992,
"step": 4260
},
{
"epoch": 0.8637974683544304,
"grad_norm": 0.1603081077337265,
"learning_rate": 5.538968062464278e-08,
"loss": 0.5699,
"step": 4265
},
{
"epoch": 0.8648101265822785,
"grad_norm": 0.1580309420824051,
"learning_rate": 5.3072671871563504e-08,
"loss": 0.5076,
"step": 4270
},
{
"epoch": 0.8658227848101265,
"grad_norm": 0.18477016687393188,
"learning_rate": 5.080484197327584e-08,
"loss": 0.4708,
"step": 4275
},
{
"epoch": 0.8668354430379747,
"grad_norm": 0.17577703297138214,
"learning_rate": 4.858621918702122e-08,
"loss": 0.4054,
"step": 4280
},
{
"epoch": 0.8678481012658228,
"grad_norm": 0.22820906341075897,
"learning_rate": 4.6416831156917744e-08,
"loss": 0.3763,
"step": 4285
},
{
"epoch": 0.8688607594936709,
"grad_norm": 0.24788212776184082,
"learning_rate": 4.4296704913618255e-08,
"loss": 0.4175,
"step": 4290
},
{
"epoch": 0.8698734177215189,
"grad_norm": 0.23278400301933289,
"learning_rate": 4.222586687397012e-08,
"loss": 0.3737,
"step": 4295
},
{
"epoch": 0.8708860759493671,
"grad_norm": 0.5961129069328308,
"learning_rate": 4.020434284068974e-08,
"loss": 0.4144,
"step": 4300
},
{
"epoch": 0.8718987341772152,
"grad_norm": 0.1277296096086502,
"learning_rate": 3.823215800203794e-08,
"loss": 0.435,
"step": 4305
},
{
"epoch": 0.8729113924050633,
"grad_norm": 0.13008461892604828,
"learning_rate": 3.630933693150817e-08,
"loss": 0.4799,
"step": 4310
},
{
"epoch": 0.8739240506329113,
"grad_norm": 0.1518930047750473,
"learning_rate": 3.443590358751924e-08,
"loss": 0.5635,
"step": 4315
},
{
"epoch": 0.8749367088607595,
"grad_norm": 0.1718684732913971,
"learning_rate": 3.2611881313117316e-08,
"loss": 0.5103,
"step": 4320
},
{
"epoch": 0.8759493670886076,
"grad_norm": 0.23731859028339386,
"learning_rate": 3.083729283568415e-08,
"loss": 0.4433,
"step": 4325
},
{
"epoch": 0.8769620253164557,
"grad_norm": 0.19327382743358612,
"learning_rate": 2.9112160266655972e-08,
"loss": 0.4194,
"step": 4330
},
{
"epoch": 0.8779746835443037,
"grad_norm": 0.20755337178707123,
"learning_rate": 2.743650510124551e-08,
"loss": 0.4052,
"step": 4335
},
{
"epoch": 0.8789873417721519,
"grad_norm": 0.27164265513420105,
"learning_rate": 2.5810348218176403e-08,
"loss": 0.3558,
"step": 4340
},
{
"epoch": 0.88,
"grad_norm": 0.29403576254844666,
"learning_rate": 2.423370987941986e-08,
"loss": 0.3889,
"step": 4345
},
{
"epoch": 0.8810126582278481,
"grad_norm": 0.45856744050979614,
"learning_rate": 2.270660972994687e-08,
"loss": 0.3908,
"step": 4350
},
{
"epoch": 0.8820253164556962,
"grad_norm": 0.11312643438577652,
"learning_rate": 2.1229066797479044e-08,
"loss": 0.424,
"step": 4355
},
{
"epoch": 0.8830379746835443,
"grad_norm": 0.1443646252155304,
"learning_rate": 1.9801099492253725e-08,
"loss": 0.4893,
"step": 4360
},
{
"epoch": 0.8840506329113924,
"grad_norm": 0.15652401745319366,
"learning_rate": 1.842272560679481e-08,
"loss": 0.4558,
"step": 4365
},
{
"epoch": 0.8850632911392405,
"grad_norm": 0.16332998871803284,
"learning_rate": 1.7093962315688936e-08,
"loss": 0.4475,
"step": 4370
},
{
"epoch": 0.8860759493670886,
"grad_norm": 0.17202718555927277,
"learning_rate": 1.5814826175374996e-08,
"loss": 0.4587,
"step": 4375
},
{
"epoch": 0.8870886075949367,
"grad_norm": 0.20603860914707184,
"learning_rate": 1.4585333123934062e-08,
"loss": 0.4442,
"step": 4380
},
{
"epoch": 0.8881012658227848,
"grad_norm": 0.2088196873664856,
"learning_rate": 1.3405498480894006e-08,
"loss": 0.3953,
"step": 4385
},
{
"epoch": 0.8891139240506329,
"grad_norm": 0.2510204315185547,
"learning_rate": 1.227533694703542e-08,
"loss": 0.439,
"step": 4390
},
{
"epoch": 0.890126582278481,
"grad_norm": 0.3226676285266876,
"learning_rate": 1.1194862604211763e-08,
"loss": 0.4086,
"step": 4395
},
{
"epoch": 0.8911392405063291,
"grad_norm": 0.5400206446647644,
"learning_rate": 1.0164088915171731e-08,
"loss": 0.3897,
"step": 4400
},
{
"epoch": 0.8921518987341772,
"grad_norm": 0.09648526459932327,
"learning_rate": 9.183028723392272e-09,
"loss": 0.3943,
"step": 4405
},
{
"epoch": 0.8931645569620253,
"grad_norm": 0.11990416049957275,
"learning_rate": 8.251694252917384e-09,
"loss": 0.4563,
"step": 4410
},
{
"epoch": 0.8941772151898734,
"grad_norm": 0.14837533235549927,
"learning_rate": 7.370097108208906e-09,
"loss": 0.5484,
"step": 4415
},
{
"epoch": 0.8951898734177215,
"grad_norm": 0.17313119769096375,
"learning_rate": 6.538248273997737e-09,
"loss": 0.5057,
"step": 4420
},
{
"epoch": 0.8962025316455696,
"grad_norm": 0.19712094962596893,
"learning_rate": 5.756158115150178e-09,
"loss": 0.4535,
"step": 4425
},
{
"epoch": 0.8972151898734178,
"grad_norm": 0.16794025897979736,
"learning_rate": 5.0238363765378044e-09,
"loss": 0.3651,
"step": 4430
},
{
"epoch": 0.8982278481012658,
"grad_norm": 0.1980295181274414,
"learning_rate": 4.3412921829149016e-09,
"loss": 0.3699,
"step": 4435
},
{
"epoch": 0.8992405063291139,
"grad_norm": 0.24372468888759613,
"learning_rate": 3.7085340388061104e-09,
"loss": 0.3597,
"step": 4440
},
{
"epoch": 0.900253164556962,
"grad_norm": 0.27603617310523987,
"learning_rate": 3.1255698284016197e-09,
"loss": 0.3631,
"step": 4445
},
{
"epoch": 0.9012658227848102,
"grad_norm": 0.4913368821144104,
"learning_rate": 2.5924068154550285e-09,
"loss": 0.3973,
"step": 4450
},
{
"epoch": 0.9022784810126582,
"grad_norm": 0.11197266727685928,
"learning_rate": 2.1090516431967463e-09,
"loss": 0.4426,
"step": 4455
},
{
"epoch": 0.9032911392405063,
"grad_norm": 0.12932994961738586,
"learning_rate": 1.6755103342500633e-09,
"loss": 0.5295,
"step": 4460
},
{
"epoch": 0.9043037974683544,
"grad_norm": 0.1532626897096634,
"learning_rate": 1.2917882905543187e-09,
"loss": 0.4934,
"step": 4465
},
{
"epoch": 0.9053164556962026,
"grad_norm": 0.17139685153961182,
"learning_rate": 9.578902932996236e-10,
"loss": 0.486,
"step": 4470
},
{
"epoch": 0.9063291139240506,
"grad_norm": 0.1812489926815033,
"learning_rate": 6.738205028677946e-10,
"loss": 0.4438,
"step": 4475
},
{
"epoch": 0.9073417721518987,
"grad_norm": 0.23248408734798431,
"learning_rate": 4.39582458778176e-10,
"loss": 0.4381,
"step": 4480
},
{
"epoch": 0.9083544303797468,
"grad_norm": 0.273440957069397,
"learning_rate": 2.55179079644563e-10,
"loss": 0.427,
"step": 4485
},
{
"epoch": 0.909367088607595,
"grad_norm": 0.23511122167110443,
"learning_rate": 1.206126631401183e-10,
"loss": 0.3707,
"step": 4490
},
{
"epoch": 0.9103797468354431,
"grad_norm": 0.3231514096260071,
"learning_rate": 3.588488596717454e-11,
"loss": 0.4062,
"step": 4495
},
{
"epoch": 0.9113924050632911,
"grad_norm": 0.7185713648796082,
"learning_rate": 9.968038368057818e-13,
"loss": 0.4184,
"step": 4500
},
{
"epoch": 0.9124050632911392,
"grad_norm": 0.10860884189605713,
"learning_rate": 6.41346604322123e-07,
"loss": 0.4172,
"step": 4505
},
{
"epoch": 0.9134177215189874,
"grad_norm": 0.15159861743450165,
"learning_rate": 6.350991992819552e-07,
"loss": 0.4867,
"step": 4510
},
{
"epoch": 0.9144303797468355,
"grad_norm": 0.1377243995666504,
"learning_rate": 6.288797466168074e-07,
"loss": 0.486,
"step": 4515
},
{
"epoch": 0.9154430379746835,
"grad_norm": 0.15941649675369263,
"learning_rate": 6.226882979919406e-07,
"loss": 0.4213,
"step": 4520
},
{
"epoch": 0.9164556962025316,
"grad_norm": 0.18730419874191284,
"learning_rate": 6.165249048399813e-07,
"loss": 0.4896,
"step": 4525
},
{
"epoch": 0.9174683544303798,
"grad_norm": 0.19335976243019104,
"learning_rate": 6.103896183605033e-07,
"loss": 0.4098,
"step": 4530
},
{
"epoch": 0.9184810126582279,
"grad_norm": 0.24511203169822693,
"learning_rate": 6.042824895195919e-07,
"loss": 0.3924,
"step": 4535
},
{
"epoch": 0.9194936708860759,
"grad_norm": 0.24723470211029053,
"learning_rate": 5.982035690494296e-07,
"loss": 0.4123,
"step": 4540
},
{
"epoch": 0.920506329113924,
"grad_norm": 0.28107598423957825,
"learning_rate": 5.921529074478662e-07,
"loss": 0.4102,
"step": 4545
},
{
"epoch": 0.9215189873417722,
"grad_norm": 0.6902102828025818,
"learning_rate": 5.861305549780105e-07,
"loss": 0.431,
"step": 4550
},
{
"epoch": 0.9225316455696203,
"grad_norm": 0.10981158912181854,
"learning_rate": 5.801365616677993e-07,
"loss": 0.416,
"step": 4555
},
{
"epoch": 0.9235443037974683,
"grad_norm": 0.12923768162727356,
"learning_rate": 5.741709773095933e-07,
"loss": 0.5106,
"step": 4560
},
{
"epoch": 0.9245569620253165,
"grad_norm": 0.171346515417099,
"learning_rate": 5.682338514597549e-07,
"loss": 0.5029,
"step": 4565
},
{
"epoch": 0.9255696202531646,
"grad_norm": 0.1871034801006317,
"learning_rate": 5.623252334382464e-07,
"loss": 0.4478,
"step": 4570
},
{
"epoch": 0.9265822784810127,
"grad_norm": 0.2404191642999649,
"learning_rate": 5.564451723282077e-07,
"loss": 0.4288,
"step": 4575
},
{
"epoch": 0.9275949367088607,
"grad_norm": 0.18959856033325195,
"learning_rate": 5.505937169755599e-07,
"loss": 0.4433,
"step": 4580
},
{
"epoch": 0.9286075949367089,
"grad_norm": 0.24367639422416687,
"learning_rate": 5.447709159885909e-07,
"loss": 0.3999,
"step": 4585
},
{
"epoch": 0.929620253164557,
"grad_norm": 0.25893864035606384,
"learning_rate": 5.389768177375611e-07,
"loss": 0.4151,
"step": 4590
},
{
"epoch": 0.9306329113924051,
"grad_norm": 0.28106915950775146,
"learning_rate": 5.332114703542903e-07,
"loss": 0.3672,
"step": 4595
},
{
"epoch": 0.9316455696202531,
"grad_norm": 0.8240860104560852,
"learning_rate": 5.274749217317671e-07,
"loss": 0.4314,
"step": 4600
},
{
"epoch": 0.9326582278481013,
"grad_norm": 0.12054485827684402,
"learning_rate": 5.217672195237446e-07,
"loss": 0.5007,
"step": 4605
},
{
"epoch": 0.9336708860759494,
"grad_norm": 0.12769296765327454,
"learning_rate": 5.160884111443531e-07,
"loss": 0.4876,
"step": 4610
},
{
"epoch": 0.9346835443037975,
"grad_norm": 0.16503486037254333,
"learning_rate": 5.104385437676932e-07,
"loss": 0.4923,
"step": 4615
},
{
"epoch": 0.9356962025316455,
"grad_norm": 0.15395459532737732,
"learning_rate": 5.048176643274584e-07,
"loss": 0.5089,
"step": 4620
},
{
"epoch": 0.9367088607594937,
"grad_norm": 0.1737959384918213,
"learning_rate": 4.99225819516531e-07,
"loss": 0.4014,
"step": 4625
},
{
"epoch": 0.9377215189873418,
"grad_norm": 0.2168796956539154,
"learning_rate": 4.936630557866093e-07,
"loss": 0.3595,
"step": 4630
},
{
"epoch": 0.9387341772151899,
"grad_norm": 0.21767963469028473,
"learning_rate": 4.881294193478069e-07,
"loss": 0.3961,
"step": 4635
},
{
"epoch": 0.9397468354430379,
"grad_norm": 0.272427499294281,
"learning_rate": 4.826249561682805e-07,
"loss": 0.4299,
"step": 4640
},
{
"epoch": 0.9407594936708861,
"grad_norm": 0.28430747985839844,
"learning_rate": 4.771497119738393e-07,
"loss": 0.3789,
"step": 4645
},
{
"epoch": 0.9417721518987342,
"grad_norm": 0.6103779673576355,
"learning_rate": 4.7170373224757387e-07,
"loss": 0.4086,
"step": 4650
},
{
"epoch": 0.9427848101265823,
"grad_norm": 0.0985502228140831,
"learning_rate": 4.6628706222946854e-07,
"loss": 0.4127,
"step": 4655
},
{
"epoch": 0.9437974683544303,
"grad_norm": 0.13281631469726562,
"learning_rate": 4.608997469160356e-07,
"loss": 0.5201,
"step": 4660
},
{
"epoch": 0.9448101265822785,
"grad_norm": 0.16024696826934814,
"learning_rate": 4.5554183105993037e-07,
"loss": 0.484,
"step": 4665
},
{
"epoch": 0.9458227848101266,
"grad_norm": 0.1749877631664276,
"learning_rate": 4.502133591695934e-07,
"loss": 0.4824,
"step": 4670
},
{
"epoch": 0.9468354430379747,
"grad_norm": 0.2039998322725296,
"learning_rate": 4.4491437550886554e-07,
"loss": 0.4035,
"step": 4675
},
{
"epoch": 0.9478481012658228,
"grad_norm": 0.22236007452011108,
"learning_rate": 4.3964492409663334e-07,
"loss": 0.3826,
"step": 4680
},
{
"epoch": 0.9488607594936709,
"grad_norm": 0.23315949738025665,
"learning_rate": 4.34405048706453e-07,
"loss": 0.4269,
"step": 4685
},
{
"epoch": 0.949873417721519,
"grad_norm": 0.28179746866226196,
"learning_rate": 4.291947928661961e-07,
"loss": 0.3772,
"step": 4690
},
{
"epoch": 0.9508860759493671,
"grad_norm": 0.46736136078834534,
"learning_rate": 4.240141998576794e-07,
"loss": 0.4178,
"step": 4695
},
{
"epoch": 0.9518987341772152,
"grad_norm": 0.4249938130378723,
"learning_rate": 4.188633127163115e-07,
"loss": 0.4064,
"step": 4700
},
{
"epoch": 0.9529113924050633,
"grad_norm": 0.11935340613126755,
"learning_rate": 4.137421742307312e-07,
"loss": 0.4997,
"step": 4705
},
{
"epoch": 0.9539240506329114,
"grad_norm": 0.16850705444812775,
"learning_rate": 4.086508269424573e-07,
"loss": 0.5142,
"step": 4710
},
{
"epoch": 0.9549367088607595,
"grad_norm": 0.15970005095005035,
"learning_rate": 4.035893131455279e-07,
"loss": 0.5048,
"step": 4715
},
{
"epoch": 0.9559493670886076,
"grad_norm": 0.16822274029254913,
"learning_rate": 3.985576748861565e-07,
"loss": 0.4233,
"step": 4720
},
{
"epoch": 0.9569620253164557,
"grad_norm": 0.22004909813404083,
"learning_rate": 3.935559539623745e-07,
"loss": 0.4302,
"step": 4725
},
{
"epoch": 0.9579746835443038,
"grad_norm": 0.2314930260181427,
"learning_rate": 3.8858419192369585e-07,
"loss": 0.434,
"step": 4730
},
{
"epoch": 0.958987341772152,
"grad_norm": 0.20686328411102295,
"learning_rate": 3.836424300707577e-07,
"loss": 0.3893,
"step": 4735
},
{
"epoch": 0.96,
"grad_norm": 0.2939393222332001,
"learning_rate": 3.7873070945498897e-07,
"loss": 0.3727,
"step": 4740
},
{
"epoch": 0.9610126582278481,
"grad_norm": 0.28259849548339844,
"learning_rate": 3.7384907087826087e-07,
"loss": 0.4401,
"step": 4745
},
{
"epoch": 0.9620253164556962,
"grad_norm": 0.4876551628112793,
"learning_rate": 3.68997554892557e-07,
"loss": 0.4029,
"step": 4750
},
{
"epoch": 0.9630379746835444,
"grad_norm": 0.09749415516853333,
"learning_rate": 3.6417620179962637e-07,
"loss": 0.4295,
"step": 4755
},
{
"epoch": 0.9640506329113924,
"grad_norm": 0.13140413165092468,
"learning_rate": 3.593850516506567e-07,
"loss": 0.5433,
"step": 4760
},
{
"epoch": 0.9650632911392405,
"grad_norm": 0.1493423581123352,
"learning_rate": 3.5462414424593637e-07,
"loss": 0.5619,
"step": 4765
},
{
"epoch": 0.9660759493670886,
"grad_norm": 0.1732608675956726,
"learning_rate": 3.498935191345276e-07,
"loss": 0.5089,
"step": 4770
},
{
"epoch": 0.9670886075949368,
"grad_norm": 0.1850598007440567,
"learning_rate": 3.451932156139352e-07,
"loss": 0.4467,
"step": 4775
},
{
"epoch": 0.9681012658227848,
"grad_norm": 0.20144909620285034,
"learning_rate": 3.405232727297833e-07,
"loss": 0.437,
"step": 4780
},
{
"epoch": 0.9691139240506329,
"grad_norm": 0.204118549823761,
"learning_rate": 3.358837292754857e-07,
"loss": 0.3854,
"step": 4785
},
{
"epoch": 0.970126582278481,
"grad_norm": 0.21772193908691406,
"learning_rate": 3.312746237919288e-07,
"loss": 0.4087,
"step": 4790
},
{
"epoch": 0.9711392405063292,
"grad_norm": 0.3913586735725403,
"learning_rate": 3.266959945671499e-07,
"loss": 0.3659,
"step": 4795
},
{
"epoch": 0.9721518987341772,
"grad_norm": 1.0478923320770264,
"learning_rate": 3.22147879636018e-07,
"loss": 0.4337,
"step": 4800
},
{
"epoch": 0.9731645569620253,
"grad_norm": 0.11660108715295792,
"learning_rate": 3.176303167799177e-07,
"loss": 0.4705,
"step": 4805
},
{
"epoch": 0.9741772151898734,
"grad_norm": 0.13056664168834686,
"learning_rate": 3.1314334352643724e-07,
"loss": 0.5276,
"step": 4810
},
{
"epoch": 0.9751898734177216,
"grad_norm": 0.14374294877052307,
"learning_rate": 3.08686997149056e-07,
"loss": 0.505,
"step": 4815
},
{
"epoch": 0.9762025316455696,
"grad_norm": 0.1691606044769287,
"learning_rate": 3.042613146668347e-07,
"loss": 0.4826,
"step": 4820
},
{
"epoch": 0.9772151898734177,
"grad_norm": 0.16940155625343323,
"learning_rate": 2.998663328441058e-07,
"loss": 0.3882,
"step": 4825
},
{
"epoch": 0.9782278481012658,
"grad_norm": 0.21718168258666992,
"learning_rate": 2.955020881901724e-07,
"loss": 0.4677,
"step": 4830
},
{
"epoch": 0.979240506329114,
"grad_norm": 0.2414480447769165,
"learning_rate": 2.9116861695900196e-07,
"loss": 0.4094,
"step": 4835
},
{
"epoch": 0.980253164556962,
"grad_norm": 0.2923324406147003,
"learning_rate": 2.868659551489241e-07,
"loss": 0.4149,
"step": 4840
},
{
"epoch": 0.9812658227848101,
"grad_norm": 0.32415714859962463,
"learning_rate": 2.825941385023358e-07,
"loss": 0.3714,
"step": 4845
},
{
"epoch": 0.9822784810126582,
"grad_norm": 0.5138719081878662,
"learning_rate": 2.7835320250540097e-07,
"loss": 0.4357,
"step": 4850
},
{
"epoch": 0.9832911392405064,
"grad_norm": 0.1094016283750534,
"learning_rate": 2.741431823877565e-07,
"loss": 0.4502,
"step": 4855
},
{
"epoch": 0.9843037974683544,
"grad_norm": 0.13364043831825256,
"learning_rate": 2.6996411312221903e-07,
"loss": 0.5117,
"step": 4860
},
{
"epoch": 0.9853164556962025,
"grad_norm": 0.14276044070720673,
"learning_rate": 2.658160294244967e-07,
"loss": 0.5047,
"step": 4865
},
{
"epoch": 0.9863291139240506,
"grad_norm": 0.16736774146556854,
"learning_rate": 2.6169896575289717e-07,
"loss": 0.5099,
"step": 4870
},
{
"epoch": 0.9873417721518988,
"grad_norm": 0.1595831662416458,
"learning_rate": 2.576129563080469e-07,
"loss": 0.4567,
"step": 4875
},
{
"epoch": 0.9883544303797468,
"grad_norm": 0.1982174515724182,
"learning_rate": 2.53558035032599e-07,
"loss": 0.4177,
"step": 4880
},
{
"epoch": 0.9893670886075949,
"grad_norm": 0.20998278260231018,
"learning_rate": 2.495342356109589e-07,
"loss": 0.4133,
"step": 4885
},
{
"epoch": 0.990379746835443,
"grad_norm": 0.24803848564624786,
"learning_rate": 2.455415914690007e-07,
"loss": 0.3886,
"step": 4890
},
{
"epoch": 0.9913924050632912,
"grad_norm": 0.37692344188690186,
"learning_rate": 2.415801357737912e-07,
"loss": 0.4002,
"step": 4895
},
{
"epoch": 0.9924050632911392,
"grad_norm": 0.6091468930244446,
"learning_rate": 2.3764990143331042e-07,
"loss": 0.4202,
"step": 4900
},
{
"epoch": 0.9934177215189873,
"grad_norm": 0.10248316079378128,
"learning_rate": 2.3375092109618521e-07,
"loss": 0.4225,
"step": 4905
},
{
"epoch": 0.9944303797468355,
"grad_norm": 0.14465945959091187,
"learning_rate": 2.2988322715141063e-07,
"loss": 0.5309,
"step": 4910
},
{
"epoch": 0.9954430379746836,
"grad_norm": 0.20383016765117645,
"learning_rate": 2.260468517280878e-07,
"loss": 0.4487,
"step": 4915
},
{
"epoch": 0.9964556962025316,
"grad_norm": 0.17198005318641663,
"learning_rate": 2.2224182669514958e-07,
"loss": 0.407,
"step": 4920
},
{
"epoch": 0.9974683544303797,
"grad_norm": 0.21341922879219055,
"learning_rate": 2.1846818366110286e-07,
"loss": 0.3467,
"step": 4925
},
{
"epoch": 0.9984810126582279,
"grad_norm": 0.31665244698524475,
"learning_rate": 2.147259539737618e-07,
"loss": 0.4034,
"step": 4930
},
{
"epoch": 0.999493670886076,
"grad_norm": 0.3315502405166626,
"learning_rate": 2.1101516871998927e-07,
"loss": 0.3793,
"step": 4935
},
{
"epoch": 1.0006075949367088,
"grad_norm": 0.09771858900785446,
"learning_rate": 2.073358587254357e-07,
"loss": 0.4936,
"step": 4940
},
{
"epoch": 1.001620253164557,
"grad_norm": 0.10631760209798813,
"learning_rate": 2.0368805455428783e-07,
"loss": 0.4839,
"step": 4945
},
{
"epoch": 1.002632911392405,
"grad_norm": 0.15343673527240753,
"learning_rate": 2.0007178650901112e-07,
"loss": 0.5751,
"step": 4950
},
{
"epoch": 1.003645569620253,
"grad_norm": 0.18907710909843445,
"learning_rate": 1.9648708463009922e-07,
"loss": 0.5181,
"step": 4955
},
{
"epoch": 1.0046582278481013,
"grad_norm": 0.19511711597442627,
"learning_rate": 1.9293397869582306e-07,
"loss": 0.5034,
"step": 4960
},
{
"epoch": 1.0056708860759493,
"grad_norm": 0.20825132727622986,
"learning_rate": 1.8941249822198625e-07,
"loss": 0.4324,
"step": 4965
},
{
"epoch": 1.0066835443037974,
"grad_norm": 0.19913554191589355,
"learning_rate": 1.8592267246167803e-07,
"loss": 0.4364,
"step": 4970
},
{
"epoch": 1.0076962025316456,
"grad_norm": 0.2075047791004181,
"learning_rate": 1.824645304050314e-07,
"loss": 0.3981,
"step": 4975
},
{
"epoch": 1.0087088607594936,
"grad_norm": 0.2356027066707611,
"learning_rate": 1.790381007789783e-07,
"loss": 0.346,
"step": 4980
},
{
"epoch": 1.0097215189873419,
"grad_norm": 0.3801570236682892,
"learning_rate": 1.7564341204701826e-07,
"loss": 0.3948,
"step": 4985
},
{
"epoch": 1.0107341772151899,
"grad_norm": 0.11079708486795425,
"learning_rate": 1.7228049240897468e-07,
"loss": 0.4755,
"step": 4990
},
{
"epoch": 1.011746835443038,
"grad_norm": 0.12703923881053925,
"learning_rate": 1.6894936980076602e-07,
"loss": 0.5031,
"step": 4995
},
{
"epoch": 1.0127594936708861,
"grad_norm": 0.1714024543762207,
"learning_rate": 1.656500718941696e-07,
"loss": 0.5946,
"step": 5000
},
{
"epoch": 1.0137721518987342,
"grad_norm": 0.19446556270122528,
"learning_rate": 1.6238262609659458e-07,
"loss": 0.5168,
"step": 5005
},
{
"epoch": 1.0147848101265822,
"grad_norm": 0.1730039119720459,
"learning_rate": 1.5914705955085305e-07,
"loss": 0.4029,
"step": 5010
},
{
"epoch": 1.0157974683544304,
"grad_norm": 0.3286128640174866,
"learning_rate": 1.5594339913493547e-07,
"loss": 0.4113,
"step": 5015
},
{
"epoch": 1.0168101265822784,
"grad_norm": 0.19967059791088104,
"learning_rate": 1.5277167146178527e-07,
"loss": 0.3663,
"step": 5020
},
{
"epoch": 1.0178227848101267,
"grad_norm": 0.20584169030189514,
"learning_rate": 1.4963190287908024e-07,
"loss": 0.3871,
"step": 5025
},
{
"epoch": 1.0188354430379747,
"grad_norm": 0.28951090574264526,
"learning_rate": 1.465241194690128e-07,
"loss": 0.4402,
"step": 5030
},
{
"epoch": 1.0198481012658227,
"grad_norm": 0.3136148154735565,
"learning_rate": 1.4344834704807274e-07,
"loss": 0.4267,
"step": 5035
},
{
"epoch": 1.020860759493671,
"grad_norm": 0.08713238686323166,
"learning_rate": 1.4040461116683244e-07,
"loss": 0.4223,
"step": 5040
},
{
"epoch": 1.021873417721519,
"grad_norm": 0.12798455357551575,
"learning_rate": 1.373929371097371e-07,
"loss": 0.4602,
"step": 5045
},
{
"epoch": 1.022886075949367,
"grad_norm": 0.17034423351287842,
"learning_rate": 1.344133498948916e-07,
"loss": 0.466,
"step": 5050
}
],
"logging_steps": 5,
"max_steps": 5500,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 50,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 2.0487900686230118e+18,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}