qwen1.5-7b-chat-sa-v0.1 / trainer_state.json
sci-m-wang's picture
Upload 16 files
1148b15 verified
raw
history blame
93.1 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 4.9976558837318334,
"eval_steps": 500,
"global_step": 2665,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.009376465072667605,
"grad_norm": 1.043060064315796,
"learning_rate": 4.999956573574533e-05,
"loss": 1.2495,
"step": 5
},
{
"epoch": 0.01875293014533521,
"grad_norm": 1.0880073308944702,
"learning_rate": 4.999826295806815e-05,
"loss": 1.0913,
"step": 10
},
{
"epoch": 0.02812939521800281,
"grad_norm": 0.6108205914497375,
"learning_rate": 4.999609171222846e-05,
"loss": 0.9163,
"step": 15
},
{
"epoch": 0.03750586029067042,
"grad_norm": 0.5320619344711304,
"learning_rate": 4.99930520736578e-05,
"loss": 0.9189,
"step": 20
},
{
"epoch": 0.04688232536333802,
"grad_norm": 0.47798001766204834,
"learning_rate": 4.998914414795668e-05,
"loss": 0.8406,
"step": 25
},
{
"epoch": 0.05625879043600562,
"grad_norm": 0.4477296769618988,
"learning_rate": 4.99843680708909e-05,
"loss": 0.7963,
"step": 30
},
{
"epoch": 0.06563525550867323,
"grad_norm": 0.527661919593811,
"learning_rate": 4.997872400838682e-05,
"loss": 0.8229,
"step": 35
},
{
"epoch": 0.07501172058134084,
"grad_norm": 0.4180678725242615,
"learning_rate": 4.997221215652562e-05,
"loss": 0.824,
"step": 40
},
{
"epoch": 0.08438818565400844,
"grad_norm": 0.33111363649368286,
"learning_rate": 4.9964832741536444e-05,
"loss": 0.8236,
"step": 45
},
{
"epoch": 0.09376465072667604,
"grad_norm": 0.29713910818099976,
"learning_rate": 4.9956586019788584e-05,
"loss": 0.7696,
"step": 50
},
{
"epoch": 0.10314111579934365,
"grad_norm": 0.34906554222106934,
"learning_rate": 4.9947472277782584e-05,
"loss": 0.7585,
"step": 55
},
{
"epoch": 0.11251758087201125,
"grad_norm": 0.447453111410141,
"learning_rate": 4.993749183214021e-05,
"loss": 0.7549,
"step": 60
},
{
"epoch": 0.12189404594467886,
"grad_norm": 0.34558868408203125,
"learning_rate": 4.992664502959351e-05,
"loss": 0.7345,
"step": 65
},
{
"epoch": 0.13127051101734646,
"grad_norm": 0.3259447515010834,
"learning_rate": 4.991493224697281e-05,
"loss": 0.7273,
"step": 70
},
{
"epoch": 0.14064697609001406,
"grad_norm": 0.39191028475761414,
"learning_rate": 4.990235389119352e-05,
"loss": 0.7803,
"step": 75
},
{
"epoch": 0.15002344116268168,
"grad_norm": 0.3894421458244324,
"learning_rate": 4.9888910399242065e-05,
"loss": 0.6957,
"step": 80
},
{
"epoch": 0.15939990623534928,
"grad_norm": 0.40983864665031433,
"learning_rate": 4.987460223816067e-05,
"loss": 0.7076,
"step": 85
},
{
"epoch": 0.16877637130801687,
"grad_norm": 0.4191415011882782,
"learning_rate": 4.985942990503119e-05,
"loss": 0.7231,
"step": 90
},
{
"epoch": 0.1781528363806845,
"grad_norm": 0.41286033391952515,
"learning_rate": 4.984339392695777e-05,
"loss": 0.6879,
"step": 95
},
{
"epoch": 0.1875293014533521,
"grad_norm": 0.40984493494033813,
"learning_rate": 4.9826494861048576e-05,
"loss": 0.665,
"step": 100
},
{
"epoch": 0.19690576652601968,
"grad_norm": 0.45545095205307007,
"learning_rate": 4.980873329439644e-05,
"loss": 0.756,
"step": 105
},
{
"epoch": 0.2062822315986873,
"grad_norm": 0.3807056248188019,
"learning_rate": 4.979010984405842e-05,
"loss": 0.6963,
"step": 110
},
{
"epoch": 0.2156586966713549,
"grad_norm": 0.4047029912471771,
"learning_rate": 4.9770625157034436e-05,
"loss": 0.731,
"step": 115
},
{
"epoch": 0.2250351617440225,
"grad_norm": 0.4204094409942627,
"learning_rate": 4.975027991024473e-05,
"loss": 0.7397,
"step": 120
},
{
"epoch": 0.23441162681669012,
"grad_norm": 0.4436323344707489,
"learning_rate": 4.972907481050637e-05,
"loss": 0.7006,
"step": 125
},
{
"epoch": 0.2437880918893577,
"grad_norm": 0.5543025732040405,
"learning_rate": 4.970701059450872e-05,
"loss": 0.7231,
"step": 130
},
{
"epoch": 0.25316455696202533,
"grad_norm": 0.3950710594654083,
"learning_rate": 4.968408802878778e-05,
"loss": 0.7323,
"step": 135
},
{
"epoch": 0.26254102203469293,
"grad_norm": 0.4335172772407532,
"learning_rate": 4.9660307909699645e-05,
"loss": 0.6721,
"step": 140
},
{
"epoch": 0.2719174871073605,
"grad_norm": 0.37986621260643005,
"learning_rate": 4.963567106339276e-05,
"loss": 0.6722,
"step": 145
},
{
"epoch": 0.2812939521800281,
"grad_norm": 0.49829936027526855,
"learning_rate": 4.961017834577927e-05,
"loss": 0.7208,
"step": 150
},
{
"epoch": 0.2906704172526957,
"grad_norm": 0.466081827878952,
"learning_rate": 4.958383064250525e-05,
"loss": 0.6621,
"step": 155
},
{
"epoch": 0.30004688232536336,
"grad_norm": 0.5066991448402405,
"learning_rate": 4.955662886891995e-05,
"loss": 0.7114,
"step": 160
},
{
"epoch": 0.30942334739803096,
"grad_norm": 0.43918004631996155,
"learning_rate": 4.952857397004401e-05,
"loss": 0.7235,
"step": 165
},
{
"epoch": 0.31879981247069855,
"grad_norm": 0.4280238747596741,
"learning_rate": 4.949966692053663e-05,
"loss": 0.6993,
"step": 170
},
{
"epoch": 0.32817627754336615,
"grad_norm": 0.36493420600891113,
"learning_rate": 4.946990872466164e-05,
"loss": 0.7186,
"step": 175
},
{
"epoch": 0.33755274261603374,
"grad_norm": 0.4803766906261444,
"learning_rate": 4.943930041625272e-05,
"loss": 0.6867,
"step": 180
},
{
"epoch": 0.34692920768870134,
"grad_norm": 0.43526196479797363,
"learning_rate": 4.940784305867741e-05,
"loss": 0.6907,
"step": 185
},
{
"epoch": 0.356305672761369,
"grad_norm": 0.4661521315574646,
"learning_rate": 4.937553774480018e-05,
"loss": 0.6693,
"step": 190
},
{
"epoch": 0.3656821378340366,
"grad_norm": 0.5094730257987976,
"learning_rate": 4.934238559694448e-05,
"loss": 0.7322,
"step": 195
},
{
"epoch": 0.3750586029067042,
"grad_norm": 0.47885701060295105,
"learning_rate": 4.9308387766853725e-05,
"loss": 0.7351,
"step": 200
},
{
"epoch": 0.38443506797937177,
"grad_norm": 0.475460022687912,
"learning_rate": 4.92735454356513e-05,
"loss": 0.6996,
"step": 205
},
{
"epoch": 0.39381153305203936,
"grad_norm": 0.46198564767837524,
"learning_rate": 4.9237859813799535e-05,
"loss": 0.7124,
"step": 210
},
{
"epoch": 0.40318799812470696,
"grad_norm": 0.5559020042419434,
"learning_rate": 4.9201332141057623e-05,
"loss": 0.6734,
"step": 215
},
{
"epoch": 0.4125644631973746,
"grad_norm": 0.5716303586959839,
"learning_rate": 4.9163963686438575e-05,
"loss": 0.6914,
"step": 220
},
{
"epoch": 0.4219409282700422,
"grad_norm": 0.4577886760234833,
"learning_rate": 4.912575574816511e-05,
"loss": 0.6584,
"step": 225
},
{
"epoch": 0.4313173933427098,
"grad_norm": 0.565101683139801,
"learning_rate": 4.908670965362457e-05,
"loss": 0.6326,
"step": 230
},
{
"epoch": 0.4406938584153774,
"grad_norm": 0.4204961657524109,
"learning_rate": 4.9046826759322825e-05,
"loss": 0.6931,
"step": 235
},
{
"epoch": 0.450070323488045,
"grad_norm": 0.6228139996528625,
"learning_rate": 4.9006108450837095e-05,
"loss": 0.6769,
"step": 240
},
{
"epoch": 0.45944678856071264,
"grad_norm": 0.4805678725242615,
"learning_rate": 4.8964556142767845e-05,
"loss": 0.6957,
"step": 245
},
{
"epoch": 0.46882325363338023,
"grad_norm": 0.5143615007400513,
"learning_rate": 4.892217127868965e-05,
"loss": 0.6949,
"step": 250
},
{
"epoch": 0.4781997187060478,
"grad_norm": 0.5050625205039978,
"learning_rate": 4.8878955331101026e-05,
"loss": 0.7084,
"step": 255
},
{
"epoch": 0.4875761837787154,
"grad_norm": 0.5915749669075012,
"learning_rate": 4.8834909801373264e-05,
"loss": 0.6439,
"step": 260
},
{
"epoch": 0.496952648851383,
"grad_norm": 0.5936545133590698,
"learning_rate": 4.879003621969831e-05,
"loss": 0.654,
"step": 265
},
{
"epoch": 0.5063291139240507,
"grad_norm": 0.575564980506897,
"learning_rate": 4.874433614503554e-05,
"loss": 0.6513,
"step": 270
},
{
"epoch": 0.5157055789967182,
"grad_norm": 0.4369766414165497,
"learning_rate": 4.869781116505768e-05,
"loss": 0.6406,
"step": 275
},
{
"epoch": 0.5250820440693859,
"grad_norm": 0.5553178787231445,
"learning_rate": 4.8650462896095597e-05,
"loss": 0.6761,
"step": 280
},
{
"epoch": 0.5344585091420534,
"grad_norm": 0.4776481091976166,
"learning_rate": 4.860229298308213e-05,
"loss": 0.644,
"step": 285
},
{
"epoch": 0.543834974214721,
"grad_norm": 0.5263465046882629,
"learning_rate": 4.8553303099495e-05,
"loss": 0.6619,
"step": 290
},
{
"epoch": 0.5532114392873887,
"grad_norm": 0.5093105435371399,
"learning_rate": 4.8503494947298634e-05,
"loss": 0.6855,
"step": 295
},
{
"epoch": 0.5625879043600562,
"grad_norm": 0.504751980304718,
"learning_rate": 4.845287025688503e-05,
"loss": 0.6716,
"step": 300
},
{
"epoch": 0.5719643694327239,
"grad_norm": 0.48812317848205566,
"learning_rate": 4.8401430787013666e-05,
"loss": 0.6784,
"step": 305
},
{
"epoch": 0.5813408345053914,
"grad_norm": 0.4794219136238098,
"learning_rate": 4.8349178324750387e-05,
"loss": 0.6587,
"step": 310
},
{
"epoch": 0.5907172995780591,
"grad_norm": 0.5530505180358887,
"learning_rate": 4.8296114685405324e-05,
"loss": 0.6411,
"step": 315
},
{
"epoch": 0.6000937646507267,
"grad_norm": 0.5403686165809631,
"learning_rate": 4.824224171246981e-05,
"loss": 0.7191,
"step": 320
},
{
"epoch": 0.6094702297233943,
"grad_norm": 0.5672193765640259,
"learning_rate": 4.8187561277552374e-05,
"loss": 0.6679,
"step": 325
},
{
"epoch": 0.6188466947960619,
"grad_norm": 0.5467439889907837,
"learning_rate": 4.813207528031366e-05,
"loss": 0.627,
"step": 330
},
{
"epoch": 0.6282231598687295,
"grad_norm": 0.5209377408027649,
"learning_rate": 4.807578564840051e-05,
"loss": 0.7084,
"step": 335
},
{
"epoch": 0.6375996249413971,
"grad_norm": 0.5048089623451233,
"learning_rate": 4.801869433737891e-05,
"loss": 0.6555,
"step": 340
},
{
"epoch": 0.6469760900140648,
"grad_norm": 0.5744394659996033,
"learning_rate": 4.796080333066613e-05,
"loss": 0.6865,
"step": 345
},
{
"epoch": 0.6563525550867323,
"grad_norm": 0.5384623408317566,
"learning_rate": 4.790211463946174e-05,
"loss": 0.6968,
"step": 350
},
{
"epoch": 0.6657290201593999,
"grad_norm": 0.591351330280304,
"learning_rate": 4.784263030267781e-05,
"loss": 0.637,
"step": 355
},
{
"epoch": 0.6751054852320675,
"grad_norm": 0.5314144492149353,
"learning_rate": 4.7782352386868035e-05,
"loss": 0.6898,
"step": 360
},
{
"epoch": 0.6844819503047351,
"grad_norm": 0.5192554593086243,
"learning_rate": 4.7721282986155945e-05,
"loss": 0.6258,
"step": 365
},
{
"epoch": 0.6938584153774027,
"grad_norm": 0.625133216381073,
"learning_rate": 4.7659424222162165e-05,
"loss": 0.6418,
"step": 370
},
{
"epoch": 0.7032348804500703,
"grad_norm": 0.5590410232543945,
"learning_rate": 4.7596778243930694e-05,
"loss": 0.6957,
"step": 375
},
{
"epoch": 0.712611345522738,
"grad_norm": 0.55343097448349,
"learning_rate": 4.7533347227854265e-05,
"loss": 0.6813,
"step": 380
},
{
"epoch": 0.7219878105954055,
"grad_norm": 0.6247820258140564,
"learning_rate": 4.7469133377598695e-05,
"loss": 0.6565,
"step": 385
},
{
"epoch": 0.7313642756680732,
"grad_norm": 0.6145781874656677,
"learning_rate": 4.740413892402639e-05,
"loss": 0.6557,
"step": 390
},
{
"epoch": 0.7407407407407407,
"grad_norm": 0.6102336645126343,
"learning_rate": 4.7338366125118775e-05,
"loss": 0.6715,
"step": 395
},
{
"epoch": 0.7501172058134083,
"grad_norm": 0.6142327189445496,
"learning_rate": 4.727181726589789e-05,
"loss": 0.6929,
"step": 400
},
{
"epoch": 0.759493670886076,
"grad_norm": 0.6012890934944153,
"learning_rate": 4.7204494658346996e-05,
"loss": 0.6474,
"step": 405
},
{
"epoch": 0.7688701359587435,
"grad_norm": 0.588411808013916,
"learning_rate": 4.713640064133025e-05,
"loss": 0.6519,
"step": 410
},
{
"epoch": 0.7782466010314112,
"grad_norm": 0.5341905951499939,
"learning_rate": 4.706753758051145e-05,
"loss": 0.66,
"step": 415
},
{
"epoch": 0.7876230661040787,
"grad_norm": 0.5620007514953613,
"learning_rate": 4.699790786827188e-05,
"loss": 0.7041,
"step": 420
},
{
"epoch": 0.7969995311767464,
"grad_norm": 0.6584540605545044,
"learning_rate": 4.6927513923627124e-05,
"loss": 0.6676,
"step": 425
},
{
"epoch": 0.8063759962494139,
"grad_norm": 0.5691075325012207,
"learning_rate": 4.68563581921431e-05,
"loss": 0.6382,
"step": 430
},
{
"epoch": 0.8157524613220816,
"grad_norm": 0.6462568640708923,
"learning_rate": 4.6784443145851074e-05,
"loss": 0.5761,
"step": 435
},
{
"epoch": 0.8251289263947492,
"grad_norm": 0.5838071703910828,
"learning_rate": 4.671177128316176e-05,
"loss": 0.6149,
"step": 440
},
{
"epoch": 0.8345053914674168,
"grad_norm": 0.5983990430831909,
"learning_rate": 4.663834512877853e-05,
"loss": 0.661,
"step": 445
},
{
"epoch": 0.8438818565400844,
"grad_norm": 0.6623389720916748,
"learning_rate": 4.6564167233609736e-05,
"loss": 0.6477,
"step": 450
},
{
"epoch": 0.853258321612752,
"grad_norm": 0.5716333389282227,
"learning_rate": 4.648924017468003e-05,
"loss": 0.6267,
"step": 455
},
{
"epoch": 0.8626347866854196,
"grad_norm": 0.6834291815757751,
"learning_rate": 4.6413566555040896e-05,
"loss": 0.6776,
"step": 460
},
{
"epoch": 0.8720112517580872,
"grad_norm": 0.6510602235794067,
"learning_rate": 4.633714900368018e-05,
"loss": 0.645,
"step": 465
},
{
"epoch": 0.8813877168307548,
"grad_norm": 0.5720119476318359,
"learning_rate": 4.625999017543075e-05,
"loss": 0.6486,
"step": 470
},
{
"epoch": 0.8907641819034224,
"grad_norm": 0.6245366930961609,
"learning_rate": 4.618209275087829e-05,
"loss": 0.6784,
"step": 475
},
{
"epoch": 0.90014064697609,
"grad_norm": 0.6153048276901245,
"learning_rate": 4.610345943626817e-05,
"loss": 0.6595,
"step": 480
},
{
"epoch": 0.9095171120487576,
"grad_norm": 0.573347270488739,
"learning_rate": 4.602409296341141e-05,
"loss": 0.6403,
"step": 485
},
{
"epoch": 0.9188935771214253,
"grad_norm": 0.5834039449691772,
"learning_rate": 4.5943996089589775e-05,
"loss": 0.6334,
"step": 490
},
{
"epoch": 0.9282700421940928,
"grad_norm": 0.6009634137153625,
"learning_rate": 4.586317159746001e-05,
"loss": 0.7385,
"step": 495
},
{
"epoch": 0.9376465072667605,
"grad_norm": 0.6226871013641357,
"learning_rate": 4.5781622294957136e-05,
"loss": 0.6463,
"step": 500
},
{
"epoch": 0.947022972339428,
"grad_norm": 0.5742005109786987,
"learning_rate": 4.569935101519692e-05,
"loss": 0.6693,
"step": 505
},
{
"epoch": 0.9563994374120957,
"grad_norm": 0.5679554343223572,
"learning_rate": 4.561636061637745e-05,
"loss": 0.7034,
"step": 510
},
{
"epoch": 0.9657759024847632,
"grad_norm": 0.5975098013877869,
"learning_rate": 4.553265398167981e-05,
"loss": 0.652,
"step": 515
},
{
"epoch": 0.9751523675574308,
"grad_norm": 0.5988393425941467,
"learning_rate": 4.5448234019167945e-05,
"loss": 0.6399,
"step": 520
},
{
"epoch": 0.9845288326300985,
"grad_norm": 0.6314676403999329,
"learning_rate": 4.536310366168763e-05,
"loss": 0.6426,
"step": 525
},
{
"epoch": 0.993905297702766,
"grad_norm": 0.5880908370018005,
"learning_rate": 4.5277265866764565e-05,
"loss": 0.66,
"step": 530
},
{
"epoch": 1.0032817627754336,
"grad_norm": 0.6495065093040466,
"learning_rate": 4.519072361650163e-05,
"loss": 0.6243,
"step": 535
},
{
"epoch": 1.0126582278481013,
"grad_norm": 0.661907970905304,
"learning_rate": 4.5103479917475286e-05,
"loss": 0.6001,
"step": 540
},
{
"epoch": 1.0220346929207689,
"grad_norm": 0.7230183482170105,
"learning_rate": 4.501553780063113e-05,
"loss": 0.6071,
"step": 545
},
{
"epoch": 1.0314111579934364,
"grad_norm": 0.5441235303878784,
"learning_rate": 4.4926900321178595e-05,
"loss": 0.6617,
"step": 550
},
{
"epoch": 1.0407876230661042,
"grad_norm": 0.6352370381355286,
"learning_rate": 4.483757055848479e-05,
"loss": 0.6291,
"step": 555
},
{
"epoch": 1.0501640881387717,
"grad_norm": 0.6081958413124084,
"learning_rate": 4.4747551615967534e-05,
"loss": 0.6011,
"step": 560
},
{
"epoch": 1.0595405532114393,
"grad_norm": 0.662145733833313,
"learning_rate": 4.4656846620987557e-05,
"loss": 0.6631,
"step": 565
},
{
"epoch": 1.0689170182841068,
"grad_norm": 0.6905146241188049,
"learning_rate": 4.4565458724739825e-05,
"loss": 0.6544,
"step": 570
},
{
"epoch": 1.0782934833567746,
"grad_norm": 0.6696512699127197,
"learning_rate": 4.447339110214405e-05,
"loss": 0.6332,
"step": 575
},
{
"epoch": 1.087669948429442,
"grad_norm": 0.6640881896018982,
"learning_rate": 4.438064695173446e-05,
"loss": 0.6426,
"step": 580
},
{
"epoch": 1.0970464135021096,
"grad_norm": 0.7038636803627014,
"learning_rate": 4.428722949554857e-05,
"loss": 0.6362,
"step": 585
},
{
"epoch": 1.1064228785747774,
"grad_norm": 0.6065195202827454,
"learning_rate": 4.419314197901537e-05,
"loss": 0.6136,
"step": 590
},
{
"epoch": 1.115799343647445,
"grad_norm": 0.6090584397315979,
"learning_rate": 4.4098387670842466e-05,
"loss": 0.636,
"step": 595
},
{
"epoch": 1.1251758087201125,
"grad_norm": 0.6664246320724487,
"learning_rate": 4.400296986290258e-05,
"loss": 0.6253,
"step": 600
},
{
"epoch": 1.1345522737927802,
"grad_norm": 0.7402456998825073,
"learning_rate": 4.390689187011917e-05,
"loss": 0.6351,
"step": 605
},
{
"epoch": 1.1439287388654478,
"grad_norm": 0.6655504107475281,
"learning_rate": 4.3810157030351276e-05,
"loss": 0.5974,
"step": 610
},
{
"epoch": 1.1533052039381153,
"grad_norm": 0.6735255122184753,
"learning_rate": 4.371276870427753e-05,
"loss": 0.6023,
"step": 615
},
{
"epoch": 1.1626816690107828,
"grad_norm": 0.6369088888168335,
"learning_rate": 4.3614730275279457e-05,
"loss": 0.6096,
"step": 620
},
{
"epoch": 1.1720581340834506,
"grad_norm": 0.7166373133659363,
"learning_rate": 4.351604514932387e-05,
"loss": 0.6156,
"step": 625
},
{
"epoch": 1.1814345991561181,
"grad_norm": 0.6789242029190063,
"learning_rate": 4.341671675484459e-05,
"loss": 0.6406,
"step": 630
},
{
"epoch": 1.1908110642287857,
"grad_norm": 0.5939568877220154,
"learning_rate": 4.331674854262331e-05,
"loss": 0.6224,
"step": 635
},
{
"epoch": 1.2001875293014534,
"grad_norm": 0.6717362403869629,
"learning_rate": 4.321614398566972e-05,
"loss": 0.6188,
"step": 640
},
{
"epoch": 1.209563994374121,
"grad_norm": 0.668659508228302,
"learning_rate": 4.3114906579100853e-05,
"loss": 0.6088,
"step": 645
},
{
"epoch": 1.2189404594467885,
"grad_norm": 0.6963931322097778,
"learning_rate": 4.301303984001967e-05,
"loss": 0.657,
"step": 650
},
{
"epoch": 1.228316924519456,
"grad_norm": 0.7464931607246399,
"learning_rate": 4.291054730739286e-05,
"loss": 0.6417,
"step": 655
},
{
"epoch": 1.2376933895921238,
"grad_norm": 0.6999815702438354,
"learning_rate": 4.2807432541927865e-05,
"loss": 0.6402,
"step": 660
},
{
"epoch": 1.2470698546647914,
"grad_norm": 0.7735899686813354,
"learning_rate": 4.2703699125949245e-05,
"loss": 0.6393,
"step": 665
},
{
"epoch": 1.256446319737459,
"grad_norm": 0.7408654093742371,
"learning_rate": 4.259935066327415e-05,
"loss": 0.7175,
"step": 670
},
{
"epoch": 1.2658227848101267,
"grad_norm": 0.5848163962364197,
"learning_rate": 4.2494390779087187e-05,
"loss": 0.68,
"step": 675
},
{
"epoch": 1.2751992498827942,
"grad_norm": 0.6354319453239441,
"learning_rate": 4.238882311981441e-05,
"loss": 0.5966,
"step": 680
},
{
"epoch": 1.2845757149554617,
"grad_norm": 0.6874629259109497,
"learning_rate": 4.228265135299669e-05,
"loss": 0.6432,
"step": 685
},
{
"epoch": 1.2939521800281293,
"grad_norm": 0.6156244874000549,
"learning_rate": 4.2175879167162304e-05,
"loss": 0.6688,
"step": 690
},
{
"epoch": 1.303328645100797,
"grad_norm": 0.6476886868476868,
"learning_rate": 4.206851027169871e-05,
"loss": 0.6407,
"step": 695
},
{
"epoch": 1.3127051101734646,
"grad_norm": 0.6186336874961853,
"learning_rate": 4.196054839672382e-05,
"loss": 0.6084,
"step": 700
},
{
"epoch": 1.3220815752461323,
"grad_norm": 0.566964328289032,
"learning_rate": 4.1851997292956255e-05,
"loss": 0.6407,
"step": 705
},
{
"epoch": 1.3314580403187999,
"grad_norm": 0.6484958529472351,
"learning_rate": 4.174286073158516e-05,
"loss": 0.643,
"step": 710
},
{
"epoch": 1.3408345053914674,
"grad_norm": 0.6172962784767151,
"learning_rate": 4.163314250413913e-05,
"loss": 0.5899,
"step": 715
},
{
"epoch": 1.350210970464135,
"grad_norm": 0.6747507452964783,
"learning_rate": 4.152284642235452e-05,
"loss": 0.6299,
"step": 720
},
{
"epoch": 1.3595874355368025,
"grad_norm": 0.7099136710166931,
"learning_rate": 4.141197631804298e-05,
"loss": 0.6462,
"step": 725
},
{
"epoch": 1.3689639006094703,
"grad_norm": 0.6501842141151428,
"learning_rate": 4.1300536042958354e-05,
"loss": 0.6558,
"step": 730
},
{
"epoch": 1.3783403656821378,
"grad_norm": 0.6962847113609314,
"learning_rate": 4.118852946866291e-05,
"loss": 0.6316,
"step": 735
},
{
"epoch": 1.3877168307548056,
"grad_norm": 0.7128638029098511,
"learning_rate": 4.107596048639274e-05,
"loss": 0.6143,
"step": 740
},
{
"epoch": 1.397093295827473,
"grad_norm": 0.6980453729629517,
"learning_rate": 4.0962833006922675e-05,
"loss": 0.6358,
"step": 745
},
{
"epoch": 1.4064697609001406,
"grad_norm": 0.6058886051177979,
"learning_rate": 4.0849150960430356e-05,
"loss": 0.6276,
"step": 750
},
{
"epoch": 1.4158462259728082,
"grad_norm": 0.6673750281333923,
"learning_rate": 4.0734918296359716e-05,
"loss": 0.6238,
"step": 755
},
{
"epoch": 1.4252226910454757,
"grad_norm": 0.6338385343551636,
"learning_rate": 4.0620138983283785e-05,
"loss": 0.6775,
"step": 760
},
{
"epoch": 1.4345991561181435,
"grad_norm": 0.6448456645011902,
"learning_rate": 4.050481700876677e-05,
"loss": 0.633,
"step": 765
},
{
"epoch": 1.443975621190811,
"grad_norm": 0.7474460005760193,
"learning_rate": 4.038895637922559e-05,
"loss": 0.6307,
"step": 770
},
{
"epoch": 1.4533520862634788,
"grad_norm": 0.7345675230026245,
"learning_rate": 4.027256111979063e-05,
"loss": 0.5959,
"step": 775
},
{
"epoch": 1.4627285513361463,
"grad_norm": 0.6653993725776672,
"learning_rate": 4.015563527416595e-05,
"loss": 0.6334,
"step": 780
},
{
"epoch": 1.4721050164088139,
"grad_norm": 0.652690052986145,
"learning_rate": 4.003818290448876e-05,
"loss": 0.644,
"step": 785
},
{
"epoch": 1.4814814814814814,
"grad_norm": 0.624614953994751,
"learning_rate": 3.992020809118832e-05,
"loss": 0.7045,
"step": 790
},
{
"epoch": 1.4908579465541492,
"grad_norm": 0.6916288137435913,
"learning_rate": 3.980171493284418e-05,
"loss": 0.6335,
"step": 795
},
{
"epoch": 1.5002344116268167,
"grad_norm": 0.6975934505462646,
"learning_rate": 3.9682707546043785e-05,
"loss": 0.6705,
"step": 800
},
{
"epoch": 1.5096108766994845,
"grad_norm": 0.7178190350532532,
"learning_rate": 3.9563190065239474e-05,
"loss": 0.6219,
"step": 805
},
{
"epoch": 1.518987341772152,
"grad_norm": 0.711218535900116,
"learning_rate": 3.9443166642604814e-05,
"loss": 0.6101,
"step": 810
},
{
"epoch": 1.5283638068448195,
"grad_norm": 0.5779418349266052,
"learning_rate": 3.932264144789038e-05,
"loss": 0.593,
"step": 815
},
{
"epoch": 1.537740271917487,
"grad_norm": 0.6230177879333496,
"learning_rate": 3.920161866827889e-05,
"loss": 0.6577,
"step": 820
},
{
"epoch": 1.5471167369901546,
"grad_norm": 0.7051926851272583,
"learning_rate": 3.908010250823972e-05,
"loss": 0.6369,
"step": 825
},
{
"epoch": 1.5564932020628222,
"grad_norm": 0.7006888389587402,
"learning_rate": 3.895809718938283e-05,
"loss": 0.6459,
"step": 830
},
{
"epoch": 1.56586966713549,
"grad_norm": 0.6218993067741394,
"learning_rate": 3.883560695031213e-05,
"loss": 0.5747,
"step": 835
},
{
"epoch": 1.5752461322081577,
"grad_norm": 0.656377911567688,
"learning_rate": 3.871263604647822e-05,
"loss": 0.6334,
"step": 840
},
{
"epoch": 1.5846225972808252,
"grad_norm": 0.6752548813819885,
"learning_rate": 3.858918875003053e-05,
"loss": 0.5922,
"step": 845
},
{
"epoch": 1.5939990623534928,
"grad_norm": 0.7023399472236633,
"learning_rate": 3.846526934966891e-05,
"loss": 0.6324,
"step": 850
},
{
"epoch": 1.6033755274261603,
"grad_norm": 0.7364317774772644,
"learning_rate": 3.834088215049464e-05,
"loss": 0.6397,
"step": 855
},
{
"epoch": 1.6127519924988278,
"grad_norm": 0.7179546356201172,
"learning_rate": 3.821603147386088e-05,
"loss": 0.6344,
"step": 860
},
{
"epoch": 1.6221284575714956,
"grad_norm": 0.6943954825401306,
"learning_rate": 3.80907216572225e-05,
"loss": 0.5903,
"step": 865
},
{
"epoch": 1.6315049226441631,
"grad_norm": 0.7795728445053101,
"learning_rate": 3.796495705398544e-05,
"loss": 0.6362,
"step": 870
},
{
"epoch": 1.640881387716831,
"grad_norm": 0.6673624515533447,
"learning_rate": 3.783874203335542e-05,
"loss": 0.6037,
"step": 875
},
{
"epoch": 1.6502578527894984,
"grad_norm": 0.7244172096252441,
"learning_rate": 3.77120809801862e-05,
"loss": 0.6131,
"step": 880
},
{
"epoch": 1.659634317862166,
"grad_norm": 0.587921679019928,
"learning_rate": 3.758497829482721e-05,
"loss": 0.6504,
"step": 885
},
{
"epoch": 1.6690107829348335,
"grad_norm": 0.6311667561531067,
"learning_rate": 3.7457438392970686e-05,
"loss": 0.5784,
"step": 890
},
{
"epoch": 1.678387248007501,
"grad_norm": 0.7019160985946655,
"learning_rate": 3.732946570549825e-05,
"loss": 0.6192,
"step": 895
},
{
"epoch": 1.6877637130801688,
"grad_norm": 0.7296169996261597,
"learning_rate": 3.720106467832701e-05,
"loss": 0.6849,
"step": 900
},
{
"epoch": 1.6971401781528364,
"grad_norm": 0.6518985033035278,
"learning_rate": 3.707223977225507e-05,
"loss": 0.6645,
"step": 905
},
{
"epoch": 1.7065166432255041,
"grad_norm": 0.629097044467926,
"learning_rate": 3.694299546280657e-05,
"loss": 0.6728,
"step": 910
},
{
"epoch": 1.7158931082981717,
"grad_norm": 0.7138874530792236,
"learning_rate": 3.681333624007623e-05,
"loss": 0.6476,
"step": 915
},
{
"epoch": 1.7252695733708392,
"grad_norm": 0.69582200050354,
"learning_rate": 3.6683266608573286e-05,
"loss": 0.7412,
"step": 920
},
{
"epoch": 1.7346460384435067,
"grad_norm": 0.7621073722839355,
"learning_rate": 3.6552791087065075e-05,
"loss": 0.6201,
"step": 925
},
{
"epoch": 1.7440225035161743,
"grad_norm": 0.6894713044166565,
"learning_rate": 3.642191420842e-05,
"loss": 0.6296,
"step": 930
},
{
"epoch": 1.753398968588842,
"grad_norm": 0.7036609649658203,
"learning_rate": 3.6290640519450074e-05,
"loss": 0.6199,
"step": 935
},
{
"epoch": 1.7627754336615096,
"grad_norm": 0.7221242189407349,
"learning_rate": 3.6158974580752954e-05,
"loss": 0.6726,
"step": 940
},
{
"epoch": 1.7721518987341773,
"grad_norm": 0.7487791180610657,
"learning_rate": 3.60269209665535e-05,
"loss": 0.6077,
"step": 945
},
{
"epoch": 1.7815283638068449,
"grad_norm": 0.7120531797409058,
"learning_rate": 3.589448426454486e-05,
"loss": 0.6546,
"step": 950
},
{
"epoch": 1.7909048288795124,
"grad_norm": 0.7374992966651917,
"learning_rate": 3.5761669075729084e-05,
"loss": 0.6134,
"step": 955
},
{
"epoch": 1.80028129395218,
"grad_norm": 0.7404727339744568,
"learning_rate": 3.562848001425729e-05,
"loss": 0.6833,
"step": 960
},
{
"epoch": 1.8096577590248475,
"grad_norm": 0.6800241470336914,
"learning_rate": 3.549492170726937e-05,
"loss": 0.6008,
"step": 965
},
{
"epoch": 1.8190342240975153,
"grad_norm": 0.736266016960144,
"learning_rate": 3.53609987947332e-05,
"loss": 0.6444,
"step": 970
},
{
"epoch": 1.8284106891701828,
"grad_norm": 0.6644765734672546,
"learning_rate": 3.5226715929283506e-05,
"loss": 0.596,
"step": 975
},
{
"epoch": 1.8377871542428506,
"grad_norm": 0.6540014147758484,
"learning_rate": 3.509207777606013e-05,
"loss": 0.5834,
"step": 980
},
{
"epoch": 1.847163619315518,
"grad_norm": 0.7583248615264893,
"learning_rate": 3.49570890125461e-05,
"loss": 0.6094,
"step": 985
},
{
"epoch": 1.8565400843881856,
"grad_norm": 0.6199962496757507,
"learning_rate": 3.482175432840495e-05,
"loss": 0.5964,
"step": 990
},
{
"epoch": 1.8659165494608532,
"grad_norm": 0.7292067408561707,
"learning_rate": 3.468607842531797e-05,
"loss": 0.618,
"step": 995
},
{
"epoch": 1.8752930145335207,
"grad_norm": 0.7406108379364014,
"learning_rate": 3.455006601682075e-05,
"loss": 0.596,
"step": 1000
},
{
"epoch": 1.8846694796061885,
"grad_norm": 0.6806167364120483,
"learning_rate": 3.441372182813946e-05,
"loss": 0.5774,
"step": 1005
},
{
"epoch": 1.8940459446788562,
"grad_norm": 0.6825852990150452,
"learning_rate": 3.427705059602671e-05,
"loss": 0.5982,
"step": 1010
},
{
"epoch": 1.9034224097515238,
"grad_norm": 0.7001626491546631,
"learning_rate": 3.414005706859693e-05,
"loss": 0.6422,
"step": 1015
},
{
"epoch": 1.9127988748241913,
"grad_norm": 0.6993778944015503,
"learning_rate": 3.400274600516152e-05,
"loss": 0.6312,
"step": 1020
},
{
"epoch": 1.9221753398968588,
"grad_norm": 0.6194260716438293,
"learning_rate": 3.386512217606339e-05,
"loss": 0.5998,
"step": 1025
},
{
"epoch": 1.9315518049695264,
"grad_norm": 0.7205774188041687,
"learning_rate": 3.372719036251132e-05,
"loss": 0.6103,
"step": 1030
},
{
"epoch": 1.9409282700421941,
"grad_norm": 0.6409869194030762,
"learning_rate": 3.3588955356413795e-05,
"loss": 0.6074,
"step": 1035
},
{
"epoch": 1.9503047351148617,
"grad_norm": 0.7475089430809021,
"learning_rate": 3.3450421960212566e-05,
"loss": 0.6142,
"step": 1040
},
{
"epoch": 1.9596812001875294,
"grad_norm": 0.6921798586845398,
"learning_rate": 3.3311594986715814e-05,
"loss": 0.6366,
"step": 1045
},
{
"epoch": 1.969057665260197,
"grad_norm": 0.7049899697303772,
"learning_rate": 3.317247925893089e-05,
"loss": 0.5871,
"step": 1050
},
{
"epoch": 1.9784341303328645,
"grad_norm": 0.7039067149162292,
"learning_rate": 3.3033079609896834e-05,
"loss": 0.601,
"step": 1055
},
{
"epoch": 1.987810595405532,
"grad_norm": 0.7747904062271118,
"learning_rate": 3.289340088251642e-05,
"loss": 0.5808,
"step": 1060
},
{
"epoch": 1.9971870604781996,
"grad_norm": 0.6818182468414307,
"learning_rate": 3.275344792938791e-05,
"loss": 0.6152,
"step": 1065
},
{
"epoch": 2.006563525550867,
"grad_norm": 0.6767409443855286,
"learning_rate": 3.2613225612636525e-05,
"loss": 0.5649,
"step": 1070
},
{
"epoch": 2.015939990623535,
"grad_norm": 0.6562533378601074,
"learning_rate": 3.247273880374542e-05,
"loss": 0.5623,
"step": 1075
},
{
"epoch": 2.0253164556962027,
"grad_norm": 0.7730954885482788,
"learning_rate": 3.2331992383386566e-05,
"loss": 0.6362,
"step": 1080
},
{
"epoch": 2.03469292076887,
"grad_norm": 0.6370785236358643,
"learning_rate": 3.21909912412511e-05,
"loss": 0.606,
"step": 1085
},
{
"epoch": 2.0440693858415377,
"grad_norm": 0.6823928952217102,
"learning_rate": 3.2049740275879493e-05,
"loss": 0.6176,
"step": 1090
},
{
"epoch": 2.0534458509142053,
"grad_norm": 0.705295205116272,
"learning_rate": 3.190824439449137e-05,
"loss": 0.6362,
"step": 1095
},
{
"epoch": 2.062822315986873,
"grad_norm": 0.7424062490463257,
"learning_rate": 3.176650851281499e-05,
"loss": 0.5969,
"step": 1100
},
{
"epoch": 2.0721987810595404,
"grad_norm": 0.7166156768798828,
"learning_rate": 3.162453755491655e-05,
"loss": 0.5587,
"step": 1105
},
{
"epoch": 2.0815752461322083,
"grad_norm": 0.6642809510231018,
"learning_rate": 3.1482336453028986e-05,
"loss": 0.5828,
"step": 1110
},
{
"epoch": 2.090951711204876,
"grad_norm": 0.7335144877433777,
"learning_rate": 3.133991014738076e-05,
"loss": 0.601,
"step": 1115
},
{
"epoch": 2.1003281762775434,
"grad_norm": 0.6921694874763489,
"learning_rate": 3.1197263586024155e-05,
"loss": 0.6003,
"step": 1120
},
{
"epoch": 2.109704641350211,
"grad_norm": 0.8270456194877625,
"learning_rate": 3.105440172466337e-05,
"loss": 0.6198,
"step": 1125
},
{
"epoch": 2.1190811064228785,
"grad_norm": 0.8127502799034119,
"learning_rate": 3.09113295264824e-05,
"loss": 0.6133,
"step": 1130
},
{
"epoch": 2.128457571495546,
"grad_norm": 0.7760539650917053,
"learning_rate": 3.076805196197255e-05,
"loss": 0.6002,
"step": 1135
},
{
"epoch": 2.1378340365682136,
"grad_norm": 0.8686167597770691,
"learning_rate": 3.0624574008759805e-05,
"loss": 0.6979,
"step": 1140
},
{
"epoch": 2.1472105016408816,
"grad_norm": 0.8145749568939209,
"learning_rate": 3.0480900651431876e-05,
"loss": 0.5819,
"step": 1145
},
{
"epoch": 2.156586966713549,
"grad_norm": 0.7867200970649719,
"learning_rate": 3.0337036881365045e-05,
"loss": 0.5804,
"step": 1150
},
{
"epoch": 2.1659634317862166,
"grad_norm": 0.7709445953369141,
"learning_rate": 3.0192987696550746e-05,
"loss": 0.5913,
"step": 1155
},
{
"epoch": 2.175339896858884,
"grad_norm": 0.7718144059181213,
"learning_rate": 3.0048758101421914e-05,
"loss": 0.5993,
"step": 1160
},
{
"epoch": 2.1847163619315517,
"grad_norm": 0.7650019526481628,
"learning_rate": 2.9904353106679178e-05,
"loss": 0.6061,
"step": 1165
},
{
"epoch": 2.1940928270042193,
"grad_norm": 0.6802553534507751,
"learning_rate": 2.975977772911671e-05,
"loss": 0.663,
"step": 1170
},
{
"epoch": 2.2034692920768872,
"grad_norm": 0.8476700186729431,
"learning_rate": 2.9615036991448015e-05,
"loss": 0.6158,
"step": 1175
},
{
"epoch": 2.212845757149555,
"grad_norm": 0.7697311639785767,
"learning_rate": 2.947013592213137e-05,
"loss": 0.6348,
"step": 1180
},
{
"epoch": 2.2222222222222223,
"grad_norm": 0.7111309170722961,
"learning_rate": 2.9325079555195163e-05,
"loss": 0.6263,
"step": 1185
},
{
"epoch": 2.23159868729489,
"grad_norm": 0.8301457166671753,
"learning_rate": 2.9179872930063e-05,
"loss": 0.5918,
"step": 1190
},
{
"epoch": 2.2409751523675574,
"grad_norm": 0.7816576957702637,
"learning_rate": 2.9034521091378635e-05,
"loss": 0.5724,
"step": 1195
},
{
"epoch": 2.250351617440225,
"grad_norm": 0.798967182636261,
"learning_rate": 2.8889029088830686e-05,
"loss": 0.6142,
"step": 1200
},
{
"epoch": 2.2597280825128925,
"grad_norm": 0.7511455416679382,
"learning_rate": 2.8743401976977257e-05,
"loss": 0.6048,
"step": 1205
},
{
"epoch": 2.2691045475855605,
"grad_norm": 0.6589820981025696,
"learning_rate": 2.8597644815070263e-05,
"loss": 0.6531,
"step": 1210
},
{
"epoch": 2.278481012658228,
"grad_norm": 0.755199670791626,
"learning_rate": 2.845176266687974e-05,
"loss": 0.5875,
"step": 1215
},
{
"epoch": 2.2878574777308955,
"grad_norm": 0.8146589398384094,
"learning_rate": 2.8305760600517862e-05,
"loss": 0.5762,
"step": 1220
},
{
"epoch": 2.297233942803563,
"grad_norm": 0.7374768853187561,
"learning_rate": 2.815964368826292e-05,
"loss": 0.6074,
"step": 1225
},
{
"epoch": 2.3066104078762306,
"grad_norm": 0.8007232546806335,
"learning_rate": 2.8013417006383076e-05,
"loss": 0.5865,
"step": 1230
},
{
"epoch": 2.315986872948898,
"grad_norm": 0.7976067662239075,
"learning_rate": 2.7867085634960016e-05,
"loss": 0.6015,
"step": 1235
},
{
"epoch": 2.3253633380215657,
"grad_norm": 0.7782787084579468,
"learning_rate": 2.772065465771244e-05,
"loss": 0.5973,
"step": 1240
},
{
"epoch": 2.3347398030942337,
"grad_norm": 0.8809172511100769,
"learning_rate": 2.7574129161819495e-05,
"loss": 0.6309,
"step": 1245
},
{
"epoch": 2.344116268166901,
"grad_norm": 0.7300681471824646,
"learning_rate": 2.7427514237744e-05,
"loss": 0.6291,
"step": 1250
},
{
"epoch": 2.3534927332395688,
"grad_norm": 0.7953155636787415,
"learning_rate": 2.7280814979055612e-05,
"loss": 0.6297,
"step": 1255
},
{
"epoch": 2.3628691983122363,
"grad_norm": 0.6899072527885437,
"learning_rate": 2.713403648225388e-05,
"loss": 0.5846,
"step": 1260
},
{
"epoch": 2.372245663384904,
"grad_norm": 0.7751972079277039,
"learning_rate": 2.698718384659114e-05,
"loss": 0.6009,
"step": 1265
},
{
"epoch": 2.3816221284575714,
"grad_norm": 0.831641674041748,
"learning_rate": 2.684026217389544e-05,
"loss": 0.6807,
"step": 1270
},
{
"epoch": 2.390998593530239,
"grad_norm": 0.7829850912094116,
"learning_rate": 2.6693276568393245e-05,
"loss": 0.5976,
"step": 1275
},
{
"epoch": 2.400375058602907,
"grad_norm": 0.7054941058158875,
"learning_rate": 2.6546232136532083e-05,
"loss": 0.6617,
"step": 1280
},
{
"epoch": 2.4097515236755744,
"grad_norm": 0.7278191447257996,
"learning_rate": 2.639913398680322e-05,
"loss": 0.6327,
"step": 1285
},
{
"epoch": 2.419127988748242,
"grad_norm": 0.6977508068084717,
"learning_rate": 2.6251987229564123e-05,
"loss": 0.6101,
"step": 1290
},
{
"epoch": 2.4285044538209095,
"grad_norm": 0.7802432775497437,
"learning_rate": 2.610479697686093e-05,
"loss": 0.5672,
"step": 1295
},
{
"epoch": 2.437880918893577,
"grad_norm": 0.8246661424636841,
"learning_rate": 2.595756834225089e-05,
"loss": 0.6172,
"step": 1300
},
{
"epoch": 2.4472573839662446,
"grad_norm": 0.8327619433403015,
"learning_rate": 2.5810306440624644e-05,
"loss": 0.6848,
"step": 1305
},
{
"epoch": 2.456633849038912,
"grad_norm": 0.8517838716506958,
"learning_rate": 2.566301638802861e-05,
"loss": 0.6313,
"step": 1310
},
{
"epoch": 2.46601031411158,
"grad_norm": 0.8053081035614014,
"learning_rate": 2.551570330148716e-05,
"loss": 0.5976,
"step": 1315
},
{
"epoch": 2.4753867791842477,
"grad_norm": 0.821284830570221,
"learning_rate": 2.5368372298824922e-05,
"loss": 0.5623,
"step": 1320
},
{
"epoch": 2.484763244256915,
"grad_norm": 0.8210687041282654,
"learning_rate": 2.5221028498488947e-05,
"loss": 0.6274,
"step": 1325
},
{
"epoch": 2.4941397093295827,
"grad_norm": 0.8597337603569031,
"learning_rate": 2.507367701937087e-05,
"loss": 0.5823,
"step": 1330
},
{
"epoch": 2.5035161744022503,
"grad_norm": 0.7325977087020874,
"learning_rate": 2.492632298062913e-05,
"loss": 0.59,
"step": 1335
},
{
"epoch": 2.512892639474918,
"grad_norm": 0.8166496157646179,
"learning_rate": 2.4778971501511063e-05,
"loss": 0.6277,
"step": 1340
},
{
"epoch": 2.5222691045475853,
"grad_norm": 0.8390632271766663,
"learning_rate": 2.4631627701175084e-05,
"loss": 0.5809,
"step": 1345
},
{
"epoch": 2.5316455696202533,
"grad_norm": 0.8438106775283813,
"learning_rate": 2.448429669851285e-05,
"loss": 0.5756,
"step": 1350
},
{
"epoch": 2.541022034692921,
"grad_norm": 0.8104760646820068,
"learning_rate": 2.43369836119714e-05,
"loss": 0.623,
"step": 1355
},
{
"epoch": 2.5503984997655884,
"grad_norm": 0.7337835431098938,
"learning_rate": 2.4189693559375365e-05,
"loss": 0.6328,
"step": 1360
},
{
"epoch": 2.559774964838256,
"grad_norm": 0.8129787445068359,
"learning_rate": 2.4042431657749117e-05,
"loss": 0.5745,
"step": 1365
},
{
"epoch": 2.5691514299109235,
"grad_norm": 0.8080322742462158,
"learning_rate": 2.3895203023139073e-05,
"loss": 0.5681,
"step": 1370
},
{
"epoch": 2.578527894983591,
"grad_norm": 0.7477867603302002,
"learning_rate": 2.3748012770435883e-05,
"loss": 0.6243,
"step": 1375
},
{
"epoch": 2.5879043600562586,
"grad_norm": 0.7538096308708191,
"learning_rate": 2.3600866013196787e-05,
"loss": 0.6017,
"step": 1380
},
{
"epoch": 2.5972808251289266,
"grad_norm": 0.7187966704368591,
"learning_rate": 2.3453767863467923e-05,
"loss": 0.59,
"step": 1385
},
{
"epoch": 2.606657290201594,
"grad_norm": 0.7694785594940186,
"learning_rate": 2.3306723431606758e-05,
"loss": 0.5715,
"step": 1390
},
{
"epoch": 2.6160337552742616,
"grad_norm": 0.7553474307060242,
"learning_rate": 2.3159737826104565e-05,
"loss": 0.6479,
"step": 1395
},
{
"epoch": 2.625410220346929,
"grad_norm": 0.7968561053276062,
"learning_rate": 2.3012816153408863e-05,
"loss": 0.6044,
"step": 1400
},
{
"epoch": 2.6347866854195967,
"grad_norm": 0.7905867099761963,
"learning_rate": 2.286596351774613e-05,
"loss": 0.5893,
"step": 1405
},
{
"epoch": 2.6441631504922647,
"grad_norm": 0.8002238869667053,
"learning_rate": 2.271918502094439e-05,
"loss": 0.5733,
"step": 1410
},
{
"epoch": 2.653539615564932,
"grad_norm": 0.7693070769309998,
"learning_rate": 2.2572485762256005e-05,
"loss": 0.5649,
"step": 1415
},
{
"epoch": 2.6629160806375998,
"grad_norm": 0.8900006413459778,
"learning_rate": 2.2425870838180507e-05,
"loss": 0.5842,
"step": 1420
},
{
"epoch": 2.6722925457102673,
"grad_norm": 0.7960033416748047,
"learning_rate": 2.2279345342287567e-05,
"loss": 0.5905,
"step": 1425
},
{
"epoch": 2.681669010782935,
"grad_norm": 0.7953693270683289,
"learning_rate": 2.2132914365039993e-05,
"loss": 0.6128,
"step": 1430
},
{
"epoch": 2.6910454758556024,
"grad_norm": 0.8477721214294434,
"learning_rate": 2.1986582993616926e-05,
"loss": 0.6135,
"step": 1435
},
{
"epoch": 2.70042194092827,
"grad_norm": 0.7555422782897949,
"learning_rate": 2.1840356311737084e-05,
"loss": 0.5995,
"step": 1440
},
{
"epoch": 2.709798406000938,
"grad_norm": 0.74222731590271,
"learning_rate": 2.169423939948215e-05,
"loss": 0.6354,
"step": 1445
},
{
"epoch": 2.719174871073605,
"grad_norm": 0.8535959124565125,
"learning_rate": 2.154823733312027e-05,
"loss": 0.6152,
"step": 1450
},
{
"epoch": 2.728551336146273,
"grad_norm": 0.8919956684112549,
"learning_rate": 2.140235518492975e-05,
"loss": 0.5884,
"step": 1455
},
{
"epoch": 2.7379278012189405,
"grad_norm": 0.8774855732917786,
"learning_rate": 2.125659802302275e-05,
"loss": 0.5956,
"step": 1460
},
{
"epoch": 2.747304266291608,
"grad_norm": 0.6897487044334412,
"learning_rate": 2.1110970911169316e-05,
"loss": 0.5954,
"step": 1465
},
{
"epoch": 2.7566807313642756,
"grad_norm": 0.714082658290863,
"learning_rate": 2.096547890862137e-05,
"loss": 0.6026,
"step": 1470
},
{
"epoch": 2.766057196436943,
"grad_norm": 0.890238881111145,
"learning_rate": 2.0820127069937008e-05,
"loss": 0.5849,
"step": 1475
},
{
"epoch": 2.775433661509611,
"grad_norm": 0.8606293201446533,
"learning_rate": 2.0674920444804847e-05,
"loss": 0.5873,
"step": 1480
},
{
"epoch": 2.7848101265822782,
"grad_norm": 0.8042623996734619,
"learning_rate": 2.0529864077868643e-05,
"loss": 0.6195,
"step": 1485
},
{
"epoch": 2.794186591654946,
"grad_norm": 0.8238710761070251,
"learning_rate": 2.0384963008551995e-05,
"loss": 0.6176,
"step": 1490
},
{
"epoch": 2.8035630567276137,
"grad_norm": 0.7856019139289856,
"learning_rate": 2.0240222270883288e-05,
"loss": 0.6364,
"step": 1495
},
{
"epoch": 2.8129395218002813,
"grad_norm": 0.8265510201454163,
"learning_rate": 2.0095646893320828e-05,
"loss": 0.5921,
"step": 1500
},
{
"epoch": 2.822315986872949,
"grad_norm": 0.7399757504463196,
"learning_rate": 1.9951241898578085e-05,
"loss": 0.5787,
"step": 1505
},
{
"epoch": 2.8316924519456164,
"grad_norm": 0.7769511938095093,
"learning_rate": 1.980701230344926e-05,
"loss": 0.5955,
"step": 1510
},
{
"epoch": 2.8410689170182843,
"grad_norm": 0.7197275757789612,
"learning_rate": 1.9662963118634954e-05,
"loss": 0.6349,
"step": 1515
},
{
"epoch": 2.8504453820909514,
"grad_norm": 0.844546914100647,
"learning_rate": 1.9519099348568127e-05,
"loss": 0.599,
"step": 1520
},
{
"epoch": 2.8598218471636194,
"grad_norm": 0.799412727355957,
"learning_rate": 1.93754259912402e-05,
"loss": 0.6072,
"step": 1525
},
{
"epoch": 2.869198312236287,
"grad_norm": 0.7316399216651917,
"learning_rate": 1.9231948038027462e-05,
"loss": 0.5714,
"step": 1530
},
{
"epoch": 2.8785747773089545,
"grad_norm": 0.7330530285835266,
"learning_rate": 1.9088670473517605e-05,
"loss": 0.6001,
"step": 1535
},
{
"epoch": 2.887951242381622,
"grad_norm": 0.7442522048950195,
"learning_rate": 1.8945598275336633e-05,
"loss": 0.5621,
"step": 1540
},
{
"epoch": 2.8973277074542896,
"grad_norm": 0.8327356576919556,
"learning_rate": 1.8802736413975844e-05,
"loss": 0.6401,
"step": 1545
},
{
"epoch": 2.9067041725269576,
"grad_norm": 0.8828593492507935,
"learning_rate": 1.866008985261924e-05,
"loss": 0.5987,
"step": 1550
},
{
"epoch": 2.916080637599625,
"grad_norm": 0.79734867811203,
"learning_rate": 1.8517663546971013e-05,
"loss": 0.5915,
"step": 1555
},
{
"epoch": 2.9254571026722926,
"grad_norm": 0.8401964902877808,
"learning_rate": 1.8375462445083464e-05,
"loss": 0.621,
"step": 1560
},
{
"epoch": 2.93483356774496,
"grad_norm": 0.7760779857635498,
"learning_rate": 1.8233491487185006e-05,
"loss": 0.6379,
"step": 1565
},
{
"epoch": 2.9442100328176277,
"grad_norm": 0.7567408084869385,
"learning_rate": 1.8091755605508643e-05,
"loss": 0.599,
"step": 1570
},
{
"epoch": 2.9535864978902953,
"grad_norm": 0.8464756608009338,
"learning_rate": 1.7950259724120512e-05,
"loss": 0.6333,
"step": 1575
},
{
"epoch": 2.962962962962963,
"grad_norm": 0.8063556551933289,
"learning_rate": 1.7809008758748913e-05,
"loss": 0.6245,
"step": 1580
},
{
"epoch": 2.972339428035631,
"grad_norm": 0.8217030167579651,
"learning_rate": 1.766800761661344e-05,
"loss": 0.575,
"step": 1585
},
{
"epoch": 2.9817158931082983,
"grad_norm": 0.9118950366973877,
"learning_rate": 1.752726119625459e-05,
"loss": 0.6453,
"step": 1590
},
{
"epoch": 2.991092358180966,
"grad_norm": 0.7912649512290955,
"learning_rate": 1.7386774387363484e-05,
"loss": 0.6524,
"step": 1595
},
{
"epoch": 3.0004688232536334,
"grad_norm": 0.8924903869628906,
"learning_rate": 1.724655207061209e-05,
"loss": 0.5959,
"step": 1600
},
{
"epoch": 3.009845288326301,
"grad_norm": 0.8312458992004395,
"learning_rate": 1.710659911748359e-05,
"loss": 0.6342,
"step": 1605
},
{
"epoch": 3.0192217533989685,
"grad_norm": 0.8554635047912598,
"learning_rate": 1.696692039010317e-05,
"loss": 0.5901,
"step": 1610
},
{
"epoch": 3.028598218471636,
"grad_norm": 0.8181942701339722,
"learning_rate": 1.6827520741069118e-05,
"loss": 0.6125,
"step": 1615
},
{
"epoch": 3.037974683544304,
"grad_norm": 0.8597581386566162,
"learning_rate": 1.6688405013284192e-05,
"loss": 0.6208,
"step": 1620
},
{
"epoch": 3.0473511486169715,
"grad_norm": 0.7262174487113953,
"learning_rate": 1.6549578039787436e-05,
"loss": 0.6135,
"step": 1625
},
{
"epoch": 3.056727613689639,
"grad_norm": 0.8130325675010681,
"learning_rate": 1.6411044643586204e-05,
"loss": 0.5635,
"step": 1630
},
{
"epoch": 3.0661040787623066,
"grad_norm": 0.8838590383529663,
"learning_rate": 1.627280963748869e-05,
"loss": 0.5519,
"step": 1635
},
{
"epoch": 3.075480543834974,
"grad_norm": 0.73687344789505,
"learning_rate": 1.613487782393661e-05,
"loss": 0.6453,
"step": 1640
},
{
"epoch": 3.0848570089076417,
"grad_norm": 0.8523545265197754,
"learning_rate": 1.5997253994838484e-05,
"loss": 0.5444,
"step": 1645
},
{
"epoch": 3.0942334739803092,
"grad_norm": 0.9110212922096252,
"learning_rate": 1.5859942931403072e-05,
"loss": 0.6103,
"step": 1650
},
{
"epoch": 3.103609939052977,
"grad_norm": 0.867031991481781,
"learning_rate": 1.5722949403973308e-05,
"loss": 0.656,
"step": 1655
},
{
"epoch": 3.1129864041256448,
"grad_norm": 0.9489920735359192,
"learning_rate": 1.5586278171860546e-05,
"loss": 0.5887,
"step": 1660
},
{
"epoch": 3.1223628691983123,
"grad_norm": 0.8050504326820374,
"learning_rate": 1.5449933983179256e-05,
"loss": 0.599,
"step": 1665
},
{
"epoch": 3.13173933427098,
"grad_norm": 0.8593913316726685,
"learning_rate": 1.5313921574682032e-05,
"loss": 0.5677,
"step": 1670
},
{
"epoch": 3.1411157993436474,
"grad_norm": 0.8273171782493591,
"learning_rate": 1.517824567159506e-05,
"loss": 0.5458,
"step": 1675
},
{
"epoch": 3.150492264416315,
"grad_norm": 0.9210759997367859,
"learning_rate": 1.5042910987453909e-05,
"loss": 0.6356,
"step": 1680
},
{
"epoch": 3.1598687294889825,
"grad_norm": 0.8576172590255737,
"learning_rate": 1.4907922223939874e-05,
"loss": 0.5678,
"step": 1685
},
{
"epoch": 3.1692451945616504,
"grad_norm": 0.838844358921051,
"learning_rate": 1.4773284070716503e-05,
"loss": 0.5883,
"step": 1690
},
{
"epoch": 3.178621659634318,
"grad_norm": 0.846047043800354,
"learning_rate": 1.4639001205266803e-05,
"loss": 0.5643,
"step": 1695
},
{
"epoch": 3.1879981247069855,
"grad_norm": 0.7580421566963196,
"learning_rate": 1.4505078292730632e-05,
"loss": 0.5889,
"step": 1700
},
{
"epoch": 3.197374589779653,
"grad_norm": 0.8859914541244507,
"learning_rate": 1.4371519985742715e-05,
"loss": 0.555,
"step": 1705
},
{
"epoch": 3.2067510548523206,
"grad_norm": 0.8607549071311951,
"learning_rate": 1.4238330924270927e-05,
"loss": 0.5856,
"step": 1710
},
{
"epoch": 3.216127519924988,
"grad_norm": 0.8387409448623657,
"learning_rate": 1.4105515735455149e-05,
"loss": 0.6232,
"step": 1715
},
{
"epoch": 3.2255039849976557,
"grad_norm": 0.8116487860679626,
"learning_rate": 1.3973079033446501e-05,
"loss": 0.6432,
"step": 1720
},
{
"epoch": 3.2348804500703237,
"grad_norm": 0.9285030364990234,
"learning_rate": 1.3841025419247045e-05,
"loss": 0.5808,
"step": 1725
},
{
"epoch": 3.244256915142991,
"grad_norm": 0.9143246412277222,
"learning_rate": 1.3709359480549932e-05,
"loss": 0.5563,
"step": 1730
},
{
"epoch": 3.2536333802156587,
"grad_norm": 0.8626976013183594,
"learning_rate": 1.3578085791580008e-05,
"loss": 0.5474,
"step": 1735
},
{
"epoch": 3.2630098452883263,
"grad_norm": 0.8373091220855713,
"learning_rate": 1.3447208912934927e-05,
"loss": 0.6451,
"step": 1740
},
{
"epoch": 3.272386310360994,
"grad_norm": 0.9622659087181091,
"learning_rate": 1.3316733391426716e-05,
"loss": 0.5905,
"step": 1745
},
{
"epoch": 3.2817627754336613,
"grad_norm": 0.8686250448226929,
"learning_rate": 1.3186663759923782e-05,
"loss": 0.5763,
"step": 1750
},
{
"epoch": 3.291139240506329,
"grad_norm": 0.8520202040672302,
"learning_rate": 1.3057004537193423e-05,
"loss": 0.5873,
"step": 1755
},
{
"epoch": 3.300515705578997,
"grad_norm": 0.8578153252601624,
"learning_rate": 1.2927760227744943e-05,
"loss": 0.5518,
"step": 1760
},
{
"epoch": 3.3098921706516644,
"grad_norm": 0.9105794429779053,
"learning_rate": 1.2798935321673e-05,
"loss": 0.5584,
"step": 1765
},
{
"epoch": 3.319268635724332,
"grad_norm": 0.7702586650848389,
"learning_rate": 1.2670534294501756e-05,
"loss": 0.5702,
"step": 1770
},
{
"epoch": 3.3286451007969995,
"grad_norm": 0.910254716873169,
"learning_rate": 1.2542561607029322e-05,
"loss": 0.5774,
"step": 1775
},
{
"epoch": 3.338021565869667,
"grad_norm": 0.8458461761474609,
"learning_rate": 1.2415021705172799e-05,
"loss": 0.6325,
"step": 1780
},
{
"epoch": 3.3473980309423346,
"grad_norm": 0.8077149987220764,
"learning_rate": 1.2287919019813807e-05,
"loss": 0.5764,
"step": 1785
},
{
"epoch": 3.356774496015002,
"grad_norm": 0.954138457775116,
"learning_rate": 1.2161257966644588e-05,
"loss": 0.6093,
"step": 1790
},
{
"epoch": 3.36615096108767,
"grad_norm": 0.8187435865402222,
"learning_rate": 1.2035042946014572e-05,
"loss": 0.6254,
"step": 1795
},
{
"epoch": 3.3755274261603376,
"grad_norm": 0.9092109799385071,
"learning_rate": 1.1909278342777513e-05,
"loss": 0.5884,
"step": 1800
},
{
"epoch": 3.384903891233005,
"grad_norm": 0.7966195940971375,
"learning_rate": 1.1783968526139121e-05,
"loss": 0.5836,
"step": 1805
},
{
"epoch": 3.3942803563056727,
"grad_norm": 0.9109008312225342,
"learning_rate": 1.1659117849505367e-05,
"loss": 0.6103,
"step": 1810
},
{
"epoch": 3.4036568213783402,
"grad_norm": 0.8773274421691895,
"learning_rate": 1.1534730650331096e-05,
"loss": 0.5888,
"step": 1815
},
{
"epoch": 3.413033286451008,
"grad_norm": 0.8986825346946716,
"learning_rate": 1.1410811249969475e-05,
"loss": 0.6512,
"step": 1820
},
{
"epoch": 3.4224097515236753,
"grad_norm": 0.8653113842010498,
"learning_rate": 1.1287363953521779e-05,
"loss": 0.591,
"step": 1825
},
{
"epoch": 3.4317862165963433,
"grad_norm": 0.8190916776657104,
"learning_rate": 1.1164393049687868e-05,
"loss": 0.5623,
"step": 1830
},
{
"epoch": 3.441162681669011,
"grad_norm": 0.7718261480331421,
"learning_rate": 1.104190281061718e-05,
"loss": 0.6128,
"step": 1835
},
{
"epoch": 3.4505391467416784,
"grad_norm": 0.8960718512535095,
"learning_rate": 1.0919897491760279e-05,
"loss": 0.526,
"step": 1840
},
{
"epoch": 3.459915611814346,
"grad_norm": 0.8175593614578247,
"learning_rate": 1.0798381331721109e-05,
"loss": 0.567,
"step": 1845
},
{
"epoch": 3.4692920768870135,
"grad_norm": 0.7988520860671997,
"learning_rate": 1.0677358552109618e-05,
"loss": 0.6084,
"step": 1850
},
{
"epoch": 3.4786685419596814,
"grad_norm": 0.7820413112640381,
"learning_rate": 1.0556833357395188e-05,
"loss": 0.574,
"step": 1855
},
{
"epoch": 3.488045007032349,
"grad_norm": 0.8565638661384583,
"learning_rate": 1.0436809934760527e-05,
"loss": 0.5817,
"step": 1860
},
{
"epoch": 3.4974214721050165,
"grad_norm": 0.8797834515571594,
"learning_rate": 1.031729245395622e-05,
"loss": 0.5398,
"step": 1865
},
{
"epoch": 3.506797937177684,
"grad_norm": 0.9399333000183105,
"learning_rate": 1.0198285067155827e-05,
"loss": 0.5761,
"step": 1870
},
{
"epoch": 3.5161744022503516,
"grad_norm": 0.8871551752090454,
"learning_rate": 1.0079791908811683e-05,
"loss": 0.5762,
"step": 1875
},
{
"epoch": 3.525550867323019,
"grad_norm": 0.8348819613456726,
"learning_rate": 9.961817095511242e-06,
"loss": 0.5656,
"step": 1880
},
{
"epoch": 3.5349273323956867,
"grad_norm": 0.7780579924583435,
"learning_rate": 9.844364725834057e-06,
"loss": 0.6402,
"step": 1885
},
{
"epoch": 3.5443037974683547,
"grad_norm": 1.0246992111206055,
"learning_rate": 9.727438880209366e-06,
"loss": 0.5627,
"step": 1890
},
{
"epoch": 3.5536802625410218,
"grad_norm": 0.9967918395996094,
"learning_rate": 9.611043620774419e-06,
"loss": 0.6265,
"step": 1895
},
{
"epoch": 3.5630567276136897,
"grad_norm": 0.803693413734436,
"learning_rate": 9.495182991233236e-06,
"loss": 0.5915,
"step": 1900
},
{
"epoch": 3.5724331926863573,
"grad_norm": 0.8501057028770447,
"learning_rate": 9.379861016716224e-06,
"loss": 0.6192,
"step": 1905
},
{
"epoch": 3.581809657759025,
"grad_norm": 0.8275097012519836,
"learning_rate": 9.265081703640285e-06,
"loss": 0.6139,
"step": 1910
},
{
"epoch": 3.5911861228316924,
"grad_norm": 0.863562822341919,
"learning_rate": 9.150849039569655e-06,
"loss": 0.5788,
"step": 1915
},
{
"epoch": 3.60056258790436,
"grad_norm": 0.777866005897522,
"learning_rate": 9.037166993077337e-06,
"loss": 0.5967,
"step": 1920
},
{
"epoch": 3.609939052977028,
"grad_norm": 0.8494200110435486,
"learning_rate": 8.92403951360726e-06,
"loss": 0.5525,
"step": 1925
},
{
"epoch": 3.6193155180496954,
"grad_norm": 0.7887282967567444,
"learning_rate": 8.811470531337102e-06,
"loss": 0.6459,
"step": 1930
},
{
"epoch": 3.628691983122363,
"grad_norm": 0.8828495144844055,
"learning_rate": 8.699463957041649e-06,
"loss": 0.6542,
"step": 1935
},
{
"epoch": 3.6380684481950305,
"grad_norm": 0.8919112682342529,
"learning_rate": 8.588023681957028e-06,
"loss": 0.6144,
"step": 1940
},
{
"epoch": 3.647444913267698,
"grad_norm": 0.9850883483886719,
"learning_rate": 8.477153577645481e-06,
"loss": 0.6243,
"step": 1945
},
{
"epoch": 3.6568213783403656,
"grad_norm": 0.7918851375579834,
"learning_rate": 8.36685749586087e-06,
"loss": 0.6097,
"step": 1950
},
{
"epoch": 3.666197843413033,
"grad_norm": 0.9201750159263611,
"learning_rate": 8.257139268414844e-06,
"loss": 0.589,
"step": 1955
},
{
"epoch": 3.675574308485701,
"grad_norm": 0.8442370891571045,
"learning_rate": 8.14800270704375e-06,
"loss": 0.6196,
"step": 1960
},
{
"epoch": 3.6849507735583686,
"grad_norm": 0.8792855143547058,
"learning_rate": 8.039451603276185e-06,
"loss": 0.5785,
"step": 1965
},
{
"epoch": 3.694327238631036,
"grad_norm": 0.760504424571991,
"learning_rate": 7.931489728301292e-06,
"loss": 0.6941,
"step": 1970
},
{
"epoch": 3.7037037037037037,
"grad_norm": 0.9026913046836853,
"learning_rate": 7.8241208328377e-06,
"loss": 0.5711,
"step": 1975
},
{
"epoch": 3.7130801687763713,
"grad_norm": 0.7041096687316895,
"learning_rate": 7.71734864700331e-06,
"loss": 0.5851,
"step": 1980
},
{
"epoch": 3.722456633849039,
"grad_norm": 0.8825089335441589,
"learning_rate": 7.611176880185597e-06,
"loss": 0.5877,
"step": 1985
},
{
"epoch": 3.7318330989217063,
"grad_norm": 0.8271408677101135,
"learning_rate": 7.505609220912821e-06,
"loss": 0.5841,
"step": 1990
},
{
"epoch": 3.7412095639943743,
"grad_norm": 0.8782235980033875,
"learning_rate": 7.4006493367258515e-06,
"loss": 0.6185,
"step": 1995
},
{
"epoch": 3.750586029067042,
"grad_norm": 0.9581049084663391,
"learning_rate": 7.2963008740507656e-06,
"loss": 0.6278,
"step": 2000
},
{
"epoch": 3.7599624941397094,
"grad_norm": 0.8480396270751953,
"learning_rate": 7.192567458072138e-06,
"loss": 0.6518,
"step": 2005
},
{
"epoch": 3.769338959212377,
"grad_norm": 0.8499250411987305,
"learning_rate": 7.089452692607146e-06,
"loss": 0.529,
"step": 2010
},
{
"epoch": 3.7787154242850445,
"grad_norm": 0.9328426122665405,
"learning_rate": 6.986960159980327e-06,
"loss": 0.554,
"step": 2015
},
{
"epoch": 3.788091889357712,
"grad_norm": 0.7562410831451416,
"learning_rate": 6.885093420899152e-06,
"loss": 0.5859,
"step": 2020
},
{
"epoch": 3.7974683544303796,
"grad_norm": 0.8860601186752319,
"learning_rate": 6.783856014330281e-06,
"loss": 0.5853,
"step": 2025
},
{
"epoch": 3.8068448195030475,
"grad_norm": 0.8630738258361816,
"learning_rate": 6.68325145737669e-06,
"loss": 0.5972,
"step": 2030
},
{
"epoch": 3.816221284575715,
"grad_norm": 0.8959819674491882,
"learning_rate": 6.583283245155414e-06,
"loss": 0.6114,
"step": 2035
},
{
"epoch": 3.8255977496483826,
"grad_norm": 1.036385178565979,
"learning_rate": 6.483954850676133e-06,
"loss": 0.5638,
"step": 2040
},
{
"epoch": 3.83497421472105,
"grad_norm": 0.830777645111084,
"learning_rate": 6.385269724720547e-06,
"loss": 0.6113,
"step": 2045
},
{
"epoch": 3.8443506797937177,
"grad_norm": 0.9295665621757507,
"learning_rate": 6.28723129572247e-06,
"loss": 0.5559,
"step": 2050
},
{
"epoch": 3.8537271448663852,
"grad_norm": 0.9256186485290527,
"learning_rate": 6.189842969648737e-06,
"loss": 0.6219,
"step": 2055
},
{
"epoch": 3.8631036099390528,
"grad_norm": 0.8613486289978027,
"learning_rate": 6.0931081298808316e-06,
"loss": 0.5907,
"step": 2060
},
{
"epoch": 3.8724800750117208,
"grad_norm": 0.8655708432197571,
"learning_rate": 5.997030137097426e-06,
"loss": 0.5563,
"step": 2065
},
{
"epoch": 3.8818565400843883,
"grad_norm": 0.864285409450531,
"learning_rate": 5.901612329157535e-06,
"loss": 0.6009,
"step": 2070
},
{
"epoch": 3.891233005157056,
"grad_norm": 0.819516122341156,
"learning_rate": 5.806858020984629e-06,
"loss": 0.6581,
"step": 2075
},
{
"epoch": 3.9006094702297234,
"grad_norm": 0.9411301016807556,
"learning_rate": 5.712770504451426e-06,
"loss": 0.5945,
"step": 2080
},
{
"epoch": 3.909985935302391,
"grad_norm": 0.8822996020317078,
"learning_rate": 5.619353048265552e-06,
"loss": 0.5586,
"step": 2085
},
{
"epoch": 3.9193624003750585,
"grad_norm": 0.8772969245910645,
"learning_rate": 5.526608897855953e-06,
"loss": 0.5624,
"step": 2090
},
{
"epoch": 3.928738865447726,
"grad_norm": 0.8968551754951477,
"learning_rate": 5.434541275260182e-06,
"loss": 0.6261,
"step": 2095
},
{
"epoch": 3.938115330520394,
"grad_norm": 0.8807035088539124,
"learning_rate": 5.343153379012444e-06,
"loss": 0.6223,
"step": 2100
},
{
"epoch": 3.9474917955930615,
"grad_norm": 0.8569791316986084,
"learning_rate": 5.252448384032471e-06,
"loss": 0.5724,
"step": 2105
},
{
"epoch": 3.956868260665729,
"grad_norm": 0.8825068473815918,
"learning_rate": 5.162429441515221e-06,
"loss": 0.5976,
"step": 2110
},
{
"epoch": 3.9662447257383966,
"grad_norm": 0.8981527090072632,
"learning_rate": 5.073099678821413e-06,
"loss": 0.5715,
"step": 2115
},
{
"epoch": 3.975621190811064,
"grad_norm": 0.8927028775215149,
"learning_rate": 4.984462199368872e-06,
"loss": 0.6233,
"step": 2120
},
{
"epoch": 3.9849976558837317,
"grad_norm": 0.8970484137535095,
"learning_rate": 4.8965200825247245e-06,
"loss": 0.574,
"step": 2125
},
{
"epoch": 3.994374120956399,
"grad_norm": 0.8048734664916992,
"learning_rate": 4.809276383498376e-06,
"loss": 0.5929,
"step": 2130
},
{
"epoch": 4.003750586029067,
"grad_norm": 0.8596838116645813,
"learning_rate": 4.722734133235438e-06,
"loss": 0.592,
"step": 2135
},
{
"epoch": 4.013127051101734,
"grad_norm": 0.8907204866409302,
"learning_rate": 4.636896338312374e-06,
"loss": 0.6445,
"step": 2140
},
{
"epoch": 4.022503516174402,
"grad_norm": 0.8903965353965759,
"learning_rate": 4.551765980832059e-06,
"loss": 0.5722,
"step": 2145
},
{
"epoch": 4.03187998124707,
"grad_norm": 0.9628483057022095,
"learning_rate": 4.467346018320198e-06,
"loss": 0.6088,
"step": 2150
},
{
"epoch": 4.041256446319737,
"grad_norm": 0.8551229238510132,
"learning_rate": 4.383639383622557e-06,
"loss": 0.5924,
"step": 2155
},
{
"epoch": 4.050632911392405,
"grad_norm": 0.9129508137702942,
"learning_rate": 4.300648984803085e-06,
"loss": 0.5758,
"step": 2160
},
{
"epoch": 4.060009376465072,
"grad_norm": 0.7754344940185547,
"learning_rate": 4.218377705042867e-06,
"loss": 0.5952,
"step": 2165
},
{
"epoch": 4.06938584153774,
"grad_norm": 0.8496968746185303,
"learning_rate": 4.1368284025399965e-06,
"loss": 0.604,
"step": 2170
},
{
"epoch": 4.0787623066104075,
"grad_norm": 0.7441787719726562,
"learning_rate": 4.0560039104102305e-06,
"loss": 0.5906,
"step": 2175
},
{
"epoch": 4.0881387716830755,
"grad_norm": 0.782251238822937,
"learning_rate": 3.975907036588594e-06,
"loss": 0.5407,
"step": 2180
},
{
"epoch": 4.0975152367557435,
"grad_norm": 0.9450968503952026,
"learning_rate": 3.8965405637318294e-06,
"loss": 0.5633,
"step": 2185
},
{
"epoch": 4.106891701828411,
"grad_norm": 0.9217153787612915,
"learning_rate": 3.817907249121713e-06,
"loss": 0.599,
"step": 2190
},
{
"epoch": 4.1162681669010786,
"grad_norm": 0.8987939953804016,
"learning_rate": 3.7400098245692572e-06,
"loss": 0.5879,
"step": 2195
},
{
"epoch": 4.125644631973746,
"grad_norm": 0.8248675465583801,
"learning_rate": 3.662850996319825e-06,
"loss": 0.562,
"step": 2200
},
{
"epoch": 4.135021097046414,
"grad_norm": 0.9806444644927979,
"learning_rate": 3.586433444959103e-06,
"loss": 0.5675,
"step": 2205
},
{
"epoch": 4.144397562119081,
"grad_norm": 0.9450485706329346,
"learning_rate": 3.5107598253199758e-06,
"loss": 0.5573,
"step": 2210
},
{
"epoch": 4.153774027191749,
"grad_norm": 0.9093477129936218,
"learning_rate": 3.4358327663902677e-06,
"loss": 0.5788,
"step": 2215
},
{
"epoch": 4.163150492264417,
"grad_norm": 0.9950726628303528,
"learning_rate": 3.3616548712214756e-06,
"loss": 0.5877,
"step": 2220
},
{
"epoch": 4.172526957337084,
"grad_norm": 0.7849723100662231,
"learning_rate": 3.288228716838246e-06,
"loss": 0.5661,
"step": 2225
},
{
"epoch": 4.181903422409752,
"grad_norm": 0.869097113609314,
"learning_rate": 3.2155568541489268e-06,
"loss": 0.6241,
"step": 2230
},
{
"epoch": 4.191279887482419,
"grad_norm": 0.8070918917655945,
"learning_rate": 3.143641807856898e-06,
"loss": 0.5932,
"step": 2235
},
{
"epoch": 4.200656352555087,
"grad_norm": 0.8659390807151794,
"learning_rate": 3.0724860763728767e-06,
"loss": 0.6129,
"step": 2240
},
{
"epoch": 4.210032817627754,
"grad_norm": 0.7739548683166504,
"learning_rate": 3.0020921317281264e-06,
"loss": 0.5467,
"step": 2245
},
{
"epoch": 4.219409282700422,
"grad_norm": 0.9735186100006104,
"learning_rate": 2.9324624194885436e-06,
"loss": 0.6347,
"step": 2250
},
{
"epoch": 4.22878574777309,
"grad_norm": 0.9059353470802307,
"learning_rate": 2.8635993586697553e-06,
"loss": 0.5657,
"step": 2255
},
{
"epoch": 4.238162212845757,
"grad_norm": 0.8660772442817688,
"learning_rate": 2.795505341653007e-06,
"loss": 0.5562,
"step": 2260
},
{
"epoch": 4.247538677918425,
"grad_norm": 0.8667048811912537,
"learning_rate": 2.728182734102111e-06,
"loss": 0.5954,
"step": 2265
},
{
"epoch": 4.256915142991092,
"grad_norm": 0.8115469217300415,
"learning_rate": 2.6616338748812255e-06,
"loss": 0.6499,
"step": 2270
},
{
"epoch": 4.26629160806376,
"grad_norm": 0.8536361455917358,
"learning_rate": 2.595861075973613e-06,
"loss": 0.5835,
"step": 2275
},
{
"epoch": 4.275668073136427,
"grad_norm": 0.8056153655052185,
"learning_rate": 2.530866622401304e-06,
"loss": 0.5804,
"step": 2280
},
{
"epoch": 4.285044538209095,
"grad_norm": 0.8950304985046387,
"learning_rate": 2.4666527721457416e-06,
"loss": 0.5838,
"step": 2285
},
{
"epoch": 4.294421003281763,
"grad_norm": 0.7784991264343262,
"learning_rate": 2.40322175606931e-06,
"loss": 0.6163,
"step": 2290
},
{
"epoch": 4.30379746835443,
"grad_norm": 0.9436336159706116,
"learning_rate": 2.3405757778378445e-06,
"loss": 0.5561,
"step": 2295
},
{
"epoch": 4.313173933427098,
"grad_norm": 0.8611324429512024,
"learning_rate": 2.278717013844059e-06,
"loss": 0.5709,
"step": 2300
},
{
"epoch": 4.322550398499765,
"grad_norm": 0.8388680219650269,
"learning_rate": 2.2176476131319707e-06,
"loss": 0.5357,
"step": 2305
},
{
"epoch": 4.331926863572433,
"grad_norm": 0.9825294613838196,
"learning_rate": 2.1573696973221922e-06,
"loss": 0.584,
"step": 2310
},
{
"epoch": 4.3413033286451,
"grad_norm": 0.8579577803611755,
"learning_rate": 2.0978853605382624e-06,
"loss": 0.5967,
"step": 2315
},
{
"epoch": 4.350679793717768,
"grad_norm": 0.8288280367851257,
"learning_rate": 2.0391966693338733e-06,
"loss": 0.5506,
"step": 2320
},
{
"epoch": 4.360056258790436,
"grad_norm": 0.8322898745536804,
"learning_rate": 1.9813056626210886e-06,
"loss": 0.5686,
"step": 2325
},
{
"epoch": 4.369432723863103,
"grad_norm": 0.7500792145729065,
"learning_rate": 1.9242143515994933e-06,
"loss": 0.5544,
"step": 2330
},
{
"epoch": 4.378809188935771,
"grad_norm": 0.9064257144927979,
"learning_rate": 1.8679247196863425e-06,
"loss": 0.5756,
"step": 2335
},
{
"epoch": 4.3881856540084385,
"grad_norm": 0.8289924263954163,
"learning_rate": 1.8124387224476347e-06,
"loss": 0.5522,
"step": 2340
},
{
"epoch": 4.3975621190811065,
"grad_norm": 0.7926067113876343,
"learning_rate": 1.757758287530195e-06,
"loss": 0.5979,
"step": 2345
},
{
"epoch": 4.4069385841537745,
"grad_norm": 0.8662146925926208,
"learning_rate": 1.7038853145946804e-06,
"loss": 0.5762,
"step": 2350
},
{
"epoch": 4.416315049226442,
"grad_norm": 0.9003971815109253,
"learning_rate": 1.6508216752496141e-06,
"loss": 0.6322,
"step": 2355
},
{
"epoch": 4.42569151429911,
"grad_norm": 0.9239695072174072,
"learning_rate": 1.5985692129863395e-06,
"loss": 0.5814,
"step": 2360
},
{
"epoch": 4.435067979371777,
"grad_norm": 0.8550184369087219,
"learning_rate": 1.547129743114978e-06,
"loss": 0.621,
"step": 2365
},
{
"epoch": 4.444444444444445,
"grad_norm": 0.8961548209190369,
"learning_rate": 1.496505052701372e-06,
"loss": 0.5462,
"step": 2370
},
{
"epoch": 4.453820909517112,
"grad_norm": 0.887492835521698,
"learning_rate": 1.4466969005050013e-06,
"loss": 0.5868,
"step": 2375
},
{
"epoch": 4.46319737458978,
"grad_norm": 0.854110598564148,
"learning_rate": 1.3977070169178763e-06,
"loss": 0.582,
"step": 2380
},
{
"epoch": 4.472573839662447,
"grad_norm": 0.806898832321167,
"learning_rate": 1.349537103904408e-06,
"loss": 0.5764,
"step": 2385
},
{
"epoch": 4.481950304735115,
"grad_norm": 0.8820757269859314,
"learning_rate": 1.3021888349423222e-06,
"loss": 0.5745,
"step": 2390
},
{
"epoch": 4.491326769807783,
"grad_norm": 0.8095480799674988,
"learning_rate": 1.2556638549644644e-06,
"loss": 0.6206,
"step": 2395
},
{
"epoch": 4.50070323488045,
"grad_norm": 0.8199688792228699,
"learning_rate": 1.2099637803016983e-06,
"loss": 0.5657,
"step": 2400
},
{
"epoch": 4.510079699953118,
"grad_norm": 0.8629977703094482,
"learning_rate": 1.1650901986267365e-06,
"loss": 0.5487,
"step": 2405
},
{
"epoch": 4.519456165025785,
"grad_norm": 1.0007340908050537,
"learning_rate": 1.1210446688989768e-06,
"loss": 0.6923,
"step": 2410
},
{
"epoch": 4.528832630098453,
"grad_norm": 0.9053314328193665,
"learning_rate": 1.0778287213103478e-06,
"loss": 0.5588,
"step": 2415
},
{
"epoch": 4.538209095171121,
"grad_norm": 0.909077525138855,
"learning_rate": 1.0354438572321546e-06,
"loss": 0.5588,
"step": 2420
},
{
"epoch": 4.547585560243788,
"grad_norm": 0.8841737508773804,
"learning_rate": 9.938915491629063e-07,
"loss": 0.5864,
"step": 2425
},
{
"epoch": 4.556962025316456,
"grad_norm": 0.9441853761672974,
"learning_rate": 9.531732406771771e-07,
"loss": 0.5896,
"step": 2430
},
{
"epoch": 4.566338490389123,
"grad_norm": 0.9060314893722534,
"learning_rate": 9.132903463754256e-07,
"loss": 0.6555,
"step": 2435
},
{
"epoch": 4.575714955461791,
"grad_norm": 0.8534815907478333,
"learning_rate": 8.742442518348965e-07,
"loss": 0.5814,
"step": 2440
},
{
"epoch": 4.585091420534458,
"grad_norm": 0.8542777895927429,
"learning_rate": 8.360363135614307e-07,
"loss": 0.5837,
"step": 2445
},
{
"epoch": 4.594467885607126,
"grad_norm": 0.8332429528236389,
"learning_rate": 7.986678589423758e-07,
"loss": 0.6351,
"step": 2450
},
{
"epoch": 4.603844350679793,
"grad_norm": 0.7473989725112915,
"learning_rate": 7.621401862004634e-07,
"loss": 0.6131,
"step": 2455
},
{
"epoch": 4.613220815752461,
"grad_norm": 0.944680392742157,
"learning_rate": 7.264545643486997e-07,
"loss": 0.5549,
"step": 2460
},
{
"epoch": 4.622597280825129,
"grad_norm": 0.8840490579605103,
"learning_rate": 6.916122331462799e-07,
"loss": 0.5807,
"step": 2465
},
{
"epoch": 4.631973745897796,
"grad_norm": 0.821358859539032,
"learning_rate": 6.576144030555259e-07,
"loss": 0.6375,
"step": 2470
},
{
"epoch": 4.641350210970464,
"grad_norm": 0.7711257934570312,
"learning_rate": 6.244622551998203e-07,
"loss": 0.58,
"step": 2475
},
{
"epoch": 4.650726676043131,
"grad_norm": 0.9030258059501648,
"learning_rate": 5.921569413225913e-07,
"loss": 0.6019,
"step": 2480
},
{
"epoch": 4.660103141115799,
"grad_norm": 0.8605978488922119,
"learning_rate": 5.606995837472817e-07,
"loss": 0.5911,
"step": 2485
},
{
"epoch": 4.669479606188467,
"grad_norm": 0.8807564377784729,
"learning_rate": 5.300912753383625e-07,
"loss": 0.5959,
"step": 2490
},
{
"epoch": 4.6788560712611345,
"grad_norm": 0.7941368222236633,
"learning_rate": 5.003330794633776e-07,
"loss": 0.5758,
"step": 2495
},
{
"epoch": 4.688232536333802,
"grad_norm": 0.9393755197525024,
"learning_rate": 4.714260299559875e-07,
"loss": 0.6281,
"step": 2500
},
{
"epoch": 4.6976090014064695,
"grad_norm": 0.9173010587692261,
"learning_rate": 4.4337113108005314e-07,
"loss": 0.5875,
"step": 2505
},
{
"epoch": 4.7069854664791375,
"grad_norm": 0.8334708213806152,
"learning_rate": 4.161693574947556e-07,
"loss": 0.6487,
"step": 2510
},
{
"epoch": 4.716361931551805,
"grad_norm": 0.8359082937240601,
"learning_rate": 3.8982165422073445e-07,
"loss": 0.558,
"step": 2515
},
{
"epoch": 4.725738396624473,
"grad_norm": 0.8860461711883545,
"learning_rate": 3.6432893660723886e-07,
"loss": 0.6207,
"step": 2520
},
{
"epoch": 4.73511486169714,
"grad_norm": 0.8132157921791077,
"learning_rate": 3.396920903003559e-07,
"loss": 0.5592,
"step": 2525
},
{
"epoch": 4.744491326769808,
"grad_norm": 0.8932085037231445,
"learning_rate": 3.1591197121222107e-07,
"loss": 0.5857,
"step": 2530
},
{
"epoch": 4.753867791842476,
"grad_norm": 0.9291250109672546,
"learning_rate": 2.9298940549128964e-07,
"loss": 0.5805,
"step": 2535
},
{
"epoch": 4.763244256915143,
"grad_norm": 0.8858462572097778,
"learning_rate": 2.7092518949362875e-07,
"loss": 0.5519,
"step": 2540
},
{
"epoch": 4.772620721987811,
"grad_norm": 0.8283771872520447,
"learning_rate": 2.4972008975527593e-07,
"loss": 0.6442,
"step": 2545
},
{
"epoch": 4.781997187060478,
"grad_norm": 0.9091749787330627,
"learning_rate": 2.2937484296556566e-07,
"loss": 0.5744,
"step": 2550
},
{
"epoch": 4.791373652133146,
"grad_norm": 0.8469679951667786,
"learning_rate": 2.0989015594158058e-07,
"loss": 0.6113,
"step": 2555
},
{
"epoch": 4.800750117205814,
"grad_norm": 0.9119710326194763,
"learning_rate": 1.9126670560356553e-07,
"loss": 0.6187,
"step": 2560
},
{
"epoch": 4.810126582278481,
"grad_norm": 0.9027541279792786,
"learning_rate": 1.735051389514214e-07,
"loss": 0.554,
"step": 2565
},
{
"epoch": 4.819503047351149,
"grad_norm": 0.8532278537750244,
"learning_rate": 1.5660607304223141e-07,
"loss": 0.5647,
"step": 2570
},
{
"epoch": 4.828879512423816,
"grad_norm": 0.9646401405334473,
"learning_rate": 1.4057009496881158e-07,
"loss": 0.6173,
"step": 2575
},
{
"epoch": 4.838255977496484,
"grad_norm": 0.8723577260971069,
"learning_rate": 1.2539776183932982e-07,
"loss": 0.5928,
"step": 2580
},
{
"epoch": 4.847632442569151,
"grad_norm": 0.8372580409049988,
"learning_rate": 1.1108960075794372e-07,
"loss": 0.5262,
"step": 2585
},
{
"epoch": 4.857008907641819,
"grad_norm": 0.8806862235069275,
"learning_rate": 9.764610880648451e-08,
"loss": 0.5449,
"step": 2590
},
{
"epoch": 4.866385372714487,
"grad_norm": 0.8421341180801392,
"learning_rate": 8.506775302719039e-08,
"loss": 0.5811,
"step": 2595
},
{
"epoch": 4.875761837787154,
"grad_norm": 0.8507823348045349,
"learning_rate": 7.335497040648898e-08,
"loss": 0.5818,
"step": 2600
},
{
"epoch": 4.885138302859822,
"grad_norm": 0.845939040184021,
"learning_rate": 6.250816785980385e-08,
"loss": 0.6438,
"step": 2605
},
{
"epoch": 4.894514767932489,
"grad_norm": 0.9025863409042358,
"learning_rate": 5.2527722217421416e-08,
"loss": 0.5808,
"step": 2610
},
{
"epoch": 4.903891233005157,
"grad_norm": 0.9419434666633606,
"learning_rate": 4.3413980211412516e-08,
"loss": 0.5931,
"step": 2615
},
{
"epoch": 4.913267698077824,
"grad_norm": 0.9738614559173584,
"learning_rate": 3.516725846355873e-08,
"loss": 0.551,
"step": 2620
},
{
"epoch": 4.922644163150492,
"grad_norm": 0.97791987657547,
"learning_rate": 2.7787843474386123e-08,
"loss": 0.5414,
"step": 2625
},
{
"epoch": 4.93202062822316,
"grad_norm": 0.8999043703079224,
"learning_rate": 2.127599161318161e-08,
"loss": 0.6081,
"step": 2630
},
{
"epoch": 4.941397093295827,
"grad_norm": 0.82914137840271,
"learning_rate": 1.5631929109102828e-08,
"loss": 0.6069,
"step": 2635
},
{
"epoch": 4.950773558368495,
"grad_norm": 0.870003879070282,
"learning_rate": 1.0855852043323289e-08,
"loss": 0.5897,
"step": 2640
},
{
"epoch": 4.960150023441162,
"grad_norm": 0.9073407053947449,
"learning_rate": 6.947926342204536e-09,
"loss": 0.558,
"step": 2645
},
{
"epoch": 4.96952648851383,
"grad_norm": 0.821584165096283,
"learning_rate": 3.908287771542396e-09,
"loss": 0.5617,
"step": 2650
},
{
"epoch": 4.978902953586498,
"grad_norm": 0.8036399483680725,
"learning_rate": 1.737041931845762e-09,
"loss": 0.6366,
"step": 2655
},
{
"epoch": 4.9882794186591655,
"grad_norm": 0.8421087861061096,
"learning_rate": 4.3426425467008035e-10,
"loss": 0.5921,
"step": 2660
},
{
"epoch": 4.9976558837318334,
"grad_norm": 0.8873931765556335,
"learning_rate": 0.0,
"loss": 0.5838,
"step": 2665
},
{
"epoch": 4.9976558837318334,
"step": 2665,
"total_flos": 2.0317707214423327e+18,
"train_loss": 0.623453339492626,
"train_runtime": 36076.8467,
"train_samples_per_second": 1.182,
"train_steps_per_second": 0.074
}
],
"logging_steps": 5,
"max_steps": 2665,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 100,
"total_flos": 2.0317707214423327e+18,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}