| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 1419, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.021141649048625793, | |
| "grad_norm": 6.24355232060165, | |
| "learning_rate": 7.042253521126762e-07, | |
| "loss": 0.5912, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.042283298097251586, | |
| "grad_norm": 2.216096271272777, | |
| "learning_rate": 1.4084507042253523e-06, | |
| "loss": 0.5024, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.06342494714587738, | |
| "grad_norm": 1.449992996844573, | |
| "learning_rate": 2.1126760563380285e-06, | |
| "loss": 0.4488, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.08456659619450317, | |
| "grad_norm": 1.5865270130811833, | |
| "learning_rate": 2.8169014084507046e-06, | |
| "loss": 0.431, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.10570824524312897, | |
| "grad_norm": 1.6309436803659814, | |
| "learning_rate": 3.5211267605633804e-06, | |
| "loss": 0.4102, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.12684989429175475, | |
| "grad_norm": 1.3774805832619317, | |
| "learning_rate": 4.225352112676057e-06, | |
| "loss": 0.3953, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.14799154334038056, | |
| "grad_norm": 1.7489511540042915, | |
| "learning_rate": 4.929577464788733e-06, | |
| "loss": 0.392, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.16913319238900634, | |
| "grad_norm": 1.4907503413861205, | |
| "learning_rate": 4.999505072506396e-06, | |
| "loss": 0.3845, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.19027484143763213, | |
| "grad_norm": 1.714741595800788, | |
| "learning_rate": 4.997794491573208e-06, | |
| "loss": 0.3754, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.21141649048625794, | |
| "grad_norm": 1.6723995412498653, | |
| "learning_rate": 4.994863075823568e-06, | |
| "loss": 0.3731, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.23255813953488372, | |
| "grad_norm": 1.559234855788911, | |
| "learning_rate": 4.990712417384526e-06, | |
| "loss": 0.3676, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.2536997885835095, | |
| "grad_norm": 1.57054985630799, | |
| "learning_rate": 4.985344770585093e-06, | |
| "loss": 0.3693, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.2748414376321353, | |
| "grad_norm": 1.6051226014645448, | |
| "learning_rate": 4.9787630507318594e-06, | |
| "loss": 0.3627, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2959830866807611, | |
| "grad_norm": 1.4668663144082323, | |
| "learning_rate": 4.9709708325256166e-06, | |
| "loss": 0.3613, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3171247357293869, | |
| "grad_norm": 1.686344018429158, | |
| "learning_rate": 4.961972348119846e-06, | |
| "loss": 0.3602, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3382663847780127, | |
| "grad_norm": 1.7636035584330838, | |
| "learning_rate": 4.951772484822128e-06, | |
| "loss": 0.3568, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.3594080338266385, | |
| "grad_norm": 2.5501698658025536, | |
| "learning_rate": 4.940376782439721e-06, | |
| "loss": 0.3596, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.38054968287526425, | |
| "grad_norm": 1.3764768803172454, | |
| "learning_rate": 4.9277914302707466e-06, | |
| "loss": 0.3574, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.40169133192389006, | |
| "grad_norm": 1.6197573568256791, | |
| "learning_rate": 4.914023263742626e-06, | |
| "loss": 0.3514, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.42283298097251587, | |
| "grad_norm": 1.7518377059862333, | |
| "learning_rate": 4.8990797606995845e-06, | |
| "loss": 0.3468, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.4439746300211416, | |
| "grad_norm": 1.507687286038259, | |
| "learning_rate": 4.882969037341239e-06, | |
| "loss": 0.3455, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.46511627906976744, | |
| "grad_norm": 1.5970818330074303, | |
| "learning_rate": 4.865699843814485e-06, | |
| "loss": 0.3428, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.48625792811839325, | |
| "grad_norm": 1.4210922325164967, | |
| "learning_rate": 4.847281559461071e-06, | |
| "loss": 0.3461, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.507399577167019, | |
| "grad_norm": 1.3912747321946244, | |
| "learning_rate": 4.827724187723427e-06, | |
| "loss": 0.3429, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5285412262156448, | |
| "grad_norm": 1.3955734367464652, | |
| "learning_rate": 4.80703835071155e-06, | |
| "loss": 0.3437, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5496828752642706, | |
| "grad_norm": 1.350657312273654, | |
| "learning_rate": 4.785235283433857e-06, | |
| "loss": 0.3393, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5708245243128964, | |
| "grad_norm": 1.2717206677797315, | |
| "learning_rate": 4.762326827695163e-06, | |
| "loss": 0.3411, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5919661733615222, | |
| "grad_norm": 1.191569222191925, | |
| "learning_rate": 4.7383254256650964e-06, | |
| "loss": 0.3385, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6131078224101479, | |
| "grad_norm": 1.1928472569401323, | |
| "learning_rate": 4.713244113120443e-06, | |
| "loss": 0.34, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6342494714587738, | |
| "grad_norm": 1.3276992158951393, | |
| "learning_rate": 4.687096512365067e-06, | |
| "loss": 0.3421, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6553911205073996, | |
| "grad_norm": 1.1887821846438988, | |
| "learning_rate": 4.659896824831302e-06, | |
| "loss": 0.3356, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6765327695560254, | |
| "grad_norm": 1.2621729084813325, | |
| "learning_rate": 4.631659823366783e-06, | |
| "loss": 0.3403, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6976744186046512, | |
| "grad_norm": 1.4771395957284508, | |
| "learning_rate": 4.6024008442109335e-06, | |
| "loss": 0.3346, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.718816067653277, | |
| "grad_norm": 1.0838005637440695, | |
| "learning_rate": 4.572135778665464e-06, | |
| "loss": 0.3361, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7399577167019028, | |
| "grad_norm": 1.1405549612151125, | |
| "learning_rate": 4.5408810644633956e-06, | |
| "loss": 0.3308, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7610993657505285, | |
| "grad_norm": 1.111444530105349, | |
| "learning_rate": 4.508653676841308e-06, | |
| "loss": 0.3337, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7822410147991543, | |
| "grad_norm": 1.1516361189573001, | |
| "learning_rate": 4.475471119319651e-06, | |
| "loss": 0.3279, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.8033826638477801, | |
| "grad_norm": 1.1394562536949298, | |
| "learning_rate": 4.441351414196139e-06, | |
| "loss": 0.329, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8245243128964059, | |
| "grad_norm": 1.2624616377095017, | |
| "learning_rate": 4.406313092757369e-06, | |
| "loss": 0.3307, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8456659619450317, | |
| "grad_norm": 1.3399500811254008, | |
| "learning_rate": 4.370375185214014e-06, | |
| "loss": 0.3308, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8668076109936576, | |
| "grad_norm": 1.2917888536055562, | |
| "learning_rate": 4.333557210365023e-06, | |
| "loss": 0.3301, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8879492600422833, | |
| "grad_norm": 1.1769502964733598, | |
| "learning_rate": 4.295879164996462e-06, | |
| "loss": 0.3271, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 1.2781090385858154, | |
| "learning_rate": 4.257361513020745e-06, | |
| "loss": 0.3242, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9302325581395349, | |
| "grad_norm": 1.2394021660894758, | |
| "learning_rate": 4.218025174362161e-06, | |
| "loss": 0.3236, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9513742071881607, | |
| "grad_norm": 1.2387248179248829, | |
| "learning_rate": 4.177891513594724e-06, | |
| "loss": 0.3249, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9725158562367865, | |
| "grad_norm": 1.2158882947276164, | |
| "learning_rate": 4.136982328338531e-06, | |
| "loss": 0.3245, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9936575052854123, | |
| "grad_norm": 1.0357890278389636, | |
| "learning_rate": 4.0953198374209045e-06, | |
| "loss": 0.3209, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.3238590359687805, | |
| "eval_runtime": 42.8216, | |
| "eval_samples_per_second": 297.607, | |
| "eval_steps_per_second": 1.168, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 1.014799154334038, | |
| "grad_norm": 1.7559135479092256, | |
| "learning_rate": 4.052926668808791e-06, | |
| "loss": 0.2656, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.0359408033826638, | |
| "grad_norm": 1.5639568698805122, | |
| "learning_rate": 4.009825847318922e-06, | |
| "loss": 0.2439, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.0570824524312896, | |
| "grad_norm": 1.1876279950755768, | |
| "learning_rate": 3.966040782112451e-06, | |
| "loss": 0.2422, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0782241014799154, | |
| "grad_norm": 1.1202067555602988, | |
| "learning_rate": 3.921595253980836e-06, | |
| "loss": 0.2425, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.0993657505285412, | |
| "grad_norm": 1.1539051669533806, | |
| "learning_rate": 3.876513402429883e-06, | |
| "loss": 0.2431, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.120507399577167, | |
| "grad_norm": 1.199763607338339, | |
| "learning_rate": 3.83081971256896e-06, | |
| "loss": 0.2413, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.1416490486257929, | |
| "grad_norm": 1.1190658369259245, | |
| "learning_rate": 3.7845390018125065e-06, | |
| "loss": 0.2454, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.1627906976744187, | |
| "grad_norm": 1.116921299899802, | |
| "learning_rate": 3.737696406401062e-06, | |
| "loss": 0.2394, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.1839323467230445, | |
| "grad_norm": 1.1477296149656728, | |
| "learning_rate": 3.6903173677491266e-06, | |
| "loss": 0.2448, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.20507399577167, | |
| "grad_norm": 1.1454465891047556, | |
| "learning_rate": 3.642427618627277e-06, | |
| "loss": 0.2437, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.226215644820296, | |
| "grad_norm": 1.1642513264879382, | |
| "learning_rate": 3.5940531691860405e-06, | |
| "loss": 0.2428, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.2473572938689217, | |
| "grad_norm": 1.3763131648109288, | |
| "learning_rate": 3.545220292829113e-06, | |
| "loss": 0.2403, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.2684989429175475, | |
| "grad_norm": 1.291772297216049, | |
| "learning_rate": 3.4959555119436033e-06, | |
| "loss": 0.2425, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.2896405919661733, | |
| "grad_norm": 1.4742209496360983, | |
| "learning_rate": 3.446285583495041e-06, | |
| "loss": 0.2419, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.3107822410147991, | |
| "grad_norm": 1.2162630598081021, | |
| "learning_rate": 3.396237484494985e-06, | |
| "loss": 0.2434, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.331923890063425, | |
| "grad_norm": 1.1668893819388124, | |
| "learning_rate": 3.345838397349115e-06, | |
| "loss": 0.2413, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.3530655391120507, | |
| "grad_norm": 1.1883780283974819, | |
| "learning_rate": 3.2951156950937728e-06, | |
| "loss": 0.243, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.3742071881606766, | |
| "grad_norm": 1.2805764821322394, | |
| "learning_rate": 3.2440969265289624e-06, | |
| "loss": 0.2421, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.3953488372093024, | |
| "grad_norm": 1.0727847278011404, | |
| "learning_rate": 3.1928098012558895e-06, | |
| "loss": 0.244, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.4164904862579282, | |
| "grad_norm": 1.1764000750965518, | |
| "learning_rate": 3.1412821746271693e-06, | |
| "loss": 0.2439, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.437632135306554, | |
| "grad_norm": 1.1953048969350113, | |
| "learning_rate": 3.0895420326178705e-06, | |
| "loss": 0.2408, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.4587737843551798, | |
| "grad_norm": 1.2778095821404496, | |
| "learning_rate": 3.03761747662561e-06, | |
| "loss": 0.2423, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.4799154334038054, | |
| "grad_norm": 1.2694623156337455, | |
| "learning_rate": 2.985536708207971e-06, | |
| "loss": 0.2401, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.5010570824524314, | |
| "grad_norm": 1.1185495126584732, | |
| "learning_rate": 2.933328013765505e-06, | |
| "loss": 0.2449, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.522198731501057, | |
| "grad_norm": 1.168267493603954, | |
| "learning_rate": 2.8810197491786714e-06, | |
| "loss": 0.2431, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.543340380549683, | |
| "grad_norm": 1.1355953180673453, | |
| "learning_rate": 2.8286403244070252e-06, | |
| "loss": 0.2414, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.5644820295983086, | |
| "grad_norm": 1.0758782490829244, | |
| "learning_rate": 2.7762181880590442e-06, | |
| "loss": 0.2415, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.5856236786469344, | |
| "grad_norm": 1.1428398992616515, | |
| "learning_rate": 2.723781811940956e-06, | |
| "loss": 0.2431, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.6067653276955602, | |
| "grad_norm": 1.0783252394855087, | |
| "learning_rate": 2.6713596755929755e-06, | |
| "loss": 0.2417, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.627906976744186, | |
| "grad_norm": 1.0764260279790863, | |
| "learning_rate": 2.6189802508213297e-06, | |
| "loss": 0.2391, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.6490486257928119, | |
| "grad_norm": 1.0866425685585581, | |
| "learning_rate": 2.5666719862344957e-06, | |
| "loss": 0.2445, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.6701902748414377, | |
| "grad_norm": 1.0529013290697624, | |
| "learning_rate": 2.51446329179203e-06, | |
| "loss": 0.2409, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.6913319238900635, | |
| "grad_norm": 1.1356539893062763, | |
| "learning_rate": 2.4623825233743896e-06, | |
| "loss": 0.241, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.712473572938689, | |
| "grad_norm": 1.1111512278149636, | |
| "learning_rate": 2.4104579673821297e-06, | |
| "loss": 0.2434, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.733615221987315, | |
| "grad_norm": 1.102236029943368, | |
| "learning_rate": 2.358717825372831e-06, | |
| "loss": 0.2413, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.7547568710359407, | |
| "grad_norm": 1.124459202450407, | |
| "learning_rate": 2.3071901987441116e-06, | |
| "loss": 0.2385, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.7758985200845667, | |
| "grad_norm": 1.2179881959913421, | |
| "learning_rate": 2.2559030734710396e-06, | |
| "loss": 0.2386, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.7970401691331923, | |
| "grad_norm": 1.0767260671283274, | |
| "learning_rate": 2.2048843049062275e-06, | |
| "loss": 0.2398, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 1.1119843325482643, | |
| "learning_rate": 2.1541616026508854e-06, | |
| "loss": 0.2413, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.839323467230444, | |
| "grad_norm": 1.0955361377120234, | |
| "learning_rate": 2.103762515505016e-06, | |
| "loss": 0.2417, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.8604651162790697, | |
| "grad_norm": 1.0754774699879763, | |
| "learning_rate": 2.0537144165049597e-06, | |
| "loss": 0.236, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.8816067653276956, | |
| "grad_norm": 1.116290127791819, | |
| "learning_rate": 2.0040444880563974e-06, | |
| "loss": 0.2396, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.9027484143763214, | |
| "grad_norm": 1.0587161825711324, | |
| "learning_rate": 1.9547797071708873e-06, | |
| "loss": 0.2402, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.9238900634249472, | |
| "grad_norm": 1.0735022958430327, | |
| "learning_rate": 1.90594683081396e-06, | |
| "loss": 0.2454, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.945031712473573, | |
| "grad_norm": 1.0625353794584775, | |
| "learning_rate": 1.8575723813727234e-06, | |
| "loss": 0.241, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.9661733615221988, | |
| "grad_norm": 1.0540753292992686, | |
| "learning_rate": 1.8096826322508745e-06, | |
| "loss": 0.2377, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.9873150105708244, | |
| "grad_norm": 1.1469855369418602, | |
| "learning_rate": 1.7623035935989388e-06, | |
| "loss": 0.2363, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.3176339268684387, | |
| "eval_runtime": 46.7927, | |
| "eval_samples_per_second": 272.35, | |
| "eval_steps_per_second": 1.069, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 2.0084566596194504, | |
| "grad_norm": 1.442414950660575, | |
| "learning_rate": 1.7154609981874945e-06, | |
| "loss": 0.2124, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 2.029598308668076, | |
| "grad_norm": 1.3377936728948667, | |
| "learning_rate": 1.6691802874310402e-06, | |
| "loss": 0.1655, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 2.050739957716702, | |
| "grad_norm": 1.3199120792646895, | |
| "learning_rate": 1.6234865975701169e-06, | |
| "loss": 0.1695, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 2.0718816067653276, | |
| "grad_norm": 1.1515710499363918, | |
| "learning_rate": 1.5784047460191638e-06, | |
| "loss": 0.1628, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 2.0930232558139537, | |
| "grad_norm": 1.0930247346459778, | |
| "learning_rate": 1.5339592178875502e-06, | |
| "loss": 0.1629, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 2.1141649048625792, | |
| "grad_norm": 1.1196935797630683, | |
| "learning_rate": 1.4901741526810793e-06, | |
| "loss": 0.1632, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.1353065539112053, | |
| "grad_norm": 1.266477815888195, | |
| "learning_rate": 1.4470733311912094e-06, | |
| "loss": 0.1648, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 2.156448202959831, | |
| "grad_norm": 1.2032366464614765, | |
| "learning_rate": 1.4046801625790954e-06, | |
| "loss": 0.1635, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 2.177589852008457, | |
| "grad_norm": 1.1118635376846853, | |
| "learning_rate": 1.3630176716614696e-06, | |
| "loss": 0.1645, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 2.1987315010570825, | |
| "grad_norm": 1.1762161662179742, | |
| "learning_rate": 1.3221084864052755e-06, | |
| "loss": 0.167, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 2.219873150105708, | |
| "grad_norm": 1.148476335813781, | |
| "learning_rate": 1.2819748256378395e-06, | |
| "loss": 0.1621, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 2.241014799154334, | |
| "grad_norm": 1.1550786530748482, | |
| "learning_rate": 1.2426384869792552e-06, | |
| "loss": 0.1626, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 2.2621564482029597, | |
| "grad_norm": 1.1566446795682568, | |
| "learning_rate": 1.2041208350035386e-06, | |
| "loss": 0.1642, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 2.2832980972515857, | |
| "grad_norm": 1.1410532150519583, | |
| "learning_rate": 1.1664427896349775e-06, | |
| "loss": 0.1642, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 2.3044397463002113, | |
| "grad_norm": 1.1400713733613699, | |
| "learning_rate": 1.129624814785987e-06, | |
| "loss": 0.1647, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 2.3255813953488373, | |
| "grad_norm": 1.0699436865402963, | |
| "learning_rate": 1.0936869072426324e-06, | |
| "loss": 0.163, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.346723044397463, | |
| "grad_norm": 1.1541436695499816, | |
| "learning_rate": 1.0586485858038623e-06, | |
| "loss": 0.1626, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 2.367864693446089, | |
| "grad_norm": 1.125483674986556, | |
| "learning_rate": 1.0245288806803492e-06, | |
| "loss": 0.1669, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 2.3890063424947146, | |
| "grad_norm": 1.1026248670403718, | |
| "learning_rate": 9.913463231586928e-07, | |
| "loss": 0.1629, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 2.41014799154334, | |
| "grad_norm": 1.159203695893513, | |
| "learning_rate": 9.591189355366054e-07, | |
| "loss": 0.1644, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 2.431289640591966, | |
| "grad_norm": 1.1077004937863641, | |
| "learning_rate": 9.278642213345369e-07, | |
| "loss": 0.167, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 2.452431289640592, | |
| "grad_norm": 1.11182862218811, | |
| "learning_rate": 8.975991557890668e-07, | |
| "loss": 0.1638, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 2.473572938689218, | |
| "grad_norm": 1.1755110000283455, | |
| "learning_rate": 8.683401766332171e-07, | |
| "loss": 0.1642, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 2.4947145877378434, | |
| "grad_norm": 1.1475624161633082, | |
| "learning_rate": 8.401031751686978e-07, | |
| "loss": 0.1676, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 2.5158562367864694, | |
| "grad_norm": 1.148787879127806, | |
| "learning_rate": 8.129034876349334e-07, | |
| "loss": 0.1644, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 2.536997885835095, | |
| "grad_norm": 1.190372203322852, | |
| "learning_rate": 7.867558868795578e-07, | |
| "loss": 0.165, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.558139534883721, | |
| "grad_norm": 1.1595867121561987, | |
| "learning_rate": 7.616745743349038e-07, | |
| "loss": 0.1638, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 2.5792811839323466, | |
| "grad_norm": 1.1432214797090672, | |
| "learning_rate": 7.376731723048383e-07, | |
| "loss": 0.1652, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 2.6004228329809727, | |
| "grad_norm": 1.1142453334609799, | |
| "learning_rate": 7.147647165661439e-07, | |
| "loss": 0.1636, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 2.6215644820295982, | |
| "grad_norm": 1.0990743293486216, | |
| "learning_rate": 6.929616492884497e-07, | |
| "loss": 0.1625, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 2.6427061310782243, | |
| "grad_norm": 1.0991366585298548, | |
| "learning_rate": 6.722758122765728e-07, | |
| "loss": 0.1638, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.66384778012685, | |
| "grad_norm": 1.1273149039763855, | |
| "learning_rate": 6.527184405389298e-07, | |
| "loss": 0.1611, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 2.6849894291754755, | |
| "grad_norm": 1.101167246100859, | |
| "learning_rate": 6.343001561855149e-07, | |
| "loss": 0.1605, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 2.7061310782241015, | |
| "grad_norm": 1.0867484989228724, | |
| "learning_rate": 6.17030962658762e-07, | |
| "loss": 0.1655, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 2.7272727272727275, | |
| "grad_norm": 1.154566269426987, | |
| "learning_rate": 6.009202393004164e-07, | |
| "loss": 0.1636, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 2.748414376321353, | |
| "grad_norm": 1.1187041716958608, | |
| "learning_rate": 5.859767362573741e-07, | |
| "loss": 0.1644, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.7695560253699787, | |
| "grad_norm": 1.1267150964681174, | |
| "learning_rate": 5.722085697292536e-07, | |
| "loss": 0.1659, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 2.7906976744186047, | |
| "grad_norm": 1.0795591650211436, | |
| "learning_rate": 5.596232175602791e-07, | |
| "loss": 0.1667, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 2.8118393234672303, | |
| "grad_norm": 1.060071714533849, | |
| "learning_rate": 5.482275151778719e-07, | |
| "loss": 0.164, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 2.8329809725158563, | |
| "grad_norm": 1.10306526157592, | |
| "learning_rate": 5.380276518801548e-07, | |
| "loss": 0.1632, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 2.854122621564482, | |
| "grad_norm": 1.0906207707452118, | |
| "learning_rate": 5.290291674743844e-07, | |
| "loss": 0.1642, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.875264270613108, | |
| "grad_norm": 1.0845536549774597, | |
| "learning_rate": 5.212369492681416e-07, | |
| "loss": 0.1641, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 2.8964059196617336, | |
| "grad_norm": 1.1569339067127147, | |
| "learning_rate": 5.146552294149075e-07, | |
| "loss": 0.1629, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 2.9175475687103596, | |
| "grad_norm": 1.064398221843151, | |
| "learning_rate": 5.092875826154744e-07, | |
| "loss": 0.1648, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 2.938689217758985, | |
| "grad_norm": 1.093531486342711, | |
| "learning_rate": 5.051369241764327e-07, | |
| "loss": 0.1621, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 2.9598308668076108, | |
| "grad_norm": 1.0796661303223472, | |
| "learning_rate": 5.022055084267933e-07, | |
| "loss": 0.1623, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.980972515856237, | |
| "grad_norm": 1.0990164115583458, | |
| "learning_rate": 5.004949274936044e-07, | |
| "loss": 0.1634, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.3471969962120056, | |
| "eval_runtime": 49.2029, | |
| "eval_samples_per_second": 259.009, | |
| "eval_steps_per_second": 1.016, | |
| "step": 1419 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 1419, | |
| "total_flos": 2376877851279360.0, | |
| "train_loss": 0.25528628041493884, | |
| "train_runtime": 8757.0427, | |
| "train_samples_per_second": 82.948, | |
| "train_steps_per_second": 0.162 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1419, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2376877851279360.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |