| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.49997953419835456, | |
| "eval_steps": 4886, | |
| "global_step": 24430, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0009823584789816217, | |
| "grad_norm": 640.0, | |
| "learning_rate": 0.001, | |
| "loss": 11885.6341, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.0019647169579632435, | |
| "grad_norm": 454.0, | |
| "learning_rate": 0.001, | |
| "loss": 8966.3691, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.0029470754369448652, | |
| "grad_norm": 572.0, | |
| "learning_rate": 0.001, | |
| "loss": 7738.5072, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.003929433915926487, | |
| "grad_norm": 636.0, | |
| "learning_rate": 0.001, | |
| "loss": 7036.127, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.004911792394908109, | |
| "grad_norm": 748.0, | |
| "learning_rate": 0.001, | |
| "loss": 6540.5202, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.0058941508738897305, | |
| "grad_norm": 556.0, | |
| "learning_rate": 0.001, | |
| "loss": 6263.0488, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.006876509352871352, | |
| "grad_norm": 532.0, | |
| "learning_rate": 0.001, | |
| "loss": 5950.1823, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.007858867831852974, | |
| "grad_norm": 928.0, | |
| "learning_rate": 0.001, | |
| "loss": 5705.2292, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.008841226310834595, | |
| "grad_norm": 444.0, | |
| "learning_rate": 0.001, | |
| "loss": 5496.4583, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.009823584789816217, | |
| "grad_norm": 656.0, | |
| "learning_rate": 0.001, | |
| "loss": 5272.5752, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.010805943268797838, | |
| "grad_norm": 612.0, | |
| "learning_rate": 0.001, | |
| "loss": 5051.2663, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 0.011788301747779461, | |
| "grad_norm": 608.0, | |
| "learning_rate": 0.001, | |
| "loss": 4938.0895, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 0.012770660226761082, | |
| "grad_norm": 466.0, | |
| "learning_rate": 0.001, | |
| "loss": 4740.762, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 0.013753018705742704, | |
| "grad_norm": 438.0, | |
| "learning_rate": 0.001, | |
| "loss": 4573.4443, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 0.014735377184724325, | |
| "grad_norm": 696.0, | |
| "learning_rate": 0.001, | |
| "loss": 4539.9521, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.015717735663705948, | |
| "grad_norm": 892.0, | |
| "learning_rate": 0.001, | |
| "loss": 4435.4001, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 0.01670009414268757, | |
| "grad_norm": 704.0, | |
| "learning_rate": 0.001, | |
| "loss": 4239.0426, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 0.01768245262166919, | |
| "grad_norm": 548.0, | |
| "learning_rate": 0.001, | |
| "loss": 4189.9281, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 0.018664811100650814, | |
| "grad_norm": 540.0, | |
| "learning_rate": 0.001, | |
| "loss": 4104.7835, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 0.019647169579632435, | |
| "grad_norm": 592.0, | |
| "learning_rate": 0.001, | |
| "loss": 4044.3369, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.020629528058614056, | |
| "grad_norm": 536.0, | |
| "learning_rate": 0.001, | |
| "loss": 3936.5283, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 0.021611886537595677, | |
| "grad_norm": 604.0, | |
| "learning_rate": 0.001, | |
| "loss": 3915.6911, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 0.0225942450165773, | |
| "grad_norm": 458.0, | |
| "learning_rate": 0.001, | |
| "loss": 3759.7747, | |
| "step": 1104 | |
| }, | |
| { | |
| "epoch": 0.023576603495558922, | |
| "grad_norm": 636.0, | |
| "learning_rate": 0.001, | |
| "loss": 3760.4476, | |
| "step": 1152 | |
| }, | |
| { | |
| "epoch": 0.024558961974540543, | |
| "grad_norm": 672.0, | |
| "learning_rate": 0.001, | |
| "loss": 3672.9059, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.025541320453522164, | |
| "grad_norm": 592.0, | |
| "learning_rate": 0.001, | |
| "loss": 3645.0697, | |
| "step": 1248 | |
| }, | |
| { | |
| "epoch": 0.026523678932503784, | |
| "grad_norm": 552.0, | |
| "learning_rate": 0.001, | |
| "loss": 3528.9896, | |
| "step": 1296 | |
| }, | |
| { | |
| "epoch": 0.02750603741148541, | |
| "grad_norm": 470.0, | |
| "learning_rate": 0.001, | |
| "loss": 3488.8187, | |
| "step": 1344 | |
| }, | |
| { | |
| "epoch": 0.02848839589046703, | |
| "grad_norm": 580.0, | |
| "learning_rate": 0.001, | |
| "loss": 3466.627, | |
| "step": 1392 | |
| }, | |
| { | |
| "epoch": 0.02947075436944865, | |
| "grad_norm": 584.0, | |
| "learning_rate": 0.001, | |
| "loss": 3399.1475, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.03045311284843027, | |
| "grad_norm": 732.0, | |
| "learning_rate": 0.001, | |
| "loss": 3363.9762, | |
| "step": 1488 | |
| }, | |
| { | |
| "epoch": 0.031435471327411896, | |
| "grad_norm": 616.0, | |
| "learning_rate": 0.001, | |
| "loss": 3337.4564, | |
| "step": 1536 | |
| }, | |
| { | |
| "epoch": 0.03241782980639352, | |
| "grad_norm": 540.0, | |
| "learning_rate": 0.001, | |
| "loss": 3298.4583, | |
| "step": 1584 | |
| }, | |
| { | |
| "epoch": 0.03340018828537514, | |
| "grad_norm": 512.0, | |
| "learning_rate": 0.001, | |
| "loss": 3212.2949, | |
| "step": 1632 | |
| }, | |
| { | |
| "epoch": 0.03438254676435676, | |
| "grad_norm": 512.0, | |
| "learning_rate": 0.001, | |
| "loss": 3217.6631, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.03536490524333838, | |
| "grad_norm": 624.0, | |
| "learning_rate": 0.001, | |
| "loss": 3175.7318, | |
| "step": 1728 | |
| }, | |
| { | |
| "epoch": 0.03634726372232, | |
| "grad_norm": 520.0, | |
| "learning_rate": 0.001, | |
| "loss": 3140.3923, | |
| "step": 1776 | |
| }, | |
| { | |
| "epoch": 0.03732962220130163, | |
| "grad_norm": 664.0, | |
| "learning_rate": 0.001, | |
| "loss": 3099.8044, | |
| "step": 1824 | |
| }, | |
| { | |
| "epoch": 0.03831198068028325, | |
| "grad_norm": 604.0, | |
| "learning_rate": 0.001, | |
| "loss": 3114.8079, | |
| "step": 1872 | |
| }, | |
| { | |
| "epoch": 0.03929433915926487, | |
| "grad_norm": 478.0, | |
| "learning_rate": 0.001, | |
| "loss": 3048.9001, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.04027669763824649, | |
| "grad_norm": 520.0, | |
| "learning_rate": 0.001, | |
| "loss": 3018.8714, | |
| "step": 1968 | |
| }, | |
| { | |
| "epoch": 0.04125905611722811, | |
| "grad_norm": 456.0, | |
| "learning_rate": 0.001, | |
| "loss": 2981.1152, | |
| "step": 2016 | |
| }, | |
| { | |
| "epoch": 0.04224141459620973, | |
| "grad_norm": 628.0, | |
| "learning_rate": 0.001, | |
| "loss": 2999.249, | |
| "step": 2064 | |
| }, | |
| { | |
| "epoch": 0.04322377307519135, | |
| "grad_norm": 804.0, | |
| "learning_rate": 0.001, | |
| "loss": 2942.3376, | |
| "step": 2112 | |
| }, | |
| { | |
| "epoch": 0.044206131554172974, | |
| "grad_norm": 600.0, | |
| "learning_rate": 0.001, | |
| "loss": 2890.7354, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 0.0451884900331546, | |
| "grad_norm": 632.0, | |
| "learning_rate": 0.001, | |
| "loss": 2896.4242, | |
| "step": 2208 | |
| }, | |
| { | |
| "epoch": 0.04617084851213622, | |
| "grad_norm": 536.0, | |
| "learning_rate": 0.001, | |
| "loss": 2874.8643, | |
| "step": 2256 | |
| }, | |
| { | |
| "epoch": 0.047153206991117844, | |
| "grad_norm": 494.0, | |
| "learning_rate": 0.001, | |
| "loss": 2807.6911, | |
| "step": 2304 | |
| }, | |
| { | |
| "epoch": 0.048135565470099465, | |
| "grad_norm": 548.0, | |
| "learning_rate": 0.001, | |
| "loss": 2820.04, | |
| "step": 2352 | |
| }, | |
| { | |
| "epoch": 0.049117923949081085, | |
| "grad_norm": 636.0, | |
| "learning_rate": 0.001, | |
| "loss": 2787.0247, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.050100282428062706, | |
| "grad_norm": 624.0, | |
| "learning_rate": 0.001, | |
| "loss": 2782.2428, | |
| "step": 2448 | |
| }, | |
| { | |
| "epoch": 0.05108264090704433, | |
| "grad_norm": 628.0, | |
| "learning_rate": 0.001, | |
| "loss": 2725.3781, | |
| "step": 2496 | |
| }, | |
| { | |
| "epoch": 0.05206499938602595, | |
| "grad_norm": 540.0, | |
| "learning_rate": 0.001, | |
| "loss": 2755.7458, | |
| "step": 2544 | |
| }, | |
| { | |
| "epoch": 0.05304735786500757, | |
| "grad_norm": 568.0, | |
| "learning_rate": 0.001, | |
| "loss": 2699.16, | |
| "step": 2592 | |
| }, | |
| { | |
| "epoch": 0.0540297163439892, | |
| "grad_norm": 808.0, | |
| "learning_rate": 0.001, | |
| "loss": 2680.3232, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 0.05501207482297082, | |
| "grad_norm": 564.0, | |
| "learning_rate": 0.001, | |
| "loss": 2669.6646, | |
| "step": 2688 | |
| }, | |
| { | |
| "epoch": 0.05599443330195244, | |
| "grad_norm": 552.0, | |
| "learning_rate": 0.001, | |
| "loss": 2683.8433, | |
| "step": 2736 | |
| }, | |
| { | |
| "epoch": 0.05697679178093406, | |
| "grad_norm": 636.0, | |
| "learning_rate": 0.001, | |
| "loss": 2643.8172, | |
| "step": 2784 | |
| }, | |
| { | |
| "epoch": 0.05795915025991568, | |
| "grad_norm": 580.0, | |
| "learning_rate": 0.001, | |
| "loss": 2649.0441, | |
| "step": 2832 | |
| }, | |
| { | |
| "epoch": 0.0589415087388973, | |
| "grad_norm": 512.0, | |
| "learning_rate": 0.001, | |
| "loss": 2615.9657, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 0.05992386721787892, | |
| "grad_norm": 688.0, | |
| "learning_rate": 0.001, | |
| "loss": 2608.1457, | |
| "step": 2928 | |
| }, | |
| { | |
| "epoch": 0.06090622569686054, | |
| "grad_norm": 696.0, | |
| "learning_rate": 0.001, | |
| "loss": 2590.1567, | |
| "step": 2976 | |
| }, | |
| { | |
| "epoch": 0.06188858417584217, | |
| "grad_norm": 704.0, | |
| "learning_rate": 0.001, | |
| "loss": 2627.8358, | |
| "step": 3024 | |
| }, | |
| { | |
| "epoch": 0.06287094265482379, | |
| "grad_norm": 704.0, | |
| "learning_rate": 0.001, | |
| "loss": 2538.9543, | |
| "step": 3072 | |
| }, | |
| { | |
| "epoch": 0.06385330113380541, | |
| "grad_norm": 568.0, | |
| "learning_rate": 0.001, | |
| "loss": 2482.9673, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 0.06483565961278703, | |
| "grad_norm": 548.0, | |
| "learning_rate": 0.001, | |
| "loss": 2530.4771, | |
| "step": 3168 | |
| }, | |
| { | |
| "epoch": 0.06581801809176865, | |
| "grad_norm": 456.0, | |
| "learning_rate": 0.001, | |
| "loss": 2496.41, | |
| "step": 3216 | |
| }, | |
| { | |
| "epoch": 0.06680037657075028, | |
| "grad_norm": 684.0, | |
| "learning_rate": 0.001, | |
| "loss": 2518.8866, | |
| "step": 3264 | |
| }, | |
| { | |
| "epoch": 0.0677827350497319, | |
| "grad_norm": 664.0, | |
| "learning_rate": 0.001, | |
| "loss": 2475.0793, | |
| "step": 3312 | |
| }, | |
| { | |
| "epoch": 0.06876509352871352, | |
| "grad_norm": 812.0, | |
| "learning_rate": 0.001, | |
| "loss": 2461.3527, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 0.06974745200769514, | |
| "grad_norm": 490.0, | |
| "learning_rate": 0.001, | |
| "loss": 2467.4508, | |
| "step": 3408 | |
| }, | |
| { | |
| "epoch": 0.07072981048667676, | |
| "grad_norm": 648.0, | |
| "learning_rate": 0.001, | |
| "loss": 2443.8037, | |
| "step": 3456 | |
| }, | |
| { | |
| "epoch": 0.07171216896565838, | |
| "grad_norm": 664.0, | |
| "learning_rate": 0.001, | |
| "loss": 2445.9336, | |
| "step": 3504 | |
| }, | |
| { | |
| "epoch": 0.07269452744464, | |
| "grad_norm": 524.0, | |
| "learning_rate": 0.001, | |
| "loss": 2411.4482, | |
| "step": 3552 | |
| }, | |
| { | |
| "epoch": 0.07367688592362164, | |
| "grad_norm": 608.0, | |
| "learning_rate": 0.001, | |
| "loss": 2417.4673, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.07465924440260326, | |
| "grad_norm": 504.0, | |
| "learning_rate": 0.001, | |
| "loss": 2420.4196, | |
| "step": 3648 | |
| }, | |
| { | |
| "epoch": 0.07564160288158488, | |
| "grad_norm": 564.0, | |
| "learning_rate": 0.001, | |
| "loss": 2390.8983, | |
| "step": 3696 | |
| }, | |
| { | |
| "epoch": 0.0766239613605665, | |
| "grad_norm": 664.0, | |
| "learning_rate": 0.001, | |
| "loss": 2377.8607, | |
| "step": 3744 | |
| }, | |
| { | |
| "epoch": 0.07760631983954812, | |
| "grad_norm": 616.0, | |
| "learning_rate": 0.001, | |
| "loss": 2359.1242, | |
| "step": 3792 | |
| }, | |
| { | |
| "epoch": 0.07858867831852974, | |
| "grad_norm": 636.0, | |
| "learning_rate": 0.001, | |
| "loss": 2385.3102, | |
| "step": 3840 | |
| }, | |
| { | |
| "epoch": 0.07957103679751136, | |
| "grad_norm": 454.0, | |
| "learning_rate": 0.001, | |
| "loss": 2373.0225, | |
| "step": 3888 | |
| }, | |
| { | |
| "epoch": 0.08055339527649298, | |
| "grad_norm": 502.0, | |
| "learning_rate": 0.001, | |
| "loss": 2361.2386, | |
| "step": 3936 | |
| }, | |
| { | |
| "epoch": 0.0815357537554746, | |
| "grad_norm": 506.0, | |
| "learning_rate": 0.001, | |
| "loss": 2341.1328, | |
| "step": 3984 | |
| }, | |
| { | |
| "epoch": 0.08251811223445622, | |
| "grad_norm": 472.0, | |
| "learning_rate": 0.001, | |
| "loss": 2308.069, | |
| "step": 4032 | |
| }, | |
| { | |
| "epoch": 0.08350047071343784, | |
| "grad_norm": 472.0, | |
| "learning_rate": 0.001, | |
| "loss": 2305.2542, | |
| "step": 4080 | |
| }, | |
| { | |
| "epoch": 0.08448282919241946, | |
| "grad_norm": 502.0, | |
| "learning_rate": 0.001, | |
| "loss": 2338.4048, | |
| "step": 4128 | |
| }, | |
| { | |
| "epoch": 0.08546518767140109, | |
| "grad_norm": 628.0, | |
| "learning_rate": 0.001, | |
| "loss": 2307.96, | |
| "step": 4176 | |
| }, | |
| { | |
| "epoch": 0.0864475461503827, | |
| "grad_norm": 516.0, | |
| "learning_rate": 0.001, | |
| "loss": 2314.313, | |
| "step": 4224 | |
| }, | |
| { | |
| "epoch": 0.08742990462936433, | |
| "grad_norm": 748.0, | |
| "learning_rate": 0.001, | |
| "loss": 2270.4917, | |
| "step": 4272 | |
| }, | |
| { | |
| "epoch": 0.08841226310834595, | |
| "grad_norm": 576.0, | |
| "learning_rate": 0.001, | |
| "loss": 2292.9497, | |
| "step": 4320 | |
| }, | |
| { | |
| "epoch": 0.08939462158732757, | |
| "grad_norm": 604.0, | |
| "learning_rate": 0.001, | |
| "loss": 2274.6584, | |
| "step": 4368 | |
| }, | |
| { | |
| "epoch": 0.0903769800663092, | |
| "grad_norm": 580.0, | |
| "learning_rate": 0.001, | |
| "loss": 2275.2266, | |
| "step": 4416 | |
| }, | |
| { | |
| "epoch": 0.09135933854529082, | |
| "grad_norm": 548.0, | |
| "learning_rate": 0.001, | |
| "loss": 2262.757, | |
| "step": 4464 | |
| }, | |
| { | |
| "epoch": 0.09234169702427245, | |
| "grad_norm": 628.0, | |
| "learning_rate": 0.001, | |
| "loss": 2257.687, | |
| "step": 4512 | |
| }, | |
| { | |
| "epoch": 0.09332405550325407, | |
| "grad_norm": 544.0, | |
| "learning_rate": 0.001, | |
| "loss": 2259.9118, | |
| "step": 4560 | |
| }, | |
| { | |
| "epoch": 0.09430641398223569, | |
| "grad_norm": 580.0, | |
| "learning_rate": 0.001, | |
| "loss": 2224.4427, | |
| "step": 4608 | |
| }, | |
| { | |
| "epoch": 0.09528877246121731, | |
| "grad_norm": 628.0, | |
| "learning_rate": 0.001, | |
| "loss": 2248.397, | |
| "step": 4656 | |
| }, | |
| { | |
| "epoch": 0.09627113094019893, | |
| "grad_norm": 600.0, | |
| "learning_rate": 0.001, | |
| "loss": 2203.2843, | |
| "step": 4704 | |
| }, | |
| { | |
| "epoch": 0.09725348941918055, | |
| "grad_norm": 780.0, | |
| "learning_rate": 0.001, | |
| "loss": 2223.5656, | |
| "step": 4752 | |
| }, | |
| { | |
| "epoch": 0.09823584789816217, | |
| "grad_norm": 536.0, | |
| "learning_rate": 0.001, | |
| "loss": 2169.4321, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.09921820637714379, | |
| "grad_norm": 516.0, | |
| "learning_rate": 0.001, | |
| "loss": 2183.4987, | |
| "step": 4848 | |
| }, | |
| { | |
| "epoch": 0.09999590683967091, | |
| "eval_loss": 2087.82763671875, | |
| "eval_runtime": 9.0001, | |
| "eval_samples_per_second": 111.11, | |
| "eval_steps_per_second": 1.444, | |
| "step": 4886 | |
| }, | |
| { | |
| "epoch": 0.10020056485612541, | |
| "grad_norm": 620.0, | |
| "learning_rate": 0.001, | |
| "loss": 2210.3151, | |
| "step": 4896 | |
| }, | |
| { | |
| "epoch": 0.10118292333510703, | |
| "grad_norm": 616.0, | |
| "learning_rate": 0.001, | |
| "loss": 2208.1779, | |
| "step": 4944 | |
| }, | |
| { | |
| "epoch": 0.10216528181408865, | |
| "grad_norm": 592.0, | |
| "learning_rate": 0.001, | |
| "loss": 2166.7116, | |
| "step": 4992 | |
| }, | |
| { | |
| "epoch": 0.10314764029307028, | |
| "grad_norm": 596.0, | |
| "learning_rate": 0.001, | |
| "loss": 2191.4295, | |
| "step": 5040 | |
| }, | |
| { | |
| "epoch": 0.1041299987720519, | |
| "grad_norm": 684.0, | |
| "learning_rate": 0.001, | |
| "loss": 2155.1141, | |
| "step": 5088 | |
| }, | |
| { | |
| "epoch": 0.10511235725103352, | |
| "grad_norm": 512.0, | |
| "learning_rate": 0.001, | |
| "loss": 2135.7635, | |
| "step": 5136 | |
| }, | |
| { | |
| "epoch": 0.10609471573001514, | |
| "grad_norm": 506.0, | |
| "learning_rate": 0.001, | |
| "loss": 2155.5701, | |
| "step": 5184 | |
| }, | |
| { | |
| "epoch": 0.10707707420899677, | |
| "grad_norm": 480.0, | |
| "learning_rate": 0.001, | |
| "loss": 2150.0086, | |
| "step": 5232 | |
| }, | |
| { | |
| "epoch": 0.1080594326879784, | |
| "grad_norm": 540.0, | |
| "learning_rate": 0.001, | |
| "loss": 2142.4181, | |
| "step": 5280 | |
| }, | |
| { | |
| "epoch": 0.10904179116696001, | |
| "grad_norm": 572.0, | |
| "learning_rate": 0.001, | |
| "loss": 2116.3011, | |
| "step": 5328 | |
| }, | |
| { | |
| "epoch": 0.11002414964594164, | |
| "grad_norm": 548.0, | |
| "learning_rate": 0.001, | |
| "loss": 2141.0239, | |
| "step": 5376 | |
| }, | |
| { | |
| "epoch": 0.11100650812492326, | |
| "grad_norm": 676.0, | |
| "learning_rate": 0.001, | |
| "loss": 2119.1307, | |
| "step": 5424 | |
| }, | |
| { | |
| "epoch": 0.11198886660390488, | |
| "grad_norm": 656.0, | |
| "learning_rate": 0.001, | |
| "loss": 2137.8016, | |
| "step": 5472 | |
| }, | |
| { | |
| "epoch": 0.1129712250828865, | |
| "grad_norm": 676.0, | |
| "learning_rate": 0.001, | |
| "loss": 2119.2923, | |
| "step": 5520 | |
| }, | |
| { | |
| "epoch": 0.11395358356186812, | |
| "grad_norm": 588.0, | |
| "learning_rate": 0.001, | |
| "loss": 2120.9912, | |
| "step": 5568 | |
| }, | |
| { | |
| "epoch": 0.11493594204084974, | |
| "grad_norm": 612.0, | |
| "learning_rate": 0.001, | |
| "loss": 2111.5037, | |
| "step": 5616 | |
| }, | |
| { | |
| "epoch": 0.11591830051983136, | |
| "grad_norm": 588.0, | |
| "learning_rate": 0.001, | |
| "loss": 2119.6444, | |
| "step": 5664 | |
| }, | |
| { | |
| "epoch": 0.11690065899881298, | |
| "grad_norm": 700.0, | |
| "learning_rate": 0.001, | |
| "loss": 2078.1807, | |
| "step": 5712 | |
| }, | |
| { | |
| "epoch": 0.1178830174777946, | |
| "grad_norm": 564.0, | |
| "learning_rate": 0.001, | |
| "loss": 2095.8706, | |
| "step": 5760 | |
| }, | |
| { | |
| "epoch": 0.11886537595677622, | |
| "grad_norm": 552.0, | |
| "learning_rate": 0.001, | |
| "loss": 2080.8527, | |
| "step": 5808 | |
| }, | |
| { | |
| "epoch": 0.11984773443575784, | |
| "grad_norm": 488.0, | |
| "learning_rate": 0.001, | |
| "loss": 2062.9159, | |
| "step": 5856 | |
| }, | |
| { | |
| "epoch": 0.12083009291473946, | |
| "grad_norm": 616.0, | |
| "learning_rate": 0.001, | |
| "loss": 2060.964, | |
| "step": 5904 | |
| }, | |
| { | |
| "epoch": 0.12181245139372109, | |
| "grad_norm": 648.0, | |
| "learning_rate": 0.001, | |
| "loss": 2088.8507, | |
| "step": 5952 | |
| }, | |
| { | |
| "epoch": 0.12279480987270272, | |
| "grad_norm": 604.0, | |
| "learning_rate": 0.001, | |
| "loss": 2052.1393, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.12377716835168434, | |
| "grad_norm": 720.0, | |
| "learning_rate": 0.001, | |
| "loss": 2043.2277, | |
| "step": 6048 | |
| }, | |
| { | |
| "epoch": 0.12475952683066596, | |
| "grad_norm": 616.0, | |
| "learning_rate": 0.001, | |
| "loss": 2043.3983, | |
| "step": 6096 | |
| }, | |
| { | |
| "epoch": 0.12574188530964758, | |
| "grad_norm": 668.0, | |
| "learning_rate": 0.001, | |
| "loss": 2080.6297, | |
| "step": 6144 | |
| }, | |
| { | |
| "epoch": 0.1267242437886292, | |
| "grad_norm": 532.0, | |
| "learning_rate": 0.001, | |
| "loss": 2059.5207, | |
| "step": 6192 | |
| }, | |
| { | |
| "epoch": 0.12770660226761082, | |
| "grad_norm": 568.0, | |
| "learning_rate": 0.001, | |
| "loss": 2030.5203, | |
| "step": 6240 | |
| }, | |
| { | |
| "epoch": 0.12868896074659245, | |
| "grad_norm": 560.0, | |
| "learning_rate": 0.001, | |
| "loss": 2047.7404, | |
| "step": 6288 | |
| }, | |
| { | |
| "epoch": 0.12967131922557407, | |
| "grad_norm": 624.0, | |
| "learning_rate": 0.001, | |
| "loss": 2043.3193, | |
| "step": 6336 | |
| }, | |
| { | |
| "epoch": 0.1306536777045557, | |
| "grad_norm": 592.0, | |
| "learning_rate": 0.001, | |
| "loss": 2051.0589, | |
| "step": 6384 | |
| }, | |
| { | |
| "epoch": 0.1316360361835373, | |
| "grad_norm": 876.0, | |
| "learning_rate": 0.001, | |
| "loss": 2054.3232, | |
| "step": 6432 | |
| }, | |
| { | |
| "epoch": 0.13261839466251893, | |
| "grad_norm": 544.0, | |
| "learning_rate": 0.001, | |
| "loss": 2047.3159, | |
| "step": 6480 | |
| }, | |
| { | |
| "epoch": 0.13360075314150055, | |
| "grad_norm": 648.0, | |
| "learning_rate": 0.001, | |
| "loss": 2029.021, | |
| "step": 6528 | |
| }, | |
| { | |
| "epoch": 0.13458311162048217, | |
| "grad_norm": 556.0, | |
| "learning_rate": 0.001, | |
| "loss": 2027.506, | |
| "step": 6576 | |
| }, | |
| { | |
| "epoch": 0.1355654700994638, | |
| "grad_norm": 672.0, | |
| "learning_rate": 0.001, | |
| "loss": 2034.3325, | |
| "step": 6624 | |
| }, | |
| { | |
| "epoch": 0.1365478285784454, | |
| "grad_norm": 648.0, | |
| "learning_rate": 0.001, | |
| "loss": 1988.6841, | |
| "step": 6672 | |
| }, | |
| { | |
| "epoch": 0.13753018705742703, | |
| "grad_norm": 592.0, | |
| "learning_rate": 0.001, | |
| "loss": 1998.0236, | |
| "step": 6720 | |
| }, | |
| { | |
| "epoch": 0.13851254553640865, | |
| "grad_norm": 552.0, | |
| "learning_rate": 0.001, | |
| "loss": 2008.8337, | |
| "step": 6768 | |
| }, | |
| { | |
| "epoch": 0.13949490401539028, | |
| "grad_norm": 780.0, | |
| "learning_rate": 0.001, | |
| "loss": 2008.4787, | |
| "step": 6816 | |
| }, | |
| { | |
| "epoch": 0.1404772624943719, | |
| "grad_norm": 716.0, | |
| "learning_rate": 0.001, | |
| "loss": 1995.237, | |
| "step": 6864 | |
| }, | |
| { | |
| "epoch": 0.14145962097335352, | |
| "grad_norm": 804.0, | |
| "learning_rate": 0.001, | |
| "loss": 1996.2018, | |
| "step": 6912 | |
| }, | |
| { | |
| "epoch": 0.14244197945233514, | |
| "grad_norm": 652.0, | |
| "learning_rate": 0.001, | |
| "loss": 1992.167, | |
| "step": 6960 | |
| }, | |
| { | |
| "epoch": 0.14342433793131676, | |
| "grad_norm": 544.0, | |
| "learning_rate": 0.001, | |
| "loss": 1985.2515, | |
| "step": 7008 | |
| }, | |
| { | |
| "epoch": 0.14440669641029838, | |
| "grad_norm": 600.0, | |
| "learning_rate": 0.001, | |
| "loss": 1989.0208, | |
| "step": 7056 | |
| }, | |
| { | |
| "epoch": 0.14538905488928, | |
| "grad_norm": 712.0, | |
| "learning_rate": 0.001, | |
| "loss": 1993.743, | |
| "step": 7104 | |
| }, | |
| { | |
| "epoch": 0.14637141336826162, | |
| "grad_norm": 580.0, | |
| "learning_rate": 0.001, | |
| "loss": 1986.2668, | |
| "step": 7152 | |
| }, | |
| { | |
| "epoch": 0.14735377184724327, | |
| "grad_norm": 716.0, | |
| "learning_rate": 0.001, | |
| "loss": 1988.6514, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 0.1483361303262249, | |
| "grad_norm": 452.0, | |
| "learning_rate": 0.001, | |
| "loss": 1971.7622, | |
| "step": 7248 | |
| }, | |
| { | |
| "epoch": 0.1493184888052065, | |
| "grad_norm": 576.0, | |
| "learning_rate": 0.001, | |
| "loss": 1977.0863, | |
| "step": 7296 | |
| }, | |
| { | |
| "epoch": 0.15030084728418813, | |
| "grad_norm": 708.0, | |
| "learning_rate": 0.001, | |
| "loss": 1968.3294, | |
| "step": 7344 | |
| }, | |
| { | |
| "epoch": 0.15128320576316975, | |
| "grad_norm": 572.0, | |
| "learning_rate": 0.001, | |
| "loss": 1981.1888, | |
| "step": 7392 | |
| }, | |
| { | |
| "epoch": 0.15226556424215137, | |
| "grad_norm": 748.0, | |
| "learning_rate": 0.001, | |
| "loss": 1937.5469, | |
| "step": 7440 | |
| }, | |
| { | |
| "epoch": 0.153247922721133, | |
| "grad_norm": 672.0, | |
| "learning_rate": 0.001, | |
| "loss": 1944.2785, | |
| "step": 7488 | |
| }, | |
| { | |
| "epoch": 0.15423028120011462, | |
| "grad_norm": 816.0, | |
| "learning_rate": 0.001, | |
| "loss": 1934.2336, | |
| "step": 7536 | |
| }, | |
| { | |
| "epoch": 0.15521263967909624, | |
| "grad_norm": 740.0, | |
| "learning_rate": 0.001, | |
| "loss": 1953.3698, | |
| "step": 7584 | |
| }, | |
| { | |
| "epoch": 0.15619499815807786, | |
| "grad_norm": 652.0, | |
| "learning_rate": 0.001, | |
| "loss": 1951.9084, | |
| "step": 7632 | |
| }, | |
| { | |
| "epoch": 0.15717735663705948, | |
| "grad_norm": 652.0, | |
| "learning_rate": 0.001, | |
| "loss": 1934.8753, | |
| "step": 7680 | |
| }, | |
| { | |
| "epoch": 0.1581597151160411, | |
| "grad_norm": 652.0, | |
| "learning_rate": 0.001, | |
| "loss": 1923.8843, | |
| "step": 7728 | |
| }, | |
| { | |
| "epoch": 0.15914207359502272, | |
| "grad_norm": 612.0, | |
| "learning_rate": 0.001, | |
| "loss": 1935.5955, | |
| "step": 7776 | |
| }, | |
| { | |
| "epoch": 0.16012443207400434, | |
| "grad_norm": 724.0, | |
| "learning_rate": 0.001, | |
| "loss": 1962.8574, | |
| "step": 7824 | |
| }, | |
| { | |
| "epoch": 0.16110679055298596, | |
| "grad_norm": 540.0, | |
| "learning_rate": 0.001, | |
| "loss": 1955.3468, | |
| "step": 7872 | |
| }, | |
| { | |
| "epoch": 0.16208914903196758, | |
| "grad_norm": 752.0, | |
| "learning_rate": 0.001, | |
| "loss": 1915.3901, | |
| "step": 7920 | |
| }, | |
| { | |
| "epoch": 0.1630715075109492, | |
| "grad_norm": 572.0, | |
| "learning_rate": 0.001, | |
| "loss": 1944.2292, | |
| "step": 7968 | |
| }, | |
| { | |
| "epoch": 0.16405386598993082, | |
| "grad_norm": 668.0, | |
| "learning_rate": 0.001, | |
| "loss": 1926.0425, | |
| "step": 8016 | |
| }, | |
| { | |
| "epoch": 0.16503622446891245, | |
| "grad_norm": 556.0, | |
| "learning_rate": 0.001, | |
| "loss": 1938.1131, | |
| "step": 8064 | |
| }, | |
| { | |
| "epoch": 0.16601858294789407, | |
| "grad_norm": 756.0, | |
| "learning_rate": 0.001, | |
| "loss": 1925.4678, | |
| "step": 8112 | |
| }, | |
| { | |
| "epoch": 0.1670009414268757, | |
| "grad_norm": 848.0, | |
| "learning_rate": 0.001, | |
| "loss": 1921.8462, | |
| "step": 8160 | |
| }, | |
| { | |
| "epoch": 0.1679832999058573, | |
| "grad_norm": 588.0, | |
| "learning_rate": 0.001, | |
| "loss": 1890.1263, | |
| "step": 8208 | |
| }, | |
| { | |
| "epoch": 0.16896565838483893, | |
| "grad_norm": 580.0, | |
| "learning_rate": 0.001, | |
| "loss": 1923.7113, | |
| "step": 8256 | |
| }, | |
| { | |
| "epoch": 0.16994801686382055, | |
| "grad_norm": 712.0, | |
| "learning_rate": 0.001, | |
| "loss": 1902.661, | |
| "step": 8304 | |
| }, | |
| { | |
| "epoch": 0.17093037534280217, | |
| "grad_norm": 676.0, | |
| "learning_rate": 0.001, | |
| "loss": 1898.4054, | |
| "step": 8352 | |
| }, | |
| { | |
| "epoch": 0.1719127338217838, | |
| "grad_norm": 604.0, | |
| "learning_rate": 0.001, | |
| "loss": 1899.0542, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 0.1728950923007654, | |
| "grad_norm": 700.0, | |
| "learning_rate": 0.001, | |
| "loss": 1906.8057, | |
| "step": 8448 | |
| }, | |
| { | |
| "epoch": 0.17387745077974703, | |
| "grad_norm": 716.0, | |
| "learning_rate": 0.001, | |
| "loss": 1908.1032, | |
| "step": 8496 | |
| }, | |
| { | |
| "epoch": 0.17485980925872865, | |
| "grad_norm": 628.0, | |
| "learning_rate": 0.001, | |
| "loss": 1928.3221, | |
| "step": 8544 | |
| }, | |
| { | |
| "epoch": 0.17584216773771028, | |
| "grad_norm": 664.0, | |
| "learning_rate": 0.001, | |
| "loss": 1890.7321, | |
| "step": 8592 | |
| }, | |
| { | |
| "epoch": 0.1768245262166919, | |
| "grad_norm": 556.0, | |
| "learning_rate": 0.001, | |
| "loss": 1910.8001, | |
| "step": 8640 | |
| }, | |
| { | |
| "epoch": 0.17780688469567352, | |
| "grad_norm": 648.0, | |
| "learning_rate": 0.001, | |
| "loss": 1908.4972, | |
| "step": 8688 | |
| }, | |
| { | |
| "epoch": 0.17878924317465514, | |
| "grad_norm": 608.0, | |
| "learning_rate": 0.001, | |
| "loss": 1870.7344, | |
| "step": 8736 | |
| }, | |
| { | |
| "epoch": 0.1797716016536368, | |
| "grad_norm": 644.0, | |
| "learning_rate": 0.001, | |
| "loss": 1901.4289, | |
| "step": 8784 | |
| }, | |
| { | |
| "epoch": 0.1807539601326184, | |
| "grad_norm": 580.0, | |
| "learning_rate": 0.001, | |
| "loss": 1883.8433, | |
| "step": 8832 | |
| }, | |
| { | |
| "epoch": 0.18173631861160003, | |
| "grad_norm": 828.0, | |
| "learning_rate": 0.001, | |
| "loss": 1869.978, | |
| "step": 8880 | |
| }, | |
| { | |
| "epoch": 0.18271867709058165, | |
| "grad_norm": 652.0, | |
| "learning_rate": 0.001, | |
| "loss": 1895.2178, | |
| "step": 8928 | |
| }, | |
| { | |
| "epoch": 0.18370103556956327, | |
| "grad_norm": 680.0, | |
| "learning_rate": 0.001, | |
| "loss": 1857.217, | |
| "step": 8976 | |
| }, | |
| { | |
| "epoch": 0.1846833940485449, | |
| "grad_norm": 608.0, | |
| "learning_rate": 0.001, | |
| "loss": 1880.6992, | |
| "step": 9024 | |
| }, | |
| { | |
| "epoch": 0.1856657525275265, | |
| "grad_norm": 664.0, | |
| "learning_rate": 0.001, | |
| "loss": 1869.5422, | |
| "step": 9072 | |
| }, | |
| { | |
| "epoch": 0.18664811100650813, | |
| "grad_norm": 720.0, | |
| "learning_rate": 0.001, | |
| "loss": 1898.1034, | |
| "step": 9120 | |
| }, | |
| { | |
| "epoch": 0.18763046948548975, | |
| "grad_norm": 604.0, | |
| "learning_rate": 0.001, | |
| "loss": 1887.1818, | |
| "step": 9168 | |
| }, | |
| { | |
| "epoch": 0.18861282796447137, | |
| "grad_norm": 672.0, | |
| "learning_rate": 0.001, | |
| "loss": 1869.6294, | |
| "step": 9216 | |
| }, | |
| { | |
| "epoch": 0.189595186443453, | |
| "grad_norm": 572.0, | |
| "learning_rate": 0.001, | |
| "loss": 1857.5962, | |
| "step": 9264 | |
| }, | |
| { | |
| "epoch": 0.19057754492243462, | |
| "grad_norm": 732.0, | |
| "learning_rate": 0.001, | |
| "loss": 1869.6444, | |
| "step": 9312 | |
| }, | |
| { | |
| "epoch": 0.19155990340141624, | |
| "grad_norm": 848.0, | |
| "learning_rate": 0.001, | |
| "loss": 1869.8807, | |
| "step": 9360 | |
| }, | |
| { | |
| "epoch": 0.19254226188039786, | |
| "grad_norm": 740.0, | |
| "learning_rate": 0.001, | |
| "loss": 1857.5882, | |
| "step": 9408 | |
| }, | |
| { | |
| "epoch": 0.19352462035937948, | |
| "grad_norm": 652.0, | |
| "learning_rate": 0.001, | |
| "loss": 1835.6294, | |
| "step": 9456 | |
| }, | |
| { | |
| "epoch": 0.1945069788383611, | |
| "grad_norm": 608.0, | |
| "learning_rate": 0.001, | |
| "loss": 1853.5081, | |
| "step": 9504 | |
| }, | |
| { | |
| "epoch": 0.19548933731734272, | |
| "grad_norm": 648.0, | |
| "learning_rate": 0.001, | |
| "loss": 1866.897, | |
| "step": 9552 | |
| }, | |
| { | |
| "epoch": 0.19647169579632434, | |
| "grad_norm": 724.0, | |
| "learning_rate": 0.001, | |
| "loss": 1848.0703, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 0.19745405427530596, | |
| "grad_norm": 628.0, | |
| "learning_rate": 0.001, | |
| "loss": 1867.1585, | |
| "step": 9648 | |
| }, | |
| { | |
| "epoch": 0.19843641275428758, | |
| "grad_norm": 612.0, | |
| "learning_rate": 0.001, | |
| "loss": 1864.8203, | |
| "step": 9696 | |
| }, | |
| { | |
| "epoch": 0.1994187712332692, | |
| "grad_norm": 572.0, | |
| "learning_rate": 0.001, | |
| "loss": 1827.0848, | |
| "step": 9744 | |
| }, | |
| { | |
| "epoch": 0.19999181367934182, | |
| "eval_loss": 1771.5172119140625, | |
| "eval_runtime": 9.0052, | |
| "eval_samples_per_second": 111.047, | |
| "eval_steps_per_second": 1.444, | |
| "step": 9772 | |
| }, | |
| { | |
| "epoch": 0.20040112971225083, | |
| "grad_norm": 784.0, | |
| "learning_rate": 0.001, | |
| "loss": 1850.5506, | |
| "step": 9792 | |
| }, | |
| { | |
| "epoch": 0.20138348819123245, | |
| "grad_norm": 728.0, | |
| "learning_rate": 0.001, | |
| "loss": 1853.3254, | |
| "step": 9840 | |
| }, | |
| { | |
| "epoch": 0.20236584667021407, | |
| "grad_norm": 744.0, | |
| "learning_rate": 0.001, | |
| "loss": 1884.8763, | |
| "step": 9888 | |
| }, | |
| { | |
| "epoch": 0.2033482051491957, | |
| "grad_norm": 740.0, | |
| "learning_rate": 0.001, | |
| "loss": 1852.5361, | |
| "step": 9936 | |
| }, | |
| { | |
| "epoch": 0.2043305636281773, | |
| "grad_norm": 780.0, | |
| "learning_rate": 0.001, | |
| "loss": 1840.8996, | |
| "step": 9984 | |
| }, | |
| { | |
| "epoch": 0.20531292210715893, | |
| "grad_norm": 676.0, | |
| "learning_rate": 0.001, | |
| "loss": 1848.5868, | |
| "step": 10032 | |
| }, | |
| { | |
| "epoch": 0.20629528058614055, | |
| "grad_norm": 764.0, | |
| "learning_rate": 0.001, | |
| "loss": 1848.7498, | |
| "step": 10080 | |
| }, | |
| { | |
| "epoch": 0.20727763906512217, | |
| "grad_norm": 856.0, | |
| "learning_rate": 0.001, | |
| "loss": 1845.7381, | |
| "step": 10128 | |
| }, | |
| { | |
| "epoch": 0.2082599975441038, | |
| "grad_norm": 740.0, | |
| "learning_rate": 0.001, | |
| "loss": 1851.979, | |
| "step": 10176 | |
| }, | |
| { | |
| "epoch": 0.2092423560230854, | |
| "grad_norm": 644.0, | |
| "learning_rate": 0.001, | |
| "loss": 1850.7716, | |
| "step": 10224 | |
| }, | |
| { | |
| "epoch": 0.21022471450206703, | |
| "grad_norm": 828.0, | |
| "learning_rate": 0.001, | |
| "loss": 1844.3057, | |
| "step": 10272 | |
| }, | |
| { | |
| "epoch": 0.21120707298104865, | |
| "grad_norm": 688.0, | |
| "learning_rate": 0.001, | |
| "loss": 1820.6678, | |
| "step": 10320 | |
| }, | |
| { | |
| "epoch": 0.21218943146003028, | |
| "grad_norm": 832.0, | |
| "learning_rate": 0.001, | |
| "loss": 1793.4041, | |
| "step": 10368 | |
| }, | |
| { | |
| "epoch": 0.21317178993901192, | |
| "grad_norm": 648.0, | |
| "learning_rate": 0.001, | |
| "loss": 1853.5828, | |
| "step": 10416 | |
| }, | |
| { | |
| "epoch": 0.21415414841799355, | |
| "grad_norm": 780.0, | |
| "learning_rate": 0.001, | |
| "loss": 1816.429, | |
| "step": 10464 | |
| }, | |
| { | |
| "epoch": 0.21513650689697517, | |
| "grad_norm": 708.0, | |
| "learning_rate": 0.001, | |
| "loss": 1827.7533, | |
| "step": 10512 | |
| }, | |
| { | |
| "epoch": 0.2161188653759568, | |
| "grad_norm": 812.0, | |
| "learning_rate": 0.001, | |
| "loss": 1807.555, | |
| "step": 10560 | |
| }, | |
| { | |
| "epoch": 0.2171012238549384, | |
| "grad_norm": 728.0, | |
| "learning_rate": 0.001, | |
| "loss": 1824.4678, | |
| "step": 10608 | |
| }, | |
| { | |
| "epoch": 0.21808358233392003, | |
| "grad_norm": 760.0, | |
| "learning_rate": 0.001, | |
| "loss": 1832.056, | |
| "step": 10656 | |
| }, | |
| { | |
| "epoch": 0.21906594081290165, | |
| "grad_norm": 672.0, | |
| "learning_rate": 0.001, | |
| "loss": 1819.7812, | |
| "step": 10704 | |
| }, | |
| { | |
| "epoch": 0.22004829929188327, | |
| "grad_norm": 808.0, | |
| "learning_rate": 0.001, | |
| "loss": 1837.8351, | |
| "step": 10752 | |
| }, | |
| { | |
| "epoch": 0.2210306577708649, | |
| "grad_norm": 716.0, | |
| "learning_rate": 0.001, | |
| "loss": 1823.1432, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 0.2220130162498465, | |
| "grad_norm": 768.0, | |
| "learning_rate": 0.001, | |
| "loss": 1810.9959, | |
| "step": 10848 | |
| }, | |
| { | |
| "epoch": 0.22299537472882813, | |
| "grad_norm": 756.0, | |
| "learning_rate": 0.001, | |
| "loss": 1848.2907, | |
| "step": 10896 | |
| }, | |
| { | |
| "epoch": 0.22397773320780975, | |
| "grad_norm": 744.0, | |
| "learning_rate": 0.001, | |
| "loss": 1786.6442, | |
| "step": 10944 | |
| }, | |
| { | |
| "epoch": 0.22496009168679137, | |
| "grad_norm": 740.0, | |
| "learning_rate": 0.001, | |
| "loss": 1804.0133, | |
| "step": 10992 | |
| }, | |
| { | |
| "epoch": 0.225942450165773, | |
| "grad_norm": 640.0, | |
| "learning_rate": 0.001, | |
| "loss": 1813.6567, | |
| "step": 11040 | |
| }, | |
| { | |
| "epoch": 0.22692480864475462, | |
| "grad_norm": 696.0, | |
| "learning_rate": 0.001, | |
| "loss": 1812.9946, | |
| "step": 11088 | |
| }, | |
| { | |
| "epoch": 0.22790716712373624, | |
| "grad_norm": 696.0, | |
| "learning_rate": 0.001, | |
| "loss": 1816.8553, | |
| "step": 11136 | |
| }, | |
| { | |
| "epoch": 0.22888952560271786, | |
| "grad_norm": 640.0, | |
| "learning_rate": 0.001, | |
| "loss": 1801.8009, | |
| "step": 11184 | |
| }, | |
| { | |
| "epoch": 0.22987188408169948, | |
| "grad_norm": 680.0, | |
| "learning_rate": 0.001, | |
| "loss": 1816.7332, | |
| "step": 11232 | |
| }, | |
| { | |
| "epoch": 0.2308542425606811, | |
| "grad_norm": 692.0, | |
| "learning_rate": 0.001, | |
| "loss": 1799.8708, | |
| "step": 11280 | |
| }, | |
| { | |
| "epoch": 0.23183660103966272, | |
| "grad_norm": 680.0, | |
| "learning_rate": 0.001, | |
| "loss": 1791.1471, | |
| "step": 11328 | |
| }, | |
| { | |
| "epoch": 0.23281895951864434, | |
| "grad_norm": 700.0, | |
| "learning_rate": 0.001, | |
| "loss": 1812.3979, | |
| "step": 11376 | |
| }, | |
| { | |
| "epoch": 0.23380131799762596, | |
| "grad_norm": 748.0, | |
| "learning_rate": 0.001, | |
| "loss": 1807.806, | |
| "step": 11424 | |
| }, | |
| { | |
| "epoch": 0.23478367647660758, | |
| "grad_norm": 716.0, | |
| "learning_rate": 0.001, | |
| "loss": 1811.6502, | |
| "step": 11472 | |
| }, | |
| { | |
| "epoch": 0.2357660349555892, | |
| "grad_norm": 736.0, | |
| "learning_rate": 0.001, | |
| "loss": 1821.8501, | |
| "step": 11520 | |
| }, | |
| { | |
| "epoch": 0.23674839343457083, | |
| "grad_norm": 648.0, | |
| "learning_rate": 0.001, | |
| "loss": 1777.6597, | |
| "step": 11568 | |
| }, | |
| { | |
| "epoch": 0.23773075191355245, | |
| "grad_norm": 768.0, | |
| "learning_rate": 0.001, | |
| "loss": 1804.2365, | |
| "step": 11616 | |
| }, | |
| { | |
| "epoch": 0.23871311039253407, | |
| "grad_norm": 672.0, | |
| "learning_rate": 0.001, | |
| "loss": 1794.9201, | |
| "step": 11664 | |
| }, | |
| { | |
| "epoch": 0.2396954688715157, | |
| "grad_norm": 696.0, | |
| "learning_rate": 0.001, | |
| "loss": 1775.2284, | |
| "step": 11712 | |
| }, | |
| { | |
| "epoch": 0.2406778273504973, | |
| "grad_norm": 700.0, | |
| "learning_rate": 0.001, | |
| "loss": 1785.5417, | |
| "step": 11760 | |
| }, | |
| { | |
| "epoch": 0.24166018582947893, | |
| "grad_norm": 840.0, | |
| "learning_rate": 0.001, | |
| "loss": 1792.2282, | |
| "step": 11808 | |
| }, | |
| { | |
| "epoch": 0.24264254430846055, | |
| "grad_norm": 992.0, | |
| "learning_rate": 0.001, | |
| "loss": 1799.9831, | |
| "step": 11856 | |
| }, | |
| { | |
| "epoch": 0.24362490278744217, | |
| "grad_norm": 872.0, | |
| "learning_rate": 0.001, | |
| "loss": 1804.1024, | |
| "step": 11904 | |
| }, | |
| { | |
| "epoch": 0.2446072612664238, | |
| "grad_norm": 668.0, | |
| "learning_rate": 0.001, | |
| "loss": 1785.5889, | |
| "step": 11952 | |
| }, | |
| { | |
| "epoch": 0.24558961974540544, | |
| "grad_norm": 872.0, | |
| "learning_rate": 0.001, | |
| "loss": 1785.6185, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.24657197822438706, | |
| "grad_norm": 784.0, | |
| "learning_rate": 0.001, | |
| "loss": 1785.6107, | |
| "step": 12048 | |
| }, | |
| { | |
| "epoch": 0.24755433670336868, | |
| "grad_norm": 644.0, | |
| "learning_rate": 0.001, | |
| "loss": 1789.2995, | |
| "step": 12096 | |
| }, | |
| { | |
| "epoch": 0.2485366951823503, | |
| "grad_norm": 772.0, | |
| "learning_rate": 0.001, | |
| "loss": 1780.3151, | |
| "step": 12144 | |
| }, | |
| { | |
| "epoch": 0.24951905366133192, | |
| "grad_norm": 728.0, | |
| "learning_rate": 0.001, | |
| "loss": 1769.0786, | |
| "step": 12192 | |
| }, | |
| { | |
| "epoch": 0.2505014121403135, | |
| "grad_norm": 752.0, | |
| "learning_rate": 0.001, | |
| "loss": 1801.0431, | |
| "step": 12240 | |
| }, | |
| { | |
| "epoch": 0.25148377061929517, | |
| "grad_norm": 844.0, | |
| "learning_rate": 0.001, | |
| "loss": 1796.9209, | |
| "step": 12288 | |
| }, | |
| { | |
| "epoch": 0.25246612909827676, | |
| "grad_norm": 1088.0, | |
| "learning_rate": 0.001, | |
| "loss": 1795.411, | |
| "step": 12336 | |
| }, | |
| { | |
| "epoch": 0.2534484875772584, | |
| "grad_norm": 1320.0, | |
| "learning_rate": 0.001, | |
| "loss": 1802.1553, | |
| "step": 12384 | |
| }, | |
| { | |
| "epoch": 0.25443084605624, | |
| "grad_norm": 716.0, | |
| "learning_rate": 0.001, | |
| "loss": 1772.7713, | |
| "step": 12432 | |
| }, | |
| { | |
| "epoch": 0.25541320453522165, | |
| "grad_norm": 744.0, | |
| "learning_rate": 0.001, | |
| "loss": 1787.7516, | |
| "step": 12480 | |
| }, | |
| { | |
| "epoch": 0.25639556301420324, | |
| "grad_norm": 744.0, | |
| "learning_rate": 0.001, | |
| "loss": 1785.005, | |
| "step": 12528 | |
| }, | |
| { | |
| "epoch": 0.2573779214931849, | |
| "grad_norm": 716.0, | |
| "learning_rate": 0.001, | |
| "loss": 1763.7612, | |
| "step": 12576 | |
| }, | |
| { | |
| "epoch": 0.2583602799721665, | |
| "grad_norm": 752.0, | |
| "learning_rate": 0.001, | |
| "loss": 1777.0677, | |
| "step": 12624 | |
| }, | |
| { | |
| "epoch": 0.25934263845114813, | |
| "grad_norm": 712.0, | |
| "learning_rate": 0.001, | |
| "loss": 1761.1235, | |
| "step": 12672 | |
| }, | |
| { | |
| "epoch": 0.2603249969301297, | |
| "grad_norm": 696.0, | |
| "learning_rate": 0.001, | |
| "loss": 1795.2174, | |
| "step": 12720 | |
| }, | |
| { | |
| "epoch": 0.2613073554091114, | |
| "grad_norm": 816.0, | |
| "learning_rate": 0.001, | |
| "loss": 1763.8905, | |
| "step": 12768 | |
| }, | |
| { | |
| "epoch": 0.262289713888093, | |
| "grad_norm": 852.0, | |
| "learning_rate": 0.001, | |
| "loss": 1761.4404, | |
| "step": 12816 | |
| }, | |
| { | |
| "epoch": 0.2632720723670746, | |
| "grad_norm": 608.0, | |
| "learning_rate": 0.001, | |
| "loss": 1762.6668, | |
| "step": 12864 | |
| }, | |
| { | |
| "epoch": 0.26425443084605627, | |
| "grad_norm": 892.0, | |
| "learning_rate": 0.001, | |
| "loss": 1725.5638, | |
| "step": 12912 | |
| }, | |
| { | |
| "epoch": 0.26523678932503786, | |
| "grad_norm": 724.0, | |
| "learning_rate": 0.001, | |
| "loss": 1762.1764, | |
| "step": 12960 | |
| }, | |
| { | |
| "epoch": 0.2662191478040195, | |
| "grad_norm": 728.0, | |
| "learning_rate": 0.001, | |
| "loss": 1764.0163, | |
| "step": 13008 | |
| }, | |
| { | |
| "epoch": 0.2672015062830011, | |
| "grad_norm": 664.0, | |
| "learning_rate": 0.001, | |
| "loss": 1754.1729, | |
| "step": 13056 | |
| }, | |
| { | |
| "epoch": 0.26818386476198275, | |
| "grad_norm": 724.0, | |
| "learning_rate": 0.001, | |
| "loss": 1772.8592, | |
| "step": 13104 | |
| }, | |
| { | |
| "epoch": 0.26916622324096434, | |
| "grad_norm": 704.0, | |
| "learning_rate": 0.001, | |
| "loss": 1780.2349, | |
| "step": 13152 | |
| }, | |
| { | |
| "epoch": 0.270148581719946, | |
| "grad_norm": 720.0, | |
| "learning_rate": 0.001, | |
| "loss": 1764.6678, | |
| "step": 13200 | |
| }, | |
| { | |
| "epoch": 0.2711309401989276, | |
| "grad_norm": 772.0, | |
| "learning_rate": 0.001, | |
| "loss": 1757.6209, | |
| "step": 13248 | |
| }, | |
| { | |
| "epoch": 0.27211329867790923, | |
| "grad_norm": 748.0, | |
| "learning_rate": 0.001, | |
| "loss": 1749.5741, | |
| "step": 13296 | |
| }, | |
| { | |
| "epoch": 0.2730956571568908, | |
| "grad_norm": 804.0, | |
| "learning_rate": 0.001, | |
| "loss": 1758.4183, | |
| "step": 13344 | |
| }, | |
| { | |
| "epoch": 0.2740780156358725, | |
| "grad_norm": 712.0, | |
| "learning_rate": 0.001, | |
| "loss": 1770.6115, | |
| "step": 13392 | |
| }, | |
| { | |
| "epoch": 0.27506037411485407, | |
| "grad_norm": 784.0, | |
| "learning_rate": 0.001, | |
| "loss": 1752.6141, | |
| "step": 13440 | |
| }, | |
| { | |
| "epoch": 0.2760427325938357, | |
| "grad_norm": 700.0, | |
| "learning_rate": 0.001, | |
| "loss": 1732.4147, | |
| "step": 13488 | |
| }, | |
| { | |
| "epoch": 0.2770250910728173, | |
| "grad_norm": 780.0, | |
| "learning_rate": 0.001, | |
| "loss": 1757.4318, | |
| "step": 13536 | |
| }, | |
| { | |
| "epoch": 0.27800744955179896, | |
| "grad_norm": 764.0, | |
| "learning_rate": 0.001, | |
| "loss": 1746.452, | |
| "step": 13584 | |
| }, | |
| { | |
| "epoch": 0.27898980803078055, | |
| "grad_norm": 696.0, | |
| "learning_rate": 0.001, | |
| "loss": 1733.8742, | |
| "step": 13632 | |
| }, | |
| { | |
| "epoch": 0.2799721665097622, | |
| "grad_norm": 824.0, | |
| "learning_rate": 0.001, | |
| "loss": 1761.4808, | |
| "step": 13680 | |
| }, | |
| { | |
| "epoch": 0.2809545249887438, | |
| "grad_norm": 664.0, | |
| "learning_rate": 0.001, | |
| "loss": 1749.8506, | |
| "step": 13728 | |
| }, | |
| { | |
| "epoch": 0.28193688346772544, | |
| "grad_norm": 776.0, | |
| "learning_rate": 0.001, | |
| "loss": 1734.1479, | |
| "step": 13776 | |
| }, | |
| { | |
| "epoch": 0.28291924194670703, | |
| "grad_norm": 708.0, | |
| "learning_rate": 0.001, | |
| "loss": 1721.951, | |
| "step": 13824 | |
| }, | |
| { | |
| "epoch": 0.2839016004256887, | |
| "grad_norm": 748.0, | |
| "learning_rate": 0.001, | |
| "loss": 1741.5046, | |
| "step": 13872 | |
| }, | |
| { | |
| "epoch": 0.2848839589046703, | |
| "grad_norm": 764.0, | |
| "learning_rate": 0.001, | |
| "loss": 1743.1763, | |
| "step": 13920 | |
| }, | |
| { | |
| "epoch": 0.2858663173836519, | |
| "grad_norm": 808.0, | |
| "learning_rate": 0.001, | |
| "loss": 1767.7448, | |
| "step": 13968 | |
| }, | |
| { | |
| "epoch": 0.2868486758626335, | |
| "grad_norm": 696.0, | |
| "learning_rate": 0.001, | |
| "loss": 1717.0291, | |
| "step": 14016 | |
| }, | |
| { | |
| "epoch": 0.28783103434161517, | |
| "grad_norm": 804.0, | |
| "learning_rate": 0.001, | |
| "loss": 1731.1566, | |
| "step": 14064 | |
| }, | |
| { | |
| "epoch": 0.28881339282059676, | |
| "grad_norm": 788.0, | |
| "learning_rate": 0.001, | |
| "loss": 1720.1551, | |
| "step": 14112 | |
| }, | |
| { | |
| "epoch": 0.2897957512995784, | |
| "grad_norm": 732.0, | |
| "learning_rate": 0.001, | |
| "loss": 1743.5467, | |
| "step": 14160 | |
| }, | |
| { | |
| "epoch": 0.29077810977856, | |
| "grad_norm": 680.0, | |
| "learning_rate": 0.001, | |
| "loss": 1736.5747, | |
| "step": 14208 | |
| }, | |
| { | |
| "epoch": 0.29176046825754165, | |
| "grad_norm": 740.0, | |
| "learning_rate": 0.001, | |
| "loss": 1737.3779, | |
| "step": 14256 | |
| }, | |
| { | |
| "epoch": 0.29274282673652324, | |
| "grad_norm": 760.0, | |
| "learning_rate": 0.001, | |
| "loss": 1718.3322, | |
| "step": 14304 | |
| }, | |
| { | |
| "epoch": 0.2937251852155049, | |
| "grad_norm": 760.0, | |
| "learning_rate": 0.001, | |
| "loss": 1736.3989, | |
| "step": 14352 | |
| }, | |
| { | |
| "epoch": 0.29470754369448654, | |
| "grad_norm": 832.0, | |
| "learning_rate": 0.001, | |
| "loss": 1738.8551, | |
| "step": 14400 | |
| }, | |
| { | |
| "epoch": 0.29568990217346813, | |
| "grad_norm": 900.0, | |
| "learning_rate": 0.001, | |
| "loss": 1711.0072, | |
| "step": 14448 | |
| }, | |
| { | |
| "epoch": 0.2966722606524498, | |
| "grad_norm": 748.0, | |
| "learning_rate": 0.001, | |
| "loss": 1709.6022, | |
| "step": 14496 | |
| }, | |
| { | |
| "epoch": 0.2976546191314314, | |
| "grad_norm": 704.0, | |
| "learning_rate": 0.001, | |
| "loss": 1741.5581, | |
| "step": 14544 | |
| }, | |
| { | |
| "epoch": 0.298636977610413, | |
| "grad_norm": 684.0, | |
| "learning_rate": 0.001, | |
| "loss": 1715.0571, | |
| "step": 14592 | |
| }, | |
| { | |
| "epoch": 0.2996193360893946, | |
| "grad_norm": 728.0, | |
| "learning_rate": 0.001, | |
| "loss": 1733.5199, | |
| "step": 14640 | |
| }, | |
| { | |
| "epoch": 0.29998772051901273, | |
| "eval_loss": 1650.6409912109375, | |
| "eval_runtime": 9.0148, | |
| "eval_samples_per_second": 110.929, | |
| "eval_steps_per_second": 1.442, | |
| "step": 14658 | |
| }, | |
| { | |
| "epoch": 0.30060169456837627, | |
| "grad_norm": 708.0, | |
| "learning_rate": 0.001, | |
| "loss": 1719.6375, | |
| "step": 14688 | |
| }, | |
| { | |
| "epoch": 0.30158405304735786, | |
| "grad_norm": 616.0, | |
| "learning_rate": 0.001, | |
| "loss": 1728.5174, | |
| "step": 14736 | |
| }, | |
| { | |
| "epoch": 0.3025664115263395, | |
| "grad_norm": 624.0, | |
| "learning_rate": 0.001, | |
| "loss": 1724.9813, | |
| "step": 14784 | |
| }, | |
| { | |
| "epoch": 0.3035487700053211, | |
| "grad_norm": 716.0, | |
| "learning_rate": 0.001, | |
| "loss": 1704.8024, | |
| "step": 14832 | |
| }, | |
| { | |
| "epoch": 0.30453112848430275, | |
| "grad_norm": 748.0, | |
| "learning_rate": 0.001, | |
| "loss": 1714.6235, | |
| "step": 14880 | |
| }, | |
| { | |
| "epoch": 0.30551348696328434, | |
| "grad_norm": 1048.0, | |
| "learning_rate": 0.001, | |
| "loss": 1734.4709, | |
| "step": 14928 | |
| }, | |
| { | |
| "epoch": 0.306495845442266, | |
| "grad_norm": 688.0, | |
| "learning_rate": 0.001, | |
| "loss": 1721.2712, | |
| "step": 14976 | |
| }, | |
| { | |
| "epoch": 0.3074782039212476, | |
| "grad_norm": 724.0, | |
| "learning_rate": 0.001, | |
| "loss": 1752.32, | |
| "step": 15024 | |
| }, | |
| { | |
| "epoch": 0.30846056240022923, | |
| "grad_norm": 792.0, | |
| "learning_rate": 0.001, | |
| "loss": 1711.9393, | |
| "step": 15072 | |
| }, | |
| { | |
| "epoch": 0.3094429208792108, | |
| "grad_norm": 904.0, | |
| "learning_rate": 0.001, | |
| "loss": 1722.3177, | |
| "step": 15120 | |
| }, | |
| { | |
| "epoch": 0.3104252793581925, | |
| "grad_norm": 768.0, | |
| "learning_rate": 0.001, | |
| "loss": 1737.7088, | |
| "step": 15168 | |
| }, | |
| { | |
| "epoch": 0.31140763783717407, | |
| "grad_norm": 768.0, | |
| "learning_rate": 0.001, | |
| "loss": 1728.0853, | |
| "step": 15216 | |
| }, | |
| { | |
| "epoch": 0.3123899963161557, | |
| "grad_norm": 776.0, | |
| "learning_rate": 0.001, | |
| "loss": 1711.749, | |
| "step": 15264 | |
| }, | |
| { | |
| "epoch": 0.3133723547951373, | |
| "grad_norm": 840.0, | |
| "learning_rate": 0.001, | |
| "loss": 1717.5446, | |
| "step": 15312 | |
| }, | |
| { | |
| "epoch": 0.31435471327411896, | |
| "grad_norm": 696.0, | |
| "learning_rate": 0.001, | |
| "loss": 1718.4888, | |
| "step": 15360 | |
| }, | |
| { | |
| "epoch": 0.31533707175310055, | |
| "grad_norm": 692.0, | |
| "learning_rate": 0.001, | |
| "loss": 1722.6672, | |
| "step": 15408 | |
| }, | |
| { | |
| "epoch": 0.3163194302320822, | |
| "grad_norm": 864.0, | |
| "learning_rate": 0.001, | |
| "loss": 1715.826, | |
| "step": 15456 | |
| }, | |
| { | |
| "epoch": 0.3173017887110638, | |
| "grad_norm": 712.0, | |
| "learning_rate": 0.001, | |
| "loss": 1714.7765, | |
| "step": 15504 | |
| }, | |
| { | |
| "epoch": 0.31828414719004544, | |
| "grad_norm": 712.0, | |
| "learning_rate": 0.001, | |
| "loss": 1718.0269, | |
| "step": 15552 | |
| }, | |
| { | |
| "epoch": 0.31926650566902703, | |
| "grad_norm": 748.0, | |
| "learning_rate": 0.001, | |
| "loss": 1691.7855, | |
| "step": 15600 | |
| }, | |
| { | |
| "epoch": 0.3202488641480087, | |
| "grad_norm": 888.0, | |
| "learning_rate": 0.001, | |
| "loss": 1717.6808, | |
| "step": 15648 | |
| }, | |
| { | |
| "epoch": 0.3212312226269903, | |
| "grad_norm": 680.0, | |
| "learning_rate": 0.001, | |
| "loss": 1719.3472, | |
| "step": 15696 | |
| }, | |
| { | |
| "epoch": 0.3222135811059719, | |
| "grad_norm": 868.0, | |
| "learning_rate": 0.001, | |
| "loss": 1691.0158, | |
| "step": 15744 | |
| }, | |
| { | |
| "epoch": 0.3231959395849535, | |
| "grad_norm": 756.0, | |
| "learning_rate": 0.001, | |
| "loss": 1704.2186, | |
| "step": 15792 | |
| }, | |
| { | |
| "epoch": 0.32417829806393517, | |
| "grad_norm": 888.0, | |
| "learning_rate": 0.001, | |
| "loss": 1723.0382, | |
| "step": 15840 | |
| }, | |
| { | |
| "epoch": 0.32516065654291676, | |
| "grad_norm": 912.0, | |
| "learning_rate": 0.001, | |
| "loss": 1702.2889, | |
| "step": 15888 | |
| }, | |
| { | |
| "epoch": 0.3261430150218984, | |
| "grad_norm": 820.0, | |
| "learning_rate": 0.001, | |
| "loss": 1728.0734, | |
| "step": 15936 | |
| }, | |
| { | |
| "epoch": 0.32712537350088006, | |
| "grad_norm": 788.0, | |
| "learning_rate": 0.001, | |
| "loss": 1720.2152, | |
| "step": 15984 | |
| }, | |
| { | |
| "epoch": 0.32810773197986165, | |
| "grad_norm": 808.0, | |
| "learning_rate": 0.001, | |
| "loss": 1702.1133, | |
| "step": 16032 | |
| }, | |
| { | |
| "epoch": 0.3290900904588433, | |
| "grad_norm": 836.0, | |
| "learning_rate": 0.001, | |
| "loss": 1720.4746, | |
| "step": 16080 | |
| }, | |
| { | |
| "epoch": 0.3300724489378249, | |
| "grad_norm": 836.0, | |
| "learning_rate": 0.001, | |
| "loss": 1689.6606, | |
| "step": 16128 | |
| }, | |
| { | |
| "epoch": 0.33105480741680654, | |
| "grad_norm": 728.0, | |
| "learning_rate": 0.001, | |
| "loss": 1689.0417, | |
| "step": 16176 | |
| }, | |
| { | |
| "epoch": 0.33203716589578813, | |
| "grad_norm": 848.0, | |
| "learning_rate": 0.001, | |
| "loss": 1703.012, | |
| "step": 16224 | |
| }, | |
| { | |
| "epoch": 0.3330195243747698, | |
| "grad_norm": 756.0, | |
| "learning_rate": 0.001, | |
| "loss": 1700.2785, | |
| "step": 16272 | |
| }, | |
| { | |
| "epoch": 0.3340018828537514, | |
| "grad_norm": 756.0, | |
| "learning_rate": 0.001, | |
| "loss": 1709.3231, | |
| "step": 16320 | |
| }, | |
| { | |
| "epoch": 0.334984241332733, | |
| "grad_norm": 960.0, | |
| "learning_rate": 0.001, | |
| "loss": 1715.8831, | |
| "step": 16368 | |
| }, | |
| { | |
| "epoch": 0.3359665998117146, | |
| "grad_norm": 692.0, | |
| "learning_rate": 0.001, | |
| "loss": 1695.813, | |
| "step": 16416 | |
| }, | |
| { | |
| "epoch": 0.33694895829069627, | |
| "grad_norm": 688.0, | |
| "learning_rate": 0.001, | |
| "loss": 1685.9803, | |
| "step": 16464 | |
| }, | |
| { | |
| "epoch": 0.33793131676967786, | |
| "grad_norm": 876.0, | |
| "learning_rate": 0.001, | |
| "loss": 1704.5868, | |
| "step": 16512 | |
| }, | |
| { | |
| "epoch": 0.3389136752486595, | |
| "grad_norm": 732.0, | |
| "learning_rate": 0.001, | |
| "loss": 1681.2751, | |
| "step": 16560 | |
| }, | |
| { | |
| "epoch": 0.3398960337276411, | |
| "grad_norm": 748.0, | |
| "learning_rate": 0.001, | |
| "loss": 1690.0252, | |
| "step": 16608 | |
| }, | |
| { | |
| "epoch": 0.34087839220662275, | |
| "grad_norm": 780.0, | |
| "learning_rate": 0.001, | |
| "loss": 1698.1932, | |
| "step": 16656 | |
| }, | |
| { | |
| "epoch": 0.34186075068560434, | |
| "grad_norm": 856.0, | |
| "learning_rate": 0.001, | |
| "loss": 1692.6128, | |
| "step": 16704 | |
| }, | |
| { | |
| "epoch": 0.342843109164586, | |
| "grad_norm": 880.0, | |
| "learning_rate": 0.001, | |
| "loss": 1696.6901, | |
| "step": 16752 | |
| }, | |
| { | |
| "epoch": 0.3438254676435676, | |
| "grad_norm": 688.0, | |
| "learning_rate": 0.001, | |
| "loss": 1693.9344, | |
| "step": 16800 | |
| }, | |
| { | |
| "epoch": 0.34480782612254923, | |
| "grad_norm": 688.0, | |
| "learning_rate": 0.001, | |
| "loss": 1704.4855, | |
| "step": 16848 | |
| }, | |
| { | |
| "epoch": 0.3457901846015308, | |
| "grad_norm": 692.0, | |
| "learning_rate": 0.001, | |
| "loss": 1705.7817, | |
| "step": 16896 | |
| }, | |
| { | |
| "epoch": 0.3467725430805125, | |
| "grad_norm": 760.0, | |
| "learning_rate": 0.001, | |
| "loss": 1690.8944, | |
| "step": 16944 | |
| }, | |
| { | |
| "epoch": 0.34775490155949407, | |
| "grad_norm": 868.0, | |
| "learning_rate": 0.001, | |
| "loss": 1685.9479, | |
| "step": 16992 | |
| }, | |
| { | |
| "epoch": 0.3487372600384757, | |
| "grad_norm": 744.0, | |
| "learning_rate": 0.001, | |
| "loss": 1698.2961, | |
| "step": 17040 | |
| }, | |
| { | |
| "epoch": 0.3497196185174573, | |
| "grad_norm": 688.0, | |
| "learning_rate": 0.001, | |
| "loss": 1693.7596, | |
| "step": 17088 | |
| }, | |
| { | |
| "epoch": 0.35070197699643896, | |
| "grad_norm": 760.0, | |
| "learning_rate": 0.001, | |
| "loss": 1702.9092, | |
| "step": 17136 | |
| }, | |
| { | |
| "epoch": 0.35168433547542055, | |
| "grad_norm": 788.0, | |
| "learning_rate": 0.001, | |
| "loss": 1672.0039, | |
| "step": 17184 | |
| }, | |
| { | |
| "epoch": 0.3526666939544022, | |
| "grad_norm": 772.0, | |
| "learning_rate": 0.001, | |
| "loss": 1673.5811, | |
| "step": 17232 | |
| }, | |
| { | |
| "epoch": 0.3536490524333838, | |
| "grad_norm": 976.0, | |
| "learning_rate": 0.001, | |
| "loss": 1697.0251, | |
| "step": 17280 | |
| }, | |
| { | |
| "epoch": 0.35463141091236544, | |
| "grad_norm": 740.0, | |
| "learning_rate": 0.001, | |
| "loss": 1699.7508, | |
| "step": 17328 | |
| }, | |
| { | |
| "epoch": 0.35561376939134703, | |
| "grad_norm": 704.0, | |
| "learning_rate": 0.001, | |
| "loss": 1708.7798, | |
| "step": 17376 | |
| }, | |
| { | |
| "epoch": 0.3565961278703287, | |
| "grad_norm": 716.0, | |
| "learning_rate": 0.001, | |
| "loss": 1693.451, | |
| "step": 17424 | |
| }, | |
| { | |
| "epoch": 0.3575784863493103, | |
| "grad_norm": 904.0, | |
| "learning_rate": 0.001, | |
| "loss": 1676.6382, | |
| "step": 17472 | |
| }, | |
| { | |
| "epoch": 0.3585608448282919, | |
| "grad_norm": 700.0, | |
| "learning_rate": 0.001, | |
| "loss": 1691.7266, | |
| "step": 17520 | |
| }, | |
| { | |
| "epoch": 0.3595432033072736, | |
| "grad_norm": 1012.0, | |
| "learning_rate": 0.001, | |
| "loss": 1666.9458, | |
| "step": 17568 | |
| }, | |
| { | |
| "epoch": 0.36052556178625517, | |
| "grad_norm": 804.0, | |
| "learning_rate": 0.001, | |
| "loss": 1675.5584, | |
| "step": 17616 | |
| }, | |
| { | |
| "epoch": 0.3615079202652368, | |
| "grad_norm": 856.0, | |
| "learning_rate": 0.001, | |
| "loss": 1685.9595, | |
| "step": 17664 | |
| }, | |
| { | |
| "epoch": 0.3624902787442184, | |
| "grad_norm": 836.0, | |
| "learning_rate": 0.001, | |
| "loss": 1695.4741, | |
| "step": 17712 | |
| }, | |
| { | |
| "epoch": 0.36347263722320006, | |
| "grad_norm": 696.0, | |
| "learning_rate": 0.001, | |
| "loss": 1689.256, | |
| "step": 17760 | |
| }, | |
| { | |
| "epoch": 0.36445499570218165, | |
| "grad_norm": 732.0, | |
| "learning_rate": 0.001, | |
| "loss": 1689.8599, | |
| "step": 17808 | |
| }, | |
| { | |
| "epoch": 0.3654373541811633, | |
| "grad_norm": 876.0, | |
| "learning_rate": 0.001, | |
| "loss": 1668.894, | |
| "step": 17856 | |
| }, | |
| { | |
| "epoch": 0.3664197126601449, | |
| "grad_norm": 780.0, | |
| "learning_rate": 0.001, | |
| "loss": 1684.5539, | |
| "step": 17904 | |
| }, | |
| { | |
| "epoch": 0.36740207113912654, | |
| "grad_norm": 952.0, | |
| "learning_rate": 0.001, | |
| "loss": 1683.5539, | |
| "step": 17952 | |
| }, | |
| { | |
| "epoch": 0.36838442961810813, | |
| "grad_norm": 860.0, | |
| "learning_rate": 0.001, | |
| "loss": 1695.5282, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 0.3693667880970898, | |
| "grad_norm": 844.0, | |
| "learning_rate": 0.001, | |
| "loss": 1667.8563, | |
| "step": 18048 | |
| }, | |
| { | |
| "epoch": 0.3703491465760714, | |
| "grad_norm": 800.0, | |
| "learning_rate": 0.001, | |
| "loss": 1671.2471, | |
| "step": 18096 | |
| }, | |
| { | |
| "epoch": 0.371331505055053, | |
| "grad_norm": 760.0, | |
| "learning_rate": 0.001, | |
| "loss": 1664.4082, | |
| "step": 18144 | |
| }, | |
| { | |
| "epoch": 0.3723138635340346, | |
| "grad_norm": 1120.0, | |
| "learning_rate": 0.001, | |
| "loss": 1666.0448, | |
| "step": 18192 | |
| }, | |
| { | |
| "epoch": 0.37329622201301627, | |
| "grad_norm": 824.0, | |
| "learning_rate": 0.001, | |
| "loss": 1665.9009, | |
| "step": 18240 | |
| }, | |
| { | |
| "epoch": 0.37427858049199786, | |
| "grad_norm": 872.0, | |
| "learning_rate": 0.001, | |
| "loss": 1663.6131, | |
| "step": 18288 | |
| }, | |
| { | |
| "epoch": 0.3752609389709795, | |
| "grad_norm": 804.0, | |
| "learning_rate": 0.001, | |
| "loss": 1665.7214, | |
| "step": 18336 | |
| }, | |
| { | |
| "epoch": 0.3762432974499611, | |
| "grad_norm": 768.0, | |
| "learning_rate": 0.001, | |
| "loss": 1663.0591, | |
| "step": 18384 | |
| }, | |
| { | |
| "epoch": 0.37722565592894275, | |
| "grad_norm": 988.0, | |
| "learning_rate": 0.001, | |
| "loss": 1683.5985, | |
| "step": 18432 | |
| }, | |
| { | |
| "epoch": 0.37820801440792434, | |
| "grad_norm": 804.0, | |
| "learning_rate": 0.001, | |
| "loss": 1661.8081, | |
| "step": 18480 | |
| }, | |
| { | |
| "epoch": 0.379190372886906, | |
| "grad_norm": 776.0, | |
| "learning_rate": 0.001, | |
| "loss": 1685.9769, | |
| "step": 18528 | |
| }, | |
| { | |
| "epoch": 0.3801727313658876, | |
| "grad_norm": 920.0, | |
| "learning_rate": 0.001, | |
| "loss": 1676.7816, | |
| "step": 18576 | |
| }, | |
| { | |
| "epoch": 0.38115508984486923, | |
| "grad_norm": 800.0, | |
| "learning_rate": 0.001, | |
| "loss": 1669.9821, | |
| "step": 18624 | |
| }, | |
| { | |
| "epoch": 0.3821374483238508, | |
| "grad_norm": 828.0, | |
| "learning_rate": 0.001, | |
| "loss": 1654.353, | |
| "step": 18672 | |
| }, | |
| { | |
| "epoch": 0.3831198068028325, | |
| "grad_norm": 756.0, | |
| "learning_rate": 0.001, | |
| "loss": 1641.133, | |
| "step": 18720 | |
| }, | |
| { | |
| "epoch": 0.38410216528181407, | |
| "grad_norm": 772.0, | |
| "learning_rate": 0.001, | |
| "loss": 1654.8257, | |
| "step": 18768 | |
| }, | |
| { | |
| "epoch": 0.3850845237607957, | |
| "grad_norm": 856.0, | |
| "learning_rate": 0.001, | |
| "loss": 1674.0685, | |
| "step": 18816 | |
| }, | |
| { | |
| "epoch": 0.3860668822397773, | |
| "grad_norm": 980.0, | |
| "learning_rate": 0.001, | |
| "loss": 1663.0804, | |
| "step": 18864 | |
| }, | |
| { | |
| "epoch": 0.38704924071875896, | |
| "grad_norm": 852.0, | |
| "learning_rate": 0.001, | |
| "loss": 1676.6375, | |
| "step": 18912 | |
| }, | |
| { | |
| "epoch": 0.38803159919774055, | |
| "grad_norm": 788.0, | |
| "learning_rate": 0.001, | |
| "loss": 1681.3968, | |
| "step": 18960 | |
| }, | |
| { | |
| "epoch": 0.3890139576767222, | |
| "grad_norm": 740.0, | |
| "learning_rate": 0.001, | |
| "loss": 1658.5428, | |
| "step": 19008 | |
| }, | |
| { | |
| "epoch": 0.3899963161557038, | |
| "grad_norm": 772.0, | |
| "learning_rate": 0.001, | |
| "loss": 1693.6683, | |
| "step": 19056 | |
| }, | |
| { | |
| "epoch": 0.39097867463468544, | |
| "grad_norm": 1224.0, | |
| "learning_rate": 0.001, | |
| "loss": 1637.9217, | |
| "step": 19104 | |
| }, | |
| { | |
| "epoch": 0.3919610331136671, | |
| "grad_norm": 896.0, | |
| "learning_rate": 0.001, | |
| "loss": 1680.59, | |
| "step": 19152 | |
| }, | |
| { | |
| "epoch": 0.3929433915926487, | |
| "grad_norm": 716.0, | |
| "learning_rate": 0.001, | |
| "loss": 1663.9777, | |
| "step": 19200 | |
| }, | |
| { | |
| "epoch": 0.39392575007163033, | |
| "grad_norm": 940.0, | |
| "learning_rate": 0.001, | |
| "loss": 1672.4491, | |
| "step": 19248 | |
| }, | |
| { | |
| "epoch": 0.3949081085506119, | |
| "grad_norm": 696.0, | |
| "learning_rate": 0.001, | |
| "loss": 1668.8294, | |
| "step": 19296 | |
| }, | |
| { | |
| "epoch": 0.3958904670295936, | |
| "grad_norm": 740.0, | |
| "learning_rate": 0.001, | |
| "loss": 1652.0028, | |
| "step": 19344 | |
| }, | |
| { | |
| "epoch": 0.39687282550857517, | |
| "grad_norm": 900.0, | |
| "learning_rate": 0.001, | |
| "loss": 1670.4543, | |
| "step": 19392 | |
| }, | |
| { | |
| "epoch": 0.3978551839875568, | |
| "grad_norm": 832.0, | |
| "learning_rate": 0.001, | |
| "loss": 1674.2799, | |
| "step": 19440 | |
| }, | |
| { | |
| "epoch": 0.3988375424665384, | |
| "grad_norm": 896.0, | |
| "learning_rate": 0.001, | |
| "loss": 1637.7557, | |
| "step": 19488 | |
| }, | |
| { | |
| "epoch": 0.39981990094552006, | |
| "grad_norm": 772.0, | |
| "learning_rate": 0.001, | |
| "loss": 1659.3999, | |
| "step": 19536 | |
| }, | |
| { | |
| "epoch": 0.39998362735868365, | |
| "eval_loss": 1588.186767578125, | |
| "eval_runtime": 9.0185, | |
| "eval_samples_per_second": 110.884, | |
| "eval_steps_per_second": 1.441, | |
| "step": 19544 | |
| }, | |
| { | |
| "epoch": 0.40080225942450165, | |
| "grad_norm": 812.0, | |
| "learning_rate": 0.001, | |
| "loss": 1663.3135, | |
| "step": 19584 | |
| }, | |
| { | |
| "epoch": 0.4017846179034833, | |
| "grad_norm": 820.0, | |
| "learning_rate": 0.001, | |
| "loss": 1648.4126, | |
| "step": 19632 | |
| }, | |
| { | |
| "epoch": 0.4027669763824649, | |
| "grad_norm": 828.0, | |
| "learning_rate": 0.001, | |
| "loss": 1629.137, | |
| "step": 19680 | |
| }, | |
| { | |
| "epoch": 0.40374933486144654, | |
| "grad_norm": 840.0, | |
| "learning_rate": 0.001, | |
| "loss": 1675.1574, | |
| "step": 19728 | |
| }, | |
| { | |
| "epoch": 0.40473169334042813, | |
| "grad_norm": 892.0, | |
| "learning_rate": 0.001, | |
| "loss": 1651.4735, | |
| "step": 19776 | |
| }, | |
| { | |
| "epoch": 0.4057140518194098, | |
| "grad_norm": 732.0, | |
| "learning_rate": 0.001, | |
| "loss": 1659.3854, | |
| "step": 19824 | |
| }, | |
| { | |
| "epoch": 0.4066964102983914, | |
| "grad_norm": 808.0, | |
| "learning_rate": 0.001, | |
| "loss": 1676.7249, | |
| "step": 19872 | |
| }, | |
| { | |
| "epoch": 0.407678768777373, | |
| "grad_norm": 1088.0, | |
| "learning_rate": 0.001, | |
| "loss": 1658.8581, | |
| "step": 19920 | |
| }, | |
| { | |
| "epoch": 0.4086611272563546, | |
| "grad_norm": 772.0, | |
| "learning_rate": 0.001, | |
| "loss": 1663.8711, | |
| "step": 19968 | |
| }, | |
| { | |
| "epoch": 0.40964348573533627, | |
| "grad_norm": 724.0, | |
| "learning_rate": 0.001, | |
| "loss": 1657.056, | |
| "step": 20016 | |
| }, | |
| { | |
| "epoch": 0.41062584421431786, | |
| "grad_norm": 920.0, | |
| "learning_rate": 0.001, | |
| "loss": 1637.993, | |
| "step": 20064 | |
| }, | |
| { | |
| "epoch": 0.4116082026932995, | |
| "grad_norm": 636.0, | |
| "learning_rate": 0.001, | |
| "loss": 1644.8833, | |
| "step": 20112 | |
| }, | |
| { | |
| "epoch": 0.4125905611722811, | |
| "grad_norm": 956.0, | |
| "learning_rate": 0.001, | |
| "loss": 1656.4339, | |
| "step": 20160 | |
| }, | |
| { | |
| "epoch": 0.41357291965126275, | |
| "grad_norm": 792.0, | |
| "learning_rate": 0.001, | |
| "loss": 1652.4888, | |
| "step": 20208 | |
| }, | |
| { | |
| "epoch": 0.41455527813024434, | |
| "grad_norm": 832.0, | |
| "learning_rate": 0.001, | |
| "loss": 1642.5142, | |
| "step": 20256 | |
| }, | |
| { | |
| "epoch": 0.415537636609226, | |
| "grad_norm": 724.0, | |
| "learning_rate": 0.001, | |
| "loss": 1668.1183, | |
| "step": 20304 | |
| }, | |
| { | |
| "epoch": 0.4165199950882076, | |
| "grad_norm": 796.0, | |
| "learning_rate": 0.001, | |
| "loss": 1660.486, | |
| "step": 20352 | |
| }, | |
| { | |
| "epoch": 0.41750235356718923, | |
| "grad_norm": 788.0, | |
| "learning_rate": 0.001, | |
| "loss": 1655.4614, | |
| "step": 20400 | |
| }, | |
| { | |
| "epoch": 0.4184847120461708, | |
| "grad_norm": 844.0, | |
| "learning_rate": 0.001, | |
| "loss": 1650.6449, | |
| "step": 20448 | |
| }, | |
| { | |
| "epoch": 0.4194670705251525, | |
| "grad_norm": 1080.0, | |
| "learning_rate": 0.001, | |
| "loss": 1639.0168, | |
| "step": 20496 | |
| }, | |
| { | |
| "epoch": 0.42044942900413407, | |
| "grad_norm": 944.0, | |
| "learning_rate": 0.001, | |
| "loss": 1655.6951, | |
| "step": 20544 | |
| }, | |
| { | |
| "epoch": 0.4214317874831157, | |
| "grad_norm": 716.0, | |
| "learning_rate": 0.001, | |
| "loss": 1644.007, | |
| "step": 20592 | |
| }, | |
| { | |
| "epoch": 0.4224141459620973, | |
| "grad_norm": 896.0, | |
| "learning_rate": 0.001, | |
| "loss": 1614.1253, | |
| "step": 20640 | |
| }, | |
| { | |
| "epoch": 0.42339650444107896, | |
| "grad_norm": 736.0, | |
| "learning_rate": 0.001, | |
| "loss": 1637.6761, | |
| "step": 20688 | |
| }, | |
| { | |
| "epoch": 0.42437886292006055, | |
| "grad_norm": 752.0, | |
| "learning_rate": 0.001, | |
| "loss": 1635.8405, | |
| "step": 20736 | |
| }, | |
| { | |
| "epoch": 0.4253612213990422, | |
| "grad_norm": 768.0, | |
| "learning_rate": 0.001, | |
| "loss": 1653.1647, | |
| "step": 20784 | |
| }, | |
| { | |
| "epoch": 0.42634357987802385, | |
| "grad_norm": 832.0, | |
| "learning_rate": 0.001, | |
| "loss": 1657.5373, | |
| "step": 20832 | |
| }, | |
| { | |
| "epoch": 0.42732593835700544, | |
| "grad_norm": 832.0, | |
| "learning_rate": 0.001, | |
| "loss": 1649.7998, | |
| "step": 20880 | |
| }, | |
| { | |
| "epoch": 0.4283082968359871, | |
| "grad_norm": 848.0, | |
| "learning_rate": 0.001, | |
| "loss": 1639.043, | |
| "step": 20928 | |
| }, | |
| { | |
| "epoch": 0.4292906553149687, | |
| "grad_norm": 716.0, | |
| "learning_rate": 0.001, | |
| "loss": 1656.0706, | |
| "step": 20976 | |
| }, | |
| { | |
| "epoch": 0.43027301379395033, | |
| "grad_norm": 924.0, | |
| "learning_rate": 0.001, | |
| "loss": 1640.4744, | |
| "step": 21024 | |
| }, | |
| { | |
| "epoch": 0.4312553722729319, | |
| "grad_norm": 832.0, | |
| "learning_rate": 0.001, | |
| "loss": 1640.8403, | |
| "step": 21072 | |
| }, | |
| { | |
| "epoch": 0.4322377307519136, | |
| "grad_norm": 828.0, | |
| "learning_rate": 0.001, | |
| "loss": 1638.4172, | |
| "step": 21120 | |
| }, | |
| { | |
| "epoch": 0.43322008923089517, | |
| "grad_norm": 732.0, | |
| "learning_rate": 0.001, | |
| "loss": 1662.0506, | |
| "step": 21168 | |
| }, | |
| { | |
| "epoch": 0.4342024477098768, | |
| "grad_norm": 752.0, | |
| "learning_rate": 0.001, | |
| "loss": 1634.0928, | |
| "step": 21216 | |
| }, | |
| { | |
| "epoch": 0.4351848061888584, | |
| "grad_norm": 976.0, | |
| "learning_rate": 0.001, | |
| "loss": 1650.1844, | |
| "step": 21264 | |
| }, | |
| { | |
| "epoch": 0.43616716466784006, | |
| "grad_norm": 864.0, | |
| "learning_rate": 0.001, | |
| "loss": 1637.9504, | |
| "step": 21312 | |
| }, | |
| { | |
| "epoch": 0.43714952314682165, | |
| "grad_norm": 660.0, | |
| "learning_rate": 0.001, | |
| "loss": 1651.0807, | |
| "step": 21360 | |
| }, | |
| { | |
| "epoch": 0.4381318816258033, | |
| "grad_norm": 884.0, | |
| "learning_rate": 0.001, | |
| "loss": 1639.6099, | |
| "step": 21408 | |
| }, | |
| { | |
| "epoch": 0.4391142401047849, | |
| "grad_norm": 752.0, | |
| "learning_rate": 0.001, | |
| "loss": 1636.7961, | |
| "step": 21456 | |
| }, | |
| { | |
| "epoch": 0.44009659858376654, | |
| "grad_norm": 704.0, | |
| "learning_rate": 0.001, | |
| "loss": 1630.2891, | |
| "step": 21504 | |
| }, | |
| { | |
| "epoch": 0.44107895706274813, | |
| "grad_norm": 868.0, | |
| "learning_rate": 0.001, | |
| "loss": 1651.1029, | |
| "step": 21552 | |
| }, | |
| { | |
| "epoch": 0.4420613155417298, | |
| "grad_norm": 1200.0, | |
| "learning_rate": 0.001, | |
| "loss": 1658.1079, | |
| "step": 21600 | |
| }, | |
| { | |
| "epoch": 0.4430436740207114, | |
| "grad_norm": 820.0, | |
| "learning_rate": 0.001, | |
| "loss": 1650.035, | |
| "step": 21648 | |
| }, | |
| { | |
| "epoch": 0.444026032499693, | |
| "grad_norm": 628.0, | |
| "learning_rate": 0.001, | |
| "loss": 1651.446, | |
| "step": 21696 | |
| }, | |
| { | |
| "epoch": 0.4450083909786746, | |
| "grad_norm": 936.0, | |
| "learning_rate": 0.001, | |
| "loss": 1651.8545, | |
| "step": 21744 | |
| }, | |
| { | |
| "epoch": 0.44599074945765627, | |
| "grad_norm": 724.0, | |
| "learning_rate": 0.001, | |
| "loss": 1633.7931, | |
| "step": 21792 | |
| }, | |
| { | |
| "epoch": 0.44697310793663786, | |
| "grad_norm": 868.0, | |
| "learning_rate": 0.001, | |
| "loss": 1643.6271, | |
| "step": 21840 | |
| }, | |
| { | |
| "epoch": 0.4479554664156195, | |
| "grad_norm": 884.0, | |
| "learning_rate": 0.001, | |
| "loss": 1628.9087, | |
| "step": 21888 | |
| }, | |
| { | |
| "epoch": 0.4489378248946011, | |
| "grad_norm": 808.0, | |
| "learning_rate": 0.001, | |
| "loss": 1633.3311, | |
| "step": 21936 | |
| }, | |
| { | |
| "epoch": 0.44992018337358275, | |
| "grad_norm": 776.0, | |
| "learning_rate": 0.001, | |
| "loss": 1636.3483, | |
| "step": 21984 | |
| }, | |
| { | |
| "epoch": 0.45090254185256434, | |
| "grad_norm": 740.0, | |
| "learning_rate": 0.001, | |
| "loss": 1627.1842, | |
| "step": 22032 | |
| }, | |
| { | |
| "epoch": 0.451884900331546, | |
| "grad_norm": 832.0, | |
| "learning_rate": 0.001, | |
| "loss": 1632.6536, | |
| "step": 22080 | |
| }, | |
| { | |
| "epoch": 0.4528672588105276, | |
| "grad_norm": 920.0, | |
| "learning_rate": 0.001, | |
| "loss": 1663.3418, | |
| "step": 22128 | |
| }, | |
| { | |
| "epoch": 0.45384961728950923, | |
| "grad_norm": 752.0, | |
| "learning_rate": 0.001, | |
| "loss": 1635.7738, | |
| "step": 22176 | |
| }, | |
| { | |
| "epoch": 0.4548319757684908, | |
| "grad_norm": 752.0, | |
| "learning_rate": 0.001, | |
| "loss": 1636.0459, | |
| "step": 22224 | |
| }, | |
| { | |
| "epoch": 0.4558143342474725, | |
| "grad_norm": 892.0, | |
| "learning_rate": 0.001, | |
| "loss": 1627.4956, | |
| "step": 22272 | |
| }, | |
| { | |
| "epoch": 0.45679669272645407, | |
| "grad_norm": 860.0, | |
| "learning_rate": 0.001, | |
| "loss": 1628.1626, | |
| "step": 22320 | |
| }, | |
| { | |
| "epoch": 0.4577790512054357, | |
| "grad_norm": 776.0, | |
| "learning_rate": 0.001, | |
| "loss": 1628.0701, | |
| "step": 22368 | |
| }, | |
| { | |
| "epoch": 0.45876140968441737, | |
| "grad_norm": 792.0, | |
| "learning_rate": 0.001, | |
| "loss": 1644.1922, | |
| "step": 22416 | |
| }, | |
| { | |
| "epoch": 0.45974376816339896, | |
| "grad_norm": 776.0, | |
| "learning_rate": 0.001, | |
| "loss": 1608.5988, | |
| "step": 22464 | |
| }, | |
| { | |
| "epoch": 0.4607261266423806, | |
| "grad_norm": 860.0, | |
| "learning_rate": 0.001, | |
| "loss": 1637.166, | |
| "step": 22512 | |
| }, | |
| { | |
| "epoch": 0.4617084851213622, | |
| "grad_norm": 924.0, | |
| "learning_rate": 0.001, | |
| "loss": 1626.6854, | |
| "step": 22560 | |
| }, | |
| { | |
| "epoch": 0.46269084360034385, | |
| "grad_norm": 876.0, | |
| "learning_rate": 0.001, | |
| "loss": 1639.1245, | |
| "step": 22608 | |
| }, | |
| { | |
| "epoch": 0.46367320207932544, | |
| "grad_norm": 936.0, | |
| "learning_rate": 0.001, | |
| "loss": 1634.3815, | |
| "step": 22656 | |
| }, | |
| { | |
| "epoch": 0.4646555605583071, | |
| "grad_norm": 912.0, | |
| "learning_rate": 0.001, | |
| "loss": 1606.1912, | |
| "step": 22704 | |
| }, | |
| { | |
| "epoch": 0.4656379190372887, | |
| "grad_norm": 952.0, | |
| "learning_rate": 0.001, | |
| "loss": 1620.5391, | |
| "step": 22752 | |
| }, | |
| { | |
| "epoch": 0.46662027751627033, | |
| "grad_norm": 960.0, | |
| "learning_rate": 0.001, | |
| "loss": 1612.0667, | |
| "step": 22800 | |
| }, | |
| { | |
| "epoch": 0.4676026359952519, | |
| "grad_norm": 832.0, | |
| "learning_rate": 0.001, | |
| "loss": 1651.9868, | |
| "step": 22848 | |
| }, | |
| { | |
| "epoch": 0.4685849944742336, | |
| "grad_norm": 712.0, | |
| "learning_rate": 0.001, | |
| "loss": 1629.237, | |
| "step": 22896 | |
| }, | |
| { | |
| "epoch": 0.46956735295321517, | |
| "grad_norm": 864.0, | |
| "learning_rate": 0.001, | |
| "loss": 1618.2004, | |
| "step": 22944 | |
| }, | |
| { | |
| "epoch": 0.4705497114321968, | |
| "grad_norm": 728.0, | |
| "learning_rate": 0.001, | |
| "loss": 1625.5379, | |
| "step": 22992 | |
| }, | |
| { | |
| "epoch": 0.4715320699111784, | |
| "grad_norm": 836.0, | |
| "learning_rate": 0.001, | |
| "loss": 1622.8146, | |
| "step": 23040 | |
| }, | |
| { | |
| "epoch": 0.47251442839016006, | |
| "grad_norm": 1064.0, | |
| "learning_rate": 0.001, | |
| "loss": 1623.9705, | |
| "step": 23088 | |
| }, | |
| { | |
| "epoch": 0.47349678686914165, | |
| "grad_norm": 860.0, | |
| "learning_rate": 0.001, | |
| "loss": 1626.2383, | |
| "step": 23136 | |
| }, | |
| { | |
| "epoch": 0.4744791453481233, | |
| "grad_norm": 1120.0, | |
| "learning_rate": 0.001, | |
| "loss": 1634.2668, | |
| "step": 23184 | |
| }, | |
| { | |
| "epoch": 0.4754615038271049, | |
| "grad_norm": 796.0, | |
| "learning_rate": 0.001, | |
| "loss": 1642.5649, | |
| "step": 23232 | |
| }, | |
| { | |
| "epoch": 0.47644386230608654, | |
| "grad_norm": 1072.0, | |
| "learning_rate": 0.001, | |
| "loss": 1633.4873, | |
| "step": 23280 | |
| }, | |
| { | |
| "epoch": 0.47742622078506813, | |
| "grad_norm": 904.0, | |
| "learning_rate": 0.001, | |
| "loss": 1604.186, | |
| "step": 23328 | |
| }, | |
| { | |
| "epoch": 0.4784085792640498, | |
| "grad_norm": 852.0, | |
| "learning_rate": 0.001, | |
| "loss": 1608.5158, | |
| "step": 23376 | |
| }, | |
| { | |
| "epoch": 0.4793909377430314, | |
| "grad_norm": 860.0, | |
| "learning_rate": 0.001, | |
| "loss": 1624.3991, | |
| "step": 23424 | |
| }, | |
| { | |
| "epoch": 0.480373296222013, | |
| "grad_norm": 820.0, | |
| "learning_rate": 0.001, | |
| "loss": 1640.4948, | |
| "step": 23472 | |
| }, | |
| { | |
| "epoch": 0.4813556547009946, | |
| "grad_norm": 700.0, | |
| "learning_rate": 0.001, | |
| "loss": 1593.3931, | |
| "step": 23520 | |
| }, | |
| { | |
| "epoch": 0.48233801317997627, | |
| "grad_norm": 824.0, | |
| "learning_rate": 0.001, | |
| "loss": 1622.2378, | |
| "step": 23568 | |
| }, | |
| { | |
| "epoch": 0.48332037165895786, | |
| "grad_norm": 936.0, | |
| "learning_rate": 0.001, | |
| "loss": 1612.6034, | |
| "step": 23616 | |
| }, | |
| { | |
| "epoch": 0.4843027301379395, | |
| "grad_norm": 732.0, | |
| "learning_rate": 0.001, | |
| "loss": 1624.4165, | |
| "step": 23664 | |
| }, | |
| { | |
| "epoch": 0.4852850886169211, | |
| "grad_norm": 820.0, | |
| "learning_rate": 0.001, | |
| "loss": 1634.2424, | |
| "step": 23712 | |
| }, | |
| { | |
| "epoch": 0.48626744709590275, | |
| "grad_norm": 896.0, | |
| "learning_rate": 0.001, | |
| "loss": 1616.617, | |
| "step": 23760 | |
| }, | |
| { | |
| "epoch": 0.48724980557488434, | |
| "grad_norm": 732.0, | |
| "learning_rate": 0.001, | |
| "loss": 1629.8065, | |
| "step": 23808 | |
| }, | |
| { | |
| "epoch": 0.488232164053866, | |
| "grad_norm": 876.0, | |
| "learning_rate": 0.001, | |
| "loss": 1611.0832, | |
| "step": 23856 | |
| }, | |
| { | |
| "epoch": 0.4892145225328476, | |
| "grad_norm": 1004.0, | |
| "learning_rate": 0.001, | |
| "loss": 1596.9705, | |
| "step": 23904 | |
| }, | |
| { | |
| "epoch": 0.49019688101182923, | |
| "grad_norm": 812.0, | |
| "learning_rate": 0.001, | |
| "loss": 1612.6437, | |
| "step": 23952 | |
| }, | |
| { | |
| "epoch": 0.4911792394908109, | |
| "grad_norm": 764.0, | |
| "learning_rate": 0.001, | |
| "loss": 1612.0643, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 0.4921615979697925, | |
| "grad_norm": 944.0, | |
| "learning_rate": 0.001, | |
| "loss": 1620.1268, | |
| "step": 24048 | |
| }, | |
| { | |
| "epoch": 0.4931439564487741, | |
| "grad_norm": 920.0, | |
| "learning_rate": 0.001, | |
| "loss": 1619.8875, | |
| "step": 24096 | |
| }, | |
| { | |
| "epoch": 0.4941263149277557, | |
| "grad_norm": 808.0, | |
| "learning_rate": 0.001, | |
| "loss": 1608.6463, | |
| "step": 24144 | |
| }, | |
| { | |
| "epoch": 0.49510867340673737, | |
| "grad_norm": 784.0, | |
| "learning_rate": 0.001, | |
| "loss": 1609.1462, | |
| "step": 24192 | |
| }, | |
| { | |
| "epoch": 0.49609103188571896, | |
| "grad_norm": 796.0, | |
| "learning_rate": 0.001, | |
| "loss": 1610.4935, | |
| "step": 24240 | |
| }, | |
| { | |
| "epoch": 0.4970733903647006, | |
| "grad_norm": 896.0, | |
| "learning_rate": 0.001, | |
| "loss": 1622.8371, | |
| "step": 24288 | |
| }, | |
| { | |
| "epoch": 0.4980557488436822, | |
| "grad_norm": 768.0, | |
| "learning_rate": 0.001, | |
| "loss": 1617.0732, | |
| "step": 24336 | |
| }, | |
| { | |
| "epoch": 0.49903810732266385, | |
| "grad_norm": 920.0, | |
| "learning_rate": 0.001, | |
| "loss": 1615.2142, | |
| "step": 24384 | |
| }, | |
| { | |
| "epoch": 0.49997953419835456, | |
| "eval_loss": 1546.7662353515625, | |
| "eval_runtime": 9.0072, | |
| "eval_samples_per_second": 111.022, | |
| "eval_steps_per_second": 1.443, | |
| "step": 24430 | |
| } | |
| ], | |
| "logging_steps": 48, | |
| "max_steps": 48862, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9223372036854775807, | |
| "save_steps": 4886, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.7859053283033743e+18, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |