| { | |
| "best_global_step": 900, | |
| "best_metric": 1.0647863328758695, | |
| "best_model_checkpoint": "./SALAMA_NEW99/checkpoint-900", | |
| "epoch": 3.0, | |
| "eval_steps": 300, | |
| "global_step": 921, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03262642740619902, | |
| "grad_norm": 4.136127471923828, | |
| "learning_rate": 1.8e-07, | |
| "loss": 0.0657, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.06525285481239804, | |
| "grad_norm": 3.8954381942749023, | |
| "learning_rate": 3.8e-07, | |
| "loss": 0.0652, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.09787928221859707, | |
| "grad_norm": 3.1246440410614014, | |
| "learning_rate": 5.800000000000001e-07, | |
| "loss": 0.0435, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.13050570962479607, | |
| "grad_norm": 5.918012619018555, | |
| "learning_rate": 7.8e-07, | |
| "loss": 0.071, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.1631321370309951, | |
| "grad_norm": 3.991386890411377, | |
| "learning_rate": 9.800000000000001e-07, | |
| "loss": 0.0554, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.19575856443719414, | |
| "grad_norm": 3.4036903381347656, | |
| "learning_rate": 1.1800000000000001e-06, | |
| "loss": 0.0729, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.22838499184339314, | |
| "grad_norm": 3.296243906021118, | |
| "learning_rate": 1.3800000000000001e-06, | |
| "loss": 0.0683, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.26101141924959215, | |
| "grad_norm": 4.985351085662842, | |
| "learning_rate": 1.5800000000000001e-06, | |
| "loss": 0.0635, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.2936378466557912, | |
| "grad_norm": 5.250389575958252, | |
| "learning_rate": 1.7800000000000001e-06, | |
| "loss": 0.0864, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.3262642740619902, | |
| "grad_norm": 3.3003408908843994, | |
| "learning_rate": 1.98e-06, | |
| "loss": 0.0582, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.35889070146818924, | |
| "grad_norm": 5.999060153961182, | |
| "learning_rate": 2.1800000000000003e-06, | |
| "loss": 0.073, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.3915171288743883, | |
| "grad_norm": 2.205695152282715, | |
| "learning_rate": 2.38e-06, | |
| "loss": 0.055, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.42414355628058725, | |
| "grad_norm": 3.037153482437134, | |
| "learning_rate": 2.5800000000000003e-06, | |
| "loss": 0.0541, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.4567699836867863, | |
| "grad_norm": 5.886940002441406, | |
| "learning_rate": 2.7800000000000005e-06, | |
| "loss": 0.0593, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.4893964110929853, | |
| "grad_norm": 3.1995198726654053, | |
| "learning_rate": 2.9800000000000003e-06, | |
| "loss": 0.0566, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5220228384991843, | |
| "grad_norm": 1.9580050706863403, | |
| "learning_rate": 3.1800000000000005e-06, | |
| "loss": 0.0533, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.5546492659053833, | |
| "grad_norm": 3.5993852615356445, | |
| "learning_rate": 3.3800000000000007e-06, | |
| "loss": 0.0723, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.5872756933115824, | |
| "grad_norm": 4.075970649719238, | |
| "learning_rate": 3.58e-06, | |
| "loss": 0.0671, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.6199021207177814, | |
| "grad_norm": 5.4311347007751465, | |
| "learning_rate": 3.7800000000000002e-06, | |
| "loss": 0.0751, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.6525285481239804, | |
| "grad_norm": 4.403566837310791, | |
| "learning_rate": 3.980000000000001e-06, | |
| "loss": 0.0571, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.6851549755301795, | |
| "grad_norm": 3.001912832260132, | |
| "learning_rate": 4.18e-06, | |
| "loss": 0.0697, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.7177814029363785, | |
| "grad_norm": 6.647680759429932, | |
| "learning_rate": 4.38e-06, | |
| "loss": 0.0636, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.7504078303425775, | |
| "grad_norm": 2.4966087341308594, | |
| "learning_rate": 4.58e-06, | |
| "loss": 0.0524, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.7830342577487766, | |
| "grad_norm": 5.768764019012451, | |
| "learning_rate": 4.78e-06, | |
| "loss": 0.042, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.8156606851549756, | |
| "grad_norm": 4.395790100097656, | |
| "learning_rate": 4.980000000000001e-06, | |
| "loss": 0.0703, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.8482871125611745, | |
| "grad_norm": 3.6472349166870117, | |
| "learning_rate": 5.18e-06, | |
| "loss": 0.055, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.8809135399673735, | |
| "grad_norm": 4.439610958099365, | |
| "learning_rate": 5.380000000000001e-06, | |
| "loss": 0.0719, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.9135399673735726, | |
| "grad_norm": 3.3142175674438477, | |
| "learning_rate": 5.580000000000001e-06, | |
| "loss": 0.0587, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.9461663947797716, | |
| "grad_norm": 2.1484858989715576, | |
| "learning_rate": 5.78e-06, | |
| "loss": 0.0491, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.9787928221859706, | |
| "grad_norm": 3.151224136352539, | |
| "learning_rate": 5.98e-06, | |
| "loss": 0.0597, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.9787928221859706, | |
| "eval_loss": 0.035220563411712646, | |
| "eval_runtime": 877.1298, | |
| "eval_samples_per_second": 2.794, | |
| "eval_steps_per_second": 0.35, | |
| "eval_wer": 2.4513747574653353, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.0097879282218598, | |
| "grad_norm": 3.2030868530273438, | |
| "learning_rate": 6.18e-06, | |
| "loss": 0.0639, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.0424143556280587, | |
| "grad_norm": 2.8345236778259277, | |
| "learning_rate": 6.380000000000001e-06, | |
| "loss": 0.0251, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.0750407830342577, | |
| "grad_norm": 0.8594768047332764, | |
| "learning_rate": 6.5800000000000005e-06, | |
| "loss": 0.0297, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.1076672104404568, | |
| "grad_norm": 4.065981388092041, | |
| "learning_rate": 6.780000000000001e-06, | |
| "loss": 0.0343, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.1402936378466557, | |
| "grad_norm": 1.896403431892395, | |
| "learning_rate": 6.98e-06, | |
| "loss": 0.0182, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.1729200652528549, | |
| "grad_norm": 1.5980521440505981, | |
| "learning_rate": 7.180000000000001e-06, | |
| "loss": 0.0319, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.2055464926590538, | |
| "grad_norm": 1.096835732460022, | |
| "learning_rate": 7.3800000000000005e-06, | |
| "loss": 0.0208, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.238172920065253, | |
| "grad_norm": 2.3664422035217285, | |
| "learning_rate": 7.58e-06, | |
| "loss": 0.0203, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.2707993474714518, | |
| "grad_norm": 3.103567600250244, | |
| "learning_rate": 7.78e-06, | |
| "loss": 0.0319, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.3034257748776508, | |
| "grad_norm": 2.2237136363983154, | |
| "learning_rate": 7.980000000000002e-06, | |
| "loss": 0.034, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.33605220228385, | |
| "grad_norm": 1.7810306549072266, | |
| "learning_rate": 8.18e-06, | |
| "loss": 0.0233, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.368678629690049, | |
| "grad_norm": 6.052729606628418, | |
| "learning_rate": 8.380000000000001e-06, | |
| "loss": 0.0205, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.401305057096248, | |
| "grad_norm": 1.5679905414581299, | |
| "learning_rate": 8.580000000000001e-06, | |
| "loss": 0.0234, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.433931484502447, | |
| "grad_norm": 2.6520698070526123, | |
| "learning_rate": 8.78e-06, | |
| "loss": 0.0212, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.466557911908646, | |
| "grad_norm": 2.1195335388183594, | |
| "learning_rate": 8.98e-06, | |
| "loss": 0.0199, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.499184339314845, | |
| "grad_norm": 2.4394052028656006, | |
| "learning_rate": 9.180000000000002e-06, | |
| "loss": 0.0173, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.531810766721044, | |
| "grad_norm": 2.404578685760498, | |
| "learning_rate": 9.38e-06, | |
| "loss": 0.0282, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.564437194127243, | |
| "grad_norm": 1.8087410926818848, | |
| "learning_rate": 9.58e-06, | |
| "loss": 0.0264, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.597063621533442, | |
| "grad_norm": 2.871264696121216, | |
| "learning_rate": 9.780000000000001e-06, | |
| "loss": 0.0379, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.629690048939641, | |
| "grad_norm": 3.031569719314575, | |
| "learning_rate": 9.980000000000001e-06, | |
| "loss": 0.0355, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.6623164763458402, | |
| "grad_norm": 2.7142934799194336, | |
| "learning_rate": 9.78622327790974e-06, | |
| "loss": 0.0399, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.6949429037520392, | |
| "grad_norm": 3.137960910797119, | |
| "learning_rate": 9.548693586698338e-06, | |
| "loss": 0.0265, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.727569331158238, | |
| "grad_norm": 3.7411792278289795, | |
| "learning_rate": 9.311163895486937e-06, | |
| "loss": 0.0407, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.7601957585644372, | |
| "grad_norm": 2.5436832904815674, | |
| "learning_rate": 9.073634204275536e-06, | |
| "loss": 0.031, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.7928221859706364, | |
| "grad_norm": 2.5608668327331543, | |
| "learning_rate": 8.836104513064134e-06, | |
| "loss": 0.0359, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.8254486133768353, | |
| "grad_norm": 2.3166933059692383, | |
| "learning_rate": 8.598574821852733e-06, | |
| "loss": 0.0394, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.8580750407830342, | |
| "grad_norm": 3.6695189476013184, | |
| "learning_rate": 8.36104513064133e-06, | |
| "loss": 0.0381, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.8907014681892331, | |
| "grad_norm": 2.352613925933838, | |
| "learning_rate": 8.12351543942993e-06, | |
| "loss": 0.0261, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.9233278955954323, | |
| "grad_norm": 2.972198963165283, | |
| "learning_rate": 7.885985748218527e-06, | |
| "loss": 0.0316, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.9559543230016314, | |
| "grad_norm": 2.503997564315796, | |
| "learning_rate": 7.648456057007126e-06, | |
| "loss": 0.0349, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.9559543230016314, | |
| "eval_loss": 0.023073038086295128, | |
| "eval_runtime": 889.2563, | |
| "eval_samples_per_second": 2.756, | |
| "eval_steps_per_second": 0.345, | |
| "eval_wer": 1.6232076096729924, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.9885807504078303, | |
| "grad_norm": 3.9330666065216064, | |
| "learning_rate": 7.410926365795725e-06, | |
| "loss": 0.02, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 2.0195758564437196, | |
| "grad_norm": 0.669647753238678, | |
| "learning_rate": 7.173396674584323e-06, | |
| "loss": 0.0263, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.0522022838499185, | |
| "grad_norm": 1.282844066619873, | |
| "learning_rate": 6.935866983372922e-06, | |
| "loss": 0.008, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.0848287112561175, | |
| "grad_norm": 1.8597607612609863, | |
| "learning_rate": 6.698337292161521e-06, | |
| "loss": 0.0108, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.1174551386623164, | |
| "grad_norm": 0.861115574836731, | |
| "learning_rate": 6.46080760095012e-06, | |
| "loss": 0.0147, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.1500815660685153, | |
| "grad_norm": 0.4870440363883972, | |
| "learning_rate": 6.223277909738717e-06, | |
| "loss": 0.0094, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.1827079934747147, | |
| "grad_norm": 2.0000433921813965, | |
| "learning_rate": 5.9857482185273165e-06, | |
| "loss": 0.01, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.2153344208809136, | |
| "grad_norm": 0.576457142829895, | |
| "learning_rate": 5.748218527315916e-06, | |
| "loss": 0.009, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.2479608482871125, | |
| "grad_norm": 2.5040650367736816, | |
| "learning_rate": 5.510688836104513e-06, | |
| "loss": 0.0073, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.2805872756933114, | |
| "grad_norm": 0.8777064085006714, | |
| "learning_rate": 5.273159144893112e-06, | |
| "loss": 0.0069, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.3132137030995104, | |
| "grad_norm": 1.4725840091705322, | |
| "learning_rate": 5.0356294536817105e-06, | |
| "loss": 0.0063, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 2.3458401305057097, | |
| "grad_norm": 4.44694185256958, | |
| "learning_rate": 4.798099762470309e-06, | |
| "loss": 0.0136, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.3784665579119086, | |
| "grad_norm": 2.1856603622436523, | |
| "learning_rate": 4.560570071258908e-06, | |
| "loss": 0.0116, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.4110929853181076, | |
| "grad_norm": 1.243648886680603, | |
| "learning_rate": 4.323040380047506e-06, | |
| "loss": 0.0073, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 2.443719412724307, | |
| "grad_norm": 1.029271125793457, | |
| "learning_rate": 4.0855106888361044e-06, | |
| "loss": 0.008, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.476345840130506, | |
| "grad_norm": 0.3295634686946869, | |
| "learning_rate": 3.8479809976247036e-06, | |
| "loss": 0.0059, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 2.5089722675367048, | |
| "grad_norm": 1.9740921258926392, | |
| "learning_rate": 3.610451306413302e-06, | |
| "loss": 0.012, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 2.5415986949429037, | |
| "grad_norm": 0.44365453720092773, | |
| "learning_rate": 3.3729216152019006e-06, | |
| "loss": 0.0145, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.5742251223491026, | |
| "grad_norm": 0.8667551875114441, | |
| "learning_rate": 3.135391923990499e-06, | |
| "loss": 0.0133, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 2.6068515497553015, | |
| "grad_norm": 1.748785138130188, | |
| "learning_rate": 2.897862232779098e-06, | |
| "loss": 0.0112, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.639477977161501, | |
| "grad_norm": 1.0357842445373535, | |
| "learning_rate": 2.6603325415676963e-06, | |
| "loss": 0.0064, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 2.6721044045677, | |
| "grad_norm": 0.541039764881134, | |
| "learning_rate": 2.422802850356295e-06, | |
| "loss": 0.0068, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.7047308319738987, | |
| "grad_norm": 0.46842265129089355, | |
| "learning_rate": 2.1852731591448932e-06, | |
| "loss": 0.0067, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 2.737357259380098, | |
| "grad_norm": 2.0844099521636963, | |
| "learning_rate": 1.947743467933492e-06, | |
| "loss": 0.0081, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 2.769983686786297, | |
| "grad_norm": 0.40694791078567505, | |
| "learning_rate": 1.7102137767220902e-06, | |
| "loss": 0.005, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.802610114192496, | |
| "grad_norm": 1.159589171409607, | |
| "learning_rate": 1.472684085510689e-06, | |
| "loss": 0.006, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 2.835236541598695, | |
| "grad_norm": 0.2513042688369751, | |
| "learning_rate": 1.2351543942992874e-06, | |
| "loss": 0.0098, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 2.867862969004894, | |
| "grad_norm": 0.3097628951072693, | |
| "learning_rate": 9.976247030878861e-07, | |
| "loss": 0.0075, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 2.9004893964110927, | |
| "grad_norm": 0.3030640780925751, | |
| "learning_rate": 7.600950118764846e-07, | |
| "loss": 0.0075, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 2.933115823817292, | |
| "grad_norm": 0.4447304904460907, | |
| "learning_rate": 5.225653206650831e-07, | |
| "loss": 0.0077, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.933115823817292, | |
| "eval_loss": 0.016508497297763824, | |
| "eval_runtime": 882.4748, | |
| "eval_samples_per_second": 2.777, | |
| "eval_steps_per_second": 0.348, | |
| "eval_wer": 1.0647863328758695, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.965742251223491, | |
| "grad_norm": 0.8142352104187012, | |
| "learning_rate": 2.8503562945368174e-07, | |
| "loss": 0.0088, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 2.99836867862969, | |
| "grad_norm": 1.0330475568771362, | |
| "learning_rate": 4.750593824228029e-08, | |
| "loss": 0.0107, | |
| "step": 920 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 921, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 300, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 8.484410769408e+18, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |