| { |
| "best_metric": 3.493159770965576, |
| "best_model_checkpoint": "/scratch/cl5625/exceptions/models/100M__495/checkpoint-30000", |
| "epoch": 3.234501347708895, |
| "eval_steps": 1000, |
| "global_step": 30000, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.005390835579514825, |
| "grad_norm": 0.9464320540428162, |
| "learning_rate": 0.000276, |
| "loss": 8.7571, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.01078167115902965, |
| "grad_norm": 1.1756318807601929, |
| "learning_rate": 0.0005759999999999999, |
| "loss": 7.0078, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.016172506738544475, |
| "grad_norm": 1.551442265510559, |
| "learning_rate": 0.000599702104695089, |
| "loss": 6.6048, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.0215633423180593, |
| "grad_norm": 0.7034249901771545, |
| "learning_rate": 0.0005993783054506205, |
| "loss": 6.3332, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.026954177897574125, |
| "grad_norm": 1.0177743434906006, |
| "learning_rate": 0.0005990545062061521, |
| "loss": 6.1389, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.03234501347708895, |
| "grad_norm": 1.3281992673873901, |
| "learning_rate": 0.0005987307069616836, |
| "loss": 6.0274, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.03773584905660377, |
| "grad_norm": 1.3769625425338745, |
| "learning_rate": 0.0005984069077172153, |
| "loss": 5.9227, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.0431266846361186, |
| "grad_norm": 1.3584883213043213, |
| "learning_rate": 0.0005980831084727469, |
| "loss": 5.8655, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.04851752021563342, |
| "grad_norm": 0.9795990586280823, |
| "learning_rate": 0.0005977593092282784, |
| "loss": 5.78, |
| "step": 450 |
| }, |
| { |
| "epoch": 0.05390835579514825, |
| "grad_norm": 1.4081448316574097, |
| "learning_rate": 0.00059743550998381, |
| "loss": 5.7304, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.05929919137466307, |
| "grad_norm": 1.7430726289749146, |
| "learning_rate": 0.0005971117107393416, |
| "loss": 5.6254, |
| "step": 550 |
| }, |
| { |
| "epoch": 0.0646900269541779, |
| "grad_norm": 1.4288103580474854, |
| "learning_rate": 0.0005967879114948732, |
| "loss": 5.5997, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.07008086253369272, |
| "grad_norm": 0.9352820515632629, |
| "learning_rate": 0.0005964641122504047, |
| "loss": 5.5009, |
| "step": 650 |
| }, |
| { |
| "epoch": 0.07547169811320754, |
| "grad_norm": 1.0235956907272339, |
| "learning_rate": 0.0005961403130059363, |
| "loss": 5.4834, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.08086253369272237, |
| "grad_norm": 1.1283656358718872, |
| "learning_rate": 0.0005958165137614678, |
| "loss": 5.4096, |
| "step": 750 |
| }, |
| { |
| "epoch": 0.0862533692722372, |
| "grad_norm": 1.1757662296295166, |
| "learning_rate": 0.0005954927145169995, |
| "loss": 5.3508, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.09164420485175202, |
| "grad_norm": 1.2652894258499146, |
| "learning_rate": 0.0005951689152725309, |
| "loss": 5.3137, |
| "step": 850 |
| }, |
| { |
| "epoch": 0.09703504043126684, |
| "grad_norm": 0.8220955729484558, |
| "learning_rate": 0.0005948451160280626, |
| "loss": 5.234, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.10242587601078167, |
| "grad_norm": 1.0178474187850952, |
| "learning_rate": 0.0005945213167835941, |
| "loss": 5.2211, |
| "step": 950 |
| }, |
| { |
| "epoch": 0.1078167115902965, |
| "grad_norm": 1.0003489255905151, |
| "learning_rate": 0.0005941975175391257, |
| "loss": 5.1665, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.1078167115902965, |
| "eval_accuracy": 0.2206364212520268, |
| "eval_loss": 5.09123420715332, |
| "eval_runtime": 146.3271, |
| "eval_samples_per_second": 123.087, |
| "eval_steps_per_second": 7.695, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.11320754716981132, |
| "grad_norm": 1.0129122734069824, |
| "learning_rate": 0.0005938737182946572, |
| "loss": 5.1253, |
| "step": 1050 |
| }, |
| { |
| "epoch": 0.11859838274932614, |
| "grad_norm": 1.2447277307510376, |
| "learning_rate": 0.0005935499190501888, |
| "loss": 5.0843, |
| "step": 1100 |
| }, |
| { |
| "epoch": 0.12398921832884097, |
| "grad_norm": 1.1244721412658691, |
| "learning_rate": 0.0005932261198057204, |
| "loss": 5.0632, |
| "step": 1150 |
| }, |
| { |
| "epoch": 0.1293800539083558, |
| "grad_norm": 1.0002185106277466, |
| "learning_rate": 0.000592902320561252, |
| "loss": 5.0409, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.1347708894878706, |
| "grad_norm": 1.237203598022461, |
| "learning_rate": 0.0005925785213167835, |
| "loss": 5.0158, |
| "step": 1250 |
| }, |
| { |
| "epoch": 0.14016172506738545, |
| "grad_norm": 0.9483816623687744, |
| "learning_rate": 0.0005922547220723151, |
| "loss": 4.9578, |
| "step": 1300 |
| }, |
| { |
| "epoch": 0.14555256064690028, |
| "grad_norm": 1.1492619514465332, |
| "learning_rate": 0.0005919309228278468, |
| "loss": 4.9422, |
| "step": 1350 |
| }, |
| { |
| "epoch": 0.1509433962264151, |
| "grad_norm": 0.8103901743888855, |
| "learning_rate": 0.0005916071235833783, |
| "loss": 4.9161, |
| "step": 1400 |
| }, |
| { |
| "epoch": 0.15633423180592992, |
| "grad_norm": 0.9849348068237305, |
| "learning_rate": 0.0005912833243389097, |
| "loss": 4.8937, |
| "step": 1450 |
| }, |
| { |
| "epoch": 0.16172506738544473, |
| "grad_norm": 1.0992136001586914, |
| "learning_rate": 0.0005909595250944414, |
| "loss": 4.8771, |
| "step": 1500 |
| }, |
| { |
| "epoch": 0.16711590296495957, |
| "grad_norm": 0.7955754995346069, |
| "learning_rate": 0.000590635725849973, |
| "loss": 4.8337, |
| "step": 1550 |
| }, |
| { |
| "epoch": 0.1725067385444744, |
| "grad_norm": 0.8510985970497131, |
| "learning_rate": 0.0005903119266055045, |
| "loss": 4.8487, |
| "step": 1600 |
| }, |
| { |
| "epoch": 0.1778975741239892, |
| "grad_norm": 0.8423263430595398, |
| "learning_rate": 0.0005899881273610361, |
| "loss": 4.8251, |
| "step": 1650 |
| }, |
| { |
| "epoch": 0.18328840970350405, |
| "grad_norm": 0.8307051658630371, |
| "learning_rate": 0.0005896643281165677, |
| "loss": 4.7528, |
| "step": 1700 |
| }, |
| { |
| "epoch": 0.18867924528301888, |
| "grad_norm": 1.171060562133789, |
| "learning_rate": 0.0005893405288720993, |
| "loss": 4.7617, |
| "step": 1750 |
| }, |
| { |
| "epoch": 0.1940700808625337, |
| "grad_norm": 1.0469765663146973, |
| "learning_rate": 0.0005890167296276308, |
| "loss": 4.7597, |
| "step": 1800 |
| }, |
| { |
| "epoch": 0.19946091644204852, |
| "grad_norm": 0.8637204170227051, |
| "learning_rate": 0.0005886929303831624, |
| "loss": 4.7128, |
| "step": 1850 |
| }, |
| { |
| "epoch": 0.20485175202156333, |
| "grad_norm": 0.9173099398612976, |
| "learning_rate": 0.0005883691311386939, |
| "loss": 4.7067, |
| "step": 1900 |
| }, |
| { |
| "epoch": 0.21024258760107817, |
| "grad_norm": 0.9003922343254089, |
| "learning_rate": 0.0005880453318942256, |
| "loss": 4.6783, |
| "step": 1950 |
| }, |
| { |
| "epoch": 0.215633423180593, |
| "grad_norm": 0.9870163798332214, |
| "learning_rate": 0.0005877215326497571, |
| "loss": 4.6555, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.215633423180593, |
| "eval_accuracy": 0.260708366848222, |
| "eval_loss": 4.6007513999938965, |
| "eval_runtime": 144.747, |
| "eval_samples_per_second": 124.431, |
| "eval_steps_per_second": 7.779, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.2210242587601078, |
| "grad_norm": 0.8778396248817444, |
| "learning_rate": 0.0005873977334052887, |
| "loss": 4.6696, |
| "step": 2050 |
| }, |
| { |
| "epoch": 0.22641509433962265, |
| "grad_norm": 0.8375086188316345, |
| "learning_rate": 0.0005870739341608202, |
| "loss": 4.6395, |
| "step": 2100 |
| }, |
| { |
| "epoch": 0.23180592991913745, |
| "grad_norm": 0.8078502416610718, |
| "learning_rate": 0.0005867501349163519, |
| "loss": 4.5823, |
| "step": 2150 |
| }, |
| { |
| "epoch": 0.2371967654986523, |
| "grad_norm": 0.8603857755661011, |
| "learning_rate": 0.0005864263356718833, |
| "loss": 4.6175, |
| "step": 2200 |
| }, |
| { |
| "epoch": 0.24258760107816713, |
| "grad_norm": 0.8307099342346191, |
| "learning_rate": 0.000586102536427415, |
| "loss": 4.5541, |
| "step": 2250 |
| }, |
| { |
| "epoch": 0.24797843665768193, |
| "grad_norm": 1.0236228704452515, |
| "learning_rate": 0.0005857787371829465, |
| "loss": 4.5368, |
| "step": 2300 |
| }, |
| { |
| "epoch": 0.25336927223719674, |
| "grad_norm": 0.9307533502578735, |
| "learning_rate": 0.0005854549379384781, |
| "loss": 4.5326, |
| "step": 2350 |
| }, |
| { |
| "epoch": 0.2587601078167116, |
| "grad_norm": 1.4191182851791382, |
| "learning_rate": 0.0005851311386940096, |
| "loss": 4.5253, |
| "step": 2400 |
| }, |
| { |
| "epoch": 0.2641509433962264, |
| "grad_norm": 1.015572428703308, |
| "learning_rate": 0.0005848073394495412, |
| "loss": 4.5272, |
| "step": 2450 |
| }, |
| { |
| "epoch": 0.2695417789757412, |
| "grad_norm": 0.8903120756149292, |
| "learning_rate": 0.0005844835402050728, |
| "loss": 4.5219, |
| "step": 2500 |
| }, |
| { |
| "epoch": 0.2749326145552561, |
| "grad_norm": 1.0123368501663208, |
| "learning_rate": 0.0005841597409606044, |
| "loss": 4.4772, |
| "step": 2550 |
| }, |
| { |
| "epoch": 0.2803234501347709, |
| "grad_norm": 0.925151526927948, |
| "learning_rate": 0.000583835941716136, |
| "loss": 4.4624, |
| "step": 2600 |
| }, |
| { |
| "epoch": 0.2857142857142857, |
| "grad_norm": 1.1478705406188965, |
| "learning_rate": 0.0005835121424716675, |
| "loss": 4.4605, |
| "step": 2650 |
| }, |
| { |
| "epoch": 0.29110512129380056, |
| "grad_norm": 1.0130943059921265, |
| "learning_rate": 0.0005831883432271992, |
| "loss": 4.4529, |
| "step": 2700 |
| }, |
| { |
| "epoch": 0.29649595687331537, |
| "grad_norm": 0.991671621799469, |
| "learning_rate": 0.0005828645439827307, |
| "loss": 4.4101, |
| "step": 2750 |
| }, |
| { |
| "epoch": 0.3018867924528302, |
| "grad_norm": 0.8699747323989868, |
| "learning_rate": 0.0005825407447382622, |
| "loss": 4.4505, |
| "step": 2800 |
| }, |
| { |
| "epoch": 0.30727762803234504, |
| "grad_norm": 1.0381041765213013, |
| "learning_rate": 0.0005822169454937938, |
| "loss": 4.4171, |
| "step": 2850 |
| }, |
| { |
| "epoch": 0.31266846361185985, |
| "grad_norm": 1.0296149253845215, |
| "learning_rate": 0.0005818931462493254, |
| "loss": 4.4014, |
| "step": 2900 |
| }, |
| { |
| "epoch": 0.31805929919137466, |
| "grad_norm": 0.9270951747894287, |
| "learning_rate": 0.0005815693470048569, |
| "loss": 4.3968, |
| "step": 2950 |
| }, |
| { |
| "epoch": 0.32345013477088946, |
| "grad_norm": 0.7531670331954956, |
| "learning_rate": 0.0005812455477603885, |
| "loss": 4.3901, |
| "step": 3000 |
| }, |
| { |
| "epoch": 0.32345013477088946, |
| "eval_accuracy": 0.29159338482103947, |
| "eval_loss": 4.296316146850586, |
| "eval_runtime": 144.5889, |
| "eval_samples_per_second": 124.567, |
| "eval_steps_per_second": 7.788, |
| "step": 3000 |
| }, |
| { |
| "epoch": 0.3288409703504043, |
| "grad_norm": 0.8791877627372742, |
| "learning_rate": 0.0005809217485159201, |
| "loss": 4.3683, |
| "step": 3050 |
| }, |
| { |
| "epoch": 0.33423180592991913, |
| "grad_norm": 0.843708872795105, |
| "learning_rate": 0.0005805979492714517, |
| "loss": 4.3643, |
| "step": 3100 |
| }, |
| { |
| "epoch": 0.33962264150943394, |
| "grad_norm": 0.8123868703842163, |
| "learning_rate": 0.0005802741500269832, |
| "loss": 4.3262, |
| "step": 3150 |
| }, |
| { |
| "epoch": 0.3450134770889488, |
| "grad_norm": 0.9056026339530945, |
| "learning_rate": 0.0005799503507825148, |
| "loss": 4.3298, |
| "step": 3200 |
| }, |
| { |
| "epoch": 0.3504043126684636, |
| "grad_norm": 0.6567044854164124, |
| "learning_rate": 0.0005796265515380463, |
| "loss": 4.306, |
| "step": 3250 |
| }, |
| { |
| "epoch": 0.3557951482479784, |
| "grad_norm": 0.9095497131347656, |
| "learning_rate": 0.000579302752293578, |
| "loss": 4.328, |
| "step": 3300 |
| }, |
| { |
| "epoch": 0.3611859838274933, |
| "grad_norm": 0.8091392517089844, |
| "learning_rate": 0.0005789789530491095, |
| "loss": 4.3044, |
| "step": 3350 |
| }, |
| { |
| "epoch": 0.3665768194070081, |
| "grad_norm": 0.9594192504882812, |
| "learning_rate": 0.0005786551538046411, |
| "loss": 4.2963, |
| "step": 3400 |
| }, |
| { |
| "epoch": 0.3719676549865229, |
| "grad_norm": 0.72095787525177, |
| "learning_rate": 0.0005783313545601726, |
| "loss": 4.2802, |
| "step": 3450 |
| }, |
| { |
| "epoch": 0.37735849056603776, |
| "grad_norm": 0.7558978199958801, |
| "learning_rate": 0.0005780075553157043, |
| "loss": 4.2589, |
| "step": 3500 |
| }, |
| { |
| "epoch": 0.38274932614555257, |
| "grad_norm": 0.7811341881752014, |
| "learning_rate": 0.0005776837560712357, |
| "loss": 4.264, |
| "step": 3550 |
| }, |
| { |
| "epoch": 0.3881401617250674, |
| "grad_norm": 0.9022195935249329, |
| "learning_rate": 0.0005773599568267673, |
| "loss": 4.2651, |
| "step": 3600 |
| }, |
| { |
| "epoch": 0.3935309973045822, |
| "grad_norm": 0.9639933109283447, |
| "learning_rate": 0.0005770361575822989, |
| "loss": 4.2789, |
| "step": 3650 |
| }, |
| { |
| "epoch": 0.39892183288409705, |
| "grad_norm": 0.9333063960075378, |
| "learning_rate": 0.0005767123583378305, |
| "loss": 4.24, |
| "step": 3700 |
| }, |
| { |
| "epoch": 0.40431266846361186, |
| "grad_norm": 0.7508504986763, |
| "learning_rate": 0.000576388559093362, |
| "loss": 4.2359, |
| "step": 3750 |
| }, |
| { |
| "epoch": 0.40970350404312667, |
| "grad_norm": 0.7458257079124451, |
| "learning_rate": 0.0005760647598488936, |
| "loss": 4.2555, |
| "step": 3800 |
| }, |
| { |
| "epoch": 0.41509433962264153, |
| "grad_norm": 0.7919742465019226, |
| "learning_rate": 0.0005757409606044253, |
| "loss": 4.2366, |
| "step": 3850 |
| }, |
| { |
| "epoch": 0.42048517520215634, |
| "grad_norm": 0.9453123211860657, |
| "learning_rate": 0.0005754171613599568, |
| "loss": 4.2055, |
| "step": 3900 |
| }, |
| { |
| "epoch": 0.42587601078167114, |
| "grad_norm": 0.6152997612953186, |
| "learning_rate": 0.0005750933621154884, |
| "loss": 4.201, |
| "step": 3950 |
| }, |
| { |
| "epoch": 0.431266846361186, |
| "grad_norm": 0.8247600197792053, |
| "learning_rate": 0.0005747695628710199, |
| "loss": 4.2075, |
| "step": 4000 |
| }, |
| { |
| "epoch": 0.431266846361186, |
| "eval_accuracy": 0.30819272110400897, |
| "eval_loss": 4.134169101715088, |
| "eval_runtime": 144.6112, |
| "eval_samples_per_second": 124.548, |
| "eval_steps_per_second": 7.786, |
| "step": 4000 |
| }, |
| { |
| "epoch": 0.4366576819407008, |
| "grad_norm": 0.8243815898895264, |
| "learning_rate": 0.0005744457636265515, |
| "loss": 4.2068, |
| "step": 4050 |
| }, |
| { |
| "epoch": 0.4420485175202156, |
| "grad_norm": 0.6849672794342041, |
| "learning_rate": 0.0005741219643820831, |
| "loss": 4.2091, |
| "step": 4100 |
| }, |
| { |
| "epoch": 0.4474393530997305, |
| "grad_norm": 0.7505493760108948, |
| "learning_rate": 0.0005737981651376146, |
| "loss": 4.193, |
| "step": 4150 |
| }, |
| { |
| "epoch": 0.4528301886792453, |
| "grad_norm": 0.6466169953346252, |
| "learning_rate": 0.0005734743658931462, |
| "loss": 4.1829, |
| "step": 4200 |
| }, |
| { |
| "epoch": 0.4582210242587601, |
| "grad_norm": 0.5937972664833069, |
| "learning_rate": 0.0005731505666486778, |
| "loss": 4.1738, |
| "step": 4250 |
| }, |
| { |
| "epoch": 0.4636118598382749, |
| "grad_norm": 0.7906216979026794, |
| "learning_rate": 0.0005728267674042093, |
| "loss": 4.1668, |
| "step": 4300 |
| }, |
| { |
| "epoch": 0.46900269541778977, |
| "grad_norm": 0.744433581829071, |
| "learning_rate": 0.0005725029681597409, |
| "loss": 4.1723, |
| "step": 4350 |
| }, |
| { |
| "epoch": 0.4743935309973046, |
| "grad_norm": 0.8073228597640991, |
| "learning_rate": 0.0005721791689152725, |
| "loss": 4.158, |
| "step": 4400 |
| }, |
| { |
| "epoch": 0.4797843665768194, |
| "grad_norm": 0.8262885808944702, |
| "learning_rate": 0.0005718553696708041, |
| "loss": 4.1602, |
| "step": 4450 |
| }, |
| { |
| "epoch": 0.48517520215633425, |
| "grad_norm": 0.6594825387001038, |
| "learning_rate": 0.0005715315704263356, |
| "loss": 4.1434, |
| "step": 4500 |
| }, |
| { |
| "epoch": 0.49056603773584906, |
| "grad_norm": 0.6674824357032776, |
| "learning_rate": 0.0005712077711818672, |
| "loss": 4.1315, |
| "step": 4550 |
| }, |
| { |
| "epoch": 0.49595687331536387, |
| "grad_norm": 0.7260637879371643, |
| "learning_rate": 0.0005708839719373987, |
| "loss": 4.1407, |
| "step": 4600 |
| }, |
| { |
| "epoch": 0.5013477088948787, |
| "grad_norm": 0.6827527284622192, |
| "learning_rate": 0.0005705601726929304, |
| "loss": 4.1196, |
| "step": 4650 |
| }, |
| { |
| "epoch": 0.5067385444743935, |
| "grad_norm": 0.774723470211029, |
| "learning_rate": 0.0005702363734484619, |
| "loss": 4.1256, |
| "step": 4700 |
| }, |
| { |
| "epoch": 0.5121293800539084, |
| "grad_norm": 0.6491437554359436, |
| "learning_rate": 0.0005699125742039935, |
| "loss": 4.1063, |
| "step": 4750 |
| }, |
| { |
| "epoch": 0.5175202156334232, |
| "grad_norm": 0.7277990579605103, |
| "learning_rate": 0.000569588774959525, |
| "loss": 4.1197, |
| "step": 4800 |
| }, |
| { |
| "epoch": 0.522911051212938, |
| "grad_norm": 0.6551647782325745, |
| "learning_rate": 0.0005692649757150567, |
| "loss": 4.1135, |
| "step": 4850 |
| }, |
| { |
| "epoch": 0.5283018867924528, |
| "grad_norm": 0.7088435292243958, |
| "learning_rate": 0.0005689411764705881, |
| "loss": 4.1188, |
| "step": 4900 |
| }, |
| { |
| "epoch": 0.5336927223719676, |
| "grad_norm": 0.7585951089859009, |
| "learning_rate": 0.0005686173772261197, |
| "loss": 4.0928, |
| "step": 4950 |
| }, |
| { |
| "epoch": 0.5390835579514824, |
| "grad_norm": 0.673933744430542, |
| "learning_rate": 0.0005682935779816514, |
| "loss": 4.0819, |
| "step": 5000 |
| }, |
| { |
| "epoch": 0.5390835579514824, |
| "eval_accuracy": 0.31859362744293795, |
| "eval_loss": 4.022838115692139, |
| "eval_runtime": 144.8217, |
| "eval_samples_per_second": 124.367, |
| "eval_steps_per_second": 7.775, |
| "step": 5000 |
| }, |
| { |
| "epoch": 0.5444743935309974, |
| "grad_norm": 0.617730438709259, |
| "learning_rate": 0.0005679697787371829, |
| "loss": 4.0793, |
| "step": 5050 |
| }, |
| { |
| "epoch": 0.5498652291105122, |
| "grad_norm": 0.6957946419715881, |
| "learning_rate": 0.0005676459794927145, |
| "loss": 4.0821, |
| "step": 5100 |
| }, |
| { |
| "epoch": 0.555256064690027, |
| "grad_norm": 0.6225258708000183, |
| "learning_rate": 0.000567322180248246, |
| "loss": 4.0873, |
| "step": 5150 |
| }, |
| { |
| "epoch": 0.5606469002695418, |
| "grad_norm": 0.6864507794380188, |
| "learning_rate": 0.0005669983810037777, |
| "loss": 4.0634, |
| "step": 5200 |
| }, |
| { |
| "epoch": 0.5660377358490566, |
| "grad_norm": 0.642590343952179, |
| "learning_rate": 0.0005666745817593092, |
| "loss": 4.0649, |
| "step": 5250 |
| }, |
| { |
| "epoch": 0.5714285714285714, |
| "grad_norm": 0.614847719669342, |
| "learning_rate": 0.0005663507825148408, |
| "loss": 4.0476, |
| "step": 5300 |
| }, |
| { |
| "epoch": 0.5768194070080862, |
| "grad_norm": 0.7870457768440247, |
| "learning_rate": 0.0005660269832703723, |
| "loss": 4.063, |
| "step": 5350 |
| }, |
| { |
| "epoch": 0.5822102425876011, |
| "grad_norm": 0.6789854764938354, |
| "learning_rate": 0.0005657031840259039, |
| "loss": 4.0518, |
| "step": 5400 |
| }, |
| { |
| "epoch": 0.5876010781671159, |
| "grad_norm": 0.6867729425430298, |
| "learning_rate": 0.0005653793847814355, |
| "loss": 4.0614, |
| "step": 5450 |
| }, |
| { |
| "epoch": 0.5929919137466307, |
| "grad_norm": 0.5899894833564758, |
| "learning_rate": 0.000565055585536967, |
| "loss": 4.0639, |
| "step": 5500 |
| }, |
| { |
| "epoch": 0.5983827493261455, |
| "grad_norm": 0.6574368476867676, |
| "learning_rate": 0.0005647317862924986, |
| "loss": 4.067, |
| "step": 5550 |
| }, |
| { |
| "epoch": 0.6037735849056604, |
| "grad_norm": 0.7431745529174805, |
| "learning_rate": 0.0005644079870480302, |
| "loss": 4.0425, |
| "step": 5600 |
| }, |
| { |
| "epoch": 0.6091644204851752, |
| "grad_norm": 0.6241595149040222, |
| "learning_rate": 0.0005640841878035617, |
| "loss": 4.0319, |
| "step": 5650 |
| }, |
| { |
| "epoch": 0.6145552560646901, |
| "grad_norm": 0.6736788749694824, |
| "learning_rate": 0.0005637603885590933, |
| "loss": 4.0366, |
| "step": 5700 |
| }, |
| { |
| "epoch": 0.6199460916442049, |
| "grad_norm": 0.6149032711982727, |
| "learning_rate": 0.0005634365893146248, |
| "loss": 4.0495, |
| "step": 5750 |
| }, |
| { |
| "epoch": 0.6253369272237197, |
| "grad_norm": 0.6543477177619934, |
| "learning_rate": 0.0005631127900701565, |
| "loss": 4.042, |
| "step": 5800 |
| }, |
| { |
| "epoch": 0.6307277628032345, |
| "grad_norm": 0.6215724945068359, |
| "learning_rate": 0.000562788990825688, |
| "loss": 4.0478, |
| "step": 5850 |
| }, |
| { |
| "epoch": 0.6361185983827493, |
| "grad_norm": 0.6606348752975464, |
| "learning_rate": 0.0005624651915812196, |
| "loss": 4.0192, |
| "step": 5900 |
| }, |
| { |
| "epoch": 0.6415094339622641, |
| "grad_norm": 0.7944669723510742, |
| "learning_rate": 0.0005621413923367511, |
| "loss": 4.012, |
| "step": 5950 |
| }, |
| { |
| "epoch": 0.6469002695417789, |
| "grad_norm": 0.6075884699821472, |
| "learning_rate": 0.0005618175930922828, |
| "loss": 4.0201, |
| "step": 6000 |
| }, |
| { |
| "epoch": 0.6469002695417789, |
| "eval_accuracy": 0.3257018080166491, |
| "eval_loss": 3.940925359725952, |
| "eval_runtime": 144.8645, |
| "eval_samples_per_second": 124.33, |
| "eval_steps_per_second": 7.773, |
| "step": 6000 |
| }, |
| { |
| "epoch": 0.6522911051212938, |
| "grad_norm": 0.7395045757293701, |
| "learning_rate": 0.0005614937938478143, |
| "loss": 3.9927, |
| "step": 6050 |
| }, |
| { |
| "epoch": 0.6576819407008087, |
| "grad_norm": 0.69764643907547, |
| "learning_rate": 0.0005611699946033459, |
| "loss": 4.0252, |
| "step": 6100 |
| }, |
| { |
| "epoch": 0.6630727762803235, |
| "grad_norm": 0.5907176733016968, |
| "learning_rate": 0.0005608461953588774, |
| "loss": 4.0071, |
| "step": 6150 |
| }, |
| { |
| "epoch": 0.6684636118598383, |
| "grad_norm": 0.7374498248100281, |
| "learning_rate": 0.000560522396114409, |
| "loss": 3.9854, |
| "step": 6200 |
| }, |
| { |
| "epoch": 0.6738544474393531, |
| "grad_norm": 0.5937222838401794, |
| "learning_rate": 0.0005601985968699405, |
| "loss": 4.003, |
| "step": 6250 |
| }, |
| { |
| "epoch": 0.6792452830188679, |
| "grad_norm": 0.6928643584251404, |
| "learning_rate": 0.0005598747976254721, |
| "loss": 4.022, |
| "step": 6300 |
| }, |
| { |
| "epoch": 0.6846361185983828, |
| "grad_norm": 0.6832301020622253, |
| "learning_rate": 0.0005595509983810038, |
| "loss": 3.9816, |
| "step": 6350 |
| }, |
| { |
| "epoch": 0.6900269541778976, |
| "grad_norm": 0.6227492690086365, |
| "learning_rate": 0.0005592271991365353, |
| "loss": 3.9855, |
| "step": 6400 |
| }, |
| { |
| "epoch": 0.6954177897574124, |
| "grad_norm": 0.6595360636711121, |
| "learning_rate": 0.0005589033998920669, |
| "loss": 3.9733, |
| "step": 6450 |
| }, |
| { |
| "epoch": 0.7008086253369272, |
| "grad_norm": 0.6538481116294861, |
| "learning_rate": 0.0005585796006475984, |
| "loss": 3.9867, |
| "step": 6500 |
| }, |
| { |
| "epoch": 0.706199460916442, |
| "grad_norm": 0.6099511384963989, |
| "learning_rate": 0.0005582558014031301, |
| "loss": 4.009, |
| "step": 6550 |
| }, |
| { |
| "epoch": 0.7115902964959568, |
| "grad_norm": 0.5673043131828308, |
| "learning_rate": 0.0005579320021586616, |
| "loss": 3.9638, |
| "step": 6600 |
| }, |
| { |
| "epoch": 0.7169811320754716, |
| "grad_norm": 0.7762152552604675, |
| "learning_rate": 0.0005576082029141932, |
| "loss": 3.9942, |
| "step": 6650 |
| }, |
| { |
| "epoch": 0.7223719676549866, |
| "grad_norm": 0.6117172241210938, |
| "learning_rate": 0.0005572844036697247, |
| "loss": 3.9672, |
| "step": 6700 |
| }, |
| { |
| "epoch": 0.7277628032345014, |
| "grad_norm": 0.6088191866874695, |
| "learning_rate": 0.0005569606044252563, |
| "loss": 3.9738, |
| "step": 6750 |
| }, |
| { |
| "epoch": 0.7331536388140162, |
| "grad_norm": 0.6414440274238586, |
| "learning_rate": 0.0005566368051807879, |
| "loss": 3.9641, |
| "step": 6800 |
| }, |
| { |
| "epoch": 0.738544474393531, |
| "grad_norm": 0.8121737241744995, |
| "learning_rate": 0.0005563130059363194, |
| "loss": 3.9423, |
| "step": 6850 |
| }, |
| { |
| "epoch": 0.7439353099730458, |
| "grad_norm": 0.7500906586647034, |
| "learning_rate": 0.000555989206691851, |
| "loss": 3.9677, |
| "step": 6900 |
| }, |
| { |
| "epoch": 0.7493261455525606, |
| "grad_norm": 0.6089574694633484, |
| "learning_rate": 0.0005556654074473826, |
| "loss": 3.9647, |
| "step": 6950 |
| }, |
| { |
| "epoch": 0.7547169811320755, |
| "grad_norm": 0.7751880288124084, |
| "learning_rate": 0.0005553416082029141, |
| "loss": 3.9519, |
| "step": 7000 |
| }, |
| { |
| "epoch": 0.7547169811320755, |
| "eval_accuracy": 0.3312385418738994, |
| "eval_loss": 3.88531494140625, |
| "eval_runtime": 144.5963, |
| "eval_samples_per_second": 124.561, |
| "eval_steps_per_second": 7.787, |
| "step": 7000 |
| }, |
| { |
| "epoch": 0.7601078167115903, |
| "grad_norm": 0.6487019658088684, |
| "learning_rate": 0.0005550178089584457, |
| "loss": 3.9428, |
| "step": 7050 |
| }, |
| { |
| "epoch": 0.7654986522911051, |
| "grad_norm": 0.6093623638153076, |
| "learning_rate": 0.0005546940097139772, |
| "loss": 3.9544, |
| "step": 7100 |
| }, |
| { |
| "epoch": 0.77088948787062, |
| "grad_norm": 0.5531768798828125, |
| "learning_rate": 0.0005543702104695089, |
| "loss": 3.9569, |
| "step": 7150 |
| }, |
| { |
| "epoch": 0.7762803234501348, |
| "grad_norm": 0.6401906609535217, |
| "learning_rate": 0.0005540464112250404, |
| "loss": 3.9564, |
| "step": 7200 |
| }, |
| { |
| "epoch": 0.7816711590296496, |
| "grad_norm": 0.5921440720558167, |
| "learning_rate": 0.000553722611980572, |
| "loss": 3.9108, |
| "step": 7250 |
| }, |
| { |
| "epoch": 0.7870619946091644, |
| "grad_norm": 0.6791409254074097, |
| "learning_rate": 0.0005533988127361035, |
| "loss": 3.9488, |
| "step": 7300 |
| }, |
| { |
| "epoch": 0.7924528301886793, |
| "grad_norm": 0.6472693681716919, |
| "learning_rate": 0.0005530750134916352, |
| "loss": 3.9341, |
| "step": 7350 |
| }, |
| { |
| "epoch": 0.7978436657681941, |
| "grad_norm": 0.6375269889831543, |
| "learning_rate": 0.0005527512142471668, |
| "loss": 3.9339, |
| "step": 7400 |
| }, |
| { |
| "epoch": 0.8032345013477089, |
| "grad_norm": 0.626977264881134, |
| "learning_rate": 0.0005524274150026982, |
| "loss": 3.9247, |
| "step": 7450 |
| }, |
| { |
| "epoch": 0.8086253369272237, |
| "grad_norm": 0.696706235408783, |
| "learning_rate": 0.0005521036157582299, |
| "loss": 3.9116, |
| "step": 7500 |
| }, |
| { |
| "epoch": 0.8140161725067385, |
| "grad_norm": 0.594398558139801, |
| "learning_rate": 0.0005517798165137614, |
| "loss": 3.9356, |
| "step": 7550 |
| }, |
| { |
| "epoch": 0.8194070080862533, |
| "grad_norm": 0.6184767484664917, |
| "learning_rate": 0.000551456017269293, |
| "loss": 3.9384, |
| "step": 7600 |
| }, |
| { |
| "epoch": 0.8247978436657682, |
| "grad_norm": 0.5797574520111084, |
| "learning_rate": 0.0005511322180248245, |
| "loss": 3.9149, |
| "step": 7650 |
| }, |
| { |
| "epoch": 0.8301886792452831, |
| "grad_norm": 0.5616925954818726, |
| "learning_rate": 0.0005508084187803562, |
| "loss": 3.9173, |
| "step": 7700 |
| }, |
| { |
| "epoch": 0.8355795148247979, |
| "grad_norm": 0.6098619103431702, |
| "learning_rate": 0.0005504846195358877, |
| "loss": 3.8999, |
| "step": 7750 |
| }, |
| { |
| "epoch": 0.8409703504043127, |
| "grad_norm": 0.6354513764381409, |
| "learning_rate": 0.0005501608202914193, |
| "loss": 3.9284, |
| "step": 7800 |
| }, |
| { |
| "epoch": 0.8463611859838275, |
| "grad_norm": 0.6673269867897034, |
| "learning_rate": 0.0005498370210469508, |
| "loss": 3.909, |
| "step": 7850 |
| }, |
| { |
| "epoch": 0.8517520215633423, |
| "grad_norm": 0.539669930934906, |
| "learning_rate": 0.0005495132218024824, |
| "loss": 3.9074, |
| "step": 7900 |
| }, |
| { |
| "epoch": 0.8571428571428571, |
| "grad_norm": 0.618360161781311, |
| "learning_rate": 0.000549189422558014, |
| "loss": 3.9046, |
| "step": 7950 |
| }, |
| { |
| "epoch": 0.862533692722372, |
| "grad_norm": 0.6503387689590454, |
| "learning_rate": 0.0005488656233135456, |
| "loss": 3.8917, |
| "step": 8000 |
| }, |
| { |
| "epoch": 0.862533692722372, |
| "eval_accuracy": 0.33594255999893957, |
| "eval_loss": 3.8360989093780518, |
| "eval_runtime": 144.8786, |
| "eval_samples_per_second": 124.318, |
| "eval_steps_per_second": 7.772, |
| "step": 8000 |
| }, |
| { |
| "epoch": 0.8679245283018868, |
| "grad_norm": 0.6546068787574768, |
| "learning_rate": 0.0005485418240690771, |
| "loss": 3.9022, |
| "step": 8050 |
| }, |
| { |
| "epoch": 0.8733153638814016, |
| "grad_norm": 0.6453102231025696, |
| "learning_rate": 0.0005482180248246087, |
| "loss": 3.8843, |
| "step": 8100 |
| }, |
| { |
| "epoch": 0.8787061994609164, |
| "grad_norm": 0.5534443259239197, |
| "learning_rate": 0.0005478942255801403, |
| "loss": 3.8988, |
| "step": 8150 |
| }, |
| { |
| "epoch": 0.8840970350404312, |
| "grad_norm": 0.6230263113975525, |
| "learning_rate": 0.0005475704263356718, |
| "loss": 3.8879, |
| "step": 8200 |
| }, |
| { |
| "epoch": 0.889487870619946, |
| "grad_norm": 0.6733997464179993, |
| "learning_rate": 0.0005472466270912034, |
| "loss": 3.8729, |
| "step": 8250 |
| }, |
| { |
| "epoch": 0.894878706199461, |
| "grad_norm": 0.6633635759353638, |
| "learning_rate": 0.000546922827846735, |
| "loss": 3.8991, |
| "step": 8300 |
| }, |
| { |
| "epoch": 0.9002695417789758, |
| "grad_norm": 0.6503227949142456, |
| "learning_rate": 0.0005465990286022665, |
| "loss": 3.8792, |
| "step": 8350 |
| }, |
| { |
| "epoch": 0.9056603773584906, |
| "grad_norm": 0.7666671276092529, |
| "learning_rate": 0.0005462752293577981, |
| "loss": 3.8944, |
| "step": 8400 |
| }, |
| { |
| "epoch": 0.9110512129380054, |
| "grad_norm": 0.6036889553070068, |
| "learning_rate": 0.0005459514301133296, |
| "loss": 3.9057, |
| "step": 8450 |
| }, |
| { |
| "epoch": 0.9164420485175202, |
| "grad_norm": 0.6154916882514954, |
| "learning_rate": 0.0005456276308688613, |
| "loss": 3.894, |
| "step": 8500 |
| }, |
| { |
| "epoch": 0.921832884097035, |
| "grad_norm": 0.5653623938560486, |
| "learning_rate": 0.0005453038316243929, |
| "loss": 3.8763, |
| "step": 8550 |
| }, |
| { |
| "epoch": 0.9272237196765498, |
| "grad_norm": 0.5695276856422424, |
| "learning_rate": 0.0005449800323799244, |
| "loss": 3.8709, |
| "step": 8600 |
| }, |
| { |
| "epoch": 0.9326145552560647, |
| "grad_norm": 0.5406414270401001, |
| "learning_rate": 0.000544656233135456, |
| "loss": 3.8623, |
| "step": 8650 |
| }, |
| { |
| "epoch": 0.9380053908355795, |
| "grad_norm": 0.5876409411430359, |
| "learning_rate": 0.0005443324338909875, |
| "loss": 3.867, |
| "step": 8700 |
| }, |
| { |
| "epoch": 0.9433962264150944, |
| "grad_norm": 0.5680667757987976, |
| "learning_rate": 0.0005440086346465192, |
| "loss": 3.8691, |
| "step": 8750 |
| }, |
| { |
| "epoch": 0.9487870619946092, |
| "grad_norm": 0.5708035230636597, |
| "learning_rate": 0.0005436848354020506, |
| "loss": 3.8589, |
| "step": 8800 |
| }, |
| { |
| "epoch": 0.954177897574124, |
| "grad_norm": 0.5115900635719299, |
| "learning_rate": 0.0005433610361575823, |
| "loss": 3.8488, |
| "step": 8850 |
| }, |
| { |
| "epoch": 0.9595687331536388, |
| "grad_norm": 0.6006278395652771, |
| "learning_rate": 0.0005430372369131138, |
| "loss": 3.8652, |
| "step": 8900 |
| }, |
| { |
| "epoch": 0.9649595687331537, |
| "grad_norm": 0.6494414210319519, |
| "learning_rate": 0.0005427134376686454, |
| "loss": 3.8716, |
| "step": 8950 |
| }, |
| { |
| "epoch": 0.9703504043126685, |
| "grad_norm": 0.5888278484344482, |
| "learning_rate": 0.0005423896384241769, |
| "loss": 3.8592, |
| "step": 9000 |
| }, |
| { |
| "epoch": 0.9703504043126685, |
| "eval_accuracy": 0.34041004078720755, |
| "eval_loss": 3.7949342727661133, |
| "eval_runtime": 144.616, |
| "eval_samples_per_second": 124.544, |
| "eval_steps_per_second": 7.786, |
| "step": 9000 |
| }, |
| { |
| "epoch": 0.9757412398921833, |
| "grad_norm": 0.5697623491287231, |
| "learning_rate": 0.0005420658391797086, |
| "loss": 3.855, |
| "step": 9050 |
| }, |
| { |
| "epoch": 0.9811320754716981, |
| "grad_norm": 0.6447634696960449, |
| "learning_rate": 0.0005417420399352401, |
| "loss": 3.866, |
| "step": 9100 |
| }, |
| { |
| "epoch": 0.9865229110512129, |
| "grad_norm": 0.583088219165802, |
| "learning_rate": 0.0005414182406907717, |
| "loss": 3.8415, |
| "step": 9150 |
| }, |
| { |
| "epoch": 0.9919137466307277, |
| "grad_norm": 0.5796465277671814, |
| "learning_rate": 0.0005410944414463032, |
| "loss": 3.8376, |
| "step": 9200 |
| }, |
| { |
| "epoch": 0.9973045822102425, |
| "grad_norm": 0.6316399574279785, |
| "learning_rate": 0.0005407706422018348, |
| "loss": 3.8524, |
| "step": 9250 |
| }, |
| { |
| "epoch": 1.0026954177897573, |
| "grad_norm": 0.5727218389511108, |
| "learning_rate": 0.0005404468429573664, |
| "loss": 3.8268, |
| "step": 9300 |
| }, |
| { |
| "epoch": 1.0080862533692723, |
| "grad_norm": 0.5713456869125366, |
| "learning_rate": 0.000540123043712898, |
| "loss": 3.7921, |
| "step": 9350 |
| }, |
| { |
| "epoch": 1.013477088948787, |
| "grad_norm": 0.6198617815971375, |
| "learning_rate": 0.0005397992444684295, |
| "loss": 3.7876, |
| "step": 9400 |
| }, |
| { |
| "epoch": 1.0188679245283019, |
| "grad_norm": 0.5770072937011719, |
| "learning_rate": 0.0005394754452239611, |
| "loss": 3.7856, |
| "step": 9450 |
| }, |
| { |
| "epoch": 1.0242587601078168, |
| "grad_norm": 0.5920796990394592, |
| "learning_rate": 0.0005391516459794927, |
| "loss": 3.7855, |
| "step": 9500 |
| }, |
| { |
| "epoch": 1.0296495956873315, |
| "grad_norm": 0.5636431574821472, |
| "learning_rate": 0.0005388278467350242, |
| "loss": 3.7857, |
| "step": 9550 |
| }, |
| { |
| "epoch": 1.0350404312668464, |
| "grad_norm": 0.669791579246521, |
| "learning_rate": 0.0005385040474905557, |
| "loss": 3.7723, |
| "step": 9600 |
| }, |
| { |
| "epoch": 1.0404312668463611, |
| "grad_norm": 0.6635991930961609, |
| "learning_rate": 0.0005381802482460874, |
| "loss": 3.8142, |
| "step": 9650 |
| }, |
| { |
| "epoch": 1.045822102425876, |
| "grad_norm": 0.601437509059906, |
| "learning_rate": 0.000537856449001619, |
| "loss": 3.798, |
| "step": 9700 |
| }, |
| { |
| "epoch": 1.0512129380053907, |
| "grad_norm": 0.5584002137184143, |
| "learning_rate": 0.0005375326497571505, |
| "loss": 3.7985, |
| "step": 9750 |
| }, |
| { |
| "epoch": 1.0566037735849056, |
| "grad_norm": 0.5526735782623291, |
| "learning_rate": 0.000537208850512682, |
| "loss": 3.8074, |
| "step": 9800 |
| }, |
| { |
| "epoch": 1.0619946091644206, |
| "grad_norm": 0.5977774858474731, |
| "learning_rate": 0.0005368850512682137, |
| "loss": 3.7989, |
| "step": 9850 |
| }, |
| { |
| "epoch": 1.0673854447439353, |
| "grad_norm": 0.5950897932052612, |
| "learning_rate": 0.0005365612520237453, |
| "loss": 3.7924, |
| "step": 9900 |
| }, |
| { |
| "epoch": 1.0727762803234502, |
| "grad_norm": 0.5768362283706665, |
| "learning_rate": 0.0005362374527792768, |
| "loss": 3.7855, |
| "step": 9950 |
| }, |
| { |
| "epoch": 1.0781671159029649, |
| "grad_norm": 0.6516294479370117, |
| "learning_rate": 0.0005359136535348084, |
| "loss": 3.7811, |
| "step": 10000 |
| }, |
| { |
| "epoch": 1.0781671159029649, |
| "eval_accuracy": 0.34365658907510427, |
| "eval_loss": 3.762051582336426, |
| "eval_runtime": 144.5563, |
| "eval_samples_per_second": 124.595, |
| "eval_steps_per_second": 7.789, |
| "step": 10000 |
| }, |
| { |
| "epoch": 1.0835579514824798, |
| "grad_norm": 0.54844731092453, |
| "learning_rate": 0.0005355898542903399, |
| "loss": 3.7802, |
| "step": 10050 |
| }, |
| { |
| "epoch": 1.0889487870619945, |
| "grad_norm": 0.6120631694793701, |
| "learning_rate": 0.0005352660550458716, |
| "loss": 3.789, |
| "step": 10100 |
| }, |
| { |
| "epoch": 1.0943396226415094, |
| "grad_norm": 0.5617305636405945, |
| "learning_rate": 0.000534942255801403, |
| "loss": 3.784, |
| "step": 10150 |
| }, |
| { |
| "epoch": 1.0997304582210243, |
| "grad_norm": 0.6198379993438721, |
| "learning_rate": 0.0005346184565569347, |
| "loss": 3.7835, |
| "step": 10200 |
| }, |
| { |
| "epoch": 1.105121293800539, |
| "grad_norm": 0.5185474753379822, |
| "learning_rate": 0.0005342946573124662, |
| "loss": 3.7928, |
| "step": 10250 |
| }, |
| { |
| "epoch": 1.110512129380054, |
| "grad_norm": 0.5625671744346619, |
| "learning_rate": 0.0005339708580679978, |
| "loss": 3.7792, |
| "step": 10300 |
| }, |
| { |
| "epoch": 1.1159029649595686, |
| "grad_norm": 0.6791179180145264, |
| "learning_rate": 0.0005336470588235293, |
| "loss": 3.7647, |
| "step": 10350 |
| }, |
| { |
| "epoch": 1.1212938005390836, |
| "grad_norm": 0.5952463746070862, |
| "learning_rate": 0.000533323259579061, |
| "loss": 3.7826, |
| "step": 10400 |
| }, |
| { |
| "epoch": 1.1266846361185983, |
| "grad_norm": 0.5367695689201355, |
| "learning_rate": 0.0005329994603345925, |
| "loss": 3.7714, |
| "step": 10450 |
| }, |
| { |
| "epoch": 1.1320754716981132, |
| "grad_norm": 0.6577426195144653, |
| "learning_rate": 0.0005326756610901241, |
| "loss": 3.7588, |
| "step": 10500 |
| }, |
| { |
| "epoch": 1.137466307277628, |
| "grad_norm": 0.5481351017951965, |
| "learning_rate": 0.0005323518618456556, |
| "loss": 3.7808, |
| "step": 10550 |
| }, |
| { |
| "epoch": 1.1428571428571428, |
| "grad_norm": 0.5973223447799683, |
| "learning_rate": 0.0005320280626011872, |
| "loss": 3.7903, |
| "step": 10600 |
| }, |
| { |
| "epoch": 1.1482479784366577, |
| "grad_norm": 0.6764339804649353, |
| "learning_rate": 0.0005317042633567188, |
| "loss": 3.7673, |
| "step": 10650 |
| }, |
| { |
| "epoch": 1.1536388140161726, |
| "grad_norm": 0.6202555894851685, |
| "learning_rate": 0.0005313804641122504, |
| "loss": 3.7633, |
| "step": 10700 |
| }, |
| { |
| "epoch": 1.1590296495956873, |
| "grad_norm": 0.6169992089271545, |
| "learning_rate": 0.0005310566648677819, |
| "loss": 3.7519, |
| "step": 10750 |
| }, |
| { |
| "epoch": 1.1644204851752022, |
| "grad_norm": 0.5893779993057251, |
| "learning_rate": 0.0005307328656233135, |
| "loss": 3.7433, |
| "step": 10800 |
| }, |
| { |
| "epoch": 1.169811320754717, |
| "grad_norm": 0.5708191394805908, |
| "learning_rate": 0.000530409066378845, |
| "loss": 3.7609, |
| "step": 10850 |
| }, |
| { |
| "epoch": 1.1752021563342319, |
| "grad_norm": 0.5436216592788696, |
| "learning_rate": 0.0005300852671343766, |
| "loss": 3.7635, |
| "step": 10900 |
| }, |
| { |
| "epoch": 1.1805929919137466, |
| "grad_norm": 0.5945119261741638, |
| "learning_rate": 0.0005297614678899081, |
| "loss": 3.7755, |
| "step": 10950 |
| }, |
| { |
| "epoch": 1.1859838274932615, |
| "grad_norm": 0.5752778053283691, |
| "learning_rate": 0.0005294376686454398, |
| "loss": 3.7908, |
| "step": 11000 |
| }, |
| { |
| "epoch": 1.1859838274932615, |
| "eval_accuracy": 0.3459605735612103, |
| "eval_loss": 3.7342617511749268, |
| "eval_runtime": 145.0651, |
| "eval_samples_per_second": 124.158, |
| "eval_steps_per_second": 7.762, |
| "step": 11000 |
| }, |
| { |
| "epoch": 1.1913746630727764, |
| "grad_norm": 0.5780414342880249, |
| "learning_rate": 0.0005291138694009714, |
| "loss": 3.7773, |
| "step": 11050 |
| }, |
| { |
| "epoch": 1.196765498652291, |
| "grad_norm": 0.6393083930015564, |
| "learning_rate": 0.0005287900701565029, |
| "loss": 3.7719, |
| "step": 11100 |
| }, |
| { |
| "epoch": 1.202156334231806, |
| "grad_norm": 0.5772005319595337, |
| "learning_rate": 0.0005284662709120345, |
| "loss": 3.7688, |
| "step": 11150 |
| }, |
| { |
| "epoch": 1.2075471698113207, |
| "grad_norm": 0.5036531686782837, |
| "learning_rate": 0.0005281424716675661, |
| "loss": 3.7578, |
| "step": 11200 |
| }, |
| { |
| "epoch": 1.2129380053908356, |
| "grad_norm": 0.56954026222229, |
| "learning_rate": 0.0005278186724230977, |
| "loss": 3.7505, |
| "step": 11250 |
| }, |
| { |
| "epoch": 1.2183288409703503, |
| "grad_norm": 0.6294519901275635, |
| "learning_rate": 0.0005274948731786292, |
| "loss": 3.7672, |
| "step": 11300 |
| }, |
| { |
| "epoch": 1.2237196765498652, |
| "grad_norm": 0.6276326775550842, |
| "learning_rate": 0.0005271710739341608, |
| "loss": 3.7523, |
| "step": 11350 |
| }, |
| { |
| "epoch": 1.2291105121293802, |
| "grad_norm": 0.6219097971916199, |
| "learning_rate": 0.0005268472746896923, |
| "loss": 3.7687, |
| "step": 11400 |
| }, |
| { |
| "epoch": 1.2345013477088949, |
| "grad_norm": 0.5939977765083313, |
| "learning_rate": 0.000526523475445224, |
| "loss": 3.7595, |
| "step": 11450 |
| }, |
| { |
| "epoch": 1.2398921832884098, |
| "grad_norm": 0.6375150680541992, |
| "learning_rate": 0.0005261996762007554, |
| "loss": 3.7498, |
| "step": 11500 |
| }, |
| { |
| "epoch": 1.2452830188679245, |
| "grad_norm": 0.534351110458374, |
| "learning_rate": 0.0005258758769562871, |
| "loss": 3.7602, |
| "step": 11550 |
| }, |
| { |
| "epoch": 1.2506738544474394, |
| "grad_norm": 0.5753054618835449, |
| "learning_rate": 0.0005255520777118186, |
| "loss": 3.7345, |
| "step": 11600 |
| }, |
| { |
| "epoch": 1.256064690026954, |
| "grad_norm": 0.5968658328056335, |
| "learning_rate": 0.0005252282784673502, |
| "loss": 3.7679, |
| "step": 11650 |
| }, |
| { |
| "epoch": 1.261455525606469, |
| "grad_norm": 0.6453511118888855, |
| "learning_rate": 0.0005249044792228817, |
| "loss": 3.7641, |
| "step": 11700 |
| }, |
| { |
| "epoch": 1.266846361185984, |
| "grad_norm": 0.5844177007675171, |
| "learning_rate": 0.0005245806799784133, |
| "loss": 3.7551, |
| "step": 11750 |
| }, |
| { |
| "epoch": 1.2722371967654986, |
| "grad_norm": 0.5597920417785645, |
| "learning_rate": 0.0005242568807339449, |
| "loss": 3.7458, |
| "step": 11800 |
| }, |
| { |
| "epoch": 1.2776280323450135, |
| "grad_norm": 0.6149951815605164, |
| "learning_rate": 0.0005239330814894765, |
| "loss": 3.7582, |
| "step": 11850 |
| }, |
| { |
| "epoch": 1.2830188679245282, |
| "grad_norm": 0.5561873316764832, |
| "learning_rate": 0.000523609282245008, |
| "loss": 3.7456, |
| "step": 11900 |
| }, |
| { |
| "epoch": 1.2884097035040432, |
| "grad_norm": 0.6077041029930115, |
| "learning_rate": 0.0005232854830005396, |
| "loss": 3.7601, |
| "step": 11950 |
| }, |
| { |
| "epoch": 1.2938005390835579, |
| "grad_norm": 0.6389844417572021, |
| "learning_rate": 0.0005229616837560712, |
| "loss": 3.7514, |
| "step": 12000 |
| }, |
| { |
| "epoch": 1.2938005390835579, |
| "eval_accuracy": 0.3482391332715678, |
| "eval_loss": 3.7092418670654297, |
| "eval_runtime": 144.795, |
| "eval_samples_per_second": 124.39, |
| "eval_steps_per_second": 7.777, |
| "step": 12000 |
| }, |
| { |
| "epoch": 1.2991913746630728, |
| "grad_norm": 0.6131287217140198, |
| "learning_rate": 0.0005226378845116028, |
| "loss": 3.7472, |
| "step": 12050 |
| }, |
| { |
| "epoch": 1.3045822102425877, |
| "grad_norm": 0.5687003135681152, |
| "learning_rate": 0.0005223140852671344, |
| "loss": 3.7541, |
| "step": 12100 |
| }, |
| { |
| "epoch": 1.3099730458221024, |
| "grad_norm": 0.603330135345459, |
| "learning_rate": 0.0005219902860226659, |
| "loss": 3.7719, |
| "step": 12150 |
| }, |
| { |
| "epoch": 1.3153638814016173, |
| "grad_norm": 0.5906956791877747, |
| "learning_rate": 0.0005216664867781975, |
| "loss": 3.7419, |
| "step": 12200 |
| }, |
| { |
| "epoch": 1.320754716981132, |
| "grad_norm": 0.5257678627967834, |
| "learning_rate": 0.000521342687533729, |
| "loss": 3.7494, |
| "step": 12250 |
| }, |
| { |
| "epoch": 1.326145552560647, |
| "grad_norm": 0.55323725938797, |
| "learning_rate": 0.0005210188882892606, |
| "loss": 3.7598, |
| "step": 12300 |
| }, |
| { |
| "epoch": 1.3315363881401616, |
| "grad_norm": 0.6195726990699768, |
| "learning_rate": 0.0005206950890447922, |
| "loss": 3.7455, |
| "step": 12350 |
| }, |
| { |
| "epoch": 1.3369272237196765, |
| "grad_norm": 0.6453518271446228, |
| "learning_rate": 0.0005203712898003238, |
| "loss": 3.7346, |
| "step": 12400 |
| }, |
| { |
| "epoch": 1.3423180592991915, |
| "grad_norm": 0.591871976852417, |
| "learning_rate": 0.0005200474905558553, |
| "loss": 3.7498, |
| "step": 12450 |
| }, |
| { |
| "epoch": 1.3477088948787062, |
| "grad_norm": 0.8153529167175293, |
| "learning_rate": 0.0005197236913113869, |
| "loss": 3.73, |
| "step": 12500 |
| }, |
| { |
| "epoch": 1.353099730458221, |
| "grad_norm": 0.6054574847221375, |
| "learning_rate": 0.0005193998920669184, |
| "loss": 3.7438, |
| "step": 12550 |
| }, |
| { |
| "epoch": 1.3584905660377358, |
| "grad_norm": 0.5890757441520691, |
| "learning_rate": 0.0005190760928224501, |
| "loss": 3.7383, |
| "step": 12600 |
| }, |
| { |
| "epoch": 1.3638814016172507, |
| "grad_norm": 0.5858767032623291, |
| "learning_rate": 0.0005187522935779816, |
| "loss": 3.7272, |
| "step": 12650 |
| }, |
| { |
| "epoch": 1.3692722371967654, |
| "grad_norm": 0.5663976073265076, |
| "learning_rate": 0.0005184284943335132, |
| "loss": 3.7232, |
| "step": 12700 |
| }, |
| { |
| "epoch": 1.3746630727762803, |
| "grad_norm": 0.6173025965690613, |
| "learning_rate": 0.0005181046950890447, |
| "loss": 3.7428, |
| "step": 12750 |
| }, |
| { |
| "epoch": 1.3800539083557952, |
| "grad_norm": 0.5534438490867615, |
| "learning_rate": 0.0005177808958445764, |
| "loss": 3.7214, |
| "step": 12800 |
| }, |
| { |
| "epoch": 1.38544474393531, |
| "grad_norm": 0.6565149426460266, |
| "learning_rate": 0.0005174570966001078, |
| "loss": 3.7343, |
| "step": 12850 |
| }, |
| { |
| "epoch": 1.3908355795148248, |
| "grad_norm": 0.518814742565155, |
| "learning_rate": 0.0005171332973556395, |
| "loss": 3.7343, |
| "step": 12900 |
| }, |
| { |
| "epoch": 1.3962264150943398, |
| "grad_norm": 0.6175557374954224, |
| "learning_rate": 0.000516809498111171, |
| "loss": 3.726, |
| "step": 12950 |
| }, |
| { |
| "epoch": 1.4016172506738545, |
| "grad_norm": 0.525554895401001, |
| "learning_rate": 0.0005164856988667026, |
| "loss": 3.7207, |
| "step": 13000 |
| }, |
| { |
| "epoch": 1.4016172506738545, |
| "eval_accuracy": 0.3513273743019188, |
| "eval_loss": 3.684098482131958, |
| "eval_runtime": 145.0832, |
| "eval_samples_per_second": 124.143, |
| "eval_steps_per_second": 7.761, |
| "step": 13000 |
| }, |
| { |
| "epoch": 1.4070080862533692, |
| "grad_norm": 0.5694339275360107, |
| "learning_rate": 0.0005161618996222341, |
| "loss": 3.7017, |
| "step": 13050 |
| }, |
| { |
| "epoch": 1.412398921832884, |
| "grad_norm": 0.5720424056053162, |
| "learning_rate": 0.0005158381003777657, |
| "loss": 3.7224, |
| "step": 13100 |
| }, |
| { |
| "epoch": 1.417789757412399, |
| "grad_norm": 0.5959078073501587, |
| "learning_rate": 0.0005155143011332973, |
| "loss": 3.7268, |
| "step": 13150 |
| }, |
| { |
| "epoch": 1.4231805929919137, |
| "grad_norm": 0.5883017182350159, |
| "learning_rate": 0.0005151905018888289, |
| "loss": 3.7077, |
| "step": 13200 |
| }, |
| { |
| "epoch": 1.4285714285714286, |
| "grad_norm": 0.5214637517929077, |
| "learning_rate": 0.0005148667026443604, |
| "loss": 3.7251, |
| "step": 13250 |
| }, |
| { |
| "epoch": 1.4339622641509435, |
| "grad_norm": 0.5803607106208801, |
| "learning_rate": 0.000514542903399892, |
| "loss": 3.7207, |
| "step": 13300 |
| }, |
| { |
| "epoch": 1.4393530997304582, |
| "grad_norm": 0.6287586688995361, |
| "learning_rate": 0.0005142191041554237, |
| "loss": 3.7256, |
| "step": 13350 |
| }, |
| { |
| "epoch": 1.444743935309973, |
| "grad_norm": 0.614562451839447, |
| "learning_rate": 0.0005138953049109552, |
| "loss": 3.7166, |
| "step": 13400 |
| }, |
| { |
| "epoch": 1.4501347708894878, |
| "grad_norm": 0.5314472317695618, |
| "learning_rate": 0.0005135715056664868, |
| "loss": 3.7338, |
| "step": 13450 |
| }, |
| { |
| "epoch": 1.4555256064690028, |
| "grad_norm": 0.6482129693031311, |
| "learning_rate": 0.0005132477064220183, |
| "loss": 3.7264, |
| "step": 13500 |
| }, |
| { |
| "epoch": 1.4609164420485174, |
| "grad_norm": 0.5918267369270325, |
| "learning_rate": 0.0005129239071775499, |
| "loss": 3.7023, |
| "step": 13550 |
| }, |
| { |
| "epoch": 1.4663072776280324, |
| "grad_norm": 0.5850944519042969, |
| "learning_rate": 0.0005126001079330814, |
| "loss": 3.7298, |
| "step": 13600 |
| }, |
| { |
| "epoch": 1.4716981132075473, |
| "grad_norm": 0.5356786847114563, |
| "learning_rate": 0.000512276308688613, |
| "loss": 3.7098, |
| "step": 13650 |
| }, |
| { |
| "epoch": 1.477088948787062, |
| "grad_norm": 0.5910783410072327, |
| "learning_rate": 0.0005119525094441446, |
| "loss": 3.7309, |
| "step": 13700 |
| }, |
| { |
| "epoch": 1.482479784366577, |
| "grad_norm": 0.5758869051933289, |
| "learning_rate": 0.0005116287101996762, |
| "loss": 3.703, |
| "step": 13750 |
| }, |
| { |
| "epoch": 1.4878706199460916, |
| "grad_norm": 0.5499842762947083, |
| "learning_rate": 0.0005113049109552077, |
| "loss": 3.7275, |
| "step": 13800 |
| }, |
| { |
| "epoch": 1.4932614555256065, |
| "grad_norm": 0.6198793053627014, |
| "learning_rate": 0.0005109811117107393, |
| "loss": 3.7118, |
| "step": 13850 |
| }, |
| { |
| "epoch": 1.4986522911051212, |
| "grad_norm": 0.5693920850753784, |
| "learning_rate": 0.0005106573124662708, |
| "loss": 3.7209, |
| "step": 13900 |
| }, |
| { |
| "epoch": 1.5040431266846361, |
| "grad_norm": 0.6063724160194397, |
| "learning_rate": 0.0005103335132218025, |
| "loss": 3.7108, |
| "step": 13950 |
| }, |
| { |
| "epoch": 1.509433962264151, |
| "grad_norm": 0.5341413021087646, |
| "learning_rate": 0.000510009713977334, |
| "loss": 3.7092, |
| "step": 14000 |
| }, |
| { |
| "epoch": 1.509433962264151, |
| "eval_accuracy": 0.35327356482845174, |
| "eval_loss": 3.66573429107666, |
| "eval_runtime": 144.4021, |
| "eval_samples_per_second": 124.728, |
| "eval_steps_per_second": 7.798, |
| "step": 14000 |
| }, |
| { |
| "epoch": 1.5148247978436657, |
| "grad_norm": 0.5940130352973938, |
| "learning_rate": 0.0005096859147328656, |
| "loss": 3.7083, |
| "step": 14050 |
| }, |
| { |
| "epoch": 1.5202156334231804, |
| "grad_norm": 0.5854769945144653, |
| "learning_rate": 0.0005093621154883971, |
| "loss": 3.7072, |
| "step": 14100 |
| }, |
| { |
| "epoch": 1.5256064690026954, |
| "grad_norm": 0.5369813442230225, |
| "learning_rate": 0.0005090383162439288, |
| "loss": 3.6994, |
| "step": 14150 |
| }, |
| { |
| "epoch": 1.5309973045822103, |
| "grad_norm": 0.6851678490638733, |
| "learning_rate": 0.0005087145169994602, |
| "loss": 3.7003, |
| "step": 14200 |
| }, |
| { |
| "epoch": 1.536388140161725, |
| "grad_norm": 0.5451446771621704, |
| "learning_rate": 0.0005083907177549918, |
| "loss": 3.7145, |
| "step": 14250 |
| }, |
| { |
| "epoch": 1.54177897574124, |
| "grad_norm": 0.550528347492218, |
| "learning_rate": 0.0005080669185105234, |
| "loss": 3.7145, |
| "step": 14300 |
| }, |
| { |
| "epoch": 1.5471698113207548, |
| "grad_norm": 0.5607097744941711, |
| "learning_rate": 0.000507743119266055, |
| "loss": 3.7074, |
| "step": 14350 |
| }, |
| { |
| "epoch": 1.5525606469002695, |
| "grad_norm": 0.5699529051780701, |
| "learning_rate": 0.0005074193200215865, |
| "loss": 3.7239, |
| "step": 14400 |
| }, |
| { |
| "epoch": 1.5579514824797842, |
| "grad_norm": 0.6193976402282715, |
| "learning_rate": 0.0005070955207771181, |
| "loss": 3.7146, |
| "step": 14450 |
| }, |
| { |
| "epoch": 1.5633423180592994, |
| "grad_norm": 0.5902137756347656, |
| "learning_rate": 0.0005067717215326498, |
| "loss": 3.6944, |
| "step": 14500 |
| }, |
| { |
| "epoch": 1.568733153638814, |
| "grad_norm": 0.5481423735618591, |
| "learning_rate": 0.0005064479222881813, |
| "loss": 3.7103, |
| "step": 14550 |
| }, |
| { |
| "epoch": 1.5741239892183287, |
| "grad_norm": 0.6323958039283752, |
| "learning_rate": 0.0005061241230437129, |
| "loss": 3.6955, |
| "step": 14600 |
| }, |
| { |
| "epoch": 1.5795148247978437, |
| "grad_norm": 0.6062991619110107, |
| "learning_rate": 0.0005058003237992444, |
| "loss": 3.6937, |
| "step": 14650 |
| }, |
| { |
| "epoch": 1.5849056603773586, |
| "grad_norm": 0.5340829491615295, |
| "learning_rate": 0.0005054830005396654, |
| "loss": 3.7098, |
| "step": 14700 |
| }, |
| { |
| "epoch": 1.5902964959568733, |
| "grad_norm": 0.590190052986145, |
| "learning_rate": 0.0005051592012951969, |
| "loss": 3.7098, |
| "step": 14750 |
| }, |
| { |
| "epoch": 1.595687331536388, |
| "grad_norm": 0.5628999471664429, |
| "learning_rate": 0.0005048354020507286, |
| "loss": 3.6799, |
| "step": 14800 |
| }, |
| { |
| "epoch": 1.6010781671159031, |
| "grad_norm": 0.5697661638259888, |
| "learning_rate": 0.00050451160280626, |
| "loss": 3.7014, |
| "step": 14850 |
| }, |
| { |
| "epoch": 1.6064690026954178, |
| "grad_norm": 0.5965352058410645, |
| "learning_rate": 0.0005041878035617917, |
| "loss": 3.7081, |
| "step": 14900 |
| }, |
| { |
| "epoch": 1.6118598382749325, |
| "grad_norm": 0.5691752433776855, |
| "learning_rate": 0.0005038640043173232, |
| "loss": 3.697, |
| "step": 14950 |
| }, |
| { |
| "epoch": 1.6172506738544474, |
| "grad_norm": 0.5844449400901794, |
| "learning_rate": 0.0005035402050728548, |
| "loss": 3.6968, |
| "step": 15000 |
| }, |
| { |
| "epoch": 1.6172506738544474, |
| "eval_accuracy": 0.35501950808274263, |
| "eval_loss": 3.6473617553710938, |
| "eval_runtime": 144.9331, |
| "eval_samples_per_second": 124.271, |
| "eval_steps_per_second": 7.769, |
| "step": 15000 |
| }, |
| { |
| "epoch": 1.6226415094339623, |
| "grad_norm": 0.5243232846260071, |
| "learning_rate": 0.0005032164058283863, |
| "loss": 3.6876, |
| "step": 15050 |
| }, |
| { |
| "epoch": 1.628032345013477, |
| "grad_norm": 0.521431565284729, |
| "learning_rate": 0.0005028926065839179, |
| "loss": 3.6816, |
| "step": 15100 |
| }, |
| { |
| "epoch": 1.633423180592992, |
| "grad_norm": 0.54181307554245, |
| "learning_rate": 0.0005025688073394495, |
| "loss": 3.681, |
| "step": 15150 |
| }, |
| { |
| "epoch": 1.6388140161725069, |
| "grad_norm": 0.5746421217918396, |
| "learning_rate": 0.0005022450080949811, |
| "loss": 3.6847, |
| "step": 15200 |
| }, |
| { |
| "epoch": 1.6442048517520216, |
| "grad_norm": 0.5637415051460266, |
| "learning_rate": 0.0005019212088505126, |
| "loss": 3.6782, |
| "step": 15250 |
| }, |
| { |
| "epoch": 1.6495956873315363, |
| "grad_norm": 0.5300856828689575, |
| "learning_rate": 0.0005015974096060442, |
| "loss": 3.6848, |
| "step": 15300 |
| }, |
| { |
| "epoch": 1.6549865229110512, |
| "grad_norm": 0.5390794277191162, |
| "learning_rate": 0.0005012736103615758, |
| "loss": 3.7087, |
| "step": 15350 |
| }, |
| { |
| "epoch": 1.6603773584905661, |
| "grad_norm": 0.6010688543319702, |
| "learning_rate": 0.0005009498111171074, |
| "loss": 3.6966, |
| "step": 15400 |
| }, |
| { |
| "epoch": 1.6657681940700808, |
| "grad_norm": 0.548942506313324, |
| "learning_rate": 0.0005006260118726389, |
| "loss": 3.6648, |
| "step": 15450 |
| }, |
| { |
| "epoch": 1.6711590296495957, |
| "grad_norm": 0.5794510245323181, |
| "learning_rate": 0.0005003022126281705, |
| "loss": 3.6827, |
| "step": 15500 |
| }, |
| { |
| "epoch": 1.6765498652291106, |
| "grad_norm": 0.5393055081367493, |
| "learning_rate": 0.000499978413383702, |
| "loss": 3.68, |
| "step": 15550 |
| }, |
| { |
| "epoch": 1.6819407008086253, |
| "grad_norm": 0.5647615194320679, |
| "learning_rate": 0.0004996546141392336, |
| "loss": 3.6677, |
| "step": 15600 |
| }, |
| { |
| "epoch": 1.68733153638814, |
| "grad_norm": 0.5918028354644775, |
| "learning_rate": 0.0004993308148947651, |
| "loss": 3.6859, |
| "step": 15650 |
| }, |
| { |
| "epoch": 1.692722371967655, |
| "grad_norm": 0.5556984543800354, |
| "learning_rate": 0.0004990070156502968, |
| "loss": 3.6733, |
| "step": 15700 |
| }, |
| { |
| "epoch": 1.6981132075471699, |
| "grad_norm": 0.5747886896133423, |
| "learning_rate": 0.0004986832164058284, |
| "loss": 3.6979, |
| "step": 15750 |
| }, |
| { |
| "epoch": 1.7035040431266846, |
| "grad_norm": 0.5632966756820679, |
| "learning_rate": 0.0004983594171613599, |
| "loss": 3.6957, |
| "step": 15800 |
| }, |
| { |
| "epoch": 1.7088948787061995, |
| "grad_norm": 0.5908682942390442, |
| "learning_rate": 0.0004980356179168915, |
| "loss": 3.6886, |
| "step": 15850 |
| }, |
| { |
| "epoch": 1.7142857142857144, |
| "grad_norm": 0.6491772532463074, |
| "learning_rate": 0.000497711818672423, |
| "loss": 3.6902, |
| "step": 15900 |
| }, |
| { |
| "epoch": 1.719676549865229, |
| "grad_norm": 0.594153106212616, |
| "learning_rate": 0.0004973880194279547, |
| "loss": 3.6752, |
| "step": 15950 |
| }, |
| { |
| "epoch": 1.7250673854447438, |
| "grad_norm": 0.6420136094093323, |
| "learning_rate": 0.0004970642201834862, |
| "loss": 3.6724, |
| "step": 16000 |
| }, |
| { |
| "epoch": 1.7250673854447438, |
| "eval_accuracy": 0.3570638121669721, |
| "eval_loss": 3.6263442039489746, |
| "eval_runtime": 144.5518, |
| "eval_samples_per_second": 124.599, |
| "eval_steps_per_second": 7.79, |
| "step": 16000 |
| }, |
| { |
| "epoch": 1.7304582210242587, |
| "grad_norm": 0.5980614423751831, |
| "learning_rate": 0.0004967404209390178, |
| "loss": 3.6727, |
| "step": 16050 |
| }, |
| { |
| "epoch": 1.7358490566037736, |
| "grad_norm": 0.5954625010490417, |
| "learning_rate": 0.0004964166216945493, |
| "loss": 3.6587, |
| "step": 16100 |
| }, |
| { |
| "epoch": 1.7412398921832883, |
| "grad_norm": 0.563701868057251, |
| "learning_rate": 0.000496092822450081, |
| "loss": 3.683, |
| "step": 16150 |
| }, |
| { |
| "epoch": 1.7466307277628033, |
| "grad_norm": 0.5353266596794128, |
| "learning_rate": 0.0004957690232056125, |
| "loss": 3.6928, |
| "step": 16200 |
| }, |
| { |
| "epoch": 1.7520215633423182, |
| "grad_norm": 0.6115831136703491, |
| "learning_rate": 0.0004954452239611441, |
| "loss": 3.6697, |
| "step": 16250 |
| }, |
| { |
| "epoch": 1.7574123989218329, |
| "grad_norm": 0.5373592376708984, |
| "learning_rate": 0.0004951214247166756, |
| "loss": 3.6907, |
| "step": 16300 |
| }, |
| { |
| "epoch": 1.7628032345013476, |
| "grad_norm": 0.7120974063873291, |
| "learning_rate": 0.0004947976254722072, |
| "loss": 3.6741, |
| "step": 16350 |
| }, |
| { |
| "epoch": 1.7681940700808625, |
| "grad_norm": 0.5365995764732361, |
| "learning_rate": 0.0004944738262277387, |
| "loss": 3.6656, |
| "step": 16400 |
| }, |
| { |
| "epoch": 1.7735849056603774, |
| "grad_norm": 0.6026771664619446, |
| "learning_rate": 0.0004941500269832703, |
| "loss": 3.6708, |
| "step": 16450 |
| }, |
| { |
| "epoch": 1.778975741239892, |
| "grad_norm": 0.5560598969459534, |
| "learning_rate": 0.0004938262277388019, |
| "loss": 3.6594, |
| "step": 16500 |
| }, |
| { |
| "epoch": 1.784366576819407, |
| "grad_norm": 0.6191685199737549, |
| "learning_rate": 0.0004935024284943335, |
| "loss": 3.685, |
| "step": 16550 |
| }, |
| { |
| "epoch": 1.789757412398922, |
| "grad_norm": 0.6867356896400452, |
| "learning_rate": 0.000493178629249865, |
| "loss": 3.6698, |
| "step": 16600 |
| }, |
| { |
| "epoch": 1.7951482479784366, |
| "grad_norm": 0.5517817139625549, |
| "learning_rate": 0.0004928548300053966, |
| "loss": 3.6725, |
| "step": 16650 |
| }, |
| { |
| "epoch": 1.8005390835579513, |
| "grad_norm": 0.5496578216552734, |
| "learning_rate": 0.0004925310307609282, |
| "loss": 3.6745, |
| "step": 16700 |
| }, |
| { |
| "epoch": 1.8059299191374663, |
| "grad_norm": 0.540912926197052, |
| "learning_rate": 0.0004922072315164598, |
| "loss": 3.6624, |
| "step": 16750 |
| }, |
| { |
| "epoch": 1.8113207547169812, |
| "grad_norm": 0.5637075304985046, |
| "learning_rate": 0.0004918834322719913, |
| "loss": 3.6658, |
| "step": 16800 |
| }, |
| { |
| "epoch": 1.8167115902964959, |
| "grad_norm": 0.5674017071723938, |
| "learning_rate": 0.0004915596330275229, |
| "loss": 3.6701, |
| "step": 16850 |
| }, |
| { |
| "epoch": 1.8221024258760108, |
| "grad_norm": 0.5619826912879944, |
| "learning_rate": 0.0004912487857528331, |
| "loss": 3.6714, |
| "step": 16900 |
| }, |
| { |
| "epoch": 1.8274932614555257, |
| "grad_norm": 0.5217414498329163, |
| "learning_rate": 0.0004909249865083648, |
| "loss": 3.6863, |
| "step": 16950 |
| }, |
| { |
| "epoch": 1.8328840970350404, |
| "grad_norm": 0.565362811088562, |
| "learning_rate": 0.0004906011872638964, |
| "loss": 3.6737, |
| "step": 17000 |
| }, |
| { |
| "epoch": 1.8328840970350404, |
| "eval_accuracy": 0.3582435652228617, |
| "eval_loss": 3.6162045001983643, |
| "eval_runtime": 144.7573, |
| "eval_samples_per_second": 124.422, |
| "eval_steps_per_second": 7.779, |
| "step": 17000 |
| }, |
| { |
| "epoch": 1.838274932614555, |
| "grad_norm": 0.5198625326156616, |
| "learning_rate": 0.0004902773880194279, |
| "loss": 3.6515, |
| "step": 17050 |
| }, |
| { |
| "epoch": 1.8436657681940702, |
| "grad_norm": 0.5708647966384888, |
| "learning_rate": 0.0004899535887749595, |
| "loss": 3.6661, |
| "step": 17100 |
| }, |
| { |
| "epoch": 1.849056603773585, |
| "grad_norm": 0.583521842956543, |
| "learning_rate": 0.000489629789530491, |
| "loss": 3.6622, |
| "step": 17150 |
| }, |
| { |
| "epoch": 1.8544474393530996, |
| "grad_norm": 0.5446178913116455, |
| "learning_rate": 0.0004893059902860227, |
| "loss": 3.647, |
| "step": 17200 |
| }, |
| { |
| "epoch": 1.8598382749326146, |
| "grad_norm": 0.5899081230163574, |
| "learning_rate": 0.0004889821910415542, |
| "loss": 3.6697, |
| "step": 17250 |
| }, |
| { |
| "epoch": 1.8652291105121295, |
| "grad_norm": 0.6142265796661377, |
| "learning_rate": 0.0004886583917970858, |
| "loss": 3.6597, |
| "step": 17300 |
| }, |
| { |
| "epoch": 1.8706199460916442, |
| "grad_norm": 0.9669179320335388, |
| "learning_rate": 0.0004883345925526173, |
| "loss": 3.6618, |
| "step": 17350 |
| }, |
| { |
| "epoch": 1.8760107816711589, |
| "grad_norm": 0.564871072769165, |
| "learning_rate": 0.00048801079330814887, |
| "loss": 3.6571, |
| "step": 17400 |
| }, |
| { |
| "epoch": 1.881401617250674, |
| "grad_norm": 0.5841614007949829, |
| "learning_rate": 0.00048768699406368047, |
| "loss": 3.6743, |
| "step": 17450 |
| }, |
| { |
| "epoch": 1.8867924528301887, |
| "grad_norm": 0.5918266773223877, |
| "learning_rate": 0.000487363194819212, |
| "loss": 3.6675, |
| "step": 17500 |
| }, |
| { |
| "epoch": 1.8921832884097034, |
| "grad_norm": 0.5641137361526489, |
| "learning_rate": 0.0004870393955747436, |
| "loss": 3.6488, |
| "step": 17550 |
| }, |
| { |
| "epoch": 1.8975741239892183, |
| "grad_norm": 0.5747765898704529, |
| "learning_rate": 0.0004867155963302752, |
| "loss": 3.6558, |
| "step": 17600 |
| }, |
| { |
| "epoch": 1.9029649595687332, |
| "grad_norm": 0.5801583528518677, |
| "learning_rate": 0.0004863917970858068, |
| "loss": 3.6348, |
| "step": 17650 |
| }, |
| { |
| "epoch": 1.908355795148248, |
| "grad_norm": 0.5788767337799072, |
| "learning_rate": 0.00048606799784133833, |
| "loss": 3.6446, |
| "step": 17700 |
| }, |
| { |
| "epoch": 1.9137466307277629, |
| "grad_norm": 0.5854235291481018, |
| "learning_rate": 0.00048574419859686994, |
| "loss": 3.648, |
| "step": 17750 |
| }, |
| { |
| "epoch": 1.9191374663072778, |
| "grad_norm": 0.5577847957611084, |
| "learning_rate": 0.0004854203993524015, |
| "loss": 3.6665, |
| "step": 17800 |
| }, |
| { |
| "epoch": 1.9245283018867925, |
| "grad_norm": 0.536088228225708, |
| "learning_rate": 0.000485096600107933, |
| "loss": 3.6671, |
| "step": 17850 |
| }, |
| { |
| "epoch": 1.9299191374663072, |
| "grad_norm": 0.5481696724891663, |
| "learning_rate": 0.00048477280086346464, |
| "loss": 3.6721, |
| "step": 17900 |
| }, |
| { |
| "epoch": 1.935309973045822, |
| "grad_norm": 0.547725260257721, |
| "learning_rate": 0.00048444900161899614, |
| "loss": 3.6622, |
| "step": 17950 |
| }, |
| { |
| "epoch": 1.940700808625337, |
| "grad_norm": 0.5937644243240356, |
| "learning_rate": 0.00048412520237452774, |
| "loss": 3.6477, |
| "step": 18000 |
| }, |
| { |
| "epoch": 1.940700808625337, |
| "eval_accuracy": 0.3597537317105952, |
| "eval_loss": 3.5971779823303223, |
| "eval_runtime": 144.7293, |
| "eval_samples_per_second": 124.446, |
| "eval_steps_per_second": 7.78, |
| "step": 18000 |
| }, |
| { |
| "epoch": 1.9460916442048517, |
| "grad_norm": 0.5931717753410339, |
| "learning_rate": 0.0004838014031300593, |
| "loss": 3.6358, |
| "step": 18050 |
| }, |
| { |
| "epoch": 1.9514824797843666, |
| "grad_norm": 0.6099575757980347, |
| "learning_rate": 0.0004834776038855909, |
| "loss": 3.644, |
| "step": 18100 |
| }, |
| { |
| "epoch": 1.9568733153638815, |
| "grad_norm": 0.5704085230827332, |
| "learning_rate": 0.00048315380464112245, |
| "loss": 3.6517, |
| "step": 18150 |
| }, |
| { |
| "epoch": 1.9622641509433962, |
| "grad_norm": 0.5262129902839661, |
| "learning_rate": 0.00048283000539665405, |
| "loss": 3.6523, |
| "step": 18200 |
| }, |
| { |
| "epoch": 1.967654986522911, |
| "grad_norm": 0.512693464756012, |
| "learning_rate": 0.0004825062061521856, |
| "loss": 3.6432, |
| "step": 18250 |
| }, |
| { |
| "epoch": 1.9730458221024259, |
| "grad_norm": 0.5602070093154907, |
| "learning_rate": 0.00048218240690771716, |
| "loss": 3.6688, |
| "step": 18300 |
| }, |
| { |
| "epoch": 1.9784366576819408, |
| "grad_norm": 0.5258756875991821, |
| "learning_rate": 0.00048185860766324876, |
| "loss": 3.6471, |
| "step": 18350 |
| }, |
| { |
| "epoch": 1.9838274932614555, |
| "grad_norm": 0.5815879106521606, |
| "learning_rate": 0.0004815348084187803, |
| "loss": 3.6496, |
| "step": 18400 |
| }, |
| { |
| "epoch": 1.9892183288409704, |
| "grad_norm": 0.5756340622901917, |
| "learning_rate": 0.0004812110091743119, |
| "loss": 3.6589, |
| "step": 18450 |
| }, |
| { |
| "epoch": 1.9946091644204853, |
| "grad_norm": 0.6157427430152893, |
| "learning_rate": 0.00048088720992984347, |
| "loss": 3.6378, |
| "step": 18500 |
| }, |
| { |
| "epoch": 2.0, |
| "grad_norm": 1.1887223720550537, |
| "learning_rate": 0.00048056341068537507, |
| "loss": 3.6418, |
| "step": 18550 |
| }, |
| { |
| "epoch": 2.0053908355795147, |
| "grad_norm": 0.5796721577644348, |
| "learning_rate": 0.00048023961144090657, |
| "loss": 3.5398, |
| "step": 18600 |
| }, |
| { |
| "epoch": 2.01078167115903, |
| "grad_norm": 0.5702629089355469, |
| "learning_rate": 0.00047991581219643817, |
| "loss": 3.5522, |
| "step": 18650 |
| }, |
| { |
| "epoch": 2.0161725067385445, |
| "grad_norm": 0.5687718391418457, |
| "learning_rate": 0.0004795920129519697, |
| "loss": 3.5469, |
| "step": 18700 |
| }, |
| { |
| "epoch": 2.0215633423180592, |
| "grad_norm": 0.6058363914489746, |
| "learning_rate": 0.0004792682137075013, |
| "loss": 3.5565, |
| "step": 18750 |
| }, |
| { |
| "epoch": 2.026954177897574, |
| "grad_norm": 0.5502753257751465, |
| "learning_rate": 0.0004789444144630329, |
| "loss": 3.5781, |
| "step": 18800 |
| }, |
| { |
| "epoch": 2.032345013477089, |
| "grad_norm": 0.5815129280090332, |
| "learning_rate": 0.00047862061521856443, |
| "loss": 3.5754, |
| "step": 18850 |
| }, |
| { |
| "epoch": 2.0377358490566038, |
| "grad_norm": 0.5740176439285278, |
| "learning_rate": 0.00047829681597409603, |
| "loss": 3.5584, |
| "step": 18900 |
| }, |
| { |
| "epoch": 2.0431266846361185, |
| "grad_norm": 0.5666323304176331, |
| "learning_rate": 0.0004779730167296276, |
| "loss": 3.5638, |
| "step": 18950 |
| }, |
| { |
| "epoch": 2.0485175202156336, |
| "grad_norm": 0.5930228233337402, |
| "learning_rate": 0.0004776492174851592, |
| "loss": 3.5922, |
| "step": 19000 |
| }, |
| { |
| "epoch": 2.0485175202156336, |
| "eval_accuracy": 0.36108874974317173, |
| "eval_loss": 3.588949203491211, |
| "eval_runtime": 144.8607, |
| "eval_samples_per_second": 124.333, |
| "eval_steps_per_second": 7.773, |
| "step": 19000 |
| }, |
| { |
| "epoch": 2.0539083557951483, |
| "grad_norm": 0.5626940727233887, |
| "learning_rate": 0.00047732541824069074, |
| "loss": 3.5695, |
| "step": 19050 |
| }, |
| { |
| "epoch": 2.059299191374663, |
| "grad_norm": 0.6210435628890991, |
| "learning_rate": 0.0004770016189962223, |
| "loss": 3.5607, |
| "step": 19100 |
| }, |
| { |
| "epoch": 2.0646900269541777, |
| "grad_norm": 0.5758371949195862, |
| "learning_rate": 0.0004766778197517539, |
| "loss": 3.5751, |
| "step": 19150 |
| }, |
| { |
| "epoch": 2.070080862533693, |
| "grad_norm": 0.6037834286689758, |
| "learning_rate": 0.0004763540205072854, |
| "loss": 3.5875, |
| "step": 19200 |
| }, |
| { |
| "epoch": 2.0754716981132075, |
| "grad_norm": 0.5686853528022766, |
| "learning_rate": 0.00047603022126281705, |
| "loss": 3.5713, |
| "step": 19250 |
| }, |
| { |
| "epoch": 2.0808625336927222, |
| "grad_norm": 0.6485291123390198, |
| "learning_rate": 0.00047570642201834855, |
| "loss": 3.5609, |
| "step": 19300 |
| }, |
| { |
| "epoch": 2.0862533692722374, |
| "grad_norm": 0.6449443697929382, |
| "learning_rate": 0.00047538262277388015, |
| "loss": 3.5789, |
| "step": 19350 |
| }, |
| { |
| "epoch": 2.091644204851752, |
| "grad_norm": 0.550469696521759, |
| "learning_rate": 0.0004750588235294117, |
| "loss": 3.554, |
| "step": 19400 |
| }, |
| { |
| "epoch": 2.0970350404312668, |
| "grad_norm": 0.5864623785018921, |
| "learning_rate": 0.0004747350242849433, |
| "loss": 3.5732, |
| "step": 19450 |
| }, |
| { |
| "epoch": 2.1024258760107815, |
| "grad_norm": 0.6241238117218018, |
| "learning_rate": 0.00047441122504047486, |
| "loss": 3.5717, |
| "step": 19500 |
| }, |
| { |
| "epoch": 2.1078167115902966, |
| "grad_norm": 0.602936327457428, |
| "learning_rate": 0.0004740874257960064, |
| "loss": 3.5604, |
| "step": 19550 |
| }, |
| { |
| "epoch": 2.1132075471698113, |
| "grad_norm": 0.5674468874931335, |
| "learning_rate": 0.000473763626551538, |
| "loss": 3.5643, |
| "step": 19600 |
| }, |
| { |
| "epoch": 2.118598382749326, |
| "grad_norm": 0.600834310054779, |
| "learning_rate": 0.00047343982730706956, |
| "loss": 3.5695, |
| "step": 19650 |
| }, |
| { |
| "epoch": 2.123989218328841, |
| "grad_norm": 0.5601726770401001, |
| "learning_rate": 0.00047311602806260117, |
| "loss": 3.5724, |
| "step": 19700 |
| }, |
| { |
| "epoch": 2.129380053908356, |
| "grad_norm": 0.5551213622093201, |
| "learning_rate": 0.0004727922288181327, |
| "loss": 3.5783, |
| "step": 19750 |
| }, |
| { |
| "epoch": 2.1347708894878705, |
| "grad_norm": 0.5737272500991821, |
| "learning_rate": 0.0004724684295736643, |
| "loss": 3.5642, |
| "step": 19800 |
| }, |
| { |
| "epoch": 2.1401617250673857, |
| "grad_norm": 0.5712146162986755, |
| "learning_rate": 0.0004721446303291959, |
| "loss": 3.5894, |
| "step": 19850 |
| }, |
| { |
| "epoch": 2.1455525606469004, |
| "grad_norm": 0.5902134776115417, |
| "learning_rate": 0.0004718208310847275, |
| "loss": 3.5903, |
| "step": 19900 |
| }, |
| { |
| "epoch": 2.150943396226415, |
| "grad_norm": 0.6139191389083862, |
| "learning_rate": 0.000471497031840259, |
| "loss": 3.5862, |
| "step": 19950 |
| }, |
| { |
| "epoch": 2.1563342318059298, |
| "grad_norm": 0.5515206456184387, |
| "learning_rate": 0.0004711732325957905, |
| "loss": 3.5706, |
| "step": 20000 |
| }, |
| { |
| "epoch": 2.1563342318059298, |
| "eval_accuracy": 0.3621358376230399, |
| "eval_loss": 3.5808968544006348, |
| "eval_runtime": 144.61, |
| "eval_samples_per_second": 124.549, |
| "eval_steps_per_second": 7.786, |
| "step": 20000 |
| }, |
| { |
| "epoch": 2.161725067385445, |
| "grad_norm": 0.6040294170379639, |
| "learning_rate": 0.00047084943335132213, |
| "loss": 3.5692, |
| "step": 20050 |
| }, |
| { |
| "epoch": 2.1671159029649596, |
| "grad_norm": 0.6468523740768433, |
| "learning_rate": 0.0004705256341068537, |
| "loss": 3.5764, |
| "step": 20100 |
| }, |
| { |
| "epoch": 2.1725067385444743, |
| "grad_norm": 0.5732126235961914, |
| "learning_rate": 0.0004702018348623853, |
| "loss": 3.572, |
| "step": 20150 |
| }, |
| { |
| "epoch": 2.177897574123989, |
| "grad_norm": 0.545032262802124, |
| "learning_rate": 0.00046987803561791684, |
| "loss": 3.5862, |
| "step": 20200 |
| }, |
| { |
| "epoch": 2.183288409703504, |
| "grad_norm": 0.5633872747421265, |
| "learning_rate": 0.00046955423637344844, |
| "loss": 3.5738, |
| "step": 20250 |
| }, |
| { |
| "epoch": 2.188679245283019, |
| "grad_norm": 0.6415254473686218, |
| "learning_rate": 0.00046923043712898, |
| "loss": 3.5789, |
| "step": 20300 |
| }, |
| { |
| "epoch": 2.1940700808625335, |
| "grad_norm": 0.6461573839187622, |
| "learning_rate": 0.0004689066378845116, |
| "loss": 3.5603, |
| "step": 20350 |
| }, |
| { |
| "epoch": 2.1994609164420487, |
| "grad_norm": 0.5884630680084229, |
| "learning_rate": 0.00046858283864004315, |
| "loss": 3.5676, |
| "step": 20400 |
| }, |
| { |
| "epoch": 2.2048517520215634, |
| "grad_norm": 0.5497499704360962, |
| "learning_rate": 0.0004682590393955747, |
| "loss": 3.5713, |
| "step": 20450 |
| }, |
| { |
| "epoch": 2.210242587601078, |
| "grad_norm": 0.5444649457931519, |
| "learning_rate": 0.0004679352401511063, |
| "loss": 3.5807, |
| "step": 20500 |
| }, |
| { |
| "epoch": 2.215633423180593, |
| "grad_norm": 0.5752549171447754, |
| "learning_rate": 0.0004676114409066378, |
| "loss": 3.576, |
| "step": 20550 |
| }, |
| { |
| "epoch": 2.221024258760108, |
| "grad_norm": 0.5602301359176636, |
| "learning_rate": 0.00046728764166216946, |
| "loss": 3.5703, |
| "step": 20600 |
| }, |
| { |
| "epoch": 2.2264150943396226, |
| "grad_norm": 0.5674179792404175, |
| "learning_rate": 0.00046696384241770095, |
| "loss": 3.5705, |
| "step": 20650 |
| }, |
| { |
| "epoch": 2.2318059299191373, |
| "grad_norm": 0.6246010661125183, |
| "learning_rate": 0.00046664004317323256, |
| "loss": 3.5675, |
| "step": 20700 |
| }, |
| { |
| "epoch": 2.2371967654986524, |
| "grad_norm": 0.5755497217178345, |
| "learning_rate": 0.0004663162439287641, |
| "loss": 3.5682, |
| "step": 20750 |
| }, |
| { |
| "epoch": 2.242587601078167, |
| "grad_norm": 0.601298987865448, |
| "learning_rate": 0.00046599244468429566, |
| "loss": 3.5731, |
| "step": 20800 |
| }, |
| { |
| "epoch": 2.247978436657682, |
| "grad_norm": 0.6204710602760315, |
| "learning_rate": 0.00046566864543982726, |
| "loss": 3.5701, |
| "step": 20850 |
| }, |
| { |
| "epoch": 2.2533692722371965, |
| "grad_norm": 0.5474254488945007, |
| "learning_rate": 0.0004653448461953588, |
| "loss": 3.5649, |
| "step": 20900 |
| }, |
| { |
| "epoch": 2.2587601078167117, |
| "grad_norm": 0.5631576776504517, |
| "learning_rate": 0.00046502752293577977, |
| "loss": 3.5743, |
| "step": 20950 |
| }, |
| { |
| "epoch": 2.2641509433962264, |
| "grad_norm": 0.5887001752853394, |
| "learning_rate": 0.00046470372369131137, |
| "loss": 3.5568, |
| "step": 21000 |
| }, |
| { |
| "epoch": 2.2641509433962264, |
| "eval_accuracy": 0.3631279471416739, |
| "eval_loss": 3.5703935623168945, |
| "eval_runtime": 144.8136, |
| "eval_samples_per_second": 124.374, |
| "eval_steps_per_second": 7.776, |
| "step": 21000 |
| }, |
| { |
| "epoch": 2.269541778975741, |
| "grad_norm": 0.5233503580093384, |
| "learning_rate": 0.0004643799244468429, |
| "loss": 3.5712, |
| "step": 21050 |
| }, |
| { |
| "epoch": 2.274932614555256, |
| "grad_norm": 0.5863848924636841, |
| "learning_rate": 0.0004640561252023745, |
| "loss": 3.5732, |
| "step": 21100 |
| }, |
| { |
| "epoch": 2.280323450134771, |
| "grad_norm": 0.610373854637146, |
| "learning_rate": 0.0004637323259579061, |
| "loss": 3.5865, |
| "step": 21150 |
| }, |
| { |
| "epoch": 2.2857142857142856, |
| "grad_norm": 0.5934001207351685, |
| "learning_rate": 0.0004634085267134376, |
| "loss": 3.5383, |
| "step": 21200 |
| }, |
| { |
| "epoch": 2.2911051212938007, |
| "grad_norm": 0.550365686416626, |
| "learning_rate": 0.00046308472746896923, |
| "loss": 3.5677, |
| "step": 21250 |
| }, |
| { |
| "epoch": 2.2964959568733154, |
| "grad_norm": 0.5905864238739014, |
| "learning_rate": 0.00046276092822450073, |
| "loss": 3.5684, |
| "step": 21300 |
| }, |
| { |
| "epoch": 2.30188679245283, |
| "grad_norm": 0.5769224166870117, |
| "learning_rate": 0.00046243712898003233, |
| "loss": 3.5621, |
| "step": 21350 |
| }, |
| { |
| "epoch": 2.3072776280323453, |
| "grad_norm": 0.5868223905563354, |
| "learning_rate": 0.0004621133297355639, |
| "loss": 3.5773, |
| "step": 21400 |
| }, |
| { |
| "epoch": 2.31266846361186, |
| "grad_norm": 0.5571302175521851, |
| "learning_rate": 0.0004617895304910955, |
| "loss": 3.5739, |
| "step": 21450 |
| }, |
| { |
| "epoch": 2.3180592991913747, |
| "grad_norm": 0.5874112844467163, |
| "learning_rate": 0.00046146573124662704, |
| "loss": 3.5778, |
| "step": 21500 |
| }, |
| { |
| "epoch": 2.3234501347708894, |
| "grad_norm": 0.6028329133987427, |
| "learning_rate": 0.00046114193200215864, |
| "loss": 3.558, |
| "step": 21550 |
| }, |
| { |
| "epoch": 2.3288409703504045, |
| "grad_norm": 0.5316165089607239, |
| "learning_rate": 0.0004608181327576902, |
| "loss": 3.5785, |
| "step": 21600 |
| }, |
| { |
| "epoch": 2.334231805929919, |
| "grad_norm": 0.5828713178634644, |
| "learning_rate": 0.00046049433351322175, |
| "loss": 3.5738, |
| "step": 21650 |
| }, |
| { |
| "epoch": 2.339622641509434, |
| "grad_norm": 0.5575670003890991, |
| "learning_rate": 0.00046017053426875335, |
| "loss": 3.5545, |
| "step": 21700 |
| }, |
| { |
| "epoch": 2.3450134770889486, |
| "grad_norm": 0.578111469745636, |
| "learning_rate": 0.0004598467350242849, |
| "loss": 3.5695, |
| "step": 21750 |
| }, |
| { |
| "epoch": 2.3504043126684637, |
| "grad_norm": 0.5924418568611145, |
| "learning_rate": 0.0004595229357798165, |
| "loss": 3.5927, |
| "step": 21800 |
| }, |
| { |
| "epoch": 2.3557951482479784, |
| "grad_norm": 0.5741672515869141, |
| "learning_rate": 0.00045919913653534806, |
| "loss": 3.576, |
| "step": 21850 |
| }, |
| { |
| "epoch": 2.361185983827493, |
| "grad_norm": 0.6206271648406982, |
| "learning_rate": 0.00045887533729087966, |
| "loss": 3.5548, |
| "step": 21900 |
| }, |
| { |
| "epoch": 2.3665768194070083, |
| "grad_norm": 0.6218240857124329, |
| "learning_rate": 0.00045855153804641116, |
| "loss": 3.5706, |
| "step": 21950 |
| }, |
| { |
| "epoch": 2.371967654986523, |
| "grad_norm": 0.5502433180809021, |
| "learning_rate": 0.0004582277388019427, |
| "loss": 3.5596, |
| "step": 22000 |
| }, |
| { |
| "epoch": 2.371967654986523, |
| "eval_accuracy": 0.36454162986502814, |
| "eval_loss": 3.558351993560791, |
| "eval_runtime": 144.2942, |
| "eval_samples_per_second": 124.821, |
| "eval_steps_per_second": 7.804, |
| "step": 22000 |
| }, |
| { |
| "epoch": 2.3773584905660377, |
| "grad_norm": 0.5923473238945007, |
| "learning_rate": 0.0004579039395574743, |
| "loss": 3.5625, |
| "step": 22050 |
| }, |
| { |
| "epoch": 2.382749326145553, |
| "grad_norm": 0.5796228051185608, |
| "learning_rate": 0.00045758014031300586, |
| "loss": 3.5745, |
| "step": 22100 |
| }, |
| { |
| "epoch": 2.3881401617250675, |
| "grad_norm": 0.5643661618232727, |
| "learning_rate": 0.00045725634106853747, |
| "loss": 3.5762, |
| "step": 22150 |
| }, |
| { |
| "epoch": 2.393530997304582, |
| "grad_norm": 0.563675582408905, |
| "learning_rate": 0.000456932541824069, |
| "loss": 3.5626, |
| "step": 22200 |
| }, |
| { |
| "epoch": 2.398921832884097, |
| "grad_norm": 0.5896977782249451, |
| "learning_rate": 0.0004566087425796006, |
| "loss": 3.5699, |
| "step": 22250 |
| }, |
| { |
| "epoch": 2.404312668463612, |
| "grad_norm": 0.5657273530960083, |
| "learning_rate": 0.0004562849433351322, |
| "loss": 3.5572, |
| "step": 22300 |
| }, |
| { |
| "epoch": 2.4097035040431267, |
| "grad_norm": 0.5821571946144104, |
| "learning_rate": 0.0004559611440906638, |
| "loss": 3.5794, |
| "step": 22350 |
| }, |
| { |
| "epoch": 2.4150943396226414, |
| "grad_norm": 0.576914370059967, |
| "learning_rate": 0.00045563734484619533, |
| "loss": 3.5643, |
| "step": 22400 |
| }, |
| { |
| "epoch": 2.420485175202156, |
| "grad_norm": 0.5930933952331543, |
| "learning_rate": 0.0004553135456017269, |
| "loss": 3.5656, |
| "step": 22450 |
| }, |
| { |
| "epoch": 2.4258760107816713, |
| "grad_norm": 0.6359766721725464, |
| "learning_rate": 0.0004549897463572585, |
| "loss": 3.5722, |
| "step": 22500 |
| }, |
| { |
| "epoch": 2.431266846361186, |
| "grad_norm": 0.5717756152153015, |
| "learning_rate": 0.00045466594711279, |
| "loss": 3.5636, |
| "step": 22550 |
| }, |
| { |
| "epoch": 2.4366576819407006, |
| "grad_norm": 0.6088978052139282, |
| "learning_rate": 0.00045434214786832164, |
| "loss": 3.5952, |
| "step": 22600 |
| }, |
| { |
| "epoch": 2.442048517520216, |
| "grad_norm": 0.596017062664032, |
| "learning_rate": 0.00045401834862385314, |
| "loss": 3.5571, |
| "step": 22650 |
| }, |
| { |
| "epoch": 2.4474393530997305, |
| "grad_norm": 0.5859695076942444, |
| "learning_rate": 0.00045369454937938474, |
| "loss": 3.5462, |
| "step": 22700 |
| }, |
| { |
| "epoch": 2.452830188679245, |
| "grad_norm": 0.5753331780433655, |
| "learning_rate": 0.0004533707501349163, |
| "loss": 3.5518, |
| "step": 22750 |
| }, |
| { |
| "epoch": 2.4582210242587603, |
| "grad_norm": 0.569882333278656, |
| "learning_rate": 0.0004530469508904479, |
| "loss": 3.5628, |
| "step": 22800 |
| }, |
| { |
| "epoch": 2.463611859838275, |
| "grad_norm": 0.604755699634552, |
| "learning_rate": 0.00045272315164597945, |
| "loss": 3.5464, |
| "step": 22850 |
| }, |
| { |
| "epoch": 2.4690026954177897, |
| "grad_norm": 0.6051239967346191, |
| "learning_rate": 0.000452399352401511, |
| "loss": 3.563, |
| "step": 22900 |
| }, |
| { |
| "epoch": 2.4743935309973044, |
| "grad_norm": 0.5919761061668396, |
| "learning_rate": 0.0004520755531570426, |
| "loss": 3.5611, |
| "step": 22950 |
| }, |
| { |
| "epoch": 2.4797843665768196, |
| "grad_norm": 0.5852389335632324, |
| "learning_rate": 0.00045175822989746355, |
| "loss": 3.5529, |
| "step": 23000 |
| }, |
| { |
| "epoch": 2.4797843665768196, |
| "eval_accuracy": 0.3654994050711128, |
| "eval_loss": 3.5451345443725586, |
| "eval_runtime": 144.5225, |
| "eval_samples_per_second": 124.624, |
| "eval_steps_per_second": 7.791, |
| "step": 23000 |
| }, |
| { |
| "epoch": 2.4851752021563343, |
| "grad_norm": 0.5408614873886108, |
| "learning_rate": 0.0004514344306529951, |
| "loss": 3.5645, |
| "step": 23050 |
| }, |
| { |
| "epoch": 2.490566037735849, |
| "grad_norm": 0.5751845836639404, |
| "learning_rate": 0.0004511106314085267, |
| "loss": 3.5671, |
| "step": 23100 |
| }, |
| { |
| "epoch": 2.4959568733153636, |
| "grad_norm": 0.5665581226348877, |
| "learning_rate": 0.00045078683216405826, |
| "loss": 3.5651, |
| "step": 23150 |
| }, |
| { |
| "epoch": 2.501347708894879, |
| "grad_norm": 0.5446679592132568, |
| "learning_rate": 0.00045046303291958976, |
| "loss": 3.5602, |
| "step": 23200 |
| }, |
| { |
| "epoch": 2.5067385444743935, |
| "grad_norm": 0.6096646189689636, |
| "learning_rate": 0.0004501392336751214, |
| "loss": 3.5606, |
| "step": 23250 |
| }, |
| { |
| "epoch": 2.512129380053908, |
| "grad_norm": 0.5711411237716675, |
| "learning_rate": 0.0004498154344306529, |
| "loss": 3.5462, |
| "step": 23300 |
| }, |
| { |
| "epoch": 2.5175202156334233, |
| "grad_norm": 0.5946795344352722, |
| "learning_rate": 0.0004494916351861845, |
| "loss": 3.5517, |
| "step": 23350 |
| }, |
| { |
| "epoch": 2.522911051212938, |
| "grad_norm": 0.6075916886329651, |
| "learning_rate": 0.00044916783594171607, |
| "loss": 3.5477, |
| "step": 23400 |
| }, |
| { |
| "epoch": 2.5283018867924527, |
| "grad_norm": 0.5496644377708435, |
| "learning_rate": 0.00044884403669724767, |
| "loss": 3.5583, |
| "step": 23450 |
| }, |
| { |
| "epoch": 2.533692722371968, |
| "grad_norm": 0.5712068676948547, |
| "learning_rate": 0.0004485202374527792, |
| "loss": 3.5607, |
| "step": 23500 |
| }, |
| { |
| "epoch": 2.5390835579514826, |
| "grad_norm": 0.5820903182029724, |
| "learning_rate": 0.00044819643820831083, |
| "loss": 3.5643, |
| "step": 23550 |
| }, |
| { |
| "epoch": 2.5444743935309972, |
| "grad_norm": 0.5886263847351074, |
| "learning_rate": 0.0004478726389638424, |
| "loss": 3.5662, |
| "step": 23600 |
| }, |
| { |
| "epoch": 2.5498652291105124, |
| "grad_norm": 0.5735592842102051, |
| "learning_rate": 0.00044754883971937393, |
| "loss": 3.5579, |
| "step": 23650 |
| }, |
| { |
| "epoch": 2.555256064690027, |
| "grad_norm": 0.6012240052223206, |
| "learning_rate": 0.00044722504047490553, |
| "loss": 3.5642, |
| "step": 23700 |
| }, |
| { |
| "epoch": 2.560646900269542, |
| "grad_norm": 0.5736075639724731, |
| "learning_rate": 0.0004469012412304371, |
| "loss": 3.5574, |
| "step": 23750 |
| }, |
| { |
| "epoch": 2.5660377358490565, |
| "grad_norm": 0.6146102547645569, |
| "learning_rate": 0.0004465774419859687, |
| "loss": 3.5704, |
| "step": 23800 |
| }, |
| { |
| "epoch": 2.571428571428571, |
| "grad_norm": 0.5984640717506409, |
| "learning_rate": 0.00044625364274150024, |
| "loss": 3.5763, |
| "step": 23850 |
| }, |
| { |
| "epoch": 2.5768194070080863, |
| "grad_norm": 0.5764756798744202, |
| "learning_rate": 0.00044592984349703184, |
| "loss": 3.5746, |
| "step": 23900 |
| }, |
| { |
| "epoch": 2.582210242587601, |
| "grad_norm": 0.6541941165924072, |
| "learning_rate": 0.00044560604425256334, |
| "loss": 3.55, |
| "step": 23950 |
| }, |
| { |
| "epoch": 2.5876010781671157, |
| "grad_norm": 0.5598556399345398, |
| "learning_rate": 0.000445282245008095, |
| "loss": 3.5466, |
| "step": 24000 |
| }, |
| { |
| "epoch": 2.5876010781671157, |
| "eval_accuracy": 0.36651433169618786, |
| "eval_loss": 3.5378851890563965, |
| "eval_runtime": 144.1764, |
| "eval_samples_per_second": 124.923, |
| "eval_steps_per_second": 7.81, |
| "step": 24000 |
| }, |
| { |
| "epoch": 2.592991913746631, |
| "grad_norm": 0.6188113689422607, |
| "learning_rate": 0.0004449584457636265, |
| "loss": 3.5458, |
| "step": 24050 |
| }, |
| { |
| "epoch": 2.5983827493261455, |
| "grad_norm": 0.5713769197463989, |
| "learning_rate": 0.00044463464651915805, |
| "loss": 3.5473, |
| "step": 24100 |
| }, |
| { |
| "epoch": 2.6037735849056602, |
| "grad_norm": 0.5993424654006958, |
| "learning_rate": 0.00044431084727468965, |
| "loss": 3.5493, |
| "step": 24150 |
| }, |
| { |
| "epoch": 2.6091644204851754, |
| "grad_norm": 0.6186326146125793, |
| "learning_rate": 0.0004439870480302212, |
| "loss": 3.5668, |
| "step": 24200 |
| }, |
| { |
| "epoch": 2.61455525606469, |
| "grad_norm": 0.5661455392837524, |
| "learning_rate": 0.0004436632487857528, |
| "loss": 3.5703, |
| "step": 24250 |
| }, |
| { |
| "epoch": 2.6199460916442048, |
| "grad_norm": 0.6343756318092346, |
| "learning_rate": 0.00044333944954128436, |
| "loss": 3.5559, |
| "step": 24300 |
| }, |
| { |
| "epoch": 2.62533692722372, |
| "grad_norm": 0.5699751377105713, |
| "learning_rate": 0.00044301565029681596, |
| "loss": 3.5509, |
| "step": 24350 |
| }, |
| { |
| "epoch": 2.6307277628032346, |
| "grad_norm": 0.5953941941261292, |
| "learning_rate": 0.0004426918510523475, |
| "loss": 3.5609, |
| "step": 24400 |
| }, |
| { |
| "epoch": 2.6361185983827493, |
| "grad_norm": 0.5832213163375854, |
| "learning_rate": 0.0004423680518078791, |
| "loss": 3.531, |
| "step": 24450 |
| }, |
| { |
| "epoch": 2.641509433962264, |
| "grad_norm": 0.6015488505363464, |
| "learning_rate": 0.00044204425256341067, |
| "loss": 3.5396, |
| "step": 24500 |
| }, |
| { |
| "epoch": 2.6469002695417787, |
| "grad_norm": 0.6897053718566895, |
| "learning_rate": 0.0004417204533189422, |
| "loss": 3.5562, |
| "step": 24550 |
| }, |
| { |
| "epoch": 2.652291105121294, |
| "grad_norm": 0.5880343317985535, |
| "learning_rate": 0.0004413966540744738, |
| "loss": 3.5371, |
| "step": 24600 |
| }, |
| { |
| "epoch": 2.6576819407008085, |
| "grad_norm": 0.574292778968811, |
| "learning_rate": 0.0004410728548300053, |
| "loss": 3.5532, |
| "step": 24650 |
| }, |
| { |
| "epoch": 2.6630727762803232, |
| "grad_norm": 0.6216583251953125, |
| "learning_rate": 0.0004407490555855369, |
| "loss": 3.5528, |
| "step": 24700 |
| }, |
| { |
| "epoch": 2.6684636118598384, |
| "grad_norm": 0.6199180483818054, |
| "learning_rate": 0.0004404252563410685, |
| "loss": 3.5568, |
| "step": 24750 |
| }, |
| { |
| "epoch": 2.673854447439353, |
| "grad_norm": 0.6043687462806702, |
| "learning_rate": 0.0004401014570966001, |
| "loss": 3.5543, |
| "step": 24800 |
| }, |
| { |
| "epoch": 2.6792452830188678, |
| "grad_norm": 0.5746572613716125, |
| "learning_rate": 0.00043977765785213163, |
| "loss": 3.5559, |
| "step": 24850 |
| }, |
| { |
| "epoch": 2.684636118598383, |
| "grad_norm": 0.5911267995834351, |
| "learning_rate": 0.0004394538586076632, |
| "loss": 3.5716, |
| "step": 24900 |
| }, |
| { |
| "epoch": 2.6900269541778976, |
| "grad_norm": 0.5925143361091614, |
| "learning_rate": 0.0004391300593631948, |
| "loss": 3.5607, |
| "step": 24950 |
| }, |
| { |
| "epoch": 2.6954177897574123, |
| "grad_norm": 0.5580176711082458, |
| "learning_rate": 0.00043880626011872634, |
| "loss": 3.555, |
| "step": 25000 |
| }, |
| { |
| "epoch": 2.6954177897574123, |
| "eval_accuracy": 0.36743233994533236, |
| "eval_loss": 3.5253796577453613, |
| "eval_runtime": 144.4202, |
| "eval_samples_per_second": 124.712, |
| "eval_steps_per_second": 7.797, |
| "step": 25000 |
| }, |
| { |
| "epoch": 2.7008086253369274, |
| "grad_norm": 0.576725959777832, |
| "learning_rate": 0.0004384889368591473, |
| "loss": 3.5449, |
| "step": 25050 |
| }, |
| { |
| "epoch": 2.706199460916442, |
| "grad_norm": 0.545190691947937, |
| "learning_rate": 0.0004381651376146789, |
| "loss": 3.5538, |
| "step": 25100 |
| }, |
| { |
| "epoch": 2.711590296495957, |
| "grad_norm": 0.5915215611457825, |
| "learning_rate": 0.00043784133837021044, |
| "loss": 3.5311, |
| "step": 25150 |
| }, |
| { |
| "epoch": 2.7169811320754715, |
| "grad_norm": 0.6212695837020874, |
| "learning_rate": 0.00043751753912574205, |
| "loss": 3.5534, |
| "step": 25200 |
| }, |
| { |
| "epoch": 2.7223719676549867, |
| "grad_norm": 0.5867090225219727, |
| "learning_rate": 0.0004371937398812736, |
| "loss": 3.5584, |
| "step": 25250 |
| }, |
| { |
| "epoch": 2.7277628032345014, |
| "grad_norm": 0.5455405116081238, |
| "learning_rate": 0.0004368699406368051, |
| "loss": 3.5605, |
| "step": 25300 |
| }, |
| { |
| "epoch": 2.733153638814016, |
| "grad_norm": 0.6148755550384521, |
| "learning_rate": 0.0004365461413923367, |
| "loss": 3.5503, |
| "step": 25350 |
| }, |
| { |
| "epoch": 2.7385444743935308, |
| "grad_norm": 0.6133694052696228, |
| "learning_rate": 0.00043622234214786825, |
| "loss": 3.5414, |
| "step": 25400 |
| }, |
| { |
| "epoch": 2.743935309973046, |
| "grad_norm": 0.5954641699790955, |
| "learning_rate": 0.00043589854290339985, |
| "loss": 3.5439, |
| "step": 25450 |
| }, |
| { |
| "epoch": 2.7493261455525606, |
| "grad_norm": 0.5887457728385925, |
| "learning_rate": 0.0004355747436589314, |
| "loss": 3.5571, |
| "step": 25500 |
| }, |
| { |
| "epoch": 2.7547169811320753, |
| "grad_norm": 0.573692798614502, |
| "learning_rate": 0.000435250944414463, |
| "loss": 3.5495, |
| "step": 25550 |
| }, |
| { |
| "epoch": 2.7601078167115904, |
| "grad_norm": 0.5801973938941956, |
| "learning_rate": 0.00043492714516999456, |
| "loss": 3.5522, |
| "step": 25600 |
| }, |
| { |
| "epoch": 2.765498652291105, |
| "grad_norm": 0.583278238773346, |
| "learning_rate": 0.0004346033459255261, |
| "loss": 3.5448, |
| "step": 25650 |
| }, |
| { |
| "epoch": 2.77088948787062, |
| "grad_norm": 0.6692058444023132, |
| "learning_rate": 0.0004342795466810577, |
| "loss": 3.5509, |
| "step": 25700 |
| }, |
| { |
| "epoch": 2.776280323450135, |
| "grad_norm": 0.609228253364563, |
| "learning_rate": 0.00043395574743658927, |
| "loss": 3.5385, |
| "step": 25750 |
| }, |
| { |
| "epoch": 2.7816711590296497, |
| "grad_norm": 0.5607845187187195, |
| "learning_rate": 0.00043363194819212087, |
| "loss": 3.5623, |
| "step": 25800 |
| }, |
| { |
| "epoch": 2.7870619946091644, |
| "grad_norm": 0.5615296363830566, |
| "learning_rate": 0.0004333081489476524, |
| "loss": 3.5353, |
| "step": 25850 |
| }, |
| { |
| "epoch": 2.7924528301886795, |
| "grad_norm": 0.5903550982475281, |
| "learning_rate": 0.000432984349703184, |
| "loss": 3.5447, |
| "step": 25900 |
| }, |
| { |
| "epoch": 2.797843665768194, |
| "grad_norm": 0.6404184103012085, |
| "learning_rate": 0.0004326605504587155, |
| "loss": 3.5401, |
| "step": 25950 |
| }, |
| { |
| "epoch": 2.803234501347709, |
| "grad_norm": 0.5671994090080261, |
| "learning_rate": 0.0004323367512142472, |
| "loss": 3.5597, |
| "step": 26000 |
| }, |
| { |
| "epoch": 2.803234501347709, |
| "eval_accuracy": 0.36840749961346736, |
| "eval_loss": 3.5180230140686035, |
| "eval_runtime": 144.5115, |
| "eval_samples_per_second": 124.634, |
| "eval_steps_per_second": 7.792, |
| "step": 26000 |
| }, |
| { |
| "epoch": 2.8086253369272236, |
| "grad_norm": 0.5972022414207458, |
| "learning_rate": 0.0004320129519697787, |
| "loss": 3.5636, |
| "step": 26050 |
| }, |
| { |
| "epoch": 2.8140161725067383, |
| "grad_norm": 0.5757322907447815, |
| "learning_rate": 0.00043168915272531023, |
| "loss": 3.5479, |
| "step": 26100 |
| }, |
| { |
| "epoch": 2.8194070080862534, |
| "grad_norm": 0.5859302282333374, |
| "learning_rate": 0.00043136535348084183, |
| "loss": 3.5425, |
| "step": 26150 |
| }, |
| { |
| "epoch": 2.824797843665768, |
| "grad_norm": 0.5637912154197693, |
| "learning_rate": 0.0004310415542363734, |
| "loss": 3.555, |
| "step": 26200 |
| }, |
| { |
| "epoch": 2.830188679245283, |
| "grad_norm": 0.5691296458244324, |
| "learning_rate": 0.000430717754991905, |
| "loss": 3.5357, |
| "step": 26250 |
| }, |
| { |
| "epoch": 2.835579514824798, |
| "grad_norm": 0.6390897631645203, |
| "learning_rate": 0.00043039395574743654, |
| "loss": 3.5369, |
| "step": 26300 |
| }, |
| { |
| "epoch": 2.8409703504043127, |
| "grad_norm": 0.6023745536804199, |
| "learning_rate": 0.00043007015650296814, |
| "loss": 3.5569, |
| "step": 26350 |
| }, |
| { |
| "epoch": 2.8463611859838274, |
| "grad_norm": 0.5968989133834839, |
| "learning_rate": 0.0004297463572584997, |
| "loss": 3.5324, |
| "step": 26400 |
| }, |
| { |
| "epoch": 2.8517520215633425, |
| "grad_norm": 0.5747265219688416, |
| "learning_rate": 0.0004294225580140313, |
| "loss": 3.5554, |
| "step": 26450 |
| }, |
| { |
| "epoch": 2.857142857142857, |
| "grad_norm": 0.5473029017448425, |
| "learning_rate": 0.00042909875876956285, |
| "loss": 3.5601, |
| "step": 26500 |
| }, |
| { |
| "epoch": 2.862533692722372, |
| "grad_norm": 0.6153762936592102, |
| "learning_rate": 0.0004287749595250944, |
| "loss": 3.5416, |
| "step": 26550 |
| }, |
| { |
| "epoch": 2.867924528301887, |
| "grad_norm": 0.5814242959022522, |
| "learning_rate": 0.000428451160280626, |
| "loss": 3.5436, |
| "step": 26600 |
| }, |
| { |
| "epoch": 2.8733153638814017, |
| "grad_norm": 0.544052243232727, |
| "learning_rate": 0.0004281273610361575, |
| "loss": 3.5283, |
| "step": 26650 |
| }, |
| { |
| "epoch": 2.8787061994609164, |
| "grad_norm": 0.580087423324585, |
| "learning_rate": 0.0004278035617916891, |
| "loss": 3.5665, |
| "step": 26700 |
| }, |
| { |
| "epoch": 2.884097035040431, |
| "grad_norm": 0.606545627117157, |
| "learning_rate": 0.00042747976254722066, |
| "loss": 3.5333, |
| "step": 26750 |
| }, |
| { |
| "epoch": 2.889487870619946, |
| "grad_norm": 0.5613415241241455, |
| "learning_rate": 0.00042715596330275226, |
| "loss": 3.5188, |
| "step": 26800 |
| }, |
| { |
| "epoch": 2.894878706199461, |
| "grad_norm": 0.5772558450698853, |
| "learning_rate": 0.0004268321640582838, |
| "loss": 3.521, |
| "step": 26850 |
| }, |
| { |
| "epoch": 2.9002695417789757, |
| "grad_norm": 0.5379123091697693, |
| "learning_rate": 0.0004265083648138154, |
| "loss": 3.5405, |
| "step": 26900 |
| }, |
| { |
| "epoch": 2.9056603773584904, |
| "grad_norm": 0.5935842394828796, |
| "learning_rate": 0.00042618456556934697, |
| "loss": 3.5612, |
| "step": 26950 |
| }, |
| { |
| "epoch": 2.9110512129380055, |
| "grad_norm": 0.622439444065094, |
| "learning_rate": 0.0004258607663248785, |
| "loss": 3.5422, |
| "step": 27000 |
| }, |
| { |
| "epoch": 2.9110512129380055, |
| "eval_accuracy": 0.3692755275342172, |
| "eval_loss": 3.508702039718628, |
| "eval_runtime": 144.9148, |
| "eval_samples_per_second": 124.287, |
| "eval_steps_per_second": 7.77, |
| "step": 27000 |
| }, |
| { |
| "epoch": 2.91644204851752, |
| "grad_norm": 0.6212618947029114, |
| "learning_rate": 0.00042554344306529947, |
| "loss": 3.531, |
| "step": 27050 |
| }, |
| { |
| "epoch": 2.921832884097035, |
| "grad_norm": 0.6010525226593018, |
| "learning_rate": 0.0004252196438208311, |
| "loss": 3.5393, |
| "step": 27100 |
| }, |
| { |
| "epoch": 2.92722371967655, |
| "grad_norm": 0.5927774310112, |
| "learning_rate": 0.0004248958445763626, |
| "loss": 3.5333, |
| "step": 27150 |
| }, |
| { |
| "epoch": 2.9326145552560647, |
| "grad_norm": 0.6531330347061157, |
| "learning_rate": 0.00042457204533189423, |
| "loss": 3.5434, |
| "step": 27200 |
| }, |
| { |
| "epoch": 2.9380053908355794, |
| "grad_norm": 0.584989607334137, |
| "learning_rate": 0.0004242482460874258, |
| "loss": 3.5498, |
| "step": 27250 |
| }, |
| { |
| "epoch": 2.9433962264150946, |
| "grad_norm": 0.6219198107719421, |
| "learning_rate": 0.0004239244468429573, |
| "loss": 3.5142, |
| "step": 27300 |
| }, |
| { |
| "epoch": 2.9487870619946093, |
| "grad_norm": 0.6168526411056519, |
| "learning_rate": 0.0004236006475984889, |
| "loss": 3.5346, |
| "step": 27350 |
| }, |
| { |
| "epoch": 2.954177897574124, |
| "grad_norm": 0.583755373954773, |
| "learning_rate": 0.00042327684835402043, |
| "loss": 3.5291, |
| "step": 27400 |
| }, |
| { |
| "epoch": 2.9595687331536387, |
| "grad_norm": 0.6100607514381409, |
| "learning_rate": 0.00042295304910955204, |
| "loss": 3.5208, |
| "step": 27450 |
| }, |
| { |
| "epoch": 2.964959568733154, |
| "grad_norm": 0.5626996159553528, |
| "learning_rate": 0.0004226292498650836, |
| "loss": 3.5461, |
| "step": 27500 |
| }, |
| { |
| "epoch": 2.9703504043126685, |
| "grad_norm": 0.5924006700515747, |
| "learning_rate": 0.0004223054506206152, |
| "loss": 3.5485, |
| "step": 27550 |
| }, |
| { |
| "epoch": 2.975741239892183, |
| "grad_norm": 0.5733758807182312, |
| "learning_rate": 0.00042198165137614674, |
| "loss": 3.5297, |
| "step": 27600 |
| }, |
| { |
| "epoch": 2.981132075471698, |
| "grad_norm": 0.6193982362747192, |
| "learning_rate": 0.00042165785213167835, |
| "loss": 3.5365, |
| "step": 27650 |
| }, |
| { |
| "epoch": 2.986522911051213, |
| "grad_norm": 0.5635408759117126, |
| "learning_rate": 0.0004213340528872099, |
| "loss": 3.5253, |
| "step": 27700 |
| }, |
| { |
| "epoch": 2.9919137466307277, |
| "grad_norm": 0.6305370926856995, |
| "learning_rate": 0.00042101025364274145, |
| "loss": 3.5239, |
| "step": 27750 |
| }, |
| { |
| "epoch": 2.9973045822102424, |
| "grad_norm": 0.581883430480957, |
| "learning_rate": 0.00042068645439827305, |
| "loss": 3.5336, |
| "step": 27800 |
| }, |
| { |
| "epoch": 3.0026954177897576, |
| "grad_norm": 0.5631284713745117, |
| "learning_rate": 0.0004203626551538046, |
| "loss": 3.4953, |
| "step": 27850 |
| }, |
| { |
| "epoch": 3.0080862533692723, |
| "grad_norm": 0.6554384827613831, |
| "learning_rate": 0.0004200388559093362, |
| "loss": 3.4436, |
| "step": 27900 |
| }, |
| { |
| "epoch": 3.013477088948787, |
| "grad_norm": 0.5941973924636841, |
| "learning_rate": 0.0004197150566648677, |
| "loss": 3.4446, |
| "step": 27950 |
| }, |
| { |
| "epoch": 3.018867924528302, |
| "grad_norm": 0.5603510141372681, |
| "learning_rate": 0.00041939125742039936, |
| "loss": 3.4422, |
| "step": 28000 |
| }, |
| { |
| "epoch": 3.018867924528302, |
| "eval_accuracy": 0.3700082826096381, |
| "eval_loss": 3.5040652751922607, |
| "eval_runtime": 144.5182, |
| "eval_samples_per_second": 124.628, |
| "eval_steps_per_second": 7.791, |
| "step": 28000 |
| }, |
| { |
| "epoch": 3.024258760107817, |
| "grad_norm": 0.58504718542099, |
| "learning_rate": 0.00041906745817593086, |
| "loss": 3.442, |
| "step": 28050 |
| }, |
| { |
| "epoch": 3.0296495956873315, |
| "grad_norm": 0.590460479259491, |
| "learning_rate": 0.00041874365893146247, |
| "loss": 3.4482, |
| "step": 28100 |
| }, |
| { |
| "epoch": 3.035040431266846, |
| "grad_norm": 0.5873590707778931, |
| "learning_rate": 0.000418419859686994, |
| "loss": 3.4321, |
| "step": 28150 |
| }, |
| { |
| "epoch": 3.0404312668463613, |
| "grad_norm": 0.5897404551506042, |
| "learning_rate": 0.00041809606044252557, |
| "loss": 3.45, |
| "step": 28200 |
| }, |
| { |
| "epoch": 3.045822102425876, |
| "grad_norm": 0.5874994397163391, |
| "learning_rate": 0.00041777226119805717, |
| "loss": 3.4424, |
| "step": 28250 |
| }, |
| { |
| "epoch": 3.0512129380053907, |
| "grad_norm": 0.5833684206008911, |
| "learning_rate": 0.0004174484619535887, |
| "loss": 3.4642, |
| "step": 28300 |
| }, |
| { |
| "epoch": 3.056603773584906, |
| "grad_norm": 0.609053909778595, |
| "learning_rate": 0.00041712466270912033, |
| "loss": 3.4483, |
| "step": 28350 |
| }, |
| { |
| "epoch": 3.0619946091644206, |
| "grad_norm": 0.5735374093055725, |
| "learning_rate": 0.0004168008634646519, |
| "loss": 3.4695, |
| "step": 28400 |
| }, |
| { |
| "epoch": 3.0673854447439353, |
| "grad_norm": 0.60845947265625, |
| "learning_rate": 0.0004164770642201835, |
| "loss": 3.4465, |
| "step": 28450 |
| }, |
| { |
| "epoch": 3.07277628032345, |
| "grad_norm": 0.6065637469291687, |
| "learning_rate": 0.00041615326497571503, |
| "loss": 3.4424, |
| "step": 28500 |
| }, |
| { |
| "epoch": 3.078167115902965, |
| "grad_norm": 0.5800220370292664, |
| "learning_rate": 0.0004158294657312466, |
| "loss": 3.443, |
| "step": 28550 |
| }, |
| { |
| "epoch": 3.08355795148248, |
| "grad_norm": 0.6139459609985352, |
| "learning_rate": 0.0004155056664867782, |
| "loss": 3.4636, |
| "step": 28600 |
| }, |
| { |
| "epoch": 3.0889487870619945, |
| "grad_norm": 0.6870418190956116, |
| "learning_rate": 0.0004151818672423097, |
| "loss": 3.4479, |
| "step": 28650 |
| }, |
| { |
| "epoch": 3.0943396226415096, |
| "grad_norm": 0.5827794075012207, |
| "learning_rate": 0.0004148580679978413, |
| "loss": 3.4704, |
| "step": 28700 |
| }, |
| { |
| "epoch": 3.0997304582210243, |
| "grad_norm": 0.5892341732978821, |
| "learning_rate": 0.00041453426875337284, |
| "loss": 3.4355, |
| "step": 28750 |
| }, |
| { |
| "epoch": 3.105121293800539, |
| "grad_norm": 0.5540435314178467, |
| "learning_rate": 0.00041421046950890445, |
| "loss": 3.4644, |
| "step": 28800 |
| }, |
| { |
| "epoch": 3.1105121293800537, |
| "grad_norm": 0.6163308024406433, |
| "learning_rate": 0.000413886670264436, |
| "loss": 3.4522, |
| "step": 28850 |
| }, |
| { |
| "epoch": 3.115902964959569, |
| "grad_norm": 0.5884636640548706, |
| "learning_rate": 0.0004135628710199676, |
| "loss": 3.4664, |
| "step": 28900 |
| }, |
| { |
| "epoch": 3.1212938005390836, |
| "grad_norm": 0.6189049482345581, |
| "learning_rate": 0.00041323907177549915, |
| "loss": 3.465, |
| "step": 28950 |
| }, |
| { |
| "epoch": 3.1266846361185983, |
| "grad_norm": 0.6232777833938599, |
| "learning_rate": 0.0004129152725310307, |
| "loss": 3.4594, |
| "step": 29000 |
| }, |
| { |
| "epoch": 3.1266846361185983, |
| "eval_accuracy": 0.37088217778633, |
| "eval_loss": 3.4994044303894043, |
| "eval_runtime": 145.0712, |
| "eval_samples_per_second": 124.153, |
| "eval_steps_per_second": 7.762, |
| "step": 29000 |
| }, |
| { |
| "epoch": 3.1320754716981134, |
| "grad_norm": 0.6272684335708618, |
| "learning_rate": 0.00041259794927145165, |
| "loss": 3.4647, |
| "step": 29050 |
| }, |
| { |
| "epoch": 3.137466307277628, |
| "grad_norm": 0.5963659882545471, |
| "learning_rate": 0.00041227415002698326, |
| "loss": 3.4641, |
| "step": 29100 |
| }, |
| { |
| "epoch": 3.142857142857143, |
| "grad_norm": 0.5980457663536072, |
| "learning_rate": 0.0004119503507825148, |
| "loss": 3.4658, |
| "step": 29150 |
| }, |
| { |
| "epoch": 3.1482479784366575, |
| "grad_norm": 0.6148801445960999, |
| "learning_rate": 0.00041163302752293576, |
| "loss": 3.4656, |
| "step": 29200 |
| }, |
| { |
| "epoch": 3.1536388140161726, |
| "grad_norm": 0.6103987693786621, |
| "learning_rate": 0.00041130922827846726, |
| "loss": 3.4807, |
| "step": 29250 |
| }, |
| { |
| "epoch": 3.1590296495956873, |
| "grad_norm": 0.6764293313026428, |
| "learning_rate": 0.0004109854290339989, |
| "loss": 3.4681, |
| "step": 29300 |
| }, |
| { |
| "epoch": 3.164420485175202, |
| "grad_norm": 0.5955349802970886, |
| "learning_rate": 0.0004106616297895304, |
| "loss": 3.4627, |
| "step": 29350 |
| }, |
| { |
| "epoch": 3.169811320754717, |
| "grad_norm": 0.5901073813438416, |
| "learning_rate": 0.000410337830545062, |
| "loss": 3.4575, |
| "step": 29400 |
| }, |
| { |
| "epoch": 3.175202156334232, |
| "grad_norm": 0.6012365818023682, |
| "learning_rate": 0.00041001403130059357, |
| "loss": 3.4659, |
| "step": 29450 |
| }, |
| { |
| "epoch": 3.1805929919137466, |
| "grad_norm": 0.6144543886184692, |
| "learning_rate": 0.00040969023205612517, |
| "loss": 3.4615, |
| "step": 29500 |
| }, |
| { |
| "epoch": 3.1859838274932613, |
| "grad_norm": 0.7040764689445496, |
| "learning_rate": 0.0004093664328116567, |
| "loss": 3.4752, |
| "step": 29550 |
| }, |
| { |
| "epoch": 3.1913746630727764, |
| "grad_norm": 0.6097145676612854, |
| "learning_rate": 0.00040904263356718833, |
| "loss": 3.4716, |
| "step": 29600 |
| }, |
| { |
| "epoch": 3.196765498652291, |
| "grad_norm": 0.6183958649635315, |
| "learning_rate": 0.0004087188343227199, |
| "loss": 3.4591, |
| "step": 29650 |
| }, |
| { |
| "epoch": 3.202156334231806, |
| "grad_norm": 0.6033417582511902, |
| "learning_rate": 0.00040839503507825143, |
| "loss": 3.4897, |
| "step": 29700 |
| }, |
| { |
| "epoch": 3.207547169811321, |
| "grad_norm": 0.6251811981201172, |
| "learning_rate": 0.00040807123583378303, |
| "loss": 3.4603, |
| "step": 29750 |
| }, |
| { |
| "epoch": 3.2129380053908356, |
| "grad_norm": 0.6118080019950867, |
| "learning_rate": 0.0004077474365893146, |
| "loss": 3.4704, |
| "step": 29800 |
| }, |
| { |
| "epoch": 3.2183288409703503, |
| "grad_norm": 0.6248761415481567, |
| "learning_rate": 0.0004074236373448462, |
| "loss": 3.4467, |
| "step": 29850 |
| }, |
| { |
| "epoch": 3.223719676549865, |
| "grad_norm": 0.6107171177864075, |
| "learning_rate": 0.00040709983810037774, |
| "loss": 3.4591, |
| "step": 29900 |
| }, |
| { |
| "epoch": 3.22911051212938, |
| "grad_norm": 0.6119734644889832, |
| "learning_rate": 0.00040677603885590934, |
| "loss": 3.4716, |
| "step": 29950 |
| }, |
| { |
| "epoch": 3.234501347708895, |
| "grad_norm": 0.6032190918922424, |
| "learning_rate": 0.00040645223961144084, |
| "loss": 3.4767, |
| "step": 30000 |
| }, |
| { |
| "epoch": 3.234501347708895, |
| "eval_accuracy": 0.37148009462797305, |
| "eval_loss": 3.493159770965576, |
| "eval_runtime": 144.5095, |
| "eval_samples_per_second": 124.635, |
| "eval_steps_per_second": 7.792, |
| "step": 30000 |
| } |
| ], |
| "logging_steps": 50, |
| "max_steps": 92750, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 10, |
| "save_steps": 10000, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 2.508207538176e+17, |
| "train_batch_size": 32, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
