| { | |
| "best_metric": 2.07828688621521, | |
| "best_model_checkpoint": "fr-mina/checkpoint-4745", | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 4745, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.026343519494204427, | |
| "grad_norm": 53.76327133178711, | |
| "learning_rate": 4.390586582367405e-07, | |
| "loss": 8.0275, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.05268703898840885, | |
| "grad_norm": 31.924686431884766, | |
| "learning_rate": 8.78117316473481e-07, | |
| "loss": 8.0678, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.07903055848261328, | |
| "grad_norm": 39.08976745605469, | |
| "learning_rate": 1.3171759747102214e-06, | |
| "loss": 7.683, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.1053740779768177, | |
| "grad_norm": 35.28089904785156, | |
| "learning_rate": 1.756234632946962e-06, | |
| "loss": 7.5655, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.13171759747102213, | |
| "grad_norm": 26.91787338256836, | |
| "learning_rate": 2.1952932911837023e-06, | |
| "loss": 7.4298, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.15806111696522657, | |
| "grad_norm": 36.86225128173828, | |
| "learning_rate": 2.634351949420443e-06, | |
| "loss": 7.1249, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.18440463645943098, | |
| "grad_norm": 35.9243278503418, | |
| "learning_rate": 3.073410607657183e-06, | |
| "loss": 6.6859, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.2107481559536354, | |
| "grad_norm": 31.12093734741211, | |
| "learning_rate": 3.512469265893924e-06, | |
| "loss": 6.5787, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.23709167544783982, | |
| "grad_norm": 21.474767684936523, | |
| "learning_rate": 3.951527924130664e-06, | |
| "loss": 6.3458, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.26343519494204426, | |
| "grad_norm": 29.76708984375, | |
| "learning_rate": 4.3905865823674045e-06, | |
| "loss": 6.2686, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.2897787144362487, | |
| "grad_norm": 26.34191131591797, | |
| "learning_rate": 4.829645240604145e-06, | |
| "loss": 5.9099, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.31612223393045313, | |
| "grad_norm": 25.640949249267578, | |
| "learning_rate": 5.268703898840886e-06, | |
| "loss": 5.1909, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.3424657534246575, | |
| "grad_norm": 28.177034378051758, | |
| "learning_rate": 5.707762557077626e-06, | |
| "loss": 5.6095, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.36880927291886195, | |
| "grad_norm": 31.351806640625, | |
| "learning_rate": 6.146821215314366e-06, | |
| "loss": 5.356, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.3951527924130664, | |
| "grad_norm": 31.123451232910156, | |
| "learning_rate": 6.585879873551107e-06, | |
| "loss": 5.3215, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.4214963119072708, | |
| "grad_norm": 25.988691329956055, | |
| "learning_rate": 7.024938531787848e-06, | |
| "loss": 4.9161, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.44783983140147526, | |
| "grad_norm": 20.850446701049805, | |
| "learning_rate": 7.463997190024588e-06, | |
| "loss": 5.0717, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.47418335089567965, | |
| "grad_norm": 28.87261962890625, | |
| "learning_rate": 7.903055848261329e-06, | |
| "loss": 4.7932, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.5005268703898841, | |
| "grad_norm": 39.34148406982422, | |
| "learning_rate": 8.342114506498068e-06, | |
| "loss": 4.9443, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.5268703898840885, | |
| "grad_norm": 24.568782806396484, | |
| "learning_rate": 8.781173164734809e-06, | |
| "loss": 4.6028, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.553213909378293, | |
| "grad_norm": 38.465816497802734, | |
| "learning_rate": 9.22023182297155e-06, | |
| "loss": 4.2893, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.5795574288724974, | |
| "grad_norm": 22.170591354370117, | |
| "learning_rate": 9.65929048120829e-06, | |
| "loss": 4.3665, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.6059009483667018, | |
| "grad_norm": 25.342151641845703, | |
| "learning_rate": 1.009834913944503e-05, | |
| "loss": 4.4401, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.6322444678609063, | |
| "grad_norm": 29.228736877441406, | |
| "learning_rate": 1.0537407797681771e-05, | |
| "loss": 4.0221, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.6585879873551106, | |
| "grad_norm": 26.896635055541992, | |
| "learning_rate": 1.0976466455918512e-05, | |
| "loss": 3.9198, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.684931506849315, | |
| "grad_norm": 22.59519386291504, | |
| "learning_rate": 1.1415525114155251e-05, | |
| "loss": 3.9569, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.7112750263435195, | |
| "grad_norm": 27.328128814697266, | |
| "learning_rate": 1.1854583772391992e-05, | |
| "loss": 3.7594, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.7376185458377239, | |
| "grad_norm": 28.952634811401367, | |
| "learning_rate": 1.2293642430628732e-05, | |
| "loss": 3.6405, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.7639620653319283, | |
| "grad_norm": 34.70003890991211, | |
| "learning_rate": 1.2732701088865473e-05, | |
| "loss": 3.5767, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.7903055848261328, | |
| "grad_norm": 31.843168258666992, | |
| "learning_rate": 1.3171759747102214e-05, | |
| "loss": 3.424, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.8166491043203372, | |
| "grad_norm": 43.431888580322266, | |
| "learning_rate": 1.3610818405338955e-05, | |
| "loss": 3.6013, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.8429926238145417, | |
| "grad_norm": 24.29863929748535, | |
| "learning_rate": 1.4049877063575696e-05, | |
| "loss": 3.6026, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.8693361433087461, | |
| "grad_norm": 18.842683792114258, | |
| "learning_rate": 1.4488935721812435e-05, | |
| "loss": 3.2906, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.8956796628029505, | |
| "grad_norm": 24.327327728271484, | |
| "learning_rate": 1.4927994380049176e-05, | |
| "loss": 3.3199, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.9220231822971549, | |
| "grad_norm": 34.937644958496094, | |
| "learning_rate": 1.536705303828592e-05, | |
| "loss": 3.3293, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.9483667017913593, | |
| "grad_norm": 20.94723129272461, | |
| "learning_rate": 1.5806111696522658e-05, | |
| "loss": 3.061, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.9747102212855637, | |
| "grad_norm": 36.616371154785156, | |
| "learning_rate": 1.6245170354759397e-05, | |
| "loss": 3.3562, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_gen_len": 29.7495, | |
| "eval_loss": 3.1300747394561768, | |
| "eval_rouge1": 14.3649, | |
| "eval_rouge2": 3.2092, | |
| "eval_rougeL": 14.1859, | |
| "eval_rougeLsum": 14.1403, | |
| "eval_runtime": 1240.6646, | |
| "eval_samples_per_second": 0.383, | |
| "eval_steps_per_second": 0.096, | |
| "step": 949 | |
| }, | |
| { | |
| "epoch": 1.0010537407797682, | |
| "grad_norm": 26.46944236755371, | |
| "learning_rate": 1.6684229012996136e-05, | |
| "loss": 3.2056, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.0273972602739727, | |
| "grad_norm": 24.677677154541016, | |
| "learning_rate": 1.7123287671232875e-05, | |
| "loss": 3.0972, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 1.053740779768177, | |
| "grad_norm": 44.66987609863281, | |
| "learning_rate": 1.7562346329469618e-05, | |
| "loss": 3.1316, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.0800842992623814, | |
| "grad_norm": 22.583946228027344, | |
| "learning_rate": 1.8001404987706357e-05, | |
| "loss": 2.9408, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 1.106427818756586, | |
| "grad_norm": 37.008148193359375, | |
| "learning_rate": 1.84404636459431e-05, | |
| "loss": 3.3225, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.1327713382507902, | |
| "grad_norm": 31.299516677856445, | |
| "learning_rate": 1.887952230417984e-05, | |
| "loss": 2.8608, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 1.1591148577449948, | |
| "grad_norm": 41.19516372680664, | |
| "learning_rate": 1.931858096241658e-05, | |
| "loss": 2.9428, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.1854583772391991, | |
| "grad_norm": 36.30160903930664, | |
| "learning_rate": 1.975763962065332e-05, | |
| "loss": 3.0541, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 1.2118018967334037, | |
| "grad_norm": 29.6497859954834, | |
| "learning_rate": 2.019669827889006e-05, | |
| "loss": 3.0771, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.238145416227608, | |
| "grad_norm": 28.226837158203125, | |
| "learning_rate": 2.0635756937126803e-05, | |
| "loss": 2.8116, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 1.2644889357218125, | |
| "grad_norm": 37.17683029174805, | |
| "learning_rate": 2.1074815595363542e-05, | |
| "loss": 2.709, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.2908324552160169, | |
| "grad_norm": 24.86113929748535, | |
| "learning_rate": 2.1513874253600282e-05, | |
| "loss": 2.8454, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 1.3171759747102212, | |
| "grad_norm": 20.32333755493164, | |
| "learning_rate": 2.1952932911837024e-05, | |
| "loss": 3.0137, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.3435194942044257, | |
| "grad_norm": 20.579383850097656, | |
| "learning_rate": 2.2391991570073764e-05, | |
| "loss": 3.064, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 1.36986301369863, | |
| "grad_norm": 34.56845474243164, | |
| "learning_rate": 2.2831050228310503e-05, | |
| "loss": 2.7602, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.3962065331928346, | |
| "grad_norm": 26.004846572875977, | |
| "learning_rate": 2.3270108886547242e-05, | |
| "loss": 2.6513, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 1.422550052687039, | |
| "grad_norm": 27.322162628173828, | |
| "learning_rate": 2.3709167544783985e-05, | |
| "loss": 3.0455, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 1.4488935721812435, | |
| "grad_norm": 25.812335968017578, | |
| "learning_rate": 2.4148226203020724e-05, | |
| "loss": 2.6421, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 1.4752370916754478, | |
| "grad_norm": 22.53229522705078, | |
| "learning_rate": 2.4587284861257463e-05, | |
| "loss": 2.6565, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.5015806111696524, | |
| "grad_norm": 21.36598777770996, | |
| "learning_rate": 2.5026343519494206e-05, | |
| "loss": 2.8773, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 1.5279241306638567, | |
| "grad_norm": 21.852357864379883, | |
| "learning_rate": 2.5465402177730945e-05, | |
| "loss": 2.5911, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 1.554267650158061, | |
| "grad_norm": 24.827373504638672, | |
| "learning_rate": 2.5904460835967688e-05, | |
| "loss": 2.8661, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 1.5806111696522656, | |
| "grad_norm": 26.9720458984375, | |
| "learning_rate": 2.6343519494204427e-05, | |
| "loss": 2.6055, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.60695468914647, | |
| "grad_norm": 23.86491584777832, | |
| "learning_rate": 2.6782578152441166e-05, | |
| "loss": 2.4583, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 1.6332982086406744, | |
| "grad_norm": 35.060543060302734, | |
| "learning_rate": 2.722163681067791e-05, | |
| "loss": 2.7086, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 1.6596417281348788, | |
| "grad_norm": 33.904869079589844, | |
| "learning_rate": 2.766069546891465e-05, | |
| "loss": 3.0039, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 1.685985247629083, | |
| "grad_norm": 16.747236251831055, | |
| "learning_rate": 2.809975412715139e-05, | |
| "loss": 2.7121, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.7123287671232876, | |
| "grad_norm": 29.914405822753906, | |
| "learning_rate": 2.853881278538813e-05, | |
| "loss": 2.4394, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 1.7386722866174922, | |
| "grad_norm": 34.94831848144531, | |
| "learning_rate": 2.897787144362487e-05, | |
| "loss": 2.2323, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 1.7650158061116965, | |
| "grad_norm": 23.3820743560791, | |
| "learning_rate": 2.9416930101861612e-05, | |
| "loss": 2.2026, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 1.7913593256059008, | |
| "grad_norm": 23.368724822998047, | |
| "learning_rate": 2.985598876009835e-05, | |
| "loss": 2.3719, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.8177028451001054, | |
| "grad_norm": 23.903703689575195, | |
| "learning_rate": 3.0295047418335094e-05, | |
| "loss": 2.6685, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 1.84404636459431, | |
| "grad_norm": 28.905014038085938, | |
| "learning_rate": 3.073410607657184e-05, | |
| "loss": 2.798, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 1.8703898840885143, | |
| "grad_norm": 37.410953521728516, | |
| "learning_rate": 3.117316473480857e-05, | |
| "loss": 2.8655, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 1.8967334035827186, | |
| "grad_norm": 30.735639572143555, | |
| "learning_rate": 3.1612223393045315e-05, | |
| "loss": 2.4698, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.9230769230769231, | |
| "grad_norm": 17.455341339111328, | |
| "learning_rate": 3.205128205128206e-05, | |
| "loss": 2.5119, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 1.9494204425711275, | |
| "grad_norm": 34.35090255737305, | |
| "learning_rate": 3.2490340709518794e-05, | |
| "loss": 2.6294, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 1.975763962065332, | |
| "grad_norm": 22.377056121826172, | |
| "learning_rate": 3.292939936775553e-05, | |
| "loss": 2.5728, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_gen_len": 8.5832, | |
| "eval_loss": 2.4636101722717285, | |
| "eval_rouge1": 25.9213, | |
| "eval_rouge2": 9.8991, | |
| "eval_rougeL": 25.4651, | |
| "eval_rougeLsum": 25.4305, | |
| "eval_runtime": 181.2174, | |
| "eval_samples_per_second": 2.621, | |
| "eval_steps_per_second": 0.657, | |
| "step": 1898 | |
| }, | |
| { | |
| "epoch": 2.0021074815595363, | |
| "grad_norm": 24.795652389526367, | |
| "learning_rate": 3.336845802599227e-05, | |
| "loss": 2.0971, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.0284510010537407, | |
| "grad_norm": 42.50971984863281, | |
| "learning_rate": 3.3807516684229015e-05, | |
| "loss": 2.2645, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 2.0547945205479454, | |
| "grad_norm": 28.93120765686035, | |
| "learning_rate": 3.424657534246575e-05, | |
| "loss": 2.2754, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 2.0811380400421498, | |
| "grad_norm": 32.35176467895508, | |
| "learning_rate": 3.4685634000702494e-05, | |
| "loss": 1.9605, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 2.107481559536354, | |
| "grad_norm": 20.049243927001953, | |
| "learning_rate": 3.5124692658939236e-05, | |
| "loss": 1.985, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.1338250790305584, | |
| "grad_norm": 21.458831787109375, | |
| "learning_rate": 3.556375131717597e-05, | |
| "loss": 1.9294, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 2.1601685985247627, | |
| "grad_norm": 29.076791763305664, | |
| "learning_rate": 3.6002809975412715e-05, | |
| "loss": 2.5018, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 2.1865121180189675, | |
| "grad_norm": 17.459365844726562, | |
| "learning_rate": 3.644186863364946e-05, | |
| "loss": 2.2102, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 2.212855637513172, | |
| "grad_norm": 35.115997314453125, | |
| "learning_rate": 3.68809272918862e-05, | |
| "loss": 2.2814, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 2.239199157007376, | |
| "grad_norm": 20.998336791992188, | |
| "learning_rate": 3.7319985950122936e-05, | |
| "loss": 2.1817, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 2.2655426765015805, | |
| "grad_norm": 24.086315155029297, | |
| "learning_rate": 3.775904460835968e-05, | |
| "loss": 2.3293, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 2.291886195995785, | |
| "grad_norm": 15.898116111755371, | |
| "learning_rate": 3.819810326659642e-05, | |
| "loss": 2.3658, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 2.3182297154899896, | |
| "grad_norm": 44.00349807739258, | |
| "learning_rate": 3.863716192483316e-05, | |
| "loss": 2.2536, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 2.344573234984194, | |
| "grad_norm": 30.016952514648438, | |
| "learning_rate": 3.90762205830699e-05, | |
| "loss": 2.1975, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 2.3709167544783982, | |
| "grad_norm": 29.138978958129883, | |
| "learning_rate": 3.951527924130664e-05, | |
| "loss": 2.1086, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 2.3972602739726026, | |
| "grad_norm": 22.48296546936035, | |
| "learning_rate": 3.995433789954338e-05, | |
| "loss": 1.8731, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 2.4236037934668073, | |
| "grad_norm": 38.969642639160156, | |
| "learning_rate": 4.039339655778012e-05, | |
| "loss": 2.4843, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 2.4499473129610116, | |
| "grad_norm": 46.12045669555664, | |
| "learning_rate": 4.0832455216016864e-05, | |
| "loss": 2.0695, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 2.476290832455216, | |
| "grad_norm": 16.667869567871094, | |
| "learning_rate": 4.1271513874253606e-05, | |
| "loss": 2.0357, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 2.5026343519494203, | |
| "grad_norm": 28.602094650268555, | |
| "learning_rate": 4.171057253249034e-05, | |
| "loss": 1.9143, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 2.528977871443625, | |
| "grad_norm": 21.20049476623535, | |
| "learning_rate": 4.2149631190727085e-05, | |
| "loss": 2.329, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 2.5553213909378294, | |
| "grad_norm": 24.602462768554688, | |
| "learning_rate": 4.258868984896383e-05, | |
| "loss": 1.9732, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 2.5816649104320337, | |
| "grad_norm": 41.749755859375, | |
| "learning_rate": 4.3027748507200563e-05, | |
| "loss": 2.3922, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 2.608008429926238, | |
| "grad_norm": 29.759376525878906, | |
| "learning_rate": 4.3466807165437306e-05, | |
| "loss": 2.1928, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 2.6343519494204424, | |
| "grad_norm": 27.628883361816406, | |
| "learning_rate": 4.390586582367405e-05, | |
| "loss": 2.6254, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.660695468914647, | |
| "grad_norm": 20.805973052978516, | |
| "learning_rate": 4.4344924481910785e-05, | |
| "loss": 1.937, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 2.6870389884088515, | |
| "grad_norm": 24.72698974609375, | |
| "learning_rate": 4.478398314014753e-05, | |
| "loss": 1.9697, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 2.713382507903056, | |
| "grad_norm": 32.659202575683594, | |
| "learning_rate": 4.522304179838426e-05, | |
| "loss": 1.945, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 2.73972602739726, | |
| "grad_norm": 30.338882446289062, | |
| "learning_rate": 4.5662100456621006e-05, | |
| "loss": 2.3917, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 2.7660695468914644, | |
| "grad_norm": 24.97830581665039, | |
| "learning_rate": 4.610115911485774e-05, | |
| "loss": 2.3004, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 2.792413066385669, | |
| "grad_norm": 20.519182205200195, | |
| "learning_rate": 4.6540217773094484e-05, | |
| "loss": 2.0284, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 2.8187565858798735, | |
| "grad_norm": 18.327741622924805, | |
| "learning_rate": 4.697927643133123e-05, | |
| "loss": 1.8822, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 2.845100105374078, | |
| "grad_norm": 28.89052963256836, | |
| "learning_rate": 4.741833508956797e-05, | |
| "loss": 2.3916, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 2.8714436248682826, | |
| "grad_norm": 17.761756896972656, | |
| "learning_rate": 4.7857393747804706e-05, | |
| "loss": 2.1041, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 2.897787144362487, | |
| "grad_norm": 16.94314193725586, | |
| "learning_rate": 4.829645240604145e-05, | |
| "loss": 2.0099, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 2.9241306638566913, | |
| "grad_norm": 28.35548973083496, | |
| "learning_rate": 4.873551106427819e-05, | |
| "loss": 1.9854, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 2.9504741833508956, | |
| "grad_norm": 20.59808921813965, | |
| "learning_rate": 4.917456972251493e-05, | |
| "loss": 2.0036, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 2.9768177028451, | |
| "grad_norm": 24.746854782104492, | |
| "learning_rate": 4.961362838075167e-05, | |
| "loss": 2.0894, | |
| "step": 2825 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_gen_len": 18.1389, | |
| "eval_loss": 2.2465784549713135, | |
| "eval_rouge1": 27.1383, | |
| "eval_rouge2": 12.4631, | |
| "eval_rougeL": 26.7601, | |
| "eval_rougeLsum": 26.8419, | |
| "eval_runtime": 963.3991, | |
| "eval_samples_per_second": 0.493, | |
| "eval_steps_per_second": 0.124, | |
| "step": 2847 | |
| }, | |
| { | |
| "epoch": 3.0031612223393047, | |
| "grad_norm": 15.351967811584473, | |
| "learning_rate": 4.9994145884556845e-05, | |
| "loss": 1.7484, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 3.029504741833509, | |
| "grad_norm": 12.023371696472168, | |
| "learning_rate": 4.994536158919721e-05, | |
| "loss": 1.6976, | |
| "step": 2875 | |
| }, | |
| { | |
| "epoch": 3.0558482613277134, | |
| "grad_norm": 16.859472274780273, | |
| "learning_rate": 4.989657729383757e-05, | |
| "loss": 1.5859, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 3.0821917808219177, | |
| "grad_norm": 11.666887283325195, | |
| "learning_rate": 4.984779299847793e-05, | |
| "loss": 1.518, | |
| "step": 2925 | |
| }, | |
| { | |
| "epoch": 3.108535300316122, | |
| "grad_norm": 16.126272201538086, | |
| "learning_rate": 4.979900870311829e-05, | |
| "loss": 1.6562, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 3.134878819810327, | |
| "grad_norm": 16.743412017822266, | |
| "learning_rate": 4.9750224407758655e-05, | |
| "loss": 1.7747, | |
| "step": 2975 | |
| }, | |
| { | |
| "epoch": 3.161222339304531, | |
| "grad_norm": 21.94359588623047, | |
| "learning_rate": 4.9701440112399017e-05, | |
| "loss": 1.3415, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 3.1875658587987354, | |
| "grad_norm": 20.92635154724121, | |
| "learning_rate": 4.965265581703938e-05, | |
| "loss": 1.5768, | |
| "step": 3025 | |
| }, | |
| { | |
| "epoch": 3.2139093782929398, | |
| "grad_norm": 19.112150192260742, | |
| "learning_rate": 4.960387152167975e-05, | |
| "loss": 1.4553, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 3.2402528977871445, | |
| "grad_norm": 14.94166088104248, | |
| "learning_rate": 4.955508722632011e-05, | |
| "loss": 1.5237, | |
| "step": 3075 | |
| }, | |
| { | |
| "epoch": 3.266596417281349, | |
| "grad_norm": 32.23541259765625, | |
| "learning_rate": 4.950630293096047e-05, | |
| "loss": 1.7323, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 3.292939936775553, | |
| "grad_norm": 15.371477127075195, | |
| "learning_rate": 4.945751863560083e-05, | |
| "loss": 1.661, | |
| "step": 3125 | |
| }, | |
| { | |
| "epoch": 3.3192834562697575, | |
| "grad_norm": 17.414264678955078, | |
| "learning_rate": 4.9408734340241195e-05, | |
| "loss": 1.8292, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 3.3456269757639623, | |
| "grad_norm": 24.73505401611328, | |
| "learning_rate": 4.935995004488155e-05, | |
| "loss": 1.4615, | |
| "step": 3175 | |
| }, | |
| { | |
| "epoch": 3.3719704952581666, | |
| "grad_norm": 33.94083023071289, | |
| "learning_rate": 4.931116574952191e-05, | |
| "loss": 1.7071, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 3.398314014752371, | |
| "grad_norm": 38.488319396972656, | |
| "learning_rate": 4.9262381454162274e-05, | |
| "loss": 1.5914, | |
| "step": 3225 | |
| }, | |
| { | |
| "epoch": 3.4246575342465753, | |
| "grad_norm": 15.29854965209961, | |
| "learning_rate": 4.9213597158802636e-05, | |
| "loss": 1.529, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 3.4510010537407796, | |
| "grad_norm": 39.15888977050781, | |
| "learning_rate": 4.9164812863443005e-05, | |
| "loss": 1.6042, | |
| "step": 3275 | |
| }, | |
| { | |
| "epoch": 3.4773445732349844, | |
| "grad_norm": 37.293724060058594, | |
| "learning_rate": 4.9116028568083366e-05, | |
| "loss": 1.4374, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 3.5036880927291887, | |
| "grad_norm": 27.48360824584961, | |
| "learning_rate": 4.906724427272373e-05, | |
| "loss": 1.5287, | |
| "step": 3325 | |
| }, | |
| { | |
| "epoch": 3.530031612223393, | |
| "grad_norm": 13.632729530334473, | |
| "learning_rate": 4.901845997736409e-05, | |
| "loss": 1.563, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 3.5563751317175973, | |
| "grad_norm": 20.885683059692383, | |
| "learning_rate": 4.896967568200445e-05, | |
| "loss": 1.638, | |
| "step": 3375 | |
| }, | |
| { | |
| "epoch": 3.5827186512118017, | |
| "grad_norm": 46.394649505615234, | |
| "learning_rate": 4.8920891386644814e-05, | |
| "loss": 1.6575, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 3.6090621707060064, | |
| "grad_norm": 37.50140380859375, | |
| "learning_rate": 4.8872107091285176e-05, | |
| "loss": 1.7124, | |
| "step": 3425 | |
| }, | |
| { | |
| "epoch": 3.6354056902002108, | |
| "grad_norm": 20.98285675048828, | |
| "learning_rate": 4.882332279592554e-05, | |
| "loss": 1.6254, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 3.661749209694415, | |
| "grad_norm": 11.166362762451172, | |
| "learning_rate": 4.87745385005659e-05, | |
| "loss": 1.69, | |
| "step": 3475 | |
| }, | |
| { | |
| "epoch": 3.68809272918862, | |
| "grad_norm": 29.99367904663086, | |
| "learning_rate": 4.872575420520626e-05, | |
| "loss": 1.5022, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 3.714436248682824, | |
| "grad_norm": 24.22369384765625, | |
| "learning_rate": 4.8676969909846624e-05, | |
| "loss": 1.8366, | |
| "step": 3525 | |
| }, | |
| { | |
| "epoch": 3.7407797681770285, | |
| "grad_norm": 18.417123794555664, | |
| "learning_rate": 4.8628185614486986e-05, | |
| "loss": 1.5953, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 3.767123287671233, | |
| "grad_norm": 31.92102813720703, | |
| "learning_rate": 4.857940131912735e-05, | |
| "loss": 1.8851, | |
| "step": 3575 | |
| }, | |
| { | |
| "epoch": 3.793466807165437, | |
| "grad_norm": 19.559165954589844, | |
| "learning_rate": 4.853061702376771e-05, | |
| "loss": 1.3543, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 3.819810326659642, | |
| "grad_norm": 27.526994705200195, | |
| "learning_rate": 4.848183272840807e-05, | |
| "loss": 1.4503, | |
| "step": 3625 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "grad_norm": 21.375267028808594, | |
| "learning_rate": 4.8433048433048433e-05, | |
| "loss": 1.7254, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 3.8724973656480506, | |
| "grad_norm": 23.377002716064453, | |
| "learning_rate": 4.8384264137688795e-05, | |
| "loss": 1.6948, | |
| "step": 3675 | |
| }, | |
| { | |
| "epoch": 3.898840885142255, | |
| "grad_norm": 32.76591110229492, | |
| "learning_rate": 4.8335479842329164e-05, | |
| "loss": 1.5097, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 3.9251844046364592, | |
| "grad_norm": 28.73731803894043, | |
| "learning_rate": 4.8286695546969526e-05, | |
| "loss": 1.3383, | |
| "step": 3725 | |
| }, | |
| { | |
| "epoch": 3.951527924130664, | |
| "grad_norm": 33.73754119873047, | |
| "learning_rate": 4.823791125160989e-05, | |
| "loss": 1.4893, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 3.9778714436248683, | |
| "grad_norm": 16.65566635131836, | |
| "learning_rate": 4.818912695625025e-05, | |
| "loss": 1.6426, | |
| "step": 3775 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_gen_len": 8.9979, | |
| "eval_loss": 2.1172335147857666, | |
| "eval_rouge1": 33.7585, | |
| "eval_rouge2": 17.8059, | |
| "eval_rougeL": 33.2997, | |
| "eval_rougeLsum": 33.2581, | |
| "eval_runtime": 180.3517, | |
| "eval_samples_per_second": 2.634, | |
| "eval_steps_per_second": 0.66, | |
| "step": 3796 | |
| }, | |
| { | |
| "epoch": 4.004214963119073, | |
| "grad_norm": 26.248737335205078, | |
| "learning_rate": 4.8140342660890605e-05, | |
| "loss": 1.4139, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 4.030558482613277, | |
| "grad_norm": 40.171348571777344, | |
| "learning_rate": 4.809155836553097e-05, | |
| "loss": 0.9346, | |
| "step": 3825 | |
| }, | |
| { | |
| "epoch": 4.056902002107481, | |
| "grad_norm": 18.67420768737793, | |
| "learning_rate": 4.804277407017133e-05, | |
| "loss": 1.0417, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 4.083245521601686, | |
| "grad_norm": 15.394213676452637, | |
| "learning_rate": 4.799398977481169e-05, | |
| "loss": 1.1243, | |
| "step": 3875 | |
| }, | |
| { | |
| "epoch": 4.109589041095891, | |
| "grad_norm": 14.950016021728516, | |
| "learning_rate": 4.794520547945205e-05, | |
| "loss": 0.9402, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 4.135932560590095, | |
| "grad_norm": 33.47651290893555, | |
| "learning_rate": 4.789642118409242e-05, | |
| "loss": 1.053, | |
| "step": 3925 | |
| }, | |
| { | |
| "epoch": 4.1622760800842995, | |
| "grad_norm": 23.766538619995117, | |
| "learning_rate": 4.784763688873278e-05, | |
| "loss": 0.9098, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 4.188619599578503, | |
| "grad_norm": 24.7260684967041, | |
| "learning_rate": 4.7798852593373145e-05, | |
| "loss": 0.8711, | |
| "step": 3975 | |
| }, | |
| { | |
| "epoch": 4.214963119072708, | |
| "grad_norm": 14.13185977935791, | |
| "learning_rate": 4.775006829801351e-05, | |
| "loss": 0.9576, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 4.241306638566913, | |
| "grad_norm": 26.87774085998535, | |
| "learning_rate": 4.770128400265387e-05, | |
| "loss": 1.1116, | |
| "step": 4025 | |
| }, | |
| { | |
| "epoch": 4.267650158061117, | |
| "grad_norm": 24.789899826049805, | |
| "learning_rate": 4.765249970729423e-05, | |
| "loss": 1.2492, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 4.293993677555322, | |
| "grad_norm": 35.61201858520508, | |
| "learning_rate": 4.760371541193459e-05, | |
| "loss": 1.0259, | |
| "step": 4075 | |
| }, | |
| { | |
| "epoch": 4.3203371970495255, | |
| "grad_norm": 18.178863525390625, | |
| "learning_rate": 4.7554931116574955e-05, | |
| "loss": 1.0043, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 4.34668071654373, | |
| "grad_norm": 18.80617904663086, | |
| "learning_rate": 4.750614682121532e-05, | |
| "loss": 1.254, | |
| "step": 4125 | |
| }, | |
| { | |
| "epoch": 4.373024236037935, | |
| "grad_norm": 23.337120056152344, | |
| "learning_rate": 4.745736252585568e-05, | |
| "loss": 1.3335, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 4.399367755532139, | |
| "grad_norm": 27.5556640625, | |
| "learning_rate": 4.740857823049604e-05, | |
| "loss": 1.1739, | |
| "step": 4175 | |
| }, | |
| { | |
| "epoch": 4.425711275026344, | |
| "grad_norm": 14.718132019042969, | |
| "learning_rate": 4.73597939351364e-05, | |
| "loss": 1.232, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 4.4520547945205475, | |
| "grad_norm": 16.896610260009766, | |
| "learning_rate": 4.7311009639776765e-05, | |
| "loss": 0.9822, | |
| "step": 4225 | |
| }, | |
| { | |
| "epoch": 4.478398314014752, | |
| "grad_norm": 32.641151428222656, | |
| "learning_rate": 4.7262225344417126e-05, | |
| "loss": 1.1221, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 4.504741833508957, | |
| "grad_norm": 8.409883499145508, | |
| "learning_rate": 4.721344104905749e-05, | |
| "loss": 1.0836, | |
| "step": 4275 | |
| }, | |
| { | |
| "epoch": 4.531085353003161, | |
| "grad_norm": 24.261465072631836, | |
| "learning_rate": 4.716465675369785e-05, | |
| "loss": 1.0266, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 4.557428872497366, | |
| "grad_norm": 27.14542579650879, | |
| "learning_rate": 4.711587245833821e-05, | |
| "loss": 0.9686, | |
| "step": 4325 | |
| }, | |
| { | |
| "epoch": 4.58377239199157, | |
| "grad_norm": 27.667455673217773, | |
| "learning_rate": 4.706708816297858e-05, | |
| "loss": 1.0164, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 4.610115911485774, | |
| "grad_norm": 22.17734718322754, | |
| "learning_rate": 4.701830386761894e-05, | |
| "loss": 1.1826, | |
| "step": 4375 | |
| }, | |
| { | |
| "epoch": 4.636459430979979, | |
| "grad_norm": 19.197708129882812, | |
| "learning_rate": 4.6969519572259305e-05, | |
| "loss": 0.8725, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 4.662802950474183, | |
| "grad_norm": 18.602115631103516, | |
| "learning_rate": 4.692073527689966e-05, | |
| "loss": 1.3157, | |
| "step": 4425 | |
| }, | |
| { | |
| "epoch": 4.689146469968388, | |
| "grad_norm": 19.589008331298828, | |
| "learning_rate": 4.687195098154002e-05, | |
| "loss": 1.2225, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 4.715489989462593, | |
| "grad_norm": 24.026233673095703, | |
| "learning_rate": 4.6823166686180384e-05, | |
| "loss": 1.1642, | |
| "step": 4475 | |
| }, | |
| { | |
| "epoch": 4.7418335089567965, | |
| "grad_norm": 16.99645233154297, | |
| "learning_rate": 4.6774382390820746e-05, | |
| "loss": 1.0141, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 4.768177028451001, | |
| "grad_norm": 15.595638275146484, | |
| "learning_rate": 4.672559809546111e-05, | |
| "loss": 1.1805, | |
| "step": 4525 | |
| }, | |
| { | |
| "epoch": 4.794520547945205, | |
| "grad_norm": 16.28839683532715, | |
| "learning_rate": 4.667681380010147e-05, | |
| "loss": 1.3, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 4.82086406743941, | |
| "grad_norm": 27.51399040222168, | |
| "learning_rate": 4.662802950474184e-05, | |
| "loss": 1.1732, | |
| "step": 4575 | |
| }, | |
| { | |
| "epoch": 4.847207586933615, | |
| "grad_norm": 30.363340377807617, | |
| "learning_rate": 4.65792452093822e-05, | |
| "loss": 1.2096, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 4.8735511064278185, | |
| "grad_norm": 27.562767028808594, | |
| "learning_rate": 4.653046091402256e-05, | |
| "loss": 1.0491, | |
| "step": 4625 | |
| }, | |
| { | |
| "epoch": 4.899894625922023, | |
| "grad_norm": 15.739079475402832, | |
| "learning_rate": 4.6481676618662924e-05, | |
| "loss": 1.0191, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 4.926238145416228, | |
| "grad_norm": 30.40022850036621, | |
| "learning_rate": 4.6432892323303286e-05, | |
| "loss": 1.2202, | |
| "step": 4675 | |
| }, | |
| { | |
| "epoch": 4.952581664910432, | |
| "grad_norm": 14.871630668640137, | |
| "learning_rate": 4.638410802794365e-05, | |
| "loss": 1.0677, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 4.978925184404637, | |
| "grad_norm": 25.979448318481445, | |
| "learning_rate": 4.633532373258401e-05, | |
| "loss": 1.2033, | |
| "step": 4725 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_gen_len": 11.7958, | |
| "eval_loss": 2.07828688621521, | |
| "eval_rouge1": 37.8388, | |
| "eval_rouge2": 19.8532, | |
| "eval_rougeL": 37.3344, | |
| "eval_rougeLsum": 37.4253, | |
| "eval_runtime": 401.1496, | |
| "eval_samples_per_second": 1.184, | |
| "eval_steps_per_second": 0.297, | |
| "step": 4745 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 28470, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 30, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.01 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 15600663134208.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |