content / trainer_state.json
RiverTest's picture
RiverMTG10
da4ad80 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 5.0,
"eval_steps": 500,
"global_step": 1620,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.02,
"learning_rate": 3.999905983014452e-05,
"loss": 1.242,
"step": 5
},
{
"epoch": 0.03,
"learning_rate": 3.999623940897003e-05,
"loss": 1.0243,
"step": 10
},
{
"epoch": 0.05,
"learning_rate": 3.9991539001644015e-05,
"loss": 1.1997,
"step": 15
},
{
"epoch": 0.06,
"learning_rate": 3.998495905008461e-05,
"loss": 1.0152,
"step": 20
},
{
"epoch": 0.08,
"learning_rate": 3.997650017291901e-05,
"loss": 0.848,
"step": 25
},
{
"epoch": 0.09,
"learning_rate": 3.996616316542537e-05,
"loss": 0.9819,
"step": 30
},
{
"epoch": 0.11,
"learning_rate": 3.995394899945796e-05,
"loss": 0.8749,
"step": 35
},
{
"epoch": 0.12,
"learning_rate": 3.993985882335584e-05,
"loss": 0.7318,
"step": 40
},
{
"epoch": 0.14,
"learning_rate": 3.9923893961834914e-05,
"loss": 0.9248,
"step": 45
},
{
"epoch": 0.15,
"learning_rate": 3.9906055915863316e-05,
"loss": 0.6812,
"step": 50
},
{
"epoch": 0.17,
"learning_rate": 3.9886346362520375e-05,
"loss": 0.9068,
"step": 55
},
{
"epoch": 0.19,
"learning_rate": 3.9864767154838864e-05,
"loss": 0.747,
"step": 60
},
{
"epoch": 0.2,
"learning_rate": 3.984132032163085e-05,
"loss": 0.6329,
"step": 65
},
{
"epoch": 0.22,
"learning_rate": 3.9816008067296905e-05,
"loss": 0.653,
"step": 70
},
{
"epoch": 0.23,
"learning_rate": 3.978883277161889e-05,
"loss": 0.851,
"step": 75
},
{
"epoch": 0.25,
"learning_rate": 3.9759796989536185e-05,
"loss": 0.7384,
"step": 80
},
{
"epoch": 0.26,
"learning_rate": 3.972890345090548e-05,
"loss": 0.7389,
"step": 85
},
{
"epoch": 0.28,
"learning_rate": 3.9696155060244166e-05,
"loss": 0.7009,
"step": 90
},
{
"epoch": 0.29,
"learning_rate": 3.9661554896457205e-05,
"loss": 0.7215,
"step": 95
},
{
"epoch": 0.31,
"learning_rate": 3.9625106212547696e-05,
"loss": 0.8377,
"step": 100
},
{
"epoch": 0.32,
"learning_rate": 3.958681243531103e-05,
"loss": 0.7017,
"step": 105
},
{
"epoch": 0.34,
"learning_rate": 3.9546677165012714e-05,
"loss": 0.6667,
"step": 110
},
{
"epoch": 0.35,
"learning_rate": 3.950470417504987e-05,
"loss": 0.7254,
"step": 115
},
{
"epoch": 0.37,
"learning_rate": 3.946089741159648e-05,
"loss": 0.7825,
"step": 120
},
{
"epoch": 0.39,
"learning_rate": 3.9415260993232405e-05,
"loss": 0.599,
"step": 125
},
{
"epoch": 0.4,
"learning_rate": 3.9367799210556124e-05,
"loss": 0.5953,
"step": 130
},
{
"epoch": 0.42,
"learning_rate": 3.931851652578137e-05,
"loss": 0.634,
"step": 135
},
{
"epoch": 0.43,
"learning_rate": 3.926741757231761e-05,
"loss": 0.6156,
"step": 140
},
{
"epoch": 0.45,
"learning_rate": 3.921450715433441e-05,
"loss": 0.6751,
"step": 145
},
{
"epoch": 0.46,
"learning_rate": 3.915979024630978e-05,
"loss": 0.659,
"step": 150
},
{
"epoch": 0.48,
"learning_rate": 3.910327199256246e-05,
"loss": 0.5869,
"step": 155
},
{
"epoch": 0.49,
"learning_rate": 3.904495770676831e-05,
"loss": 0.6645,
"step": 160
},
{
"epoch": 0.51,
"learning_rate": 3.898485287146068e-05,
"loss": 0.6884,
"step": 165
},
{
"epoch": 0.52,
"learning_rate": 3.892296313751502e-05,
"loss": 0.5636,
"step": 170
},
{
"epoch": 0.54,
"learning_rate": 3.885929432361752e-05,
"loss": 0.5711,
"step": 175
},
{
"epoch": 0.56,
"learning_rate": 3.879385241571817e-05,
"loss": 0.6194,
"step": 180
},
{
"epoch": 0.57,
"learning_rate": 3.8726643566467866e-05,
"loss": 0.5763,
"step": 185
},
{
"epoch": 0.59,
"learning_rate": 3.865767409464002e-05,
"loss": 0.5505,
"step": 190
},
{
"epoch": 0.6,
"learning_rate": 3.858695048453645e-05,
"loss": 0.5694,
"step": 195
},
{
"epoch": 0.62,
"learning_rate": 3.8514479385377813e-05,
"loss": 0.6624,
"step": 200
},
{
"epoch": 0.63,
"learning_rate": 3.844026761067837e-05,
"loss": 0.6462,
"step": 205
},
{
"epoch": 0.65,
"learning_rate": 3.8364322137605484e-05,
"loss": 0.8272,
"step": 210
},
{
"epoch": 0.66,
"learning_rate": 3.828665010632359e-05,
"loss": 0.5884,
"step": 215
},
{
"epoch": 0.68,
"learning_rate": 3.8207258819322936e-05,
"loss": 0.8304,
"step": 220
},
{
"epoch": 0.69,
"learning_rate": 3.812615574073301e-05,
"loss": 0.6169,
"step": 225
},
{
"epoch": 0.71,
"learning_rate": 3.804334849562076e-05,
"loss": 0.4681,
"step": 230
},
{
"epoch": 0.73,
"learning_rate": 3.795884486927376e-05,
"loss": 0.7102,
"step": 235
},
{
"epoch": 0.74,
"learning_rate": 3.787265280646825e-05,
"loss": 0.6784,
"step": 240
},
{
"epoch": 0.76,
"learning_rate": 3.7784780410722125e-05,
"loss": 0.6555,
"step": 245
},
{
"epoch": 0.77,
"learning_rate": 3.7695235943533155e-05,
"loss": 0.7551,
"step": 250
},
{
"epoch": 0.79,
"learning_rate": 3.760402782360222e-05,
"loss": 0.6307,
"step": 255
},
{
"epoch": 0.8,
"learning_rate": 3.7511164626041823e-05,
"loss": 0.5889,
"step": 260
},
{
"epoch": 0.82,
"learning_rate": 3.741665508156984e-05,
"loss": 0.6239,
"step": 265
},
{
"epoch": 0.83,
"learning_rate": 3.732050807568878e-05,
"loss": 0.5586,
"step": 270
},
{
"epoch": 0.85,
"learning_rate": 3.7222732647850276e-05,
"loss": 0.6846,
"step": 275
},
{
"epoch": 0.86,
"learning_rate": 3.7123337990605335e-05,
"loss": 0.6084,
"step": 280
},
{
"epoch": 0.88,
"learning_rate": 3.702233344873999e-05,
"loss": 0.6484,
"step": 285
},
{
"epoch": 0.9,
"learning_rate": 3.691972851839682e-05,
"loss": 0.6492,
"step": 290
},
{
"epoch": 0.91,
"learning_rate": 3.6815532846182066e-05,
"loss": 0.5321,
"step": 295
},
{
"epoch": 0.93,
"learning_rate": 3.6709756228258735e-05,
"loss": 0.6356,
"step": 300
},
{
"epoch": 0.94,
"learning_rate": 3.660240860942558e-05,
"loss": 0.6797,
"step": 305
},
{
"epoch": 0.96,
"learning_rate": 3.649350008218214e-05,
"loss": 0.7649,
"step": 310
},
{
"epoch": 0.97,
"learning_rate": 3.638304088577984e-05,
"loss": 0.5038,
"step": 315
},
{
"epoch": 0.99,
"learning_rate": 3.6271041405259354e-05,
"loss": 0.649,
"step": 320
},
{
"epoch": 1.0,
"eval_loss": 0.6880638003349304,
"eval_runtime": 15.1875,
"eval_samples_per_second": 2.239,
"eval_steps_per_second": 2.239,
"step": 324
},
{
"epoch": 1.0,
"learning_rate": 3.6157512170474224e-05,
"loss": 0.6886,
"step": 325
},
{
"epoch": 1.02,
"learning_rate": 3.604246385510088e-05,
"loss": 0.6532,
"step": 330
},
{
"epoch": 1.03,
"learning_rate": 3.592590727563511e-05,
"loss": 0.5473,
"step": 335
},
{
"epoch": 1.05,
"learning_rate": 3.580785339037519e-05,
"loss": 0.6395,
"step": 340
},
{
"epoch": 1.06,
"learning_rate": 3.568831329839152e-05,
"loss": 0.6133,
"step": 345
},
{
"epoch": 1.08,
"learning_rate": 3.5567298238483206e-05,
"loss": 0.5942,
"step": 350
},
{
"epoch": 1.1,
"learning_rate": 3.5444819588121384e-05,
"loss": 0.5419,
"step": 355
},
{
"epoch": 1.11,
"learning_rate": 3.532088886237956e-05,
"loss": 0.4829,
"step": 360
},
{
"epoch": 1.13,
"learning_rate": 3.5195517712850994e-05,
"loss": 0.6501,
"step": 365
},
{
"epoch": 1.14,
"learning_rate": 3.506871792655321e-05,
"loss": 0.5375,
"step": 370
},
{
"epoch": 1.16,
"learning_rate": 3.4940501424819927e-05,
"loss": 0.4726,
"step": 375
},
{
"epoch": 1.17,
"learning_rate": 3.48108802621801e-05,
"loss": 0.6864,
"step": 380
},
{
"epoch": 1.19,
"learning_rate": 3.4679866625224705e-05,
"loss": 0.5066,
"step": 385
},
{
"epoch": 1.2,
"learning_rate": 3.4547472831460976e-05,
"loss": 0.6395,
"step": 390
},
{
"epoch": 1.22,
"learning_rate": 3.4413711328154295e-05,
"loss": 0.5156,
"step": 395
},
{
"epoch": 1.23,
"learning_rate": 3.4278594691157985e-05,
"loss": 0.5499,
"step": 400
},
{
"epoch": 1.25,
"learning_rate": 3.4142135623730954e-05,
"loss": 0.4902,
"step": 405
},
{
"epoch": 1.27,
"learning_rate": 3.400434695534337e-05,
"loss": 0.5851,
"step": 410
},
{
"epoch": 1.28,
"learning_rate": 3.3865241640470486e-05,
"loss": 0.6611,
"step": 415
},
{
"epoch": 1.3,
"learning_rate": 3.372483275737468e-05,
"loss": 0.555,
"step": 420
},
{
"epoch": 1.31,
"learning_rate": 3.358313350687587e-05,
"loss": 0.525,
"step": 425
},
{
"epoch": 1.33,
"learning_rate": 3.3440157211110454e-05,
"loss": 0.5132,
"step": 430
},
{
"epoch": 1.34,
"learning_rate": 3.3295917312278754e-05,
"loss": 0.4506,
"step": 435
},
{
"epoch": 1.36,
"learning_rate": 3.315042737138128e-05,
"loss": 0.7732,
"step": 440
},
{
"epoch": 1.37,
"learning_rate": 3.300370106694368e-05,
"loss": 0.5592,
"step": 445
},
{
"epoch": 1.39,
"learning_rate": 3.285575219373079e-05,
"loss": 0.3999,
"step": 450
},
{
"epoch": 1.4,
"learning_rate": 3.27065946614497e-05,
"loss": 0.5542,
"step": 455
},
{
"epoch": 1.42,
"learning_rate": 3.255624249344198e-05,
"loss": 0.5484,
"step": 460
},
{
"epoch": 1.44,
"learning_rate": 3.2404709825365204e-05,
"loss": 0.5777,
"step": 465
},
{
"epoch": 1.45,
"learning_rate": 3.2252010903864057e-05,
"loss": 0.6279,
"step": 470
},
{
"epoch": 1.47,
"learning_rate": 3.2098160085230835e-05,
"loss": 0.4809,
"step": 475
},
{
"epoch": 1.48,
"learning_rate": 3.194317183405573e-05,
"loss": 0.5332,
"step": 480
},
{
"epoch": 1.5,
"learning_rate": 3.17870607218669e-05,
"loss": 0.6713,
"step": 485
},
{
"epoch": 1.51,
"learning_rate": 3.1629841425760534e-05,
"loss": 0.5631,
"step": 490
},
{
"epoch": 1.53,
"learning_rate": 3.147152872702092e-05,
"loss": 0.5472,
"step": 495
},
{
"epoch": 1.54,
"learning_rate": 3.1312137509730776e-05,
"loss": 0.4905,
"step": 500
},
{
"epoch": 1.56,
"learning_rate": 3.115168275937186e-05,
"loss": 0.5958,
"step": 505
},
{
"epoch": 1.57,
"learning_rate": 3.0990179561416124e-05,
"loss": 0.4714,
"step": 510
},
{
"epoch": 1.59,
"learning_rate": 3.0827643099907394e-05,
"loss": 0.5457,
"step": 515
},
{
"epoch": 1.6,
"learning_rate": 3.066408865603383e-05,
"loss": 0.6976,
"step": 520
},
{
"epoch": 1.62,
"learning_rate": 3.0499531606691204e-05,
"loss": 0.5061,
"step": 525
},
{
"epoch": 1.64,
"learning_rate": 3.0333987423037262e-05,
"loss": 0.4926,
"step": 530
},
{
"epoch": 1.65,
"learning_rate": 3.016747166903711e-05,
"loss": 0.4324,
"step": 535
},
{
"epoch": 1.67,
"learning_rate": 3.0000000000000004e-05,
"loss": 0.5261,
"step": 540
},
{
"epoch": 1.68,
"learning_rate": 2.9831588161107413e-05,
"loss": 0.4908,
"step": 545
},
{
"epoch": 1.7,
"learning_rate": 2.9662251985932773e-05,
"loss": 0.4972,
"step": 550
},
{
"epoch": 1.71,
"learning_rate": 2.9492007394952812e-05,
"loss": 0.5168,
"step": 555
},
{
"epoch": 1.73,
"learning_rate": 2.9320870394050783e-05,
"loss": 0.6196,
"step": 560
},
{
"epoch": 1.74,
"learning_rate": 2.9148857073011618e-05,
"loss": 0.5633,
"step": 565
},
{
"epoch": 1.76,
"learning_rate": 2.897598360400925e-05,
"loss": 0.6044,
"step": 570
},
{
"epoch": 1.77,
"learning_rate": 2.88022662400861e-05,
"loss": 0.5007,
"step": 575
},
{
"epoch": 1.79,
"learning_rate": 2.8627721313625073e-05,
"loss": 0.47,
"step": 580
},
{
"epoch": 1.81,
"learning_rate": 2.8452365234813992e-05,
"loss": 0.5468,
"step": 585
},
{
"epoch": 1.82,
"learning_rate": 2.8276214490102788e-05,
"loss": 0.5174,
"step": 590
},
{
"epoch": 1.84,
"learning_rate": 2.8099285640653474e-05,
"loss": 0.5872,
"step": 595
},
{
"epoch": 1.85,
"learning_rate": 2.792159532078314e-05,
"loss": 0.5265,
"step": 600
},
{
"epoch": 1.87,
"learning_rate": 2.7743160236400014e-05,
"loss": 0.4837,
"step": 605
},
{
"epoch": 1.88,
"learning_rate": 2.7563997163432853e-05,
"loss": 0.555,
"step": 610
},
{
"epoch": 1.9,
"learning_rate": 2.738412294625369e-05,
"loss": 0.4858,
"step": 615
},
{
"epoch": 1.91,
"learning_rate": 2.720355449609421e-05,
"loss": 0.607,
"step": 620
},
{
"epoch": 1.93,
"learning_rate": 2.702230878945578e-05,
"loss": 0.4843,
"step": 625
},
{
"epoch": 1.94,
"learning_rate": 2.684040286651338e-05,
"loss": 0.3911,
"step": 630
},
{
"epoch": 1.96,
"learning_rate": 2.665785382951353e-05,
"loss": 0.4817,
"step": 635
},
{
"epoch": 1.98,
"learning_rate": 2.6474678841166426e-05,
"loss": 0.5283,
"step": 640
},
{
"epoch": 1.99,
"learning_rate": 2.6290895123032277e-05,
"loss": 0.4481,
"step": 645
},
{
"epoch": 2.0,
"eval_loss": 0.617668628692627,
"eval_runtime": 15.1901,
"eval_samples_per_second": 2.238,
"eval_steps_per_second": 2.238,
"step": 648
},
{
"epoch": 2.01,
"learning_rate": 2.6106519953902268e-05,
"loss": 0.4244,
"step": 650
},
{
"epoch": 2.02,
"learning_rate": 2.5921570668174e-05,
"loss": 0.4139,
"step": 655
},
{
"epoch": 2.04,
"learning_rate": 2.5736064654221808e-05,
"loss": 0.3932,
"step": 660
},
{
"epoch": 2.05,
"learning_rate": 2.5550019352761917e-05,
"loss": 0.352,
"step": 665
},
{
"epoch": 2.07,
"learning_rate": 2.536345225521275e-05,
"loss": 0.4855,
"step": 670
},
{
"epoch": 2.08,
"learning_rate": 2.5176380902050418e-05,
"loss": 0.3717,
"step": 675
},
{
"epoch": 2.1,
"learning_rate": 2.4988822881159627e-05,
"loss": 0.4027,
"step": 680
},
{
"epoch": 2.11,
"learning_rate": 2.4800795826180122e-05,
"loss": 0.5418,
"step": 685
},
{
"epoch": 2.13,
"learning_rate": 2.4612317414848804e-05,
"loss": 0.4436,
"step": 690
},
{
"epoch": 2.15,
"learning_rate": 2.4423405367337756e-05,
"loss": 0.5756,
"step": 695
},
{
"epoch": 2.16,
"learning_rate": 2.423407744458822e-05,
"loss": 0.4582,
"step": 700
},
{
"epoch": 2.18,
"learning_rate": 2.4044351446640763e-05,
"loss": 0.4207,
"step": 705
},
{
"epoch": 2.19,
"learning_rate": 2.3854245210961798e-05,
"loss": 0.4545,
"step": 710
},
{
"epoch": 2.21,
"learning_rate": 2.3663776610766537e-05,
"loss": 0.4234,
"step": 715
},
{
"epoch": 2.22,
"learning_rate": 2.3472963553338614e-05,
"loss": 0.5984,
"step": 720
},
{
"epoch": 2.24,
"learning_rate": 2.3281823978346484e-05,
"loss": 0.4888,
"step": 725
},
{
"epoch": 2.25,
"learning_rate": 2.3090375856156813e-05,
"loss": 0.4076,
"step": 730
},
{
"epoch": 2.27,
"learning_rate": 2.2898637186144935e-05,
"loss": 0.5183,
"step": 735
},
{
"epoch": 2.28,
"learning_rate": 2.2706625995002626e-05,
"loss": 0.491,
"step": 740
},
{
"epoch": 2.3,
"learning_rate": 2.251436033504326e-05,
"loss": 0.4578,
"step": 745
},
{
"epoch": 2.31,
"learning_rate": 2.2321858282504606e-05,
"loss": 0.5541,
"step": 750
},
{
"epoch": 2.33,
"learning_rate": 2.2129137935849365e-05,
"loss": 0.6285,
"step": 755
},
{
"epoch": 2.35,
"learning_rate": 2.1936217414063584e-05,
"loss": 0.4705,
"step": 760
},
{
"epoch": 2.36,
"learning_rate": 2.174311485495317e-05,
"loss": 0.4543,
"step": 765
},
{
"epoch": 2.38,
"learning_rate": 2.154984841343862e-05,
"loss": 0.4023,
"step": 770
},
{
"epoch": 2.39,
"learning_rate": 2.1356436259848186e-05,
"loss": 0.4061,
"step": 775
},
{
"epoch": 2.41,
"learning_rate": 2.1162896578209517e-05,
"loss": 0.4692,
"step": 780
},
{
"epoch": 2.42,
"learning_rate": 2.0969247564540064e-05,
"loss": 0.4699,
"step": 785
},
{
"epoch": 2.44,
"learning_rate": 2.077550742513634e-05,
"loss": 0.4712,
"step": 790
},
{
"epoch": 2.45,
"learning_rate": 2.058169437486223e-05,
"loss": 0.4355,
"step": 795
},
{
"epoch": 2.47,
"learning_rate": 2.038782663543649e-05,
"loss": 0.4162,
"step": 800
},
{
"epoch": 2.48,
"learning_rate": 2.0193922433719572e-05,
"loss": 0.4225,
"step": 805
},
{
"epoch": 2.5,
"learning_rate": 2e-05,
"loss": 0.3708,
"step": 810
},
{
"epoch": 2.52,
"learning_rate": 1.980607756628043e-05,
"loss": 0.4406,
"step": 815
},
{
"epoch": 2.53,
"learning_rate": 1.9612173364563517e-05,
"loss": 0.4062,
"step": 820
},
{
"epoch": 2.55,
"learning_rate": 1.9418305625137773e-05,
"loss": 0.4487,
"step": 825
},
{
"epoch": 2.56,
"learning_rate": 1.9224492574863663e-05,
"loss": 0.482,
"step": 830
},
{
"epoch": 2.58,
"learning_rate": 1.9030752435459946e-05,
"loss": 0.4464,
"step": 835
},
{
"epoch": 2.59,
"learning_rate": 1.8837103421790486e-05,
"loss": 0.4163,
"step": 840
},
{
"epoch": 2.61,
"learning_rate": 1.8643563740151817e-05,
"loss": 0.3538,
"step": 845
},
{
"epoch": 2.62,
"learning_rate": 1.8450151586561386e-05,
"loss": 0.4511,
"step": 850
},
{
"epoch": 2.64,
"learning_rate": 1.8256885145046837e-05,
"loss": 0.5241,
"step": 855
},
{
"epoch": 2.65,
"learning_rate": 1.806378258593642e-05,
"loss": 0.4675,
"step": 860
},
{
"epoch": 2.67,
"learning_rate": 1.7870862064150635e-05,
"loss": 0.4342,
"step": 865
},
{
"epoch": 2.69,
"learning_rate": 1.7678141717495394e-05,
"loss": 0.4943,
"step": 870
},
{
"epoch": 2.7,
"learning_rate": 1.748563966495675e-05,
"loss": 0.4316,
"step": 875
},
{
"epoch": 2.72,
"learning_rate": 1.7293374004997384e-05,
"loss": 0.4722,
"step": 880
},
{
"epoch": 2.73,
"learning_rate": 1.7101362813855068e-05,
"loss": 0.4893,
"step": 885
},
{
"epoch": 2.75,
"learning_rate": 1.6909624143843193e-05,
"loss": 0.4276,
"step": 890
},
{
"epoch": 2.76,
"learning_rate": 1.671817602165352e-05,
"loss": 0.618,
"step": 895
},
{
"epoch": 2.78,
"learning_rate": 1.6527036446661396e-05,
"loss": 0.3951,
"step": 900
},
{
"epoch": 2.79,
"learning_rate": 1.6336223389233466e-05,
"loss": 0.4054,
"step": 905
},
{
"epoch": 2.81,
"learning_rate": 1.6145754789038205e-05,
"loss": 0.366,
"step": 910
},
{
"epoch": 2.82,
"learning_rate": 1.5955648553359247e-05,
"loss": 0.4008,
"step": 915
},
{
"epoch": 2.84,
"learning_rate": 1.5765922555411793e-05,
"loss": 0.3746,
"step": 920
},
{
"epoch": 2.85,
"learning_rate": 1.5576594632662247e-05,
"loss": 0.4082,
"step": 925
},
{
"epoch": 2.87,
"learning_rate": 1.53876825851512e-05,
"loss": 0.373,
"step": 930
},
{
"epoch": 2.89,
"learning_rate": 1.5199204173819886e-05,
"loss": 0.4913,
"step": 935
},
{
"epoch": 2.9,
"learning_rate": 1.5011177118840376e-05,
"loss": 0.3975,
"step": 940
},
{
"epoch": 2.92,
"learning_rate": 1.4823619097949584e-05,
"loss": 0.3414,
"step": 945
},
{
"epoch": 2.93,
"learning_rate": 1.4636547744787251e-05,
"loss": 0.5242,
"step": 950
},
{
"epoch": 2.95,
"learning_rate": 1.444998064723809e-05,
"loss": 0.4427,
"step": 955
},
{
"epoch": 2.96,
"learning_rate": 1.4263935345778202e-05,
"loss": 0.4831,
"step": 960
},
{
"epoch": 2.98,
"learning_rate": 1.4078429331826007e-05,
"loss": 0.4786,
"step": 965
},
{
"epoch": 2.99,
"learning_rate": 1.389348004609774e-05,
"loss": 0.4109,
"step": 970
},
{
"epoch": 3.0,
"eval_loss": 0.5816357731819153,
"eval_runtime": 15.1822,
"eval_samples_per_second": 2.239,
"eval_steps_per_second": 2.239,
"step": 972
},
{
"epoch": 3.01,
"learning_rate": 1.3709104876967732e-05,
"loss": 0.3918,
"step": 975
},
{
"epoch": 3.02,
"learning_rate": 1.3525321158833582e-05,
"loss": 0.3377,
"step": 980
},
{
"epoch": 3.04,
"learning_rate": 1.3342146170486468e-05,
"loss": 0.4046,
"step": 985
},
{
"epoch": 3.06,
"learning_rate": 1.3159597133486628e-05,
"loss": 0.3737,
"step": 990
},
{
"epoch": 3.07,
"learning_rate": 1.2977691210544228e-05,
"loss": 0.4188,
"step": 995
},
{
"epoch": 3.09,
"learning_rate": 1.2796445503905797e-05,
"loss": 0.3644,
"step": 1000
},
{
"epoch": 3.1,
"learning_rate": 1.2615877053746315e-05,
"loss": 0.5003,
"step": 1005
},
{
"epoch": 3.12,
"learning_rate": 1.2436002836567154e-05,
"loss": 0.3132,
"step": 1010
},
{
"epoch": 3.13,
"learning_rate": 1.2256839763599993e-05,
"loss": 0.3906,
"step": 1015
},
{
"epoch": 3.15,
"learning_rate": 1.2078404679216864e-05,
"loss": 0.3351,
"step": 1020
},
{
"epoch": 3.16,
"learning_rate": 1.1900714359346527e-05,
"loss": 0.4456,
"step": 1025
},
{
"epoch": 3.18,
"learning_rate": 1.1723785509897219e-05,
"loss": 0.4693,
"step": 1030
},
{
"epoch": 3.19,
"learning_rate": 1.1547634765186016e-05,
"loss": 0.342,
"step": 1035
},
{
"epoch": 3.21,
"learning_rate": 1.1372278686374935e-05,
"loss": 0.411,
"step": 1040
},
{
"epoch": 3.23,
"learning_rate": 1.119773375991391e-05,
"loss": 0.4285,
"step": 1045
},
{
"epoch": 3.24,
"learning_rate": 1.1024016395990758e-05,
"loss": 0.4612,
"step": 1050
},
{
"epoch": 3.26,
"learning_rate": 1.0851142926988387e-05,
"loss": 0.473,
"step": 1055
},
{
"epoch": 3.27,
"learning_rate": 1.067912960594923e-05,
"loss": 0.3247,
"step": 1060
},
{
"epoch": 3.29,
"learning_rate": 1.0507992605047193e-05,
"loss": 0.3601,
"step": 1065
},
{
"epoch": 3.3,
"learning_rate": 1.033774801406723e-05,
"loss": 0.3196,
"step": 1070
},
{
"epoch": 3.32,
"learning_rate": 1.016841183889259e-05,
"loss": 0.415,
"step": 1075
},
{
"epoch": 3.33,
"learning_rate": 1.0000000000000006e-05,
"loss": 0.4202,
"step": 1080
},
{
"epoch": 3.35,
"learning_rate": 9.832528330962896e-06,
"loss": 0.3907,
"step": 1085
},
{
"epoch": 3.36,
"learning_rate": 9.666012576962743e-06,
"loss": 0.3335,
"step": 1090
},
{
"epoch": 3.38,
"learning_rate": 9.5004683933088e-06,
"loss": 0.362,
"step": 1095
},
{
"epoch": 3.4,
"learning_rate": 9.33591134396618e-06,
"loss": 0.4044,
"step": 1100
},
{
"epoch": 3.41,
"learning_rate": 9.172356900092607e-06,
"loss": 0.4207,
"step": 1105
},
{
"epoch": 3.43,
"learning_rate": 9.009820438583881e-06,
"loss": 0.3547,
"step": 1110
},
{
"epoch": 3.44,
"learning_rate": 8.848317240628146e-06,
"loss": 0.3789,
"step": 1115
},
{
"epoch": 3.46,
"learning_rate": 8.687862490269232e-06,
"loss": 0.3858,
"step": 1120
},
{
"epoch": 3.47,
"learning_rate": 8.528471272979083e-06,
"loss": 0.4855,
"step": 1125
},
{
"epoch": 3.49,
"learning_rate": 8.370158574239466e-06,
"loss": 0.3644,
"step": 1130
},
{
"epoch": 3.5,
"learning_rate": 8.212939278133105e-06,
"loss": 0.5069,
"step": 1135
},
{
"epoch": 3.52,
"learning_rate": 8.056828165944282e-06,
"loss": 0.371,
"step": 1140
},
{
"epoch": 3.53,
"learning_rate": 7.901839914769165e-06,
"loss": 0.4089,
"step": 1145
},
{
"epoch": 3.55,
"learning_rate": 7.747989096135943e-06,
"loss": 0.3389,
"step": 1150
},
{
"epoch": 3.56,
"learning_rate": 7.595290174634802e-06,
"loss": 0.5253,
"step": 1155
},
{
"epoch": 3.58,
"learning_rate": 7.443757506558033e-06,
"loss": 0.4673,
"step": 1160
},
{
"epoch": 3.6,
"learning_rate": 7.293405338550304e-06,
"loss": 0.3363,
"step": 1165
},
{
"epoch": 3.61,
"learning_rate": 7.1442478062692135e-06,
"loss": 0.3789,
"step": 1170
},
{
"epoch": 3.63,
"learning_rate": 6.996298933056333e-06,
"loss": 0.3884,
"step": 1175
},
{
"epoch": 3.64,
"learning_rate": 6.84957262861873e-06,
"loss": 0.3993,
"step": 1180
},
{
"epoch": 3.66,
"learning_rate": 6.704082687721243e-06,
"loss": 0.3588,
"step": 1185
},
{
"epoch": 3.67,
"learning_rate": 6.559842788889552e-06,
"loss": 0.5273,
"step": 1190
},
{
"epoch": 3.69,
"learning_rate": 6.4168664931241384e-06,
"loss": 0.513,
"step": 1195
},
{
"epoch": 3.7,
"learning_rate": 6.275167242625331e-06,
"loss": 0.4905,
"step": 1200
},
{
"epoch": 3.72,
"learning_rate": 6.1347583595295205e-06,
"loss": 0.3483,
"step": 1205
},
{
"epoch": 3.73,
"learning_rate": 5.9956530446566305e-06,
"loss": 0.5998,
"step": 1210
},
{
"epoch": 3.75,
"learning_rate": 5.857864376269051e-06,
"loss": 0.3786,
"step": 1215
},
{
"epoch": 3.77,
"learning_rate": 5.721405308842023e-06,
"loss": 0.3979,
"step": 1220
},
{
"epoch": 3.78,
"learning_rate": 5.586288671845708e-06,
"loss": 0.3185,
"step": 1225
},
{
"epoch": 3.8,
"learning_rate": 5.452527168539026e-06,
"loss": 0.3171,
"step": 1230
},
{
"epoch": 3.81,
"learning_rate": 5.320133374775296e-06,
"loss": 0.4599,
"step": 1235
},
{
"epoch": 3.83,
"learning_rate": 5.189119737819912e-06,
"loss": 0.4121,
"step": 1240
},
{
"epoch": 3.84,
"learning_rate": 5.059498575180084e-06,
"loss": 0.3233,
"step": 1245
},
{
"epoch": 3.86,
"learning_rate": 4.9312820734467855e-06,
"loss": 0.5224,
"step": 1250
},
{
"epoch": 3.87,
"learning_rate": 4.804482287149015e-06,
"loss": 0.4146,
"step": 1255
},
{
"epoch": 3.89,
"learning_rate": 4.679111137620442e-06,
"loss": 0.3294,
"step": 1260
},
{
"epoch": 3.9,
"learning_rate": 4.555180411878617e-06,
"loss": 0.3016,
"step": 1265
},
{
"epoch": 3.92,
"learning_rate": 4.4327017615168e-06,
"loss": 0.2814,
"step": 1270
},
{
"epoch": 3.94,
"learning_rate": 4.311686701608486e-06,
"loss": 0.4891,
"step": 1275
},
{
"epoch": 3.95,
"learning_rate": 4.1921466096248164e-06,
"loss": 0.3668,
"step": 1280
},
{
"epoch": 3.97,
"learning_rate": 4.074092724364889e-06,
"loss": 0.3226,
"step": 1285
},
{
"epoch": 3.98,
"learning_rate": 3.957536144899123e-06,
"loss": 0.3381,
"step": 1290
},
{
"epoch": 4.0,
"learning_rate": 3.842487829525779e-06,
"loss": 0.3175,
"step": 1295
},
{
"epoch": 4.0,
"eval_loss": 0.5760383605957031,
"eval_runtime": 15.1722,
"eval_samples_per_second": 2.241,
"eval_steps_per_second": 2.241,
"step": 1296
},
{
"epoch": 4.01,
"learning_rate": 3.7289585947406504e-06,
"loss": 0.3827,
"step": 1300
},
{
"epoch": 4.03,
"learning_rate": 3.616959114220162e-06,
"loss": 0.33,
"step": 1305
},
{
"epoch": 4.04,
"learning_rate": 3.5064999178178648e-06,
"loss": 0.4641,
"step": 1310
},
{
"epoch": 4.06,
"learning_rate": 3.397591390574424e-06,
"loss": 0.2855,
"step": 1315
},
{
"epoch": 4.07,
"learning_rate": 3.290243771741275e-06,
"loss": 0.492,
"step": 1320
},
{
"epoch": 4.09,
"learning_rate": 3.1844671538179407e-06,
"loss": 0.3964,
"step": 1325
},
{
"epoch": 4.1,
"learning_rate": 3.0802714816031787e-06,
"loss": 0.4128,
"step": 1330
},
{
"epoch": 4.12,
"learning_rate": 2.9776665512600054e-06,
"loss": 0.5308,
"step": 1335
},
{
"epoch": 4.14,
"learning_rate": 2.876662009394673e-06,
"loss": 0.3281,
"step": 1340
},
{
"epoch": 4.15,
"learning_rate": 2.777267352149724e-06,
"loss": 0.4674,
"step": 1345
},
{
"epoch": 4.17,
"learning_rate": 2.679491924311226e-06,
"loss": 0.3746,
"step": 1350
},
{
"epoch": 4.18,
"learning_rate": 2.5833449184301597e-06,
"loss": 0.4369,
"step": 1355
},
{
"epoch": 4.2,
"learning_rate": 2.488835373958185e-06,
"loss": 0.3269,
"step": 1360
},
{
"epoch": 4.21,
"learning_rate": 2.3959721763977805e-06,
"loss": 0.4396,
"step": 1365
},
{
"epoch": 4.23,
"learning_rate": 2.304764056466844e-06,
"loss": 0.3521,
"step": 1370
},
{
"epoch": 4.24,
"learning_rate": 2.2152195892778795e-06,
"loss": 0.4001,
"step": 1375
},
{
"epoch": 4.26,
"learning_rate": 2.127347193531757e-06,
"loss": 0.3788,
"step": 1380
},
{
"epoch": 4.27,
"learning_rate": 2.0411551307262356e-06,
"loss": 0.4339,
"step": 1385
},
{
"epoch": 4.29,
"learning_rate": 1.9566515043792455e-06,
"loss": 0.3463,
"step": 1390
},
{
"epoch": 4.31,
"learning_rate": 1.8738442592670014e-06,
"loss": 0.4395,
"step": 1395
},
{
"epoch": 4.32,
"learning_rate": 1.792741180677069e-06,
"loss": 0.3108,
"step": 1400
},
{
"epoch": 4.34,
"learning_rate": 1.713349893676415e-06,
"loss": 0.322,
"step": 1405
},
{
"epoch": 4.35,
"learning_rate": 1.6356778623945223e-06,
"loss": 0.3564,
"step": 1410
},
{
"epoch": 4.37,
"learning_rate": 1.5597323893216332e-06,
"loss": 0.4093,
"step": 1415
},
{
"epoch": 4.38,
"learning_rate": 1.4855206146221934e-06,
"loss": 0.4037,
"step": 1420
},
{
"epoch": 4.4,
"learning_rate": 1.4130495154635494e-06,
"loss": 0.3104,
"step": 1425
},
{
"epoch": 4.41,
"learning_rate": 1.3423259053599891e-06,
"loss": 0.4232,
"step": 1430
},
{
"epoch": 4.43,
"learning_rate": 1.273356433532138e-06,
"loss": 0.3957,
"step": 1435
},
{
"epoch": 4.44,
"learning_rate": 1.2061475842818337e-06,
"loss": 0.4254,
"step": 1440
},
{
"epoch": 4.46,
"learning_rate": 1.140705676382483e-06,
"loss": 0.3837,
"step": 1445
},
{
"epoch": 4.48,
"learning_rate": 1.0770368624849947e-06,
"loss": 0.3663,
"step": 1450
},
{
"epoch": 4.49,
"learning_rate": 1.0151471285393223e-06,
"loss": 0.3837,
"step": 1455
},
{
"epoch": 4.51,
"learning_rate": 9.550422932316938e-07,
"loss": 0.331,
"step": 1460
},
{
"epoch": 4.52,
"learning_rate": 8.967280074375395e-07,
"loss": 0.3507,
"step": 1465
},
{
"epoch": 4.54,
"learning_rate": 8.402097536902221e-07,
"loss": 0.2632,
"step": 1470
},
{
"epoch": 4.55,
"learning_rate": 7.854928456655897e-07,
"loss": 0.45,
"step": 1475
},
{
"epoch": 4.57,
"learning_rate": 7.325824276823934e-07,
"loss": 0.3298,
"step": 1480
},
{
"epoch": 4.58,
"learning_rate": 6.814834742186361e-07,
"loss": 0.2389,
"step": 1485
},
{
"epoch": 4.6,
"learning_rate": 6.322007894438842e-07,
"loss": 0.3721,
"step": 1490
},
{
"epoch": 4.61,
"learning_rate": 5.847390067676007e-07,
"loss": 0.442,
"step": 1495
},
{
"epoch": 4.63,
"learning_rate": 5.391025884035239e-07,
"loss": 0.3195,
"step": 1500
},
{
"epoch": 4.65,
"learning_rate": 4.952958249501393e-07,
"loss": 0.3131,
"step": 1505
},
{
"epoch": 4.66,
"learning_rate": 4.533228349872887e-07,
"loss": 0.4478,
"step": 1510
},
{
"epoch": 4.68,
"learning_rate": 4.1318756468897047e-07,
"loss": 0.4988,
"step": 1515
},
{
"epoch": 4.69,
"learning_rate": 3.748937874523062e-07,
"loss": 0.3354,
"step": 1520
},
{
"epoch": 4.71,
"learning_rate": 3.3844510354280157e-07,
"loss": 0.3146,
"step": 1525
},
{
"epoch": 4.72,
"learning_rate": 3.038449397558396e-07,
"loss": 0.3662,
"step": 1530
},
{
"epoch": 4.74,
"learning_rate": 2.7109654909452275e-07,
"loss": 0.5044,
"step": 1535
},
{
"epoch": 4.75,
"learning_rate": 2.402030104638198e-07,
"loss": 0.3398,
"step": 1540
},
{
"epoch": 4.77,
"learning_rate": 2.111672283811106e-07,
"loss": 0.3719,
"step": 1545
},
{
"epoch": 4.78,
"learning_rate": 1.839919327030937e-07,
"loss": 0.3409,
"step": 1550
},
{
"epoch": 4.8,
"learning_rate": 1.5867967836915354e-07,
"loss": 0.3881,
"step": 1555
},
{
"epoch": 4.81,
"learning_rate": 1.3523284516113955e-07,
"loss": 0.3081,
"step": 1560
},
{
"epoch": 4.83,
"learning_rate": 1.1365363747963155e-07,
"loss": 0.4667,
"step": 1565
},
{
"epoch": 4.85,
"learning_rate": 9.394408413668343e-08,
"loss": 0.4495,
"step": 1570
},
{
"epoch": 4.86,
"learning_rate": 7.61060381650891e-08,
"loss": 0.3963,
"step": 1575
},
{
"epoch": 4.88,
"learning_rate": 6.014117664415953e-08,
"loss": 0.5192,
"step": 1580
},
{
"epoch": 4.89,
"learning_rate": 4.60510005420467e-08,
"loss": 0.4057,
"step": 1585
},
{
"epoch": 4.91,
"learning_rate": 3.383683457463649e-08,
"loss": 0.3126,
"step": 1590
},
{
"epoch": 4.92,
"learning_rate": 2.3499827080992744e-08,
"loss": 0.3974,
"step": 1595
},
{
"epoch": 4.94,
"learning_rate": 1.5040949915399173e-08,
"loss": 0.3476,
"step": 1600
},
{
"epoch": 4.95,
"learning_rate": 8.460998355988014e-09,
"loss": 0.3137,
"step": 1605
},
{
"epoch": 4.97,
"learning_rate": 3.760591029973171e-09,
"loss": 0.3195,
"step": 1610
},
{
"epoch": 4.98,
"learning_rate": 9.401698554767358e-10,
"loss": 0.3808,
"step": 1615
},
{
"epoch": 5.0,
"learning_rate": 0.0,
"loss": 0.4407,
"step": 1620
},
{
"epoch": 5.0,
"eval_loss": 0.5766593813896179,
"eval_runtime": 15.1796,
"eval_samples_per_second": 2.24,
"eval_steps_per_second": 2.24,
"step": 1620
},
{
"epoch": 5.0,
"step": 1620,
"total_flos": 1.426343690502144e+17,
"train_loss": 0.49629129811569495,
"train_runtime": 2296.2906,
"train_samples_per_second": 0.705,
"train_steps_per_second": 0.705
}
],
"logging_steps": 5,
"max_steps": 1620,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 500,
"total_flos": 1.426343690502144e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}